vxEPA
United States
Environmental Protection
Agency
  Fish Physiology, Toxicology,
        and Water Quality

  Proceedings of the Ninth International
        Symposium, Capri, Italy,
           April 24-28, 2006
       RESEARCH AND DEVELOPMENT

-------
                                            EPA/600/R-07/010
                                               February 2007
   Fish  Physiology, Toxicology, and
                 Water Quality
Proceedings of the Ninth International Symposium,
           Capri, Italy, April 24-28, 2006
                         Edited by

        lin J. Brauner, fern Suvajdzic, 2Goran Nilsson, *David Randall
       Department of Zoology, University of British Columbia, Canada
                  2University of Oslo, Norway
                       Published by

                  Ecosystems Research Division
                    Athens, Georgia 30605
               U.S. Environmental Protection Agency
                Office of Research and Development
                    Washington, DC 20460

-------
                                   NOTICE

       The views expressed in these Proceedings are those of the individual authors and
do not necessarily reflect the views and policies of the U.S. Environmental Protection
Agency (EPA).  Mention of trade names or commercial products does not constitute
endorsement or recommendation for use by EPA.
                                       11

-------
                                  ABSTRACT

       Scientists from Europe, North America and South America convened in Capri,
Italy, April 24-28, 2006 for the Ninth International Symposium on Fish Physiology,
Toxicology, and Water Quality. The subject of the meeting was "Eutrophi cation: The
toxic effects of ammonia, nitrite and the detrimental effects of hypoxia on fish." These
proceedings include 22 papers presented over a 3-day period and discuss eutrophication,
ammonia and nitrite toxicity and the effects of hypoxia on fish with the aim of
understanding the effects of eutrophication on fish. The ever increasing human
population and the animals raised for human consumption discharge their sewage into
rivers and coastal waters worldwide. This is resulting in eutrophication of rivers and
coastal waters everywhere. Eutrophication is associated with elevated ammonia and
nitrite levels, both of which are toxic, and the water often becomes hypoxic. Aquatic
hypoxia has been shown to reduce species diversity and reduce total biomass.
                                        in

-------
                                    CONTENTS

NOTICE	ii
ABSTRACT	iii
ACKNOWLEDGEMENTS	vi
FOREWORD	vii
      ENVIRONMENTAL EFFECTS OF EUTROPHICATION AND HYPOXIA

Environmental eutrophication and its effects on fish of the Amazon.
  A.L. Veil, M.N. Paula da Silva, and V.M.F. Almeida-Val	1

Biochemical responses to hypoxia: The case of amazon fishes.
  V.M.F. Almeida-Val, A.R. Chippari-Gomes, N.P. Lopes, R. Araujo, S.R. Nozawa, M.S.
  Ferreira-Nozawa, M. de Nazare Paula-Silva, andA.L. Val	13

Swimming performance as a practical and effective biomarker of pollution exposure in
fish.
  E. W. Taylor andD.J. McKenzie	25

Management of eutrophication.
  M.J. Gromiec	41
                     AMMONIA AND NITRITE TOXICITY

Effects of ammonia on locomotor performance in fishes.
  D.J. McKenzie, A. Shingles, P. Domenici, andE.W. Taylor	49

Ammonia and salinity tolerance in the California Mozambique tilapia.
  K. Suvajdzic, B. Sardella, andC.J. Brauner	65

Nitrite toxicity to fishes.
  R.C.Russo	73

The formation of S-nitrosoglutathione in conditions mimicking hypoxia and acidosis.
  L. Grossi	89

Nitrite regulates hypoxic responses and protects against ischemia/reperfusion injury.
  S. Shiva, C. Dezfulian, andM.T. Gladwin	97

Influence of nitrite on the mechanical performance offish and mammalian hearts.
  D. Pellegrino, T. Angelone, B. Tola, andM.T. Gladwin	107
                                      IV

-------
Potential NO-routes in the piscine circulation.
  C. Agnisola	113
Physiological effects of nitrite: Balancing the knife's edge between toxic disruption of
functions and potential beneficial effects.
  F.B.Jensen	119
                           RESPONSES TO HYPOXIA

Hypoxia in fish.
  D.J.Randall, W.L. Poon, C.C.Y. Hung, andT.K.N. Tsui	133

Gene expression profiles of common carp, Cyprinus carpio, during prolonged starvation
and hypoxia reflect differences in hypometabolism.
  C.C.Y. Hung, A.R Cossins, A.Y. Gracey, and D.J.Randall	139

Studies of gene expression in brain of anoxic crucian carp.
  S. EHefsen ,  G.K. Sandvik, D.A. SteenhoffHov, T. Kristensen, andG.E. Nilsson	151

Doing the impossible: Anoxic cell division in the crucian carp (Carassius carassius)
  C. S0rensen, J. Sollid, andG.E. Nilsson	155

Hypoxia-tolerance in a tropical elasmobranch: does adenosine trigger a multi-system
protective response?
  G.M.C. Renshaw	165

Behavioural, respiratory, ionoregulatory, and N-metabolic adaptations to low
environmental O2, and the influence of body size in the hypoxia-tolerant Amazonian
oscar (Astronotus ocellatus).
  C.M. Wood, K.A. Sloman, M. Kajimura, O.E.  Johannsson, P.J. Walsh, G. Scott,_S.
  Wood, V.M.F. Almeida-Val, andA.L.  Val	.".	179

Effect of winter hypoxia on fish in small lakes with different water quality.
  A. Tuvikene, L. Tuvikene, andM. Viik	201

Interdemic variation in gill morphology  of a eurytopic African cichlid.
  L.J. Chapman, T. DeWitt, V. Tzaneva, and J. Pater son	209

Hypoxia tolerance in coral  reef teleosts.
  G.E. Nilsson and S. Ostlund-Nilsson	227

Oxygen consumption offish exposed to hypoxia: Are they all oxyregulators or are any
oxyconformers?
  J.F. Steffensen	239

-------
                          ACKNOWLEDGEMENTS

       The 9th International Symposium was organized by Drs. Rosemarie Russo and
Brenda Rashleigh, US Environmental Protection Agency (EPA) Ecosystems Research
Division, Athens, Georgia, Drs. Colin Brauner and David Randall from the Department
of Zoology, University of British Columbia, Dr. Bruno Tota, University of Calabria,
Italy, and Dr. Goran Nilsson, University of Oslo, Norway. The meeting was held in the
Hotel St. Michele on the Isle of Capri, Italy and managed by Ms. Laurajean Carbonaro
who did an excellent job. The meeting was funded by the U.S. EPA (EPA-ORD-25489)
and the University of British Columbia, Canada, the "A. Dohrn Zoological Station of
Naples" and the Doctorate School of Animal Biology of the University of Calabria, Italy.
We were advised throughout by Drs. Brenda Rashleigh and Rosemarie Russo. Ms. Kim
Suvajdzic assisted greatly in the preparation of this document and Dan Baker, Clarice Fu,
Matthew Regan, and Jodie Rummer helped with many of the reviews. Finally, I attribute
the  success of this Symposium to the many researchers who came to Capri to present
their findings.
Department of Zoology,
University of British Columbia
Vancouver, BC, Canada

Colin Brauner
                                      VI

-------
                                 FOREWORD

Symposia on Fish Physiology, Toxicology, and Water Quality; a Brief History

       Vance Thurston and Rosemarie Russo were good friends long before I knew
them, and I have known Vance since the 1970's. We would go to Russia together and we
had a number of joint research projects. In the early 1980's Rose and I were driving from
Athens to Atlanta when she told me she had been asked to organize a science exchange
program with several research institutes in China.  She asked if I knew anyone in China
because she knew I had been a guest of the central government in China and had worked
at Zhongshan University in Guangzhou, PRC for several months. The end result was that
UBC held the cooperative agreement to encourage research collaboration between
environmental scientists from the Peoples' Republic of China and scientists from North
America and Europe, under the USA-PRC environmental protection agreement.

       The First Symposium was held at Zhongshan University, Guangzhou, PRC, in
September 1988, with the help of Professor Lin Hao-ran of Zhongshan University, and
attracted scientists from Europe, Canada, and the U.S., as well as many scientists from
the PRC and Hong Kong. This was the beginning of the series of international symposia
organized by Vance, Rose, and myself, sponsored by the US Environmental Protection
Agency through the Athens Laboratory. The Guangzhou Symposium was memorable for
its audio equipment: the sound was such that the lectures could be heard by people on
boats passing down the Pearl River. The Second Symposium was held two years later in
September 1990 in  Sacramento, California, with the help of Professor Joe Cech. We had
an excellent dinner in the Train Museum; the positive response from all participants
illustrated the world wide acceptance of Californian cuisine. The location for the Third
Symposium was Nanjing University  in Nanjing, PRC, in November 1992 and this time
we had the skilled help of Professor Jin Hong-jun. Vance, the great entertainer, sang
songs during dinner and as usual brought us together as a group. Vance made a special
effort for the Fourth Symposium, held in Bozeman, Montana, in September 1995, to
encourage participation of scientists from Europe, as well as from the PRC, North
America, and Mexico.  Vance drove  a van  from Bozeman to San Francisco with Chinese
delegates on board to show them various sites of ecological interest in North America.
We returned to China for the Fifth Symposium, which was held at the City University of
Hong Kong, in November 1998 with the able help of Professor Rudolf Wu. The Sixth
Symposium was held in La Paz, Mexico, in January 2001.  In addition to attracting a large
audience  from the Mexican scientific community, the 30 papers accepted for presentation
represented 15 countries, more than any previous Symposium. The Seventh Symposium,
in Tallinn, Estonia, was affectionately dedicated to the memory of Robert Vance
Thurston, who died unexpectedly on February  16th, 2002,  at the age of 75. Dr. Arvo
Tuvikene was a great help in putting that Symposium together, along with Gretchen Rupp
from Montana State University. Vance was very active in international environmental
research projects in the Baltic republics, the former Soviet Union, and Mexico. He was
practically an honorary citizen of Lithuania for the many projects he had there and the
computers and other equipment he provided to their scientists. Rose and I organized the
Eighth Symposium in Chongqing, China in October 2004 in association with Dr. George
                                      vn

-------
Bailey of Athens EPA and the able help of Professor Gao Yuqi of the 3rd Military
Medical University in Chongqing.

       The functions of these Symposia are twofold, the first to exchange scientific
information and the second to remove political barriers between scientists from different
countries and promote collaboration. This has been achieved. Now we come to the Ninth
Symposium in this long-standing series, to be held in Capri, Italy with the able assistance
of Professor Bruno Tota and Laurajean Carbonaro.
David Randall
University of British Columbia
Vancouver, BC, Canada
                                       Vlll

-------
Attendees of the Ninth International Symposium on Fish Physiology, Toxicology, and
Water Quality, Capri, Italy, April 24-28, 2006.
                                        IX

-------
 Environmental eutrophication and  its effects on fish of
                               the Amazon

                                       by

               A.L. Val1, M.N. Paula da Silva and V.M.F. Almeida-Val

Introduction

       Eutrophication is a natural process whereby lakes, estuaries and slow-moving
streams receive excess nutrients as a consequence of weathering of rocks and soils from
the surrounding watershed. Increased nutrient inputs, particularly phosphorus and
nitrogen, result in increased growth of aquatic plants and organic production of the water
body. Young water bodies (lakes and man made reservoirs) usually are oligotrophic as
they have low levels of nutrients and correspondingly low levels of biological activity. In
contrast, old water bodies possess high biological activity as a consequence of high
nutrient levels. These are referred to as eutrophic water bodies. The natural time scale
from being oligotrophic to eutrophic is in the order of thousands of years, depending on
the levels of encrusted minerals and on the rate of watershed weathering, among other
environmental  characteristics (Wetzel, 1975). These terms were first applied to lakes by
Naumann in early 1900  (Naumann, 1919, 1927) noting that oligotrophic lakes contained
modest levels of algae and were often found in igneous rock areas while eutrophic lakes
contained high amounts of algae and were found in more fertile lowland regions. The
author concluded that within a normal thermal range, levels of phosphorus, nitrogen and
calcium are the primary determining factors of lake trophic status.

       There is no single or simple definition of eutrophy or oligotrophy. These states are
the extremes along an axis defining the trophic state of a given water mass. Based on
phosphorus concentrations, Mueller and Helsel (1999) classified lakes with
concentrations  belowlOugP/L as oligotrophic, those with concentrations betweenlO-
20ugP/L as mesotrophic, and those lakes with concentrations exceeding 20ug/L as
eutrophic. The  relationship between mineralization and production is also an important
tool defining the trophic state of water bodies as, in general, in oligotrophic lakes
production and mineralization are closely coupled, while in eutrophic lakes production
far exceeds mineralization resulting in accumulation of organic matter and a depletion of
oxygen (Niell et a/., 2005). Regardless of the definition employed, the evolution of a
water body from oligotrophic to eutrophic state is characterized by an overload of
nutrients, mainly nitrogen and phosphate, that displaces the system from the  equilibrium.

       In 1934, Alfred Redfield proposed that the N:P ratio in the interior of all major
oceans were remarkably similar,  based on empirical observation.  This ratio is 16:1 (N:P)
and is similar to that of plankton from all over the world. Today, the residence time of
1 Laboratory of Ecophysiology and Molecular Evolution, National Institute for Research
in the Amazon, Brazil, e-mail: dalval@inpa.gov.br.

-------
these elements relative to the ocean's circulation time is though to be the basis of the
Redfield ratio. Large lakes, though much more variable due to the effects of surrounding
ecosystems, display similar elemental composition ratios (Falkowski and Davis, 2004).
Indeed, some variability is expected due to several internal processes such as recycling,
sedimentation, resuspension or release from the bottom, nitrogen fixation, temperature,
among other factors, in both oceans and lakes (Tundisi and Matsumura-Tundisi, 1984;
Howarth, 1988).

       More recently, eutrophication rates have increased dramatically as a consequence
of alterations in nutrient cycles related to land-use changes, i.e., related to human inputs
of urban and agricultural waste, including sewage and fertilizers.  Analyses of rivers in the
temperate zone indicate an increase of 3-20 fold of river nitrogen export in developed
areas since industrialization. Since 1960, our population has doubled (we were six billion
people in 2000) which has necessitated a doubling of food production.  However, the use
of N and P  has increased at a much higher rate: the use of N as fertilizer increased 8.8
fold (from  10 to 88 million metric tons) and the use of P increased 4.4  fold (from 9 to 40
million metric tons). There is a projection of a further 50% increase in  the use of these
fertilizers in decade of 2030-2040 (Vance, 2001). These changes  lead to the concept of
cultural eutrophi cation where sudden environmental changes (10  years or less, a time
frame that contrasts to that of natural eutrophi cation of 1,000-10,000 years) displace the
ecosystem into a state of un-compensated disequilibrium (Stirn, 1987).

       In contrast to temperate water bodies, nutrient fluxes in tropical water bodies are
less well documented, limiting projections of aquatic disturbances caused by
anthropogenic factors such as increased silt, deforestation, nutrient loads and land-use
perturbations. In the Amazon, these perturbations are exacerbated by two important
environmental factors: a) increased incidence of ultraviolet radiation, and b) increased
temperature. The purpose of this review is to analyze the current  status of eutrophi cation
in the Amazon, both natural and cultural in origin, and the effects of eutrophication on
fish of the Amazon.

Eutrophication in the Amazon

       In general, there is a clear trend between pristine N-export and  water runoff which
is much higher for tropical rivers compared to temperate ones. There are two exceptions:
the  Rio Negro, a tributary of the Amazon, that is exceptionally low in sediment load, rich
in dissolved organic carbon, poor in ions and behaves as most pristine  temperate rivers;
and Mackenzie and Lena Rivers that are large  high latitude rivers carrying large sediment
loads and behaving as most tropical pristine rivers (Sioli,  1984; Furch  and Junk, 1997;
Downing etal., 1999).

       The major ecological driving force in the Amazon is the annual river water level
oscillation that affects nearly all organic-aquatic environment interactions (Fink and Fink,
1979; Val and Almeida-Val, 1995). The annual regular flood pulse (Junk et al., 1989)
extends over significant parts of nearby rivers flooding a myriad of ria and varzea lakes.
As the flooding occurs for few months every year, the lakes respond with bursts of

-------
production which are swept into the rivers (Rai and Hill, 1984). However, as the flood
pulses occur at different periods of the year within the region, the overall production in
the main river channel is low.

                               ;«-- •#„•'-*•.'• -'' *-:^3O* ;"..„'    .   ' - •'
                                               f •• i - ^;'' '^ >. *
                   *-~M- JtefvA  •\ff-l*t-nJ.  .,/.;•:-.:. '  , t , .PB*%,^
                   ^••^^m^'b?^'^-;,'-.^; s*/*y^. *^-f
            Figure 1.  The nutrients unloaded into the vdrzea as the river water level
                    increases result in an extensive propagation of aquatic plants that
                    cover extensive areas water surface limiting light penetration and
                    so causing reduced photosynthesis. As the water recedes the
                    aquatic plants decomposes what results in a further deterioration of
                    water quality.

       As the river water level increases it floods extensive areas unloading its nutrient
rich sediment which causes excessive growth of phytoplankton, algae and rooted aquatic
plants (macrophytes). These plants cover the entire water surface in many places (Fig. 1)
reducing light levels and rates of photosynthesis in the water column. When the water
stops flowing into these areas, there is a decrease in available nutrients causing extensive
plant decay. Subsequently, the water recedes leaving behind an enormous amount of dead
aquatic plants that lead to anoxia, high levels of hydrogen sulfide, methane and ammonia.
The naturally fertilized soil left behind is then used by  locals for production of vegetables

-------
before the next flood. So, the cycles of natural eutrophication in the pristine areas of the
Amazon have many social, economical and ecological implications (Junk, 1984; Val and
Almeida-Val, 1995; Junk, 1996).

       In general, tropical freshwaters are more nitrogen limited than temperate
freshwaters while phosphorus is frequently more limiting in tropical marine systems
(Downing et a/., 1999). These authors suggest that disturbances to pristine tropical land
will lead to profound freshwater disturbances. They have analyzed the effects of cultural
eutrophi cation, from deforestation, atypical disturbance of all phases of disturbances in
tropical regions, up to urban and industrial development and hypothesized for tropical
aquatic systems, including the Amazon, a significant increase of N-export and a decrease
of N:P ratio (see Downing et a/., 1999).

       Eutrophi cation, regardless of origin, causes significant changes within aquatic
communities. Fish for example experience a change in food availability as the food web
is dramatically affected by changes in water quality. Deterioration of water quality
imposes additional physiological challenges to fish that are naturally exposed to periodic
episodes of hypoxia, hydrogen sulfide and ammonia. In some cases the animals are
exposed to extreme conditions, hypoxia and hyperoxia, for example, within short periods
of time. In other cases, eutrophi cation may over expose fish to uncommon environmental
conditions, as ultraviolet, as under hypoxia many fish species breathe at the water-air
interface.

Effects of eutrophication on fish of the Amazon

       Natural eutrophication is a cyclic process that occurs every year in the Amazon.
Amazonian fish have evolved a myriad of adjustments to survive such environmental
conditions.

Oxygen

       The fish of the Amazon face low dissolved oxygen, a regular environmental
constraint since the formation of the Amazon basin. In fact, oxygen levels below the
present atmospheric levels existed when the major fish groups appeared during the
Devonian (Acanthopterygii), the Triassic (Teleostei), and the Jurassic (Euteleostei)
(Dudley, 1998).  Adaptations to low 62 levels occur at the behavioral, morphological,
physiological, and biochemical level and adjustments can be made as soon as the animal
senses hypoxia. Oxygen sensing has been analyzed under a variety of conditions and in a
diversity of animals and plants (Hochachka, 1996; Sundin etal., 2000; Bailey-Serres and
Chang, 2005). However, we are far from a clear picture of this issue for fish of the
Amazon.

       The first line of defense against hypoxia is behavioral. At least two behavioral
changes have been observed among the  fish of the Amazon: a position change within the
water column and lateral migration. When exposed to hypoxia, many fish species move
to the upper region of the water column, close to the air-water interface, where more

-------
dissolve oxygen is naturally available. While some fish species expand their lower lip to
facilitate surface skimming, as many species of serrasalmids, others gulp air and water
into the digestive system, as some loricariids (Gradwell, 1971), to aid in oxygen uptake
(Val, 1995). If access to the upper region of the water column is denied, there is a
significant decrease of blood oxygenation as observed for Pterygoplichthys multiradiatus
(Val, 1995). Lateral migration refers to a movement between the main river channel,
where dissolved oxygen is more stable, and the flooded forest, where food is available
(Junk etal., 1983). The sunset is accompanied by a significant decrease in dissolved
oxygen due to a reduction of photosynthesis and an increase of respiration and serves as
cue for many fish species to migrate back to the main river. Early in the morning these
fish migrate again back to the flooded forest to feed. Lateral migration has been described
for fish species in the Amazon (Lowe McConnell, 1987), Pantanal (Antunes de Moura,
2000),  and in Africa (Benech and Quensiere,  1982). These behavioral changes have been
reported for several  fish species and have been described as adaptive convergence
(Brauner and Val, 2006).

       Air-breathing for fish means independence from fluctuations of dissolved oxygen.
In addition to the obligatory air-breathers, such as Arapaima and Lepidosiren, there are
several groups of facultative air-breathers (Table I). Facultative air-breathers use many
structures to take up oxygen directly from air and are able to switch from water- to air-
breathing according to dissolve oxygen availability. In general, under normoxia these
animals are aquatic breathers while under hypoxia they rely on some degree of aerial
respiration, using a variety of air-breathing organs (ABO) (Val, 1999). Indeed, facultative
air-breathers switching to air-breathing experiences some extra physiological adjustments
related to blood chemistry, as blood oxygenation occur at the ABO and carbon dioxide
excretion takes  place at the gills. Aerial exposure may also increase the risk of predation.

       Table I.  Major air-breathing fish families of the Amazon. O=obligatory air-
               breather; F-facultative air-breather; L=lung; SB=swim bladder; Sk=skin;
               S/I=stomach and intestine; PBM=pharyngeal, branchial and mouth
               diverticula.
Fish Family
Lepidoseriniformes
Arapaimidae
Erythrinidae
Doradidade
Callichthyidae
Loricariidae
Rhamphi chthy i dae
Electrophoridae
Synbranchidae
Type
O
O
F
F
F
F
F
O
F
ABO
L
SB
Sk, SB, S/I
S/I
S/I
S/I
PBM
PBM
PBM

-------
       If hypoxia cannot be avoided, fish make adjustments either directed towards
increasing oxygen transfer to tissues or reducing oxygen consumption through metabolic
depression. Adjustments to blood characteristics are very common when fish are exposed
to stress which includes adrenergically mediated red blood cell swelling and release of
red blood cells from the spleen (Val, 1993; Randall and Perry, 1994) as reported in
tambaqui (Moura, 1994). In general, hypoxia causes a reduction of spontaneous activity
and metabolic consumption of oxygen (Brauner et a/., 1995; Almeida-Val et a/., 2000),
an increase in gill ventilation rate (Rantin et a/., 1992) and bradycardia (Rantin et a/.,
1995).

       In addition to behavioral and morphological adjustments, fish are able to increase
blood-oxygen affinity by adjusting the levels of organic phosphates within the
erythrocytes (Val, 2000). ATP and GTP are the major organic phosphates  detected in fish
erythrocytes and both are negative modulators of Hb-C>2 affinity. The concentration of
these phosphates within erythrocytes is regulated according to dissolved oxygen, i.e., the
lower the oxygen availability, the lower the levels of these phosphates. The reduction of
erythrocytic levels of ATP and GTP results in an increase of Hb-C>2 affinity, safeguarding
oxygen loading at the gills. In general the regulation of GTP is faster than  the regulation
of ATP. In addition to ATP and GTP, other phosphates have been detected in the
erythrocytes offish of the Amazon, namely 2,3DPG m Hoplosternum littorale, IPP in
Arapaima gigas, and I?2 (inositol diphosphate) and UTP (uridine-5'-triphosphate), in
Lepidosiren (Val, 2000).
                 1.6  -


                 1.2  -
                 0.0
=i  Normoxia
a  Hypoxia
   Hyperoxia
                                 ATP
                    GTP
                  Figure 2. Erythrocytic ATP and GTP levels in specimens of
                          Pygocentrus nattereri, a species of piranha, exposed to
                          normoxia, hypoxia and hyperoxia.

-------
       Eutrophication may lead some water bodies to oxygen super-saturation during the
day; contrasting to hypoxia caused by increased respiration at night. To prevent tissue
damage caused by excess oxygen, Pygocentrus nattereri, a species of piranha, is able to
increase the concentration of erythrocytic ATP and GTP (Fig. 2), decreasing Hb-O2
affinity. This daily oxygen oscillation (over-saturation during the day and anoxia at night)
imposes rapid adjustments directed towards oxygen transfer to tissues. Analysis of ATP
and GTP levels over time in the erythrocytes of the same species of piranha above
mentioned revealed a 40% reduction of these organic phosphates during the first ten
minutes of exposure to deep hypoxia.

       So, fish of the Amazon have developed a suite of adjustments to face low oxygen
that can be used during cyclic periods of natural eutrophication.  Cultural  eutrophication
processes, however, induce additional challenges that need to be further analyzed.

Ammonia/Nitrite

       In addition to changes in available oxygen, eutrophication causes  significant
accumulation of ammonia, inorganic phosphorus, nitrate and nitrite, particularly with
poor tidal flushing and high stocking density (Wu et a/., 1994; Gonzalez  et a/., 2004).
The toxic effect of ammonia on temperate fish has been extensively discussed elsewhere
(Tomasso etal., 1980; Randall and Wright, 1987; Shingles etal., 2001) but is unknown
in fish of the Amazon. The ammonia released to water is readily converted to nitrite by
Nitrosomonas. Nitrite together with other nitrogen compounds increases  in vdrzea lakes
as result of plant decomposition and diffuses across the fish gills and then into the
erythrocytes, there converting hemoglobin to methemoglobin. Methemoglobin is unable
to bind reversibly to oxygen which further aggravates the effects of eutrophication  on
fish.

Hydrogen sulfide

       Exceptionally high levels of hydrogen sulfide (HS) occur at the vdrzea lakes due
to the circulation of the water column, particularly after eutrophication and plant decay.
Together with HS, oxygen poor water is displaced from the bottom. Hydrogen sulfide
inhibits a series of enzymes, including many related to oxidative phosphorylation,
causing metabolic impairment and generation of oxygen radicals (Nichols and Kim,
1982; Hill etal., 1984). Thus, fishes exposed to HS are unable to maintain regular
biological activities. As hydrogen sulfide occurs together with low oxygen, many fish
alter conditions to increase oxygen transfer to tissues, resulting in an even greater HS
transfer to tissues  (Val, 1999).

       In general, the great majority offish that survive in varzea lakes during periods of
high HS/low oxygen are air-breathers. Analysis of the effect of HS on the facultative  air-
breather Hoplosternum littorale indicate that air-breathing  represents an important
adaptation to reduce HS  transfer to tissues in this animal (Brauner et a/.,  1995). The
authors reported that animals exposed to  acidic water (pH 3.8, Po2=155 mmHg) increased
air-breathing frequency to levels displayed by animals exposed to mild hypoxia (28 air

-------
breaths/h (ab/h)). Interestingly, air-breathing frequency increased up to 40ab/h in animals
exposed to buffered HS, suggesting that water-air breathing transition may be an
adaptation to acidic and high HS waters. In specimens denied access to air, low levels of
HS are lethal.

Anthropogenic activities and eutrophication

       Cultural eutrophication is a world wide matter of concern. In the Amazon, though
it does not represent a major issue due to the shear volume of the system, many
anthropogenic activities are taking place, all potentially able to displace the system from
equilibrium. Deforestation, fire, siltation, mineral and petroleum mining, river damming
and urbanization are, among others, causes of concern. These activities result in an
increase of nutrients and energy being transported to the water systems increasing aquatic
plant growth and changing water quality. In conjunction with other environmental
changes, cultural eutrophication is potentially dangerous to fish. For example,
deforestation and fire causes a removal of natural protection against solar radiation
(shadowing) in addition to increased siltation. As many fish species of the Amazon
respond to low dissolved oxygen by switching from water- to air-breathing, which may
increase UV exposure. Exposure to UV causes massive kills and DNA breakdown. Using
the comet assay we have observed a significant increase of DNA damage over time of
exposure of tambaqui to UVR (Groff etal., unpublished data). Crude oil  spills close to
eutrophic water bodies also create large challenges for fish. The water-air breathing
transition results in an increase of crude oil taken in inducing a series of physiological
disturbances. Thus, while the fish of the Amazon areprepared to face the regular
constraints caused by natural eutrophication, they are not prepared to face cultural
eutrophication, in particular when it is associated with other extreme environmental
changes such as increased UV and crude oil.

Conclusions

       Organisms  are expected to respond to novel events as if they are familiar events.
Fish of the Amazon have developed a suite of adjustments to face natural constraints of
their environment,  including those related to cyclic eutrophication process. Hypoxia, for
example, is a familiar event in the Amazon and so fish respond to it with adaptations
shaped over their existence, but in the case  of anthropogenic pressures these adaptations
may have negative consequences. Cultural eutrophication is becoming a matter of
concern, particularly around the major cities of the Amazon and its evolution should be
followed over  the next decades to support regional procedures.

References

Almeida-Val, V.M.F., A.L.  Val, W.P. Duncan, F.C.A. Souza, M.N. Paula-Silva, and S.
       Land. 2000. Scaling effects on hypoxia tolerance in the Amazon fish Astronotus
       ocellatus (Perciformes, Cichlidae): contribution of tissue enzyme levels. Comp.
       Biochem. Physiol.  1258:21-226.

-------
Antunes de Moura, N. 2000. Influencia de fatores fisico-quimicos e recursos alimentares
       na migra9ao lateral de peixes no lago Chocorore, Pantanal de Barao de Melga9o,
       estado de Mato Grosso. MSc Thesis, INPA/UFAM, Manaus, AM.

Bailey-Serres, J. and R. Chang. 2005. Sensing and signalling in response to oxygen
       deprivation in plants and other organisms. Ann. Bot. 96: 507-518.

Benech, V. and J. Quensiere.  1982. Migrations de poisson vers le lac Tchad a la decrue
       de la plaine inondee du Nord Cameroum. Rev. Hydrobiol. Trop. 15: 253-270.

Brauner, C.J., C.L. Ballantyne, DJ. Randall, and A.L. Val. 1995. Air breathing in the
       armoured catfish (Hoplosternum littorale} as an adaptation to hypoxic, acid, and
       hydrogen sulphide rich waters. Can.  J. Zool. 73: 739-744.

Brauner, C. J. and A.L. Val. 2006. Oxygen transfer. Pages 277-306 In: The Physiology of
       Tropical Fish. Vol. 21, A.L. Val, V.M.F. Almeida-Val, and DJ. Randall (Eds).
       Elsevier/Academic Press, San Diego.

Downing, J.A., M. MacClain, R.  Twilley, J.M. Melack, J. Elser, N.N. Rabalais, W.M.
       Lewis Jr., R.E. Turner, J.  Corredor, D.  Soto, A. Yanez-Arancibia, J.A. Kopaska,
       and R.W. Howarth.  1999. The impact of accelarating land-use change on the N-
       cycle of tropical aquatic ecosystems: Current conditions and projected changes.
       Biogeochem. 46: 109-148.

Dudley, R.  1998. Atmospheric oxygen, giant paleozoic insects and the evolution of aerial
       locomotor performance. J. Exp. Biol. 201: 1043-1050.

Falkowski, P.O. and C.S. Davis. 2004. Natural proportions. Nature. 431: 131.

Fink, W.L.  and S.V. Fink. 1979. Central Amazonia and its fishes. Comp. Biochem.
       Physiol., 62A: 13-29.

Furch, K. and WJ. Junk. 1997. Physicochemical conditions in the floodplains. Pages 69-
       108 In: The Central Amazon floodplain. Ecology of a pulsing system. Vol. 126.
       W. J. Junk (Ed). Springer Verlag, Heidelberg.

Gonzalez, E.J., M. Ortaz, C. Pefiaherrera, and A. Infante. 2004. Physical and chemical
       features of a tropical hypertrophic reservoir permanently stratified.
       Hydrobiologia. 522: 301-310.

Gradwell, N. 1971. A photographic analysis of the air breathing behavior of the catfish,
       Plecostomuspunctatus. Can. J. Zool. 49:  1089-1094.

Hill, B.C., T.C. Woon, P.D. Nichols, J. Peterson, C. Greenwood, and A.J. Thomson.
       1984. Interactions of sulphide and other ligands with cytochrome c oxidase.
       Biochem. J. 224: 591-600.

-------
Hochachka, P.W. 1996. Oxygen sensing and metabolic regulation: Short, intermediate
       and long term roles. Pages 233-256 In: Physiology and Biochemistry of the fishes
       of the Amazon. A.L. Val, V.M.F. Almeida-Val, and DJ. Randall (Eds). INPA,
       Manaus.

Howarth, R.W. 1988. Nutrient limitation of net productivity in marine ecosystems. Annu.
       Rev. Ecol. Syst. 19: 89-110.

Junk, WJ. 1984. Ecology of the varzea, floodplain of Amazonian Whitewater rivers.
       Pages 215-244 In: The Amazon. Limnology and landscape ecology of a mighty
       tropical river and  its basin. H. Sioli (Ed). Springer, Dordrecht.

Junk, W.J., P.B. Bayley, and R.E. Sparks.  1989. The flood pulse concept in River-
       Floodplain Systems. Pages 110-127 In: Proceedings of the International Large
       River Symposium. Vol. 106.  D.P. Dodge (Ed). Can. Spec. Publ. Fish. Aquat. Sci.,
       Canada.

Junk, W.J., M.G. Scares,  and F.M. Carvalho.  1983. Distribution offish species in a lake
       of the Amazon river floodplain near Manaus (lago Camaleao), with special
       reference to extreme oxygen  conditions. Amazoniana. 7:397-431.

Junk, W.L. 1996. Ecology of floodplains - a challenge for tropical limnology. In
       Perspectives in tropical limnology. F.  Schiemer and E.J. Boland, editors. SBP
       Academic Publishing bv, Amsterdam. 255-265.

Lowe McConnell, R.H. 1987. Page 382 In: Ecological studies in tropical fish
       communities. Cambridge University Press, Cambridge.

Moura, M.A.F. 1994. Efeito da anemia, do exercicio fisico e da adrenalina sobre o ba9o e
       eritrocitos de Colossoma macropomum (Pisces). Thesis, PPG INPA/FUA.

Mueller, D. and D. Helsel. 1999. Nutrients in the nation's waters—too much of a good
       thing? In: US Geological Survey Circular.  1136.

Naumann, E. 1919. Nagra synpunker agaende planktons okologi. Svensk Botanisk
       Tidskrift. 13: 129-158.

Naumann, E. 1927. Ziel und hauptprobleme der regionaler Limnologie. Botanisk Notiser.
       1927:81-103.

Nichols, P. and J.K. Kim. 1982.  Sulphide as an inhibitor and donor for the cytochrome c
       oxidase system. Can. J. Biochem. 60: 613-623.

Niell, F.X., A. Aviles, and V. Clavero. 2005. Effects of eutrophication on Biological
       systems. Pages  9-11 In: The concept of eutrophi cation. R.C. Russo (Ed).
                                       10

-------
Rai, H. and G. Hill. 1984. Primary production in the Amazonian aquatic ecosystem.
       Pages 311-335 In: The Amazon. Limnology and landscape ecology of a mighty
       tropical river ans its basin. H. Sioli (Ed). Dr W. Junk Publishes, Dordrecht.

Randall, DJ. and S.F. Perry. 1994. Catecholamines. Pages 255-300 In: Fish Physiology,
       Volume XIIB. W.S. Hoar, DJ. Randall, and A.P. Farrell (Eds). Academic Press,
       New York.

Randall, DJ. and P.A. Wright. 1987. Ammonia distribution and excretion in fish. Fish
       Physiol. Biochem. 3: 107-120.

Rantin, F.T., A.L. Kalinin, M.L. Glass, andM.N. Fernandes. 1992. Respiratory responses
       to hypoxia in relation to mode of life of two erythrinid species (Hoplias
       malabaricus and Hoplias lacerdae). J. Fish Biol. 41: 805-812.

Rantin, F.T., A.L. Kalinin, C. Guerra, M. Maricondi-Massari, andR. Verzola. 1995.
       Electroncardiographic characterization of myocardial function in normoxic and
       hypoxic teleosts. Braz. J. Med. Biol. Res. 28: 1277-1289.

Shingles, A., DJ. McKenzie, E.W. Taylor, A. Moretti, PJ. Butler, and S. Ceradini. 2001.
       Effects of sublethal ammonia exposure on swimming performance in rainbow
       trout (Oncorhynchus mykiss). J. Exp. Biol. 204: 2691-2698.

Sioli, H. 1984. The Amazon and its main affluents: Hydrogeography, morphology of the
       river courses and river types. Pages 127-165 In: The Amazon. Limnology and
       landscape ecology of a mighty tropical river and its basin. H.  Sioli (Ed). Dr. W.
       Junk Publishers, Dordrecht.

Stirn, J. 1987. Eutrophication in the Mediterranean Sea. Pages 131-187 In:
       Eutrophication in the Mediterranean Sea: receiving capacity and monitoring of
       long-term effects. Unesco, FAO, UNEP, Regione Emilia Romagna and University
       of Bologna, Bologna.

Sundin, L., S.G. Reid, F.T. Rantin, and W.K. Milson. 2000. Branchial receptors and
       cardiorespiratory reflexes in a neotropical fish, the tambaqui (Colossoma
       macropomum). J. Exp.  Biol. 203: 1225-1239.

Tomasso, J.R., C.A. Gouldie, B.A. Simco, and K.B. Davis. 1980. Effects of
       environmental pH and calcium on ammonia toxicity in channel catfish. Trans.
       Amer. Fish. Soc. 109: 229-234.

Tundisi, J.G. and T. Matsumura-Tundisi. 1984. Comparative limnological studies at three
       lakes in tropical Brazil. Verh. Internal. Verein. Limnol. 22: 1310-1314.
                                       11

-------
Val, A.L. 1993. Adaptations of fishes to extreme conditions in fresh waters. Pages 43-53
       In: The vertebrate gas transport cascade. Adaptations to environment and mode of
       life. J.E.P.W. Bicudo (Ed). CRC Press, Boca Raton.

Val, A.L. 1995. Oxygen transfer in fish: morphological and molecular adjustments. Braz.
       J. Med. Biol. Res. 28: 1119-1127.

Val, A.L. 1999. Water-air breathing transition in fishes of the Amazon. Pages 145-161
       In: Water/Air Transition in Biology. A.K. Mittal, D.E. Eddy, and IS. Datta
       Munshi (Eds). Science Publishers, Inc, Enfield.

Val, A.L. 2000. Organic phosphates in the red blood cells offish. Comp. Biochem.
       Physiol. 125A: 417-435.

Val, A.L. and V.M.F. Almeida-Val. 1995. Fishes of the Amazon and their environments.
       Page 224 In: Physiological and biochemical features. Springer Verlag,
       Heidelberg.

Vance, C.P. 2001. Symbiotic nitrogen fixation and phosphorus acquisition. Plant
       nutrition in a world of declining renewable resources.  Plant Physiol. 127: 390-
       397.

Wetzel, R.G. 1975. Page 743 In: Limnology.  W B Saunder Co, London.

Wu, R.S.S., K.S. Lam, D.W. MacKay, T.C. Lau, and V. Yam. 1994. Impact of marine
       fish farming on water quality and bottom sediment: A case study in the sub-
       tropical environment. Mar. Env. Res. 38: 115-145.
                                       12

-------
               Biochemical responses to  hypoxia:
                     The case of amazon fishes

                                     by

 V.M.F. Almeida-Val1, A.R. Chippari-Gomes1, N.P. Lopes2, R. Araujo1, S.R. Nozawa3,
           M.S. Ferreira-Nozawa3, M. de Nazare Paula-Silva1 and A.L. Val1

       The Amazon basin is the result of geological and climatic phenomena that took
place during the different eras. Hypoxic and anoxic conditions were prevalent in the
aquatic environment during the Cambrian period, owing to the low atmospheric oxygen
levels at that time. Since the Cambrian geological period, oxygen depletion has been a
limiting factor for aquatic life in general (Randall etal., 1981; Almeida-Val and Farias,
1996; Almeida-Val et a/., 1999). After the break up in the southern hemisphere that
caused the appearance of South America and Africa during the Cretaceous period, the
main geological phenomenon causing the Amazonian hydrographic basin formation was
the Andean Mountains uplift, occurring in the Tertiary period. This fact caused an
enormous change in the region, cutting off the Pacific Ocean drainage of the upper
tributaries of Amazon River and changing the whole orientation of the Amazon basin
towards the Atlantic Ocean (Reviewed by Val and Almeida-Val, 1995). Natural episodes
of hypoxia occur globally and have different causes  and effects.  The poorly oxygenated
waters of the Amazon basin result from a number of phenomena. Annual flood pulses
produce an average crest of 10 meters between November and June in central Amazonia
and the flooding of the jungle results in a complete new set of habitats becoming
available each year, causing large changes in water physical-chemical parameters,
including pH, water density, conductivity, temperature, and dissolved oxygen. Currently,
these annual flood pulses cause oscillations in oxygen availability with interspersed
episodes of severe hypoxia and anoxia (Fig.  1).
1 Laboratory for Ecophysiology and Molecular Evolution, Department of Ecology,
National Institute for Amazon Research (INPA), Manaus, AM, Brazil.
2 University of North (UNINORTE), Manaus, AM, Brazil.
3 University Nilton Lins (UNINILTONLINS), Manaus, AM, Brazil.
                                      13

-------
August 2000 March 2001
02 (mg / L) 02 (mg / L)
•p" 1 0 cm
£ 25 cm
Q.
Ol
^ 50 cm
Si
| 100 cm
125cm
01 234567
}f
J?
' '<' f /^ • '• ' 8 pm
\ * ^r ' midnight
4am
Sam
10cm
| 25 cm
^ 50 cm
TL
£ 75 cm
^ 1 00 cm
re
S 1 25 cm
1 50 cm
01234
V
f
j • noon
-j — .«~ 4 pm
" 8 pm
"ip " midnight
4 am
8 am
     Figure 1. Natural changes in dissolved oxygen in Catalao Lake, located in
             front of Manaus, where the Negro and Solimoes rivers mix.
             Panels A and B show diurnal and spatial changes from high and
             low water level seasons, respectively.

       Besides the natural phenomena, there are other causes for the changes in
dissolved oxygen, which are common in many water bodies and are caused mainly by
human activities. Acute pollution episodes may cause mortality or permanent damage to
aquatic organisms. Constant pollution activities also may induce a chronic decrease in
oxygen availability, which may result in alterations in species distribution or their
occurrence, or cause severe  decreases in population size. Interestingly, animals that
survive such conditions can adapt to these new hypoxic environments once their
evolutionary histories have provided them with many adaptive traits to survive hypoxic
conditions. For instance, fish of the Amazon have developed a series of coordinated
metabolic adjustments, which, combined with morphological and anatomical changes,
have resulted in a number of solutions to avoid or minimize the stress caused by hypoxia
(Val and Almeida-Val, 1995). In addition, long and short-term changes in oxygen are
determinants  offish distribution in Amazonian water bodies (Almeida-Val et a/., 1999;
Fig. 2).
               6  -

               5  -

               4  -

               3  -

               2  -

               1

               o  H
                 | N urn ber os species
                       Oxygen (mg/L)
3
                                _a
                                                                  160
                                              120
                                              80
                                              40
       Figure 2. Relationship between oxygen distribution (circle symbols) during the
                year, August, 2000 to July, 2001 (values obtained at noon) and the
                abundance (number of fishes - empty bars) and numbers of cichlid
                species (gray bars) captured near Catalao lake.
                                        14

-------
       Hypoxic episodes may be devastating for most ecosystems because it may cause
mass mortality, defaunation of benthic populations, declines in fisheries production,
changes in community composition, and, as an ultimate consequence, a decrease in
animal diversity. In the Amazon, chronic hypoxic situations are common, and have
resulted in fish adaptation at different levels of biological organization, thereby inducing
increased species diversity.  Seasonal changes in species composition may occur as the
result of different oxygen distribution in the environment. Junk etal. (1983) showed that
low oxygen levels are coincident with a selective occurrence of air-breathing fish species
in a varzea lake. The hypoxia tolerant cichlids are among the water-breathing fish
remaining in the lake during low oxygen periods.

       Some field studies have suggested that fish distribution in aquatic ecosystems of
the Amazon is the consequence of aquatic oxygen contents (Junk et a/., 1983; Crampton,
1998; Chippari-Gomes, 2002). Low oxygen environments are known to be the ideal
places for hypoxia tolerant species to avoid predation, competition for food, and other
constraints, since during low oxygen season, these places hold few species. Nevertheless,
varzea lakes, which are commonly hypoxic, are considered the main nursery site for
several Amazon fish  species due to the presence of high levels of organic matter. The
high abundance offish fmgerlings and juveniles in varzea lakes are, thus, a paradox. To
address this issue, we have investigated one of the most anoxia tolerant fish ever
described in a tropical region (Muusze et al.,  1998): the Oscar, Astronotus ocellatus
(Cichlid: Perciformes).

       Studies conducted in our laboratory have demonstrated that both anaerobic power
(the ability of the animal to  activate anaerobic metabolism, as reflected by LDH absolute
levels in skeletal muscle) and hypoxia survivorship are a function of body mass
(Almeida-Val etal., 1999; 2000). As fish size increases, the ability to survive under
severe hypoxia increases  as well. The total amount of time required to reach
disequilibrium (loss of orientation that precedes death) increases as the animal  increases
in size, suggesting that the Oscar shows an increase in hypoxia tolerance with age (Fig.
3).

                     ~  50
                     £_
                     "aT  30
                     E
                     D
                     V)
                     D)
                     O
15
 9

 5
 3
                               5     15     50    150
                                 log - body mass (g)
                              350
      Figure 3. Relationship between body mass (g) of Astronotus ocellatus and its
               ability to survive hypoxia (hours). The log-log regression shows a
               close relationship (r=0.98) (Almeida-Val etal., 1999).
                                        15

-------
       Recent studies comparing the behavior of small and adult Oscars when exposed to
hypoxia revealed different strategies between the two groups (Sloman et a/., 2006).
Specific metabolic rates are higher in small animals (Almeida-Val et a/., 1999, 2000;
Sloman et a/., 2006).  Adult Oscars are able to regulate their oxygen consumption by
adjustments in respiration and circulation to a lower oxygen threshold than juveniles (50
Torr compared to 70 Torr in the latter). Adult Oscars also have a much greater ability to
survive exposure to extreme hypoxia as already mentioned. Fish around 16 g in weight
(equivalent to 'small') survived extreme hypoxia for about 9 h (Almeida-Val et al., 2000)
and the respective larger individuals weighing 230 g, survived for approximately 35 h.
Thus while adult Oscars appear to tolerate falling PO2 slightly better than their juvenile
counterparts this difference is magnified considerably once extreme hypoxia is reached
and anaerobic metabolism becomes necessary. The greater anaerobic potential of adults,
as indicated by higher concentrations of lactate dehydrogenase and malate
dehydrogenase, (Almeida-Val et a/., 2000) also fits with this scenario. Thus, unlike other
temperate species (e.g. yellow perch, Percaflavescens, Robb and Abrahams, 2003;
largemouth bass, Micropterus salmoides; Burleson et a/., 2001) the Oscar shows a
positive relationship between physiological tolerance of hypoxia and size.

       Besides the increase in anaerobic potential  as animals become bigger (Almeida-
Val et a/., 2000), Oscars also have a low metabolic rate, undoubtedly lower than most
fish species, as demonstrated by Almeida-Val et al. (1999). Among Amazon fishes,
metabolic rates of exclusively water-breathing fishes vary related to a spectrum of
sluggish to "athletic"  type behavior patterns. As would be expected, comparisons
between Amazon and temperate fish species suggest that the more sluggish the fish the
less the amount of oxygen consumed per unit weight (reviewed by Val and Almeida-Val,
1995).

       Metabolic rate plotted versus total fish mass shows the relationship between body
mass (g) and organism oxygen uptake (mg oxygen per fish per hour). The allometric
relationship is described as: VOi = aM6, where a = log mass coefficient, M= log body
mass and b = mass exponent. The value of the exponent b obtained for Oscar is 0.52.
Reviewing this exponent for several species of tropical fishes, Hammer and Purps (1996)
reported a mean value of 0.73. Compared to other tropical fishes, Oscar has one of the
lowest exponents reported. Former studies with temperate fish species described an
exponent of 0.86 (Glass,  1969). It is evident that tropical fishes have lower metabolic
rates, in general, than temperate fishes, and that, among troical fishes, Oscars have the
lowest metabolic rate among all species studied  (Fig. 4).
                                       16

-------
CD
+-»
2
o
55  J3
S  04
0  O
         O
            an
         &D  vK
         M)
         O
                                                             "r"Fishes in general


                                                             "r" Tropical fishes


                                                              T  Oscar
                                         =0.59
                               log - body mass (g)
      Figure 4. Relationship between body mass (g) and whole organism oxygen
               uptake ofAstronotus ocellatus (Oscar) compared with the same
               relationship for tropical fishes and fishes in general.

       We can summarize the Oscar's responses to hypoxia as follows: 1) escaping the
hypoxic water or skimming the water surface; 2) reducing metabolic rate; 3) activating
anaerobic glycolysis; 4) depressing metabolic rate below standard rates; 5) gene
regulation and signal transduction.

       Most studies on whole animals conducted in our laboratory with Amazon fish
species subjected to some level of oxygen depletion (acute hypoxia, graded hypoxia or
anoxia). These studies revealed that the animals  showed alterations in plasma glucose and
lactate levels resulting from glucose reserve mobilization and anaerobic-based lactate
production (Fig. 5 and 6). From this data, it becomes clear that anaerobic glycolysis takes
place in most species. This response can be combined with metabolic depression in some
species, mainly in those that are already known to be hypoxia-tolerant, such as the
cichlids, the group that includes the Oscar.
                                        17

-------
     300
    |>250-
    3>200H
    ro

    | 100-


       50-
       0
               Normoxia
               Hypoxia
            II
i
                 Catfishes
                     Cichlids
Figure 5. Plasma glucose levels from different species under normoxia and
         acute hypoxia. The two groups are composed of species of catfishes
         (Siluriformes) and cichlids (Perciformes) (Data obtained from
         Almeida-Val et al., 2005).
\
£
1
a
V) K .
_a 3
Q.
M-4-
o
I2"
1 •
0-
^^1 Normoxia
i i Hypoxia

,





JU,
••






1,















i
'•

















n



































i








s






Catfishes Cichlids
 Figure 6. Plasma lactate levels from different species under normoxia and
          acute hypoxia. The two groups are composed of species of catfishes
          (Siluriformes) and cichlids (Perciformes). (Data obtained from
          Almeida-Val et al., 2005)
                                 18

-------
       Glucose mobilization may occur even in facultative air-breathers such as the
armored catfish Glyptoperychthys gibbceps, which is not necessarily related to activation
of anaerobic metabolism, since lactate levels are significantly decreased (Lopes, 2003,
Fig. 6). When denied aerial respiration under hypoxia in laboratory aquaria, the armored
catfish Glyptoperychthys gibbceps, increased gill ventilation rates, but no alteration was
detected in heart rate, suggesting that bradycardia is not one of their strategies against
hypoxia. Thus, it is not possible to affirm that all species respond to hypoxia with similar
metabolic adjustment, i.e., the generalization that anaerobic metabolism is activated and
aerobic metabolism is suppressed during hypoxia is  not true, at least not to all fish species
from the Amazon. Nevertheless, cichlids are uniform in their responses.

       The ability of the organisms to deal with environmental change depends on the
magnitude of the change, the time frame in which the change occurs, and the individual
genetic constitution, which may be altered over generations by the selection of genetic
variants that are better suited to cope with the new environmental situation. As a
consequence, environmental  stress has been considered to be among the most important
triggers of change in biological organization and functioning during evolution (Almeida-
Val etal., 1999). Wilson (1976) called attention to the importance of regulatory genes in
the evolution of plants and animals. This author stated that "although definitive
conclusions are not possible at present, it seems likely  that evolution at the organism
level depends predominantly on regulatory gene mutations. Structural gene mutations
may have a secondary role in organism evolution". So, changes in form, color,
morphology, physiology, and metabolism of many organisms may occur according to
environmental changes and the investigations about the kind of genetic (or metabolic)
control over phenotypes  under different environmental conditions have revealed that
some genes are turned on or off accordingly (Walker, 1979; Smith, 1990; De Jong, 1995;
Land and Hochachka, 1995; Hochachka, 1996; Walker, 1997;  Hochachka etal., 1998).
As we have mentioned on different occasions (Almeida-Val et a/., 1993; 1999), long-
term adaptive responses  to low-oxygen environments involves oxidative metabolic
suppression in fish of the Amazon, as first suggested by Hochachka and Randall (1978)
and corroborated by Driedzic and Almeida-Val (1996) and West etal. (1999). However
the immediate hypoxia responses from fish of the Amazon have been barely studied from
the evolutionary  point of view (Almeida-Val et a/., 1999).

       Oxygen sensing and its physiological and biochemical  consequences in cells are
not fully understood yet, despite the fact that some mechanisms have been extensively
studied in isolated cell models. There are many reviews on this subject (Gracey etal,
2001; Yu etal., 2001). Genes are coordinated and individually regulated during hypoxia
by a variety of hypoxia-responsive transcription factors including HIF-la (Webster,
2003). This system regulates many glycolytic genes, inducing  anaerobic glycolysis and
down regulating many genes of aerobic metabolism. Investigations of this system
indicates that this pathway developed in the Silurian period, 500 MY A, when highly
mobile sea and land species were evolving. In fact, this period is coincident with the high
DNA duplication rates (polyploidy) which induced the radiation of vertebrates and the
appearance of new duplicated genes, giving rise to most gene families and pathway
systems currently known to exist in vertebrates. The gene family of LDH is one of the
                                        19

-------
best studied gene groups, at both the functional and molecular levels. LDH-A* is one of
the genes that is up regulated under hypoxia. Studies with cDNA microarrays in fish
(Gillichthys mirabilis) by Gracey and co-workers (2001) revealed which genes were up-
regulated and which were down-regulated. Based on that, they suggested that hypoxia
survival may involve three molecular strategies: (i) down-regulating genes for protein
synthesis and locomotion to reduce energy consumption; (ii) up-regulating genes for
anaerobic ATP production and gluconeogenesis, and (iii) suppressing cell growth and
channeling energy to essential metabolic processes. Observed changes in gene expression
was tissue-specific and reflected metabolic roles.

       Our first experiments conducted with LDH-A* regulation in Oscar showed that
hypoxia may induce or suppress its expression according to the size of the animal.
Changes in absolute LDH activities may be the result of gene expression and post
translational processes that take place after the protein is synthesized.
The first results of LDH-A* expression in response to different oxygen tensions and
anoxia have showed some variation in LDH-A* mRNA levels (Fig. 7). Changes in LDH-
A* expression may also be related to animal size, tissue sample, and pre-acclimated
condition.
                            LDH (RT-PCR)
                          C   40% 20% 4%  10%
        Figure 7. mRNA expressed for LDH-A* levels in skeletal muscle of
                Oscar juveniles submitted to different levels of hypoxia for
                2 hours. Animals were pre-acclimated to normoxia for 24
                hours.
                                       20

-------
Concluding remarks

       In summary, we conclude that the ability of many fishes of the Amazon to survive
daily extreme changes in oxygen, especially if they spend part of their life cycle in
floodplain areas, known as varzea, is the result of behavioural, physiological,
biochemical, and molecular adaptations to hypoxia. These responses are integrated in
order to respond and survive to environmental oscillations. Nowadays we can recognize
an enormous diversity of adaptations to hypoxia in fish of the Amazon. Air-breathing,
aquatic surface respiration (ASR), and metabolic depression, are all related to the
machinery that connects environmental changes with metabolism through a series of
signals that promote responses at biochemical, physiological, and molecular levels. The
survivorship of fishes exposed to hypoxia and anoxia such as the water-breathing Oscar,
which do not have the morphological changes to breathe air, depends upon the
coordination of aquatic  surface respiration and metabolic depression, affecting a whole
suite of strategies to optimize oxygen use. This species can be used as an anaerobic
model in further studies of oxygen sensing and molecular gene regulation. The
comparison of young and  adults have shown several adaptive responses related to their
size; young animals are less tolerant than adult animals, which may stand anoxia for 6
hours at 28°C. The perfect combination of metabolic depression and activation of
anaerobic metabolism allows the animals to increase their tolerance to hypoxia as they
grow. Anaerobic machinery is activated through a series of gene regulation, which is
similar in all vertebrates. Analysis of LDH-A * gene expression in Oscars exposed to
hypoxia and anoxia revealed that changes in muscle LDH levels are due to gene
regulation, but possible posttranscriptional changes due to other endogenous conditions
such as the substrate amount in muscle tissue still remains unknown.

References

Almeida-Val, V.M.F. and IP. Farias. 1996. Respiration in fish of the Amazon metabolic
       adjustments to chronic hypoxia. Pages 257-271 In: Physiology and Bichemistry of
       the fishes of the Amazon. A.L. Val, V.M.F. Almeida-Val and DJ. Randall (Eds).
       INPA, Manaus.

Almeida-Val, V.M.F., A.L. Val,  and P.W. Hochachka. 1993. Hypoxia Tolerance in
       Amazon Fishes: Status of an Under- Explored  Biological "Goldmine". Pages 435-
       445 In: Surviving Hypoxia: Mechanisms of Control versus Adaptation. P.W.
       Hochachka, G. Van den Thillart and P. Lutz (Eds). CRC Press, Boca Raton.

Almeida-Val, V.M.F., A.L. Val, W.P. Duncan, F.C. Souza, M.N. Paula-Silva, and S.
       Land. 2000. Scaling effects on hypoxia tolerance in the Amazon fish Astronotus
       ocellatus (Perciformes: Cichlidae): contribution of tissue enzyme levels. Comp.
       Biochem. Physiol. 125B: 219-226.
                                       21

-------
Almeida-Val, V.M.F., A.L. Val, and I Walker. 1999. Long- and short-term adaptation of
       Amazon fishes to varying (^-levels: intra-specific phenotypic plasticity and inter-
       specific variation. Pages 185-206 In: Biology of Tropical Fishes. AL. Val and
       V.M.F. Almeida-Val (Eds). INPA, Manaus.

Burleson, M.L., D.R. Wilhelm, andNJ. Smatresk. 2001. The influence offish size on the
       avoidance of hypoxia and oxygen selection by largemouth bass. J. Fish Biol. 59:
       1336-1349.

Chippari-Gomes, A.R. 2002. Adapta9oes metabolicas dos ciclideos aos ambientes
       hipoxicos da Amazonia. Tese de Doutorado, Institute Nacional de Pesquisas da
       Amazonia/ Funda9ao Universidade do Amazonas. Manaus, Amazonas. 148pp.

Crampton, W.G.R.  1998. Effects of anoxia on the distribution, respiratory strategies and
       electric signal diversity of gymnotiform fishes. J. Fish Biol. 53 (Supplement A):
       307-330.

De Jong, G. 1995. Phenotypic plasticity as a product of selection in a variable
       environment. Am. Nat. 145: 493-512.

Driedzic, W.R. and V.M.F. Almeida-Val. 1996. Enzymes of Cardiac Energy Metabolism
       in Amazonian Teleosts and the Fresh-Water Stingray (Potamotrygon hystrix). J.
       Exp. Zool._274: 327-333.

Glass, N.R. 1969. Discussion of calculation of power function with special reference to
       respiratory metabolism in fish. J.  Fish Res. Board Can. 26: 2807-2821.

Gracey, A.Y., J.V. Troll, and G.N.  Somero. 2001. Hypoxia-induced gene expression
       profiling in the euryoxic fish Gillichthys mirabilis. Proc. Natl. Acad. Sci. USA
       98(4): 1993-1998.

Hammer, C. and M. Purps. 1996. The metabolic exponent of Hoplostermm littorale in
       comparison with Indian air breathing catfish, with methodological investigation
       on the nature of metabolic exponent. Pages 283-297 In: Physiology and
       Biochemistry of the fishes of the Amazon, A.L. Val and V.M.F. Almeida-Val
       (Eds). INPA, Manaus.

Hochachka, P.W. 1996. Oxygen sensing  and metabolic regulation: short, intermediate,
       and long term roles. Pages 233-256 In: Physiology and Biochemistry of the fishes
       of the Amazon, A.L. Val, V.M.F. Almeida-Val and D.J. Randall (Eds). INPA,
       Manaus.

Hochachka, P.W., G.B. McClelland, G.P. Burness, J.F. Staples, R.K. Suarez. 1998.
       Integrating metabolic pathway fluxes with gene-to-enzyme expression rates.
       Comp. Biochem. Physiol. 120B: 17-26.
                                       22

-------
Hochachka, P.W. and DJ. Randall. 1978. Alpha-Helix Amazon expedition, September-
       October 1976. Can J. Zool. 56: 713-716.

Junk, W.J., G.M. Scares, P.M. Carvalho. 1983. Distribution offish species in a lake of
       the Amazon river floodplain near Manaus (Lago Camaleao), with special
       reference to extreme oxygen conditions. Amazoniana 7(4): 39-431.

Land, S.C. and P.W. Hochachka. 1995. A heme-protein-based oxygen-sensing
       mechanism controls the expression and suppression of multiple protein in anoxia-
       tolerant turtle hepatocytes. Proc. Natl. Acad. Sci. USA 92:  7505-7509.

Lopes, N.P. 2003. Ajustes metabolicos em sete especies de Siluriformes sob condi9oes
       hipoxicas:  aspectos adaptativos. Biologia de Agua Doce e Pesca Interior, PIPG-
       BTRN. INPA/UFAM, Manaus: 168.

Muusze, B., J. Marcon, G. van den Thillart, and V. Almeida-Val. 1998. Hypoxia
       tolerance of Amazon fish: respirometry and energy metabolism of the cichlid
       Astronotus ocellatus. Comp. Biochem. Physiol. 120A: 151-156.

Randall, D.J., W.W. Burggren, A.P. Parrel, and M.S. Haswell.  1981. The evolution of
       air-breathing vertebrates. Cambridge University Press, Cambridge,.

Robb, T. and M.V. Abrahams. 2003. Variation in tolerance to hypoxia in a predator and
       prey species: an ecological advantage of being small? J. Fish Biol. 62: 1067-1081.

Sloman, K.A., C.M. Wood, G.R. Scott, S. Wood, M. Kajimura, O.E. Johannsson, V.M.F.
       Almeida-Val, and A.L. Val. 2006. Tribute to R.G. Boutilier: The effect of size on
       the physiological and behavioural responses of Oscar, Astronotus ocellatus, to
       hypoxia. J. Exp. Biol. 209: 1197-1205.

Smith, H. 1990. Signal perception, differential expression within multigene families and
       the molecular basis of phenotypic plasticity. Plant Cell Environ. 13: 585-594.

Val, A.L. and V.M.F. Almeida-Val. 1995. Fishes of the Amazon and their environments.
       Physiological and biochemical features. Springer Verlag, Heidelberg.

Walker, I. 1979. The mechanical properties of proteins determine the laws of
       evolutionary change. Acta biotheor. 28: 239-282.

Walker, I. 1997. Prediction or Evolution? Somatic plasticity as a basic, physiological
       condition for the viability of genetic mutations. Acta biotheor. 44: 165-168.

Webster, K.A. 2003. Evolution of the coordinate regulation of glycolytic enzyme genes
       by hypoxia. J. Exp. Biol. 206: 2911-2922.
                                       23

-------
West, J.L., J.R. Bailey, V.M.F. Almeida-Val, A.L. Val, B.D. Sidell, and W.R. Driedzic.
       1999. Activity levels of enzymes of energy metabolism in heart and red muscle
       are higher in north-temperate-zone than in Amazonian teleosts. Can. J. Zool.
       77(5): 690-696.

Wilson, A.C. 1976. Gene Regulation in Evolution. Pages 225-235 In: Molecular
       Evolution. F. J. Ayala (Ed.). Sinauer Associates Inc., Sunderland.

Yu, F., S.B. White, Q. Zhao, and F.S. Lee. 2001. HIF-la binding to VHL is regulated by
       stimulus-sensitive proline hydroxylation. Proc. Natl. Acad. Sci. USA 98: 9630-
       9635.
                                       24

-------
    Swimming performance as a practical and effective
            biomarker of pollution  exposure in fish

                                      by

                        E.W. Taylor1 and DJ. McKenzie1'2

Introduction

       It is commonly accepted that physiological adaptation by fishes to their
environment will influence the success with which they can colonise particular habitats
(Fry, 1947; 1971; Prosser, 1950). Fish species that pursue an active lifestyle perform
sustained aerobic exercise to forage, to migrate and to maintain position against currents.
This requires the coordinated activity of systems at various levels of organismal
organisation  (Brett, 1958; Randall, 1982; Moyes and West, 1995) and a single unifying
trait known as maximum sustainable aerobic swimming speed (Ucrit, as conceived by
Brett, 1964) has proven to be sensitive to many environmental stressors (Randall and
Brauner 1991) including pollutants such as low pH (Ye and Randall, 1991; Butler etal.,
1992), dissolved metals (e.g. Waiwood and Beamish, 1978; Wilson et a/., 1994;
Beaumont etal, 1995a,b; 2003), ammonia (Shingles etal, 2001; Wicks etal, 2002;
McKenzie et al, 2003), and various other toxic chemicals such as organo-phosphate
pesticides (Peterson, 1974).

       Another complex trait that has been proposed as a potentially valuable indicator
of sub-lethal  pollution is routine aerobic metabolic rate, measured as rates of oxygen
consumption (Sprague, 1971; Fry,  1971; Rice, 1990). Metabolic rate can be considered
the unifying currency of adaptation to the environment (Wikelski and Rickleff, 2001) and
can be linked to the increased energetic costs associated with occupying polluted habitats
(Rice, 1990). There is much evidence to indicate that increased metabolic rate is a general
indicator of stress in fish (Wendelaar Bonga, 1997) and studies have shown it to be raised
by exposure to various pollutants such as organophosphate pesticides (Holmberg and
Saunders, 1979; Farrell etal, 1998), methylmercury (Rodgers and Beamish, 1981) and
various specific herbicides (Johansen and Geen, 1990; Janz et al, 1991).

       The literature investigating the effects of pollutants upon fish exercise
performance and metabolic rates is, however, almost exclusively laboratory-based, and
mainly comprises the exposure of salmonid species to single toxicants. Relatively little is
known about the potential physiological effects of exposure to polluted natural
environments (Farrell et al, 2004). Many aquatic habitats are continuously loaded with
mixtures of chemicals released by human communities and industries, and many of these
pollutants have been shown to exert adverse impacts upon the resident biota. The last few
decades have, therefore, seen an increasing interest in the use of "biomarkers" to
1 School of Biosciences, University of Birmingham, Birmingham, United Kingdom.
2 CNRS UMR 5171, Station Mediterraneenne de I'Environnement Littoral, Sete, France.
                                      25

-------
establish early-warning signals of exposure and toxic effects of specific pollutants or
pollutant classes (reviewed by van der Oost et al., 2003). This term is most commonly
used to refer to measurements in body fluids, cells or tissues, which are indicative of
bioaccumulation of toxic chemicals, biochemical and cellular modifications provoked by
specific toxicants, or secondary responses of host tissues to these toxicants (van der Oost
et al., 2003). The assumption is that such modifications at these lower orders of
biological organisation are indicative  of, or directly linked to, modifications in systemic
and organismal function which, in turn,  lead to changes in the populations and
communities that comprise the ecosystem.

       These assumptions, and in particular the link between expression of such
biomarkers and the functional integrity of the whole organism remain, however, to be
proven (van der Oost et al., 2003).  Both swimming performance and metabolic rate may
be valuable in this sense because they directly reflect the functional integrity offish and
are also of immediate relevance to their ecology (Sprague, 1971; Rice, 1990; MacKinnon
andFarrell, 1992).

       The current study used custom-built portable swim-tunnel respirometers to
compare exercise performance (Ucrit) and associated aerobic metabolism offish exposed
in cages for three to four weeks at either clean or polluted sites on three urban European
river systems in different seasons (spring, summer, and winter). The rivers studied were
the Lambro (Milan, Italy), the Blythe/Cole/Tame confluence (Birmingham, United
Kingdom), and the Amstel (Amsterdam, The Netherlands). Two species of cyprinid were
chosen as models, the chub (Leuciscus cephalus) was studied in Italy  and the UK
whereas carp (Cyprinus carpio) were  studied in the Netherlands. Jain et al. (1998)
demonstrated that the ability of the fish to perform two sequential exercise tests with  a
brief intervening recovery interval could provide significantly more sensitive information
about fish health and water quality than a single exercise test alone. This "repeat-
exercise" protocol was, therefore, adopted in the current study. Measurements were made
of routine rates of oxygen uptake under  standardised conditions of sub-maximal exercise,
to investigate sensitivity of this trait to the prevailing water chemical quality. Other
metabolic traits, such as the maximum metabolic rate, measured as oxygen uptake, during
exercise and aerobic metabolic scope, were also derived during the swim tests, to gain
insight into proximate physiological mechanisms that underlie impaired exercise
performance (Beamish, 1978; Wilson etal.,  1994; McKenzie et al., 2003; Pane et al.,
2004).

Methods

River sites

       Italy

       The river Lambro rises in the foothills of the Alps and flows in a southerly
direction through the provinces and cities of Monza and Milan, which are major centres
of population and industry. Two sites were studied: a relatively clean  upstream site at
                                        26

-------
Merone and a polluted downsteam site just north of the city of Milan at Brugherio. Water
sampling campaigns used diffusive gradient in thin film (DGT) and semi-permeable
membrane devices (SPMD) passive samplers to monitor total bioavailable heavy metal
and organic pollutants respectively (Garofalo etal., 2004; Garofalo and Ceradini,
unpublished observations). The Brugherio site was polluted by bioavailable copper,
nickel and zinc plus bioavailable organics such as polycyclic aromatic hydrocarbons
(PAHs), polychlorinated biphenyls (PCBs) and organochlorine pesticides (OCPs). There
were no differences in water temperature, pH, dissolved oxygen and conductivity
between sites. During the reported campaign in September 2001 water temperature varied
closely around 20°C.

       United Kingdom

       The confluent Blythe, Cole and Tame rivers lie within the Birmingham
conurbation in the West Midlands, a major centre of industry and population with a
consequent legacy of pollution. They have a history of relatively good, intermediate and
poor chemical water qualities respectively (Winter etal., 2004; 2005). The Tame is
polluted with a complex mixture of metals, largely copper, nickel and zinc; the Cole was
also polluted with these metals but to a lesser extent, whereas the Blythe exhibited only a
low level of bioavailable zinc (Garofalo etal., 2004). Sampling of bioavailable organics
with SPMDs revealed that the Tame had four to five fold higher levels of PAHs, PCBs
and OCPs than the Blythe, which had low levels of all these contaminants. The Cole
exhibited similar levels of PAHs to the Tame but levels of PCBs and OCPs intermediate
between the Tame and the Blythe (Winter et al., 2005). During the sampling campaign in
June/July, 2002 water temperatures were 13-15°C.

       The Netherlands

       The Amstel River is 40 km long, running south to north from a rural area into and
through the centre of Amsterdam where it becomes heavily polluted by industrial and
domestic effluents . A number of study sites on the river have been described in detail by
Verweij et al. (2004). Three were identified for the current investigation: a clean site in
the southern rural area,  at Vrouwenakker; a polluted site in downtown Amsterdam and a
heavily polluted landfill site, Volgermeerpolder. These sites presented very little
bioavailable heavy metal pollution (Garofalo et al., 2004) but extremely high levels of
organic pollution (Verweij et a/., 2004). Water temperature was 18-19°C.

Experimental animals

       Two species of cyprinid were studied. The chub, Leuciscus cephalus, is common
in running water throughout Europe and is native to the rivers that were studied in Italy
and the UK. It is an omnivorous species, eating invertebrates and some plant material
when young and becoming an active piscivore as it grows to maturity. The carp, Cyprinus
carpio, is native to Eastern Europe and Asia but has been introduced in many other
countries, including the Netherlands. It occupies ponds, lakes and slow-flowing rivers
                                       27

-------
such as the river Amstel. It is also an omnivorous species, feeding on bottom-dwelling
invertebrates and plant material.

      Italy

      Wild chub were captured by electrofishing in the river Lambro, at the clean
Merone site. The fish were transported live to the La Casella Fluvial Hydrobiology
Station (via Argine del Ballottino, 29010 Sarmato [PC], Italy), where they were stocked
in 4m2 fibreglass tanks provided with a flow of water within a recirculating biofiltered
system (vol. 90 m3), at prevailing seasonal ambient temperatures and photoperiods. The
animals were maintained under these conditions for two weeks and fed daily with
pelleted feed prior to use in any experiments.

      United Kingdom

      Farm-reared chub were obtained  from the Environment Agency fish farm
(Calverton, Nottinghamshire) and transported to the animal holding facilities at the
School of Biosciences, University of Birmingham. The fish were held in laboratory
aquaria in biofiltered dechlorinated Birmingham tapwater for at least two weeks prior to
use in experiments, at prevailing seasonal temperatures and photoperiod and were fed
daily with a pelleted feed.

      The Netherlands

       Caging experiments (see below) were performed with genetically identical male
carp from a cultured Fl hybrid fish line produced and maintained at the Agricultural
University of Wageningen (van der Oost, 1998; Verweij etal., 2004). This experimental
group offered the clear advantage of reduced variability between individuals in the
toxicity  tests.

Caging exposures

        The caging offish on the bed of the river Lambro in Italy was described by
McKenzie et al. (2006) and for the rivers in the UK by Winter et al. (2005). Briefly, two
cages were transported to each river site  and anchored on the riverbed, in areas of gentle
flow. Ten chub were placed in each  cage for at least 3 weeks during which time they had
access to both the water column and the  riverbed. The fish were observed visually to feed
upon naturally  available food. Data are reported for caging experiments performed in
Italy in September 2001 and in the UK in June/July 2002. In the Netherlands, caging
experiments were performed as described by van der Oost et al. (1998). Briefly, twenty
carp were placed in submerged cages anchored in the water column at each site, for at
least 3 weeks. No attempt was made to feed the fish during the exposure  protocol. Data
are reported for a caging experiment performed in September 2002 (summer).
                                       28

-------
Exercise respirometry

       Exercise performance and metabolism were measured with custom-built portable
swimming respirometers, designed to exercise individual fish in a non-turbulent water
flow with a uniform velocity profile (Steffensen et al., 1984). One respirometer
constructed of PVC, with a cross-sectional area of 225 cm2 to the swimming chamber and
a respirometric volume of 49.0 L, has been described in detail previously (McKenzie et
a/., 2001). The second was of similar design but had a cross-sectional area of 100 cm2 to
the swim chamber, a respirometric volume of 13.4 L and was constructed of Plexiglass.
In each, water flow was generated by a stainless steel propeller attached to a variable
speed DC permanent magnet motor. Motor speed was controlled by a PC and Labview
software (National Instruments Inc.), calibrated to deliver water velocities in cm s"1 and,
hence, swimming speeds corrected for the solid blocking effect of the fish (Bell and
Terhune, 1970). The respirometer chambers were thermostatted by immersion in larger
outer tanks that received a constant flow of the appropriate source of water.

       Use of these respirometers in the field was described in detail by McKenzie et al.
(2006). Fish were transferred to the respirometer without air-exposure, then permitted 4h
recovery from handling stress while swimming gently at a current speed of 20 cm s"1,
prior to testing their repeat swimming performance (Jain et a/., 1998). All fish were
exercised by progressive increments in swimming speed of 10 cm s"1 every 30 min until
fatigue. Fatigue was unambiguous in both the chub and the carp, they swam vigorously
until they collapsed against the back screen and would not resume in response to gentle
prodding or further increases in current velocity. The fish were then allowed 40 min
recovery from the first swim test (Tl) after which they were exposed to exactly the same
protocol a second time (T2). Maximum sustainable aerobic  swimming speed (Ucrit) for
both Tl and T2 was calculated in BL s"1  as described by Brett (1964). The repeat
performance ratio was calculated as T2/T1 (Jain et a/., 1998).

       Instantaneous C>2 uptake (Mo2) was measured at each swimming speed by
intermittent flow-through respirometry (Steffensen, 1989) controlled by a PC and
Labview software as described in detail by McKenzie et al.  (2001). For both Tl and T2,
active (maximum) metabolic rate (AMR) was estimated as the M02 measured at highest
swimming speed immediately prior to fatigue (Fry, 1971; Beamish, 1978). For both Tl
and T2, an estimate of functional aerobic scope for activity was also calculated by
subtracting rates of 62 uptake measured  at the lowest swimming speed (20 cms"1, defined
as routine metabolic rate, RMR) from the measured AMR.

Statistics

       To analyse the effect of the repeated exercise protocol upon measured variables,
these were compared between groups by two-way analysis of variance (ANOVA) for
repeated samples, where the interacting factors were the group (i.e. caging sites for the
field studies) and the repetition of the exercise protocol. To  compare the repeat
performance ratio between groups a T-test was used to compare the two Italian sites,
whereas a one-way ANOVA was used to compare three sites in the UK and the
                                       29

-------
Netherlands. Holm-Sidak post-hoc tests were used to identify differences amongst means.
In all cases, p< 0.05 was taken as the fiducial level for statistical significance.
             'w
                        D Merone
                        • Brugherio
                 0.0
                                               1.2
                                               1.0
                                               0.8
                                               0.6
                                               0.4
                                               0.2
0.0
I






t
T



Mer Brug
River Site
                         T1        T2
                           Swim Test

       Figure 1.   Swimming performance of chub (Leuciscus cephalus)
                 following exposure in submerged cages to two sites on the
                 river Lambro, Italy, in summer 2001.  The graphs show mean
                 (± SEM) critical swimming speed (Ucrit) as measured twice,
                 with the second swim test (T2) measured 40 min following
                 fatigue in the first swim test (Tl), and the corresponding repeat
                 performance ratio (Ucrit T2 / Ucrit Tl). n = 6 in all cases, *
                 denotes a significant difference between Tl and T2 for that
                 river site, ^ denotes a significant difference between the two
                 sites for the relevant variable, Mer, Merone; Brug, Brugherio.

Results

       Italy

       Figure 1 shows the swimming performance of chub in Merone and Brugherio, in
late summer at a water temperature of 20°C. At both sites, the fish (chub, mean mass 200
± 25g) swam equally well in Tl. The fish from the clean site at Merone were able to
repeat this performance in T2 but those from the polluted site at Brugherio were not, and
exhibited a significant decline in their Ucrit. As a result, the fish from Merone had a
repeat ratio that was not significantly different from 1, whereas the fish from Brugherio
had a significantly lower ratio of around 0.7 (Fig. 1).

       Figure 2 shows the respirometric measures taken during these swim tests.  Prior to
Tl, RMR was not significantly different between Merone and Brugherio. In both groups,
Tl caused a significant increase in metabolic rate, with AMR and functional aerobic
scope for activity being similar in both groups (Fig. 2).  During T2 the fish from Merone
                                       30

-------
showed a similar AMR to that in Tl and, therefore, a functional aerobic scope in T2 that
was similar to that measured in Tl. The chub from Brugherio exhibited a reduced AMR
during T2, and a highly significant decline in functional aerobic scope for activity in T2
when compared with Tl (Fig. 2).
20 -,
5 15 -
"5
E
.§ 10 -
CD
cc
-l-»
Q_
3
CM 5 -
O

n









River Lambro, Merone 20 n
1 I ^ 15 "


*
T

^
I !
I I






















O
I s. 10 -










1





CD
3
Q.
3
CM 5 -
0

n
River Lambro, Brugherio



1




- I










I








I







1


*t
111 T

                   RMR
AMR
ASA
RMR
AMR
ASA
       Figure 2. Swimming respirometry of chub (Leuciscus cephalus) following
                 exposure in submerged cages to two sites on the river Lambro,
                 Italy, in summer 2001. The graphs show mean (± SEM) routine
                 metabolic rate (RMR) offish swimming steadily at 20 cm s"1 prior
                 to the exercise challenge; the active metabolic rate achieved during
                 exercise (AMR) and the functional aerobic scope for activity
                 (ASA) calculated as AMR-RMR. Each variable was measured
                 twice, with the second swim test (T2, filled columns) measured 40
                 min following fatigue in the first swim test (Tl, open columns), n
                 = 6 in all cases, * denotes a significant difference between Tl and
                 T2 for that river site, ^ denotes a significant difference between the
                 two sites for the relevant variable.

       United Kingdom

       Figure 3 shows the swimming performance of chub (mean mass 62 ± 7g) caged in
the Blythe, Cole and Tame, as measured in summer at water temperatures between 13
and 15 °C. At all sites, the fish swam equally well in Tl. The fish from the clean Blythe
site were able to repeat this performance in T2. Those from the polluted Cole and Tame
sites were not, and exhibited a significant decline in their Ucrit relative to Tl. As a result,
the fish from the Blythe had a repeat ratio that was not significantly different from 1,
whereas the fish from the Cole and Tame had a significantly lower ratio of around 0.7.
                                       31

-------

                                          1.2 -i
                                      CM
                                       CD
                                       a:
                                       -i—<
                                       CD
                                       CD
                                       O.
                                       CD
                                       a:
       O
           0.0
0.4
                                          0.2
0.0
                   T1         T2
                     Swim Test
     Ely  Col  Tarn
        River Site
Figure 3. Swimming performance of chub (Leuciscus cephalus) following
          exposure in submerged cages to three river sites on the
          confluent Blythe, Cole and Tame rivers in the UK, in summer
          2002. The graphs show mean (± SEM) critical swimming
          speed (Ucrit) as measured twice, with the second swim test
          (T2) measured 40 min following fatigue in the first swim test
          (Tl), and the corresponding repeat performance ratio (Ucrit T2
          / Ucrit Tl). n = 6 in all cases, * denotes a significant
          difference between Tl and T2 for that river site; Ely, Blythe;
          Col, Cole; Tarn, Tame.
                                 32

-------
25 -,
T^ 20
.c
O)
^x"
o 15
E
,§
"S.
(N
0 5
n
Rive


I
1

















r Blythe







i
























i



25 -,
-P 20
.c
O)
^x"
o 15 -
E
,§
1 10 "
"a.
(N
0 5
n
River
*


t
i









Cole 25 -i
L r°"







I

o 15 -
I
1 10-
•5.
* ^
— L n
RiverTame

t
I


I





I



I






ft
          RMR  AMR ASA
RMR AMR ASA
RMR AMR  ASA
       Figure  4.   Swimming respirometry of chub following exposure in submerged
                 cages to three river sites on the confluent Blythe, Cole and Tame rivers
                 in the UK, in summer 2002. Graphs show mean routine metabolic rate
                 (RMR) of fish swimming steadily at 20 cm s"1  prior to the exercise
                 challenge; active metabolic rate achieved during exercise (AMR) and
                 functional aerobic  scope for  activity (ASA).  Each  variable  was
                 measured  twice, with  the  second swim test (T2, filled columns)
                 measured 40  min following fatigue in the first  swim test (Tl, open
                 columns),   n =  6 in all cases, * denotes  a significant difference
                 between Tl and T2 for a river site, ^ denotes a significant difference
                 between the Tame and the other sites for that variable.

       Figure 4 shows the respirometric measures taken during these performance tests.
Prior to Tl, the fish from the Tame exhibited elevated RMR relative to those from the
Blythe. All fish, however, showed a statistically similar AMR and functional aerobic
scope in Tl, although the mean value appears visibly lower in the Tame, where there was
much variability in exercise metabolism amongst the fish. In T2, chub from  the Blythe
achieved similar AMRs and functional aerobic scopes to those measured in Tl. The fish
from the Cole and Tame, however, showed significant declines in their functional scope
in T2 relative to Tl (Fig. 4).

       The Netherlands

       There were no differences in performance between the carp caged in the
Vrouwenakker, Amsterdam and Volgermeerpolder sites with different pollution status
(data not shown). All animals exhibited a repeat performance ratio that was  statistically
identical to 1. Despite the absence of any differences in swimming performance between
the groups, there were differences in their respiratory metabolism (Fig. 5). Prior to Tl the
groups differed in their RMR, which was lower at the Vrouwenakker and Amsterdam
sites when compared to the severely polluted Volgermeerpolder site. Despite this
                                       33

-------
apparent metabolic loading in the animals in the heavily polluted Volgermeerpolder site,
there were no differences in AMR or functional aerobic scope for activity at any site, in
either Tl or T2.
20 -i Vrouwenakker 20 n Amsterdam

_—
.C 1 ^ -
1 \J
'O)
"o
E
£ 10
o>
~m
~Q.
d 5
n









t
I







*t

T

Tr^










4i
in;j«
i|j;5«

lli=o<
|i!i = 0<
|i|i!0<



lli = 0<
|i|i!0<

lli = 0<
|i!i = 0<
|i|i!0<
lli = 0<
Ijij1'^










_—

u I1"









1
I


I







o
E
£ 10
o>
Is
"S.
6 5
n

TT c





*
T
t
•^r
,10





















I









^_
O)
^x"
ilil _ E







111
111



T E,





O)
ro
Q.
(N
O

                                                           20 -\
                                                           10
                                                            5
Volgermeerpolder


T
JL







I




I

            RMR AMR  ASA
RMR AMR  ASA
RMR AMR  ASA
       Figure 5.  Swimming respirometry of chub following exposure in submerged
                 cages to three river sites on the confluent Blythe, Cole and Tame
                 rivers in the UK, in summer 2002. Graphs show mean routine
                 metabolic rate (RMR) offish swimming steadily at 20 cm s"1 prior to
                 the exercise challenge; active metabolic rate achieved during
                 exercise (AMR) and functional aerobic scope for activity (ASA).
                 Each variable was measured twice, with the second swim test (T2,
                 filled columns) measured 40 min following fatigue in the first swim
                 test (Tl, open columns), n = 6 in all cases, * denotes a significant
                 difference between Tl and T2 for that river site, ^ denotes a
                 significant difference from the Volgermeerpolder site for the
                 relevant variable.

Discussion

       The caging studies demonstrated that measured values of the exercise
performance and metabolic rate offish vary with the recorded conditions in their sites of
exposure so that these indices of physiological performance can be used to demonstrate
sub-lethal effects of the complex mixtures of chemicals which prevail in polluted urban
rivers. In both Italy and the UK, chub exposed to sites polluted with bioavailable heavy
metals (primarily Cu, Ni,  and Zn) and organics (PAHs, PCBs, and OCPs) exhibited
impairments to their exercise physiology, revealed as a reduced ability to repeat their
swimming performance in a standard Ucrit test. Such impairments were not observed in
the carp exposed in the Netherlands to sites that were heavily polluted with bioavailable
organics. In both chub and carp, however, there was evidence of metabolic disruption
following exposure to polluted sites, with fish exhibiting elevated RMR. However, there
                                        34

-------
was no measurable effect of water chemical quality upon exercise performance or
exercise metabolism of chub during the winter campaigns in both Italy and the UK
(McKenzie etal., 2006), possibly because the sub-lethal toxic effects of the pollutants at
each site were less pronounced when physico-chemical activities and fish metabolism
were depressed by low water temperatures (Beaumont et al., 1995b; Taylor et al., 1996).
These results indicate that traits of performance such as the ability to repeat a swim test,
and traits of metabolism  such as routine metabolic rate during sustained low-intensity
aerobic exercise, can be employed as physiological biomarkers of sub-lethal aquatic
pollution. This utility may be confined to spring and summer, when pollution events are
likely to be most critical, though more work is required to examine responses to winter
spates when maximum swimming speeds may be exceeded.

              The swimming performance studies upon the chub in Italy and the UK
provide strong support for the  proposal of Jain etal. (1998) that a protocol of repeated
exercise performance can provide more sensitive information about fish health and water
quality than a standard single Ucrit test. Evidence that this is the case and the proximate
mechanisms underlying the impaired repeat performance in chub exposed to polluted
river sites was discussed by McKenzie et al. (2006).

       In brown trout, Salmo trutta, copper impairs performance by interfering with
ammonia excretion, causing plasma ammonia accumulation and a consequent
depolarisation of white muscle (Beaumont et al., 2000a,b),  which compromises white
muscle recruitment and therefore swimming performance at the highest speeds
(Beaumont et al., 2003; McKenzie et al.,  2003). Zinc  may have a similar mode of action
(Alsop etal., 1999). The absence of any impairments  to the exercise performance of carp
exposed to the polluted sites on the river Amstel may  have been due to the absence of
significant bioavailable metals. An impact of pollution was, however, visible in both
chub and carp as an increase in metabolic rate during  low-level sustained aerobic exercise
at a water speed of 20 cm s"1 (defined as RMR for the purposes of the current study).  The
elevated RMR that was measured prior to Tl in the chub and carp exposed to polluted
sites may have derived from a metabolic load imposed upon fish exposed to polluted
sites. All of the polluted  sites at which fish exhibited elevated RMR also had significant
bioavailable OCP levels, particularly the Vogermeerpolder site in the Netherlands
(Verweij et al., 2003; Winter et al., 2005; Garofalo and Ceradini, unpublished
observations). Exposure to pentachlorophenol leads to profound increases in routine
oxygen consumption in the American eel, Anguilla rostrata (Holmberg and Saunders,
1979) and in sockeye salmon (Farrell et al., 1998). There is also quite a large body of
evidence to suggest that elevated metabolic rates in fish can be an indicator of chronic
aspecific stress (Schreck,1990; Wendelaar Bonga, 1996).

       Whatever the mechanisms by which exposure  to polluted river sites caused a
reduced ability to repeat strenuous exercise and/or elevated routine metabolic rates, these
traits do seem to offer some potential as physiological biomarkers of sublethal toxic
stress.  McKenzie et al. (2006) considered the extent to which these measures of
swimming performance satisfied the set of six criteria used to evaluate the strengths or
weaknesses of a biomarker, proposed by Stegeman et al. (1992) and van der Oost et al.
                                       35

-------
(2003). They concluded that most were met so that this approach can be recommended to
agencies responsible for maintaining the health of rivers and their fish populations.
Acknowledgements

       This work was funded by a contract from the Commission of the European
Community (EVK-CT1999 - 00009) in support of a project entitled: Ecological Quality
of Urban Rivers: Environmental Factors Limiting Restoration of Fish Populations, and
codenamed "CITYFISH". The CITYFISH team comprised: Birmingham: Ted Taylor,
David McKenzie, Matt Winter, Ruth Hayes, Norman Day, Pat Butler, Kevin Chipman;
Milan: (CESI) S. Ceradini, E. Garofalo, B. Stoppelli; Amsterdam: (OMEGAM) F.
Verweij, R. van der Oost; Berlin: (Technical University) E. Unruh, P Hansen. We are
particularly indebted to the late Mr Ted Betham-Evans, a member of the mechanical
workshop team in Birmingham and to his team leader, the late Chris Hardeman, who
together designed and built the swimming respirometers to our specifications. Our work,
using this equipment, is dedicated to their memories and to that of our colleague the late
Frank Verweij.

References

Alsop, D.H., J.C. McGeer, D.G. McDonald, and C.W. Wood. 1999. Costs of chronic
       waterborne zinc exposure and the consequences of zinc acclimation on the
       gill/zinc interactions of rainbow trout in hard and soft water. Environ. Toxicol.
       Chem. 18: 1014-1025.

Beamish, F.W.H. 1978. Swimming Capacity. Pages 101-187 In: Fish Physiology Vol.
       VII. W.S. Hoar and DJ. Randall (Eds.). Academic Press, New York.

Beaumont, M.W., PJ. Butler, and E.W. Taylor. 1995a. Exposure of brown trout, Salmo
       trutta, to sub-lethal copper concentrations in soft acidic water and its effects upon
       sustained swimming performance. Aquatic Toxicol. 33:  45-63.

Beaumont, M.W., PJ. Butler, and E.W. Taylor. 1995b. Plasma ammonia concentration in
       brown trout (Salmo truttd) exposed to acidic water and sublethal copper
       concentrations and its relationship to decreased swimming performance. J. Exp.
       Biol.  198: 2213-2220.

Beaumont, M.W., E.W. Taylor, and PJ. Butler. 2000a. The resting membrane potential
       of white muscle from brown trout (Salmo truttd) exposed to copper in soft, acidic
       water. J. Exp.  Biol.  203: 2229-2236.

Beaumont, M.W., P J. Butler, and E.W. Taylor. 2000b. Exposure of brown trout, Salmo
       trutta, to a sub-lethal concentration of copper in soft acidic water: effects upon
       muscle metabolism and membrane potential. Aq. Toxicol. 51: 259-272.
                                      36

-------
Beaumont, M.W., PJ. Butler, and E.W. Taylor. 2003. Exposure of brown trout, Salmo
       trutta, to a sub-lethal concentration of copper in soft acidic water: effects upon
       gas exchange and ammonia accumulation. J. Exp. Biol. 206:  153-162.

Bell, W.H. and L.D.B. Terhune. 1970. Water tunnel design for fisheries research. Fish.
       Res. Bd. Can. Tech. Report 195: 1-169.

Brett, J.R. 1958. Implications and assessments of environmental stress. Pages 69-93 In:
       The Investigation of Fish-Power Problems. P. A. Larkin  (Ed.). University of BC:
       Institute of Fisheries.

Brett, J.R. 1964. The respiratory metabolism and swimming performance of young
       sockeye salmon. J. Fish. Res. Board. Can. 21: 1183-1226.

Butler, P.J., N. Day, and K. Namba. 1992. Interactive effects of seasonal temperature and
       low pH on resting oxygen uptake and swimming performance of adult brown
       trout Salmo trutta. J. Exp. Biol. 165: 195-212.

Farrell, A.P., A.K. Gamperl, and I.K.  Birtwell. 1998. Prolonged swimming, recovery and
       repeat swimming performance of mature sockeye salmon Oncorhynchus nerka
       exposed to moderate hypoxia and pentachlorophenol. J. Exp. Biol. 201: 2183-
       2193.

Farrell, A.P., CJ. Kennedy, and A. Kolok. 2004. Effects of wastewater from oil refining
       operations on survival, haematology, gill histology and swimming performance of
       fathead minnows.  Can. J. Zool. 82: 1519-1527.

Fry, F.E.J. 1947. The effects of the environment on animal activity.  Univ. Toronto Stud.,
       Biol. Ser. 55: 1-62.

Fry, F.E.J. 1971. The effect of environmental factors on the physiology offish. Pages 1-
       98 In: Fish Physiology Vol. VI. W.S. Hoar and D.J. Randall  (Eds.). Academic
       Press, New York.

Garofalo, E., S. Ceradini, and M.J. Winter. 2004. The use of diffusive gradients in thin-
       film (DGT) passive samplers for the measurement of bioavailable metals in river
       water. Annali di Chimica 94: 515-20.

 Holmberg, B. andR.L.  Saunders. 1979. The effects of pentachlorophenol on swimming
       performance and oxygen consumption in the American eel (Anguilla rostratd).
       Rapp. P.-v. Reun. Cons. Int. Explor. Mer. 174: 144-149.

Jain, K.E., I.K. Birtwell, and A. Farrell. 1998. Repeat swimming performance of mature
       sockeye salmon  following a brief recovery period: a proposed measure offish
       health and water quality. Can. J. Zool. 76: 1488-1496.
                                       37

-------
Janz, D.M., A.P. Farrell, J.D. Morgan, and G.A. Vigers. 1991. Acute physiological stress
       responses of juvenile coho salmon (Oncorhynchus kisutch) to sub-lethal
       concentrations of Garlon-4, Garlon-3A and Vision herbicides. Environ. Toxicol.
       Chem. 10: 81-90.

Johansen, J.A. and G.H. Geen. 1990. Sublethal and acute toxicity of the ethylene glycol
       butyl ether ester formulation of triclopyr to juvenile coho salmon (Oncorhynchus
       kisutch). Arch. Environ. Contam. Toxicol. 19: 610-616.

McKenzie, D.J., E. Cataldi, S. Owen, E.W. Taylor, and P. Bronzi. 2001. Effects of
       acclimation to brackish water on the growth, respiratory metabolism and exercise
       performance of Adriatic sturgeon (Acipenser naccarif). Can. J. Fish. Aquat. Sci.
       58: 1104-1112.

McKenzie, D.J., A. Shingles, and E.W. Taylor. 2003. Sub-lethal plasma ammonia
       accumulation  and the swimming performance of salmonids. Comp. Biochem.
       Physiol.  135A: 515-526.

MacKinnon, D.L. and A.P. Farrell. 1992. The effects of
       2(thiocyanomethylthio)benzothiazole on juvenile coho salmon (Oncorhynchus
       kisutch)  - sublethal toxicity testing. Environ. Toxicol. Chem. 11: 1541-1548.

Moyes, C.D. and T.G. West.  1995. Exercise metabolism offish. Pages 367-392 In:
       Biochemistry  and Molecular Biology of Fishes, Vol. 4. P.W. Hochachka and T.P.
       Mommsen (Eds.). Elsevier Science, Amsterdam.

Pane, E.F., A. Haque, G.G. Goss, and C.W. Wood. 2004. The physiological
       consequences of exposure to chronic, sub-lethal waterborne nickel in rainbow
       trout (Oncorhynchus mykiss): exercise vs resting physiology. J. Exp. Biol. 207:
       1249-1261.

Petersen, R.H. 1974. Influence of fenitrothion on swimming velocities of brook trout
       (Salvelinusfontinalis). J. Fish. Res. Board.  Can. 31: 1757-1762.

Prosser, C.L. 1950. Comparative Animal Physiology. 1st edn. Philadelphia and London:
       W.B. Saunders.

Randall, DJ. 1982. The control of respiration and circulation in fish during exercise and
       hypoxia. J. Exp. Biol. 100: 175-188.

Randall, DJ. and C. Brauner. 1991. Effects of environmental factors on exercise in fish.
       J. Exp. Biol. 160: 113-126.
                                       38

-------
Rice, J.A. 1990. Bioenergetics modelling approaches to evaluation of stress in fishes.
       Am. Fish. Soc. Symp. 8: 90-92.

Rodgers, D.W. and F.W.H. Beamish. 1981. Uptake of waterborne methylmercury by
       rainbow trout (Salmo gairdnerf) in relation to oxygen consumption and
       methylmercury concentration. Can. J. Fish. Aquat. Sci. 38: 1309-1315.

Schreck, C.B. 1990. Physiological, behavioural and performance indicators of stress.
       Am. Fish. Soc. Symp. 8: 29-37.

Shingles, A., DJ. McKenzie, E.W. Taylor, A. Moretti, P.J. Butler, and S. Ceradini. 2001.
       Effects of sub-lethal ammonia exposure on swimming performance in rainbow
       trout (Oncorhynchus mykiss). J. Exp. Biol. 204: 2699-2707.

Sprague, J.B.  1971. Measurement of pollutant toxicity to fish - III. Sublethal effects and
       "safe" concentrations. Water Res. 5: 245-266.

Steffensen, J.F. 1989. Some errors in the respirometry of water breathers: how to avoid
       and correct for them. Fish Physiol. Biochem. 6: 49-59

Steffensen, J.F., K. Johansen, and P.G. Bushnell. 1984. An automated swimming
       respirometer. Comp. Biochem. Physiol. 79A: 437-440.

Stegeman, J.J., K.W. Rento, B.R. Woodin, Y.-S. Zhang, and R.F. Addison.  1992.
       Molecular responses to environmental contamination: enzyme and protein
       systems as indicators of chemical exposure and effect. Pages 235-335 In:
       Biomarkers: Biochemical, Physiological and Histological Markers of
       Anthropogenic Stress. R.J. Hugget, R.A. Kimerly, M. Mehrle and H.L. Bergman
       (Eds.). Lewis Publishers, Chelsea.

Taylor, S.E., S. Egginton, and E.W. Taylor. 1996. Seasonal temperature acclimatisation
       of rainbow trout: cardiovascular and morphometric influences on maximum
       sustainable exercise level.  J. Exp. Biol. 199: 835-845.

Van der Cost, R., J. Beyer, and N.P.E. Vermeulen. 2003. Fish bioaccumulation and
       biomarkers in environmental risk assessment: a review. Env. Toxicol. Pharmacol.
       13: 57-149.

Verweij F., K. Booij, K. Satumalay, N. van der Molen, and R. van der Cost. 2004.
       Assessment of bioavailable PAH, PCB and OCP concentrations in water, using
       semipermeable membrane devices (SPMDs) sediments and caged carp.
       Chemosphere 54: 1675-1689.
                                      39

-------
Waiwood, K.G. and F.W.H. Beamish. 1978. Effects of copper, pH and hardness on the
       critical swimming performance of rainbow trout (Salmo gairdneri Richardson).
       Water Res. 12: 611-619.

Wendelaar Bonga, S.E. 1997. The stress response in fish. Physiol. Rev. 77: 591-625.

Wicks, B.J., R. Joensen, Q. Tang, and DJ. Randall. 2002. Swimming and ammonia
       toxicity in salmonids: the effect of sub lethal exposure on the swimming
       performance of coho salmon and the acute toxicity of ammonia in swimming and
       resting rainbow trout. Aquatic Toxicol. 59: 55-69.

Wikelski, M. and R.E. Ricklefs. 2001. The physiology of life histories. Trends Ecol.
       Evol. 16:479-481.

Wilson, R.W., H.L. Bergman, and C.M. Wood. 1994. Metabolic costs and physiological
       consequences of acclimation to aluminium in juvenile rainbow trout
       (Oncorhynchus mykiss). 2. Gill morphology, swimming performance, and aerobic
       scope. Can.  J. Fish. Aquat. Sci. 51: 527-535.

Winter, M.J., N. Day, R.A. Hayes, P.J. Butler, E.W. Taylor, and J.K. Chipman. 2004.
       DNA strand breaks and adducts determined in feral and caged chub (Leuciscus
       cephalus) exposed to rivers exhibiting variable water quality around Birmingham,
       UK. Mutat. Res. 552: 163-175.

Winter, M.J., F. Verweij, E. Garofalo, S. Ceradini, DJ. McKenzie, M.A. Williams, E.W.
       Taylor, P.J. Butler, R. van der Oost, and J.K. Chipman. 2005. Tissue levels and
       biomarkers of organic contaminants in feral and caged chub (Leuciscus cephalus)
       from rivers in the West Midlands, UK. Aquatic Toxicol. 73: 394-405.

Ye, X. and DJ. Randall.  1991. The effect of water pH on swimming performance in
       rainbow trout (Salmo gairdneri R.). Fish Physiol. Biochem. 9: 15-21.
                                       40

-------
                  Management of eutrophication

                                       by

                                  MJ. Gromiec1

Introduction

       Introduction of excess nutrients, as well as other pollutants, into water bodies are
causing many changes in aquatic environments, since they accelerate the process of
eutrophication. The term "eutrophication", usually refers to the natural or artificial
addition of nutrients to water bodies and to the effects of these added nutrients. It is
possible that either nitrogen or phosphorus will be the limiting nutrient controlling the
process of eutrophication in a water body. It is recognized that phosphorus is typically the
limiting nutrient in freshwaters, whereas nitrogen is typically limiting in estuarine or
marine waters. However, the relationships are more complex than this. Environmental
conditions in most water bodies are dominated by many factors, including seasonal
changes and interactions with bottom sediment. The dynamics of limiting nutrients are
not well known in most water bodies. Currently, the best nutrient management policy is
based on simultaneous reduction of both nitrogen and phosphorus inputs to water bodies.

Causes and effects of eutrophication

       The causes of eutrophication are related to meteorological and climatic status,
anthropogenic causes and features and characteristics of water bodies.
The anthropogenically induced nutrient loads represent the  main problem and include the
indirect loads which originate from atmospheric deposition or rivers and wastewaters
including that from agriculture which represent the main point sources. A fourth cause is
an internal nutrient load, which is related to release from sediments during hypoxic and
anoxic conditions. Increased concentrations of nitrogen  and phosphorus are the main
primary causes of eutrophication. However, substances other than inorganic phosphorus
and nitrogen compounds can also contribute to eutrophication. It should be stressed that
the whole population is contributing to the problem of eutrophication by our life style.

       The complex relation between natural and anthropogenic processes in relation to
nutrient dynamics make it difficult to understand cause and effect relationships.
Therefore, only some examples of eutrophication are presented below. The primary
effects of eutrophication are related to biological, chemical  and physical disturbances.
The increasing production of algae biomass, the decreasing amount of silica, and the
increasing turbidity of the water, are well known examples. Cyanobacterial blooms may
cause damage to organisms and result in odour problems. Sedimentation associated with
primary production and occurrence of hypoxia and anoxia are well known examples of
secondary and tertiary effects caused by eutrophication (Lundberg, 2005).
1 Department of Water Management, Institute of Meteorology and Water Management,
Warsaw, Poland
                                       41

-------
Water basin policy and nutrient management

       It is well recognised that river basin management is the key for sustainability in
water and land use management. In the European Union (EU), a Water Framework
Directive (WFD) is the most significant legislative instrument of policy in the water
management field (Chave, 2001). This legal instrument provides a framework for each
member state to develop a common basis for the sustainable use of water and for the
protection of water. The main requirement of the WFD is that EU member states ensure
that all waters are in good status by the end of 2015.

       The most important features of WFD are that it aims to manage the water
environment as a whole on a river basin basis. In addition, it calls to use a combined
approach to pollution control, setting limit values to control emissions from individual
point sources, and establishing water quality standards to  limit the cumulative impact of
emissions and diffuse sources of pollution.

       Therefore, in nutrient management it is necessary to assess the impact of human
activities on water bodies in each river basin, taking into consideration nutrient inputs
from point sources and from diffuse sources and other human activities that may impact
water status. Furthermore, it is necessary to  establish and  implement a legally binding
program of measures. This program to achieve the defined quality objectives will have to
follow the above mentioned combined approach, using the setting of emission limit
values and water quality standards.

Water quality criteria and effluent standards for nutrient  management

       Population growth in cities has resulted in an effort to reduce nutrient loads in
municipal wastewaters on receiving water bodies. A number of options exist for removal
of nitrogen and phosphorus from wastewaters. The strategies to achieve low
concentrations of nitrogen and phosphorus are usually based on water quality criteria and
effluent standards (Barnard and Steichen, 2006).

       In 2001, the United States Environmental Protection Agency published water
quality criteria with stringent nitrogen and phosphorus requirements. Depending on the
eco-region, the water quality criteria are as follows (USEPA, 2001):

•  total nitrogen (TN)             —     from   0.12mgTN/l
                                        to     2.18mgTN/l
•  total phosphorus (TP)           —     from   10 ug TP/1
                                        to     76 ug TP/1
These water quality criteria for 17 eco-regions of the USA are currently being evaluated.
There are also stringent requirements for low levels of effluent nutrient concentrations in
many regions of the world. In Europe, according to the EU urban wastewater treatment
directive (91/27I/EC) the effluents also have to meet guidelines on the content of
                                       42

-------
nitrogen and phosphorus. However, when a territory is designated as a sensitive area,
more advanced treatment of wastewater with nutrient removal should be provided
according to article 5.4 of the directive. This guideline restricts the following nutrients
concentrations in the effluent, for the sensitive areas, (where p.e. (population equivalent)
is equal to 60 g BOD5/d):

•  total nitrogen            —    15 mg TN/1 for 10 000-100 000 p.e.
                                  10 mg TN/1 for >100 000 p.e.
•  total phosphorus         —    2 mg TP/1 for 10  000-100 000 p.e.
                                  1 mg TP/1 for >100 000 p.e.

       The minimum percentage reduction of the overall load of phosphorus is at least
80% and a minimum of 70 to 80% for nitrogen. However, there is presently a great
emphasis on reaching very low levels of effluent phosphorus. For example, in Germany
(Berlin area), limiting phosphorus removal to  0.05 mg  TP/1 is required for the larger
plants to further reduce eutrophication of the local surface bodies.

       The above criteria and requirements require technologies to achieve low nitrogen
and phosphorus concentrations. Obviously, there are some limits to obtaining very low
total nitrogen concentrations in the treated effluents. Presently, a challenge is to remove
the dissolved organic nitrogen from wastewaters. Since the residual effluent dissolved
organic nitrogen (DON)  is composed of non-biodegraded nitrogen forms, a method is
needed to assess the biodegradability of effluent DON  (Pagilla etal., 2006). It should be
stressed, however, that the biological and chemical technologies for nutrient removal are
rather well established. A problem still exists with efficiently controlling nutrient loads
from diffused sources.

Non-point nutrient management in a catchment- An example

       Non-point source nutrients resulting from agricultural practices can be analyzed
with the assistance of mathematical models. The application of these models should be
based on reliable,  complete and correct data. The project "Controlling non-point pollution
in Polish catchments" deals with control measures in relation to agricultural practices and
management that minimizes nutrient leakage from non-point sources (DHI/IMWM,
2005). The project was implemented in the Pasleka River Catchment in the northern part
of Poland. The following main institutions participated in the project: Institute of
Meteorology  and Water Management (IMWM), Poland and DHI- Water and
Environment, Denmark.  The project has been sponsored by the Polish National Fund of
Environmental Protection and Water Management (NFOSGW) and the Danish
DANCEE.

       The Pasleka River (Fig. 1) discharges through the Vistula Lagoon into the Baltic
Sea. In the Vistula Lagoon some eutrophication  problems have been observed, among
others, due to excess discharge of nutrients from the catchment (Fig. 2). Data and
information were collected concerning soil types, land-use patterns and general
                                       43

-------
agricultural practice. In addition, monitoring of flow, water quality parameters and
meteorological conditions was performed.
              Land Use
              ^^| Discon. urban area
              |    | Arable
              ^^| Pastures
              |    | Natural vegetation
                 J Broad-leaved forest
              |    | Coniferous forest
              |    | Mixed forest
              |    | Marshes
              ^^1 Lake and river
             Figure 1. Land cover - Pasleka River Catchment
                                        44

-------
                 Total Nitrogen
Total phosphorus
            Diffuse load
                          Point
                         sources
                            Retention
                                           Diffuse
                                            load
                                                           Retention
                    Figure 2. Load assessment - Pasleka River

The applied modeling tools (Fig. 3) were as follows:

•  DAISY - an agricultural field model describing the relation between nutrient runoff
   from the root ozone and agricultural practice.
•  MIKE BASIN - a catchment model suitable for describing the overall transport of
   water and nutrients through the river basin including different nutrient concentrations
   in different water compartments.
•  MIKE 11HD, WQ and WET - a full dynamic hydrodynamic and water quality model
   system for rivers, lakes and wetland areas describing water flow, transport, and
   transformation of organic material and nutrient.
                        Figure 3. Outline of use modeling tools

       Future load scenarios have been defined based on collected information on point
sources, present agricultural practice and evaluation of potential development in
agricultural activity. The following five scenarios were simulated by the model system
covering the Pasleka River catchment:
                                        45

-------
•  Scenario 00 - Existing conditions.
•  Scenario 01 - Existing management with increasing animal production
•  Scenario 02 - Existing condition with grass in rotation instead of permanent
                grasslands
•  Scenario 03 - Existing management and animal production at EU level
                (characterized as "worst case").
•  Scenario 04 - Animal productivity of EU level plus optimal handling of slurry and
                crop rotation.
•  Scenario 05 - animal productivity of EU level plus optimal handling of slurry and
                crop rotation plus catch crop.

Tables I and II summarize the DAISY and MIKE BASIN modeling results from the
scenarios 0-5.

Table I.   Summary of loss of N03-N per ha (leaving the root zone)
Scenario
00
01
02
03
04
05
Average loss
kg N03 - N/ha
11
13
16
21
10
6
Difference in % of Existing
Conditions
0
18
45
90
-9
-45
Table II. Summary of total nitrogen transport and area specific transport in Pasleka
       River system
Scenario
00
01
02
03
04
05
TN-Transport
tons/year
931
1016
1393
1654
841
567
Area specific transport
kg/ha
4.0
4.4
6.0
7.1
3.6
2.4
•  Numbers include point sources and retention of nitrogen in rivers, lakes and
   groundwater

       The study shows in the case where animal productivity is increased to levels
approaching 1.5 animal units (AU) per ha without changing the general agricultural
management practice (scenario 3), a significant difference in the nitrogen loss can be
expected. An increased loss of nitrate from the root zone of 90% is expected.
                                       46

-------
Furthermore, the study shows that through optimal agricultural management, using catch
crops and extensive straw incorporation thus increasing animal productivity (scenario 5),
it is even possible to achieve a reduction of nitrate loss compared to the existing situation.
A 45 % reduction in nitrate loss from the catchment to the Baltic Sea has been simulated
(scenario 05 compared to 00).

       In addition, the study demonstrated that the existing riparian wetland zone along
the Pasleka River has a high potential for protection against nitrogen to the river
environment and thereby to the Baltic Sea. The planning and management tools (the
modeling software) were installed at the Regional Boards of Water Management in
Gdansk.

Conclusions

       Eutrophication management should be based on controlling both point and non-
point nutrients sources in a given watershed. Management of nutrients is only possible if
all sources of the nutrients as well as their fate in the catchment are known.

       As far as the nutrient point sources are concerned, there is presently emphasis on
reaching very low levels of effluent nitrogen and phosphorus. However, successful
strategies for nutrient management additionally need to be based also on economic
considerations. Especially, the cost of removing the remaining soluble organic nitrogen
from wastewaters must be justified by the environmental impact. Therefore, cost -
effectiveness and sustainability of the existing technologies for nutrient removal should
be evaluated.

       In successful strategies for nutrient management it is also necessary to address
non-point sources of nutrients, particularly from agriculture. The contributions of
agriculture are dependent on many factors, such as type and volume of agricultural
production, geologic, morphologic and climatic conditions, as well as agricultural
practice (density of animals, intensity of crop production, etc.). The optimal application
of scientific knowledge for nutrient management can be achieved by the use of modeling
tools.

References

Barnard,  J.L. and M.T. Steichen. 2006. Where is biological nutrient removal going now?
       Water Sci. Technol. 53(3): 155-164.

Chave, P. 2001. The EU Water Framework Directive-An Introduction. IWA Publishing,
       London.

Lundberg, C. 2005.Conceptualizing the Baltic Sea Ecosystem- An interdisciplinary tool
       for environmental decision making.  Ambio 34(6): 433-439.
                                        47

-------
Pagilla, K.R., M. Urgun-Demitras, and R. Ramanii. 2006. Low effluent nutrient
      technologies for wastewater treatment. Water Sci. Technol. 5(3): 165-172.

USEPA. 2001. Integrated Water Quality Monitoring and Assessment Report Guidance
      Memorandum.

DHI/IMWM. 2005. Controlling Non-Point Sources in Polish Catchments. Final report for
      DANCEE/NFOSGW, Copenhagen-Warsaw.
                                     48

-------
       Effects of ammonia on locomotor  performance

                                  in fishes

                                      by

           D.J. McKenzie1'2, A. Shingles1'3, P. Domenici3 and E.W. Taylor1

       Ammonia, which is toxic to all vertebrates, has become a pervasive pollutant of
aquatic habitats emanating from point sources such as sewage treatment plants, or diffuse
sources such as agricultural and urban storm runoff (API, 1981). In aqueous solution,
ammonia exists as two species, unionised ammonia gas (NH3) and the ammonium ion
(NH4+), with a pKa of approximately 9.5 in freshwater. Here, the term "ammonia" refers
to total ammonia, the sum of NH3 and NH4+. In freshwater fishes, ammonia is thought to
traverse the gill epithelium almost exclusively via passive diffusion of ammonia as a gas
in solution (Randall and Wright, 1987; Ip etal, 2001), so that water NH3 concentration
(partial pressure) determines the potential for toxicity. In seawater fish, gill permeability
to the ammonium ion may also contribute to toxicity (Randall and Tsui, 2002). Ammonia
is also a metabolite, produced in fish as an end product of protein and purine metabolism
and then excreted predominantly by passive diffusion of NH3 across the gill epithelium
(Randall and Wright, 1987; Wright, 1995). In teleost fish ammonia can, therefore,
accumulate to toxic levels, either as a consequence of exposure to elevated water
ammonia concentrations or when excretion of the endogenous metabolite is inhibited.
Within fish, the primary form of total body ammonia  at physiological pH (7.0 to 8.0) is
NH4+, and it is this chemical species that is responsible for toxic effects (Smart, 1976;
Hillaby and Randall, 1979). The toxic effects of ammonia in teleosts have been the
subject of recent reviews (Ip et al, 2001; Randall and Tsui, 2002).  In most completely
aquatic teleosts, immersed in unpolluted water, plasma ammonia is typically between 150
and 300 uM (Walsh, 1998).  The current paper will review briefly some studies of sub-
lethal toxicological effects on locomotor performance of increased plasma ammonia
concentrations.

Effects of ammonia on critical speed swimming

       A number of investigators have demonstrated  that sub-lethal increases in the
concentration of ammonia in the plasma can impair the ability of freshwater salmonids to
perform in a critical swimming speed (Ucrit) test (Beaumont et al, 1995a; Shingles et al.
2001; Wicks et al. 2002; McKenzie et al. 2003).  This test involves exposing the fishes to
sequential increments in current velocity in a swim tunnel until they fatigue (Brett,  1964).
During this test, the fish will initially  power swimming activity with slow-twitch
oxidative "red" muscle but, as faster current velocities are imposed, it will eventually be
obliged to recruit fast-twitch glycolytic "white" muscle to achieve the highest tailbeat
1 School of Biosciences, University of Birmingham, Birmingham, United Kingdom.
2 CNRS UMR 5171, Station Mediterraneenne de I'Environnement Littoral, Sete, France.
3 International Marine Centre, Oristano, Italy.
                                       49

-------
frequencies and swimming speeds, and this will lead relatively rapidly to fatigue (Brett,
1964; Beamish, 1978). A negative linear relation has been described between plasma
ammonia concentration and Ucrit, as shown in Figure 1. In the first study to reveal this
effect (Beaumont et al, 1995a), an increase in plasma ammonia occurred in brown trout
(Salmo trutta) as a consequence of exposure to sublethal concentrations of copper in soft
acidic water (pH 5). Beaumont et al. (1995b; 2000a,b; 2003) attributed the impairment of
swimming performance to plasma ammonia accumulation because there was no evidence
of major disruptions to cardiorespiratory metabolism, particularly such as problems with
62 uptake and cardiovascular convection. Figure 1 shows that exposure of brown trout to
ammonia alone (Shingles, McKenzie and Taylor, unpublished data, reported in Shingles
2002 and McKenzie et al., 2003) impairs Ucrit in a very similar fashion to the impairment
observed by Beaumont et al. (1995a) in trout exposed to copper, thereby supporting these
latter authors' conclusion that the effects of the heavy metal were mediated through
ammonia accumulation (Beaumont, 1995b; 2000a,b; 2003; McKenzie et al, 2003).
                        200
400
600
800
1000
                                 plasma total ammonia u^ mol
                                                           -i
     Figure 1. Linear relationships between plasma ammonia concentration and
             Ucrit in brown trout (Salmo trutta) and rainbow trout
             (Oncorhynchus mykiss). The blue symbols are data for individual
             brown trout in which plasma ammonia accumulated following
             exposure to sub-lethal concentrations of copper in soft acidic
             water, replotted from Beaumont etal. (1995a). BL, bodylength.
             The blue line describes a least squares linear regression equation
             whereby Ucrit = -0.0020 * [ammonia] + 2.089 (R2 = 0.670, n =
                                       50

-------
             30). The large black symbols are data for brown trout exposed to
             three water concentrations of ammonia, replotted from
             McKenzie et al. (2003). Plasma ammonia and Ucrit were
             measured on separate groups offish (n = 6 or 7), and the black
             line describes a least squares linear regression equation whereby
             mean Ucrit = -0.0018 * mean[ammonia]) + 2.347 (R2 = 0.903, n =
             3). The red symbols are data for individual rainbow trout
             exposed to elevated water ammonia, plotted from data reported
             in Shingles et al. (2001).  The red line describes a least squares
             linear regression equation whereby Ucrit = -0.0024 * [ammonia]
             + 2.677 (R2 = 0.590, n = 12).

       One of the major toxic effects of ammonia is that, as the ammonium ion, it can
substitute for potassium at vertebrate muscle  and nerve membranes, thereby
compromising their function (Raabe and Lin, 1985; Randall and Tsui, 2002). Wright et
al. (1988) demonstrated that the distributions of NH4+ between extracellular  and
intracellular compartments could be used with the Nernst equation to calculate membrane
potential (EM) in fish. Beaumont et al. (1995b; 2000a,b) found that the relative plasma to
white muscle (WM) ammonia distributions predicted a significant depolarisation of the
tissue in brown trout that were hyperammonemic following exposure to copper in acidic
water.  Beaumont et al.  (2000c) then made direct measurements of EM in WM of
hyperammonemic brown trout, and confirmed the depolarisation, presumably due to the
replacement of K+ with NH4+ (Table I).  Since then, in all studies investigating effects of
ammonia on Ucrit performance in salmonids, application of the Nernst equation
consistently predicts a significant depolarisation of WM (Table I).  The reduced WM  EM
(Beaumont etal.,l995b; 2000a,b;  Shingles et al. 2001; Wicks et al. 2002 McKenzie et al.
2003) was judged to be sufficient to cause a complete loss of electrical excitability in  that
tissue (Jenerick, 1956), such that it could not be recruited to power anaerobic burst
swimming at the highest speeds.
                                       51

-------
 Table I. Plasma and white muscle ammonia concentrations, and membrane potential, in
         various salmonid species during exposure to sublethal concentrations of either
         copper in soft acidic water or water ammonia gas. Taken from McKenzie et al.
         (2003).

 Species      Exposure conditions  PI. Tamm  PI. NH4+  PI.      WM  WM   WM   WMEM  WM EM
 (live mass)	NH3    Tamm  NH4+  NH3    (calc.)    (meas.)
 Salmotruttc? Softwater (0.05 mM  109 ±    107 ±    1.4 ±   3168  3162   5.9 ±   -92.8  ±  -86.5  ±
 (300-600g)   Ca++) at 10°C, pH    21       21       0.3     ±223 ±223  0.4    3.4       2.9
             7.0
             96h exposure to     652    ±  645   ±  6.9  ±  4248  4248   6.9 ±  -54.5  ±  -52.2  ±
             0.08 umol'1 Cu at    95       94       1.0     ±616 ±616  1.0    5.1       4.9
	pH5.0	
 S.tmttab     Hardwater (280 mM  134 ±    132 ±    2.0 ±   2097  2092   4.2 ±   -71.5 ±   NA
 (370-740g)   Ca++) at 15°C, pH    29       29       0.4     ±544 ±543  0.8    7.5
             8.4
             24h exposure to 7    386 ±    380 ±    5.9 ±   2735  2728   7.8 ±  -49.9 ±   NA
             uMNH3           42       42       0.5     ±612 ±611  1.7    8.3

             24h exposure to 14   771    ±  757 ±    14.6 ±  4249  4240   9.3 ±  -43.4  ±  NA
             uMNH3           92       91       1.6     ±315 ±314  2.1    1.6

 Oncorhynch  Hardwater (280 mM  183 ±    180 ±    2~7±1750  1748   22±   -60.3 ±   NA
 usmykissc    Ca++) at 14°C, pH    30       29       0.6     ±360 ±360  0.4    6.8
 (380-790g)   8.4
             24h exposure to 20   436    ±  429   ±  7.0  ±  1950  1948   2.3 ±  -34.9  ±  NA
	uMNH3	34       34       0.5     ±240 ±240  0.3    2.8	
 O.kisutchd   Softwater (0.12 mM  5300   ±  5273  ±  26.7 ±  NA   NA   NA    -53.0  ±  NA
 (350±65g)   Ca++)at9-12°C,pH  700      696      3.5                         3.0
             6.0
             4h exposure to 1.1    6800   ±  6771  ±  28.6 ±  NA   NA   NA    -41.0  ±  NA
             uMNH3           900      896      3.8                         4.0
             4h exposure to 2.2    8500   ±  8466  ±  34.1 ±  NA   NA   NA    -25.0  ±  NA
             uMNH3           1100     1096     4.4                         3.0
             4h exposure to 4.4    9500   ±  9453  ±  46.8 ±  NA   NA   NA    -20.0  ±  NA
	uMNH3	1500     1493     7.4	4.0	
 PL,  plasma; Tamm, total ammonia; WM,  white muscle; EM, membrane potential; calc.,
 calculated with the Nernst equation; meas., measured directly; NA, not available.
 All ammonia concentrations are uM, EM is in mV.
 a As reported by Beaumont et al.. (2000c). Sampled from  fish at rest, arterial plasma
 obtained via chronic indwelling dorsal aortic cannula
 b As reported by Shingles (2002) and McKenzie et al. (2003). Sampled from fish at rest,
 arterial plasma obtained via chronic indwelling dorsal aortic cannula
 c As reported by Shingles et al. (2001).  Sampled  from fish following exercise to Ucrit,
 arterial plasma obtained via chronic indwelling dorsal aortic cannula.
 d As reported by Wicks et al. (2002). Sampled from fish  following  exercise to Ucrit,
 venous plasma obtained by caudal puncture.
                                         52

-------
       This depolarisation of WM may be the primary means by which ammonia impairs
Ucrit performance in salmonids. During a Ucrit protocol, the incremental increases in
aerobic red muscle (RM) work lead to an exponential increase in whole-animal oxygen
demand, and this demand is met by increased cardiac work and internal oxygen
convection (Jones and Randall, 1978). There is evidence that these processes function
equally well in control and hyperammonemic brown trout. Beaumont etal. (2003) found
no evidence that the reduced Ucrit of hyperammonemic brown trout could be attributed to
impaired cardiac performance. Exposure of brown trout for 24h to two sub-lethal
concentrations of ammonia (100 umol I"1 and 200 umol I"1) reduced Ucrit but this was not
linked to reduced maximum rates of oxygen uptake nor aerobic scope during the Ucrit test
(Table II, Shingles et al, unpublished data). On the other hand, the hyperammonemic
trout could not achieve the same maximum tailbeat frequencies and, at fatigue, WM of
hyperammonemic trout did not exhibit the  large lactate accumulation observed in  control
trout (Table II). Thus, these observations are consistent with the hypothesis that
hyperammonemic brown trout have reduced Ucrit because glycolytic WM is depolarised
and therefore cannot be recruited to achieve the highest tailbeat frequencies and
swimming speeds.

Table II.  Mean (+  SEM) values for selected exercise-related respiratory and performance
        variables in  brown trout following exposure to one  of three water ammonia
        concentrations:  ambient (control); 98 ± 6 umol I"1 NlHLiCl (low ammonia), or
        210  ±  11 umol I"1 NlHLtCl  (high ammonia). Ucrit, critical swimming speed; MMR,
        maintenance metabolic rate; AMR, active metabolic rate as maximum measured
        M02;/TB max., maximum tailbeat frequency; RM red muscle; WM, white
        muscle. N = 6 in all cases, * significantly different from all other groups, a
        different superscipt indicates a significant difference between the means for that
        variable,§ indicates a significant difference between rest and fatigue (Tukey test
        post-hoc to two-way ANOVA, p < 0.05).

Ucrit (bodylengths. s"1)
MMR (mmol O2 kg'1 h'1)
AMR (mmol O2 kg'1 h'1)
Aerobic scope (mmol O2 kg"1 h"1)
/TB max. (Hz)
WM lactate at rest (mmol g"1)
RM lactate at rest (mmol g"1)
WM lactate at fatigue (mmol g"1)
RM lactate at fatigue (mmol g"1)
control
2.24 ±0.15*
2.62 ±0.39
12.32 ±2.50
9.70 ±1.00
4.90 ±0.31*
10.4±2.1a'b
4.6 ±1.6
21.3±2.2a§
8.9±1.4a
low ammonia
1.46 ±0.09
2.43 ± 0.49
10.82 ± 1.87
8.39 ±0.85
2.85 ±0.40
7.5±1.7b
3.3 ±1.0
11.2±3.3b
3.1±1.2b
high ammonia
1.08 ±0.16
3. 15 ±0.90
11.24 ±1.98
8.10±0.87
2.60 ±0.25
14.6 ± 1.7s
5.0± 1.3
11.7±3.1b
8.3±2.1a
       Table III shows the ammonia accumulation in the tissues of the brown trout
studied by Shingles et al. (unpublished data). Both WM and RM exhibited significant
ammonia accumulation. As already stated, the accumulation in the WM was enough to
cause a significant predicted depolarisation (Table I). Shingles etal.  (unpublished data)
did not find evidence for a significant depolarisation of RM, despite  the ammonia
accumulation in the tissue. The means by which RM might be able to maintain function
while WM cannot is worthy of further investigation. The most striking data in Table III,
                                       53

-------
however, are the ammonia levels in the heart and brain. In the hyperammonemic groups,
these tissues did not exhibit a significant increase in their ammonia content despite the
large plasma accumulation.  As shown in Figure 2, for the trout exposed to the highest
concentration of ammonia, the distribution of ammonia between plasma and the
intracellular compartment of these tissues was significantly less than would be predicted
if it were distributed according either to EM or pH. Thus, the heart and brain would
appear to be protected from ammonia, and this must contribute to the ability of the trout
to co-ordinate and perform prolonged aerobic exercise during the Ucrit tests. Tsui et al.
(2004) found that the performance of isolated ventricular trabeculae from rainbow trout
was not significantly affected by extracellular ammonia concentrations exceeding 1 mM,
significantly higher than the plasma levels that impaired Ucrit performance in either
rainbow trout (Shingles et al., 2001) or brown trout (Beaumont et al., 1995a,b; 2000a,b;
2003; Shingles et al, unpublished data). Tsui et al. 2004 attributed this protective effect
to background K+ channels that were not as sensitive to NH4+ in the myocardium as they
were in WM. The brain, on the other hand, may protect itself from accumulation through
detoxification of ammonia to glutamine (Wicks and Randall, 2002; Randall and Tsui,
2002). Wicks and Randall (2002); Randall and Tsui (2002) and Tsui et al. (2004) suggest
that these mechanisms for protecting heart and brain function may have evolved to cope
with the surges of plasma ammonia that occur in salmonids after feeding. It is known that
rainbow trout perform poorly in Ucrit tests following feeding (Alsop and Wood, 1997)
although the potential role of ammonia in this response remains to be explored.

Table III. Mean (+  SEM) total ammonia concentrations in plasma (umol I"1) and selected
         tissues (umol g"1) of brown trout following 24h exposure to one of three water
         ammonia concentrations: ambient (control); 98 ± 6 umol I"1 NFLjCl (low
         ammonia), or 210 ± 11  umol I"1 NFLjCl (high ammonia). Data are shown for
         animals at rest (A) and  at fatigue following sustained exercise (B). N = 6 in all
         cases, plasma measurements are not available for animals at fatigue. Where
         indicated, a different superscript indicates a significant difference between the
         means for that tissue (Tukey test post-hoc to two-way ANOVA, p < 0.05).
         *significant difference between animals at rest and at fatigue for a given
         ammonia exposure regime (Tukey test post-hoc to two-way ANOVA, p <
         0.05).

(A) At rest
plasma
white muscle
red muscle
heart
brain
(B) At fatigue
white muscle
red muscle
heart
brain
Control
133.6±29.2a
1.2±0.3a
1.4 ±0.4
1.3 ±0.5
1.1±0.3
2.9 ±0.5*
2.8 ±0.6^
1.4 ±0.4
1.9±0.6a
Low ammonia
386.0 ± 41. 5b
2.0±0.4a
2.1 ±0.5
1.8±0.6
1.1±0.2
3. 5 ±0.8
2.1±0.3a
2.1 ±0.7
1.6±0.5a
High ammonia
771.3±92.2C
3.1±0.2b
2.7 ±0.2
1.5 ±0.5
1.8±0.5
3.7±0.6
4.7±0.4b*
2.7 ±0.4
2.8±0.2a
                                       54

-------
    heart
                           brain
                                               UJ


                                               o
                                               U-J
                                               3
                                               .Q
                                               T3
                                               CO

                                               O
                                               E
                                               E
                                               CD
                                                                       by EM
                                                                       bypH
                                                                       measured
            control
   100
treatment
200
control
   100
treatment
200
   Figure 2.  Ammonia distributions between the extracellular compartment
            (plasma) and the intracellular compartment of the heart and brain in
            brown trout (Salmo trutta) following 24h exposure to one of three
            water ammonia concentrations: ambient (control); 98 ± 6 umol I"1
            NH4C1 ("100"), or 210 ± 11 umol I'1 NH4C1 ("200"). Distributions
            are given as the ratio of intra- to extra-cellular concentrations, when
            predicted from the plasma concentration and either the membrane
            potential (EM) or the pH of the two compartments, or as directly
            measured. N = 6 in all cases, for pH and measured distributions,
            values are mean ± SEM.

Effects of ammonia on performance of the escape response

       Startle escape responses are used by many fish as the main defence against
predator attacks, and their kinematics, performance, behaviour and physiology have been
studied extensively (see Domenici and Blake,  1997 for a review). Escape responses
comprise a sudden and brief acceleration, typically in a direction away from the startling
stimulus (Domenici and Blake,  1993). They are usually triggered by one of a pair of
Mauthner cells, giant neurons that permit a rapid response in the order of a few
milliseconds (Eaton and Hackett, 1984). The response involves unilateral contraction of
the axial musculature contralateral to the stimulus, so that the fish bend into a
characteristic C-shape. This represents stage 1 and it can be followed by a second
contraction, stage 2, on the opposite side of the body (Foreman and Eaton, 1993;
Domenici and Blake,  1997). These movements are all exclusively dependant upon fast-
twitch glycolytic WM function (Webb, 1998). Beyond stage 2, locomotor behaviour is
                                       55

-------
variable and can include either steady swimming or just coasting (Weihs, 1973). The
success of an escape response for predator avoidance depends, therefore, on both sensory
and locomotor performance, although little is known about how these might be affected
by environmental factors (e.g. Webb, 1978; Webb and Zhang, 1994; Lefran9ois etal.,
2005). There is no knowledge of how they might be affected by sub-lethal concentrations
of pollutants such as ammonia.

       A series of experiments were performed to investigate the effects of sub-lethal
concentrations of ammonia upon performance of the startle reflex by a marine teleost, the
golden grey mullet, Liza aurata (Shingles, McKenzie, Claireaux and Domenici,
unpublished data). Figure 3 shows the effects of two water ammonia concentrations
(nominally 400 umol I"1 and 1600 umol I"1) on ammonia levels in the venous plasma,
WM and brain following 24h exposure. The mullet were extremely tolerant of ammonia
by comparison with salmonids. Initial 96h tests revealed that 2 mM total ammonia in
seawater did not cause any fatalities. Both WM and brain exhibited large and significant
increases in ammonia content, with little evidence that the brain was protected against
ammonia accumulation in the manner observed in brown trout.
                  6000
• plasma
D muscle
n brain
                             control
                400

             Treatment
1600
             Figure 3. Mean (± SEM) total ammonia (Tamm) concentrations in
                      the plasma (umol I"1), the white muscle and the brain
                      (umol g"1) of golden grey mullet (Liza aurata) at 24h
                      following exposure to one of three water ammonia
                      concentrations: ambient (control); 400 umol I"1 NFLtCl, or
                                1-1
                      1600 umol 1  NH4C1. N = 8 in all cases.

       Individual reflex responses by mullet to a mechanical stimulus, a small weight
dropped into their tank, were filmed with a high speed camera (500 frames/sec). Digital
video sequences were exported to tracking software for analysis and calculation of the
various events and dependent variables that comprise the startle reflex. Detailed materials
and methods are as reported in Le Fra^ois et al. (2005) with the exception that fish were
allowed 48h recovery from anaesthesia. Figure 4 shows the effects of ammonia upon a
series of "non-locomotor" variables. There was no effect of ammonia on overall
                                       56

-------
responsiveness as shown by the percentage of animals that responded to the stimulus.
This is an indicator of the fish's acoustic and/or visual sensitivity and its motivation to
escape (Domenici and Blake, 1997). There was, however, a direct effect of ammonia
upon response latency, the interval between stimulus onset to the first detectable
movement leading to the escape of the animal (Fig. 4). Response latency will be a
function of nervous performance throughout the reflex arc. Thus, these results indicate
that the observed accumulation of ammonia in the brain (and presumably also peripheral
nervous tissues) did not affect overall sensitivity and motivation to escape, but did have
negative effects upon those elements of nervous performance which determine latency.
                    n  non responders
                    H  responders
       0)
       0)
       .2
       "£
       0)
          100%
          75%
          50%
          25%
           0%
m     m
                                            0)
                                            E
0.50


0.40


0.30


0.20


0.10 •


0.00
                                                                      •x
                                           I
                 control    400     1600
                                   control
                                                              400     1600
                       Treatment
                                          Treatment
             Figure 4. Responsiveness (percentage responding) and mean (±
                      SEM) response latency to a startling stimulus by golden
                      grey mullet (Liza aurata) at 24h following exposure to
                      one of three water ammonia concentrations: ambient
                      (control); 400 umol I'1 NH4C1, or 1600 umol I'1 NH4C1.
                      N = 12 for responsiveness, N = at least 8 for latency. An
                      asterisk indicates a significant difference from the control
                      value (Tukey test post-hoc to one-way ANOVA, p <
                      0.05).

       Once the escape response is triggered, "prey" locomotor performance can  be
evaluated with variables such as maximum turning rate during the C-bend, the maximum
acceleration and maximum velocity achieved during the subsequent escape phase, and the
total cumulative distance covered as a result (Domenici and Blake, 1993; 1997; Le
Fran9ois et al, 2005). Figure 5 shows that maximum turning rate during the C-bend was
significantly lower in mullet exposed to the highest concentration of ammonia. Maximum
acceleration was not significantly different between ammonia-exposed animals and the
controls, although there is some evidence of a trend that is consistent with the trends
observed in the other locomotor performance variables. That is, the maximum velocity
                                       57

-------
achieved was significantly affected by both concentrations of ammonia, with an
apparently direct relation between water (and therefore tissue) ammonia levels and
performance. The cumulative distance covered was significantly reduced by exposure to
the highest ammonia concentration. All of these locomotor variables are, of course, a
direct result of glycolytic WM performance and, therefore, the results indicate that
ammonia exposure has impaired WM function.

       Further analysis of the data can provide a proximate reason for the impaired
locomotor performance during the escape response. Figure 6 shows that the mullet
exposed to ammonia typically only performed a single-bend reflex, hence an abbreviated
and partial response by comparison to that observed in the control animals. A direct
comparison of single versus double bend responses confirmed that the former generate
lower acceleration, lower maximum velocity and, therefore, lower distance swum. This is
another clear indication that WM function was impaired in the mullet exposed to
ammonia. Unfortunately, it was not posible to calculate WM EM in the hyperammonemic
muscle because measurements of plasma and tissue pH were not made. However, it
seems probable that the WM was depolarised in the hyperammonemic mullet, as has been
observed in hyperammonemic salmonids (Beaumont et al, 2000c).
    
-------
                      exposure to one of three water ammonia concentrations:
                      ambient (control); 400 umol I"1 NH4C1, or 1600 umol I"1
                      NH4C1. N = at least 8. An asterisk indicates a significant
                      difference from the control value (Tukey test post-hoc to
                      one-way ANOVA, p < 0.05).
                   O)
                       100%
                                                  D single bend
                                                  • double bend
                       75% -
                   s    50% ^
                   D.
                       25%
                        0%
                                control
  400
Treatment
1600
                   Figure 6. Percentage of either single or double bend
                            responses to a startling stimulus by golden grey
                            mullet (Liza aurata) at 24h following exposure to
                            one of three water ammonia concentrations:
                            ambient (control); 400 umol I"1 NH4C1, or 1600
                            umol I'1 NH4C1.  N = at least 8.

Conclusions and perspectives

       Their is now good evidence to indicate that exposure to sub-lethal concentrations
of ammonia impairs the function of WM, such that fish cannot recruit the tissue to
generate high tailbeat frequencies and high swimming speeds, or to power effective
startle escape responses. This may have a number of implications for the animals in their
natural environment. Fish recruit WM to negotiate velocity barriers (Peake and Farrell,
2004). This may be particularly important for salmonids during their spawning migration
(Standen et al, 2004), but also for all fish that might negotiate such barriers as part of
their routine activities (Castro-Santos, 2005) or  when seeking flow refuges during, for
example, seasonal storms and floods. Thus, sub-lethal pollution by ammonia may not
only impair the ability of salmonids to migrate (Wicks et al., 2002), but also the ability of
all fish species to successfully inhabit particular areas. The impairment of the startle
response would presumably put the fishes at greater risk of predation, in particular, from
aerial predators such a birds, that would not experience the toxic effects of increased
water ammonia.  The mullet was exceptionally tolerant of ammonia, but the fishes are
known to colonise polluted habitats such as harbours. Salmonids might exhibit impaired
escape responses at much lower ambient ammonia concentrations.
                                        59

-------
Acknowledgements

This work was partially funded by a NERC studentship to A.S. and contracts from the
Commission of the European Community  in  support  of  the  projects  Cityfish and
Ethofish.
References

Alsop, D.H. and C.M. Wood. 1997. The interactive effects of feeding and exercise on
   oxygen consumption, swimming performance and protein usage in juvenile rainbow
   trout (Oncorhynchus mykiss). J. Exp. Biol. 200: 2337-2346.

API 1981. The sources, chemistry, fate and effects of ammonia in aquatic environments.
   American Petroleum Institute, pp.  145. Washington D.C.

Beamish, F.W.H. 1978. Swimming Capacity. Pages 101-187 In: Fish Physiology, Vol. 7.
   W.S. Hoar and DJ. Randall (Eds.). Academic Press, New York.

Beaumont, M.W.,  PJ. Butler, and E.W. Taylor. 1995a. Exposure of brown trout, Salmo
   trutta, to sub-lethal copper concentrations in soft acidic water and its effects upon
   sustained swimming performance. Aquat. Toxicol. 33: 45-63.

Beaumont, M.W., PJ. Butler, and E.W. Taylor. 1995b.  Plasma ammonia concentration in
   brown trout (Salmo trutta) exposed to acidic water and sublethal copper
   concentrations and its relationship to decreased swimming performance. J. Exp. Biol.
   198: 2213-2220.

Beaumont, M.W., Butler, P.J.. and Taylor, E.W. 2000a. Tissue ammonia levels and
   swimming performance of brown trout exposed to copper in soft, acidic water. Pages
   51-68 In: Fish  Physiology, Fish Toxicology and Fisheries Management. R.V.
   Thurston (Ed.) EPA/600/R-00/015. United States Environmental Protection Agency,
   Athens, Georgia.

Beaumont, M.W., Butler, and E.W. Taylor. 2000b. Exposure of brown trout, Salmo
   trutta, to a sub-lethal concentration of copper in soft acidic water: effects upon
   muscle metabolism and membrane potential. Aquat. Toxicol. 51: 259-272.

Beaumont, M.W., E.W. Taylor, and PJ. Butler. 2000c.  The resting membrane potential
   of white muscle from brown trout  (Salmo trutta) exposed to copper in soft, acidic
   water. J. Exp. Biol. 203: 2229-2236.

Beaumont, M.W., PJ. Butler, and E.W. Taylor. 2003. Exposure of brown trout, Salmo
   trutta, to a sub-lethal concentration of copper in soft acidic water: effects upon gas
   exchange and ammonia accumulation. J. Exp. Biol.  206: 153-162.
                                      60

-------
Brett, J.R. 1964. The respiratory metabolism and swimming performance of young
   sockeye salmon. J. Fish. Res. Board. Can. 21: 1183-1226.

Castro-Santos, T. 2005. Optimal swim speeds for traversing velocity barriers: an analysis
   of volitional high-speed swimming behavior of migratory fishes. J. Exp. Biol. 208:
   421-432.

Domenici, P. and R.W. Blake. 1993. Escape trajectories in angelfish (Pterophyllum
   eimekei). J. Exp. Biol. 177: 253-272.

Domenici, P. and R.W. Blake. 1997. Fish fast start kinematics and performance. J. Exp.
   Biol. 200: 1165-1178.

Eaton, R.C. and J.T. Hackett. 1984. The role of Mauthner cells in fast-starts involving
   escape in teleost fish. Pages 213-266 In: Neural Mechanisms of Startle Behavior.
   R.C. Eaton (Ed.). Plenum Press, New York.

Foreman, M.B. and R.C. Eaton. 1993.  The direction change concept for reticulospinal
   control of goldfish escape. J. Neurosci. 13: 4101-4133.

Hillaby, B.A. and DJ. Randall. 1979. Acute ammonia toxicity and ammonia excretion in
   Rainbow trout (Salmo gairdnerf). J. Fish. Res. Bd. Can.  36: 621-629.

Ip, Y.K., S.F. Chew, and DJ. Randall. 2001. Ammonia toxicity, tolerance and excretion.
   Pages 10-148 In Fish Physiology Vol. 20. P.A. Wright and P.M. Anderson (Eds.).
   Academic Press, San Diego.

Jenerick, H.P. 1956. The relations between prepotential, resting potential and latent
   period in frog muscle fibres. J. Gen. Physiol. 39: 773-787.

Jones, D.R. and DJ. Randall. 1978. The respiratory and circulatory systems during
   exercise. Pages 425-501 In: Fish Physiology, Vol. 7. W. S. Hoar and D. J. Randall
   (Eds.). Academic Press, New York.

Le Fran9ois, C., A. Shingles, and P. Domenici. 2005. The effect of hypoxia on locomotor
   performance and behaviour during escape in Liza aurata. J. Fish. Biol. 67: 1711-
   1729.

McKenzie, D J., A. Shingles, and E.W. Taylor.  2003. Sub-lethal plasma ammonia
   accumulation and the swimming performance of salmonids. Comp. Biochem.
   Physiol. 135A: 515-526.

Peake S J. and A.P. Farrell. 2004. Locomotory behaviour  and post-exercise physiology in
   relation to swimming speed, gait transition and metabolism in free-swimming
   smallmouth bass (Micropterus dolomieu). J. Exp. Biol. 207: 1563-1575
                                       61

-------
Raabe, W. and S. Lin. 1985. Pathophysiology of ammonia intoxication. Exp. Neurol. 87:
   519-532.

Randall DJ. and T.K.N. Tsui. 2002. Ammonia toxicity in fish. Mar. Poll. Bull. 45: 17-
   23.

Randall, DJ. and P.A. Wright. 1987. Ammonia distribution and excretion in fish. Fish
   Physiol. Biochem. 3:  107-120.

Shingles, A. 2002. The effects of some environmental pollutants on ammonia excretion
   and swimming performance of trout. Ph.D. Thesis, University of Birmingham, United
   Kingdom.

Shingles, A., DJ. McKenzie, E.W.  Taylor, A. Moretti, PJ. Butler, and S. Ceradini. 2001.
   Effects of sub-lethal ammonia exposure on swimming performance in rainbow trout
   (Oncorhynchus mykiss). J. Exp. Biol.  204, 2699-2707.

Smart, G.  1976. The effect of ammonia exposure on gill structure of the rainbow trout
   (Salmogairdnerf). J. Fish Biol.  8: 471-475.

Standen, E.M., S.G. Hinch, and P.S. Rand. 2004. Influence of river speed on path
   selection by migrating adult sockeye salmon {Oncorhynchus nerka). Can. J. Fish.
   Aquat. Sci. 61: 905-912.

Tsui T.K.N., D J. Randall, L. Hanson, A.P. Farrell, S.F. Chew, and Y.K. Ip. 2004.
   Dogmas and controversies in the handling of nitrogenous wastes: Ammonia tolerance
   in the oriental weatherloachMisgurmis onguillicaudatus. J. Exp. Biol. 207: 1977-
   1983.

Walsh, P J. 1998. Nitrogen excretion and metabolism. Pages 201-216 In:  The Physiology
   of Fishes, 2nd edn. D.H. Evans (Ed.). CRC Press, Boca Raton.

Webb, P.W. 1978. Temperature effects on acceleration of rainbow trout Salmo gairdneri.
   J. Fish. Res. Bd. Can. 35:  1417-1422.

Webb, P.W. and H. Zhang.  1994. The relationship between responsiveness and
   elusiveness of heat-shocked goldfish (Carassius auratus) to attacks by rainbow trout
   (Oncorhynchus mykiss). Can. J. Zool. 72: 423-426.

Webb, P.W. 1998. Swimming. Pages 3-24 In: The Physiology of Fishes, 2nd edn. D.H.
   Evans (Ed.). CRC Press, Boca Raton.

Weihs, D. 1973. The mechanism of rapid starting of slender fish. Biorheology 10:  343-
   350.
                                       62

-------
Wicks, BJ. and DJ. Randall. 2002. The effect of sub-lethal ammonia exposure on fed
   and unfed rainbow trout: the role of glutamine in regulation of ammonia. Comp.
   Biochem. Physiol. 132A: 275-285.

Wicks, B.J., R. Joensen, Q. Tang, and DJ. Randall. 2002. Swimming and ammonia
   toxicity in salmonids: the effect of sub lethal exposure on the swimming performance
   of coho salmon and the acute toxicity of ammonia in swimming and resting rainbow
   trout. Aquat. Toxicol. 59: 55-69.

Wright, P.A. 1995. Nitrogen excretion: three end products, many physiological roles. J.
   Exp. Biol. 198:273-281.
                                      63

-------
64

-------
     Ammonia and salinity tolerance in the California

                         Mozambique tilapia.

                                       by

                    K. Suvajdzic1, B. Sardella and CJ. Brauner

Introduction

       The Salton Sea is a large inland lake (980 km2) that was created in 1905, when
water from the Colorado River was accidentally diverted, and flooded the Imperial and
Coachella Valleys of SE California. The lake resides at 200 ft below sea level in a desert,
and there is no outflow.  However there is substantial evaporative water loss that
approximately matches water input which is comprised largely of municipal and
agricultural run-off. Consequently, in the last 100 years, the salinity of the Salton Sea has
risen from freshwater to a current salinity of 43 g I"1, and the salinity is currently
increasing at a rate of about 0.3 g I"1 annually (Gonzalez etal, 1998).

       When the salinity of the Salton Sea  reached that of seawater, a number offish
species were transplanted from the Sea of Cortez in the hopes of establishing a
recreational fishery. Of those introduced, sargo (Anisotremus davidsoni), gulf croaker
(Bardiella icistius\ and  orange mouth corvina (Cynoscion xanthulus), were very
successful. Following this, the California Mozambique tilapia (Oreochromis
massambicus x O. urolepis hornorum) was  accidentally introduced, presumably from
local fish farms, and more recently, has become the dominant fish species in the sea.
Because of the ever increasing salinity, at some point an upper salinity limit will be
reached beyond which sustainable fisheries for these species will not be possible. For this
reason there has been interest in assessing the salinity tolerance of these species from a
fisheries management perspective.

       While salinities above that of seawater represent a considerable  challenge to most
fish, there are additional environmental factors that interact with salinity to affect a
fishery, such as temperature, oxygen,  ammonia, selenium, arsenic and sulfide (Watts et
al, 1991). We have been conducting experiments to investigate the salinity tolerance of
the California Mozambique tilapia, and the influence that temperature and ammonia may
have upon ionoregulation at salinities greater than seawater.

Salinity tolerance and the effect of temperature on California Mozambique tilapia

       California Mozambique tilapia are remarkably tolerant of salinities greater than
seawater. In fish that were acclimated to seawater (35ppt) at 25°C and then transferred
every 5 days to a 10 g I"1 increase in salinity (at constant temperature), there were no
changes in plasma Na+,  Cl", osmolality or gill Na+,K+-ATPase activity up to a salinity of
1 Department of Zoology, University of British Columbia
                                       65

-------
65 g I"1. These data indicate that these fish have a remarkable ability to tolerate high
                                                               1-1
salinity at this temperature (Sardella et al, 2004a). From 65 to 95 g 1" , there was a
progressive increase in all of these parameters. Despite an elevation in drinking rate and
gill Na+,K+-ATPase activity (the predominant driving force for ion excretion), there was
an elevation in plasma [Na+], [Cl"]and osmolality, at 75 g I"1 indicating that the tilapia
were experiencing some degree of osmoregulatory stress at this temperature (Fig.  1).
While fish in these experiments were only exposed to each salinity for 5 days before
being subjected to the next salinity increase, longer duration exposure (28 days) to similar
salinities at 25°C result in a qualitatively similar pattern (Sardella and Brauner,
unpublished data).
           4.5
        ro
        .c
        O
4.0 -


3.5 -

3.0 -


2.5 -

2.0 -


1.5 -

1.0 -


0.5
                    Plasma Sodium
                    Plasma Chloride
                    Plasma Osmolality
                    Na+, K+-ATPase activity
                    Drinking Rate
             30
                     40
                            50
                                    60      70

                                    Salinity (g/l)
                                                   80
                                                          90
                                                                  100
Figure 1. Indices of change relative to values measured in seawater-acclimated animals
         for Plasma osmolality, [Na+], and [Cl~], Na+,K+-ATPase activity and drinking
         rate (* absolute values statistically significant from seawater-acclimated fish, p<
         0.05; modified from Sardella etal, 2004a).

       The Salton Sea experiences large seasonal fluctuations in temperature (13-35°C;
Watts et al., 2001) with little in the way of temperature refugia for fish, and thus it is very
likely that fish in the Salton Sea will experience a great range of temperatures. When
California Mozambique tilapia were acclimated to 35 g I"1 and then directly transferred to
15°C for 24 h, there were substantial increases in plasma Na+,  Cl" and osmolarity, and a
virtual elimination of gill Na+,K+-ATPase activity (Sardella et al., 2004b). Simultaneous
direct transfer from 35 g I"1 at 25°C to a salinity of 51 or 60 g I"1 at 15°C resulted in 100%
mortality. At a constant temperature of 25°C, all tilapia survive direct transfer from 35 to
65 g I"1 with minimal osmoregulatory disturbances at 24 h and complete recovery by  5
                                         66

-------
days. Taken together, these data indicate that salinity tolerance is greatly reduced at 15
relative to 25°C. Thus, despite the tilapia's amazing ability to tolerate salinities greater
than seawater at 25°C, they may be near to their salinity tolerance during the winter
months when the water temperature drops below 15°C and large fish kills are observed in
the Salton Sea (Sardella etal, 2006).

       Ammonia levels in the Salton Sea are also very high, and the following
experiments were designed to assess ammonia tolerance of the California Mozambique
tilapia and determine whether there are interactions between ammonia and salinity
tolerance.

Ammonia toxicity and the effect of ammonia on salinity tolerance

       Ammonia can exist in the ionized (NH4+) and unionized (NHa) form, the latter
being the most toxic to fish. Unionized ammonia levels have been reported to be as high
as 1.2 mg I"1 in the Salton Sea, which exceeds the US EPA water quality criterion. To
determine whether environmental ammonia levels could have an affect on survival offish
in the Salton Sea, we assessed the ammonia tolerance of California Mozambique tilapia,
determined whether ammonia has an influence on salinity tolerance, and investigated the
basis for ammonia tolerance; specifically  whether this tilapia species is capable of
producing urea during exposure to elevated water ammonia levels. It is known that some
fish produce urea as a method of dealing with high ammonia levels.

       California Mozambique tilapia hybrids (5-10 g) were exposed to a range of
environmental ammonia levels for 96h at 25°C, at an average water pH of 8  following
acclimation to either 35g I"1 or 44g l^salinity. Plasma ion levels and gill Na+,K+-ATPase
activity were measured in surviving fish from these trials to determine whether ammonia
exposure was associated with ionoregulatory disturbances. Total ammonia was measured
and unionized NHs was calculated using the measured pH and a pKa of 9.354 in 35g I"1
and 9.374 in 44g I"1 seawater (Khoo et al, 1977).

       As seen  in Figure 2, the 96hLCso at 35 g I"1 was 9.75mM total ammonia or
0.48mM (8.26mg I"1) unionized NHa. In fish transferred to 44g I"1  during exposure to a
range of environmental ammonia levels, the 96hLCso was 0.49mM (8.27mg I"1) NH3 or
13.0 mM total ammonia, nearly identical to the 35g I"1 values. This is approximately
seven fold greater than the unionized ammonia levels reported in the Salton  Sea.  In
comparison with other fish, the California Mozambique tilapia are remarkably ammonia
tolerant. Typical ammoniotelic teleosts have LCso values for unionized ammonia less
than 0.10 mMNH3 (Walsh etal., 1993; Wang and Walsh, 2000; Person-Le Ruyet etal.,
1995). High ammonia tolerance has been  observed in a number of air-breathing fishes (Ip
et al, 2004; Wood et al, 2005, Randall et al, 2004), including the oriental weatherloach
(Tsui et al, 2004), the toadfish (Wand and Walsh, 2000), and others, and this is often
associated with  the ability to  produce urea as discussed below.

       In some  marine fish, ionoregulatory disturbances have been observed during
exposure to lethal ammonia concentrations (18.2 jimol l^NHsin Atlantic salmon), but
                                       67

-------
not at sub-lethal ammonia concentrations (Knoph and Thorud, 1996). Plasma from the
California Mozambique tilapia that were transferred to 44g I"1 and survived the 96h
ammonia exposures above did not exhibit any significant changes in plasma [Na+] or
[Cl"] (Fig. 3). Minor elevations in gill Na+,K+-ATPase activity were observed at higher
ammonia exposures (Fig. 4). In general, there appears to be little effect of ammonia on
ionoregulation in California Mozambique tilapia, at least over this salinity range.
              100 -
              80 -
              60 -
              40 -
        . Survival
              20 -
               o -
                       0       2       4       6      8      10

                       Total Unionized Ammonia (NH3-N mg/l) at 35 ppt
                                                                 12
Figure 2. Percent survival of 35g 1" acclimated fish following exposure to NHa (mg 1" )
         for 96 hours.
         g
         ro
        ^o
        0_
240 -


220 -


200 -


180 -


160 -


140 -


120 -
             100
                      0      2      4      6      8      10

                          Total Unionized Ammonia (NH3-N mg/l)
                                                                12
Figure 3. Plasma Na+ and Cl" concentrations (mM) in surviving fish following 96 h
         exposure to ammonia in 44g I"1 and 25°C.
                                        68

-------
            1
            Q.
            D) '
16
            "5
              12
              10 -
            ro
            Q_
              6 -
                      0      2      4      6      8      10

                          Total Unionized Ammonia (NH3-N mg/l)
                                                                12
Figure 4. Branchial Na+/K+ ATPase activity in surviving fish following 96 h exposure to
         ammonia in 35g I"1 or 44g I"1 and 25°C.

Urea production

       Fish that can survive exposure to high ammonia levels employ a range of
strategies to survive what would otherwise be toxic conditions (Ip et al., 2004; Randall et
al, 2004; Wang and Walsh, 2000; Tsui etal, 2004; Wilkie, 2002; Randall etal, 1989;
Wood et al, 2005). The giant mudskipper Periophthalmodon schlosseri is able to
decrease amino acid catabolism, or undergo partial amino acid catabolism thus
decreasing ammonia production (Ip et al, 2004; Randall et al, 2004). Fish can also alter
body surface pH to facilitate NHa volatilization as is seen in the Oriental weatherloach
Misgurnus anguillicaudatus (Ip et al, 2004; Tsui etal, 2004). The slender African
lungfish Protopterus dolloi expresses ornithine-urea cycle enzymes, and converts
ammonia to urea (Ip et al, 2004; Wood et al, 2005), as does the Toadfish Opsanus beta
and O. tau (Wand and Walsh, 2000). Of the mechanisms employed by fish to tolerate
high internal ammonia levels, the most common is the production of urea (Wilkie, 2002),
and this strategy has been observed in the Lake Magadi tilapia (Oreochromis alcalicus
grahamf) to tolerate high environmental pH (Randall et al, 1989).

       To determine whether the basis for California Mozambique tilapia to tolerate high
environmental ammonia levels  was associated with the ability to produce  urea, fish were
exposed to sub-lethal ammonia concentrations and urea appearance in the water was
                                       69

-------
measured. When tilapia hybrids were exposed to a sub-lethal ammonia concentration
(8mM TAmm, or 4.4 mg I"1 unionized ammonia) minor changes in urea production were
observed (Fig. 5). At no time did urea production ever exceed 15% of total nitrogenous
waste excretion. Thus, although the California Mozambique tilapia appear to be
extremely ammonia tolerant,  they only produce limited amounts of urea and thus this is
probably not their main mechanism of ammonia detoxification or basis for their
exceptional ammonia tolerance.
              b, 3(H
              o
                 20 -
               03
                                 a,b
                                        a,b
                                                 ft
                                  48       72

                                    Time (h)
       Figure 5. Urea production during a 96h exposure at 35 ppt to 0 and 8mM Total
               Ammonia. Letters (a, b, etc.) indicate significant difference within a
               treatment group. Asterisks (*) indicate significant difference at a given
               time between treatments.

       In summary, California Mozambique tilapia hybrids are extremely ammonia
tolerant with 96 h LC50s of 0.48mM (8.26 mg I"1) NH3 at 35 g I"1. This ability to
withstand elevated ammonia levels is independent of salinity as the 96 h LCso at 44g I"1
was nearly identical to that at 35 g I"1 (0.49mM/8.27 mg I"1 NHa). Sublethal increases in
environmental ammonia do not appear to have significant ionoregulatory consequences
as there was no  significant change in plasma [Na+]or [Cl"]concentration and only minor
elevations in Na+,K+-ATPase. While some species offish produce urea as a method of
ammonia detoxification, this does not appear to be the case in these tilapia as urea
production only increases slightly during sub-lethal ammonia exposure and never
accounts for greater than 15% of total nitrogenous waste excretion. Although the
tolerance of California Mozambique tilapia hybrids is far in excess of levels measured in
the Salton Sea, how ammonia interacts with temperature and other abiotic factors has not
been investigated and ammonia should not be discounted as a potential factor reducing
survival in this species in the Salton Sea. Ammonia tolerance in other species of the
Salton Sea is not known.

Acknowledgements:
       This research was  supported by the U.S. Bureau of Reclamation and NSERC
Discovery Grant to CJB. We wish to thank Doug Barnum and Ray Stendell of the Salton
                                       70

-------
Sea Science Office for their input and general interest in the project. Finally, we greatly
thank Colin Bornia and Bill Engler of Pacific Aquafarms for their kind donation offish.
References

Gonzalez, M.R., C.M. Hart, J.R. Verfaillie, and S.H. Hurlbert. 1998. Salinity and fish
       effects on Salton Sea microecosystems: water chemistry and nutrient cycling.
       Hydrobiologia381: 105-128.

Ip, Y.K., S.F. Chew, DJ. Randall. 2004. Five tropical air-breathing fishes, six different
       strategies to defend against ammonia toxicity on land. Physiol. Biochem. Zool.
       77(5): 768-782.

Khoo, K.H., C.H. Culberson, and R.G. Bates. 1977. Thermodynamics of the dissociation
       of ammonium ion in seawater from 5 to 40°C. J. Sol. Chem. 6(4) 281-290.

Knoph, M.B. and K. Thorud. 1996. Toxicity of Ammonia to Atlantic Salmon (Saho
       Sabr L.) in Seawater -Effects on Plasma Osmolality, Ion, Ammonia, Urea and
       Glucose Levels and Hematologic Parameters. Comp. Biochem. Physiol. 113A(4):
       375-381.

Person-Le Ruyet, J., J.H. Chartois, and L.  Quemener. 1995. Comparative acute ammonia
       toxicity in marine fish and plasma  ammonia response. Aquaculture 136: 181-194.

Randall, D.J., C.M. Wood, S.F. Perry, H. Bergman, G.M.O. Maloiy, T.P.  Mommsen, and
       P. A. Wright. 1989. Urea excretion as a strategy for survival in a fish living in a
       very alkaline environment. Nature 337: 165-166.

Randall, D.J., Y.K. Ip, S.F. Chew, and J.M. Wilson. 2004. Air breathing and ammonia
       excretion in the giant mudskipper,  Periophthalmodon schlosseri. Physiol.
       Biochem. Zool. 77(5): 783-788.

Sardella, B.A., V. Matey, J. Cooper, R.  Gonzalez, and CJ. Brauner. 2004a.
       Physiological, biochemical and morphological indicators of osmoregulatory stress
       in "California" Mozambique tilapia (Oreochromis mossambicus x O. urolepsis
       hornorum) exposed to hypersaline water. J. Exp. Biol. 207(8):  1399-1413.

Sardella, B.A.,  J. Cooper, R. Gonzalez,  and CJ. Brauner. 2004b. The effect of
       temperature on the salinity tolerance of juvenile Mozambique tilapia hybrids
       (Oreochromis mossambicus x O. urolepis hornorum). Comp. Biochem. Physiol.
       137(4): 621-629.

Sardella, B.A., V. Matey, and CJ. Brauner. 2006. Coping with multiple stressors:
       Physiological mechanisms and strategies in fishes of the Salton Sea.
       Hydrobiologia (In Press).
                                       71

-------
Tsui, T.K.N., DJ. Randall, L. Hanson, A.P. Farrell, S.F. Chew, and Y.K. Ip. 2004.
       Dogmas and controversies in the handling of nitrogenous wastes: Ammonia
       tolerance in the oriental weatherloack Misgurnus anguillicaudatus. J. Exp. Biol.
       204:  1977-1983.

Walsh, P.J., H.L. Bergman, A. Narahara, C.M. Wood, P.A. Wright, DJ. Randall, J.N.
       Maina, and P. Laurent. 1993. Effects of ammonia on survival, swimming and
       activities of enzymes of nitrogen metabolism in the Lake Magadi tilapia,
       Oreochromis alcalicus grahami. J. Exp. Biol. 180: 323-327.

Wang, Y. and PJ. Walsh. 2000. High ammonia tolerance in fishes of the family
       Batrachoididae (toadfish and midshipmen). Aqat. Toxicol. 50: 205-219.

Watts, J.M.,  B.K. Swan, M.A. Tiffany, and S.H. Hurlbert. 2001. Thermal, mixing, and
       oxygen regimes in the Salton Sea, California,  1997-1999. Hydrobiologia (in
       press).


Wilkie, M.P. 2002.  Ammonia excretion and urea handling by fish gills: Present
       understanding and future research challenges. J.  Exp. Zool. 293: 284-301.

Wood, C.M., PJ. Walsh, S.F. Chew, and Y.K. Ip. 2005. Ammonia tolerance in the
       slender lungfish (Protoptems dolloi): the importance of environmental
       acidification. Can. J. Zool. 83: 507-517.
                                       72

-------
                        Nitrite toxicity to fishes

                                       by

                                   R.C. Russo1

Introduction

       Nitrite concentrations can build up in the aquatic environment from point sources
such as fish culture systems with recirculated water; septic tanks; industrial effluents
from metal, dye, and celluloid industries; and wastewater treatment plants if there is an
imbalance among species of nitrifying bacteria. It can also enter aquatic systems from
nonpoint sources such as fertilizer and animal wastes; feedlot discharges; nitric oxide and
nitrite discharges from automobile exhausts;  and leachates from waste disposal dumps.
Nitrite is extremely toxic to many aquatic organisms, as shown by numerous studies of its
toxicity and physiological effects.

Factors Affecting Nitrite Toxicity

       There are differences in species sensitivities to nitrite, as shown in the toxicity
data compilation in Table I. In general, saltwater fish species appear to be more tolerant
of nitrite than are fresh water species. Salmonids are the most sensitive species, and
centrarchids are the most resistant. Most reported toxicity  tests were static bioassays.
One research group (Tilak etal., 2002) performed both static and continuous flow tests
and obtained very similar results; static values were consistently higher, but not
significantly so. The U.S. Environmental Protection Agency EcoTox database has
additional data records (U.S. EPA, 2002).

       Several investigators have studied acute nitrite toxicity on fish of different sizes.
Alcaraz and Espina (1995) reported that larger juvenile grass carp (Ctenopharyngodon
idelld) are more tolerant than smaller ones. Almendras (1987) found that smaller juvenile
milkfish (Chanos chanos Farsskae)  are more tolerant. Atwood etal. (200Ib)  reported that
smaller Nile tilapia (Oreochromis niloticus) were significantly more tolerant  of nitrite
than larger specimens. Palachek and Tomasso (1984a) found that smaller fathead
minnows (Pimephalespromelas) were more tolerant than  larger ones. Russo  (1980), who
tested rainbow trout (Oncorhynchus mykiss) of sizes 2-387 g, found no differences
related to fish size; however, Lewis and Morris (1986), using a different statistical
procedure on the same data, reported that small fish had significantly higher LCsos than
large fish. Hilmy etal. (1987) found a modest difference in susceptibility between 65-
and 166-g Glorias lazera, with the larger fish somewhat more susceptible than the
smaller ones. No significant difference due to fish size was found between channel
catfish (Ictaluruspunctatus) of 3.0 g and 80.2 g or between largemouth bass
(Micropterus salmoides) of 2.8 g and 36.3 g (Palachek and Tomasso, 1984b;  Tomasso,
1986).
1 190 West Huntington Road, Bogart Georgia, USA
                                       73

-------
       Water chemistry conditions affect nitrite toxicity. The aqueous nitrite equilibrium
is pH-dependent, with relative concentrations of ionized nitrite (NCV) and un-ionized
nitrous acid (HNO2) dependent on the pH of the system. The concentration of HNC>2 is 4-
5 orders of magnitude less than the concentration of NO2" in the pH range 7.5-8.5 (Russo
et al., 1981). It has been shown that over the pH range 6.4-9.0 the toxicity of total nitrite
on rainbow trout (Oncorhynchus mykiss) decreases as pH increases. As pH increases,
NO2-N toxicity decreases and HNO2-N toxicity increases (Russo et al., 1981).
Furthermore, Bowser et al. (1983) found that dissolved oxygen affects the toxicity of
nitrite: an oxygen concentration of 5 mg/L in the presence of nitrite was insufficient for
channel catfish (Ictaluruspunctatus), even though this species would normally tolerate
lower oxygen concentrations.

       In studies on juvenile grass carp (Ctenopharyngodon idelld) at 3 temperatures,
Alcaraz and Espina (1995) found lower toxicity on larger fish (7.60 g) at 29 °C than at 24
°C or 32 °C. A temperature of 29 °C had a "protective effect" for the larger fish; the
authors suggested that this could be attributed to the fact that biochemical and enzymatic
processes are sensitive to temperature, being more efficient in the temperature range
corresponding to the preferred temperature of the species. Adult grass carp generally
prefer 29 °C. At a temperature of 24 °C nitrite was less toxic to smaller (0.02 g) fish than
at 29 °C or 32 °C. The authors suggested the possibility that smaller fish have a different
interval of preferred temperature or perhaps the acclimation mechanisms in smaller fish
have not developed completely.

       Watenpaugh et al. (1985) investigated the temperature tolerance of channel
catfish (Ictaluruspunctatus) exposed to sublethal concentrations of nitrite for 24 hours.
The critical thermal maximum (a measure of the upper limit of thermal tolerance) was
inversely related to  nitrite concentration. Percent methemoglobin was correlated with
nitrite concentration and was inversely correlated with the critical thermal maximum.
The authors concluded that nitrite exposure had the potential of adversely affecting the
productivity of high density channel catfish aquaculture systems at higher temperatures.

       Crawford and Allen (1977) studied the toxicity of nitrite on chinook salmon
(Oncorhynchus tshawytscha) and found that the toxicity of nitrite in seawater was
markedly less than that in fresh water and that increasing the calcium concentration  in
both fresh water and seawater decreased the toxicity of nitrite. However, Atwood et al.
(200la) tested southern flounder (Paralichthys lethostigma) in fresh and brackish water
and found similar mortality at similar nitrite concentrations. In both fresh and brackish
water, plasma nitrite concentrations were well below environmental concentrations.
Plasma nitrite concentrations increased significantly with increasing environmental nitrite
concentrations in both fresh and brackish water, but fish did not appear to concentrate
nitrite in plasma. Grosell and Jensen (1999), in studies with European flounder
(Platichthysflesus), reported concentration of nitrite in plasma. In studies with juvenile
mullet (Mugilplatanus) Sampaio et al. (2002) found that fish acclimated to fresh water
were significantly more sensitive to nitrite than those held at higher salinities. Increased
nitrite toxicity in fresh water vs. salt water has also been reported for European eel
Anguilla anguilla (Saroglia et al., 1981) and for red drum Sciaenops ocellatus (Wise and
Tomasso, 1989).
                                        74

-------
       The presence of other chemicals affects nitrite toxicity. It is well known that
chloride ions inhibit nitrite toxicity in fishes (Russo and Thurston, 1977; Perrone and
Meade, 1977; Wedemeyer and Yasutake, 1978; Tomasso etal.,  1979; Tomasso, 1986).
Tomasso (1986), however, found no effect of chloride on toxicity of nitrite in largemouth
bass (Micropterus salmoides). Atwood et a/., (2001a) reported that environmental
chloride in the form of salts had little effect on either survival or uptake of nitrite by
southern flounder (Paralichthys lethostigmd). The high serum chloride concentrations of
saltwater species (e.g., pinfish) (Folmar et a/.,  1993) may help explain their greater
tolerance to nitrite than freshwater species. Atwood et al. (200Ib) tested Nile tilapia
(Oreochromis niloticus) and found increased survival with added chloride. The plasma
nitrite values in the tests with chloride added were well below the environmental
concentrations. They also reported that there was no difference whether the chloride was
added as calcium chloride or sodium chloride,  whereas others (Wedemeyer and
Yasutake, 1978 with rainbow trout (Oncorhynchus mykiss); and Mazik et al., 1991 with
striped bass (Morone saxatilis)) reported that calcium chloride is a more effective nitrite
toxicity inhibitor than sodium chloride. Chloride was more effective at inhibiting the
uptake of environmental nitrite by shortnose sturgeon fingerlings (Acipenser
brevirostruni) when chloride was added as calcium chloride rather than as sodium
chloride (Fontenot etal., 1999). They found that shortnose sturgeon fingerlings
previously exposed to nitrite (for 2 days) had significantly lowered plasma NO2-N levels
when calcium chloride was added to the water. They suggested calcium chloride could be
an effective treatment for nitrite toxicity for shortnose sturgeon

       Most channel catfish farmers in the southeastern United  States routinely add
sodium chloride to their ponds to prevent methemoglobinemia. In a survey of Alabama
catfish farmers, Sipauba Tavares and Boyd (2003) found most farmers maintain a
chloride concentration of 50-100 mg/L through annual sodium chloride applications.

       Other substances including bromide, sulfate, phosphate,  nitrate, methylene blue,
ascorbic acid, and uric acid have been reported to inhibit nitrite toxicity to different
degrees.

Sublethal Effects

       Clinical observations in fishes of stress induced by nitrite exposure reported by
Das et al. (2004c) included: erratic swimming, frequent opercular movement, surfacing,
resting on bottom of tank with irregular opercular movements, loss of equilibrium, and
lying on their sides. Reduction in food consumption and growth rate have  also been
observed (Kumta and Gaikwad, 1997).

       Stormer et al. (1996) exposed rainbow trout (Oncorhynchus mykiss) to nitrite
concentrations of 1 mM for 8 days. They found nitrite accumulation in plasma to
concentrations above the environmental level.  Plasma nitrate concentration increased in
parallel with the accumulation of nitrite. Hematocrit and blood hemoglobin decreased;
the concentration of chloride and bromide in plasma decreased;  K+ concentration in
caudal  muscle tissue decreased significantly and water content of muscle tissue also
decreased. Tested fish fell into two distinctly different groups: those that died between 1
                                        75

-------
and 2 days of exposure and those that survived up to day 4 but died before day 6. Nitrite
accumulation was more rapid in the first group.

       In experiments using isolated rainbow trout (Oncorhynchus mykiss) hepatocytes,
Doblander and Lackner (1996) found that the hepatocytes detoxified nitrite through
oxidation to nitrate. Detoxification was inhibited by bumetanide and furosemide, but was
strongly accelerated by uric acid. Uric acid also enhanced detoxification activity in trout
liver.

       Das et al. (2004a) studied the effects of nitrite on mrigal (Cirrhinus mrigala
(Ham.)),  a fish species of major importance for aquaculture in India. Exposure of mrigal
fingerlings to sublethal nitrite levels (1-8 mg/L NO2-N) in static toxicity tests caused
progressive reduction in total erythrocyte count, hemoglobin, and serum protein content.
Both decrease and increase in total erythrocyte count were observed, depending on
concentration and exposure period. Blood glucose decreased up to 24-h of exposure at all
concentrations, but then increased until 96 h of exposure. Similar results were reported
for catla (Catla catla (Ham.)) by Das  et al.  (2004b) and Das et al. (2004c).

       In a 96-h study of the effects on  enzymes of sublethal nitrite exposures to catla
(Catla catla)., rohu (Labeo rohita\ and mrigal (Cirrhinus mrigala) Das et al. (2004c)
observed similar responses in all three species. As nitrite concentration increased from 1
to 10.4 mg/L, reduction in activities was observed in acetylcholinesterase (AChE) in
brain and liver; alkaline phosphatase (ALP) in serum, brain, and gill; and acid
phosphatase (ACP) in gill. There was also a progressive increase in alanine
aminotransferase (ALAT) and aspartate aminotransferase (ASAT) activities in brain, gill,
and serum; and ACP activity in serum and brain in response to increasing nitrite
concentrations. Lactate dehydrogenase (LDH) activity increased in gill, liver, kidney,
brain, and serum up to 8 mg/L nitrite with a reduction in activity at the highest (10.4
mg/L) concentration tested. AChE activity reduction was indicative of liver tissue
damage. LDH activity increase was attributable to prevalence of anoxia. The increase in
ALAT and ASAT activities in serum and the brain is an indicator of tissue damage. They
suggested that the measurement of LDH activity in tissues  (but not serum) of the Indian
major carp fingerlings could be a biomarker for nitrite toxicity.

       In 15-day exposures to sublethal nitrite levels with grass carp (Ctenopharyngodon
idella) Alcaraz and Espina (1997) found that ingestion rates were not affected, but
assimilation efficiency and assimilation were reduced at concentrations of 1.6 and 2.5
mg/L NO2-N. Respiration rates decreased and nitrogen excretion rate increased at those
levels. The sublethal nitrite exposures reduced the scope for growth of the fish.

       Kumta and Gaikwad  (1997) conducted sublethal exposures to 1.25 mg/L of
NaNC>2 [0.2 mg/L NO2-N] in mosquitofish (Gambusia affinis) for one month. Reported
fecundity was significantly reduced in exposed fish. The gonadosomatic index was also
reduced in exposed fish. These results demonstrate the adverse effect of nitrite on
reproduction. Reductions in food consumption (67.3%) and growth rate (59.5%) were
also observed.
                                        76

-------
       In 5-d exposures of shortnose sturgeon (Acipenser brevirostum) fmgerlings to
nitrite Fontenot et al. (1999) found that plasma nitrite concentrations were more than 63
times the environmental concentration. Other species have also been reported to
concentrate nitrite in plasma to concentrations greater than environmental exposure
concentrations, including channel catfish Ictaluruspunctatus (Tomasso et al., 1979;
Palachek and Tomasso, 1984b; Tomasso, 1994); blue tilapia Tilapia aurea (Palachek and
Tomasso, 1984b); and rainbow trout Oncorhynchus mykiss (Eddy et al., 1983).

       In studies on sea bass (Lates calcarifer) exposed to 30 to 80 mg/L NO2-N for 4
days, Woo and Chiu (1997) observed osmoregulatory dysfunction with elevated  serum
Na+ and Cl" levels and reduced branchial Na+-K+-ATPase activity. Serum lactate levels
were significantly elevated and serum protein levels reduced at only the 80 mg/L
exposure. Increased serum ammonia and urea and decreased serum glucose and liver
glycogen levels were observed at lower nitrite levels. Significant decreases were found  in
activities of glycogen phosphorylase a, glutamate-oxaloacetate transaminase, and
glutamate dehydrogenase in the liver. Exposure to 50 mg/L induced an increase in the
rate of ammonia excretion.

       Pinfish (Lagodon rhomboides) were exposed to nitrite for 96 h by Folmar et al.
(1993), and 20 serum chemistry parameters were measured. They used concentrations of
NaNC>2 from 1.2-11 mg/L [0.24-2.2 mg/L NO2-N]. They analyzed for sodium, potassium,
chloride, calcium, magnesium, iron, inorganic phosphorus, alkaline phosphatase, alanine
aminotransferase, aspartate aminotransferase, lactate dehydrogenase,  creatine kinase,
blood urea nitrogen, creatinine, albumin, cholesterol, glucose, total  protein, triglyceride,
and uric acid. The only significant change observed was a decrease in uric acid at 48 h,
while no difference was observed at 96 h. There  were no other alterations in electrolyte
balance or serum chemistries.

       A recent review by Jensen (2003) provides a more comprehensive discussion of
the physiological effects of nitrite on fishes and other aquatic animals, including ion
regulatory, respiratory, cardiovascular, endocrine, and excretory processes.

       Increase in susceptibility to infection can occur in fishes subjected to sublethal
concentrations of some toxicants. Carballo and Mufioz (1991) investigated the
susceptibility of rainbow trout (Oncorhynchus mykiss) to infection by fungus Saprolegnia
parasitica (sin. S. diclina type 1). Only Saprolegnia spp. have been found on live
salmonids, even though the fish come in contact with many fungal spores. Fish infected
with these spores experience a breakdown in their osmoregulatory mechanism and, unless
they can be treated,  the condition is fatal. The authors tested the effects of ammonia,
nitrite, copper, and cyanide on susceptibility to infection. In 10-d exposures to 0.05 mg/L
NH3-N or 0.12 mg/L NO2-N at inoculum concentrations of 1.4xl06, 9.75xl05, and 5xl05
zoospores/L they found that 75% of ammonia-exposed fish tested at the highest inoculum
concentration and 50% of the nitrite-exposed fish at the highest inoculum concentration
developed infection. Infection occurred in 20% of ammonia-exposed fish at the medium
fungal dose. No infection occurred in nitrite-exposed fish at the medium dose, and no
infection occurred in either ammonia- or nitrite-exposed fish at the lowest fungal dose
tested. No infection was found with copper and cyanide exposures.  Their experiments
confirmed that ammonia and nitrite predispose fish to saprolegniosis,  and they concluded

                                        77

-------
that fish in fish farms that are subjected to elevated ammonia and nitrite levels are at
increased risk of saprolegniosis.
Long-term nitrite toxicity

       Wedemeyer and Yasutake (1978) tested 5- to 10-g steelhead trout (Oncorhynchus
mykiss) atNO2-N concentrations of 0, 0.015, 0.030, and 0.060 mg/L in a flow-through
system over a 6-month rearing period. The exposed fish exhibited mild
methemoglobinemia. There was no effect on growth, and transfer to 30% seawater over a
48-h period caused no mortalities. No hematological abnormalities were observed. At the
0.060 mg/L exposure concentration, there was minimal hypertrophy, hyperplasia and
lamellar separation in gill epithelium. This decreased over time, and after 28 weeks of
exposure most of the test fish had recovered.

       Hilmy et al. (1987) carried out 6-month exposures on juvenile Clarias lazera at a
NO2-N concentration of 1/10 of the 96-h LC50 values (28 mg/L for 65-g fish and 32 mg/L
for 166-g fish). Erythrocyte count, hemoglobin concentration, and hematocrit values were
decreased from the beginning in nitrite-exposed fish, reaching minimum levels after 4
months exposure; after this, the values began to increase during the remainder of the
exposure period. The nitrite exposure caused a mild but statistically significant
methemoglobinemia. There was a significant decrease of serum total proteins in the first
4 months followed by an elevation during the remainder of the experiment; however, the
values were still less than those of control fish. The authors surmised that this trend in the
5th and 6th months represents an adaptation response to nitrite. Histological effects of gill
hypertrophy and hyperplasia were observed (Michael etal., 1987)

Joint toxicity of ammonia and nitrite

       In exposures of rainbow trout (Oncorhynchus mykiss) to nitrite and ammonia
simultaneously for 4 days Vedel et al. (1998) observed high mortality (68%) in fish
exposed to 500 uM ammonia [7 mg/L NH3-N] plus 600 uM nitrite [8.4 mg/L NO2-N], the
highest concentration combination tested. Nitrite and ammonia interactive effects (either
synergism or antagonism) were not observed in the physiological parameters measured.
Nitrite was accumulated in plasma to about twice the exposure concentration. Ammonia
and nitrite exposure both caused a significant and additive increase in muscle potassium
concentrations. There was also an increase in methemoglobin concentration by nitrite.
Ammonia exposure decreased brain glutamate concentration while glutamine
concentration increased. The  authors suggested that ammonia was detoxified by reacting
with glutamate to form glutamine.

Nitrite Toxicity to Invertebrates

       Yildiz and Benli (2004) studied the effects of nitrite on narrow-clawed crayfish
(Astacus leptodactylus) and found nitrite to be acutely toxic at a concentration range of
22-70 mg/L NO2" [6.7-21.3 mg/L NO2-N] for 48-h exposures. The toxic concentration
was increased to 31-80 mg/L NO2" [9.4-24.3 mg/L NO2-N] when 100 mg/L chloride was
added to test solutions. They measured hemolymph nitrite, total hemocyte counts (THC),

                                       78

-------
and hemolymph glucose at sublethal nitrite concentrations. They found accumulation of
nitrite in hemolymph increased directly with water nitrite concentration. The hemolymph
nitrite values were significantly higher than the exposure levels, up to 34 times the
exposure level of 25 mg/L NCV [7.6 mg/L NO2-N]. After a 24-h recovery period in
nitrite-free water, hemolymph nitrite decreased. Nitrite was completely eliminated from
the hemolymph only when the lowest (9 mg/L NCV, 2.7 mg/L NO2-N) environmental
concentration was used. THC decreased in nitrite-exposed crayfish and increased after
24-h recovery. Hemolymph glucose levels increased with increasing nitrite exposure
concentrations. They also tested the effect of chloride addition and found the
accumulation of nitrite in hemolymph to be low relative to the nitrite-only tests. THC
increased in nitrite-plus-chloride tests and remained elevated after return to nitrite-free
water. Hemolymph glucose levels did not change in nitrite-plus-chloride tests. The
addition of environmental chloride itself caused a decrease in THC,  indicating stress.

       In studies where freshwater crayfish Astacus astacus were exposed to 0.8 mM
nitrite [11.2 mg/L NO2-N] for 7  days Jensen (1990) found a rapid accumulation of nitrite
in hemolymph to concentrations much higher than the exposure concentration,  plateauing
at an internal/external ratio of about 10. There was essentially no mortality at the test
concentration. Hemocyanin concentration of the hemolymph decreased significantly, and
oxygen tension increased. Chloride and sodium concentrations in hemolymph decreased
and plateaued after 2 days. Potassium,  calcium, and magnesium concentrations of the
hemolymph remained essentially constant throughout the 7-d test period.

       A number of toxicity studies have been conducted on shrimp, and reported 96-h
LCso values for shrimp are summarized in Table II.

       Shrimp Litopenaeus vannamei were exposed for 4 days to nitrite concentrations
of <0.1 to 8.8 mg/L NO2-N under different conditions of salinity,  chloride and calcium
levels (Sowers etal., 2004). Hemolymph nitrite increased significantly and was dose-
dependent and higher than environmental nitrite concentrations. Hemolymph nitrite
concentrations were inversely related to the level of dissolved solids and to the chloride
concentrations. Nitrite uptake did not interfere with normal water and ion balance. 96-h
LCso values ranged from 8.4 to 30 mg/L MVN under conditions of total dissolved solids
of 2 to 10 g/L. Increasing total dissolved salt and chloride concentration resulted in
reduced toxicity. They concluded that large quantities of salts would be required to
manage water quality in shrimp ponds and suggested that a more economical approach
would be to reduce pond inputs of nitrogen through stocking  and feeding rates and feed
formulations, and removal of pond wastes.

       Lin and Chen (2003) conducted acute toxicity studies on Litopenaeus vannamei
juveniles at salinity levels of 15, 25, and 35 %o. The 96-h LCso values were 76.5 mg/L
NO2-N at 15 %o, 178.3 mg/L at 25 %o, and 321.7 mg/L at 35 %o. As the salinity decreased
from 35 %o to 15 %o, susceptibility to nitrite increased by 421% after 96 h exposure. The
tests were continued to 144 h, with the LCso values found at 144 h being 61.1, 152.4,
257.2 mg/L MVN at 15, 25, and 35 %o, respectively. The authors suggested that a "safe
level" for culturing L. Vannamei juveniles would be 6.1, 15.2, and 25.7 mg/L NO2-N in
15,25, and35%0 salinity.
                                        79

-------
       The toxicity of nitrite on Penaeus monodon adolescents in 20 ppt seawater was
reported by Chen etal. (1990a) to be 218, 193, 171, 140, 128, and 106 mg/L NO2-N LC50
values at 24, 48, 96, 144, 192, and 240 h, respectively. The toxicity curve approached an
asymptote at 240 h. The authors suggested that a safe value for culturing adolescent P.
monodon at 20 ppt salinity, 7.57 pH, and 24.5 °C was 10.50 mg/L NO2-N. In similar
tests on Penaeus chinensis in 33 %o seawater (Chen et a/., 1990b), the LCso values at 24,
96, 120, 144, and 192 h were 339, 37.71, 29.18, 26.98, and 22.95 mg/L, respectively. The
asymptotic LC50 was 22.95 mg/L NO2-N at 192 h. The authors suggested that a "safe
value" for P. chinensis juveniles was 2.30 mg/L NO2-N.

       Chen and Cheng (1995) studied sublethal effects of nitrite exposure in Penaeus
japonicus. Nitrite exposures ranged from 5.12 to 50.86 mg/L NO2-N in 30 ppt seawater
for 24 h. Nitrite and urea accumulated in the hemolymph and excretion of ammonia
increased with increasing nitrite over the entire concentration range tested. Nitrite
exposure decreased the oxyhemocyanin and the ratio of oxyhemocyanin/protein. Cheng
and Chen (2001) also investigated the time-course change of nitrogenous excretion in
Penaeus japonicus after 48-h exposure to nitrite concentrations ranging from 0.076 to
1.433 mM at 30 ppt salinity. Hemolymph nitrite and urea increased with nitrite
concentration and exposure time,  whereas hemolymph ammonia was inversely related to
nitrite concentration and exposure time.  Excretion of total nitrogen, ammonia nitrogen,
urea nitrogen, and organic nitrogen  increased with nitrite concentration and exposure
time. The authors found that exposure of P. japonicus for 24 h to nitrite concentrations as
low as 0.076 mM increased ammonia nitrogen excretion by a factor of 1.9, urea nitrogen
excretion by 200, and organic nitrogen excretion by 37 as compared to controls.

       Sublethal effects of nitrite exposure to Penaeus monodon were studied by Chen
and Cheng (2000). Exposure concentration was 0.72 mM nitrite at pH values of 6.8, 8.2,
and 9.8, and exposure time was 48 h. Nitrite influx, hemolymph nitrite and osmotic
differential increased with exposure time and were higher at pH 6.8. Water nitrite
concentration, oxyhemocyanin, protein,  the oxyhemocyanin/protein ratio, and
hemolymph osmolality decreased with exposure time and were lower at pH 6.8. The
same parameters were measured for shrimp placed in nitrite-free water after 3,6,  12, and
24 h following 48-h exposure to nitrite. Water nitrite concentration increased with
depuration time and was higher at pH 6.8, whereas hemolymph oxyhemocyanin,
oxyhemocyanin/protein ratio, and hemolymph osmolality increased with depuration time
and were higher at pH 9.8. Hemolymph  nitrite decreased with depuration time and after
24 h, depuration was undetectable, with 72-88% recovery of oxyhemocyanin.
       Penaeus setiferus postlarvae were exposed to NO2-N concentrations of 25, 50,
and 100 mg/L for 72 h in an investigation (Alcaraz etal., 1997) of effects of ammonia,
nitrite, and their combination to thermal tolerance  of the shrimp. Mortalities of 10, 20,
and 35% occurred at nitrite concentrations of 25, 50, and 100 mg/L, respectively. The 72-
h LCso was 172.8 mg/L NO2-N. The temperature tolerance of the tested shrimp
postlarvae was unaffected by the 25 mg/L NO2-N exposure. However, the temperature
tolerance was reduced by 5.0 and 8.4% at exposures of 50 and 100 mg/L, respectively.
No difference was observed in critical thermal maximum values between organisms
exposed to 50 and 100 mg/L. Four different ammonia-nitrite combination exposures were
tested, with concentrations being based on the individual toxicant's LCso value. The
researchers found no relationship between the mortality rate and the toxicity ratio of

                                        80

-------
ammonia and nitrite mixtures. The authors concluded that mortality rates were related to
ammonia concentration and not to the toxicity ratios of ammonia and nitrite mixtures.
However, the thermal tolerance of the shrimp exposed to ammonia-nitrite mixtures was
related to the toxicity ratio of the mixtures. The critical thermal maximum did not
decrease in organisms exposed to a joint concentration of 0.8 mg/L NH3-N and 60 mg/L
NO2-N, but decreased significantly compared to controls in the higher joint
concentrations tested. The authors concluded that the toxic contribution of both
compounds determined the thermal response of the shrimp.

       Koo etal. (2005) investigated the effect of nitrite on juvenile tiger crab (Orithyia
sinicd) after 30-d exposures to nitrite concentrations of 50, 100, 150, 200, and 250 mg/L
NO2-N. Survival rates decreased linearly with concentration and exposure time. The
growth rate of the crabs decreased at 150, 200, and 250 mg/L nitrite. The intermolt period
of the crabs was shortened between the first  and second molt, and the numbers of
moltings of crabs exposed to higher  concentrations were significantly higher than that of
controls.

       Toxicity tests on seven species of aquatic invertebrates have been carried out by
Russo and coworkers (unpublished data):  two species of Ephemeroptera (mayfly),
Ephemerella doddsi and Ephemerella grandis; three species of Plecoptera (stonefly),
Arcynopteryxparallela, Pteronarcella badia., andlsoperlafulva; one species of
Tricoptera (caddisfly), Arctopsyche grandis; and one species of Diptera (true fly), Atherix
variegata. All of these species are common to cold water environments in the Western
United States. The 96-h LCso values  for the mayflies ranged from 0.52-2.00 mg/L NC>2-
N; for the stoneflies from 0.25-0.46 mg/L; and for the caddisfly 1.02-2.43 mg/L. No LC50
was obtainable for the true fly Atherix variegata because fewer than 50% of the test
organisms died at the highest concentrations used in two tests; there was 38% mortality at
123 mg/L NO2-N, the highest concentration  tested. E. grandis and P. badia were also
tested for nitrite toxicity with added  sodium  chloride to investigate whether added
chloride reduced nitrite toxicity in these insect species, as is commonly observed with
fishes. The addition of chloride ion greatly reduced nitrite toxicity in E. grandis and P.
badia. The LCso with added chloride was  3.5-10 times greater than without chloride for
E. grandis; the LCso was 30-50 times greater than without chloride for P. badia.
                                        81

-------
Table I.  Toxicity of Nitrite in Different Fish Species.
Species
Rainbow trout
Oncorhynchus mykiss
Cutthroat trout
Salmo clarkii
Channel catfish
Ictalurus punctatus
Walking catfish
Glorias lazera
Grass carp
Ctenopharyngodon idella
Mrigal
Cirrhinus mrigala (Ham.)
Indian major carp
Catla catla Hamilton
Goldfish
Carassius auratus
Fathead minnow
Pimephales promelas
Mosquitofish
Gambusia qfftnis
Striped bass
Morone saxatilis
Green sunfish
Lepomis cyanellus
Bluegill
Lepomis macrochirus
Largemouth bass
Micropterus salmoides
Smallmouth bass
Micropterus dolomieui
Guapote tigre
Cichlasoma managuense
Siberian sturgeon
Acipenser baeri,Rrandt
Nile tilapia
Oreochromis niloticus
Blue tilapia
Oreochromis aureus
Mullet
Mugil platanus
Southern flounder
Paralichthys lethostigma
Mottled sculpin
Coitus bairdi
96-h LC50 (mg/L N)
0.2-0.3
0.5-0.6
7.1
28-32
6-13
10.4
120.84:24h static
17.43:24hflwth
52
2.3-3.0
1.5
50 (24 h)
160
80
140.2
160
20.4
130 (72 h)
(130.5 mg/L Cl')
8 to 81
16.2
35.9-36.2 SW
1.51FW
35.2
>67
Reference
Russo etal. (1974);
Russo and Thurston (1977)
Thurston et al. (1978)
Tomasso (1986)
Hilmy etal. (1987)
Alcaraz and Espina (1995)
Das et al. (2004a)
Tilak et al. (2002)
Tomasso (1986)
Russo and Thurston (1977)
Wallen etal. (1957)
Mazik etal. (1991)
Tomasso (1986)
Tomasso (1986)
Palachek and Tomasso
(1984b)
Tomasso (1986)
Chin and Shy ong ( 1998)
Huertas et al. (2002)
Atwoode^a/. (200 Ib)
Palachek and Tomasso
(1984b)
Sampaio et al. (2002)
Atwood etal. (200 la)
Russo and Thurston (1977)
                                          82

-------
Table II.  Nitrite Toxicity in Shrimp
Species
Litopenaeus
vannamei
Penaeus monodon
Penaeus chinensis
Penaeus penicillatus
Penaeus setiferus
96-h LC50
(mg/L NO2-N)
8.4-30
77-322
171
38
39
173 (72-h)
Salinity
(g/L)
2-10
15-35
20
33
25
25
Reference
Sowers et al. (2004)
Lin and Chen
(2003)
Chenetal. (1990a)
Chenetal. (1990b)
Chen and Lin
(1991)
Alcaraz ^a/. (1997)
References

Alcaraz, G., X. Chiappa-Carrara, and C. Vanegas. 1997. Temperature tolerance of
       Penaeus setiferus postlarvae exposed to ammonia and nitrite. Aquat. Toxicol. 39:
       345-353.

Alcaraz, G. and S. Espina. 1995. Acute toxicity of nitrite in juvenile grass carp modified
       by weight and temperature. Bull. Environ. Contam. Toxicol. 55: 473-478.

Alcaraz, G. and S. Espina. 1997. Scope for growth of juvenile grass carp
       Ctenopharyngoda idella exposed to nitrite. Comp. Biochem. Physiol. 116C: 85-
Almendras, J.M.E. 1987. Acute toxicity and methemoglobinemia in juvenile milkfish
       (Chanos chanos Forsskae). Aquaculture 61: 33-40.

Atwood, H.L., J.R. Tomasso Jr., and T.I.J. Smith. 2001a. Nitrite toxicity to southern
       flounder Paralichthys lethostigma in fresh and brackish water. J. World
       Aquaculture Society 32(3): 348-351.

Atwood, H.L., Q.C. Fontenot, J.R. Tomasso, and J.J. Isely. 2001b. Toxicity of nitrite to
       Nile tilapia: effect offish size and environmental chloride. North American
       Journal of Aquaculture 63: 49-51.

Bowser, P.R., W.W. Falls, J. VanZandt, N. Collier, and J.D. Phillips. 1983.
       Methemoglobinemia in channel catfish: methods of prevention. Prog. Fish-Cult.
       45: 154-158.

Carballo, M. and M.J. Munoz. 1991. Effect of sublethal concentrations of four chemicals
       on susceptibility of juvenile rainbow trout (Oncorhynchus mykiss) to
       saprolegniosis. Appl. Environ. Microbiol. 57: 1813-1816.
                                       83

-------
Chen, J-C., P-C. Liu, and S-C. Lei. 1990a. Toxicities of ammonia and nitrite io Penaeus
       monodon adolescents. Aquaculture 89: 127-137.

Chen, J-C., Y-Y. Ting, J-N. Lin, and M-N. Lin.  1990b. Lethal effects of ammonia and
       nitrite onPenaeus chinensis juveniles. Mar. Biol. 107: 427-431.

Chen, J-C. and C-Y. Lin. 1991. Lethal effects of ammonia and nitrite on P. Penicillatus
       juveniles at two salinity levels. Comp. Biochem. Physiol. 100C: 477-482.

Chen, J-C. and S-Y.  Cheng. 1995. Changes of oxyhemocyanin and protein levels in the
       hemolymph of Penaeus japonicus exposed to ambient nitrite. Aquat. Toxicol. 33:
       215-226.

Chen, J-C. and S-Y.  Cheng. 2000. Recovery of Penaeus monodon from functional
       anaemia after exposure to sublethal concentration of nitrite at different pH levels.
       Aquat.Toxicol. 50: 73-83.

Cheng, S-Y. and J-C. Chen. 2001. The time-course change of nitrogenous excretion in
       the Kuruma shrimp Penaeus japonicus following nitrite exposure. Aquat.Toxicol.
       51:443-454.

Chin, T-S. and W-J.  Shyong. 1998. Short-term toxicity of nitrogenous compounds to the
       fry of guapote tigre, Cichlasoma managuense. J. Fish. Soc. Taiwan 25: 295-302.

Crawford, R.E. and G.H. Allen. 1977. Seawater inhibition of nitrite toxicity to chinook
       salmon. Trans. Amer. Fish. Soc. 106:  105-109.

Das, P.C., S. Ayyappan, J.K. Jena, and B.K. Das. 2004a. Nitrite toxicity in Cirrhinus
       mrigala (Ham.): acute toxicity and sub-lethal effect on selected haematological
       parameters. Aquaculture 235:  633-644.

Das, P.C., S. Ayyappan, J.K. Jena, and B.K. Das. 2004b. Effect of sub-lethal nitrite on
       selected haematological parameters in fmgeiiing Catla catla (Hamilton). Aquae.
       Res. 35: 874-880.

Das, P.C., S. Ayyappan, B.K. Das, and J.K. Jena. 2004c. Nitrite toxicity in Indian major
       carps: sublethal effect on selected enzymes in fingerlings of Catla catla, Labeo
       rohita, and Cirrhinus mrigala. Comp. Biochem. Physiol 138C: 3-10.

Doblander, C. and R. Lackner. 1996. Metabolism and detoxification of nitrite by trout
       hepatocytes. Biochim. Biophys. Acta  1289: 270-274.

Eddy, F.B., P.A. Kunzlik, and R.N. Bath. 1983. Uptake and loss of nitrite from the blood
       of rainbow trout, Salmo gairdneri Richardson, and Atlantic salmon, Salmo salar
       L. in fresh water and in dilute sea water. J. Fish Biol. 23: 105-116.
                                       84

-------
Folmar, L.C., S. Bonomelli, T. Moody, and J. Gibson. 1993. The effect of short-term
       exposure to three chemicals on the blood chemistry of the pinfish (Lagodon
       rhomboides).  Arch. Environ. Contam. Toxicol. 24: 83-86.

Fontenot, Q.C., JJ. Isely, and J.R. Tomasso. 1999. Characterization and inhibition of
       nitrite uptake in shortnose sturgeon fingerlings. J. Aquat. Anim. Health 11: 76-80.

Grosell, M. and F.B. Jensen. 1999. NCV uptake and HCCV excretion in the intestine of
       the European flounder (Platichthysflesus). J. Exp. Biol. 202: 2103-2110.

Hilmy, A.M., N.A. El-Domiaty, and K. Wershana. 1987. Acute and chronic toxicity of
       nitrite to Clarias lazera. Comp. Biochem. Physiol. 86C:  247-253.

Huertas, M., E. Gisbert, A. Rodriguez, L. Cardona, P.  Williot, and F. Castello-Orvay.
       2002. Acute exposure of Siberian sturgeon (Acipemer baeri, Brandt) to nitrite:
       median-lethal concentration (LC50) determination, haematological changes and
       nitrite accumulation in selected tissues. Aquat. Toxicol. 57: 257-266.

Jensen, F.B. 1990. Sublethal physiological changes in freshwater crayfish, Astacus
       astacus, exposed to nitrite: haemolymph and muscle tissue electrolyte status, and
       haemolymph acid-base balance and gas transport. Aquat. Toxicol. 18: 51-60.

Jensen, F.B. 2003. Nitrite disrupts multiple physiological functions in aquatic animals.
       Comp. Biochem. Physiol. 135A: 9-24.

Koo, J-G, S-G Kim, J-H. Jee, J-M.  Kim, S.C. Bai, and J-C. Kang. 2005. Effects of
       ammonia and nitrite on survival, growth and moulting in juvenile tiger crab,
       Orithyia sinica (Linnaeus). Aquae. Res. 36: 79-85.

Kumta A. and S.A. Gaikwad. 1997. Fecundity and gonadosomatic index  ofGambusia
       affinis affected by nitrite toxicity. Environ. Ecol. 15: 834-837.

Lewis, W.M. Jr. and D.P. Morris. 1986. Toxicity  of nitrite to fish:  a review. Trans.
       American Fish. Soc.  115: 183-195.

Lin, Y-C. and J-C. Chen. 2003. Acute toxicity of nitrite on Litopenaeus vannamei
       (Boone) juveniles at different salinity levels. Aquaculture 224: 193-201.

Mazik, P.M., M.L. Hinman, D.A. Winkelmann, S.J. Klaine, B.A. Simco, andN.C.
       Parker. 1991. Influence of nitrite and chloride concentrations on survival and
       hematological profiles of striped bass. Trans. American Fish. Soc. 120: 247-254.

Michael, M.L, A.M. Hilmy, N.A. El-Domiaty, and K.  Wershana. 1987. Serum
       transaminases activity and histopathological changes in Clarias lazera chronically
       exposed to nitrite. Comp. Biochem. Physiol. 86C: 255-262.

Palachek, R.M. and J.R. Tomasso. 1984a. Nitrite  toxicity to fathead minnows: effect of
       fish weight. Bull. Environ. Contam. Toxicol. 32: 238-242.

                                       85

-------
Palachek, R.M. and J.R. Tomasso. 1984b. Toxicity of nitrite to channel catfish (Ictalurus
      punctatus), tilapia (Tilapia aurea), and largemouth bass (Micropterus salmoides):
       evidence for a nitrite exclusion mechanism. Can. J. Fish. Aquat. Sci. 41: 1739-
       1744.

Perrone, S.J. and T.L. Meade. 1977. Protective effect of chloride on nitrite toxicity to
       coho salmon (Oncorhynchus kisutch).  J. Fish. Res. Board Canada 34: 486-492.

Russo, R.C. 1980. Recent advances in the study of nitrite toxicity to fishes. Pages 226-
       240 In: Proceedings of the Third USA-USSR Symposium on the Effects of
       Pollutants upon Aquatic Ecosystems. July 2-6, 1979, Borok, Jaroslavl Oblast,
       USSR. W.R. Swain and V.R.  Shannon (Eds.). EPAEcol.  Res. Ser. EPA-600/9-
       80-034. U.S. Enviromental Protection Agency, Duluth, MN.

Russo, R.C. and R.V. Thurston. 1977. The acute toxicity of nitrite to fishes. Pages 118-
       131 In: Recent Advances in Fish Toxicology R.A. Tubb (Ed.). EPA Ecol. Res.
       Ser. EPA-600/3-77-085. U.S. Environmental Protection Agency, Corvallis, OR.

Russo, R.C., R.V. Thurston, and K. Emerson.  1981. Acute toxicity of nitrite to rainbow
       trout (Salmo gairdneri): Effects of pH, nitrite species, and anion species. Can. J.
       Fish. Aquat. Sci. 38: 387-393.

Russo, R.C., C.E. Smith, and R.V. Thurston. 1974. Acute toxicity of nitrite to rainbow
       trout (Salmo gairdneri). J. Fish. Res. Board Canada 31:1653-1655.

Sampaio, L.A., W. Wasielesky , and K.C. Miranda-Filho. 2002. Effect of salinity on
       acute toxicity of ammonia and nitrite to juvenile Mugilplatanus. Bull. Env.
       Contam. Toxicol. 68: 668-674.

Saroglia, M.G., G. Scarano, and E. Tibladi.  1981. Acute toxicity  of nitrite to sea bass
       (Dicentrarchus labrax) and European eel (Anguilla anguilla). J. World
       Mariculture Soc. 12: 121-126.

Sipauba Tavares, L.H. and C.E. Boyd. 2003. Possible effects of sodium chloride
       treatment on quality of effluents from Alabama channel catfish ponds. J. World
       Aquaculture Society 34: 217-222.

Sowers, A., S.P. Young, JJ. Isely, C.L. Browdy, and J.R. Tomasso Jr. 2004. Nitrite
       toxicity to Litopenaeus vannamei in water containing low concentrations of sea
       salt or mixed salts. J. World Aquaculture Society 35: 445-451.

Stormer, J., F.B. Jensen, and J.C. Rankin. 1996. Uptake of nitrite, nitrate and bromide in
       rainbow trout, Onchorhynchus mykiss: effects on ionic balance. Can. J. Fish.
       Aquat. Sci. 53: 1943-1950.

Thurston, R.V., R.C. Russo, and C.E. Smith. 1978. Acute toxicity of ammonia and nitrite
       to cutthroat trout fry. Trans. American Fish. Soc. 107: 361-368.

                                       86

-------
Tilak, K.S., S. J. Lakshmi, and T. A. Susan. 2002. The toxicity of ammonia, nitrite and
       nitrate to the fish, Catla catla (Hamilton). J. Environ. Biol. 23: 147-149.

Tomasso, J.R. 1986. Comparative toxicity of nitrite to freshwater fishes. Aquat. Toxicol.
       8: 129-137.

Tomasso, J.R. 1994. Toxicity of nitrogenous wastes to aquaculture animals. Reviews in
       Fisheries Science 2: 291-314.

Tomasso, J.R., B.A. Simco, and K.B. Davis. 1979. Chloride inhibition of nitrite-induced
       methemoglobinemia in channel catfish (Ictaluruspunctatus).  J. Fish. Res. Board
       Canada 36: 1141-1144.

U.S. EPA (U.S. Environmental Protection Agency). 2002. ECOTOX User Guide:
       ECOTOXicology Database System. Version 3.0. Available at:
       http :/mountain. epa.gov/ecotox/.

Vedel, N.E., B. Korsgaard, and F.B. Jensen. 1998. Isolated and combined exposure to
       ammonia and nitrite in rainbow trout (Oncorhynchus mykiss}: effects on
       electrolyte status, blood respiratory properties and brain glutamine/glutamate
       concentrations. Aquat. Toxicol. 41: 325-342.

Wallen, I.E., W.C. Greer, and R. Lasater. 1957. Toxicity to Gambusia affinis of certain
       pure chemicals in turbid waters. Sewage Ind. Wastes 29: 695-711.

Watenpaugh, D.E., T.L. Beitinger, and D.W. Huey. 1985. Temperature tolerance of
       nitrite-exposed channel  catfish. Trans. American Fish. Soc. 114: 274-278.

Wedemeyer, G.A. and W.T. Yasutake. 1978. Prevention and treatment of nitrite toxicity
       in juvenile steelhead trout (Salmo gairdneri).  J. Fish. Res. Bd. Canada 35: 822-
       827.

Wise, D.J. and J.R. Tomasso. 1989. Acute toxicity of nitrite to red drum Sciaenops
       ocellatus: effect of salinity. J. World Aquaculture Soc. 20: 193-198.

Woo, N.Y.S.  and F.S. Chiu. 1997. Metabolic and osmoregulatory responses of the  sea
       bass Lates calcarifer to  nitrite exposure. Environ. Toxicol. Water Qual. 12: 257-
       264.

Yildiz H.Y. and A.C.K. Benli. 2004. Nitrite toxicity to crayfish, Astacus leptodactylus,
       the effects of sublethal nitrite exposure on hemolymph nitrite, total hemocyte
       counts, and hemolymph glucose. Ecotoxicology and Environmental Safety  59:
       370-375.
                                       87

-------
88

-------
   The formation of S-nitrosoglutathione in conditions

                 mimicking hypoxia and acidosis

                                      by

                                   L. Gross!1


Introduction

      Nitric Oxide (NO) is synthesized from L-arginine by nitric oxide synthase (NOS)
with the involvement, among other species, of molecular oxygen. Nitrite and nitrate are
believed to be the waste forms of NO. During hypoxia/ischemia elevation of NO
production occurs which, in principle, contrasts with the enzymatic mechanism which
requires the presence of oxygen. A question arises from this situation; are the waste
forms of NO produced in an irreversible process? It would seem more reasonable to think
that NO metabolites can be recycled back to bioactive NO again. During renal vascular
occlusion (Okamoto et a/., 2005) as well as in an ischemic heart (Zweier et a/., 1995), the
pH rapidly decreases, and the production of NO increases sharply, independent of L-
arginine administration. This shows that the direct reduction of nitrite (or its derivatives)
to NO, under acidic conditions, has been invoked in a NOS-independent manner.

Discussion

      The knowledge that in acidic conditions inorganic nitrites can dilate vessels is
well known, and different concentrations of these species are present in vivo (100-500
DM).  In these conditions, the ac/'-form of the nitrous  acid, and consequently the amount
of nitrous anhydride N2O3 increases (reactions [1] and [2]).
                                                  [1]

                                                  [2]

       To account for the formation of NO, the reduction of HNO2, as well as the
spontaneous decomposition of N2Os (Lundberg and Weitzberg, 2005), into NO and other
NOx, has been invoked (reactions [3] and [4]).
                                                  [3]

                                                  [4]
1 Dipartimento di Chimica Organica "A. Mangini". Universita di Bologna - Viale
Risorgimento, 4 1-40136 Bologna, Italy: loris.grossi@unibo.it.

                                      89

-------
       Concerning the latter mechanism, recently, Butler and Ridd (2004) have discussed
the mechanism of formation of NO from nitrous acid under conditions of low pH. They
concluded that the formation of N2Os is the rate determining step and its homolytic
cleavage, to form NO, is very slow. In particular, also at the lowest physiological pH (e.g.
in ischemic situations), the amount of NO produced via such a process has been found to
be lower than that produced in vivo. So,  the extremely low rate of conversion of a nitrite
derivative, via homolysis, into NO and NO2, leads us to consider the key step in the NO
production, a reduction process. In fact,  it is reported that the amount of NO generated is
dependent not only on the nitrite concentration and the pH, but also on the presence of
reducing agents,  such as vitamin C (Heller etal., 1999); proximity to heme groups,
proteins and thiols; and oxygen tension.

       Several processes involving NO which take place in hypoxic and/or acidotic
conditions, could be due to the reaction of different reducing agents with N2Os  and
supports the idea that the nitrite anion is a non enzymatic reservoir of NO. The in vivo
increase of NO production with decreasing pH, reaches values similar to those  seen
during hypoxia. In this condition, an  increase of the amount of N2Os can be hypothesized
and its interaction with Hb-Fe (reaction  [5]) could lead to met-Hb, free NO and the nitrite
anion (or HNO2).
       Hb-Fe   	-      Hb-Fe    + NO + NO2"                    [5]

       According to this mechanism, as hemoglobin deoxygenates, vacant hemes
become nitrite reductase systems generating methemoglobin and NO (Henry, 1999). This
reaction, in principle, can account for several findings. It reduces the amount of Hb,
thereby increasing the amount of met-Hb and thus justifying the unequal amounts of
MetHb-NO and Hb-NO experimentally detected. This leads to the hypothesis that
significant concentrations of NO can avoid the hemoglobin-trap, accounting thus for the
vasodilation (Doylee^a/., 1981)

       In this view, the mechanism of formation in vivo (Spencer etal., 2000) of Hb-
SNO, and the functional consequences of this modification, could also be involved. The
S-nitrosated Hb is a modified form of Hb that has been postulated to be implied in a
dynamic cycle of intravascular NO uptake, transport, and delivery, and its potential role
as an NO transporter is very attractive. However, the environment of |3-93Cys is sensitive
to the R <-»• T conformational equilibrium of Hb and studies have shown that the |3-93Cys
residues, at which NO can be bound and then released, are more accessible in the high-
affinity conformation of oxy (R-state) Hb than  in deoxy (T-state) Hb (Bonaventura  et a/.,
2004). In particular, the normal deoxy conformation of Hb (T-state) cannot be assumed
with NO on |3-93Cys, even if proteins could in  principle have sufficient conformational
flexibility to accommodate a possible S-NO linkage. The hemoglobin is forced into the
R-state when these thiols are modified, as evidenced by the easier formation of Hb-SNO
with oxygenated Hb (arterial blood), than with  deoxygenated Hb (venous blood).

       The role of Hb-SNO, and other S-nitrosothiols, for example, as NO releasers via
homolysis,  or their ability to perform transnitrosation  (Crawford, White and Patel, 2003),

                                        90

-------
would seem unusual in a non pathological situation. But, the primary area of debate is in
trying to identify the extent to which the destabilized T-state of Hb-SNO enhances NO
release and/or the transnitrosation process (McMahon, Gow and Stammler, 2000). In
particular, it has been hypothesized that Hb-SNO elicits NO-dependent vasodilation in
environments of low oxygen tension thereby stimulating blood flow to these regions.
Deoxygenation of Hb-SNO is followed by the transfer of the NO group via a
transnitrosation reaction to thiols such as glutathione (GSH).

       The biochemical alterations induced by NO consumption show a possible
pathogenesis of a variety of diseases via stimulation of a NO-derivative involved in blood
flow regulation which is not susceptible to rapid reactions with oxyhemoglobin, and that
can be readily converted back to NO when required (Webb et a/., 2004). Thus, besides S-
nitrosothiols (Gaston,  1999), could nitrite (Cosby etal., 2003;  Gladwin and Schechter,
2004) also be considered a regulator of vasodilation in non-severe pathological
conditions such as a slightly lower physiological pH? If, even  a small to gentle fall in pH
corresponds to an increase in the concentration of nitrous acid, and then nitrous
anhydride, the latter could react with reducing species present  in the medium such as
vitamin C producing free NO. A possible indirect action of NO as a regulator could be
carried out by thiol derivatives, which, reacting via an in cage  process with the nitrous
anhydride, lead to S-nitrosothiols thus avoiding the formation  of free NO. This is
confirmed from studies on NO-hemoglobin biochemistry which show that the proteins S-
nitrosation in red blood cells occurs, if only at low levels and despite the presence of high
affinity heme sinks for NO (Gladwin etal, 2003; Gladwin etal, 2002; Jia etal, 1996).

       In a previous study (Grossi and Montevecchi, 2002) on the nitrosation of cysteine
in acidic conditions, the final derived kinetic expression showed that the NO+ nitrosating
agent could be involved only in very acidic conditions (pH<3.5). At a higher pH (>3.5),
the rate of nitrosation achieved depended on [HNO2]2 and was independent of the
cysteine concentration which suggested that the formation of N2Os was the rate
determining step, and was the nitrosating species, not the nitrous acid itself (Williams,
1985, 2003). However, the detection of CySNO failed at pH > 5.3 and therefore no
information about its possible formation at pHs closer to physiological pH could be
obtained. In light of these results, we considered it of interest to verify the behavior of
glutathione, probably one of the molecules responsible for the  transport of NO  as GSNO
in vivo. GSNO is known to be a stable S-nitroso derivative, thus it was interesting to
verify if it could be formed by direct nitrosation at a higher pH compared to those used
for obtaining CySNO. To verify this hypothesis, experiments were conducted with
glutathione and NaNO2, in buffered deoxygenated aqueous solution, in a pH range from 4
to 6.86, mimicking physiological conditions of acidosis and hypoxia. The formation of
GSNO was detectable throughout the entire pH range.
      2NO~+2H+      »W  2HNO2^i^i  N2O3

                           H+
           N2O3  +GSH 	*- GSNO+ HNO2                       [6]

       But, the detection of GSNO at pH 6.86, a physiological pH, was unexpected and
definitely a remarkable result (Fig. 1).

                                        91

-------
                0,5 -
                                                             pH4.0




                                                             pH5.0

                                                             pH6.0
                                                             pH6.86
                                       10         15
                                     T (min.)
20
             Figure 1. Plots of A vs. time (min) for reactions of 0.04 M GSH with 0.04
                       M NaNC>2 in buffered deoxygenated aqueous solution, pH 4.0 -
                       6.86, at37°C.

       The direct formation via nitrosation of an S-nitrosothiol at so high a pH has not
been reported before, except to invoke transnitrosation. This result led us to formulate a
reaction mechanism excluding the involvement of NO+ (or H2O-NO+), because it can be
formed only at very low pH, and NO, because it cannot act as a direct nitrosating agent.
As for the reaction between cysteine and NaNC>2,  in mildly acidic conditions an Electron
Transfer process (SET) between the glutathione and N2Os occurs. Via the formation of a
radical ion pair in cage a S-nitroso derivative is formed followed by the loss of HNO2
(reaction [6]), underlining that ^63 can act efficiently as the nitrosating species at low
concentration (Hughes, Ingold and Ridd, 1958). This result was also in agreement with
the mechanism we hypothesized (Grossi, Montevecchi and Strazzari, 200la and 200Ib)
for the nitrosation of alcohols in organic solvents, and for the nitrosation ofp-cresol and
styrene, in neutral or acidic conditions, by ^63 (NO/Ch). In light of these findings,
preliminary studies on the kinetics of this process show the linear dependence of the rate
of formation of GSNO on pH (Fig. 2).
                                        92

-------
             3,5x10"2-
             3,0x1 0"2-
             2,5x1 0"2-
             2,0x1 0"2-
             1,5x10-2-
           CO

          >  5,0x10"3-

                0,0-
                      4,0
4,5
5,0
5,5

PH
6,0
6,5
7,0
Figure 2. Plot of FGSNo vs. pH for reactions of 0.04 M GSH with 0.04 M NaNO2, at 37°C.

Conclusion

       These results seem to stress that the reaction between GSH and NaNO2 in
anaerobic and aqueous buffer solutions can lead to the detection of the corresponding
GSNO at pH mimicking physiological conditions. Preliminary results on the kinetic
behavior of this reaction led to the hypothesis that N2Os is the oxidant species, whose
amount depends on its equilibrium with HNO2. It also allows us to hypothesize that the
interaction between N2Os and haemoglobin is the process leading to the formation of free
NO, responsible for vasodilation in hypoxia and/or acidotic conditions.

Experimental

       The reactions were conducted in a thermostated bath at 37°C. The formation of
GSNO was constantly monitored by measuring the absorbance at A=543 nm (s = 16.34
M"1 cm"1), pumping the reacting solution using a peristaltic pump through a UV/vis
spectrophotometer equipped with a flow-cell, and a computer for the continuous
acquisition  of data; the absorbance was detected every three seconds. The solutions were
prepared using buffer standard solution carefully deoxygenated by bubbling with pure
nitrogen.

Acknowledgment

       This work was financially supported by the Ministry of the University and
Scientific and Technological Research (MURST), Rome (funds 60% and 40%).
                                       93

-------
References
Bonaventura, C., A. Fago, R. Henkens, and A.L. Crumbliss. 2004. Antioxid. Redox
       Signal. 6:979-991.

Butler, A.R. and J.H. Ridd. 2004. Formation of nitric oxide from nitrous acid in ischemic
       tissue and skin. Nitric Oxide 10: 20-24.

Cosby, K., K.S. Partovi, J.H. Crawford, R.P. Patel, C.D. Eiter, S. Martyr, B.K. Yang,
       M.A. Waclawiw, G. Zalos, X. Xu, K.T. Huang, H. Shields, D.B. Kim-Shapiro,
       A.N. Schechter, R.O. Cannon, and M.T. Gladwin. 2003. Nitrite reduction to nitric
       oxide by deoxyhemoglobin vasodilates the human circulation. Nat. Med. 9: 1498-
       1505.

Crawford, J.H., C.R. White, and R.P. Patel. 2003. Vasoactivity of S-nitrosohemoglobin:
       role of oxygen, heme, and NO oxidation states. Blood. 101: 4408-4415.

Doyle, M.P., R.A. Pickering, T.M. DeWeert, J.W. Hoekstra, and D. Pater.  1981. Kinetics
       and  mechanisms of the oxidation of human deoxyhemoglobin by nitrites. J. Biol.
       Chem. 256: 12393- 12398.

Gaston, B. 1999. Nitric oxide and thiol groups. Biochim. Biophys. Acta. 1411: 323-33.

Gladwin, M.T. and A.N. Schechter. 2004. Nitrite Versus ^-Nitroso-Hemoglobin. Circ.
       Res. 94: 851-856.

Gladwin, M.T., J. R. Lancaster, B.A. Freeman, and A.N. Schechter. 2003. Nitric oxide's
       reaction with hemoglobin: a view through the SNOstorm. Nat. Med. 9: 496- 500.

Gladwin, M.T., X. Wang, C.D. Reiter, B.K. Yang, E. X. Vivas, C. Bonaventura, and
       A.N. Schechter. 2002. S-Nitrosohemoglobin is unstable in the reductive red cell
       environment and lacks  O2/NO-linked allosteric function. J. Biol. Chem. 21:
       27818-27828.

Grossi, L. and P.C. Montevecchi. 2002. ^-Nitrosocysteine and cystine from reaction of
       cysteine with nitrous acid. A kinetic investigation. J. Org. Chem. 67: 8625-8930.

Grossi, L., P.C. Montevecchi, and S. Strazzari. 2001a. Decomposition of ^-Nitrosothiols:
       Unimolecular versus Autocatalytic Mechanism. J. Am. Chem. Soc. 123: 4853-
       4854.

Grossi, L., P.C. Montevecchi, and S. Strazzari. 2001b. The Chemistry of Peroxynitrite:
       Involvement of an ET Process in the Radical Nitration of Unsaturated and
       Aromatic Systems. Eur. J. Org. Chem. 2001(4): 741-748.

Heller, R., F. Munscher-Paulig, R. Grabner and U.Till. 1999. J. Bio.  Chem. 274: 8254-
       8260.

                                       94

-------
Henry, Y.A. 1999. Effect of nitric oxide on red blood cell. Pages 87-97 In: Nitric Oxide
       Research from Chemistry to Biology: EPR Spectroscopy of Nitrosylated
       Compounds, A.Y. Henry A. Guissani and B. Ducastel, (Eds.). R.G. Landes
       Company, Austin.

Hughes, E.D., C.K. Ingold, and J.H. Ridd. 1958. Nitrosation, diazotisation, and
       deamination. Part I. Principle, background, and method for the kinetic study of
       diazotisation. J.Chem. Soc. 58-65.

Jia, L., C. Bonaventura, J. Bonaventura, and J.S. Stamler. 1996. S-Nitrosohaemoglobin: a
       dynamic activity of blood involved in vascular control. Nature 380: 221-226.

Lundberg, J.O. and E. Weitzberg.  2005. NO Generation from nitrite  and its role in
       vascular control. Arterioscler. Thromb. Vase. Biol. 25: 915 - 922.

McMahon, T.J., A. J. Gow, and J. Stammler. 2000. The respiratory cycle: a three-gas
       system, pp.243-249 In: Nitric Oxide - Biology and Pathology, LJ. Ignarro (Ed.).
       Academic Press Inc., San Diego.

Okamoto, M., K. Tsuchiya, Y. Kanematsu, Y. Izawa, M. Yoshizumi, S. Kagawa, and T.
       Tamaki. 2005. Nitrite-derived nitric oxide formation following ischemia-
       reperfusion injury in kidney. Am. J. Physiol. Renal. Physiol.  288: F182-F187.

Spencer, N.Y., H. Zeng, R.P. Patel, and N. Hogg. The Journal of Biological Chemistry,
       H. Zeng, R.P. Patel and N. Hogg. 2000. Reaction of S-Nitrosoglutathione with the
       Heme Group of Deoxyhemoglobin. J. Biol. Chem.. 275: 36562-36567.

Webb, A., R. Bond, P. McLean, R.Uppal, N. Benjamin, and A. Ahluwalia. 2004.
       Reduction of nitrite to nitric oxide during ischemia protects against myocardial
       ischemia-reperfusion damage. Proc. Natl. Acad. Sci. USA. 101: 13683-13688.

Williams, D.L.H..  1985. S-Nitrosation and the reactions of S-nitroso compounds. Chem.
       Soc. Rev. 14: 171-196.

Williams, D.L.H.. 2003. A chemist's view of the nitric oxide story. Org. Biomol. Chem.
       1:441-449.

Zweier, J.L., P. Wang , A. Samouilov , P. Kuppusamy. 1995. Enzyme-independent
       formation of nitric oxide in biological tissues. Nat. Med. 1: 804-809.
                                       95

-------
96

-------
Nitrite regulates hypoxic responses and  protects against
                    ischemia/reperfusion injury

                                      by

                   S. Shiva1, C. Dezfulian1'2 and M.T. Gladwin1'2

Introduction

       The elucidation of the nitric oxide (NO) vasodilatory pathway initiated the field of
NO physiology which has since rapidly expanded.  Conventionally, NO is thought to be a
paracrine signaling molecule, with the enzyme nitric oxide synthase generating NO by
the oxidation of arginine to citrulline, using tetrahydrobiopterin  and oxygen as substrates.
In the vasculature, NO produced in the endothelium diffuses to underlying smooth
muscle cells where it binds the heme group of soluble guanylate cyclase to enable the
enzyme to produce cGMP and initiate a signaling cascade that ultimately results in
relaxation of the smooth muscle and vasodilation (Ignarro, Buga et al, 1987; Palmer,
Ferrige et al, 1987). Beyond this classical cGMP mediated vasodilatory pathway, NO is
now known to be an integral signaling molecule in a number of vascular responses
including platelet activation, thrombosis, and angiogenesis.

       Initially NO was thought to only mediate local paracrine signaling, and this idea
was supported by studies measuring the halflife of NO in blood  to be less than one
second (Liu, Miller et al, 1998). However, in the last decade, a number of studies have
suggested that NO may not only be a paracrine signaling molecule as first thought, but
may also be stabilized and transported in the circulation to mediate endocrine signaling at
a later time and farther away from the site of NO production (Fox-Robichaud, Payne et
al, 1998; Cannon, Schechter et al, 2001; Rassaf, Preik et al, 2002; Ng, Jourd'heuil et al,
2004). For example, in the human forearm, inhaled NO (80 ppm) has been shown to
increase blood flow in the human forearm when NOS is inhibited (Cannon, Schechter et
al, 2001) and NO solution infused in one arm has been shown to increase blood flow in
the other arm (Rassaf, Preik et al, 2002).  While several molecules, including iron-
nitrosyl hemoglobin (Gladwin, Ognibene et al, 2000a), S-nitrosoalbumin (Ng ,
Jourd'heuil et al, 2004),  and S-nitrosohemoglobin (Stamler, Jia et al, 1997) have been
proposed to be the endocrine transporters of NO, we have recently proposed  that nitrite
(MV) is the largest vascular storage form of NO (Cosby, Partovi et al, 2003). We have
found that nitrite functions as a store of NO that can be bioactivated during to mediate
hypoxic vasodilation, modulation of hypoxic mitochondrial respiration and
cytoprotection from ischemia/ reperfusion (I/R) injury.
1 Vascular Medicine Branch, National Heart Lung and Blood Institute
2 Critical Care Medicine Department, Clinical Center, National Institutes of Health,
Bethesda, Maryland, 20892


                                      97

-------
Nitrite as a vasodilator

       Nitrite, generated predominantly by the auto-oxidation of NO, has long been
considered a physiologically inert product of NO metabolism in mammalian systems.
Early in vitro experiments by Furchgott and colleagues demonstrated that nitrite may
mediate vasodilation when applied to isolated aortic rings at high (lOOuM-lmM)
concentrations (Furchgott and Bhadrakom, 1953).  However, since the concentration of
nitrite in blood was found to be less than one micromolar (Dejam, Hunter et al, 2005),
this was thought to preclude a role for nitrite as a physiological vasodilator (Lauer, Preik
et al, 2001). More recently, our lab has observed artery-to-vein gradients of nitrite in the
human circulation and has measured increases in the consumption of nitrite during
exercise (Gladwin, Shelhamer et al, 2000b; Cannon,  Schechter et al, 2001).  Furthermore,
in healthy volunteers, we observed that inhaled NO gas mediated endocrine vasodilation,
which was accompanied by a significant rise in plasma nitrite concentrations (Cannon,
Schechter et al,  2001).  Taken together, these data led to the hypothesis that nitrite may
be an intravascular store of NO that contributes to blood flow regulation in vivo.

       To test the vasoactivity of nitrite, near physiological levels of nitrite were infused
into the human brachial artery while forearm blood flow was measured. Infusion of nitrite
mediated vasodilation while subjects were at rest, in the presence of nitric oxide synthase
inhibition by L-NMMA, as well as during exercise. This nitrite-dependent vasodilation,
was accompanied by an increase in the formation of NO-modified hemoglobin (iron-
nitrosyl and S-nitrosothiol) (Cosby, Partovi et al, 2003).

       Several mechanisms of nitrite reduction to NO have been described, including
acidic disproportionation (Zweier, Wang et al, 1995) and enzymatic reduction by
xanthine oxidoreductase (Millar, Stevens et al, 1998). However, the correlation  of
vasodilation with the formation of modified hemoglobin led to the investigation  of
hemoglobin as a nitrite reductase, an idea that was supported by the previously
characterized reaction of deoxyhemoglobin with nitrite, (equation 1):

Nitrite + DeoxyHb + H+  -> NO + MetHb + OH                (1)

By this reaction, deoxygenated hemoglobin (deoxyHb) reduces nitrite to NO while being
oxidized to methemoglobin (metHb) (Doyle, Pickering et al, 1981).  The NO generated
from this reaction would then react with unreacted deoxyHb to form iron-nitrosyl
hemoglobin.

       The capability of this reaction to mediate vasodilation was tested in a series of
aortic ring experiments in which rat aortic rings were treated with nitrite (0.1-lOOOuM) in
the presence and absence of red blood cells (0.3% hematocrit) (Cosby, Partovi et al,
2003). While nitrite alone dilated vessels only at high concentrations, consistent with
Furchgott's early experiments; in the presence of red blood cells physiological
concentrations of nitrite mediated vasodilation, with  lOOnM nitrite mediating
vasodilation when the oxygen tension of the vessel bath was dropped to ISmmHg.
Indeed, in these experiments cyclic GMP was generated in the aortic rings, an effect that
                                        98

-------
was inhibited by the NO scavenger PTIO, confirming that the reduction of nitrite to
bioavailable NO was responsible for cGMP dependent vasodilation (Cosby, Partovi et al,
2003; Crawford, Isbell et al, 2006).

       To further understand the nitrite reductase activity of hemoglobin, biochemical
aspects of the deoxyHb-nitrite reductase reaction previously characterized by Doyle and
colleagues (Equation 1) was re-examined. Interestingly, kinetic analysis of this reaction
showed that hemoglobin dependent nitrite reductase activity was regulated by the
allosteric structural transition of hemoglobin from the deoxy (T) to oxy (R) state, with the
maximal reductase activity and NO generation coinciding with the p50 of hemoglobin
(Huang, Shiva et al, 2005b). This was found to be due to the differing characteristics of
the two allosteric conformations of hemoglobin.  Deoxy (T) state hemoglobin has free
heme groups, and hence many available heme groups to react with nitrite. But as nitrite
reacts with T state hemoglobin, the number of free hemes decreases, which decelerates
the nitrite-hemoglobin reaction.  However, this is countered by the reaction of nitrite with
oxy (R) state hemoglobin which has a lower reduction potential (is a better electron
donor) and hence accelerates the reaction.  These two opposing processes reach an
optimal balance at the p50 of hemoglobin. Consistent with this chemistry, the rate of the
reaction of nitrite with purified hemoglobin with increasing oxygen saturation is
sigmoidal, with maximum rate (and maximal NO production) occurring when
hemoglobin oxygen saturation is between 40 and 60% (Huang, Shiva et al, 2005b).

       In addition, as shown in Equation 1 above, hemoglobin-dependent nitrite
reduction requires a proton to proceed; hence we investigated whether this reaction was
regulated not only by hemoglobin allostery, but also by pH. Indeed, decreases in pH
increased hemoglobing-dependent nitrite reduction rate (Huang, Shiva et al, 2005b).
These data clearly demonstrate that hemoglobin is a nitrite reductase capable of
producing bioavailable NO in a manner that is regulated by the allosteric transition of the
hemoglobin molecule as well as pH.  This regulation of nitrite reduction to NO by
oxygen and pH suggested that this nitrite may be an  important source of NO in an acidic
environment where oxygen is limited, such as tissue ischemia.

Nitrite mediates cytoprotection during lischemia/Rreperfusion

       Recent studies utilizing several different models have now demonstrated that
nitrite is cytoprotective against injury induced by ischemia/reperfusion (il/R) (Webb,
Bond et al, 2004; Duranski, Greer et al, 2005). Utilizing a Langendorff perfusion model
of myocardial I/R, Webb and colleagues showed that low micromolar levels of nitrite
protect the rat heart from injury (Webb, Bond et al, 2004).  In vivo, Duranski et al,
demonstrated that nanomolar increases (200nM) in the concentration of plasma nitrite
reduced infarct size in a murine model of myocardial infarction at a dose as low as
1.2nmoles.  A dose of 48 nmoles increased plasma nitrite levels by 200nM and reduced
infarct size maximally (over 71% versus control). Nitrite has also been shown to have
similar effects in a murine model of hepatic I/R, in which levels of circulating liver
enzymes, ALT and AST, were reduced by 67% (as compared to the nitrate control) with
nitrite (48 nmoles) treatment (Duranski, Greer et al,  2005). Interestingly, in both murine
                                       99

-------
and Langendorff models of I/R, nitrite-dependent cytoprotection was inhibited by the NO
scavenger PTIO, suggesting that cytoprotection occurs secondary to nitrite reduction.

       The mechanism by which nitrite is reduced to NO during ischemia/reperfusion
remains unclear. Several potential nitrite reductases in tissue may be responsible.  Webb
and colleagues show that in heart homogenates, addition of nitrite produces NO, an effect
that is partially blocked in the presence of allopurinol, an inhibitor of xanthine oxidase
(Webb, Bond et al, 2004).  The mitochondrion has  been implicated in nitrite conversion
to NO (Kozlov, Staniek et al,  1999; Castello, David et al, 2006), and in the heart,
myoglobin may play this role  (Huang, Shiva et al, 2005b). It is also unknown whether
cytoprotection is dependent solely on NO production or whether nitrite may mediate
effects independent of NO formation.

       Previous studies have shown that nitrite is able to mediate post-translational
modification of proteins (Bryan, Rassaf et al, 2004; Duranski, Greer et al, 2005), metal
centers (Huang, Keszler et al,  2005a,; Huang, Shiva et al, 2005b) and lipids (O'Donnell,
Eiserich et al, 1999), as well as regulate gene expression (Bryan, Fernandez et al, 2005).
Several of these modifications, including nitrosylation of heme and S-nitrosation of
thiols, have been measured and associated with nitrite-dependent cytoprotection in the
murine models of liver and heart I/R (Duranski, Greer et al, 2005).   The specific
identities of these modified proteins are yet to be identified, although several potential
target proteins exist. S-nitrosation of key regulatory proteins involved in apoptosis has
been shown to be cytoprotective in  several models  of NO-mediated cytoprotection from
I/R injury.  One such example is the downstream apoptotic mediator caspase-3, which is
inhibited by S-nitrosation, subsequently inhibiting  apoptosis (Rossig, Fichtlscherer et al,
1999; Maejima, Adachi et al, 2005).

       Another possible mechanism of cytoprotection involves regulation of reactive
oxygen species (ROS) generation, particularly from the mitochondrion.  Small amounts
of ROS generation have been  determined to be a necessary component of mitochondrial
signaling in cytoprotection (Chandel, Maltepe  et al, 1998; Vanden Hoek, Becker et al,
1998).  However, the large burst of ROS generated after reperfusion from ischemia is
believed to be one of the mechanisms whereby cellular injury and necrosis occurs
(Saikumar, Dong et al, 1998; Di Lisa, 2001; Sorescu and Griendling, 2002). NO is a
known regulator of mitochondrial function, with low concentrations reversibly inhibiting
complex IV (Shiva, Oh et al, 2005) and higher concentrations mediating S-nitrosation of
complex I (Clementi, Brown et al, 1998; Burwell, Nadtochiy et al, 2006). Regulation of
electron transport by nitrite in this manner may decrease electron transport through the
chain during ischemia, preventing the formation of damaging ROS production by the
mitochondrion upon reperfusion.

       Although the specific molecular mechanism of nitrite dependent cytoprotection
remains unknown, it is clear that nitrite is an important hypoxic signaling molecule
which, in addition to mediating physiological effects, could potentially be used
therapeutically for the  prevention of I/R induced injury (Figure. 1). Current research
                                       100

-------
focuses on further characterizing the effects of nitrite in I/R and other pathological
models as well as determining the molecular mechanisms of its actions.
                                      NO
                                         Hemoglobin/
                                          Myoglobin
                             Xantnine
                          Oxidoreductase
              H"*7Ascorbate
                                        Heme and thiol
                                          containing
                                           enzymes
                 ",';»!i»i :.•&•••'
                  f
                           N-nitrosamines
                           S-nitros-othiols
   NO
               Iron-nitrosyl
               complexes
                                                      Nitrated Eipids
                     Modulation of
                    cytochrome p45O
                   heme oxygenase-1
                  heat shock protein-1
cGMP dependent
    hypoxic
  vasodilation
                 Ischemia-reperfusion
                    cytoprotection
                                  Nitrite Therapeutics
         Gastric mucosal
            protection
                                                                 Ischemia reperfusion:
                                                                    m heart
                                                                      solid organ
                                                                      transplantation
                    Gastric host defense
                    antibacterial effects
                                         Cerebral
                                        vasospasm
                                                        Pulmonary
                                                       hypertension
       Figure 1. Nitrite is a hypoxic signaling molecule. Nitrite is reduced to nitric
                oxide along a physiological oxygen and pH gradient by a number of
                mechanisms. The NO results in modification of iron centers, thiols,
                amines and lipids to mediate biological responses that may be utilized
                therapeutically. (Taken with permission from Nature Chemical
                Biology  (2006)).

References

Bryan, N.S., T. Rassaf, R.E. Maloney, C.M. Rodriguez, F. Saijo, J.R. Rodriguez, and M.
       Feelisch. 2004. Cellular targets and mechanisms of nitros(yl)ation: an insight into
       their nature and kinetics in vivo.  Proc. Natl. Acad. Sci. USA 101(12): 4308-
       4313.

Bryan, N. S., B. O. Fernandez, S.M. Bauer, M.F. Garcia-Saura, A.B. Milsom, T. Rassaf,
       R.E. Maloney, A.  Bharti, J. Rodriguez, and M. Feelisch. 2005. "Nitrite is a
       signaling molecule and regulator of gene expression in mammalian tissues." Nat.
       Chem. Biol. 1(5):  290-297.
                                         101

-------
Bryan, N. S., T. Rassaf, R.E. Maloney, C.M. Rodriguez, F. Saijo, J.R. Rodriguez, and M.
       Feelisch. 2004. "Cellular targets and mechanisms of nitros(yl)ation: an insight
       into their nature and kinetics in vivo." Proc Natl Acad Sci U S A 101(12): 4308-
       13.

Burwell, L. S., S. M. Nadtochiy, AJ. Tompkins, S. Young, andP.S. Brookes. 2006.
       "Direct evidence for S-nitrosation of mitochondrial complex I." Biochem. J.
       394(Pt 3): 627-634.

Cannon III, R. O., A. N. Schechter, J. A. Panza, P.P. Ognibene, M.E. Pease-Fye, M.A.
       Waclawiw, J.H. Shelhamer, andM.T. Gladwin. 2001. "Effects of inhaled nitric
       oxide on regional blood flow are consistent with intravascular nitric oxide
       delivery." J. Clin. Invest. 108(2): 279-287.

Castello, P. R., P. S. David, T. McClure, Z. Crook, and R.O. Poyton. 2006.
       "Mitochondrial cytochrome oxidase produces nitric oxide under hypoxic
       conditions: implications for oxygen sensing and hypoxic signaling in eukaryotes."
       Cell. Metab. 3(4): 277-287.

Chandel, N. S., E. Maltepe, E. Goldwasser, C.E. Mathieu, M.C. Simon, and P.T.
       Schumacker. 1998. "Mitochondrial reactive oxygen species trigger hypoxia-
       induced  transcription." Proc. Natl. Acad. Sci. USA 95(20): 11715-11720.

Clementi, E., G. C. Brown, M. Feelisch, and S. Moncada.  1998. "Persistent inhibition of
       cell respiration by nitric oxide: crucial role of S-nitrosylation of mitochondrial
       complex I and protective action of glutathione." Proc. Natl. Acad. Sci. USA
       95(13): 7631-7636.

Cosby, K., K. S. Partovi, J.H. Crawford, R.P. Patel, C.D. Reiter, S. Martyr, B.K. Yang,
       M.A. Waclawiw, G. Zalos, X. Xu, K.T. Huang, H.  Shields, D.B. Kim-Shapiro,
       A.N. Schechter, R.O. Cannon 3rd, and M.T. Gladwin. 2003. "Nitrite reduction to
       nitric oxide by deoxyhemoglobin vasodilates the human circulation." Nat. Med.
       9(12): 1498-1505.

Crawford, J. H., T. S. Isbell, Z. Huang, S. Shiva, B.K. Chacko, A.N. Schechter, V.M.
       Darley-Usmar, J.D. Kerby, J.D. Lang Jr, D. Kraus, C. Ho, M.T. Gladwin, and
       R.P. Patel. 2006. "Hypoxia, red blood cells, and nitrite regulate NO-dependent
       hypoxic vasodilation." Blood 107(2): 566-574.

Dejam, A., C. J. Hunter, M.M. Pelletier, L.L. Hsu, R.F. Machado, S. Shiva, G.G. Power,
       M. Kelm, M.T. Gladwin, and A.N. Schechter.  2005. "Erythrocytes are the major
       intravascular storage sites of nitrite in human blood." Blood 106(2): 734-739.

Di Lisa, F. 2001. "Mitochondrial contribution in the progression of cardiac ischemic
       injury." IUBMB Life 52(3-5): 255-261.
                                       102

-------
Doyle, M. P., R. A. Pickering, T.M. DeWeert, J.W. Hoekstra, and D. Pater. 1981.
       "Kinetics and mechanism of the oxidation of human deoxyhemoglobin by nitrites.
       " J. Biol. Chem. 256(23): 12393-12398.

Duranski,  M. R., J. J. Greer, A. Dejam, S. Jaganmohan, N. Hogg, W. Langston, R.P.
       Patel, S.F. Yet, X. Wang, C.G. Kevil, M.T. Gladwin, and DJ. Lefer. 2005. "
       Cytoprotective effects of nitrite during in vivo ischemia-reperfusion of the heart
       and liver." J. Clin. Invest. 115(5): 1232-1240.

Fox-Robichaud, A., D. Payne, S.U. Hasan, L. Ostrovsky, T. Fairhead, P. Reinhardt, and
       P. Kubes. 1998. "Inhaled NO as a viable antiadhesive therapy for
       ischemia/reperfusion injury of distal microvascular beds." J. Clin. Invest.  101(11):
       2497-2505.

Furchgott, R. F. and S. Bhadrakom 1953. "Reactions of strips of rabbit aorta to
       epinephrine, isopropylarterenol, sodium nitrite and other drugs." J. Pharmacol.
       Exp.  Ther. 108(2): 129-143.

Gladwin, M. T., F. P. Ognibene, L.K. Pannell, IS. Nichols, M.E. Pease-Fye, J.H.
       Shelhamer, and A.N. Schechter. 2000a. "Relative role of heme nitrosylation and
       beta-cysteine 93 nitrosation in the transport and metabolism of nitric oxide by
       hemoglobin in the human circulation." Proc. Natl. Acad. Sci. USA 97(18): 9943-
       9948.

Gladwin, M. T., J. H. Shelhamer, A.N. Schechter, M.E. Pease-Fye, M.A. Waclawiw,  J.A.
       Panza, F.P. Ognibene, and R.O. Cannon III. 2000b. "Role of circulating nitrite
       and S-nitrosohemoglobin in the regulation of regional blood flow in humans."
       Proc. Natl. Acad. Sci. USA 97(21): 11482-11487.

Huang, K. T., A. Keszler, N. Patel, R.P. Patel, M.T. Gladwin, D.B. Kim-Shapiro, andN.
       Hogg. 2005a. "The reaction between nitrite and deoxyhemoglobin. Reassessment
       of reaction kinetics and stoichiometry." J. Biol. Chem. 280(35): 31126-31131.

Huang, Z., S. Shiva, D.B. Kim-Shapiro, R.P. Patel, L.A. Ringwood, C.E. Irby, K.T.
       Huang, C. Ho, N. Hogg, A.N. Schechter, and M.T. Gladwin. 2005b. "Enzymatic
       function of hemoglobin as a nitrite reductase that produces NO under allosteric
       control. " J. Clin. Invest. 115(8): 2099-2107.

Ignarro, L. J., G. M. Buga, K.S. Wood, R.E. Byrns, and G. Chaudhuri.  1987.
       "Endothelium-derived relaxing factor produced and released from artery and vein
       is nitric oxide." Proc. Natl. Acad. Sci. USA 84(24):  9265-9269.

Kozlov, A. V., K. Staniek, and H. Nohl.  1999. "Nitrite reductase activity is a novel
       function of mammalian mitochondria." FEES Lett. 454(1-2): 127-130.
                                       103

-------
Lauer, T., M. Preik, T. Rassaf, B.E. Strauer, A. Deussen, M. Feelisch, and M. Kelm.
       2001. "Plasma nitrite rather than nitrate reflects regional endothelial nitric oxide
       synthase activity but lacks intrinsic vasodilator action." Proc. Natl. Acad. Sci. U S
       A 98(22): 12814-12819.

Liu, X., M. J. Miller, M.S. Joshi, H. Sadowska-Krowicka, D.A. Clark, and J.R. Lancaster
       Jr. 1998. "Diffusion-limited reaction of free nitric oxide with erythrocytes." J.
       Biol. Chem. 273(30): 18709-18713.

Maejima, Y., S. Adachi, K. Morikawa, H. Ito, and M. Isobe. 2005. "Nitric oxide inhibits
       myocardial apoptosis by preventing caspase-3 activity via S-nitrosylation." J.
       Mol. Cell. Cardiol. 38(1): 163-174.

Millar, T. M., C. R. Stevens, N. Benjamin, R. Eisenthal, R. Harrison, and D.R. Blake.
       1998. "Xanthine oxidoreductase catalyses the reduction of nitrates and nitrite to
       nitric oxide under hypoxic conditions." FEES Lett. 427(2): 225-228.

Ng, E. S., D. Jourd'heuil, J.M. McCord, D. Hernandez, M. Yasui, D. Knight, and P.
       Kubes. 2004. "Enhanced S-nitroso-albumin formation from inhaled NO  during
       ischemia/reperfusion." Circ. Res. 94(4): 559-565.

O'Donnell, V. B., J. P. Eiserich, P.H. Chumley, M.J. Jablonsky, N.R. Krishna, M. Kirk,
       S. Barnes, V.M. Darley-Usmar, amdB.A. Freeman. 1999. "Nitration of
       unsaturated fatty acids by nitric oxide-derived reactive nitrogen species
       peroxynitrite, nitrous acid, nitrogen dioxide, and nitronium ion." Chem. Res.
       Toxicol. 12(1): 83-92.

Palmer, R. M., A. G. Ferrige, and S. Moncada. 1987. "Nitric oxide release accounts for
       the biological activity of endothelium-derived relaxing factor." Nature 327(6122):
       524-526.

Rassaf, T., M. Preik, P. Kleinbongard, T. Lauer, C. Heiss, B.E. Strauer, M. Feelisch, and
       M. Kelm. 2002. "Evidence for in vivo transport of bioactive nitric oxide in human
       plasma." J. Clin. Invest. 109(9):  1241-8.

Rossig, L., B. Fichtlscherer, K. Breitschopf, J. Haendeler, A.M. Zeiher, A. Mulsch, and
       S. Dimmeler. 1999. "Nitric oxide inhibits caspase-3 by S-nitrosation in vivo." J.
       Biol. Chem. 274(11): 6823-6826.

Saikumar, P., Z. Dong, J.M. Weinberg, and M.A. Venkatachalam. 1998. "Mechanisms of
       cell death in hypoxia/reoxygenation injury." Oncogene 17(25): 3341-3349.

Shiva, S., J. Y. Oh, A.L. Landar, E. Ulasova, A. Venkatraman, S.M.  Bailey, and V.M.
       Darley-Usmar. 2005. "Nitroxia: the pathological consequence of dysfunction in
       the nitric oxide-cytochrome c oxidase signaling pathway." Free Radic. Biol. Med.
       38(3): 297-306.
                                       104

-------
Sorescu, D. andK. K. Griendling. 2002. "Reactive oxygen species, mitochondria, and
       NAD(P)H oxidases in the development and progression of heart failure." Congest.
       Heart Fail. 8(3): 132-140.

Stamler, J. S., L. Jia, J.P. Eu, TJ. McMahon, IT. Demchenko, J. Bonaventura, K.
       Gernert, and C.A. Piantadosi. 1997. "Blood flow regulation by S-
       nitrosohemoglobin in the physiological oxygen gradient." Science 276(5321):
       2034-2037.

Vanden Hoek, T. L., L. B. Becker, Z. Shao, C. Li, and P.T. Schumacker.  1998. "Reactive
       oxygen species released from mitochondria during brief hypoxia induce
       preconditioning in cardiomyocytes." J. Biol. Chem. 273(29):  18092-18098.

Webb, A., R. Bond, P. McLean, R. Uppal, N. Benjamin, and A. Ahluwalia. 2004.
       "Reduction of nitrite to nitric oxide during ischemia protects against myocardial
       ischemia-reperfusion damage." Proc. Natl. Acad. Sci. USA 101(37): 13683-
       13688.

Zweier, J. L., P. Wang, et al. A. Samouilov, and P. Kuppusamy. 1995. "Enzyme-
       independent formation of nitric oxide in biological tissues." Nat. Med. 1(8): 804-
       809.
                                      105

-------
106

-------
  Influence of nitrite on the mechanical performance of
                    fish and mammalian hearts

                                      by

             D. Pellegrino1'2, T. Angelone u, B. Tota 2'3, M.T. Gladwin4

       The free radical nitric oxide (NO) is an important signalling molecule involved in
many physiological and pathological processes in biological systems. The high
membrane permeability and short life make this molecule ideal as a short distance
intercellular and intracellular messenger. In particular, NO represents one of the most
important regulators of the cardiovascular system.

       It is well acknowledged that NO is generated in biological tissues by specific
nitric oxide synthases (NOSs) which metabolize arginine to citrulline with the formation
of NO. However, over the last years, a growing body of evidence suggests that the nitrite
anion (NXV) may represent the largest form of intravascular and tissue storage of NO
(Gladwin et a/., 2004). Nitrite anion is relatively abundant in blood and tissues: it is
found in human plasma in vivo, arteries (540±74 nM) and veins (466±79 nM) (Gladwin
et al., 2000); as well as in rat plasma in vivo (150 to 1.000 nM) and aorta (>10 microM)
(Rodriguez etal, 2003).

       Mechanisms for the in vivo conversion of nitrite to NO may involve either
enzymatic reduction or non-enzymatic reduction (Cosby et a/., 2003 and references
therein). Some proteins show nitrite reductase capacity, i.e. glutathione-S-transferases,
xanthine oxidoreductase,  deoxy-hemoglobin, cytochrome P-450 enzymes, and, recently,
also eNOS (Gautier et a/., 2006 and references therein). Each mechanism would occur
preferentially during pathological hypoxia and acidosis (Gladwin etal., 2005), which
occur in disease states, such as ischemia (Duranski et a/., 2005). Because the generation
of NO from L-arginine by NOS enzymes  depends on oxygen, this alternative method of
NO production represents an important protective mechanism in ischemic conditions
where oxygen is rapidly depleted (Duranski et al., 2005).

       Recently, some authors have reported that nitrite has a distinct and important
signalling role under normal physiological conditions. Nitrite is capable of modulating
many important signalling pathways, including soluble guanylyl cyclase (sGC)
stimulation, cytochrome P-450 activity and the expression of two archetypical proteins,
heat shock protein 70  (Hsp 70) and heme  oxygenase-1 (HO-1) (Bryan etal., 2005).
1 Departments of Pharmaco-Biology, University of Calabria, Italy
2 Departments of Cellular Biology, University of Calabria, Italy
3 Zoological Station "A: Dohrn", Naples, Italy
4 National Heart Lung and Blood Institute, National Institutes of Health, Bethesda, MD,
USA
                                      107

-------
       The aim of this research was to explore the biological activity of nitrite as a
putative signaling molecule under basal conditions in both mammalian (rat) and non-
mammalian (temperate and Antarctic teleost and frog) heart models. Using isolated and
perfused heart preparations offish (Anguilla anguilla and the hemoglobinless Antarctic
icefish Chionodraco hamatus\ frog (Rana esculenta) and rat (Langendorff technique),
we have assessed the effects of nitrite on myocardial contractility. The use of hearts
without coronary vessels (icefish and frog) or poorly vascularized hearts (eel), allowed
the evaluation of direct myocardial effects of nitrite, independent from effects on vascular
reactivity. The working heart preparations, standardized for these cold-blooded animals,
mimic the physiological conditions of the heart in vivo (Gattuso etal., 1999; Imbrogno et
a/., 2001). At the same time, these preparations permit the analysis of the mechanical
performance free from extrinsic neuronal and endocrine influences. The advantages of
the classic Langendorff rat heart preparation have been illustrated by Legssyer et al.
(1997). The nitrite dose-response curves (nanomolar- micromolar range) revealed a
remarkable negative inotropism of nitrite in eel, frog and rat hearts. In contrast, nitrite
induced significant positive inotropic effect in the icefish (Fig.  1). Notably, the nitrite-
dependent inotropic effect in these species correlates with the effect observed with NO
donors, i.e. negative inotropism in eel  and rat, positive inotropism in icefish (Fig. 2).

       To analyze if the nitrite effects involved enzymatic NO  generation, nitrite dose-
response curves in the presence of NOS inhibitors (L-NMMA and L-NIO) and sGC
inhibitor (ODQ) were tested in the hearts of icefish and rat. The two NOS inhibitors
completely blocked nitrite-dependent positive inotropism  in icefish, while in the rat heart
they did not modify nitrite-dependent negative inotropism (data not shown). In both rat
and icefish, ODQ pretreatment completely blocked nitrite-dependent inotropism. These
results indicate that in icefish the nitrite positive inotropic effect needs a functional NOS
system while in rat the nitrite inotropic effects were NOS-independent; in both species,
nitrite effect is cGMP-dependent.

       Our data support the hypothesis that the biological activity of nitrite, which is
already significant at nanomolar concentrations, may be of relevance in modulating the
"normal" function of the vertebrate heart. At the same time, the comparative analysis
emphasizes the importance of the species-specific cardiac biochemical conditions, which
underlies the myocardial action of nitrite.
                                        108

-------
   EEL
          -7
m -
-20 -
-30 -
9





1
*
[Nitrite], log M

1
*

               FROG
                   0
                                               -7
                                                              -5
                                         -10 -
                                       g -20-
                                         -30 -
                                                      [Nitrite], log M
                                        RAT
                                             -10   -9   -8   -7    -6   -5   -4
                 -6      -5
                 [Nitrite], log M
                                                     n
                                                      4.    J-
                                                              **   **
                              [Nitrite], log M
  Figure  1. Dose-response curves for nitrite on stroke volume (Vs) and left ventricular
          pressure (LVP) in isolated and perfused hearts of eel, frog,  icefish and rat.
          Percentage changes were  evaluated as means ±  S.E.M.  (n= 3-5; *p<0.05,
          **p<0.025).
       30
       20
       10

        °
       -10
       -20
       -30
       -40
eel
 frog
         **
                 * *
                                I
                       **
               L-arg
    SIN-1
L-NIO
ODQ      SBrcGMP
Figure 2. Effect of NO-cGMP in eel (Imbrogno et al., 2001), frog (Sys et al., 1997) and
         icefish (Pellegrino et al, 2004).
                                       109

-------
References
Bryan, N.S., B.O. Fernandez, S.M. Bauer, M.F. Garcia-Saura, A.B. Milsom, T. Rassaf,
      R.E. Maloney, A. Bharti, J. Rodriguez, and M. Feelisch. 2005. Nitrite is a
      signaling molecule and regulator of gene expression in mammalian tissues. Nat.
      Chem. Biol. 1:290-297.

Cosby, K., K.S. Partovi, J.H. Crawford, R.P. Patel, C.D. Reiter, S. Martyr, B.K. Yang,
      M.A. Waclawiw, G. Zalos, X. Xu, K.T. Huang, H. Shields, D.B. Kim-Shapiro,
      A.N. Schechter, R.O. Cannon III, and M.T. Gladwin. 2003. Nitrite reduction to
      nitric oxide by deoxyhemoglobin vasodilates the human circulation. Nat. Med.
      9(12): 1498-1505.

Duranski, M.R., J.J.M. Greer, A. Dejam, S. Jaganmohan, N. Hogg, W. Langston, R.P.
      Patel, S-F. Yet, X. Wang, C.G. Kevil, M.T. Gladwin, and D.J. Lefer. 2005.
      Cytoprotective effects of nitrite during in vivo ischemia-reperfusion of the heart
      and liver. J. Clin. Invest. 115(5): 1232-1240.

Gattuso, A., R. Mazza, D. Pellegrino, and B. Tota. 1999. Endocardial endothelium
      mediates luminal ACh-NO signaling in isolated frog heart. Am. J. Physiol.  276:
      H633-H641.

Gladwin, M.T., A.N. Schechter, D.B. Kim-Shapiro, R.P. Patel, N. Hogg, S. Shiva,  R.O.
      Cannon III, M. Kelm, D.A. Wink, M.G. Espey, E.H. Oldfield, R.M. Pluta, B.A.
      Freeman, J.R. Lancaster Jr, M. Feelisch, and J.O. Lundberg. 2005. The emerging
      biology of the nitrite anion. Nat. Chem. Biol.  1(6): 308-314.

Gladwin, M.T., J.H. Crawford, and R.P. Patel. 2004.  The biochemistry of nitric oxide,
      nitrite, and hemoglobin: role in blood flow regulation. Free Rad. Biol. Med.
      36(6): 707-717.

Gladwin, M.T., J.H. Shelhamer, A.N. Schechter, M.E. Pease-Fye, M.A. Waclawiw, J.A.
      Panza, F.P. Ognibene, and R.O. Cannon III. 2000. Role of circulating nitrite and
      S-nitrosohemoglobin in the regulation of regional blood flow in humans. Proc.
      Natl. Acad. Sci. USA 97(21): 11482-11487.

Gautier, C.,  E. van Faassen, I. Mikula, P. Martasek, and A. Slama-Schwok. 2006.
      Endothelial nitric oxide synthase reduces nitrite anions to NO under anoxia.
      Biochem. Biophys. Res. Commun. 341(3): 816-821.

Imbrogno, S., L. De luri, R. Mazza,  and B. Tota. 2001. Nitric oxide modulates cardiac
      performance in the heart of Anguilla anguilla. J. Exp. Biol.  204(10): 1719-1727.

Legssyer, A., L. Hove-Madsen, J. Hoerter, and R. Fischmeister. 1997. Sympathetic
      Modulation of the Effect of Nifedipine on Myocardial Contraction and Ca Current
      in the Rat. J. Mol.Cell Cardiol. 29: 579-591.
                                      110

-------
Pellegrino, D., C. A. Palmerini, and B. Tota. 2004. No hemoglobin but NO: the icefish
       (Chionodraco hamatus) heart as a paradigm. J. Exp. Biol. 207: 3855-3864.

Rodriguez, J., R.E. Maloney, T. Rassaf, N.S. Bryan, and M. Feelisch. 2003. Chemical
       nature of nitric oxide storage forms in rat vascular tissue.  Proc. Natl. Acad. Sci.
       USA 100(1): 336-341.

Sys, S.U., D. Pellegrino, R. Mazza, A. Gattuso, LJ. Andries, and L. Tota. 1997.
       Endocardial endothelium in the avascular heart of the frog: morphology and role
       of nitric oxide. J. Exp. Biol. 200(24): 3109-3118.
                                       Ill

-------
112

-------
        Potential NO-routes in the piscine  circulation

                                       by

                                   C. Agnisola1

       Nitric oxide (NO) is a well sutdied and ubiquitous regulatory biological
compound that acts as a specific mediator in various physiological and patho-
physiological processes in cardiovascular, nervous and immunological systems. Being a
highly diffusible, membrane-permeant and short living gas molecule, it is ideal for short
distance intercellular and intracellular signalling. In particular, NO represents one of the
most important regulators of vascular resistance, by acting mainly via soluble guanylate-
cyclase (Dattilo etal, 1997; Shah etal, 2000).

       Nitric oxide appeared as a signalling molecule before the radiation of metazoans.
Accumulating evidence reveals that NO is used as a signalling molecule in a wide variety
of invertebrate and vertebrate animals. Despite the fact that fish constitute a group of
animals with high diversity, and they have colonized a wide variety  of environments,
information on the role of NO in the regulation offish circulation is relatively scant.

       Nitric oxide is synthesized by nitric oxide synthase (NOS) isoforms, eNOS
(endothelial, calcium-dependent), nNOS (neuronal, calcium-dependent) and iNOS
(inducible,  calcium-independent) (Alderton et al, 2001), which utilize L-arginine,
oxygen, and NADPH as substrates, and require FAD, FMN, calmodulin, and
tetrahydrobiopterin as cofactors (Mayer, 1995). Plasma NO is mainly derived from
endothelial NOS (Dattilo et al, 1997; Bredt, 1999). Endothelial NO production is
stimulated by mechanical forces (shear stress and cyclic strain) and a variety of humoral
factors, including acetylcholine, vascular endothelial growth factor (VEGF), bradykinin,
estrogen, sphingosine 1-phosphate, H2O2, angiotensin II, and ATP (Arnal etal, 1999;
Boo etal, 2003; Cai etal, 2003).

       Vascular endothelium is continuously exposed to shear stress (a frictional force
exerted on the vessel surface per unit area by blood, as it flows at a constant flow rate in
the vessel). It is well known that, in mammals, shear stress controls vascular tone and
diameter, vessel wall remodeling, hemostasis, and inflammatory responses (Davies
1995). Laminar shear stress has antiatherogenic effects (Boo et al, 2003). NO produced
by eNOS is involved in this role of shear stress. NO-mediated responses to shear stress
include vessel relaxation, inhibition of apoptosis, and platelet and monocyte adhesion
triggered by a variety of pro-atherogenic factors (Boo et al,  2003). eNOS stimulation by
a step increase in shear stress occurs in two phases: an initial burst phase (lasting from
seconds up to 30 min), which is Ca2+/calmodulin dependent, followed by a Ca2+-
independent phase, in which NO production rate is maintained at a lower level, that
involves protein kinases and eNOS phosphorylation (Boo et al, 2003).  The exact
mechanism is still under debate, and several phosphorylation sites have been identified.
1 Department of Biological Sciences, University of Naples Federico II, Napoli, Italy


                                       113

-------
Chronic changes in shear stress may also modify eNOS expression level by both
transcriptional induction and stabilization of mRNA (Davis et al, 2001). It is possible
that the acute, robust NO production due to a step increase in shear stress may play a
critical role in vessel relaxation, whereas the chronic, low level of NO production due to
the steady laminar shear stress may play a critical role as an antiatherogenic and anti-
inflammatory molecule  in cardiovascular biology and pathobiology.

       Information on the role of shear stress in the regulation of blood flow in fish is
lacking. A putative role  of shear stress in trout coronary circulation has been proposed by
Mustafa and Agnisola (1998), where a flow-dependent NO release was involved in the
vasodilatory response to adenosine. This stretch-dependent effect has been recently
confirmed in the trout coronary circulation (Agnisola and Mustafa, unpublished), where a
significant inhibition of adenosine vasodilatory response by gadolinium (Gd3+), a known
inhibitor of stretch-activated ion channels (Hamill et al., 1996), was observed. Moreover,
using a constant-flow preparation of the isolated and perfused trout heart, it has been
possible to demonstrate  that NO release from the preparation is directly related to
perfusion flow, independent from the presence of adenosine (Fig. 1, left panel). At a
constant flow, NO release in the presence of adenosine is half that of the control (Fig. 1,
right panel). These results can be explained assuming that in trout coronaries there is a
shear-stress dependent nitric oxide release from the endothelium.
        80-.

        70-

        GO-

        50-
     re
     o>
     a>  30-|
      CM
     O  20-
        10-
-O-CONTROL

—•—Adenosine 10"6 M
60-
                               £  40-
                               o>
                               to
                               re
                               0)
                               a>
                               ,*-  20H
                               ' CM
                               O
                                                          i
          0.0  0.1  0.2  0.3  0.4  0.5  0.6
                 Coronary flow
                   (ml min"1)
                                    0.0
          0.5
1.0
1.5
                                       Resistance (TPa s m  )
Figure 1. NO production (measured as NO2" in the effluent from the preparation) in the
         coronary system of the isolated trout (Oncorhynchus mykiss) heart under
         constant flow perfusion conditions. Left: relationship between NO release and
         coronary flow in the absence and presence of adenosine, a known vasodilator.
         Right: effect of the reduction of coronary resistance induced by adenosine 10"6
                                        114

-------
         M on NO release under constant flow perfusion conditions; flow = 0.34 ± 0.04
                "1
         ml min". Data are means ± SE of 5 determinations.
       Some information is available on the endothelial NO release induced by
neurohumoral factors in fish. NO signalling has been reported to be involved in the
angiotensin II effect on the eel heart (Imbrogno et al, 2003). Oxytocin induced, NO
mediated vasodilation has been reported in rainbow trout (Haraldsen et al., 2002). NO
has also been involved in vasodilation induced by serotonin and acetylcholine on the
coronary vasculature of the trout heart (Mustafa et al.,  1997). A NO-dependent
mechanism for acetylcholine-induced vasodilation has been reported in crucian carp
          ^a/., 1995) and trout (Soderstrom et al., 1995).
       NO appears to also be involved in the role of red blood cells (RBC) as O2 sensors.
One main pathway through which RBCs contribute to the regulation of blood flow and
O2 delivery is via ATP release, depending on the oxygenation state of haemoglobin.
Erythrocytes release  ATP when haemoglobin is deoxygenated (Sprague etal, 2003).
ATP, acting via endothelial P2Y receptors, stimulates vasodilatation through the release of
nitric oxide (NO) (Takemura et al., 1998), prostaglandins, and endothelium-derived
hyperpolarizing factor (EDHF) (Wang et al., 2005). Preliminary experiments on the
isolated and perfused trout heart have shown that ATP exerts vasodilatory effects on
coronary resistance in the concentration range of 10"9-10"6 M; however, this effect
appears weak if compared with adenosine vasodilation, and disappears at higher
concentrations (Fig. 2).  Apparently, there is little aim in exploiting the putative
erythrocyte and NO-dependent role of plasma ATP in this fish model.
            0)
            o
            ra  20H
            w
            a>
            ra
              -20-
o

.E  -40H
0)
2-60H
re
            o
              -80J
                           -ATP
                           •Adenosine
                   10
                     -10
               10-9     10-8
 10-
mol I
10-6     10-5
                                             -1
     Figure 2.  Dose response curve of ATP and adenosine in the perfused
               coronary system of the isolated, non-working trout (Oncorhynchus
               mykiss) heart. Data are means ± SE of 5 determinations.
                                       115

-------
       Another possible mechanism for the role of RBCs as C>2 sensors involves a NOS-
independent pathway for NO release. Nitrite anion (NCV), which can be relatively
abundant in blood and tissues (up to |jM levels in mammals, Gladwin et al, 2000;
Rodriguez et al, 2003), has been proposed as the largest intravascular and tissue storage
form of NO. Mechanisms for the in vivo conversion of nitrite to NO may involve either
enzymatic,  non-enzymatic, or acidic reduction (Cosby et al., 2003 and references
therein). Several proteins show nitrite reductase capacity: glutathione-S-transferases,
xanthine oxidoreductase, deoxy-hemoglobin, cytochrome P-450 enzymes, and, recently,
also eNOS  (Gautier et al., 2006 and references therein). This mechanism would occur
preferentially during pathological hypoxia and acidosis, when NOS is inactive (Webb et
al., 2004; Duranski et al., 2005), but recently some authors reported that nitrite is a
signalling molecule also under physiological conditions (Bryan et al., 2005).

       This NOS-independent role of NO may also occur in fish. Jensen and Agnisola
(2005) have recently demonstrated in the coronary circulation of the isolated trout heart
perfused with erythrocyte suspensions, that nitrite is converted to NO in a process that is
not inhibited by the NOS inhibitor L-NA. This supports the possibility of deoxyHb-
mediated reduction of nitrite to NO, a mechanism that would be  significant under
hypoxic conditions to help maintain oxygen supply to tissues.

References

Alderton, W.K., C.E. Cooper, andR.G. Knowles. 2001. Nitric oxide synthases:  structure,
       function and inhibition.  Biochem. J. 357: 593-615.

Arnal, J.F., A.T. Dinh-Xuan, M. Pueyo, B. Darblade, and J. Rami. 1999. Endothelium-
       derived nitric oxide and vascular physiology and pathology. Cell Mol. Life Sci.
       55:  1078-1087.

Boo, Y.C. and H. Jo. 2003. Flow-dependent regulation of endothelial nitric oxide
       synthase: role of protein kinases. Am. J. Physiol. Cell Physiol. 285: C499-508.

Bredt, D.S. 1999. Endogenous nitric oxide synthesis: biological functions and
       pathophysiology. Free Radic. Res. 31: 577-596.

Bryan, N.S., B.O. Fernandez, S.M. Bauer, M.F. Garcia-Saura, A.B. Milsom, T. Rassaf,
       R.E. Maloney, A. Bharti, J. Rodriguez, and M. Feelisch. 2005. Nitrite is a
       signaling molecule and regulator of gene expression in mammalian tissues. Nat.
       Chem. Biol. 1:290-297.

Cai, H., Z. Li, M.E. Davis, W. Kanner, D.G. Harrison, and S.C.J. Dudley. 2003. Akt-
       dependent phosphorylation of serine 1179 and mitogen-activated protein kinase
       kinase/extracellular signal-regulated kinase 1/2 cooperatively mediate activation
       of the endothelial nitric-oxide synthase by hydrogen peroxide. Mol. Pharmacol.
       63:325-331.
                                       116

-------
Cosby, K., K.S. Partovi, J.H. Crawford, R.P. Patel, C.D. Reiter, S. Martyr, B.K. Yang,
       M.A. Waclawiw, G. Zalos, X. Xu, K.T. Huang, H. Shields, D.B. Kim-Shapiro,
       and A.N. Schechter. 2003. Nitrite reduction to nitric oxide by deoxyhemoglobin
       vasodilates the human circulation. Nat. Med. 9: 1498-1505.

Dattilo, J.B. and R.G. Makhoul. 1997. The role of nitric oxide in vascular biology and
       pathobiology. Ann. Vase. Surg. 11: 307-314.

Davies, P.P.  1995. Flow-mediated endothelial mechanotransduction. Physiol. Reviews
       75: 519-555.

Davis, M.E., H. Cai, G.R.  Drummond, and D.G. Harrison. 2001. Shear stress regulates
       endothelial nitric oxide synthase expression through c-Src by divergent signalling
       pathways. Circ. Res. 89: 1073-1080.

Duranski, M.R., JJ. Greer, A. Dejam, S. Jaganmohan, N. Hogg, W. Langston, R.P. Patel,
       S.F. Yet, X. Wang, C.G Kevil, M.T. Gladwin, and D.J. Lefer. 2005.
       Cytoprotective effects of nitrite during in vivo ischemia-reperfusion of the heart
       and liver. Clin. Invest. 115: 1232-1240.

Gautier, C., E. van Faassen, I. Mikula, P. Martasek, andA. Slama-Schwok. 2006.
       Endothelial nitric oxide synthase reduces nitrite anions to NO under anoxia.
       Biochem. Biophys. Res. Commun. 341, 816-821.

Gladwin, M.T.,  J.H. Shelhamer, A.N. Schechter, M.E.Pease-Fye, M.A. Waclawiw, J.A.
       Panza, F.P. Ognibene, and R.O. Cannon. 2000. Role of circulating nitrite and S-
       nitrosohemoglobin in the regulation of regional blood flow in humans. Proc. Natl.
       Acad. Sci. U S A 97: 11482-11487.

Hamill, O.P. and D.W. McBride Jr. 1996. The pharmacology of mechanogated
       membrane ion channels. Pharmacol. Rev. 48:231-252.

Haraldsen, L., V. Soderstrom-Lauritzsen,  and G.E. Nilsson. 2002. Oxytocin stimulates
       cerebral  blood flow in rainbow trout (Oncorhynchus mykiss) through a nitric
       oxide dependent mechanism. Brain Res. 929:  10-14.

Hylland, P. and G.E. Nilsson. 1995. Evidence that acetylcholine mediates increased
       cerebral  blood flow velocity in crucian carp through a nitric oxide-dependent
       mechanism. J. Cerebr. Blood Flow Metab. 15: 519-524.

Imbrogno, S., M.C. Cerra, and B.  Tota. 2003. Angiotensin II-induced inotropism  requires
       an endocardial endothelium-nitric oxide mechanism in the in-vitro heart of
       Anguilla anguilla.  J. Exp. Biol. 206: 2675-2684.
                                       117

-------
Jensen, F.B and C. Agnisola. 2005. Perfusion of the isolated trout heart coronary
       circulation with red blood cells effects of oxygen supply and nitrite on coronary
       flow and myocardial oxygen consumption. J. Exp. Biol. 208: 3665-3674.

Mayer, B. 1995. Biochemistry and molecular pharmacology of nitric oxide synthase In:
       Nitric oxide in the nervous system. S. Vincent (Ed.). Academic Press, London.

Mustafa, T., C. Agnisola, and J.K. Hansen. 1997. Evidence for NO-dependent
       vasodilation in the trout (Oncorhynchus mykiss) coronary system. J Comp
       Physiol B-Biochem. Syst. Environ. Physiol. 167: 98-104.

Mustafa, T. and C. Agnisola. 1998. Vasoactivity of adenosine in the trout (Oncorhynchus
       mykiss) coronary system: Involvement of nitric oxide and interaction with
       noradrenaline. J. Exp. Biol. 201: 3075-3083.

Rodriguez, J., R. Maloney, T. Rassaf, N. Bryan, and M. Feelisch. 2003. Chemical nature
       of nitric oxide storage forms in rat vascular tissue. Proc. Natl. Acad. Sci. USA
       100:336-341.

Shah, A.M. and P. A. MacCarthy. 2000. Paracrine and autocrine effects of nitric oxide on
       myocardial function. Pharmacol. Ther. 86: 49-86.

Soderstrom, V., P. Hylland, and G.E. Nilsson. 1995. Nitric oxide synthase inhibitor
       blocks acetylcholine induced increase in brain blood flow in rainbow trout.
       Neurosci. Lett. 197:  191-194.

Sprague, R.S., JJ. Olearczyk, D.M.  Spence, A.H. Stephenson, R.W. Sprung, and AJ.
       Lonigro. 2003. Extracellular ATP signaling in the rabbit lung: erythrocytes as
       determinants of vascular resistance. Am. J. Physiol. Heart Circ. Physiol. 285:
       H693-700.

Takemura, S., Y. Minamiyama, N. Kawada, M. Inoue, S. Kubo, K. Hirohashi, and H.
       Kinoshita. 1998. Extracellular nucleotides modulate the portal circulation with
       generation of nitric oxide. Hepatol. Res. 13: 29-36.

Wang, L., G.  Olivecrona, M. Gotberg, M.L. Olsson, M.S. Winzell, and D. Erlinge. 2005.
       ADP acting on P2Y13 receptors is a negative feedback pathway for ATP release
       from human red blood cells. Circ.  Res. 96: 189-196.

Webb, A., R.  Bond, P. McLean, R. Uppal, N. Benjamin, and A.  Ahluwalia. 2004.
       Reduction of nitrite to nitric oxide during ischemia protects against myocardial
       ischemia-reperfusion damage. Proc. Natl. Acad. Sci. USA 101: 13683-13688.
                                       118

-------
                  Physiological effects of nitrite:
  Balancing the knife's edge between toxic disruption of
           functions and potential beneficial effects

                                       by

                                  F.B.Jensen1
       In relation to vertebrate biology, nitrite has mainly been known for its toxic
effects. In human health, main concerns has been with the potential formation of possible
carcinogenic nitrosamines (an issue still not settled) and methemoglobin (metHb)
formation following excess intake of nitrite or nitrate (that can be reduced to nitrite in the
digestive system) in the diet. Aquatic animals, like fish, are more prone to experience
high nitrite concentrations than terrestrial animals. Freshwater fish actively take up nitrite
across the gills, resulting in well-documented toxicity (Lewis and Morris, 1986). Several
physiological disturbances have been discovered in aquatic animals that add to the
standard methemoglobinemia caused by nitrite, and the picture is emerging that toxicity
results from the impact of multiple nitrite-induced physiological disturbances (Jensen,
2003).

       In recent years the conception that nitrite is exclusively a toxicant has been
refined, and important biological functions of nitrite are beginning to emerge (Gladwin et
a/., 2005). Nitrite is naturally present at low concentrations (0.15-0.6 jiM in plasma of
mammals), which is due to the endogenous production of nitrite as an oxidative
metabolite of the important messenger molecule nitric oxide (Kleinbongard et a/., 2003).
Additional sources  are intake of nitrite via the diet, and reduction of nitrate to nitrite by
bacteria in the oral  cavity (Lundberg and Weitzberg, 2005). At the low natural
concentration, nitrite is far from biologically inert. Nitric oxide can be re-generated from
nitrite, whereby nitrite may function as a vascular storage pool of NO activity that
participates in blood flow regulation and other functions (Cosby etal., 2003). Nitrite may
also act as a signaling molecule on its own and regulate gene expression (Bryan et a/.,
2005). Thus, nitrite appears to have vital functions that have been carried through in
evolution. There seems to be a schism  between beneficial effects of nitrite at low
concentrations and  harmful effects at high concentrations.

       The physiological effects of nitrite in aquatic animals was recently
comprehensively reviewed (Jensen, 2003). This presentation gives an overview of
physiological disturbances in nitrite-exposed fish and additionally describes recent
developments that suggest nitrite to be an important compound in animal biology at its
natural low endogenous concentration.
1 Institute of Biology, University of Southern Denmark, Odense, DK-5230 Odense M,
Denmark. E-mail: fbj@biology.sdu.dk
                                      119

-------
Nitrite in the environment

       Nitrite is part of the nitrogen cycle in ecosystems, where it is an intermediate in
bacterial nitrification and denitrification processes. The concentration of nitrite in
unpolluted waters is normally low (well below 1 |iM), but an imbalance in either
nitrification or denitrification can lead to a build up of nitrite. Elevated nitrite levels
pertain to habitats receiving nitrogenous waste and various hypoxic habitats (Eddy and
Williams, 1987), including the oxygen minimum zone of the oceans (Anderson etal.,
1982). Apart from its environmental relevance, nitrite is a matter of concern in
aquaculture industry, where episodic nitrite poisoning can cause extensive fish death
(Svobodova et a/., 2005). Aquaculture facilities have a high density offish and a large
production of waste products (including ammonia excreted by the fish). Discharge of
ammonia and establishment of nitrification is likely to cause a transient increase in nitrite
because Nitrosomonas (that oxidize ammonia to nitrite) tend to grow faster than
Nitrobacter (that oxidize nitrite to nitrate), and because Nitrobacter is inhibited by
elevated free NH3 levels (Balmelle et a/., 1992).  Nitrite concentrations of 1 mM or more
can be reached in the water in such circumstances (Jensen, 2003).

Nitrite uptake in fish

       When exposed to raised environmental [NCV], freshwater fish accumulate nitrite
in blood plasma to much higher concentrations than in the ambient water (Bath and Eddy,
1980; Margiocco et a/., 1983; Jensen etal., 1987). To understand why, a brief look at the
ion uptake mechanism in the freshwater fish gill is useful (Fig. 1). Freshwater fish need
an active uptake of ions to compensate for passive ion losses. The current view is that a
proton pump in the apical membrane creates the driving force for Na+ uptake through
sodium channels (Perry, 1997; Randall  and Brauner, 1998; Marshall, 2002; Evans etal.,
2005). The protons come from hydration of CC>2 (catalyzed by epithelial carbonic
anhydrase),  and the bicarbonate formed subsequently serves as counter ion for Cl" uptake
via an anion exchange mechanism. Nitrite has an affinity for this uptake mechanism, so
whenever nitrite is present in the water, part of the Cl" uptake will be shifted to nitrite
uptake (Fig. 1), and internal concentrations in the millimolar range can eventually
develop.
                                        120

-------
Freshwater
         Na
                                Gill epithelium
Blood
                                     Carbonic
                                     anhydrase
   Figure 1. Current view on mechanisms of ion uptake in the freshwater fish gill,
            illustrating the ability of nitrite to compete with chloride for the Cl"
            uptake mechanism. An apical membrane H+-ATPase creates the driving
            force for Na+ uptake via a sodium channel. Na+ subsequently exits the
            basolateral membrane via the Na+-K+-ATPase. Hydration of CC>2,
            catalyzed by epithelial carbonic anhydrase, supplies H+ for the proton
            pump and delivers HCCV for Cl" uptake via an apical membrane Cl"
            /HCCV exchanger. The proton  pump energizes Cl" uptake by raising
            cytosol [HCCV] and by acidifying the boundary layer, lowering external
            [HCCV]. Cl" (or nitrite) presumably exits the basolateral membrane via a
            chloride channel. Further details on gill ion transport and involved
            epithelial cells can be found in  Perry (1997), Marshall (2002) and Evans
            etal. (2005).

       Several observations are in agreement with nitrite being transported by the
branchial Cl" uptake mechanism. Nitrite influx and Cl" influx show saturation kinetics
(supporting carrier-mediated uptake), and NCV is a competitive inhibitor of Cl" uptake
and vice versa (Williams and Eddy, 1986; Tomasso and Grosell, 2005). Elevation of
ambient [Cl"] is known to effectively protect against nitrite uptake (Bath and Eddy, 1980)
and nitrite toxicity (Perrone and Meade, 1977; Russo etal., 1981). The branchial Cl"
uptake rate varies between species, and species with low uptake rates (e.g. eel, tench  and
carp) are less sensitive to nitrite than species with high uptake rates (rainbow trout, perch,
pike) (Williams and Eddy,  1986). Most freshwater fishes accumulate nitrite in blood
plasma when exposed to this anion, but a few species (e.g. bluegill, largemouth bass and
striped bass) do not concentrate nitrite in plasma (Tomasso, 1986; Palacheck and
Tomasso, 1984; Mazik et a/., 1991). This puzzling observation appears to be explained
by exceptionally  low (barely detectable) branchial Cl" influx rates, as reported in eel
(Hyde  and Perry, 1989) and bluegill (Tomasso and Grosell, 2005).

       In addition to the variation in nitrite uptake between species, some species may
show intraspecific variation, as reported in rainbow trout (Margiocco et a/., 1983;
Williams and Eddy, 1988; Stormer et a/.,  1996; Aggergaard and Jensen, 2001). Two
studies that exposed rainbow trout to 1 mM ambient nitrite in hard freshwater of the same
                                       121

-------
composition, suggest that the splitting of rainbow trout into groups with different
sensitivity to nitrite is highly reproducible (Stormer et a/., 1996; Aggergaard and Jensen,
2001). In both studies the rainbow trout could be divided into two distinct groups.
Sensitive trout showed fast accumulation and died between 1 and 2 days of exposure,
whereas less sensitive trout accumulated nitrite slower and survived more than 4 days.
The nitrite accumulation pattern and the nitrite levels reached in the two groups were
very similar in both studies, as was the physiological disturbances (Stormer et a/., 1996;
Aggergaard and Jensen, 2001). The separation into two groups raised the idea that this
might be gender-related. To test this hypothesis, individual rainbow trout were exposed to
1 mM nitrite for 18 hours (allowing the splitting up into two groups without causing
mortality) and then returned to nitrite-free water (to study recovery  from nitrite
exposure). The fish indeed could be divided into two groups based on their degree of
nitrite accumulation; but males and females were present in both groups (Fig. 2).
                          Nitrite exposure
Recovery in nitrite-free water
             O
                                                       Sensitive fish
                                                       More tolerant fish
                                    20         40
                                         Time (h)

       Figure 2. Time-dependent changes in plasma nitrite concentration in six male and
                five female rainbow trout during 18 h exposure to 1 mM nitrite and
                subsequent recovery in nitrite-free water. Successive blood samples
                were drawn through indwelling dorsal aortic catheters, and results are
                depicted for each individual fish. Sensitive fish reached plasma [NCV]
                values well above 2 mM during nitrite exposure, whereas more tolerant
                fish attained significantly lower values. Males and females were
                present in both groups. When the fish were returned to nitrite-free
                water the nitrite build-up was reversed (F. Zachariasen and F.B. Jensen,
                unpublished data).

Thus, the intraspecific variation was not gender related. The most sensitive fish were
shown to have a significantly higher branchial nitrite influx than the less sensitive fish (F.
Zachariasen and F.B. Jensen, unpublished). Chloride cells are the likely site of Cl" uptake
(Perry, 1997) and are additionally known to proliferate upon nitrite exposure (Williams
and Eddy, 1988). Thus, a relevant hypothesis for future testing is that chloride cell
                                        122

-------
number and/or the degree of chloride cell proliferation differ between the two groups of
rainbow trout.

       Seawater fish are hypo-osmotic regulators and secrete NaCl across the gills. Thus,
the problem of nitrite uptake via the active Cl" uptake route that prevails in freshwater
fish is eliminated. However, nitrite can still be taken up passively across the gills.
Further, marine fish drink seawater taking up ions and water across their intestine. The
European flounder has been shown to take up nitrite across the gut epithelium if nitrite is
present in the lumen. The transport is inhibited by bumetanide, suggesting that part of the
uptake is mediated via an affinity of nitrite for the Na+/K+/2Cl" cotransporter that is
involved in intestinal ion uptake (Grosell and Jensen, 1999). When exposed to
waterborne nitrite it appears that some 66% of whole-body NO2" uptake is via the
intestinal route (Grosell and Jensen, 2000). Nitrite uptake is, however, lower in seawater
than in freshwater fish. In the European flounder plasma [NCV] stays below the ambient
concentration (Grosell and Jensen, 2000).

       The transport of nitrite from the extracellular space into intracellular
compartments has been studied in some detail for red blood cells (Jensen, 2003), whereas
little is know about uptake in other cell types. Nitrite is, however, known to concentrate
in liver, brain, gill and muscle tissue offish (Margiocco etal., 1983).

Nitrite-induced physiological disturbances

       When active branchial Cl" uptake partly shifts to nitrite uptake and passive Cl"
efflux persists, extracellular [Cl"] can be predicted to decrease, as is also observed in carp
and rainbow trout (Jensen etal., 1987; Stormer etal., 1996).  Proliferation of chloride
cells (e.g. Williams and Eddy, 1988) may be considered a response aimed at restoring
chloride balance, but in this particular circumstance it appears maladaptive, because an
increased capacity for Cl" uptake also increases nitrite uptake. While nitrite is
accumulated in plasma, plasma nitrate increases in parallel, because some of the nitrite is
endogenously converted (detoxified) to nitrate (Stormer et a/., 1996; Doblander and
Lackner, 1997). Plasma lactate also increases, as tissue 62 delivery becomes
compromised by high methemoglobin levels (Jensen etal., 1987). The sum of
extracellular anion, however, stays constant during nitrite exposure (Jensen etal., 1987;
Stormer etal, 1996).

       Among the cations, the most prominent effect concerns potassium. Extracellular
[K+] increases (Jensen et a/., 1987) whilst the intracellular K+ content decreases in red
blood cells (Jensen,  1990) and skeletal muscle (Stormer et a/., 1996; Knudsen and
Jensen, 1997). The K+ efflux from red blood cells results from activation of K+/C1"
cotransport and leads to erythrocyte shrinkage in carp (Jensen, 1990, 1992). The K+ loss
from skeletal muscle is substantial and could increase extracellular [K+] to much higher
levels than observed, suggesting further transport of K+ to the ambient water with the
development of an overall potassium deficit (Knudsen and Jensen, 1997). Interference
with K+ balance has been observed in a number offish species and also in invertebrates
(freshwater crayfish),  suggesting that it is a general effect  of nitrite (Jensen, 2003).
                                        123

-------
Elevated extracellular [K+] is unfavorable for heart and nerve function by causing
depolarization, and the reduced intracellular K+ content may critically influence muscular
metabolism.

       Nitrite reacts with both oxygenated hemoglobin (oxyHb) and deoxygenated
hemoglobin (deoxyHb) to form metHb, but the reaction mechanisms, reaction kinetics
and reaction products differ. The stoichiometry for the reaction between oxyHb and
nitrite is (Kosaka and Tyuma, 1987):

4Hb(Fe2+)O2 + 4NCV + 4H+  -> 4Hb(Fe3+) + 4NO3" + O2 + 2H2O       (1)

Thus, in parallel with the oxidation of heme iron (Fe2+ —»• Fe3+) to form metHb, nitrite is
oxidized to nitrate (see Jensen, 2003 for further details). In the reaction with deoxyHb,
nitrite is reduced to nitric oxide, whereby deoxyHb functions as a nitrite reductase (Cosby
etal, 2003):

Hb(Fe2+) + NCV + H+  ->  Hb(Fe3+) + NO + OH'                      (2)

       The formation of metHb, which is non-functional  in regards to O2 transport, can
be countered by metHb reductase activity inside the erythrocytes. At low nitrite
concentrations, metHb levels need not  increase significantly, but at elevated nitrite
concentrations metHb levels increase. At a given nitrite load, a balance is established
between nitrite-induced oxidation of hemoglobin and reduction to functional Hb via
metHb reductase systems (Jensen, 1990, 1992). The continuous increase in nitrite load in
nitrite-exposed fish, however, forces metHb levels upwards,  eventually reaching levels of
70-85 % of the total Hb (Eddy and Williams, 1987; Jensen etal, 1987; Aggergaard and
Jensen, 2001). Such high metHb levels drastically decreases the arterial O2 content and
result in a severe tissue O2 shortage that becomes reflected in elevated plasma lactate
levels (Jensen etal, 1987). Methemoglobinemia is the prime reason for the disruption of
O2 transport, but additional contributions may come from decreases in blood O2 affinity
(related to erythrocyte shrinkage in carp), total Hb content (possibly via increased
removal of damaged erythrocytes), and interaction of nitrite with cellular heme proteins
such as myoglobin and cytochromes (cf. Jensen, 2003). Fish can tolerate relatively high
blood metHb  levels at rest, but their swimming performance is impaired (Brauner et al.,
1993) as is their ability to handle environmental hypoxia.  An increased ventilatory
activity (Aggergaard and Jensen, 2001) can be seen as an  attempt to ameliorate O2
conditions, but during continued nitrite exposure the decline in blood O2 capacitance
becomes critical and ultimately lethal. However, if the fish are returned to nitrite-free
water the metHb build-up is reversed (Jensen, 2003) in parallel with the elimination of
nitrite from the blood (Fig. 2).

       Nitrite exposure is associated with a rapid and lasting increase in heart rate in
rainbow trout (Aggergaard and Jensen, 2001). It is interesting that this increase in heart
rate occurs very early during the exposure, before any significant increases in metHb and
plasma [K+] have developed. The effect on heart rate appears related to the mere
appearance of nitrite in the blood. It was suggested that nitric oxide was generated from
                                       124

-------
nitrite, and that the vasodilation caused by this was countered by an increased cardiac
pumping to re-establish blood pressure (Aggergaard and Jensen, 2001). Indeed, NO can
be generated from nitrite in various ways, including non-enzymatic acidic reduction and
enzymatic reduction by xanthine oxidoreductase or by deoxygenated Hb (Gladwin et a/.,
2005). Such nitric oxide generation from nitrite may play a physiological role in nitric
oxide homeostasis at the low nitrite levels seen under normal (non-toxic) circumstances
(cf. below). During nitrite exposure in fish, however, where nitrite concentrations can rise
to the millimolar level, an excess production of NO may ensue.  As NO is an important
signal molecule that is involved in processes ranging from blood flow regulation to
neurotransmission, such abnormal NO levels may have a number of critical effects.
The putative physiological role of nitrite in nitric oxide homeostasis

       Nitric oxide produced in vascular endothelial cells exerts its function by diffusing
into underlying vascular smooth muscle, causing its relaxation, which results in local
vasodilation and increased blood flow. Nitric oxide is a free radical with a short life time
in blood. NO can be inactivated inside erythrocytes (e.g. by reacting with oxyHb to form
metHb and nitrate), and in plasma its reaction with O2 forms nitrite. Nitrite and nitrate
were long considered relatively inert metabolites of NO. However, the possibility that
NO can be re-generated from nitrite has recently attracted considerable attention. The
formation of NO from nitrite by enzymatic and non-enzymatic means is favored by low
pH and/or low Po2 and is believed to be of physiological importance in humans,
particular in hypoxia and during ischemia (Gladwin et a/., 2005).

       The idea has emerged that nitrite at its natural low concentration in mammalian
blood functions as a vascular storage pool of nitric oxide that can be activated in a
physiological appropriate way (Cosby etal., 2003). The reaction of nitrite with deoxyHb
(equation 2 above) has attracted particular interest, because this  leads to an NO
production that is linked to the degree of Hb deoxygenation, which may supply a
mechanism for matching blood flow to O2 conditions (Cosby et  al., 2003; Nagababu et
al., 2003). Thus, increased deoxygenation of Hb in the microcirculation of hypoxic
tissues can be predicted to elevate erythrocyte NO formation from nitrite. The formed NO
can react with unoxidized heme groups to form iron-nitrosyl-hemoglobin, Hb(Fe2+)-NO,
but escape of some of the NO from the erythrocytes should suffice to cause vasodilation
and elevate blood flow and tissue oxygenation. The mechanism  is supported by nitrite
infusion-caused vasodilation in the human forearm at near-physiological nitrite
concentrations and by hypoxic vasodilation experiments using aortic rings (Cosby et al.,
2003; Crawford etal., 2006). The function of Hb as a nitrite reductase  is modulated by
both heme deoxygenation and heme redox potential and shows maximal activity around
50 % Hb oxygen saturation (Huang et al., 2005; Crawford et al., 2006).

       Nitrite entry into red blood cells is an essential first step  of the proposed
mechanism. In carp, nitrite transport across the red blood cell membrane is strongly
oxygenation-dependent. At physiological pH, nitrite extensively enters deoxygenated red
                                        125

-------
blood cells, whereas it hardly permeates oxygenated cells (Jensen, 1990, 1992). When pH
is lowered, nitrite starts to enter oxygenated red blood cells, but at much lower rates than
in deoxygenated cells (Jensen, 1992). A similar oxygenation dependency pertains to
tench and whitefish erythrocytes (Jensen, 2003). Thus, in these fish, nitrite preferentially
permeates erythrocytes with low oxygen saturation, thus apparently supplying nitrite for
deoxyHb-mediated NO generation in an appropriate manner. Further, in carp, as nitrite
enters erythrocytes with low C>2 saturation, a subsequent full oxygenation stops nitrite
influx (Knudsen and Jensen, 1997), showing that oxygenation functions as a switch that
obliterates further nitrite entry.

       There is limited information on the oxygenation dependency of nitrite transport in
mammalian erythrocytes. When nitrite is added to a suspension of pig erythrocytes at
physiological pH, nitrite quickly permeates and equilibrates across the membrane, and
then continues to enter the cells as result of intracellular nitrite removal (via its reactions
with hemoglobin), but the membrane permeation shows little oxygenation dependency
(Jensen, 2005). There is an extensive entry of nitrite into both oxygenated and
deoxygenated pig erythrocytes, which contrasts sharply with the oxygenation-
dependency of nitrite fluxes in carp, tench and whitefish. This difference is not just some
simple difference between non-nucleated mammalian erythrocytes and nucleated
erythrocytes of lower vertebrates,  because in rainbow trout (like in pig) there is no
significant oxygenation dependency of erythrocyte nitrite entry (Jensen and Agnisola,
2005). It appears that in  carp, tench and whitefish, erythrocyte nitrite transport is
governed by a major oxygenation-dependent change in the membrane permeability to
nitrite (high P at low O2 saturation and low P at high O2 saturation). In pig and rainbow
trout nitrite permeation is similar in oxygenated and deoxygenated erythrocytes, and
nitrite quickly equilibrates across the membrane,  after which further entry is governed by
the chemical reactions that remove nitrite inside the cells (Jensen, 2005).
                                        126

-------
Figure 3. Nitrite entering erythrocytes can react with deoxygenated heme groups to form
         nitric oxide that may escape from the cells and partake in local blood flow
         regulation. The reaction of nitrite with oxygenated heme groups can be
         considered a nitrite detoxification mechanism, because it converts nitrite to
         non-toxic nitrate, while the oxidized heme groups (metHb) can be reduced to
         functional Hb via metHb reductase. See text for further details.

       A preferential entry of nitrite into erythrocytes at low O2 saturation would seem
ideal for its subsequent reaction with deoxyHb to form NO that via its release from the
RBCs could participate in blood flow regulation (Fig. 3, right). The reaction with oxyHb
is, however, also important, as it may form a defense against inappropriate high levels of
nitrite (Fig. 3, left). Thus, nitrite reacting with oxyHb is detoxified to non-toxic nitrate
(Doblander and Lackner, 1997; Jensen, 2003). Of course, the concomitant formation of
metFIb is a toxic effect of nitrite, but this can be countered by metFIb reductase activity
that regenerates functional Fib. In this way each Hb molecule can participate in several
oxidation-reduction cycles that eliminate excess nitrite, and blood  metFIb levels need not
increase significantly. A dynamic balance may accordingly exist between the need for
NO release from partly deoxygenated erythrocytes to promote blood flow at low O2
tension, and the need for detoxification of nitrite in oxygenated erythrocytes at
inappropriate high nitrite concentrations (Jensen, 2005). The entry of nitrite into both
oxygenated and deoxygenated red blood cells may support both these functions. The
situation with the endogenous levels of both NO  and nitrite is like  balancing on a knife's
edge between potential beneficial effects at low levels and toxic effects at high levels.

       The idea that nitrite is converted to NO by deoxyHb has mainly been examined in
mammalian models. A recent study tested the hypothesis that nitrite is converted to
vasoactive NO in the coronary circulation of the isolated trout heart. Perfusion of the
                                        127

-------
coronary vessels with hypoxic saline elicited NO production. The nitric oxide synthase
inhibitor L-NA inhibited this NO production and decreased coronary flow, showing that
NO produced in the endothelium is vasoactive (Jensen and Agnisola, 2005). A switch to
perfusion with an erythrocyte suspension caused an increased NO signal that was also
inhibited by L-NA. The change in NO production upon subsequent nitrite addition, in
contrast, was not inhibited by L-NA, suggesting that it may have occurred via deoxyHb-
mediated reduction of nitrite to NO. All prerequisites for a conversion of nitrite to NO
inside the red blood cells appeared fulfilled: (1) nitrite rapidly permeated the erythrocyte
membrane, (2) there was a significant decrease in HbO2 saturation in the coronary
circulation, (3) there was a gradient in [NO2"] and a rise in metHb between the input and
output of the coronary circulation, and (4) a nitric oxide signal was registered (Jensen and
Agnisola, 2005). Thus,  the study supports the idea that NO can be produced from nitrite
in the erythrocytes, but it cannot be excluded that the heart itself may generate NO from
nitrite by means of cellular heme proteins or xanthine oxidoreductase activity (Jensen and
Agnisola, 2005). The NO formation associated with nitrite was without effect on
coronary flow. So, apparently NO was produced from nitrite without causing
vasodilation. This may  reflect that the NO was produced in the capillaries after the
resistance vessels, and that the signal was not conducted to upstream arterioles (Jensen
and Agnisola, 2005). Further research is required to fully uncover the role of nitrite-
erythrocyte interactions in blood flow regulation.
Acknowledgment

     The work was supported, in part, by the Danish Natural Science Research Council.

References

Aggergaard, S. and F.B. Jensen. 2001. Cardiovascular changes and physiological
       response during nitrite exposure in rainbow trout. J. Fish Biol. 59: 13-27.

Anderson, J.J., A. Okubo, A.S. Robbins, and F.A. Richards.  1982. A model for nitrite and
       nitrate distributions in oceanic oxygen minimum zones. Deep Sea Res. 29: 1113-
       1140.

Balmelle, B., K.M. Nguyen, B. Capdeville, J.C. Cornier, and A. Deguin. 1992. Study of
       factors controlling nitrite build-up in biological processes for water nitrification.
       Water Sci. Tech. 26: 1017-1025.

Bath, R.N., and F.B. Eddy. 1980. Transport of nitrite across fish gills. J. Exp. Zool. 214,
       119-121.

Brauner, C.J., A.L. Val, and D.J. Randall. 1993. The effect of graded methaemoglobin
       levels on the swimming performance  of Chinook salmon (Oncorhynchus
       tshawytscha). J. Exp. Biol. 185: 121-135.
                                        128

-------
Bryan, N.S., B.O. Fernandez, S.M. Bauer, M.F. Garcia-Saura, A.B. Milsom, T. Rassaf,
      R.E. Maloney, A. Bharti, J. Rodriguez, and M. Feelisch. 2005. Nitrite is a
      signaling molecule and regulator of gene expression in mammalian tissues. Nat.
      Chem. Biol. 1:290-297.

Cosby, K., K.S. Partovi, J.H. Crawford, R.P. Patel, C.D. Reiter, S. Martyr, B.K. Yang,
      M.A. Waclawiw, G. Zalos, X. Xu, K.T. Huang, H. Shields, D.B. Kim-Shapiro,
      A.N. Schechter, R.O. Cannon III, and M.T. Gladwin. 2003. Nitrite reduction to
      nitric oxide by deoxyhemoglobin vasodilates the human circulation. Nat. Med. 9:
       1498-1505.

Crawford, J.H., T.S. Isbell, Z. Huang, S. Shiva, B.K. Chacko, A.N. Schechter, V.M.
      Darley-Usmar, J.D. Kerby, J.D. Lang Jr, D. Kraus, C. Ho, M.T. Gladwin, and
      R.P. Patel. 2006. Hypoxia, red blood cells, and nitrite regulate NO-dependent
      hypoxic vasodilation. Blood 107: 566-574.

Doblander, C. and R. Lackner. 1997. Oxidation of nitrite to nitrate in isolated
      erythrocytes: a possible mechanism for adaptation to environmental nitrite. Can.
      J. Fish. Aquat. Sci. 54: 157-161.

Eddy, F.B. and E.M. Williams. 1987. Nitrite and freshwater fish. Chem. Ecol. 3:  1-38.

Evans, D.H., P.M. Piermarini, and K.P. Choe. 2005. The multifunctional fish gill:
      dominant site of gas exchange, osmoregulation, acid-base regulation, and
      excretion of nitrogenous waste. Physiol. Rev. 85: 97-177'.

Gladwin, M.T., A.N. Schechter, D.B. Kim-Shapiro, R.P. Patel, N. Hogg, S. Shiva, R.O.
      Cannon III, M. Kelm, D.A. Wink, M.G. Espey, E.H. Oldfield, R.M. Pluta, B.A.
      Freeman, J.R. Lancaster Jr, M. Feelisch, J.O. Lundberg. 2005. The emerging
      biology of the nitrite anion. Nat. Chem. Biol. 1: 308-314.

Grosell, M. and F.B. Jensen. 1999. NCV uptake and HCOs"  excretion in the intestine of
      the European flounder (Platichthysflesus). J. Exp. Biol. 202: 2103-2110.

Grosell, M. and F.B. Jensen. 2000. Uptake and effects of nitrite in the marine teleosts fish
      Platichthysflesus. Aquat. Toxicol. 50: 97-107.

Huang, Z., S. Shiva, D.B. Kim-Shapiro, R.P. Patel, L.A. Ringwood, C.E. Irby, K.T.
      Huang, C. Ho, N. Hogg, A.N. Schechter, and M.T. Gladwin. 2005. Enzymatic
      function of hemoglobin as a nitrite reductase that produces NO under allosteric
      control. J. Clin. Invest. 115: 2099-2107.

Hyde, D.A. and S.F. Perry. 1989. Differential approaches to blood acid-base regulation
      during exposure to prolonged hypercapnia in two freshwater teleosts: the rainbow
      trout (Salmo gairdneri) and the American eel (Anguilla rostrata). Physiol. Zool.
      62: 1164-1186.
                                       129

-------
Jensen, F.B. 1990. Nitrite and red cell function in carp: control factors for nitrite entry,
       membrane potassium ion permeation, oxygen affinity and methaemoglobin
       formation. J. Exp. Biol. 152: 149-166.

Jensen, F.B. 1992. Influence of haemoglobin conformation, nitrite and eicosanoids on K+
       transport across the carp red blood cell membrane. J. Exp. Biol. 171: 349-371.

Jensen, F.B. 2003. Nitrite disrupts multiple physiological functions in aquatic animals.
       Comp. Biochem. Physiol. 135A: 9-24.

Jensen, F.B. 2005. Nitrite transport into pig erythrocytes and its potential biological role.
       Acta Physiol. Scand.  184: 243-251.

Jensen, F.B., and C. Agnisola. 2005. Perfusion of the isolated trout heart coronary
       circulation with red blood cells: effects of oxygen supply and nitrite on coronary
       flow and myocardial oxygen consumption. J. Exp. Biol. 208:  3665-3674.

Jensen, F.B., N.A. Andersen, and N. Heisler. 1987. Effects of nitrite exposure on blood
       respiratory properties, acid-base and electrolyte regulation in the carp (Cyprinus
       carpio). J. Comp. Physiol. 157B: 533-541.

Kleinbongard, P., A. Dejam, T. Lauer, T. Rassaf, A. Schindler, O. Picker, T. Scheeren, A.
       Godecke, J. Schrader, R. Schulz, G. Heusch, G.A. Schaub, N.S. Bryan, M.
       Feelisch, and M. Kelm. 2003. Plasma nitrite reflects constitutive nitric oxide
       synthase activity in mammals. Free Radic. Biol. Med. 35: 790-796.

Knudsen, P.K. and F.B. Jensen. 1997. Recovery from nitrite-induced
       methaemoglobinaemia and potassium balance disturbances in carp. Fish Physiol.
       Biochem. 16: 1-10.

Kosaka, H. and I. Tyuma. 1987. Mechanism of autocatalytic oxidation of oxyhemoglobin
       by nitrite. Environ. Health Perspect. 73: 147-151.

Lewis, W.M. and D.P. Morris. 1986. Toxicity of nitrite to fish: a review. Trans. Am.
       Fish. Soc. 115: 183-195.

Lundberg, J.O. and E. Weitzberg. 2005. NO generation from nitrite and  its role in
       vascular control. Arterioscler.  Thromb. Vase. Biol.  25: 915-922.

Margiocco, C., A. Arillo, P. Mensi,  and G. Schenone. 1983. Nitrite bioaccumulation in
       Salmo gairdneri Rich, and hematological consequences. Aquat. Toxicol. 3: 261-
       270.

Marshall, W.S. 2002. Na+, Cl", Ca2+ and Zn2+ transport by fish gills: retrospective review
       and prospective synthesis. J. Exp. Zool. 293: 264-283.
                                       130

-------
Mazik, P.M., M.L. Hinman, D.A. Winkelmann, S.J. Klaine, andB.A. Simco. 1991.
       Influence of nitrite and chloride concentrations on survival and hematological
       profiles of striped bass. Trans. Am. Fish. Soc. 120: 247-254.

Nagababu, E.,  S. Ramasamy, D.R. Abernethy, and J.M. Rifkind. 2003. Active nitric
       oxide produced in the red cell under hypoxic conditions by deoxyhemoglobin-
       mediated nitrite reduction. J. Biol. Chem. 278: 46349-46356.

Palachek, R.M. and J.R.  Tomasso. 1984. Toxicity of nitrite to channel catfish (Ictalurus
      punctatus), tilapia (Tilapia aurea), and largemouth bass (Micropterus salmoides):
       evidence for a nitrite exclusion mechanism. Can. J. Fish. Aquat. Sci. 41:  1739-
       1744.

Perrone,  S.J. and T.L. Meade. 1977. Protective effect of chloride on nitrite toxicity to
       coho salmon (Oncorhynchus kisutch). J. Fish. Res. Bd. Can. 34: 486-492.

Perry, S.F. 1997. The chloride cell: structure and function in the gills of freshwater fishes.
       Annu. Rev. Physiol. 59: 325-347.

Randall, DJ. and C. Brauner. 1998. Interactions between ion and gas transfer in
       freshwater teleost fish. Comp. Biochem. Physiol. 119A: 3-8.

Russo, R.C., R.V. Thurston, and K. Emerson. 1981. Acute toxicity of nitrite to rainbow
       trout (Salmo gairdneri): effects of pH, nitrite species, and anion species.  Can. J.
       Fish. Aquat. Sci.  38: 387-393.

Stormer,  J., F.B. Jensen,  and J.C. Rankin. 1996. Uptake of nitrite, nitrate, and bromide in
       rainbow trout, Oncorhynchus mykiss: effects on ionic balance. Can. J. Fish.
       Aquat.  Sci. 53: 1943-1950.

Svobodova, Z., J. Machova, G. Poleszczuk, J. Huda, J. Hamackova, and H. Kroupova.
       2005. Nitrite poisoning offish in aquaculture facilities with water-recirculating
       systems. Acta Vet. Brbo 74: 129-137.

Tomasso, J.R.  1986. Comparative toxicity of nitrite to freshwater fishes. Aquat.  Toxicol.
       8: 129-137.

Tomasso, J.R.  and M. Grosell. 2005.  Physiological basis for large differences in
       resistance to nitrite among freshwater and freshwater-acclimated euryhaline
       fishes. Environ. Sci. Technol. 39: 98-102.

Williams, E.M. and F.B. Eddy. 1986. Chloride uptake in freshwater teleosts and its
       relationship to nitrite uptake and toxicity. J. Comp. Physiol. 156B: 867-872.
                                       131

-------
Williams, E.M. and F.B. Eddy. 1988. Anion transport, chloride cell number and nitrite-
       induced methaemoglobinaemia in rainbow trout (Salmo gairdnerf) and carp
       (Cyprinus carpio). Aquat. Toxicol. 13: 29-42.
                                      132

-------
                             Hypoxia in fish

                                      by

              DJ. Randall1, W.L. Poon1, C.C.Y. Hung1 and T.K.N. Tsui1

Introduction

       Vertebrates try to maintain oxygen delivery in the face of reductions in water
oxygen levels.  If oxygen delivery is compromised and tissue oxygen levels fall then
energy expenditure is reduced and anaerobic metabolism is up-regulated (Boutilier etal.,
1987). Fish reduce energy expenditure during hypoxia by inhibiting feeding and
reproduction, moving to a lower temperature and reducing swimming activity (see
Randall, 2004, for review). These energy savings are considerable and genes associated
with aerobic metabolism are down-regulated, probably in response to the reduction in
aerobic energy expenditure (Hung, 2005). Anaerobic metabolism is up-regulated to
maintain function in the face of limitations in aerobic energy production. The liver plays
a central role in these responses, but studies of the effects of hypoxia on liver cellular
changes in vivo are rare compared with in vitro studies. Hypoxia causes DNA damage
and apoptosis in mammalian cell lines (Thompson, 1998; Bras etal., 2005), however, in
vivo responses to DNA damage are known for only a few mammals and nothing is
known of in vivo responses offish liver to hypoxic DNA damage.

       In studies of the in vivo responses of the liver of common carp, Cyprinus carpio
L, to hypoxia we observed extensive DNA damage during the first days of hypoxic
exposure, as indicated by Terminal transferase mediated dutp Nick End Labelling
(TUNEL). TUNEL labeling was very high (found in around 60% of the liver cells)
during hypoxia, especially after four days of exposure to aquatic hypoxia at 0.5 mg O2.
L-l (Poon and  Randall, 2003). The level  of TUNEL  staining was reduced after about a
week of hypoxic exposure. Such extensive DNA damage will often lead to programmed
cell death or apoptosis. In fact TUNEL is often used to indicate  apoptosis.
1 Department of Biology and Chemistry, City University of Hong Kong, Kowloon, Hong
Kong, S.A.R., China.
                                      133

-------
                 0-day normoxic control
4-day hypoxic (0.3mg02/L)
               42-day normoxic control
42-day hypoxic (0.5mg02/L)
       Figure 1. Sections of normoxic and hypoxic carp liver. Cells stained by TUNEL
               and examined with confocal microscopy are shown in fluorescent green.
               The 4-day hypoxic liver was exposed to 0.3mgo2/L whereas the 42-day
               hypoxic liver was exposed to 0.5mgO2/L. Scale bar 20um.

       If the TUNEL signal was indicative of rates of apoptosis in the in vivo hypoxic
carp liver then, in the face of low rates of cell proliferation, the carp liver should have
been reduced in size after six weeks of hypoxia (0.5 mg O2. L-l), however, both the size
of the liver and the number of liver cells did not change significantly during these 42 days
of hypoxia. Cell-proliferation rates were always low, as indicated by the protein
expression level of a cell mitosis indicator, proliferating cell nuclear antigen (PCNA), and
by flow cytometry. Thus rates of apoptosis, which were low in the normoxic liver,
appeared not to increase during hypoxia. The absence of any increase in apoptosis during
hypoxia was also supported by results using a single strand DNA (ssdna) antibody to
assay for single strands of DNA, and a DNA fragmentation assay, as well as flow
cytometric analysis of normoxic and hypoxic liver cells. We also measured the activities
of caspase-3,  an enzyme involved in one of the pathways of programmed cell death, and
found no change in activity. A TEM investigation of carp liver cells indicated no
inflammation, and no necrosis or apoptosis during hypoxia. Thus  liver cells were stained
by TUNEL during the first days of hypoxia, indicating DNA damage, but there was no
apoptosis. As the level of TUNEL staining decreased during prolonged hypoxia, DNA
                                       134

-------
nick ends must have been repaired during hypoxia, probably after the initial stress had
been ameliorated by various cellular changes.
Table I. There is no increase in apoptosis in carp liver, in vivo, when exposed to hypoxia.
        Various methods were used to determine the extent of apoptosis and the results
        are summaries in the text (from Poon and Randall, 2006).
Experiment
Anti-ssdna staining
Caspase-3 activity
DNA fragmentation
Flow cytometry
Liver size
Mrna level of Bel -2
Number of cells per mm2
(cell size)
PCNA
TEM
TUNEL
Result
No observable difference at 0-, 4-, 8-, 16-, 28-, and 42-day
hypoxia (0.5mgO2/L)
No significant increase at 4-day hypoxia (0.3mgO2/L)
No DNA laddering observed at 42-day hypoxia (0.5mgo2/L)
No significant increase percentage of cells under sub-Gl peak
at 7-day hypoxia (0.5mgO2/L)
No significant decrease at 42-day hypoxia (0.5mgO2/L)
A significant increase at 4-day hypoxia (0.5mgO2/L)
No significant change at 0-, 4-, 8-, 16-, 28-, and 42-day hypoxia
(0.5mgo2/L)
No significant change at 42-day hypoxia (0.5mgO2/L)
No significant increase number of irregular shaped nuclei at 2-
day hypoxia (0.5mgo2/L)
More than 5-fold increase was observed at 4-day hypoxia
(0.3mgo2/L) and 4-, 16-, 28-, and 42-day hypoxia (0.5mgo2/L)
       In order to repair DNA, cell cycle arrest must be maintained. During hypoxia
there was an initial increase in protein levels of the cell cycle inhibitor, p27, and the
increase in this protein was associated with the increased levels of DNA damage. In
addition, there were increases in the levels of a number of anti-apoptotic genes, including
Bcl2 and erythropoietin (EPO), and the down regulation of pro-apoptotic genes such as
Tetraspanin 5 and Cell death activator in the in vivo carp liver. The anti-apoptotic factor,
heat shock protein 70 (HSP70) decreased during hypoxia (Poon etal., 2006, unpublished
data). Double immunological staining of liver cells for TUNEL and HSP70 indicated that
only cells with no TUNEL staining showed high levels of HSP70. Thus HSP70 was
associated with undamaged cells.

       We conclude that the carp liver, in vivo, is maintained in a quiescent state during
hypoxia with no change in apoptotic rate and little or no cell division, maintained by a
variety of mechanisms, including p27, and the up-regulation of a number of anti-
apoptotic genes and the down regulation of a number of pro-apoptotic genes. The cellular
mechanisms seem to be directed towards preventing apoptosis in the face of DNA
damage and promoting DNA repair.

       When common carp were exposed to hypoxia of 0.5 mg O2 L-l for six weeks,
most of the changes in gene expression occurred during the first few day after the onset
of hypoxia (Hung, 2005). These changes were not related to starvation as starvation alone
                                       135

-------
is associated with only minor changes in gene expression during the first week. The
changes in expression appear to be a response to, rather than a cause of, the reduction in
metabolism. Hypoxia inducing factor 1 (HIF1) expression increased early and was then
reduced during long term exposure to hypoxia. HIF2 and HIF4 expression also increased.
The relative role of these various hifs in fish is not clear.

       There was also a large increase in the gene expression and level of uncoupling
proteins (ucps) in carp liver, in vivo, during hypoxia. In mammals it is clear that ucps are
inserted into the inner mitochondrial membrane and will short circuit the proton gradient
generated by NADH oxidation (see Krauss et a/., 2005 for review). In the presence of
oxygen there will be a futile cycle of protons across the mitochondrial membrane the
consequence of which is the production of heat. This is important in regulating heat
production in mammals, but why are there uncoupling proteins in fish, when clearly fish
are not homeothermic endotherms like mammals? Increases in heat production in fish
would indeed be futile. The rate of production of reactive oxygen species (ROS) is
related to mitochondrial membrane potential, at least in the rat (Korshunov et a/., 1998).
The production of reactive oxygen species decreases with mitochondrial membrane
potential and UCP lowers mitochondrial membrane potential and, therefore, ROS
production.

       UCP2 mrna was up-regulated more than six fold in carp liver during hypoxia
exposure and, using isolated mitochondrial preparations, we were able to show palmitate
stimulation of UCP activity. In addition, palmitate reduces H2O2 production in isolated
fish mitochondria, presumably by activating UCP. Free fatty acid levels increase during
hypoxia in the carp liver and this may stimulate UCP2 to decrease mitochondrial
membrane potential and, thus, decrease ROS production. UCP appears not to be inhibited
by ATP in carp and  so the control of UCP may be somewhat different from that of
mammals, which is known to be inhibited by ATP (Hagen et a/.,  2000). In summary it is
possible that ucps can rapidly decrease the production of ROS in  fish liver during
hypoxia.

       ROS levels in tissues can be regulated either by reducing  production or increasing
rates of removal by increasing antioxidant capacity. We found large increases in both
UCP1 and UCP2 mrna levels in carp liver during hypoxia. There were much smaller but
significant increases in UCP1 mrna levels in brain tissue but there was only a small initial
increase and then a decrease in UCP gene expression in the kidney during hypoxia.
Increasing rates of ROS removal during hypoxia do not seem to be an option  as several
antioxidant genes were suppressed in both liver and kidney during hypoxia. The kidney
has a high antioxidant capacity normally and perhaps this suffices during hypoxia and up
regulation of kidney ucps is not required. The liver, on the other hand,  appears to control
production rather than removal during hypoxia. The kidney presumably attempts to
maintain aerobic capacity as most of the changes in gene expression returned to levels
close to control levels during hypoxia whereas those in the liver do not. Thus responses to
hypoxia at the tissue level are complex and are probably tissue specific.
                                       136

-------
Acknowledgement

       This work was supported by the Research Grants Council of the Hong Kong
Special Administrative Region, China. Project number 9040658-740 and by the Areas of
Excellence Scheme established under the University Grants Committee of the Hong
Kong Special Administrative Region, China (Project No. Aoe/P-04/2004). Common carp
used in this work was supplied by the Au Tau Station, Agriculture, Fisheries and
Conservation Department, Hong Kong Special Administrative Region Government.

References

Boutilier, R.G., G. Dobson, U. Hoeger, and DJ. Randall. 1987. Acute exposure to graded
       levels of hypoxia in rainbow trout (Salmo gairdneri): metabolic and respiratory
       adaptations. Resp. Physiol. 71: 69 82.

Bras, M., B.  Queenan, and S.A.  Susin. 2005. Programmed cell death via mitochondria:
       different modes of dying. Biochemistry-Moscow. 70: 231-239.

Hagen, T., C.Y. Zhang, C.R. Vianna,  and B.B. Lowell. 2000. Uncoupling proteins 1  and
       3 are regulated differently. Biochemistry 39: 5845-5851.

Hung, C.C.Y. 2005. Survival strategies of Common Carp for prolonged starvation and
       hypoxia. Ph.D. Thesis, City University of Hong Kong.

Korshunov, S.S., O.V. Korkina, E.K. Ruuge, V.P. Skulachev, and A.A. Starkov. 1998.
       Fatty acids as natural uncouplers preventing generation of O2- and H2O2 by
       mitochondria in the resting state. FEES Lett. 435: 215-218.

Krauss, S., C.Y. Zhang, and B.B. Lowell. 2005.  The mitochondrial uncoupling-protein
       homologues. Nat. Rev. Mol. Cell Biol. 6: 248-261.

Poon, W.L. and D. J. Randall. 2006. Hypoxia does not induce apoptosis in common carp
       (Cyprinus carpio L.) Liver in vivo. Submitted for publication.

Poon, W.L., K. Nakano, C.C.Y. Hung, DJ. Randall. 2006. Hypoxia-induced cell cycle
       arrest in common carp (Cyprinus carpio L.) Liver in vivo, unpublished results

Poon, W.L. and DJ. Randall. 2003. Effect of hypoxia in common carp: cell life or death.
       The 7th International Symposium on Fish Physiology, Toxicology and Water
       Quality, Tallinn, Estonia, 12-15 May, 2003.

Randall, D J. 2004. Hypoxia in the aquatic environment. The 8th International
       Symposium on Fish Physiology, Toxicology and Water Quality, Chongqing,
       China, 12th-14th October 2004.

Thompson, E.B. 1998. Special topic: apoptosis. Annu. Rev. Physiol. 60: 525-532.
                                      137

-------
138

-------
   Gene expression  profiles of common carp, Cyprinus

     carpio, during prolonged starvation and hypoxia

             reflect differences in hypometabolism

                                     by

            C.C.Y. Hung1, A.R Cossins2, A.Y. Gracey3 and DJ. Randall1


Introduction: Metabolic Depression

      Metabolic depression in response to environmental stress has been reported in
both invertebrates and vertebrates. A diverse range of environmental stressors that result
in metabolic depression in animals have been studied, including anoxia / hypoxia (e.g.
crucian carp and coral-reef shark; Nilsson  and Renshaw, 2004), food deprivation (torpor
in salamander; Hervant et a/., 2001), dehydration (e.g. aestivation in snails and lungfish;
Guppy etal., 2000; Chew et a/., 2004) and diapause (embryos of annual killifish;
Podrabsky and Hand, 1999). Much work has been devoted to understanding metabolic
depression by means of suppressing energy production (e.g. oxidative phosphorylation)
and energy consuming processes (e.g. ion pumping, protein synthesis etc) in a
coordinated fashion, and many reviews have been published on this subject (Hochachka
etal., 1996; Hand, 1996; Guppy, 2004; Storey and Storey, 1990, 2004). These reviews
tend to concentrate on biochemical mechanisms of metabolic depression and ignore
behavioural and physiological strategies such as moving to a lower temperature, reduced
activity, and inhibition of feeding and reproduction. Several studies investigating the
responses to environmental stresses by means of global gene expression screening have
been carried out in the last decade (O'Hara et al., 1999; Gracey et al., 2001, 2004).
However, our knowledge of molecular adaptation to hypoxia is still very limited and the
molecular mechanisms associated with physiological adaptations have yet to be
delineated.

      Survival of animals subjected to  environmental stress depends on several factors
centred on matching energy expenditure to reduced energy availability through energy
cost saving. Common carp, although not an anoxia tolerant species, are able to survive a
substantial period of hypoxia exposure and starvation. There are many similarities in the
responses of animals to hypoxia and starvation such as depressed metabolism, reduced
locomotion and impaired reproductive ability. Nevertheless, not all responses are the
same. During hypoxia, ATP supply is limited and therefore a quick response to maintain
ATP balance is followed by regulated hypometabolism (Boutilier, 2001). During
Department of Biology and Chemistry, City University of Hong Kong, Hong Kong PRC
2School of Biological Sciences, Crown Street, University of Liverpool, Liverpool L69
7ZB, UK
3Department of Biology and Chemistry, City University of Hong Kong, Hong Kong PRC
                                     139

-------
starvation, however, the rate of energy store utilization becomes the deciding factor in the
survival offish. The mechanisms of gene expression by which cells respond and adapt to
hypoxic conditions have recently been studied (Semenza, 2001; Gracey etal., 2001), but
less is known about cellular responses to starvation in fish. The processes of coordinated
gene expression that mediate the whole cellular response towards energy conservation are
virtually unknown. Therefore, we hypothesized that there are similarities in the gene
expression profiles during starvation and hypoxia, in both cases directed towards energy
conservation. In this study, we explored the underlying mechanism of hypoxia tolerance
in common carp and their survival during starvation at the molecular level, and to relate
the changes of gene expression with changes in stored substrates.

Materials and Methods

       Carp were exposed to prolonged (six weeks) exposure to hypoxia or starvation.
All fish survived these treatments. Water temperature was maintained between 19±1°C
with a constant photoperiod 12L:12D. The dissolved oxygen level (DO) was maintained
at approximately 7.0mgO2/L for the starved group and between 0.5-0.65mgO2/L for the
hypoxic fish. Hypoxic fish were fed daily but starved fish were deprived of food
completely. Liver and kidney samples were collected at day 0 (control), 4, 8, 16, 28, 42
and frozen immediately in liquid nitrogen until processing. Total RNA was extracted
using Trizol and further purified using QIAgen RNeasy mini kit. Total RNA from all
time points were reverse-transcribed, labeled and hybridized to carp microarray (Gracey
et a/., 2004). Data analysis and statistics were done using the software GeneSpring.
Hepatic glycogen, protein and lipid levels were also measured.

Results and Discussion

       Metabolic depression during food deprivation and hypoxia, although not
measured in this study, has been reported in several fish species including starved perch
(Mehner and Weiser, 1994) and rainbow trout (Lauff and Wood, 1996), as well as
hypoxic carp (Van Ginneken etal., 1998). Liver and kidney gene expression profiles
reflected the metabolic depression observed in both starved and hypoxic carp; however,
the responses were somewhat different, with a different time frame between starvation
and hypoxia and between tissues. Table I shows the general gene  expression patterns of
common carp during hypoxia and starvation. In general, carp kidney genes respond to
starvation much faster than liver.  The expressions of liver genes remain relatively
unchanged until day 16 during starvation, indicating that this period of starvation does
not have any impact on the hepatic metabolism in carp, and this animal might be
employing a "wait and see" strategy during this time. Unlike starvation, carp liver and
kidney responded to hypoxia similarly and more acutely (i.e. within the first few days of
exposure).
                                       140

-------
Glycolysis and oxidative phosphorylation

       During starvation, hepatic glycogen dropped rapidly at day 4 and this level was
maintained until day 42 when another significant decrease was detected (Fig. 1). At the
molecular level, induction of expression of numerous glycolytic genes including pyruvate
kinase (PK), 6-phosphofructokinase (PFK), L-lactate dehydrogenase A (LDH-A), P-
enolase and fructose-bisphosphate aldolase B were detected after 42 days of starvation.
Similar temporal induction was also observed in tricarboxylic cycle (TCA) genes (citrate
synthase and isocitrate dehydrogenase) and gluconeogenic genes (G6Pase, F-1,6-BP,
phenylalanine catabolic genes, 4-hydroxyphenylpyruvate dehydrogenase and
homogentisatel,2-deoxygenase). This induction of glycolytic genes coincides with a
decrease in hepatic glycogen content at day 42 (Fig. 1). Carp kidney, on the contrary,
sustained suppression of genes encoding enzymes for glycolysis, gluconeogenesis, and
TCA cycle.

       Reduction in resting metabolism may have occurred prior to day 4, and pre-
existing glycolytic enzyme activity was sufficient to drive the usage of glycogen without
having to increase the expression of glycolytic or glycogenolysis genes. The early decline
in glycogen concentrations seen in this study might be glucagon-driven as suggested by
Moon and Foster (1995), and the late-mobilization of hepatic glycogen also suggests that
glycogen is either not the preferred fuel or it is reserved during starvation in fish. One
probable explanation is that hypoxia is a common environmental stress,  and during
oxygen limited  periods, anaerobic respiration is the only means of ATP production.
Therefore, it is crucial for carp to reserve glycogen which aids its survival during
episodes of oxygen shortage.
 OS)
 ~o
 5
 3.
                     10     15    20    25    30     35
                            Days after starvation
40
45
Figure 1. Changes of glycogen content of carp liver during starvation.
         Data are presented as Mean + SE (One-Way ANOVA, p<0.05).
         a: significantly different from Day
         b: significantly different from Day 0, 4, 8, 16, 28.
                                       141

-------
       On the other hand, when carp were exposed to hypoxia, hepatic glycogen declined
rapidly at day 4 and then gradually returned to control levels from day 16 onwards (Fig.
2). The initial decline of hepatic glycogen is probably related to the induction of
numerous glycolytic genes (fructose-bisphosphate aldolases, hexokinase, glyceraldehydes
3-phosphate dehydrogenase, phosphoglycerate mutases, beta-enolase) during early
hypoxia exposure. Similar patterns of glycolytic gene induction are seen in the liver of G.
mirabilis exposed to hypoxia (Gracey etal., 2001). The up-regulation of glycolytic genes,
in turn, was probably mediated by hypoxia-inducible factor la (HIFla) which showed a
significant 1.2-fold (p<0.05) increase at day 4 of hypoxia. Congruent with glycolytic
gene expression, gluconeogenic genes (fructose-1, 6-bisphosphatase and fructose-1,6-
bisphosphatase isozyme 2) and phenylalanine catabolism genes, whose pathway is solely
cytosolic, were also induced. A similar induction of gene expression was also observed in
the carp kidney during hypoxia. The degree of induction of the above genes gradually
decreased as hypoxia continued. Changes in gene expression and in hepatic glycogen
levels indicate that common carp were probably well adjusted to the hypoxic
environment after 2 weeks of exposure and able to replenish glycogen stores from
consumed food even though they ingested only half of the normoxic food intake. Rapid
restoration of hepatic glycogen further indicates that glycogen is important as an energy
source and necessary for carp to survive hypoxia.
                                       rb,c
                                10    15   20   25   30
                                       Days after hypoxia
35
40
45
              Figure 2. Change of hepatic glycogen of carp during hypoxia exposure.
                      Data are presented as Mean + SE (One-Way ANOVA, p<0.05).
                      a: significantly different from Day 0
                      b: significantly different from Day 4
                      c: significantly different from Day 8
                      d: significantly different from Day 16.
Oxidative phosphorylation and ATP production

       Starvation did not induce congruent patterns of differential expression of genes
involved in aerobic respiration and the electron transport chain in the carp liver. At the
                                       142

-------
same time, expression of beta-oxidation genes remained unaltered, which implies that
mitochondrial function was not impaired during food withdrawal. In the carp kidney,
however, sustained suppression of genes encoding enzymes for TCA cycle, ETC, beta-
oxidation and ATP synthases was observed. Because oxygen is not limited during food
withdrawal, mitochondria can remain functional and supply energy via aerobic
respiration. The general suppression of genes involved in ATP production in carp kidney
indicates that energy demand during starvation has decreased substantially.

       During hypoxia, suppression of various ETC genes (cytochrome P450 3A40 and
2J2, NADH-ubiquinone oxidoreductase subunitB14.5b) and ATP synthases (ATP
synthase beta, delta chains and ATP synthase lipid-binding protein) was detected in the
carp liver and a greater number of ETC and ATP synthase genes were found suppressed
in the carp kidney.  Carp are oxyconformers meaning that their energy demands decline
quickly with decreasing external  oxygen supply resulting in their entering a state of
hypometabolism. The lack of oxygen resulted in an inhibition of ATP supply and
probably a depletion of ATP stores; hence the animal is forced to enter hypometabolism.
This forced hypometabolism probably occurred at the onset of hypoxia exposure. With
prolonged hypoxia, hypometabolism was associated with suppression of a large number
of genes involved in ATP  consumption and turnover.

       In addition to controlling  ATP production by ETC, production of adenosine has
been proposed to contribute to the regulation of depressed metabolism and its elevation
has been reported during times of energy deficiency in fish, including hypoxia. The
actions of adenosine include 1) stimulation of glycogenolysis to fuel glycolysis with
glucose (Magistretti et a/., 1986); 2) inhibition of ATP uptake during protein synthesis
(Tinton and Buc-Calderon, 1995); 3) a relative reduction in anaerobic respiration (Bernier
et a/., 1996); 4) impairment of protein synthesis (Krumschnabel et a/., 2000); 5)
depression of cardiac activity (MacCormack and Driedzic, 2004); as well as 6) channel
arrest in the brain of anoxic turtle (Buck, 2004). In other words, adenosine aids in
hypoxia and anoxia survival by reducing ATP consumption to match production and
hence results in metabolic depression (Nilsson and Renshaw, 2004). In this  study,
sustained induction of hepatic S-adenosylhomocysteinase 2 during 6 weeks of hypoxia
exposure might be related to adenosine production (including the net breakdown of
phosphorylated adenylates ATP,  ADP and AMP). S-adenosylhomocysteinase 2 catalyzes
the reversible hydrolysis of S-adenosylhomocysteine to homocysteine and adenosine
without the requirement of cofactors (Palmer and Abeles, 1979). At the same time,
suppression of genes encoding S-adenosylmethionine synthetase alpha and beta forms
was detected in the kidney. This  enzyme prevents the coupling of methionine and
adenosine, hence sparing one molecule of ATP and adenosine. The overall reduction in
gene expression during prolonged hypoxia in both liver and kidney in this study in
hypoxic carp was probably in response to metabolic depression. No differential
expression of S-adenosylhomocysteinase and S-adenosylmethionine synthetase alpha and
beta forms was found in the starved carp tissues, indicating that different control
mechanisms were involved in metabolic  depression during hypoxia and starvation.
                                       143

-------
Lipid and Protein Metabolism

       Synthesis of macromolecules requires ATP, and not surprisingly, lipid synthesis
was suppressed during both starvation and hypoxia in carp. Suppression of lipogenesis
genes was observed at day 4 in liver and kidney of hypoxic carp and at day 8 and day 16
in starved carp kidney and liver, respectively. It was interesting to note that genes
encoding ribosomal proteins were suppressed only in carp kidney during starvation but
remained relatively unchanged in the liver. This indicates the important role of carp liver
as the main metabolic organ during starvation.

       Translation machinery, on the other hand, remained fairly active during hypoxia
in both liver and kidney, which might be related to the requirement for rapid production
of glycolytic enzymes in the two organs studied. Interestingly, the magnitude of the
increase in expression of translational factors and ribosomal proteins declined in the
kidney as hypoxia persisted, however induction remained high in the liver, indicating that
the liver is the main ATP supplier as the fish adjusted to hypoxic stress. Likewise,
suppression  of ribosomal gene expression was detected only in skeletal and cardiac
muscles, but not liver, of hypoxic goby fish (Gracey etal., 2001). In extreme cases of
anoxia, half-lives of proteins increased by 40% to 50%, respectively (Land and
Hochachka,  1994). The changes in expression of numerous signaling molecules and
transcriptional/translational modulators during starvation and hypoxia also reflect rapid
re-organization of cellular functions during these two stresses in carp.

       In addition to synthesis, degradation of macromolecules is also energy-expensive.
For example, protein degradation by ubiquitin-proteasome pathway is ATP-dependent.
Therefore, it becomes economical to have proteins preserved rather than degraded during
energetically stressed period such as hypoxia and starvation. During starvation, ubiquitin-
proteasome genes remained unchanged during the first 2 weeks of starvation but showed
congruent induction at day 28, but no  change in total hepatic protein was detected. This
could be due to either the amount of protein change being too little to be detected, or the
action of protein degradation being  counteracted by the up-regulation of ubiquitin
carboxyl-terminal hydrolase isoenzyme LI, a deubiquitining enzyme that prevents
protein degradation from taking place.

       In the case of oxygen shortage, it has been demonstrated that proteolysis is
inhibited in isolated, anoxic turtle hepatocytes (Land and Hochachka, 1994), as well as
anoxic A. franciscana embryos (Anchordoguy and Hand,  1994) and hibernating ground
squirrels (van Breukelen and Carey, 2002). Suppression of a large number of ubiquitin-
proteasome genes at day 4 in the carp kidney and thereafter, was subsequently
ameliorated. This is probably due to the adaptive strategy of carp to depress metabolism
during hypoxic conditions. Hence, proteolysis continued to maintain basal cellular
protein turnover. On the other hand, hypoxia had no effect on ubiquitin-proteasome gene
expression in the carp liver, indicating that this ATP-dependent pathway is still
functioning during hypoxia in the carp liver, probably related to the re-organisation of
cellular proteins. However, as observed during starvation, no change in total hepatic
                                       144

-------
protein content was observed during hypoxia. In the case of hypoxia, genes involved in
lipid and protein metabolic pathways were generally down-regulated. On the other hand,
no change in hepatic lipid content was observed in carp (data not shown) after 6 weeks of
starvation. Lipid metabolic genes showed varied changes in expression in both starved
carp liver and kidney.

Interactions between Starvation and Hypoxia in Common Carp

       During the first 2 days of hypoxia, common carp were reluctant to take food
pellets, but appetite resumed after 2 days, probably when the fish had adjusted to the
hypoxic condition. However, the amount of food intake during hypoxia was
approximately half the usual daily intake. Reduced food intake may lead to slower
growth rates of common carp under hypoxic conditions. Reduction of food intake and
retardation of growth is indeed commonly observed in fish during hypoxia exposure
(Secor and Gunderson, 1998; Pichavant et al, 2000; Taylor and Miller, 2001; Foss et al.,
2002). Although hepatosomatic index (HSI) of common carp after 6 weeks hypoxia was
not different from day  0 (data not shown), HSI  of common carp might have dropped (but
no measurements were made) during the initial phase of hypoxia exposure due to more
than 30% reduction of hepatic glycogen stores at day 4 and day 8. HSI of the parallel
control fish was significantly higher than the starved fish indicating that hepatic storage
has probably increased under fed condition. Hepatic glycogen stores  of the hypoxia fish,
however, returned to levels that were close to the time 0 level from day 16 onwards,
indicating that common carp were probably well adjusted to the hypoxic environment
after about 2 weeks of exposure and were able to replenish the glycogen store from
consumed food, even though ingestion was about half of the normoxic food intake.
Although there was a reduction of food intake by carp during the first couple days of
hypoxia exposure, its effect on gene expression was minimal, as changes in gene
expression were evident only after 16 days of starvation in the carp liver. Furthermore,
feeding is an energy-expensive process. Fish eat when energy is available to process
food. Therefore, some changes in gene expression late in hypoxia probably relate to the
restoration of feeding.

Conclusion

       Many physiological activities such as reproduction and locomotion are suppressed
during both hypoxia and starvation, and energy utilization is reduced. Responses to
hypoxia can occur rapidly, within hours, whereas changes associated with starvation are
much slower. During hypoxia, oxygen limitations resulted in a decrease in oxidative
phosphorylation. This led to a reduction of the ATP pool, which in turn resulted in
suppression of gene expression. Changes in gene expression appear to follow, rather than
direct, the changes in metabolism during hypoxia. In the case of starvation, oxygen is
readily available, and it seems that the liver continues as normal and  other tissues
contribute to the general reduction  of metabolism.  One major group of genes that were
suppressed in liver was digestive and lipogenesis enzymes; whereas severe suppression
of ATP production pathways, protein biosynthesis and many other cellular functions were
detected in kidney. With decreased ATP turnover during food deprivation, the liver was
                                       145

-------
able to supply ATP required for the animal, and only minimum alterations were seen in
the gene expression patterns. During prolonged hypoxia, in contrast, liver metabolism
had to be reorganized. Anaerobic respiration, mediated by HIF, was enhanced to sustain
the hypoxia-driven, reduced ATP turnover. The liver, with strong induction of anaerobic
respiration, appeared to play a more prominent role as an ATP contributor than kidney.
Gene expression during both starvation and hypoxia in carp is directed towards metabolic
depression. In the liver, genes were greatly suppressed in response to hypoxia, but not
many changes were seen in response to starvation. In the kidney, common suppression of
genes was seen during early hypoxia and starvation. However, suppression (or further
suppression) of genes was observed in prolonged starvation; whereas during hypoxia,
initial suppression was gradually lessened, probably reflecting the carp's ability to adjust
to the surrounding reduced oxygen level.

   In addition, the changes of gene expression did not appear to be substrate-driven in
both cases. Although hepatic glycogen was depleted during early  exposure to hypoxia, it
was later replenished as carp settled well into prolonged hypoxia exposire. During
starvation, carp appeared to prolong the period of substrate availability by conserving
stored substrates and consuming them at very slow rates. Hence, expressions of substrate
mobilizing-associated genes was maintained in the liver but suppressed in the carp
kidney. Nonetheless, both hypoxia and starvation resulted in reduction in reproduction-
associated gene expression. Energy was channeled to  self-survival during stressed
periods rather than to gamete production. During starvation, energy was probably
conserved to ensure survival, but in the case of hypoxia, inhibition of reproduction  may
be initially via the effects of cortisol, followed by suppression of reproduction-associated
gene expression. Clearly, in this study, metabolic depression in response to starvation is
greater in the carp kidney than liver. In fact, other less vital and energy-demanding
organs may show even more drastic reduction in energy use.

       The gene expression pattern of carp in response to starvation and hypoxia are
clearly different with little overlap. However, how metabolic depression is initiated
during exposure to hypoxia and starvation awaits elucidation.
                                        146

-------
Table I. General Gene Expression Patterns of Common Carp in Response to
       Prolonged Starvation and Hypoxia.

Metabolism /ATP
production
Glycolysis
Gluconeogenesis
TCA cycle
Oxidative
phosphorylation
and ATP
synthases
p-oxidation
Lipid transport &
metabolism
Proteasome-
ubiquitin pathway
Other proteases
and peptidases
Liver
Starvation
Induced after 28 days
Suppressed and then
returned to control
levels at day 42
Induced after 28 days
No change
No change
Varied expression
Induced after 28 days
Suppressed from day
16 onwards
(Digestive enzymes)
Hypoxia
Induced
Cytosolic
gluconeogenic genes:
induced
Suppressed at day 4 &
some genes gradually
returned to near time 0
levels by day 42
Suppressed
Suppressed
Suppressed
Varied expression
Varied expression
Kidney
Starvation
Suppressed
Suppressed
Suppressed
Suppressed
Suppressed or no
change initially,
increased at day 28
Induced from day 8
onwards
Varied expression
Varied expression
Hypoxia
Induced
Induced (cytosolic
gluconeogenic genes)
Suppressed and
returned to control
level by 42 days
Suppressed and
returned to control
levels by day 42
Suppressed and
returned to control
levels by day 42
Varied expression
Suppressed and
gradually returned to
control levels by day
28
Suppressed
Macromolecules Biosynthesis
Fatty acid / lipid
Protein:
Transcriptional
factors
Protein:
Translation
Suppressed from day
16 onwards
Varied expression
Translational factors:
varied expression;
Ribosomes: no
change
Suppressed
Poly(A)-binding
protein: induced
Ribosome genes: no
change
Suppressed from day
8 onwards
Increased expression
Translational factors:
varied expression;
Ribosomes:
suppressed from day
8 onwards
Suppressed
Varied expression
Varied expression
Others
Transporters
Oxidative stress
response
Immuno-related
Steroidogenesis
and reproduction
related
Varied expression
UCP1: suppressed
Antioxidant genes:
varied expression
Varied expression
Suppressed
O2 transporters-
associated genes:
varied expression;
intra-/intercellular
transporters:
suppressed;
UCP1: induced
Antioxidant genes:
suppressed
Varied expression
Suppressed
O2 transporters:
suppressed;
Others transporters:
varied expressions
UCP1: induced;
Antioxidant genes:
suppressed
MHC/ blood
coagulation factor:
induced; Cytokines /
inflammatory
response: suppressed
Not detected
Suppressed in
general; myoglobin
induced strongly
UCP1: induced at
day 4 only;
Antioxidant genes:
suppressed &
returned to time 0
levels
Immuno-related:
varied expression
Not detected
                                   147

-------
Acknowledgements
The work described in this abstract was fully supported by a grant from the University
Grants Committee of the Hong Kong Special Administrative Region, China (Project No
UGC 1224/02M / CREG Grant No.:  9040759)

Reference

Anchordoguy TJ. and Hand S.C. 1994. Acute blockage of the ubiquitin-mediated
       proteolytic pathway during invertebrate quiescence. Am. J. Physiol. 267: R895-
       900.

Bernier, N.J., J. Harris, J. Lessard, and DJ. Randall. 1996. Adenosine receptor blocker
       and hypoxia-tolerance in rainbow trout and Pacific hagfish. J. Exp. Biol. 199:
       485-495.

Boutilier, R.G. 2001. Mechanisms of cell  survival in hypoxia and  hypothermia. J. Exp.
       Biol. 204: 3171-3181.

Chew, S.F., N.K. Chan, A.M. Loong, K.C.  Hiong, W.L. Tarn, and Y.K. Ip. 2004.
       Nitrogen metabolism in the African lungfish (Protopterous dollof) aestivating in a
       mucus cocoon on land. J. Exp. Biol. 207: 777-786.

Foss, A., T.H. Evensen, and V. Oiestad. 2002. Effects of hypoxia and hyperoxia on
       growth and food conversion efficiency in the spotted wolfish Anarhichas minor
       (Olafsen). Aquae. Res. 33: 437-444.

Gracey, A.Y., J.V.  Troll,  and G.N. Somero. 2001. Hypoxia-induced gene expression
       profiling in the euryoxic fish Gillichthys mirabilis. Proc. Natl.  Acad. Sci. U.S.A.
       98: 1993-1998.

Gracey, A.Y.,  I.E. Fraser,  W. Li, Y. Fang, R.R. Taylor, J. Rogers, A. Brass, and A.R.
       Cossins.  2004.  Coping  with cold:  an  integrative, multitissue  analysis  of
       transcriptome of a poikilothermic vertebrate. Proc. Natl. Acad. Sci.  U.S.A.  101:
       16970-16975.

Guppy, M., D.C. Reeves, T. Bishop, P. Withers, J.A. Buckingham, and M.D.  Brand.
       2000. Intrinsic metabolic depression in cells isolated from the hepatopancreas
       festivating snails. FASEB J. 14: 999-1004.

Guppy, M. 2004. The biochemistry of metabolic depression: a history of perceptions.
       Comp. Biochem. Physiol. 139B: 435-442.

Hand, S.C. 1996. Downregulation of cellular metabolism during environmental stress:
       mechanisms and implications. Annu. Rev. Physiol. 58: 539-563.
                                      148

-------
Hervant, F., J. Mathieu, and J. Durand. 2001. Behavioural, physiological and metabolic
       responses to long-term starvation and refeeding in a blind cave-dwelling (Proteus
       anguinus) and a surface-dwelling (Euproctus asper) salamander. J. Exp. Biol.
       204:269-281.

Hochachka, P.W., L.T. Buck, CJ. Doll, and S.C. Land. 1996. Unifying theory of hypoxia
       tolerance: molecular / metabolic defense and rescue mechanisms for surviving
       oxygen lack. Proc. Natl. Acad. Sci. U.S.A. 93: 9493-9498.

Krumschnabel, G., C. Biasi, and W. Wieser. 2000. Action of adenosine on energetics,
       protein synthesis and K+ homeostasis in teleost hepatocytes. J. Exp. Biol. 203:
       2657-2665.

Land, S.C. and P.W. Hochachka. 1994. Protein turnover during metabolic arrest in turtle
       hepatocytes: role and energy dependence of proteolysis. Am. J. Physiol. 266:
       C1028-1036.

Lauff, R.F. and C. Wood.  1996. Respiratory gas exchange, nitrogenous waste excretion,
       and fuel usage during starvation in juvenile rainbow trout, Oncorhynchus mykiss.
       J. Comp. Physiol. 165B: 542-551.

MacCormack, T. J. and W.R. Driedzic. 2004. Cardiorespiratory and tissue adenosine
       responses to hypoxia and reoxygenation in short-horned sculpin Myoxocephalus
       scorpius. J. Exp. Biol. 207: 4157-4164.

Mehner, T. and W. Weiser.  1994.  Energetics and metabolic correlates to starvation in
       juvenile perch (Percafluviatilis). J. Fish Biol. 45: 325-333.

Moon, T. W. and G.D. Foster. 1995. Tissue carbohydrate metabolism, gluconeogenesis
       and hormonal and environmental influences. Pages 65-100 In: Biochemistry and
       molecular biology  of fishes, vol. 4. Metabolic biochemistry. P.W. Hochachka and
       T.P. Mommsen (Eds.). Elsevier, Amsterdam.

Nilsson, G.E. and  G.M. Renshaw. 2004. Hypoxia survival strategies in two fishes:
       extreme anoxia tolerance in the North European crucian carp and natural hypoxic
       preconditioning in a coral-reef shark. J.  Exp. Biol. 207: 3131-3139.

O'Hara, B. F., F.L. Walson, H.K.  Srere, H. Kumar,  S.W. Wiler, S.K. Welch, L. Bitting,
       H.C. Heller, and T.S. Kilduff. 1999. Gene expression in the brain across the
       hibernation cycle. J. Neurosci. 19: 3781-3790.

Palmer, J.L. and R.H. Abeles. 1979. The mechanism of action of S-
       adenosylhomocysteinase. J. Biol. Chem. 254: 1217-1226.
                                       149

-------
Pichavant, K., J. Person-Le-Ruyet, N. Le Bayon, A. Severe, A.L. Le Roux, L. Quemener,
       V. Maxime, G. Nonnotte, and G. Boeuf. 2000. Effects of hypoxia on growth and
       metabolism of juvenile turbot. Aquaculture. 188: 103-114.

Podrabsky, I.E. and S.C. Hand. 1999. The bioenergetics of embryonic diapause in an
       annual kiftifish Austrofundiilus limnaeus. J. Exp. Biol. 202: 2567-2580.

Secor, D.H. and I.E. Gunderson.  1998. Effects of hypoxia and temperature on survival,
       growth, and respiration of juvenile Atlantic sturgeon, Acipenser oxyrinchus. Fish.
       Bull. 96: 603-613.

Semenza, G.L. 2001. HIF-1 and mechanisms of hypoxia sensing. Curr. Opin. Cell Biol..
       13:  167-71.

Storey, K.B. and J.M. Storey. 1990. Metabolic rate depression and biochemical
       adaptation in anaerobiosis, hibernation and estivation.  Q. Rev. Biol. 65: 145-174.

Storey, K.B. and J.M. Storey. 2004. Metabolic rate depression in animals: transcriptional
       and translational controls. Biol. Rev. Camb. Philos. Soc. 79: 207-233.

Taylor, J.C. and J.M. Miller. 2001. Physiological performance of juvenile southern
       flounder, Paralichthys lethostigma, in chronic and episodic hypoxia. J. Exp. Mar.
       Biol. Ecol. 258:  195-214.

Tinton, S. and P. Buc-Calderon. 1995. Inhibition of protein synthesis induced by adenine
       nucleotides requires their metabolism into adenosine. Biochem. Pharmacol. 50:
       481-488.

Van Breukelen, F. and H.V. Carey. 2002. Ubiquitin conjugate dynamics in the gut and
       liver of hibernating ground squirrels. J. Comp. Physiol. 172B: 269-273.

Van Ginneken, V.J.T., P. van Caubergh, M. Nieveen, P. Balm, G. van den Thillart, and
       A. Addink. 1998. Influence of hypoxia exposure on the energy metabolism of
       common carp (Cyprimis Carpio L.). Neth. J. Zool. 48: 65-82.
                                       150

-------
   Studies of gene expression  in brain  of anoxic crucian
                                     carp

                                       by

    S. Ellefsen1, G.K. Sandvik, D.A. Steenhoff Hov, T. Kristensen and G.E. Nilsson

Introduction
       Vertebrates are obligated oxygen-consumers, and without proper oxygenation
ATP demands rapidly exceed ATP production, leading to cell death. However, there are
exceptions to this rule, and the crucian carp, Carassius carassius, survives months of
anoxia at temperatures close to 0 °C. This implies that it manages to cope with the
encountered energetic problems, through an increase in the rate of anaerobic ATP
production and/or a depression of the metabolic rate (hypometabolism). Indeed, both
processes appear to play a role in the anoxic survival of crucian carp (reviewed by
Nilsson, 2001).

       The vertebrate brain has a high rate of ATP use, most of which is associated with
the ion pumping needed to sustain ion gradients across the neurolemma. Thus, lowering
of ion-fluxes through ion-channels represent a potential way of reducing neuronal ATP
needs. Indeed, "channel-arrest" has been hypothesized to be an important mechanism for
energy conservation in brain tissue of anoxia-tolerant vertebrates such as crucian carp and
several species of freshwater turtles (Hochachka, 1988; Lutz etal.,  1985). Channel arrest
has been demonstrated in the anoxic turtle brain, which, among other adaptations, show
reduced NMDA receptor function (reviewed by Bickler and Buck,  1998). In this study
we assessed the role of channel arrest in anoxic crucian carp brain tissue by investigating
the mRNA expression of the glutamatergic AMP A- and NMDA receptors.

       AMPA receptors (AMPARs) are excitatory, ionotropic glutamate receptors that
usually have low Ca2+permeability. They are important mediators of excitatory
neurotransmission and play vital roles in synaptogenesis and synaptic plasticity (Tanaka
et al, 2000). Moreover, AMPARs are central in ischemic neuronal cell death, being
partially responsible for the excitotoxic events that follow ATP depletion (Arundine and
Tymianski, 2003). Until now, four AMPAR subunits have been characterized in
mammals, GluRl-4, while eight subunits have been characterized in Danio rerio
(zebrafish), GluRla,b-GluR4a,b. Functional AMPARs are tetrameric complexes
combining the different subunits, and the subunit composition is a major determinant of
receptor function. Numerous studies have reported changes in AMPAR function
mediated by changes in subunit composition, and such alterations are known to occur in
response to ischemic insults  (review by Tanaka et al., 2000).
1 Department of Molecular Biosciences, University of Oslo, 0316 Oslo, Norway
stian.ellefsen@imbv.uio.no
                                      151

-------
       NMDA receptors (NMDARs) are excitatory, ionotropic glutamate receptors that
have high Ca2+permeability. They are key players of synaptic plasticity (Collingridge et
al, 1983; Wenthold et al, 2003) and they are the main mediators of neuronal cell death
in ischemia (Arundine and Tymianski, 2003). So far, seven NMDAR subunits have been
characterized in mammals; NR1, NR2A-D and NR3A-B. Functional NMDARs are
believed to consist of two NR1 subunits and two NR2 subunits, and receptor diversity is
mainly decided by the NR2 subunit composition (Dingledine et al., 1999). One way of
studying properties of NMDARs in an organism would therefore be to study NR2 subunit
composition.

       We hypothesized that the properties of AMP A- and NMDA receptors change in
crucian carp brain tissue upon anoxia-exposure, and that these changes are caused by
alterations in the expression of the various receptor subunits. Here we report the cloning
of subunits of AMP A- (GluRla,b-GluR4a,b) and NMDA (NR1 and NR2A-NR2D)
receptors in crucian carp, and show their expression in normoxic vs. anoxic  brain tissue.
Four experimental groups were investigated; normoxia 7 days, anoxia 1 day, anoxia 7
days and anoxia 7 days/normoxia 7 days.
Results

       Judging from their expression, the composition of AMPARs and NMDARs seem
to be fairly stable during anoxia. Only GluR3a, NR1 and NR2C mRNA levels were
significantly changed, all being decreased (Table I).

Table I.  Anoxic mRNA levels of AMPAR subunits (GluRla,b-GluR4a,b) and NMDAR
         subunits (NR1, NR2A-NR2D) in crucian carp brain at 10 °C, standardized to
         normoxic levels. Significant changes are indicated by arrows, and were found
         for GluR3a (A7), NR1 (Al, A7 and A7R7) and NR2C (Al).
mRNA
GluRla
GluRlb
GluR2a
GluR2b
GluR3a
GluR3b
GluR4a
GluR4b
Anoxia 1 day (Al)
-
-
-
-
-
-
-
-
Anoxia 7 days (A7)
-
-
-
-
1
-
-
-
Anoxia 7 days; Reox
7 days (A7R7)
-
-
-
-
-
-
-
-
NR1
NR2A
NR2B
NR2C
NR2D
1
-
-
1
-
1
-
-
-
-
1
-
-
-
-
                                      152

-------
Discussion

       The subunit expression of AMPARs and NMDARs in crucian carp brain tissue
was strikingly well preserved during such a drastic insult as one week of anoxia. Thus,
the anoxic crucian carp appears to keep its glutamatergic system in a steady-state,
retaining its functionality. This tentative conclusion goes well with the suggestion that
crucian carp needs to retain its brain tissue ion conductance to be able to survive anoxia
in an active state (Nilsson, 2001). Thus, after finding no effects of anoxia on K+ and Ca2+
permeability in crucian carp brain slices it was postulated that channel arrest is not an
important component of the anoxic survival strategy utilized by the crucian carp (Nilsson,
2001). Having said this, based on the current knowledge it is not possible to fully exclude
channel-arrest as a mean of lowering neuronal energy expenditure in anoxic crucian carp
brain. There are at least two reasons for this. Firstly, proteins such as AMPARs and
NMDARs may be altered at the post-translational level rather than at the transcriptional
level (e.g. by controlling the state of phosphorylation). Such post-translational changes
have been shown to occur in NMDARs of anoxic turtle brain tissue (Bickler et al, 2000).
Secondly, electric activity and the formation of action potentials are influenced by
numerous factors other than properties of AMPARs and NMDARs. Major sites for ion
fluxes in brain are sodium channels, and how they are influenced by anoxia has so far has
not been examined in crucian carp brain.

       Still, the glutamate receptors did show some changes in expression during anoxia.
Interestingly, the mRNA level of the GluRSa AMPAR subunit was significantly lowered
after 7 days of anoxia (Table I). In mammals, AMPARs containing GluR3 subunits are
thought to be important in long-term depression (LTD) events of mammalian
hippocampal  neurons, and they are thought to target and stabilize AMPARs to synapses
(Meng et al,  2003). Moreover, in 2006 Satake et al. ascribed a role for GluR2/GluR3-
containing AMPARs in modulation of presynaptic function. They found such AMPARs
in presynaptic membranes of GAB Aergic interneurons, and by activating them they
managed to inhibit GABA-release into corresponding synapses. Speculatively, a lowering
of GluR3a in anoxic crucian carp may  serve to lower the presence of GluR2/GluR3a
AMPARs in presynaptic membranes. Eventually this may lead to increased GABAergic
activity, which in turn will have inhibitory effects on the electric activity of the brain.
Increased levels of extracellular GAB A is seen in the anoxic crucian carp brain, and is
thought to function as a metabolic depressant by reducing electric activity (Hylland and
Nilsson,  1999).

       We found lowered mRNA levels of the NR1 NMDAR subunit in the anoxic
groups (Table I). Intuitively, this could indicate a general lowering of functional
NMDARs in  anoxic crucian carp brain, since NR1 is thought to be an essential part of all
NMDAR complexes. However, NR1 is known to be expressed in a surplus stock, and the
level of functional NMDARs has been suggested to be decided by the availability of NR2
subunits (Wenthold et al., 2003). Thus, the  lowering of NR1 expression seen in anoxic
crucian carp brain does not necessary mean a reduced number of functional NMDARs.
                                       153

-------
References

Arundine, M. andM. Tymianski. 2003. Molecular mechanisms of calcium-dependent
        neurodegeneration in excitotoxicity. Cell Calcium 34: 325-337.

Bickler, P.E. and L.T. Buck. 1998. Adaptations of vertebrate neurons to hypoxia and
        anoxia: maintaining critical Ca2+ concentrations. J. Exp. Biol. 201: 1141-1152.

Bickler, P.E., P.H. Donohoe, and L.T. Buck. 2000. Hypoxia-induced silencing of NMDA
        receptors in turtle neurons. J. Neurosci. 20: 3522-3528.

Collingridge, G.L., SJ. Kehl, and H. McLennan. 1983. Excitatory amino acids in
        synaptic transmission in the Schaffer collateral-commissural pathway of the rat
        hippocampus. J. Physiol. 334: 33-46.

Dingledine, R., K. Borges, D. Bowie, and S.F. Traynelis. 1999. The glutamate receptor
        ion channels. Pharmacol. Rev. 51: 7-61.

Hochachka, P.W. 1988. Metabolic, channel-, and pump-coupled functions: constraints
        and compromises of coadaptions. Can. J. Zool. 66: 1015-1027.

Hylland, P. and G.E. Nilsson. 1999. Extracellular levels of amino acid neurotransmitters
        during anoxia and forced energy deficiency in crucian carp brain. Brain Res.
        823:49-58.

Lutz, P.L., M. Rosenthal, and TJ. Sick. 1985. Living without oxygen: turtle brain as a
        model for anaerobic metabolism. Mol. Physiol. 8: 411-425.

Meng, Y., Y. Zhang,  and Z. Jia. 2003. Synaptic transmission and plasticity in the absence
        of AMP A glutamate receptor GluR2 and GluR3. Neuron 39: 163-176.

Nilsson, G.E. 2001. Surviving anoxia with the brain turned on. News Physiol. Sci.16:
        217-221.

Satake, S., S.Y. Song, Q. Cao, H. Satoh, D.A. Rusakov, Y. Yanagawa, E.A. Ling, K.
        Imoto, and S. Konishi. 2006. Characterization of AMP A receptors targeted by
        the climbing fiber transmitter mediating presynaptic inhibition of GAB Aergic
        transmission at cerebellar interneuron-Purkinje cell synapses. J. Neurosci. 26:
        2278-2289.

Tanaka, H., S.Y. Grooms, M.V. Bennett, and R.S.  Zukin. 2000. The AMPAR subunit
        GluR2: still front and center-stage. Brain Res. 886: 190-207.

Wenthold, R.J., K. Prybylowski, S. Standley, N. Sans, and R.S. Petralia. 2003.
        Trafficking of NMDA receptors. Annu. Rev. Pharmacol. Toxicol. 43: 335-358.
                                       154

-------
                         Doing the impossible:

     Anoxic  cell division in  the crucian carp (Carassius

                                 carassius)

                                      By

                      C. S0rensen1, J.  Sollid, and G.E. Nilsson

       Vertebrates depend on an uninterrupted supply of oxygen to maintain energy
production. Some species, however, are able to survive severe hypoxia for several hours,
including many fishes of the Amazon, like the oscar cichlid (Astronotus ocellatus)
(Muuscze etal., 1998), and several coral reef fishes (Wise etal., 1998; Nilsson and
Nilsson, 2004). However,  the true master of hypoxic and anoxic survival appear to be the
crucian carp (Carassius carassius). It lives in lakes and ponds in northern Europe and
Asia, where it is able to survive under the ice cover in the winter with little or no oxygen
for several months (Holopainen, 1986).

       The crucian carp has solved the problem of living without oxygen in a very exotic
manner. When oxygen levels drop, the crucian carp up-regulates glycolysis (Storey,
1987),  but in order to avoid self-pollution by increased lactate levels, lactate is converted
to ethanol, which leaves the fish over the gills (Johnston and Bernard, 1983). In this way
the crucian carp is able to  maintain energy production in the absence of oxygen as long as
there is glycogen present in the liver (Nilsson, 1990). Still, glycolysis is much less
effective than aerobic respiration and releasing an energy-rich compound like ethanol is
energetically wasteful. It is therefore advantageous for the crucian carp to minimize the
energy consumption while in anoxia, something it does by reducing brain activity and
locomotor activity (Nilsson, 1992, Nilsson et al, 1993).

       It is also beneficial for the crucian carp to postpone the onset of anaerobic
respiration for as long as possible. The crucian carp does this by being very efficient at
taking up oxygen from its  environment: its haemoglobin has an extremely high oxygen
affinity (Sollid et al., 2005a), allowing it to take up oxygen at normal rates even if
ambient oxygen falls to 10% of air saturation (Sollid et al.,  2003).

       The high oxygen affinity of the haemoglobin might be partially responsible for
bringing about a peculiar gill morphology in the crucian carp: under normoxic conditions,
its gills completely lack protruding lamellae (Fig.  la) (Sollid etal., 2003). This is
exceptional as the lamellae are the primary site of gas exchange, making up most of the
respiratory  surface area offish gills. However, due to the high oxygen affinity of the
crucian carp haemoglobin, the fish is still able to respire normally with only a fraction of
the respiratory surface area of other fish. Exposing a minimal surface area to the exterior
1 Department of Molecular Biosceinces, University of Oslo, POBox 1041, NO-0316
Oslo, Norway. Christina.Sorensen@imbv.uio.no
                                      155

-------
milieu has the potential for limiting the uptake of pathogens and toxic substances, and
reducing energy costs associated with excessive water influx and ion loss, a challenge
faced by all freshwater fish (Sollid et al, 2003). Thus, in normoxic water this gill
morphology has probably several advantages.

       However, under hypoxic conditions, the crucian carp gills change morphology,
and protruding lamellae appear (Fig.  Ib-e), increasing the respiratory surface of the gills.
Interestingly, this change of morphology is completely reversible, and the lamellae
disappear again on re-exposure to normoxic conditions (Fig. If).
Figure 1. Scanning electron micrographs from the 2nd gill arch of crucian carp kept in
         normoxic or hypoxic water. In normoxia, the gill filaments have no protruding
         lamellae (a), but after 1, 3, 7 and 14 day of hypoxia exposure the lamellae
         gradually appear (b,c,d,e). After a further 7 days of recovery in normoxic
         water, the gills again lack protruding lamellae (f). Scale bar is 50 jim. Adapted
         from Sollid et al. (2003).

       Sollid et al. (2003) studied this phenomenon immunohistochemically in order to
reveal the underlying cellular mechanisms. They found that under normoxic conditions
the spaces between the gill lamellae were filled with a mass of cells they named the
interlamellar cell mass (ILCM). This cell mass would be reduced in size when fish were
being held in hypoxia (Fig. 2). Further, mitotic (S-phase) cells and apoptotic cells in the
ILCM were identified using immunohistochemistry for BrdU (Bromodeoxyuridine - an
externally introduced marker of DNA synthesis, and thus an S-phase marker) and
TUNEL staining (Fig. 2).  The number of cells in S-phase dropped significantly, and the
number of TUNEL positive apoptotic cells increased transiently after exposure to
hypoxia (Fig. 3). These effects combined appeared to bring about the change in gill
morphology.
                                       156

-------
                                                    c. 7 3aj js of hypoxia ,

Figure 2.  Light micrographs of gills stained for S-phase cells (BrdU) (a-c) and apoptotic
          cells (TUNEL) (d-f) in normoxia (a,d), 3 days of hypoxia (b,e) and 7 days of
          hypoxia (c,f). Arrows point out some of the positively stained cells. ILCM;
          interlamellar cell mass. Scale bar, 50 jim. Adapted from Sollid etal. (2003).

       An interesting point is that the cell proliferation did not drop to zero; 0.58% of the
cells were still actively dividing after 7 days of hypoxia. This is indeed surprising as the
animals had been kept at oxygen levels far below the point where mammalian cells are
known to halt cell cycle progression (Pettersen and Lindmo,  1983).
                                                a % Apoptotic cells
                                                o % Mitotic cells
                                                # Cells in the ILCM
                                 3                 7
                                 Days in hypoxia
14
Figure 3.  The percentage of apoptotic cells and S-phase cells (left y-axis) and the
          total cell number in a central cross-section of the interlamellar cell mass
          (ILCM; right y-axis) after hypoxia exposure. Values are means ± S.E.M.
          Values that are significantly different from previous or later time point are
          marked with an asterisk (P<0.01). From Sollid et al.  (2003).
                                        157

-------
       Sollid et al. (2005b) also studied the gill morphology of crucian carp exposed to
anoxia. Interestingly there was no difference in morphology between the gills of
normoxic fish and fish that had spent 7 days in an anoxic environment (Fig. 4). This does
make sense, as there is no oxygen to extract from the environment in this case, and
maintaining the ILCM saves energy on osmoregulation and ion transport.
Figure 4. Scanning micrograph of crucian carp gill filament in normoxia (a) and after 7
         days of anoxia exposure (b). There were no apparent differences in gill
         morphology between the two exposure groups. Scale bar = 50 jim. From Sollid
         etal. 2005b.
       Once again the gills were studied on a cellular level with immunohistochemistry
for BrdU. Proliferative activity was significantly down-regulated, but after 7 days
completely without oxygen, 5.0% of the ILCM cells were still going through S-phase
(Fig. 5) (this number is higher than that of the hypoxic experiment, but the anoxia
experiment was conducted at a slightly higher temperature, and a higher general
metabolism and cell proliferation rate is to be expected). The same study (Sollid et al,
2005b) also showed that cell proliferation continued in liver and intestine during anoxia.
        a, BrdU
Figure 5. Light micrographs of gills stained for BrdU from fish taken from normoxic (a)
         and 7 days of anoxia (b). Scale bars = 50 jim. Adapted from Sollid et al.
         (2005b).
                                       158

-------
       In mammals, hypoxia and anoxia have profound effects on cell cycle progression.
Exposure of mammalian cell cultures to an atmosphere containing < 4 ppm C>2, under
which respiration is inhibited (Froese, 1962) results in immediate arrest of cells in S-
phase. Several proteins are involved in the oxygen dependent regulation of the cell cycle,
but one of those with a direct effect on cell cycle progression is ribonucleotide reducase
(RNR). RNR is the enzyme responsible for converting the four standard ribonucleotides
to deoxyribonucleotides needed for DNA synthesis (Thelander and Reichard, 1979;
Eklund et al, 2001). This reaction is the rate-limiting step of DNA synthesis (Eriksson et
al,  1984; Engstrom et al., 1985), and thus inhibition of RNR halts progression through
the  S-phase. The vertebrate RNR consists of 4 subunits (0,2 $2) (Figure 6). The two |3-
subunits (R2), harbours a di-iron centre each, involved in formation of a tyrosyl radical
responsible for the reduction of ribonucleotides. The formation of this radical is an
oxygen dependent reaction, and in the absence of oxygen the radical has a half-life of 30-
60 minutes at room temperature (Chimploy et al., 2000; Nyholm et al., 1993). Thus, in
the  absence of oxygen, RNR activity is rapidly lost and DNA synthesis stops and cell
cycle progression is halted (Graff et al., 2004; Probst et al., 1984; Probst et al., 1988).
The levels of Rl  are stable throughout the cell cycle due to a half life  of more than 20
hours, but the R2 subunit shows S-phase specific expression and has a half life of only 3
hours, probably regulated by controlled degradation (Chabes and Thelander, 2000).
                           Activity
                          ATPdATP
                        Specificity
                        ATPdATP
                        tfTTPdGTP
<2 x 85.7 kD)
                                                     R2
                                                  {2 x 43.4 KD)
Figure 6. An overview of RNR showing the dimmer of the Rl (0,2) and R2 (02) subunits.
         In the R2 subunit the tyrosyl radicals are indicated. From Rei chard 1993.

       As DNA synthesis continues in anoxic crucian carp, the crucian carp RNR might
have bypassed the oxygen dependence of the tyrosyl radical formation. Sollid et al.
(2005b) partially cloned the crucian carp RNR R2 subunit and aligned it with the same
sequences from zebra fish (Danio rerio), Xenopus laevis and mouse (Mus musculus) (Fig.
7). There was a high degree of sequence homology, and all the amino acids involved in
coordinating and creating the tyrosyl radical, as identified in a 3D model of the mouse
RNR R2 were conserved in all four species compared. Thus, the three-dimensional
structure of the crucian carp RNR R2 is almost identical to its mammalian counterpart,
which is known to be oxygen dependent.
                                       159

-------
                                  139
                                  I
Figure 7. Amino acid sequences of the RNR R2 subunits of CC: crucian carp (Carassius
         carassius), DR: zebrafish (Danio rerio\ XL: African clawed frog (Xenopus
         laevis) and MM: mouse (Mus musculus). Amino acids holding the iron centre
         are highlighted in red. The radical is generated on Tyrl77 (yellow). Amino
         acids involved in transport of the radical to the active site on the Rl subunit are
         highlighted in blue, while amino acids involved in generating the entrance to
         and surrounding the radical site are highlighted in grey. From Sollid etal.
         (2005b).
       Sollid et al. (2005b) also quantified the amount of RNR R2 mRNA in crucian
carp gills in normoxia and anoxia, performing real time PCR, and found no significant
difference between the groups, even though the number of S-phase cells was reduced.
There does not appear to be any anoxia-induced regulation of R2 on the transcriptional
level.

       It thus appears that the crucian carp, in contrast to other investigated vertebrate
species, is able to maintain DNA synthesis in the absence of oxygen. Surprisingly, amino
acid residues, especially those involved in coordinating the tyrosyl radical, are identical
in crucian carp and mouse RNR R2. Also there  is no change in RNR R2 mRNA in
crucian carp exposed to anoxia for 7 days. There is a possibility that the crucian carp
RNR R2 is more effective in keeping  the tyrosyl radical stable. However, a difference in
stability from 3 hours in mammals at 25 °C to several days in crucian carp  at 10 °C is not
likely.  Still, since no structural or spectroscopic data are presently available for crucian
carp RNR, no direct measurements of the radical stability in this species exist.

       Another possibility is that the  crucian carp uses another RNR or R2 subunit for
ribonucleotide reduction either at all times or when exposed to anoxia. The cloning was
                                       160

-------
done based on sequences of RNRs of other species, and might not have captured
alternative RNR variants. The crucian carp genome is tetraploid, so it harbours multiple
versions of several proteins, where the different versions often have divergent properties.

       In conclusion, the crucian carp is capable of maintaining DNA synthesis and cell
proliferation in the absence of oxygen. How this is done, however, still remains a
mystery.

References

Chabes, A.  and L. Thelander. 2000. Controlled protein  degradation regulates
       ribonuleotide reductase activity in proliferating  mammalian cells during the
       normal cell cycle and  in response to DNA damage and replication blocks. J. Biol.
       Chem. 275: 17747-17753.

Chimploy, K., M.L.  Tassotto, and C.K. Mathews. 2000. Ribonucleotide reductase, a
       possible agent in deoxyribonucleotide pool asymmetries induced by hypoxia. J.
       Biol. Chem. 275: 39267-39271.

Eklund, H., U. Uhlin, M. Farnegardh, D.T. Logan, and  P. Nordlund. 2001. Structure and
       function of the radical enzyme ribonucleotide reductase. Prog. Biophys. Mol.
       Biol. 77: 177-268.

Engstrom, Y., S. Eriksson, I.  Jildevik, S. Skog, L. Thelander, andB. Tribukait. 1985. Cell
       Cycle-Dependent Expression of Mammalian Ribonucleotide Reductase -
       Differential Regulation of the 2 Subunits. J. Biol. Chem. 260: 9114-9116.

Eriksson, S., A.  Graslund, S.  Skog, L. Thelander,  andB. Tribukait. 1984. Cell Cycle-
       Dependent Regulation of Mammalian Ribonucleotide Reductase - the S-Phase-
       Correlated Increase in Subunit-M2 Is Regulated by Denovo Protein-Synthesis. J.
       Biol. Chem. 259: 1695-1700.

Froese, G. 1962. Respiration  of Ascites Tumour Cells at Low Oxygen Concentrations.
       Biochim. Biophys. Acta 57: 509-519.

Graff, P., J. Seim, O. Amellem, H. Arakawa, Y. Nakamura, K.K.  Andersson, T. Stokke,
       and E.O. Pettersen. 2004. Counteraction of pRb-dependent protection after
       extreme hypoxia by elevated ribonucleotide reductase. Cell Prolif 37: 367-383.

Holopainen, I.J., H. Hyvarinen, and J. Piironen. 1986. Anaerobic  Wintering of Crucian
       Carp (Carassius-Carassius L)  .2. Metabolic Products. Comp. Biochem. Physiol.
       83A(2): 239-242.

Johnston, LA. and L.M. Bernard. 1983. Utilization of the Ethanol Pathway in Carp
       Following Exposure to Anoxia. J. Exp. Biol. 104: 73-78.
                                       161

-------
Muusze, B., J. Marcon, G. van den Thillart, and V. Almeida-Val. 1998. Hypoxia
       tolerance of Amazon fish Respirometry and energy metabolism of the cichlid
       Astronotus ocellatus. Comp. Biochem. Physiol. 120A(1): 151-156.

Nilsson, G.E. 1990. Long-Term Anoxia in Crucian Carp - Changes in the Levels of
       Amino-Acid and Monoamine Neurotransmitters in the Brain, Catecholamines in
       Chromaffm Tissue, and Liver-Glycogen. J. Exp. Biol.  150: 295-320.

Nilsson, G.E. 1992. Evidence for a Role of Gaba in Metabolic Depression During Anoxia
       in Crucian Carp (Carassius-Carassius). J. Exp. Biol. 164: 243-259.

Nilsson, G.E. and S. Ostlund-Nilsson. 2004. Hypoxia in paradise: widespread hypoxia
       tolerance in coral reef fishes. Proc. R. Soc. Lond. Ser. B. 271:  S30-S33.

Nilsson, G.E., P. Rosen, and D. Johansson. 1993. Anoxic Depression of Spontaneous
       Locomotor-Activity in Crucian Carp Quantified by a Computerized Imaging
       Technique. J. Exp. Biol. 180: 153-162.

Nyholm, S., GJ. Mann, A.G. Johansson, R.J. Bergeron, A. Graslund,  and L.  Thelander.
       1993.  Role of Ribonucleotide Reductase  in Inhibition of Mammalian-Cell Growth
       by Potent Iron Chelators. J. Biol. Chem. 268:  26200-26205.

Pettersen, E.O. and T. Lindmo. 1983. Inhibition  of Cell-Cycle Progression by Acute
       Treatment with Various Degrees of Hypoxia - Modifications Induced by Low
       Concentrations of Misonidazole Present During Hypoxia. Br. J. Cancer 48: 809-
       817.

Probst, H., V. Gekeler, and E. Helftenbein. 1984. Oxygen Dependence of Nuclear-DNA
       Replication in Ehrlich Ascites-Cells. Exp. Cell Res.  154: 327-341.

Probst, H., H. Schiffer, V. Gekeler, H. Kienzlepfeilsticker, U.  Stropp, K.E. Stotzer, and I.
       Frenzelstotzer. 1988. Oxygen Dependent Regulation of DNA-Synthesis and
       Growth of Ehrlich Ascites Tumor-Cells Invitro and Invivo. Cancer Res. 48: 2053-
       2060.

Reichard, P. 1993. From RNA to DNA, why so many ribonucleotide reductases. Science
       260: 1773-1777.

Sollid, J., P. De Angelis, K. Gundersen, and G.E. Nilsson. 2003. Hypoxia induces
       adaptive and reversible gross morphological changes in crucian carp gills. J. Exp.
       Biol. 206(20): 3667-3673.

Sollid, J., R.E. Weber, and G.E. Nilsson. 2005a.  Temperature alters the respiratory
       surface area of crucian carp Carassius carassius and goldfish Carassius auratus. J.
       Exp. Biol. 208: 1109-1116.
                                       162

-------
Sollid, 1, A. Kjernsli, P.M. De Angelis, A.K. R0hr, and G.E. Nilsson. 2005b. Cell
      proliferation and gill morphology in anoxic crucian carp. Am. J. Physiol. 289:
      R1196-R1201

Storey, K.B. 1987. Tissue-Specific Controls on Carbohydrate Catabolism During Anoxia
      in Goldfish. Physiol. Zool. 60: 601-607.

Thelander, L. and P. Reichard. 1979. Reduction of Ribonucleotides. Annu. Rev.
      Biochem. 48: 133-158

Wise, G., J.M. Mulvey, and G.M.C. Renshaw. 1998. Hypoxia tolerance in the epaulette
      shark (Hemiscyllium ocellatum). J. Exp. Zool. 281: 1-5.
                                      163

-------
164

-------
   Hypoxia-tolerance in a tropical elasmobranch: does
  adenosine trigger a  multi-system protective response?
                                      by

                                G.M.C. Renshaw1.

       Life in a marine setting poses a particular set of physiological challenges to be
met and responded to, whether it is surviving on the reef or surviving man-made changes
to the environment. Few vertebrates can survive prolonged hypoxia or anoxia and most of
those studied evolved their tolerance at temperatures close to freezing when metabolic
rates are lower. A physiologically more severe challenge occurs when normoxia or
anoxia is encountered at normothermic temperatures. Activation of retaliatory or pre-
emptive protective mechanisms in response to a physiological stressor can confer cross
protection to other types of physiological stressors,  particularly if molecular chaperones
such as heat shock proteins are upregulated. So it is possible that some of the protective
mechanisms elicited in fish to prolong survival in hypoxia and anoxia may be also
switched on to prolong their survival in temporarily toxic environments. Consequently, a
comparative examination of the repertoire of physiological mechanisms that have
evolved to prolong survival may be useful not only  in providing a deeper understanding
of evolutionary processes but also help to develop conservation strategies to counter a
variety of environmental stressors such as eutrophication and pollution which provide
increasingly serious risks to fish populations.

       When O2 supply is diminished, high energy  purines such as ATP, ADP and AMP
can not be re-synthesised at a rate to match their usage with the result that increased
adenosine is formed from AMP or via the IMP and  inosine pathways in what Lutz et al.
(2003) term the "energetically compromised brain." The rising adenosine level can be
both friend  and foe. While a rise in adenosine can signal imminent destruction of tissue
via necrosis or apoptosis, with the most vulnerable tissues being the brain and heart, an
elevated level of adenosine can act as a switch to conserve energy. In hypoxia- and
anoxia-tolerant animals adenosine may reduce metabolic rate and neuronal activity, as
well as stimulate glycolysis to increase available energy which ultimately delays the
onset of tissue damage (Lutz etal., 2003). In the brain, the action of adenosine on its
receptor conserves neuronal energy because it clamps the resting membrane potential and
inhibits transmitter release making it less likely that the neuron will respond to or
generate an action potential resulting in decreased energy utilisation.
Hypoxia and Ischemia Research Unit, Heart Foundation Research Centre & School of
Physiotherapy  and Exercise Science, Griffith University, PMB  50  Gold Coast Mail
Centre, Queensland, 9726, Australia.
                                      165

-------
       Stimulation of adenosine receptors increases hypoxia and anoxia tolerance by
triggering metabolic depression in vertebrates that evolved their tolerance at 0°C (Lutz et
a/., 2003). Adenosine plays a key role in hypoxia and anoxia tolerant teleosts such as
Crucian carp (Carassus carrassus) that evolved their tolerance to diminished oxygen in
conjunction with 'over wintering' at 0°C, making it extremely difficult to disentangle
cytoprotective strategies from a metabolic shut-down triggered in response to freezing. It
is easier for an organism to tolerate anoxia when freezing temperatures have already
depressed energy consumption. Hypoxia and anoxia provide a greater challenge at
tropical temperatures; a few species offish in Brazil (Almeida-Val, 1995; Val and
Almeida-Val, 1995; Chippari-Gomez etal., 2005) and Australia (Renshaw and Dyson,
1999; Renshaw et al, 2002; Nilsson and Ostlund-Nilsson, 2004) have evolved
mechanisms to enable them to successfully cope with severe hypoxia and anoxia.

       Some reef platforms on the Great Barrier Reef in Australia  can be subject to
extreme fluctuations in dissolved oxygen levels. Heron Island reef platform (23°27'S,
151°55"E) is surrounded by a fringing reef and dissolved oxygen levels range from over
150% saturation at midday to 30% saturation on some nocturnal low tides (Kinsey and
Kinsey, 1967). During nocturnal low tides when the water on the reef platform is cut off
from the  surrounding ocean water by a fringing reef and  prevailing wind conditions do
not cause mixing and re-oxygenation of the surface waters, dissolved oxygen levels can
fall to 19% saturation (Renshaw, unpublished observations). This extreme environment
provides  cycles of nocturnal low tides that could potentially pre-condition its inhabitants
to hypoxia. The epaulette shark (Hemiscyllium ocellatum) is a nocturnal feeder that
successfully exploits this sheltered habitat. In previous studies we have shown that the
epaulette survives severe hypoxia, 0.39mg O2 I"1 for 2 hours, without delayed neuronal
apoptosis (Renshaw and Dyson, 1999) and at least one hour of anoxia (Renshaw et a/.,
2002) without a deleterious decrease in brain energy charge. More  recent experiments
have established that animals recover from 5-6 hours of anoxia (Chapman and Renshaw,
unpublished results). The metabolic and ventilatory depression that occurs in response to
hypoxia (Routley et al., 2002) and anoxia (Renshaw et al.,  2002) may serve to match
energy consumption to reduced ATP generation in the epaulette shark because blockade
of adenosine receptors with aminophylline, a non-specific adenosine antagonist, lowered
brain energy charge (Renshaw et a/., 2002). We demonstrated that  hypoxia- and anoxia-
tolerance could be increased in the epaulette by prior exposure to preconditioning
episodes  of sub-lethal hypoxic (Routley et a/., 2002) or anoxic exposure (Renshaw et a/.,
2002), a phenomenon first described in goldfish (Prosser et a/., 1957).

       This paper  examines  the  evidence  that  adenosine  switches  on  potentially
protective mechanisms in the  tropical hypoxia and anoxia tolerant epaulette shark and
makes an appreciable difference  in conserving  brain  energy  charge during  anoxia  at
tropical temperatures. In a series  of  separate studies, we focused on  the effect  of
adenosine on the control of: i) ventilation rate; ii) heart rate; iii) blood supply to the gills,
heart  and brain;  iv) neuronal activity and brain  energy charge; and v) the level  of
molecular chaperones. It is clear that adenosine is  capable of triggering a multi-system
response to diminished oxygen levels. However the full  extent of the protective extent of
adenosine, during exposure  to hypoxic  and  anoxic  conditions,  is  not completely
understood. In some but not all systems, adenosine appears to  be  a major retaliatory
                                       166

-------
molecule to pre-empt energy failure and increase natural repair systems both of which
would serve to reduce cell damage.

The effect of hypoxia and adenosine on ventilation rate

       To examine the role of adenosine in controlling heart rate and blood flow to the
heart, gills and brain during a hypoxic response, anaesthetised epaulette sharks were
exposed to either  hypoxic (0.34  mg Oi I"1) or normoxic conditions with and  without
pharmacological intervention.  Separate groups of unanaesthetised animals were exposed
to progressive hypoxia in a closed system respirometer or to sudden anoxia.

       Increasing the rate of buccal pumping serves to increase the speed and volume at
which water passes over the epithelium of the secondary lamellae, ultimately maintaining
the oxygen gradient between the water surrounding the gills and the blood (Randall and
Daxbroek, 1982). However during hypoxia and anoxia, this increased ventilatory activity
reduces the energy budget at a time when oxidative phosphorylation is limited due to
hypoxemia. In such environments, ventilatory depression is adaptive and prolongs
survival time. Examination of the changes in ventilatory  rate during progressive hypoxia
revealed that the [O2]crit of the epaulette shark occurred at 2.30 mg O21"1 and ventilatory
depression occurred between 1.5 and 1.0 mg O2 I"1. Furthermore, ventilatory depression
coincided with a significant elevation in lactate (Routley et a/., 2002). An increase in the
level of the anaerobic byproduct, lactate, suggests that dependence on anaerobic
pathways was increased as oxidative pathways slowed due to the diminished availability
of oxygen. Under such conditions, it is likely that adenosine levels rose and thus could be
implicated in triggering a decrease in the ventilation rate.
                     30-i A
                                **T
                     28 -
                     26 -
                     24
                     22 "
                     2O-
                     18 -
                     16 -
                     14
                       -1O
                     26 -
                     24 -
                     22 -
                     2O-
                     18 -
                     1 6 -
                     14 -
                     12 -
                     1O
                          B       _
                                    1O     2O    30
                                       Time (min)
                                                      4O
                                                            5O
                         -6O
                                    6O
                                         12O 2OO 3OO 4OO 5(K) 6OO
                                         Time (s)
  Figure 1. Effects of (A) hypoxia and (B) adenosine (1 umol kg" ) on ventilation
           frequency. The horizontal line indicates a significant time interval that
           differed from the last normoxic value; non-parametric ANOVA (Freidman
                                       167

-------
           test) with Dunn post-test. Values are mean + S.E.M. (Adapted from
           Stenslokken et al., 2004)

       A bi-phasic ventilatory response was observed during the exposure of
anaesthetised sharks to hypoxia (0.35 mg O2 I"1) (Fig. 1A). Initially, ventilatory
movements steadily increased and then significantly decreased, indicating that ventilatory
depression had occurred in response to hypoxia (Stenslokken et al., 2004).  While this
confirms previous findings in un-anaesthetized epaulette sharks (Routley etal., 2002), it
was noted that ventilatory movements were at a lower frequency in anaesthetised
animals.

       However, when we examined the  effect of adenosine administration (1 umol kg"1)
on anaesthetised animals, the ventilatory response elicited by adenosine administration
was the reverse of that observed in response to hypoxic exposure. Instead of an initial
increase in ventilation, there was an initial pronounced decrease followed by a delayed
increase (Fig. IB). This is the first indication of the involvement of adenosine in a
respiratory reflex in any fish (Stenslokken etal.,  2004).
       Furthermore, all ventilatory movement ceased when aminophylline  (30 mg kg"1)
was injected into anaesthetised sharks. This finding is in sharp contrast to the lack of
effect of aminophylline on the ventilation rate of unanaesthetised sharks exposed to
anoxia (Fig. 2). Both aminophylline (30 mg kg"1) and saline treated controls exhibited a
similar immediate and  significant decrease in ventilatory rate that was unaffected by
aminophylline (Renshaw etal., 2002).
        90
        80
        70
        60
        50
        40
        30
        20
--*•-• Anoxia + Saline
—•—Anoxia + Aminophylline
-•*- Normoxia
                       10
                                   20           30

                                       Duration (mins)
                                                             40
                                                                          50
Figure 2.  The ventilation rate of anaesthetised epaulette sharks held at normoxia or at
          anoxia after the prior administration of aminophylline (30 mg kg"1 in saline) or
          saline alone. After initial an initial increase in ventilation rate most likely due
          to capture stress, the steady decline in ventilation rate was similar for
                                        168

-------
         aminophylline and saline treated animals. (Kerrisk, Nilsson and Renshaw,
         unpublished results).
       The discrepancy between the effect of aminophylline on anaesthetised and
unanaesthetised  sharks may have occurred due to the interaction of the Benzocaine
anaesthetic, used in the study by Stenslokken et al. (2004), with GABA receptors present
in cardiorespiratory centres (Mulvey and Renshaw, unpublished). It is suggested here that
if GABA receptors are blocked by anaesthetic, then adenosine may be needed to maintain
respiratory drive. Furthermore, we have shown that there is an overall decrease in the
level of neuronal activity in the brainstem in response to hypoxic preconditioning
(Mulvey and Renshaw, 2000), and a significant concomitant increase in GABA in
hypometabolic brain nuclei (Mulvey and Renshaw, unpublished observations). More
specifically, hypometabolism and a concomitant increase in GABA were evident in an
important cardiorespiratory centre, the dorsal vagal nucleus. Since adenosine levels are
likely to increase during hypoxia, the potential synergism of adenosine and GABA in the
ventilatory control of these sharks during hypoxia needs further attention using specific
adenosine receptor subtypes to determine under what conditions adenosine plays a  role in
controlling ventilation.

The effect of hypoxia and adenosine on heart rate

       Hypoxia induced bradycardia in the epaulette shark (Soderstrom et a/.,  1999). The
heart rate decreased from 58.44 + 0.97 beats per min (bpm) to 39.03 + 2.61 bpm over a
15 minute period with a significant concomitant drop in blood pressure in the dorsal and
ventral aorta and an 84% decrease in erythrocyte velocity indicating a reduced cardiac
output (Stenslokken et a/.,  2004). In contrast, bradycardia was evident within 1 min of
hypoxic exposure in the non-hypoxia tolerant spiny dogfish (Scyliorhinus canicula)
(Taylor etal., 1977). The slow onset of hypoxia-induced bradycardia in the epaulette
shark suggests that there is a non-nervous component to the response (Stenslokken et al.,
2004). While aminophylline alone  did not block bradycardia in response to hypoxia,
neither did muscarinic antagonists revealing that this shark differs from other
elasmobranchs and teleosts studied  so far. However, the administration of adenosine (1
umol kg"1) during normoxia did mimic bradycardia (Fig. 3) and its action could be
blocked by aminophylline  (10 mg kg"1) indicating that at present, we cannot rule out a
cardioprotective role for adenosine  during hypoxia.  Further work is needed to clarify
whether specific adenosine receptors have a  cardioprotective role, during a hypoxic
exposure or subsequent reperfusion, or whether there is a specific time point at which a
short-term effect of adenosine could be observed.
                                       169

-------
                                  Ado
                                  Atlu after •mlDOphylHne
                             123
4    5    6    7   8   9
 (mill)
Figure 3. Adenosine administration (1 mmol kg "*), during normoxia in anaesthetised
         epaulette sharks, mimicked the effects of hypoxia. Heart rate fell significantly
         (PO.05) from 58.60 + 1.01 beats min "* to 48.87+4.49 beats min"1 and could be
         blocked by aminophylline (10 mg kg"1). Time dependant changes were tested
         using a repeated measures ANOVA with a Dunnet post-test. Lines indicate the
         time periods that differ significantly from the last normoxic value (P<0.05).
         (From Stenslokken etal, 2004).

The effect of hypoxia and adenosine on the regulation of blood supply to the gills,
heart and brain.

      In teleosts, adenosine is one of the key regulators of blood flow to the gills and
acts to constrict branchial circulation (Colin and Leray, 1981; Sundin and Nilsson, 1996).
Both hypoxic exposure and adenosine administration during normoxia caused
bradycardia and a fall in the blood pressure in the dorsal and ventral aorta (Stenslokken et
a/., 2004). Epi-illumination microscopy revealed that after 4-5 minutes, two parallel
longitudinal blood vessels opened. Figure 4 shows a video micrograph of the free tip of a
gill filament showing one of the longitudinal vessels, which opened when blood pressure
dropped during hypoxia. The direction of blood flow (black arrows) in the longitudinal
vessel (outlined in black) was toward the base of the filament. These longitudinal vessels
extended from  the filament base to its tip, where the free tip attached to the septum. The
white arrows indicate the position of anastomoses where the blood started flowing in
response to hypoxia or adenosine injections (1 umol kg "1). Such anastamoses are also
present in the spiny dogfish (Olsen and Kent, 1980) but not in the small spotted cat shark
                                       170

-------
(Laurent, 1984; Randall, 1985). After approximately 12-15 min of hypoxic exposure,
which is the approximate time course for the development of hypoxia-induced
bradycardia, blood flow stopped in the longitudinal vessels even though the openings
were still visible at high magnification. While the control mechanisms to open these
anastomoses are not known, it was suggested that they could be under adenosinergic
control (Stenslokken etal., 2004).
Figure 4. The longitudinal vessel is outlined in black with the direction of blood flow
         indicated by black arrows and was toward the base of the filament. These
         longitudinal vessels extended from the filament base to its tip, where the free tip
         attaches to the septum. White arrows indicate the position of anastomoses where
         the blood started flowing in response to hypoxia or adenosine injections (lumol
         kg"1). (From Stenslokken etal, 2004).

       Since blood in the longitudinal vessels drains directly into the heart via the
arterio-venous circulation, the initial effect of hypoxia or adenosine in recruiting these
normally collapsed longitudinal vessels could be expected to provide a short term
increase in the blood supply to the heart and protect it before hypoxia- and anoxia-
induced bradycardia spare cardiac energy consumption (Stenslokken etal., 2004). This
indicates a potential role for adenosine in achieving  hypoxia and anoxia tolerance in this
successfully tolerant species.

       While  blood  pressure dropped  by 50 % in  the epaulette shark during  severe
hypoxia  (0.35 mgO2 I"1  at 24°C), cerebral  blood flow, measured by  epi-illumination
microscopy, was unaffected indicating that compensatory vasodilation or autoregulation
was switched on in response to hypoxia (Soderstrom  et al., 1999). During normoxia, we
observed  an increase in cerebral  blood  flow velocity  when the normoxic brain was
superfused with adenosine and this increase  could  be blocked with aminophylline. This
finding reveals that sharks  may be the oldest vertebrate group in which adenosinergic
control of cerebral blood flow occurs.  However, during hypoxia neither adenosine  nor
aminophylline had an effect upon the maintenance of cerebral blood flow (Soderstrom et
al., 1999). There was no evidence of an adenosine-mediated increase in cerebral blood
flow in the epaulette shark that corresponded to that observed in freshwater turtles and
cyprinid  fish (Lutz  et al., 2003),  so our present results  are in contrast to the effect of
                                       171

-------
adenosine in other hypoxia- and anoxia-tolerant species that evolved their tolerance at
temperatures close to freezing. Further investigation of the key neurotransmitters and
neuromodulators involved in maintaining cerebral blood flow is needed with particular
attention paid to more specific antagonists of adenosine receptors.

The effect of hypoxia and adenosine on the maintenance of brain energy charge.

       To examine the role of adenosine in maintaining brain energy charge in anoxia-
preconditioned sharks during an anoxic challenge, unanaesthetised epaulette sharks were
placed into sudden anoxia (< 0.02 mg O2 I"1) until they lost their righting reflex, this
constituted the preconditioning phase (episode  1, El). The time to loss of righting reflex
was measured and used as the end point for episode 1.  Sharks were returned to a
normoxic holding pool and allowed to recover in it for 24 hours. Sharks were pair
matched for  endurance time and 15  min prior to a second anoxic challenge (episode 2,
E2), one shark from each pair was injected IP with aminophylline (30 mg kg"1 in saline)
and the other with saline alone, and then the time to loss  of righting reflex was measured.

       The mean time to loss of righting reflex in the first episode (El) of anoxia was 47
minutes. The mean time to loss of righting reflex varied significantly with treatment and
Figure 5 shows the percentage change in time to loss of righting reflex in E2 in relation to
the mean time to loss of righting reflex in El. Saline treated sharks lost responsiveness
and their righting reflex 66% faster  in E2 than El (P<0.001).  While aminophylline
treated sharks remained alert and retained their righting reflex 143% longer than they had
in El  (PO.001).
a 150
I
^ *I/Y"I _
S3 1OU
^c
i
1= en .
Time to loss of righting re
to E1 (%)
8 & o I
Saline Arrinophylline
treated treated

n



n
	 1



MM1
™ • •
S1 S2 S3 S4 S5 A1
Animal number
A2
A3 A4 A5
Figure 5. The time to loss of righting reflex for 5 pairs of sharks in episode 2 (E2). Each
         histogram represents the percentage of time to loss of righting reflex in E2
         relative to episode 1 (El) shown as the baseline.  Sharks were pair matched for
         their time to loss of righting reflex in El, allowed to recover in a normoxic
         holding tank then 24 hours later one animal from each pair was injected with
         aminophylline (30 mg kg"1 in saline) and the other with saline alone. Animals
         were moved to sudden anoxia 15 minutes after injection and the time to loss of
         right reflex was measured.  (Adapted from  Renshaw et a/., 2002).
                                       172

-------
       Since the righting reflex is under cerebellar control it is likely that the loss of the
righting reflex indicated the time at which neuronal activity in the cerebellum was
significantly reduced. This appears to have occurred more rapidly when adenosine was
allowed to act at its receptor in saline treated sharks than when aminophylline blockade
was present. Decreased neuronal activity is an energy sparing adaptive mechanism that
appears to be under adenosinergic control in the epaulette shark and mirrors the
adaptation of other anoxia tolerant vertebrates (reviewed by Lutz et a/., 2003).

       In a parallel set of experiments, the time in episode 2 was for a set period of 50
minutes and a group of untreated sharks were held at normoxia for corresponding periods
of time in El and E2. At the conclusion of this experiment, anoxic animals, their pair
matched controls and untreated normoxic sharks were anaesthetised and the brainstems
were rapidly frozen and then used to measure energy charge and adenosine levels (via
HPLC) or to measure molecular chaperone Heat shock protein 70 (Hsp70) levels (via
semi-quantitative western blotting).

       Adenosine levels were 3.5 fold higher in the animals exposed to 50 minutes of
anoxia in E2 than in controls exposed to normoxia for 50 minutes demonstrating that the
level of the anoxic challenge was sufficient to deplete high-energy purines.
Measurements of brain energy charge (Fig. 6) showed that while the energy charge in
both saline and aminophylline treated brains was significantly lower than in untreated
animals exposed to anoxia  for the same period of time, the greatest decrease occurred
when aminophylline blockade was in place. The mean energy charge of anoxia treated
sharks was 0.736 and of aminophylline treated sharks was 0.687 + 0.008 which was
significantly lower than untreated sharks held at normoxia (P<0.001) and saline treated
controls exposed to anoxia (P<0.05). These results revealed that the delayed time to loss
of righting reflex detrimentally affected the energy budget and that adenosine has an
effect in suppressing neuronal metabolism to conserve energy charge. Therefore, the loss
of the righting reflex in this species, may be an energy conserving strategy.
                    Energy Charge in the Epaulette Shark Brain
Anoxia + Saline      Anoxia
                  Normoxia
                                  Experimental
Figure 6. The mean energy charge of brains from epaulette sharks exposed to two
         episodes of normoxia or anoxia 24 hours apart. The mean energy charge of
         aminophylline treated sharks was 0.687+0.008 which was significantly lower
                                       173

-------
        than untreated sharks held at normoxia (P<0.001) and saline treated controls
        exposed to anoxia (P<0.05). (From Renshaw et a/., 2002)

The effect of hypoxia and adenosine on the level of a neuroprotective molecular
chaperone: heat shock protein 70

       Changes in gene expression are associated with switching to a protected
phenotype in response to environmental and/or physiological stress. Ubiquitous
molecular chaperones from the heat shock protein superfamily confer neuronal protection
that can be blocked by anti Hsp70 antibodies (Nakata et a/., 1993). Activation of the
Hsp70 promoter responds to negative cellular energy balance (Kiang and Tsokos,  1998)
and oxidative stress (Das etal., 1995).

       We examined the effect of anoxia and aminophylline treatment of the level of
Hsp70, using the methods  described above. Figure 7 shows that while the constitutive
level of Hsp70 in untreated controls was low, the level of Hsp70 was significantly higher
in the brain of anoxia treated epaulette compared to untreated controls (P<0.005)
(Renshaw et a/., 2004). When aminophylline was administered 15 minutes prior to the
second episode of anoxia, the level of Hsp70 was significantly higher than  in anoxia
treated animals (P<0.01) and represented a 7.4 fold increase above the mean constitutive
level of Hsp70 in untreated controls (Renshaw et al.,  2004). This indicates  that Hsp70
may also act as an energy sensor and a useful marker of reduced energy charge.
   O  300000
   C
       200000
   O  100000
   (0

                     Hsp70 levels after anoxic exposure
                  C1   A1   S1   C2   A2   S2
A3   S3   C4   A4   S4
Figure 7. The level of Hspyo in the brain of epaulette sharks after exposure to normoxia
        (series C) or two episodes of normoxia or anoxia 24 hours apart. Sharks were
        injected with aminophylline in saline (series A) or saline alone (series S) 15
        minutes prior to the second anoxic exposure, Adapted from Renshaw et a/.,
        2004).
                                      174

-------
Conclusions

       While this series of studies has demonstrated that elevated adenosine levels can
initiate protective responses such as ventilatory depression, bradycardia, increased
cerebral blood flow and increased blood flow to the heart, the regulatory role of elevated
adenosine levels during hypoxia needs further clarification. The use of selective
adenosine antagonists prior to hypoxic exposure is expected to provide a better
understanding of the retaliatory and pre-emptive effects of adenosine.

       There is compelling evidence that elevated adenosine levels, via acting on its
receptor in the brain, resulted in the temporary loss of cerebellar responsiveness because
the righting reflex was lost 66% earlier after anoxic-preconditioning. The loss of the
righting reflex in response to anoxia appears to have a neuroprotective functional
correlate because the administration of aminophylline prolonged the time to loss of the
righting reflex with a significant decrease in brain energy charge. Furthermore, the
significantly higher levels of neuroprotective Hsp70 in aminophylline treated animals,
which had significantly lower brain energy charge than controls, indicate that increased
cellular stress occurred when adenosine receptors were blocked. Taken together these
results reveal that elevated adenosine associated with anoxic exposure did provide a pre-
emptive state of metabolic depression, which served to conserve ATP in this successfully
hypoxia and anoxia tolerant species.

References

Almeida-Val,  V.M.F.,  IP. Farias, M.N.P. Silva, W.P. Duncan, and A.L. Val. 1995.
       Biochemical adjustments  to hypoxia by amazon cichlids. Braz. J. Med. Biol. Res.
       28(11-12): 1257-1263.
Chippari-Gomez, A.R., L.C. Gomes, N.P. Lopes, A.L. Val, and V.M.F. Almeida-Val.
       2005. Metabolic adjustments in two Amazonian cichlids exposed to hypoxia and
       anoxia. Comp. Biochem. Physiol. 141B(3): 347-55.
Colin, D.A. and C. Leray. 1981. Vasoactivities of adenosine analogues in trout gill
       (Salmo gardneiri R.). Biochem. Pharmacol. 30: 2971-2977.
Das, D.K., N. Haulik, and I.I. Moraru. 1995. Gene expression in acute myocardial stress.
       Induction by hypoxia, ishemia, reperfusion, hyperthermia and oxidative stress.  J.
       Mol. Cell. Cardiol. 27:181-193.
Kiang, J.G. and G.C.  Tsokos.  1998. Heat shock protein 70 kDa: Molecular biology,
       biochemistry and physiology. Pharmacol. Therapeutics 80: 183-201.
Kinsey, D.W. and B.E.  Kinsey. 1967.  Diurnal changes in oxygen content of the water over
       the coral reef platform at Heron I, Aust. J. Mar. Freshwater. Res.  18:23-34.
Laurent, P. 1984. Gill internal morphology. Pages 73-172 In Fish Physiology, Vol. 10.
       W.S. Hoar and DJ. Randall (Eds.). Academic Press, Orlando
                                       175

-------
Lutz P.L., G.E. Nilson, and H.M. Prentice. 2003. The brain without oxygen: causes of
       failure -physiological and molecular mechanisms for survival. Kluwer Academic
       Publishers, Dordrecht, Boston, London..Mulvey J., and G.M.C. Renshaw. 2000.
       Neuronal oxidative hypometabolism in the brainstem of the epaulette shark
       (Hemiscyllium ocellatum) in response to hypoxic pre-conditioning. Neurosci.
       Lett. 290: 1-4.
Nakata, N., H. Kato, and K.  Kogure. 1993. Inhibition of ischemic tolerance in the gerbil
       hippocampus quercetin and anti-heat shock protein-70 antibody. Neuroreport
       4:695-8.
Nilsson, G.E. and S. Ostlund-Nilsson.  2004. Hypoxia in paradise: widespread hypoxia
       tolerance in coral reef fishes. Proc. Royal Soc. Lond. Suppl. 271: S30-S33.
Olsen, K.R. and B.  Kent. 1980. The microvasculature of the elasmobranch gill. Cell
       Tissue Res.  209:49-63.
Prosser, C.L., L.M. Barr, R.D. Pine, and C.Y. Lauer. 1957. Acclimation of goldfish to
       low concentrations of oxygen.  Physiol. Zool. 30: 137-141.
Randall, DJ. 1985. Shunts in fish gills. In Cardiovascular shunts: Phylogenetic,
       Ontogenetic and Clinical Aspects. Alfred Benson Symposium Vol. 21. K.
       Johansen and W. Burggren (Eds.). Munksgaard, Copenhagen.
Randall, DJ. and C. Daxboeck. 1982.  Cardiovascular changes in the rainbow trout
       (Salmo gairdneri Richardson) during exercise. Can. J. Zool. 60:  1135-1140.
Renshaw, G.M.C. and S.E. Dyson. 1999. Increased nitric oxide synthase in the
       vasculature  of the epaulette shark brain following hypoxia. Neuroreport 10: 1-6.
Renshaw, G.M.C., C.B. Kerrisk, and G.E. Nilsson. 2002. The role of adenosine in the
       anoxic survival of the epaulette  shark, Hemiscyllium ocellatum. Comp. Biochem.
       Physiol. 131B: 133-141.
Renshaw, G.M.C., J. Warburton, and A. Girjes. 2004. Oxygen  sensors and energy sensors
       ct synergistically to achieve  a graded alteration  in gene  expression: Consequences
       for assessing the level of neuroprotection in response to stressors. Frontiers in
       Bioscience9:  110-116.
Routley, M.H., G.E. Nilsson, and G.M.C. Renshaw. 2002. Exposure to hypoxia primes
       the respiratory and metabolic responses of the epaulette shark to progressive
       hypoxia. Comp. Biochem. Physiol 131 A: 313-321.
Soderstrom, V., G.M.C. Renshaw, and  G.E. Nilsson. 1999. Brain blood flow and blood
       pressure during hypoxia in the epaulette shark {Hemiscyllium ocellatum). J. Exp.
       Biol. 202: 829-835.
Stensl0kken, K-O., L. Sundin, G.M.C. Renshaw, and G.E. Nilsson. 2004. Adenosinergic
       and cholinergic control mechanisms during hypoxia in the epaulette shark
       (Hemiscyllium ocellatum), with emphasis on branchial circulation. J. Exp. Biol.
       207:4451-4461.
Sundin, L. and G.E. Nilsson. 1996. Branchial and systemic roles of adenosine receptors
       in rainbow trout:  an in vivo microscopy study. Am. J. Physiol. 271: R666-R669.
                                       176

-------
Taylor, E.W., S. Short, and PJ. Butler. 1977. The role of the cardiac vagus in the
       response of the dogfish Scyliorhinus canicula to hypoxia. J. Exp. Biol. 70: 57-75.

Val, A.L., V.M.F. Almeida-Val. 1995. Fishes of the Amazon and their environment.
       Springer, Berlin, Heidelberg.
                                       177

-------
178

-------
     Behavioural, respiratory, ionoregulatory, and N-

 metabolic adaptations to  low environmental O2, and the

        influence of body size in the hypoxia-tolerant

           Amazonian oscar (Astronotus ocellatus )

                                     by

 C.M. Wood1'2, K.A. Sloman3, M. Kajimura1, O.E. Johannsson4, PJ. Walsh2, G. Scott5,
                  S. Wood1, V.M.F. Almeida-Val6 and A.L. Val6.

Introduction

      The Brazilian Amazon, a "giant piece of amphibian land" (Val and Almeida-Val,
1995), has been an evolutionary testing ground for strategies of hypoxia-tolerance in fish.
Due to the dramatic annual cycles in water level, fish entering the flooded jungle
("igapo") and grasslands ("varzea") to feed and reproduce encounter environments that
fluctuate greatly in oxygen levels due to photosynthesis of algae  and respiration of
submerged organic matter (Fink and Fink, 1979; Junk et a/., 1983; Val and Almeida-Val,
1995). As the water recedes, many become trapped and have to migrate over land back
to the main river, activities facilitated by the evolution of a variety of air-breathing organs
and locomotory strategies (Graham, 1997). Others survive by metabolic and behavioural
adaptations that allow them to cope with severely hypoxic conditions without having to
actually enter the aerial environment. These include down-regulation of metabolic rate,
anaerobic metabolism, seeking out less hypoxic environments, periodic bouts of aquatic
surface respiration (ASR, "skimming") to exploit the more oxygen-rich surface film
(Kramer and Mehegan, 1981; Kramer and McClure,  1982; Kramer, 1987), and changes in
social and foraging behaviour (Val and Almeida-Val, 1995). These adaptations may be
viewed as compromises to maximize fitness at times of adversity, as they may increase
susceptibility to predation and/or decrease feeding or reproductive success, while at the
same time increasing the chances of surviving oxygen lack so  as to be able to grow and
reproduce in the future. One such species is the oscar (Acara-a9u; Astronotus ocellatus;
Cichlidae) which is renowned for its hypoxia tolerance; adults are reported to survive up
to 6h of complete anoxia, and can tolerate levels of 5-20 % air saturation for 20 -50 h
(Muusze etal., 1998; Almeida-Val et a/., 2000; J.G.  Richards, pers. comm.). We have
1 Department of Biology, McMaster University, Hamilton, Ontario, Canada, L8S 4K1
2 Rosenstiel School of Marine and Atmospheric Sciences, University of Miami, Florida
33149, U.S.A.
3 School of Biological Sciences, University of Plymouth, Devon, U.K., PL4 8AA
4 Dept. of Fisheries and Oceans, Canada Centre for Inland Waters, Burlington, Ontario,
Canada L7R 3 A6
5 Dept. of Zoology, University of British Columbia, Vancouver, B.C., Canada V6T 1Z4
6 Laboratory of Ecophysiology and Molecular Evolution, INPA, Manaus, Brazil
                                    179

-------
used this species to investigate a number of different aspects of physiology and behaviour
in response to environmental hypoxia.

       Throughout these studies, a particular theme has been the influence of body size
on hypoxia tolerance. Within a number of teleost species, smaller individuals are more
tolerant of hypoxia than larger individuals (Smale and Rabeni, 1995; Burleson et al.,
2001; Robb and Abrahams, 2003). This observation has given rise to the idea that
hypoxia can serve as an ecological refuge where smaller individuals can avoid larger
predatory fish which are less tolerant of low oxygen levels (Kolar and Rahel, 1993;
Chapman et a/., 1996; Robb and Abrahams,  2003). However, this may not always be the
case, and for the oscar, exactly the opposite may be true. Almeida-Val et al. (2000)
reported that small oscars (-16 g) survived fairly severe hypoxia (-30 torr) for only
about 9 h whereas larger individuals (-230 g) survived under identical conditions for
approximately 35 h. The greater anaerobic potential of larger fish was indicated by higher
concentrations of lactate dehydrogenase and malate dehydrogenase in a variety of tissues
(Almeida-Val et a/., 2000). Thus a scaling effect on hypoxia tolerance has been proposed
where larger oscars are better physiologically equipped for coping with hypoxic
conditions. We therefore hypothesized that larger individuals would be able to preserve
various aspects of physiological  homeostasis for a longer period and/or down to a lower
PO2 threshold than smaller individuals. Similarly, we hypothesized that larger oscars
would postpone "risky" behaviours such as ASR or leaving shelter to search for better
oxygenated waters to a similar degree.

       A second theme has been the influence of hypoxia on two key aspects of gill
function - ionoregulation and nitrogenous waste excretion. To date, impacts of hypoxia
on these processes have received little attention, but there are several reasons for
believing that these functions may be particularly sensitive to low environmental oxygen.

       Firstly, there is a well-documented respiratory-osmoregulatory compromise at the
gills, such that effective gill area and diffusion distance are adjusted to provide the
permeability required for gas exchange, while minimizing the permeability for diffusive
ion losses and osmotic water gain (Wood and Randall, 1973a,b; Gonzalez and
McDonald, 1992). It is probable that lamellar recruitment and gill vasodilation would
occur during hypoxia to help sustain oxygen uptake (Holeton and Randall, 1967a,b). We
therefore hypothesized that increases in ion efflux rates to the water,  and a dilution of
plasma ion levels would also occur.

       Secondly, ionoregulation is a costly process in fish, with estimates generally
falling in the range of 2-20% of resting metabolism at the whole animal level (reviewed
by Febry and Lutz, 1987).  The very dilute nature of many Amazonian waters ("slightly
contaminated distilled water" - Sioli, 1984) may exacerbate these costs,  and a general
tendency for reduced ion levels in the plasma of Amazonian teleosts has been noted
(Mangum et al.,  1978). At the cellular level, it is generally accepted that the cost of ion-
pumping is second only to that of protein synthesis, and both  processes are markedly
turned down during hypoxia in model species such as the turtle and crucian carp which
are capable of severe hypometabolism (reviewed by Hochachka and Lutz, 2001;
                                       180

-------
Boutilier, 2001). This is accomplished by both channel arrest and down-regulation of
Na+,K+-ATPase activity. Furthermore, gill ionocytes which are directly exposed to the
external water will be on the front line of hypoxia exposure and the very first cells to
experience the oxygen deficit. Studies with a water-ventilated and saline-perfused trout
gill preparation (Wood et a/., 1978) demonstrated that gill tissue normally acquires about
50% of its oxygen uptake directly from the external water, and the other 50% from the
perfusate, but it must be remembered that this perfusate has itself just been oxygenated in
the arterial-arterial pathway of the gill lamellae before perfusing the filaments where the
majority of the ionocytes are located (Olson, 2002; Wilson and Laurent, 2002).
Therefore, it seems likely that ionocytes will receive little benefit from internal oxygen
stores in the bloodstream during hypoxia. In total, the rate of oxygen utilization by
perfused and ventilated gill tissue amounts to about 4 -12% of resting oxygen uptake by
the whole animal (Wood et al., 1978; Lyndon, 1994; Morgan and Iwama,  1999) - i.e. in
the same range as estimates of the costs of ionoregulation. Therefore, we hypothesized
that active ion influx rates from the water would fall during hypoxia, perhaps in a two-
step process, the first in response to oxygen starvation of the working ionocytes, and the
second as a result of down-regulation of uptake channels and Na+,K+-ATPase activity to
save metabolic costs.

       Lastly, the gills excrete more than 80% of the metabolic ammonia production in
fish and the mechanism is thought to be linked in some way to the active uptake of Na+
(discussed by Wood, 1993; Wilson, 1996; and Wilkie, 1997, 2002). Original ideas about
direct Na+/NH4+ exchange coupling in the gill cells (e.g. Krogh, 1939; Maetz and Garcia-
Romeu, 1964; Wright and Wood, 1985), while not entirely disproven, have given way in
recent years to the concept that the coupling is indirect (Avella and Bornancin, 1989; Lin
and Randall, 1995).  Thus an H+ pump on the apical membrane provides the electrical
gradient needed to drive Na+ uptake from the water through coupled Na+-channels, and
the associated acidification of the gill boundary layer enhances the "diffusion-trapping"
of NHa as NFLt+, thus sustaining the PNHa gradient for diffusive NHa efflux (Wilson et
a/., 1994; Clarke and Potts, 1998).  If hypoxia interferes with the Na+ uptake-H+ pumping
mechanism (or Na+/NH4+ exchange),  ammonia excretion may be inhibited. At the same
time however, the rate of ammonia production may be reduced by down-regulation of
aerobic metabolism  (i.e. less deamination of amino acids for fuel; Kutty, 1972; van
Waarde, 1983). Therefore we hypothesized that ammonia excretion would fall during
hypoxia and become uncoupled from Na+ uptake, and that changes in plasma ammonia
levels would indicate which of the excretion or the production processes were impaired to
the greater extent.

       The present paper provides an overview of the experiments performed on oscar to
test these hypotheses, some of which were confirmed, while others were disproven or
modified as outlined below. The studies are reported in detail in Sloman et al. (2006) and
Wood et al. (2007),  and all methodological information is given in these two papers.
When reference is made to "small" and "large" oscar, the animals were about 20 g and
200 g respectively. Experimental temperature was 28 + 1.5°C, and all tests were
performed in very dilute, very soft water taken from  a well on the INPA campus in
Manaus. The water is typical of the Amazon region, with the following composition: Na+
                                       181

-------
                        ,2+
= 19, Cr = 21, K = 16, Ca/+ =11, Mg/+ = 2 umol.L , pH = 6.5, dissolved organic
                   -i
carbon = 0.6 mg C L" .

Respiratory Responses to Hypoxia - the Influence of Body Size
       Under normoxic conditions, routine mass-specific MC>2 was more than twice as
high in small oscars than large oscars (Fig.  1). However these rates were only about half
of those recorded for teleosts in general at comparable size and temperature (Clarke and
Johnston, 1999), indicating that low metabolic rate in itself may serve as a general
adaptation to frequently hypoxic conditions in this species. When environmental PC>2
was progressively reduced in a closed system, large fish exhibited no change in MO2 until
PC>2 fell to a threshold of approximately 50 torr. Similarly,  Muusze et al. (1998) working
with adult oscars of unspecified size found  a threshold of about 30 torr. In contrast, small
oscars showed no clear threshold with MO2 falling progressively with PC>2 right from
normoxia, though the first statistically significant decrease  in MO2 occurred at about 70
torr (Fig. 1). By 10 torr, MO2 had fallen to values of about 20% and 10% of normoxic
rates in large and small fish respectively. Thus in accord with our initial hypothesis,
larger fish were able to maintain 02- independent respiration (Hughes, 1981) down to a
lower threshold PO2 than  smaller fish, and support it to a higher degree below that point.
Additional fish were available to extend the mass range from 9g to 308g, and allowed the
derivation of a three-dimensional model, predicting MO2 (|J,mol C>2 h"1) as a function of
body mass (kg) and environmental PC>2 (torr) at 28°C:

Log MO2 = 3.20686 + 0.786*log(mass)-0.88913*0.96371p°2 (0.05 - 0.3 kg) (r2 = 0.948)
Log MO2 = 3.28076 + 0.786*log(mass)- 0.92331*0.97024
                               62 Consumption Rate
                                                     PO2
(0.01-0.05 kg) (r = 0.928)
6000 -
5000 -

4000 -


3000 -
2000 -
1000 -
n -

1

















|

I








\ 	 \ SmaU Fish
_,_ v//\ Large Fish





\

\





T




\






*



]

T



*


1

JL




1





*
HL
\





it
T
                        160-140 140-120 120-100 100-80  80-60 60-40  40-20  20-0

                                        PO2 (torr)
   Figure 1.   The influence of water PC>2 on MO2 in large versus small oscars. (Data
              from Sloman et al., 2006).
                                       182

-------
       Thus not only do large oscars have a much greater anaerobic capacity and
survival time during severe hypoxia (Almeida-Val et a/., 2000), they also have a greater
ability to maintain MO2 under more moderate hypoxia, thus avoiding or postponing the
need to resort to more costly or dangerous metabolic or behavioural survival strategies.
Unlike many other species (e.g. yellow perch, Percaflavescens, Robb and Abrahams,
2003; largemouth bass, Micropterus salmoides; Burleson et a/., 2001), Astronotus
ocellatus shows a positive relationship between physiological tolerance of hypoxia and
mass.
Behavioural Responses to Hypoxia - the Influence of Body Size

       In light of the above findings, it was initially surprising that when progressive
hypoxia trials were run in an apparatus that allowed the fish to voluntarily move to the
surface to exploit ASR, individual fish mass was positively correlated, and not negatively
correlated with the PO2 at which this behaviour first occurred (p=0.04), in contrast to one
of our original hypotheses. Indeed, on average, this behaviour was postponed to a much
lower environmental PC>2 (22.3 ±3.7 torr) in small oscars than large ones (49.6 ± 9.8
torr). Therefore the PC>2 threshold for ASR corresponded almost perfectly with the PC>2
threshold for MO2 decline in large fish (i.e. 50 torr versus 50 torr) , and not at all for
small fish, where there was a large discrepancy (70 torr versus 22 torr ) (Fig. 1).

       However a likely explanation for this difference, was provided by a second
behavioural experiment run in a long shallow arena tank. An oxygen gradient was set up
by bubbling nitrogen into one end and air into the other, as illustrated in Fig. 2A. At the
most hypoxic end of the tank (approximately 30 torr), shelter was created by the addition
of some floating  plants, Pistia stratiotes (shaded  area). Fish which were naive to the
arena were  added individually at the middle of the gradient (at -65 torr), and observed for
10 minutes. Large oscar spent approximately equal amounts of time under the hypoxic
shelter, and exploring the less hypoxic (but exposed) parts of the arena (Fig. 2B).  In
contrast,  small oscars spent essentially the entire  period under the hypoxic shelter and did
not explore the less hypoxic exposed areas. When the fish were pre-exposed to this same
level of hypoxia  (30 torr) in a separate tank for 60 min prior to the test, the pattern of
exploratory behaviour by the large fish did not change significantly, whereas the small
fish abandoned the hypoxic shelter, spending  about 80% of their time in the more
normoxic but exposed areas (Fig. 2B).

       A possible interpretation of these results is that at least down to a PC>2 of 22 torr,
ASR has more negative consequences over the short term for fitness of the small fish
than remaining in a hypoxic environment because it exposes them to aerial predators
(Kramer  etal., 1983; Randle and Chapman, 2005) and other predatory air-breathing fish
(Wolf, 1985). Therefore they choose to remain under the shelter for virtually the entire
experimental period and accept the associated physiological cost of exposure to hypoxia.
However, when pre-exposed for one hour at 30 torr, their "reserves" of physiological
tolerance, which are lower in small oscar because of their lower anaerobic (Almeida-Val
et a/., 2000) and  oxygen regulation capacities (Fig. 1), become exhausted so that they are
                                       183

-------
forced to choose more oxygenated waters at the sacrifice of shelter. Indeed, Shingles et
al. (2005) recently demonstrated that ASR is a behavioural O2-chemoreflex that can be
modified by the risk of predation in the flathead grey mullet, Mugil cephalus. These
observations fit with the ideas of Claireaux etal. (1995) who studied the behaviour and
physiology of cod in fluctuating salinity and oxygen conditions, and concluded that in
responding to environmental factors, fish may simply be constrained into choosing the
lesser of two evils.
               A
                     120 n
                     100 -
                      80 -
                      40 -
               B
                  CO
                  H
700


600


500


400


300


200 -


100


  0
                            123456789    10

                                 Equidistance Positions Along the Tank
Prior Exposure to Normoxia
Prior Exposure to Hypoxia
                                       1
                                   Large                    Small
                                             Fish Size


Figure 2.   (A) Gradient of PC>2 in the test arena: the cross-hatched area represents shelter
           by floating plants.  (B) Time spent under shelter at the most hypoxic end of
           the gradient by large versus small fish with 1 hour prior exposure to either
           normoxia (130 torr, open bars) or hypoxia (40 torr, hatched bars).  (Data from
           Sloman et al., 2006).
                                        184

-------
       Another behavioural test examined the effect of progressive hypoxia on
spontaneous activity level. Social groups of four fish in a large aquarium were established
for a one hour period under normoxia. Behaviour was then examined, firstly under
normoxia (136 torr), then at 80 torr, and finally at 40 torr, with the reductions between
each level occurring steadily over 60 min periods. Vertical and horizontal movements as
            A
                 60 -,
                 50 -
                 30 -
              £  20
              HH
                 10 -
            B
                 25 n
    S
    ^  20 H
                 15 -
              §  10
    .§   5
    o
                            Large Fish
                            A
                           1
                                             B
                                                             B
                 136               80               40

                       Oxygen Concentration (Torr)
                            Small Fish
                                                             Y
                                            XY
                           X
                            T
                           136
                                                             40
                                  Oxygen Concentration (Torr)
Figure 3.
                     (A) Change in horizontal activity in large oscars during progressive
                     hypoxia. (B) Change in vertical activity in small oscars during
                     progressive hypoxia. (Data from Sloman et al., 2006).
                                        185

-------
well as acts of aggression were scored for 10 min periods at each level. Large oscars
were more active than small oscars under normoxic conditions, and formed clear social
hierarchies with many acts of aggression, whereas none were observed in the small fish.
Aggression was not influenced by hypoxia, but large oscars exhibited a decrease in
movement in the horizontal plane with decreasing oxygen tensions (Fig. 3 A), whereas
small fish showed an increase in activity in the vertical plane during hypoxia (Fig. 3B);
activities in the two reciprocal planes were not affected. However the increase in vertical
activity in smaller fish occurred only up to the 50 % depth line, not up to the water
surface (as might be expected at tensions less than 22 Torr). Thus it seems likely that
larger fish reduce their level of activity to aid metabolic suppression (Boutilier and St-
Pierre, 2000) but do not sacrifice the behaviour needed to maintain their social status,
whereas smaller fish, which are not reproductively mature, do not "waste" metabolic
reserves on aggression, but do increase their activity during moderate hypoxia,
potentially in the hope of finding  areas less devoid of oxygen (Domenici et al., 2000).
Clearly, the trade-offs between behaviour and physiology in relation to ultimate fitness
during hypoxic exposure are both size-related and complex.
lonoregulatory Responses to Hypoxia - the Influence of Body Size

       In these experiments, unidirectional and net fluxes of Na+ between the water and
the fish were measured using standard radio-isotopic (22Na) techniques whereby uptake
(influx) is determined by disappearance of the radio-label from the water, net flux by
change in total Na+ in the external water, and efflux by difference (Wood and Randall,
1973a,b;  Kirschner, 1970; Wood, 1992). Water-to-fish volume ratios, radio-isotopic
specific activity, and measurement periods were optimized as a compromise between
measurement sensitivity and the need to avoid errors due to isotopic-recycling
("backflux"). In practice, this meant that the period of total measurement was limited to
7h, so in  various experiments, "snap-shots" of different periods during hypoxic exposure
and recovery were taken. Our initial intention was to expose both small fish and large fish
to the same severe hypoxia (-10 torr) for 3 h, but in preliminary experiments it was
found that some small oscars succumbed at this level, so a less severe hypoxia (-20 torr)
was used for the latter.

       In contrast to rates of oxygen uptake (Fig.  1) and ammonia excretion (Fig. 4C,F),
mass-specific unidirectional Na+ flux rates were of similar magnitude in small (Fig. 4B)
and large fish (Fig. 4E), though net balance tended to be more negative in the former.  The
fact that mass-specific ion-turnover rates are not greater in the expected fashion (e.g.
Bianchini et a/., 2002) may in itself represent a cost-saving adaptive strategy for these
small fish in ion-poor water. The fact that these rates are low in both large and small fish
relative to many other temperate and tropical species appears to be typical of Amazonian
cichlids,  and reflects a low gill permeability which is adaptive to ion-poor and acidic
waters (reviewed by Gonzalez et a/., 2005), as well as to periodic hypoxia.  In both large
and small oscars, Na+ influx rates declined substantially during hypoxia (Fig. 4B,E), in
accord with our original hypothesis that this expensive process would be limited by
oxygen availability.
                                       186

-------
160
120
° 80
40

'
100 -,
50
s —
n\
f-50-
^100 -
-150
onn
OIIIOMIIOII A /j/"\
A Water P02 16° '




B J



^^_


I 120-
o 80 -
40 -
_ , » . m
• « • — —
-
\la in
n
'//,

i
Na o

-ZUU i
700 -i

600 -
^500 -
Jc
7400 -
O)
-^
|300 -

200 -
100 -
0 -
C J

n
ty
XX
y/
ut
rh
%
i
T
100 -
T * * '
rh -r * = 50 -
v 'Y/ ^ ^ ^o
^ 1 * f 1-50-
-150 -
Ofifi
D Water POo
-^- j IT



^
E JNain


Jh
JL

T
Ja


OL
F-

/ /
T
t



JL

T
*

&
*


Amm
r'n 100 -
I















1







1















T_ 80 -
,
5* 60 -

* o
n ^ fl f 40 "
20 -

	 o -1
F JAmm
1 .
























T















I













^
f

234567 1234
_, _



^

*
-^i
*













5




,

*

^










-p
^

6
^_










IT]
*









^


+



7
Time(h) Time(h)
Figure 4. The responses of small oscars (left panels) and large oscars (right panels) to a
         step induction of severe hypoxia for 3 hours followed by a step restoration of
         normoxia. (A,D) Water C>2 tension; (B, E) unidirectional influx (upward bars),
         efflux (downward bars), and net flux (hatched bars) of Na+ across the gills; and
         (C, F) net excretion rate of ammonia. (Data from Wood et a/., 2007).
                                       187

-------
       However, rather than increasing during hypoxia, unidirectional Na+ efflux rates
also declined in both large and small fish, so that net balance remained unaltered (Fig.
4B,E). Furthermore in a separate experimental series with large fish, rather than the
predicted decrease in plasma ions, a slight but significant increase in plasma Na+ and Cl"
concentrations was seen during hypoxia (Fig. 5B), perhaps attributable to
hemoconcentration due to a water shift into lactate-producing tissues as occurs after
exhaustive exercise (e.g. Wang etal., 1994). These observations on Na+ efflux rates and
on plasma electrolytes directly contrast with our hypothesis based on the respiratory-
osmoregulatory compromise, and suggest that other factors come into play.

       The reductions in Na+ influx and efflux rates during hypoxia became significant
more rapidly in large fish (first hour versus second or third hour; Fig. 4B,E), but in view
of the different hypoxia exposure levels and data variability, it is difficult to say whether
this is a true size-specific difference. To pursue this further, more gradual, graded
hypoxia exposure trials were performed. In large fish, reductions in both Na+ influx and
efflux rates first became significant at a threshold PO2 of about 40 torr, while in small
fish the PC>2 threshold  for significant reduction was about 20 torr, though a non-
significant reduction was seen at 40 torr (data not shown). Thus it appears that large
oscars are at least slightly better able to implement this cost-saving response earlier, at a
higher PO2, without compromising ionic homeostasis (i.e. no change in net Na+ balance),
another indicator of their better hypoxia tolerance relative to small oscars.

       Three possible explanations come to mind for the observed decreases in Na+
efflux rates during hypoxia (Fig. 4B,E). The first is that exchange diffusion is turned
down during hypoxia,  so that the reduced Na+ efflux is directly coupled to the reduced
Na+influx. In  the exchange diffusion phenomenon, the same transport mechanism may
perform both "futile" Na+/Na+ self-exchange and vectorial transport (e.g.  Na+ /H+
exchange). Exchange diffusion has been seen during normoxia in many freshwater
teleosts and crustaceans (e.g. Shaw, 1959; Wood and Randall, 1973b), including about
half of the Amazonian teleosts surveyed by Gonzalez et al. (2002), but to our knowledge,
has never been studied during hypoxia. However, as pointed out by Potts (1994), the
phenomenon can be equally well explained by an exchange protein or a selective channel
mechanism linked to an electrogenic pump, such as the H+-pump, Na+-channel model of
active Na+ uptake discussed earlier (Avella and Bornancin, 1989; Lin and Randall, 1995).
This brings us  to the second possibility, that channel arrest, a well-documented
phenomenon at the cellular level in hypoxia-tolerant species (Hochachka and Lutz, 2001;
Boutilier, 2001), could also occur at the gills, thereby reducing both Na+ influx and Na+
efflux. Surprisingly, this has not been previously investigated in gill tissues during
hypoxia exposure to our knowledge, but will be discussed in greater detail subsequently.
The third possibility is that a hypoxia-tolerant species such as the oscar actually reduce
gill area and permeability during hypoxia by reducing lamellar perfusion so as to reduce
ionoregulatory costs in a situation where the potential for oxygen uptake from the water
has become very slight.
                                       188

-------
160 -I
120 •
80 •
40 -
n .
A







I 	 _
m 	 1
Water PO2




                    C   0
1    2
5    6
                                         Gill Na+ K+ - ATPase
Figure 5. Changes in (B) plasma Na+ and Cl" concentrations; (C) gill Na+, K+-
         ATPase activity; and (D) plasma total ammonia concentration in large
         oscars subjected to 3 hours of severe hypoxia followed by 3 hours of
         normoxic recovery as illustrated in (A). (Data from Wood et a/., 2007).
                                  189

-------
       It seems unlikely that this final idea is the complete explanation, because both Na+
influx and efflux rates remained depressed during the first hour of return to normoxia
(Fig. 4B,E), despite the fact that blood measurements demonstrated almost complete
clearance of the blood lactate load during this first hour (data not shown). To allay
concerns that this was a not a measurement artifact in the final hour of the "flux
window", an additional 7h experiment was performed on large fish, focussing on the
recovery period as its mid-point, and confirmed that restoration of Na+ influx and efflux
rates was delayed for one hour following the re-establishment of hypoxia (data not
shown).

       Several other lines of evidence point to  at least a partial temporal disconnection
between the oxygen regime and the simultaneously measured ion fluxes, suggesting that
two (or more) mechanisms may be causing the  reduction in Na+ turnover during hypoxia,
in accord with one of our initial hypotheses. The first is that in the large fish experiment
of Fig. 5, branchial Na+,K+-ATPase activity was fully maintained at  the end of 1 h  of
hypoxia (Fig. 5C), despite the fact that Na+ influx and efflux rates had already declined
(Fig. 4 E). Na+,K+-ATPase activity did fall greatly (by about 60%) by 3h of hypoxia, but
had fully recovered by Ih of normoxia re-establishment (Fig. 5C), yet unidirectional flux
rates remained depressed during this first hour of recovery (Fig.  4E). Secondly, in a
separate experiment on large fish, unidirectional Na+ flux rates were measured during
hours 1 and 2 of hypoxia, and again at hour 8 of continued hypoxia. Both influx and
efflux values were further reduced at this latter time,  again pointing to the involvement of
more than one mechanism (data not shown).  Thirdly, the dependence of Na+influx rate on
external Na+concentration ("kinetics"; Wood and Goss, 1990; Potts,  1994) was measured
during normoxia and again during prolonged hypoxia (2h to  lOh at ~ 10 torr) in the same
large fish (Fig. 6A). The maximum Na+ influx rate (Jmax) was significantly depressed by
about 60 % during hypoxia as might be expected from the observed 60% reduction in
branchial Na+,K+-ATPase activity (Fig. 5C).  There was no significant change in affinity
(Km; Fig. 6A). Notably, these Km values are high relative to many Amazonian teleosts
surveyed by Gonzalez etal. (2002), and this low affinity (i.e. high Km's) is in accord
with the low permeability, low Na+ turnover, and high Km values recorded in other
cichlids collected in this region (reviewed by Gonzalez et a/., 2005).

       In summary, the data are consistent with down-regulation of Na+ influx (and
efflux) rates by at least two mechanisms during hypoxia. One is clearly by a delayed
reduction in Na+,K+-ATPase activity. Notably,  this appears to occur by post-translational
modification of enzyme activity, because specific Na+,K+-ATPase mRNA and protein
abundance did not fall in a comparable experiment on Astronotus ocellatus in which
branchial Na+,K+-ATPase activity was depressed by about 50% after 4h of hypoxia (J.G.
Richards, pers. comm.). There is also clearly another mechanism (or mechanisms) which
is/are more rapid and more persistent, but additional work will be required to determine
the relative contributions of oxygen starvation,  channel closing,  changes in lamellar
perfusion, and alterations in the activity of other proteins (e.g. H+-ATPase, Na+/H+ and
Na+/Na+ exchangers) in the observed responses. Regardless,  the bottom line is that
Astronotus ocellatus can withstand severe hypoxia without a marked disturbance of
internal ion status, by simultaneously reducing  Na+ pumping and leak rates at the gills.
                                       190

-------
As with other adaptations, this ability appears to be at least slightly better developed in
large fish.

Nitrogen Metabolism Responses to Hypoxia - the Influence of Body Size

       Similar to most other teleosts, Astronotus ocellatus is strongly ammoniotelic;
urea-N excretion rates are less than 10% of ammonia-N excretion rates. As with MC>2
(Fig. 1), mass-specific ammonia excretion rates (JAmm) were much higher in small fish
than in large fish (Fig. 4C, F). In general, the relative difference was even greater than in
MC>2, suggesting that small oscars rely on protein oxidation to a greater extent (van den
Thillart and Kesbeke,  1978; Lauff and Wood, 1996). Ammonia excretion rates (JAmm)
were also much higher on an absolute basis than Na+ influx rates, especially in small fish
(Fig. 4B,C,E,F), so any coupling of ammonia excretion to Na+ uptake, if it occurs, must
be indirect. Nevertheless, during normoxia, there was clear evidence of some sort of
coupling, because JAmm exhibited a Michaelis-Menten type dependency on external Na+
concentration in both large oscars (Fig. 6B) and small oscars (data not shown).
Interestingly however, the apparent Km was significantly lower (i.e. affinity for Na+ was
higher) than for the simultaneously determined Michaelis-Menten dependence of Na+
influx on external Na+ concentration, whereas the Jmax values were similar (Fig. 6A,B).
              A   J Ma in - Large Fish
                                    B  J Arm- Large Fish
   500n


   400-
  •300-
 D)
   100-

    0
                             600n
                             500-
Norrroxia
Km = 780 urrol I/1
Jmax = 502 urrol kg"1 h"1
r2 = 0.96
           Hypoxia
              = 470urrolL~1       IQQ
              = 218urralkg-1h-1*
           [2 = 0.96
                                                 0
            Norrroxia
            Km=103urrolL-1*
            Jmax = 434 urrol kg
      0    200    400    600   800
                   Na+ (umol L'1)
                  1000   1200
200   400   600   800   1000  1200
            Na+ (umol. L'1)
Figure 6. (A) Michaelis-Menton kinetics of Na+ influx as a function of external Na+
          concentration in large oscars during normoxia or severe hypoxia. Note the
          decrease in Jmax during hypoxia. (B) Michaelis-Menton kinetics of ammonia
          excretion as a function of external Na+ concentration in large oscars during
          normoxia. Note the lack of a Michaelis-Menton relationship during severe
          hypoxia. Note also the difference in Km between the Na+ influx relationship
          (A) and the ammonia excretion relationship (B) in normoxia. (Data from Wood
          etal, 2007).
                                        191

-------
       During hypoxia, JAmm was reduced in a similar fashion to Na+ influx and efflux
rates in both small fish (Fig. 4B,C) and large fish (Fig. 4E,F). Again the reduction
occurred more quickly in large oscars. Notably, as with Na+ influx, recovery did not
occur during the first hour of restoration of normoxia (Fig. 4C,F).  Furthermore, in the
prolonged hypoxia experiment with large fish, the reduction in JAmm at hour 8 was
significantly greater than during hours 1 and 2 (data not shown). Overall, these responses
were very similar to those seen in unidirectional Na+ fluxes, arguing for some sort of
common mechanism. However, the apparent kinetic coupling of JAmm to external Na+
concentration was completely lost during prolonged hypoxia (Fig. 6B). While all these
observations were in accord with our original hypotheses, it remains to be determined
whether the declines in JAmm were driven primarily by a down-regulation of the ammonia
production rate, or by specific hypoxia-induced blockade of a branchial ammonia
excretion mechanism, such as "diffusion-trapping" linked to an H+-pump/Na+-channel
system or Na+/H+ exchange (see Introduction).

       For several reasons, it appears likely that both phenomena were involved, and that
the latter predominated. Urea-N excretion, although it represented only a small fraction of
N-waste excretion, was reduced whenever JAmm was reduced during hypoxic exposures
(Fig. 7). Urea arises from different metabolic pathways than ammonia in teleost fish
(uricolysis or arginolysis rather than trans-deamination or adenylate breakdown; see
Wood, 1993, Wilkie, 2002 for reviews). This suggests that a general reduction in
metabolic N-waste production occurs during hypoxia, in accord with a general
suppression of metabolic rate. However the slope of the regression relating the relative
reduction in urea-N excretion to that in ammonia-N excretion was only 0.57, and the
intercept was significantly greater than zero, indicating that ammonia-N excretion was
more strongly depressed at any given level (Fig. 7). Secondly, if the reduction in JAmm
were simply a consequence of reduced production, it seems unlikely that it should persist
during the first hour of normoxia restoration when aerobic metabolism was likely
restored (Fig. 4C,F). And most cogently, plasma total ammonia concentration increased
significantly by 3h of hypoxic exposure in the blood sampling experiment (Fig. 5D),
suggesting that the excretion mechanism was inhibited to a greater extent than the
production mechanism. This conclusion is in accord with early observations that
ammonia-N production is reduced to a lesser extent than aerobic metabolic rate when
teleosts are exposed to hypoxia (Kutty, 1972; van den Thillart and Kesbeke, 1978; van
Waarde, 1983). Notably  however, the change in plasma ammonia was not large, so again
the oscar appears to very good at maintaining internal homeostasis during severe
hypoxia.

Concluding Remarks
       In general, research on hypoxia in fish has emphasized the respiratory and
metabolic responses, with less concern for how behaviour and other physiological
systems may be impacted. By focusing on an extremely hypoxia-tolerant Amazonian
species (Astronotus ocellatus\ we have found important responses in behaviour, in
                                       192

-------
ionoregulation, and in N-waste excretion which help to preserve homeostasis and thereby
ensure survival during severe environmental hypoxia. Many of these adaptations are
                120 -i
                100 -
              8
              ro
              5
                 60 -
                 40 -
                 20 -
large fish, gradual hypoxia
large fish, acute hypoxia
small fish, gradual hypoxia
small fish, acute hypoxia
                                                 urea-N (%) = 0.57 Amm-N (%)+ 23
                                                 r=0.75, P=< 0.0001
                           20       40       60       80

                                     Amm - N (% of control)
                                                            100
                                                                     120
        Figure 7.  The relationship between the mean relative urea-N excretion rate (as a
                  percentage of the normoxic control value) and the mean relative
                  ammonia-N excretion rate in individual periods of various hypoxic
                  exposure experiments with large and small oscars. (Data from Wood
                  etal., 2007).

size-dependent. In future, it will be of interest to examine these same areas during more
moderate hypoxia exposure in hypoxia-intolerant model species, such as salmonids.

Acknowledgements

       Financial support was provided by a Natural Sciences and Engineering Research
Council of Canada (NSERC) Discovery Grant to CMW, by a National Research Council
(CNPq) of Brazil/Amazon State Research Foundation (FAPEAM) PRONEX grant to
ALV, and by a National Science Foundation U.S.A. (NSF) Grant JOB 0455904 to PJW.
CMW is supported by the Canada Research Chair Program and VMFAV and ALV are
supported by CNPq and FAPEAM, and are recipients of research fellowships from
CNPq. KAS received a travel grant from the the Royal Society, and MK received travel
grants from the Society of Experimental Biology, and the American Fisheries Society.
Special thanks to Maria de Nazere Paula da Silva, Gudrun de Boeck, Marisa Fernandes
Castilho, Ana Cristina Leite Menezes, Linda Diao, and Sunita Nadella for help with
experiments and analyses.
                                       193

-------
References

Almeida-Val, V. M.F., A.L. Val, W.P. Duncan, F.C.A. Souza, M.N. Paula-Silva, and A.
       Land. 2000. Scaling effects on hypoxia tolerance in the Amazon fish Astronotus
       ocellatus (Perciformes: Cichlidae): contribution of tissue enzyme levels. Comp.
       Biochem. Physiol. 125B: 219-226.

Avella, M. and M. Bornancin. 1989. A new analysis of ammonia and sodium transport
       through the gills of the freshwater rainbow trout (Salmo gairdnerf). J. Exp. Biol.
       142:155-175.

Bianchini, A., M. Grosell, S.M. Gregory, and C.M. Wood. 2002. Acute silver toxicity in
       aquatic animals is a function of sodium uptake rate. Environ. Sci. Technol. 36:
       1763-1766.

Boutilier, R.G. 2001. Mechanisms of cell survival in hypoxia and hypothermia. J. Exp.
       Biol. 204: 371-381.

Boutilier, R.G. and J. St-Pierre. 2000.  Surviving hypoxia without really dying. Comp.
       Biochem. Physiol. 126A: 481-490.

Burleson, M.L., D.R. Wilhelm, andNJ.  Smatresk. 2001. The influence offish size on the
       avoidance of hypoxia and oxygen selection by largemouth bass. J. Fish Biol.  59:
       1336-1349.

Chapman, L.J., C.A. Chapman, and M. Chandler. 1996. Wetland ecotones as refugia for
       endangered fishes. Biol. Cons. 78: 263-270.

Claireaux, G., D.M. Webber, S.R. Kerr, and R.G. Boutilier. 1995. Physiology and
       behaviour of free-swimming Atlantic cod (Gadus morhud) facing fluctuating
       salinity and oxygenation conditions. J.  Exp. Biol. 198:  61-69.

Clarke, A. and N.M. Johnston. 1999. Scaling of metabolic rate with body mass and
       temperature in teleost fish. J. Animal Ecol. 68: 893-905.

Clarke, A.P. and W.T.W. Potts. 1998.  Sodium, net acid, and ammonia fluxes in
       freshwater-adapted European flounder (Platichthys flesus L.). Pharmacological
       inhibition and effects on gill ventilation volume. J. Zool., Lond. 246: 427-432.

Domenici, P., J.F. Steffensen, and R.S. Batty. 2000. The effect of progressive hypoxia on
       swimming activity and schooling in Atlantic herring. J. Fish Biol. 57: 1526-1538.

Febry R. and P. Lutz. 1987. Energy partioning in fish: the activity-related cost of
       osmoregulation in a euryhaline cichlid. J. Exp. Biol. 128: 63-85.
                                       194

-------
Fink W.L. and S.V. Fink. 1979. Central Amazonia and its fishes. Comp. Biochem.
       Physiol. 62A: 13-29.

Gonzalez, RJ. and D.G. McDonald. 1992. The relationship between oxygen consumption
       and ion loss in a freshwater fish. J. Exp. Biol. 163: 317-332.

Gonzalez, R.J., R.W. Wilson, C.M. Wood, M.L. Patrick, and A.L. Val. 2002. Diverse
       strategies for ion regulation in fish collected from the ion-poor, acidic Rio Negro.
       Physiol. Biochem. Zool. 75: 37-47.

Gonzalez, R.J., R.W. Wilson, and C.M. Wood. 2005. lonoregulation in tropical fish from
       ion-poor,  acidic blackwaters. In: The Physiology of Tropical Fish, Fish
       Physiology, Vol. 22. A.L. Val, V.M. Almeida-Val, and DJ. Randall (Eds.).
       Academic Press, San Diego.

Graham, J.B.  1997. Air Breathing Fishes: Evolution, Diversity and Adaptation. Boston,
       MA: Academic Press.

Holeton, G.F. and DJ. Randall. 1967a. Changes in blood pressure in the rainbow trout
       during hypoxia. J. Exp. Biol. 46: 297-305.

Holeton, G.F. and DJ. Randall. 1967b. The effect of hypoxia on the partial pressures of
       gases in the blood and water sfferent and efferent to the gills of rainbow trout. J.
       Exp. Biol. 46:317-327.

Hochachka, P.W. and P.L. Lutz. 2001. Mechanism, origin, and evolution of anoxia
       tolerance  in animals. Comp. Biochem. Physiol. 130B: 435-459.

Hughes, G.M. 1981. Effects of low oxygen and pollution on the respiratory systems of
       fish. Pages 121-144 In: Stress and Fish. A.D. Pickering(Ed.). Academic Press,
       New York.

Junk, W.J., G.M. Scares, and F.M. Carvalho. 1983. Distribution offish species in a lake
       of the  Amazon river floodplain near Manaus (Lago Camaleao), with special
       reference to extreme oxygen conditions. Amazoniana 7: 397-431.

Kirschner, L.B. 1970. The study of NaCl transport in aquatic animals. Am. Zool.  10: 365-
       376.

Kolar, C.S. and F J. Rahel.  1993. Interaction of a biotic factor (predator presence) and an
       abiotic factor (low oxygen) as an influence on benthic invertebrate communities.
       Oecologia95: 210-219.

Kramer, D.L.  1987. Dissolved oxygen and fish behaviour. Environ. Biol. Fish. 18: 81-92.
                                      195

-------
Kramer, D.L. and M. McClure. 1982. Aquatic surface respiration, a widespread
       adaptation to hypoxia in tropical freshwater fishes. Environ. Biol. Fish. 7: 47-55.

Kramer, D.L. and J.P. Mehegan. 1981. Aquatic surface respiration, an adaptive response
       to hypoxia in the guppy, Poecilia reticulata (Pisces, Poeciliidae). Environ. Biol.
       Fish. 6: 99-313.

Kramer, D. L., D. Manley, and R. Bourgeois. 1983. The effect of respiratory mode and
       oxygen concentration on the risk of aerial predation in fishes. Can. J. Zool. 61:
       653-665.

Krogh, A. 1939. Osmotic Regulation in Aquatic Animals. Cambridge University Press,
       Cambridge.

Kutty, M.N. 1972. Respiratory quotient and ammonia excretion in Tilapia mossambica.
       Mar. Biol. 16: 126-133.

Lauff,  R.F. and C.M. Wood. 1996. Respiratory gas exchange, nitrogenous waste
       excretion, and fuel usage during starvation in juvenile rainbow trout,
       Oncorhynchus mykiss. J. Comp. Physiol. 165B: 542-551.

Lin, H. and DJ. Randall. 1995. Proton pumps in fish gills. Pages 229-255 In: Cellular
       and Molecular Approaches to Fish Ionic Regulation. C.M. Wood and TJ.
       Shuttleworth (Eds.). Academic Press, London.

Lyndon, A.R. 1994. A method for measuring oxygen consumption in isolated perfused
       gills. J. Fish. Biol. 44:707-715.

Maetz, J. and F. Garcia-Romeu. 1964.  The mechanism of sodium and chloride uptake by
       the gills of a fresh-water fish Carassius auratus. J. Gen. Physiol. 47:  1209-1226.

Mangum, C.P., M.S. Haswell, J. Johansen, and D.W. Towle. 1978. Inorganic ions and pH
       in the body fluids of Amazon animals. Can. J. Zool. 56: 907 - 916.

Morgan, J.D. and G.K. Iwama. 1999. Energy cost of NaCl transport in isolated gills of
       cuthroat trout. Am.J. Physiol.: R631-R639.

Muusze, B., J. Marcon, G. van den  Thillart, and V. Almeida-Val. 1998.  Hypoxia
       tolerance of Amazon fish. Respirometry and energy metabolism of the cichlid
       Astronotus ocellatus. Comp. Biochem. Physiol. 120A:  151-156.

Olson,  K.R. 2002. Gill circulation: regulation of perfusion distribution and metabolism of
       regulatory molecules. J. Exp. Zool. 293: 320-335.

Potts, W.T.W. 1994. Kinetics of sodium uptake in freshwater animals:  A comparison of
       ion exchange and proton pump hypotheses. Am. J. Physiol. 266: R315 - R320.
                                      196

-------
Randle, A.M. and LJ. Chapman. 2005. Air-breathing behaviour of the African
       anabantoid fish Ctenopoma muriei. J. Fish Biol. 67: 292-298.

Robb, T. and M.V. Abrahams. 2003. Variation in tolerance to hypoxia in a predator and
       prey species: an ecological advantage of being small? J. Fish Biol. 62:1067-1081.

Shaw. J. 1959. The absorption of sodium ions by the crayfish, Astacuspallipes
       Lereboullet. I. The effect of external and internal sodium concentrations. J. Exp.
       Biol. 36: 126-144.

Shingles, A., DJ. McKenzie, G. Claireaux, and P. Domenici. 2005. Reflex
       cardioventilatory responses to hypoxia in the flathead gray mullet (Mugil
       cephalus) and their behavioural modulation by perceived threat of predation and
       water turbidity. Physiol. Biochem. Zool. 78: 744-755.

Sioli, H. 1984. The Amazon and its main affluents:  hydrography, morphology of the river
       courses, and river types. Pages 127-166 In: The Amazon: Limnology and
       Landscape Ecology of a Mighty Tropical River and its Basin. H. Sioli (Ed.).
       Dr.W. Junk Publishers, Dordrecht.

Sloman, K.A., C.M. Wood, G.R. Scott, S. Wood, K. Kajimura, O.E. Johannsson, V.M.F.
       Almeida-Val, and A.L. Val. 2006. Tribute to R.G. Boutilier: The effect of size on
       the physiological and behavioural responses of oscar, Astronotus ocellatus, to
       hypoxia. J. Exp. Biol. 209: 1197-1205.

Smale, M.A. and C.F. Rabeni. 1995. Hypoxia and hypothermia tolerances of headwater
       stream fishes. Trans. Am. Fish. Soc. 124:  698-710.

Val, A.L. and V.M.F. de Almeida-Val. 1995. Fishes of the Amazon and Their
       Environment. 224 pp. Springer, Berlin.

van den Thillart G. and F. Kesbeke. 1978. Anaerobic production of carbon dioxide  and
       ammonia by goldfish, Carassius auratus L.  Comp. Biochem Physiol. 59A: 393-
       400.

van Waarde, A. 1983. Aerobic and anaerobic ammonia production by fish. Comp.
       Biochem. Physiol. 74B: 675-684.

Wang, Y., G.J.F. Heigenhauser, and C.M. Wood. 1994. Integrated responses to
       exhaustive exercise and recovery in rainbow trout white muscle: acid-base,
       phosphogen,  carbohydrate, lipid, ammonia,  fluid volume and electrolyte
       metabolism.  J. Exp. Biol. 195: 227-258.

Wilkie, M.P. 1997. Mechanisms of ammonia excretion across fish gills. Comp. Biochem.
       Physiol. 118A: 39-50.
                                      197

-------
Wilkie, M.P. 2002. Ammonia excretion and urea handling by fish gills: present
      understanding and future research challenges. J. Exp. Biol. 293: 284 -301.

Wilson, J.M. and P. Laurent. 2002. Fish gill morphology: inside out. J. Exp. Zool. 293:
      192-213.

Wilson, R.W. 1996. Ammonia excretion in fish adapted to an ion-poor environment.
      Pages 123-138 In: Physiology and Biochemistry of the Fishes of the Amazon.
      A.L. Val, V.M. Almeida-Val and D.J. Randall (Eds.). INPA Press, Manaus.

Wilson, R.W., P.M. Wright, S. Munger, and C.M. Wood. 1994. Ammonia excretion in
      rainbow trout Oncorhynchus mykiss:  the importance of gill boundary layer
      acidification: Lack of evidence for Na+/NH4+ exchange. J. Exp. Biol. 191: 37-58.

Wilson, R.W. 1996^ Ammonia excretion in fish adapted to an ion-poor environment.
      Pages 123-138 In: Physiology and Biochemistry of the Fishes of the Amazon.
      A.L. Val, V.M. Almeida-Val and D. J. Randall (Eds.). INPA Press, Manaus.

Wolf, N.G. 1985. Air breathing and risk of aquatic predation in the dwarf gourami Colisa
      Mia. Am. Zool. 25: 89A.

Wood, C.M.  1992. Flux measurements as indices of H+ and metal effects on freshwater
      fish. Aquat. Toxicol. 22: 239-264.

Wood, C.M.  1993. Ammonia and urea metabolism and excretion. Pages 379-425. In: The
      Physiology of Fishes. D. Evans (Ed.). CRC Press, Boca Raton.

Wood, C.M.  and G.G. Goss.  1990. Kinetic analysis of the relationships between ion
      exchange and acid-base regulation at the gills of freshwater fish. Pages 119-136
      In: Animal Nutrition and Transport Processes, Vol. 2. J.P. Truchot and B. Lahlou
      (Eds.).  Karger, Basel.

Wood, C.M.  and DJ. Randall. 1973a. The influence of swimming activity on sodium
      balance in the rainbow trout (Salmo gairdnerf). J. Comp. Physiol. 82: 207-233.

Wood, C.M.  and DJ. Randall. 1973b. Sodium balance in the rainbow trout (Salmo
      gairdnerf) during extended exercise. J. Comp. Physiol. 82: 235-256.

Wood, C.M., B.R. McMahon, and D.G. McDonald. 1978. Oxygen exchange and vascular
      resistance in the totally perfused rainbow trout. Am.  J. Physiol. 234: R201-R208.

Wood, C.M., M. Kajimura, K.A. Sloman,  GR. Scott, S. Wood, PJ. Walsh, V.M.F.
      Almeida-Val, and A.L. Val. 2007. Down-regulation  of Na+ fluxes andN-waste
      excretion in response to severe environmental hypoxia in the Amazonian oscar,
      Astronotus ocellatus.  In Preparation.
                                      198

-------
Wright, P. A. and C.M. Wood. 1985. An analysis of branchial ammonia excretion in the
       freshwater rainbow trout: effects of environmental pH change and sodium uptake
       blockaden. J. Exp. Biol.  114: 329-353.
                                      199

-------
200

-------
    Effect of winter hypoxia  on fish in small lakes with
                        different water quality

                                      by

                      A. Tuvikene1, L. Tuvikene and M. Viik

Introduction
       The most common cause of natural fish kills is lack of oxygen in surface water. A
lack of oxygen can asphyxiate most fish species. Oxygen deficiency may occur in
shallow lakes with thick ice and snow cover. Light is shut off by ice and snow, and the
plants, including planktonic algae, consume more oxygen than they produce. Due to
temperature stratification in the water under ice cover, and the most intensive degradation
of organic matter at the bottom water layer, the concentration of dissolved oxygen starts
to decrease first at near-bottom water layers. Usually in such lakes, temperate zone
oxygen deficiency extends to the whole water column by the second half of winter -
February or March.  Massive fish deaths occur most often in shallow eutrophic lakes
where a high content of organic matter results in extensive oxygen depletion. For
example, in shallow eutrophic Lake Vortsjarv, the rate of mean oxygen decrease under
the ice cover can reach 100 mg/m2/day in March (Tuvikene et al, 2002). Substances
released into the water after massive fish deaths affect local water quality, and can
temporarily intensify eutrophication (Kasumyan and Tuvikene, 2004).

       Low oxygen concentrations are known to modify metabolism, growth rate and
feeding efficiency offish (Pichavant et al., 2000). At low concentrations, oxygen can act
as a limiting factor for growth and, in some conditions, oxygen may be a more important
limiting factor than  food (Kramer, 1987).

       During the severe winter of 2002/2003 fish populations in many Estonian lakes
suffered from hypoxia, and massive fish kills happened.  During 2003-2005 we performed
a fish survey with multi-size gillnets on several small lakes with different water qualities
(conductivity, nutrients) to assess fish numbers and relative biomass (CPUE) as well as
species composition and condition factor (CF). The aim  of the survey was to estimate the
natural recovery offish populations in different lakes within 3 years after the massive fish
kill.

Material and methods

       During 2003-2005 we performed a fish survey with multi-size gillnets to assess
fish numbers and relative biomass (CPUE, catch per unit effort), species composition and
1 Centre for Limnology, Institute of Agricultural and Environmental Sciences, Estonian
University of Life Sciences, Rannu 61101, Tartu County, Estonia
                                      201

-------
condition factors (CF, Fulton's coefficient of condition, CF = total weight •  standard
Iength3/100). The lakes studied were: Prossa (surface area 33 ha, average depth 2.2 m),
Jarvi Pikkjarv (5 ha, 1.8 m), Pillejarv (7 ha, 3.6 m), Kahala (346 ha,  1.0 m), Endla (180
ha, 2.4 m), Lahepera (100 ha, 2.4 m) and Eistvere (24 ha, 1.5 m). The lakes differ from
each other mainly by the mineralization level (Fig. 1). Fish were caught with multisize
monitoring gillnets (14 different sections with mesh size between 6.25 and 75 mm).
Calibration studies with repeated fish seining and fish marking, and following recapturing
showed that 1000 g CPUE corresponds roughly 100 kg fish/ha.
          u
          (0
                    Eistvere
Prossa
               Figure 1. Water conductivity in lakes during winter 2005.

       To assess the dependence offish hypoxia tolerance on low water mineralization,
we investigated, in the laboratory, the surface respiratory behavior of the roach (Rutilus
rutilus) adapted during 24 hours in Lake Jarvi Pikkjarv water (low mineral content) and
Lake Kahala water (high mineral content). In the aquarium experiments, we registered
the oxygen content in water when fish switched to continuous surface respiration. We
also investigated the emission of nutrients into the water from dead fish (killed by
asphyxia) during the first 24 h. Emissions of nutrients from dead fish were measured at
25 °C and 4 °C, representing summer and winter condition, respectively.

Results and discussion

Oxygen conditions in studied lakes

       Winter hypoxia occurs most often in small and shallow lakes, as seen in Lake
Prossa and Lake Eistvere (Fig. 2).
                                       202

-------
                             1m         2m
                                Depth (m)
                                                  3m
 Figure 2.  Content of dissolved oxygen under the ice in Estonian small
                lakes in March 2004.

Biomass offish in studied lakes

       The highest fish biomasses were recorded in the larger lakes Endla and Lahepera
(Fig. 3) where the biggest changes of biomass also occurred. The lowest CPUE was
recorded in the soft water lake Jarvi Pikkjarv. The prehypoxia levels of biomass were
achieved in some lakes within three years of recovery. The very soft water lake Jarvi
Pikkjarv (conductivity 30 uS/cm), suffered the most severe fish population decline with
deaths of hypoxia tolerant crucian carp (Carassius carassius) and gibel carp (C. gibelio)
recorded (Fig. 3). Perhaps this is due to a low ion content of the water resulting in a
greater osmoregulatory energy cost. The number offish species in the lakes varied
between 3 and 9, being lower in the lakes suffering most often from hypoxia (3 in Lake
Prossa) and low mineralization (3 in Jarvi Pikkjarv). In larger lakes the number offish
species was 9 in Lahepera, 8 in Endla, and 7 in Kahala.
            8000


            6000


            4000


            2000
                  I
n-BI
                 Eistvere   Endla    Jarvi   Kahala Lahepera   Pille   Prossa
                               Pikkjarv
   Figure 3. Relative biomass offish (CPUE) in studied lakes in different years
                                       203

-------
   Condition factor offish

       The condition factor (CF) offish is a frequently used index in fish studies. It
provides important information related to fish physiological state, based on the principle
that individuals of a given length, exhibiting higher weight, are in better condition. CF
provides a relatively simple and rapid indication of how well a fish copes with its
environment. A decline in CF may reflect a change in feeding pattern, which could be a
behavioral response to low oxygen, or an increase in metabolic rate in response to stress
caused by hypoxia. Hypoxia has been shown to reduce food intake, thus resulting in
slower growth (Randall and Yang, 2004) and a smaller CF.

       The CF of many fish species is lower in lakes which are often hypoxic when
compared to the same species in lakes where hypoxia is  rare (Table I). Condition factors
were also low in the soft water lake Jarvi Pikkjarv. This  could be due to elevated osmotic
stress, resulting in less energy for growth. It was also found that the number offish
species was reduced in lakes which were often hypoxic (Table I).

Table I. Condition factors offish.
Lake
Gibel
carp,
Carassius
auratus
Crucian
carp,
Carassius
carassms
Roach,
Rutilus
rutilus
Perch,
Perca
fluviatilis
Pike,
Esox
lucius
Eistvere
3.2







1.9





0.9


Endla








1.9


1.6





Jarvi
2.5



2.8






1.6





Kahala
3.0



3.2



2.0


2.1


1.0


Lahepera
3.2







1.7


1.9


0.9


Pille
3.7







1.8


1.9


0.9


Prossa








1.6


1.7





Hypoxia tolerance in water with low mineralization

       All freshwater fishes devote a significant portion of their basal metabolic rate to
maintain their internal salts and other dissolved substances at concentrations different
than those in their environment. Salts are very efficiently resorbed from the urine of
freshwater teleosts, e.g., 99.9% of the Na+ and Cl~ ions (Bone et al, 1995). Freshwater
teleosts also have a high-affinity salt-uptake mechanism at the gills. The efficiency of this
                                       204

-------
mechanism becomes obvious from the low rate of salt loss in freshwater teleosts in
comparison with marine teleosts, and by the accumulation of ions from very dilute
solution, e.g., Na+ >10'4 M or 2.3 mg/1 (Bone etal, 1995).

       The energy costs of osmotic and ionic regulation mean that less energy can be
allocated to growth (Matey, 1996). In water with low mineralization, fishes experience
osmotic stress. Large water uptake results in high rates of dilute urine excretion, and ion
losses must be compensated for by energetically costly ion pumping (Sollid et al, 2003).

       When acclimated to soft water, fish experience a proliferation of the ion-
transporting chloride cells of the branchial epithelium, which contributes to the
maintenance of ionic homeostasis by enhancing branchial ion uptake (Matey, 1996).

       In spring 2003 many crucian carps and gibel carps died in lake Jarvi Pikkjarv.
These fish species are known to be highly tolerant to hypoxia. In addition to hypoxia, the
reason for their death in lake Jarvi Pikkjarv could be the osmotic stress resulting from a
very low ion content in the lake water (Table II). Lake Jarvi Pikkjarv is a forest lake and
there is no evidence of any direct anthropogenic influence on this lake. An alkalinity of at
least 20 mg/1 is good for the well-being offish (Wedemeyer and Yasutake, 1977).
Table II. Water parameters in soft water lake Jarvi Pikkjarv and in hard water lake
        Kahala.

Na+
K+
Ca2+
Alkalinity
Conductivity
Unit
mg/1
mg/1
mg/1
mg/1
[j,S/cm
Jarvi Pikkjarv
2.1
0.73
5
15
29-53
Kahala
8.4
3
70
156
470
       The use of the aquatic surface respiration is one of the few alternatives to aerial
respiration, which allow fish to survive in conditions of extreme hypoxia. Surface
respiration behavioral studies were conducted with roach Rutilus rutilus exposed to the
soft water of lake Jarvi Pikkjarv and the hard water of lake Kahala water. Fish adapted in
the water of lake Jarvi Pikkjarv showed higher thresholds for aquatic surface respiration,
than fish adapted in water of lake Kahala (Fig. 4). This was most likely due to a higher
osmotic stress in water with low mineralization.
                                       205

-------
                   1,5
                6
                u>
                   0,5
                                Kahala
Jarvi
                                            Lakes
       Figure 4. Oxygen content at 25 °C when roach switched to surface respiration
                (statistically significant difference, P<0.05, N=3).

Nutrients from dead fish

       Dead fish are a source of chemical substances that appear in the surrounding
water shortly after fish deaths (Kasumyan and Tuvikene, 2004). Dead fish may cause a
rapid increase in the content of dissolved organic matter in the surrounding water (Table
III). In turn, this can result in a dissolved oxygen deficit that aggravates other unfavorable
conditions, and can lead to massive fish mortality. Nutrients from dead fish accelerate
eutrophication of a water body.

Table III. Average emission of nutrients (|ig/g/h) from dead fish during first 24 hours (see
         text for more details).

Total nitrogen
Nitrate
Nitrite
Ammonia
Total phosphorus
Reactive phosphorus
4°C
6.1
0
0.05
2.0
1.0
0.5
25°C
116.2
0
0.75
40.2
27.0
21.9
Difference (folds)
20
0
15
20
27
44
       Average fish biomass in Estonian lakes is around 200 kg/ha (10 g/m3 if the lake is
2 m deep). If all the fish in a lake die, then of all of the nutrients released from dead fish
at 25 °C (Table III), only total and reactive phosphorus are important from the point of
view of eutrophication. They can raise the phosphorus content in water (Table IV) by
30%. The nitrate, nitrite, and ammonia is negligible (2% increase in ammonia, other
nitrogen compounds less). In winter conditions, at 4°C, the quantity of released nutrients
from dead fish is negligible (all nutrients less than 1%).
                                        206

-------
Table IV. Chemical parameters of the upper water layer measured in spring 2006.

N-tot (mg/L)
NH4-N (mg/L)
NO2-N (mg/L)
NO3-N (mg/L)
P-tot (mg/L)
PO4-P (mg/L)
BOD5 (mg O/L)
Jarvi Pikkjarv
1.28
0.68
<0.01
0.10
0.03
0.005
1.5
Kahala
1.47
0.48
0.01
0.01
0.04
0.015
1.7
Summary

       During the severe winter of 2002/2003 in Estonia, fish populations in many lakes
suffered from hypoxia/anoxia, and massive fish kills occurred. In several lakes the fish
populations recovered within the following three years. In a very soft water lake even
hypoxia tolerant crucian carp and gibel carp died from hypoxia in conjunction with
elevated osmotic stress due to very low ionic content of the lake water. In the summer,
only the amounts of total and reactive phosphorus released from dead fish are significant
from the point of view of eutrophication while the quantity of nitrogen ions is negligible.
Under the ice cover, the quantity of released nutrients from dead fish is negligible.
Hypoxia or/and low mineralization of lake water result in unfavorable conditions for fish.

Acknowledgements

       This study was supported by grant No 008910 from Estonian Ministry of the
Environment and by the projects No 0362107s02 and No 0362480s03 from Ministry of
Education and Research  of Estonia.

References

Bone, Q., N.B. Marshall, and J.H.S. Blaxter.  1995. Page 332 In: Biology of Fishes.
       Second edition. Blackie Academic and Professional. Glasgow.

Kasumyan, A. and A. Tuvikene. 2004. Composition and rate of release of chemical
       substances into the water by dead fish. Pages 263-275 In: Fish physiology,
       toxicology, and water quality. Proceedings of the seventh international
       symposium, Tallinn, Estonia, May 12-25, 2003. G. Rupp and M. D. White (Eds.).

Kramer, D.L.  1987. Dissolved oxygen and fish behaviour. Environ. Biol. Fishes 18: 81-
       92.

Matey, V. E. 1996. Page 205 In: Gills of freshwater teleost fishes (in Russian). Nauka
       Publishing, St. Petersburg.
                                       207

-------
Pichavant, K., J. Jerson-Le-Ruyet, N. Le Bayon, A. Severe, A. Le Roux, L. Quemener, V.
       Maxime, G. Nonnotte, and G. Boeuf. 2000. Effects of hypoxia on growth and
       metabolism of juvenile turbot. Aquaculture 188(1-2): 103-114.

Randall, D., and H. Yang. 2004. The role of hypoxia, starvation, p-naphthoflavone and
       the aryl hydrocarbon receptor nuclear translocator in the inhibition of
       reproduction in fish. Pages 253-261 In: Fish physiology, toxicology, and water
       quality. Proceedings of the seventh international symposium, Tallinn, Estonia,
       May 12-25, 2003. G. G. Rupp and M. D. White (Eds.).

Sollid, J., P. De Angelis, K. Gundersen, and G.E. Nilsson. 2003. Hypoxia induces
       adaptive and reversible gross morphological changes in crucian carp gills. J. Exp.
       Biol. 206: 3667-3673.

Tuvikene, L., P. Noges, and T. Noges 2002. Hypoxia/anoxia  in Lake Vortsjarv, Estonia.
       Pages 163-169 In: Fish physiology, toxicology, and water quality. V. Thurston
       (Ed.). Proceedings of the sixth international symposium, LaPaz, B.C.S, Mexico,
       January 22-26, 2001.

Wedemeyer, G. and W.T. Yasutake. 1977. Clinical methods for the assessment of the
       effects of environmental stress on fish health. Technical Papers of the U.S. Fish
       and Wildlife Service. 89: 18.
                                       208

-------
  Interdemic variation in gill  morphology of a eurytopic

                             African cichlid

                                      by

              LJ. Chapman1'2, T. DeWitt3, V. Tzaneva1 and J. Paterson1

Introduction

       The physical environment has a major influence on the ecology of organisms. For
fishes, the availability of dissolved oxygen (DO) is one abiotic factor that can exert a strong
selective force by affecting habitat quality, growth, survival, and reproduction. Oxygen
scarcity (hypoxia) occurs naturally in systems characterized by low levels of ambient light and
mixing. For example, heavily vegetated swamps, flooded forests, and the hypolimnion of deep
lakes are habitats particularly prone to oxygen limitation. Low DO can be acute in (but not
limited to) tropical waters where high temperatures elevate rates of organic decomposition and
reduce oxygen tensions in the water. Unfortunately, environmental degradation is increasing
the occurrence of hypoxia as the influx of municipal wastes and fertilizer runoff accelerates
eutrophication and pollution of water bodies (Prepas and Charette, 2003).  Increasing hypoxia
is a threat to fresh waters and coastal waters worldwide; and oxygen depletion in deeper
waters, one side effect of this process, can lead to fish kills and a massive reshaping offish
communities (Prepas and Charette, 2003). It has therefore become increasingly important to
understand the effects of hypoxia on aquatic organisms.

       Despite much interest in the physiological and biochemical adaptations of fishes to
hypoxic stress, the significance of dissolved oxygen in driving divergence among populations
remains largely unexplored.  Strong selection for hypoxia tolerance may lead to variation
among populations that experience divergent aquatic oxygen environments.  This may lead to
further diversification if the benefits accrued by higher respiratory performance in hypoxic
habitats lead to sub-optimal performance in normoxic waters. Our studies of East African
fishes have demonstrated that alternative dissolved oxygen (DO) environments provide a
strong predictor of interpopulational (interdemic) variation, particularly with respect to
respiratory traits (e.g., gill size) and associated characters. In a series of studies comparing
populations from low- and high-oxygen environments, we found total gill size (surface area
and/or total gill filament length) to be larger in swamp-dwelling populations of the cyprinid
Barbus neumayeri (Chapman etal., 1999; Schaack and Chapman, 2003), the mormyrids
Gnathonemus victoriae and Petrocephalus catostoma (Chapman and Hulen, 2001), the cichlid
1 Department of Biology, McGill University, 1205 Avenue Docteur Penfield, Montreal, PQ,
Canada H3A IB 1
2 Wildlife Conservation Society, 2300 Southern Boulevard, Bronx, New York, USA 10460
3 Departments of Wildlife & Fisheries Sciences and Plant Pathology & Microbiology, Texas
A&M University, 2258 TAMU, College Station, Texas, USA 77843-2258
Corresponding Author. Lauren Chapman, Fax: 514-398-5069; Tel. 514-398-6431; e-mail:
Lauren. chapman@mcgill. ca
                                      209

-------
Pseudocrenilabrus multicolor victoriae (Chapman et al, 2000; Chapman et al., 2002), and the
air-breathing African catfish Glorias liocephalus (McCue, 2001) relative to open-water
populations. This interdemic variation in gill traits could be purely genetically based or due to
environmental influences on gene expression (i.e. phenotypic plasticity; DeWitt and Scheiner,
2004). If there is phenotypic plasticity, it may be determined either at a critical period of
ontogeny, or be a phenotypic response that remains labile throughout an individual's lifetime.

       Recent studies support an element of environmentally-induced plasticity in gill
morphology in response to DO regime. In a study of the African cichlid P. multicolor, we
compared gill size of a population from a stable hypoxic habitat with one of a stable well-
oxygenated habitat (Chapman et al, 2000). In addition, we compared siblings (split-brood)
raised under hypoxic or well-oxygenated conditions.  The response to hypoxia was an increase
in gill area, both in the field (29%) and in the plasticity experiment (18%). The difference in
the magnitude of the response between field and experimental fishes may reflect differences in
selection pressures between populations, and/or a combination of inherited changes and
plasticity (Chapman et al, 2000). For the sea bass (Dicentrarchus labrax) Saroglia et al.
(2002) reported higher gill surface area associated with lower oxygen partial pressure of the
water in which the bass were reared for 3 months, providing further evidence of phenotypic
plasticity in fish gills in response to DO availability. The maintenance of plasticity in gill
morphology of divergent populations may preserve the possibility  for future evolutionary
responses. This could be particularly relevant in aquatic systems of the Lake Victoria basin of
East Africa that is characterized by  a volatile history, peppered by  volcanic explosions and
changing lake levels (Beadle, 1981; Stager et al,  1986; Johnson et al,  1996).

       The species flock of endemic haplochromine cichlids in Lake Victoria represents one of
the most rapid, extensive, and recent radiations of vertebrates known (Kaufman 1992;
Kaufman et al., 1997; Seehausen et al., 2002). By contrast, a small number of eurytopic
cichlid species inhabit abroad range of habitats (rivers, streams, lakes,  and swamps)
throughout the Lake Victoria watershed and adjacent areas.  Mechanisms facilitating the
eurytopic distribution of these species remain largely unknown; however, strong patterns of
morphological variation across populations suggest locally adapted phenotypes (Smits et al,
1996; Chapman et al., 2000).  In this study, we quantified interdemic variation in the gill
morphology of the widespread African cichlid Astatoreochromis alluaudi. We compared gill
metrics of a population from a hypoxic lake habitat with one of a well-oxygenated hyper-
eutrophic lake (Chapman etal, 2000). In addition, we compared siblings  (split-brood) reared
under hypoxic or well-oxygenated conditions for one population.

Study Site and Species

     Astatoreochromis alluaudi is  a  widespread haplochromine cichlid that can be found in a
range of habitats in the Lake Victoria basin of East Africa including fast flowing rivers, lakes,
wetlands, and streams (Greenwood, 1959; Chapman etal., 1996a,b). It is a species well  known
for a high degree  of plasticity in the pharyngeal jaw apparatus and associated  muscles in
response to the nature of its prey base. When introduced to a hard prey diet (e.g., mollusks) A.
alluaudi develops a large pharyngeal mill with hypertrophied muscles; while a softer diet is
associated with  a  smaller jaw and trophic  muscles  (Greenwood,  1965a;  Huysseune  et al,
1994). Smits et al. (1996) reported that A alluaudi feeding on snails showed a 31% increase in
                                       210

-------
head volume compared to fish that fed on insects.  Given the widespread distribution of this
species  and the  extraordinary plasticity of its pharyngeal jaw apparatus, we anticipated a
similarly high level of variation in its gill apparatus in response to variation in aquatic oxygen.
We selected two field populations of A. alluaudi from Uganda. Both are lake populations but
one (Lake Nabugabo) experiences very low oxygen conditions in dense wetland bays and the
other (Lake Saka) is found in areas of high oxygen.

     Lake Nabugabo, a satellite of Lake Victoria (24 km2, approximately 0°45' S and 31°45'
E), is characterized by an extensive stretch of shoreline macrophytes (mainly Miscanthidium
violaceum and Vossia cuspidata), interrupted by stretches of forests (dominated by Ficus spp.)
and sand beaches. Small bays surround the east side of the lake and provide an ideal
environment for the development of the bladder-wort Utricularia.  The lake lies within an
extensive wetland that was formerly a bay on the western shore of Lake Victoria (Fig. 1,
Greenwood, 1965b).  Long shore bars that isolate Lake Nabugabo from Lake Victoria were
created during water-level fluctuations about 4,000 years ago (Greenwood, 1965b). Today,
water from Lake Nabugabo drains southeastward via the Lwamunda swamp before it seeps
through the sand bar into Lake Victoria. Astatoreochromis alluaudi was first reported from
Lake Nabugabo by Greenwood (1965b) based on a survey conducted by a Cambridge
expedition. However, only one specimen was collected at that time with no records of habitat
use.  In recent studies, A. alluaudi has been reported in the wetland ecotones of the lake and the
papyrus-choked Juma River, the main tributary to the lake (Chapman et a/., 1996a,b;
Rosenberger and Chapman, 1999; Schofield and Chapman,  1999). In a quantitative survey of
habitat use of fishes in Lake Nabugabo conducted in 2000 (see Chapman et a/., 2003 for
details), we found that A alluaudi were restricted in their distribution to wetland ecotones,
primarily in small bays characterized by dense growth of Ceratophyllum and bordered by
emergent wetland grasses.  In the 2000 survey, the DO level where A. alluaudi were captured
averaged 3.5 mg I"1 (upper 50 cm of the water column) with an average water temperature of
24.4 °C.  This current distribution pattern  of this species may reflect the introduction of the
predatory Nile perch in the early 1960's.  Many haplochromine cichlids were largely restricted
to wetland ecotones, which serve as  refuges from Nile perch predation (Chapman et a/.,
1996a,b, 2002, 2003).

      Lake Saka is found at the northernmost extreme of the crater lakes in western Uganda
(0° 40'N and 30° 15'E, elevation of  1,520 m) and has a maximum length of 1.4 km and
maximum width of 1.0 km. A small crater forms an embayment at the southeast corner of the
lake (Melack, 1978).  Several wetlands drain into Lake Saka, and the drainage basin has been
almost totally cleared for agricultural production by subsistence farmers and a large Ugandan
government prison farm (Crisman et a/., 2001). Unlike Lake Nabugabo, this is a highly
eutrophic lake characterized by supersaturated DO conditions during the peak light of the day.
Crisman et al. (2001) reported DO in surface waters often exceeding 15 mg I"1 (180%
saturation) during 1995-1998. In a quantitative survey of the fish fauna of Lake Saka
conducted in 2000, we found that A. alluaudi was most abundant in wetland ecotones with
average surface DO concentration of 10.4 mg I"1. Nile perch were also stocked into Lake Saka
in the early 1970s, but currently persist at low numbers.
                                      211

-------
                                            p
            Lake
            Nabugabo
                      Lake Saka
Lake
Nabugabo
Lake Saka
Figure 1. Mean levels of dissolved oxygen concentration (mg I"1) and water temperature (°C)
         measured in the upper 50 cm of the water column at sites where Astatoreochromis
         alluaudi were captured in a survey (2000) of Lake Nabugabo and Lake Saka,
         Uganda. (Unpublished data from Chapman and colleagues.)

Methods

Field Collections
traps.
 In both lakes (Nabugabo and Saka), A. alluaudi were captured using metal minnow
Fish were euthanized with an overdose of MS222 in the field and preserved in buffered
paraformaldehyde (40 g 1" ).

Rearing Experiment

       As part of a larger lab-rearing experiment to explore direct and indirect tradeoffs
between trophic structures and the respiratory apparatus of A. alluaudi, Fl offspring from Lake
Nabugabo were raised under low (-1.3 mg I"1) and high (-7.4 mg I"1) DO. Water temperature
averaged 24.5 °C. We used a split brood design with the Fl offspring of three sets of parents to
provide family level replication; but the number of families was limited due to complexity of
key target traits (gill surface area). Brooding pairs were held in separate, normoxic aquaria
until young were released from the female's mouth.  When a brood was released from a
female, each brood was divided into two groups of 10 individuals and groups randomly
allocated to one of 6 aquaria (20-1) between two treatments.  After 2 months, each tank was
cropped to 6 individuals, by randomly removing and euthanizing offspring. Astaoreochromis
alluaudi in Lake Nabugabo are not molluscivorous, as snails are extremely rare in this system
(Beadle, 1981; Efitre et a/., 2001). In the rearing experiment reported in this paper, fish from
Lake Nabugabo were fed tetramin food flakes once per day. Fish were raised for
approximately 1 year  and then euthanized (MS222)  and preserved in buffered
paraformaldehyde.
                                      212

-------
Gill morphometry

        Gill metrics were estimated for 10 specimens from each field population. We
measured the following gill characters:  total gill filament length (TGFL), average lamellar
density (ALD), average lamellar area (ALA), total hemibranch area (THA), and total gill
surface area (TGSA). For the rearing experiment, we measured three specimens from each
family, and we included only measures for TGFL and THA. TGSA and lamellar characters on
lab-reared fish will be the focus of future studies.  Total gill filament length was measured
using standard methods modified after Muir and Hughes (1969) and Hughes (1984a). For each
fish, the branchial basket was removed, and the four gill arches from the left side  of the basket
were separated.  For each hemibranch of the gill arches, the length of every 5th gill filament
was measured (Fig. 2a).  Two successive measurements along a hemibranch were averaged and
multiplied by the number of filaments in the section between the two filaments. Filament
lengths were summed for the four hemibranchs and multiplied by 2 to produce an estimate of
total gill filament length (TGFL).  Lamellar density was measured in the dorsal, middle, and
ventral parts of every 10th filament of the second gill arch on the left side (Fig. 2b). The total
number of lamellae (on one side of the filament) and average lamellar density (ALD) were
estimated using a weighted mean method that takes into account the difference in length of
different filaments (Muir and Hughes, 1969; Hughes and Morgan, 1973).  For every 10th
filament starting at filament 5, the length and  height of 5 secondary lamellae was  measured at
the top, middle, and bottom sections of the filament (Fig. 2b).  Average values of these
characters for each filament were converted to estimates of lamellar area using a regression
determined through the dissection of several lamellae from various sections of the second gill
arch from  two or more specimens from each population.  The sum of the total lamellar area for
all sections of the  second arch was divided by the total number of lamellae and multiplied by 2
to produce a weighted average bilateral  surface area on one side of the filament (ALA). Total
gill surface area (TGSA) was determined using the formula: TGSA = TGFL x 2 x ALD x
ALA.  To estimate hemibranch area, we digitized the area of the gill filaments on the 8
hemibranchs on 1  side of the fish. This was multiplied by 2 to produce an estimate of the total
hemibranch area (THA, not including the bony arches, Fig. 2a).  For all characters, images
were captured with a Leica stereoscope  and Infinity I camera, and linear and areal dimensions
measured with Motic Images software version 2.0.
                                      213

-------
                            total nemibranch area
                            (enclosed by outer line)
                                                                   1.5mm
                             lamellar density
                                           amellar length-
             (b)
Figure 2. Illustration of measurements used for analyses of gill size and shape of
         Astatoreochromis alluaudi. (a) The length of filaments was used to estimate total gill
         filament length (TGFL); the total hemibranch area (THA) was estimated from the
         outer line, (b) Average lamellar density (ALD) was estimated by measuring the
         length encompassed by 10 filaments at the base, middle, and upper part of selected
         filaments. Lamellar length (shown here) and height (not shown here) were measured
         for 5 lamellae at the base, middle, and upper part of selected filaments and used in
         the calculation of average lamellar area (ALA).
                                      214

-------
       Palzenberger and Pohla (1992) reviewed the literature on gill morphometry of fishes.
From their data set for 28 non-air-breathing freshwater species (with multiple populations for
eight species), they extracted the mean slope of significant regressions for gill morphometric
parameters and body weight. They set the lowest and highest mean values within each
parameter range to 0% and 100% respectively to create a range of values for each gill
character. This permitted them to express the values of a species as a percentage within the
range of values for freshwater fishes.  We used their parameter estimates for total gill surface
area and total gill filament length to estimate these characters for each field population
expressed as a percentage of freshwater fishes.

Results

Field populations

       Gill characters were measured on 10 fish from each field population selected to
maximize range in body mass (Lake Nabugabo: mean body mass = 5.20 g, range = 1.427 g to
11.277 g; Lake Saka: mean = 10.14 g, range= 2.486 to 15.06 g). For the two populations, total
gill filament length, average lamellar area, hemibranch area, and total gill surface area were
positively correlated with body size (Table I). Average lamellar density was negatively related
to body size in both groups (Table I).  ANCOVA indicated no difference in the slopes of the
bilogarithmic relationships between A. alluaudi of the Lake Nabugabo and Lake  Saka
populations for total gill filament length, total hemibranch  area, lamellar density, or lamellar
area, though there was trend toward heterogeneity in slopes for the latter (Table I). Intercepts
differed for all four of these characters between field populations (Table I). When adjusted for
body mass total gill filament length, hemibranch area, and  lamellar area were greater in fish
from the hypoxic waters of the Lake Nabugabo  wetland, than in fish from the well-oxygenated
waters of Lake Saka (Table I, Fig. 3a,b). Interestingly, gill lamellar density was lower in A
alluaudi from Lake Nabugabo (Table I, Fig. 3c). The slopes of the bilogarithmic relationship
between total gill surface  area and body mass differed between the two populations, with gill
surface area increasing more slowly with body size in the Lake Saka population.  Given
heterogeneity in the slopes, we did not test for a difference in the intercepts; however, for a fish
of the average size of the two populations, and using the independent regressions for each
population (5.8 g), total gill surface area was estimated at 35.87 cm2 for A alluaudi from Lake
Nabugabo and 15.65 cm2  for A alluaudi from Lake Saka (Fig. 3d).

       Data on A. alluaudi were converted into a percentage of the estimated range of all
species following Palzenberger and Pohla (1992) for total gill filament length and total gill
surface area.  Total gill filament length, expressed as a percentage of freshwater fishes
averaged 67% for Lake Nabugabo, and 56% for Lake Saka. For total gill surface area, A.
alluaudi from Lake Saka averaged only 20% of the range for freshwater fishes, while those
from Nabugabo averaged 54% of the range.
                                       215

-------
Table I. Summary of linear regression analyses and analyses of covariance (ANCOVA of
         relationships between gill morphometric characters and body mass (g) for
         Astaoreochromis alluaudi from Lake Nabugabo (low-oxygen site) and Lake Saka
         (high-oxygen site).  Both gill characters and body mass were logio transformed.  The
         mean values represent antilogged adjusted means calculated from the ANCOVA
         analyses (sample means adjusted for a common mean body mass of 6 g and a
         common regression line). If slopes were heterogeneous, then we did not test for a
         difference in intercepts (Int.).  *For total gill surface area, the means represented the
         predicted value from the population-specific regression lines, since slopes were
         heterogeneous.
Character
Total gill
filament
length (mm)
Hemibranch
Area
(mm2)
Lamellar
Density
(no. per mm)
Lamellar
area (mm2)
Total gill
Surface
area (cm2)
Site
Nabugabo
Saka

Nabugabo
Saka

Nabugabo
Saka

Nabugabo
Saka
Nabugabo
Saka

n
10
10

10
10

10
9

10
9
10
9

Slope
0.641
0.471

0.671
0.526

-0.17
-0.14

0.641
0.339
0.723
0.529

Int.
2.012
2.891

2.184
3.002

1.589
1.677

-2.467
-2.488
1.003
0.791

r
0.978
0.979

0.995
0.994

0.807
0.786

0.893
0.869
0.97
0.971

P
0.001
O.001

O.001
0.001

0.008
0.007

0.001
0.001
0.001
0.001

ANCOVA Slope ANCOVA Int. Adj.
F p F p means
2.008 0.176 146.665 O.001 2552.70
1782.38

0.342 0.567 150.676 O.001 504.66
321.36

0.195 0.665 29.269 O.001 28.71
36.90

4.429 0.053 23.83 O.001 0.0099
0.0055
8.669 0.01 ***** ***** 30.76*
15.49*

                                      216

-------

    c    3.5
    0)
         3.3
    I
         3.1
 ,° 2.9
          0.0   0.2  0.4  0.6   0.8  1.0   1.2  1.4
                      Log mass (g)
                                               Low DO
                                               High DO

                                                <    1.3
             '-'IP.  D
           0.0  0.2   0.4  0.6   0.8  1.0   1.2  1.4
                      Log mass (g)
                                                     0)
                                                     U
                                                     O)
                                                     o
                                                         1.4
                                                     O)
                                                         "
                                                         0.6
                                                      0.0   0.2  0.4   0.6  0.8   1.0  1.2  1.4
                                                                  Log mass (g)
Figure 3. Bilogarithmic plots of gill metrics and body mass for Astatoreochromis alluaudi from
         two field populations: Lake Nabugabo (low oxygen), Lake Saka (high oxygen), (a)
         total gill filament length (mm), (b) total hemibranch area (mm2), (c) average lamellar
         density (number of lamellae per mm), and (d) total gill surface area (cm2).

Lab-rearing study

       For our lab-rearing study, full siblings whose parents originated from Lake Nabugabo
were raised under low- and high-DO.  Total gill filament length and total hemibranch area were
measured for 3 fish per family per treatment selected to maximize range in body size
(normoxia: mean body weight=5.4 g,  range=1.7 to 10.4 g; hypoxia:  mean=4.8 g, range=3.4 to
9.2 g). For both the normoxia- and hypoxia-raised fish, total gill filament number and
hemibranch area were positively related to body mass (Table II). Analyses of covariance
indicated no difference in the slopes of the bilogarithmic relationships between the hypoxia and
normoxia groups for total gill filament length and hemibranch area (Table II); however, the
intercepts differed between groups.  For a fish of a given body mass, total gill  filament length
                                        217

-------
and hemibranch area were greater in fish grown under extreme hypoxia than in fish grown
under normoxia (Table II, Fig. 4a,b).

Table II. Summary of linear regression analyses and analyses of covariance (ANCOVA) of
         relationships between gill morphometric characters and body mass (g) for Fl
         offspring of Astaoreochromis alluaudi from Lake Nabugabo. Both gill characters
         and body mass were logio transformed. The mean values represent antilogged
         adjusted means calculated from the ANCOVA analyses (sample means adjusted for a
         common mean body mass of 6 g and a common regression line).
 Character
      DO
n  Slope  Intercept   r
ANCOVA  Slope   ANCOVA  Intercept  Adj.
Fp	F	pmeans
Total gill
Filament
length (mm)
Hemibranch
area (mm2)
Hypoxia
Normoxia

Hypoxia
Normoxia
9
9

9
9
0.422
0.48

0.528
0.677
3.094
2.92

2.302
2.021
0.905
0.985

0.902
0.994
<0.001 0.537
<0.001

<0.001 2.904
O.001
0.476 122.776 O.001 2371.4
1733.8

0.11 156.36 <0.001 449.78
295.80
 E
 E
 O)
 a>
 TO
 O)

 1
3.6

3.5

3.4

3.3

3.2

3.1
      3.0
        0.0   0.2   0.4    0.6   0.8    1.0
                   Log mass (g)
                                  1.2
                                                Low DO
                                                High DO
                                                ~
                           O
                           c
                           TO
                           0)
                                               3.0
                                                —   2.8

                                                §!
                                                TO
 2.6


 2.4


 2.2
                                                     2.0
                                                         B
                                 0.0    0.2   0.4    0.6   0.8    1.0   1.2
                                            Log mass (g)
Figure 4. Bilogarithmic plots of gill metrics and body mass for Fl offspring Astatoreochromis
         alluaudi from Lake Nabugabo raised under low- and high-oxygen conditions,  (a)
         total gill filament length (mm) and (b) total hemibranch area (mm2).
                                       218

-------
Discussion

Aquatic Oxygen as a Predictor of Morphological Divergence

       Several studies based on interspecific comparisons of non-air-breathing fishes have
suggested that large gill respiratory surface may reflect hypoxic conditions in their
environment (Gibbs and Hurwitz, 1967; Galis and Barel; 1980, Fernandes etal., 1994; Mazon
et a/., 1998).  There is now a growing body of evidence to support similar patterns of variation
within species. Significant variation in total gill surface area and/or other metrics  of gill size
has been reported among populations of widespread species that inhabit alternative oxygen
environments, including representatives of several families: Cichlidae, Cyprinidae,
Mormyridae, and Poeciliidae (Chapman et al., 1999, 2000, 2002; Chapman and Hulen, 2001;
Schaack and Chapman, 2003; Timmerman and Chapman, 2004).  In A. alluaudi, patterns of
variation in gill metrics between populations from alternative aquatic oxygen environments
support this trend.  Astatoreochromis alluaudi is one of the few extremely widespread
haplochromine cichlids in East Africa, in stark contrast to the enormous number of stenotypic
endemic cichlids in great lakes of the region.  The remarkable morphological variation in its
trophic and respiratory characters may have facilitated its eurytopic distribution. The
haplochromine cichlid Pseudocrenilabrus multicolor inhabits a similarly broad range of
habitats (rivers, streams, lakes, and swamps), and is also characterized by strong interdemic
variation in gill morphology, trophic morphology, and other morphological traits (Chapman et
al., 2000, 2002).

       The difference in total gill filament length between A. alluaudi from lakes Nabugabo
and Saka (30%) was much lower than the difference in total gill surface area (50%), and
notably the total gill surface area offish from Lake Saka fell within the lower range (20th
percentile) of freshwater fishes (derived from equation in Palzenberger and Pohla, 1992).
Lakes Nabugabo and Saka differ not only in dissolved oxygen availability but also in other site
characters that may contribute to differences in gill morphology between lakes. For example,
Lake Saka is  subject to very high concentrations of a number of potentially toxic blue-green
algae (aka cyanobacteria), including Microcyctis aeruginosa, Oscillatoria sp. and
Cylindrospermopsis sp., withM aeruginosa the dominant species by biomass (E.  Phlips and L.
Chapman unpubl. data).  Since the gill comprises over half the body surface area of a fish and
is characterized by a thin barrier between the blood and the water, most chemical transfer
between the fish and the aquatic environment occurs across the gills (Hughes, 1984b; Wood
and Soivio, 1991; Randall and Brauner,  1993). Thus, fishes living in waters with high levels of
algal toxins may be selected for decreased gill surface area. To provide additional support for
the hypothesis of oxygen caused population differentiation, we performed the rearing
experiment to directly test for oxygen effects on gill morphology while holding other
environmental parameters constant.

Developmental Plasticity in Fish Gills

        Phenotypic plasticity often evolves because it allows organisms to mitigate
environmental variation (DeWitt and Scheiner, 2004).  We found a strong element of
developmental plasticity in total gill filament length and total hemibranch area in A. alluaudi in
response to the dissolved oxygen environment in which it was raised.  Astatoreochromis
                                       219

-------
alluaudi grown under normoxia exhibited a total gill filament length and a hemibranch area
smaller (27% and 34%, respectively) than fish raised under hypoxic conditions. Two lines of
evidence to suggest this plasticity in gill morphology is adaptive. First the induced
morphological responses are in the direction one would predict to increase oxygen uptake
capacity.  Second, the response was similar in the lab-reared fish to that observed between the
two field populations from alternative oxygen environments. The next steps in this work will
focus on quantification of gill surface area for this rearing experiment and the interaction of
environmentally-induced variation and population effects by comparing fish from both lakes
Nabugabo and Saka reared under low- and high-oxygen conditions.

       The maintenance of plasticity in these divergent populations may preserve the
possibility for future evolutionary responses and foster population colonization and persistence
in novel environments (Schlichting and Pigliucci, 1998; Yeh and Price, 2004). Swamps grade
into lakes and rivers in the Lake Victoria basin, and small changes in water levels can produce
large changes in available habitat and in connectivity.  Thus, fish lineages may experience
alternative oxygen environments either within or among generations.  We found high levels of
developmental plasticity in the gill morphology of two other species of East African fishes that
persist in variable DO environments.  These other species include the cichlid P. multicolor
(Chapman et al., 2000) and the cyprinid Barbus neumayeri (L. Chapman unpubl. data).  And,
other studies have demonstrated high levels of plasticity in gill traits in fishes (Schwartz, 1995;
Sargolia etal., 2002) and larval  salamanders (Bond, 1960;  Burggren and Mwalukoma, 1983).
Thus, environmentally-induced gill proliferation may be a widespread response to sub-lethal
hypoxic stress.

       The degree of developmental plasticity in response to alternative DO environments in
A. alluaudi may differ depending on other features  of the environment or the natal history of
the population.  When fed on hard prey (e.g., molluscs) A. alluaudi will develop a massive
pharyngeal mill with hypertrophied muscles, whereas a softer diet leads to reduction in
pharyngeal jaw size and associated musculature (Greenwood, 1965a; Huysseune etal., 1994).
Smits et al. (1996) found that the total head volume in snail-eating A.  alluaudi was 31% larger
than in fish from an insect-eating population, and they reported internal reallocations of the
respiratory apparatus (change in the shape of the gills). Thus, gill proliferation may be
compromised to some degree when A alluaudi is faced with dual challenge of hypoxia and a
mollusk-dominated diet. We are currently exploring this interaction in A alluaudi (Chapman,
Galis, and DeWitt, unpubl. data).

       Despite an apparent advantage to  gill proliferation in response to hypoxic stress, an
important issue is understanding what maintains these divergent respiratory phenotypes in the
field.  Why not have large gills in all environments? Fitness trade-offs, whereby the phenotype
with the highest performance in one habitat performs sub-optimally in the alternative
environment, may contribute to the maintenance of variation among field populations (Van
Buskirk etal., 1997; DeWitt and Scheiner, 2004). Trade-offs between feeding and respiratory
structures seem very likely in fishes because of their generally compact, laterally compressed
head morphology. Our studies on two East African fishes, the cichlid P. multicolor and the
cyprinid B. neumayeri,  suggest potential trade-offs  between respiratory phenotypes.  For
example, we demonstrated that adaptive change in gill size (large gills) in fish from hypoxic
                                       220

-------
waters correlates with reduced size of key trophic muscles and feeding performance relative to
small-gilled conspecifics (Chapman et a/., 2000; Schaack and Chapman, 2003). These trade-offs
may lead to fitness costs in the field that impose habitat-specific selection pressures on
dispersers.

       Recent models of the role of phenotypic plasticity in driving genetic evolution argue that
moderate levels of adaptive plasticity are optimal for evolution in novel environments by
enhancing population persistence and placing populations under directional selection leading to
potentially  higher adaptive peaks (Price et a/., 2003). The widespread distribution of some East
Africa cichlids such as A. allauadi and P. multicolor may reflect broad environmental tolerances
due to their phenotypically plastic responses to environmental variation. However, these plastic
responses may be assimilated genetically on the long term if populations are under directional
selection towards new adaptive peaks (Price etal., 2003; West-Eberhard, 2005).

Summary

       The significance of variation in dissolved oxygen in driving phenotypic divergence is a
largely unexplored aspect of aquatic biodiversity. However, there is now strong evidence that
alternative  oxygen environments are a strong predictor of intraspecific variation in fishes,
particularly in respiratory traits and associated characters.  Developmental plasticity seems to
play a large role in explaining variation in gill morphology among populations, and may be an
important mechanism contributing to the widespread distribution of species that cross strong
dissolved oxygen gradients.  Future studies on widespread African cichlids that explore the
interaction  of genetic and environmentally-induced morphological variation in multiple
populations should elucidate the potential that an initially plastic response to a novel oxygen
environment may be followed by genetic changes in the same direction.

       One of the many challenges facing freshwater fishes is the  increasing occurrence of
hypoxia, which has lead to fish kills, changes in fish distribution, and a massive reshaping of
some fish communities.  Thus, it has become increasingly important to understand the
consequences of low-oxygen stress on fish populations. Interpopulational variation and
phenotypic plasticity in respiratory traits may contribute to species persistence in the face of
environmental change.
Acknowledgments

       Funding for studies described in this chapter was provided from the National Science
Foundation (IBN-0094393), the Wildlife Conservation Society, the National Geographic
Society, the Natural Sciences and Engineering Research Council of Canada (Discovery grant to
LJC), McGill University, and the Canada Research Chair program.  Permission to conduct
research in Uganda was acquired from the Uganda National Council for Science and
Technology and Makerere University.  We thank Colin Chapman, Erin Reardon, Brian
Langerhans, Ole Seehausen, and field assistants at the Makerere University Biological Field
Station and Lake Nabugabo for assistance with various aspects of this project.
                                       221

-------
References

Beadle, L.C. 1981.  The inland waters of tropical Africa. Longman, London.

Bond, A.N. 1960. An analysis of the response of salamander gills to changes in the oxygen
      concentration of the medium. Dev. Biol. 2:1-20.

Burggren, W.W. and A. Mwalukoma. 1983. Respiration during chronic hypoxia and hyperoxia
      in larval and adult bullfrogs (Rana catesbeiand). I  Morphological responses of lungs,
      skin and gills.  J. Exp. Biol. 105: 191-203.

Chapman, LJ. and K. Hulen. 2001. Implications of hypoxia for the brain size and gill surface
      area of mormyrid fishes.  J. Zool. 254:461-472.

Chapman, L.J., C.A. Chapman, and M. Chandler.  1996a. Wetland ecotones as refugia for
      endangered fishes.  Biol. Cons.  78:263-270.

Chapman, L.J., C.A. Chapman, R. Ogutu-Ohwayo, M. Chandler, L. Kaufman, and A.E. Keiter.
      1996b. Refugia for endangered fishes from an introduced predator in Lake Nabugabo,
      Uganda. Cons. Biol.  10:554-561.

Chapman, L.J., C.A. Chapman, D. Brazeau, B. McGlaughlin, and M. Jordan. 1999. Papyrus
      swamps and faunal diversification: Geographical variation among populations of the
      African cyprinid Barbus neumayeri. J. Fish Biol.  54: 310-327.

Chapman, L.J., C.A. Chapman, J.P. Olowo, PJ. Schofield, L.S. Kaufman, O. Seehausen, and
      R. Ogutu-Ohwayo. 2003.  Fish Faunal Resurgence in Lake Nabugabo, East Africa.
      Cons. Biol.  17: 500-511.

Chapman, L.J., F. Galis, and J. Shinn. 2000. Phenotypic plasticity  and the possible role of
      genetic assimilation:  Hypoxia-induced trade-offs in the morphological traits of an
      African cichlid. Ecol. Lett. 3:387-393.

Chapman, L.J., C.A. Chapman, F.G. Nordlie, and A.E. Rosenberger. 2002. Physiological
      refugia: Swamps, hypoxia tolerance, and maintenance offish biodiversity in the Lake
      Victoria Region. Comp. Biochem. Physiol.  133A: 421-437.

Crisman, T.L., LJ. Chapman, C.A. Chapman, and J. Prenger.  2001. Cultural eutrophication of
      a Ugandan highland crater lake:  A twenty-five year comparison of limnological
      parameters.  Verhandlungen Internationale Vereinigung Limnologie  27: 3574-3578.

DeWitt,  TJ. and S. M. Scheiner. 2004. Phenotypic plasticity. Functional and conceptual
      approaches.  Oxford University Press, New York.
                                      222

-------
Efitre, J., LJ. Chapman, and B. Makanga. 2001. The inshore benthic macroinvertebrates of
      Lake Nabugabo, Uganda: Seasonal and spatial patterns. African Zoology 36:205-
      216.

Fernandes, M.N., F.T. Rantin, A.L. Kalinin, and S.E. Moron. 1994.  Comparative study of gill
      dimensions of three erythrinid species in relation to their respiratory function. Can. J.
      Zool.  72: 160-165.

Galis, F. and C.D.N. Barel. 1980.  Comparative functional morphology of the gills of African
      lacustrine Cichlidae (Pisces, Teleostei):  An ecomorphological approach. Neth. J. Zool.
      30: 392-430.

Gibbs, R.H. and B.A. Hurwitz. 1967. Systematics and zoogeography of the stomiatoid fishes,
      Chauliodus pammelas and C. sloani, of the Indian Ocean.  Copeia 1967: 798-805.

Greenwood, P.H. 1959. The monotypic genera of cichlid fishes in Lake Victoria. Part II.
      Bull. Brit. Mus. Nat. Hist. (Zoology) 7:163-177.

Greenwood, P.H. 1965a.  Environmental effects on the pharyngeal mill of a cichlid fish,
      Astatoreochromis alluaudi, and their taxonomic implications. Proc. Linn. Soc. Lond.
       176:  1-10.

Greenwood, P.H. 1965b.  The cichlid fishes of Lake Nabugabo, Uganda. Bull. Brit. Mus.
      Nat. Hist. (Zoology) 12: 315-357.

Hughes, G.M. 1984a. Measurement of gill area in fishes: Practices and problems. J. Mar.
      Biol. Assoc. (U.K.) 64: 637-655.

Hughes, G.M. 1984b. General anatomy of the gills. Pages 1-72 In: Fish Physiology, Volume
       10A. W.S. Hoar and DJ. Randall (Eds). Academic Press, Inc., New York.

Hughes, G.M. and M. Morgan. 1973. The structure offish gills in relation to their respiratory
      function.  Biol. Rev. 48: 419-475.

Huysseune, A., J.-Y. Sire, and FJ. Meunier. 1994.  Comparative study of lower pharyngeal jaw
      structure in two phenotypes of Astatoreochromis alluaudi (Teleostei: Cichlidae). J.
      Morph. 221:25-43.

Johnson, T.C., C.A. Scholz, M.R. Talbot, K. Kelts,  R.D. Ricketts, G. Ngobi, K. Beuning, I.
      Ssemmanda, and J.W. McGill. 1996. Late Pleistocene desiccation of Lake Victoria and
      rapid evolution of cichlid fishes. Science 273: 1091-1093.

Kaufman, L.S. 1992. Catastrophic change in species-rich freshwater ecosystems: The lessons
      of Lake Victoria.  BioScience 42:846-858.
                                      223

-------
Kaufman, L.S., LJ. Chapman, and C.A. Chapman. 1997. Evolution in fast forward:
       Haplochromine fishes of the Lake Victoria Region. Endeavour 21:23-30.

Mazon, A. de F., M.N. Fernandes, M.A. Nolasco, and W. Severi. 1998.  Functional
       morphology of gills and respiratory area of two active rheophilic fish species,
       Plagioscion squamosissimus and Prochilodus scrofa. J. Fish. Biol.  52: 50-61.

McCue, M. 2001. African swamps and respiratory tradeoffs: Interdemic variation in the air-
       breathing catfish (Glorias liocephalus). Honours thesis, University  of Florida,
       Gainesville, Florida.

Melack, J.M. 1978. Morphometric, physical and chemical features of the volcanic crater lakes
       of western Uganda. Arch. Hydrobiol.  84:430-453.

Muir, B.S. and G.M. Hughes. 1969. Gill dimensions for three species of tunny. J. Exp. Biol.
       51:271-285.

Palzenberger, M. and H. Pohla. 1992.  Gill surface area of water-breathing freshwater fish.
       Rev. Fish Biol. Fisheries 2: 187-216.

Prepas, E.E. and T. Charette. 2003. Worldwide eutrophi cation of water bodies: Causes,
       concerns, controls. Pages 311-331 In: Treatise on Geochemistry, vol. 9, H.D. Holland
       and K.K.Terekian (Eds). Elsevier Ltd., Science Direct online version.

Price, T.D., A. Qvarnstrom, and D.E. Irwin. 2003. The role of phenotypic plasticity in driving
       genetic evolution.  Proc. Roy. Soc. Lond.  270 :1433-1440.

Randall, DJ. and CJ. Brauner. 1993.  Toxicant uptake across fish gills.  Pages 109-116.  In:
       Fish Physiology, Toxicology, and Water Quality Management.  Proceedings.
       Environmental Research Laboratory, U.S. Environmental Protection Agency, Athens,
       Georgia.  EPA/600/R-93/157.

Rosenberger, A.E. and LJ. Chapman. 1999. Hypoxic wetland tributaries as faunal refugia
       from an introduced predator. Ecol. Freshwat. Fish 8: 22-34.

Saroglia, M., G. Terova, A. De Stradis, and A. Caputo. 2002. Morphometric adaptations of sea
       bass gills to different dissolved oxygen partial pressures.  J.  Fish Biol. 60: 1423-1430.

Schaack, S.R. and LJ. Chapman. 2003. Interdemic variation in the African cyprinid Barbus
       neumayeri: Correlations among hypoxia, morphology, and feeding  performance. Can.
       J. Zool.  81:430-440.

Schwartz, F J.  1995. Gill filament responses and modifications during spawning by a mouth
       brooder and substratum brooder (Tilapiine, Pisces) cichlids. Acta Universitatis
       Carolinae Biologica 39:231-242.
                                       224

-------
Schlichting, C.D. and M. Pigliucci. 1998. Phenotypic evolution. A reaction norm perspective.
       Sinauer Association, Sunderland.

Schofield, PJ. and LJ. Chapman. 1999. Interactions between Nile perch, Lates niloticus, and
       other fishes in Lake Nabugabo, Uganda. Env. Biol. Fish. 55:343-358.

Seehausen, O. 2002. Patterns in fish radiation are compatible with Pleistocene desiccation of
       Lake Victoria and 14,600 year history for its cichlid species flock. Proc. Roy. Soc.
       Lond. B 269: 491-497.

Smits, J.D., F. Witte, and F.G. Van Veen. 1996. Functional changes in the anatomy of the
       pharyngeal jaw apparatus of Astatoreochromis alluaudi (Pisces, Cichlidae), and their
       effects on adjacent structures.  Biol. J. Linn. Soc. 59: 389-409.

Stager, J.C., P.N. Reinthal, and D.A.  Livingstone. 1986.  A 25,000-year history for Lake
       Victoria, East Africa, and some comments on its significance for the evolution of
       cichlid fishes. Freshwat. Biol. 16: 15-19.

Timmerman,  C.M. and L. J. Chapman. 2004. Hypoxia and interdemic variation in the sailfm
       molly (Poecilia latipinna).  J. Fish Biol. 65:635-650.

Van Buskirk, J., S.A. McCollum, and E.E. Werner. 1997. Natural selection for
       environmentally induced phenotypes in tadpoles. Evolution 51: 1983-1992.

West-Eberhard, MJ. 2005. Developmental plasticity and the origin of species differences.
       Proc. Nat. Acad. Sci.  102: 6543-6549.

Wood, C.M. and A.  Soivio.  1991. Environmental effects on gill function:  An introduction.
       Physiol. Zool.  64: 1-3.

Yeh, PJ. and T.D. Price. 2004.  Adaptive phenotypic plasticity and the successful colonization
       of a novel environment. Am. Nat.  164:531-542.
                                       225

-------
226

-------
                  Hypoxia tolerance in  coral reef teleosts

                                            by

                        G.E. Nilsson1 and S. Ostlund-Nilsson
Introduction
       Water-breathing animals run a much greater risk of experiencing hypoxia than air-
breathers, as the concentration of oxygen in air-saturated water is only about 3 - 5 % of
that in air, and because oxygen diffuses about 10 000 times faster in air than in water. In
many situations, aquatic organisms may use up the oxygen in the water before it is
replenished by diffusion or photosynthesis. Hypoxia is particularly likely to occur at
night, when both plants and animals have to rely on respiration for their energy supply.
While tropical freshwater habitats are well known to expose their inhabitants to hypoxia,
it has only recently become apparent that hypoxia is a major abiotic factor shaping the
teleost fauna on coral reefs.

Hypoxia tolerance on the Reef

       A few years ago, we used closed respirometry to survey hypoxia tolerance,
indicated by the critical oxygen concentrations ([O2]crit), in 31 species of teleost fish
representing seven families (Apogonidae, Blennidae, Gobiidae, Labridae,
Monacanthidae, Nemipteridae, and Pomacentridae) on the coral reef at Lizard Island ,
Great Barrier Reef, (Table I) (Nilsson and Ostlund-Nilsson, 2004). [O2]crit is the lowest
oxygen level where a fish is able to maintain its resting rate of oxygen consumption
(Prosser and Brown, 1961). All species examined were found to be strikingly hypoxia
tolerant, showing [O2]crit  values of 13 - 34 % of air saturation (mean being ca 24 %.).
Behavioural signs of hypoxic stress were generally not seen until the O2 level  in the
closed respirometers fell below 10 % of air  saturation, indicating that the fishes were able
to compensate for the reduced oxidative ATP production by boosting anaerobic ATP
production (i.e. glycolysis), or by reducing ATP demand (metabolic depression), or both.

       Until this recent study, hypoxia tolerance was generally not thought to be of
importance for coral reef fishes,  except in the special case of the epaulette shark (recently
reviewed by Nilsson and Renshaw, 2004). Still, the teleosts we examined were a
representative selection of the eye catching piscine beauties that people generally
associate with coral reefs, all  caught in close proximity to living coral in 2 - 5  m deep
water.
1 Department of Molecular Biosciences, University of Oslo, P. O. Box 1041, NO-0316
Oslo, Norway, g.e.nilsson@imbv.uio.no
                                       227

-------
       That coral reef fishes maintain O2 uptake in hypoxia at a temperature of about 30
°C can be viewed as a considerable physiological achievement, due to the combined
effects of a low solubility of O2 in warm sea water, and the high rate of oxygen
consumption that small fish have at such a high temperature. Most fishes examined
weighed less than 10 grams and had resting rates of oxygen consumption (MO2) of 200 -

700 mg O2 kg"  h" , which is several times higher than that of fishes in cold temperate
water. In fact, the [O2]crit values shown by the coral reef fishes are similar to those of
fishes inhabiting hypoxic tropical freshwaters.  African cichlid species like tilapia,
Oreochromis niloticus, for example, display [O2]crit values of about 20 % of air
saturation at 25 °C (Verheyen etal., 1994; Chapman etal., 1995).

When and where do they encounter hypoxia?

       During the light hours, hypoxia is probably very rare or absent on coral reefs.
However, the situation may change drastically  during the night. We now know that coral
reef fishes are exposed to hypoxia  either when  they hide from predators at night by
moving into the coral colonies and residing between coral branches, or when they get
trapped in tide pools  at nocturnal low tides (Nilsson and Ostlund-Nilsson, 2004). In
Acropora nasuta colonies from Lizard Island, kept in outdoor tanks, we found that the
average oxygen level between the branches fell to 20 % of air saturation just before
sunrise (Figure 1), and oxygen levels as low as 2 % were occasionally measured (Nilsson
et al., 2004). Similarly, Goldshmid et al. (2004) found that water oxygen levels fell to 10
- 20 % of air saturation in colonies of Stylophorapistillata (Esper, 1797) from the Red
Sea kept dark in the laboratory. In  the same study, Goldshmid et al., (2004)  obtained
results indicating that damselfishes (Chromis viridis, Dascyllus aruanus and D.
marginatus) inhabiting the Stylophora colonies perform nocturnal "sleep-swimming" to
increase the water flow through the coral in order to reduce the nocturnal hypoxia.

       We recently examined oxygen levels on the reef near Lizard Island on low
nocturnal tides in calm weather between 2.00 - 5.00 am (G. E. Nilsson, S. Ostlund-
Nilsson, and J.-P. Hobbs, unpublished observations). When pushing an oxygen electrode
some 10 cm into coral colonies where numerous fishes were seen to hide, oxygen levels
between 10 and 20 % of air saturation were measured. We also found that in an area of
about 1  m2  of living coral, there could be hundreds of fishes hiding under these severely
hypoxic conditions. This habit of hiding in coral at night has previously been described
(Fishelson etal., 1974; Goldshmid etal., 2004), and is well known to many night divers.
During reef walks at  nocturnal low tides, we also observed living fishes trapped in tide
pools with water oxygen levels as low as 8 - 12 % of air saturation. The fishes observed
to reside under the severely hypoxic conditions in coral and in tide pools included
surgeonfishes, emperors, coral breams, rockcods, damselfishes, butterflyfishes, wrasses,
gobiids, sandperches and cardinalfishes.
                                       228

-------
Table I. Hypoxia tolerance at 30 °C of Fishes at Lizard Island, Great Barrier Reef.
Family / Species
Cardinalfishes (Apogonidae)
Apogon compressus
Apogon cyanosoma
Apogon doederleini
Apogon exostigma
Apogon fragilis
Apogon leptacanthus
Archamia fucata
Cheilodipterus quinquelineatus
Sphaeramia nematoptera
Damselfishes (Pomacentridae)
Acanthochromis polyacanthus
Chromis atripectoralis
Chromis viridis
Chrysiptera flavipinnis
Dascyllus aruanus
Neoglyphidodon melas
Neoglyphidodon nigroris
Neopomacentrus azysron
Pomacentrus ambionensis
Pomacentrus bankanensis
Pomacentrus coelestis
Pomacentrus lepidogenys
Pomacentrus moluccensis
Pomacentrus philippinus
Gobies (Gobiidae)
Amblygobius phalaena
Asteropteryx semipunctatus
Gobiodon histrio
Blennies (Blennidae)
A trosalarias fuscus
Atrosalarias fuscus juvinile
Filefishes (Monacanthidae)
Paramonacanthusjaponicus
Breams (Nemipteridae)
Scolopsis bilineata juvenile
Wrasses (Labridae)
Halichoeres melanurus
Labroides dimidiatus juvenile
N

4
1
1
1
14
14
1
2
1

1
5
6
1
3
6
6
1
4
1
6
5
4
2

1
1
10

3
1

1

1

1
3
Weight Normoxic MO2
(g) (mgkg-lh-1)

7.0±1.2
2.2
4.4
3.7
1.9±0.1
1.5±0.1
5.8
1.8-7.4
7.3

15.4
8.4±2.5
2.5±1.1
2.4
4.1±1.3
32.1±8.8
14.9±2.4
3.2
12.6±1.7
7.8
7.8±3.0
3.1±0.6
5.2±4.0
2.2-6.9

2.4
1.4
1.2±0.2

7.3±1.9
0.29

1.7

1.9

1.8
0.56±0.09
(Normoxic MO2 = rate of C>2 consumption at a water [02] >70% air saturation, [O2]cr;t
no longer independent of ambient [02], [O2]out
fish are means ± SD. Taxonomy follows Randal]

179±67
259
288
218
255±17
239±19
225
244-263
131

197
358±84
555±108
384
306±37
216±32
162±21
493
201±11
237
387±85
516±73
397±85
320-348

333
403
248±31

208±34
552

486

375

394
736±35
= critical [02], below this level
= [02] at which the fish showed signs of agitation or balance problems.
etal. (1997)
[O2]crit [O2]out
(%ofair (%ofair
saturation) saturation

19±5
30
31
26
17±1
19±1
34
23-31
17

26
22±2
23±1.2
30
19±0
25±2
22±3
32
22±4
19
22±4
31±2
25±3
26-33

21
26
18±1

18±2
13

23

28

25
24±5
MO2 starts falling and is
Values for 3 or more

6.7±1.9


11.4
7.2±1.0
7.0±1.2

7.2-11.1
10.0

6.5
8.8±0.8
7.4±0.9
12.0
5.9±0.6
5.6±0.7
8.9±1.5

7.1±2.0

9.3±1.0
12.5±1.1
10.4±1.9
9.3-10.5

2.8
1.4
2.8±0.5

1.6±07
1.5

9.5

12.8

6.8
7.8±0.9


From Nilsson and Osltlund-Nilsson (2004Y
                                              229

-------
                 A. Water outside cora
                 B. Water between coral branches
            0
              18   19  20  21  22  23  24  01   02  03  04  05  06  07  08  09
                                     Time of day (hour)

Figure 1. Tracing of the oxygen level (%air saturation) inside the coral Acropora nasuta,
         the habitat where some coral dwelling gobies spend their whole adult life.
         From Nilsson et al. (2004).

Thus, from being a well oxygenized paradise during the day, at night the reef becomes a
world where the smaller fish hide between branches in coral colonies, having to endure
severe hypoxia.

The extreme coral dwellers

       The species of the genera Gobiodon and Paragobiodon are all obligate coral-
dwellers. Thus, they spend most of their lives in narrow spaces formed between the
branches of coral colonies (Randall etal., 1997). The Gobiodon species are also arguably
exceptionally cowardly fishes. Although they are practically inaccessible to predatory
fish in their coral home, they still secrete a poisonous mucus that should make them
                                      230

-------
inedible to most predators. Fish fed with pieces of Gobiodon die within a few minutes
(Schubert etal., 2003). Moreover, these gobies have the ability to repeatedly change sex,
so that they do not need to leave the coral to find a mating partner (Munday et a/., 1998).
Figure 2.  The scaleless skin of the coral dwelling goby Gobiodon histrio through which
          it can take up oxygen during air exposure. Small circular structures that cover
          the whole surface are poison secreting cells. The width of the picture is 5 mm.
          Photo by Goran E. Nilsson.

       Coral gobies often inhabit coral habitats in shallow water which may become air
exposed for several hours at the lowest tides. A study on two Gobiodon species revealed
not only a higher degree of hypoxia tolerance that the average of coral reef fish, but also
excellent air breathing capabilities (Nilsson et a/., 2004). This study provided the first
evidence that some fishes intimately connected to living coral have evolved excellent air-
breathing abilities. It is well known that many fishes living in tropical freshwaters or
estuarine habitats have evolved air-breathing capacities to cope with hypoxia or air
exposure (Graham, 1997). Apparently, in some situations, it is not  enough to be hypoxia
tolerant for fishes to survive in a coral habitat, they may also need to have the ability to
breathe air. The selection pressure that has given rise to air-breathing in the genus
Gobiodon is clearly the air exposure that they experience when their coral colonies are
out of the water a low tides. By breathing air, they do not have to leave the coral even if it
is completely out  of the water. Interestingly, continued studies (Nilsson, J.-P. Hobbs, S.
Ostlund-Nilsson, and P. L. Munday, unpublished) have shown that the  air breathing
capacities varies with the depth range of the gobiid species. Those  that  live in shallow-
                                        231

-------
water corals can maintain the rate of oxygen uptake in air at the same level as in water for
up to four hours, while those living deeper can only air breath for about an hour. In fact,
the deeper living Paragobiodon xanthosomus was completely unable to breath air, which
coincided with having a body covered with scales. All the Gobiodon species have
scaleless bodies, and the major route of oxygen uptake in these fishes appear to be over
their skin (Fig. 2).

Cardinalfishes: Fitness versus hypoxic survival

       Male cardinalfishes (Apogonidae), one of the more species rich fish families on
coral reefs, are faced with a particular respiratory problem. They brood the egg clutch
they have received from the females in their mouths for approximately two weeks after
fertilization. This egg mass can make up a quarter of their body mass (Ostlund-Nilsson
and Nilsson, 2004), filling up most of the oral cavity, which makes up about 20-30 % of
their body volume (Barnett and Bellwood, 2005). Obviously, one may assume that the
egg mass will reduce the ability of male cardinalfish to ventilate their gills. Like other
coral reef fish, cardinalfishes generally seek shelter in coral colonies at night and will
therefore have to cope with hypoxia. This poses the question: how do male cardinalfish
reconcile mouthbrooding with hypoxia? To try to answer this, we compared the
respiratory consequences of mouth brooding in two cardinalfish species occurring at the
Lizard Island reef: Apogon fragilis  and A leptacanthus (Ostlund-Nilsson and Nilsson,
2004).

       Mouthbrooding did not seem to be particularly costly by itself, since the resting
metabolic rate was not significantly affected by the presence of the egg clutch in the
mouth (after accounting for the oxygen consumption of the clutch it self). By contrast,
[O2]crit of the mouthbrooding  males was about 32% of air saturation, as  compared to 18
% in non-brooding males or females, revealing a significantly reduced hypoxia tolerance
during mouthbrooding.

       Interestingly, the two species studied were found to differ in the mean egg-clutch
mass. The males of A. fragilis and A leptacanthus had broods corresponding to 20.0 %
and 14.4 % of their body mass, respectively, a difference that clearly affected their ability
to cope with hypoxia. Thus, when faced with falling oxygen levels in the closed
respirometer, both species  eventually spat out the clutch, apparently to  save their own
life. However, in A. fragilis (the species with the larger egg mass), the brood spitting
occurred at a higher oxygen level, 21.7 % of air saturation, compared to A leptacanthus.
The latter species spat the eggs at 13.0 % of air saturation. Moreover, in contrast to A
leptacanthus, mouthbrooding A. fragilis were running at their maximal ventilatory
frequency already during normoxic conditions, and they were therefore unable to increase
ventilation when exposed to hypoxia. Apparently, there is a trade-off situation between
brood size and hypoxia tolerance. The successful brooding of a larger clutch should mean
a correspondingly larger increase in fitness, and A. fargilis seems to be gambling on a
brooding period without any severely hypoxic episodes to maximize the fitness gain from
each brood. By  contrast, A. leptacanthus does not seem to take this risk. These different
strategies do not seem to be the result of different environmental constrains in the
                                       232

-------
preferred habitat, because both species often school together in the same habitat, close to
branching coral. Still, it is likely that the mouthbrooding strategy will have to influence
the choice of nocturnal shelter (more or less deep within coral colonies) or willingness to
move out of hypoxic corals. Thus, the mouthbrooding A. fragilis males may have to
spend their night on the outskirts of coral colonies and the fitness they gain from a large
broad may be balanced by a higher risk of being predated.
Landing on the reef: From record swimming to hypoxia tolerance

       Coral reef fishes generally start their life as planktonic larvae (Thresher, 1984),
spending a few weeks drifting in the open water before they settle on the reef. At the end
of their pelagic phase coral reef fish larvae develop extremely impressive capacities for
high-speed sustained swimming (i.e. for hours or even days), which they need to reach
suitable coral habitats (Stobutzki and Bellwood, 1994; Leis and Carson-Ewart, 1997).
Indeed, they appear to be the fastest swimmers of all fishes, since many of them are
capable of reaching maximal sustained swimming speeds of 30 - 50 body lengths per
second (BL/s) (Stobutzki and Bellwood, 1994;  Leis and Carson-Ewart, 1997; Fisher et
a/., 2005). For comparison, it can be mentioned that larvae of temperate fishes do not
usually reach sustainable swimming speeds higher than 4-5 BL/s (Blaxter, 1986; Meng,
1993), while the adult fishes best known for exceptional swimming performance, which
include swordfish (Xiphias\ tunas (Thunnus and Euthynnus) the inconnu (Stenodus
leucichthys) do not exceed 20 BL/s (Aleyev, 1977; Beamish, 1978).
Figure 3. The damselfish, Pomacentrus ambionensis, in the late plaktonic stage, just
         before settlement on a coral reef. This larvae is among fastest swimming
                                       233

-------
         fish, capable of sustaining a speed of about 30 body lengths per second, and
         during swimming it also displays one of the highest rates of oxygen uptake measured
         in any fish. The fish is 12 mm long. Photo by Goran E. Nilsson.

       Swimming can only be sustained if it is fully aerobic and does not lead to a build
up of lactate (Goolish, 1991). We recently constructed a miniature swim respirometer that
allowed us to measure maximal rates of oxygen uptake (MC^max) during high speed
swimming in  pre- and post-settlement larvae of two species of damselfish (Chromis
atripectoralis and Pomacentrus ambionensis} (G. E. Nilsson, S. Ostlund-Nilsson and A.
S. Grutter, unpublished). Our measurements showed that C. atripectoralis and P.
ambionensis reach MC^max of about 5000 and 4000 mg C>2 kg"1 h"1, respectively, when
swimming at maximal sustained speeds. This is, to our knowledge, the highest MO2max
values ever measured in cold blooded vertebrates. Leis and Carson-Ewart (1997)
examined swimming performance in pre-settlement larvae of 17 damselfish species, and
found that they reached an  average sustained speed of 34 BL/s, with C. atripectoralis
larvae being the fastest swimmers observed, reaching maximal sustained swimming
speeds of 53 BL/s. Pre-settlement larvae of P. ambionensis (Figure 3) can be considered
to be more average performer among damselfish larvae, reaching sustained speeds of 30
BL/s (Stobutzki andBellwood,  1994).

       Our results allow us to conclude that the extraordinary high sustained swimming
speeds of pre-settlement damselfish larvae are paralleled by extraordinary high capacities
for rapid oxygen uptake.  Obviously, these traits are important for larvae to reach a
suitable reef at the end of their planktonic period.

       As most fish biologists know, high aerobic capacities of very active fish species
preclude hypoxia tolerance, and vice versa (Burggren et al., 1991, for review). The
underlying reason is probably the opposing demands that a high MO2inax and hypoxia
tolerance put on the oxygen carrying properties of haemoglobin. Oxygen uptake in
hypoxia requires a haemoglobin with a high oxygen affinity. However, this leads to
relatively slow rates of oxygen downloading in the tissues, because oxygen has to be
downloaded at a low partial pressure, resulting in a small pressure gradient from blood
into the mitochondria and therefore a slow oxygen delivery. Consequently, the
haemoglobins of highly active fish show lower oxygen affinities than those of sedentary
species (reviewed by Burggren etal., 1991).

       As we have discussed, adult coral reef fishes are hypoxia tolerant, probably
because they need to cope with hypoxia at night, when they avoid predators by moving
into coral. Our respirometry studies of post-settlement and juvenile individuals of C.
atripectoralis and P. ambionensis (G. E. Nilsson, S. Ostlund-Nilsson and A. Grutter,
unpublished)  show a striking, almost transient, reduction in MO2inax and [O2]crit within
the first 5-10 days of settlement. Apparently, the larval C. atripectoralis and P.
ambionensis,  have to adjust their respiratory capacities to accommodate hypoxia
tolerance at the expense of high rates of oxygen uptake. The result is that they are no
longer exceptionally fast swimmers, but they should, after this transition, be able to seek
shelter within coral colonies at night. It is likely that the transition involves changes in
                                       234

-------
haemoglobin oxygen affinity, and there are examples of ontogenetic changes in
haemoglobin isoform expression in fishes (reviewed by Jensen et a/., 1998).
Conclusions

       For coral reef fishes, as for all animals that strive to survive hypoxia, the key issue
is the same: to ensure that ATP levels are maintained so that the integrity of cellular
function is not compromised. This means that they have to be able to take up oxygen
from the water at as low oxygen levels as possible, and that they can defend their ATP
levels with glycolysis combined with reduced ATP utilization, if aerobic metabolism can
no longer be sustained. Teleosts  on coral reefs are faced with hypoxia when they venture
into coral colonies to seek shelter from predators at night,  or are trapped in tide pools
during nocturnal low tides. While virtually all fishes seen on a coral reef are strikingly
hypoxia tolerant, the species most intimately connected to coral are also those that are
most hypoxia tolerant, and some of these even have excellent air-breathing capabilities.
Most coral reef fishes have planktonic larvae with outstanding swimming abilities. These
will have to rapidly transform their respiratory properties when they settle on the reef,
from being the fastest swimmers with the highest rates of oxygen uptake of any fishes, to
becoming hypoxia tolerant. Since coral reefs contain one of the most biodiverse and
fastest evolving vertebrate communities, we can be certain that a wealth of respiratory
adaptations remain to be discovered and explored among coral reef fishes.

References

Aleyev, Y.G. 1977. Nekton. Dr.  W. Junkb. v. Publishers,  The Hague.

Barnett, A. and D.R. Bellwood. 2005. Sexual dimorphism in the buccal cavity of paternal
       mouthbrooding cardinalfishes (Pisces : Apogonidae), Mar. Biol. 148: 205-212.

Beamish, F.W.H. 1978. Swimming capacity. Pages 101-187 In: Fish Physioloy VII.
       Locomotion, W.S. Hoar and DJ. Randall (Eds). Academic Press, New York.

Blaxter, J.H.S. 1986. Development of sense organs and behaviour of teleost fish larvae
       with special reference to feeding and predator avoidance. Trans. Am. Fish.  Soc.
        115:98-114.

Burggren, W., B. McMahon, and D. Powers. 1991. Respiratory functions of blood. Pages
       437-508 In: Environmental and Metabolic Animal Physiology, C. L. Prosser
       (Ed). Wiley-Liss, New York.

Chapman, L.J., L.S Kaufman, C.A. Chapman, and F.E. McKenzie. 1995. Hypoxia
       tolerance in twelve species of East African cichlids: potential for low oxygen
       refugia in Lake Victoria. Conserv. Biol. 9: 1274-1288.
                                       235

-------
Fishelson L, D. Popper, and A. Avidor. 1974. Biosociology and ecology of pomacentrid
        fishes around the Sinai Peninsula (northern Red Sea). J. Fish Biol. 6: 119-133.

Fisher R, J.M. Leis, D.L. Clark, and S.K. Wilson. 2005. Critical swimming speeds of late-
        stage coral reef fish larvae: variation within species, among species and between
        locations. Mar. Biol. 147:  1201-1212.

Goldshmid, R., R. Holzman, D. Weihs, and A. Genin. 2004. Aeration of coals by  sleep-
        swimming fish. Limnol. Oceanogr. 49: 1832-1839.

Goolish, E.M. 1991. Aerobic and anaerobic scaling in fish. Biol. Rev. 66:  33-56.

Graham, J.B. 1997. Air-Breathing Fishes: Evolution, Diversity, and Adaptation. Academic
        Press, San Diego.

Jensen, F.B., A. Fago, and R.E. Weber. 1998. Hemoglobin structure and functions. In: Fish
        Physiology Vol. 17, Fish Respiration, S.F. Perry and B. Tufts (eds.), Academic
        Press, San Diego. Pp. 1-40.

Leis, J.M. and B.M. Carson-Ewart. 1997. In situ swimming speeds of the late pelagic larvae
        of some Indo-Pacific coral-reef fishes. Marine Ecology Progress Series 159:165-174.

Meng, L. 1993. Sustainable swimming speeds of striped bass larvae. Transactions of the
        American Fisheries Society 122:702-708.

Munday, P.L., M.J. Caley, and G.P. Jones. 1998. Bi-directional sex change in a coral-
        dwelling goby. Behavioral Ecology and Sociobiology  43:371-377.

Nilsson, G.E. and S. Ostlund-Nilsson. 2004. Hypoxia in paradise, widespread hypoxia
        tolerance in coral reef fishes. Proceedings of The Royal Society Series B
        (Biology Letters Supplement) 271:830-833.

Nilsson, G.E. and G.M.C. Renshaw. 2004. Hypoxic survival strategies in two fishes:
        extreme anoxia tolerance in the North European crucian carp and natural  hypoxic
        preconditioning in a coral-reef shark. Journal of Experimental Biology  207:3131-
Nilsson,  G.E, J.-P. Hobbs,  P.L. Munday,  and 8. Ostlund-Nilsson. 2004.  Coward or
       braveheart: extreme  habitat fidelity through  hypoxia tolerance  in  a coral-
       dwelling goby. Journal of Experimental Biology 207:33-39.

Ostlund-Nilsson, S. and G.E. Nilsson. 2004. A mouth full of eggs: respiratory
       consequences of mouthbrooding in cardinalfishes. Proceedings of The Royal
       Society Series B 271:1015-1022.
                                       236

-------
Prosser, C.L. andF.A. Brown. 1961. Comparative Animal Physiology. W. B. Saunders
        Co., Philadelphia.

Randall, J.E, G.R. Allen, and R.C. Steene. 1997. Fishes of the Great Barrier Reef and
        Coral Sea, 2nd ed. Crawford House Press, Bathurst.
Schubert, M., P.L. Munday, MJ. Caley, G.P. Jones, and L.E. Llewellyn. 2003. The
        toxicity of skin secretions from coral-dwelling gobies and their potential role as a
        predator deterrent. Environmental Biology of Fishes 67:359-367.
Stobutzki, 1C. and D.R. Bellwood. 1994. An analysis of the sustained swimming abilities of
        pre- and post-settlement coral reef fishes. Journal of Experimental Marine Biology
        and Ecology 175:275-286.

Thresher, R.E. 1984 Reproduction in Reef Fishes. T. F. H. Publications, Neptune City, NJ.

Verheyen, R., R. Blust, and W. Decleir.  1994. Metabolic rate, hypoxia tolerance and aquatic
        surface respiration of some lacustrine and riverine African cichlid fishes.
        Comparative Biochemistry and Physiology 107A:403-411.
                                       237

-------
238

-------
   Oxygen consumption offish exposed to hypoxia: Are

     they all oxyregulators or are any oxyconformers?

                                     by

                                J.F. Steffensen1

      Animals exposed to a decreasing partial pressure of oxygen - hypoxia - are
traditionally categorized as either oxygen regulators or conformers. Regulators can
maintain their standard metabolic rate when exposed to hypoxia until the so-called
critical partial pressure of oxygen is reached. At further decreasing partial pressures of
oxygen, beyond the critical point, the metabolic rate will decrease and the animal will
start utilizing anaerobic metabolism with a concurrent accumulation of lactate (R in Fig.
1). Conformers, on the other hand, are not able to maintain standard metabolic rate during
hypoxia, and it will decrease linearly with decreasing partial pressure of oxygen (C in
Fig. 1). Metabolic rate versus hypoxia may also have a trace like T in Figure 1. - neither
a conformer nor a regulator, but a Typical trace that can be seen in many publications.
Reasons for this shape will be discussed below.
   100 -i


    90 -


    80


    70 -


 ffl  60 -
 ro
 o
 '•Q  50 -
 J2
 ra
 IS
 E  40 -


    30 -


    20 -


    10
     0
T


R
      0       20      40       60      80      100     120      140      160
                           Partial pressure of oxygen, mmHg

Figure 1. Example of an oxygen conformer (C), an oxygen regulator, and an example
        of an animal with an intermediate typical (T) metabolic rate vs. hypoxia.
1 Marine Biological Laboratory, University of Copenhagen, Strandpromenaden 5, DK-
3000 Helsing0r DENMARK JFSteffensen@bi.ku.dk
                                     239

-------
       The sluggish toadfish (Opsanus tail) is the classic example of an oxygen
conformer (Hall, 1929) and has often been quoted as such for more than 5 decades. Keys
(1930), however, questioned Hall's results due to the experimental design, which was
based on a modified version of a flow-through respirometer described by Ege and Krogh
(1914). While oxygen partial pressure decreased from normoxia to severe hypoxia as the
flow of water through the respirometer was decreased, partial pressure of CC>2 increased,
but this was ignored by Hall. Keys (1930) also considered that stratification affiliated
with this particular flow-through respirometry led to serious errors.
            ec
          =;
          5
          K
          I*
          O
          A
            4C
          II
          o
          o

          5 20
          X
          O
                                 •:
                 Seup
                                  1
                     xt"
                    <
                                 40        60         60
                                  cxrsa:  ?£KSicK   HH.  Kg,
100
120
       Figure 2. Oxygen consumption of the scup, puffer and toadfish, at different
               oxygen tensions. The scup and puffer oxygen consumption vs. hypoxia
               is T-shaped, while the toadfish has a typical conformer C-shape. From
               Hall (1929).
                                      240

-------
       The sturgeon Acipenser transmontanus is another classic example of an oxygen
conformer, as reported by Burggren and Randall (1978). They found that gill stroke
volume and ventilation rate, and hence gill ventilation decreased with decreasing partial
pressure of oxygen (Fig. 3). Since oxygen extraction also decreased, oxygen consumption
decreased as well, and it was concluded that the sturgeon clearly was an oxygen
conformer. Compared to other teleosts it is unusual that gill ventilation volume decreases
with decreasing oxygen partial pressures. It is also unusual that the gill oxygen extraction
is highest at the highest gill ventilation volume - usually oxygen extraction is highest in
normoxia at which the ventilation volume is normally the lowest. Oxygen consumption
was reported to decrease to less than 5 % of the normoxic rates during hypoxia exposure,
and because there was no sign of anaerobiosis, it was concluded that this ancient fish
reduces total energy expenditure during hypoxic exposure,  rather than switching from
aerobic to anaerobic metabolism - it simply shuts down metabolism.
                         400
                         200
                          80

                          GO

                          40

                          2O
                                    40      80      120     160
              oxygen coowmpflon
                (ml
% utf zoticn
                                    40      80     I2O     160
                          80 r   ventilation rate (beots/mln)
                          60

                          4O

                          20
 heort rote(beotsAnin)
                                    40      80      120
                                       Pin (mmHg)
                           160
                                         z
Figure 3. Ventilatory parameters and heart rate in Acipenser transmontanus during
         exposure to 4 levels of hypoxia. (Mean + SD, n=6). From Burggren and Randall
         (1978).
                                       241

-------
        Even though neither Keys' (1930) remarks on the effects of increased CC>2 nor
the stratification problems in Hall's (1929) experiments were never specifically
addressed, Ultsch etal. (1981) revisited the toadfish experiments nonetheless. They also
used flow-through respirometry, but rather than causing hypoxia by decreasing the flow
through the respirometer, they closely and continuously controlled and regulated oxygen
saturation in the water supplying the respirometers.  Hence their experimental design
among others minimized possible  problems with pCC>2. Ultsh etal. (1981) hypothesized
that there was no reason why any vertebrate with a significant amount of functioning
respiratory pigment and a reasonable matching of ventilation and perfusion at the gas
exchange surface should not be able to regulate its standard metabolic rate during a
considerable decrease in partial pressure of oxygen. The results (Fig.  4) showed  that they
were correct and that the toadfish actually was a regulator similar to most other fishes
(Ultsch etal, 1981).
                              sun , Ji.i
                               % • 328
                               * • SB
                               S • 33
                                                       H . J4.S
                                                       »• 1*9
                                                       »• SJ
        o
        (-
        Q.
        5
        D
       O
       u
bJ
o
>-
X
o
                                        mo   0
                                          JS-,
                       tu*> 31.4
                        f, * JS.T
                                                      SUR . (0-»
                                                       fj «212
                                                       w • 314
                                                       S • 56
     0
   sv
                        .«* ,	-*i  .
                   /7^—"  " .. '"..""•—
                                  .
                              ^ " 41-4
                               » -or
                               S- SB
           »' C
                                                                ' J4«
                                                                •3T
                               OXYGEN TENSION  (mmHg)

Figure 4. Resting rates of oxygen consumption of toadfish, maintained and tested at 22
         °C, as a function of the ambient pC>2. From Ultsch etal. (1981).

       Later, a study on two species of sluggish flatfish, the plaice (Pleuronectes
platessa) and flounder (Platichthys flesus) concluded that the former was a regulator
while the latter was an oxygen conformer (Steffensen et al.,  1982) as shown in Figure 5.
In contrast to the sturgeon mentioned above, however, both species of flatfish increased
the gill ventilation volume with respect to decreasing oxygen, and the highest oxygen
extraction was found at the lowest gill ventilation volume - which was in normoxia as
usual.
                                        242

-------
          ,600-
          500-
       'c  400
       I
          .300-
          .200-
       2
       ? .100-
       o
                 • Platichthys  flesus
                 o Reuronectes platessa
                     20      40      60      80     100
                             Water oxygen tension , mm Hg
120
       Figure 5. Oxygen uptake across the gills for plaice and flounder at different
               oxygen tensions. From Steffensen et al. (1982).

       In 1989  Steffensen described some problems and possible errors with different
respirometry methods of aquatic breathers, and how to avoid and correct for them. Only 1
methods, or a combination of these, are normally used - 1) open or flow-through
respirometry, or 2) intermittent flow-through respirometry. Ege and Krogh (1914)
introduced open or flow-through respirometry. Based on the difference in oxygen content
in the water running in and out of the respirometer, and the flow of water through the
respirometer, the oxygen consumption could be calculated. Difficulties with correcting
for the exponential wash-out if the system came out of steady state, however, is the main
problem with this method (Steffensen, 1989). By decreasing the flow of water through
the respirometer the oxygen tension inside the respirometer can be forced to decline, and
the animal exposed to hypoxia. Unfortunately, however, neither Hall (1929) nor several
other investigators  at that time considered the dilution factor and affiliated lag-time and
the erroneous oxygen consumption values that would be the results (Steffensen, 1989).

       Figure 6 shows an example of a somewhat similar erroneous use of flow-through
respirometry. The graph on the far left side shows a steady state situation with a constant
flow of water through the respirometer and a theoretical fish with constant oxygen
consumption. Acutely, the oxygen tension in the inspired water is then changed stepwise,
but the expired water will not be in steady state again before more than 50 min later.  If
oxygen consumption (uncritically) is calculated as the difference in oxygen content of the
                                       243

-------
in- and out-flowing water multiplied with the flow rate, the interpretation will be that
acute hypoxia initially depresses the metabolic rate severely (+ in Fig. 5.). In addition it
will be concluded that the fish slowly acclimate to the hypoxic level, and after > 50 min
the fish have the same oxygen consumption as they initially had under normoxic
conditions. When normoxia is reestablished, the erroneous result for the uncritical
experimenter will be that the fish initially has an oxygen consumption of about twice the
consumption in steady state in normoxia, and that this is due to an oxygen debt that is
paid back during the following 50 min (+).
           160

           uo

           120

           100
        Q?  80
            60
            20
          *•*** •
                                                         . •» i 4 *•* <
                          0    10   20  30   40   0   10  20   30  40
                                 Time, min

       Figure 6. Theoretical example of a flow-through respirometer experiment when
               exposing a fish to acute hypoxia and return to normoxia. For further
               details see Steffensen (1989).

       A similar use of flow-through respirometry may explain why Hall (1929) reported
the toadfish to be a conformer, in contrast to Ultsch et al. (1991).

       In 1938, van Dam made a modification to the flow-through systems involving a
rubber membrane separation of the head and mouth in one compartment and the
operculum and the remainder of the fish in another compartment.  In the initial experiment
oxygen consumption of rainbow trout was studied with respect to hypoxia that was
induced by decreasing the inspired water oxygen tension, as shown in Figure 5.  The
result was that oxygen consumption increased 79 % compared to the normoxic value.
This increase was ascribed to added cost of ventilating the gills. Later it has been
suggested that a cost of ventilation less than 10 % of the total oxygen consumption
(Steffensen,  1985), and the high value found by van Dam can probably be ascribed to
improper use of flow- through respirometry.
                                      244

-------
       Theoretically a correction for the wash-out and lag-time can be carried out
according to Niimi (1978), and the calculated oxygen consumption (x in Fig. 6) is
identical to the theoretical. In practice, however, it often results in quite fluctuating
oxygen consumption values due to the exponential correction, and the method can not be
recommended.

       The flow-through method was also used by Burggren and Randall (1978) when
studying the Acipenser transmontanus. Steffensen (1989) suggested that the response
observed by Burggren and Randall (1978) may have been an experimental artifact
affiliated with the problem described above.

       Are there any other reasons why some earlier studies reported fish to be oxygen
conformers? Spontaneous activity which causes an increase in metabolic rate can also be
part of the problem, as pointed out by Ultsch et al. (1981), particularly since spontaneous
activity often decreases with decreasing partial pressure of oxygen, as shown by
Schurmann and Steffensen (1997). The result of spontaneous activity and hypoxia can
lead to the T-curve shown above in Figure 1.  Handling stress affiliated with the transfer
of a fish to the respirometer can also be part of the reason, and can also result in a T-
curve.

       Overall, most fish species behave as perfect oxygen regulators. Steffensen etal's
flatfish experiment (1982) was repeated with intermittent flow-through respirometry to
address the idea that the reason they did not behave as oxygen regulators could be
ascribed to handling stress causing erroneously high oxygen consumption at normoxia.
The results showed that the plaice, Pleuronectesplatessa, and the dab, Limanda limada,
are both perfect regulators, as  can be seen in Fig.  7. (S0borg et al., 1993), and hance
Steffensen et al. (1982) were wrong.
                                       245

-------
        70
        60
        50
        40
        30
        20
        10
           oxygen consumption (VO2),mg O2/kg/h
                                                Dab, l.imnntia limiindu

                                                Plaice, Plcuroncctcs plsitessa.
               7
               *
Plaice, SMR

Dab, SMR
                       1
           0      20     40    60     80     100    120   140    160
               ambient oxygen tension (PO2), mmHg
     Figure 1. Representative oxygen consumption of 2 species of flatfish exposed to
             hypoxia. Both species behave as oxygen regulators. At intermediate
             hypoxia both species have an increased oxygen consumption that probably
             can be ascribed to a combination of cost of ventilation and spontaneous
             activity. From S0eborg et al. (1993).

      Another example of a perfect oxygen regulator is the pikeperch, Stizostedion
lucioperca, as seen in Figure 8 (Jungersen and Steffensen, submitted). The pikeperch is
wide-spread in freshwaters of Northern and central Europe, and inhabits areas that
regularly reach severe hypoxia. The critical oxygen partial pressure is approximately 25
mmHg, but this species can tolerate oxygen tensions as low as 10 mmHg for short
periods without mortality.
                                    246

-------
                70
                20
                                            N-30
                                                      1SX10
                         20
                                40
                                       60
                                                    100
                                                                  140
         Figure 8. Oxygen consumption of pikeperch,, Stizostedion lucioperca, exposed
                 to hypoxia. From Jungersen and Steffensen (Submitted).

       Yet another example of a perfect oxygen regulator is the lesser sandell,
Ammodytes tobianus, shown in Figure 9. This species is abundant in coastal waters in the
North Sea, and it is believed to bury in hypoxic or anoxic sediments for long periods,
even during the entire winter season. The lesser sandell does not have a swim bladder,
and hence has to swim continuously to keep position in the water column. When the
sandell isn't swimming it is either lying restless on the seafloor or buried in the sediment.
                     100
                      80
                   o>
                   E
                                                                SMR
                         'i
                                          0   50   100  150  26ft  250  3m
                       0   2    4    6   8   10   12   14  16  18   20   22   24
                                            P02 (kPa)


              Figure 9. Oxygen consumption of lesser sandell, exposed to hypoxia.
                       From Behrens and Steffensen (Submitted).
                                       247

-------
       What about the sturgeons? Are they conformers as described by Burggren and
Randall (1978), who also suggested that this response was an adaptive metabolic
depression? A recent study by McKenzie et al. (in press) pointed out that it is an unusual
behaviour for a sturgeon to be resting as in a respirometer since they normally are
cruising around at swimming speeds of approximately 1 body length per second.
McKenzie further suggested it could be a result of not being able to swim and ram
ventilate. The results showed that resting Adriatic sturgeon behave as oxygen conformers
when exposed to hypoxia, and that lactate levels increase while arterial oxygen content
decreases in response to progressive hypoxia. While behaving as an oxygen regulator
when allowed to swim at the preferred speed, neither lactate nor arterial oxygen content
change in response to progressive hypoxia (McKenzie et al.  in press).
        6 i
        4 -
  o
  E
        3
   S
                y = 0.4642x + 1.3877
                    R2 = D.9662
                                   10           15

                                 P02 (kPa)
20
Figure 10.  Oxygen consumption of resting (filled circles) and swimming (closed circles)
          Adriatic sturgeon exposed to hypoxia. From McKenzie et al. (in press).

Conclusions

       In conclusion I believe that most fish species behave as oxygen regulators when
exposed to hypoxia within their natural temperature regime, and there are no a priori
reasons why any vertebrate with a significant amount of functional respiratory pigment
and a reasonable matching of ventilation and perfusion at the gas exchange surface
                                       248

-------
should not be able to regulate its standard metabolic rate over a wide range of oxygen
tensions from normoxia to severe hypoxia. The reason that many investigators have
shown T-curves can probably be ascribed to spontaneous activity, handling stress and or
erroneous use of flow-through respirometry. It must be accepted, however, that at least
some of the sturgeons are oxygen conformers when resting, as described by Burggren and
Randall (1978), but may behave as oxygen regulators when allowed to swim at their
preferred swimming speed according to McKenzie etal. (in press). Are there any other
potential oxygen conformers? What about icefish that do not have any red blood cells and
no functional respiratory pigment? Experiments in the near future will hopefully solve
this question!

References

Behrens, J. and J.F. Steffensen. (Submitted). The effect  of hypoxia on behavioural and
       physiological aspects of lesser sandeel, Ammodytes tobianus.

Burggren, W.W. and DJ. Randall. 1978. Oxygen uptake and transport during hypoxic
       exposure in the sturgeon Acipenser transmontanus. Resp. Physiol. 34: 171-183.

Ege, R. and A. Krogh. 1914. On the relation between the temperature and the respiratory
       exchange in fishes. Int. Rev. Hydrobiol. Hydrograph. 7: 48-55.

Hall, F.F.G.  1929. The influence of varying oxygen tension upon the rate of oxygen
       consumption in marine fishes.  Am. J. Physiol. 88: 212-218.

Jungersen, M. and J.F. Steffensen. (submitted). Pikeperch Stizostedion lucioperca L.
       resting and  exercise metabolism and the effect of graded hypoxia on resting
       metabolism.

Keys, A. 1930. The relationship of the oxygen consumption in the external respiratory
       medium to the oxygen consumption of fishes. Science 71:  195-196.

McKenzie, D., J.F. Steffensen, K. Korsmeyer, N.M. Whiteley, and E.W. Taylor. (In
       press). The  Adriatic sturgeon (Acipenser naccarii) swims to regulate aerobic
       metabolism in hypoxia.  J. Fish Biol.

Niimi, A.J.  1978. Lag adjustment between estimated and actual physiological responses
       conducted in flow-through systems. J. Fish. Res. Bd. Can.  35: 1265-1269.

Schurmann, H. and J.F. Steffensen. 1997. Effects of temperature, hypoxia and activity on
       the metabolism of Atlantic cod, Gadus morhua. J. Fish Biol. 50: 1166-1180.

Steffensen, J.F., J.P. Lomholt, and K.  Johansen.  1982. Gill ventilation and O2 extraction
       during graded hypoxia in two ecologically distinct species of flatfish, the flounder
       (Platichthys flesus) and the plaice (Pleuronectes  platessa). Env. Biol. Fish. 7:  157-
       163.
                                       249

-------
Steffensen, J.F. 1985. The transition from active to ram ventilation in fishes: energetic
       consequences and dependence on water oxygen tension. J. Exp. Biol. 114: 141-
       150.

Steffensen, J.F. 1989. Some errors in respirometry of aquatic breathers: how to avoid and
       correct for them. Fish. Physiol. Biochem. 6: 49 - 59.

S0eborg, B., L.K. J0rgensen, and J.F. Steffensen. 1993. Oxygen consumption of plaice,
       Pleuronectes platessa, and dab, Limanda limanda, exposed to hypoxia. M.Sc.-
       thesis, University of Copenhagen. 44p.

Ultsch, G., D.C. Jackson, and R. Moalli. 1981. Metabolic oxygen conformity among
       lower vertebrates: Thetoadfish revisited. J. Comp. Physiol. 142: 439-443.
                                       250

-------