The Medaka as a Model for Studying Germ-CellMutagenesis and Genomic Instability

Akihiro Shima1,2,* and Atsuko Shimada2

1Department of Integrated Biosciences, Graduate School of Frontier Sciences, The University of Tokyo, Tokyo, Japan 113-00332Department of Biological Sciences, School of Science, The University of Tokyo, Tokyo, Japan 113-0033

Abstract: To study germ-cell mutagenesis from the viewpoint of biodiversity, we developed a nonmammalian

specific-locus test system using the Japanese medaka, Oryzias latipes. The genetic end points available are

dominant lethal mutations, total specific-locus mutations, and viable specific-locus mutations. We examined

1,091,824 F1 embryos that correspond to 3,135,999 genetic loci using gamma rays and ethylnitrosourea as

mutagens. The results indicated the feasibility of using the medaka test system to detect genotoxic substances

in the aquatic environment. The data also indicated the presence of in vivo safeguards against germ-cell

mutagenesis. We present a brief summary of our medaka specific-locus test system raising perspectives on its

value.

Key words: medaka, Oryzias latipes, germ cells, mutagenesis, safeguard.

INTRODUCTION

The heritable genetic effects of environmental mutagens are

of great concern in assessing risks to humans. Among vari-

ous mutagens, the mutagenic effects of radiation were first

recognized in the 1920s using Drosophila, indicating that

ionizing radiation damages the genetic material in germ

cells, and this damage is transmitted from generation to

generation. Next to tests with Drosophila, the mouse spe-

cific-locus test (SLT), established primarily by Russell

(1951) and Russell and Kelly (1982), has been used exten-

sively for assessing the genetic risks in humans from envi-

ronmental mutagens. Hence, the mouse data often have

been regarded as being representative of the nonhuman

vertebrates.

To address the issue of how best to assess the effects of

genotoxic agents on ecosystems, however, we believe it im-

portant to remember the importance of biodiversity. Pio-

neering attempts were made to use other vertebrate spe-

cies—the guppy (Schroeder, 1969) and the zebrafish

(Chakrabarti et al., 1983). However, the results obtained

from these fish were not thorough enough to establish a

SLT. In 1985, we began to develop a nonmammalian SLT

system using the Japanese medaka, Oryzias latipes. Here we

give a short review of our work on the medaka SLT system,

which benefited greatly from the experiences of its prede-

cessor, Russell’s mouse SLT system.

DEVELOPMENT OF A NONMAMMALIANSPECIFIC-LOCUS TEST SYSTEMUSING MEDAKA

The medaka, Oryzias latipes, is a small oviparous freshwater

teleost native to Asian countries, such as Japan, China, and

Received January 31, 2001; accepted March 30, 2001

*Corresponding author: telephone and fax +81 3 5802 2911; e-mail: shima@biol.s.u-

tokyo.ac.jp

Mar. Biotechnol. 3, S162–S167, 2001DOI: 10.1007/s10126-001-0038-X

© 2001 Springer-Verlag New York Inc.

Korea. The facts that the habitats of the medaka range

widely from tropical to temperate regions in Asia, and that

Yamamoto (1975) listed 10 species belonging to the genus

Oryzias indicate that the medaka is a useful model for

studying biodiversity and even the evolution of species. The

advantages of using the medaka include the following: (1)

its basic biology, including basic genetics, is well established

(Yamamoto, 1975), (2) its ease in culture and breeding, (3)

its relatively short life-cycle (about 2 months), and (4) the

availability of different wild populations, spontaneous mu-

tants, and inbred strains.

The basic principle we adopted was a specific-locus test

(SLT) that was initially developed primarily by Russell as

the mouse 7-locus system, which allows the detection in F1

progeny of recessive visible mutations induced at seven

marker loci in parental germ cells. The mouse 7-locus sys-

tem was quite successful, and the vast amount of experi-

mental information obtained from it has been the main

source for assessing the genetic effects of ionizing radiation

in humans. Because we were interested in determining the

validity of the medaka as a nonmammalian SLT model, we

performed preliminary studies on gamma-ray-induced mu-

tations in male germ cells (sperm, spermatids, and sper-

matogonia) at a specific locus b. The results indicated that

many color mutant embryos, whose mutant phenotype

could be recognized very early after fertilization by the ab-

sence of melanotic melanophores, died during development

before hatching (Shima and Shimada, 1988). We then pro-

posed the term “total mutations” for specific-locus muta-

tions that were phenotypically detectable during early de-

velopment, a unique genetic end point available only in

oviparous organisms, and the term “viable mutations” for

hatched viable mutants (Shima and Shimada, 1988, 1991).

This distinction between the two kinds of mutations, total

and viable, inevitably arose from two facts: (1) medaka eggs

and embryos have transparent membranes through which

embryonic development can be easily observed under a ste-

reoscopic microscope, and (2) phenotypes of all marker loci

can be recognized during early developmental stages. Thus,

we felt that the medaka was a promising candidate for SLT.

In the meantime, by then we had succeeded in establishing

a tester strain with multiple recessive marker loci (Shimada

et al., 1988). By using a tester stock that is homozygous

recessive at the three marker loci, b, lf, and gu, and by

irradiating the wild-type males and then crossing them to

tester females, we could examine in more detail the dose–

response relations for male germ cells between the gamma-

ray dose and three genetic end points: dominant lethal mu-

tations, total mutations, and viable mutations. With the

reservation that we then had detected only one spontaneous

viable mutation, we concluded that the quantitative data,

including the doubling doses for dominant lethal and viable

mutation obtained from the medaka SLT, are quite com-

parable to those from the mouse SLT and, hence, indicated

the validity of the medaka SLT as a possible nonmammalian

test system for studying germ-cell mutagenesis. In addition,

we noted that because the spontaneous and radiation-

induced “total mutation” frequencies are almost one order

of magnitude higher than those for the “viable mutation,”

the doubling dose for total mutations, i.e., the amount of

radiation needed to double the natural incidence that can be

operationally calculated by dividing the spontaneous inci-

dence by the induced rate per unit dose, appeared to be

almost the same as those for dominant lethal and viable

mutations (Shima and Shimada, 1991). Therefore, we con-

cluded that although total mutation was a unique genetic

end point proposed on the basis of recognizing the mutant

phenotypes during the embryonic development of an

oviparous vertebrate, this end point could be used as a

quantitatively feasible genetic end point in a study of mu-

tagenesis. Table 1 shows the number of embryos and genetic

loci scored up to May 9, 2000 (1,091,824 embryos corre-

sponding to 3,135,999 loci). For details, see Shima and Shi-

mada (1991) and Shimada and Shima (1998).

In parallel with accumulation of data from gamma-

irradiation experiments, we began to examine effects of eth-

ylnitrosourea (ENU), a potent mutagenic alkylating agent.

Coinciding with the previous finding for gamma rays, that

irrespective of the stage of spermatogenesis at the time of

exposure approximately 90% of spontaneous and gamma-

ray-induced total mutants died during development, expo-

sure of sperm and spermatids to ENU also resulted in em-

bryonic death of approximately 90% of the total mutants.

In sharp contrast, however, approximately 90% of total mu-

tants obtained from ENU-exposed spermatogonia became

viable mutants (Shima and Shimada, 1994; Shimada and

Shima, 1998). These results indicated that for gamma rays

the quantitative relationship between induction of specific-

locus mutations and dominant lethal remains the same

among all stages of spermatogenesis (i.e., sperm, sperma-

tids, and spermatogonia), while for ENU-treated spermato-

gonia, it is biased primarily toward the induction of spe-

cific-locus mutations. Our next goal is to understand the

molecular mechanisms underlying this differential response

to damage between gamma rays and ENU, as well as among

spermatogenic stages. These comprehensive basic data on

Germ-cell mutagenesis in medaka S163

Table 1. *

Stage Mutagens

Fertile

eggs

Dead

embryos

Total

mutations

Total effec.

loci

Viable

mutations

Viable effec.

loci

Sperm Hdr 0.64 Gy Sa 7092 760 19 20433 2 18996

Hdr 1.90 Gy Sa 4667 768 34 13255 2 11697

Hdr 1.90 Gy H 400 37 7 1185 1 1089

Hdr 2.40 Gy H 2268 304 37 11255 1 9820

Hdr 4.75 Gy Sa 13046 4527 194 30878 19 21747

Hdr 4.75 Gy H 4015 1205 119 16669 9 12240

Hdr 7.10 Gy H 1270 548 71 6194 3 3610

Hdr 9.50 Gy Sa 9573 5691 235 15309 10 6948

Hdr 9.50 Gy H 1380 817 112 6417 10 2800

Ldr 4.75 Gy Sa 6563 2241 94 17973 9 12966

Ldr 9.50 Gy Sa 3303 1950 75 8326 5 4059

React 0.31 Gy Sa 7704 787 18 22577 2 21051

React 0.63 Gy Sa 8243 1237 33 23553 5 21018

ENU 0.1 mM Sa 2462 84 2 7328 1 7134

ENU 0.5 mM Sa 7165 951 15 21113 1 19452

ENU 1.0 mM Sa 5124 2227 19 13644 1 8736

Com 4.75/0.5 Sa 2495 1369 113 6654 4 3378

PBTA1 10 ug/H 940 17 0 4668 0 4615

Spermatids Hdr 0.64 Gy H 13620 1076 17 39624 1 37632

Hdr 1.90 Gy H 9079 1319 32 26002 3 23280

Hdr 1.90 Gy H 658 35 3 1971 0 1869

Hdr 4.75 Gy H 21372 4802 175 46306 13 38195

Hdr 4.75 Gy H 3634 545 82 15467 7 13301

Hdr 9.50 Gy Sa 15523 5881 217 25477 10 12903

Hdr 9.50 Gy H 1016 480 49 4792 4 2680

Ldr 4.75 Gy Sa 10361 1760 61 29294 3 25803

Ldr 9.50 Gy Sa 4987 1354 53 13647 3 10827

React 0.31 Gy Sa 14808 844 19 43597 1 41892

React 0.63 Gy Sa 18433 1794 55 53336 7 49917

ENU 0.1 mM Sa 5687 245 2 16876 1 16326

ENU 0.5 mM Sa 14832 3167 20 43097 3 34995

ENU 1.0 mM Sa 11396 6349 29 27675 2 15111

Com 4.75/0.5 Sa 5506 2860 105 15120 6 7929

PBTA1 10 ug/H 2150 77 1 10637 0 10375

Spermatocytes Hdr 4.75 Gy H 4656 819 32 13363 1 11508

Hdr 4.75 Gy H 1853 349 46 9172 6 7535

ENU 0.1 mM Sa 1278 82 1 3794 0 3615

ENU 0.5 mM Sa 8498 824 14 24881 4 22227

ENU 1.0 mM Sa 5200 2803 13 12424 2 7341

Com 4.75/0.5 Sa 2448 980 29 6947 2 4404

Differentiating spermatogonia Hdr 4.75 Gy Sa 6214 779 32 17975 4 16299

Hdr 4.75 Gy H 1807 246 32 8956 0 7805

ENU 0.1 mM Sa 3572 259 0 10559 0 9939

ENU 0.5 mM Sa 17842 993 18 52469 10 50547

ENU 1.0 mM Sa 10324 1531 18 29168 18 26379

Com 4.75/0.5 Sa 3859 591 31 11254 3 9804

S164 Akihiro Shima and Atsuko Shimada

the medaka SLT led us to more mechanistic studies of the

mechanisms of germ-cell mutagenesis using various mo-

lecular biological procedures (Kubota, et al., 1992, 1995;

Shimada and Shima, 1998; Fukamachi et al., 2001).

SAFEGUARDS AGAINST GERM-CELLMUTAGENESIS—PRESENCE OFTWOFOLD CHECKS

Germ cells are responsible for transmitting genetic infor-

mation of a species from generation to generation. As the

genomic stability of somatic cells is essential for an organ-

ism to survive, the continuity per se of a species depends

primarily on the stability of genetic information borne by

germ cells. Therefore, it can be speculated that mutations

induced in germ cells should be subjected to some kind of

surveillance that operates primarily to guarantee the geno-

mic stability and sustainability of a species.

Results gathered over the last 15 years from our

medaka SLT studies suggest that the initial genomic changes

induced in male germ cells would not straightforwardly

manifest themselves as phenotypic effects in F1 progeny but

that safeguards against germ-cell mutagenesis should oper-

ate to restore or ameliorate genomic damages (Figure 1).

Twofold checks, one a prefertilization check in the gonads

using DNA repair machinery as well as a cell-suicide re-

sponse (check 1), and the other a postfertilization check in

developing embryos through dominant lethal effects (check

2), could exemplify the presence of such safeguards against

germ-cell mutagenesis. Sperm are the most mature male

haploid gametes that can directly take part in fertilization

and are considered to be deficient in DNA repair and cell-

suicide capabilities. Therefore, sperm are assumed to evade

the prefertilization check (check 1), and in sperm the pre-

check 1 component of the genomic alterations should be-

come the postcheck 1 component that is then subjected to

check 2. By contrast, spermatogonia, which are stem cells

Table 1. * (Continued)

Stage Mutagens

Fertile

eggs

Dead

embryos

Total

mutations

Total effec.

loci

Viable

mutations

Viable effec.

loci

Spermatogonia Hdr 0.64 Gy Sa 48783 2411 10 142149 1 125547

Hdr 1.90 Gy H 35532 771 28 124999 3 122637

Hdr 4.75 Gy Sa 76020 5954 74 174472 8 166964

Hdr 4.75 Gy H 15310 773 104 63339 33 60656

Hdr 4.75 Gy Ngt 4788 880 5 21240 0 19540

Hdr 9.50 Gy Sa 43731 3481 51 89085 7 85413

Ldr 1.90 Gy H 20792 478 7 78468 1 77038

Ldr 4.75 Gy Sa 48738 3583 33 140475 6 135465

Ldr 4.75 Gy H 4579 183 9 13408 2 13047

Ldr 9.50 Gy Sa 48656 3996 93 139641 9 133980

React 0.31 Gy Sa 17726 741 2 43443 1 41955

React 0.63 Gy Sa 52206 3419 13 152161 2 138192

ENU 0.1 mM Sa 24033 1637 14 70239 3 67209

ENU 0.5 mM Sa 33042 1926 22 96767 17 93348

ENU 1.0 mM Sa 33165 1986 33 96687 30 93537

Com 4.75/0.5 Sa 17061 832 28 50382 9 48687

PBTA1 10 ug/H 6246 169 7 30925 0 30385

Control (males) Sa(m)xT(f) 200000 11109 19 523809 2 501570

Control (males) H(m)xT(f) 26259 643 4 114044 1 112184

Control (females) T(m)xSa(f) 5135 74 1 15345 1 15183

Control (females) T(m)xH(f) 21699 510 20 93610 1 91767

Total (05/09/00) 1091824 113920 2960 3135999 336 2886098

*Wild-type males were mutagenized with gamma-rays or chemicals. Sa: Sakura strain; H: HNI strain; T: tester strain; Ngt: Niigata strain.

Germ-cell mutagenesis in medaka S165

equipped with DNA repair machinery together with a cell-

suicide capability, would be subjected to both these safe-

guards. Our data (Table 1) suggest that as a result the mu-

tation inducibility in spermatogonia is reduced to about

1/30 for gamma rays and about 1/5 for ENU compared to

those in sperm (Table 1 and check 1 in Figure 1).

Male gametes, after having evaded or been subjected to

the prefertilization check 1, then fertilize the eggs. After

fertilization, the DNA damage originally induced in a male

gamete but thereafter processed through the prefertilization

check 1, i.e., gametic DNA damage, is now converted to

DNA damage in a developing embryo, i.e., zygotic DNA

damage. The zygotic DNA damage thus generated could be

the target for DNA damage-sensing and damage-

responding factors that are assumed to be initially under the

control of material inheritance, but later under the control

of zygotic gene expression.

Although we do not yet know what kind of mechanism

is responsible for the postfertilization check in developing

embryos, we propose, on the basis of our mutagenesis data

(Table 1), that a phenomenon well known as dominant

lethal should be operating and presumably implemented as

the postfertilization check (check 2) by reducing total mu-

tations down to viable mutations. In fact, about 90% of the

total mutants are eliminated as dominant lethals during

embryonic development and only the remaining 10% can

become viable mutants for all spermatogenic stages for

gamma rays and for sperm/spermatids for ENU. It was

interesting to note that almost all total mutants derived

from the ENU-treated spermatogonia became viable mu-

tants, indicating that the ENU-treated spermatogonia evade

check 2. Here again, there was a differential response of

germ cells to gamma rays and ENU. Elucidation of the

underlying molecular mechanisms is under way.

The safeguards against germ-cell mutagenesis could be

twofold; prefertilization and postfertilization checks. Nev-

ertheless, some germ-cell mutations might slip through.

Germ-cell mutations in the elements of the safeguards,

themselves, which presumably include DNA damage-

sensing and damage-responding mechanisms (repair and

cell suicide) and hence are responsible for genomic stability,

would, in turn, result in genomic instability of germ cells,

which might eventually endanger a species to the point of

extinction.

ACKNOWLEDGMENTS

This research was supported by grants-in-aid from the Min-

istry of Education, Science, Sports, and Culture, Japan, and

also from the Ministry of Health and Welfare, Japan. The

assistance by Shizuko Takada in fish care is greatly appre-

ciated. Thanks are due to Hoshio Eguchi at the Research

Center for Nuclear Science and Engineering of the Univer-

sity of Tokyo for his assistance in operating the irradiation

facility with an 80-TBq 137Cs source.

REFERENCES

Chakrabarti, S., Streisinger, G., Singer, F., and Walker, C. (1983).

Frequency of g-ray induced specific locus and recessive lethal mu-

Figure 1. Safeguards against events

involved in germ-cell mutagenesis

from initial genomic changes

induced in male germ cells to

phenotypic effects manifested in F1

progeny.

S166 Akihiro Shima and Atsuko Shimada

tations in mature germ cells of the zebrafish, Brachydanio rerio.

Genetics 103:109–123.

Fukamachi, S., Shimada, A., Naruse, K., and Shima, A. (2001).

Genomic analysis of gamma-ray-induced germ-cell mutations at

the b locus recovered from the medaka specific-locus test. Mutat

Res (in press).

Kubota, Y, Shimada, A., and Shima, A. (1992). Detection of

gamma-ray-induced DNA damages in malformed dominant lethal

embryos of the Japanese medaka (Oryzias latipes) using AP-PCR

fingerprinting. Mutat Res 283:263–270.

Kubota, Y, Shimada, A., and Shima, A. (1995). DNA alterations

detected in the progeny of paternally irradiated Japanese medaka

fish (Oryzias latipes). Proc Natl Acad Sci USA 92:330–334.

Russell, W.L. (1951). X-ray induced mutations in mice. Cold

Spring Harbor Symp Quant Biol 16:327–336.

Russell, W.L., and Kelly, E.M. (1982). Mutation frequencies in

male mice and the estimation of genetic hazards of radiation in

men. Proc Natl Acad Sci USA 79:542–544.

Schroeder, J.H. (1969). X-ray-induced mutations in the poeciliid

fish, Lebistes reticulatus. Mutat Res 7:75–90.

Shima, A., and Shimada, A. (1988). Induction of mutations in

males of the fish Oryzias latipes at a specific locus after gamma-

irradiation. Mutat Res 198:93–98.

Shima, A., and Shimada, A. (1991). Development of a possible

nonmammalian test system for radiation-induced germ-cell mu-

tagenesis using a fish, the Japanese medaka (Oryzias latipes). Proc

Natl Acad Sci 88:2545–2549.

Shima, A., and Shimada, A. (1994). The Japanese medaka, Oryzias

latipes, as a new model organism for studying environmental

germ-cell mutagenesis. Environ Health Perspect 102(suppl 12):33–

35.

Shimada, A., and Shima, A. (1998). Combination of genomic

DNA fingerprinting into the medaka specific-locus test system for

studying environmental germ-line mutagenesis. Mutat Res 399:

149–165.

Shimada, A., Shima, A., and Egami, N. (1988). Establishment of a

multiple recessive tester stock in the fish Oryzias latipes. Zool Sci

5:897–900.

Yamamoto, T. (1975). Medaka (Killifish), Biology and Strains. To-

kyo: Keigaku.

Germ-cell mutagenesis in medaka S167