Gonads

Wild type adult male and female zebrafish, Danio rerio, were obtained from a commercial supplier (Ekkwill, Gibsonton, FL) and maintained in 30 gal aquaria at 28°C on a 14:10 light-dark cycle. Fertilized eggs were collected after natural spawning, washed, and distributed into 20 × 100 mm culture plates (Fisher Scientific). Embryos (150 embryos/50 ml egg water) were allowed to develop at 28°C on a 14L:10D cycle [36 ]. For developmental expression analysis embryos were collected after timed intervals: 2, 6, 12, 24, 48, 72, and 120 hours post-fertilization (hpf), quick-frozen on dry ice, and stored at -70°C until analysis (3 independent embryo pools, 50 embryos per pool, per time point from the same spawning group). For treatment expression analysis embryos were left untreated until 24 hpf and then exposed to 17β-estradiol (E2; 0.1 μM), testosterone (T; 1 μM), ICI 182,780 (ICI; 10 μM; Tocris Bioscience, Ellisville, MO), β-napthaflavone (BNF; 10 nM), or 2,3,7,8, tetrachlodibenzo-p-dioxin (TCDD; 1 nM; Ultra Scientific, N. Kingstown, RI) dissolved in dimethyl sulfoxide (DMSO). All chemicals were obtained from Sigma-Aldrich (St. Louis, MO) unless otherwise noted. Stock solutions of chemicals were added directly to egg water and replaced daily. In addition, embryos were treated with DMSO alone (final concentration, 0.0006%), EtOH alone (final concentration 0.0005%), or left untreated as a control. Embryos were collected at 96 hpf, quick-frozen on dry ice, and stored at -70°C until analysis (3 independent embryo pools per treatment). Treated embryo RNAs were used for both housekeeping gene expression analysis (Table 3) and gene of interest normalization (Figure 2). Tissues (brain, eye, heart, liver, muscle, gonad) were collected from adult male and female zebrafish, pooled by sex (3 pools per tissue type/sex, 5 fish per pool), quick-frozen on dry ice, and stored at -70°C. Adult fish were reproductively active stock from our breeding colony.

Expression levels of the individual TE superfamilies were calculated by averaging the TPM values among replicates of each sex and then summing the average TPM of all contigs annotated to each superfamily. For TE superfamilies detected in both the genomic and transcriptomic datasets, we tested for a relationship between genomic abundance and expression levels in the gonads of each sex using linear regression on log-transformed data.
To identify differentially expressed contigs between testes and ovaries, DESeq2 (Love et al., 2014 (link)) was used with an adjusted p-value cut off of 0.05. Among the 15,011 total differentially expressed transcripts between testes and ovaries (including TEs, endogenous genes, and unannotated contigs), 869 were TEs, representing 18 superfamilies and other unknown TEs. Superfamilies with fewer than 10 differentially expressed transcripts between testes and ovaries were removed, leaving nine superfamilies; for each, we tested for a difference in expression between testes and ovaries using a t-test.

In all experiments, sacrifices were carried out in the morning (from 9 am to 1 pm). Mice were anesthetized, exsanguinated by heart puncture, and transcardially perfused with PBS. For plasma metabolite analysis, blood was collected in tubes containing 20 μL heparin and spun at 3,000 g for 15 min at 4 °C. Supernatant plasma was aliquoted and snap frozen in liquid nitrogen (LN). For immune profiling, peripheral blood mononuclear cells (PBMCs) were isolated by density gradient centrifugation protocol (Cytiva™ – density 1.084 ± 0.001 g/ml). After dissection, the following brain regions were excised: for histology, left hemisphere, post-fixed in 4% paraformaldehyde (PFA)/PBS; for sNuc-Seq, left hippocampus, snap frozen in LN; for ELISA and biochemistry, right cortex and right hippocampus, snap frozen in LN. The spleen was mashed with the plunger of a syringe against a 70 mm strainer and treated with ammonium-chloride-potassium (ACK) lysis buffer (Gibco™) to remove erythrocytes. Splenocytes were then used immediately for flow cytometry and pan-T-cell cultures, while for CyTOF, aliquots were resuspended in cell freezing medium (Sigma-Aldrich) and frozen in a Mr. Frosty container (Thermo Fisher) at −80 °C. The left gonadal fat pad was snap frozen in LN for sNuc-Seq.

The first crossing experiments were performed as previously reported^{31 (link)}. In short, three J4 juvenile females of dioecious species and six young-adult males of P. pacificus were crossed on Nematode Growth Medium (NGM) plates with a 50 µl OP50 lawn (N = 4 in each species). The following strains were used: P. pacificus, PS312; P. exspectatus, RS5522; P. occultus, RS5811; P. sikae, RS5901; P. arcanus, RS5527; P. kurosawai, RS5914; P. taiwanensis, RS5797; P. maxplancki, RS5594. Males were removed after two days and females were removed after four days to avoid backcross. Hybrids were allowed to cross on the same plate and cultured. To prevent starvation, at least 50 worms were transferred to a new plate to continue the culture. If newborn larvae continued to be produced for one month (~6 generations), we considered the hybrids to propagate. To confirm the reproducibility of the original cross between P. pacificus and P. exspectatus, we made eight additional replicates (Supplementary Table 6). To prevent starvation, 25% of the individuals on the plate were transferred before starvation (after day 5). Thus, the number of animals counted on day ten is reduced relative to day five because 75% of the animals were not transferred. Subsequently, >95% of worms were transferred in the second or later transfers. We counted the number of individuals all 5 days for 30 days, indicating the numbers of J2/J3 and J4 juveniles, as well as adults for both, hermaphrodites and males. Note that eggs were not counted. We found continuous production of juveniles in four of the eight replicates (50%). Note that the number of animals declines over time because some hybrid progeny die without producing a large number of progeny. Importantly, however, new juveniles were constantly observed throughout the duration of the experiment, and no trend of change of sex ratio was observed (Supplementary Table 6).
For quantitative reproduction tests, one virgin female and one young-adult male were mated on the NGM plate with a 50 µl OP50 lawn with egg laying for six days. Parents were transferred to new plates every second day. Progeny were grown for three to four days on these plates. The number of males, females (or hermaphrodites) and immature progeny were counted on the basis of their morphology. Because the hermaphrodites have the same morphology as females, we do not distinguish these two sexes. When the two-arm gonad and the vulva were observed, the worm was categorized as a female or hermaphrodite. When the one-arm gonad and connection of the gonad to the spicule were observed, the worm was categorized as a male. When these reproductive traits were not observed, the worm was categorized as an immature animal. The type strain of P. pacificus, PS312, and an inbred line of the type strain of P. exspectatus, RS5522B, were used in this experiment. For hybrid crosses, old females or hermaphrodites (four days after J4 stage) were used for mating to let hermaphrodites run out of self-sperm. We also tested the number of progeny of old P. pacificus hermaphrodites without mating at the same time (N = 18). Because only one progeny was found from all hermaphrodites (0.056 progeny per replication on average), the self-progeny is negligible in the analysis. For backcrossing, we first prepared F1 animals produced by P. exspectatus dam and P. pacificus sire or F1 animals produced by P. pacificus dam and P. exspectatus sire using the experimental set-up described in the preceding and backcrossed them with animals of the pure species. We used young J4-stage females or hermaphrodites for backcrossing. We did not test the backcross with P. pacificus hermaphrodites that produce ~200 self-progeny because that makes the interpretation difficult. For the test of hermaphroditic reproduction of F1 animals, F1 female or hermaphrodite was placed on the NGM plate with a 50 µl OP50 lawn without males. We tested the wild-type cross of P. exspectatus in each backcross as control. For intercrosses of F1 animals, F1 hybrids were crossed to each other to avoid the effect of inbreeding. The sample number of each experiment is listed in Supplementary Table 1. Asymptotic Wilcoxon–Mann–Whitney test was performed using wilcox_test function of an R package, coin.
The reproductive capacity of F1 males was compared between crosses of P. pacificus and different dioecious species using the same experimental scheme. The F1 males were produced by crosses between females of dioecious species and P. pacificus male and backcrossed with parental dioecious species. We used wild isolates of six dioecious species. Only presence or absence of progeny (BC1) was analysed in these experiments (N = 25 each).