fog-2 and the Evolution of Self-Fertile Hermaphroditism in Caenorhabditis

doi:10.1371/journal.pbio.0030006

Journal List > PLoS Biol > v.3(1); Jan 2005

PLoS Biol. 2005 January; 3(1): e6.

Published online 2004 December 28. doi: 10.1371/journal.pbio.0030006.

PMCID: PMC539060

Copyright : © 2004 Nayak et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited

fog-2 and the Evolution of Self-Fertile Hermaphroditism in Caenorhabditis

Sudhir Nayak,¹ Johnathan Goree,¹ and Tim Schedl¹

1Department of Genetics, Washington University School of Medicine, St. Louis, Missouri, United States of America

Barbara Meyer, Academic Editor

University of California at Berkeley, United States of America

Corresponding author.

Tim Schedl: ts/at/genetics.wustl.edu

Received July 23, 2004; Accepted October 16, 2004.

See "The Evolution of Self-Fertile Hermaphroditism: The Fog Is Clearing" , e30.

This article has been cited by other articles in PMC.

Abstract

Somatic and germline sex determination pathways have diverged significantly in animals, making comparisons between taxa difficult. To overcome this difficulty, we compared the genes in the germline sex determination pathways of Caenorhabditis elegans and C. briggsae, two Caenorhabditis species with similar reproductive systems and sequenced genomes. We demonstrate that C. briggsae has orthologs of all known C. elegans sex determination genes with one exception: fog-2. Hermaphroditic nematodes are essentially females that produce sperm early in life, which they use for self fertilization. In C. elegans, this brief period of spermatogenesis requires FOG-2 and the RNA-binding protein GLD-1, which together repress translation of the tra-2 mRNA. FOG-2 is part of a large C. elegans FOG-2-related protein family defined by the presence of an F-box and Duf38/FOG-2 homogy domain. A fog-2-related gene family is also present in C. briggsae, however, the branch containing fog-2 appears to have arisen relatively recently in C. elegans, post-speciation. The C-terminus of FOG-2 is rapidly evolving, is required for GLD-1 interaction, and is likely critical for the role of FOG-2 in sex determination. In addition, C. briggsae gld-1 appears to play the opposite role in sex determination (promoting the female fate) while maintaining conserved roles in meiotic progression during oogenesis. Our data indicate that the regulation of the hermaphrodite germline sex determination pathway at the level of FOG-2/GLD-1/tra-2 mRNA is fundamentally different between C. elegans and C. briggsae, providing functional evidence in support of the independent evolution of self-fertile hermaphroditism. We speculate on the convergent evolution of hermaphroditism in Caenorhabditis based on the plasticity of the C. elegans germline sex determination cascade, in which multiple mutant paths yield self fertility.

Introduction

Sex determination is an ancient and universal feature in metazoans. In spite of this, comparison of distantly related species such as Caenorhabditis elegans and Drosophila melanogaster has revealed little about the evolution of the complex pathways that mediate the sexual fate decision in the soma and germline [1,2,3]. This is likely due to the combination of gross morphological, functional, and behavioral dissimilarity and extensive sequence divergence. Thus, if we wish to clarify the etiology of diverged sex determination pathways, an alternative approach is required.

One approach is to perform comparative analysis of sex determination genes in species separated by sufficient evolutionary time to allow for changes in pathway components yet retain comparable somatic and germline morphology and function. The clade containing C. elegans and C. briggsae represents an ideal case for this type of study, as the sex determination pathway has been well studied in C. elegans and an abundance of sequence information is available for both species [4,5].

C. elegans and C. briggsae, while sharing very similar germline and somatic morphology, are separated by approximately 100 million years and are members of a clade that employs multiple mating systems [5,6,7,8,9,10]. C. elegans and C. briggsae are self-fertile hermaphrodites that maintain males at a low frequency (androdioecious), whereas the morphologically similar C. remanei and C. sp. CB5161 are obligate female/male (gonochoristic) species [6,7,10]. Phylogenetic analysis of the four closely related Caenorhabditis species suggests that self-fertile hermaphroditism has evolved independently in C. elegans and C. briggsae from an ancestral male/female state [10,11]. Importantly, a transition in mating system from female/male to hermaphroditic (or hermaphroditic to male/female) requires that one or more changes in the sex determination pathway have occurred.

C. elegans and C. briggsae, like many other animals, have two sexes specified by the ratio of X chromosomes to sets of autosomes [8,12,13]. In both species, XX animals are somatically female while the germline is hermaphroditic. Self fertility is achieved by a transient period of spermatogenesis beginning in the third larval (L3) stage before the organism switches to the production of oocytes in the L4 stage, which continues throughout adulthood [14,15]. In both species, XO males begin sperm production in the L3 stage and continue spermatogenesis throughout their reproductive lives [14,16,17].

A major determinant of germline sexual fate in C. elegans is the relative activity of two key regulators: tra-2, which promotes the female fate (oocyte), and fem-3, which promotes the male fate (sperm) [18,19] (Figure 1

A). The activities of tra-2 and fem-3 must be regulated in both males and hermaphrodites to allow spermatogenesis to occur, however the mechanisms by which this regulation occurs differs between the two sexes. In males, her-1 represses tra-2 feminizing activity and raises the relative level of fem-3 activity so that spermatogenesis is continuous [20,21]. Since null mutations in her-1 have no effect on hermaphrodites and her-1 is not expressed in XX animals, a different mechanism is used to allow for the transient production of sperm [22,23].

Figure 1

The C. elegans XX Hermaphrodite Germline Sex Determination Pathway

Self fertility in C. elegans hermaphrodites is achieved by an early period of spermatogenesis followed by a later period of oogenesis (Figure 1

A). The promotion of spermatogenesis during the L3 stage (early) is achieved by translational repression of the tra-2 mRNA mediated by gld-1 (“defective in germline development”) and fog-2 (“feminization of germline”)[24,25] (Figure 1

A and and1B).1

B). The transient reduction in the level of tra-2 feminizing activity raises the relative level of fem-3 masculinizing activity to promote spermatogenesis (Figure 1

A). Later in L4 and adult animals, oogenesis is promoted by relieving the fog-2/gld-1-mediated repression of tra-2 feminizing activity combined with repression of fem-3 masculinizing activity by mog-1 to mog-6, fbf-1 and fbf-2, and nos-1 to nos-3 [18,19,26].

Central to this work are the genes fog-2 and gld-1. fog-2 is required for hermaphrodite, but not male, spermatogenesis in C. elegans, as XX animals that lack fog-2 produce only oocytes, resulting in functional females, whereas XO males are unaffected [27]. Similarly, loss-of-function mutations in gld-1 result in the feminization of the hermaphrodite germline without affecting males [28,29]. Both fog-2 and gld-1 are germline-specific regulators of sexual fate, since they do not appear to be expressed in the soma, and null mutations in either gene do not affect somatic sexual fate [25,27,28,29,30].

C. elegans gld-1 is a germline-specific tumor suppressor that is indispensable for oogenesis [28,29] and encodes a conserved KH-type RNA-binding protein [30]. GLD-1 is a translational repressor that binds to multiple mRNA targets [31], including tra-2, where it binds as a dimer to each of two tra-2 and GLI elements (TGEs) present on the 3′ untranslated region (UTR) of the tra-2 mRNA [24,32] (Figure 1

B). Deletion of the tra-2 TGEs results in a loss of GLD-1-mediated translational control and feminization of the germline, such that only oocytes are produced [20,25,33,34].

C. elegans FOG-2 was identified as a GLD-1-interacting protein with a structure similar to canonical F-box proteins; it has an N-terminal F-box and a C-terminal protein–protein interaction domain. In the case of FOG-2 the putative protein–protein interaction domain is referred to as Duf38 (Pfam in [35]) or FOG-2 homology domain (FTH) [25]. F-box proteins are often core components of the Skp1/Cullin/F-box-type E3 ubiquitin ligase complexes, and they serve to link specific substrates to the ubiquitin ligase machinery for subsequent proteolysis [36]. However, FOG-2 cannot target GLD-1 for degradation since both function to promote hermaphrodite spermatogenesis [25] (Figure 1

A). Current data suggest that the formation of a FOG-2/GLD-1/tra-2 mRNA ternary complex mediates translational repression of tra-2 and a corresponding reduction in feminizing activity to allow hermaphrodite spermatogenesis [24,25] (Figure 1

B).

The completion of the C. elegans genome sequence [4] and the 10X sequence (representing more than 98% coverage) of the closely related species C. briggsae [5] permits studies of the evolution of sex determination and the inception of hermaphrodite spermatogenesis in morphologically comparable species. Here, we pose the question, do C. elegans and C. briggsae specify male sexual fate in the hermaphrodite germline similarly?

We find that 30 of 31 C. elegans sex determination genes have C. briggsae orthologs, indicating that there is extensive conservation of sex determination pathway components; the lone exception is fog-2. We provide evidence that the essential role of FOG-2 in C. elegans hermaphrodite spermatogenesis evolved from post-speciation duplication and divergence of the fog-2-related (FTR) gene family and that a fog-2 gene is not present in C. briggsae. Furthermore, double-stranded-RNA-mediated interference (RNAi) of the gld-1 ortholog in C. briggsae results in masculinization of the germline instead of the feminization of the germline phenotype observed in C. elegans. The lack of a potential C. briggsae fog-2 combined with the opposite sex determination function of GLD-1 in C. briggsae indicate that the control of hermaphrodite spermatogenesis, while using most of the same gene products, is fundamentally different between the species and is likely to have evolved independently.

Results

Components of Sex Determination Pathway Are Conserved between C. elegans and C. briggsae

To survey conservation in the sex determination pathway between C. elegans and C. briggsae we used reciprocal best BLAST [37,38,39] to identify potential C. briggsae orthologs of 31 known C. elegans sex determination genes, some of which have been previously identified. The 31 genes included 16 that function only in germline sex determination, seven that function in both somatic and germline sex determination, two that function only in somatic sex determination, and six that coordinate sex determination and dosage compensation. We found that 30 of 31 genes have C. elegans–to–C. briggsae reciprocal best BLAST hits and alignments consistent with a high level of conservation (Table 1). Using this method, putative orthologs of all known sex determination genes, including less conserved members, and previously identified genes were recovered [17,26,40,41,42,43,44], with the notable exception of fog-2.

Table 1

Comparative Analysis of Sex Determination Genes in C. elegans and C. briggsae

The functions of seven C. briggsae sex determination genes have been tested, and current data indicate that these genes exhibit similar and possibly identical functions in C. elegans and C. briggsae (her-1 [43], tra-2 [21], fem-1 [A. Spence, personal communication], fem-2 [45], fem-3 [41], fog-3 [42], and tra-1 [17]). Importantly, the epistatic relationship and function of two key regulators of sex determination, tra-2 and fem-3, are essentially intact between the sister species in somatic sex determination [21,41] (Figure 1

A). At first glance, given the conservation of 30/31 sex determination genes, similar or identical functions for 7/7 genes tested, and maintenance of a key epistatic relationship, it would appear that the sex determination pathway is generally conserved between C. elegans and C. briggsae. However, genetic and molecular studies will be required to determine whether the C. briggsae orthologs are functionally equivalent to their C. elegans counterparts.

A single FOG-2 ortholog could not be resolved by reciprocal best BLAST or by using the reciprocal smallest distance algorithm [46], which uses global sequence alignment and maximum likelihood estimation of evolutionary distances, to infer putative orthologs (data not shown). This indicates that fog-2 is either highly diverged, present in an unsequenced portion (<2%) of the C. briggsae genome, or potentially a C. elegans–specific adaptation not present in C. briggsae.

fog-2 Is a C. elegans–Specific Adaptation

FOG-2 is part of a large, highly diverged F-box- and DUF38/FTH-containing protein family in C. elegans with more than 100 members referred to as FTR proteins [25,36]. The FTR family is also expanded in C. briggsae, making the identification of a single functionally equivalent ortholog from a large number of paralogs difficult. Therefore, to discern the relationships among C. elegans and C. briggsae FTR family members, 30 C. elegans and C. briggsae FTR proteins or protein predictions closely related to FOG-2 were used to generate a neighbor-joining phylogeny. The remaining, more diverged FTR members from either species were not included in the phylogeny to avoid long branch attraction [47].

The C. elegans and C. briggsae FTR phylogeny reveals that all of the C. elegans FOG-2 relatives form a single clade and all of the C. briggsae relatives a distinct clade. An unrooted radial phylogram illustrating C. elegans and C. briggsae FTR relationships is presented in Figure 2

, and a rectangular representation of the same phylogeny with bootstrap support information is shown in Figure S1. If a closely related homolog of C. elegans FOG-2 were present in C. briggsae the expectation is that it would have clustered with the C. elegans proteins. Contrary to this, the phylogenetic separation of C. elegans and C. briggsae FTR family members into distinct lineages indicates that extensive expansion in the FTR family occurred post-speciation and that C. elegans and C. briggsae FTR genes do not have one-to-one orthologous relationships.

Figure 2

The FTR Gene Family in C. elegans and C. briggsae

The above results could be misleading if a closely related C. briggsae fog-2 homolog were present in the less than 2% of the genome sequence that is not present in the final assembly or if the fog-2 ortholog diverged sufficiently such that the computational methods were not able to distinguish between orthologous and paralogous relationships. To address these possibilities we used low-stringency cross-species Southern blotting in an effort to identify closely related fog-2-like sequences in unsequenced portions of the C. briggsae genome, and we used conserved synteny in an attempt to identify a diverged fog-2 ortholog that might reside in the same genomic location. Both approaches were used to effectively identify other diverged sex determination genes from C. briggsae (tra-2, her-1, and fem-2) prior to the release of the C. briggsae genome sequence [40,43,44].

For low-stringency Southern blotting we used a C. elegans fog-2 probe and a fem-2 positive control probe against C. briggsae genomic DNA. Under conditions that detected cross-species hybridization with the C. elegans fem-2 probe against C. briggsae genomic DNA [40], no C. briggsae signal was observed with the C. elegans fog-2 probe (Figure 3

A). This suggests either that a close fog-2 relative is not present in the less than 2% of the C. briggsae genome that is unsequenced or that it has diverged significantly beyond the level of fem-2.

Figure 3

fog-2 Is Likely Absent in C. briggsae

For analysis of conserved syntenic relationships, five conserved C. elegans genes surrounding fog-2 (srg-34, sec-23, psa-1, Y113G7A.14, and Y113G7B.15) were used to query C. briggsae contigs. The genes srg-34, sec-23, and psa-1 are highly conserved across metazoans and have reciprocal best BLAST hits in C. briggsae (Figure 3

B and and3C,3

C, small arrow heads). The genes Y113G7A.14 and Y113G7B.15 flank the gene-dense C. elegans fog-2 region and also have reciprocal best BLAST hits in C. briggsae (Figure 3

B and and3C,3

C, large arrow heads). All five genes were found to be represented on a single C. briggsae contig, suggesting that the global syntenic relationships are conserved, but with detailed analysis revealing a number of differences in gene order (Figure 3

B and and3C).3

C). However, fog-2, its four adjacent close FTR relatives, and 16 surrounding genes in an 82.6-kb region were absent from this C. briggsae contig, while the conserved genes on either side were present (Table S1 and S2).

The closest relative of fog-2 is the gene ftr-1, which is part of a group of five closely related ftr genes that are colinear in C. elegans and not present in C. briggsae [25] (Figures 2

and and3).3

). If fog-2 and ftr-1 are the result of a “recent” post-speciation duplication within the C. elegans lineage, as suggested by the phylogeny, then we would expect that fewer synonymous substitutions (K_s) have occurred between fog-2 and ftr-1 relative to other C. elegans/C. briggsae best BLAST orthologs. Consistent with a recent duplication, the K_s for fog-2/ftr-1 is not saturated (K_s = 0.36) whereas the average K_s for reciprocal best BLAST hits between C. elegans and C. briggse is saturated (K_s = 1.72) [5].

The finding that fog-2 and ftr-1 arose from a relatively recent local duplication within C. elegans strongly supports the contention that fog-2 is not present in C. briggsae. These results imply that C. briggsae must regulate hermaphrodite spermatogenesis differently than C. elegans.

The Diverged C-Terminal of FOG-2 Is Necessary for GLD-1 Binding

Previous work has shown that FOG-2 is an integral part of the tra-2 3′ UTR translational repression complex. The RNA-binding protein GLD-1 makes direct contact with the tra-2 3′ UTR, and FOG-2 is recruited to the complex via its interaction with GLD-1 [24,25]. In spite of the high similarity between fog-2 and ftr-1 (Figure 4

), ftr-1 cannot compensate for fog-2 in the promotion of hermaphrodite spermatogenesis [25]. This indicates that fog-2 must contain unique sequences that allow it to function in sex determination.

Figure 4

The Highly Diverged FOG-2 C-Terminal Region Is Responsible for GLD-1 Interaction in C. elegans

Pairwise comparisons between FOG-2 and FTR-1 reveal a highly diverged C-terminal region encoded by the final exon (exon 4) (Figure 4

A–4

C). Before the C-terminal region of low similarity, the relative reading frames of fog-2 and ftr-1 are conserved with all insertions and deletions in three nucleotide multiples and an overall amino acid identity of 70%. Within the final exon, multiple amino acid substitutions, insertions, and deletions have occurred, resulting in a region of low nucleotide and amino acid identity (Figure 4

B and and4C).4

C). For example, an indel (deletion relative to fog-2) at nucleotide 805 shifts the reading frame of FOG-2 relative to FTR-1 and results in a region of low similarity between the proteins (Figure 4

B). A second indel at position 819 restores the reading frame but additional substitutions result in a diverged amino acid sequence (Figure 4

C).

The dramatic differences between the FOG-2 and FTR-1 C-terminal regions suggested a connection between the unique functionality of FOG-2 in sex determination and the highly diverged C-terminal region. Since FOG-2 interacts with GLD-1 and both are required for the promotion of the male germ cell fate in the hermaphrodite, we determined whether the diverged FOG-2 C-terminal region was necessary for its interaction with GLD-1 (Figure 4

). Progressive C-terminal deletions of FOG-2 were tested for their ability to interact with GLD-1 in the yeast two-hybrid system (Figure 4

D). Full-length FOG-2 interacts with GLD-1 [25]; however, C-terminal deletions of nine and 28 aa in FOG-2 reduced the interaction, and deletion of 64 and 76 aa (essentially all of exon 4) eliminated the interaction (Figure 4

D), indicating that the highly divergent C-terminal region is necessary for GLD-1 binding. All full-length and deletion constructs were tested against the Skp1 homolog SKR-1 as a positive control for functionality in the two-hybrid system (see Materials and Methods).

To determine whether the C-terminal region of FOG-2 is sufficient to confer GLD-1 interaction, an FTR-1/FOG-2 exon 4 chimera was generated and assayed for its ability to interact with GLD-1. Normally FTR-1 lacks the ability to interact with GLD-1 [25] (Figure 4

D). The replacement of exon 4 from ftr-1 with exon 4 from fog-2 allowed the chimera to interact with GLD-1 (Figure 4

D). Thus, the C-terminal 74aa region of FOG-2, when in the context of the FTR-1 F-box and Duf38/FTH sequences, is sufficient to confer GLD-1 binding.

FOG-2/GLD-1 Interaction Evolved Rapidly in C. elegans

Gene duplication provides the raw material for the evolution of novel adaptations, having been implicated in the diversity of the host–pathogen immune response, rapid onset of insecticide resistance, and diversity of vertebrate body plans [48]. Rapidly evolving genes, or portions of genes, under positive selection can be identified by comparison of nucleotide alterations that result in amino acid changes (non-synonymous substitutions [K_a]) to alterations that do not change the amino acid (K_s) [49,50]. K_a/K_s ratios that are equal to or less than one are indicative of neutral or purifying selection, where substitutions that change amino acids offer no fitness advantage or result in lowered fitness. In contrast, K_a/K_s ratios greater than one, common in rapidly evolving genes, are indicative of positive selection, where non-synonymous changes offer some fitness advantage and are fixed at a higher rate than synonymous substitutions [51].

To determine the selection acting on the fog-2/ftr-1 duplication we compared K_a/K_s ratios between fog-2, ftr-1, and the five FTR genes closest to fog-2 in C. elegans. Pairwise comparisons of codon-delimited full-length coding sequences closely related to fog-2 suggest that purifying selection dominates along the fog-2 branch, as all comparisons produced K_a/K_s ratios less than one (mean = 0.46). However, while the overall K_a/K_s ratio for fog-2/ftr-1 is not indicative of positive selection (mean = 0.58), sliding-window K_a/K_s ratio estimates [52] for fog-2 and ftr-1 indicate that the highly diverged C-terminal region of FOG-2/FTR-1 contains residues under positive selection (K_a/K_s = 1.98 for nucleotides 777–987, windows 33–37) (Figure 4

). An alternate method using maximum likelihood estimation of K_a/K_s (PAML and codeml [53]) confirmed the presence of residues under positive selection within the C-terminal region (see Materials and Methods). Thus, the primary differences between FOG-2 and FTR-1 are localized to the rapidly evolving C-terminus of FOG-2 that is required for GLD-1 binding and is under positive selection.

The yeast two-hybrid data, together with the genetics of fog-2 [25], indicate that FOG-2 is unique among C. elegans FTR genes in functioning with GLD-1 in germline sex determination. Given the specificity of the FOG-2/GLD-1 interaction in C. elegans, phylogenetic analysis of FTR proteins (see Figure 2

), and additional experiments (see Figures 3

and and4)4

) that indicate that there are no close relatives of fog-2 among C. briggsae FTR genes, it is unlikely that any C. briggsae FTR protein functions with C. briggsae GLD-1 in sex determination.

In contrast with FOG-2, a highly conserved GLD-1 ortholog is present in C. briggsae (Table 1) and has a germline expression pattern essentially identical to that of C. elegans (Figure 5

A, top right and middle right). In fact, C. elegans GLD-1 and C. briggsae GLD-1 share 81% amino acid identity overall and more than 90% in the maxi-KH RNA-binding region. Since FOG-2 and GLD-1 function together to promote the male germ cell fate in C. elegans hermaphrodites, this raised the question of what role, if any, C. briggsae GLD-1 plays in C. briggsae germline sex determination.

Figure 5

GLD-1 Has the Opposite Sex Determination Function in C. elegans and C. briggsae

GLD-1 Has Distinct Functions in C. elegans and C. briggsae Germline Sex Determination

To examine C. briggsae GLD-1 function in sex determination we performed RNAi [54] by injecting double-stranded C. briggsae gld-1 RNA into C. briggsae adult hermaphrodites followed by phenotypic analysis of F1 self progeny (see Materials and Methods). From genetic analysis of C. elegans gld-1 [28,29] there are two functions relevant to this study. First, C. elegans GLD-1 has an essential function in meiotic prophase progression during oogenesis. In null mutant hermaphrodites oogenic germ cells progress to pachytene and then return to the mitotic cell cycle, giving rise to ectopic proliferation and a germline tumor [28]. For this function C. elegans GLD-1 acts to spatially restrict the translation of multiple target mRNAs during oogenesis. GLD-1 oogenic target mRNAs are repressed during early meiotic prophase, but then are translated during late meiotic prophase following the loss of GLD-1 at the end of pachytene [30,31,55]. Second, C. elegans GLD-1 is necessary for the specification of the male sexual fate in the hermaphrodite germline. This function is most simply revealed as a haplo-insufficient feminization of the hermaphrodite germline [28,29]. C. elegans gld-1 has no known essential functions in male meiotic prophase progression or in XO male germline sex determination as C. elegans null males are wild-type [28,29].

C. briggsae GLD-1 may still function as a translational repressor of C. briggsae tra-2 mRNA even in the absence of a FOG-2 ortholog. This is a possibility because FOG-2 is not required for C. elegans GLD-1 binding to the C. elegans tra-2 mRNA in vitro [25], and some conservation is preserved between the C. elegans and C. briggsae tra-2 3′ UTRs [34]. In this case, RNAi of GLD-1 in C. briggsae might feminize the germline given that C. briggsae tra-2 promotes female development in both the germline and soma [21]. Alternatively, C. briggsae GLD-1 might have no role in germline sex determination, in which case RNAi would not result in a sex determination phenotype.

Surprisingly, C. briggsae gld-1 RNAi resulted in a masculinized germline (Figure 5

A, bottom; Table 2), with no effect on the soma. Staining with 4′,6′-diamidino-2-phenylindole hydrochloride (DAPI) and anti–major sperm protein (MSP) (see Materials and Methods) revealed continuous spermatogenesis leading to a vast excess of sperm at the expense of oogenesis. Anti-GLD-1 antibody staining of gld-1 RNAi F1 gonad arms indicated that the level of GLD-1 protein was reduced to below detectable limits (Figure 5

A, bottom right). C. briggsae control hermaphrodites injected with double-stranded RNA for green fluorescent protein (GFP) had gonad morphology identical to wild-type (Figure 5

A, top left and middle left). The masculinized phenotype of gld-1 RNAi in C. briggsae indicates that the wild-type function of GLD-1 in C. briggsae is to promote the female germ cell fate, likely by the translational repression of an mRNA that encodes a masculinizing gene product. This function is in direct contrast to that of C. elegans GLD-1, which promotes the male germ cell fate by translational repression of the feminizing tra-2 mRNA.

Table 2

Summary of GLD-1 RNAi Germline Phenotype in C. elegans and C. briggsae

GLD-1 Function in Meiotic Prophase Progression during Oogenesis Is Conserved

Given the difference in sex determination function, it is possible that C. elegans and C. briggsae GLD-1 have few conserved functions in germline development. To investigate this we took advantage of well-defined activities of gld-1 in C. elegans such as its essential function in female meiotic prophase progression and in the translational repression of the evolutionarily conserved yolk receptor mRNA encoded by the rme-2 locus [28,31].

The gld-1-null tumorous phenotype results from aberrant oogenic prophase progression and a return to mitosis [28,29]. This phenotype is dependent on germline sex because a tumor only occurs when germ cell fate is set to female [28,29]. The masculinized phenotype caused by gld-1 RNAi in C. briggsae is likely to preclude the detection of this function as the C. elegans gld-1-null tumorous phenotype is suppressed by mutations that cause masculinization of the germline [29]. To overcome the masculinization we combined fog-3 RNAi with gld-1 RNAi in C. briggsae. Since C. elegans fog-3 functions near the end of the sex determination pathway and in C. briggsae fog-3 RNAi results in feminization of the germline [42], we predicted that C. briggsae fog-3 RNAi would be epistatic to the masculinization of the germline of C. briggsae gld-1 RNAi.

Similar to the C. elegans gld-1-null, RNAi of gld-1 or gld-1 and fog-3 in C. elegans and double RNAi of gld-1 and fog-3 in C. briggsae resulted in a robust proximal germline tumor (Figure 5

B; Table 2). Control RNAi with fog-3 alone resulted in feminized germlines in both species [42]. Both the mitotic zone and transition zone appear to have roughly normal nuclear morphology, with more proximal nuclei having abnormal pachytene morphology (Figure 5

B), suggesting that germ cells are entering meiosis but progressing aberrantly before returning to mitosis. The return-to-mitosis tumorous phenotype in each species was confirmed using phosphohistone H3 staining, a mitotic proliferation marker [56]. We cannot rule out the possibility that the C. briggsae phenotypes observed, masculinization of the germline with gld-1 RNAi alone and tumorous germline with gld-1 and fog-3 RNAi, are the result of incomplete knockdown leading to partial gld-1 loss of function.

The rme-2 yolk receptor mRNA is a known target of GLD-1-mediated translational repression in C. elegans [31]. In C. elegans, GLD-1 and RME-2 have mutually exclusive expression patterns because rme-2 mRNA is translationally repressed in the transition zone and pachytene region, where GLD-1 levels are high, and translated in oocytes, where GLD-1 levels are low [31]. In C. elegans gld-1-null germlines RME-2 is ectopically expressed in the transition zone and pachytene region owing to loss of GLD-1-mediated translational repression of the rme-2 mRNA [31].

A similar, mutually exclusive accumulation pattern in C. briggsae suggests that C. briggsae GLD-1 is a translational repressor of C. briggsae rme-2 mRNA (Figure 6

). To determine whether C. briggsae GLD-1 represses the rme-2 mRNA, double RNAi of gld-1 and fog-3 was performed in both species, and gonad arms were stained for RME-2 protein [57]. Reduction of GLD-1 and FOG-3 by RNAi results in the ectopic accumulation of RME-2 protein in both C. elegans and C. briggsae (Figure 6

), indicating that the role of GLD-1 in the translational repression of the rme-2 mRNA is conserved. Thus, despite the opposite roles of GLD-1 in sex determination, essential functions of GLD-1 in oogenesis are conserved between the species.

Figure 6

GLD-1-Mediated Translational Repression of rme-2 mRNA in C. elegans and C. briggsae

Discussion

Our results indicate that the control of hermaphrodite spermatogenesis is fundamentally different between the sister species C. elegans and C. briggsae at the level of FOG-2/GLD-1/tra-2 mRNA regulation. While FOG-2 is essential for self-fertile hermaphroditism in C. elegans, a closely related homolog of FOG-2 could not be recovered in C. briggsae by reciprocal best BLAST, phylogenetic inference, low-stringency hybridization, or analysis of conserved synteny. Comparison of synonymous changes between fog-2 and its closest relative, ftr-1, indicates that fog-2 is the product of a recent expansion “specific” to C. elegans in the FTR gene family and implies that the evolution of FOG-2 and its incorporation into the sex determination pathway occurred post-speciation. Consistent with this, the C-terminal region of FOG-2 required for binding to GLD-1 was found to be highly diverged and “unique” to FOG-2 in C. elegans. Interestingly, GLD-1 was found to have a sex determination function in C. briggsae opposite that in C. elegans while retaining similar functions in female meiotic prophase progression and oogenesis. The absence of FOG-2, and the opposite sex determination function of GLD-1, provides evidence for the independent evolution of hermaphroditism in C. elegans and C. briggsae.

General Conservation of the Sex Determination Pathway

Reciprocal best BLAST indicates that C. elegans and C. briggsae have orthologs of 30 of 31 known sex determination pathway genes. Conserved functions for C. briggsae her-1, tra-2, fem-1, fem-2, fem-3, fog-3, and tra-1 have been demonstrated by transgene rescue of C. elegans mutations or similarity of RNAi loss-of-function phenotype [17,21,26,41,42,43,45]. The general conservation of genes that govern sex determination suggests that the underlying pathway remains largely intact between the species.

RNAi and transgenic experiments have suggested that while fem-2 and fem-3 have conserved roles in the somatic sex determination of both species, they may play diminished roles in C. briggsae germline sex determination [41,45]. There are two possibilities that could explain these results. One is that there are inherent species-specific differences in susceptibility to RNAi or in the ability to reconstitute complete gene function by transgene rescue. The other is that differences in C. elegans and C. briggsae phenotypes reveal functional divergence in sex determination pathway components. Analysis of null mutations in C. briggsae orthologs of C. elegans sex determination genes will help to distinguish between these possibilities. While some functional differences may turn out to be valid, tra-2 (feminizing) and fem-3 (masculinizing) apparently play the same somatic roles in both species, and their epistatic relationship appears to be conserved [41].

fog-2 Is Unique to C. elegans

Within the context of general conservation of sex determination pathway components and conserved key epistatic relationships, the absence of fog-2 in C. briggsae is intriguing. fog-2 arose as a consequence of recent C. elegans–specific gene duplication events, and none of the closely related C. elegans fog-2 paralogs can compensate for loss of fog-2 in sex determination [25]. Thus, it is unlikely that more distantly related C. briggsae FTRs are involved in GLD-1/tra-2-mRNA-mediated promotion of hermaphrodite spermatogenesis. Since fog-2 is essential for the promotion of spermatogenesis in C. elegans hermaphrodites and is not present in C. briggsae, the direct implication is that specification of the male germ cell fate in C. briggsae hermaphrodites is fundamentally different from that in C. elegans and that it evolved independently.

The highly diverged C-terminus of FOG-2 is under positive selection and is necessary and sufficient for GLD-1 binding within the context of an F-box and FTH domain (see Figure 4

). Acquiring the diverged C-terminus was crucial in FOG-2 becoming incorporated into the sex determination pathway. With respect to the C. elegans lineage, it is unclear whether fog-2 retains an ancestral function in sex determination and ftr-1 has changed/drifted away or, alternatively, whether ftr-1 represents the ancestral function and fog-2 has recently evolved a role in sex determination (also see Figure S2). The ftr-1 gene is expressed, though its function is currently unknown. RNAi of ftr-1 into the fog-2 null did not reveal any obvious phenotypes beyond feminization of the germline [25].

Conserved GLD-1 Functions in C. elegans and C. briggsae Meiotic Prophase during Oogenesis

GLD-1 function in meiotic prophase progression and oogenesis shows substantial conservation between the species (see Figures 5

and and6),6

), which is not surprising given the high level of sequence conservation between C. elegans and C. briggsae GLD-1. This is illustrated by the rme-2 yolk receptor mRNA being regulated similarly between the species (Figure 6

). Current data indicate that C. elegans GLD-1 binds to, and likely represses translation of, more than 100 mRNA targets [31,55] (M.-H. Lee, V. Reinke, and T. Schedl, unpublished data). The C. elegans gld-1-null tumorous phenotype likely results from misregulation of multiple mRNA targets [31]. While the identity of the misregulated mRNA targets causing the gld-1-null tumorous phenotype are currently unknown, the fact that C. briggsae gld-1 and fog-3 RNAi results in a similar tumorous phenotype suggests that a similar, if not identical, set of C. briggsae GLD-1 mRNA targets are misregulated. The absence of a FOG-2 ortholog in C. briggsae is unlikely to have a major effect on GLD-1-mediated translational control since FOG-2 appears to be required only as a cofactor for tra-2 repression [25,27,31,55,58]. Thus, it is possible that the majority of GLD-1 mRNA targets involved in prophase progression and oogenesis are regulated similarly between species.

Divergent GLD-1 Function in C. elegans and C. briggsae Sex Determination

Genetic analysis reveals that C. elegans and C. briggsae GLD-1 have opposite functions in germline sex determination; C. elegans GLD-1 promotes spermatogenesis while C. briggsae GLD-1 promotes oogenesis. This indicates that the major sex determination function of C. briggsae GLD-1 is not translational repression of tra-2 feminizing activity. C. elegans GLD-1 binds two 28 nucleotide direct repeat elements on the C. elegans tra-2 mRNA 3′ UTR to mediate translational repression [24]. Somatic reporter gene assays in C. elegans and C. briggsae have suggested that the tra-2 3′ UTRs of both species are able to function in translational repression [34], with the implication being that the C. elegans and C. briggsae 3′ UTRs are regulated similarly. However, these data are difficult to interpret in the context of germline sex determination, as GLD-1 and FOG-2 are not natively expressed in the soma and neither GLD-1 nor FOG-2 have essential functions in somatic sex determination [25,27,28,29,30].

One hypothesis to explain our results is that C. briggsae GLD-1 binds to the C. briggsae tra-2 mRNA but is necessary for translational activation instead of translational repression as in C. elegans. However, for all characterized C. elegans GLD-1 targets, and C. briggsae rme-2 mRNA, GLD-1 acts as a translational repressor [2,31,55,58,59]. We currently do not understand how FOG-2 acts with GLD-1 in translational repression of C. elegans tra-2 mRNA. In C. elegans, GLD-1 can bind the tra-2 mRNA in the absence of fog-2 in worm extracts but cannot properly repress its translation in vivo [25]. This suggests that the role of FOG-2 may be to recruit additional factors specific to the C. elegans tra-2 mRNA 3′ UTR that allow for efficient GLD-1 translational repression. Assuming C. briggsae GLD-1 binds C. briggsae tra-2 mRNA in vivo, given the absence of a FOG-2 ortholog, there may be no regulatory consequence of this binding.

Another possibility is that C. briggsae GLD-1 binds and translationally represses an mRNA that promotes spermatogenesis. This could occur if a masculinizing sex determination gene, either present in both species or unique to C. briggsae, has come under GLD-1 control in C. briggsae. Given the conservation of GLD-1 and its regulation of at least some common targets (e.g., rme-2) it is unlikely that changes in GLD-1 are responsible for a new mRNA target in C. briggsae. Instead, it is more likely that one or more new target mRNAs have acquired sequences that direct GLD-1 binding and translational repression. The requirements for GLD-1 binding are only just being elucidated, with a hexanucleotide sequence being one important feature amid otherwise diverse GLD-1 binding regions [32,55]. Thus, small numbers of changes in UTRs are likely to be sufficient for new mRNAs to come under GLD-1-mediated regulation.

Evolution of Self-Fertile Hermaphroditism

Current phylogenetic data suggest that hermaphroditism evolved independently in Caenorhabditis and other lineages of Rhabditid nematodes from an ancestral female/male state [5,6,7,10,11,60]. This is consistent with our results showing that control of hermaphrodite spermatogenesis at the level of FOG-2/GLD-1/tra-2 mRNA is fundamentally different between C. elegans and C. briggsae. This raises the question, how might the transition from the ancestral female/male to hermaphrodite/male system of reproduction have occurred multiple times within the Caenorhabditis clade?

The anatomy and reproductive physiology of C. elegans allow both sperm that is introduced by mating and sperm that develops within the female gonad of the hermaphrodite to be effectively used in reproduction [14,61,62]. Either source of sperm generates a MSP-derived signal that is required for full-grown oocytes to undergo meiotic maturation, ovulation, and fertilization in the spermatheca [62,63]. Not only is the anatomy conserved but an MSP-derived sperm signal also appears to be utilized by both C. briggsae and C. remanei (a female/male species) to induce oocyte maturation and ovulation [63,64]. This conservation within Caenorhabditis indicates that major changes in anatomy and reproductive physiology are not necessary in the transition from female/male to hermaphrodite/male reproduction.

The relative ease with which mutants and mutant combinations can alter the sex determination system in C. elegans has suggested that transitions between mating systems may not be difficult and that the overall sex determination pathway reflects selection for a particular mating system rather than a constant regulatory mechanism [65]. The hermaphrodite pattern of spermatogenesis first then oogenesis is achieved by high masculinizing/low feminizing activity in early larvae followed by low masculinizing/high feminizing activity in late larvae/adults (see Figure 1

; reviewed in [18,19,26,66]). Lowering or eliminating germline masculinizing activity in XX animals can convert C. elegans from hermaphrodite/male to female/male reproduction (Table 3, and references therein [20,27,28,29,66,67,68,69]). For example, fog-2-null mutations result in strains that reproduce as XX females and XO males. The mutant female/male strains can be converted back to hermaphrodite reproduction by introducing masculinizing mutations in certain genes (e.g., fog-2-null; fem-3-gf; Table 3). The generality of high masculinizing/low feminizing activity early followed by low masculinizing/high feminizing activity late is borne out by other sets of mutually suppressing feminizing-plus-masculinizing combinations in which the double mutants are self-fertile while each single mutant is usually self-sterile (e.g., tra-1-gf; fem-3-gf; Table 3). Thus, multiple genetic states can yield self-fertile hermaphrodite/male and male/female reproduction in C. elegans.

Table 3

C. elegans Sex Determination Mutants That Yield Female/Male Reproduction and Mutually Suppressed Hermaphrodite Reproduction

Given the conservation of anatomy and reproductive physiology, an initial conversion from an ancestral Caenorhabditis female/male species to a hermaphrodite/male mode of reproduction may only require a genetic event that results in a transient increase in germline masculinizing activity in early larvae to produce sperm. As long as this change does not interfere with the higher level of feminizing activity (oogenesis) in late larvae/adults, self fertility would be possible. After the establishment of self fertility, there would likely be strong selection for additional genetic events that would optimize self-fertile brood size [70] and result in a clean transition from sperm to oocyte development so that wasteful intersexual gametes are not formed (Table 3). Thus, it is very likely that multiple genetic events now define the differences in the C. elegans and C. briggsae germline sex determination pathways.

In C. elegans, the relative levels of TRA-2 feminizing to FEM-3 masculinizing activity appear to be the major regulatory point for the sperm-then-oocyte pattern. There is no a priori reason for TRA-2 or FEM-3 to be the major focus of regulation to achieve hermaphroditism in C. briggsae; if one of these is the focus, then at least some of the regulation must differ between C. elegans and C. briggsae, given the absence of fog-2 and the changed role of GLD-1. Since the last common ancestor of C. briggsae and C. elegans must have contained orthologs of 30 of 31 C. elegans sex determination genes, a change in the regulation of one or more of these genes might be responsible. Alternatively, since much of the regulation of C. elegans germline sex determination is by translational control, mutations in UTRs of mRNAs may result in new genes coming under the control of GLD-1 or another RNA sex determination gene regulator (Table 1). Additionally, duplication and divergence, analogous to what we have found for FOG-2 in C. elegans, may have resulted in a new gene being incorporated into the germline sex determination pathway. To move beyond speculation, the forward genetic analysis currently in progress (R. Ellis and E. Haag, personal communication) will be important for the identification of C. briggsae–specific genes, analogous to fog-2, that are necessary for self-fertile hermaphroditism.

Materials and Methods

Sex determination pathway conservation

Protein coding sequences of cloned C. elegans sex determination genes were obtained from Wormbase (http://www.wormbase.org; WormPep release 112). C. briggsae genomic sequence was obtained from The Sanger Institute (Cambridge, United Kingdom) or the Genome Sequencing Center (St. Louis, Missouri, United States), and protein sequences were obtained from either Wormbase or Ensemble (http://www.ensembl.org/; version 17.25.1). Best BLAST orthologs of C. briggsae sex determination proteins were obtained using C. elegans sex determination protein sequences as queries against C. briggsae predicted proteins and six-frame translated C. briggsae genomic sequence. C. briggsae proteins obtained at an e-value cutoff of 1 × 10⁻⁵⁰ reciprocal best hits were recovered for 26 of 31 C. elegans proteins. NOS-1 and XOL-1 orthologs were identified at an e-value cutoff of 1 × 10⁻²⁰ and were also reciprocal best BLAST hits between species. In each case a single reciprocal best hit was identified for each component of the sex determination pathway with the exception of FBF-1 and FBF-2, which returned the same best BLAST hit, and FOG-2. Searches of the non-redundant National Center for Biotechnology Information protein database (GenBank CDS+PDB+SwissProt+PIR+WormPep) with full-length FOG-2 as query revealed only weak similarity to the F-box motif for non–C. elegans or –C. briggsae sequences. Using the highly diverged C-terminal end of FOG-2, including a portion of the Duf38/FTH, or the GLD-1 interaction region of FOG-2 as query did not reveal any hits below an e-value of 0.01 in C. elegans or C. briggsae other than FOG-2 and FTR-1.

Identification of FTR family members

FTR family members are defined by the presence of an N-terminal F-box and C-terminal Duf38/FTH domain (FTR) [25]. C. elegans FTR family members were identified using FOG-2 as a query against WormPep release 112. Each potential FTR was scanned for an N-terminal F-box motif and C-terminal Duf38/FTH domain using the hidden Markov models (HMMs) for each domain (HMMER 2.3.2) [35]. Similarly, C. briggsae FTR family members were identified using FOG-2 as a BLAST query and HMMs. In C. elegans, fog-2 (Y113G7B.5), ftr-1 (Y113G7B.4), CE35646 (Y113G7B.1), CE24144 (Y113G7B.3), CE23289 (Y113G7B.6), and CE23288 (Y113G7B.7) are closely related and tightly linked on Chromosome 5. CE35646 was not included in later analysis because of a divergent N-terminal structure.

An FTR family also appears to be present and expanded in the obligate male/female species C. remanei based on the currently sequence assembly (Genome Sequencing Center, Washington University, St. Louis, Missouri, United States; 16 September 2004, BLASTn and tBLASTn; ftp://genome.wustl.edu/pub/seqmgr/remanei/plasmid_assembly). Our preliminary analysis suggests that closest FOG-2 homologs from C. remanei have diverged from C. elegans approximately to the same level as the FTR genes in C. briggsae. A comprehensive phylogenetic analysis to resolve the relationships between C. elegans, C. briggsae, and C. remanei FTR family members will await accurate C. remanei protein predictions and a complete C. remanei assembly.

Sequence alignments and analysis

Alignments were generated using CLUSTALW, and conserved residues were identified with the Lasergene MEGALIGN (DNASTAR, Madison, Wisconsin, United States) package and Dialign [71,72], which was also used to identify conserved regions for subsequent phylogenetic analysis. The best BLAST C. briggsae hit to each C. elegans FTR protein used in the phylogeny was included in order to identify any potential one-to-one orthologous pairs along the FOG-2 branch. Non-homologous N- and C-terminal extensions were trimmed, and extremely distant family members unlikely to be functional FOG-2 orthologs were excluded to avoid long branch attraction [47]. Phylogenetic inference was performed using the neighbor-joining (neighbor) program in the PHYLIP package (Phylogeny Inference Package version 3.5c; Department of Genetics, University of Washington, Seattle, Washington, United States) using the BLOSUM45 distance matrix. Trees with and without gaps were generated, and comparison revealed some differences in branching order, but only within the species. For the tree presented here, positions with gaps were excluded and all non-homologous or highly divergent sequences trimmed. The topology of the tree structure was tested by bootstrapping with 1,000 replicates and by analysis of the alignment using protpars from the PHYLIP package (a maximum parsimony method), which produced a tree with a similar branching order. Trees were processed using TreeView [73].

Codon-restricted alignments for K_a/K_s calculation were generated using Se-Al (a sequence alignment editor by A. Rambaut, version 2; available at http://evolve.zoo.ox.ac.uk/software.html?id=seal) to modify CLUSTALW-aligned cDNA or predicted cDNA sequences, and all gaps and frame-shifted regions were removed. Sliding-window K_a and K_s estimates [74] were generated using DNASP (version 3) [52], and codon-based analysis was performed using PAML (codeml) [53] (HKY substitution model) to confirm the presence of codons under positive selection (95% confidence) within the sliding windows.

Worm culture and RNAi

C. elegans (N2, Bristol, United Kingdom) and C. briggsae (AF16) were obtained from the Caenorhabditis Genetics Center University of Minnesota, Minneapolis, Minnesota, United States. Cultures of both were maintained on Escherichia coli OP50 on NGM plates at 20 °C as previously described [75]. RNAi was performed by injection in C. elegans and C. briggsae essentially as described previously [54]. Double-stranded RNAs for species-specific gld-1 and fog-3 were generated by PCR amplification of cDNA with SP6 (5′) and T7 (3′) linkers, gel purified, sequenced, and used in RNA synthesis reaction using the appropriate Ambion kit (MEGAscript SP6 or T7; Austin, Texas, United States). Double-stranded RNAs were injected at 0.5 mg/ml into young adult N2 animals and F1 progeny collected 12–48 h post injection and matured to 24 h post L4 stage before gonads were dissected, fixed, and stained to score for abnormal phenotypes.

Staining

Dissection, antibody, and DAPI staining of C. elegans and C. briggsae gonads were performed essentially as previously described with fixation in 3% formaldehyde, 80% methanol, and 100 mM dibasic potassium phosphate [29,30]. Affinity purified rabbit polyclonal anti-GLD-1 antibodies were used at 1:50, and MSP mouse monoclonal antibody was used at 1:2,000, both with overnight incubation at room temperature (anti-MSP antibody was the kind gift of M. Kosinski and D. Greenstein, Vanderbilt University School of Medicine, Nashville, Tennessee, United States). Texas Red or Alexa488 secondary antibodies were used to detect staining, and DAPI was used visualize DNA morphology. Epifluorescent images were captured with a Zeiss (Oberkochen, Germany) Axioskop coupled to a Hamamatsu Photonics (Hamamatsu City, Japan) digital CCD camera, and processed with Photoshop 7.0 (Adobe, San Jose, California, United States). All image post-processing (brightness, contrast, pseudo-color, unsharp mask) was performed identically for each image.

Constructs and transformation

GLD-1 and FOG-2 yeast two-hybrid binding assays were performed as previously described [25] with the inclusion of 20 mM 3-amino-triazole. Progressive C-terminal deletions in FOG-2 and FTR-1/FOG-2 chimeric constructs were generated using PCR amplification of the appropriate coding sequences (FOG-2 full-length [327 aa], 318 aa, 299 aa, 263 aa, or exon 4 [251aa], or FTR-1 full-length [318 aa]) and cloned by recombination in yeast. In each case GLD-1 was used as bait in the pAS1 vector (DNA binding) and FOG-2 deletion constructs in the pACTII vector (activation). FOG-2 was found to exhibit low levels of auto-activation in the pAS1 (DNA binding) vector, so binding assays were performed in only one direction to avoid background and using high levels of 3-amino-triazole. The constructs were sequenced, and the Skp1-related F-box-binding protein SKR-1 (in pAS1) was used as a positive control for interaction [76,77].

Supporting Information

Figure S1: Phylogenetic Relationships of 30 C. elegans and C. briggsae FTR Genes Closely Related to FOG-2 Presented as a Rectangular Phylogram

A clear separation of C. elegans and C. briggsae FTR genes (C. briggsae is in grey shade) is suggested by the phylogeny. The branch containing FOG-2 and FTR-1 is in bold. Tree is unrooted, and branch lengths are proportional to divergence. Bar represents 0.1 substitutions per site. Bootstrap support for separation of C. elegans and C. briggsae sequences is indicated at the node (black dot) and at each node for the C. elegans FOG-2 branch.

(34.1 MB TIF).

Click here for additional data file.^{(33M, tif)}

Figure S2: Alignments of FTR-1 and FOG-2 C-Terminal Regions to Other Closely related C. elegans FTR Family Members

(A) FTR-1 and FTR family alignment. Residues identical to FTR-1 are shaded black, and residues identical between all FTR family members tested are shaded red. Average pairwise identity to FTR-1 is 48%.

(B) FOG-2 and FTR family alignment. Residues identical to FOG-2 are shaded black, and residues identical between all FTR family members tested are shaded red. Average pairwise identity to FOG-2 is 22%.

(15.6 MB TIF).

Click here for additional data file.^{(15M, tif)}

Table S1: Analysis of Genes in the fog-2 Cluster

(59 KB PDF).

Click here for additional data file.^{(59K, pdf)}

Table S2: Analysis of Genes Surrounding Y113G7B.11 in C. briggsae

(59 KB PDF).

Click here for additional data file.^{(59K, pdf)}

Acknowledgments

This work was supported by National Institutes of Health (NIH) grant GM63310 to TS and NIH National Research Service Award GM20864 to SN. JG was supported in part by Howard Hughes Medical Institute grant 52003842 through the Undergraduate Biological Sciences Education Program to Washington University. Some nematode strains used in this work were provided by the Caenorhabditis Genetics Center, which is funded by the NIH National Center for Research Resources. We would like to thank Mary E. Kosinski and David Greenstein for the anti-MSP antibody. We would like to thank Justin Fay for important suggestions and assistance with the work on positive selection. We would like to thank Eric Haag, Ronald Ellis, and members of the Schedl lab for helpful discussions and Dave Hansen, Jim Skeath, Sean Eddy, and Susan Dutcher and the three anonymous referees for comments on the manuscript. Finally, we would like to thank the Consortium at Washington University, St. Louis, and at the Sanger Institute for the high-quality genome sequence of C. elegans and C. briggsae that made this project possible.

Competing interests. The authors have declared that no competing interests exist.

Abbreviations


DAPI	4′,6′-diamidino-2-phenylindole hydrochloride

FTH	FOG-2 homology domain

FTR	fog-2 related (F-box and FTH)

GFP	green fluorescent protein

HMM	hidden Markov model

K_a	non-synonymous substitutions

K_s	synonymous substitutions

L[number]	[number] larval

MSP	major sperm protein

RNAi	double-stranded-RNA-mediated interference

TGE	tra-2 and GLI element

UTR	untranslated region

Footnotes

Author contributions. SN and TS conceived and designed the experiments. SN and JG performed the experiments. SN and TS analyzed the data. SN and TS wrote the paper.

Citation: Nayak S, Goree J, Schedl T (2004) fog-2 and the evolution of self-fertile hermaphroditism in Caenorhabditis. PLoS Biol 3(1): e6.

References

Cline TW, Meyer BJ. Vive la difference: Males vs females in flies vs worms. Annu Rev Genet. 1996;30:637–702. [PubMed]
Marin I, Baker BS. The evolutionary dynamics of sex determination. Science. 1998;281:1990–1994. [PubMed]
Zarkower D. Invertebrates may not be so different after all. Novartis Found Symp. 2002;244:115–126. [PubMed]
C. elegans Sequencing Consortium. Genome sequence of the nematode C. elegans A platform for investigating biology. The C. elegans Sequencing Consortium. Science. 1998;282:2012–2018. [PubMed]
Stein LD, Bao Z, Blasiar D, Blumenthal T, Brent MR, et al. The genome sequence of Caenorhabditis briggsae A platform for comparative genomics. PLoS Biol. 2003;1:e45. [PubMed]
Fitch DH, Bugaj-Gaweda B, Emmons SW. 18S ribosomal RNA gene phylogeny for some Rhabditidae related to Caenorhabditis. Mol Biol Evol. 1995;12:346–358. [PubMed]
Fitch DH. Evolution of male tail development in rhabditid nematodes related to Caenorhabditis elegans. Syst Biol. 1997;46:145–179. [PubMed]
Baird SE. Haldane's rule by sexual transformation in Caenorhabditis. Genetics. 2002;161:1349–1353. [PubMed]
Coghlan A, Wolfe KH. Fourfold faster rate of genome rearrangement in nematodes than in Drosophila. Genome Res. 2002;12:857–867. [PubMed]
Kiontke K, Gavin NP, Raynes Y, Roehrig C, Piano F, et al. Caenorhabditis phylogeny predicts convergence of hermaphroditism and extensive intron loss. Proc Natl Acad Sci U S A. 2004;101:9003–9008. [PubMed]
Cho S, Jin SW, Cohen A, Ellis RE. A phylogeny of Caenorhabditis reveals frequent loss of introns during nematode evolution. Genome Res. 2004;14:1207–1220. [PubMed]
Madl JE, Herman RK. Polyploids and sex determination in Caenorhabditis elegans. Genetics. 1979;93:393–402. [PubMed]
Hodgkin J. Two types of sex determination in a nematode. Nature. 1983;304:267–268. [PubMed]
Hirsh D, Oppenheim D, Klass M. Development of the reproductive system of Caenorhabditis elegans. Dev Biol. 1976;49:200–219. [PubMed]
Kimble J. Genetic control of sex determination in the germ line of Caenorhabditis elegans. Philos Trans R Soc Lond B Biol Sci. 1988;322:11–18. [PubMed]
Klass M, Wolf N, Hirsh D. Development of the male reproductive system and sexual transformation in the nematode Caenorhabditis elegans. Dev Biol. 1976;52:1–18. [PubMed]
de Bono M, Hodgkin J. Evolution of sex determination in Caenorhabditis unusually high divergence of tra-1 and its functional consequences. Genetics. 1996;144:587–595. [PubMed]
Puoti A, Pugnale P, Belfiore M, Schlappi AC, Saudan Z. RNA and sex determination in Caenorhabditis elegans. Post-transcriptional regulation of the sex-determining tra-2 and fem-3 mRNAs in the Caenorhabditis elegans hermaphrodite. EMBO Rep. 2001;2:899–904. [PubMed]
Kuwabara PE, Perry MD. It ain't over till it's ova: Germline sex determination in C. elegans. Bioessays. 2001;23:596–604. [PubMed]
Doniach T. Activity of the sex-determining gene tra-2 is modulated to allow spermatogenesis in the C. elegans hermaphrodite. Genetics. 1986;114:53–76. [PubMed]
Kuwabara PE. Interspecies comparison reveals evolution of control regions in the nematode sex-determining gene tra-2. Genetics. 1996;144:597–607. [PubMed]
Hodgkin J. More sex-determination mutants of Caenorhabditis elegans. Genetics. 1980;96:649–664. [PubMed]
Perry MD, Trent C, Robertson B, Chamblin C, Wood WB. Sequenced alleles of the Caenorhabditis elegans sex-determining gene her-1 include a novel class of conditional promoter mutations. Genetics. 1994;138:317–327. [PubMed]
Jan E, Motzny CK, Graves LE, Goodwin EB. The STAR protein, GLD-1, is a translational regulator of sexual identity in Caenorhabditis elegans. EMBO J. 1999;18:258–269. [PubMed]
Clifford R, Lee MH, Nayak S, Ohmachi M, Giorgini F, et al. FOG-2, a novel F-box containing protein, associates with the GLD-1 RNA binding protein and directs male sex determination in the C. elegans hermaphrodite germline. Development. 2000;127:5265–5276. [PubMed]
Stothard P, Pilgrim D. Sex-determination gene and pathway evolution in nematodes. Bioessays. 2003;25:221–231. [PubMed]
Schedl T, Kimble J. fog-2, a germ-line-specific sex determination gene required for hermaphrodite spermatogenesis in Caenorhabditis elegans. Genetics. 1988;119:43–61. [PubMed]
Francis R, Barton MK, Kimble J, Schedl T. gld-1, a tumor suppressor gene required for oocyte development in Caenorhabditis elegans. Genetics. 1995;139:579–606. [PubMed]
Francis R, Maine E, Schedl T. Analysis of the multiple roles of gld-1 in germline development: Interactions with the sex determination cascade and the glp-1 signaling pathway. Genetics. 1995;139:607–630. [PubMed]
Jones AR, Francis R, Schedl T. GLD-1, a cytoplasmic protein essential for oocyte differentiation, shows stage- and sex-specific expression during Caenorhabditis elegans germline development. Dev Biol. 1996;180:165–183. [PubMed]
Lee MH, Schedl T. Identification of in vivo mRNA targets of GLD-1, a maxi-KH motif containing protein required for C. elegans germ cell development. Genes Dev. 2001;15:2408–2420. [PubMed]
Ryder SP, Frater LA, Abramovitz DL, Goodwin EB, Williamson JR. RNA target specificity of the STAR/GSG domain post-transcriptional regulatory protein GLD-1. Nat Struct Mol Biol. 2004;11:20–28. [PubMed]
Okkema PG, Kimble J. Molecular analysis of tra-2, a sex determining gene in C.elegans. EMBO J. 1991;10:171–176. [PubMed]
Jan E, Yoon JW, Walterhouse D, Iannaccone P, Goodwin EB. Conservation of the C.elegans tra-2 3′UTR translational control. EMBO J. 1997;16:6301–6313. [PubMed]
Bateman A, Birney E, Durbin R, Eddy SR, Howe KL, et al. The Pfam protein families database. Nucleic Acids Res. 2000;28:263–266. [PubMed]
Kipreos ET, Pagano M. The F-box protein family. Genome Biol. 2000;1:REVIEWS3002. [PubMed]
Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, et al. Gapped BLAST and PSI-BLAST: A new generation of protein database search programs. Nucleic Acids Res. 1997;25:3389–3402. [PubMed]
Tatusov RL, Koonin EV, Lipman DJ. A genomic perspective on protein families. Science. 1997;278:631–637. [PubMed]
Hirsh AE, Fraser HB. Protein dispensability and rate of evolution. Nature. 2001;411:1046–1049. [PubMed]
Hansen D, Pilgrim D. Molecular evolution of a sex determination protein. FEM-2 (pp2c) in Caenorhabditis. Genetics. 1998;149:1353–1362. [PubMed]
Haag ES, Wang S, Kimble J. Rapid coevolution of the nematode sex-determining genes fem-3 and tra-2. Curr Biol. 2002;12:2035–2041. [PubMed]
Chen PJ, Cho S, Jin SW, Ellis RE. Specification of germ cell fates by FOG-3 has been conserved during nematode evolution. Genetics. 2001;158:1513–1525. [PubMed]
Streit A, Li W, Robertson B, Schein J, Kamal IH, et al. Homologs of the Caenorhabditis elegans masculinizing gene her-1 in C. briggsae and the filarial parasite Brugia malayi. Genetics. 1999;152:1573–1584. [PubMed]
Kuwabara PE, Shah S. Cloning by synteny: Identifying C. briggsae homologues of C. elegans genes. Nucleic Acids Res. 1994;22:4414–4418. [PubMed]
Stothard P, Hansen D, Pilgrim D. Evolution of the PP2C family in Caenorhabditis Rapid divergence of the sex-determining protein FEM-2. J Mol Evol. 2002;54:267–282. [PubMed]
Wall DP, Fraser HB, Hirsh AE. Detecting putative orthologs. Bioinformatics. 2003;19:1710–1711. [PubMed]
Felsenstein J. Cases in which parsimony or compatibility methods will be positively misleading. Syst Zool. 1978;27:401–410.
Otto SP, Yong P. The evolution of gene duplicates. Adv Genet. 2002;46:451–483. [PubMed]
Fares MA, Elena SF, Ortiz J, Moya A, Barrio E. A sliding window-based method to detect selective constraints in protein-coding genes and its application to RNA viruses. J Mol Evol. 2002;55:509–521. [PubMed]
Bielawski JP, Yang Z. Maximum likelihood methods for detecting adaptive evolution after gene duplication. J Struct Funct Genomics. 2003;3:201–212. [PubMed]
Hurst LD. The Ka/Ks ratio: Diagnosing the form of sequence evolution. Trends Genet. 2002;18:486. [PubMed]
Rozas J, Rozas R. DnaSP version 3: An integrated program for molecular population genetics and molecular evolution analysis. Bioinformatics. 1999;15:174–175. [PubMed]
Yang Z. PAML: A program package for phylogenetic analysis by maximum likelihood. Comput Appl Biosci. 1997;13:555–556. [PubMed]
Fire A, Xu S, Montgomery MK, Kostas SA, Driver SE, et al. Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature. 1998;391:806–811. [PubMed]
Lee MH, Schedl T. Translation repression by GLD-1 protects its mRNA targets from nonsense-mediated mRNA decay in C. elegans. Genes Dev. 2004;18:1047–1059. [PubMed]
Ajiro K, Yoda K, Utsumi K, Nishikawa Y. Alteration of cell cycle-dependent histone phosphorylations by okadaic acid. Induction of mitosis-specific H3 phosphorylation and chromatin condensation in mammalian interphase cells. J Biol Chem. 1996;271:13197–13201. [PubMed]
Grant B, Hirsh D. Receptor-mediated endocytosis in the Caenorhabditis elegans oocyte. Mol Biol Cell. 1999;10:4311–4326. [PubMed]
Mootz D, Ho DM, Hunter CP. The STAR/Maxi-KH domain protein GLD-1 mediates a developmental switch in the translational control of C. elegans PAL-1. Development. 2004;131:3263–3272. [PubMed]
Xu L, Paulsen J, Yoo Y, Goodwin EB, Strome S. Caenorhabditis elegans MES-3 is a target of GLD-1 and functions epigenetically in germline development. Genetics. 2001;159:1007–1017. [PubMed]
Haag ES, Kimble J. Regulatory elements required for development of Caenorhabditis elegans hermaphrodites are conserved in the tra-2 homologue of C. remanei a male/female sister species. Genetics. 2000;155:105–116. [PubMed]
Ward S, Carrel JS. Fertilization and sperm competition in the nematode Caenorhabditis elegans. Dev Biol. 1979;73:304–321. [PubMed]
McCarter J, Bartlett B, Dang T, Schedl T. On the control of oocyte meiotic maturation and ovulation in Caenorhabditis elegans. Dev Biol. 1999;205:111–128. [PubMed]
Miller MA, Nguyen VQ, Lee MH, Kosinski M, Schedl T, et al. A sperm cytoskeletal protein that signals oocyte meiotic maturation and ovulation. Science. 2001;291:2144–2147. [PubMed]
Hill KL, L'Hernault SW. Analyses of reproductive interactions that occur after heterospecific matings within the genus Caenorhabditis. Dev Biol. 2001;232:105–114. [PubMed]
Hodgkin J. Exploring the envelope. Systematic alteration in the sex-determination system of the nematode Caenorhabditis elegans. Genetics. 2002;162:767–780. [PubMed]
Goodwin EB, Hofstra K, Hurney CA, Mango S, Kimble J. A genetic pathway for regulation of tra-2 translation. Development. 1997;124:749–758. [PubMed]
Maine E, Hansen D, Springer D, Vought VE. Caenorhabditis elegans atx-2 promotes germline proliferation and oocyte fate. Genetics. 2004;168:817–830. [PubMed]
Barton MK, Schedl TB, Kimble J. Gain-of-function mutations of fem-3, a sex-determination gene in Caenorhabditis elegans. Genetics. 1987;115:107–119. [PubMed]
Barton MK, Kimble J. fog-1, a regulatory gene required for specification of spermatogenesis in the germ line of Caenorhabditis elegans. Genetics. 1990;125:29–39. [PubMed]
Hodgkin J, Barnes TM. More is not better: Brood size and population growth in a self-fertilizing nematode. Proc R Soc Lond B Biol Sci. 1991;246:19–24.
Morgenstern B. DIALIGN 2: Improvement of the segment-to-segment approach to multiple sequence alignment. Bioinformatics. 1999;15:211–218. [PubMed]
Lassmann T, Sonnhammer EL. Quality assessment of multiple alignment programs. FEBS Lett. 2002;529:126–130. [PubMed]
Page RD. TreeView: An application to display phylogenetic trees on personal computers. Comput Appl Biosci. 1996;12:357–358. [PubMed]
Nei M, Gojobori T. Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions. Mol Biol Evol. 1986;3:418–426. [PubMed]
Brenner S. The genetics of Caenorhabditis elegans. Genetics. 1974;77:71–94. [PubMed]
Nayak S, Santiago FE, Jin H, Lin D, Schedl T, et al. The Caenorhabditis elegans Skp1-related gene family: Diverse functions in cell proliferation, morphogenesis, and meiosis. Curr Biol. 2002;12:277–287. [PubMed]
Yamanaka A, Yada M, Imaki H, Koga M, Ohshima Y, et al. Multiple Skp1-related proteins in Caenorhabditis elegans Diverse patterns of interaction with Cullins and F-box proteins. Curr Biol. 2002;12:267–275. [PubMed]
Schedl T, Graham PL, Barton MK, Kimble J. Analysis of the role of tra-1 in germline sex determination in the nematode Caenorhabditis elegans. Genetics. 1989;123:755–769. [PubMed]

Formats:

PubMed articles by these authors

PubMed related articles

Recent Activity

Links

Simple NCBI Directory

Getting Started

Resources

Popular

Featured

NCBI Information