The Journal of Heredity 2002:93(2)
© 2002 The American Genetic Association 93:91-99
Mutation in a Sex-Determining Gene in Rainbow Trout: Detection and Genetic Analysis
From Unité ‘Génétique des Poissons’, Institut National de la Recherche Agronomique, 78 352 Jouy-en-Josas cedex, France (Quillet and Aubard), and Salmoniculture Expérimentale Marine IFREMER-INRA, Le Drennec, BP 17, 29 450 Sizun, France (Quéau).
Address correspondence to E. Quillet, Unité de Génétique des Poissons, INRA, Domaine de Vilvert, 78 352 Jouy-en-Josas cedex, France, or e-mail: edwige.quillet{at}diamant.jouy.inra.fr.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In rainbow trout (Oncorhynchus mykiss), the acknowledged sex-determining system is genetic sex determination (GSD) with female homogamety (
XX-
XY). Subsequently, mitotic gynogens are all expected to be females. Unexpected maleness was fortuitously observed in a mitotic gynogenetic family of rainbow trout (13 males out of 27). An equal ratio of males and females suggested the possible segregation of some Mendelian sex-influencing factor. In order to perform a comprehensive analysis of the inheritance and expression of the factor involved, the transmission of maleness was studied across the next three generations, using both conventional and/or meiotic and mitotic gynogenetic offspring. On the whole, males as well as intersexes were observed in crosses between two expected carrier parents, and in gynogenetic offspring of expected carrier females, but not in crosses between one expected carrier parent and one normal XX control. Sex ratios in the different crosses often fitted Mendelian proportions, but not always. Both excess and lack of maleness were observed. The simplest hypothesis consistent with most results is a one-locus model, assuming the existence of a mutation (termed mal) of a sex-determining gene, which is able to override the primary XX mechanism of sex determination and to induce the development of testicular tissue in the gonads of expected XX individuals. The one-locus model requires that the mal mutation usually, but not systematically, behave as a recessive mutation and have a limited penetrance, that is, heterozygous (mal/+) may be sex reversed, homozygous (mal/mal) may remain female, and carrier individuals may undergo partial masculinization alone (many intersexes were recorded). Inconsistency in sex ratios among offspring of parents expected to respond the same way was recorded, indicating that other modifier loci may also be involved. Finally, the occurrence of both males and females in clonal progenies showed that epigenetic factors also likely influence the expression of maleness. The effects of the mal mutation are compared to similar mutations recently described in other fish species. The nature and location of the mal gene (carried by heterochromosomes or an autosomal pair) is briefly discussed in view of the knowledge recently acquired on the subject. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The acknowledged system of sex determination in rainbow trout (Oncorhynchus mykiss) is a GSD one (genetic sex determination versus ESD, environmental sex determination) with female homogamety (
XX-XY) as in most other salmonids. Still, sex chromosomes are hardly recognizable on conventional karyotypes (Chourrout and Happe 1986; Thorgaard 1983). It has been possible to recognize the sex chromosome pair only recently using more sophisticated techniques such as fluorescence in situ hybridization FISH (Iturra et al. 1998; Moran et al. 1996). To our knowledge, there is as yet no unequivocally sex-specific molecular marker or probe available in rainbow trout (Allendorf et al. 1994; Iturra et al. 1998; Nakayama et al. 1994; Sakamoto et al. 2000). The first evidence of female homogamety came from investigations carried out in progenies of hormonally sex-reversed individuals (XX males, XY females, or induced hermaphrodites; see review by Chourrout 1988). The analysis of sex ratios in uniparental families (gynogens or androgens) further supported the model. As a matter of fact, gynogenetic offspring are usually all female, and studies of androgens have confirmed the existence of YY males (Parsons and Thorgaard 1985; Scheerer et al. 1991). Yet scattered observations that deviate from the simple XX-XY model have been reported in rainbow trout. Males have occasionally been detected among meiotic gynogens (1 out of 50 in Chourrout and Quillet 1982), as well as in the offspring of assumed XX males (Feist et al. 1995; Okada et al. 1979; Tsumura et al. 1991). On the other hand, females have sometimes been recorded in the offspring of expected YY supermales. Chevassus et al. (1988) found 2 females out of 399 progenies of 4 supermales obtained by self-fertilization. Scheerer et al. (1991) recorded 12 and 20 females out of 16 and 22 offspring, respectively, of two homozygous androgenetic males. Data from other salmonid species confirm that female homogamety is the usual rule in the family, and it is difficult to explain the observations made in trout from those observations alone.
Such deviations from the monofactorial XX-XY system (as well as from the ZZ-ZW system) have been described in other fish species. Some are well documented, and several explanations could be proposed. The existence of mutant heterochromosomes (Y* or W* with reduced dominance) has been evidenced from progeny testing in the platyfish (Xiphophorus maculatus; Kallman 1984) and suggested in Leporinus elongatus from the use of molecular probes (Baroiller et al. 1996; Nakayama et al. 1994).
Another explanation proposed is the presence of Mendelian sex-determining "autosomal" factors. Though authors have often termed those factors "autosomal," in our opinion, data do not always unequivocally demonstrate that they are carried by autosomes, and the model with multiple allelic heterochromosomes should not be formally excluded. The existence of such factors was first identified in the guppy (Poecilia reticulata) by Winge and Ditlevsen (1947): after a few generations of selection from spontaneous XX males, the authors obtained an XX strain with a balanced sex ratio. They explained their result as the accumulation of numerous autosomal male-determining genes overriding the sex genes of the X chromosome. The same explanation was proposed in more recent studies, especially in carp (Cyprinus carpio) and tilapias (Oreochromis sp.). Studying mitotic gynogens and their progenies and backcrosses in carp, Komen et al. (1992a,b, 1995) described a mutation (that they termed mas-1) that induces maleness in XX females. In tilapias, a number of articles by different authors support the conclusion that in Oreochromis niloticus, there are one or more minor sex-influencing loci that cause female to male sex reversal in XX individuals (Hussain et al. 1994; Mair et al. 1991a; Müller Belecke and Hörstgen-Schwark 1995; Sarder et al. 1999), as well as occasional females in the progeny of "YY" males (Mair et al. 1997). Similarly, a mutation preventing ZZ individuals from becoming males was suspected in Oreochromis aureus (Mair et al. 1991b).
Finally, the role of environmental factors in sex determination in fish is evidenced in an increasing number of studies. Environmental factors, and especially temperature, may play a crucial role in sex differentiation in some species, though this influence is highly variable from one species to the next (see Baroiller et al. [1999] for a recent review). A few studies on clonal fish have clearly demonstrated the role of epigenetic factors. Three of the four clones of carp tested by Komen et al. (1992a) contained a small proportion of intersexes or males, despite the genetic identity among individuals. Similarly Sarder et al. (1999) recorded males in one of six clones of Nile tilapia. In salmonids, GSD seems to be the rule, and very few reports on ESD modification of sex are available (Craig et al. 1996; Nagler et al. 2001).
In the course of a research program on mitotic gynogenesis in rainbow trout, we fortuitously recorded an unexpected maleness in one mitotic family (13 phenotypic males out of 27 mitogens). Though the number of fish was quite small, the 1:1 sex ratio might suggest the Mendelian segregation of some sex-modifying genetic factor in that family. In order to analyze the inheritance and expression of the factor involved, we analyzed the sex ratios during the next three generations, using both conventional and/or meiotic and mitotic gynogenetic offspring as experimental groups.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Experimental Fish
Rainbow trout (O. mykiss) breeders belonged to the SY strain reared at the Institut National de la Recherche Agronomique (INRA) experimental farm (Gournay-sur-Aronde, Oise, France). The strain is the result of the initial introduction of several domestic populations from the United States and France that were pooled and then maintained as a single population by random mating over several generations. Unless otherwise specified, the different experiments were carried out at this farm.
The first experiment (Exp.A0) was performed in November 1991 as part of a program of production of homozygous mitotic gynogens. Twenty-two individual SY females were used to produce both mitotic gynogenetic lines and control families (fertilization of ova of each SY female with a mixture of milt from 7–12 males). When fish were 1 year old, mitotic lines (EM) were mixed after fin clipping, while control families were gathered anonymously into three subgroups. Sixteen mitotic lines were studied for sex ratio and other traits. Fish from one particular mitotic line with a distinctive sex ratio (line EM1, descended from female SY1, 13 phenotypic males out of 27 two-year-old individuals) were used as breeders for the present study. Experimental crosses were then performed across the next three generations.
Experiment A1 (Exp.A1) was designed to study the sex ratio in offspring of EM1 males when mated with different types of females. The crosses were performed in November 1993 and 1994. In Exp.A1a, a single EM1 male was mated with a pool of 15 unrelated control females. In Exp.A1b, two EM1 males were mated with two of their mitotic sisters. As EM1 fish were homozygous mitotic gynogens, the resulting families were four clones reared all together as of the eyed stage. Sex ratio was then recorded within the whole mixed group which was called EM1G2. Each sexed fish was assigned to its parents from genotypes at the microsatellite loci. In Exp.A1c, 7 EM1 males were mated with 18 individual females from the SY control subgroup that contained their control sisters. These females were termed D, as daughters of SY mothers, that is, control daughters of SY1 anonymously mixed with the daughters of five other SY females. Several factorial crosses involving individual EM1 males and D females were performed. Because of experimental limitations, offspring were sent at the eyed stage to another experimental farm (SEMII, Finistère, France) where they were partially mixed (47 families distributed into 10 identical tanks) after adjustment for an equal number of eggs. Rearing conditions were controlled and standardized among tanks (water flow, density, feeding rate). Sex ratios were recorded for juveniles within each tank, and the parents that the fish originated from were assessed from the analyses of microsatellite markers in SY females, EM1 males, D females, and offspring.
Experiment A2 (Exp.A2) was designed more precisely to study the offspring of D females that had given birth to males in Exp.A1c. Female D12 was the only one still available at the time of the experiment (December 1996). It was mated with normal control males and XX unrelated neomales (XX sex-reversed females), and reproduced by both meiotic and mitotic gynogenesis. Sex ratios were checked in the four types of progenies.
Experiment A3 (Exp.A3) was carried out to analyze the progenies of EM1G2 fish. In December 1995, seven EM1G2 males were individually mated with a pool of three control females. One year later, five EM1G2 females were also tested. They were crossed with one unrelated XX sex-reversed male and one of the EM1G2 males tested previously. They were also reproduced by both meiotic and mitotic gynogenesis. Yet, survival was quite low in most of the crosses (very likely because of the high consanguinity of the females themselves and of the resulting progenies), and only two EM1G2 females could be used in the end.
Lastly, experiment A4 (Exp.A4) was performed in November 1999. The goal was to use progeny testing to select females that carried the masculinizing factor in the meiotic gynogens of female D12 in order to maintain a carrying line. Eight gynogenetic females were used for a new round of mitotic gynogenesis, but only two produced mitotic offspring large enough to be kept until sexing (F59 and F62).
Mitotic and Meiotic Gynogenesis
Meiotic gynogenesis was induced using the methods described in Chourrout and Quillet (1982): sperm was ultraviolet (UV) light inactivated and early heat shock (26.5°C for 20 min) was applied to prevent polar body extrusion. Mitotic gynogenesis was induced using late heat shock (31°C for 5 min) as described in Diter et al. (1993). In each case, donor males used for UV irradiation were homozygous for the "golden" dominant gene of depigmentation, and haploid control batches (ova fertilized with irradiated sperm and no thermal shock) were produced for every single female.
Genetic Controls and Parental Assignment
In Exp.A0, the mitotic gynogenetic origin of EM fish was checked with six polymorphic isozymes: esterase (EST-1), isocitrate dehydrogenase (sIDHP-1 and sIDHP-2), malate dehydrogenase (sMDH-B1,2), phosphoglucomutase (PGM2), and superoxide dismutase (sSOD1). Three microsatellites were analyzed as an additional control of the origin of individual fish that had a male phenotype: Str15INRA, Str60INRA, and Str73INRA (GenBank accession numbers AB001058, AB001057, and AB001056, respectively). In the EM1 line, every adult fish was checked for five microsatellite loci, whatever its sexual phenotype: OmyFGT2Tuf (previously FGT2; Sakamoto et al. 1994a), Str2INRA (Estoup et al. 1998), Omy18INRA (Gharbi K, unpublished), and OmyPuPuPyDU and Omy77DU (Morris et al. 1996).
Microsatellites were used to assign each individual progeny to its parent couple when families had been mixed. Two microsatellites in Exp.A1b (OmyFGT2Tuf and Omy77DU) and three in Exp.A1c [FGT1 (Sakamoto et al. 1994b), OmyFGT2Tuf, and Omy77DU] were enough to trace the genetic origin of the fish in the required groups.
The electrophoretic procedures for horizontal starch-gel electrophoresis were as described in Guyomard and Krieg (1983) and Krieg and Guyomard (1985). The genetic determinism of the electrophoretic variation in isozyme loci was demonstrated in Diter et al. (1988) and Guyomard (1984).
Microsatellite polymorphism was analyzed using the usual techniques: DNA extraction (from blood or fin tissue), polymerase chain reaction (PCR) amplification, electrophoresis on standard DNA sequencing gel, and autoradiography. Primers, sequences, and specific conditions of PCRs have been published in the above-mentioned publications and in Estoup et al. (1993).
Determination of Sex Ratios and Statistical Analyses
The sex ratios were most usually determined for juveniles (6–11 months old). Fish were sacrificed and both gonads were dissected, slightly squashed, and directly examined over their entire length under a microscope. The presence of ovarian lamellae was the criterion used to attribute female sex, whether oocytes were present or not. Fish were scored as intersexes when at least one of the gonads was an ovotestis or when one ovary and one testis were observed simultaneously. In some experiments, sex ratio was determined at adulthood: in these cases, sex determination relied on phenotype only (secondary sexual traits and/or spontaneous production of gametes by abdominal pressure).
Comparisons of observed sex ratios to expected proportions were performed using the chi-squared test corrected for continuity (Sokal and Rohlf 1981). Comparisons among sex ratios were performed using the two-way tables test of independence.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Experiment A0
Sex ratios recorded in the different EM lines are presented in Table 1. Considering the genetic determination of sex in rainbow trout (XX-XY), only females were expected in gynogenetic progenies. Yet at 8 months, six of the nine sampled lines contained fish with completely or partially male gonads. Some of the females (about 4%) exhibited reduced fertility (few or no visible oocytes). At adulthood, fish of male appearance were again observed in several lines: 5 of the 13 lines contained males or intersexes. Only fish with a clear male phenotype were sacrificed (except in line EM1, where they were all kept for further crosses) and their gonads dissected. It is therefore possible that intersexes of ambiguous phenotype or limited male appearance were not identified. On the whole, males or intersexes were found in 7 of 16 EM lines. The mean proportion was about 10% in juveniles and 5% at adulthood (which may be an underestimation, as some intersexes may not have been detected by phenotype).
|
Residual paternal contribution would be the most obvious explanation for the presence of males in gynogenetic offspring. Yet all of the 99 fry that hatched in the 16 haploid control batches (16,000 ova overall) were normally pigmented, and no golden fry were observed in the mitotic batches themselves, which supports high-efficiency sperm irradiation. The normal fry (0.6%) very likely came from spontaneously unreduced diploid eggs. In the same way, controls performed at the isozymes and microsatellite markers in juveniles and 2-year-old adults showed that males and intersexes were all homozygous and carried only alleles present in the mother. These controls do not definitely rule out the risk of some paternal contribution, as genetic markers were not analyzed in golden donor males that may have shared common alleles with females. Yet EM fish were all homozygous: in that case, paternal contribution is not detected in each given offspring, only when it supplies the same allele as that supplied by the mother. For a given locus, the probability decreases when the probability that parents are heterozygous increases, which is the case with microsatellites which are far more variable than isozymes. For a given fish, the probability decreases as the number of markers increases. Considering the number of markers that were used, it seems unlikely that a substantial paternal contribution alone explains the results observed.
We then focused on line EM1 which had quite a large number of male progenies: 43% of the total sample (versus 5% as a mean value in other lines), which was not significantly different from a 1:1 sex ratio (
21df = 0.73; P = .39). Males from other lines were not studied further.
Experiment A1
Experiment A1 was designed to check the assumption that SY1 was heterozygous for some sex-determining locus and, consequently, that EM1 males were homozygous at that locus. Mating of EM1 males with control females (Exp.A1a) or with their EM1 mitotic sisters (Exp.A1b) was expected to produce offspring of the same genotype as the SY1 mother, that is, all females. The sex ratios observed in the two experiments are presented in Table 2. In Exp.A1a, only females were recorded, though some were partially (greatly reduced number of oocytes) or completely sterile (1 and 2 of 91, respectively). In contrast, in Exp.A1b, many fish (16 of 24) displayed a male phenotype. Yet 8 of the 16 phenotypic males did not produce milt by abdominal pressure, and may have been intersexes. As survival in this experiment was very low (2% from the eyed stage to maturation at 2 years), we carried out microsatellite analyses to check the status of survivors and to assess the actual contribution of the four initial homozygous EM1 parents. Twenty-two of the 24 remaining fish had the same EM1 mother (6), while the 2 EM1 males (3 and 6) were equally represented. The two cloned progenies from 6 contained both females and males. The two remaining fish belonged to the other two clones (43 and 46) and were phenotypic males.
|
The purpose of Exp.A1c was to check whether SY1 had actually transmitted some masculinizing gene to its control daughters. Such daughters would be heterozygous for the gene, and when mated with EM1 spontaneous males should give birth to 50% heterozygotes (remaining female, as did their mother and SY1 grandmother) and 50% homozygous offspring that would express maleness. As the experiment was not initially designed for this purpose, daughters of SY1 were anonymously mixed with daughters of five other SY females. There were 18 D females available in the control group: thus the expected number of SY1 daughters was three, and the number of heterozygous daughters for the gene of interest was half of that (i.e., one or two females in all probability), which was quite low. Consequently every D female was used in the experiment, and microsatellites were used for assignment to SY mothers. Sex ratios are reported in Tables 3 and 4. Males or intersexes were observed in 4 of the 10 tanks. Clearly, maleness depended on the D female rather than on males. Besides the fact that tanks were managed under controlled conditions, the results in tank 1 strongly support the idea that expression of maleness is due to a female effect rather than a tank effect: D1 and D2 progenies were incubated in the same recirculating unit (10°C) and then reared in the same tank as the eyed stage, and only one contained males and intersexes (13 of 18 for D2 versus 0 of 20 for D1). Parental assignment was performed in tanks 1 and 6 only (detailed in Table 4). Three females had males and intersex offspring (D2, D11, and D12). It is striking to note that all three were putative daughters of SY1. The frequency of sex-reversed fish was 72%, 64%, and 57% for D2, D11, and D12, respectively, which was not significantly different (
21df < 0.50; P > .48). Similarly, no effect of EM1 males on sex ratio could be detected (highest 21df = 0.44; P = .51). The overall mean frequency of males was not different from 0.5 (33 of 66). Yet the total frequency of sex-reversed fish ( + I = 64%) was different from 0.5 (21df = 4.38; P = .0364).
|
|
Families of tanks 3 and 9, containing only a few males or intersexes, were not studied further. Yet one should emphasize the fact that, in both tanks, female SY5 was involved as the putative mother and remember that line EM5 also contained males or intersexes (see Table 1). All these results show that SY1 (and SY5 very likely) has actually transmitted some masculinizing factor to its daughters through normal mating, and they rule out uncontrolled parental contributions after gynogenesis as an explanation for the maleness observed in Exp.A0.
Experiment A2
Female D12 (assumed to be heterozygous for the masculinizing factor, just like SY1) was used for further analysis of the transmission of male phenotype in normal and gynogenetic progenies. The aim was to confirm the results obtained with SY1 using additional controls and to compare the frequency of maleness in the two types of gynogenetic progenies (results in Table 5). The expected 1:1 sex ratio was observed in the cross with control males (21df = 0.48; P = .49), and only females were detected in the cross with XX males. Homozygous mitotic progenies contained 58% partially or completely male fish, a proportion that is not different from a 1:1 ratio (21df = 1.07; P = .30). Sex reversal ( + I) was lower in meiotic gynogens (18%) than in mitotic gynogens (21df = 18.8; P = 2.10-5).
|
Experiment A3
As maleness in the EM1G2 group was unexpectedly high, additional crosses were performed to study transmission and expression in the next generation. Two EM1G2 females (no. 6 from clone
63 and no. 7 from clone 66) and seven males (from the same clones plus clone 46) were used as breeders. In that particular experiment, gynogenesis was not completely successful, as two golden fry were observed in the haploid control (more than 500 ova). Yet all the gynogenetic groups were fertilized with the same pool of irradiated sperm, which made it possible to make between-group comparisons.
Crossing EM1G2 males with control females produced only females (60 progenies per male; i.e., 420 fish sexed). Five had either no visible or few oocytes. The sex ratios of the offspring of the two EM1G2 females are reported in Table 6. Only females were recorded in the crosses with the control XX neomale (110 females, of which 5 were partially or completely sterile). Males and intersexes were observed in all other groups. In female 6, sex ratios were not different from 1:3 in the cross with the EM1G2 male and from 1:1 in mitotic progeny, respectively. On the other hand, sex ratios were female biased in progenies of female 7 (Table 6), though the differences between the two females were not significant (21df = 0.96 and 0.66; P = .33 and .42 in the two types of crosses, respectively). The frequencies of ( + I) in meiotic gynogenetic progenies were lower than in the mitotic groups in female 6 (21df = 9.92; P = .0017) and were the same in the two females (21df = 2.73; P = .098).
|
Experiment A4
Sex reversal was usually lower in meiogens than in mitogens of females assumed to be heterozygous for the masculinizing factor (Exp.A2 and Exp.A3). One possible explanation is that meiotic recombination at the presumed locus produces heterozygotes (assumed to remain female), which limits the expression of the trait to the residual frequency of homozygotes. Under this assumption, females in the meiogenetic progeny of D12 were expected to be either homozygous for the wild-type allele or heterozygous. Sex ratios observed in the mitogens from F59 and F62 are reported in Table 7. Offspring of F62 actually had an equal ratio of males and females (
+ I = 58%; 21df = 1.82; P = .18), indicating a probable heterozygous status. Unexpectedly, offspring of F59 were almost sex reversed ( + I = 83%; 21df = 34.26; P < 10-6). Moreover, only 7 of the 14 females had ovaries containing oocytes. These results strongly suggest that female F59 was homozygous for the mutation despite its phenotypic sex (gonads were not examined). Similarly its mitotic offspring were homozygous, but few remained female nevertheless.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A Model to Explain the Data
Unexpected maleness was observed in several mitotic gynogenetic lines of rainbow trout. The transmission of maleness through normal mating and across generations meant it was possible to rule out paternal contribution to gynogenetic progeny as the reason behind the observations. Inbreeding and the associated reduction in developmental homeostasis [according to the theory of Lerner (1954)] could also be proposed as an explanation for unexpected sexual phenotypes, and could account for occasional male mitogens in some lines. Yet the inbreeding effect per se (i.e., not caused by homozygosity at some unknown modifying loci) could not easily explain the maleness we observed steadily across generations.
The simplest hypothesis to explain the results overall is a one-locus model in which the SY1 female was heterozygous at some sex-determining locus and carried a recessive allele (that we shall call mal hereafter) able to override the usual XX mechanism of sex determination and induce the development of testicular tissue in the gonads of expected XX individuals. The results of Exp.A1a are consistent with the hypothesis that EM1 males were homozygous XX (mal/mal). When crossed with normal females (+/+), they produced only heterozygous progeny, that were indeed all female, as was the SY1 mother. Data from Exp.A1c also support the one-locus model: D2, D11, and D12 females were heterozygous (mal/+) daughters of SY1, and female D14, the last putative daughter of SY1 whose progeny contained only females, should therefore have received the wild-type allele from its mother. When crossed with any of the homozygous (mal/mal) EM1 males, heterozygous D females actually produced progenies of which about 50% reversed to males. Finally, the 1:1 sex ratio in mitogens of several females expected to be heterozygous (mal/+) from their parentage is consistent with this assumption (D12 in Exp.A2, EM1G2 no. 6 in Exp.A3, and F62 in Exp.A4).
Yet in a number of cases, the observed sex ratios did not fit the ones expected with the one-locus model alone. In some cases, an excess of males (intersexes included) was observed, indicating that heterozygotes (mal/+) can also be sex reversed. The proportion of such reversals was estimated at 27% in progenies of D females (Exp.A1c) and up to 67% in Exp.A1b, where EM1 males (mal/mal) were mated with their EM1 "sisters" expected to be (+/+). On the other hand, sex ratios recorded in Exp.A3 (female EM1G2 no. 7) and Exp.A4 (F59 and its progeny) suggest that the opposite phenomenon can also occur, that is, homozygotes (mal/mal) may sometimes keep the primarily all-female gonads.
Because of the pedigree of our fish (a single carrier ancestor), we monitored the transmission and expression of the same allelic form of the gene across the different generations and types of crosses. Therefore, under the one-locus model, the conclusions are that the mal mutation usually, but not systematically, behaves as a recessive mutation and has limited penetrance. At the population level, the consequence is that there is no unequivocal relationship between the genotype at the locus involved and gonadal sex. At the individual level, it results in some fish undergoing only partial sex reversal and remaining intersexes. A few females with apparently abnormal ovaries were also observed throughout the study. In the crosses with control parents, they may have been sterile triploids, resulting from spontaneous diploid eggs. In other groups, inbreeding resulting from gynogenesis and/or mating between relatives is likely to account for reduced fertility or a delay in gonad development (indeed, all tested fish have SY1 as a common ancestor). Yet one can also assume that in some cases the normal expression of female phenotype was restricted by the presence of the masculinizing factor in carrier females.
Although a one-locus model accounts for most results, it is possible that one or more other modifier loci may also be involved in the expression of maleness. As a matter of fact, the results of Exp.A3 suggested a possible effect of parents on sex reversal in offspring expected to respond in the same way with regard to their assumed genotype at the mal locus. While sex ratios fitted the expected ratios in offspring of female 6 (1:3 in the cross with the heterozygous EM1G2 male and 1:1 in mitotic gynogens), they were systematically skewed toward females in the same groups from female 7. Thus specific genetic background may contribute to the inconsistency of maleness expression between and within groups. Yet in the particular case of clonal progeny (Exp.A1b), genetic factors should not account for the occurrence of both males and females. Such a result indicates the role of epigenetic factors in the expression of maleness. The possible role of environmental factors on sex reversal was not investigated in the present study, but may explain some inconsistencies with data of experiments that were not carried out at the same time or same place. In particular, environmental differences may have contributed to the occurrence of apparently opposing penetrances in different experiments (maleness in mal/+ individuals versus the occurrence of mal/mal females).
Comparison of Results with Literature
The mutation we identified here in trout appears quite similar in terms of nature and effects to those described in other species. Mutations have essentially been described as recessive. The primary argument is that the detection of mutations usually relies on the analysis of heterozygous individuals that remained female. The second argument is that maleness is generally reduced or absent when carrier individuals are crossed with nonrelated individuals. An additional argument is that the proportion of males is very often lower in meiogens than in mitogens from the same mother. This is interpreted as the result of recombination between the minor sex-determining locus and the centromere, which produces heterozygosity at the gene and decreases the occurrence of expected males. For instance, in carp, Komen et al. (1992a, b) found 6.6% males in meiogens versus 47.7% in mitogens of the same parents. In tilapia, published data are quite consistent: 7.5% versus 47.5% with one female (Hussain et al. 1994), 0% versus 33.3% and 66.6% for two females (Müller-Belecke and Hörstgen-Schwark 1995), 12% versus 20% (Sarder et al. 1999) for the mean of four females. In the present study, frequencies were 14% versus 40% (mean of three females).
Yet mutations often show incomplete penetrance rather than strict recessiveness. Some heterozygous progeny can undergo sex reversal, as was confirmed by Komen et al. (1992a,b, 1995) in carp and in the present study. On the other hand, trout homozygous for the mutation may keep female gonads. There is no indication of that in carp, but in tilapia, the ratio of males is significantly lower than 50% in some, but not all, mitotic lines (Sarder et al. 1999). Similarly in this study we observed a lack of males in mitotic gynogens of female 7 (Exp.A3). In mitotic lines, selective mortality might also contribute to unbalanced sex ratios. Yet results obtained with female F59 support the hypothesis of incomplete penetrance in some cases fairly well.
Limited penetrance also results in the association between the mal mutation and the development of intersex individuals. Species we refer to in the present article (carp, tilapia, and trout) are all gonochoristic species, and hermaphrodites or intersexes have been exceptionally observed under natural or normal breeding conditions. In salmonids, very few individuals were reported: one in chum salmon (Oncorhynchus keta; Hikita and Hashimoto 1978) and another, more recently in a wild Scottish population of Arctic char (Salvelinus alpinus; Fraser 1997). Very few intersexes were observed in normal progenies of O. niloticus (9 of 2999 fish belonging to 59 families; Mair et al. 1991a), but in the same study, one particular female gave a progeny containing a high proportion of intersexes (5.8–25.6%). Of interest is that the experimental history led the authors to assume that intersexes were genetically female. In tilapia (O. aureus), Penman et al. (1987) also recorded intersexes in one out of six meiogenetic offspring. Yet in other studies dealing with genetic sex determination in tilapias, intersexes were not recorded. In the course of their work on the mas-1 mutation in carp, Komen et al. (1992a, b) observed a fairly high proportion of intersexes in meiogens (3%) and to a greater extent in mitogens (25%). These were usually predominantly male, and in all cases were classified as homozygous for the mutation. In trout, intersexes were usually not observed in crosses with controls, but in all other cases. More generally, we found intersexes associated with most cases of unexpected maleness, whatever the mitotic line affected (Exp.A0).
The effects of environmental factors on sex determination in fish are well documented (Baroiller et al. 1999), but in salmonids we only found two reports dealing with this. Female-biased sex ratios were observed by Craig et al. (1996) in two genetically distinct forms of sockeye salmon (Oncorhynchus nerka) and were parsimoniously attributed to the effect of temperature shifts during embryonic development. Recently Nagler et al. (2001) reported a high incidence of a male-specific DNA marker in wild phenotypic females of chinook salmon (Oncorhynchus tshawytcha) from the Columbia River: a survey of the local environment led the authors to conclude that genetic males had been sex reversed by exposure to adverse conditions during incubation (exposure to estrogens or to fluctuating temperatures, which would be more relevant for the present purpose). To our knowledge, there are no indications of such effects in rainbow trout. Conversely, several studies demonstrate the considerable stability of sex ratios under various thermal conditions (Baroiller et al. 1999; Van den Hurk and Lambert 1982). Nevertheless, environmental impacts on sex ratios are likely to occur when the primary sex-determining factors are bypassed (XX-mal system). For instance, Winge and Ditlevsen (1947) mentioned that the sex balance in the line selected from spontaneous XX males of guppy was more susceptible to external factors than in the XY system. Komen et al. (1995) also alluded to the influence of temperature and feeding level on the expression of the mas-1 mutation in carp.
Frequency and Nature of the Gene
In our strain we found 1 female out of 16 that unequivocally carried the mutation. Yet because of incomplete penetrance, other EM lines with some spontaneous males may also have originated from females carrying the same mutation or similar ones. In carp, Komen et al. (1992a) identified two nonallelic masculinizing factors, and the existence of such multiple mutations (allelic or not) in trout cannot be excluded. Moreover, deviations from expected sex ratios were hardly noticeable in adulthood, while the systematic observation of juvenile gonads revealed that maleness (associated to intersexuality) was not exceptional at all. Thus the actual frequency of mutation(s) may be largely underestimated in rainbow trout. As a matter of fact, the observations of Scheerer et al. (1991) match the XX-mal model we propose here quite well (the authors had themselves suggested that their data might reflect autosomal gene influences): the androgenetic male parents they tested, and that gave birth to almost all female offspring when mated with control females, behaved as would XX (mal/mal) individuals (the few males would be heterozygous, having undergone sex reversal). It is therefore very likely that the mutation (or a similar one) was also present in their original strain, which would indicate that the phenomenon is not that exceptional in rainbow trout. Yet the occurrence of deviations from the strict monofactorial XX-XY model of sex determination seems far less frequent in trout than in tilapia. In O. niloticus, for instance, different authors found 4 females out of 8 (Mair et al. 1991a), 10 out of 14 (Müller-Belecke and Hörstgen-Schwark 1995), and 4 out of 4 (Sarder et al. 1999) that gave male mitogens, but all worked with the same population originating from Lake Manzala (Egypt) and kept at the Institute of Aquaculture of Stirling, which may limit the implications of the data. The frequency of mas-1 in carp was not estimated [the single female reported in the experiment of Komen et al. (1992b) was heterozygous].
Meiotic offspring were produced in the present work with the hope that their sex ratios would provide an estimation of the recombination rate of the mal locus, and possibly an indication of whether it is carried by the Y chromosome (a "weak" Y) or by an autosomal one. Yet because of incomplete penetrance, phenotypic data are unreliable and recombination rates may be biased high. A tentative correction can be made from data of F59 mitogens: using the observed value of penetrance (0.83), the frequency of actual homozygotes can be reestimated from the observed sex reversal in the three meiotic gynogenetic families of Exp.A2 and Exp.A3 (6–19%, intersexes included). Corrected estimations of recombination rates remain quite high (more than 50%), which may indicate that the mal locus is quite distant from its centromere, as least in female meiosis.
On the other hand, genetic analyses of male meioses mapped the primary sex-determining locus very close to the centromere [Allendorf et al. (1994) with enzyme loci, Sakamoto et al. (2000) with microsatellites], while a more distal position (about 20 cM) was observed by Young et al. (1998) from amplified fragment length polymorphism (AFLP) data. Yet there are considerable sex-specific differences in recombination rates (Sakamoto et al. 2000), with map distances usually being higher in female meiosis. Therefore no reliable conclusion can be drawn concerning the relative location of the mal locus and the primary sex-determining factor. Genome scanning with autosomal and sex-linked microsatellites is now planned to address the issue.
|
Acknowledgments |
|---|
The authors thank J. P. Hiseux and E. Huchet for looking after the fish at the INRA trout farm; A. Fauré for having kindly accepting the fish of Exp.A1c in SEMII facilities and the farm staff for their friendly and efficient help; M. Andriamanga, A. Estoup, R. Guyomard, and K. Gharbi for their helpful advice with regard to enzyme and microsatellite analyses; and three anonymous referees for constructive criticism of the manuscript. Rearing costs of Exp.A1c were jointly supported by IFREMER and INRA.
|
Footnotes |
|---|
Corresponding Editor: Lisa Seeb
Received October 14, 2000
Accepted December 31, 2001
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-
Allendorf FW, Gelman WA, and Thorgaard GH, 1994. Sex-linkage of two enzyme loci in Oncorhynchus mykiss (rainbow trout). Heredity 72:498–507.
Baroiller JF, Guiguen Y, and Fostier A, 1999. Endocrine and environmental aspects of sex differentiation in fish. Cell Mol Life Sci 55:910–931.[CrossRef]
Baroiller JF, Nakayama I, Foresti F, and Chourrout D, 1996. Sex determination studies in two species of teleost fish, Oreochromis niloticus and Leporinus elongatus. Zool Stud 35:279–285.
Chevassus B, Devaux A, Chourrout D, and Jalabert B, 1988. Production of YY rainbow trout males by self-fertilization of induced hermaphrodites. J Hered 79:89–92.
Chourrout D, 1988. Revue sur le déterminisme génétique du sexe des poissons téléostéens. Bull Soc Zool Fr 113:123–144.
Chourrout D and Happe A, 1986. Improved methods of direct chromosome preparation in rainbow trout, Salmo gairdneri. Aquaculture 52:255–261.[CrossRef]
Chourrout D and Quillet E, 1982. Induced gynogenesis in the rainbow trout: sex and survival of progenies. Production of all-triploid populations. Theor Appl Genet 63:201–205.[CrossRef]
Craig JK, Foote CJ, and Wood CC, 1996. Evidence for temperature-dependant sex determination in sockeye salmon (Oncorhynchus nerka). Can J Fish Aquat Sci 53:141–147.[CrossRef]
Diter A, Guyomard R, and Chourrout D, 1988. Gene segregation in induced tetraploid rainbow trout: genetic evidence of preferential pairing of homologous chromosomes. Genome 30:547–553.[Medline]
Diter A, Quillet E, and Chourrout D, 1993. Suppression of first egg mitosis induced by heat shocks in the rainbow trout. J Fish Biol 42:777–786.[CrossRef]
Estoup A, Gharbi K, SanCristobal M, Chevalet C, Haffray P, and Guyomard R, 1998. Parentage assignment using microsatellites in turbot (Scophtalmus maximus) and rainbow trout (Oncorhynchus mykiss) hatchery populations. Can J Fish Aquat Sci 55:715–725.[CrossRef]
Estoup A, Presa P, Krieg F, Vaiman D, and Guyomard R, 1993. (CT)n and (GT)n microsatellites: a new class of genetic markers for Salmo trutta L (brown trout). Heredity 71:488–496.
Feist G, Yeoh CG, Fitzpatrick MS, and Schreck CB, 1995. The production of functional sex-reversed male rainbow trout with 17
-methyltestosterone and 11ß-hydroxyandrostenedione. Aquaculture 131:145–152.[CrossRef]
Fraser D, 1997. A hermaphrodite Arctic char from Loch Rannoch, Scotland. J Fish Biol 50:1358–1359.[CrossRef]
Guyomard R, 1884. High level of residual heterozygosity in gynogenetic rainbow trout, Salmo gairdneri Richardson. Theor Appl Genet 67:307–316.[CrossRef]
Guyomard R and Krieg F, 1983. Electrophoretic variations in six populations of brown trout (Salmo trutta L.). Can J Genet Cytol 25:403–413.[Web of Science][Medline]
Hikita T and Hashimoto S, 1978. A hermaphroditic chum salmon, Oncorhynchus keta, from the Chitose River with an example of its self-fertilization. Sci Rep Hokkaido Salmon Hatch 32:61–64.
Hussain MG, McAndrew BJ, Penman DJ, and Sodsuk P, 1994. Estimating gene-centromere recombination frequencies in gynogenetic diploids of Oreochromis niloticus L., using allozymes, skin color and a putative sex-determination locus (SDL-2). In: Genetics and evolution of aquatic organisms (Beaumont AR, ed). London: Chapman & Hall 502–509.
Iturra P, Medrano J, Bagley M, Lam N, Vergara N, and Marin JC, 1998. Identification of sex chromosome molecular markers using RAPDs and fluorescent in situ hybridization in rainbow trout. Genetica 101:209–213.[CrossRef]
Kallman KD, 1984. A new look at sex determination in poeciliid fishes. In: Evolutionary genetics of fishes (Turner BJ, ed.). New York: Plenum; 95–171.
Komen J, de Boer P, and Richter CJJ, 1992a. Male sex reversal in gynogenetic XX females of common carp (Cyprinus carpio L.) by a recessive mutation in a sex-determining gene. J Hered 83:431–434.
Komen J, Spaans P, Zhang DM, and Richter CJJ, 1995. Sex determination in common carp Cyprinus carpio L [abstract]. Aquaculture 137:106.[CrossRef]
Komen J, Wiegertjes GF, Van Ginneken VJT, Eding EH, and Richter CJJ, 1992b. Gynogenesis in common carp (Cyprinus carpio L.). III. The effect of inbreeding on gonadal development of heterozygous and homozygous offspring. Aquaculture 104:51–66.
Krieg F and Guyomard R, 1985. Population genetics of French brown trout (Salmo trutta L.): large geographical differentiation of wild populations and high similarity of domesticated stocks. Genet Sel Evol 17:225–242.
Lerner IM, 1954. Genetic homeostasis. Edinburgh: Oliver & Boyd.
Mair GC, Abucay JS, Skinbinski DOF, Abella TA, and Beardmore J, 1997. Genetic manipulation of sex-ratio for the large-scale production on all-male tilapia, Oreochromis niloticus. Can J Fish Aquat Sci 54:396–404.[CrossRef]
Mair GC, Scott AG, Penman DJ, Beardmore JA, and Skibinski DOF, 1991a. Sex determination in the genus Oreochromis. 1. Sex reversal, gynogenesis and triploidy in O. niloticus (L.). Theor Appl Genet 82:144–152.[CrossRef]
Mair GC, Scott AG, Penman DJ, Skibinski DOF, and Beardmore JA, 1991b. Sex determination in the genus Oreochromis. 2. Sex reversal, hybridisation, gynogenesis and triploidy in O. aureus Steindachner. Theor Appl Genet 82:153–160.
Moran P, Martinez JL, Garcia-Vazquez E, and Pendas AM, 1996. Sex chromosome linkage of 5S rDNA in rainbow trout (Oncorhynchus mykiss). Cytogenet Cell Genet 75:145–150.[Web of Science][Medline]
Morris DB, Richard KR, and Wright JM, 1996. Microsatellites from rainbow trout (Oncorhynchus mykiss) and their use for genetic study of salmonids. Can J Fish Aquat Sci 53:120–126.[CrossRef]
Müller-Belecke A and Hörstgen-Schwark G, 1995. Sex determination in tilapia (Oreochromis niloticus) sex ratios in homozygous gynogenetic progeny and their offspring. Aquaculture 137:57–75.[CrossRef]
Nagler JJ, Bouma J, Thorgaard GH, Dauble DD, 2001. High incidence of male specific genetic marker in phenotypic female chinook salmon from the Columbia River. Environ Health Perspect 109:67–69.[Web of Science][Medline]
Nakayama I, Foresti F, Tewari R, Schartl M, and Chourrout D, 1994. Sex chromosome polymorphism and heterogametic males revealed by two cloned DNA probes in the ZW/ZZ fish Leporinus elongatus. Chromosoma 103:31–39.[CrossRef][Web of Science][Medline]
Okada H, Matumoto H, and Yamazaki F, 1979. Functional masculinisation of genetic females in rainbow trout. Bull Jpn Soc Sci Fish 45:413–419.
Parsons JE and Thorgaard GH, 1985. Induced androgenesis in rainbow trout. J Exp Zool 231:407–412.[CrossRef]
Penman DJ, Shah MS, Beardmore JA, and Skibinski DOF, 1987. Sex ratio of gynogenetic and triploid tilapia. In: Proceedings of the World Symposium on Selection, Hybridisation and Genetic Engineering in Aquaculture, Bordeaux, France, May 27–30, 1986 (Tiews K, ed). Berlin: Heeneman Verlagsgesellschaft mbH; 267–276.
Sakamoto T, Danzman RG, Gharbi K, Howard P, Ozaki A, Sokkean K, Woram RA, Okamoto N, Ferguson MM, Holm LE, Guyomard R, and Hoyeim B, 2000. A microsatellite linkage map of rainbow trout (Oncorhynchus mykiss) characterized by large sex-specific differences in recombination rates. Genetics 155:1331–1345.
Sakamoto T, Okamoto N, and Ikeda Y, 1994a. Dinucleotide repeat polymorphism of rainbow trout, FGT2. J Anim Sci 72:2765.[Web of Science][Medline]
Sakamoto T, Okamoto N, Ikeda Y, Nakamura Y, and Sato T, 1994b. Dinucleotide-repeat polymorphism in DNA of rainbow trout, Oncorhynchus mykiss, and its application in fisheries science. J Fish Biol 44:1093–1096.[CrossRef]
Sarder MRI, Penman DJ, Myers JM, and McAndrew BJ, 1999. Production and propagation of fully inbred clonal lines in the Nile tilapia (Oreochromis niloticus L.). J Exp Zool 284:675–685.[CrossRef][Web of Science][Medline]
Scheerer PD, Thorgaard GH, and Allendorf FW, 1991. Genetic analysis of androgenetic rainbow trout. J Exp Zool 260:382–390.[CrossRef][Web of Science][Medline]
Sokal RR and Rohlf FJ, 1981. Biometry: the principles and practice of statistics in biological research, 2nd ed. San Francisco: WH Freeman.
Thorgaard GH, 1983. Chromosomal differences among rainbow trout Salmo gairdneri populations. Copeia 1983:650–662.[CrossRef]
Tsumara K, Blann VE, and Lamont C, 1991. Progeny test of masculinized female rainbow trout having functional gonoducts. Prog Fish Cult 53:45–47.[CrossRef]
Van den Hurk R and Lambert JGD, 1982. Temperature and steroid effects on gonadal sex differentiation in rainbow trout. In: Proceedings of the International Symposium on the Reproductive Physiology of Fish, Wageningen, The Netherlands, August 2–6, 1981 (Richter CJJ and Goos HJT, eds). Wageningen, The Netherlands: Pudoc; 69–72.
Young PW, Wheeler PA, Coryell VH, Keim P and Thorgaard GH, 1998. A detailed map of rainbow trout produced using doubled haploids. Genetics 148:1–13.
Winge Ø and Ditlevsen E, 1947. Colour inheritance and sex determination in Lebistes. Heredity 1:65–83.[Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||