Robust sex determination in the Caenorhabditis nigoni germ line

Abstract

Sexual characteristics and reproductive systems are dynamic traits in many taxa, but the developmental modifications that allow change and innovation are largely unknown. A leading model for this process is the evolution of self-fertile hermaphrodites from male/female ancestors. However, these studies require direct analysis of sex determination in male/female species, as well as in the hermaphroditic species that are related to them. In Caenorhabditis nematodes, this has only become possible recently, with the discovery of new species. Here, we use gene editing to characterize major sex determination genes in Caenorhabditis nigoni, a sister to the widely studied hermaphroditic species Caenorhabditis briggsae. These 2 species are close enough to mate and form partially fertile hybrids. First, we find that tra-1 functions as the master regulator of sex in C. nigoni, in both the soma and the germ line. Surprisingly, these mutants make only sperm, in contrast to tra-1 mutants in related hermaphroditic species. Moreover, the XX mutants display a unique defect in somatic gonad development that is not seen elsewhere in the genus. Second, the fem-3 gene acts upstream of tra-1 in C. nigoni, and the mutants are females, unlike in the sister species C. briggsae, where they develop as hermaphrodites. This result points to a divergence in the role of fem-3 in the germ line of these species. Third, tra-2 encodes a transmembrane receptor that acts upstream of fem-3 in C. nigoni. Outside of the germ line, tra-2 mutations in all species cause a similar pattern of partial masculinization. However, heterozygosity for tra-2 does not alter germ cell fates in C. nigoni, as it can in sensitized backgrounds of 2 hermaphroditic species of Caenorhabditis. Finally, the epistatic relationships point to a simple, linear germline pathway in which tra-2 regulates fem-3 which regulates tra-1, unlike the more complex relationships seen in hermaphrodite germ cell development. Taking these results together, the regulation of sex determination is more robust and streamlined in the male/female species C. nigoni than in related species that make self-fertile hermaphrodites, a conclusion supported by studies of interspecies hybrids using sex determination mutations. Thus, we infer that the origin of self-fertility not only required mutations that activated the spermatogenesis program in XX germ lines, but prior to these there must have been mutations that decanalized the sex determination process, allowing for subsequent changes to germ cell fates.

Evolution can dramatically change a species. To learn how these changes occur, Harbin et al. study how roundworms became self-fertile. Because this transformation has happened several times, the authors thought that related male/female species would have flexible sexual development. Thus, they used gene editing to study sex in the male/female nematode C. nigoni, a close relative of a self-fertile species. Surprisingly, it shows robust control of all sexual development. Thus, a key step in evolution must be mutations that allow flexibility in development.

Introduction

Evolution of mating systems

Reproduction plays a central role in evolution (Schwander et al. 2014). Hence, changes in mating systems not only involve numerous sexual traits, but also shape the future course of evolution. For example, many plant (Pannell 2015) and animal species (Weeks 2012) have undergone transitions from obligate female/male systems to ones involving self-fertile hermaphrodites and males. Such changes in mating systems influence many other traits, such as sex ratios (Van Goor et al. 2021), outcrossing (Cutter et al. 2019), dispersal (Trochet et al. 2016), sperm storage (Orr and Brennan 2015), and genome size (Fierst et al. 2015; Yin et al. 2018).

The importance of these evolutionary transitions has led to a focus on the genetic mechanisms that underlie them in certain groups, including plants of the Brassicaceae (Vekemans et al. 2014) and Caenorhabditis nematodes (Ellis and Lin 2014; Ellis 2017). In Caenorhabditis, 3 species independently evolved self-fertility (Fig. 1a), providing rich opportunities for study and comparison (Kiontke et al. 2011).

Fig. 1.

Evolution of mating systems in Caenorhabditis. a) Phylogeny of the elegans group of Caenorhabditis, based on work by Kiontke et al. (2011). Gray font indicates male/female species. b) Schematic diagrams of the sexes discussed in this paper. Males make sperm that remain inactive until mating (small blue cells). Virgin females only make oocytes, which they retain until mating (large pink cells). Hermaphrodites resemble females but produce sperm in larval development, which they activate and store for use in fertilizing oocytes they make as adults (embryos in the uterus are white). c) Outline of gene regulation in the C. briggsae germ line. CAPITALS indicate protein names and lowercase italics gene names. Items in blue generally promote spermatogenesis and those in red oogenesis. Black lines denote interactions that favor L4 spermatogenesis, and gray lines denote ones that favor oogenesis. Arrows indicate positive regulation and “—|” negative regulation. For details, see text. d) Outline of gene regulation in C. elegans. e) DIC photomicrographs of a C. nigoni JU1422 female (left) and male (right). Visible portions of the posterior arm of the gonad are outlined with dotted white lines in the female and of the whole gonad in the male. Two oocytes are outlined in yellow, and one primary spermatocyte in blue. A yellow arrow marks the vulva, green arrows mark residual bodies, and blue arrows mark sperm. White dots indicate the 14 out of 18 rays visible in this plane of focus, and a yellow oval marks the spicules.

Self-fertile hermaphrodites

Caenorhabditis hermaphrodites are essentially XX females that have been altered to produce, store, and use sperm (Fig. 1b, Ellis and Lin 2014; Ellis 2017). Since they lack other male somatic traits, they can only self-fertilize or mate with males. This genetic system is particularly easy to work with in the laboratory (Brenner 1974), leading to the wide adoption of Caenorhabditis elegans as a model for studying developmental biology and neurobiology.

Although there are numerous differences between XX females and XX hermaphrodites in this genus, such as genome size and attraction to mates, the initial transformation to self-fertility might have involved as few as 2 genes, one controlling the sex determination of germ cells and the other controlling sperm activation (Baldi et al. 2009). The ability of hermaphrodites to activate their own sperm, allowing self-fertilization, was shown to require the co-option of one of the redundant activation signals used by males (Wei et al. 2014b). The other change appears more complex, involving alterations to the regulatory pathway that controls sex determination in the germ line (Ellis 2022).

Nematode sex determination

In Caenorhabditis nematodes, sex is determined by a signal transduction pathway that responds to the ratio of X chromosomes to autosomes (Meyer 2022). In C. elegans, this pathway acts through HER-1, a secreted protein that promotes male fates in both the germ line and soma (Hodgkin 1980; Hunter and Wood 1992; Perry et al. 1993), with the ultimate target being the Gli transcription factor TRA-1 (Zarkower and Hodgkin 1992). In XO animals, HER-1 binds to and inactivates the TRA-2 receptor, preventing its cleavage by TRA-3 (reviewed by Zarkower 2006). This allows a complex of 3 FEM proteins to work with a cullin to ubiquitinate TRA-1, targeting it for destruction (Starostina et al. 2007). As a result, TRA-1 repressor is not made, and male genes can be expressed, leading to male development. By contrast, in XX animals TRA-1 is not degraded, but is instead cleaved (Schvarzstein and Spence 2006) to form a repressor of male genes throughout the body (Berkseth et al. 2013) and germ line (Chen and Ellis 2000).

Modifications to genetic pathways in the hermaphrodite germ line

Regulation in the germ line is more complex, probably because XX animals need to produce both sperm and oocytes in a female body (Fig. 1c and d, reviewed by Ellis 2022). Although the core pathway is the same as in the soma, several additional genes are involved, including novel ones like fog-2 in C. elegans (Schedl and Kimble 1988; Clifford et al. 2000; Nayak et al. 2005) and she-1 in Caenorhabditis briggsae (Guo et al. 2009), which were each formed by gene duplication and divergence.

Perhaps because the regulatory pathway has been adapted for self-fertility, mutations in some of the core sex determination genes produce complex or intersexual phenotypes.

tra-2

Two aspects of tra-2 mutants are particularly striking. First, the decision of germ cells to form sperm or oocytes in C. elegans and C. briggsae responds very strongly to small changes in tra-2 activity. For example, gain-of-function mutations in the C. elegans tra-2 3′-UTR cause all XX germ cells to develop as oocytes (Doniach 1986; Goodwin et al. 1993). Furthermore, mutations that prevent TRA-2 from binding TRA-1 cause XX animals to make only oocytes in C. elegans (Doniach 1986; Lum et al. 2000; Wang and Kimble 2001) but extra sperm in C. briggsae (Shen et al. 2024a). Moreover, mutations in fog-2 that create a female/male population of C. elegans are often suppressed by haploinsufficiency for tra-2, and even by weak, cryptic tra-2 mutations (Schedl and Kimble 1988; Hu et al. 2019). Similarly, she-1 mutations that create a female/male population of C. briggsae are strongly suppressed by haploinsufficiency for tra-2 (Guo et al. 2009).

Second, the somatic effects of tra-2 are complex. Most of the soma and all of the germ cells undergo male development in C. elegans and C. briggsae tra-2 null mutants (Hodgkin and Brenner 1977; Kelleher et al. 2008). However, in both species these XX mutants make defective male tails and are not capable of mating with hermaphrodites.

fem-3

In C. elegans, fem-3 acts upstream of tra-1 in somatic development but is also required downstream of tra-1 to control germ cell fates (Hodgkin 1986; Chen and Ellis 2000). By contrast, C. briggsae fem-3 is not required for spermatogenesis, and both XX and XO fem-3 mutants are hermaphrodites (Hill et al. 2006). However, loss of fem-3 function is nonetheless partially epistatic to tra-2 mutations in the germ line (Hill et al. 2006) and increases the likelihood that tra-1 mutants make oocytes (Hill and Haag 2009).

tra-1

In both C. elegans and C. briggsae, tra-1 mutations often cause animals to produce sperm and then switch to oogenesis (Hodgkin 1987; Schedl et al. 1989; Kelleher et al. 2008). This is true for both XX animals (which normally make sperm and oocytes) and XO males (which normally make only sperm). Thus, tra-1 appears necessary to sustain spermatogenesis, even though its main role is to prevent male gene expression.

These traits might reflect an ancient propensity for the pathway to switch from one sexual state to the other, which could have been a precondition for the evolution of self-fertility in this genus. On the other hand, the ancestral pathway might have been robust, and these unusual phenotypes might have been caused by the modifications that were needed to produce self-fertility.

C. nigoni is the ideal species for studying male/female nematodes

Addressing these questions about how sex determination worked in the male/female ancestor of Caenorhabditis requires studying the pathway in a male/female species from the genus. The ideal place to begin is with C. nigoni, an XX female/XO male species (Kiontke et al. 2011). These animals are so closely related to C. briggsae that they can mate with them and produce fertile hybrids (Woodruff et al. 2010). Thus, studying C. nigoni should provide a window into what the sex determination pathway might have looked like before C. briggsae began the path to self-fertility. Finally, the XX hybrids develop as females and so provide a method for testing theories about the control of sex determination in each parent species.

Materials and methods

Strains

We used C. nigoni inbred strains JU1422 (Félix et al. 2014) and CP168 (below), as well as: LGII: tra-2(v498), unc-104(v494), dpy-10(v541); LGIII: tra-1(v481), tra-1(v538), dpy-18(v484); LGIV: fem-3(v496), fem-3(v542), fem-3(v395gf), unc-129(v495), unc-33(v543). All mutations were first described here except for dpy-18(v484) (Harbin and Ellis 2023).

For C. briggsae, we used the wild-type strain AF16 (Fodor et al. 1983) and the mutations tra-1(v181) III (Shen et al. 2024b), tra-2(v440) IV (Shen et al. 2024), tra-2(v417ts) IV (this paper), fem-3(nm63) IV (Hill et al. 2006), and unc-7(v271) X (Wei et al. 2014a).

Genetics

Unless stated otherwise, worms were raised at 25°C.

Microscopy

Differential interference contrast (DIC) microscopy was performed by standard methods (Wood 1988). Images were captured with a Zeiss Axiocam digital camera and AxioVision 4.8 software.

Single-worm PCR and RT-PCR analyses

For genotyping, individual worms were picked into 2.5 µl worm lysis buffer (Fay and Bender 2008), frozen at −80°C, and incubated at 65°C for 1 h to release the DNA. When multiple reactions had to be run from the same template, additional lysis buffer was used to harvest the animal, and the sample split into fractions after incubation. Finally, PCR reactions were run in standard conditions using hot-start Taq polymerase (New England Biolabs). Individual worms were analyzed by RT-PCR as described by Ly et al. (2015). Primers are listed in Table 1.

Table 1.

Primers used in these experiments.

Use	Primer name	Sequence
Genotyping Cbr-tra-2(v440)	JH9-Cbr-Tra2-E-F	AAAGAAGCCAGAAGCAGACC
	JH10-Cbr-Tra2-E-R	GTGTTCCTTCAAGAAATGGAGC
Genotyping Cni-tra-2	JH65-Cni-tra-2-F	TGGAGCAATGAAGTTGGCAT
	JH66-Cni-tra-2-R	AATAGAGAAGCCAGAAGCAGACC
Genotyping Cni-dpy-10	Cni-dpy-10-F	AACTATTCGCGTCAGATGACGTA
	Cni-dpy-10-R	TTGAGTGGACAGGTCTGATTTGG
Genotyping Cni-unc-104	JH71-Cni-unc-104-F	AAAACCCAGGGATGGACTCG
	JH72-Cni-unc-104-R	ACGTATTCGGAACCTTGACCA
Genotyping Cni-tra-1(v481)	JH26-Cni-tra1-F	CTATCTGGCACCACCACAT
	JH27-Cni-tra1-R	GCATCACACAGAGGGAGTAT
Genotyping Cni-tra-1(v538)	JH116-Cni-tra-1.1 F	GCCTTGTTCCCAATAGCTTG
	JH117-Cni-tra-1.1 R	TTGATGGCTCAGCTGTTGAA
Genotyping Cni-dpy-18	JH34-Cni-dpy18-F	TAGCACTTCTCGTGCTGGC
	JH35-Cni-dpy18-R	GTTGCTGCATATCCGCAATCG
Genotyping Cni-fem-3	JH75-Cni-fem-3-F	TACCGGATGATGTGGAACCC
	JH76-Cni-fem-3-R	TCCAGGCCATTTATTCCCCC
Genotyping Cni-unc-129	JH78-Cni-unc-129-F	GACGACTCCCAATCGTCTCG
	JH79-Cni-unc-129-R	ACTCACTTGGATCCGAACTCT
Genotyping Cni-unc-33	JH128-Cni-unc-33 F	GATGCGTCTGGAAAACTTCT
	JH129-Cni-unc-33 R	GCATCCAGTTGAAATATCGTC
Identifying JU1422 vs CP168 X chromosomes	JH112-Cni-SNPX1 F	AAGATCAACTTCCTGCAAGC
	JH113-Cni-SNPX1 R	TCAGAGCTCATACATTCCCA
Identifying Cni vs Cbr DNA	JH114-Cni-SNPX2 F	TGATTTCGTGTGGTGCTCTA
	JH115-Cni-SNPX2 R	ACGGGTTTTTCCAAATCACA
Cni-her-1 RT-PCR	JH88-Her-1.2 F	CGTTACGAATGTTGTATGGATATGAT
	JH89-Her-1.2 R	AATTGCATGCAACAGGAACT

Production of CP168

The C. nigoni wild isolate EG5268 was originally isolated by Joel Ehrenkranz, from human-disturbed soil in 2008 in Shinkolobwe, Katanga Province, Democratic Republic of the Congo, and was obtained from the Caenorhabditis Genetics Center at the University of Minnesota (https://cgc.umn.edu/strain/EG5268). It was inbred as follows: 3 individual female adults that had already mated (as judged by the presence of a copulatory plug and embryos) were used to establish 3 isofemale lines on 2.2% agar NGM plates. The number of sires of was unknown. After 2 to 3 generations of growth, a single mated female was again chosen to found the next plate. This process was continued for 10 rounds of single-female bottlenecks. All 3 lines persisted, and 1 was chosen to be CP168.

Yeast two-hybrid analysis

Clones were constructed and tested as described by Shen et al. (2024a) using pGBKT7 as a bait vector and pGADT7 as a prey vector.

Genome analysis

High-molecular-weight genomic DNA was extracted from a mixed-stage culture of CP168 using a conventional proteinase K/phenol–chloroform procedure, with modifications to prevent shearing. These were described in detail by Yin et al. (2018). CP168 genomic DNA was sequenced in paired-end 2 × 150-nt Illumina reads by Novogene USA; separate runs were pooled into a single set of read pairs. Pooled reads were quality-filtered and trimmed with fastp 0.23.1 (Chen 2023) using the arguments “–length_required 75 –max_len1 150 –detect_adapter_for_pe –dedup –merge –correction.” This yielded a filtered set of paired-end, merged, and unpaired reads with a genome coverage of 26.3×. The fastp-filtered and merged reads were assembled with SPAdes 3.15.3 (Prjibelski et al. 2020) using the arguments “–isolate –pe1-1 [pair-end 1 read file] –pe1-2 [pair-end read 2 file] –pe1-m [merged read file] –pe1-s [unpaired read file].” Reference-based chromosomal scaffolding of the SPAdes contigs was done with RagTag 2.1.0 (Alonge et al. 2022) using the arguments “scaffold –r” and using the C. nigoni JU1422 assembly (Yin et al. 2018) as a reference genome. Protein-coding gene annotations were lifted over from the JU1422 genome assembly to the final CP168 genome assembly with LiftOff 1.6.2 (Shumate and Salzberg 2021) using the arguments “-copies -polish –cds.” Protein and DNA sequences were extracted from the lifted-over gene annotations and the final CP168 genome assembly with gffread 0.12.7 (Pertea and Pertea 2020) using the arguments “-W -C -x [CDS DNA FASTA file] -y [translated protein FASTA file] -o/dev/null.”

Results

Isolation of C. nigoni sex determination mutants

To learn how sex determination is regulated in male/female species of nematodes, we identified the C. nigoni orthologs of 3 critical sex determination genes known from the male/hermaphrodite species C. elegans and C. briggsae (Supplementary Figs. 1–3, Ellis 2022) and used gene editing to make null alleles in each (Harbin and Ellis 2023). Because we suspected these mutations could not be maintained as homozygous strains, we also made mutations in nearby marker genes that could be used as genetic balancers (Wei et al. 2014a).

C. nigoni tra-1 promotes female development in both the soma and germ line

TRA-1 is a Gli transcription factor with 5 conserved zinc fingers, and directly controls the expression of hundreds of genes responsible for male development in C. elegans (Fig. 2a, Berkseth et al. 2013). To produce C. nigoni tra-1 null mutations, we made frameshift alleles just before or after the zinc fingers. Similar mutations behave as null alleles in C. elegans and C. briggsae (Hodgkin and Brenner 1977; Hodgkin 1987; Zarkower and Hodgkin 1992; Kelleher et al. 2008).

Fig. 2.

C. nigoni tra-1 mutations transform XX animals into males with underdeveloped gonads. a) Comparison of C. briggsae and C. nigoni TRA-1 proteins, showing regions of high homology as colored domains. The C. nigoni TRA-1 mutant proteins are indicated below, with gray highlighting sequences after the frameshift but before the stop codon. b) Mating strategy to generate homozygous mutants from a balanced strain. Light blue indicates JU1422 sequences. c) DIC photomicrographs of: (top) a tra-1(v481) XX animal displaying typical male soma with tiny gonad and (bottom) dpy-18(v484) homozygote. Anterior is left and ventral up. d) Single-worm RT-PCR analysis to identify karyotype by the level of her-1 expression. e) Interstrain cross to produce tra-1 null mutants in a uniform JU1422/CP168 background. Dark blue marks CP168 sequences. f) Phenotypes of tra-1 mutants in interstrain hybrids. XX animals are males with small gonads (top); the few germ cells develop as sperm (blue arrows, inset). XO animals are normal males (bottom). Anterior is left and ventral up.

We began by studying the C. nigoni strain JU1422, which had been produced from the wild isolate JU1325 by 25 generations of inbreeding (Félix et al. 2014). After injecting gravid females with CRISPR/Cas9/sgRNA solutions (Harbin and Ellis 2023), we identified a tra-1 mutation among the progeny by PCR analysis. Because this mutation could not reproduce as a homozygous strain, we used dpy-18 (Harbin and Ellis 2023) to balance it (Fig. 2b). This allele, cni-tra-1(v481), causes a frame shift and truncation shortly after the zinc fingers (Supplementary Fig. 1).

To analyze homozygous tra-1 mutants, we studied the progeny produced by our balanced strain and then determined which were tra-1 XX or XO animals. First, we used PCR to determine genotype and then single-animal RT-PCR analysis to determine karyotype. Individual XX animals had low her-1 transcript levels, and XO animals had high levels (Figs. 2c and d, Trent et al. 1991; Ly et al. 2015). The tra-1 XX animals developed male bodies with fully developed male tails that included 18 rays and 2 spicules (Fig. 2c). However, their gonads failed to develop (N = 28/28). By contrast, the homozygous tra-1 XO animals developed as normal males with mature gonads and made sperm throughout their lives (N = 17/17 5-to-7-day-old adults, 25°C). These results suggest that C. nigoni TRA-1 is not required to sustain spermatogenesis, unlike TRA-1 in C. elegans (Hodgkin 1987; Schedl et al. 1989) or C. briggsae (Kelleher et al. 2008).

The gonad problems in these tra-1 XX mutants were unexpected. However, male/female species of Caenorhabditis have a high degree of standing genetic variation (e.g. Dey et al. 2013), and the inbreeding of large haplotypes required to make a mutation homozygous in a new strain can produce artificial phenotypes unrelated to the focal mutation itself (Yin et al. 2018). We thus wondered if this phenotype had been influenced by the genetic background of the inbred JU1422 strain. To address this concern, we made another inbred C. nigoni strain with a different genetic origin, CP168, for comparative analysis. We then assembled and annotated a draft genome of CP168 to facilitate the design of guide RNAs and primers.

Finally, we made a new tra-1 null allele in this strain, Cni-tra-1(v538), which causes a frame shift and a truncation before the zinc fingers (Supplementary Fig. 1). Once again, we saw that tra-1 XX mutants had male bodies (N = 32/32), but that their gonads failed to develop normally. However, these XX tra-1 gonads appeared slightly larger than in the JU1422 genetic background (Supplementary Fig. 4). The tra-1 XO animals were normal males and made only sperm (N = 14/14, 6-day-old adults, 25°C).

To confirm our identification of the tra-1 null phenotype, we studied JU1422/CP168 interstrain hybrids that were heterozygous for these 2 null alleles (Fig. 2e). This approach provides a uniform genetic background without the potential complication of homozygous recessive mutations from inbreeding. Furthermore, it allowed us to determine karyotypes with a simple PCR test (e.g. Fig. 3c). In these hybrids, the tra-1 XX animals still developed male bodies with defective gonads (N = 8/8). However, these gonads were slightly larger than in either JU1422 or CP168 and occasionally produced sperm (Fig. 2f). These rare sperm in the mutants show that tra-1 normally promotes oogenesis as well as female somatic fates in C. nigoni XX animals.

Fig. 3.

C. nigoni FEM-3 directs male development in both the germ line and soma. a) Diagram comparing wild-type C. nigoni and C. briggsae FEM-3 proteins. Darker shades of blue indicate higher levels of conservation. The sizes of both C. nigoni mutant proteins are indicated below, with gray highlighting sequences after the frameshift but before the stop codon. b) Cross for producing interstrain hybrid mutants shown below. c) PCR tests to confirm fem-3 genotype (left) and karyotype (right). d) Phenotypes of fem-3 mutants in interstrain hybrids. XX animals are normal females (top). Anterior is left and ventral is down. XO animals are females, but do not produce progeny in crosses (bottom). Anterior is left and ventral is facing up. e) Identification of conserved PUF regulatory sites in the 3′-UTR of C. nigoni fem-3 mRNAs. A gray box highlights the site first defined in C. elegans (Ahringer and Kimble 1991; Bernstein et al. 2005) and gray shading a duplicate site located adjacent to it in C. nigoni. Critical residues are in red. These two sites are destroyed by the cni-fem-3(v395) mutation. f) The fem-3(v395) XX animals develop as normal females. Anterior is left and ventral down.

By contrast, the hybrid XO animals were normal males that made only sperm (5/5, 7-day-old adults, 25°C), as we had observed for tra-1(v481) and tra-1(v538) homozygous XO mutants. Thus, tra-1 is not needed in C. nigoni XO animals to sustain spermatogenesis. This result suggests that the role of TRA-1 in determining germ cell fates differs between the female/male species C. nigoni and the 2 hermaphrodite/male species studied previously.

C. nigoni fem-3 promotes male development in both soma and germ line

TRA-1's stability is controlled by a ubiquitin ligase complex that includes 3 FEM proteins and CUL-2 (Starostina et al. 2007). We selected FEM-3 for detailed analysis because it is essential for the complex to function and does not have pleiotropic activities like FEM-2. To work with C. nigoni fem-3(null) mutants, we used unc-129 as a balancer in the JU1422 background, and unc-33 in the CP168 background. The unc-129 mutants were extremely sick when homozygous, producing mainly dead larvae. The unc-33 homozygotes were healthier and often survived to adulthood, but the females did not mate efficiently.

Both the JU1422 fem-3(v496) mutation and the CP168 fem-3(v542) mutation are early frameshifts that eliminate most of the protein (Fig. 3a;Supplementary Fig. 2). Furthermore, both mutations completely transformed XO individuals into females (N = 19/19), but did not affect XX animals (N = 31/31) (Supplementary Fig. 5). We confirmed these phenotypes by studying interstrain hybrids (Fig. 3b) and identified the genotypes and karyotypes of individual animals using PCR (Fig. 3c).

The fem-3 mutants not only developed female bodies, but even fem-3 XO animals produced oocytes throughout their lives (N = 11/11, 7-day-old adults, 25°C). By contrast, in the sister species C. briggsae, fem-3 XX and XO animals are self-fertile hermaphrodites (Hill et al. 2006). Thus, fem-3 completely controls germ cell fates in C. nigoni, whereas it is dispensable for spermatogenesis in C. briggsae.

We tested young adult females for their ability to reproduce by mating individual animals with 3 wild-type males from strain JU1422. Although the homozygous XX animals could mate and produce progeny, the homozygous XO animals did not produce offspring in male/female crosses (N = 29/29), despite developing as females. This fertility problem resembles that of C. briggsae fem-3 XO mutants (Hill et al. 2006) and of C. elegans her-1 XO mutants (Hodgkin 1980), which both produce tiny self-broods. Thus, we suspect that the lack of progeny from these C. nigoni fem-3 XO animals reflects the harmful effects of an XO karyotype on the expression of oogenesis genes (Strome et al. 2014).

While studying C. nigoni fem-3, we identified a conserved site in its 3′-UTR, which is orthologous to the FBF binding site in C. elegans fem-3 (Ahringer and Kimble 1991; Zhang et al. 1997; Bernstein et al. 2005). In C. elegans, mutations at this site cause XX animals to make sperm instead of oocytes (Barton et al. 1987; Ahringer and Kimble 1991), so we used gene editing to remove it (Fig. 3e). Despite this change, the Cni-fem-3(v395) mutants developed as XX females (Fig. 3f) and XO males. The females made only oocytes and did not display any signs of incipient spermatogenesis, such as laying unfertilized oocytes (N = 14/14). Thus, this regulatory site is conserved in sequence, but appears to have different functions in C. nigoni and C. elegans.

C. nigoni tra-2 promotes female development in soma and germ line

The activities of both FEM-3 and TRA-1 are regulated by binding a fragment of the TRA-2 receptor, which is related to Patched (Kuwabara et al. 1992; Kagawa et al. 2011). Thus, the final gene we targeted was Cni-tra-2. We made 2 alleles in the JU1422 background, v498 and v499 (Fig. 4a;Supplementary Fig. 3). Since both cause an early frameshift and truncation that removes most of the protein, we focused our analyses on v498. In the sister species C. briggsae, a similar mutation behaves as if it causes a complete loss of function (Shen et al. 2024a).

Fig. 4.

C. nigoni TRA-2 directs female development in both germ line and soma. a) Diagram showing wild-type and mutant C. nigoni TRA-2, with white highlighting sequences after the frameshift but before the stop codon. Arrows indicate predicted transmembrane domains, with up indicating those pointing out of the cell. b) DIC photomicrograph of C. nigoni tra-2(v498) XX pseudomale. Anterior is left and ventral facing out. The dotted line outlines the visible part of the male gonad. c) DIC photomicrographs of male tails in wild-type XO and tra-1 XX males and defective tails in tra-2 XX mutants, which have abnormal or missing rays. d) Haploinsufficiency for tra-2 induces spermatogenesis in XX females for C. elegans fog-2 and C. briggsae she-1 mutants, but not for C. nigoni strains.

To balance tra-2(v498), we generated Cni-unc-104(v494) in the JU1422 background. UNC-104 is a kinesin involved in synaptic vesicle transport within axons (Hall and Hedgecock 1991), and the cni-unc-104 mutants often coiled or arrested in early larval development. Finally, the 2 lines were crossed to create the balanced JU1422 strain: Cni-tra-2(v498) + / + Cni-unc-104(v494). We also backcrossed tra-2(v498) 10 times into the CP168 genetic background. To balance it, we generated a recessive mutation of C. nigoni dpy-10 in CP168.

Finally, we studied C. nigoni tra-2(v498) in the JU1422 background (Fig. 4b), the CP168 background, and hybrids. In all cases, it partially masculinized somatic tissues and completely masculinized the germ cells of XX animals (N = 33/33, 19/19, and 5/5, respectively). The partial masculinization was restricted to the tail, which had reduced or incomplete male rays, a smaller fan and stubbier tail section than in XO males. Indeed, these stunted tails resembled those of C. elegans and C. briggsae tra-2 XX mutants (Fig. 4c). Because of these defects, tra-2 XX animals cannot mate and are referred to as pseudomales (Hodgkin and Brenner 1977). By contrast, the C. nigoni tra-2 XO mutants were normal males. Thus, C. nigoni TRA-2 promotes female cell fates, but does so imperfectly in the tail. We infer that some aspects of tra-1 activity in the tail do not require tra-2.

Although the germ line is completely masculinized in C. nigoni tra-2 XX homozygotes, we saw no masculinization in tra-2 heterozygotes (Fig. 4d). By contrast, when C. elegans or C. briggsae XX animals are converted into true females by the loss of fog-2 or she-1, respectively, heterozygosity for tra-2 restores spermatogenesis (Schedl and Kimble 1988; Guo et al. 2009). Finally, we tested C. nigoni tra-2 heterozygotes for signs of partial masculinization of the germ line by seeing if any young adult virgin females laid unfertilized oocytes, which would indicate the presence of sperm proteins even in the absence of functional sperm, but saw no effect (N = 15/15).

Epistasis tests show that sex determination is the same in the C. nigoni soma and germ line

Our results with tra-1, fem-3, and tra-2 suggested that control of sex determination in the germ line is more robust in C. nigoni than in either of the hermaphroditic species. To test this idea, we examined double mutants, since some double mutants have differing effects on germ cells and the soma in the 2 hermaphroditic species (Hodgkin 1986; Hill et al. 2006).

The fact that C. nigoni is a female/male species meant that we could not maintain homozygous double mutant strains. Furthermore, making strains in which 2 sex determination genes were present and both balanced by marker mutations would be difficult. Thus, we set up crosses between animals heterozygous for 2 sex determination genes and then studied their progeny by DIC microscopy (Fig. 5). Afterward, we determined the genotype and karyotype of each animal by PCR. To simplify karyotyping, we used the JU1422 strain to introduce one gene into the cross and CP168 to introduce the other gene. This procedure only allowed us to distinguish XO and XX animals among half of the progeny, and the others remained ambiguous.

Fig. 5.

In C. nigoni, sex determination occurs in a linear regulatory pathway. a) Cross showing how tra-1; fem-3 double mutants were generated, and the sexual phenotypes of all the progeny. b) DIC photomicrograph of a tra-1; fem-3 XX double mutant, showing normal male development and spermatogenesis. Anterior is left and ventral up. The karyotype was inferred from the observation that the animal lacks a gonad. Expanded views of spermatogenesis in both tra-1; fem-3 XX and XO animals (right). c) Cross showing how tra-2; fem-3 double mutants were generated, and the sexual phenotypes of all the progeny. d) DIC photomicrograph of a tra-2; fem-3 double mutant, showing normal female development and a gonad populated only with oocytes. Anterior is left and ventral up.

First, we studied tra-1; fem-3 double mutants. In C. elegans, these animals develop male bodies but produce only oocytes (Hodgkin 1986). In C. briggsae, they develop male bodies and usually make both sperm and oocytes (Hill et al. 2006). However, in C. nigoni they formed male bodies that only made sperm (Fig. 5a and b). Since the fem-3 mutation did not rescue gonad development in the XX double mutants, we saw both tra-1; fem-3 animals with small gonads and sperm (presumed to be XX) and tra-1; fem-3 animals with normal male gonads and sperm (some presumed and some proven to be XO, Fig. 5b). Thus, tra-1 is completely epistatic to fem-3 in both the soma and germ line of C. nigoni. This relationship differs from that seen in both of the hermaphroditic species.

Next, we studied tra-2; fem-3 double mutants. Although fem-3 is completely epistatic to tra-2 in C. elegans (Hodgkin 1986), the tra-2; fem-3 double mutants are self-fertile hermaphrodites in C. briggsae (Hill et al. 2006), which is hard to interpret on its own. However, further tests showed that C. briggsae she-1 fem-3 XX animals are female, but tra-2; she-1 fem-3 animals are self-fertile hermaphrodites (Guo et al. 2009; Shen et al. 2024a), suggesting that tra-2 acts downstream of she-1 and fem-3 in the germ line.

When we studied C. nigoni, we saw that fem-3 was completely epistatic to tra-2 in both the soma and the germ line (Fig. 5c and d). This result was true for both XX and XO double mutants. Thus, in C. nigoni, it appears that TRA-2 acts through FEM-3 to control germ cell fates. By contrast, in C. briggsae, the critical step is the direct interaction of TRA-2 with TRA-1 (Shen et al. 2024a). Putting these results together, it appears that sex determination in the C. nigoni germ line is not only robust, but that the critical interactions occur in a simple linear pathway.

Heterozygosity for C. nigoni mutations does not affect germ cells in interspecies hybrids

A simple explanation for the robust regulation of sexual development we observe in C. nigoni would be that TRA-1 or TRA-2 activity is very high in XX animals relative to that of FEM-3 and its partners, but very low in XO animals. To probe this idea, we studied the effects of heterozygosity for tra-1 or tra-2 in interspecies hybrids between C. nigoni and C. briggsae. Normally, these XX animals make only oocytes, just like C. nigoni XX animals (Fig. 6a;Woodruff et al. 2010). We wondered if high TRA-2 or TRA-1 activity from the C. nigoni genes was the reason these XX animals only produce oocytes.

Fig. 6.

C. nigoni/C. briggsae XX hybrids have robust sex determination in germ cells. a) Summary of crosses that produce intraspecies hybrids (Woodruff et al. 2010). Genes derived from C. nigoni are in blue and those from C. briggsae in red. “St” denotes sterile. b) Diagram of crosses that produce intraspecies hybrids lacking a functional C. nigoni tra-2 allele. c) DIC photomicrograph of a hybrid XX animal lacking C. nigoni tra-2. Note that it develops a female body and female gonad (outlined in white) and produces oocytes (four outlined in yellow). The red arrow indicates the vulva. Anterior is left and ventral down. d) Diagram of crosses that produce intraspecies hybrids lacking a functional C. nigoni tra-1 allele. e) DIC photomicrograph of a hybrid XX animal that is Cni-tra-1(null)/Cbr-tra-1(v197 v383). Note that it develops a female body with an abnormally shaped gonad (outlined in white) but does produce 2 oocytes (outlined in yellow). Anterior is left and ventral down.

First, we crossed C. briggsae wild-type XO males with C. nigoni tra-2/+ females, studied their progeny, and used PCR to confirm the genotype and karyotype of each hybrid. Both the +/+ and the Cni-tra-2/+ XX animals were female (N = 78/78 and 48/48, respectively, Fig. 6b and c). As expected, those XO animals that survived were sterile males. Thus, even hybrids that only express C. briggsae tra-2 make no sperm and develop as females. This result implies that high C. nigoni TRA-2 levels are not the only factor causing oogenesis in the interspecies hybrids.

Second, we crossed C. briggsae XO males with C. nigoni tra-1/+ females. Here too, we saw that both the +/+ and Cni-tra-1/+ XX animals were female (N = 17/17 and 9/9, respectively, Fig. 6d). Again, the XO animals that survived were sterile males. Thus, hybrids that express only C. briggsae tra-1 are also female and do not make sperm. Since tra-1 acts at the end of the sex determination pathway, this result implies that upstream factors from C. nigoni are able to drive oogenesis in the hybrid XX animals, by acting through C. briggsae tra-1.

Since these results implied that C. nigoni TRA-1 should be able to interact with C. briggsae TRA-2, and vice versa, we tested protein interactions for both C. nigoni and for cross-species binding using the yeast two-hybrid system. We detected TRA-1/TRA-2 and TRA-2/FEM-3 interactions in C. nigoni and in all the possible cross-species arrangements (Supplementary Fig. 6).

Third, we studied hybrids in which the only tra-1 allele present is one that promotes extra spermatogenesis in C. briggsae XX animals (Shen et al. 2024a). Even among these hybrid progeny, we saw only one animal that might have made sperm. Most had abnormal gonads and made either germline tumors, or oocytes, or both (Supplementary Fig. 7; Fig. 6e).

Finally, we carried out a series of experiments using a temperature-sensitive allele of C. briggsae tra-2 that we produced by gene editing. Cbr-tra-2(v417ts) is a missense mutation that changes A2629 to C, causing the substitution Aspartic Acid 1210 to Alanine. This alters a rare, conserved residue in an otherwise rapidly evolving part of the FEM-3-binding region (Fig. 7a).

Fig. 7.

The temperature-sensitive mutation tra-2(v417) is suppressed in hybrids. a) Location of the Cbr-tra-2(v417) mutation. Alignment of a portion of the TRA-2 proteins from 5 Caenorhabditis species, with residues conserved between C. nigoni and any of the other species highlighted in gray. The residue affected by v417 in C. briggsae is in red. The left inset shows the position of TRA-2 in the membrane and the release of its intracellular fragment by cleavage. The right inset shows the location of the residue altered by v417 on the intracellular fragment, along with the FEM-3 and TRA-1 binding domains. b) DIC photomicrograph of a Cbr-tra-2(v417ts) XX animal at 25°C. The visible parts of the male gonad are outlined in white, and 3 sperm are indicated with blue arrows. Anterior is left and ventral down. The frames showing the head and tail were taken in a different plane of focus from the others. c) DIC photomicrograph of a Cbr-tra-2(v417ts)/Cbr-tra-2(v440) XX mutant. The abnormal gonad is outlined in white, and 2 sperm are indicated with blue arrows. The red arrow marks the protruding vulva. d) DIC photomicrograph of a Cbr-tra-2(v417ts)/Cni-tra-2(v498) XX mutant. Note that it develops a female body and female gonad (outlined in white) and produces oocytes (2 outlined in yellow). The red arrow indicates the vulva. Anterior is left and ventral up.

C. briggsae tra-2(v417ts) XX animals were normal hermaphrodites at 20°C, but showed strong masculinization at 25°C. Most notably, at the restrictive temperature all animals showed partially masculinized tails (N = 11/11). Their gonad development ranged from hermaphroditic (N = 1/11) to fully male (N = 2/11; Fig. 7b) with the remainder being severely disorganized. As a consequence of the partial transformation of the gonad, we observed a protruding vulva in some animals, which would be induced by a hermaphroditic anchor cell (N = 4/11). Finally, all animals made sperm, but only a few also made oocytes (N = 3/11).

We observed a similar range of phenotypes in C. briggsae tra-2(v417ts)/tra-2(v440null) XX mutants (Fig. 7c). Thus, we propose that v417 causes a strong but incomplete reduction of TRA-2 function at the restrictive temperature of 25°C.

To study the effects of low TRA-2 activity in intraspecies hybrids, we examined Cbr-tra-2(v417ts)/Cni-tra-2(v498null) XX animals at 25°C (Fig. 7d). Surprisingly, all developed as normal females and produced only oocytes (N = 9/9). Thus, we infer that when Cbr-TRA-2(v417) works in the context of C. briggsae development, it is unable to block FEM activity at 25°C. However, when it works in a hybrid context, the robust regulation from the C. nigoni members of the regulatory pathway restores normal sex determination.

Discussion

First sex determination mutants described in a male/female species of nematode

Because of the power and simplicity of hermaphroditic genetics, enormous progress has been made using nematodes to model animal development and behavior (see WormBook.org). However, these studies cannot address critical aspects of female development and behavior, or male/female interactions. They also focus on species that are highly inbred (Dolgin et al. 2007; Gimond et al. 2013), which could have produced unusual genetic traits or regulatory systems.

Addressing this problem requires studying female/male species. Previously, such studies depended on imprecise techniques like RNA interference (Haag and Kimble 2000; Chen et al. 2001). However, the ability to make mutations and genetic balancers using gene editing has made more precise genetic studies possible (e.g. Yin et al. 2018). Here, we significantly extend these analyses by isolating and studying single and double mutants for 3 critical sex determination genes, none of which could be maintained as a homozygous line. We anticipate a flowering of similar work on female/male species.

C. nigoni has a robust sex determination pathway for germ cells

Both C. elegans and C. briggsae produce XX animals with male and female germ cells, making them self-fertile. Moreover, mutations in the genes that control this sexual trait reveal considerable flexibility in the decision to produce sperm or oocytes. For example, tra-1 null mutations cause both spermatogenesis and oogenesis. Furthermore, tra-2 mutations reveal extreme sensitivity to gene dosage. And finally, the fem-3 mutations reveal an essential role in spermatogenesis for C. elegans, but almost no role in C. briggsae.

By contrast, our C. nigoni mutations have clear, simple germline phenotypes. The tra-2 null mutants make only sperm. The fem-3 null mutants and tra-2; fem-3 double mutants make only oocytes. Finally, the tra-1 null mutants and tra-1; fem-3 double mutants make only sperm. Moreover, a fem-3 regulatory mutation that causes C. elegans to produce sperm has no effect on C. nigoni. These results suggest that the binary nature of sex determination in the C. nigoni germ line is inherently robust (Fig. 8). As a further test for robustness of sex determination in C. nigoni germ cells, we studied hybrids with the sister species C. briggsae. Even those XX hybrids with only C. briggsae tra-1 activity, or with only diminished C. briggsae tra-2 activity, developed as females. Thus, the C. nigoni sex determination pathway still causes all XX germ cells to become oocytes in these hybrids, even when key C. nigoni genes are not present.

Fig. 8.

Sex determination in C. nigoni. a) Outline of gene regulation in the C. nigoni soma. CAPITALS indicate protein names. Items in blue generally promote male fates and those in red female fates. Larger fonts and black lines denote strong activity, whereas smaller fonts and gray line denote weak activity. Arrows indicate positive regulation and “—|” negative regulation. Because C. nigoni tra-1(lf) XX males have incomplete gonadogenesis, there must be a separate regulatory interaction that promotes male gonad development, in addition to that of TRA-1. b) Outline of gene regulation in the C. nigoni germ line.

Robustness is a critical trait for many biological systems (reviewed by Kitano 2004; Félix and Barkoulas 2015), including the robust suppression of variable phenotypes (Hong et al. 2016; Katsanos et al. 2017). Thus, regulatory processes are often structured so that small changes in gene activity do not result in large changes in phenotype, or in extremely variable phenotypes. As a result, development and homeostasis are buffered from minor genetic variation or environmental effects. However, such robustness might also be an impediment to the origin of new traits during evolution. We have been studying the frequent, independent origin of self-fertile hermaphrodites in nematodes as a model for this process.

One major question for the origin of self-fertility has been which preexisting traits made this transition possible and capable of independent repetition. One such trait has been known for decades—the use of an XX:XO system to determine sex. Since the XX females have all the genetic information needed for male development in their genomes, regulatory changes that activate male programs in the opposite sex are possible. By contrast, in XX:XY systems, the XX females typically lack many Y genes needed for reproduction (e.g. Hennig et al. 1989; Chandley 1998), so only a complex series of genetic changes could (in theory) lead to sperm production. Here, we tested the possibility that flexibility in the sex determination pathway also provided a precondition for the origin of self-fertility. For example, if low tra-1 activity always caused animals to make both sperm and oocytes, even in female/male species, then small regulatory changes that affected tra-1 function might have precipitated the origin of hermaphrodites.

Our observations do not support this model. Although changes in the sex determination pathway are clearly a major part of the evolution of self-fertility, the natural robustness of the pathway in C. nigoni implies that these changes must fall into 2 categories. First, some changes impaired the robust sexual canalization of germ cells, making it easier for individual animals to switch to producing both sperm and oocytes. Second, other changes directly altered germ cells fates in XX animals, so that they make sperm as well as oocytes. Each of these classes might have involved multiple mutations and a long process of refinement.

This result helps clarify the function of genes like fog-2 in C. elegans and she-1 in C briggsae. Both genes are unique and were produced by recent gene duplications, and both are required for XX animals to develop as hermaphrodites rather than females (Schedl and Kimble 1988; Clifford et al. 2000; Nayak et al. 2005; Guo et al. 2009). As a result, fog-2 and she-1 strains reproduce as XX females and XO males, which suggested that each gene might be the key factor responsible for self-fertility in its respective species. However, we have shown here that haploinsufficiency for tra-2 does not cause spermatogenesis in C. nigoni XX animals, whereas it does so in C. elegans fog-2 strains and C. briggsae she-1 strains. Hence, we propose that tra-2 activity is much higher relative to that of its competitors in C. nigoni germ cells than in either hermaphroditic species, which implies that lowering tra-2 activity might have been crucial for allowing genes like fog-2 and she-1 to evolve and function. Consistent with this model, levels of the TRA-2 protein are so low in C. elegans germ cells as to be almost undetectable (Hu et al. 2019).

tra-1 plays a conserved role in gonad development that depends on genetic background

In C. nigoni, tra-1 XX null mutants have an unexpected phenotype—gonad development is severely compromised. Although C. elegans tra-1 XX mutants often make fully functional males with normal gonads, they sometimes produce defective gonads, and the tra-1 XO mutants occasionally make defective gonads too (Hodgkin 1987). These defects stem from problems with the positioning of the somatic gonad precursors within the gonad primordium of young larvae, as well as defects in the polarity of the Z1 cell division (Mathies et al. 2004). In C. briggsae, the strong tra-1 mutant nm2 produces XX males with normal somatic gonads, although they initially have defects in polarity of the Z1 cell division (Kelleher et al. 2008). The XO phenotype has not been studied.

By contrast, our results reveal a more severe tra-1 gonadal defect than seen in either C. elegans or C. briggsae. However, they also show strong dependence on karyotype—the XX animals all have defective gonads, whereas the XO animals are normal males, which we have used successfully in crosses. This result implies that the X:Autosome ratio controls gonadal sex determination in 2 ways: by acting directly through the sex determination pathway to control tra-1 activity and by acting in parallel to this pathway on gonad development. For example, C. nigoni tra-1 might act in conjunction with another gene to control early development of the gonad in XX animals. One possibility for future exploration is fkh-6, which works with tra-1 to control male gonad development in C. elegans (Chang et al. 2004) and which has orthologs in other Caenorhabditis species.

The X:Autosome ratio can influence sex outside of the normal sex determination pathway

These tra-1 results establish two roles for the X:Autosome ratio in C. nigoni sexual development. One study in C. elegans also raised the possibility of split regulatory processes: the xol-1 gene, which directly interprets the X:Autosome ratio, acts both upstream and downstream of tra-2 to regulate male tail development (Miller et al. 1988). In these 2 cases, these effects might represent distinct gene regulatory pathways, or the influence of changes in dosage compensation on particularly sensitive genes in the sex determination pathway.

Hybrid mutant analysis facilitates the study of rapidly evolving traits

To date, most work on interspecies hybrids has focused on the genes responsible for incompatibility. These studies are most advanced in Drosophila (reviewed by Presgraves and Meiklejohn 2021), but have made significant progress in Caenorhabditis as well (Bi et al. 2019; Xie et al. 2022). However, speciation involves not only barriers to hybridization, but also rapid changes in traits critical for each new species to adapt to its environment. We have used mutants in hybrid nematodes to elucidate key changes in sex determination during recent evolution and anticipate that a similar use of hybrids will be fruitful for many additional studies, now that the necessary mutations can be produced by gene editing. Furthermore, genetic studies like we present here can be combined with the analysis of hybrid transcriptomes (Sánchez-Ramírez et al. 2021) to deepen our understanding of regulatory networks.

Model for the evolution of self-fertile hermaphrodites

Putting our current results in the context of previous work in our field, we propose that self-fertility involved a series of regulatory changes. (1) One of the two male sperm activation programs could have been activated in otherwise female animals (Baldi et al. 2009; Wei et al. 2014b). (2) Mutations in some sex determination genes might have compromised the robustness of the process in germ cells, without themselves creating XX hermaphrodites. Such changes would create the ground state needed for co-option of male programs to occur. Kitano (2004) notes that robustness is facilitated by redundancy. Because redundant interactions can be shed or augmented in different ways without altering the phenotype, robustness could facilitate the evolution of the system (Haag 2007). Such cryptic evolution has been dubbed developmental systems drift (True and Haag 2001). Sex determination systems, which evolve to produce exactly two alternative outcomes, may be especially prone to this type of developmental system drift. (3) A precipitating mutation could then have tilted the regulatory balance so that XX animals made some sperm as well as oocytes. (4) Finally, multiple mutations would refine this process, optimizing its efficiency and the number of self-sperm.

Once self-fertile hermaphrodites existed, they might have spread easily because of their natural efficiency at colonization (Pannell 2015). At the same time, the new population would have gone through inbreeding crises before eventually stabilizing, once many harmful recessive mutations were purged (Dolgin et al. 2007). It would eventually reach a point where outcrossing might even be detrimental (Gimond et al. 2013). Meanwhile, a host of associated changes in male traits and genome size would occur as the change in mating systems altered the way natural selection affected the 2 sexes (Yin et al. 2018). Given the richness and scope of this model for evolutionary change and the numerous developments in experimental evolution in nematodes (Cutter et al. 2019), we anticipate the repetition of these natural experiments in laboratory simulations, where every detail can be monitored.

Data availability

For the C. nigoni CP168 genome, its raw sequencing reads, genome assembly, and protein-coding gene annotations are available under GenBank BioProject accession number PRJNA1143586. These data are also archived in the Open Science Framework (https://osf.io/qzupj).

Supplemental material available at GENETICS online.

Funding

The authors thank National Institutes of Health grants R01GM118836 and R01GM121688 for funding to REE and the American Cancer Society for a Postdoctoral Award to support S-YL (126627-PF-15-228-01-DDC).