The nematode C. elegans naturally develops as either an XO male or XX hermaphrodite. The sex-determining gene, tra-2, promotes hermaphrodite development in XX animals. This gene encodes a predicted membrane protein, named TRA-2A, which has been proposed to provide the primary feminising activity of the tra-2 locus. Here, we show that transgenic TRA-2A driven from a heat shock promoter can fully feminise the somatic tissues of XX tra-2 loss-of-function mutants, which would otherwise develop as male. TRA-2A is thus likely to provide a component of the tra-2 locus that is both necessary and sufficient to promote female somatic development. Transgenic TRA-2A driven by the heat shock promoter can also transform XO animals from male to self-fertile hermaphrodite. This result establishes the role of tra-2 as a developmental switch that controls somatic sexual cell fate. We show that a carboxy-terminal region of TRA-2A, predicted to be intra-cellular, can partially feminise XX tra-2 loss-of-function mutants and XO tra-2(+) males. We suggest that this intra-cellular domain of TRA-2A promotes hermaphrodite development by negatively regulating the FEM proteins.

Sex determination in the nematode C. elegans has been extensively characterised at the genetic level. Our current understanding of the genetic pathway that controls sexual fate is summarised in Fig. 1A. The primary determinant of sex is the ratio of X chromosomes to sets of autosomes (X:A ratio) (Madl and Herman, 1979). Diploid XX animals develop as hermaphrodites, whereas XO animals become males. A C. elegans hermaphrodite is essentially a self-fertile female: the hermaphrodite soma is indistinguishable from that of closely related female nematodes (Baird et al., 1994), but its germ line produces sperm first and then oocytes. A number of genes control sexual fate in response to the X:A ratio (refer to Fig. 1A, for details). These genes have been ordered into a regulatory hierarchy, whereby each gene negatively regulates the activity of genes positioned immediately downstream in the pathway. Genes at the beginning of the pathway control both sex determination and dosage compensation. The pathway then bifurcates; one branch controls sexual phenotype, while the other branch controls dosage compensation. This paper focuses on the pathway controlling sexual phenotype, and more specifically on the role of the tra-2 gene.

Fig. 1.

Regulation of somatic sexual fate in the nematode C. elegans. (A) Genetic pathway of dosage compensation and somatic sex determination (from Hodgkin, 1990; Villeneuve and Meyer, 1990). The X:A ratio is the primary determinant of sex – XX animals develop as hermaphrodites and XO animals develop as males (Madl and Herman, 1979). In response to the X:A ratio, genes that regulate dosage compensation and somatic sex determination function as a series of HIGH/LOW switches. At the beginning of the pathway, xol-1 and sdc-1, sdc-2 and sdc-3 control both sex determination and dosage compensation. Details of the dosage compensation pathway are beyond the scope of this paper (readers are referred to Villeneuve and Meyer, 1990, for details). In XX hermaphrodites, high sdc levels negatively regulate her-1. In turn, tra-2 and tra-3 negatively regulate the fem genes. Low fem activity permits tra-1 to promote hermaphrodite development. In somatic tissues, tra-1 is the terminal regulator of somatic sex determination: high levels of tra-1 promote hermaphrodite development, whereas low levels result in male development. Included in this model is a proposed positive feedback loop in which tra-1 activates tra-2 and reinforces a commitment to the hermaphrodite pathway of differentiation. In XO males, xol-1 negatively regulates the sdc genes, resulting in high her-1 activity. In turn, her-1 negatively regulates tra-2 and tra-3. Consequently, the fem genes are free to negatively regulate tra-1 and male development ensues. The germline pathway of sex determination involves the same genes described above and additional genes that have germline-specific activities (see Schedl, 1991; Ellis and Kimble, 1994, for details of the germline pathway). (B) Speculative model of protein-protein interactions controlling somatic sex determination, focusing on the role of TRA-2A. Central to this model is the prediction that the 4.7 kb tra-2 RNA encodes a transmembrane protein, TRA-2A, which provides the primary feminising activity of tra-2 (Kuwabara et al., 1992). In XX animals, TRA-2A negatively regulates the activity of one or more FEM proteins, thereby allowing TRA-1, a predicted transcription factor (Zarkower and Hodgkin, 1992) to promote hermaphrodite development. In XO males, the activity of TRA-2A is predicted to be negatively regulated by binding HER-1, which functions cell non-autonomously and encodes a secreted protein (Perry et al., 1993, Hunter and Wood, 1992). TRA-2A and HER-1 are thus postulated to mediate cell-to-cell signalling (Kuwabara et al., 1992, Hunter and Wood, 1992). In turn, the FEM proteins repress TRA-1 and male development ensues. In this working model, we suggest that negative regulation of the FEM proteins by TRA-2A and TRA-1 by the FEM proteins may involve sequestration, however, other models exist. A role for TRA-3 in our model for sex determination has been omitted, because tra-3 has been postulated to be an almost dispensable positive co-factor of tra-2 (Hodgkin, 1980).

Fig. 1.

Regulation of somatic sexual fate in the nematode C. elegans. (A) Genetic pathway of dosage compensation and somatic sex determination (from Hodgkin, 1990; Villeneuve and Meyer, 1990). The X:A ratio is the primary determinant of sex – XX animals develop as hermaphrodites and XO animals develop as males (Madl and Herman, 1979). In response to the X:A ratio, genes that regulate dosage compensation and somatic sex determination function as a series of HIGH/LOW switches. At the beginning of the pathway, xol-1 and sdc-1, sdc-2 and sdc-3 control both sex determination and dosage compensation. Details of the dosage compensation pathway are beyond the scope of this paper (readers are referred to Villeneuve and Meyer, 1990, for details). In XX hermaphrodites, high sdc levels negatively regulate her-1. In turn, tra-2 and tra-3 negatively regulate the fem genes. Low fem activity permits tra-1 to promote hermaphrodite development. In somatic tissues, tra-1 is the terminal regulator of somatic sex determination: high levels of tra-1 promote hermaphrodite development, whereas low levels result in male development. Included in this model is a proposed positive feedback loop in which tra-1 activates tra-2 and reinforces a commitment to the hermaphrodite pathway of differentiation. In XO males, xol-1 negatively regulates the sdc genes, resulting in high her-1 activity. In turn, her-1 negatively regulates tra-2 and tra-3. Consequently, the fem genes are free to negatively regulate tra-1 and male development ensues. The germline pathway of sex determination involves the same genes described above and additional genes that have germline-specific activities (see Schedl, 1991; Ellis and Kimble, 1994, for details of the germline pathway). (B) Speculative model of protein-protein interactions controlling somatic sex determination, focusing on the role of TRA-2A. Central to this model is the prediction that the 4.7 kb tra-2 RNA encodes a transmembrane protein, TRA-2A, which provides the primary feminising activity of tra-2 (Kuwabara et al., 1992). In XX animals, TRA-2A negatively regulates the activity of one or more FEM proteins, thereby allowing TRA-1, a predicted transcription factor (Zarkower and Hodgkin, 1992) to promote hermaphrodite development. In XO males, the activity of TRA-2A is predicted to be negatively regulated by binding HER-1, which functions cell non-autonomously and encodes a secreted protein (Perry et al., 1993, Hunter and Wood, 1992). TRA-2A and HER-1 are thus postulated to mediate cell-to-cell signalling (Kuwabara et al., 1992, Hunter and Wood, 1992). In turn, the FEM proteins repress TRA-1 and male development ensues. In this working model, we suggest that negative regulation of the FEM proteins by TRA-2A and TRA-1 by the FEM proteins may involve sequestration, however, other models exist. A role for TRA-3 in our model for sex determination has been omitted, because tra-3 has been postulated to be an almost dispensable positive co-factor of tra-2 (Hodgkin, 1980).

The tra-2 gene promotes female development in an XX her-maphrodite (Fig. 1A) (Klass et al., 1976, Hodgkin and Brenner, 1977). In the absence of wild-type tra-2 activity, XX animals are sexually transformed from hermaphrodite to male; XO tra-2 mutants develop normally as males. tra-2 activity is required throughout XX larval development (Klass et al., 1976). In somatic tissues, tra-2 directs XX hermaphrodite development, which is the same as female development. However, the situation is more complex in the germ line, because the feminising activity of tra-2 must be negatively regulated to allow the onset of XX hermaphrodite spermatogenesis (Doniach, 1986, Schedl and Kimble, 1988). Dominant mutations of tra-2, which escape this germline negative regulation, transform the XX germline from hermaphrodite to female. Therefore, tra-2 can promote female development in both somatic and germline tissues, and must be regulated to achieve hermaphrodite germline development.

A molecular analysis of the tra-2 locus reveals a complex developmental pattern of transcripts (Okkema and Kimble, 1991). A 4.7 kb tra-2 mRNA is detected in both XO and XX animals throughout development; however, it is 15-fold more abundant in XX than in XO animals (Okkema and Kimble, 1991). This sex-specific difference in the amount of 4.7 kb tra-2 mRNA is dependent on the activity of the downstream tra-1 gene, and provides evidence for a positive feedback loop in the pathway (Fig. 1A) (Okkema and Kimble, 1991). Two other tra-2 transcripts have also been detected. One is a 1.8 kb tra-2 mRNA found in XX animals, which appears to be germlinespecific in L4 and adult hermaphrodites and is also present in early embryos. The other is a 1.9 kb tra-2 mRNA found in adult XO males and during larval stages of XX hermaphrodite development when the germline is undergoing spermatogenesis. The structures and possible functions of the 1.8 kb and 1.9 kb tra-2 mRNAs will be discussed elsewhere (P. Kuwabara, P. Okkema, and J. Kimble, in preparation). We have proposed that the 4.7 kb tra-2 mRNA provides the primary feminising component of the tra-2 locus, based on the developmental pattern of tra-2 transcripts and on the analysis of tra-2(lf) mutations (Kuwabara et al., 1992). The 4.7 kb tra-2 mRNA encodes a predicted membrane protein called TRA-2A (Kuwabara et al., 1992).

A molecular model for the control of somatic sexual phenotype has been proposed (Fig. 1B; for review, see Kuwabara and Kimble, 1992). The main features of this model are based on the deduced amino acid sequences of cloned genes (tra-2, Kuwabara et al., 1992; her-1, Perry et al., 1993; fem-1, Spence et al., 1990; fem-3, Ahringer et al., 1992; tra-1, Zarkower and Hodgkin, 1992) and on evidence that specification of sexual fate depends on cell-to-cell communication (Villeneuve and Meyer, 1990; Schedin et al., 1991; Hunter and Wood, 1992). In XX animals, the predicted membrane protein TRA-2A is proposed to promote female development constitutively, because HER-1 is absent (refer to Fig. 1A,B) (Kuwabara et al., 1992; Trent et al., 1991). An intracellular carboxyterminal region of TRA-2A is postulated to bind and to suppress one or more of the FEM proteins, perhaps through sequestration (Fig. 1B) (Kuwabara et al., 1992). As a consequence, TRA-1, a zinc finger protein and putative transcriptional regulator (Zarkower and Hodgkin, 1992), is free to direct hermaphrodite development.

In XO males, genetic arguments predict that tra-2 is negatively regulated by her-1 (Fig. 1A) (Hodgkin, 1980). The her-1 gene functions cell non-autonomously to promote XO male development (Hunter and Wood, 1992). In addition, the HER-1 protein appears to be secreted (Perry et al., 1993). Therefore, HER-1 is an excellent candidate for an antagonist that directly binds to and negatively regulates TRA-2A. This interaction ensures that all cells in a region follow one of the two sexual fates. Inactivation of TRA-2A removes the inhibition of the FEM proteins, and hence allows the FEM proteins to negatively regulate TRA-1 and to promote male development (Fig. 1B).

This paper tests three hypotheses on which the model in Fig. 1B is based. First, we have proposed that the 4.7 kb tra-2 mRNA provides the primary feminising component of the tra-2 locus (Kuwabara et al., 1992). We demonstrate here that TRA-2A is both necessary and sufficient to provide tra-2 somatic feminising activity. Transgenic expression of a full-length cDNA corresponding to the 4.7 kb tra-2 transcript feminises the somatic tissues of XX animals that lack wild-type tra-2 activity. Second, we postulate that an increased level of TRA-2A activity is sufficient to transform XO animals into hermaphrodites. We show that expression of TRA-2A driven from a strong promoter can fully transform XO animals from male to hermaphrodite. This suggests that the relative amounts of HER-1 and TRA-2A are crucial for sex determination. Third, we test the hypothesis that an intracellular domain of TRA-2A might, by itself, have feminising activity. We find that an intracellular domain of TRA-2A can indeed partially feminise XX tra-2 and XO tra-2(+) males. Therefore, the TRA-2A intracellular domain appears to be an essential part for regulating the FEM proteins.

Nematode culture, strains and general handling methods

General methods for genetic manipulation, culturing and microscopy of nematodes have been described (Sulston and Hodgkin, 1988). Standard nomenclature is used in this paper (Horvitz et al., 1979). The suffix gf designates gain-of-function and unless otherwise stated it is implicit that all other alleles are loss-of-function (lf). Extrachromosomal arrays were transferred to different genetic backgrounds using standard genetic techniques. For information regarding C. elegans genes and alleles refer to (Hodgkin et al., 1988). A brief description of genes and alleles used in this paper follows.

tra-2(e1095) II: a putative null allele of the tra-2 gene (Hodgkin and Brenner, 1977), encodes a TRA-2A protein with a nonsense mutation at amino acid 1290 (Kuwabara et al., 1992). This nonsense change is also present in all other TRA-2 proteins (P. Kuwabara, P. Okkema, and J. Kimble, in preparation).

unc-4(e120) II: a closely linked gene to tra-2. Most tra-2(lf) strains were maintained as tra-2unc-4/mnC1. Hence, tra-2unc-4 homozygotes have an uncoordinated (Unc) phenotype, which permits simple recognition of tra-2 homozygotes regardless of the sexual phenotype. mnC1: a rearrangement of chromosome II, which suppresses recombination. All strains heterozygous for tra-2(lf) were maintained as tra-2/mnC1.

him-8(e1489) IV: increases the frequency of XO male progeny from 0.2 to 37%.

dpy-21(e428) V: distinguishes XX from XO animals. XX are dumpy (Dpy), while XO have a normal wild-type length.

Construction of expression plasmids

Methods used for manipulating nucleic acids are described by Sambrook et al. (1989). To construct pJK349, a fragment containing a full-length wild-type 4.7 kb tra-2 cDNA with 5 bp of 5′ UTR and the entire 3′ untranslated region (3′ UTR) (Kuwabara et al., 1992), flanked by XbaI and SmaI restriction sites, was ligated into the NheI-EcoRV site of pPD49.83 (D. Dixon, S. White-Harrison, and A. Fire, unpublished data). Translation is predicted to begin at the TRA-2A initiation codon (Kuwabara et al., 1992). A similar tra-2 expression clone was also constructed in the heat shock vector, pPD49.79 (D. Dixon, S. White-Harrison, and A. Fire, unpublished data). However, this clone and pJK349 function similarly in transgenic animals, hence only results with pJK349 are described. pPK81 is a derivative of pJK349 that replaces the wild-type tra-2 3′ UTR with the tra-2(e2020gf) 3′ UTR (Goodwin et al., 1993). pPK83 is a derivative of pPK81 that replaces the SphI-XmaI fragment of the heat shock promoter with a 2.9 kb SphI-XmaI her-1 P2 promoter fragment from pWLG1(Perry et al., 1993). The SphI-XmaI fragment provides her-1 P2 promoter activity, although it removes ∼560 bp from the 5′ end of the originally defined her-1 P2 promoter (Perry et al., 1993). pPK58 was constructed by ligating a KpnI-SmaI restriction fragment, containing bases 2787-4464 of the 4.7 kb tra-2 cDNA coding sequence (Kuwabara et al., 1992), into the KpnI-EcoRV site of pPD49.83. Expression of pPK58 is predicted to produce a 387 amino acid protein named TRA-2B, which starts at M1089, based on the numbering of amino acids in TRA-2A (Kuwabara et al., 1992). All plasmids contain both a tra-2 and unc-54 3′ UTR and an artificial intron between the promoter and coding region (D. Dixon, S. White-Harrison, and A. Fire, unpublished data).

Generation of transgenic nematodes

Standard methods were used to generate transgenic worms (Fire, 1986; Mello et al., 1991). All DNA injection solutions contained 100 μg/ml pRF4 and 10-20 μg/ml of appropriate expression plasmid. pRF4 carries a dominant marker, rol-6(su1006), which is used to identify transgenic animals, because they have a Rol phenotype (Mello et al., 1991). Only transgenic animals that heritably transmit extrachromosomal arrays to F2 progeny were maintained for further analysis. Single-worm PCR, using a unique plasmid primer and tra-2 primer, was used to verify that transgenic animals carried not only pRF4, but also the expression plasmid (Williams et al., 1992). The composition of extrachromosomal arrays is described below (see Fig. 2 and above for description of plasmids).

Fig. 2.

Transgenes carrying tra-2 cDNA sequences. Above, Kyte-Doolittle hydropathy plot (Kyte and Doolittle, 1982) indicates that TRA-2A is likely to be a transmembrane protein, because it contains a potential secretion signal sequence and transmembrane helices (Kuwabara et al., 1992). Below, tra-2 transgenes (for details on construction, see Materials and Methods). The HS-TRA-2A transgene expresses a full-length TRA-2A protein driven by the C. elegans heat shock promoter hsp16 (Stringham et al., 1992; D. Dixon, S. White-Harrison, and A. Fire, unpublished data). The HS-TRA-2A (3′ UTR gf) transgene is identical to HS-TRA-2A, except that it carries the tra-2(e2020gf) 3′ UTR (Goodwin et al., 1993), in place of a wild-type tra-2 3′ UTR. P2-TRA-2A expresses a full-length TRA-2A protein driven by the C. elegans her-1 P2 promoter, which is XO-specific (Perry et al., 1993). The HS-TRA-2B transgene expresses the carboxy-terminal 387 amino acids of TRA-2A driven from the C. elegans heat shock promoter hsp-16. Hydropathy analysis indicates that HS-TRA-2B is likely to be cytoplasmic, because it lacks any hydrophobic domains. This construct has been named HS-TRA-2B, because it is predicted to have the same sequence as TRA-2B, the product of the 1.8 kb tra-2 mRNA (Kuwabara, Okkema, and Kimble, unpublished data). The extrachromosomal array that was used to study the expression of a specific transgene is listed in parentheses. All constructs carry both a tra-2 3′ UTR and unc-54 3′ UTR, and an artificial intron placed between the promoter and tra-2 coding sequence. P, promoter with arrow indicating direction of transcription. Stippled boxes, TRA-2A coding sequence: unshaded boxes, untranslated regions; hatched box, unc-54 3′ UTR.

Fig. 2.

Transgenes carrying tra-2 cDNA sequences. Above, Kyte-Doolittle hydropathy plot (Kyte and Doolittle, 1982) indicates that TRA-2A is likely to be a transmembrane protein, because it contains a potential secretion signal sequence and transmembrane helices (Kuwabara et al., 1992). Below, tra-2 transgenes (for details on construction, see Materials and Methods). The HS-TRA-2A transgene expresses a full-length TRA-2A protein driven by the C. elegans heat shock promoter hsp16 (Stringham et al., 1992; D. Dixon, S. White-Harrison, and A. Fire, unpublished data). The HS-TRA-2A (3′ UTR gf) transgene is identical to HS-TRA-2A, except that it carries the tra-2(e2020gf) 3′ UTR (Goodwin et al., 1993), in place of a wild-type tra-2 3′ UTR. P2-TRA-2A expresses a full-length TRA-2A protein driven by the C. elegans her-1 P2 promoter, which is XO-specific (Perry et al., 1993). The HS-TRA-2B transgene expresses the carboxy-terminal 387 amino acids of TRA-2A driven from the C. elegans heat shock promoter hsp-16. Hydropathy analysis indicates that HS-TRA-2B is likely to be cytoplasmic, because it lacks any hydrophobic domains. This construct has been named HS-TRA-2B, because it is predicted to have the same sequence as TRA-2B, the product of the 1.8 kb tra-2 mRNA (Kuwabara, Okkema, and Kimble, unpublished data). The extrachromosomal array that was used to study the expression of a specific transgene is listed in parentheses. All constructs carry both a tra-2 3′ UTR and unc-54 3′ UTR, and an artificial intron placed between the promoter and tra-2 coding sequence. P, promoter with arrow indicating direction of transcription. Stippled boxes, TRA-2A coding sequence: unshaded boxes, untranslated regions; hatched box, unc-54 3′ UTR.

Heat shock and phenotypic analysis of transgenic nematodes

Unless otherwise stated, the progeny of gravid Rol hermaphrodites, carrying heat shock driven transgenes, were subjected to a total of three heat shocks beginning at the late embryo/early L1 stage of development. Each heat shock consisted of a 2 hour incubation at 33 °C, followed by a recovery period of 24 hours at 23 °C. The effects of single heat shocks were observed by subjecting a mixed-stage population to a single 2 hour incubation at 33 °C and examining animals of the appropriate genotype when they reached the adult stage. The sexual phenotypes of the somatic gonad, tail, hypodermis, intestine, and germ line were scored in adult transgenic animals using Nomarski DIC optics (400× or 630×). Somatic tissues were considered feminised if they displayed one or more of the following hermaphroditic or intersexual characteristics.

somatic gonad: complete or partial bi-lobed arms.

tail: hermaphrodite spike; truncated or missing fan, rays, or spicules.

hypodermis: complete or partial vulval induction. intestine: yolk protein accumulation in the pseudocoelom. germ line: oocytes.

For some experiments, adult male animals were heat shocked once at 33 °C for 2 hours and scored for germline and intestinal phenotypes 24 hours after heat shock.

Controls: feminisation by tra-2 transgenes is heat shock dependent (n>>100), except pPK83, which is driven by the her-1 P2 promoter. Heat shock does not feminise either XX tra-2 or tra-2(+) XO transgenic animals that carry a heat shock driven lacZ transgene, which has no tra-2 activity (kindly provided by A. Fire) (n>>100).

SDS polyacrylamide gel electrophoresis

A 7% polyacrylamide gel with a 4.75% stacking gel was prepared as described (Sambrook et al., 1989). Samples were prepared by washing hand-picked worms in M9 salts three times before resuspending each worm pellet in 2× SDS gel sample buffer. Prior to loading, samples were heated to 95 °C for 10 minutes. Gels were stained after electrophoresis with Coomassie Blue.

HS-TRA-2A feminises the soma of XX tra-2 mutants

To test whether TRA-2A promotes hermaphrodite development when introduced as a transgene, we examined the effect of HS-TRA-2A (Fig. 2) on the sexual phenotype among the self-progeny of XX tra-2unc-4/ + +;qEx32 hermaphrodites. Without heat shock, transgenic roller animals are either non-Unc hermaphrodites of genotype tra-2unc-4/++;qEx32 or ++/++;qEx32 or Unc pseudomales of genotype tra-2unc-4;qEx32 (n>>100). However, after a series of heat shocks, many of the XX tra-2unc-4;qEx32 homozygotes (identified by their uncoordinated phenotype) are clearly feminised (see Materials and Methods for heat shock regime and scoring criteria for sexual phenotypes). HS-TRA-2A extensively feminises the gonad, tail, hypodermis, and intestine of XX tra-2unc-4;qEx32 animals, but does not feminise the germline of XX tra-2unc-4;qEx32 animals: the somatic gonad contains sperm but not oocytes (Table 1, line 1). An example of an XX tra-2unc-4;qEx32 transformant is shown in Fig. 3A. This transformant has a virtually wild-type hermaphrodite soma, yet its germ line produces only sperm. XX tra-2unc-4;qEx32 mutants that receive only a single heat shock (see Materials and Methods) also show somatic feminisation. However, these animals are not as extensively feminised as animals that receive multiple heat shocks (data not shown).

Table 1.

HS-TRA-2A feminises the soma of XX and XO tra-2 mutants and sexually transforms XO tra-2(+) nematodes from male to hermaphrodite

HS-TRA-2A feminises the soma of XX and XO tra-2 mutants and sexually transforms XO tra-2(+) nematodes from male to hermaphrodite
HS-TRA-2A feminises the soma of XX and XO tra-2 mutants and sexually transforms XO tra-2(+) nematodes from male to hermaphrodite
Fig. 3.

Transgenic TRA-2A promotes feminisation of XX tra-2 and XO tra-2(+) animals. Top, Nomarski DIC photomicrograph (630×) of adult transgenic animal, lateral view. Bottom, schematic representation of photomicrograph, except panel A. (A) Adult XX tra-2unc-4;qEx32 mutant transformed from pseudomale to hermaphrodite by HS-TRA-2A after a series of heat shocks. Feminised somatic tissues include: bi-lobed somatic gonad, vulva, spiked tail, and intestine (yolk). Sperm, but not oocytes are present in each lobe of the somatic gonad. Middle, schematic representation of photomicrograph. Bottom, enlargement of photomicrograph, which corresponds to boxed region in middle panel, details the presence of sperm, but not oocytes in the germ line. (B) Adult XO tra-2(+);dpy-21;him-8;qEx32 animal transformed by HS-TRA-2A from male to fertile hermaphrodite after a series of heat shocks. Feminised somatic tissues include: bi-lobed somatic gonad, vulva, spiked tail, and intestine (yolk). The snub tail is incompletely feminised. Sperm and oocytes are present in each lobe of the somatic gonad. (C) Adult XO tra-2(+);him-8;crEx1 intersexual animal. The somatic gonad, vulva, tail, and intestine (yolk) are partially feminised by P2-HER-1, however, masculinised tail structures such as a ray and truncated fan are also visible. Sperm, but not oocytes are visible. Scale bar, 10 μm.

Fig. 3.

Transgenic TRA-2A promotes feminisation of XX tra-2 and XO tra-2(+) animals. Top, Nomarski DIC photomicrograph (630×) of adult transgenic animal, lateral view. Bottom, schematic representation of photomicrograph, except panel A. (A) Adult XX tra-2unc-4;qEx32 mutant transformed from pseudomale to hermaphrodite by HS-TRA-2A after a series of heat shocks. Feminised somatic tissues include: bi-lobed somatic gonad, vulva, spiked tail, and intestine (yolk). Sperm, but not oocytes are present in each lobe of the somatic gonad. Middle, schematic representation of photomicrograph. Bottom, enlargement of photomicrograph, which corresponds to boxed region in middle panel, details the presence of sperm, but not oocytes in the germ line. (B) Adult XO tra-2(+);dpy-21;him-8;qEx32 animal transformed by HS-TRA-2A from male to fertile hermaphrodite after a series of heat shocks. Feminised somatic tissues include: bi-lobed somatic gonad, vulva, spiked tail, and intestine (yolk). The snub tail is incompletely feminised. Sperm and oocytes are present in each lobe of the somatic gonad. (C) Adult XO tra-2(+);him-8;crEx1 intersexual animal. The somatic gonad, vulva, tail, and intestine (yolk) are partially feminised by P2-HER-1, however, masculinised tail structures such as a ray and truncated fan are also visible. Sperm, but not oocytes are visible. Scale bar, 10 μm.

Two additional approaches were taken in an attempt to see an effect of HS-TRA-2A on the XX germ line. First, we searched among XX tra-2unc-4/++;qEx32 and +/+;qEx32 animals after heat shock for the presence of phenotypic females. We reasoned that the combination of endogenous tra-2(+) activity plus transgenic HS-TRA-2A might produce a dominant gain-of-function phenotype similar to that observed in tra-2(gf) mutants (Doniach, 1986; Schedl and Kimble, 1988). However, no females were detected (n>>100). Second, we generated the transgene HS-TRA-2A(3′ UTRgf) by methods similar to those used to generate HS-TRA-2A (see Materials and Methods). HS-TRA-2A(3′ UTRgf) carries a deletion within the tra-2 3′ UTR, which permits tra-2 to escape negative translational control (Goodwin et al., 1993). We found that the soma of XX tra-2unc-4;crEx2 homozygous animals expressing HS-TRA-2A(3′ UTRgf) is feminised to the same extent as XX tra-2unc-4;qEx32 transgenic animals (data not shown), and the germ line is not feminised. HS-TRA-2A(3′ UTRgf) also fails to feminise the germ line of wild-type animals (eg. ++/++; crEx2).

We conclude that HS-TRA2A can provide the major somatic feminising activity of tra-2. The failure to detect oocytes in XX tra-2unc-4;qEx32 animals suggests that either the heat shock promoter does not function in the germ line (Stringham et al., 1992) or that an additional tra-2 gene product is required for oogenesis.

Transgenic TRA-2A transforms XO males into hermaphrodites

In XO males, secreted HER-1 is postulated to bind and to inactivate TRA-2A (Fig. 1B) (Kuwabara et al., 1992, Hunter and Wood, 1992). From this model, we predict that an elevated level of TRA-2A might escape HER-1 regulation and hence feminise XO animals (Kuwabara et al., 1992). To test this hypothesis, we asked if HS-TRA-2A could feminise XO animals. The strain constructed for this experiment was dpy-21;him-8;qEx32. This strain carries dpy-21 to permit us to distinguish XX (Dpy) from XO (non-Dpy) animals (Hodgkin, 1983), him-8, which generates 37% XO animals, and qEx32, the extrachromosomal array bearing HS-TRA-2A. After a series of heat shocks, we examined non-Dpy adult XO animals by Nomarski DIC optics. We found that HS-TRA-2A feminised not only somatic tissues, but also the germ line of XO animals (Table 1, line 2). All XO animals with feminised germ lines produced sperm first, then oocytes-indicating that HS-TRA-2A expression results in hermaphrodite rather than female germline development. These animals are often selffertile, albeit with low brood sizes <10. Many of the brood die as embryos, but occasional animals develop into adult males (data not shown). An example of an XO tra-2(+);qEx32 transgenic animal is shown in Fig. 3B. This XO animal is fully transformed from a male to a self-fertile hermaphrodite by HS-TRA-2A, although it has a slightly snubbed tail. We conclude that HS-TRA-2A is capable of feminising all XO tissues, including the germ line.

In a separate set of experiments, the 4.7 kb tra-2 cDNA was expressed in XO animals under control of the her-1 P2 promoter (P2-TRA-2A) (Fig. 2). The purpose of this experiment was to verify that expression of P2-TRA-2A from the extrachromosomal array, crEx1, could feminise XO males under non-heat shock conditions. P2-TRA-2A was not expected to transform XO males completely into hermaphrodites. tra-2 activity is required throughout larval development (Klass et al., 1976), whereas the her-1 P2 promoter is active primarily during early stages of XO, but not XX development (Perry et al., 1993). An example of an XO tra-2(+);crEx1 male with intersexual somatic development is provided in Fig. 3. Thus, we have shown in two independent experiments that transgenic TRA-2A feminises somatic tissues of XO animals, presumably by overcoming negative regulation by HER-1.

Feminising activity associated with a putative intracellular carboxy-terminal domain of TRA-2A

We have hypothesised that a carboxy-terminal domain of TRA-2A negatively regulates one or more of the FEM proteins through a protein-protein interaction in the cytoplasm (Fig. 1B) (Kuwabara et al., 1992). If true, then expression of the carboxy-terminal region of TRA-2A, by itself, might have feminising activity. To test this possibility, we asked if HS-TRA-2B could feminise XX tra-2unc-4;qEx35 mutants. HS-TRA-2B consists of the carboxy-terminal 387 amino acids of TRA-2A and is predicted to be cytoplasmic, because it lacks any hydrophobic region that might function as a signal sequence or as a membrane spanning domain (Fig. 2). HS-TRA-2B is also predicted to encode the same TRA-2B protein as the 1.8 kb tra-2 mRNA (P. Kuwabara, P. Okkema, and J. Kimble, in preparation). We found that a number of XX tra-2unc-4;qEx35 Unc homozygotes had partially feminised tails after a series of heat shocks. However, these feminised animals did not also display the Rol phenotype, which is diagnostic of animals carrying an extrachromosomal array. Nomarski DIC optics revealed that these feminised animals probably failed to roll because they suffered from severe defecation defects, the result of intersexual tail development. Therefore, we initially examined all Unc animals (tra-2unc-4;qEx35 and tra-2unc-4) to determine the range of XX tra-2unc-4;qEx35 phenotypes, although not all Unc animals are expected to express the transgene. We found that 9/41 animals had partially feminised tails or defecation defects arising from intersexual tail development, 14/41 animals accumulated yolk in the pseudocoelom, which is indicative of intestinal feminisation, and 2/41 had partially feminised tails and accumulated yolk. In addition, 16/41 Unc animals were not feminised, as might be expected if they did not carry or express the transgene. None of the animals examined showed feminisation of the germ line.

The number of XX tra-2unc-4;qEx35 animals accumulating yolk in response to heat shock was probably underestimated, because it is difficult to score visually for yolk in nematodes with severe defecation defects. Therefore, in a second experiment, we used Nomarski DIC optics and SDS polyacrylamide gel electrophoresis to examine transgenic animals that were clearly XX tra-2;qEx35 Unc rollers and not defecation defective. We found that 26/30 XX tra-2unc-4;qEx35 Unc rollers accumulated yolk after heat shock (Table 2, line 1). Analysis by SDS polyacrylamide electrophoresis verified that XX tra-2unc-4;qEx35 mutants produce yolk only in response to heat shock (Fig. 4, compare lanes 3,4). We also found that HS-TRA-2B induced yolk accumulation in 35/35 XO tra-2(+);him-8;qEx35 males (Table 2, line 2). Again, yolk accumulation in these animals is heat shock dependent (Fig. 4, compare lanes 1, 2). In addition, none of the XO tra-2(+);him-8;qEx35 males showed feminisation of the germ line. We conclude that ectopic expression of the carboxy terminus of TRA-2A has feminising activity in both XX tra-2unc-4;qEx35 and XO tra-2(+);him-8;qEx35 males. Therefore, these results support the hypothesis that the carboxy-terminal region of TRA-2A contains a domain involved in negatively regulating the FEM proteins.

Table 2.

HS-TRA-2B feminises the intestine of XX tra-2 and XO tra-2(+) males

HS-TRA-2B feminises the intestine of XX tra-2 and XO tra-2(+) males
HS-TRA-2B feminises the intestine of XX tra-2 and XO tra-2(+) males
Fig. 4.

HS-TRA-2B promotes partial feminisation of XO tra-2(+);qEx35 and XX tra-2unc-4;qEx35 transgenic nematodes. SDS polyacrylamide gel showing the accumulation of four yolk proteins: yp170A, yp170B, yp115, and yp88 (Sharrock, 1983) in response to heat shock driven expression of HS-TRA-2B. Each lane contains 30 animals of the specified genotype. The position and molecular mass (×10−3) of each yolk protein is marked by an arrow. (+/−) indicates with or without heat shock.

Fig. 4.

HS-TRA-2B promotes partial feminisation of XO tra-2(+);qEx35 and XX tra-2unc-4;qEx35 transgenic nematodes. SDS polyacrylamide gel showing the accumulation of four yolk proteins: yp170A, yp170B, yp115, and yp88 (Sharrock, 1983) in response to heat shock driven expression of HS-TRA-2B. Each lane contains 30 animals of the specified genotype. The position and molecular mass (×10−3) of each yolk protein is marked by an arrow. (+/−) indicates with or without heat shock.

HS-TRA-2A requires an endogenous wild-type tra-2 gene to promote germline feminisation

It might be predicted that HS-TRA2A should feminise XX tra-2 animals more efficiently than XO tra-2(+) animals, because TRA-2A is not inactivated by HER-1 in XX animals. Therefore, it was a surprise to find that HS-TRA-2A feminised both the soma and germ line of XO tra-2(+);qEx32 transgenic animals, yet failed to feminise the germ line of XX tra-2;qEx32 mutants. These results suggested that the feminising effects of HS-TRA-2A were more extensive in XO tra-2(+);qEx32 animals than in XX tra-2; qEx32 mutants, because the former carried a wild-type tra-2 gene. To test this idea, we asked if HS-TRA-2A could feminise the germline of XO mutants that lack a wild-type tra-2 gene. For this study, we constructed the strain tra-2;tra-1(e1575gf)/+;qEx32, using methods similar to those described by Hodgkin (1980). This strain produces two kinds of males: XO tra-2;qEx32 and XX tra-2;qEx32. An XO tra-2;qEx32 male can be distinguished from an XX tra-2;qEx32 male by adult tail morphology and mating behaviour. Therefore, for this experiment we selected adult males of the appropriate genotype and subjected them to a single heat shock. First, we established that applying a single heat shock to an adult XO tra-2(+);qEx32 or XX tra-2;qEx32 male has the same effect on germline and intestinal phenotype as applying a series of heat shocks throughout development. We found that after a single heat shock, both the intestine and germ line of adult XO tra-2(+);him-8; qEx32 males were feminised by HS-TRA-2A (Table 1, line 3). An example of an adult XO tra-2(+);him-8;qEx32 male that produced both yolk and oocytes in response to HS-TRA-2A is shown in Fig. 5. Under the same conditions, HS-TRA-2A feminised the intestine (yolk) of adult XX tra-2;qEx32 males, but again failed to feminise the germ line (Table 1, line 4). Thus, applying a single heat shock to an adult, which carries the qEx32 transgene, appears to have the same effect on the germline and intestinal phenotype as applying a series of developmental heat shocks. Next, we examined the effect of HS-TRA-2A on the phenotype of adult XO tra-2;qEx32 males, which lack a wild-type tra-2 gene. We found that HS-TRA-2A feminised the intestine (yolk) of XO tra-2;qEx32 males, but failed to feminise the germ line (Table 1, line 5). Therefore, we conclude that HS-TRA-2A cannot feminise the germ line if an endogenous wild-type tra-2 gene is absent. This suggests that additional wild-type tra-2 products may be required to elicit germ line feminisation. These results also indicate that HS-TRA-2A can reverse sexual cell fate decisions in animals that are already committed to following the male fate. A similar plasticity in sexual cell fate maintenance was also noted by Schedin et al. (1994). They found that the intestine and germline of adult XO her-1(ts) males could be feminised by shifting animals from permissive to restrictive temperature.

Fig. 5.

HS-TRA-2A feminises the intestine and germ line of adult XO males. Top, Nomarski DIC photomicrograph (630×) of adult XO him-8;qEx32 male, oblique view. Bottom, schematic representation of photomicrograph. Feminised intestine (yolk) and germline (oocytes) are indicated. No other changes in somatic structures are observed. Scale bar, 10 μm.

Fig. 5.

HS-TRA-2A feminises the intestine and germ line of adult XO males. Top, Nomarski DIC photomicrograph (630×) of adult XO him-8;qEx32 male, oblique view. Bottom, schematic representation of photomicrograph. Feminised intestine (yolk) and germline (oocytes) are indicated. No other changes in somatic structures are observed. Scale bar, 10 μm.

TRA-2A is necessary and sufficient to promote feminisation of XX somatic tissues

The tra-2 locus expresses multiple transcripts (Okkema and Kimble, 1991). Our model for sex determination (Fig. 1B) proposes that the predicted membrane protein, TRA-2A, encoded by the 4.7 kb tra-2 mRNA, provides the primary feminising activity of the tra-2 locus (Kuwabara et al., 1992). Here, we show that transgenic TRA-2A does provide a tra-2(+) activity that directs somatic cells of XX tra-2 mutants to follow the hermaphrodite fate. We argue that TRA-2A is both a necessary and sufficient component of somatic tra-2 feminising activity for three reasons. First, mutations that disrupt only the TRA-2A coding sequence and not other predicted TRA-2 proteins abolish tra-2 activity (Okkema and Kimble, 1991; Kuwabara et al., 1992). Second, HS-TRA-2A alone is sufficient to feminise the soma of XX tra-2 null mutants (this study). Finally, like tra-2(+), which is required throughout hermaphrodite larval development (Klass et al., 1976), ectopic TRA-2A can affect sexual cell fates at multiple points during development.

Cell-to-cell signalling mediated by HER-1 and TRA-2A

Cell-to-cell signalling is an important mechanism for regulating cell fates during the development of many organisms. In C. elegans, we have proposed that TRA-2A and HER-1 mediate cell-to-cell communication to regulate sexual cell fate decisions and to ensure that all cells follow the same sexual fate (Kuwabara et al., 1992). tra-2 mRNAs are found in both XX hermaphrodites and in adult XO males, however, tra-2 mRNA levels are 15-fold lower in XO males than in XX hermaphrodites (Okkema and Kimble, 1991). We have suggested that HER-1 functions as a TRA-2A antagonist to ensure that even low levels of TRA-2A remain inactive in XO males (Kuwabara et al., 1992). Otherwise, inappropriate TRA-2A activity in XO animals might activate a positive feed-back loop that leads to increased tra-2 mRNA steady-state levels and probably TRA-2A protein (Fig. 1A) (Okkema and Kimble, 1991). As a result, an XO cell might be driven to follow the hermaphrodite fate, if insufficient HER-1 is present to negatively regulate TRA-2A (Kuwabara et al., 1992). We have shown that HS-TRA-2A driven from a strong promoter does indeed transform XO animals into fertile hermaphrodites, although HER-1 is presumably present in these animals. We suggest that the level of HS-TRA-2A is sufficiently elevated to titrate HER-1 and to allow some HS-TRA-2A activity to escape negative regulation, because the transformation of XO males into hermaphrodites mimics the XO her-1 loss-of-function phenotype. Therefore, the relative ratio of HER-1 to TRA-2A may be crucial in determining sexual cell fate. It might also be predicted that mutant TRA-2A proteins, which are essentially wild-type in activity except that they are insensitive to negative regulation by HER-1, would also transform XO animals to the hermaphrodite fate. tra-2 alleles with such properties have been identified and their characterisation will be reported elsewhere (J. Hodgkin, submitted; P. Kuwabara, submitted).

The carboxy-terminal domain of TRA-2A contains feminising activity that may mediate signal transduction

It has been hypothesised that TRA-2A promotes XX hermaphrodite development by negatively regulating one or more of the predicted cytoplasmic FEM proteins (Spence et al., 1990; Ahringer et al., 1992). We have proposed that an intra-cellular carboxy-terminal region of TRA-2A plays a crucial role in this regulation (Kuwabara et al., 1992). In this study, we have demonstrated that HS-TRA-2B, which contains only a carboxy-terminal region of TRA-2A, has feminising activity on its own. HS-TRA-2B is so named because it is identical in sequence to TRA-2B, the predicted protein encoded by the 1.8 kb tra-2 mRNA. The normal role of the 1.8 kb tra-2 mRNA in C. elegans sex determination will be discussed elsewhere (P. Kuwabara, P. Okkema, and J. Kimble, in preparation). HS-TRA-2B is likely to be cytoplasmic because it lacks any hydrophobic domains or other sub-cellular localisation signals (Fig. 2). We found that HS-TRA-2B expression in XX tra-2;qEx35 mutants led to intersexual tail development and yolk protein accumulation. In addition, HS-TRA-2B induced yolk accumulation in XO tra-2(+);qEx35 males. These results indicate that the TRA-2A carboxy terminus probably contains a regulatory domain that represses the activity of one or more of the cytoplasmic FEM proteins. This interaction is proposed to occur when TRA-2A is not repressed by HER-1 and implies that TRA-2A is constitutively active in a signal transduction process during XX hermaphrodite somatic development.

HS-TRA-2B does not feminise the soma of animals to the same extent as HS-TRA-2A. This difference can be attributed to a number of factors such as protein topology, intracellular localisation, or protein stability. For example, TRA-2A might be better at sequestering the FEM proteins, because its carboxy terminus is anchored to the membrane; in contrast, HS-TRA-2B is likely to be freely cytoplasmic. In addition, HS-TRA-2B does not feminise the germline, possibly because the heat shock promoter fails to function in the germ line (Stringham et al., 1992; see below).

HS-TRA-2A may reinforce a commitment to the hermaphrodite fate

We have shown that HS-TRA-2A does not feminise the germ line of XX tra-2 mutants. This could be because the heat shock promoter does not function in the germ line (Stringham et al. 1992) or because TRA-2A is not the only tra-2 gene product needed to support hermaphrodite germline development. However, if the heat shock promoter does not function in the germline, it becomes necessary to explain how the germline of XO tra-2(+) animals can be feminised by HS-TRA-2A. One possibility is that somatic HS-TRA-2A titrates HER-1 protein, which would otherwise bind to and repress endogenous germline TRA-2A. As a consequence, the endogenous germline TRA-2 proteins are freed from repression and can promote hermaphrodite germline development. This model would be consistent with the finding that in XO animals mosaic for her-1, certain her-1(+) cells can be induced to follow the female fate (Hunter and Wood, 1992), presumably because of influences exerted by neighbouring cells (Kuwabara and Kimble, 1992).

Alternatively, it remains possible that HS-TRA-2A is expressed in the germline, but that in addition, the endogenous tra-2 gene products are required to promote hermaphrodite germline development. Evidence that the heat shock promoter does not function in the germline is based primarily on the failure to observe germline lacZ reporter activity (Stringham et al., 1992); this does not rule out the possibility that the heat shock promoter may function in the germ line, but at a level lower than that found in somatic tissues. It is tempting to speculate that HS-TRA-2A may indeed be expressed in the germ line, albeit poorly, and that HS-TRA-2A can thereby recruit endogenous tra-2 gene products by activating the same positive feedback loop that is likely to be responsible for the sex-specific differences in tra-2 mRNA levels (Fig. 1A) (Okkema and Kimble, 1991). In this scenario, HS-TRA-2A expression is predicted to elevate the steady-state levels of both the 4.7 kb and 1.8 kb tra-2 mRNAs. Either or both of these tra-2 mRNAs might play an important role in promoting hermaphrodite germline development.

Our results indicate that TRA-2A plays a central role in regulating sexual fate decisions in both XX and XO animals. Now that all of the known major regulatory genes that control sex determination in C. elegans have been cloned, we have the tools to investigate how sexual cell fate decisions are controlled at the biochemical level. Future experiments will focus on demonstrating whether a direct binding interaction can be detected between TRA-2A and HER-1. In addition, it should be possible to determine how the intracellular carboxyterminal domain of TRA-2A interacts with one or more of the FEM proteins to mediate signal transduction.

We thank A. Fire and M. Perry for sending plasmids. We are grateful to M. Bretscher and J. Hodgkin for critical reading of the manuscript. We thank S. Ingham, A. Lenton and B. Pashley for help with illustrations. This work was supported by the Howard Hughes Medical Institute and NIH grant HD24663 to J.K. and funding from the Medical Research Council of Great Britain to P. E. K.

Ahringer
,
J.
,
Rosenquist
,
T. A.
,
Lawson
,
D. N.
and
Kimble
,
J.
(
1992
).
The Caenorhabditis elegans sex determining gene fem-3 is regulated post-transcriptionally
.
EMBO J
.
11
,
2303
2310
.
Baird
,
S. E.
,
Fitch
,
D. H. A.
and
Emmons
,
S. W.
(
1994
).
Caenorhabditis vulgaris Sp.N. (Nematoda: Rhabditidae): A necromenic associate of pill bugs and snails
.
Nematologica
40
,
1
11
.
Doniach
,
T.
(
1986
).
Activity of the sex-determining gene tra-2 is modulated to allow spermatogenesis in the C. elegans hermaphrodite
.
Genetics
114
,
53
76
.
Ellis
,
R. E.
and
Kimble
,
J.
(
1994
).
Control of germ cell differentiation in Caenorhabditis elegans
.
CIBA Foundation Symposium
182
,
179
192
.
Fire
,
A.
(
1986
).
Integrative transformation of Caenorhabditis elegans
.
EMBO J
.
5
,
2673
2680
.
Goodwin
,
E. B.
,
Okkema
,
P. G.
,
Evans
,
T. C.
and
Kimble
,
J.
(
1993
).
Translational regulation of tra-2 by its 3′ untranslated region controls sexual identity in C. elegans
.
Cell
75
,
329
339
.
Hodgkin
,
J.
(
1980
).
More sex-determination mutants of Caenorhabditis elegans
.
Genetics
96
,
649
664
.
Hodgkin
,
J.
(
1983
).
X chromosome dosage and gene expression in Caenorhabditis elegans
.
Genetics
.
96
,
649
664
.
Hodgkin
,
J.
(
1990
).
Sex determination compared in Drosophila and Caenorhabditis
.
Nature
344
,
721
8
.
Hodgkin
,
J.
and
Brenner
,
S.
(
1977
).
Mutations causing transformation of sexual phenotype in the nematode Caenorhabditis elegans
.
Genetics
86
,
275
287
.
Hodgkin
,
J.
,
Edgley
,
M.
,
Riddle
,
D.
and
Albertson
,
D. G.
(
1988
).
Genetics appendix
.
In The Nematode Caenorhabditis elegans
, (ed.
W. B.
Wood
), pp.
491
584
.
Cold Spring Harbor, NY
:
Cold Spring Harbor Laboratory Press
.
Horvitz
,
R.
,
Brenner
,
S.
,
Hodgkin
,
J.
and
Herman
,
R.
(
1979
).
A uniform genetic nomenclature for the nematode Caenorhabditis elegans
.
Mol. Gen. Genet
.
175
,
129
133
.
Hunter
,
C. P.
and
Wood
,
W. B.
(
1992
).
Evidence from mosaic analysis of the masculinizing gene her-1 for cell interactions in C. elegans sex determination
.
Nature
355
,
551
555
.
Klass
,
M.
,
Wolf
,
N.
and
Hirsh
,
D.
(
1976
).
Development of the male reproductive system and sexual transformation in the nematode Caenorhabditis elegans
.
Dev. Biol
.
69
,
329
335
.
Kuwabara
,
P. E.
and
Kimble
,
J.
(
1992
).
Molecular genetics of sex determination in C. elegans
.
Trends Genet
.
8
,
164
168
.
Kuwabara
,
P. E.
,
Okkema
,
P. G.
and
Kimble
,
J.
(
1992
).
tra-2 encodes a membrane protein and may mediate cell communication in the Caenorhabditis elegans sex determination pathway
.
Mol. Biol. Cell
3
,
461
473
.
Kyte
,
J.
and
Doolittle
,
R. F.
(
1982
).
A simple method for displaying the hydropathic character of a protein
.
J. Mol. Biol
.
157
,
133
148
.
Madl
,
J. E.
and
Herman
,
R. K.
(
1979
).
Polyploids and sex determination in Caenorhabditis elegans
.
Genetics
93
,
393
402
.
Mello
,
C. C.
,
Kramer
,
J. M.
,
Stinchcomb
,
D.
and
Ambros
,
V.
(
1991
).
Efficient gene transfer in C. elegans: extrachromosomal maintenance and integration of transforming sequences
.
EMBO J
.
10
,
3959
3970
.
Okkema
,
P. G.
and
Kimble
,
J.
(
1991
).
Molecular analysis of tra-2, a sex determining gene in C. elegans
.
EMBO J
.
10
,
171
176
.
Perry
,
M. D.
,
Li
,
W.
,
Trent
,
C.
,
Robertson
,
B.
,
Fire
,
A.
,
Hageman
,
J. M.
and
Wood
,
W. B.
(
1993
).
Molecular characterization of the her-1 gene suggests a direct role in cell signaling during Caenorhabditis elegans sex determination
.
Genes Dev
.
7
,
216
228
.
Sambrook
,
J.
Fritsch
,
E. F.
and
Maniatis
,
T.
(
1989
).
Molecular Cloning: A Laboratory Manual
.
Cold Spring Harbor, New York
:
Cold Spring Harbor Laboratory
.
Schedin
,
P.
,
Hunter
,
C. P.
and
Wood
,
W. B.
(
1991
).
Autonomy and nonautonomy of sex determination in triploid intersex mosaics of C. elegans
.
Development
112
,
833
879
.
Schedin
,
P.
,
Jonas
,
P.
and
Wood
,
W. B.
(
1994
).
Function of the her-1 gene is required for maintenance of male differentiation in adult tissues of C. elegans
.
Dev. Genet
.
15
,
231
239
.
Schedl
,
T.
(
1991
).
The role of cell-cell interactions in postembryonic developmpent of the Caenorhabditis elegans germ line
.
Curr. Opin. Genet. Dev
.
1
,
185
190
.
Schedl
,
T.
and
Kimble
,
J.
(
1988
).
fog-2, a germ-line-specific sex determination gene required for hermaphrodite spermatogenesis in Caenorhabditis elegans
.
Genetics
123
,
755
769
.
Sharrock
,
W.
(
1983
).
Yolk proteins of Caenorhabditis elegans
.
Dev. Biol
.
96
,
182
188
.
Spence
,
A. M.
,
Coulson
,
A.
and
Hodgkin
,
J.
(
1990
).
The product of fem-1, a nematode sex-determining gene, contains a motif found in cell cycle control proteins and receptors for cell-cell interactions
.
Cell
60
,
981
90
.
Stringham
,
E. G.
,
Dixon
,
D. K.
,
Jones
,
D.
and
Candido
,
E. P. M.
(
1992
).
Temporal and spatial expression patterns of the small heat shock (hsp16) genes in transgenic Caenorhabditis elegans
.
Mol. Biol. Cell
3
,
221
233
.
Sulston
,
J. E.
and
Hodgkin
,
J.
(
1988
).
Methods
.
In The Nematode Caenorhabditis elegans
(ed.
W. B.
Wood
), pp.
587
606
.
Cold Spring Harbor, NY
:
Cold Spring Harbor Laboratory Press
.
Trent
,
C.
,
Purnell
,
B.
,
Gavinski
,
S.
,
Hageman
,
J.
,
Chamblin
,
C.
and
Wood
,
W. B.
(
1991
).
Sex-specific transcriptional regulation of the C. elegans sex-determining gene her-1
.
Mech. Dev
.
34
,
43
56
.
Villeneuve
,
A. M.
and
Meyer
,
B. J.
(
1990
).
The regulatory hierarchy controlling sex determination and dosage compensation in Caenorhabditis elegans
.
Adv. Genet
.
27
,
117
88
.
Villeneuve
,
A. M.
and
Meyer
,
B. J.
(
1990
).
The role of sdc-1 in the sex determination and dosage compensation decisions in Caenorhabditis elegans
.
Genetics
124
,
91
114
.
Williams
,
B. D.
,
Schrank
,
B.
,
Huynh
,
C.
,
Shownkeen
,
R.
and
Waterston
,
R. H.
(
1992
).
A genetic mapping system in Caenorhabditis elegans based on polymorphic sequence tagged sites
.
Genetics
131
,
609
624
.
Zarkower
,
D.
and
Hodgkin
,
J.
(
1992
).
Molecular analysis of the C. elegans sex-determining gene tra-1: a gene encoding two zinc finger proteins
.
Cell
70
,
237
249
.