Abstract
Gap-junctional signaling mediates myriad cellular interactions in metazoans. Yet, how gap junctions control the positioning of cells in organs is not well understood. Innexins compose gap junctions in invertebrates and affect organ architecture. Here, we investigate the roles of gap-junctions in controlling distal somatic gonad architecture and its relationship to underlying germline stem cells in the nematode Caenorhabditis elegans. We show that a reduction of soma-germline gap-junctional signaling causes displacement of distal sheath cells (Sh1) towards the distal end of the gonad. We show that a somatically expressed innexin fusion protein, which was used as marker in a prior study asserting that the wild type lacked a bare region between the distal tip cell (DTC) and Sh1, encodes a poisonous gap junction subunit. We determine that, contrary to the model put forth in the prior study based on this marker, Sh1 mispositioning does not markedly alter the position of the borders of the stem cell pool or of the progenitor cell pool. Together, these results demonstrate that gap junctions can control the position of Sh1, but that Sh1 position is neither relevant for GLP-1/Notch signaling nor for the exit of germ cells from the stem cell pool.
Introduction
The relative positions of certain cells within larger organ structures are often important for organ function. Yet the mechanisms by which cells reach and maintain their precise relative positions within organs are poorly defined. Gap junctions act as conduits for small molecules passed between cells and/or as rivets to ensure adhesion between cells (reviewed by (Skerrett and Williams, 2017)). They have also been implicated in cell morphology within organs (reviewed by (Phelan, 2005)), however this latter role is less well characterized. Here we take advantage of well-characterized and stereotypical morphology, interactions and relationships among cells in Caenorhabditis elegans to investigate the role gap junctions play in somatic gonad architecture and its consequences for the underlying germ line stem cells.
The C. elegans hermaphrodite gonad provides a premier system for studying organogenesis and stem cell behavior (reviewed by (Hubbard and Greenstein, 2000; Hubbard and Schedl, 2019)). Two gonad arms, anterior or posterior of a central uterus and vulva, are each capped by a single somatic cell, the distal tip cell (DTC) that establishes a stem cell niche (Figure 1). Germline stem cells and their proliferative progeny, which together are referred to as progenitors, are maintained by GLP-1/Notch mediated signaling in the germ line in response to DSL family ligands LAG-2 and APX-1 produced by the DTC (Austin and Kimble, 1987; Berry et al., 1997; Henderson et al., 1994; Nadarajan et al., 2009; Yochem and Greenwald, 1989). Proximal to the DTC, five pairs of sheath cells (named as pairs Sh1 to Sh5, distal to proximal) provide additional support (DTC and Sh1 shown in Fig. 1). In particular, Sh1 is implicated in promoting germline progenitor cell proliferation (Killian and Hubbard, 2005; McCarter et al., 1997). Although the molecular and cellular mechanisms by which Sh1 promotes germline proliferation remain to be fully elucidated, it is clear that one mechanism for the function of these cells involves the formation of gap junctions with germ cells (Starich et al., 2014).
Invertebrate gap junctions are formed from octameric hemichannels of innexin proteins (Oshima et al., 2016). In C. elegans, INX-8 and INX-9 associate to form hemichannels in the hermaphrodite somatic gonad which couple to germline innexin hemichannels (INX-14 with INX-21 or INX-22) to promote germline proliferation and inhibit meiotic maturation, respectively (Figure 1A-B; (Starich et al., 2014)). Phenotypic analysis of reduction-of-function mutants in inx-8 recently led to the discovery of malonyl-CoA as a key cargo that traverses the soma-germline junction to ensure timely gametogenesis and proper embryogenesis (Starich et al., 2020) .
In the distal gonad, the somatic gonadal hemichannel components inx-8 and inx-9 are required redundantly for germ cell proliferation and differentiation. Loss of both components renders the germline devoid of all but a handful of germ cells, which fail to undergo gametogenesis. Restoration of inx-8 either to the DTC or to Sh1 rescues the severe germline proliferation defect of the inx-8(0) inx-9(0) double mutant, while a reduction of inx-8 and inx-9 via hypomorphic alleles or by RNAi limits expansion of the pool of proliferative germ cells (Dalfo et al., 2020; Starich and Greenstein, 2020; Starich et al., 2014).
Several observations point to a role for innexins in overall somatic gonad architecture. In young adult hermaphrodites, the DTC forms long extending processes reaching proximally towards Sh1, while the distal border of Sh1 is more regular, with filopodia extending distally towards the DTC. Extensive ultrastructural and both fixed and live image analysis demonstrated the existence of a “bare region” in the adult hermaphrodite gonad in which germ cells are covered only by a basal lamina in the region between the proximal extending DTC processes and the distal extending filopodia of Sh1 ((Hall et al., 1999); Figure 1). Interestingly, the hypomorphic allele inx-14(ag17) (Miyata et al., 2008) causes Sh1 to reach almost to the distal end of the gonad (Starich et al., 2014), obliterating the bare region between the DTC and Sh1. Similarly, a loss of the bare region was observed in inx-8(0) inx-9(0) double mutants in which germline proliferation was largely restored through expression of an inx-8::gfp transgene in the DTC only (Starich et al., 2014). This latter result suggested that if Sh1 cannot form gap junctions with germ cells, it extends distally. However, the consequences of this mis-positioning and the accompanying loss of the bare region for germline stem cells has not been previously explored.
This spatial relationship between the DTC, Sh1 and the germ line was recently challenged, and an hypothesis put forth that Sh1 might guide an oriented and asymmetric division of stem cells, such that a daughter cell in contact with Sh1 enters the differentiation pathway while the other, in contact with the DTC, remains a stem cell (Gordon et al., 2020). However, given that much of the analysis was performed using an INX-8 fusion protein marker that could conceivably alter the position of Sh1, and given that the precise relationship between the position of Sh1 vis-à-vis the border of the stem cell pool was not directly investigated, we wished to determine how hypomorphic innexin alleles alter the position of Sh1 in live worms and whether the position of Sh1 influences the germline stem or progenitor pools.
In short, our results confirmed the presence of a bare region in the wild type and showed that reducing soma-germline gap junction coupling causes Sh1 to be mispositioned distally. Further, we determined that absence of the bare region by Sh1 distal mispositioning does not markedly alter the position of the borders of the stem or progenitor cell pools. In addition, we show that the marker used in the previous study (Gordon et al., 2020) encodes a poisonous allele of inx-8 that also causes distal mispositioning of Sh1. Together, these results demonstrate that the position of the distal border of Sh1 is not relevant for GLP-1/Notch signaling nor for germ cells to exit the stem cell pool.
Results
Distal somatic gonad architecture is dictated by both somatic- and germline-expressed innexins
Previously, in fixed preparations, we observed that the distal edge of Sh1 was shifted almost all the way to the distal end of the gonad in worms bearing a hypomorphic mutation in the germline innexin inx-14(ag17) (Starich et al., 2014). We further investigated the position of the DTC relative to Sh1 in inx-14(ag17) using live imaging of intact young adult hermaphrodites bearing contrasting markers for Sh1 and the DTC (see Materials and Methods for details on markers used). We found that, in contrast with inx-14(+), the distal edge of Sh1 in inx-14(ag17) extends to the distal end of the gonad, well distal to the average position of the DTC processes (Figure 1C-E). We note that this allele only moderately impairs fertility; inx-14(ag17) hermaphrodites display a slightly reduced average brood size of 230 progeny without appreciable embryonic lethality (Table 1).
To determine whether the position of Sh1 is also shifted distally upon reduction of the somatic innexins, we investigated the position of Sh1 relative to the distal end of the gonad in a well-characterized compound inx-8 mutant ((Starich et al., 2020); Figure 1B-E). We found that Sh1 in worms bearing one partially functional somatic gonad innexin encoded by inx-8(tn1513tn1555) in an otherwise null inx-9(ok1502) background (hereafter referred to as “inx-8(rf)”; (Starich et al., 2020)), is also distally positioned, similar to inx-14(ag17) (Figure 1C-E). The existence of the bare region in wild-type worms, as well as the distally altered Sh1 position in the inx-14 and inx-8 mutants were consistent in live images of worms bearing different DTC and sheath markers (Figure 1––figure supplement 1).
An mKate2::INX-8 fusion encodes a poisonous INX-8 protein
We extended our analysis to inx-8(qy78[mKate2::INX-8]), an allele that encodes a fusion protein of mKate2 and INX-8 that was used to mark Sh1 in a prior study (Gordon et al., 2020). We found that, like inx-14(ag17) and inx-8(rf), inx-8(qy78) caused a distal shift in Sh1 (Figure 2). In addition, this allele causes a severe reduction in brood size and highly penetrant embryonic lethality (Table 1). We also observed that this same deletion in the background of the inx-9 null mutant (inx-8(qy102), see Materials and Methods) shifts Sh1 even more distally, whereas loss of inx-9 alone does not significantly affect Sh1 position (Figure 2––figure supplement 1).
To ensure that the apparent distal shift of Sh1 in the inx-8(qy78[mKate2::INX-8]) background did not reflect a disparity between the expression patterns of mKate2::INX-8 and either of the lim-7p-driven Sh1 markers, we examined the overlap between the mKate2::INX-8 and lim-7p-driven markers in strains expressing both inx-8(qy78[mKate2::INX-8]) and GFP markers encoded by tnIs6 [lim-7p::GFP] or bcIs39 [lim-7p::CED-1::GFP] (Figure 2––figure supplement 2). In short, in over 85% of gonad arms examined, the overlap was complete. In short, in over 85% of gonad arms examined, the overlap was complete. In both cases, the remaining gonads displayed reduced Sh1 expression, which may be the result of stochastic transgene downregulation.
Based on our observations that inx-8(qy78[mKate2::INX-8]) displays embryonic lethality and a distal shift in the border of Sh1, we hypothesized that inx-8(qy78) might encode a poisonous INX-8 protein. If so, we would predict that the distal shift of Sh1, the reduced brood size, and embryonic lethality seen with this allele would be dependent on the presence of the INX-8 coding region. To test this hypothesis, we used CRISPR-Cas9 genome editing to generate inx-8 null alleles in both the inx-8(qy78[mKate2::inx-8]) and wild-type genetic backgrounds. We generated deletions with identical breakpoints in the inx-8 locus in both genetic backgrounds [e.g., inx-8(qy78tn2031) and inx-8(tn2034)] starting 136 bp upstream of the wild-type inx-8 ATG start codon and extending 221 bp into inx-8 exon 3 (Figure 2 and Figure 2—figure supplement 1). In the inx-8(qy78[mKate2::inx-8]) context, this deletion also removes the mKate2 moiety. These deletions are expected to constitute inx-8 null alleles because, in addition to removing the start codon, they delete amino acids 1–349 (out of 382 amino acids), including virtually all residues essential for spanning the plasma membrane and forming a channel (Starich and Greenstein, 2020). These deletions must not appreciably perturb the function of inx-9 because they exhibit nearly normal brood sizes (Table 1). Unlike the original inx-8(qy78) allele, the compound mutant inx-8(qy78tn2031) almost completely restores the DTC-Sh1 positional relationship with a substantial return of the bare region (Figure 2). Further, inx-8(qy78 tn2031) exhibits a nearly normal brood size and suppresses the embryonic lethality observed in the inx-8(qy78) starting strain (Table 1). Likewise, the identical deletion generated in the wild-type genetic background [e.g., inx-8(tn2034)] also exhibits a substantial bare region (Figure 2––figure supplement 1) with a normal brood size and negligible embryonic lethality (Table 1). Thus, we conclude that inx-8(qy78) encodes a poisonous mKate2::INX-8 product that interferes with the normal channel and/or rivet functions of soma-germline gap junctions.
A surprising observation was that both the inx-14(ag17) mutation and the inx-9(ok1502) null mutation could individually suppress the embryonic lethality caused by an mKate2::INX-8 fusion protein (Table 1). Because it has been shown that gap junctions in the proximal gonad are required for embryonic development by virtue of their function to deliver malonyl-CoA to developing oocytes (Starich et al., 2020), one possibility is that mKate2::INX-8-containing channels are constitutively or too-widely open such that embryos receive inappropriately large amounts of transiting biomolecules, and that this can be compensated by reducing channel function. Alternatively, possible delays in gametogenesis that may occur in the double mutants might effectively increase oocyte quality by providing additional time for levels of needed biomolecules to build up in the germ line. Perhaps favoring this second possibility is an unusual behavior of inx-8(qy78) itself: the first embryos produced in the brood display heightened embryonic lethality, suggesting that if key limiting substances fail to accumulate early, a time-dependent or later mechanism may compensate.
In any case, the genetic behavior of inx-8(qy78) suggests that this mutant allele confers both loss-of-function and antimorphic properties to soma-germline gap junctions. We infer loss-of- function behavior since other loss-of-function mutations affecting soma-germline gap junctions, such as inx-14(ag17), also cause a loss of the bare region, though inx-14(ag17) does so without reducing brood size or embryonic viability (Table 1). We infer antimorphic behavior of inx-8(qy78) since removing the entire protein suppresses all defects, including loss of the bare region, brood size and embryonic viability.
The distal position of Sh1 does not influence the position of the stem cell pool border
A recent model proposed that the position of the Sh1 border influences the stem/non-stem decision in underlying germ cells (Gordon et al., 2020). However, because the previous study did not examine the position of stem or progenitor cells, and because the model was based on results using the poisonous inx-8(qy78) allele, we investigated this relationship.
In its simplest form, the model predicts that when the distal edge of Sh1 is positioned distally, the stem/non-stem border should similarly shift distally. The SYGL-1 protein serves as a stem cell marker as sygl-1 is a direct transcriptional target of GLP-1/Notch in the germ line (Brenner and Schedl, 2016; Chen et al., 2020; Kershner et al., 2014; Lee et al., 2019; Lee et al., 2016; Shin et al., 2017). We analyzed the proximal extent of the pool of SYGL-1-positive cells bearing a well-characterized OLLAS epitope tag on SYGL-1 and compared that boundary relative to the distal Sh1 border (Figure 3 and Materials and Methods). In the case of inx-14(ag17), though the distal border of Sh1 was shifted drastically and significantly, there was no significant change in the size of the SYGL-1-positive stem cell pool. In the case of the inx-8(qy78[mKate2::inx-8]) allele, the border of the SYGL-1-positive pool was marginally shifted distally relative to the wild type, though not commensurate with the extent to which Sh1 shifted distally in this background. Furthermore, the shifted border of the stem cell pool was suppressed when inx-8 was deleted, either in inx-8(qy78tn2031) on in inx-8(tn2034), suggesting that such a defect was due to the altered function of mKate2::INX-8, rather than due to the position of Sh1 (Figure 3A-C). To detect any subtle correlation between the proximal end of the SYGL-1 pool and the distal extent of Sh1, we plotted these against each other and computed an R value (Figure 3D). By Pearson correlation, there is no significant relationship in any genotype examined between the position of the sheath cell, and the extent of the SYGL-1(+) stem cell pool.
The recent model also proposed that Sh1 controls spindle orientation at the stem/progenitor border. However, we found that in the wild type (marker-only) strain, the distal position of Sh1 was proximal to the proximal SYGL-1-positive border in 86% of the gonads (276/320), with the distance 5 cell diameters or greater in 67% of gonads (215/320) (Figure 3 and Figure 3—figure supplement 1). This 5 cell-diameter distance is not consistent with the hypothesis that Sh1 is controlling spindle orientation at the border as such control would be expected to occur over a distance of 1-2 cell diameters.
We conclude that there is no correlation between the position of the distal border of Sh1 and the proximal border of the SYGL-1-positive stem cell pool and that if spindle-oriented divisions occur at the Sh1 border, they are not influencing cell fate.
The Distal position of Sh1 does not influence the position of the progenitor pool border
Although we found no correlation between the stem cell pool border and Sh1 position, we wondered whether altered Sh1 position might nevertheless influence the position of the border between the progenitor zone (PZ) and the transition zone that marks overt meiotic entry. In wild type, we found that the distal position of Sh1 can be either distal or proximal of the PZ border, using the length of the CYE-1-positive region to define the PZ border, following CYE-1 and pSUN-1(S8) co-staining (Figure 4 and Materials and Methods; Mohammad et al., 2018). We found that although there is a subtle shift in the PZ border in inx-8(qy78) and inx-14(ag17), it does not correlate with the dramatic shift in Sh1 position seen in these mutants (Figure 4).
Discussion
These studies show that impaired innexin function distally displaces sheath cells, but that this displacement does not similarly shift the proximal border of the stem cell pool (Figure 5). We show that a gap normally exists between the DTC and Sh1, but that this gap can be closed with reduced innexin activity either in the soma or the germ line. These results contradict a previous observation that relied on a marker that was itself interfering with innexin function. The cautionary tale is that fusion proteins used as markers, even when they are generated by CRISPR/Cas9 genome editing in the context of the endogenous locus and therefore are not likely mis-expressed or overexpressed, may nevertheless generate poisonous proteins. Here, the mKate2::INX-8 fusion protein caused a distal shift in the Sh1 position due to its apparent antimorphic effect on gap junctions. Because of the redundant function of INX-9, we were able to remove the offending mKate2::INX-8 protein entirely and restore the bare region. Our finding that the germline response to signaling from the DTC, as measured by expression of the GLP-1/Notch target, SYGL-1, is independent of interactions with distal sheath cells also meshes with the finding that males, which lack distal sheath cells altogether, exhibit similarly sized stem cell pools (Crittenden et al., 2019). In addition, the distal border of Sh1 relative to the proximal stem cell border in the wild type is ≥5 cell diameters in the majority of gonads examined. Thus, the previous model that Sh1 acts to orient divisions of stem cells to thereby direct their fate is also called into question by our results. Finally, using alleles that dramatically alter the position of Sh1, we found no evidence supporting the prediction that the stem/non-stem border is coincident with the Sh1 border. Together, these results indicate that Sh1 is not involved in the germline stem-progenitor fate decision.
Our studies also provide evidence that innexin gap junctions not only serve as communication and adhesion junctions, but that in the context of an organ system they contribute to the positioning of cells relative to each other. How might gap junctions influence the relative position of the DTC and Sh1 in the distal gonad arm? The DTC also forms gap junctions with germ cells, which must be disassembled as germ cells enter the bare region, only to be reassembled again when in contact with Sh1 (and then again with the more proximal sheath cells). A detailed TEM analysis of the gonad (Hall et al., 1999) led to the consideration that a constant interplay of association and dissociation likely also occurs between Sh1 and the underlying germ cells that migrate proximally along the arm: as germ cell flux continually moves germ cells towards the proximal end of the gonad, the Sh1 cells presumably extend their filopodia distally and form new gap junction connections with incoming germ cells. Otherwise, the bare region would increase in size. The relative steady-state positions of the DTC and Sh1 may therefore be determined by the rate of germ cell proliferation as well as by the strength of interaction between Sh1 and germ cells, and gap junctions could contribute to both.
To complement the role of gap junctions in promoting robust proliferation, the kinetics of gap junction coupling between the somatic gonad and germ cells may play a role in determining the strength of the interactions between the two cell types. Unlike sheath-oocyte junctions, which form large plaques containing many functional gap junctions, the gap junctions formed in the distal arm appear to represent looser associations of a few gap-junction channels (Starich et al., 2014). Nonetheless, these associations may be sufficient to maintain adhesion with the underlying germline, functioning like regularly spaced rivets, albeit dynamic and removeable ones.
Disentangling the adhesive and channel functions of gap junctions is a complex issue. The mutants used in this study are competent to form gap junctions. However, they may do so less efficiently than their wild-type counterparts. For example, the pattern of localization of gap junction puncta in inx-8(rf) and inx-14(ag17) appears more diffuse than in the wild type (Starich et al., 2014);(Starich et al., 2020). Alternatively or additionally, the mutants in this study may assemble into hemichannels as readily as the wild-type, but the pairing or opening of gap junction channels may be compromised. Studies of connexin gap-junction channels in paired Xenopus oocytes strongly suggest that opening of hemichannels facilitates their assembly into gap junctions. That study proposed hemichannel opening collapses the intermembrane space between juxtaposed cells to allow the extracellular loops of connexins to dock into gap junctions (Beahm and Hall, 2004). If a similar model applies to innexin-containing gap junctions, then rivet and channel function would be coupled.
How could impaired innexin function cause Sh1 to creep more distally? One hypothesis is that when fewer junctions are made, reduced adhesion or reduced recognition occurs. This scenario would also be consistent with Sh1 extending to the distal end when no gap junctions can form between sheath cells and germ cells (Starich et al., 2014). At the same time, it is not possible to exclude the formal possibility that the DTC and Sh1 engage in an active repellent or passive space-excluding interaction that somehow involves gap junction function. Another possibility is that a deficit in gap-junctions might be sensed by Sh1, which then responds by extending its coverage of the germ line to increase the surface area over which junctions may form to supply more of the active biomolecules that transit through these junctions. Nevertheless, our studies show that the position of germline stem cells is independent of the position of Sh1.
Materials and Methods
Live imaging and image analysis of live samples
Live specimens were grown at 20°C, and staged by picking mid-L4 larvae, then allowing them to grow at 20°C until imaging them 24 hours later. Animals were immobilized using 10mM Levamisole (Sigma T1512) in M9 buffer. Imaging was carried out on a Nikon W1 spinning disk confocal microscope.
Image analysis was carried out on 2-dimensional maximum-projection Z-stack images of 3D confocal data. The distance from the distal end of the gonad to the end of each DTC process was measured along a line drawn from the end of each process parallel to the distal-proximal axis to a line drawn tangent to the distal end, orthogonal to the distal-proximal axis line (Figure 1—figure supplement 1). The distance between the distal end of the gonad and the most distal extent of Sh1 was measured in the same way. All data points were recorded for each sample and used to calculate the mean values presented in Figures 1 and 2.
Sh1 visualization in live worms: bcIs39 [lim-7p::CED-1::GFP] (Zhou et al., 2001) encodes a functional membrane-localized fusion to CED-1. tnIs5 and tnIs6 encode an identical non-functional fusion to the first 61 amino acids of LIM-7 (tnIs5 or tnIs6) denoted here as “lim-7p::GFP” that includes 2.23kb upstream, the first two exons, and the first intron of lim-7 fused to GFP (Hall et al., 1999).
DTC visualization in live worms: naIs37 [lag-2p::mCherry-PH] encodes mCherry fused to the PH domain of rat phospholipase C delta (Pekar et al., 2017) and qIs154 [lag-2p::MYR-tdTomato] encodes a src kinase myristoylation sequence fused to tdTomato (Byrd et al., 2014).
Strains
C. elegans strains (Table S1) were grown on standard NGM media [containing 6.25 mg/ml Nystatin (added after autoclaving) and 200 mg/ml streptomycin sulphate (added before autoclaving)] with E. coli strain OP50-1 as food source. Similar results were obtained on NGM medium with OP50 as food source and without inclusion of streptomycin sulphate in the media. Strains were grown at 20°C. In addition to the wild-type strain N2, the following alleles, described in WormBase (www.wormbase.org) or in the cited references, were used: Chr. I—inx-14(ag17) (Miyata et al., 2008; Starich et al., 2014), sygl-1(q983[3xOLLAS::sygl-1]) (Shin et al., 2017) .
Chr. IV—inx-8(qy78[mKate2::inx-8]) (Gordon et al., 2020), inx-8(qy78 tn2031) (this work), inx-8(tn2034) (this work), inx-9(ok1502), inx-8(qy102[mKate2::inx-8)] inx-9(ok1502) (Gordon et al., 2020), inx-8(tn1513tn1553) inx-9(ok1502) (Starich and Greenstein, 2020), inx-8(tn1513tn1555) inx-9(ok1502) (Starich et al., 2020).
Balancer chromosomes (Dejima et al., 2018) used included: tmC18 [dpy-5(tmIs1236)] I, tmC27[tmIs1239] I, tmC5[tmIs1220] IV.
Integrated transgenes included: mIs11[myo-2p::gfp + pes-10p::gfp + gut promoter::gfp] IV, bcIs39[lim-7p::ced-1::gfp + lin-15(+)] V (Zhou et al., 2001), qIs154[lag-2p::MYR::tdTomato +ttx-3p::gfp] V (Byrd et al., 2014), tnIs5[lim-7p::gfp + rol-6(su1006)] X, tnIs6[lim-7p::gfp + rol-6(su1006)] X (Hall et al., 1999), cpIs122[lag-2p::mNeonGreen::plcdeltaPH] (Gordon et al., 2020), naIs37[lag-2p::mCherry:: plcdeltaPH + unc-119(+)] (Pekar et al., 2017).
Extrachromosomal arrays used included: tnEx42[acy-4::gfp + rol-6(su1006)] (Govindan et al., 2009).
Strain constructions
Multiply mutant strains were constructed in a straightforward manner (Huang and Sternberg, 1995). tmC18 was used as a balancer chromosome for inx-14(ag17). tmC27 was used as a balancer chromosome for sygl-1(q983). tmC5 or mIs11 were used as balancer chromosomes for inx-8 and inx-9 mutant alleles. The presence of inx-14(ag17) in strains was verified by PCR and DNA sequencing. The ag17 allele was originally described as an Arg to His change in the second extracellular loop of INX-14, but the exact residue position was not specified (Miyata et al., 2008). A 1.2-kb PCR fragment covering this region was amplified with primers inx-14delF and inx-14delR (see Table S1 for the sequence of oligonucleotides used in this study). The PCR fragment was sequenced with the inx-14delR primer. No sequence changes were found in Arg residues predicted to occupy the second extracellular loop. However, a CGT to CAT (R326H) change was identified at a residue position predicted to lie near the cytoplasmic end of the fourth transmembrane domain, and we surmise that this change represents the original ag17 mutation. The presence of the sygl-1(q983[3xOLLAS::sygl-1]) mutation in strains was verified by PCR with primers sygl1-F and sygl1-R, which produce a 216 bp product in the wild type and a 348 bp product in sygl-1(q983) and by anti-OLLAS staining. The presence of inx-8(qy78 tn2031) and inx-8(tn2034) in strains was verified by PCR with oligonucleotide primers inx8_delta.F and inx8_delta.R.
Brood counts and Embryonic lethality measurements
L4-stage hermaphrodites were cultured individually and transferred approximately every 24 hours until they stopped producing embryos (4–6 days). Worms that crawled off the media and died were redacted (varied from 0–10% depending on the experiment). Embryos that failed to hatch after 24–36 hours were counted and scored as dead. In the majority of cases, these embryos exhibited morphological abnormalities. Control experiments demonstrated that these embryos were not simply delayed and never hatched. Embryos that hatched were counted and scored as viable. This includes embryos in inx-8(qy78) that died shortly after hatching, arrested as larvae, and/or exhibited morphological abnormalities.
Genome editing
CRISPR-Cas9 genome editing was used to generate inx-8 null alleles in both the inx-8(qy78[mKate2::inx-8]) and wild-type genetic backgrounds. The approach taken generated identical 1524 bp deletions within the inx-8 locus in both genetic backgrounds starting 136 bp upstream of the wild-type inx-8 ATG start codon and extending 221 bp into inx-8 exon 3. In the inx-8(qy78[mKate2::inx-8]) context, this edit removes both the mKate2 moiety and inx-8. The deletions are expected to constitute inx-8 null alleles because, in addition to removing the start codon, they delete amino acids 1–349 (out of 382 amino acids), including virtually all residues essential for spanning the plasma membrane and forming a channel (Starich and Greenstein, 2020). The approach used pRB1017 to express two single guide RNAs (sgRNAs) under control of the C. elegans U6 promoter (Arribere et al., 2014). Oligonucleotides inx8_us_sgRNA1.F and inx8_us_sgRNA1.R were annealed and used to generate the plasmid inx8_us_sgRNA1 to direct Cas9 cleavage 136 bp upstream of the ATG initiator codon (Table S1 lists the sequences of all oligonucleotides used in this study). Oligonucleotides inx8_sgRNA1.F and inx8_sgRNA1.R were annealed and used to generate the plasmid inx8_sgRNA1 to direct Cas9 cleavage in exon 3. To generate sgRNA clones, annealed oligonucleotides were ligated to BsaI-digested pRB1017 plasmid vector, and the resulting plasmids were verified by Sanger sequencing. pDD162 served as the source of Cas9 expressed under control of the eef-1A.1/eft-3 promoter (Dickinson et al., 2013). The repair template oligonucleotide used was inx8_rpr. Genome editing employed the dpy-10 co-conversion method (Arribere et al., 2014). The injection mix contained pJA58 (7.5 ng/μl), AF-ZF-827 (500 nM), inx8_us sgRNA1 (25 ng/μl), inx8_sgRNA1 (25 ng/μl), inx8_rpr (500 nM), and pDD162 (50 ng/μl) and was injected into adult hermaphrodites from strains DG5131 inx-8(qy78[mKate2::inx-8]) IV; bcIs39[lim-7p::ced-1::gfp + lin-15(+)] V; naIs37[lag-2p::mCherry::PH + unc-119(+)] and DG5020 bcIs39[lim-7p::ced-1::gfp + lin-15(+)]V; naIs37[plag-2::mCherryPH + unc-119(+)]. Correct targeting was verified by conducting PCR with primer pairs inx8_delta.F and inx8_delta.R followed by DNA sequencing. Three deletion alleles were recovered from the injections into DG5131 (qy78tn2031, qy78tn2032, and qy78tn2033), and two deletion alleles were recovered from the injections into DG5020 (tn2034 and tn2035). The deletion alleles were outcrossed to tmC5(tmIs1220[pmyo-2::Venus])/+ IV; bcIs39[lim-7p::ced-1::gfp + lin-15(+)/+V; naIs37[plag-2::mCherryPH + unc-119(+)]/+ males. Homozygous strains were analyzed by confocal microscopy.
Immunostaining and image analysis of fixed samples
Immunostaining was carried out as described (Mohammad et al., 2018). Briefly, synchronized adult hermaphrodites, 24-hr past mid-L4, were dissected in PBST (PBS with 0.1% Tween 20), with 0.2 mM levamisole to extrude the gonads. The gonads were fixed in 3% paraformaldehyde solution for 10 min and then post-fixed in −20° chilled methanol for 10 min. After 3x 10-min washes with PBST, they were blocked in 30% goat serum for 30 min at RT. The gonads were then incubated with the desired primary antibodies diluted (see below) in 30% goat serum at 4° overnight. The next day, after 3x 10-min PBST washes, the gonads were further incubated with appropriate secondary antibodies, diluted in 30% goat serum, at 4° overnight. The gonads were washed 3 times with PBST, then incubated with 0.1 g/ml DAPI in PBST for 30 min. After removal of excess liquid, the gonads were mixed with anti-fading agent (Vectashield) and transferred to an agarose pad on a slide. Hyperstack images were captured using a spinning disk confocal microscope (PerkinElmer-Cetus, Norwalk, CT). Two overlapping hyperstack images were captured for each gonad arm to obtain coverage of >50 cell diameters from the distal end of the gonad. Images were further processed in Fiji, and DAPI stained nuclei were used to mark the number of cell diameters from the distal end. Employing pixel to micron ratio, specific to the images captured, cell diameters were converted into microns where required.
SYGL-1 zone length assessment: OLLAS staining was used to assess 3xOLLAS::SYGL-1 accumulation (Shin et al., 2017). In wild-type young adults, SYGL-1 accumulates at the distal end of the germline and is downregulated around 10 cell diameters from the distal tip (Kocsisova et al., 2019; Shin et al., 2017). Cell diameters were counted from the distal end of the germline up to the row where SYGL-1 is no longer visible by eye. OLLAS staining in the wild type worms without OLLAS tag was used to differentiate staining from the background. To confirm the accuracy of our visual assessment, we quantified the intensity of SYGL-1 accumulation in the distal germline, employing methods similar to Chen et al., 2020 in the same set of germlines where the SYGL-1 zone was visually evaluated. We found that the cell diameter position called as the end of the SYGL-1 zone consistently corresponded to 6 – 9% of peak SYGL-1 intensity, for each genotype. These results indicate that the SYGL-1 zone length visual assessment was reproducible and consistent.
Progenitor zone length assessment: The gonads were stained with a progenitor zone marker, CYE-1, and an early meiotic prophase marker, pSUN-1, (anti-SUN-1 S8-Pi) (Mohammad et al., 2018). For assessing the progenitor zone length, cell diameters (rows) were counted from the distal end of the germline, where all cells are CYE-1 positive, till the point after which the majority of the cells in a row have switched from staining for CYE-1 to pSUN-1. Note that pSUN-1 staining is not shown in the figures though it was used to assess the PZ border.
Assessment of distal position of Sh1: anti-GFP antibody staining was used to visualize the sheath, where cell diameters were counted from distal end to the point where GFP staining became prominent.
Primary antibodies used: mouse anti-CYE-1 (1:100; (Brodigan et al., 2003)); guinea pig anti-SUN-1 S8-Pi (1:1000; (Penkner et al., 2009)); rat anti-OLLAS (1:2000; Novus Biological); rabbit anti-GFP (1:200; from Swathi Arur, MD Anderson Cancer Center).
Secondary antibodies used: Alexa Fluor 647 goat anti-mouse (Life Technologies), Alexa Fluor 594 goat anti-guinea pig (Invitrogen), Alexa Fluor 594 donkey anti-rat (Invitrogen), Alexa Fluor 488 goat anti-rabbit (Invitrogen).
Funding
ACS grant PF-19-231-01-CSM to TT
NIH R35GM134876 and R01AG065672 to EJAH
R01 GM100756 to T. Schedl
NIH GM57173 to DG
Contributions
TT, EJAH, wrote the manuscript with contributions from all authors. TT, AM, DG, T Starich performed analyzed and interpreted experiments. EJAH, T Schedl, DG oversaw experiments and analysis
Competing interests
E.J.A.H. holds US patent 6,087,153.
Acknowledgements
We thank Gabriela Huelgas-Morales for discussions and constructive suggestions during the course of this work, and we thank Swathi Arur for the affinity purified anti-GFP antibody. We thank WormBase. The CGC, which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440), provided some starting strains.