Proteasomal ubiquitin receptor RPN-10 controls sex determination in Caenorhabditis elegans

Abstract

The ubiquitin-binding RPN-10 protein serves as a ubiquitin receptor that delivers client proteins to the 26S proteasome. Although ubiquitin recognition is an essential step for proteasomal destruction, deletion of the rpn-10 gene in yeast does not influence viability, indicating redundancy of the substrate delivery pathway. However, their specificity and biological relevance in higher eukaryotes is still enigmatic. We report herein that knockdown of the rpn-10 gene, but not any other proteasome subunit genes, sexually transforms hermaphrodites to females by eliminating hermaphrodite spermatogenesis in Caenorhabditis elegans. The feminization phenotype induced by deletion of the rpn-10 gene was rescued by knockdown of tra-2, one of sexual fate decision genes promoting female development, and its downstream target tra-1, indicating that the TRA-2-mediated sex determination pathway is crucial for the Delta rpn-10-induced sterile phenotype. Intriguingly, we found that co-knockdown of rpn-10 and functionally related ubiquitin ligase ufd-2 overcomes the germline-musculinizing effect of fem-3(gf). Furthermore, TRA-2 proteins accumulated in rpn-10-defective worms. Our results show that the RPN-10-mediated ubiquitin pathway is indispensable for control of the TRA-2-mediated sex-determining pathway.

Figure 1.

(A) Schematic representation of the 26S proteasome. The 26S proteasome is composed of the catalytic 20S proteasome and a regulatory particle termed PA700, which can be subdivided into the Base and the Lid complex. RPN-10 is an intrinsic ubiquitin receptor of 26S proteasomes and is highlighted in the figure. It was also reported that there are alternative ubiquitin receptors of 26S proteasome, UBL-UBA domain proteins, and that they interact with RPN-1 and possibly RPN-2 subunits of the Base complex of 26S proteasome. (B) Schematic diagram of RPN-10 ubiquitin receptor. Previous studies have shown that RPN-10 can bind to the Base complex via N-terminal VWA domain and bind to polyubiquitinated substrates via its C-terminal domain, including polyubiquitin (poly-Ub)-binding UIMs.

Figure 2.

RPN-10 knockdown results in F1 sterility in C. elegans. (A) Schematic diagram of RNAi treatment. (B) Effects of RNAi on the reproduction of P0 parents. The average numbers of F1 progeny per a P0 parent are indicated. rpn-1 (RNAi) was used as a positive control for this RNAi experiment. (C) Effects of RNAi on the reproduction of F1 parents. The average numbers of F2 progeny per an F1 parent are indicated. Note that double RNAi treatment of rpn-10 and ufd-2 results in a nearly complete sterile phenotype. (D) Sedimentation velocity analysis of the 26S proteasome. Worm extracts were fractionated by glycerol density-gradient centrifugation and assayed for 26S proteasomal chymotrypsin-like activity. Elution points of 20S and 26S proteasomes are indicated by arrows. (E) Proteins in the 26S proteasome fraction (Fr. 21) were precipitated with acetone and subjected to immunoblotting with anti-RPN-10 antibody. (F) Expression of RPN-10 protein is suppressed in rpn-10 (RNAi) worms. Immunosignal of the 20S proteasome C2 subunit was used as a control. (G) RPN-10 knockdown results in accumulation of polyubiquitinated (poly-Ub) proteins. α-tubulin as a loading control.

Figure 3.

RPN-10 knockdown induces characteristic reproductive defects. Compared with the WT gonad (A), typical rpn-10 (RNAi) worms show expanded oocytes (indicated by a red box) and nearly empty uterus (indicated by a yellow box) (C), representing similar morphology to oocytes-defective oma-2 (tm911);oma-1 (RNAi) worms (F). Double RNAi of rpn-10 and ufd-2 results in more penetration but essentially the same phenotype as in C (B). ufd-2 (RNAi) alone shows no abnormality in gonads (D). Knockdown of the expression of RPN-1, a constitutive subunit of the 26S proteasome, induces immediate growth arrest (E). Feminized fog-2(oz40) (G) and fem-2(b245ts) mutant (H), show accumulation of unfertilized oocytes in the uterus due to their stochastic maturation. Images on the left-top side correspond to the ventral uterus. Bar, 50 μm.

Figure 4.

The primary defect in RPN-10-knockdown worm is sperm formation. (A) Mating experiments with normal males and sterile worms. With excess males, feminized fem-2(b245ts) mutants were able to produce progeny, whereas oocyte-defective oma-1;oma-2 (RNAi) worms were not. rpn-10;ufd-2 (RNAi) sterile worms can be cross-fertilized by wild-type (N2) males to produce normal viable F2 embryos. (B) Hoechst staining of dissected gonads from WT, rpn-10;ufd-2 (RNAi), fem-2(b245ts), and oma-1;oma-2 (RNAi) worms. A number of sperm signals are evident in WT and oma-1;oma-2 (RNAi) worms as indicated but are absent in fem-2(b245ts) and rpn-10;ufd-2 (RNAi) worms. Bar, 10 μm. (C) RT-PCR analysis of the sperm-specific marker msp-77. Transcript of msp-77 is not expressed in fem-2(b245ts) and rpn-10;ufd-2 (RNAi) worms. inf-1 transcript (encoding translation initiation factor CeIF) amplified under similar conditions was used as a control.

Figure 5.

tm1349, a null allele of rpn-10, shows a feminized phenotype. (A) Schematic diagram of wild-type and tm1349 mutant allele of the rpn-10 gene. C. elegans rpn-10 gene is composed of four exons, and the tm1349 mutant completely lacked the first and second exons with partial deletion of the third exon. Boxes, exons; lines, introns. Extent of the 882-base pair rpn-10 (tm1349) deletion is indicated by a gap. (B) Nucleotide sequence of the mutated region of the rpn-10 gene in tm1349 allele. rpn-10 cDNA (genome) obtained from a tm1349 homozygote was sequenced, and a 882-base pair deletion and 8-base pair insertion were identified in the region indicated. No sequence difference was identified around this region before and after backcrossing. The numbers of the sequence denote the wildtype genome nucleotide number of chromosome I. (C) Genomic PCR analysis confirming deletion of the rpn-10 gene in rpn-10 (tm1349) heterozygote and homozygote worms. (D) The average F1 sterilities at 25°C are indicated. Note that F1 of the rpn-10 (tm1349) homozygote shows nearly complete sterility. (E) rpn-10 (tm1349) mutant adult worms show expanded oocytes (indicated by a red box) and a nearly empty uterus (indicated by a yellow box). Hoechst staining of dissected gonads shows essentially no sperm in rpn-10 (tm1349). Spermatheca is indicated by a white box. Photographs were taken at the adult stage (54 h from L1 at 25°C). (F) Nomarski observations of gametogenesis at the young adult stage of an rpn-10 (tm1349) homozygote worm. Proximal oocytes and primary and secondary spermatocytes (labeled “Primary Sp.” and “Secondary Sp.,” respectively) are indicated. Bar, 10 μm.

Figure 6.

Sterility induced by deletion of the rpn-10 gene was rescued by knockdown of tra-2. Schematic diagram of the somatic (A) and germline (B) sex determination pathway in C. elegans. Accumulation of TRA-2 and TRA-1 proteins, indicated by red, induces the feminization phenotype. (C) Results of a “rescue” experiment. RNAi of either tra-1 or tra-2 rescued the sterile phenotype of rpn-10 (tm1349) to produce normal viable F2 embryos. Control vector RNAi did not influence the sterile phenotype. (D) Oocyte maturation defects of rpn-10 (tm1349) can be restored by knockdown of the expression of tra-2 and tra-1. Note that the expanded oocytes and empty uterus observed in rpn-10 (tm1349) worms were restored in tra-2 (RNAi) individuals. Hoechst staining of dissected gonads from rpn-10 (tm1349);vector control (RNAi) and rpn-10 (tm1349);tra-2 (RNAi) worms. Restored spermatheca is indicated by a white box. Bar, 10 μm. (E) At P0 stage, all worms were treated with firstRNAi (rpn-10;ufd-2). This procedure ensures F1 sterility even in the absence of RNAi treatment at the F1 stage. From the young adult stage of P0 to F1 larval development, individuals were treated with second RNAi with a vector control, sdc-1 (RNAi), tra-1 (RNAi), and tra-2 (RNAi), respectively. The resulting number of F2 progeny per an F1 parent was counted. Second RNAi of tra-2 and tra-1 restored sperm formation. (F) RT-PCR analysis of the spermspecific marker msp-77. Expression of msp-77 is restored in rpn-10; ufd-2;tra-2 (RNAi) worms but not in rpn-10;ufd-2;vector control (RNAi) worms. inf-1 transcript amplified under similar conditions was used as a control.

Figure 7.

RPN-10 acts cell-autonomously in the germline. (A) RNAi analysis under an rrf-1 (pk1417) genetic background, which is fully proficient for RNAi responses in the germline but not the soma. Nomarski observations indicate a typical sign of feminization of the germline in rrf-1 (pk1417) homozygotes that were treated with rpn-10; ufd-2 (RNAi). Inset, Hoechst staining of dissected gonads show absence of sperm in the spermatheca. Bar, 10 μm. (B) PCR verification of a double homozygote mutant, rpn-10 (tm1349); rrf-1 (pk1417). (C) Results of a “rescue” experiment. RNAi of either tra-1 or tra-2 rescued the sterile phenotype of rpn-10 (tm1349); rrf-1 (pk1417) to produce viable embryos. Control vector RNAi did not influence the sterile phenotype.

Figure 8.

Co-knockdown of rpn-10 and ufd-2 overcomes the germline-masculinizing effect of fem-3 (gf). P0 larvae of fem-3 (gf) XX homozygous mutants were treated with either rpn-10;ufd-2 (RNAi) or control vector (RNAi) at a permissive temperature (20°C), and resulting F1 larvae were synchronized and continued to treated with respective RNAi at a restrictive temperature (25°C). Resulting F1 adult worms (55 h after hatching) were observed (A). In control vector (RNAi) worms, complete germline masculinization was observed, whereas >70% of rpn-10;ufd-2(RNAi)-treated F1 showed mutual suppression of germline-masculinizing phenotypes of fem-3 (gf), resulting in self-fertility at a restrictive temperature (25°C) (B).

Figure 9.

Co-knockdown of rpn-10 and ufd-2 cause weak XO intestinal feminization. (A) Tail morphologies of rpn-10;ufd-2 (RNAi) and wild-type XO males (a and b). Hoechst staining of XO male gonads from rpn-10;ufd-2 (RNAi) and wild-type worms (c and d). Bar, 10 μm. (B) RT-PCR quantification of the somatic feminization marker vit-2 (422 base pairs, exons 4 and 5) and the sperm-specific marker msp-77 (360 base pairs) from a single individual worm. Note that the expression levels of msp-77 signal in hermaphrodites were too low to detect under the conditions used in this experiment. inf-1 transcript was used as a amplification control. (C) More than 70% individual males treated with rpn-10;ufd-2 (RNAi) expressed increased amounts of vit-2, whereas no wild-type male express it.

Figure 10.

Establish of anti-TRA-2 ICD antibody. (A) RT-PCR-based expression analysis of tra-2 transcript (top; 519 base pairs, exons 22 and 23). inf-1 transcript (586 base pairs, exons 3–5) amplified under similar conditions was used as a control (bottom). (B) FLAG-tagged TRA-2 protein (fragments of amino acids 1135-1475 and 1135–1413, respectively) was expressed in COS7 cells, and the extracts were blotted with either an antibody against the C-terminal fragment (ICD) of TRA-2 or anti-FLAG M2 antibody (Sigma-Aldrich). (C) Fifty individual animals showing an appropriate phenotype were picked up and boiled with SDS sample buffer and then subjected to immunoblot experiments. The anti-TRA-2 ICD antibody can detect the endogenous 54-kDa band (indicated by an arrow) as well as several lower-molecular-weight fragments in the extracts. A tra-2 (e1095) putative null mutant was used as a negative control. Note that the 85-kDa signal is not specific because preimmune control serum stained the same band as well.

Figure 11.

TRA-2 protein accumulates in RPN-10-defective worms. (A) fem-3 (e1996lf) is a putative null allele and developed as a true female regardless of whether tra-2 gene was knocked down or not. Even in a background of fem-3 (e1996lf), rpn-10;ufd-2 (RNAi) enhanced the nuclear accumulation of TRA-2, whereas tra-2 (RNAi) completely remove the staining. (B) The intestinal nuclei of rpn-10;ufd-2 (RNAi)-treated worms as well as gain-of-function mutant tra-2 (e2020) were stained more strongly than those of wild-type worms (vector RNAi control), whereas nuclei of loss-of-function mutant tra-2 (e1095) and wild-type male were scarcely immunostained with anti-TRA-2 ICD antibody. Note that all photographs were taken with identical exposure time. Bar, 10 μm.

Figure 12.

rpn-10 mutant worm tm1180 shows significantly increased immunosignal of TRA-2 in nuclei of the intestine. The nuclear staining was eliminated by potent tra-2 (RNAi) treatment. Note that worms used in this experiment were fixed at the late larval stage (L4 to young adult). All photographs were taken with identical exposure time. Bar, 10 μm.