A role for BYN-1/bystin in cellular uptake and clearance of residual bodies in the Caenorhabditis elegans germline Free

GLH/Vasa/DDX4 plays a crucial role in germline development of metazoans, yet it is just one of several DEAD-box RNA helicases present in an animal's genome. Setting Vasa apart from other DEAD-box proteins are three additional domains surrounding its structured helicase: a glycine-rich repeat motif, a flanking domain wrapping around the helicase, and a negatively charged domain preceding a terminal tryptophan (Sengoku et al., 2006). In Drosophila, Vasa's helicase domain interacts with Piwi Argonaute proteins in a transient piRNA amplifier complex (Xiol et al., 2014). Similarly, in Caenorhabditis elegans, the Argonaute proteins CSR-1 and PRG-1 have been identified in GLH immunoprecipitation and proximity labeling studies, indicating that they likely form a complex analogous to Drosophila's transient piRNA amplifier complex (Chen et al., 2020; Marnik et al., 2019; Price et al., 2021). However, little is known about germline proteins that interface with the Vasa-defining peripheral domains.

In C. elegans, the paralogs GLH-1 and GLH-2 are the only DEAD-box helicases possessing all three Vasa-defining peripheral domains, and they share redundancy in maintaining the germline. glh-1 null mutants remain fertile at 20°C, displaying minor defects in sperm activation and germline expression, but at 26°C lose fertility, showing reduced germ cells and often lacking sperm (Rochester et al., 2022; Spike et al., 2008). glh-2 null mutants show wild-type fertility at 20°C and 26°C (Spike et al., 2008). In contrast, glh-2 glh-1 double mutants lose fertility at both 20°C and 26°C (Marnik et al., 2019). The redundancy among GLH proteins enables the evaluation of domain-specific contributions to germline development in fertile and living animals. Precision edits of endogenous glh-1 show that disrupting any of the three Vasa-defining peripheral domains induces temperature-sensitive sterility, suggesting that the peripheral domains of GLH-1 can be as important to germline development as its helicase activity (Marnik et al., 2019).

In this study, we used a yeast-two-hybrid (Y2H) screen to complement previous approaches in identifying GLH-1-interacting proteins. An earlier Y2H screen used only the helicase domain of GLH-1 as bait, potentially missing crucial interactions with Vasa-defining peripheral domains (Smith et al., 2002). Here, we used full-length GLH-1, to identify BYN-1, the C. elegans homolog of bystin, as one of six high-confidence interactions. Curiously, GLH/Vasa/DDX4 is found in cytoplasmic germ granules, whereas bystin is typically found in the nucleolus where it is involved in 40S ribosomal subunit assembly (Fukuda et al., 2008; Miyoshi et al., 2007; Roos et al., 1997). However, bystin's localization can also be dynamic. For instance, under nucleolar stress in human cancer cell lines, bystin mobilizes from the nucleolus into the cytoplasm (Ayala et al., 2006; Miyoshi et al., 2007). Bystin (Bysl) mRNA expression in the mouse is upregulated in activated early embryo blastocysts (Aoki et al., 2006; Hamatani et al., 2004). During embryonic implantation in humans, bystin is expressed in the trophectoderm and interacts with a protein called trophinin in the cytoplasm. This interaction modulates epidermal growth factor receptor activation, stimulating the epithelial-to-mesenchymal transition of trophectoderm into syncytiotrophoblasts that can invade the uterine wall (Fukuda and Nozawa, 1999; Fukuda and Sugihara, 2008; Nakayama et al., 2003; Sugihara et al., 2007, 2008; Suzuki et al., 1998). Disruption of bystin in mice results in embryonic lethality soon after implantation (Aoki et al., 2006), leaving the molecular and developmental mechanisms of cytoplasmic bystin–trophinin interactions unresolved. Interestingly, GLH-1 in C. elegans and trophinin in placental mammals have very similar decapeptide phenylalanine-glycine (FG)-repeat domains, suggesting potential sites of interaction with BYN-1/bystin in the cytoplasm.

Here, we show that BYN-1 and GLH-1 are subcellularly partitioned from each other through most of C. elegans germline development. The two proteins ultimately colocalize when BYN-1 is redistributed during nucleolar breakdown in spermatocytes towards the end of meiotic prophase, when it leaves the nucleus to reassemble in GLH-1-containing perinuclear germ granules. The interaction between GLH-1 and BYN-1 persists through spermatogenesis and cues the CED-1-dependent uptake of anucleate residual bodies (RBs) by overlying gonadal epithelial sheath cells. Although the extent to which RB uptake resembles cellular infiltration in cancer or human embryonic implantation has yet to be determined, our results implicate BYN-1 in a cellular uptake signaling cascade the genetics of which can be illuminated in C. elegans.

To investigate the roles of GLH-1 and its binding partners in regulating germ cell development, we analyzed 80,000,000 potential interactions in a Y2H screen using full-length GLH-1 as bait. BYN-1, the ortholog of human BYSL (bystin like), was one of six high-confidence interactions identified, appearing in six sequenced clones (Fig. 1A, Table S1). Bystin is evolutionarily conserved from yeast to humans (Fig. S1A). In Saccharomyces cerevisiae, Drosophila, HeLa cells, COS-7 cells, mouse embryo, and mouse embryonic stem cells, bystin homologs are mainly nucleolar, with roles in 18S ribosomal RNA and 40S subunit assembly (Adachi et al., 2007; Miyoshi et al., 2007; Roos et al., 1997; Stewart and Nordquist, 2005). Bystin is cytoplasmic in human trophoblasts, forming a complex with the cell adhesion molecule trophinin to mediate the initial attachment of the blastocyst to uterine epithelial cells during implantation (Sugihara et al., 2007; Suzuki et al., 1998). C. elegans does not have a homolog of trophinin, but GLH-1 and trophinin share a unique and strikingly similar glycine-rich decapeptide FG-repeat domain (Fig. 1B, Fig. S1B,C).

The FG-repeat domains of GLH proteins help distribute germ granules around the nuclear periphery by tethering to peripheral FG nucleoporins (FG-Nups) (Updike et al., 2011). Removing the FG-repeats from GLH-1 and GLH-2 interferes with the wetting-like distribution of these proteins around the nuclear periphery of primordial germ cells, resulting in germ granules that are still attached but less evenly distributed and larger and more spherical in shape (Marnik et al., 2019). A directed 1×1 Y2H assay was used to determine whether the interaction with BYN-1 is mediated through the FG-repeat domain of GLH-1 (Fig. 1C). This assay confirmed that the C-terminus of BYN-1 (aa 372-449) interacts with full-length GLH-1 (Fig. 1C, column 4) and GLH-1 ΔFG, which is missing 15 of its 19 FG-repeats from the N terminus (Fig. 1C, column 8). Increasing amounts of 3-aminotriazole, an inhibitor of the HIS3 Y2H reporter, showed that the strength of this interaction is severely compromised with the removal of these FG-repeats. However, the Y2H assay also showed that these 15 FG-repeats are not sufficient to bind BYN-1 alone (Fig. 1C, column 6). These data suggest that other GLH-1 domains are likely required to seed contact with BYN-1, but that the FG-repeat domain stabilizes the interaction.

Large-scale RNA interference (RNAi) screens in C. elegans have shown that BYN-1 is essential. Depletion of byn-1 mRNA causes phenotypes ranging from embryonic lethality, larval arrest, sterility, and aberrant germline development (Green et al., 2011, 2024; Kamath et al., 2003; Rual et al., 2004; Sönnichsen et al., 2005; Sun et al., 2011). We confirmed these phenotypes using two different byn-1 RNAi feeding constructs, one from the Vidal ORFeome-based RNAi library (Rual et al., 2004) and another directed at the byn-1-specific sequence in its N terminus (Fig. 2A). Closer examination revealed a drastic reduction in the brood size of adult animals subjected to byn-1 RNAi feeding in the first larval stage (L1) (Fig. 2B). When byn-1 RNAi feeding was initiated in the fourth larval stage (L4), brood sizes in that generation were still reduced (Fig. 2B), which may reflect a decrease in germ cell proliferation (Fig. S2A) and an alteration in physiological apoptosis at 48 h post-L4 stage (Fig. S2B,C). Embryonic progeny of these byn-1 RNAi-treated nematodes exhibited increased lethality (Fig. 2C), were often ruptured or misshapen (Fig. 2D), and showed permeability defects (Fig. 2E). byn-1 RNAi embryos also showed defects in GLH-1 partitioning to P blastomeres (Fig. 2F), developmental delays in hatched progeny, and L1-arrested development of hatched progeny in the RNAi-sensitive rrf-3 mutant background (Fig. 2G). To assess the null phenotype, CRISPR gene editing was used to create a precise byn-1 deletion over a balancer chromosome marked with myo-2::GFP (pharynx fluorescence). Homozygous animals carrying the byn-1 deletion (no green pharynx) arrested in the first larval stage, as determined by size and the GLH-1::GFP pattern (Fig. 2H). In yeast, the BYN-1 homolog has been shown to be essential for 18S rRNA production for 40S subunit assembly (Roos et al., 1997). The L1 arrest we observed is identical to mutant phenotypes of other C. elegans genes required for 18S rRNA and 40S subunit assembly and translation (Curran and Ruvkun, 2007; Dalton and Curran, 2018; Hansen et al., 2007; Pan et al., 2007; Saijou et al., 2004), suggesting a similar role for BYN-1 in 40S subunit synthesis.

To determine the subcellular localization of BYN-1 and where it might overlap with GLH-1, CRISPR/Cas9 was used to place a C-terminal mCherry::V5 tag on endogenous BYN-1 in a strain in which endogenous GLH-1 was tagged with GFP::3xFLAG (S3A). BYN-1 was expressed in most cell types of adult hermaphrodites and males (Fig. S3B), and the tag presented no deleterious effects on fertility or development (Fig. S3C,D). Live imaging of germ cells showed BYN-1::mCherry::V5 in the nucleus and enriched in nucleoli, partitioned away from GLH-1 residing on the cytoplasmic surface of the nuclear periphery (Fig. 3A, Fig. S4A,B). The partitioning of BYN-1::mCherry::V5 from GLH-1::GFP::3xFLAG is seemingly at odds with the Y2H results. However, during spermatogenesis, the nucleolus breaks down in the condensation zone during late diplotene (Shakes et al., 2009). At this stage, we observed BYN-1 to disperse throughout the nucleoplasm and then enter the cytoplasm at the karyosome stage of spermatogenesis, where it colocalized with GLH-1 (Fig. 3B,C). A putative monopartite nuclear localization signal (EKQKRKR) is found within the disordered N terminus of BYN-1, but a putative nuclear export sequence was not evident. Therefore, the entry of BYN-1 into the cytoplasm may require the initiation of nuclear envelope breakdown. Unlike in spermatogenesis, nucleoli remain intact throughout the diplotene and early diakinesis stages of oogenesis, when BYN-1 remains in the nucleolus (Fig. 3D, Fig. S4B). BYN-1 also colocalized with GLH-1 in L4 hermaphrodites as germ cells entered spermatogenesis (Fig. S4C). BYN-1 and GLH-1 remained colocalized throughout meiotic divisions until both proteins were finally packaged into the RB upon spermatid budding (Fig. 3E). These RBs are then engulfed and cleared out by the surrounding somatic gonad. Neither GLH-1 nor BYN-1 was observed in spermatids. This narrow spatiotemporal window of interaction, as evidenced by the Y2H results, likely explains why BYN-1 does not substantially co-immunoprecipitate with GLH-1 in adult hermaphrodite lysates (Fig. S5) and why BYN-1 was not identified in previous GLH-1 immunoprecipitation-qMS experiments (Marnik et al., 2019), but is able to physically bind GLH-1 when in the same cellular compartment.

To investigate further the role of BYN-1 in spermatogenesis, the stage at which BYN-1 and GLH-1 colocalize, we used RNAi to knock down byn-1 in L1-staged males and examined their DNA as adults by DAPI staining (Fig. 4, Fig. S6). Compared with control (empty L4440 vector RNAi), byn-1 RNAi males showed similar spermatogenic developmental zones and produced mature spermatids (Fig. 4A). However, GLH-1::GFP expression looked markedly different following byn-1 RNAi. GLH-1 is deposited in RBs upon spermatid budding and lost when RBs are cleared by engulfment from the surrounding somatic gonad (Huang et al., 2012). Following byn-1 RNAi, GLH-1::GFP positive RBs accumulated, pointing to defects in RB clearance (Fig. 4B). To confirm this observation, control and byn-1 RNAi males were co-immunostained with anti-phosphohistone (pH3) antibody (which marks late G2 and M phases) and anti-α-tubulin antibody (which marks spindle microtubules later deposited in RBs). The expression of pH3-positive cells in byn-1 RNAi animals was normal; however, RBs positive for α-tubulin were significantly increased in the post-meiotic spermatid region in byn-1 RNAi males (Fig. 4C). To confirm RB accumulation in byn-1 RNAi males, we dissected out sperm and stained with Annexin V, which is known to bind the phosphatidylserine (PS) expressed on the surface of RBs that serves as an ‘eat me’ signal for clearance (Huang et al., 2012). We observed an increase in RBs, all of which were Annexin-V positive, demonstrating that the PS uptake signal is present in byn-1 RNAi males but the RBs are not being engulfed (Fig. 4D). These results suggest that BYN-1 functions in RB clearance during spermatogenesis in C. elegans downstream of the PS ‘eat me’ signal. It has been reported that RB accumulation affects sperm counts and male mating efficiency (Huang et al., 2012). Indeed, compared with the control at 8 h post-L4 stage, the number of sperm in byn-1 RNAi males was significantly reduced (Fig. 4E). This reduction in sperm number was accompanied by a reduction in the mating efficiency of byn-1 RNAi males with fog-2 females (Fig. 4F). We further tested whether the abnormal RB phenotype is caused by BYN-1 in the germline or soma. We observed the RB phenotype for byn-1 using soma or germline-biased RNAi strains. RNAi depletion of byn-1 in both germline and soma-biased strains increased RB accumulation, but significantly less than in the wild-type background (Fig. 4G,H). These results suggest that RB uptake requires BYN-1 in the germline, but may also require cooperation of BYN-1 in both germline and soma. Taken together, we conclude that BYN-1 affects sperm counts and male mating efficiency by ensuring the clearance of RBs in the seminal vesicle.

There are five sheath cell pairs surrounding each gonadal arm in hermaphrodites that serve as phagocytic cells for clearing germ cell corpses during oogenesis (Gumienny et al., 1999). The gonadal sheath cells are also the main phagocytic cells that clear RBs during spermatogenesis (Huang et al., 2012). CED-1 is a transmembrane scavenger-receptor-like protein homologous to mammalian LRP1 and MEGF10, expression of which in gonad sheath cells triggers the clearance of RBs in both hermaphrodites and males. The ABC transporter CED-7, which clusters around cell corpses that are recognized by CED-1, is thought to be dispensable for RB clearance in males, whereas the parallel CED-12 pathway plays a more minor role in RB clearance (Fig. 5A) (Huang et al., 2012). To understand how BYN-1 regulates RB clearance during spermatogenesis, we observed hermaphrodites, in which RB clearance occurs during a limited time window as the germline switches from sperm to oocyte production (Fig. 5B). A CED-1::GFP transgene readily labels cells of the somatic gonad as they engulf RBs (Fig. 5C). Compared with the control, byn-1 RNAi significantly reduced CED-1::GFP-positive cells surrounding the RBs in 6 h post-L4 animals (Fig. 5C,D). In adults (24 h post-L4), some CED-1 positive cells persisted near the spermatheca of byn-1 RNAi animals (Fig. 5C, white arrows; Fig. 5E), but in the loop region, where physiological apoptosis occurs, the number of CED-1 positive cells was similar between byn-1 and control RNAi in 24 h post-L4 stage hermaphrodites (Fig. S7). These results suggest that CED-1-mediated engulfment is not increased in more distal germ cells where BYN-1 is nucleolar. Rather, CED-1-mediated engulfment increases when BYN-1 is cytoplasmic during RB clearance. byn-1 RNAi was performed in both ced-1 and ced-12 mutant backgrounds, in which the accumulation of RBs shows that their clearance is compromised. byn-1 RNAi did not further enhance RB accumulation over the control in these backgrounds (Fig. 5F-H). We also observed that the number of RBs surrounded by CED-1::GFP in males was significantly reduced by byn-1 RNAi (Fig. S8), consistent with the decrease in CED-1 activity caused by byn-1 RNAi in hermaphrodites. We further investigated whether sperm from byn-1 RNAi-treated males are fully functional or whether they exhibit other defects. Upon in vitro sperm activation with pronase, a subtle but significant decrease in pseudopod length was observed (Fig. S9). This result suggests that BYN-1 may also play a role in spermiogenesis. Altogether, these results suggest that BYN-1 contributes to RB clearance upstream of the transmembrane engulfment receptor CED-1.

The FG-repeats of C. elegans GLH-1 are most similar to the FG-repeats of the bystin-interacting protein trophinin (Fig. S1B; Suzuki et al., 1998). Trophinin-like FG-repeats are also found in the germ-granule proteins GLH-2, GLH-4, RDE-12 and DDX-19 (Sheth et al., 2010). We have previously shown the impact on germ granule size and shape when FG-repeat domains are deleted from GLH-1 (ΔFG) and both GLH-1 and GLH-2 (2ΔFG) (Fig. 6A; Marnik et al., 2019). 2ΔFG animals have reduced broods compared with wild type, but to a lesser extent than observed with byn-1 RNAi (Fig. S10A). To determine whether the FG-repeats of GLH-1 and GLH-2 help facilitate RB uptake, RB accumulation was quantified in wild-type, ΔFG and 2ΔFG animals. At both 6 and 24 h post-L4, 2ΔFG hermaphrodites showed a significant accumulation of RBs compared with the control (Fig. 6B). This increase in RB accumulation was accompanied by a significant decrease in the number of sperm (Fig. 6C). These results suggest that the FG-repeat domains of GLH-1 and GLH-2 contribute to RB clearance.

We observed that RB accumulation is higher with byn-1 RNAi than it is in 2ΔFG animals. This difference in accumulation may be due to the presence of other perinuclear FG-repeat proteins or the ability of BYN-1 to bind elsewhere on GLH-1 or GLH-2, using the FG-repeats to stabilize the association. However, RB accumulation in byn-1 RNAi animals was not enhanced by the 2ΔFG background, suggesting that the FG-repeats of GLH-1 and GLH-2 act in the same pathway as BYN-1 for RB clearance (Fig. 6D). The same conclusion can be drawn when comparing the decrease in sperm (Fig. 6E). 2ΔFG animals showed reduced colocalization between GLH-1 and BYN-1 during spermatogenesis (Fig. 6F,G), suggesting that the efficient clearance of RBs requires BYN-1 to interact with the FG-repeats of GLH proteins in these cytoplasmic foci. We also observed that the size of GLH-1::GFP granules within the condensation zone was markedly increased in 2ΔFG animals compared with the wild type (Fig. 6H), reflecting the size increase previously reported in 2ΔFG embryos (Chen et al., 2020; Marnik et al., 2019). Within RBs, BYN-1 intensity and colocalization with GLH-1 were reduced in 2ΔFG mutants compared with wild type (Fig. 6I-K). These data suggest that, in RBs, BYN-1 is stabilized by FG-repeat interactions within GLH assemblies, perhaps to cue CED-1-dependent uptake of the RBs by the surrounding somatic gonad.

To analyze further the dependence of the RB accumulation phenotype on GLH FG-repeats, we examined RB accumulation in animals carrying glh-1 and glh-2 temperature-sensitive sterile alleles that still retain their FG-repeat domains (Marnik et al., 2019). One of these strains carries a mutation that deletes the zinc-finger domain from GLH-1 but leaves its association with germ granules intact. A second strain carries a T-to-A mutation in the C-terminal helicase of GLH-1 (aa 663) to inhibit its ATPase activity, dispersing GLH-1 into the cytoplasm. The third strain carries E-to-A mutations in the DEAD-box of both GLH-1 and GLH-2 that disperses them into the cytoplasm and likely inhibits their helicase activity. RB accumulation phenotypes were not observed in these three strains (Fig. S10B,C), pointing to the significance of BYN-1–FG-repeat interactions for RB clearance during spermatogenesis.

The phenotypic analysis of byn-1 RNAi animals suggests that the activity of BYN-1 in the nucleolus is crucial for cell proliferation and survival, whereas its activity in the cytoplasm mediates cell engulfment. These observations provide added insight into a dual function for BYN-1. By imaging cells in live animals, the localization and expression of BYN-1 during germline differentiation revealed previously unseen dynamic processes that drive RB clearance during spermatogenesis. BYN-1 is localized in the nucleus and enriched in nucleoli, partitioned away from GLH-1 and the FG-repeat-rich germ granules residing on the cytoplasmic surface of the nuclear envelope (Fig. 7A). In the diplotene stage of spermatogenesis, as nucleoli disassemble, BYN-1 first disperses throughout the nucleoplasm and then enters the cytoplasm where it encounters GLH-1. As spermatids are formed, BYN-1 and GLH-1 are partitioned into anucleate RBs and engulfed by the surrounding somatic gonad. The correlation of these activities suggests that the translocation of BYN-1 and colocalization with FG-repeat domain-containing GLH proteins triggers cell engulfment.

For comparison, during oogenesis nucleoli remain intact through the diplotene stage and do not disassemble until the end of diakinesis. As a result, nucleolar BYN-1 is retained in maturing oocytes until after most germ granules disperse. In addition, the activity of CED-1 was not significantly affected by the depletion of byn-1 during physiological apoptotic events in oogenesis. These results suggest that BYN-1 does not function in cell migration and engulfment while it is retained in nucleoli. Therefore, BYN-1-mediated engulfment is likely dependent on its subcellular location. BYN-1 does not appear to encounter GLH-1 during oogenesis as it does during spermatogenesis. It will be important to look closer at the timing of these events, as the cytoplasmic colocalization of BYN-1 and GLH-1 initiates at the karyosome stage, and cell engulfment takes place moments later when spermatids bud. Determining whether engulfment can occur in the absence of budding will provide insight into the activation time and checkpoints of CED-1 pathways.

The results presented in this study point to BYN-1 acting upstream of CED-1 to promote programmed cell death. Interestingly, BYN-1/bystin is found in a rare syntenic group of tightly clustered orthologous genes present from single-cell eukaryotes to humans (Trachtulec and Forejt, 2001). The syntenic group includes programmed cell death 2 (PDCD2) and Unc93, an inducer of apoptosis, which might indicate a link between bystin and cell death programs. In hepatocellular carcinoma cells, BYSL is crucial for cell proliferation in vitro and in vivo; BYSL depletion increases apoptosis and suppresses nucleoli assembly after mitosis (Wang et al., 2009). Similarly, our study showed that byn-1 RNAi also decreased germline stem cell proliferation, and increased physiological germ cell apoptosis in 48 h post L4-staged worms (Fig. S2B). It is unclear how the cytoplasmic interaction between BYN-1 and GLH-1 promotes RB internalization. In both trophoblasts and glioblastoma cells, cytoplasmic bystin promotes an epithelial-to-mesenchymal transition to mobilize cells (Sha et al., 2020; Sugihara et al., 2007). By extension, BYN-1 may help to mobilize or position RBs for uptake by the somatic gonad (Fig. 7B).

FG-repeat domains have been extensively studied within the context of the nuclear pore complex (NPC). In the NPC, unstructured FG-repeats extend into the central channel to establish a size-exclusion barrier between the nucleus and cytoplasm (Kim et al., 2018; Yamada et al., 2010). This size exclusion barrier is created by regularly spaced phenylalanines that interact to form a mesh or smart sieve (Schmidt and Görlich, 2015). Proteins smaller than 40 kDa diffuse through that sieve, whereas larger proteins and complexes must attach to a karyopherin to facilitate nucleocytoplasmic transport. Of the 15 longest FG-repeat proteins in C. elegans, just six are components of the NPC (Shinkai et al., 2021). Another five (GLH-1, GLH-2, GLH-4, RDE-12 and DDX-19) are found in germ granules that cluster on the cytoplasmic surface of germ cell NPCs (Sheth et al., 2010), and another (PQN-75) is expressed in pharyngeal gland cells (Rochester et al., 2017). We previously showed that FG-repeat-rich germ granules extend the size exclusion barrier of NPCs further into the cytoplasm and, like FG-Nups, are disassembled with low concentrations of 1,6-hexanediol (Updike et al., 2011). GLH-containing granules can be detached from the periphery by removing their FG-repeat domains or by RNAi depletion of FG-Nup-encoding mRNAs (Marnik et al., 2019; Updike and Strome, 2009; Voronina and Seydoux, 2010).

In humans, seven of the 15 longest FG-repeat proteins are components of the NPC (Shinkai et al., 2021). Another seven are associated with keratin formation in the epidermis. The remaining protein is trophinin, which boasts the highest number of FG-repeats in the human proteome. Trophinin was initially shown to be, and primarily thought of as, an apical membrane protein in trophoblasts, but curiously it lacks the signal sequence for insertion into the plasma membrane. In fact, studies in which human trophoblastic teratocarcinoma HT-H cells were permeabilized to improve trophinin antibody accessibility, as well as COS-1 cells expressing recombinant trophinin, show the protein localizing to the cytoplasmic surface of the nuclear periphery (Aoyama et al., 2005). The distribution of cytoplasmic trophinin at the nuclear envelope is reminiscent of the perinuclear distribution of GLH proteins in C. elegans germ cells (Aoyama et al., 2005). By extension, trophinin may also be using its FG-repeats to extend the NPC size exclusion barrier into the cytoplasm of trophoblasts.

The cytoplasmic location of trophinin around the nuclear periphery could be where bystin first encounters its FG-repeats during embryonic implantation, mirroring the subcellular location where BYN-1 encounters GLH-1. Trophinin homologs are not found outside of placental mammals, and the protein itself could have resulted from a fusion between a MAGE domain and an FG-repeat protein. Fusions of the FG-repeats of Nup98 are commonly found in acute myeloid leukemias and include fusions to DDX10, a DEAD-box helicase similar to GLH/DDX4 (Arai et al., 1997; Gough et al., 2011). However, the FG-repeats of trophinin are more similar to those of nematode GLH homologs than those of known mammalian FG-Nups (Fig. S1C). The affinity of BYN-1 for FG-repeat proteins may reflect an evolutionarily conserved function. However, considering the rapid evolution of sperm genes, the interaction between BYN-1 and GLH-1 could also represent a novel, sperm-specific role that evolved in C. elegans. In either scenario, C. elegans serves as a valuable model for deciphering the developmental, genetic and molecular events that occur when BYN-1/bystin encounters cytoplasmic FG-repeats. The dual-functional nature of BYN-1 appears to be determined and activated by subcellular localization, and C. elegans development provides a novel context for evaluating and discriminating these functions. Observations presented in this study suggest that efficient RB uptake and clearance may require BYN-1 in both the germline and soma, making it essential to determine whether BYN-1 binds other FG-repeat proteins in somatic cells where GLH proteins are not expressed. In doing so, the roles played by bystin, trophinin, and FG-repeat fusions in both embryonic implantation and cancer will become more apparent.

All strains of C. elegans were maintained and cultured on Nematode Growth Medium (NGM) agar plates at 20°C with standard protocols (Brenner, 1974). N2 (C. elegans wild isolate, Bristol variety) strain was used as wild-type control for this study. The following strains were created for this study:

DUP261: glh-1(sam24[glh-1::gfp::3xFlag]) I; byn-1(sam159[byn-1::mCherry::V5]) III.

DUP263: glh-2(sam83[ΔFGglh-2]) glh-1(sam72[ΔFGglh-1::gfp::3xFlag]) I; byn-1[sam159(byn-1::mCherry::V5)] III.

DUP313: byn-1[sam159(byn-1::mCherry::V5)] III.

DUP315: glh-1(sam24[glh-1::gfp::3xFlag])/hT2g [bli-4(e937) let-?(q782) qIs48] I; byn-1*sam188([Δbyn-1::mCherry::V5])/hT2 [bli-4(e937) let-?(q782) qIs48] III.

GenBank-formatted sequence files for the two CRISPR-generated byn-1 alleles, sam159 and sam188, are available on Figshare (https://doi.org/10.6084/m9.figshare.26132659.v1).

Strains that have been described previously (Marnik et al., 2019) were:

DUP64: glh-1(sam24[glh-1::gfp::3xFlag]) I.

DUP163: glh-1(sam72[ΔFGglh-1::gfp::3xFlag]) I; itIs37[pie-1p::mCherry::H2B::pie-1 3′UTR, unc-119(+)] IV.

DUP165: glh-2(sam83[ΔFGglh-2]) glh-1(sam72[ΔFGglh-1::gfp::3xFlag]) I; itIs37[pie-1p::mCherry::H2B::pie-1 3′UTR, unc-119(+)] IV.

DUP143: glh-1(sam64[T663Aglh-1::gfp::3xFlag]) I.

SS940: bnIs1(pie-1p::GFP::PGL-1) I; fog-2(q71) V.

DUP151: glh-1(sam71[ΔZFglh-1::gfp::3xFlag]) I.

DUP170: glh-2(sam89[DEADtoDAADglh-2) glh-1(sam90[DEADtoDAADglh-1::gfp::3xFlag]) I/hT2[bli-4(e937) let-?(q782) qIs48] (I;III).

Strains obtained from the CGC were:

DCL569: mkcSi13 [sun-1p::rde-1::sun-1 3′UTR+unc-119(+)] II; rde-1(mkc36) V.

NL3511: ppw-1(pk1425) I.

NL2099: rrf-3(pk1426) II.

CU1546: smIs34[ced-1p::ced-1::GFP+rol-6(su1006)].

CB3203: ced-1(e1735) I.

MT11068: ced-12(n3261) I.

CRISPR/Cas9 editing was performed as described by Ghanta and Mello (2020). byn-1[sam159(byn-1::mCherry::V5] was created using the 20 bp crRNA AGGCCGAAAGACGGATAGTA and a homologous repair (HR) template to insert mCherry::V5 in-frame before the stop codon. byn-1[sam188(Δbyn-1::mCherry::V5)] was created using a left-flanking crRNA ATTCCTTTTCACCATGTTAC, a right-flanking crRNA ATAAATAGCTGCCTTCTGCT, and an HR template GTAATTCTGCTAATTTTTAATTATGAATTTCCAGTAGGCAGCTATTTATGAGCTGATCAGAGTAAGTT to create a 1608 bp deletion in the byn-1[sam159(byn-1::mCherry::V5] background, starting 2 bp upstream of the ATG start codon through most of the ninth exon, also turning off any mCherry expression. This deletion does not disrupt F57B9.8, which is encoded on the opposite strand of the last intron of byn-1 (Fig. S3A).

ULTImate Y2H™ screening was performed by Hybrigenics Services, S.A.S., Evry, France (http://www.hybrigenics-services.com). The coding sequence for full-length C. elegans glh-1 (NM_059562.6) was PCR-amplified and cloned into pB29 as an N-terminal fusion to LexA (glh-1-LexA). The construct was checked by sequencing the entire insert and used as a bait to screen a random-primed C. elegans mixed-stages cDNA library constructed into pP6. pB29 and pP6 derive from the original pBTM116 and pGADGH plasmids, respectively (Béranger et al., 1997; Vojtek and Hollenberg, 1995). Eighty million clones (eightfold the complexity of the library) were screened using a mating approach with YHGX13 (Y187 ade2-101::loxP-kanMX-loxP, matα) and L40ΔGal4 (mata) yeast strains as previously described (Fromont-Racine et al., 1997) and 232 His⁺ colonies were selected on a medium lacking tryptophan, leucine and histidine. The prey fragments of the positive clones were amplified by PCR and sequenced at their 5′ and 3′ junctions. The resulting sequences were used to identify the corresponding interacting proteins in the GenBank database (NCBI) using a fully automated procedure. A confidence score (PBS, for Predicted Biological Score) was attributed to each interaction as previously described (Formstecher et al., 2005). The PBS relies on two different levels of analysis. First, a local score takes into account the redundancy and independency of prey fragments, as well as the distribution of reading frames and stop codons in overlapping fragments. Second, a global score takes into account the interactions found in all the screens performed at Hybrigenics using the same library. This global score represents the probability of an interaction being nonspecific. For practical use, the scores were divided into four categories, from ‘A’ (highest confidence) to ‘D’ (lowest confidence). A fifth category (‘E’) specifically flags interactions involving highly connected prey domains previously found several times in screens performed on libraries derived from the same organism. Finally, several of these highly connected domains have been confirmed as false positives of the technique and are now tagged as ‘F’. The PBS scores have been shown to positively correlate with the biological significance of interactions (Rain et al., 2001; Wojcik et al., 2002). Of the 232 processed clones, 117 received high (‘B’) or very high (‘A’) PBS scores in fragments encoded by six different genes (Table S1). Six of the 117 corresponded to a C-terminal fragment (aa 377-449) of BYN-1.

The coding sequences for the fragments containing amino acids 1-152 and 153-763 of C. elegans glh-1 (NM_059562.6) were PCR-amplified and cloned into pB29 as N-terminal fusion to LexA (glh-1-fragments-LexA). The constructs were checked by sequencing the entire inserts. The fragment corresponding to amino acids 372-449 of byn-1 was extracted from the ULTImate Y2H™ screening of full-length glh-1 with the C. elegans mixed-stages library. It was cloned in-frame with the Gal4 activation domain (AD) into plasmid pP6. The AD construct was checked by sequencing. Bait and prey constructs were transformed in the yeast haploid cells L40ΔGal4 (mata) and YHGX13 (Y187 ade2-101::loxP-kanMX-loxP, matα), respectively. The diploid yeast cells were obtained using a mating protocol with both yeast strains (Fromont-Racine et al., 1997). These assays are based on the HIS3 reporter gene (growth assay without histidine). As negative controls, the bait plasmids were tested in the presence of empty prey vector (pP6) and the prey plasmid was tested with an empty bait vector that allows the expression of LexA (pB27). The interaction between SMAD and SMURF was used as positive control (Colland et al., 2004). Controls and interactions were first tested in the form of streaks of three independent yeast clones on DO-2 and DO-3 selective media. The DO-2 selective medium lacking tryptophan and leucine was used as a growth control and to verify the presence of the bait and prey plasmids. The DO-3 selective medium without tryptophan, leucine and histidine selects for the interaction between bait and prey. Interaction pairs were then tested in duplicate in the form of calibrated drops as two independent clones were picked for the growth assay. For each interaction, several dilutions (undiluted, 10^–1, 10^–2, 10^–3) of the diploid yeast cells (culture normalized at 5×10⁷ cells) and expressing both bait and prey constructs were spotted on several selective media. The DO-2 selective medium lacking tryptophan and leucine was used as a growth control and to verify the presence of both the bait and prey plasmids. The different dilutions were also spotted on a selective medium without tryptophan, leucine and histidine (DO-3). Four different concentrations of 3-aminotriazole (1, 5, 10 and 50 mM), an inhibitor of the HIS3 gene product, were added to the DO-3 plates to increase stringency.

Depletion of byn-1 was performed using the bacteria-mediated (feeding) method as described previously (Kamath et al., 2003). Escherichia coli strain HT115(DE3) containing F57B9.5 (byn-1) inserted L4440 vector was obtained from the Vidal library (Rual et al., 2004). Sequencing of this first byn-1 RNAi clone revealed it contains a 139 bp insertion targeting part of exon 9 and all of exon 10. The second RNAi clone, pDU103, was generated by inserting 270 bp of the byn-1 mRNA sequence (overlapping the first two exons) into HindIII sites of the L4440 plasmid, which was then re-transformed into HT115(DE). L4440 in HT115(DE3) was used as the RNAi control. Synchronized L1 or L4 larval-stage worms were placed on feeding RNAi plates and were incubated at 20°C for phenotypic analysis.

Three-hundred adult hermaphrodites were collected in TRIzol (Invitrogen) for each condition, and total RNA was extracted using a phase lock gel (MaXtract High Density, QIAGEN). cDNA was synthesized using oligo-dT primer and M-MLV reverse transcriptase (Invitrogen). qRT-PCR assays were performed using SYBR Green PCR Master Mix (Applied Biosystems). The final PCR volume was 10 µl. act-1 mRNA was used as an endogenous control for data normalization. The primers used for the measurement of expression of the byn-1 gene were as follows: byn-1 forward, 5′-AAGATGCTATCCAAATGTCCAC-3′; byn-1 reverse, 5′-GTCACGCATATTAAACTCATTCGG-3′; act-1 forward, 5′-CCAGGAATTGCTGATCGTATGCAGAA-3′; act-1 reverse, 5′-TGGAGAGGGAAGCGAGGATAG-3′.

Synchronized L4-stage hermaphrodites were moved to individual plates seeded with control RNAi or byn-1 RNAi bacteria and allowed to lay eggs for 24 h at 20°C. Each worm was transferred to a new plate every 24 h for 4 days. The number of progeny was calculated as the total number of non-hatched and hatched embryos produced by a single mother. Embryonic lethality was calculated as the percentage of non-hatched embryos out of the total number of embryos produced for 4 days. To investigate eggshell integrity, membrane permeability was assessed using FM1-43 dye (Sigma-Aldrich), as previously described (Nakayama et al., 2009). In brief, either control or byn-1 RNAi-treated mothers were dissected in 150 mM KCl with 30 mM of FM1-43 dye to observe embryos. The proportion of embryos infiltrated by FM1-43 was measured using a Leica Thunder Imager Live Cell microscope with an HC PL APO 40×/0.95 DRY objective. For each case, three independent experiments were performed. Percentage larval development was calculated as the percentage of larvae out of the total number of hatched embryos that reached each developmental stage, as previously described (Min et al., 2015).

Fluorescence microscopy and image processing were performed as previously described (Rochester et al., 2022; Spaulding et al., 2022). In brief, GFP- and mCherry-tagged fluorescent proteins were visualized in living worms or dissected embryos by mounting on 2% agarose pads with M9 buffer with 10 mM levamisole. Live fluorescent images were captured using a point scanning confocal unit (LSM980, Carl Zeiss Microscopy) on a Zeiss Axio Examiner Z1 upright microscope stand equipped with a Zeiss Plan Apo 63×/1.4 NA Oil (420782-9900-799) objective and Immersol 518F Immersion Oil. sGFP and mCherry fluorescence were excited with the 488 nm diode (0.2% laser power) and the 561 nm DPSS laser (2% power), and collected using Airyscan2 with the following detection wavelengths: sGFP (499-548 nm) and mCherry (574-627 nm). Images were sequentially acquired in Super Resolution mode (SR) at zoom 10, with a line average of 1, a resolution of 292×292 pixels, 0.043×0.043 μm pixel size, a pixel time of 0.69 μs, in 16-bit, and in bidirectional mode. z-stack images were collected with a step size of 0.150 μm, 3.3 μm deep, with the Motorized Scanning Stage 130×85 PIEZO (Carl Zeiss Microscopy) mounted on Z-piezo stage insert WSB500. Airyscan images were processed using the ‘auto’ mode in Zen Blue Software (Zen Pro 3.1) and saved in CZI format. Live image processing and quantification of fluorescent proteins were performed using Fiji and ImageJ. Single-slice images of gonads and images from maximum intensity projections of z-series of gonads were used for quantification. Image thresholds were set manually, regions of interest were selected, and areas of regions of interest were measured. To quantify the colocalization of GLH-1 and BYN-1 during sperm differentiation, we used ImageJ with the plugin JACoP to calculate Pearson's colocalization coefficient. To analyze the intensity of BYN-1 in RBs, the measure tool was applied to measure integrated density. Mean pixel intensity was subtracted from the background to calculate the corrected total cell fluorescence [CTCF=integrated density – (area of cell×mean background fluorescence)] using ImageJ. Fixed immunofluorescent images and FM1-43 staining were observed on a Leica Thunder Imager Live Cell microscope with an HC PL APO 63×/1.47 Oil CORR TIRF objective and DAPI, GFP and TXR filter sets. Images were acquired with a Leica DFC9000 CT deep-cooled sCMOS camera and Leica LAS X imaging software. To observe the progression of spermatogenesis in males, worms were dissected and fixed with 100% methanol at −20°C for 10 min, followed by 100% acetone fixation at −20°C for 10 min. The specimens were further counterstained with DAPI to stain DNA. The following primary and secondary antibodies were used in this study: mouse monoclonal anti-a-tubulin (1:500; Sigma-Aldrich, T9026), rabbit anti-phospho-histone H3 (Ser10) (1:500; EMD Millipore, 06-570), Alexa Fluor 488-conjugated goat anti-mouse IgG (1:500; Invitrogen, A32723), Alexa Fluor 594-conjugated goat anti-rabbit IgG (1:500; Invitrogen, A32740).

Annexin V staining and quantification of RBs were performed as previously described with some modifications (Huang et al., 2012). In brief, single males were placed in 3×3 drops (15 μl) of sperm medium (SM) buffer (50 mM HEPES, 25 mM KCl, 45 mM MgSO₄, 5 mM CaCl₂ and 10 mM dextrose, pH 7.8) on GCP-coated slides [0.2% gelatin (USP), 0.002% chrome alum (Sigma-Aldrich), 1 mg/ml poly-L-lysine (Sigma-Aldrich)]. Males were dissected in a drop of SM buffer containing 0.5 µl Alexa Fluor 488-conjugated Annexin V (Molecular Probes) on a GCP slide to release RBs and sperm and incubated in a wet chamber for 20 min at room temperature. A coverslip was gently overlaid and sealed with Vaseline. The samples were observed under Nomarski differential interference contrast (DIC) microscopy, and RBs with Annexin V were counted using a Leica DFC9000 GT microscope. In hermaphrodites, RBs were counted in the proximal gonad arm at two different time points (6 h and 24 h post-L4 stage) using Nomarski optics.

Spermatid quantification assay was performed as previously described with minor modifications (Huang et al., 2012). Briefly, a single hermaphrodite or a single male (5 h post-L4 stage) was picked into a small drop of fixation buffer on a slide and the gonad dissected to release sperm. The sample was then fixed with 4% paraformaldehyde for 20 min at room temperature, washed three times with PBST (1.4 M NaCl, 0.03 M KCl, 0.1 M Na₂HPO₄, 0.02 M KH₂PO₄, 0.2% Tween 20, pH 7.2), and further fixed with 70% ethanol at 4°C overnight. Then, the sample was washed twice with PBST and stained with DAPI. Sperm nuclei were counted in each gonad arm, and at least ten worms were scored for each genotype.

Synchronized adult males were mated with fog-2 females in a 3:1 ratio for 6 h. Mated fog-2 females were subsequently transferred to new NGM plates, and the fertilized eggs in the uterus were quantified.

Anti-FLAG-M2 magnetic beads (Sigma-Aldrich, M8823) were resuspended by gentle inversion, washed in cold TBS three times and left in TBS on ice during the following sample preparation. Approximately 400-500 µl of synchronized young adult DUP64, DUP261 and DUP313 worms were washed off plates with dH₂O and flash-frozen in liquid nitrogen. Frozen worm pellets were ground into a fine powder using a mortar and pestle in liquid nitrogen. The powder was resuspended in 1 ml of lysis buffer [25 mM HEPES-KOH, pH 7.4, 10 mM potassium acetate (KOAc), 2 mM Mg(OAc)₂, 100 mM KCl, 0.25% Triton X-100, 1 mM PMSF, one cOmplete protease inhibitor tablet]. Samples were spun at 12,000 g for 5 min at 4°C then 1 ml of supernatant was added to 100 µl of beads and rotated at 4°C overnight. Beads were washed three times with 1 ml of cold TBS. Protein was eluted by incubating beads in 250 µl of 0.1 M glycine HCl, pH 3.0, at room temperature for 5 min with gentle shaking. Eluate was transferred to fresh tubes containing 10 μl of 0.5 M Tris-HCl, pH 7.4, with 1.5 M NaCl. Elution was repeated and the second eluate was added to the same tube (total eluate is 500 µl). Eluate was added 1:1 to 2× Laemmli buffer (10% BME, 4% SDS, 20% glycerol, 0.004% Bromophenol Blue, 0.125 M Tris-Cl pH 6.8) and boiled for 5 min. Then, 40 μl of sample was loaded onto a mini-PROTEAN TGX stain-free gel (Bio-Rad, 4568084) and run at 200 V for 25 min in SDS running buffer (0.2501 M Tris base, 1.924 M glycine, 0.0347 M SDS). The contents of the gel were transferred using a trans-blot turbo transfer pack (Bio-Rad, 1704156) on the Bio-Rad Trans-Blot Turbo Transfer System. The PVDF membrane was blocked in 5% nonfat milk in TBST (20 mM Tris-HCl pH 7.4, 150 mM NaCl, 0.1% Tween 20) at room temperature for 1 h. The membrane was incubated overnight at 4°C with mouse monoclonal anti-V5 (Thermo Fisher Scientific, R960; 1:5000) and mouse monoclonal anti-FLAG M2 (Sigma-Aldrich, F1804; 1:5000) in 5% milk and then washed six times, 10 min each wash, in TBST at room temperature. The membrane was then incubated in goat anti-mouse IgG-HRP (Invitrogen, 31430; 1:10,000) in 5% milk for 1 h at room temperature and washed six times, 10 min each wash, in TBST. Finally, the membrane was developed using Clarity Western ECL Substrate (Bio-Rad, 1705060S) and imaged on a Syngene G:Box gel and blot imaging system.

To observe the mitotic germ cells at the distal region of gonad, worms were dissected and fixed with 100% methanol at −20°C for 10 min, followed by 100% acetone fixation at −20°C for 10 min. The specimens were immunostained using rabbit anti-phospho-histone H3 (Ser10) (1:500; EMD Millipore, 06-570) and Alexa Fluor 594-conjugated goat anti-rabbit IgG (1:500; Invitrogen, A32740).

Germ cell corpses were counted under Nomarski DIC microscopy at 24 h and 48 h post-L4 stages of hermaphrodites as previously described (Gumienny et al., 1999). Apoptotic germ cells were visualized by Acridine Orange vital staining as previously described (Navarro et al., 2001). Briefly, 24 h and 48 h post-L4 stage worms were stained with 25 μg/ml Acridine Orange in M9 buffer for 1 h in the dark, allowed to recover on new NGM plates for 30 min and observed under fluorescence microscopy to count the number of Acridine Orange-positive germ cells per gonad arm. Only one gonad arm was scored for each observed animal, and 15-20 animals were examined for each condition.

Phalloidin staining was performed as described previously with modifications (Huang et al., 2012). Males were fixed with 4% paraformaldehyde at room temperature for 20 min and further fixed with 70% ethanol. The samples were rinsed three times for 10 min each in PBST. The samples were incubated with rhodamine-conjugated phalloidin (Molecular Probes, Life Technologies) for 2 h, and then washed with PBST for 10 min. Fixed fluorescent images were observed on a Leica Thunder Imager microscope with an HC PL APO 63×/1.47 Oil CORR TIRF objective. Images were acquired with a Leica DFC9000 CT deep-cooled sCMOS camera and Leica LAS X imaging software.

Statistical analysis was performed using GraphPad Prism 9.3.1 software. P-values lower than 0.05 were considered statistically significant. Data graphs were plotted in GraphPad Prism as scatter dot plots with error bar or bar graphs displaying mean±s.d. P-values are represented by asterisks and n values for each experiment are denoted in the corresponding figure panel, figure legend or bar graph.

We acknowledge support from Chris Smith in the MDI Biological Laboratory (MDIBL) Sequencing Facility for Sanger sequencing services; Frederic Bonnet in the MDIBL Light Microscopy Facility for training, assistance and use of super-resolution image acquisition; and plate pouring services provided by the MDIBL Animal Resources Core. MDIBL Cores and Facilities receive support from the National Institutes of Health [Maine INBRE GM103423, MDIBL COBRE GM104318]. Some of the strains for this study were provided by the CGC, funded by National Institutes of Health Office of Research Infrastructure Programs OD010440.