|Home | About | Journals | Submit | Contact Us | Français|
C. elegans male sexual behaviors include chemotaxis and response to hermaphrodites, backing/turning, vulva location, spicule insertion and sperm transfer, culminating in cross fertilization of hermaphrodite oocytes with male sperm. The LOV-1 and PKD-2 transient receptor potential polycystin (TRPP) complex localizes to ciliated endings of C. elegans male-specific sensory neurons and mediates several aspects of male mating behavior. TRPP complex ciliary localization and sensory function is evolutionarily conserved. A genetic screen for C. elegans mutants with PKD-2 ciliary localization (Cil) defects led to the isolation of a mutation in the cil-1 gene.
Here, we report that a phosphoinositide (PI) 5-phosphatase CIL-1 regulates TRPP complex ciliary receptor localization and sperm activation. cil-1 does not regulate the localization of other ciliary proteins, including intraflagellar transport (IFT) components, sensory receptors, or other TRP channels in different cell types. Rather, cil-1 specifically controls TRPP complex trafficking in male-specific sensory neurons and does so in a cell autonomous fashion. In these cells, cil-1 is required for normal PI(3)P distribution, indicating that a balance between PI(3,5)P2 and PI(3)P is important for TRPP localization. cil-1 mutants are infertile due to sperm activation and motility defects. In sperm, the CIL-1 5-phosphatase and a wortmannin sensitive PI 3-kinase act antagonistically to regulate the conversion of sessile spermatids into motile spermatozoa, implicating PI(3,4,5)P3 signaling in nematode sperm activation.
Our studies identify the CIL-1 5-phosphatase as key regulator of PI metabolism in cell types that are important in several aspects of male reproductive biology.
Phosphoinositides (PIs) and their phosphatases and kinases play pivotal roles in receptor trafficking as well as in membrane organelle biogenesis/transport, endocytosis, cytoskeleton dynamics, signal transduction, cell motility, and channel activity modulation [1, 2]. Many sensory receptors localize to ciliary membrane, which serve as cellular antennae and function in development, signaling, and physiology. Whether PIs and PI-generating enzymes regulate ciliary receptor trafficking is unknown.
The nematode Caenorhabditis elegans is a powerful model to study molecular mechanisms required for ciliary receptor trafficking. The C. elegans transient receptor potential polycystin (TRPP) complex proteins LOV-1 (TRPP1) and PKD-2 (TRPP2) localize to sensory cilia [3, 4]. Autosomal dominant polycystic kidney disease (ADPKD) is caused by mutations in the TRPP1 and TRPP2 genes . Since the discovery of LOV-1 and PKD-2 in cilia of C. elegans male-specific sensory neurons, TRPPs have been found in primary cilia or flagella in organisms ranging from alga to man, hinting at an evolutionarily conserved mechanism regulating TRPP ciliary localization (reviewed in ).
In diverse species, TRPP family proteins serve reproductive functions [7-12]. C. elegans LOV-1 and PKD-2 function in male mating behaviors. pkd-2 appears to be expressed solely in the male nervous system and not sperm, as judged by antibody staining . In contrast to mammalian and Drosophila sperm, nematode sperm are not flagellated and do not possess an actin/tubulin cytoskeleton (reviewed in ). Instead, each sperm extends a single pseudopod enriched with major sperm protein (MSP). MSP is used to assemble a filamentous network required for amoeboid motility and fertilization. Screens have identified genes required for spermatogenesis, hermaphrodite sperm activation, and sperm function. Signaling pathways involved in sperm activation in both males and hermaphrodites are not known.
PKD-2 ciliary localization requires vesicular trafficking and utilizes both general and cell-type specific factors . General factors include the clathrin coated vesicle adaptor protein-1 (AP-1) UNC-101 and the intraflagellar transport (IFT) machinery. Cell-type specific PKD-2 localization factors include LOV-1 and STAM-Hrs (Signal-transducing adaptor molecule and hepatocyte growth factor regulated tyrosine kinase substrate) complex [14, 15]. We performed a genetic screen for C. elegans mutants with PKD-2 ciliary localization (Cil) defects . Here, we identify CIL-1, a PI 5-phosphatase that regulates TRPP complex localization. cil-1 is also a positive regulator of sperm activation and motility. Using genetically encoded PI-indicators and pharmacological approaches, we determine that CIL-1 hydrolyses PI(3,5)P2 and PI(3,4,5)P3 in male-specific sensory neurons and sperm, respectively. We conclude that CIL-1 acts in multiple tissues that are important for male reproductive biology, thereby controlling diverse cellular processes as an in vivo PI-metabolizing enzyme.
In wild type (WT), PKD-2GFP localizes to cell bodies and ciliary endings of 21 male-specific neurons in the head (cephalic CEMs) and tail (ray RnBs and hook HOB) (, Figure 1A-D). In my15 mutants, PKD-2GFP is distributed throughout these male-specific neurons including dendrites, axons, cell bodies, and cilia (Figure 1E-F). cil-1 is not required for neuronal cell fate or development of pkd-2 expressing neurons as judged by reporters including transcriptional and soluble Ppkd-2GFP, OSM-6GFP (data not shown), Ppkd-2:SNB-1:GFP (Supplemental Figure 1, SF1), and fluorescent protein tagged PI-markers (SF4). Ppkd-2GFP and endogenous pkd-2 mRNA levels are unaltered in my15 animals as judged by qRT-PCR (data not shown), indicating that my15 may affect PKD-2 protein expression or stability, but not gene expression. In WT and my15 dendrites, small PKD-2GFP particles move bidirectionally (Supplemental Movies 1 and 2), indicating that a pool of PKD-2GFP is properly trafficked in my15 dendrites. Moreover, PKD-2GFP distribution in the my15 ciliary region appears normal, despite abnormally increased dendritic and axonal distributions (Figure 1E-F, ).
We examined the distribution of additional GFP-tagged ciliary proteins, including functional LOV-1GFP (, SF1A-D), a TRP-vanilloid OSM-9, a G-protein coupled receptor (GPCR) ODR-10, an IFT B-complex polypeptide OSM-6, and IFT modulator BBS-5. Only LOV-1GFP is abnormally distributed to dendritic and axonal processes in my15 sensory neurons (SF1C-D). We also examined the localization of the presynaptic marker synaptobrevin Ppkd-2SNB-1GFP . In WT and my15 males, SNB-1GFP labels presynaptic puncta along axonal processes (SF1E-F), indicating that cil-1 does not grossly affect axonal targeting. my15 mutants are normal in lipophilic DiI dye-filling of ciliated sensory neurons, chemotaxis to diacetyl, dauer formation, and osmotic avoidance (data not shown). We conclude that cil-1 is specifically required for localization of the TRPP complex, but not for general receptor trafficking, ciliogenesis, neuronal polarization, or sensation.
lov-1, pkd-2, and a subset of Cil mutants, are response and location of vulva (Lov) defective during male mating [3, 4, 16]. my15 males exhibit normal response and vulva location behaviors (SF1G, ), which may be explained by the presence of PKD-2GFP in cilia. cil-1 double or triple homozygous or trans heterozygous mutants with pkd-2 and lov-1 did not alter male mating efficiency, indicating that cil-1 is not a genetic modifier of the C. elegans TRPP genes (data not shown). Although mating behaviors appear normal, my15 males are largely infertile and produce few offspring due to a sperm (Spe) defect, which will be discussed later.
We previously showed that TRPP complex formation is important for trafficking . In a lov-1 mutant, PKD-2GFP forms aggregates in cell bodies and localizes to cilia at a reduced level (Figure 1G-H). In lov-1; cil-1 CEM neurons (Fig 1I), the PKD-2GFP localization phenotype is additive: bright aggregates in the cell bodies (lov-1 phenotype; white arrows) and mislocalization to the dendrites and axons (cil-l phenotype), albeit at reduced levels (red arrows). In lov-1; cil-1 RnB neurons (Figure 1J), the PKD-2GFP localization phenotype resembles that of lov-1 but not cil-1: aggregates in the cell body and absence from dendritic and ciliary compartments. This strict requirement of lov-1 in RnB as compared to CEM neurons has been previously shown for PKD-2 ciliary targeting . The basis of this cell type specificity is unknown, but may be due to differences in protein trafficking mechanisms or structural differences between CEM and RnB neurons.
PKD-2 ciliary abundance is tightly controlled. STAM and Hrs mediate PKD-2 and LOV-1 downregulation via transport from early endosomes to endosomal sorting complexes (ESCRT) . Reducing stam-1 or hgrs-1 function results in PKD-2GFP and LOV-1GFP accumulation at the ciliary base (Figure 1K-L, ). In CEM neurons of a stam-1(ok406); cil-1(my15) double mutant (Figure 1M), the PKD-2GFP localization phenotype is additive. In stam-1; cil-1 RnB neurons (Figure 1N), PKD-2GFP is distributed evenly throughout, similar to cil-1 single mutants. We conclude that, in RnB neurons, cil-1 acts after lov-1 but before stam-1. In CEMs, the lov-1; cil-1 and cil-1; stam-1 phenotype is complex, making it difficult to place lov-1, cil-1, and stam-1 in a linear or parallel pathway.
The smallest rescuing genomic fragment for both the Cil and Spe phenotypes (8160K-1, Figure 2A-B) contains two genes, C50C3.7 and bath-42 in the operon CEOP3484. The bath-42(tm2360) deletion mutant is nonCil and nonSpe (data not shown). A construct containing 2.2kb 5′ UTR of the bath-42 and the C50C3.7 genomic region including the 3′UTR rescues both Cil and Spe phenotypes of my15 (PCR-SOEed C50C3.7, Figure 2A-B). Sequencing analysis of my15 genomic DNA identified a G→A transition that converts Trp301 (TGG) to a stop codon (TAG) in the 5th exon of C50C3.7 (Figure 2A). From both WT and my15 cDNA pools, RT (reverse transcriptase)-PCR identified two cil-1 cDNAs (long C50C3.7a and short C50C3.7b). The short form (C50C3.7b) is generated by alternative splicing within the 3rd intron, introducing a stop codon at 124th amino acid before the my15 induced lesion. Ppkd-2CIL-1a(C50C3.7a)tdTomato fully rescues the cil-1(my15) Cil but not Spe phenotype (Figure 2A, B), which is not surprising given that pkd-2 is not expressed in sperm. In contrast, expression of C50C3.7a using an intestinal promoter fails to rescue the Cil phenotype (data not shown). We did not attempt to rescue the cil-1(my15) Spe phenotype using cell-type specific promoters as transgenes are often silenced in the germline. We conclude that C50C3.7 is the gene mutated in cil-1(my15) animals and that CIL-1 acts autonomously in male-specific neurons to control TRPP localization.
cil-1 encodes a phosphoinositide (PI) 5-phosphatase (referred as 5-phosphatase), which removes the D-5 phosphate from the inositol ring of membrane associated PI or soluble inositol phosphates. The 5-phosphatase family comprises ten mammalian, four yeast, and five C. elegans enzymes. Phylogenetic analysis of the 5-phosphatase catalytic domain and/or of the presence or absence of adjacent domains reveals that CIL-1 is closely associated with two mammalian 5-phosphatases: SKIP (skeletal muscle and kidney enriched inositol phosphatase) and PIPP (proline rich inositol polyphosphate phosphatase). CIL-1, SKIP, and PIPP belong to SKICH (SKIP carboxyl homology) subfamily, which contains a C-terminal SKICH-like domain (SF2A, D) and mediates protein localization .
The C. elegans genome encodes five 5-phosphatase genes: ipp-5, unc-26, ocrl-1, cil-1/C50C3.7, and T25B9.10. ipp-5 (Type I) negatively regulates ovulation by inhibiting inositol 1,4,5-triphosphate (IP3) signaling in the spermatheca . IP3 signaling also regulates mating behavior steps of turning, spicule insertion, and sperm transfer . unc-26 (synaptojanin) is required for synaptic vesicle endocytosis and recycling, with mutants exhibiting uncoordinated movements . Neither ipp-5 nor unc-26 mutant is Cil or Spe defective, and cil-1(my15) mutants are normal in ovulation, locomotion, male turning, spicule insertion, and sperm transfer, ruling out possible overlapping functions (data not shown).
CIL-1 contains two conserved 5-phosphatase motifs in its catalytic domain (SF2B). We introduced a missense mutation in a known critical residue of 5-phosphatase motif (CIL-1N175A, SF2B, red arrowhead). Ppkd-2 CIL-1N175AtdTomato failed to rescue the my15 Cil phenotype, indicating that phosphatase catalytic activity is required for CIL-1 function.
In male neurons, the rescuing Ppkd-2CIL-1tdTomato is distributed in cilia, dendrites, axons, cell bodies with occasional small puncta, and weakly in nuclei (SF3D), and is often visible as bright dots at ciliary bases of ray neurons (SF3E), suggesting CIL-1 function in ciliary regions. In the intestine, Pvha-6CIL-1GFP localizes to cytoplasmic reticular structures (SF3F).
Seven PI species are generated by the reversible phosphorylation and dephosphorylation. To determine what PI species are CIL-1 substrates, we expressed genetically encoded biosensors to detect changes in specific PI lipid concentrations in male-specific sensory neurons and the intestine. We observed obvious differences in PI(3)P Hrs(2×FYVE) and PI(3,4)P2/PI(3,4,5)P3 AKT(PH), but not PI(4,5)P2 (PH domain of PLC-delta) markers between WT and cil-1(my15) animals. In the WT intestine, PI(3)P is primarily found in tubulovesicular structures without any plasma membrane (PM) enrichment, similar to the CIL-1 distribution pattern (compare Figure 3A with SF3F). In the cil-1 intestine, PI(3)P is severely disrupted, with a diffuse pattern in the cytoplasm (Figure 3D). In WT intestine, PI(3,4)P2/PI(3,4,5)P3, is found in tubulovesicular structures and enriched at the PM (Figure 3B, arrowheads and arrow). In cil-1(my15) mutants, the PI(3,4)P2/PI(3,4,5)P3 marker labels no distinct structure, appearing diffuse in the cytoplasm with no PM enrichment. A similar pattern has been reported in C. elegans let-512/vps-34 PI 3-kinase mutants, in which PI(3)P generation is reduced . In both WT and cil-1(my15) intestine, PI(4,5)P2 is enriched at the apical PM lining the intestinal lumen as well as basolateral PM (Figure 3C-F). These data indicate that CIL-1 displays in vivo phosphatase activity towards PI(3,4,5)P3 and PI(3,5)P2 in the intestine.
In male-specific sensory neurons, we observe differences in PI(3)P distribution (Figure 3G-H), but not PI(3,4)P2/PI(3,4,5)P3 or PI(4,5)P2 markers in cil-1(my15) males (SF4). In WT neurons, the PI(3)P marker is enriched in nuclei, small puncta in the cell bodies, but rarely in dendritic and ciliary regions (Figure 3G, G′), with only 1/13 animals displaying detectable expression in dendrites and cilia. In cil-1(my15), the PI(3)P marker decorates dendritic processes and cilia (16/27 animals) in addition to the enrichment in the nuclei and cell bodies (Figure 3H, 4H′). The PI(3)P marker is distinctly bright at cil-1(my15) ciliary bases (Figure 3H inset), hinting that loss of CIL-1 perturbs PI(3)P distribution in this region. These data suggest that CIL-1 displays a tissue-specific substrate preference towards PI(3,5)P2 in male sensory neurons.
Since PI(3)P localizes to early endosomes, we examined early endosomal protein distribution. RAB-5 is an early endosomal protein that recruits and activates PI(3)P generating PI 3-kinases . STAM-1 colocalizes with RAB-5 in C. elegans male specific sensory neurons and promotes polycystin trafficking from early endosomes to the ESCRT complex . In WT and cil-1(my15) animals, RAB-5 and STAM-1 localize to small puncta in the cell bodies, axons, and dendrites (SF4G-J), indicating that cil-1 does not modify the overall organization of early endosomes.
cil-1(my15) hermaphrodites exhibit a drastic reduction in brood size (4.85% of WT) (SF5A), which is due to a spermatogenesis defect (Spe) based on the following observations. (1) a my15 hermaphrodite lays 200-300 unfertilized eggs, which is comparable to the number of WT fertilized eggs (SF5A), (2) my15 male germline architecture and early stages of spermatogenesis (data not shown), hermaphrodite oocyte maturation, and ovulation appear normal (Suppl. Figure 4D-E), (3) my15 hermaphrodites contain endomitotic cells without eggshells in the uterus (SF5F-G), (4) the my15 fertility defect is completely rescued when mated with WT males (SF5B), (5) my15 males fail to sire cross progeny without overt behavioral defects in mating behaviors (SF5C and SF1G). We conclude that cil-1 is required for sperm function in both hermaphrodites and males.
A Spe phenotype may arise from developmental defects in spermatogenesis, sperm activation (spermiogenesis), sperm motility, or sperm-egg interactions. my15 male gonads have normal DAPI staining patterns for each meiotic stage in the gonad and normal number of spermatids (inactive 1N sperm) (data not shown). Thus, cil-1 is not required for early spermatogenesis up to spermatid production. Defective sperm-egg interactions are the basis of Spe phenotypes in spe-9, spe-38, trp-3/spe-41, and spe-42 (Figure 4A) (reviewed by ). In these Spe mutants, male-derived sperm normally develop, activate, crawl, and compete with endogenous hermaphroditic sperm, but cannot fertilize an oocyte. However, my15 male-derived sperm do not compete with endogenous hermaphrodite-derived sperm, as reflected by a large number of self-progeny and extremely low mating efficiency (ME) (1.9%, SF5C). ME of WT males is higher with my15 hermaphrodites (~95%) than control hermaphrodites (58%) (SF5C), illustrating that my15 endogenous sperm are nearly incapable of competing with WT male-derived sperm. Hence cil-1 acts in events between spermatid production and sperm-egg interactions.
During sperm activation, a round spermatid develops into a motile spermatozoon with a pseudopod (Figure 4A). This process can be mimicked in vitro by chemical activators such as the ionophore monensin or Pronase (reviewed by ). Within 15 minutes of Pronase application, the majority of WT male spermatids (63.4%, n=347) extend pseudopods (Figure 4B-B′) and the average length of spermatozoa is 8.13 ± 1.11 um (±stdev, n=95) (Figure 4B′). In contrast, only 47.8% (n=128) of my15 male spermatids develop pseudopods after Pronase activation (Figure 4C-C′) with a significantly shorter spermatozoa length (6.37 ± 0.61um (n=112), p =1.09E-28, Figure 4C′). The diameter of my15 spermatids is also slightly smaller (6.22 ± 0.54 um, n=136) than WT (6.73 ± 0.65 um, n=110, p=8.08E-07). We also observe in vivo sperm activation defects in my15 hermaphrodites. 83.3% of WT spermatozoa possess pseudopods (n=36, Figure 4B″) whereas only 8.3% do in my15 mutants (n=72, Figure 4C″). When present, my15 pseudopods are significantly shorter (average length of longest axes of sperm; 4.32 ± 0.38 um, n=6) than WT (5.43 ± 0.62 um, n=37, p=1.14E-4). Thus, cil-1 positively regulates sperm activation in vitro and in vivo.
To measure my15 sperm motility, we performed a time-lapse sperm tracking assay. WT male sperm crawl from the uterus to the spermatheca over time, with the majority localizing at the spermatheca in 16 hours (Figure 4D). In contrast, my15 sperm are not observed in the spermatheca at 16 hours despite being present at 4 hours (Figure 4D). These data suggest that my15 sperm display reduced motility and, consequently, are not retained in the hermaphrodite spermatheca. After in vitro chemical activation with monensin, WT sperm crawl at the rate of 0.33 ± 0.04 μm/sec (n=7) on glass slides (Suppl. Movie 3) whereas isolated my15 sperm are immotile (Suppl. Movie 4).
Sperm activation involves fusion of membranous organelles (MOs) to the PM, exocytosis, and pseudopod extension (reviewed by ). To examine MOs, we used the MO specific monoclonal antibody 1CB4  and the lipophilic dye FM1-43 . In WT, the majority of MOs are found at the cell periphery of round spermatids (Figure 4H). After activation, MOs are excluded from the pseudopod (Figure 4H′). In my15 sperm, 1CB4 staining for MO morphology and localization before and after activation appears WT (Figure 4I, I′). In both WT and my15, FM1-43 labels the PM of the round spermatids (Figure 4F-G) and, after activation, concentrates at the site membrane fusion on the spermatozoon cell body (Figure 4F′, G′, arrowheads). cil-1 is not essential for MO morphology, localization, or fusion.
The ultrastructure of my15 sperm is largely unaffected as determined by transmission electron microscopy (TEM). The early stages of spermatogenesis in WT and my15 appear very similar (data not shown). Occasionally, dissected males will produce spontaneously activated spermatozoa, and we examined such cells in both WT and my15 (Figure 4E, E′). In both cases, a pseudopod is extended and separated from the cell body by laminar membranes. MOs successfully fuse with the PM in my15, consistent with immunostaining and MO fusion assay (Figure 4G′, I′).
cil-1 may regulate membrane receptor localization in sperm. TRP-3/SPE-41 acts in sperm, translocates from MOs to the PM upon activation, and is required for fertilization . In WT and my15 spermatids, anti-TRP-3 labels cytoplasmic puncta partially overlapping with MOs (Figure 4H, I). In WT and my15 spermatozoa, TRP-3 is detected ubiquitously on the PM of the pseudopod and cell body (Figure 5H″, I″). cil-1 is not required for TRP-3/TRPC translocation to the PM.
Since PI biosensors show that CIL-1 hydrolyzes PI(3,4,5)P3 and PI(3,5)P2, we asked whether 3-kinase activity antagonizes CIL-1 5-phosphatase function. We applied 1nM, 10nM, and 100nM wortmannin, a pharmacological inhibitor of PI 3-kinases that produce PI(3,4,5)P3 from PI(4,5)P2 , to WT and my15 spermatids. At all concentrations tested, wortmannin acts as an in vitro activator of WT but not my15 spermatids, with 100nM being the most effective (compare Figure 4B with 4C). The low dose (1nM) of wortmannin suggests specificity, and indicates PI 3-kinase activity and cil-1 act antagonistically in sperm activation (Figure 5C, D).
The PI 5-phosphatase CIL-1 mediates multiple aspects of C. elegans male reproductive biology (Figure 5). In male neurons, cil-1 acts cell autonomously to control LOV-1 and PKD-2 localization but not function. Our data supports a model whereby cil-1 acts in early steps in TRPP complex downregulation by regulating PI(3)P distribution but not RAB-5 and STAM localization (Figure 5A, B). In sperm, cil-1 positively regulates sperm activation and motility without affecting major membrane trafficking events. The 3-kinase inhibitor wortmannin is a potent sperm activator, which indicates that PI(3,4,5)P3 is a major CIL-1 substrate in sperm. We propose that one of the three 3-kinase encoded in the C. elegans genome (VPS-34, AGE-1, or F39B1.1) acts antagonistically to CIL-1 in a sperm activation pathway (Figure 5C, D).
cil-1 may be required for cellular polarity as distinct PI enrichment is a hallmark of polarized plasma membrane domains [28, 29]. However, cil-1 mutants properly localize several other ciliary and presynaptic markers (SF1), and exhibit normal sensory behaviors, reflecting properly polarized and functional neurons. Alternatively, cil-1 may modulate the activity of the TRPP complex. PI(4,5)P2 is a regulator of both TRP channel activity and trafficking . Since cil-1 mutant males exhibit normal mating behaviors, cil-1 appears to regulate TRPP trafficking but not activity.
In neurons, cil-1 regulates the balance between PI species that is important for intracellular polycystin trafficking. While cil-1 may negatively regulate polycystin insertion into the plasma membrane after initial targeting, our data are inconsistent with this model. A similar PKD-2GFP Cil phenotype is observed in the adaptor protein 1 (AP-1) unc-101(m1) mutant background . However, PKD-2GFP particles are visibly moving along dendrites of both WT and my15 (Suppl Movies 1 and 2), but not unc-101 male sensory neurons (data not shown). Finally, PKD-2 abundance within ciliary regions is comparable between WT and my15 (Figure 1, ).
Our data are most consistent with a requirement for CIL-1 in TRP-polycystin trafficking to early endosomes after endocytosis (Figure 5A, B). In RnB neurons, cil-1 acts before STAM/HRS-mediated receptor downregulation, which transport PKD-2 and LOV-1 from early endosomes to ESCRT . In CEMs, it is difficult to place lov-1, cil-1, and stam-1 in a pathway due to additive phenotypes of double mutants. PI(3)P and PI(3,5)P2 balance in endocytic compartments is important for organelle maturation and receptor trafficking. In cil-1(my15), pre-early endosomal vesicles lacking the normal destination, the PI(3)P enriched early endosomes, may simply accumulate and disperse along sensory neurons. Alternatively, cil-1 may be required for maturation of endocytic compartments by affecting PI(3)P distribution, although this is unlikely given the normal localization of RAB-5 and STAM-1 in my15 male sensory neurons. We propose that maintaining the balance between PI(3)P and PI(3,5)P2 by the CIL-1 5-phosphatase plays an essential role in polycystin trafficking between endocytic compartments.
Our data also reveal that C. elegans sperm activation is coordinated by the antagonistic actions of the CIL-1 5-phosphatase and an unidentified wortmannin sensitive target. While we cannot rule out off-target effects of wortmannin, we propose that CIL-1 and a PI3-kinase control PI(3,4,5)P3 levels. PI(3,4,5)P3 recruits and activates various effectors such as serine/threonine kinases and tyrosine kinases . Cytoplasmic protein kinases and protein phosphatases are significantly overrepresented in C. elegans sperm . Molecules such as SPE-6 serine threonine kinase  may regulate sperm activation downstream of CIL-1 mediated turnover of PI(3,4,5)P3.
TRPP family members play important roles in mating and fertilization of numerous species [3, 7-12]. ADPKD male patients may exhibit infertility due to sperm immotility . Sea urchin suREJ3 (receptor for egg jelly, a polycystin-1 homolog) and suPC2 (polycystin-2) localize to the PM proximal to acrosomal vesicles in sperm [8, 34]. In mice, PKD-REJ controls postcopulatory reproductive selection via effects on sperm transport and exocytic competence during the acrosome reaction . In the hermaphroditic chordate Ciona intestinalis, the PKD1 homolog is expressed in testes and may act on sperm to control self-incompatibility . In Drosophila, PKD2 may mediate sperm directional movement [9, 10]. Intriguingly, mutations in the Type IV 5-phosphatase INPP5E gene cause human ciliary diseases [35, 36]. According to our phylogeny tree (SF2A), cil-1 is the closest C. elegans homolog of INPP5E. Further experiments are needed to determine whether CIL-1 plays an evolutionarily conserved role in TRPP complex trafficking and sperm function in other species.
Nancy L'Hernault provided assistance with electron microscopy. We thank Natalia Morsci, Dr. Jinghua Hu, and members of Barr laboratory, Dr. John Archibald, Ching Kung, and Andy Singson for helpful discussion, Drs. David Sherwood and Joshua Ziel for critical reading of the manuscript, Dr. Barth Grant for intestinal PI markers, Dr. Shawn X-Y. Xu for the anti-TRP-3 antibody, and Dr. Sam Ward for sharing unpublished data. Some nematode strains used in this work were provided by the Caenorhabditis Genetics Center, which is funded by the NIH National Center for Research Resources (NCRR), and by Dr. Shohei Mitani of the National Bioresource Project for the Nematode (Japan). This work was supported by the NIH (DK059418 and DK074746 to MMB, GM082932 to SLH), NSF (0131532 to SLH), and the PKD Foundation (to MMB).
Experimental Procedures: Refer to Supplement
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.