| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a University of Notre Dame, Department of Biological Sciences, Notre Dame, Indiana 46556
ABSTRACT
A mutagenesis screen was conducted on zebrafish using N-ethyl N-nitrosourea as a mutagen and an F2 crossing scheme to obtain homozygous mutants in the F3 generation. Whole abdomens of 3-mo-old F3 zebrafish progeny were fixed and mass-embedded in paraffin blocks. Blocks were cut with a microtome to obtain cross-sections of the entire body cavity that included the ovaries and testes. Slides of the cross-sections were analyzed for alterations in gonadal structure and gametogenesis and were compared with gonads of wild-type fish. A total of 125 mutagenized genomes in 81 families were screened and 11 mutations were observed that produced visible phenotypes in only one sex per family. Male mutations included testes without mature sperm that contained either predominantly spermatocytes or spermatogonia. Female mutations included ovaries containing 1) degenerating oocytes surrounded by hypertrophied follicle walls or stroma, 2) extrafollicular tissue proliferation, 3) proliferating postovulatory follicle walls, and 4) large numbers of degenerating preovulatory and postovulatory oocytes. While past screens on zebrafish have concentrated on early developmental mutations, the results of this study demonstrate for the first time that mutagenesis can be used with zebrafish to study reproduction in adult animals.
follicle, gamete biology, ovary, ovum, sperm, testis
INTRODUCTION
In recent years, the zebrafish (Danio rerio) has become a popular research model for the study of early development and embryogenesis. There have been three large-scale mutagenesis studies performed on zebrafish to identify important genes in those two processes. Two of the screens made use of the chemical mutagen, N-ethyl-N-nitrosourea (ENU) [1, 2], while the other used retroviral vectors that were randomly inserted into the genome [3]. The popularity of zebrafish for these types of studies can be traced to several factors. Zebrafish are small (3–4 cm), so many fish can be maintained and raised in a relatively small facility. They reach sexual maturity quickly and females can continuously produce large numbers of progeny (80–100) over short time intervals. Zebrafish embryos are translucent, which allows the early development of the brain, eyes, trunk musculature, gut, heart, and nervous system to be observed without dissection. Finally, because zebrafish are vertebrates, they may be more applicable as a biomedical model for humans than are popular invertebrate genetic models.
A few studies have examined mutations generated by ENU in older zebrafish. For example, a screen was carried out to obtain mutations that affected fin regeneration in zebrafish at 6–8 wk of age [4]. In addition, some early developmental screens have uncovered embryonic mutations that are not lethal and that manifest a specific phenotype in adults (see for example [5]). Studies on vertebrate physiology have rarely used mutagenic approaches. However, many of the same traits that make zebrafish an attractive model system for the study of development also make it attractive for the study of physiological systems or processes, including reproduction. A great deal of information has been published on the physiology and endocrinology of the reproductive system in vertebrates. However, much of our understanding of vertebrate reproduction is based largely on correlative data, or on the effects of agonists and antagonists that frequently exhibit wide variations in specificity. Studies using reverse genetics have provided more conclusive information on the reproductive role of some gene products (see for example [6–8]), yet these strategies are based on a single predetermined gene. In contrast, the identification of genes that are responsible for ovarian or testicular mutations produced randomly by a mutagen may serve to substantiate conclusions drawn from more traditional physiological studies or, more importantly, bring to light controls not previously known. For this reason we used chemical mutagenesis and an F2 crossing scheme to uncover mutations that affect the ovaries and testes of adult zebrafish. To our knowledge, this is the first time that a mutagenesis approach has been used to study the reproductive system of vertebrates.
MATERIALS AND METHODS
Mutagenesis Scheme
Protocols for the treatment and care of zebrafish used in this study were reviewed and approved by the University of Notre Dame Institutional Animal Care and Use Committee. For chemical mutagenesis, 140 adult (6-mo-old) male AB zebrafish (wild-type line provided by D. Grunwald, University of Utah) were incubated one or two times per week for 1 h in a 3 mM ENU (Sigma, St. Louis, MO) solution for a total of four treatments. Treatment and post-treatment handling of fish was carried out as previously described [9, 10]. The mutagenic efficiency of this protocol was tested by determining the frequency of induced mutations at two different pigmentation loci; albino (alb) and golden (gol). These specific locus tests were conducted in single pair matings by crossing mutagenized males with homozygous tester females carrying either the alb or gol alleles and determining the frequency with which new alleles were generated at either loci. Of the 1827 progeny from those crosses, approximately 0.22% represented new alb or gol alleles, which is consistent with the mutagenic rates reported in earlier studies [10].
Three weeks to 4 mo after the mutagenesis treatment, fixed germline mutations (those mutations affecting premeiotic germ cells at the time of treatment) were recovered from the males by breeding them to wild-type AB females (Fig. 1). The resulting F1 generation consisted of approximately 1000 progeny that were heterozygous (+/M) at a given mutagenized locus. These mutations were distributed to the F2 generation in one of two types of single pair matings; either by crossing two heterozygous F1 fish (+/M x +/M), or by crossing a heterozygous F1 with a wild-type AB fish (+/M x +/+). Forty-four families from +/M x +/M crosses and 37 families from +/M x +/+ crosses were raised, representing a total of 125 mutagenized genomes. To screen for mutations affecting the reproductive system, F3 progeny were produced through individual pair matings within a given F2 family, giving rise to four possible genotypes (+/+, +/M, M/+, M/M). At least six crosses that generated progeny were produced for each F2 family.
|
Histological Screen
At 3 mo of age (sexual maturity in zebrafish), families of F3 progeny were killed. The heads and tails were removed with a scalpel and the remaining bodies were placed in Altmanns fixative [11] for a minimum of 2 days. The bodies were washed in tap water for 24 h, decalcified (CalEx; Fisher, Pittsburgh, PA) for 24 h, washed again in tap water for 24 h, and placed in 50% alcohol until further processing. Using a Shandon Citadel 2000 automatic tissue processor, the bodies were dehydrated in a series of increasing alcohol, cleared in toluene, and then infiltrated with paraffin. Processed bodies from individual pair matings were mass embedded in parallel in paraffin blocks (>15 bodies/block; 3.0 x 2.5 cm block; Fig. 2A). Blocks were initially trimmed 2100 µm in from the face and placed in tap water for 2 days to soften. Sections (7 µm) were then taken at the face, and at 350 and 700 µm further into the block. These three sections were sufficient to observe the gonads from all of the individuals in a block regardless of size. A single slide (75 x 50 mm) was made from each region (Fig. 2B), stained with hematoxylin/eosin, and visualized under a compound microscope. Because of the parallel orientation of the bodies, slides contained cross-sections through the body cavity of all fish within a block (Fig. 2B). The cross-sections of the ovaries and testes in each individual (Fig. 2C) were analyzed for alterations in gonadal structure and gametogenesis and compared with gonads of wild-type fish. The cellular classification of the stages for spermatogenesis and oogenesis in wild-type zebrafish gonads was based on descriptions published for several other fish species [12, 13].
|
Both parents from F3 families containing individuals with altered gonads were outcrossed to wild-type fish. After 3 mo the progeny from the outcross were randomly pair-mated and the progeny of that incross were screened histologically as described above. Only altered gonadal phenotypes that appeared after the outcross/incross were considered to be the result of mutations in the germline. Mutant alleles are being maintained as heterozygous stocks that have been re-established periodically by outcrossing, followed by random incross mating and histological screening to identify new carrier parents.
RESULTS
Using the histological screen, several mutations in the gonads were observed (Table 1). In general, there were fewer mutations observed in the testes than in the ovaries and all mutations had visible phenotypes in only one sex per family.
|
The testes of zebrafish are arranged in lobules that contain germ cells undergoing spermatogenesis. Within a lobule, spermatogenesis occurs synchronously and therefore, lobules containing only spermatogonia, spermatocytes, or spermatids can be observed (Fig. 3, D and E). Following maturation of the spermatids, the connective tissue borders of lobules break down, forming large areas of mature sperm (Fig. 3E) that are contiguous with the sperm ducts. Testes of 3-mo-old wild-type zebrafish contain all stages of spermatogenesis and in general are cytologically consistent from one individual to another, regardless of body size. Two testicular mutations were observed that had distinct phenotypes characterized by the stage of spermatogenesis. In one mutant, the testes contain large numbers of spermatogonia with very few spermatocytes (Fig. 4, A and B; Table 1, 4003). In the second mutation, the testes contain primarily spermatocytes with a few nests of spermatogonia (Fig. 4, C and D; Table 1, 4113). In both mutations, the testes were devoid of mature sperm.
|
|
Zebrafish have "asynchronous" ovaries and at 3 mo of age, ovaries of medium to large wild-type females normally contain a hierarchy of follicles (i.e., follicle wall and enclosed oocyte) with oocytes ranging from the early perinucleolus stage to yolk globule stages (Fig. 3, A–C). If sampled at the correct time of day, ovulated oocytes may also be observed in the ovaries (not shown). Unlike the situation in males, small zebrafish females normally do not contain oocytes in later stages of oogenesis. Thus, identification of mutant phenotypes involving oogenesis is problematic. As a result, mutations affecting the progression of oogenesis were not pursued. In addition, all mutations in the ovaries were confirmed to be present in at least some large females in a family.
Several mutations were observed in the follicles, in the tissues that surround the follicles, or both. In the ovaries of wild-type females, the oocyte is enclosed by a thin follicle wall that is only a few cell layers thick (Fig. 3C). In one mutation, either the follicle wall or ovarian stroma proliferates and encircles developing oocytes (Fig. 5, A–F; Table 1, 4090). The result is the degeneration and resorption of the oocyte, which appears to be replaced by a tissue mass of unknown origin (Fig. 5, D–F). In another mutation, large accumulations of homogenous extrafollicular tissue are observed in areas within the ovary (Fig. 5, G–I; Table 1, 4086). Unlike the previous mutation, this mass of tissue does not specifically encircle follicles and areas of the ovary outside of this tissue appear normal.
|
Two mutations appeared to be related specifically to the follicle wall structure or the entire follicle. In one mutant, the ovary contains a large number of empty follicles that are lined with inclusions containing a pink staining substance (Fig. 6, A–C; Table 1, 3088). Given the coloration, it is likely that this substance is the cytoplasm or yolk of oocytes being reabsorbed. In another mutant, it appears that postovulatory follicles (presumed) are proliferating, resulting in very large numbers of dark staining cells in the presumptive follicle wall (Fig. 6, D–F; Table 1, 3124).
|
Several ovarian mutations were observed in which extensive oocyte degeneration was occurring (Fig. 7, A–I; Table 1, 3027, 4163, 3150B). In these mutants, ovaries were frequently characterized by large accumulations of chorions (Fig. 7, C and F) and large amounts of pink staining fluid (Fig. 7, D and G) that presumably represents the cytoplasm of degenerating oocytes. It is likely that the yolk and cytoplasm of the degenerating oocytes is being reabsorbed in these mutants because masses of cells that are presumed to be leukocytes could also be observed in these ovaries (Fig. 7, C and I). While some of the characteristics of the degenerating ovaries were similar, there were also differences. For example, in one of the mutations it was clear that intrafollicular oocytes were degenerating because a nucleus and follicle wall could still be observed in some of the degenerating oocytes (Fig. 7A). In contrast, based again on the structure of the degenerating oocytes, it appeared that others involved postovulatory oocytes (Fig. 7, D and G).
|
DISCUSSION
In the present study, we demonstrated that a general histological screen could be used with adult zebrafish to uncover gonadal mutations induced by ENU. This screen was small (<5%) compared with earlier developmental screens [1–3], but was intended to show that the zebrafish model can be used to study physiological processes in adult fish. Many of the traits that make zebrafish conducive to molecular genetic studies in embryos also facilitate physiological screens in adults. However, there are some difficulties in using the zebrafish to study reproduction or, generally, any physiological process in adults. Because adult fish are screened, the crossing schemes require an additional 3 mo beyond that observed for any early developmental screens. In addition, it is not possible to employ some of the haploid methodology that has been used to shorten the screening time in studies on development. Haploid embryos can develop abnormally and the number that survive to adulthood is much lower [14]. Also, if reproductive mutations affect only one phenotypic sex (as we observed), then only 12.5% of the progeny from any one heterozygous pair would be phenotypically mutant. This means that fewer of the F3 generation will express the mutation. In addition, because the entire genome is mutagenized, many families that carry reproductive mutations also carry embryonic lethal mutations and do not survive to 3 mo of age. This can reduce the number of progeny within a F3 family, at times below the number that would be reasonable for screening. Therefore, because of inherent problems in obtaining large numbers of reproductive mutant individuals, fine-scale mapping of the reproductive mutations obtained using ENU may be difficult. Instead, a mutagenesis approach involving retroviruses [3] or transposons [15], combined with our histological screen, could be used more efficiently to uncover genes that are responsible for reproductive mutations.
The mutations observed in this screen appeared to alter spermatogenesis, and various aspects of ovarian morphology and cytology. During the screen, other reproductive abnormalities were observed, but the characteristics of those abnormalities indicated that they were not the result of a mutation, or made it difficult to determine if they actually were mutations. For example, several families were found to have apparent mutations affecting oogenesis. These families produced offspring that had ovaries containing only early stage oocytes. However, unlike males, in which body size generally appeared to have little affect on testis development, ovarian development appears directly linked to size. This makes identifying a potential oogenesis mutation difficult unless the mutation is seen in a number of large fish. Frequently, this was not the case and, therefore, these mutations were not pursued. There were also several families that produced offspring that possessed an ovotestis, a gonad containing at least one mature or maturing oocyte surrounded by testicular tissue. However, compared with the mutations described here, the occurrence of this structure was more frequent. In addition, when it was observed in a family, it occurred in only one individual. The presence of an ovotestis can probably be attributed more to the plasticity of the phenotypic sex in certain fish species than the result of a mutation. For example, in several fish species the temperature of the water in which young fish are reared can strongly influence the phenotypic sex of resulting progeny [16, 17].
Several mutant phenotypes, including the testis mutations arrested at the spermatocyte stage (3150A and 4113), and mutations involving the degeneration of oocytes (4163, 4011, 3027, and 3150B), have appeared in more than one family. We have not performed complementation analyses on families with similar phenotypes. Therefore, at this point it is unknown whether these represent mutations in the same or different genes. However, it is unlikely that all of the degenerating oocyte phenotypes represent the same mutation because there are obvious differences in the type of oocytes that are being reabsorbed (e.g., preovulatory versus postovulatory oocytes).
The lack of mutations affecting both sexes indicates that the mutations observed in this study probably involve processes acting at the gonadal level. Mutations acting at the hypothalamic-hypophyseal level would be more likely to affect the phenotypes of both males and females, unless the mutation alters a mediator that is still able to act in a sex-specific manner. Mutations could be the result of defects in various cellular mediators of gametogenesis or gonadal maturation such as cell cycle regulators or hormones. However, given the complexity of the reproductive system, some of the mutations we observed could result from abnormalities at several different levels. For example, in some of the degenerating oocyte phenotypes, it is possible that a mutation that affects the structure of the ovary or even reproductive behavior could ultimately block the release of oocytes from the ovary and result in their degeneration.
The zebrafish has the potential to be a valuable model for the study of a number of physiological processes. We have shown that the zebrafish can be chemically mutagenized and used to study mutations that affect the reproductive systems of adults. However, the production of reproductive mutants that are tagged may be necessary to isolate the specific genes responsible for these phenotypes.
FOOTNOTES
1 This research was supported by grant 99-35203-7718 from the U.S. Department of Agriculture to F.W.G. 
2 Correspondence: Frederick William Goetz, Department of Biological Sciences, University of Notre Dame, P.O. Box 369, Notre Dame, IN 46556-0369. FAX: 219 631 7413; goetz.1{at}nd.edu
Accepted: September 19, 2000.
Received: June 7, 2000.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  I M McGonnell and R C Fowkes Fishing for gene function - endocrine modelling in the zebrafish. J. Endocrinol., June 1, 2006; 189(3): 425 - 439. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. King Heiden, M. J. Carvan III, and R. J. Hutz Inhibition of Follicular Development, Vitellogenesis, and Serum 17{beta}-Estradiol Concentrations in Zebrafish Following Chronic, Sublethal Dietary Exposure to 2,3,7,8-Tetrachlorodibenzo-p-Dioxin Toxicol. Sci., April 1, 2006; 90(2): 490 - 499. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Sakai Transmeiotic differentiation of zebrafish germ cells into functional sperm in culture Development, March 9, 2003; 129(14): 3359 - 3365. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Spitsbergen and M. L. Kent The State of the Art of the Zebrafish Model for Toxicology and Toxicologic Pathology Research--Advantages and Current Limitations Toxicol Pathol, January 1, 2003; 31(1_suppl): 62 - 87. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
