The Utility of the Guppy (Poecilia reticulata) and Medaka
(Oryzias latipes) in Evaluation of Chemicals for Carcinogenicity
Grace E. Kissling,*,1Naomi J. Bernheim,† William E. Hawkins,‡ Marilyn J. Wolfe,§ Micheal P. Jokinen,¶
Cynthia S. Smith,† Ronald A. Herbert,† and Gary A. Boorman†
*Environmental Medicine and Diseases Program and †Environmental Toxicology Program, National Institute of Environmental Health Sciences,
Research Triangle Park, North Carolina 27709; ‡Gulf Coast Research Laboratory, University of Southern Mississippi, Ocean Springs, Mississippi 39566;
§Experimental Pathology Laboratories, Inc., Herndon, Virginia 20172; and¶Pathology Associates, Inc., Cary, North Carolina 27513
Received February 15, 2006; accepted March 29, 2006
There has been considerable interest in the use of small fish
models for detecting potential environmental carcinogens. In this
study, both guppies (Poecilia reticulata) and medaka (Oryzias
latipes) were exposed in the aquaria water to three known rodent
carcinogens for up to 16 months. Nitromethane, which caused
mammary gland tumors by inhalation exposure in female rats,
harderian gland and lung tumors in male and female mice, and
liver tumors in female mice by inhalation, failed to increase
tumors in either guppies or medaka. Propanediol, which when
given in the feed was a multisite carcinogen in both sexes of rats
and mice, caused increased liver tumors in male guppies and male
medaka. There was reduced survival in female guppies and no
increased tumors in female medaka. 1,2,3-Trichloropropane,
which when administered by oral gavage was a multisite carcin-
ogen in both sexes of rats and mice, caused an increased incidence
of tumors in the liver of both male and female guppies and
medaka and in the gallbladder of male and female medaka. The
results of this study demonstrate that for these three chemicals,
under these specific exposure conditions, the fish appear less
sensitive and have a narrower spectrum of tissues affected than
rodents. These results suggest that fish models are of limited util-
ity in screening unknown chemicals for potential carcinogenicity.
Key Words: bioassays; small fish models; medaka; guppy; rodent
carcinogens; nitromethane; propanediol; trichloropropane.
The existence of a sensitive inexpensivevertebrate model for
screening for potential carcinogens would have many advantages
where there are more chemicals to evaluate than resources.
Small fish models have been suggested as more sensitive, less
costly, and quicker than traditional rodent models (Bailey et al.,
1984; Ishikawa and Takayama, 1979; Reddy et al., 1999;
Simon and Lapis, 1984; Sinnhuber et al., 1978; Walker et al.,
1985). The National Toxicology Program (NTP) has found that
other types of short-term tests have limitations for predicting
carcinogenicity when systemically evaluated under controlled
conditions in a contract laboratory situation (Zeiger et al.,
1990). Most fish cancer studies use potent carcinogens that also
cause liver tumors in rodents as well as in the fish models
(Brown-Peterson et al., 1999; Hawkins et al., 1998; Law et al.,
1998; Liu et al., 2003; Okihiro and Hinton, 1999). However,
environmental chemicals may affect a variety of tissues in
rodents, and often this spectrum provides clues as to the nature
and mechanism of the carcinogenicity of the test chemical.
Therefore, we decided to evaluate three chemicals that
caused cancer in rodents, but generally not primarily of the
liver, using two common small fish models. Two chemicals,
1,2,3-trichloropropane (TCP) and 2,2-bis(bromomethyl)-1,3-
propanediol (BMP), are mutagenic and cause a wide spectrum
of tumors in rodents. TCP when given by oral gavage in corn
oil caused increased incidences of tumors of the oral cavity,
forestomach, kidney, harderian gland, Zymbal gland, liver,
uterus, pancreas, and other sites in rodents (NTP, 1993). BMP
when given in the feed caused increased incidences of tumors
of the mammary gland, skin, oral cavity, forestomach, intes-
tines, harderian gland, Zymbal gland, lung, kidney, urinary
bladder, and other sites in rodents (NTP, 1996). To determine
how these fish models would respond to a less-potent carcino-
gen, nitromethane (NM), a nonmutagen with a more modest
response in rodents was selected. NM did not cause tumors of
male rats, but there was clearly a carcinogenic effect in female
rats based on increased incidence of mammary gland tumors
and clearly a carcinogenic effect in mice based on increased
incidences of lung and harderian gland tumors (NTP, 1997).
NM exposure was also associated with an increased incidence
of liver tumors in female mice. Since chemical evaluations in
rodents have benefited from careful attention to study details,
we attempted to follow NTP procedures for chemistry, pathology,
statistics, and QA evaluations. We also evaluated the data for
carcinogenicity determinations as would be done in rodent
studies. Thus, the result for each chemical was judged to be
1To whom correspondence should be addressed at Environmental Medicine
and Diseases Program, National Institute of Environmental Heath Sciences,
MD A3-03, P.O. Box 12233, 111 T. W. Alexander Drive, Research Triangle
Park, NC 27709. Fax: (919) 541-4311. E-mail: firstname.lastname@example.org.
Published by Oxford University Press 2006.
TOXICOLOGICAL SCIENCES 92(1), 143–156 (2006)
Advance Access publication March 31, 2006
by guest on June 1, 2013
positive, negative, equivocal, or inadequate for each sex and
Our results indicate that for the three chemicals evaluated
under maximally tolerateddoses,the medaka and guppy appear
less sensitive for a carcinogenic response than did the rodents.
Further, while a variety of tissues showed a carcinogenic
response in rodents, in the fish model only the liver and biliary
system showed increased tumor rates with exposure. This
suggests that caution is warranted in using these model systems
for evaluating chemicals for which the carcinogenic potential is
MATERIALS AND METHODS
Fish. Guppy (Poecilia reticulata) and medaka (Oryzias latipes) fry used in
the studies were obtained from established Gulf Coast Research Laboratory
(GCRL) cultures and were approximately 14–16 days old at study start. The
fish were maintained in glass aquaria partially submerged in a singlewater bath
to maintain temperature. Fish were approximately equally allocated to each of
two duplicate aquaria per dose group. Aquaria were cleaned once weekly
except for NM, where the aquaria were cleaned two (guppy) or three (medaka)
times per week. Animals were fed flake fish food (Aqua-Tox Flake, Ziegler
Brothers, Inc., Gardners, PA) and brine shrimp (Artemia) larvae (Aquarium
Products, Glen Burnie, MD) once per day; shrimp were not fed to the fish
during the last week of the study. The fish had a 16-h light period/8-h dark
period with a 30-min transition period to simulate dawn and dusk. The aquaria
temperature was 26 ± 1?C with an aquaria pH between 8.6 and 9.0. More
completedetailsonthe fishandanimalcare proceduresare availablein theNTP
Technical Report (NTP, 2005), which is also available at http://ntp-server.
Chemicals. BMP, NM, and TCP were obtained from Aldrich Chemical
Company (Milwaukee, WI), and each was found to be 99% pure. All three of
the bulk chemicals were stored at approximately 4?C during the study. The
stability of the bulk chemical was checked monthly during the study using gas
chromatography (GC), and no contaminants were found. More details on the
chemical analysis and spectra are available in the NTP Technical Report (NTP,
Exposure solution generation. Exposure was intermittent flow-through
and was conducted in a closed chamber similar to that described by Walker
et al. (1985). Stock solutions of BMP, NM, and TCP were prepared by adding
neat chemical to filtered and ultraviolet-sterilized well water in glass carboys.
Dispensing pumps injected the stock solution into glass mixing/splitting boxes
prior to the delivery of 2 l of filtered and ultraviolet-sterilized water to produce
the required concentrations for the exposure aquaria. A water dispenser was
timer regulated to perform at least five volume additions per day to each
exposure aquarium. Filtered and ultraviolet-sterilized well water alone was
delivered to the control aquaria. Aquaria volumes were maintained by an
overflow drain siphon designed to remove water from near the bottom of each
aquarium. The exposure delivery system has been described in detail (NTP,
Exposure characterization. Exposure characterizations were performed
prior to the start of the 14- and 16-month studies. The generation of the target
concentrations of BMP in exposure aquaria by intermittent flow-through was
determined with and without fish present. Duplicate aquaria target concen-
trations of 10, 35, and 100 mg/l were sampled prior to the first injection and 3,
6, 9, 12, 14, and 24 h after the initial injection. Target concentrations were
reached by 24 h. Uniformity of BMP concentrations was measured by sampling
nine locations at 2 cm below the surface and nine locations within 2 cm of the
bottom of the aquaria. GC-analyzed samples demonstrated BMP uniformity.
More detailed methods and results are available in the NTP Technical Report
NM exposure characterizations were similar to BMP. Duplicate aquaria
target concentrations of 8.6, 24.5, and 70 mg/l were sampled prior to the first
injection and every 2 h after the initial injection. Target concentrations were
reached by 14 h. GC-analyzed samples demonstrated NM uniformity (NTP,
TCP exposure characterizations were similar to BMP using similar
concentrations and sampling times. Target concentrations were reached by
24 h. GC-analyzed samples demonstrated TCP uniformity (NTP, 2005).
Concentrations of BMP, NM, and TCP in the water of the exposure aquaria
were monitored using GC, approximately three times each week. Duplicate
sampleswere analyzed fromeach aquarium. Table 1summarizesthe dosesused
for each chemical in the fish studies, as well as in comparative rodent studies.
Pathology examination. All animals were observed twice daily. Visual
external findings were recorded daily, and body weights and lengths were
recorded at sacrifice after overexposure to a lethal concentration of tricaine
methanesulfonate (MSZ22). Whole fish with the tails removed were fixed in
Bouin’s fixative for 96 h and then transferred to 10% neutral buffered formalin,
processed, and embedded in paraffin. Five longitudinal step sections and two
serials of each step were made, stained with hematoxyl and eosin and examined
histologically by a study pathologist. The organs and tissues examined
histologically in fish and in comparative rodent studies are listed in Table 2. A
second pathologist reviewed all diagnoses in 40% of controls and fish from
thehigh exposure concentration. Subsequently, all neoplasms in the target tissue,
liver, were reviewed. Hepatocellular adenomas and hepatocellular carcinomas
in fish share many features with rodent liver neoplasia, and similar diagnostic
criteria were used for the medaka lesions (Boorman et al., 1997). A pathology
working group consisting of pathologists experienced in rodent and/or fish
pathology reviewed selected hepatic neoplasms. Details of specimen preparation
and examination are included in the NTP Technical Report (NTP, 2005).
Statistical analysis. The probability of survival was estimated by the
product-limit procedure of Kaplan and Meier (1958). Fish that were sacrificed
or transferred to another tank were censored; fish found dead before the 9-
month and terminal sacrifices were not censored. Possible dose-related trends
in survival were tested with Tarone’s (1975) life table test; pairwise
comparisons with the control group were made with Cox’s (1972) method
for testing the equality of two groups. For each chemical, data were collected
Doses and Routes of Exposure Used in the Fish and Rodent Studies
0, 24, 50, 150 mg/l in water
0, 10, 30, 70 mg/l in water
0, 4.5, 9, 18 mg/l in water
0, 24, 60, 150 mg/l in water
0, 10, 20, 40 mg/l in water
0, 4.5, 9, 18 mg/l in water
0, 2500, 5000, 10,000 ppm in feed
0, 94, 188, 375 ppm by inhalation
0, 3, 10, 30 mg/kg by gavage
0, 312, 625, 1250 ppm in feed
0, 188, 375, 750 ppm by inhalation
0, 6, 20, 60 mg/kg by gavage
KISSLING ET AL.
by guest on June 1, 2013
The chemicals were selected because they caused cancer in
a variety of tissues and not just the liver. Wewere surprised that
the cancer response in fish was restricted to the liver except for
TCP that also caused a modest increase in tumors of the bile
duct. The limited fish cancer response appears less informative
than the wide spectrum of tumors found in rodent studies.
The loss of specimens during the study limited the in-
formation that was available for analysis. For example, time to
tumor and presence of preneoplastic lesions could not be
analyzed with the present study design. Multiple interim
sacrifices may have been helpful. However, the 9-month
sacrifice showed few or no tumors with these three chemicals,
and the medaka had to be terminated at 13 or 14 months (16
months for guppy), providing a very small window of time in
which to follow tumor development. One concern that was
difficult to address was the possibility that some fish that
developed tumors died early and were lost to the study. For
example, all three guppy studies had reduced survival in the
high-dose groups after 9 months when the groups were split
into continued exposure and stop-exposure groups. The
percentage of fish examined at final sacrifice varied between
55% in the NM study and 71% in the BMP study. It is not
known if this resulted in an underestimation of the true tumor
incidence or whether this affected the sensitivity of the model
for detecting carcinogenicity.
This study based on a limited number of chemicals suggests
to us that routine use of small fish models in waterborne assays
is likely to underestimate the number of chemicals that would
demonstrate carcinogenic activity if evaluated in standard
rodent studies. Small fish species continue to be excellent
research models to address mechanistic questions for a variety
of disease processes including carcinogenicity. This is espe-
cially true for zebra fish (Danio rerio) where the genome has
been sequenced (Jekosch, 2004), and resources such as the
Zebrafish Information Network (Rasooly et al., 2003) are avail-
able to support investigators who address mechanistic questions.
National Institutes of Health and National Institutes of Environmemtal Health
Sciences. The in-life portion of the study was conducted under NIEHS contract
NO1-ES-35371 to Gulf Coast Research Laboratories. The authors thank Drs
Michael Cunningham and Robert Maronpot for their helpful suggestions.
Alden, C. L., Smith, P. F., Piper, C. E., and Brej, L. (1996). A critical appraisal
of the value of the mouse cancer bioassay in safety assessment. Toxicol.
Pathol. 24, 722–725.
Bailey, G. S., Hendricks, J. D., Nixon, J. E., and Pawlowski, N. E. (1984). The
sensitivity of rainbow trout and other fish to carcinogens. Drug Metab. Rev.
Bennett, L. M., and Davis, B. J. (2002). Identification of mammary carcinogens
in rodent bioassays. Environ. Mol. Mutagen. 398, 150–157.
Boorman, G. A., Botts, S., Bunton, T. E., Fournie, J. W., Harshbarger, J. C.,
Hawkins, W. E.,Hinton, D.E.,Jokinen, M.P., Okihiro, M.S.,andWolfe,M.J.
(1997). Diagnostic criteria for degenerative, inflammatory, proliferative
nonneoplastic and neoplastic liver lesions in medaka (Oryzias latipes):
Consensus of a National Toxicology Program Pathology Working Group.
Toxicol. Pathol. 25, 202–210.
Brown-Peterson, N. J., Krol, R. M., Zhu, Y., and Hawkins, W. E. (1999). N-
nitrosodiethylamine initiation of carcinogenesis in Japanese medaka (Or-
yzias latipes): Hepatocellular proliferation, toxicity, and neoplastic lesions
resulting from short term, low level exposure. Toxicol. Sci. 50, 186–194.
Davis, C. R., Okihiro, M. S., and Hinton, D. E. (2002). Effects of husbandry
practices, gender and normal physiological variation on growth and re-
production of Japanese medaka, Oryzias latipes. Aquat. Toxicol. 60, 185–
Eastin, W. C. (1998). The U. S. National Toxicology Program evaluation of
transgenic mice as predictive models for identifying carcinogens. Environ.
Health Perspect. 106, 81–84.
Gardner, H. S. J., Brennan, L. M., Toussaint, M. W., Rosencrance, A. B.,
Boncavage-Hennessey, E. M., and Wolfe, M. J. (1998). Environmental complex
mixture toxicity assessment. Environ. Health Perspect. 106, 1299–1305.
Geter, D. R., Winn, R. N., Fournie, J. W., Norris, M. B., DeAngelo, A. B., and
Hawkins, W. E. (2004). MX93-chloro-4-(dichloromethyl)-5-hydro-2(5H)-
furanole),a drinking-watercarcinogen,does notinducemutationsinthe liver
of cII transgenic medaka (Oryzias latipes). J. Toxicol. Environ. Health A 67,
Haseman, J. K. (2000). Using the NTP database to assess the value of
rodent carcinogenicity studies for determining human cancer risk. Drug Metab.
Rev. 32, 169–186.
Hawkins, W. E., Overstreet, R. M., Fournie, J. W., and Walker, W. W. (1985).
Development of aquarium fish models for environmental carcinogenesis:
Tumor induction in seven species. J. Appl. Toxicol. 5, 261–264.
Hawkins, W. E., Walker, W. W., Fournie, J. W., Manning, C. S., and Krol, R. M.
(2003). Use of the Japanese medaka (Oryzias latipes) and guppy (Poecilia
reticulata) in carcinogenesis testing under National Toxicology Program
protocols. Toxicol. Pathol. 31, 88–91.
Hawkins, W. E., Walker, W. W., James, M. O., Manning, C. S., Barnes, D. H.,
Hearn, C. S., and Overstreet, R. M. (1998). Carcinogenic effects of 1, 2-
dibromoethane (ethylene dibromide; EDB) in Japanese medaka (Oryzias
latipes). Mutat. Res. 399, 221–232.
Hinton, D. E., Lantz, R. C., and Hampton, J. A. (1984). Effect of age and
exposure to a carcinogen on the structure of the medaka liver: A
morphometric study. J. Natl. Cancer Inst. 65, 239–249.
Ishikawa, T., Masahito, P., and Takayama, S. (1984). Usefulness of the medaka,
Oryzias latipes, as a test animal: DNA repair processes in medaka exposed to
carcinogens. J. Natl. Cancer Inst. 65, 35–43.
Ishikawa, T., and Takayama, S. (1979). Importance of hepatic neoplasms in
lower vertebrate animals as a tool in cancer research. J. Toxicol. Environ.
Health 5, 537–550.
Jekosch, K. (2004). The zebrafish genome project: Sequence analysis and
annotation. Methods Cell Biol. 77, 225–239.
Kaplan, E. L., and Meier, P. (1958). Nonparametric estimation from incomplete
observations. J. Am. Stat. Assoc. 53, 457–481.
Komulainen, H., Hakulinen, P., Servomaa, K., Makkonen, K., Vasara, R.,
Maki-Paakkanen, J., and Kosma, V. M. (2000). No consistent pattern of muta-
tions in p53 and ras genes in liver tumors of rat treated with the drinking
(MX). Environ. Mol. Mutagen. 36, 292–300.
Komulainen, H., Kosma, V. M., Vaittinen, S. L., Vartiainen, T., Kaliste-
Korhonen, E., Lotjonen, S., Tuominen, R. K., and Tuomisto, J. (1997).
USE OF SMALL FISH MODELS IN BIOASSAYS
by guest on June 1, 2013
Carcinogenicity of the drinking water mutagen 3-chloro-4-(dichloromethyl)-
5-hydroxy-2(5H)-furanone in the rat. J. Natl. Cancer Inst. 89, 848–856.
Law, J. M., Bull, M., Nakamura, J., and Swenberg, J. A. (1998). Molecular
dosimetry of DNA adducts in the medaka small fish model. Carcinogenesis
Liu, Z., Kullman,S. W., Bencic, D. S., Torten,M., and Hinton, D. E. (2003). ras
Oncogene mutations in diethylnitrosamine-induced hepatic tumors in
medaka (Oryzias latipes), a teleost fish. Mutat. Res. 539, 43–53.
National Toxicology Program (NTP). (1993). Toxicology and Carcinogenesis
Studies of 1,2,3-Trichloropropane (CAS No. 96-18-4) in F344/N Rats and
B6C3F1 Mice (Gavage Study). NTP Technical Report No. 384. U.S.
Department of Health and Human Services, Public Health Service, National
Institutes of Health, Research Triangle Park, NC.
National Toxicology Program (NTP). (1996). Toxicology and Carcinogenesis
Studies of 2,2-Bis(Bromomethyl)-1,3-Propanediol (FR-1138?) (CAS. No.
3296-90-0)in F344/NRatsand B6C3F1Mice(FeedStudies), NTPTechnical
Report No. 452. National Institute of Environmental Health Sciences, Public
Health Service, National Institutes of Health, Research Triangle Park, NC.
National Toxicology Program (NTP). (1997). Toxicology and Carcinogenesis
Studies of Nitromethane (CAS No. 75-52-5) in F344/N Rats and B6C3F1
Mice (Inhalation Study). NTP Technical Report No. 461. U.S. Department of
Health and Human Services, Public Health Service, National Institutes of
Health, Research Triangle Park, NC.
National Toxicology Program (NTP). (2005). Toxicology and Carcinogenesis
Studies of 2,2-Bis(Bromomethyl)-1,3-Propanediol, Nitromethane, and 1,2,3-
Trichloropropane (CAS Nos. 3296-90-0, 75-52-5 and 96-18-4) in Guppies
(Poecilia reticulata) and Medaka (Oryzias latipes) (Waterborne Studies).
NTP Technical Report No. 528. U.S. Department of Health and Human
Services, Public Health Service, National Institutes of Health, Research
Triangle Park, NC.
Okihiro, M. S., and Hinton, D. E. (1999). Progression of hepatic neoplasia in
medaka (Oryzias latipes) exposed to diethylnitrosamine. Carcinogenesis 20,
Rasooly, R. S., Henken, D., Freeman, N., Thompkins, L., Badman, D., Briggs, J.,
and Hewitt, A. T. (2003). Genetic and genomic tools for zebrafish research:
The NIH zebrafish initiative. Dev. Dyn. 228, 490–496.
Reddy, A. P., Spitsbergen, J. M., Mathews, C., Hendricks, J. D., and Bailey,
G. S. (1999). Experimental hepatic tumorigenicity by environmental hydro-
carbondibenzo[a,1]pyrene.J. Environ. Pathol.Toxicol.Oncol.18,261–269.
Salem, H., and Katz, S. (1998). Advances in Animal Alternatives for Safety and
Efficacy Testing. Taylor and Francis, Philadelphia, PA.
Simon, K., and Lapis, K. (1984). Carcinogenesis studies on guppies. Natl.
Cancer Inst. Monogr. 65, 71–81.
Sinnhuber, R. O., Hendricks, J. D., Wales, J. H., and Putnam, G. B. (1978).
Neoplasms in rainbow trout, a sensitive animal model for environmental
carcinogenesis. Ann. N. Y. Acad. Sci. 298, 389–408.
Tarone, R. E. (1975). Tests for trend in life table analysis. Biometrika 62,
Toussaint, M. W., Rosencrance, A. B., Brennan, L. M., Denns, W. E., Beaman,
J. R., Wolfe, M. J., Hoffmann, F. J., and Gardner, H. S. J. (2001). Chronic
toxicity of bromodichloromethane to the Japanese medaka (Oryzias latipes).
Toxicol. Pathol. 29, 662–669.
Toussaint, M. W., Wolfe, M. J., Burton,D. T., Hoffmann,F. J., Shedd,T. R., and
Gardner, H. S. J. (1999). Histopathology of Japanese medaka (Oryzias
latipes) chronically exposed to a complex environmental mixture. Toxicol.
Pathol. 27, 652–663.
Walker, W. W., Manning, C. S., Overstreet, R. M., and Hawkins, W. E. (1985).
Development of aquarium fish models for environmental carcinogenesis: An
intermittent-flow exposure system for volatile hydrophobic chemicals. J.
Appl. Toxicol. 5, 255–260.
Zeiger, E., Haseman, J. K., Shelby, M. D., Margolin, B. H., and Tennant, R. W.
(1990). Evaluation of four in vitro genetic toxicity tests for predicting rodent
carcinogenicity: Confirmation of earlier results with 41 additionalchemicals.
Environ. Mol. Mutagen. 16, 1–14.
Zimmerer, E. J. (1984). Usefulness of the guppy, Poecilia reticulata, in
carcinogenicity testing: Special advantages and problems. J. Natl. Cancer
Inst. 65, 59–64.
KISSLING ET AL.
by guest on June 1, 2013