United States
Environmental Protection
Agency
Health Effects Research
Laboratory
Cincinnati OH 45268
EPA-600/1-80-005
January 1980
Research and Development
Infectivity and
Pathogenicity of
Enteroviruses
Ingested with
Drinking Water
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL HEALTH EFFECTS RE-
SEARCH series. This series describes projects and studies relating to the toler-
ances of man for unhealthful substances or conditions. This work is generally
assessed from a medical viewpoint, including physiological or psychological
studies. In addition to toxicology and other medical specialities, study areas in-
clude biomedical instrumentation and health research techniques utilizing ani-
mals — but always with intended application to human health measures.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/1-80-005
January 1980
INFECTIVITY AND PATHOGENICITY OF ENTEROVIRUSES
INGESTED WITH DRINKING WATER
by
Dean 0. Cliver
Food Research Institute and
World Health Organization Collaborating Centre on Food Virology,
University of Wisconsin-Madison
Madison, Wisconsin 53706
Grant Number R803986
Project Officer
Elmer W. Akin
Water Quality Division
Health Effec'ts Research Laboratory
Cincinnati, Ohio 45268
HEALTH EFFECTS RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Health Effects Research Laboratory,
U.S. Environmental Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the views and policies
of the U.S. Environmental Protection Agency, nor does mention of trade names
or commercial products constitute endorsement or recommendation for use.
ii
-------
FOREWORD
The U. S. Environmental Protection Agency was created because of
increasing public and government concern about the dangers of pollution
to the health and welfare of the American people. Noxious air, foul
water, and spoiled land are tragic testimony to the deterioration of our
natural environment. The complexity of that environment and the inter-
play between its components require a concentrated and integrated attack
on the problem.
Research and development is that necessary first step in problem
solution and it involves defining the problem, measuring its impact, and
searching for solutions. The primary mission of the Health Effects
Research Laboratory in Cincinnati (HERL) is to provide a sound health
effects data base in support of the regulatory activities of the EPA.
To this end, HERL conducts a research program to identify, characterize,
and quantitate harmful effects of pollutants that may result from ex-
posure to chemical, physical, or biological agents found in the environ-
ment. In addition to valuable health information generated by these
activities, new research techniques and methods are being developed that
contribute to a better understanding of human biochemical and physio-
logical functions, and how these functions are altered by low level
insults.
This report provides an evaluation of the number of viral units
required to initiate infection in an animal model. Some aspects of this
study precluded the use of human subjects. Therefore, inferential
judgments on the application of the data to man must be made.
Certain environmental exposures to viruses are subject to man's
control and zero exposure from some sources is technically possible,
e.g. drinking water and wastewater used for irrigation. However,
treatment required to eliminate these potential sources of viral ex-
posure is costly and may be economically unsound unless a significant
health risk exists. Infective dose data is basic to determining the
risk of viral infections from such environmental exposures. This work
has provided a significant contribution to the data base needed to make
sound risk judgments in this area.
R. J. Garner
Director
Health Effects Research Laboratory
iii
-------
PREFACE
Snow1 s epidemiologic studies on cholera in mid-19th-century London
proved that disease could be transmitted through drinking water. Viruses and
their ability to cause human disease were discovered near the turn of the cen-
tury. Later, epidemiologic studies showed that some virus diseases, such as
poliomyelitis and hepatitis A, were also transmissible through drinking water.
Increasingly sensitive methods for detecting viruses (especially entero-
viruses) in drinking water have been developed in recent years, but it is not
yet clear whether the viruses that may thus be detected denote a threat to
human health.
The present study looks at waterborne enteroviruses as potential causes
of infection and disease. Swine and their homologous enteroviruses were
chosen as the model system to produce the most information applicable to
human health with the least cost and risk to research personnel. Devising
the needed experimental procedures took much time and effort, since a study
of this type and design appears to be unprecedented. Thus, in addition to
the insights gained into the initiation of infections by waterborne entero-
viruses, the study has produced a set of techniques applicable to research on
the transmission of viruses and other infectious agents through water and,
perhaps, via other environmental routes. We are grateful to those at the
U.S. Environmental Protection Agency who were willing to risk funding an
unconventional project. We believe that this report vindicates their
judgment.
iv
-------
ABSTRACT
This study was designed to examine the relationship of waterborne entero-
viruses to infections and disease. Young weanling swine and their homolo-
gous enteroviruses were chosen as the model system: The porcine digestive
tract is like that of man, but pigs can be handled under more closely stan-
dardized conditions than humans or other primates. Porcine enteroviruses
resemble those of man in every way, but they infect swine so specifically
that handling the most virulent of the porcine agents is apparently no threat
to the health of research personnel. Known quantities (as measured by the
plaque technique in tissue cultures) of two enteroviruses were administered
in 5 ml of drinking water in such a way that the subjects were obliged to
swallow all of it. The host's body was found to be about 1000 times (600 to
750 for one virus and 1800 to 2500 for the other) less likely than the tissue
cultures to be infected by a given quantity of enterovirus. The ratio did
not depend on whether the animals were fed just before challenge. The prob-
ability of infection was cumulative with iterated small doses: this indi-
cated that there was, in the strict sense, no minimum infectious dose. None
of the infected animals became ill, despite the reported virulence of the
challenge viruses. Chlorine treatment of a concentrated virus suspension,
which reduced infectivity to a level detectable by cytopathic effect but not
plaque formation in tissue culture, left enough virus to infect one of five
challenged subjects. Neither of two colostrum-deprived pigs, challenged by
stomach tube with 20 plaque-forming units of enterovirus at 1% hr of age,
became infected.
This report was submitted in fulfillment of Grant No. R803986 by the
University of Wisconsin under the partial sponsorship of the U.S. Environmen-
tal Protection Agency. This report covers a period from November 1, 1975 to
October 31, 1978, and work was completed as of October 31, 1978.
-------
CONTENTS
Foreword ill
Preface iv
Abstract v
Acknowledgments viii
1. Introduction 1
2. Conclusions 3
3. Recommendations 4
4. Materials and Methods 5
Viruses 5
Cell Cultures 5
Swine 6
Fostering 7
Experimental Housing and Feeding 7
Challenge 9
Observation 9
Testing Fecal Samples 10
Serology 11
5. Experimental Procedures 12
Peroral Infectivity Experiments 12
Peroral Pathogenicity Experiments 12
Empty Versus Full Stomach 12
Iterated Doses 13
Chlorine Disinfection 13
Gavage Challenge of Newborns 14
6. Results 15
Peroral Infectivity Experiments 15
Peroral Pathogenicity Experiments 19
Empty Versus Full Stomach 20
Iterated Doses 20
Chlorine Disinfection | 21
Gavage Challenge of Newborns 21
7. Discussion 22
References 27
vii
-------
ACKNOWLEDGMENTS
The work described here was done by Sandra Wolens, Beverly M. Thompson,
Jack Handley, Wendy L. Schell, and Theodore J. McKenna, assisted by Diane
Bishop and Mary Hugdahl. Ancillary data were contributed by Jean Jensen
Kostenbader, Kenneth D. Kostenbader, Jr., and Gregory S. Naze.
Experimental materials, advice, and techniques were provided by Donald
B. Nelson, Jack Rutledge, Jan Rapacz, Edward H. Bohl, Louis Kasza, Thomas
Yuill, Ralph Anslow, and Norlin Benevenga.
viii
-------
SECTION 1
INTRODUCTION
Once a virus infection has begun, little can be done therapeutically to
influence its outcome. Therefore it is important that virus infections,
including those that might be transmitted via drinking water, be prevented
when possible.
Viruses with a known potential for transmission via drinking water are
those that emanate from the human intestinal tract. They are shed in feces,
are relatively stable outside the host, and must be ingested to produce an
infection. Although the host that produces the virus may not be perceptibly
ill as a result of the infection, such viruses cannot be regarded as part of
the normal intestinal flora—most persons in countries such as the United
States do not generally harbor viruses in their intestines. Thus there is a
finite limit to the amount of virus that might contaminate the environment by
way of feces and sewage. Infection with an enteric virus usually results in
durable immunity; insofar as people in the U.S. are relatively unlikely to
have been infected previously with any given enteric virus, they are more
susceptible than most of the world's population.
Some of these viruses are shed from the pharynx, as well as from the
intestines in feces. The most common route by which enteric viruses are
transmitted is directly from person to person, but contamination of water or
food with feces or wastewater has led to transmission of intestinal virus
infections on occasion. Whether the virus is transmitted by personal contact
or through water or food, little is known of how infection begins in the
human body. That is, neither the identity of the cells in the digestive
tract with which ingested virus first interacts nor the precise mode of
interaction between the virus and these cells has yet been determined.
In the present study, enteroviruses (the most numerous group of the
human intestinal viruses) were selected for peroral administration in drink-
ing water. The objects of the study were to learn how much virus had to be
ingested to cause infection and whether a greater quantity was necessary to
cause disease. Because the aim was to produce disease, virulent virus had
to be used and human subjects were not appropriate. Instead, we chose to
work with swine because their digestive tract is very similar to that of man
and because their homologous enteroviruses resemble those of man in every way
except for their host specificity.
Two serotypes of porcine enteroviruses were used in these experiments.
Doses were administered perorally with drinking water to determine (a) how
much of each serotype, measured in plaque-forming units (PFU), constituted an
oral infectious dose (OID, or dose required to produce infection in the host)
-------
and (b) whether the peroral pathogenic dose (OPD, or dose required to produce
symptoms of disease in the host) was the same or larger. We also studied the
effect of presence or absence of food in the stomach, of administration of
the virus in repeated small doses, of disinfection of virus with chlorine,
and of challenge of newborn pigs by gavage. Important insights into these
questions were gained, even though not all of them were answered explicitly.
-------
SECTION 2
CONCLUSIONS
A given quantity of enterovirus is more likely to produce an infection
if inoculated into a susceptible tissue culture than if ingested by a poten-
tial host organism with drinking water. The OID for waterborne enterovirus
ingested by susceptible young pigs was calculated with the aid of the Poisson
formula to be approximately 1800 to 2500 PFU of porcine enterovirus type 3
(strain ECPO-6) and 600 to 750 PFU of type 7 (strain 05i). The PFUrOID ratio
is a measure of the relative efficiency with which a virus infects a tissue
culture and a whole host organism. It obviously varied with the virus type
or strain, but it did not seem to depend on the host organism. Results of
successive experiments with a given virus strain were comparable, even when
groups of pigs were obtained from different herds. This result suggests that
in the absence of detectable antibody against the virus, the infection of
individuals at low doses of virus depends on chance rather than on individual
differences in susceptibility. In a single experiment with animals that had
received half their breakfast, instead of fasting for at least 12 hr before
being challenged with waterborne virus, the proportion of animals that
became infected was unchanged. Thus virus infectivity might be the same
whether the virus is ingested in water with a meal or on an empty stomach,
and perhaps the same even if the virus is present in the food rather than in
the water. Even though both strains of virus used in this study were pur-
portedly virulent, no perceptible illness resulted from the virus infections
produced with dosages up to 10^ OID.
The risk of becoming infected appeared to cumulate with ingestion of
repeated, small doses of virus. This result indicates that the likelihood
of infection can be predicted by simple probability statistics and that
there is no absolute minimum dose that must be ingested at a single instance
to produce infection.
i
Virus that had been disinfected by chlorine evidently was not reacti-
vated after ingestion. Millions of PFU of enterovirus were reduced to a
level of <1 PFU that was just detectable by cytopathic effect in tissue cul-
tures. Only one of five pigs challenged with this chlorine-treated prepara-
tion became infected.
Two colostrum-deprived newborn pigs, challenged via stomach tube with 20
PFU of porcine enterovirus type 3 at 1% hr of age and then reared with bovine
colostrum in which no antibody against the virus could be detected, did not
become infected. This contrasts with previously published reports that new-
born humans became infected when challenged in this manner with as little as
1 TCD5Q of vaccine poliovirus.
-------
SECTION 3
RECOMMENDATIONS
The results of the present study indicate that enteroviruses infect
their homologous host perorally with lower efficiency than they infect mono-
layer cell cultures of the same species. The risk of infection appeared to
cumulate with ingestion of repeated small doses: this merits further re-
search. The finding that enteroviruses infect tissue cultures with greater
efficiency than they infect the intestines after ingestion with drinking water
is surprising and merits further study. Nothing in the evolutionary history
of enteroviruses should cause them to favor tissue cultures over the intes-
tines as hosts. More should be learned about the barriers to virus infection
that operate at the cell, tissue, and organism levels. The virus type that
appears to be more efficient when peroral infectious doses are expressed as
PFU might instead simply be found to be a very inefficient initiator of
plaques if the number of virus particles inoculated were known in each in-
stance. Therefore, the fate of virus particles in tissue cultures and in the
intestines should be investigated.
Most intestinal virus infections fail to cause overt-illness, as did all
of the infections produced in the present study. Still, the possibility that
a virus infection will cause disease is the ultimate reason for wanting to
prevent transmission of viruses in drinking water. The success of the oral
polio vaccine indicates that there is reasonable empiric knowledge of the
virus-dependent aspects of pathogenesis. What was fcmnd lacking in the pres-
ent study was knowledge of the host-dependent aspects of the disease process.
Among other things, investigation is needed of the influence of age and of
other past and present insults to the host organism on the probability that
illness will result from an intestinal virus infection.
Finally, the infected intestines should be studied as a source of virus
shed from the host organism. The duration of shedding and the quantity of
virus produced, as well as its possible association with sloughed cells,
other fecal solids, and coproantibody, need to be investigated. The influ-
ence of these factors on contact transmission and on transmission through
food and water of virus produced by an infected host needs to be better
understood if control measures are to be applied appropriately and in pro-
portion to their relative significance to public health.
-------
SECTION 4
MATERIALS AND METHODS
Many of the methods used in this investigation evolved significantly
over the 3-year study period. If the method that finally resulted is wholly
adequate to its task, the method may be regarded as one of the products of
the study. However, this was not intended to be primarily a methodology
study, so the methods described are ordinarily reported in their most repre-
sentative form (that is, the version by which another investigator could most
reasonably expect to reproduce the reported results). Suggestions for fur-
ther improvement of methods are included where appropriate.
VIRUSES
Eight serologic types of porcine enteroviruses have been recognized by
the World Health Organization's Programme on Comparative Virology (1). In
surveys of sera obtained principally from swine at slaughter, we learned that
porcine enterovirus type 3 (PE3) and later type 7 (PE7) did not occur with
detectable frequency in the area of Madison, Wisconsin.
PE3 (strain ECPO-6) was supplied by Prof. Edward H. Bohl of the Ohio
Agricultural Research and Development Center at Wooster as the first cell
culture passage from the lumbar cord of a gnotobiotic pig that had paralysis
of the rear legs. The pig had been infected with PE3 at the fourteenth cell
culture passage level.
PE7 (strain 05i) was obtained from Dr. Louis Kasza, now of the Depart-
ment of Pathology, Bureau of Foods, U.S. Food and Drug Administration. Like
the PE3 strain that we used, this agent had been isolated from swine on farms
in Ohio (2). The preparation that we received had been passed seven times in
swine kidney cell cultures at ithe Department of Veterinary Pathology, Ohio
State University, Columbus.
Poliovirus type 1 (P01), strain CHAT, was used as an analog of the por-
cine enteroviruses. It was obtained from the American Type Culture Collec-
tion (3) and had been used as a model virus in our laboratory for several
years.
CELL CULTURES
The porcine enteroviruses are specific for swine cells. Primary cell
cultures (called PSK) were prepared from kidneys of (a) pig fetuses collected
-------
at a local slaughterhouse, (b) young weanling pigs obtained along with those
used in experiments, and (c) uninoculated control animals after experiments.
Animals other than fetuses were killed by electrocution before the kidneys
were collected. Cells were freed from minced and washed kidney cortices by
batch treatment with 0.25% trypsin solution. Washed cells were suspended in
Eagle's minimum essential medium (MEM) with nonessential amino acids and 10%
fetal calf serum, and they were ordinarily seeded in 25-cm2 plastic flasks.
Secondary cultures were produced in some instances by dividing cells harvested
with trypsin from confluent primary cultures.
We also used cultures from an established line of minipig kidney (MPK)
cells obtained from t he American Type Culture Collection (3). Seed cell
stock was propagated in 150-cm2 plastic flasks or in plastic roller bottles.
Virus tests were conducted with confluent monolayers in 25-cm2 plastic flasks.
HeLa cells, prepared in similar fashion, were used when the virus was P01.
Viral cytopathic effects (CPE) were detected by microscopic examination
of cells maintained at 36° to 37°C with 5 ml per 25-cm2 flask of MEM con-
taining nonessential amino acids and 5% fetal calf serum. Plaque assay was
done in 25-cm2 flasks from which the medium had been discarded. The volume
of inoculum was 0.5 ml per flask, and the adsorption period was 90 min at 36°
to 37°C. The residual inoculum was decanted, and the cells were overlaid with
5 ml of MEM plus nonessential amino acids, 5% fetal calf serum, 0.02 mg/ml
neutral red, and 0.7% BBL or Difco purified agar. Flasks in which the agar
medium had solidified were incubated cell-side-up at 36° to 37°C and observed
daily for plaque formation on the basis of the neutral red vital dye. On
occasion, plaques were demonstrated or contrasted semipermanently by removing
the agar overlay and staining the remaining cells with crystal violet in
formaldehyde solution (345 ml deionized water, 4.25 g NaCl, 130 ml 95% ethanol,
2.5 g crystal violet, and 25 ml 37% formaldehyde solution).
SWINE
The animals used during these studies were obtained from two University
of Wisconsin experimental farms and three commercial farms. At least one
other commercial farm had been tried and eliminated as an animal source be-
fore this project began. The essential problem is that porcine enteroviruses
are highly enzootic in our area; the same may be true for other porcine
enteric viruses and opportunistic bacterial pathogens. Only heroic sanita-
tion measures were able to minimize the incidence of these viruses. No farm
was apparently free of porcine enteroviruses all of the time.
Rearing practices and general sanitation differed from farm to farm. In
a well-run farrowing system, pigs would have access only to their own dam
(who would not ordinarily be infected with enterovirus) and could be expected
to remain virus-free during the first few days of their lives (4). At the
University's Mandt Farm, which practiced rigorous sanitation and took ex-
treme measures to avoid the introduction of infectious agents to the herd,
pigs were often (but not always) virus-free when 3 to 3% weeks old, the age at
which we would have preferred to obtain them. The likelihood of obtaining
virus-free pigs of that age from other farms was even smaller.
6
-------
Faces from each animal after it arrived at our animal quarters were
tested for enteroviruses. The time that elapsed before this testing (includ-
ing at least one blind passage) was completed was often 2 weeks, which meant
that ordinarily the animal's 3-day acclimation period, administration of the
challenge virus dose, and most of the subsequent observation period were over
before the results of the first tests were known. Of eight groups of animals
obtained at the age of at least 3 weeks, five groups included enterovirus-
infected animals. Obviously, this percentage was not good enough.
FOSTERING
We were able to borrow a unit in which to raise younger swine. The unit
had been designed and built originally by the University's Department of Meat
and Animal Science. It consisted of a box inside of which were a slatted,
false floor, a thermostatically controlled ventilation system, and dispensing
troughs for milk replacer. Access and observation were permitted by hinged
acrylic panels at the top. We added meshwork partitions that subdivided the
interior into 10 sections. Pigs were left with their dam after birth, to
ensure that they got a full feed of colostrum, and were moved into the fos-
tering unit in our animal quarters at age 2 to 8 days. Even with the colos-
trum for passive immunity, as well as the antibiotic and supportive therapy
by our staff and by University veterinary personnel, diarrheal illness oc-
curred frequently, and it was not always possible to keep the pigs healthy
and normal (or at times, even alive) to the age of 3% weeks.
Nevertheless, the fostering unit accomplished its intended task of remov-
ing the pigs from potential sources of enterovirus infection long enough to
complete testing of the initial fecal specimens before the animals were chal-
lenged experimentally with enterovirus in water. No pig that was used exper-
imentally after residence in the fostering unit was ever found later to have
been infected with an enterovirus at the time of challenge. We usually ob-
tained 10 animals at a time, in the expectation that at least eight animals
would remain available by the date that challenge was scheduled. Any extra
animals could be used as a source of kidneys for cell cultures.
EXPERIMENTAL HOUSING AND FEEDING
The animals were to be housed during the actual experimental period
under conditions that would preclude introduction of viruses other than at
the time of challenge and would prevent the transfer of virus from one animal
to another. Control animals, which had not been challenged with virus, were
included in each experiment in a way that we thought would maximize the like-
lihood of detecting any lapse in technique or in sequestration of the animals.
Four monkey cages were initially adapted in our own laboratory for use in
housing pigs. The cages had been fabricated of stainless steel, part of which
was expanded mesh joined by welding. To limit air intake, the sides and bot-
tom were covered with acrylic and with tempered hardboard, and sealed to the
metal frame and solid panels with gun caulking compound. The mesh top of the
cage was covered with four thicknesses of Owens-Corning glass fiber filter
-------
medium FM-004 (for a total thickness of approximately 5 cm) and sealed at the
edges. An acrylic front panel with an exhaust duct connector was fitted to
each cage by means of four toggle clasps. Room air was drawn through the
glass fiber filter, into the cage, and then out through connector and flexi-
ble ducting to a blower on a manifold mounted on the exhaust duct leading
from the animal room. Air that had been through a cage went directly out of
the entire building; the intake filter system was comparable to that used
(and recommended) by the University of Wisconsin (UW) Gnotobiote Laboratory.
Food and water containers were located entirely inside the enclosure. A
light was hung on the back of each cage so that it would shine in through an
acrylic back panel and enable observation of the animal that was inside.
These enclosures were all that we had during the first four experimental runs
in this project. They represented a great improvement over some that had
been made for us on a trial basis before the project, but they were still
somewhat inconvenient to enter when the animals were fed and watered twice
daily and when the litter pans were cleaned.
Soon after the project was funded, two racks of two enclosures each were
ordered from Germfree Laboratories, Inc., of Miami, Florida. These were
molded as one seamless piece, including top, bottom, sides, and back, of
glass-fiber-reinforced polyester gel. An acrylic door was hinged to the
front, and provision was made for sealing around its edges. Each unit had
its own prefitted glass fiber intake and exhaust filters and an exhaust
blower, which we subsequently connected by flexible ducting to the animal
room's air exhaust duct. When the units were received, they were fitted for
gnotobiotic operation but were not adapted to house any particular species of
animal. Internal arrangements for containment, feeding, and watering of the
animal; lighting; and litter removal were all made to our specifications by
the sheet metal shop of the UW physical plant. These enclosures were first
occupied 8 months after the project began. A few further modifications have
since been made to them.
Pigs in the early runs were weighed when they arrived. Weight gains
could be determined anytime until virus was administered and the individual
enclosures had to be operated on a sealed basis. Food solids were appor-
tioned at the rate of approximately 5% of the animal's weight per day, and
water at a minimum of 10% per day (5). Initially, a standard breakfast com-
prised a premix of 50 g instant nonfat dry milk, 150 g dry instant oatmeal,
0.6 g iodized salt, and one crushed multiple vitamin tablet, to which were
added 14 g peanut butter, 1 medium (VL50 g) banana, one medium hardboiled
egg, and ^250 ml water before serving. A standard supper was made up of 30 g
peanut butter, 130 g cooked ground beef, 75 g instant nonfat dry milk, one
medium (^150 g) banana, and ^250 ml of water. Additional water was available
ad libitum. Later versions of the ration sometimes substituted pasteurized
process cheese for ground beef and an apple for one of the bananas. Feeding
was ordinarily done at approximately 8 a.m. and 4 p.m. each day.
The cages were numbered sequentially (1 through 4 for the adapted monkey
cages and 5 through 8 for the germfree enclosures) and were always entered in
numerical order. Only four enclosures were available during the first four
experimental runs. Enclosure 4 was always the one that contained the control
animal that had not been challenged with virus. When the number of enclosures
8
-------
increased to eight, one control animal was always in enclosure 8, and another
was placed in an enclosure whose number had been selected randomly for that
experimental run only. This scheme was intended to afford maximum likelihood
that any lapse of technique, either in disinfecting enclosures after the
previous run or in twice-daily entries for feeding and watering, would be
revealed. Sterile, disposable gloves were worn and discarded after each
animal was fed, and those who did the feeding had at least a B.S. degree and
experience in microbiology. Nevertheless, control animals were found to be
infected on three different occasions. Two of these three animals were found
to have virus in their 14-day, but not their 7-day, fecal specimens (counted
from the day on which virus challenges had been administered). This result
suggested that lapses had occurred during feeding rather than in disinfecting
the enclosures. However, the animal in position 8, which should have been at
greatest risk from this standpoint, was never the one involved.
CHALLENGE
Except where noted, challenge virus was administered in Madison tap
water (5 ml per animal) that had been boiled previously to eliminate residual
chlorine, as indicated by absence of detectable antiviral activity. In the
first seven runs, the virus was simply thawed on the day of challenge and
diluted to the desired level. Thereafter, in the course of dilution on the
day of challenge, the virus was filtered through a Nuclepore polycarbonate
filter of 50 nm porosity to remove or break up any aggregates that might be
present. Previous testing of stock virus suspensions by filtration at 50 nm
had not produced any decrease in plaque titer, so aggregation of the virus
may not have been a problem under these conditions.
The level of virus in a dose was ordinarily determined on the basis of
dilution from standard virus suspensions. The titers of the standard sus-
pensions were verified periodically bythe plaque technique. A formula pro-
vided by Sobsey (12) yielded an estimate of ±25% of the mean (e.g., 1000 ±
250 PFU) for the 95% confidence interval of these titers.
The animals in each group were challenged in numerical order. The con-
trol pigs received 5 ml of water to which virus had not been added. Each
animal was upended and the head tilted backward. The 5-ml dose was delivered
at the back of the tongue, and the mouth was held shut so that the animal
was obliged to swallow all of It. The enclosure was "sealed" after the
animal was returned to it; thereafter, the inside of the isolator was under
negative pressure with respect to the room, and all air was taken in through
filter, except during feeding, watering, and cleaning.
OBSERVATION
The condition of the animals and whether they had eaten their food was
ordinarily noted at each feeding. A blood sample was collected from the
anterior vena cava (5) either when the animal was put into its isolator at
approximately 3% weeks of age or before challenge 3 days later. Serum from
this sample was used to test for pre-existing passive immunity to the challenge
-------
virus. Fecal samples were ordinarily collected before challenge and on days
7 and 14 after challenge. Another blood sample was collected on the last day
of the run (ordinarily day 14).
We wanted to determine whether the animals became infected by the virus
that was Administered and whether illness resulted from the infection. The
criteria for infection, assuming the absence of virus and of antiviral serum
antibody from before-challenge samples, were to be the presence of virus in
the stool sample from day 7 or 14, or the presence of serum antibody by day
14. As will be discussed below, the virus criterion proved more useful than
the antibody criterion. The criteria for disease were to be signs of central
nervous system involvement (including incoordination, paralysis, or death) or
a rectal temperature in excess of 40.5°C. No animals met these criteria fol-
lowing virus challenge during the entire course of this study.
We were most anxious that neither of our model virus serotypes be intro-
duced to the on-farm swine population in our area: Whatever might have been
the consequences to animal health, introduction of these viruses would have
further complicated procurement of experimental animals. For this reason,
animals from all runs were presumed infected and were killed by electrocution
after the observation period ended. Useful portions of the body were some-
times collected (e.g., kidneys from control animals for tissue culture), and
the remainder was bagged and frozen for later incineration by the UW research
animal resources facility.
TESTING FECAL SAMPLES
Fecal samples were collected from the isolators and frozen at -20°C to
await testing. It was important to obtain samples that were as freshly
voided as possible because feces dried rapidly in the litter pans because of
the constant flow of air through the isolators. Drying, to the extent that
it did occur in these experiments, probably did not affect virus infectivity;
but it sometimes did complicate suspension of the fecal mass in diluent.
Feces (1 to 2 g) were placed in a disposable plastic (styrene) centrifuge
tube, and sufficient phosphate-buffered saline (PBS), to which 2% fetal calf
serum had been added, was placed in the tube to bring the total volume to
10 ml. The tube was agitated repeatedly with a Vortex mixer and refrigerated
(^"C) to allow the fecal mass to soften. Enough 0.1 IT NaOH was added to
bring the pH to 8.8 to 9 (sometimes 0.1 Jtf HC1 to pH 3.5, depending on whether
coproantibody or adsorption to solids seemed of greater concern), and the
fecal mass was broken up with a sterile tongue depressor and agitated several
times with the Vortex mixer. Tubes were then placed in icewater in a sonic
cleaning bath and treated at 20 kHz for 15 min. Fecal solids were sedimented
by 30 min of centrifugation at 16,000 rpm. The supernatant fluid was ad-
justed to pH 7 and passed through Gelman cellulose triacetate filters of 0.45
and 0.20 urn nominal porosity (the latter was sterile). Confluent swine cell
cultures (25 cm2) were inoculated with 0.5 ml or 0.1 ml (two of each) of the
filtered extract. If CPE were not seen after these cultures had been incu-
bated at 36° to 37°C until the uninoculated control cultures deteriorated,
the cultures were frozen and thawed twice, sometimes filtered at 0.20 ym
10
-------
nominal porosity, and inoculated into additional cultures at 0.5 ml per flask.
SEROLOGY
As stated earlier, blood for serological testing was ordinarily drawn
from the pig's anterior vena cava. The blood was allowed to clot for 1 hr at
room temperature and held in the refrigerator for approximately 24 hr more.
The serum was aspirated from the retracted clot and centrifuged for 30 min at
2500 rpm. The supernatant serum from this step was frozen at -20°C in small
volumes for later testing.
The basic test for antibody against virus was plaque-reduction neutrali-
zation. Serum samples were ordinarily tested at dilutions of 1:20, 1:40, and
1:80 in PBS. Virus was diluted to an estimated 100 PFU/ml in PBS with 2%
fetal calf serum. Sometimes serum was diluted 1:50 and a constant serum-
varying virus procedure was used instead. In either case, equal volumes of
diluted serum and virus were mixed and incubated at 37°C for 1 hr in a water
bath. Then the mixtures were inoculated at 0.5 ml per culture into duplicate
25-cm2 cultures and tested by the plaque technique, which was described above.
Sometimes the virus/serum mixtures were tested instead for their ability to
cause CPE in cell cultures maintained with fluid medium.
11
-------
SECTION 5
EXPERIMENTAL PROCEDURES
PERORAL INFECTIVITY EXPERIMENTS
The general purpose of many of the experiments was to determine the OID
for virus in drinking water. The quantity of virus administered to all
animals in a given experimental run was the same. Runs I, II, III, IV, V,
VI, and VIII were intended primarily to determine the OID of PE3 (strain
ECPO-6). The animals were challenged with 1, 10, 100, 1000, 1000, 250, and
250 PFU of PE3, respectively, in this series of experimental runs. Runs IX,
X, XI, and XV were intended to determine the OID of PE7 (strain 05i); doses
were 350, 1000, 250, and 250 PFU per animal, respectively.
PERORAL PATHOGENICITY EXPERIMENTS
The second stated objective of this study was to determine the quantity
of virus that had to be ingested with drinking water to cause disease (the
OPD). Experiments intended to reveal the OPD included run VII for PE3 and
run XIII for PE7. The dose of PE3 per animal was 2.5 x 107 PFU of virus that
had been extracted from the brain of a pig inoculated intracerebrally with
the virus in an attempt to enhance its virulence. Because we had great
difficulty producing useful quantities of either virus by extraction from
brains of intracerebrally inoculated pigs (2,6), the PE7 virus administered
in run XIII was an extract (supernatant fluid after 30 min of centrifugation
at 13000 rpm, filtered through a sterile, 0.2-ym porosity, cellulose tri-
acetate membrane) of a 10% suspension of fourteenth-day feces from run XI.
Each of the challenged animals received approximately 105 PFU of PE7.
EMPTY VERSUS FULL STOMACH
In nearly all experiments conducted during this study, the animals were
deprived of food for approximately 16 hr before the virus dose was adminis-
tered. This empty-stomach condition might have influenced the efficiency
with which virus ingested with drinking water would initiate infection, al-
though it was impossible to predict whether the viurs would be helped or
hindered. Furthermore, the empty-stomach condition would not necessarily be
typical of a human that might ingest virus with drinking water.
12
-------
This question was addressed in experimental run XII. Eight 30-day-old
pigs were fed approximately half their breakfasts (37.5 g dry milk, 10 ml
peanut oil, 57 g cheese, and 500 ml water). When this amount had been con-
sumed, six of the animals were challenged with 5 ml of water containing 1000
PFU of PE3; the two control animals received 5 ml of virus-free water. The
animals were each fed 37.5 g dry milk, 10 ml peanut oil, 1 egg, and 500 ml of
water upon being returned to their isolators.
ITERATED DOSES
The most frequently used test dose of PE3, 1000 PFU, produced infections
in about one-third of the animals challenged, whereas no infections were
shown to result from a single dose of 250 PFU of PE3. This result suggests
that there might be some level of virus below which infection cannot take
place (a minimum infectious dose; 7). On the other hand, a corollary to the
assumption that these viruses showed a Poisson distribution was that the risk
of infection should cumulate with iterated doses. This question is exceed-
ingly important from the standpoint of public policy. The matter was exam-
ined in experimental runs XIV and XVI.
The challenge protocol was the same for both runs, but the numbers of
animals differed. Because of supply problems, only six animals were avail-
able for run XIV. At that time, it was uncertain that an infection could
result from ingestion of as few as 250 PFU of PE3, so it seemed more impor-
tant to expose as many animals to this dose as possible, at the expense of
including the usual two controls. Five of the animals were challenged, and
only one served as a control. On each of four successive days, each of the
five animals received 250 PFU of PE3 in 5 ml of water, and the control pig
received only the water. Once it was established that infection could result
from repeated doses of this size, we scheduled run XVI as a full-scale experi-
ment to confirm this finding. Six of eight animals received the 250 PFU of
PE3 in 5 ml of drinking water on four successive days, whereas the two con-
trol animals received only the 5 ml of water on each of those days.
CHLORINE DISINFECTION
Chlorine is the disinfectant against which the efficiency of others is
measured. Nevertheless, there, has sometimes been concern that inactivation,
as it is demonstrated in tissue culture, might be insufficient to prevent
infection when the virus is ingested by a complex organism such as the human
body. Run XVII examined this question.
PE3 (1 ml containing 3.6 x 107 PFU) was treated in a sonic bath, passed
through a 50-nm porosity polycarbonate filter membrane, and added to 75 ml of
0.2% Clorox solution. This amounted to an applied dose of 50 mg of chlorine
per liter in unbuffered solution at room temperature. The mixture was stirred
for 5 min, and to it was added 24 ml of 0.1 M ^28203. This was stirred for
an additional 10 min and served as the challenge virus suspension for run
XVII. Only seven pigs were available for this experiment, so five were chal-
lenged, and the other two served as controls and received water only. The
13
-------
dose administered (5 ml of the chlorine-treated suspension) contained what
would have been 1.8 x 106 PFU of PE3 in the absence of chemical inactivation.
Repeated testing of the portion that had not been administered to the pigs
revealed some residual infectivity detectable without dilution, but not at
10"1. Titration of the residual virus by the plaque technique was not
successful.
GAVAGE CHALLENGE OF NEWBORNS
Plotkin and Katz (8) reported having produced infections in newborn
humans with as little as 1 TCD5Q of vaccine poliovirus administered by gavage.
We attempted to duplicate this experiment, except that newborn pigs served as
subjects.
Given the difficulty that we had experienced in fostering pigs that had
been weaned after as few as 2 days with the sow, there would have been no
reasonable prospect of success in experimenting with newborn animals had it
not been for the reported finding of Prof. Jan Rapacz (UW Department of Meat
and Animal Science) that newborn pigs could be reared entirely on cow colos-
trum (9). We obtained surplus bovine colostrum from the UW Department of
Dairy Science and stored it frozen while determining that it had no antibody
against PE3. We also had the four newer isolators modified by installation
of a small infrared lamp in each. This modification enabled the animals to
move about and find a place where the temperature suited them. Because of
the amount of individual handling that was required, run XVIII included only
four pigs, two of which were challenged with virus by stomach tube (8). The
inoculum consisted of 2.5 ml of sterile deionized water (to which 2% fetal
calf serum had been added in an attempt to limit losses of virus by sorption
to equipment surfaces) containing a total of 20 PFU of PE3. The animals were
collected as they were born, without allowing them to nurse or to ingest any-
thing. Challenge took place about 1^ hr after the pigs were born. The first
feeding of bovine colostrum was attempted 30 min later. Although newborn
pigs are supposed to be able to drink from a dish, ours did not do well at it
and were bottle-fed instead every 2 hr. Some diarrhea occurred in all four
animals during the 2-wk observation period. The controls had a few more
problems than the challenged animals, but all survived and were in apparently
good condition when the experimental run ended.
14
-------
SECTION 6
RESULTS
The results of experimental runs reported here are divided according to
the kinds of information they were intended to produce. The Roman numeral
designations indicate the true chronological sequence in which the experi-
ments were done. Given the innovative nature of this study, what we learned
from each of the experiments could not always be predicted on the basis of
the stated purpose of the experiment.
PERORAL INFECTIVITY EXPERIMENTS
Run I
Four pigs were obtained from the UW Arlington Experimental Farm. The
project officially began on November 1, 1975. These animals were born on
September 5 and 6, 1975, and were among the last of the Arlington Farm's fall
pigs. Because they were more than 6 weeks old by the time we received them
on November 20, the pigs were larger than we wanted; but no others were avail-
able at the time. We did not have a scale on which animals of their size
could be weighed. Blood samples were taken while the pigs were still at the
farm; a definitive test for antibody against PE3 was not accomplished. Stool
samples taken from all of the animals before challenge were shown to contain
virus, as did those taken on days 7 and 14 after challenge. Because of the
lapse of time involved in detecting the preexisting infection, we did not
know about it in time to avoid doing the challenge. The challenge dose for
run I was 1 PFU per animal, given on November 24. Although this run told us
nothing about the effect of that size dose, we decided to increase the dose
for the next run.
Run II ,
The pigs were born at the Arlington Farm on December 9, 1975 and
received at our Institute December 30 at precisely 3 weeks of age. Their
weights ranged from 5.15 to 6.95 kg. They were challenged on January 2,
1976, with 10 PFU of PE3, which was administered to the animals in isolators
1, 2, and 3. The pig in unit 4 served as the control. Before-challenge
stool samples from pigs 2, 3, and 4 were later shown to contain virus; the
apparent absence of virus from the stool of pig 1 may well have been due to
a false negative test result. Virus was present in fecal samples from all
animals on days 7 and 14 after challenge. Because we had no antiserum with
which to identify PE3 at that time, all virus was assumed to be wild-type
from the Arlington Farm. Again, definitive tests for antibody in the blood
15
-------
serum were not accomplished. This run revealed nothing about the 10-PFU
challenge dose. However, it did lead us to conclude that we would have
difficulty getting virus-free pigs from the Arlington Farm, so no further
animals were obtained from that source.
Run III
The four pigs were born at the UW Mandt Farm on February 20, 1976. They
arrived at our facility on March 13 and ranged in weight from 5.5 to 8 kg.
Three of the pigs were challenged on March 15 with 100 PFU of PE3. No virus
was detected in the before-challenge stool samples or in those taken on days
7 and 14 after challenge. No antibody to PE3 was detected in blood serum
from before challenge and from day 14 after challenge. This was the first
experimental run that produced a clear result: We concluded that the DID
was greater than 100 PFU for PE3.
Run IV
The four pigs were born at the Mandt Farm on March 22, 1976, and were
received at our facility on April 12. They weighed from 7.35 to 9.1 kg. The
challenge dose of 1000 PFU of PE3 was given to animals 1, 2, and 3 on April
14. These animals were found to have been infected with a virus that re-
quired repeated passage in cell culture for positive detection. Virus was
eventually confirmed in before-challenge stool samples from animals 1, 2, and
4, in day-7 (after challenge) stool specimens, from animals 2 and 4, and in
day-14 stool specimens from animals 3 and 4. Very possibly all four animals
were infected with this unknown virus throughout the run. Antibody against
PE3 was absent from the before-challenge and day-14 blood samples. On day 14
(April 28), we thought there was reason to believe that pig 3 had been infec-
ted by the challenge dose. The other three pigs were sacrificed on that day,
and pig 3 was given 'VLO6 PFU of PE3 intraperitoneally and the same amount of
virus per os in an effort to evoke a high titer of antibody against PE3 in
this animal. Blood drawn on day 28 appeared to contain antibody against PE3,
but because of the possible preexisting infection, this could not be assumed
to be a source of monospecific antiserum for virus identification.
Run V
This was the first run in which we could accommodate eight pigs. They
were born on the Mandt Farm on June 11, 1976, and arrived at our facility on
July 1. Their weights, when they arrived, ranged from 4 to 7 kg. The
animals in enclosures 4 and 8-served as controls; on July 5, the other six
pigs were challenged with 1000 PFU each of PE3. Serum antibody against PE3
was absent at challenge, as was any virus in the feces of these pigs. Two
of the challenged animals (numbers 5 and 7) were shedding virus on day 7, and
virus was detected in the feces of pig 7 on day 14 after challenge as well.
Results of tests for antibody against PE3 in the day-14 blood were equivocal.
Run V, then, was the first that began with all animals uninfected and pro-
duced some clear-cut infections as a result of the challenge. Two of the six
animals that had received 1000 PFU of PE3 were infected, whereas 100 PFU of
the same virus had not infected any of the three pigs challenged in run III.
A simple formula for estimating the OID is that given by Sobsey (12) for
16
-------
estimating a most probable number of virus from a data set such as this. It
derives from the Poisson formula and is written: m = -ln(q/n), which is to
say that the most probable number of OID's present is the negative Napierian
logarithm of the ratio between the number of negative results (q) and the
number of animals inoculated (n). In this instance, the ratio of 4 animals
uninfected among 6 inoculated indicates that the 1000 PFU dose was equivalent
to 0.4 OID and, therefore, that the OID is approximately 2500 PFU of PE3.
The next experiment was to be done with 0.1 OID, or 250 PFU of PE3 per animal.
Run VI
The eight pigs were born'at the Mandt Farm on September 14, 1976.
They weighed from 5.7 to 6.9 kg when we received them on October 5. The
animals in isolators 6 and 8 served as controls in this run. The other six
pigs were challenged on October 8 with 250 PFU of PE3 each. By agreement
with the project officer, these animals were to be observed for 4 weeks after
challenge. Unfortunately, virus was found to be present in feces collected
from all eight pigs 2 days before challenge. Although a few of the inter-
mediate samples did not contain detectable virus, feces from all eight
animals had virus at the end of the 4-week period. We did not then have
monospecific antiserum with which to determine whether any of this end-of-
run virus was PE3. Obviously, the experiment needed to be repeated.
Run VIII
The eight pigs were the first obtained from a commercial source: They
came from the Kaltenberg Farm near Waunakee, Wisconsin. Four of them had
been born on January 24, 1977, and the other four on January 27. They were
4 and 7 days old, respectively, when we received them on January 31. Their
weights were not recorded. They were held in the four newer, larger iso-
lators (two pigs per unit) until challenge on February 24, when they were
separated into all eight units. The pigs in units 3 and 8 were used as
controls, and the other six received 250 PFU of PE3. Fecal specimens taken
5 days before challenge, as well as those taken weekly for 4 weeks after chal-
lenge, were negative for virus. Though the evidence that this series of runs
supplies is a limited basis for the estimation of an OID for PE3, runs XII,
XIV, and XVI, described in the following sections, are also pertinent to this
question.
Run IX
This was the first effort to determine the peroral infectious dose for
PE7. Eight pigs born on April 5, 1977, at the Mandt Farm were received on
May 2 at 4 weeks of age. Weights were not recorded. The pigs in enclosures
4 and 8 were used as controls, and the remaining animals were challenged on
May 6 with an estimated 350 PFU of PE7 (there were problems with the assay of
this particular preparation). Three animals were later shown to have been
infected at the time of challenge and to have remained so during the follow-
ing 2 weeks: virus was present in the feces of animals 1, 2, and 5 on the
challenge day and on days 7 and 14 thereafter. These three pigs apparently
had been infected before they left the Mandt Farm. None of the other pigs
became infected as a result of the experimental dose. This was the last
17
-------
experiment in which we attempted to obtain animals 3 or more weeks of age and
challenge them within a few days after they arrived at our facility.
Run X
Eight pigs born on August 19 and 20, 1977 at the Kaltenberg Farm were
received on August 22 and placed in the fostering unit. Various health prob-
lems occurred during fostering, culminating in the death of one of the
animals. At least those still alive on the challenge day '(September 15)
were without virus. The control animals for this run were in isolators 7 and
8, and isolator 3 was vacant. Pigs 1, 2, 4, 5, and 6 received 1000 PFU of
PE7. Virus was found in the feces of each of these five animals on days 7
and 14 after challenge. This instance was one of the very few in which
definitive immunologic responses could be demonstrated: None of the seven
pigs had detectable serum antibody before challenge, and all of the five
animals that were shown to have been infected on the basis of virus in their
feces were also found by day 14 to have produced antibody that reacted
strongly in the plaque-reduction neutralization test. Two of these animals
received another 10 PFU of PE7 intramuscularly on day 14 and were bled from
the heart on October 18 as sources of typing antiserum for this virus.
Run XI
The eight animals were born at the Mandt Farm on October 5, 1977, and
received on October 7. These animals prospered in the fostering unit. They
were transferred to the isolators, and those in units 7 and 8 were selected
as controls. The others received 250 PFU of PE7 on October 28, at which time
none was evidently infected. On day 7 after challenge, virus was detected
in the feces of pigs 3 and 6, and on day 14, of pigs 3, 5, 6, and 7 (the
latter was a control). The neutralization test results were anomalous. It
seems likely that at least pigs 3 and 6 were infected by the challenge dose,
that pig 7 was infected as a result of cross-contamination during the second
week, and that pig 5 may have been infected in either way.
Run XV
The pigs were born on April 14, 1978, at the Mandt Farm. All of 10
animals survived the fostering period, which began April 19, but only eight
were used (one per isolator). The animals in units 4 and 8 served as con-
trols; the others were challenged with 250 PFU of PE7 on May 17. No virus
was detected in the pre-challenge stools. Virus was present in the fecal
samples on day 7 from pigs 6 and 7 and on day 14 from pig 4 (a control) and
pig 7. The problem of cross-contamination leading to the infection of an
uninoculated control animal was not encountered in runs XII, XIII, or XIV or
in any subsequent run, although there had been a similar occurrence in run
VII, which is described below. Two of the six animals challenged with 250
PFU of PE7 seem certainly to have been infected as a result. The results of
this run and the others in this series indicate that the OID of PE7 is ap-
proximately 600 to 750 PFU.
18
-------
PERORAL PATHOGENICITY EXPERIMENTS
Experimental runs VII and XIII were designed explicitly to induce dis-
ease in the pigs. Specific means of enhancing viral pathogenicity were
tested, inasmuch as mere infection with the waterborne virus had not resulted
in illness. The histories of the two model viruses indicated that they might
produce central nervous system symptoms.
Run VII
The eight pigs were born on the Mandt Farm on October 7, 1976, and were
received at our facility on November 8 at the age of 4% weeks. The inoculum
was derived from the cerebellum and medulla of a colostrum-deprived pig that
had been inoculated intracerebrally by Thomas Yuill of the UW Department of
Veterinary Science at 1 day of age and sacrificed at 7 days of age. The
tissue suspension was treated with chloroform, with sodium dodecyl sulfate,
and with ultrasound, then clarified by passage through a 5-um porosity cellu-
lose triacetate filter and purified on an Amicon PM30 ultrafilter membrane.
Titrations of this preparation in MPK and PSK cell cultures yielded estimates
of 4.7 x 106 and 5.0 x 106 PFU/ml, respectively. Thus the animals challenged
with 5 ml of this preparation on November 10 received approximately 2.5 x 10
PFU, which is 10,000 times our estimate of the OID for PE3. None of the pigs
was shedding virus in its feces before challenge. The animals in enclosures
5 and 8 served as the controls in this run. On day 7 and again on day 14
after challenge, virus was present in the feces of all six challenged animals
and in that of control 5. This run (VII) was the first instance in which a
supposedly uninoculated control became infected, and the only instance in
which the infection was detected as early as day 7. Given the high level of
virus used in this experiment, it seems possible that this animal was ex-
posed while it was receiving its sham challenge, rather than as a result of
cross contamination during its stay in the isolator. More important, some
diarrhea and vomiting were seen on occasion in some of these animals during
the 2-week observation period, but no signs were observed that were specifi-
cally indicative of central nervous system involvement.
Run XIII
Healthy, virus-free pigs were very difficult to obtain in January and
February of 1978. Eleven 2- to 3-day-old pigs were obtained on January 25
from a commercial (Noltner) farm. By February 9, all but three of these had
died in the fostering unit. The three survivors were moved to temporary
quarters, and on February 10, 10 2-day-old pigs were received from another
commercial (Double H) farm. By March 7, all but two of these had died in the
fostering unit. These five survivors made up the group. The older animals
were housed in isolators 1, 2, and 4, and the younger two were put into units
3 and 8 (the latter was the uninoculated control). The challenge preparation
of PE7 was extracted from day-14 feces of run XI as described above. This
procedure was chosen because we wanted to use virus passed directly from
animal to animal, without intervening cell culture passages, and we had been
unable to produce a satisfactory preparation of PE7 by intracerebral inocu-
lation and extraction of the brain. The titer of the fecal extract was
approximately 105 PFU per ml, and each of the four challenged animals received
19
-------
5 ml of this, which is at least 3000 times the DID for PE7. None of the
animals was shedding virus in its feces when challenged. All four challenged
animals (but not the control) shed virus on days 7 and 14 thereafter. No
significant illness was noted in any of these animals during the 2-week
observation period, despite the large quantity of virus that had been admin-
istered.
EMPTY VERSUS FULL STOMACH
Run XII
The eight pigs were born on the Kaltenberg Farm on December 4, 1977. Of
10 animals received on December 7, one died by December 11, and the other
nine grew well until the challenge day, January 3, 1978. The animals in
isolators 2 and 8 were the controls for this experiment. The specifics of
the challenge are described in the previous section. The challenge dose
consisted of 1000 PFU of PE3 in 5 ml of water, administered between halves
of the pigs' breakfast on January 3. No virus was found in the preinoculation
fecal specimens from any of these pigs; virus was present in the feces of pig
5 on post-challenge day 7 and of pigs 3 and 5 on post-challenge day 14. This
finding of infection in two of six pigs challenged when they had food in
their stomachs is identical to that obtained in run V, when 1000 PFU of PE3
was administered to six pigs with empty stomachs.
ITERATED DOSES
Run XIV
The animal health problems mentioned earlier continued into March 1978.
Ten pigs born March 20 and 21 on the Mandt Farm were received on March 22 and
had all died in the fostering unit or been sacrificed moribund by March 27.
The fostering unit was disinfected and refilled with 10 pigs (born March 26
at Mandt) on March 28. Six of these survived. The challenge procedure was
as described in the preceding section. The pig in unit 8 served as a control,
and the other five received 250 PFU of PE3 per animal on April 24, 25, 26,
and 27. No virus was detected in the prechallenge fecal specimens. Post-
challenge days 7 and 14 were numbered from the last of the four daily dosings.
Virus was found in the 7-day samples from animals 1 and 7 and in the 14-day
sample from pig 7. Thus two of five animals challenged with iterated doses
of 250 PFU of PE3 became infected, whereas, none of six animals challenged
once with 250 PFU of PE3 had been infected in experimental run VIII.
Run XVI
The animals were born at the Kaltenberg Farm on July 9 and 10, 1978.
Eleven animals were received in the fostering unit on July 17. All of them
survived until the time of challenge, although only eight could be used. The
animals in enclosures 5 and 8 served as controls, and the other pigs were
challenged in the same manner as in run XIV on August 7, 8, 9, and 10. None
of the pigs was shedding virus before challenge. On day 7 after the last
20
-------
dose, there was virus in the feces of animals 1, 4, and 6; these and pig 7
were found to be shedding virus on day 14. Clearly, iteration of small virus
doses can result in infections.
CHLORINE DISINFECTION
Run XVII
Twelve pigs born on August 1, 4, and 5, 1978, were received on August 9.
Five of these died during the period in the fostering unit, but the seven
remaining were sufficient to do almost a full-scale experiment. Preparation
of the chlorine-treated PE3 suspension is described in the preceding section.
In the absence of chlorine treatment, the dose per animal would have been
1.8 x 106 PFU. The suspension was administered to the pigs in all units but
1 and 8 (which were occupied by the controls) and 5 (which was vacant). None
of the animals was shedding virus at the time of challenge, but virus was
found in the feces of pig 3 on days 7 and 14 thereafter. A trace of residual
infectivity was detected in the undiluted inoculum by CPE, but not by plaque
formation, in cell culture. This was the only instance in which the sen-
sitivity of the whole organism seemed to approach that of cell cultures.
GAVAGE CHALLENGE OF NEWBOPJSS
Run XVIII
The four pigs were born at approximately 10 a.m. on October 2, 1978, at
the Kaltenberg Farm. They were brought immediately to our facility, without
nursing, and two of them were dosed by stomach tube with 20 PFU of PE3 at
11:30 a.m. These were placed in enclosures 5 and 7, and the sham-inoculated
controls were housed in units 6 and 8. The animals were reared principally
on cow colostrum, as described in the previous section. There were no pre-
challenge fecal specimens, and no virus was detected in the samples collected
on days 7 and 14. If newborns are inordinately sensitive to challenge by
gavage, this admittedly limited experiment did not show it.
21
-------
SECTION 7
DISCUSSION
The goal of this study was to produce information that would help to
assess the significance of viruses in drinking water as a threat to human
health. It has been established epidemiologically that viruses are occasion-
ally transmitted through drinking water: Reported outbreaks of illness
associated with public water supplies have been due to "breaks" in treating
water derived from known contaminated sources or to contamination of finished
water during distribution (10). Nevertheless, there has been manifest con-
cern that viruses could be transmitted through drinking water and cause water-
borne disease that could go unrecognized and unreported (11). Therefore, it
is most unfortunate that studies done years ago concerning peroral infection
by the polioviruses were discontinued once the procedures for administering
the oral polio vaccine were established. Many questions significant to
environmental virus transmission were left unanswered.
The major objectives of the study were to determine the OID for entero-
viruses ingested with drinking water and to determine whether the OPD was
greater. Some corollaries to these objectives were also included by agree-
ment with the project officer. Because we had not developed the capability
to enumerate virus particles, virus doses were expressed in PFU measured in
monolayer cell cultures. Host responses to the challenges would be quantal
(infection or none, disease or none), so the Poisson distribution that is
used in most probable number (MPN) statistics was assumed (12). Though a
variable cannot be proven to obey the Poisson distribution (at least not in
a study of this scope), no evidence was adduced to the contrary in the course
of these investigations.
A few peroral infection studies have been conducted with the attenuated
oral poliomyelitis vaccine; the statistical approach in these was not stated
explicitly (10). However, it has been suggested that the infecting effi-
ciency of the poliovirus was impaired by attenuation (8). This appears to
be an after-the-fact assertion based on no direct experimental evidence, but
an experimental test of the hypothesis would require administration of viru-
lent virus and would essentially rule out the use of human subjects. We
chose swine and their homologous enteroviruses as our experimental system
because swine of known background could be obtained locally, whereas most
experimental primates (if not embargoed as endangered species) are captured
in the wild in tropical countries and have an extremely varied virologic
history by the time they can be brought to an American laboratory. A great
advantage of using porcine enteroviruses in these studies is that the swine
enteroviruses do not infect human or primate cells and are therefore no
threat to the personnel handling the virulent viruses or the infected swine.
22
-------
Research with swine did pose some problems that we had not anticipated.
We had intended to get conventionally reared pigs of approximately 3% weeks
of age, acclimate them to the experimental isolators, and then challenge them.
Conventional rearing, especially nursing the sow at birth to get the colos-
trum, seemed important because some evidence exists that pigs deprived of
colostrum are immunologically impaired (13). Enteroviruses abound among
swine in the Midwest (14), and though the same may not hold true in more arid
areas of the United States, we found that virus-free swine were very hard to
obtain at 3% weeks of age, even from farms that appeared to practice quite
rigorous sanitation. Furthermore, pigs were not always available at 2 to 8
days of age from farms that we knew practiced good sanitation. When such pigs
could be gotten, they did not always survive in our fostering unit to an age
where they could be used experimentally. We received a great deal of veteri-
nary consultation when our pigs were dying, but this proved to be of little
use with regard to diagnosis, treatment, or prevention. The use of the fos-
tering unit proved valuable in that we no longer had to conduct our experi-
ments with pigs that would later be found to have been infected with virus
before they were challenged. It would have been very helpful if the protec-
tive effects of bovine colostrum had come to our attention sooner, but the
popular report we cite on this subject did not appear until June, 1978 (9).
Our original estimate of the OID for porcine enterovirus 3 (PE3) was
based on the observations that no animals became infected after single doses
of 100 or 250 PFU (runs III and VIII, respectively), but that two of six
animals challenged with 1000 PFU (run V) became infected. As explained
earlier, the most probable number of OID's in a given dose can be calculated
as the negative Napierian logarithm of the proportion of recipient animals
which does not become infected (12). The proportion of animals that did not
become infected upon ingesting 1000 PFU of PE3 in run V was 0.67, which
indicates that this virus dose was 0.4 OID, and that the OID is 2500 PFU for
PE3. This is a small number of observations on which to base the estimate;
but run XII, in which two of six pigs became infected when challenged with
1000 PFU of PE3 after eating half their breakfast, appears to support it. A
dose of 0.1 OID (250 PFU, in this case) should have a 0.9 probability of not
infecting the recipient. However, if a host receives this dose on four suc-
cessive occasions, the cumulated probability of not becoming infected becomes
(0.9)\ or approximately 0.66. Of 11 pigs challenged this way in runs XIV
and XVI, six became infected; so the portion that did not was 0.45. This
does not differ significantly from the prediction that was based on an esti-
mated OID of 2500 PFU for PE3. I But if the results were pooled for all experi-
mental runs in which the total dose of PE3 was 1000 PFU (viz., runs V, XII,
XIV, and XVI), and if a new estimate were made based on the findings that 10
out of 23 challenged animals became infected, the new OID estimate would be
1800. The exponential relationship of dose to response in the Poisson formula
is such that only order-of-magnitude differences need be taken seriously:
1800 and 2500 PFU are not very different quantities of virus. On the other
hand, the OID estimate for PE7, which is 600 to 750 PFU based on runs X, XI,
and XV, may well be significantly lower. Because we did not count virus
particles in this study, it is hard to be sure of the reason that PE7 appears
to cause infection more efficiently than PE3. It is possible that the number
of viral particles per OID is the same for both viruses and that PE7 seems
more efficient in causing peroral infections only because it is less efficient
in infecting tissue cultures (i.e., if there are more particles of PE7 than
23
-------
of PE3 per PFU, the number of particles per DID may be the same even though
more PFU of PE3 are required.
Though we did not undertake a count of virus particles, we believe that
the challenge suspensions were essentially monodisperse. Tests of the virus
preparation used in early runs showed no significant loss of titer upon fil-
tration through a 50-nm porosity polycarbonate filter. Thereafter, each dose
was filtered at 50 nm porosity at the time of challenge. Aggregated virus
would perhaps have been more representative of natural conditions, but we
were unable to produce a stably aggregated suspension with which to do com-
parative tests.
We did not choose to assume at the outset of the study that cell cultures
were more sensitive than the whole host organism to enteroviruses, even
though there were indications in the literature that this was so (10). Now
that yet another manifestation of a difference has been seen, we would like
to understand the basis of the difference. Within reasonable limits, litter-
to-litter, farm-to-farm, and perhaps individual variations (excluding immu-
nity) seem to exert little influence, nor does the use of pigs with full or
empty stomachs. The difference between the two porcine enteroviruses may
well have been significant, but we cannot tell whether two strains of a
single serotype would have differed as much in efficiency as these two, which
happened to be of dissimilar serotypes. The effects of extremely early
exposure to virus, colostrum deprivation, and challenge by gavage to a pris-
tine stomach remain to be demonstrated. If these conditions are important,
our small-scale trial (run XVIII) did not show it. Although the evidence
from our study is somewhat limited, it seems to indicate that the OID for
enterovirus ingested with drinking water is virus-dependent and host-indepen-
dent, at least in the absence of antibody against the virus. Unfortunately,
our efforts to measure the animals' antibody responses to infection yielded
erratic results. The absence of passively acquired antibody to the experi-
mental virus before challenge was general, but active antibody responses were
difficult to demonstrate within 2 weeks. Because testing for virus in the
stool provided a straightforward basis for determining infection, tests for
serologic response were not included in the last few runs. We had hoped that
providing each animal with colostrum would guarantee a prompt, active anti-
body response; but there is some doubt that this is relevant. In any case,
ingested antibody may not provide even local protection in the intestine to
animals 4 to 6 weeks old. Contents of stomach and intestine from sacrificed
control pigs were able to reactivate poliovirus that had been neutralized
with coproantibody (15): This result indicated that ingested antibody in the
lumen of the digestive tract is likely to afford very limited protection.
Animals held beyond the 2-week observation period and exposed to further
large quantities of the same virus provided antisera for use in virus identi-
fication. Also, antibody against PE3, labeled with fluorescein isothio-
cyanate, was used to locate virus in frozen sections of intestine from in-
fected pigs and of infected ileal explant cultures (16,17). Virus antigen
appeared to be located in the lamina propria and not in the epithelium of the
villi. This finding suggested that the cells the virus would infect were not
directly exposed in the intestinal lumen, and it might partly explain the
inefficiency with which virus infection occurs upon ingestion (J. Jensen
24
-------
Kostenbader and D. 0. diver, unpublished data, 1977). Alternate observa-
tions have been reported, however (17).
The effort to determine the OPD was quite unsuccessful. Although both
viruses were able to cause paralysis in colostrum-deprived, germfree pigs
(2; E. H. Bohl, Ohio Agricultural Research and Development Station, Wooster,
1976), none of the animals infected under the conditions of our experiments
became ill. Passage of PE3 in the brain of a pig did not have any percep-
tible effect on pathogenicity via the digestive tract. Our hopes in this
regard smack of Lysenkoism and probably do not have a well-founded theoretical
or experimental basis. Neither was dose demonstrably a factor. The 2.5 x
107 PFU of PE3 administered to each animal in run VII could be the equivalent
of ingesting at least 2.5 g of feces from an infected animal (18). If patho-
genesis required a dose larger than this, enteroviral illnesses would be even
rarer than they are. It seems reasonable to infer that even virulent virus
can cause disease upon ingestion only if the host organism is being insulted
in some other way at the time. We are now doing some exploratory experiments
on this subject which we hope will lead to another project.
The irrevocability of chlorine inactivation was not as neatly demon-
strated as we had hoped. The dose per animal, had it not been for the
chlorine treatment in run XVII, was 1.8 x 106 PFU of PE3 per pig. If virus
that was inactivated by chlorine in the sense that it could no longer infect
a cell culture were still infectious per os, as has been suggested (19), all
five of the challenged pigs could have been infected. Instead, one of the
five became infected, which suggested that a fraction of an OID was present
in the inoculum. Residual infectivity was detectable in tissue cultures, but
not at a level that could be measured by the plaque technique. It does not
seem reasonable to attribute this result to the presence of aggregates in the
suspension, inasmuch as the suspension had been passed through a 50-nm filter
membrane just before the chlorine treatment. The most conservative e:q>lana-
tion seems to be that chlorine inactivation is irrevocable and that this
single infection was a chance result at this extremely low level of virus
infectivity. Further studies in this area are needed.
Our results appear to show that ingestion with drinking water (and,
perhaps, food) is not a very efficient way for enterovirus to initiate infec-
tion in vivo. Assuming that an infected individual might shed the virus at
approximately 106 PFU per gram of feces and taking 103 PFU as a value inter-
mediate between the two experimentally-determined OID's, the occurrence of a
common source outbreak (one in which several persons are ill at the same
time) would require that the vehicle be contaminated to a degree that each
susceptible consumer would ingest the equivalent of at least 1 mg of feces.
This level of contamination is conceivable in food mishandled during final
preparation, but it seems unlikely in public water supplies, except in
instances of gross post-treatment contamination during distribution of fin-
ished water.
On the other hand, an isolated enterovirus infection (and perhaps ill-
ness) might occur in a consumer of drinking water contaminated at levels far
below that detectable by the most powerful concentration methods available
(10,20). Furthermore, this hypothetical infection might give rise to others
25
-------
by direct (contact) transmission. However, contact transmission appears to
be the most prevalent (and perhaps least understood) mode of transfer for
enterovirus infections (21); and the vast majority of virus donors in in-
stances of contact transmission have undoubtedly been infected by contact
with a previously infected person, rather than by ingestion of virus in water,
Thus a susceptible member of a community is probably at greater risk of con-
tracting an infection directly from a family member, a neighbor, or an in-
fected visitor from another community than from a community water supply
hypothetically contaminated with virus that has escaped inactivation as the
virus level in the treated water approached zero asymptotically. An infected
member of a community is a far greater threat to his contacts than to those
to whom his infection could be transmitted only via the water route (21).
This fact is important because a community has limited resources for pre-
venting virus transmission. It would be detrimental to the public health
for the community to dissipate its resources on extraordinary measures in
drinking water surveillance and treatment that go beyond the relative hazard
that drinking water represents. We believe that our findings provide the
beginning of a new basis for judging relative hazards.
26
-------
REFERENCES
1. Cottral, G. E. , ed. Picornaviruses. In: Manual of Standardized Methods
for Veterinary Microbiology. Cornell University Press, London, 1978.
pp. 249-272.
2. Kasza, L. Swine Polioencephalomyelitis Viruses Isolated from the Brains
and Intestines of Pigs. Am. J. Vet. Res. 26(110):131-137, 1965.
3. American Type Culture Collection. Catalog of Strains-II. First Edition.
American Type Culture Collection, Rockville, Maryland, 1975. 241 pp.
4. Singh, K. V., and E. H. Bohl. The Pattern of Enteroviral Infections in
a Herd of Swine. Can. J. Comp. Med. 36(3):243-248, 1972.
5, Mount, L. E., and D. L. Ingram. The Pig as a Laboratory Animal.
Academic Press, London and New York, 1971. 175 pp.
6. Kasza, L., J. Holman, and A. Koestner. Swine Polioencephalomyelitis
Virus in Germfree Pigs: Viral Isolation, Immunoreaction, and Serum
Electrophoresis. Am. J. Vet. Res. 28(123):461-467, 1967.
7. Westwood, J. C. N., and S. A. Sattar. The Minimal Infective Dose. In:
Viruses in Water. G. Berg, H. L. Bodily, E. H. Lennette, J. L. Melnick,
and T. G. Metcalf, eds. American Public Health Association, Washington,
D. C., 1976. pp. 61-69.
8. Plotkin, S. A., and M. Katz. Minimal Infective Doses of Viruses for
Man by the Oral Route. In: Transmission of Viruses by the Water Route.
G. Berg, ed. Interscience (John Wiley & Sons), New York, 1967. pp.
151-166.
9. Anonymous. Cow's Colostrum Can Save Runts, Extra Pigs in Big Litters.
National Hog Farmer, June, 1978. p. 57.
10. Safe Drinking Water Committee (National Research Council). Drinking
Water and Health. National Academy of Sciences, Washington, D. C., 1977.
939 pp.
11. Goldfield, M. Epidemiological Indicators for Transmission of Viruses
by Water. In: Viruses in Water. G. Berg, H. L. Bodily, E. H.
Lennette, J. L. Melnick, and T. G. Metcalf, eds. American Public Health
Association, Washington, D.C., 1976. pp. 70-85.
27
-------
12. Sobsey, M. D. Field Monitoring Techniques and Data Analysis. In:
Virus Aspects of Applying Municipal Waste to Land. L. B. Baldwin, J.
M. Davidson, and J. F. Gerber, eds. University of Florida, Gainesville,
Florida, 1976. pp. 87-96.
13. Bustad, L. K. Pigs in the Laboratory. Scientific American 214(6) :94-
100, 1966.
14. Kostenbader, K. D., Jr., and D. 0. Oliver. Quest for Viruses Asso-
ciated with Our Food Supply. J. Food Sci. 42(5):1253-1257, 1268, 1977.
15. Cliver, D. 0., and K. D. Kostenbader, Jr. Antiviral Effectiveness of
Grape Juice. J. Food Protection 42(2):100-104, 1979.
16. Derbyshire, J. B., and A. P. Collins. The Multiplication of Talfan
Virus in Pig Intestinal Organ Cultures. Res. Vet. Sci. 12(4):367-370,
1971.
17. Dolin, R., N. R. Blacklow, R. G. Wyatt, and M. W. Sereno. Detection and
Localization of Viruses in Human Fetal Intestinal Organ Cultures by
Immunoflorescence. Infect. Immun. 6(6):958-964, 1972.
18. Singh, K. V., S. J. McConnell, E. H. Bohl, and" J. M. Birkeland. Fecal
Excretion of Porcine Enteroviruses following Experimental Infection of
Pigs. Virology 12(1):139-141, 1960.
19. Melnick, J. L. Discussion. In: Virus and Water Quality: Occurrence
and Control. V. Snoeyink and V. Griffin, eds. University of Illinois
Engineering Publications, Urbana, Illinois, 1971. p. 111.
20. Melnick, J. L. Detection of Virus Spread by the Water Route. In:
Virus and Water Quality: Occurrence and Control. V. Snoeyink and V.
Griffin, eds. University of Illinois Engineering Publications, Urbana,
Illinois, 1971. pp. 114-125.
21. Lennette, E. H. Problems Posed to Man by Viruses in Municipal Wastes.
In: Virus Aspects of Applying Municipal Waste to Land. L. B. Baldwin,
J. M. Davidson, and J. F. Gerber, eds. University of Florida,
Gainesville, Florida, 1976. pp. 1-7.
28
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
i. REPORT NO.
EPA-600/1-80-005
3. RECIPIENT'S ACCESSIOf*NO.
4. TITLE AND SUBTITLE
5. REPORT DATE
Infectivity and Pathogenicity of Enteroviruses
Ingested With Drinking Water
.January 1_980 i_s_sui_ng da_te_
6. PERFORMING"oR~GANIZATlbfi CODE
7. AUTHOR(S)
Dean 0. Oliver
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Food Research Institute & World Health Organization
Collaborating Centre on Food Virology
University of Wisconsin-Madison
Madison. Wisconsin 53706
10. PROGRAM ELEMENT NO.
1CC824
11. CONTRACT/GRANT NO.
R-8Q3986
12. SPONSORING AGENCY NAME AND ADDRESS
Health Effects Research Laboratory _ cinn OH
Office of Research and Development
U. S. Environmental Protection Agency
Cincinnati. Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final; 10/6/7S-1 /n/7Q
14. SPONSORING AGENCY CODE
EPA/600/10
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This study was designed to examine the relationship of waterborne enteroviruses to
infections and disease. Young weanling swine and their homologous enteroviruses were
chosen as the model system: The porcine digestive tract is like that of man, but pigs
can be handled under more closely standardized conditions than humans or other primates
Known quantities of two enteroviruses were administered in 5 ml of drinking water in
such a way that the subjects were obliged to swallow all of it. The intact animal was
found to be about 1000 times (600 to 750 for one virus and 1800 to 2500 for the other)
less likely than the tissue cultures to be infected by a given quantity of enterovirus.
The ratio did not depend on whether the animals were fed just before challenge. The
probability of infection was cumulative with iterated small doses: this indicated that
there was, in the strict sense, no minimum infectious dose. None of the infected
animals became ill, despite the reported virulence of the challenge viruses. Chlorine
treatment of a concentrated virus suspension, which reduced infectivity to a level de-
tectable by cytopathic effect but not plaque formation in tissue culture, left enough
virus to infect one of five challenged subjects. Neither of two colostrum-deprived
pigs, challenged by stomach tube with 20 plaque-forming units of enterovirus at 1% hr
of age, became infected.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
enterovirus
water supply
public health
virus infective dose
minimum infective dose
pig model system
06M
57K
68D
18. DISTRIBUTION STATEMENT
Release to public
19. SECURITY CLASS (ThisReport)
ed-
11T Y C LASS (TMS
21. NO. OF PAGES
37
20. SECURITY CLASS (ThTspage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
: «o -45Y-146/SS*
29
------- |