r
EPA/600/A-94/202
A Fish Model as an Indicator for Teratogenic Substances
by Douglas P. Middaugh and Michael J. Hemmer
U.S. Environmental Protection Agency, Environmental Research Laboratuj,
Sabine Island, Gulf Breeze, Florida 32561, U.S.A.
A recent expansion has occurred in development of in vivo methods for teratogenesis
(Birge etal., 1983; Whitby and Flyim, 1987). This effort has focused on a number of
potential nonmammalian systems, and fish have received considerable attention (Laale
and Lemer, 1981; Collins, 1987; Weis and Weis, 1987; Solomon and Faustman,
1987). We have developed a fish embryo model as an indicator for teratogenic
substances. Our procedure utilizes naturally spawned embryonic inland silversides,
MenkUa beryllina. This euryhaline fish is ubiquitous in estuaries, coastal rivers and
numerous lakes in the southeastern United States and is also found in the Mississippi
River basin (Sisk and Stephens, 1964; Gomez and Lindsay, 1972; Johnson, 1975;
Chernoff etal., 1981).
Embryo Acquisition
Adult M. beryllina are easily maintained in the laboratory. A population of 25 males
and 25 females collected from a low salinity (0-5 °/oo) estuarine site in early April
(20°C) and maintained at 5°/oo 25°C produced 300 to 2000 embryos daily for
approximately 5 months (Middaugh ex al., 1985). Thereafter a new population of
young adults collected in late September, produced embryos throughout the winter.
Holding tank configuration, environmental and nutritional requirements for brood
stock adults have been summarized by Middaugh etal. (1986).
Test Procedure
Natural unpolluted seawater used in all tests is treated by pre-filtration through a 20 jim
polypropylene filter. It is then passed through a 6 ^im filter, diluted to 5°/oo salinity
with deionized water and autoclaved. Stock solutions are prepared by diluting the
analytical grade test chemical in deionized water, acetone, or another appropriate
solvent.
A range finder test may be employed using a reduced number of embryos and a wide
range of concentrations to bracket the potential teratogenic response levels. It is
preferable that a concentration range be chosen that limits the lethal response at the
highest treatment levels while enhancing the teratogenic response at the lower treatment
levels.
Just prior to beginning a test, aliquots of stock solution are added to 1.0 litre of
autoclaved saline water (5°/oo) to yield the desired nominal exposure concentrations.
We generally use 5 concentrations of the suspected teratogen plus a seawater control If
116

-------
an organic solvent such as acetone is employed, an additional earner control must be
maintained with the concentration of the solvent equal to that in the highest exposure
concentration of teratogen. The quantity of solvent added to each treatment should be
equal to that in the highest exposure concentration of toxicant and in the carrier-control;
and should not exceed concentrations recommended by ASTM (1980).
Exposures are conducted in borosilicate glass tissue culture tubes 16 x 93 mm
(window size 11 x 55 mm). Twenty tubes per concentration are used, including 20
controls and 20 carrier controls (if required). Numbered tubes are randomized prior to
addition of embryos.
Early blastula embryos, rinsed 3 times with autoclaved 5°/oo water, are placed in a
crystallization dish under a dissecting scope (15-20x). Single embryos are drawn into a
pasteur pipette and placed in an empty tube with ~ 0.1 ml of 5°/oo water from the
crystallization dish. The randomized numbered tubes, each containing a single embryo,
are then quickly reordered (1-20, 21-40, 41-60, 61-80, 81-100, 101-120, 121-140)
and 6 ml of respective exposure solutions (or clean 5°/oo saline water for controls) are
added to each tube. Tubes, each with an airspace of 7 cc, are sealed with teflon-lined
caps, placed in racks and stored in a horizontal position in an incubator at 25+ 1°C with
a 14L:10 D photoperiod. Cool-white fluorescent lamps provide 500 lux illumination
during the light phase.
Dissolved oxygen (D.O.) and pH should be checked in 3 to 5 control(s) and respective
treatment tubes at the end of each test. A Lazar Model DO 166* oxygen probe and
Orion Model SA 520 pH meter, or equivalent, are suitable for measurements. In
conducting ~ 50 tests with a variety of teratogens, including aliphatic and aromatic
hydrocarbons, measured pH values have ranged from 6.1 to 7.3 and D.O. from 5.1 to
7.6 mg/1. Maintenance of adequate D.O. concentrations is essential since low D.O. can
also cause terata.
Quantification of Responses
Individual control and chemically-exposed embryos are examined daily for 7 days until
death or hatching occurs. The criterion for a successful test is 85% or greater hatching
in the control and carrier control. A Zeiss Invertoscope D Microscope, or equivalent, is
used to categorize responses. Observed terata are scored each day using a numerical
severity index for craniofacial (CR), cardiovascular (CV), and skeletal (SK) defects.
Numerical scores (Table 1) are based upon classification schemes devised by Weis and
Weis (1977), Weis etal. (1981), and Weis and Weis (1982).
Data Analysis
Upon completion of a test embryo deaths are categorized based on percentage mortality
and analysed, by either probit analysis to derive an LC50 or by a one-way analysis of
variance (ANOVA) and Duncans multiple range test, using arcsine transformed data to
•Mention of tradenames does not imply endorsement by the U.S. Environmental Protection Agency.
117

-------
indicate differences between control(s) and treatment responses (Finney, 1978; Steel
and Torrie, I960; Sokal and Rohlf, 1969). It is preferable that a concentration range be
chosen that limits the lethal response at the highest treatment level while enhancing the
teratogenic response.
Analysis of the total teratogenic response is performed as follows: for each tube, CR,
CV, and SK values (Table 1) are summed to provide a total daily score which can
range from 0 for a normally developing embryo to 13 for a severely deformed embryo
(i.e., CR 3 + CV 5 + SK 5 = 13). If an embryo dies before the end of the test, a value
of 14 is assigned to the tube daily from the date of death until test termination. The
daily scores are then summed to generate a total severity index for each embryo. By
assigning a value of 14 to exposed and control embryos that die, the uncertainty of
causal effect for death is accounted for without biasing the total severity-indices for
terata (Middaugh etal., 1988). To test for differences in terata between control(s) and
treatment concentrations, nonparametric Van der Waerden normal scores tests (SAS,
1985) and Van der Waerden post-hoc multiple comparison procedures, a = 0.05
(Marascuilo and McSweeney, 1977) are conducted, using total severity-indices.
Analysis of individual response categories, CR, CV and SK (Table 1), may be
conducted as follows: for each embryo, the daily score for the CR index could range
from 0 to 3; for CV, 0 to 5; and for SK, 0 to 5. If an embryo dies before the end of the
test, a daily value of 4 is assigned to the CR index; the CV and SK indices are assigned
a value of 6. As with total severity-indices, assignment of a daily value for embryo
death in individual categories enables one to account for mortalities without
introduction of bias by mortalities of unknown cause (Middaugh et ai, 1986).
Respective categorical indices are then analysed using Van der Waerden normal scores
tests (SAS, 1985) and post-hoc multiple comparison procedures (Marascuilo and
McSweeney, 1977).
With some compounds, the teratogenic expression is limited to one dysmorphic
response, e.g., skeletal defect. In this case, the data analysis is greatly simplified. The
teratogenic response is analysed by calculating the percentage of embryos that elicit a
response at each treatment level. The data can then be presented graphically or arcsine
transformed and analysed by one-way ANOVA and post-hoc procedures, if
appropriate. If categories based on the severity of the response can be distinguished
with the single morphogenic response, then the analysis is the same as described above
for individual response categories (CR, CV and SK).
Utilization of the system described above provides for quantification of
teratogenic/toxic effects of chemicals during embryonic development. Because the
exposure tubes are sealed for the duration of the test, the risk of human exposure to the
test compounds is greatly reduced. Furthermore, embryos from other species such as
the Japanese medaka, Oryzias latipes, and zebra fish, Brachydanio rerio, probably
could be tested in this system.
118

-------
Table 1. Menidia beryllina, Synopsis of observed teratologics! responses in embryos
and numerical severity-Index for craniofacial (CK), cardiovascular (CV), and skeletal
(SK) defects. Adapted in part from Weis and Weis (1977), Weis et al. (1981), and
Weis and Weis (1982).
Craniofacial
Cardiovascular
Skeletal
Value Effect
Value Effect
Value Effect
0 none observed
0 none observed
0 none observed
1 slight defect in
1 slight defect in
1 slight bend or kink
structure or size
structure or function
including scoliosis.

including reduced
lordosis

circulation

2 moderate defect in
2 tube heart, beating
2 major bend or kink
structure or size
with or without
(greater than 90° angle
including
circulation
or more than one bend)
synophthalmia


3 severe defect in
3 tube heart, not beating
3 stunted
structure or size


including
4 beating tissue, but no
4 very stunted, but axis
microphthalmia,
heart structure
discernible
anophthalmia, or


anencephaly
5 no discernible heart
5 no axis discernible
References
American Society for Testing and Materials. (1980) Standard practices for conducting acute toxicity tests
with fishes, macroinvertebrates and amphibians. In: Annual Book of ASTM Standards, pp. 1-25.
ASTM, Philadelphia.
Birge, WJ., Black, J.A., Westerman, A.G., Ramey, B.A. (1983) Fish and amphibian embryos - a model
system for evaluating teratogenicity. Fundamentals Appl. Toxicol. 3:237-242.
Chemoff, B., Conner, J.V., Bryan, C.F. (1981) Systematica of the Menidia beryllina complex (Pisces:
Atherinidae) from the Gulf of Mexico and its tributaries. Copeia. 1981:319-335.
Collins, T.F.X. (1987) Teratological research using in vitro systems. V. Nonmammalian model systems.
Environ. Health Perspectives. 72:237-249.
Finney, DJ. (1978) Statistical Methods in Biological Assay, 3rd Edition. Charles Griffin and Co. Ltd.,
London. 508 pp.
Gomez, R„ Lindsay, Hi,. Jr (1972) Occurrence of the Mississippi silverside, Menidia audens (Hay) in
Keystone Reservoir and the Arkansas River. Proc. Okla. Acad. Sci. 52:16-18.
Johnson, M.S. (1975) Biochemical systematics of the Atherinid genus Menidia. Copeia 1975: 662-691.
Laale, H.W., Lemer, W. (1981) Teratology and early fish development. Am. Zool. 21:517-533.
119

-------
Marascuilo, L.A., McSwceney, M. (1977) Nonpararmtric and Distribution - Free Methods for the
Social Sciences. Brooks/Cole, Monterey, CA. 556 p.
Middaugh, DP., Hester, P.G., Meish, M.V., Stark, P.M. (1985) Preliminary data on use of the inland
silverside, Menidia beryllina, to control mosquito larvae. J. Am. Mosquito Control Assoc. 1:435-
441.
Middaugh, DP., Hemmer, MJ„ Rose, YX. (1986) Laboratory spawning of the inland silverside,
Menidia beryllina, and Tidewater Silverside, Menidia peninsulae, with notes on survival and
growth of larvae at different salinities. Environ. Biol. Fishes. 15:107-117.
Middaugh, D.P., Hemmer, MJ., Lores, E.M. (1988) Teratological effects of 2,4-dinitrophenol,
"produced water" and naphthalene on embryos of the inland silverside, Menidia beryllina. Dis.
aquat. Org. 4. 53-65.
SAS Institute Inc. (1985) SAS User's Guide: Statistics, Version 5 Edition. Carey, NC. 956 p.
Sisk, M.E., Stephens, R.E. (1964) Menidia audens (Pisces: Atherinidae) in Boomer Lake, Oklahoma,
and its possible spread in the Arkansas River System. Proc. Olda. Acad, of Sci. 44:71-73.
Sokal, R.R., Rohlf, FJ. (1969) Biometry. The Principles and Practice of Statistics in Biological
Research. W.H. Freeman Co., San Francisco. 776 pp.
Solomon, F.P., Faustman, E.M. (1987) Developmental toxicity of four model alkylating agents on
Japanese medaka fish (Oryzias latipes) embryos. Environ. Toxicol. Chem. 6:747-753.
Steel, R.G.D., Torrie, J.H. (1960) Principles and Procedures in Statistics. McGraw-Hill, New York.
481 p.
Weis, J. S., Weis, P. (1977) Effects of heavy metals on development of the killifish, Fundulus
heteroclitus, /. Fish. Biol. 11:49-54.
Weis, J.S., Weis, P., Ricci, J.L. (1981) Effects of cadmium, zinc, salinity and temperature on the
teratogenicity of methylmercury to the killifish, Fundulus heteroclitus. Rapp. P.-V. Reun. Cons.
Int. Explor, Mer. 178:64-70.
Weis, P., Weis, J.S. (1982) Toxicity of methylmercury, mercuric chloride and lead in killifish, Fundulus
heteroclitus, from Southampton, New York. Environ. Res. 28:364-374.
Weis, J.S.. Weis, P. (1987) Pollutants as developmental toxicants in aquatic organisms. Environ. Health
Perspect. 71:77-85.
Whitby, K£„ Fiynn, TJ. (1987) Introduction - teratological research using in vitro systems. Environ.
Health Perspect. 72:201
120

-------

TECHNICAL REPORT DATA
(I'kase read Instructions on the reverse before completing)
1. REPORT NO.
EPA/60Q/A-94/2Q2
4. TITLE AND SUBTITLE
A FISH MODEL AS AN INDICATION FOR TERATOGENIC SUBSTANCES
3. RECIPI
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
D.P. Middaugh and M.J. Hemmer
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. ENVIRONMENTAL PROTECTION AGENCY
ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
GUI..F BREF7F. Fl OR IDA 37561	
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
In: Biological Monitoring of the Environment: h Manual of Methods. J. Salanki,
D. Jcffrev and G. M. Huqhes (od.), C/U3 Int'l., Wallinyford, ENG. p. 116-120,
16. ABSTRACT
A fish model, suitable for use as an indicator for teratogenic substances, is describe<
Individual blastula stage embroys of the inland si Iverside, Menidia beryl!ina, are
exposed to teratogens in sealed tissue culture tubes containing 6 m» of saline test
media, 5°/°° salinity and 25 + PC.
enumerated" usi ng a system that ranks
skeletal (SK) responses. Procedures
Individual embryos are examined daily and tetra
craniofacial (CR), cardiovascular (CV) and
for statistical analysis of data are described.
17.
KEY WOODS AND DOCUMENT ANALYSIS
c. COSATI Field/Group
J.iDENTiF ICRS/OPEN ENDED TERM
DESCRIPTOR
TO. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19.	SECUHITY CLASS (this Report I
.UNCLASSIFIED		
20.	SECURITY CI.ASS tTlltS paxel
UNCLASSIFIED
21. NO. OF PAGES
22. PRICE
EPA Form 2220-1 (Rev. 4 —77) phcvious edition is oit.o. i/'i
I

-------