Revisiting Suppression of Interspecies Hybrid Male Lethality in Caenorhabditis Nematodes

Within the nematode genus Caenorhabditis, Caenorhabditis briggsae and C. nigoni are among the most closely related species known. They differ in sexual mode, with C. nigoni retaining the ancestral XO male–XX female outcrossing system, while C. briggsae recently evolved self-fertility and an XX-biased sex ratio. Wild-type C. briggsae and C. nigoni can produce fertile hybrid XX female progeny, but XO progeny are either 100% inviable (when C. briggsae is the mother) or viable but sterile (when C. nigoni is the mother). A recent study provided evidence suggesting that loss of the Cbr-him-8 meiotic regulator in C. briggsae hermaphrodites allowed them to produce viable and fertile hybrid XO male progeny when mated to C. nigoni. Because such males would be useful for a variety of genetic experiments, we sought to verify this result. Preliminary crosses with wild-type C. briggsae hermaphrodites occasionally produced fertile males, but they could not be confirmed to be interspecies hybrids. Using an RNA interference (RNAi) protocol that eliminates any possibility of self-progeny in Cbr-him-8 hermaphrodites, we found sterile males bearing the C. nigoni X chromosome, but no fertile males bearing the C. briggsae X, as in wild-type crosses. Our results suggest that the apparent rescue of XO hybrid viability and fertility is due to incomplete purging of self-sperm prior to mating.


Haldane's rule X chromosome hybrids Genetics of Sex
Interspecies hybrids can provide insight into the genetic mechanisms behind the diversity of organisms, speciation, and the arising of novel traits. Reproductive barriers limit gene flow between species and can be pre-or postzygotic. Prezygotic barriers include behavioral isolation or gametic incompatibility. Postzygotic reproductive barriers include hybrid lethality (at any prereproductive developmental stage) or sterility. It is this latter barrier that is the focus here. The Bateson-Dobzhansky-Muller model proposes a genetic basis for hybrid incompatibility, whereby incompatibilities between heterospecific loci result in impaired function or nonfunction (Dobzhansky 1936). In addition, interspecies hybrids often manifest Haldane's rule, where the heterogametic sex is more severely impacted, presumably because of sex chromosome hemizygosity (Orr and Turelli 2001). Darwin's corollary to Haldane's rule is also observed in many cases, where there is asymmetric male viability in reciprocal crosses (Turelli and Moyle 2007).
Caenorhabditis nematodes are an excellent system to study the genetic basis of reproductive diversity. The genus contains both gonochoristic (male/female) and androdioecious (male/hermaphrodite) species, making it possible to study the variation of reproductive mode. The essence of hermaphroditism is limited spermatogenesis in the context of the XX female ovary. How development of the bisexual germline is regulated has been studied heavily in Caenorhabditis elegans (e.g., Doniach 1986;Goodwin et al. 1993;Ellis and Kimble 1995;Francis et al. 1995;Zhang et al. 1997;Chen et al. 2000;Clifford et al. 2000;Luitjens et al. 2000) and, more recently, in C. briggsae (Chen et al. 2001Hill et al. 2006;Guo et al. 2009;Hill and Haag 2009;Beadell et al. 2011;Liu et al. 2012). These two selfing species, while superficially similar, evolved self-fertility independently (Kiontke et al. 2011) and it was achieved via distinct modifications of the global sex determination pathway (Hill et al. 2006;Guo et al. 2009;Hill and Haag 2009;Chen et al. 2014).
The first studies comparing sex determination in hermaphrodite and female relatives of C. elegans used candidate gene approaches (de Bono and Hodgkin 1996;Kuwabara 1996;Streit et al. 1999;Haag and Kimble 2000;Haag et al. 2002;Hill et al. 2006;Hill and Haag 2009;Beadell et al. 2011;Liu et al. 2012) and forward genetic screens (Hill et al. 2006;Kelleher et al. 2008;Guo et al. 2009Guo et al. , 2013. The C. briggsae and C. nigoni system has opened the tantalizing possibility of using hybrids between them to identify factors distinguishing hermaphrodite and female germline sex determination (Woodruff et al. 2010). However, these efforts have been thwarted by extensive genetic incompatibilities. C. briggsae · C. nigoni hybrids are subject to both Haldane's rule and Darwin's corollary to Haldane's rule (Woodruff et al. 2010;Kozlowska et al. 2012). Specifically, no viable male F1 are found when wild-type C. briggsae hermaphrodites are mated with C. nigoni males, but viable yet sterile males are produced when C. nigoni females are crossed with C. briggsae males. Surprisingly, after laying a few hybrid progeny, C. briggsae hermaphrodites mated with C. nigoni males are sterilized by the aggressive C. nigoni sperm (Ting et al. 2014).
F1 females from both possible C. briggsae · C. nigoni crosses produce viable progeny only when backcrossed to C. nigoni (Woodruff et al. 2010). This has allowed introgression of marked C. briggsae chromosomal segments into C. nigoni (Yan et al. 2012;Bi et al. 2015). These segments remain large in spite of multiple backcrosses, and harbor a number of inviability and sterility loci, some of which impact germline small RNA pathways (Li et al. 2016). In addition, polymorphisms within C. briggsae and C. nigoni can impact the severity of F1 hybrid phenotypes (Kozlowska et al. 2012).
As for its ortholog in C. elegans (Phillips et al. 2005), the C. briggsae high incidence of males 8 (him-8) gene is required for faithful segregation of the X chromosome during meiosis (Wei et al. 2014). This, in turn, greatly elevates the spontaneous production of XO self-progeny, which are male. Thus, while an unmated C. briggsae hermaphrodite will produce , 1% males naturally, Cbr-him-8 mutants are Him, producing 15% males (Wei et al. 2014). Recently, Ragavapuram et al. (2016) reported that loss of Cbr-him-8 function can rescue both lethality and sterility in male hybrids bearing the C. briggsae X chromosome. As such males might allow F1 intercrosses and new backcross types, we sought to verify and extend these results. Surprisingly, using the same strains as Ragavapuram et al. (2016) but with methods that eliminate the possibility of males arising from selfing, we find no evidence of hybrid male rescue.

Cbr-fog-3 RNAi
A Cbr-fog-3 template for in vitro transcription was made with the polymerase chain reaction (PCR) using Taq DNA polymerase (New England Biolabs) with recommended concentrations of dNTPs and primers. Primer sequences (including the underlined T7 promoter required for subsequent in vitro transcription) are: forward: 59-TAATACGACTCACTATAGGGAGCCGACGAAGTTCTTGAAA-39; reverse: 59-TAATACGACTCACTATAGGGCCCACCATGGTCTGCA GATC-39. The PCR product was purified using a QIAquick PCR purification kit (QIAGEN). Next, 780 ng of Cbr-fog-3 PCR product was used as template in a Megascript T7 in vitro transcription reaction (Thermo Fisher Scientific) following the provided protocol. The resulting dsRNA was purified by ammonium acetate and ethanol precipitation. PB192 hermaphrodites were picked onto a separate plate at the L4 stage 12 hr prior to injection. Forty adult worms were mounted on agar pads and injected with Cbr-fog-3 dsRNA at a concentration of 3000 ng/ml. Injected animals were moved to individual NGM plates seeded with OP50 E. coli plates after 8 hr of recovery time.

Crosses
For interspecies crosses not employing Cbr-fog-3(RNAi), PB192 hermaphrodites were purged by several days of serial transfer to NGM plates, similar to the method of Ragavapuram et al. (2016).
C. briggsae mothers who could not self-fertilize were produced through RNAi targeting of Cbr-fog-3 (Chen et al. 2001). Progeny of injected PB192 mothers that appeared to have the feminization of germline (Fog) oocyte stacking phenotype after reaching adulthood were moved in small groups to new agar plates seeded with OP50 E. coli, and allowed to sit for 6 hr to ensure that no self-progeny were produced. C. nigoni EG5268 males were added at a 2:1 ratio of males/ hermaphrodites and allowed to mate for 4 hr. All plugged animals were moved to individual plates. Cbr-fog-3(RNAi) PB192 pseudofemales were also crossed to C. briggsae AF16 males to verify that they had normal fecundity.

Microscopy
Routine maintenance and crosses were performed using a Leica MZ125 stereoscope. Analysis of male germline morphology used differential interference contrast optics on a Zeiss Axioskop 2 plus at 400 · magnification.

Data availability
All strains utilized are available from the Caenorhabditis Genetics Center (https://cbs.umn.edu/cgc/home) or from the authors by request.

RESULTS
We performed preliminary experiments with C. briggsae him-8; myo-2::GFP X hermaphrodites (strain PB192) that had been ostensibly purged of self-sperm by serial transfer over several days until embryo laying stopped, as done by Ragavapuram et al. (2016). The expectation was that roughly 15% of progeny from mating such purged hermaphrodites with C. nigoni males would be fertile males. Using the C. nigoni strain JU1422 in three different trials with 6-7 mothers each, 0/47, 3/50, and 0/116 progeny were male (1.4%).
Because Ragavapuram et al. (2016) used C. nigoni males of the African EG5268 strain, we next considered the possibility that the unexpectedly infrequent males obtained above were a strain effect. Using the purging approach, 7/66 progeny were male in the first experiment with EG5268 males (11%). Of these, two were extremely small and n Table 1 Phenotypes of progeny from Cbr-him-8; myo-2::GFP; Cbr-fog-3(RNAi) mothers mated to C. nigoni EG5268 wild-type males infertile, similar to those observed when F1 males have a C. nigoni X chromosome (Woodruff et al. 2010). Because him-8 mothers produce nullo-X oocytes at an appreciable frequency, this is expected (Ragavapuram et al. 2016). Five others were GFP+, robust, and fertile, and thus candidates for rescued F1 hybrid males. However, plates bearing these putative hybrid males subsequently gave rise to vigorous populations of uniformly GFP+ animals. Individual virgin hermaphrodites isolated from these plates were invariably Him. This suggested that, despite purging, C. briggsae PB192 mothers could occasionally produce self-progeny after mating with C. nigoni males, perhaps via residual self-sperm that were resistant to purging. A second attempt to generate hybrid males with EG5268 sires produced three males, all GFP2 and small. The above preliminary crosses produced fertile males at a frequency lower than the expected 15%. They also indicated that complete purging in C. briggsae may be more difficult to achieve than had been previously appreciated. To ensure that all progeny being scored were interspecies hybrids, and to allow the use of younger, healthier mothers, self-sperm were ablated in PB192 by Cbr-fog-3(RNAi) via maternal injection (Chen et al. 2001). From over 2000 hybrid F1 embryos laid, 16 viable male adults were obtained (Table 1). These males had fully formed tails and exhibited mating behavior, but were GFP2 and unusually small. Attempts to backcross them to their siblings failed to produce any embryos. Consistent with this apparent sterility, all of these males lacked a fully formed germline (often apparently completely absent).
The above results are consistent with all of the F1 hybrid males obtained in the crosses being derived from fertilization of a nullo-X C. briggsae oocyte by a C. nigoni male X-bearing sperm. This produces an X Cni O genotype known to produce sterile males (Woodruff et al. 2010). They also indicate a lack of any viability or fertility of F1 X Cbr O males, contrary to the interpretation of Ragavapuram et al. (2016). To be sure that the germline feminization of P0 C. briggsae hermaphrodites by RNAi did not suppress normal fertility in their sons, we verified that male offspring of Cbr-fog-3(RNAi) Fog mothers have normal fertility with conspecific matings (data not shown).

DISCUSSION
In both C. elegans and C. briggsae, loss of him-8 function specifically impairs X chromosome pairing (Phillips et al. 2005;Wei et al. 2014), and unpaired C. elegans chromosomes are subject to meiotic silencing (Bean et al. 2004). In addition, hemizygosity of the X chromosome is thought to underlie Haldane's rule in male-heterogametic systems. These observations led Ragavapuram et al. (2016) to hypothesize that Cbr-him-8 mutant hermaphrodites produce X-bearing oocytes with altered X-linked gene expression that rescues X Cbr O F1 viability. As plausible as this mechanism is, we were unable to replicate the rescue of these males in our own hybrid crosses when all possibility of selfing was eliminated via Cbr-fog-3(RNAi). The rare sterile males we did observe resulted from a C. nigoni X-bearing sperm fertilizing a nullo-X oocyte of the C. briggsae hermaphrodite, which often occurs in the Cbr-him-8 PB192 strain.
It is conceivable that the use of Cbr-fog-3(RNAi) somehow blocks suppression of X Cbr O lethality that would otherwise be provided by the Cbr-him-8 mutations. However, we also saw few or no fertile males in crosses with purged hermaphrodites lacking this treatment, and those that were produced appeared not to be interspecies hybrids. Therefore, we conclude that the previous report of X Cbr O hybrid male rescue was premature, and may be the result of incomplete purging of C. briggsae mothers. Perhaps importantly, the PB192 strain produces male self-progeny at a rate similar to that reported by Ragavapuram et al. (2016) as F1 hybrids.
We see some evidence for cryptic retention of a small number of selfsperm in putatively purged C. briggsae hermaphrodites. This may occur if sperm-derived major sperm protein-mediated signaling is insufficient to stimulate ovulation (Miller et al. 2001). Subsequent mating with C. nigoni males could stimulate resumption of ovulation, allowing the remaining conspecific sperm to be used. In addition, C. briggsae X-bearing sperm contributed by males are preferentially used by hermaphrodites over those lacking an X (LaMunyon and Ward 1997), such that the last sperm used are highly enriched for nullo-X gametes. Because Cbr-him-8 hermaphrodites produce an unusual number of nullo-X self-sperm (Wei et al. 2014), a similar mechanism could produce C. briggsae PB192 XO males that are both fertile and GFP+ from only a handful of residual self-sperm. However, this preferential fertilization by X-bearing sperm is not seen with another C. briggsae mutation that induces nullo-X self-sperm at rates comparable to the him-8 mutations employed here (LaMunyon and Ward 1997). Whatever the mechanism involved, the production of rare selfed progeny from nominally purged animals underscores the need for caution.
It remains possible that, under some conditions, fertile C. briggsae · C. nigoni hybrid males may yet be produced. If so, it will be crucial to demonstrate their hybrid nature by genotyping assays. Sufficient genome data now exist for C. briggsae (Ross et al. 2011) and C. nigoni (Li et al. 2016) to make such assays, such as PCR amplification of indel polymorphisms (Koboldt et al. 2010), rapid and inexpensive.