In oma-1(zu405) embryos, the C blastomere is transformed to the ethyl methanesulfonate (EMS) blastomere fate at a restrictive temperature, resulting in temperature sensitive embryonic lethality (Figure 1A; Lin, 2003). We performed the suppressor screen to identify mutations rescuing embryonic lethality of the oma-1(zu405) strain. During the screen, F2 progenies of EMS/N-ethyl-N-nitrosourea (ENU) treated oma-1(zu405) worms were cultured at 25°C to determine whether F3 eggs were hatched or not. All viable F3 worms were collected for further study (Figure 1B). A total of 0.6 × 10⁵ F2s (i.e., 1.2 × 10⁵ genomes) were screened and 23 mutant strains were isolated. We used snip-SNPs as molecular markers to roughly map mutations in different chromosomes, then we performed whole genome sequencing (WGS) for oma-1(zu405) and all mutants, and analyzed the mutated genes with reference to the corresponding mapped chromosomes (Figure 1B). By comparing mutated genes in different strains, we identified three strains that had independent mutations in mat-1, two strains that had independent mutations in mat-3, and four strains that had independent mutations in fzy-1. Interestingly, MAT-1, MAT3, and FZY-1 all belong to subunits of the APC/C complex (Kitagawa et al., 2002; Shakes et al., 2003; Garbe et al., 2004). APC/C is a complex that has E3 ubiquitin ligase activity and plays a central role in the metaphase to anaphase transition and chromosome segregation steps during mitosis. To date, 12 subunits have been identified in yeast, 11 subunits have been identified in humans, and 10 subunits have been identified in C. elegans (Figure 1C; Furuta et al., 2000; Golden et al., 2000; Davis et al., 2002; Peters, 2002; Rappleye et al., 2002; Shakes et al., 2003; Passmore, 2004; Yeong, 2004; Pines, 2006). These subunits are organized into three functionally related subcomplexes: the platform (MAT-2/Apc1, EMB-30/Apc4, and SUCH-1/Apc5), the TPR lobe (MAT-1/Apc3, EMB-27/Apc6, MAT-3/Apc8, and CDC-26/Cdc26), and the catalytic core (APC-2/Apc2, APC-10/Apc10, and APC-11/Apc-11). There are also two activators, FZY-1/Cdc20 and FZR-1/Cdh1, that determine the target specificity of APC/C (Figure 1D).

We next asked whether other genes coding APC/C subunits have mutations in these strains. Eventually we identified one strain that had a mutation in mat-2 and one strain that had a mutation in emb-30 (Figure 2A). All mutations we identified in APC/C genes are recessive, since only homozygous, but not heterozygous, mutants suppressed lethality of oma-1(zu405).

To confirm whether these candidate genes are functional genes that suppress lethality of oma-1(zu405), we performed two sets of experiments and investigated animal viability by scoring brood size: (1) RNAi treatment of candidate genes in oma-1(zu405), and (2) regeneration of exactly the same mutations of candidate genes in oma-1(zu405) identified in EMS screens using the clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein 9 (Cas9) gene editing system. We first scored brood size for all strains that contained the APC/C related alleles from our forward genetic suppressor screen (Figure 2B). oma-1(zu405) has a very tight phenotype of absolutely no progenies that can escape the lethality phenotype at 25°C (Lin, 2003). Although not all alleles restored the brood size of oma-1(zu405) to the wild type, they all suppressed lethality of oma-1(zu405) and restored the brood size from ∼20 to ∼140 offspring per animal. When wild type worms were treated with mat-1, mat-2, mat-3, emb-30, and fzy-1 RNAi, we found that knockdown of mat-1, mat-3, and fzy-1 caused lethality of wild type worms, while knockdown of mat-2 and emb-30 only had a mild effect on the viability of wild type animals (Figure 2C). Previous studies indicate that known alleles of these genes are either lethal or temperature sensitive lethal (Furuta et al., 2000; Golden et al., 2000; Kitagawa et al., 2002), suggesting RNAi of mat-2 and emb-30 did not completely deplete functional proteins. To assess RNAi efficiency, we measured mRNA levels of corresponding genes by quantitative real-time PCR (qRT-PCR), and found that, in RNAi treated strains, mRNA levels were knocked down to 15–40% of wild type RNA levels (Supplementary Figure 1A). In addition, we treated a strain expressing OMA-1::GFP (see below) with oma-1 RNAi, and found that the OMA-1 protein level was reduced to ∼20% of the wild type protein level (Supplementary Figure 1B). We next scored brood size after treating the oma-1(zu405) strain with mat-1, mat-2, mat-3, emb-30, and fzy-1 RNAi. We found all RNAi treatments restored fertility in oma-1(zu405) to very low levels. RNAi treatment of emb-30 on oma-1(zu405) produced the largest brood size, which was only about 5 (Figure 2D).

Next, we reconstituted exact mutations in the APC/C genes identified in EMS screens by the CRISPR/Cas9 gene editing system. The mutations we regenerated included mat-1^R496C, mat-1^A580T, mat-2^A1048V, mat-3^A453V, emb-30^N58S, and fzy-1^D433N (Supplementary Figure 1C). We first generated these mutations in the wild type N2 background, and we found strains introduced by these mutations had relatively large brood sizes, from the smallest of 12 (fzy-1^D433N) to the largest of more than 175 (mat-1^R496C) (Figure 2E). Considering other known mutants of these genes are lethal, these data suggest that mutants we identified contain hypomorph alleles of APC/C genes. We crossed these CRISPR strains with oma-1(zu405) and scored brood sizes in double mutants. All regenerated mutations by CRISPR suppressed lethality of oma-1(zu405) to similar levels of EMS mutants of corresponding genes (Figure 2F). These results demonstrated that mutations in key APC/C subunits are the basis for the phenotypes observed in the mutants we isolated from the genetic screen.

It should be noted that the CRISPR generated fzy-1^D433N mutant only had a very small brood size; however, the fzy-1^D433N;oma-1(zu405) double mutant had a brood size of more than 90. This means that fzy-1^D433N and oma-1(zu405) mutually suppress each other’s lethality phenotype, which further implied that OMA-1-regulated the embryogenesis pathway and this pathway and the APC/C complex have strong genetic interactions.

Since 10 subunits and two activators of APC/C have been identified in C. elegans, we asked whether other components of APC/C suppress lethality of oma-1(zu405). We treated oma-1(zu405) with RNAi for apc-2, such-1, emb-27, apc-10, apc-11, cdc-26, and fzr-1, and then scored brood sizes. We found knockdown of apc-2, such-1, and cdc-26 restored the brood size of oma-1(zu405) from 0 to around 6, which was very similar to RNAi treatment for APC/C genes identified by genetic screens, while knockdown of other genes, including emb-27, apc-10, apc-11, and fzr-1, did not suppress oma-1(zu405) (Figures 2C,D). Considering the observation that RNAi treatment of APC/C genes we identified by genetic screen only mildly suppress oma-1(zu405), but hypomorph alleles of these genes efficiently restored viability of oma-1(zu405) (Figures 2C,D), we suspected that apc-2, such-1, and cdc-26 could also have the same genetic interactions with oma-1(zu405). RNAi of emb-27 and apc-11 caused complete lethality of wild type animals, indicating RNAi of these two genes were efficient. Although mRNA levels of fzy-1 and apc-10 were reduced to similar mRNAs levels of other APC/C genes after RNAi treatments (Supplementary Figure 1A), we cannot completely rule out that RNAi of fzr-1 and apc-10 depleted all function proteins because fzr-1 RNAi and apc-10 RNAi animals are not lethal. Among all genes that were investigated, fzy-1 and fzr-1 encoded the only two activators of the APC/C complex in C. elegans, which were homologs of yeast Cdc20 and Cdh1, respectively (Fay et al., 2002; Kitagawa et al., 2002). The main substrate of APC/C^Cdc20 is securin, which inhibits separase and prevents chromosome segregation, and the main substrate of APC/C^Cdh1 is cyclin B, which when removed can promote metaphase to anaphase during mitosis. It is interesting that only mutants of fzy-1, but not fzr-1, suppressed lethality of oma-1(zu405), indicating the APC/C^Cdc20 complex plays a particular role in the oma-1-regulated oocyte-to-zygote transition and oocyte maturation processes. Since among all mutants of APC/C genes, fzy-1^D433N restored the brood size of oma-1(zu405) the best, we next used fzy-1^D433N as the representative mutant of APC/C genes in the following studies.

To better characterize how mutations in APC/C complex suppress oma-1(zu405) lethality, we investigated transcriptome changes in fzy-1^D433N;oma-1(zu405) double mutants versus the oma-1(zu405) single mutant. For this purpose, we isolated early embryos of corresponding strains, extracted total RNA, and performed mRNA sequencing (mRNA-seq) analysis. The isolated embryos were all from synchronized early young adults, and almost all were ∼30-cell or earlier stage embryos. We compared transcriptomes of oma-1(zu405) and fzy-1^D433N;oma-1(zu405) to the wild type strain N2. Differentially expressed genes for each comparison were counted and identified by DESeq2. We identified 1,222 mRNAs that were upregulated and 4,656 mRNAs that were downregulated in the oma-1(zu405) mutant compared to the wild type N2 strain (| log₂(fold change)| > 2, p < 0.05, Benjamini-Hochberg adjusted p-value). Based on the same criteria, mis-regulated expressions of mRNAs were largely corrected to normal levels, with only 266 mRNAs that were upregulated and eight mRNAs that were downregulated in the fzy-1^D433N;oma-1(zu405) double mutant (Figure 3A). Venn diagrams showed that 192 mRNAs were commonly upregulated in both oma-1(zu405) and the fzy-1^D433N;oma-1(zu405) mutants, while only six mRNAs were commonly downregulated (Figure 3B). These mRNA-seq data indicated mutations in APC/C genes can in general restore the overall disordered transcriptome in the oma-1(zu405) mutant to the wild type.

To better understand our mRNA-seq data, we used k-means clustering to assign gene expression levels into groups, and aimed to identify which group of genes were most affected when introducing fzy-1^D433N into the oma-1(zu405) mutant. One group containing 318 genes was particularly notable, since in this group, expression levels of genes were similar in the wild type and fzy-1^D433N;oma-1(zu405) mutant, implying expression of this group of genes is best representative of the overall gene expression levels (Figure 3C). We next performed Gene Ontology (GO) enrichment analysis for these 318 genes and determined the most significant terms in biological process (BP), cellular compartment (CC), and molecular function (MF). The terms were ordered from most enriched (up) to least enriched (bottom) in each category (Figure 3D). Interestingly, the terms assigned to biological processes with the most enrichment were involved in germ cells and embryo development; the terms assigned to CCs included protein containing complexes, P granules, and cytoplasmic ribonucleoprotein (RNP) granules; and the terms assigned into CCs were involved in protein and nucleic acid binding, especially mRNA and mRNA 3′UTR binding (Figure 3D). The data from the GO enrichment analysis further implicated mutations in APC/C genes promote normal embryonic development of the oma-1(zu405) strain, which may be caused by regulating cellular distributions of OMA-1 and its downstream cell fate determinants, including RNA and transcription factor-binding proteins.

We chose 80 genes that were assigned to the terms of germ cell development, embryo development, protein binding, and RNA binding and drew a heatmap of expression levels of these genes. These genes are among the genes with the most relevant functions during OMA-1-regulated embryogenesis and have the most dramatic differential expression levels, including zif-1, pos-1, meg-1, meg-2, and skn-1. Again, the heatmap showed that introducing fzy-1^D433N could change expression levels of these genes in the oma-1(zu405) mutant to wild type (Figure 3E). Finally, we chose 23 genes from these 80 genes and performed qRT-PCR to evaluate expression levels. qRT-PCR data confirmed that changes in gene expression levels in oma-1(zu405), fzy-1^D433N;oma-1(zu405), and wild type strains were consistent with mRNA-seq data (Figure 3F).

To investigate whether APC/C mutants suppress lethality of oma-1(zu405) by regulating cellular distribution of OMA-1, we created in situ knock-in strains expressing OMA-1::GFP and OMA-1^P240L::GFP using the CRISPR/Cas9 gene editing system. Brood size of OMA-1::GFP strain was similar as the untagged strain either in wild type or oma-2 mutant background, indicating tagging green fluorescent protein (GFP) into C-terminal OMA-1 did not affect OMA-1 function (Supplementary Figure 1D). As in previous studies, OMA-1::GFP accumulated in oocytes and very early embryos and started to be degraded after the first embryonic cell division and disappeared after 4-cell embryos (Figures 4A,B). OMA-1^P240L::GFP fluorescence was still detected easily in 4-cell embryos and signals in P2 cells were stronger than those in somatic cells. OMA-1^P240L::GFP fluorescence was eventually only detectable in P4 cells (Figures 4A,B).

Next, we introduced fzy-1^D433N into OMA-1::GFP and OMA-1^P240L::GFP transgenic strains to investigate whether OMA-1^P240L::GFP is properly degraded in early embryos of the fzy-1^D433N mutant. To rule out the possibility that introducing fzy-1^D433N caused changes of overall OMA-1 protein levels, we compared OMA-1::GFP levels in wild type and fzy-1^D433N embryos at the same developmental stage. By measuring fluorescent signals within whole embryos, we found that fzy-1^D433N did not affect overall OMA-1 expression at any developmental stages we investigated (Figures 4C–E). We found fzy-1^D433N mutants affected the distribution of wild type OMA-1::GFP in early embryos (Figures 4A,B). Especially in 2- and 4-cell embryos, OMA-1::GFP was less localized into the puncta-like structure P granules (Figure 4B and see below). Quantification in the 4-cell embryos showed reduced P granule localization was statistically significant (Figure 4F). In addition, OMA-1^P240L::GFP was also altered when APC/C mutants were introduced. Although OMA-1^P240L:GFP fluorescence was still easily detected in 2-, 4-, or 20-cell embryos, its accumulation in P granules was reduced compared to OMA-1^P240L::GFP in the APC/C wild type cells (Figure 4B and see below).

Previous studies showed that OMA-1 accumulates at P granules in early embryos (Shimada et al., 2006). Considering untranslated maternal mRNAs accumulate at P granules during early embryogenesis (Noble et al., 2008), and OMA-1 binds to maternal mRNA to repress translation, it is reasonable to determine whether the APC/C complex regulates OMA-1 accumulation at P granules. We used mRuby::PGL-1 as the P granule marker, since PGL-1 is a typical component of P granules (Kawasaki et al., 1998). We first confirmed P granule accumulation of OMA-1 by showing that OMA-1::GFP and mRuby::PGL-1 were co-localized at P lineage cells in the wild type embryos (Figures 5A,C). By introducing the fzy-1^D433N mutant, as we observed above, GFP:OMA-1 accumulated less to mRuby::PGL-1 labeled P granules, and Pearson’s correlation coefficients (PCCs) also indicated the reduction of co-localization between OMA-1::GFP and mRuby::PGL-1 in fzy-1^D433N mutant (Figures 5A,C). Notably, highly enriched OMA-1^P240L:GFP in P cells was also co-localized with mRuby::PGL-1, suggesting OMA-1^P240L::GFP was indeed accumulated at P granules (Figure 5B). Surprisingly, we did not observe an obvious reduction of co-localization between OMA-1^P240L:GFP and mRuby::PGL-1 in fzy-1^D433N mutant (Figure 5C). It is possible that PGL-1 was also delocalized to P granules in OMA-1^P240L::GFP genetic background. Another possibility is that mRuby::PGL-1 is overexpressed in this strain (Yang et al., 2014), which does not fully behave like the endogenous PGL-1 under certain circumstances. All the data above indicated APC/C mutants suppress oma-1(zu405) not by promoting degradation of OMA-1^P240L proteins, but possibly by restricting localization of OMA-1^P240L at P granules in early embryos. Less accumulation of OMA-1^P240L at P granules may reflect that OMA-1^P240L bound to different groups of mRNAs in APC/C mutants.

To investigate whether APC/C complex regulate OMA-1-mRNA association, we immunopurified OMA-1::GFP by GFP nanobody and sequenced OMA-1-associated mRNAs in duplicate from young adults of oma-1:gfp, oma-1^P240L:gfp, and fzy-1^D433N;oma-1^P240L:gfp strains. We also sequenced total RNAs in corresponding strains as input controls. OMA-1-associated mRNAs were identified using DESeq2 with a cutoff of | log2(fold change)| > 1 and Benjamini-Hochberg adjusted p-value < 0.05. Eventually, 1139 mRNAs were identified as OMA-1-associated mRNAs. To validate our analysis, we referred to published OMA-1 RIP-seq data (Tsukamoto et al., 2017), and found that 808/1139 (71%) mRNAs identified in our study were identified as OMA-1-assoicated mRNAs previously (Figure 4G), indicating our OMA-1 RIP-seq analysis was reliable. Based on the same criteria, 985 and 1081 mRNAs were identified as OMA-1^P240L-associated mRNAs in the wild type and fzy-1^D433N mutant, respectively. OMA-1 and OMA-1^P240L shared 496 mRNAs (44% of OMA-1- and 50% of OMA-1^P240L-associated mRNAs), whereas OMA-1 and OMA-1^P240L in fzy-1^D433N background shared 890 mRNAs (78% of OMA-1- and 82% of OMA-1^P240L-assoicated mRNAs). Our data indicated that depletion of APC/C subunits affected OMA-1-RNA association, and turned mRNAs associated with OMA-1^P240L to be more similar to mRNAs associated with wild type OMA-1. It should be noted that numbers of mRNAs bound to OMA-1 and OMA-1^P240L were almost the same, implying that although mutation P240L altered OMA-1 binding affinity to untranslated mRNAs, the overall number of mRNAs bound to OMA-1 was largely unchanged. Developmental progresses during embryogenesis are largely determined by OMA-1 and OMA-1^P240L specifically associated RNAs.

During embryogenesis, one of the mRNA targets OMA-1/2 binds to is zif-1. Binding of OMA-1/2 represses translation of zif-1 (Guven-Ozkan et al., 2010). ZIF-1 encodes a subunit of E3 ligase whose direct substrate is PIE-1, an important cell fate determinant that segregates with the germ lineage for germ cell development during embryogenesis (Tenenhaus et al., 1998, 2001). Degradation of PIE-1 in somatic cells is delayed in oma-1(zu405) mutant embryos, possibly by delayed expression of ZIF-1. Degradation of several other maternal cell fate determinants, including MEX-1, MEX-3, MEX-5, POS-1, and MEG-1, was also delayed in oma-1(zu405) mutant embryos (Lin, 2003). To test whether APC/C mutants suppress delayed degradation of these cell fate determinants in the oma-1(zu405) strain, we investigated the distribution of CRISPR-mediated knock-in transgenic PIE-1::GFP, POS-1:GPF, and GFP::MEG-1, and a MosSCI-mediated single copy inserted transgenic GFP::MEX-3 in the oma-1(zu405) mutant and double mutants containing mutations of selected APC/C genes.

PIE-1 is a CCCH zinc finger protein that binds to the 3′UTR of maternal mRNAs to serve as an essential regulator of germ cell fate. Previous studies showed that PIE-1 is accumulated at P cells; i.e., germline precursor cells (Tenenhaus et al., 1998; Oldenbroek et al., 2013). We also observed that PIE-1:GFP asymmetrically localized at the posterior pole in 1-cell embryos so that PIE-1::GFP fluorescence in the anterior AB cell was weaker than that in the posterior P1 cell in 2-cell embryos. PIE-1::GFP further accumulated in P1, P2, and P4 cells as granule-like structures and in the nucleus, and at the same time, PIE-1:GFP in the somatic cells (i.e., non-P lineage cells) was rapidly degraded to levels below detection threshold (Figure 6A). On the other hand, in the oma-1(zu405) mutant, degradation of PIE-1::GFP in somatic cells was severely delayed, resulting in obvious nuclear located PIE-1::GFP signals in somatic cells at ∼20-cell embryos, although fluorescence in somatic cells was weaker than in P cells (Figure 6A). fzy-1^D433N, mat-1^A580T, or mat-2^A1048V single mutants did not affect the dynamic distribution pattern of PIE-1::GFP in wild type animals. However, in the oma-1(zu405) mutant, fzy-1^D433N, mat-1^A580T, or mat-2^A1048V mutants all promoted degradation of PIE-1::GFP to various levels. Specifically, somatic cell accumulation of PIE-1::GFP completely disappeared in fzy-1^D433N;oma-1(zu405) and mat-1^A580T;oma-1(zu405) double mutants, but only weak suppression of delayed degradation of PIE-1::GFP was shown in the mat-2^A1048V;oma-1(zu405) double mutant since trace fluorescence signals were still detectable in some of the soma cells (Figure 6A). It should be noted that PIE-1::GFP only accumulated at the nucleus but not at cytoplasmic granules in somatic cells. The delayed degradation of nuclear PIE-1::GFP in the oma-1(zu405) mutant was also suppressed by APC/C mutants.

Similar to PIE-1, POS-1 is another CCCH zinc finger protein and cell fate determinant for germ cells, whose proper localization is disrupted in the oma-1(zu405) mutant (Lin, 2003; Farley et al., 2008). As with PIE-1::GFP, POS-1::GFP asymmetrically accumulates in granules at the posterior pole of 1-cell embryos, and eventually accumulates at P4 in ∼20-cell embryos after several rounds of asymmetric cell division (Figure 6B). In the oma-1(zu405) mutant, POS-1::GFP was still detectable in somatic cells rather than only in P4 in ∼20-cell embryos. Introduction of fzy-1^D433N in oma-1(zu405) eliminated the remaining POS-1::GFP in somatic cells, recovering the defective distribution of POS-1::GFP to the wild type (Figure 6B).

We also investigated two other cell fate determinants; i.e., GFP::MEX-3 and GFP::MEG-1, which are essential for germ cell development. In the wild type, GFP::MEX-3 and GFP::MEG-1 only accumulated at P4 in ∼20-cell embryos after several rounds of asymmetric cell division. In the oma-1(zu405) mutant, GFP::MEX-3 and GFP::MEG-1 remained in the somatic cells, and most were cytoplasmic and a small portion accumulated at granules (Figures 6C,D). fzy-1^D433N could fully remove the abnormal somatic remaining GFP::MEX-3 and GFP::MEG-1 in the oma-1(zu405) mutant. In the fzy-1^D433N;oma-1(zu405) double mutant, localization of GFP::MEX-3 and GFP::MEG-1 behaved as in the wild type animals (Figures 6C,D).

All these results together indicated that disrupting APC/C functions could rescue delayed degradation of downstream cell fate determinants of OMA-1 caused by abnormal prolonged accumulation of OMA-1 protein during embryogenesis in the oma-1(zu405) mutant. This further suggested APC/C may regulate the binding ability of OMA-1 to RNAs or to proteins so that even OMA-1 is not degraded properly in the oma-1(zu405) mutant; it loses the ability to repress transcription or translation in early embryos. However, we cannot completely rule out the other possibility that APC/C indirectly regulates pathways that could bypass functions of OMA-1 to promote proper embryogenesis.

Animals were grown on standard nematode growth media (NGM) plates seeded with the Escherichia coli OP50 strain at 16, 20, or 25°C where indicated. Some of the C. elegans strains used in this study were purchased from the Caenorhabditis Genetics Center (CGC). Other strains were obtained through genetic screens or generated by the CRISPR/Cas9 gene editing system in this study. The strains that were used are listed in Supplementary Table 1.

RNA interference experiments were performed as described previously (Timmons and Fire, 1998). Monoclonal bacteria of HT115 (DE3) were inoculated and incubated overnight in LB medium supplemented with 100 μg/mL ampicillin at 37°C. Fresh cultures were seeded on NGM plates supplemented with 100 μg/mL ampicillin and 1 mM IPTG, and incubated at room temperature for 24 h before use. RNAi feeding bacterial strains were obtained from the C. elegans RNAi Collection (Ahringer; Source BioScience). L1–L3 stage animals were fed RNAi feed bacteria expressing corresponding double-stranded RNA at the normal culture temperature for 24 h, then shifted to 25°C for phenotype scoring. HT115 (DE3) expressing empty RNAi vector L4440 was used as the control.

Forward genetic screens were performed as described (Brenner, 1974). EMS and ENU were combined to serve as mutagens in forward genetic screens (final concentration: 47 mM of EMS and 0.98 mM of ENU). Generally, oma-1(zu405) animals were treated with mutagens and mutants that could survive at 25°C were isolated. Synchronized L4 animals of the oma-1(zu405) strain were collected and incubated with EMS/ENU mutagens for 4 h at 16°C on a head-to-toe rotator. Animals were recovered and washed with M9 buffer (22 mM KH₂PO₄, 42 mM Na₂HPO₄, and 86 mM NaCl) five times and cultured at 16°C to adults. Animals were then bleached and synchronized F1s were collected. F1 progeny were cultured at 16°C with a density of around 300 animals per plate. After laying eggs, F1s were washed away with M9 buffer, and only F2 eggs were retained on plates. F2 was transferred to 25°C after reaching the L1 or L2 stage for further culture. Plates with significant amounts of F3 and F4 progenies were retained and mutants from those plates were isolated.

We first sequenced the oma-1 locus to exclude any mutation in oma-1 coding sequences in each mutant strain. Mutants were crossed with the Hawaiian strain CB4856 and the Snip-SNP strategy was used to determine the genetic location of candidate genes, as described previously (Minevich et al., 2012). Next, genomic DNA was extracted from the mutants and oma-1(zu405) strain, and WGS was performed. WGS data from mutants and oma-1(zu405) were compared using homemade scripts. Combining the methods described above, candidate genes were determined. Reconstruction of candidate mutations were generated by the CRISPR/Cas9 gene editing system to eventually confirm the candidate genes.

Single hermaphrodite L4s (P0s) were placed onto individual freshly seeded NGM plates and allowed to grow for 24 h at 25°C. P0 adults were transferred to new NGM plates every 24 h until they no longer laid eggs. All the F1 progeny on each plate were counted. The brood size of each P0 animal was the total sum of F1s for all plates where the P0 animal laid eggs. 10–15 P0 animals were used to calculated the brood size for each strain. Data were analyzed by 1-way ANOVA followed by Dunnett’s multiple comparisons.

Single guide RNAs (sgRNAs) were designed with the online tool CHOPCHOP.¹ sgRNA sequences used in this study are listed in Supplementary Table 2. Two strategies were used to obtain sgRNA. (1) The DNA sequences encoding sgRNAs were cloned into pDD162 by overlapping polymerase chain reaction (PCR) using appropriate primers and pDD162 as the PCR template (Dickinson et al., 2013). Overlapping PCR products were inserted into pDD162 linearized with SpeI/BsrBI digestion with a ClonExpress Ultra One Step Cloning Kit (Vazyme #C115-01). (2) sgRNAs were synthesized and purified in vitro with a HiScribe Quick T7 High Yield RNA Synthesis Kit (New England Biolabs #E2050).

To generate the donor construct for gfp knock-in into the oma-1 locus, 500 bp upstream and downstream DNA of the oma-1 3′UTR and GFP coding sequences were amplified by PCR using N2 genomic DNA or plasmids containing GFP coding sequences as templates. Overlapping PCR fragments were linearized with HindIII/KpnI digestion with a ClonExpress Ultra One Step Cloning Kit (Vazyme #C115-01) and inserted into pUC19. Single-stranded oligonucleotides (ssODN) with 30 bp upstream and 30 bp downstream homology sequences and desired mutations were used as donors to introduce point mutations.

CRISPR experiments were conducted with co-CRISPR or Cas9 RNP strategies (Dokshin et al., 2018). For the co-CRISPR strategy, DNA mixtures were introduced into the germline of C. elegans young adults by microinjection. Final concentrations of plasmids in the injection mixtures were as follows: 50 ng/mL of pCCM935 unc-22 sgRNA, 50 ng/mL of pDD162 Cas9 + sgRNA, 30 ng/mL of pRF4 rol-6 (dm) and 50 ng/mL of single-stranded DNA (ssDNA) donor oligo (or 50 ng/mL of plasmid-based donor). F1 twitchers or rollers were isolated, followed by genotyping of the desired mutations. When applicable, transgenic strains were outcrossed to remove unc-22 mutations. For the RNP strategy, final concentrations of injection components were as follows: 250 ng/mL of Alt-R Cas9 protein (IDT), 100 ng/μL of Alt-R tracrRNA (IDT), and 56 ng/μL of customized Alt-R crRNA (IDT) or 200 ng/μL of customized Alt-R sgRNA (IDT). All components were mixed and incubated at 37°C for 10 min, then 110 ng/μL of ssDNA donor oligo or 200 ng/μL of dsDNA donor and 40 ng/μL of pRF4 rol-6 (dm) were added. Injection mixtures were introduced into the germline of C. elegans young adults by microinjection. The F1 rollers were isolated and genotyped to identify the desired mutations.

GFP- and mRuby-tagged fluorescent proteins were visualized in living nematodes or dissected embryos by mounting young adult animals on 2% agarose pads with M9 buffer (22 mM KH₂PO₄, 42 mM Na₂HPO₄, and 86 mM NaCl) with 10–50 mM levamisole, or mounting one-cell embryos on 2% agarose pads by dissecting gravid hermaphrodites into egg salt buffer (5 mM HEPES pH = 7.4, 118 mM NaCl, 40 mM KCl, 3.4 mM MgCl₂, and CaCl₂ 3.4 mM). Fluorescent images were captured using a Zeiss LSM800 confocal microscope with a Plan-Apochromat 63X/1.4 Oil DIC M27 objective.

The quantification of fluorescent puncta (i.e., OMA-1::GFP and OMA-1^P240L::GFP) was performed using ImageJ. Maximum intensity projections of z-series were obtained. Image thresholds were set manually and fluorescent puncta were selected. Integrated intensity of all puncta in a single embryo was measured and summed together. Images of 8–12 embryos were collected and quantified. Data were analyzed by 2-way ANOVA followed by Dunnett’s multiple comparisons.

The quantification of overall fluorescence of OMA-1::GFP and OMA-1^P240L::GFP in embryos or germlines was performed using ImageJ. Maximum intensity projections of z-series were obtained. Region of interest (ROIs) was selected and integrated intensity of all pixels within an ROI was measured. Images of 8–12 embryos or germlines were collected and quantified. For quantification of OMA-1::GFP and OMA-1^P240L::GFP in embryos, data were analyzed by 2-way ANOVA followed by Dunnett’s multiple comparisons. For quantification of OMA-1::GFP in germlines, data were analyzed by student’s t-test.

For quantitative co-localization analysis, all image manipulations were performed with ImageJ using the Coloc 2 plugin. PCCs were calculated. Data were analyzed by 1-way ANOVA followed by Dunnett’s multiple comparisons. N = 8 for each strains.

Synchronized L1 animals were cultured at 25°C until they grew into early young adults, which were then collected and washed with M9 buffer (22 mM KH₂PO₄, 42 mM Na₂HPO₄, and 86 mM NaCl) several times. Early embryos were obtained by bleaching and washing with M9 buffer five times and ice water for the last time. Total RNA was extracted using the standard method with TRIzol reagent (Invitrogen), and an mRNA library was generated for high throughput sequencing using the VAHTS Universal V8 RNA-seq Library Prep Kit for Illumina (Vazyme #NR605-01). Sequencing reads were generated with an Illumina NovaSeq 6000 system. All trimmed reads were aligned to the reference genome using HISAT2 (Kim et al., 2019). The read counts mapped to each transcript were calculated using HTSeq (Anders et al., 2015), and then normalized to reads per kilobases per million reads (RPKM). The results were used for the analysis of differentially expressed genes. All differentially expressed genes were identified using DESeq2 (Love et al., 2014) with a cutoff of | log₂(fold change)| > 2 and Benjamini-Hochberg adjusted p-value < 0.05. k-means clustering analysis of differentially expressed genes was performed using iDEP 0.93² (Ge et al., 2018). GO enrichment analysis of differentially expressed genes was performed with the GO Enrichment Analysis tool at https://www.omicshare.com/tools/.

A total of 100,000 synchronized young adult animals were frozen in liquid nitrogen and stored at −80°C. Pellets were resuspended in equal volumes of immunoprecipitation buffer [20 mM Tris–HCl pH 7.5, 150 mM NaCl, 2.5 mM MgCl₂, 0.5% NP-40, 80 U ml^–1 RNase Inhibitor (Thermo), 1 mM dithiothreitol, and protease inhibitor cocktail without EDTA (Promega)], and grinded in a glass grinder for 8–10 times. The whole grinding process should not exceed 10 min. Lysates were clarified by spinning down at 15000 rpm, 4°C, for 15 min. Supernatants were incubated with the GFP-Trap magnetic agarose beads (ChromoTek) at 4°C for 1 h. Beads were washed with IP wash buffer (20 mM Tris–HCl pH 7.5, 150 mM NaCl, 2.5 mM MgCl₂, 0.5% NP-40, and 1 mM dithiothreitol) six times, and then resuspended in TBS buffer for RNA extraction. Total RNA was extracted using the standard method with TRIzol reagent (Invitrogen). The mRNA library was generated, sequenced and analyzed as transcriptome analysis by mRNA-seq in this study described above to get RPKM values. RPKM values from IP samples and input samples were used for defining OMA-1-associated mRNAs by DESeq2 (Love et al., 2014) with a cutoff of | log₂(fold change)| > 1 and Benjamini-Hochberg adjusted p-value < 0.05.

RNA was extracted from the early embryos as described above. cDNA was then synthesized using HiScript II Q Select RT SuperMix for qPCR with gDNA wiper (Vazyme #R233-01) according to the manufacturer’s instructions. Real-time PCR was performed on a CFX Connect Thermal Cycler (Bio-Rad) with ChamQ SYBR qPCR Master Mix (Vazyme # Q311-02). Amplification was performed with a two-step reaction at 95°C for 3 min and 40 cycles at 95°C for 15 s and at 60°C for 30 s. A 20 μL PCR mixture included 10 μL ChamQ SYBR qPCR Master Mix, 2 μL primer mix (10 μM each), 3 μL diluted template cDNA, and 5 μL deionized distilled water. The relative fold changes in related genes were normalized to expression levels of actin, and analyzed by 2-way ANOVA followed by Dunnett’s multiple comparisons. Each experiment was repeated four times. The PCR primers used in this study are listed in Supplementary Table 2.