PAB-1, a Caenorhabditis elegans Poly(A)-Binding Protein, Regulates mRNA Metabolism in germline by Interacting with CGH-1 and CAR-1

Poly(A)-binding proteins are highly conserved among eukaryotes and regulate stability of mRNA and translation. Among C. elegans homologues, pab-1 mutants showed defects in germline mitotic proliferation. Unlike pab-1 mutants, pab-1 RNAi at every larval stage caused arrest of germline development at the following stage, indicating that pab-1 is required for the entire postembryonic germline development. This idea is supported by the observations that the mRNA level of pab-1 increased throughout postembryonic development and its protein expression was germline-enriched. PAB-1 localized to P granules and the cytoplasm in the germline. PAB-1 colocalized with CGH-1 and CAR-1 and affected their localization, suggesting that PAB-1 is a component of processing (P)-bodies that interacts with them. The mRNA and protein levels of representative germline genes, rec-8, GLP-1, rme-2, and msp-152, were decreased after pab-1 RNAi. Although the mRNA level of msp-152 was increased in cgh-1 mutant, it was also significantly reduced by pab-1 RNAi. Our results suggest that PAB-1 positively regulates the mRNA levels of germline genes, which is likely facilitated by the interaction of PAB-1 with other P-body components, CGH-1 and CAR-1.


Introduction
Poly(A)-binding proteins (PABPs) have been identified in many organisms, from yeast, flies, and mice to human [1][2][3][4]. PABPs are classified into 2 categories: nuclear PABPs, which are necessary for mRNA maturation; and cytoplasmic PABPs, which are involved in mRNA metabolism [5,6]. PABPs are composed of RNA recognition motifs and a C-terminal domain that is necessary for protein-protein interaction [7]. PABPs influence many aspects of mRNA metabolism, such as transport from the nucleus to the cytoplasm, protection of mRNAs, and translation through the formation of a "closed loop" structure [8][9][10]. PABPs are also components of the stress granules in which some mRNAs are transiently sequestered during stress in mammals [11]. Stress granules interact with processing (P)-bodies in mammalian cells [12]. Pbodies are cytoplasmic aggregates of RNA and proteins that contain translational repressors and mRNA decay machinery [13,14]. P-bodies are concentrated into foci and some components are diffused throughout the cytoplasm [15]. In Caenorhabditis elegans, stress granules and P-bodies are colocalized and a subset of genes that enhances formation of P-bodies also regulates the formation of stress granules [16]. Germ granules are germline-specific cytoplasmic organelles found in the germ cells of many species including C. elegans, where they are called as P granules [17]. P granules consist of mRNAs and multiple RNA-binding proteins, some of which were shown to be essential for the germline development. It was also shown that some of C. elegans P-body component orthologs, including CGH-1 and CAR-1, colocalize to P granules in the germ cells [15,[18][19][20]. In C. elegans, 2 cytoplasmic poly(A)-binding proteins, PAB-1 and PAB-2, and a nuclear poly(A)-binding protein, PABP-2 (also known as PAB-3), have been identified. PAB-1 and PAB-2 are functionally redundant in the soma, but PAB-1 is essential for the germline development [21][22][23].
C. elegans adult hermaphrodite gonads contain mitotic germline stem cells, meiotic germ cells, and differentiating gametes from the distal to proximal orientation [24,25]. In the first larval stage (L1), the gonad primordium contains 2 somatic gonadal precursor cells, Z1 and Z4, and 2 primordial germ cells, Z2 and Z3. The number of germ cells gradually increases during larval development to as many as 1000 germ cells in each of the 2 gonadal arms and they form syncytia in an adult hermaphrodite [24,25]. During the course of development, all the germ cells, except mature sperm, contain P granules in their perinuclear region, which eventually disperse into the cytoplasm during oogenesis [17,26]. In the gonad arm, most germline mRNAs are transcribed primarily in the pachytene stage germ cells and are transported through the nucleus-free core of germline syncytial cytoplasm, called the rachis, to the proximal region; translation occurs in a temporally and spatially modulated fashion depending on the needs of each protein product [27][28][29][30][31]. During these processes, multiple germlineenriched RNA-binding proteins play critical roles [32]. Therefore, elucidating the regulatory mechanisms of mRNA metabolism by these RNA-binding proteins is essential to understand the process of C. elegans germline development at the molecular level.
pab-1 mutants show mitotic proliferation defects in the germline [21][22][23]. Nevertheless, it has not been clear whether pab-1, a seemingly general mRNA regulator, has a specific function only for mitotic proliferation; or pab-1 also has essential functions in later stages of germline development, but those phenotypes are masked by the proliferation defects. In this study, we examined possible functions of pab-1 in the later stages of germline development by treating synchronized worms with pab-1 RNAi at each larval stage. We also examined possible functional association of PAB-1 with other RNA-binding proteins. Here, we show that PAB-1 colocalizes and interacts with P-body components, CGH-1 and CAR-1, and promotes the entire processes of postembryonic germline development by maintaining the mRNA levels of germline genes. An earlier study demonstrated that PAB-1 is actively involved in translation of germline proteins through its cosedimentation with polysomes [33]. Our data obtained in this study reveal that PAB-1 is also involved in mRNA metabolism through its association with P-body components.

pab-1 is required throughout germline development
In the previous study, we demonstrated that pab-1 mutants were defective in germline stem cell proliferation [22]. To understand further the functions of PAB-1 during germline development, we treated the rrf-1(pk1417) mutant, in which RNAi is effective only in the germline and limited somatic tissues [34], with pab-1 dsRNA at each larval stage for 24 hours. The pab-1 RNAi-treated worms were then recovered to seeded NGM plates, allowed to develop to adult stage, and their germline was observed after double immunostaining with anti-PGL-1, a germline-specific P-granule marker ( Figure 1A1, B1, C1, D1, E1), and SP56 [35], a sperm marker ( Figure 1A2, B2, C2, D2, E2), along with TO-PRO-3 nuclear staining ( Figure  1A3, B3, C3, D3, E3). pab-1 RNAi treatment at the L1 stage resulted in germline proliferation defects, as observed in pab-1 mutants ( Figure 1B1-B3). We found that these germ cells were arrested before entering meiosis because they were negative for immunostaining with anti-HIM-3, a meiotic marker (Figure S1A-F) [36], indicating that pab-1 RNAi phenotype is different from that of glp-1(lf) mutants, which exhibit premature meiotic entry without mitotic proliferation [37]. When pab-1 RNAi was treated at the L2 stage, germ cells of the worms were moderately proliferated and entered meiosis (Figure S1G-I), but sperm were not produced ( Figure 1C1-C3). When worms were treated with pab-1 RNAi at the L3 stage, the worms produced sperm but not oocytes ( Figure 1D1-D3). These worms, which were pab-1 RNAi treated at the L1, L2, or L3 stage, all became sterile adults. pab-1 RNAi at the L4 stage resulted in defective oogenesis ( Figure 1E1-E3), as well as severely reduced brood size (17.1 ± 2.6, n=15) and high embryonic lethality of their progeny (50.8%, n=15). Moreover, the hatched embryos became sterile adults (data not shown).
These results indicate that PAB-1 is required throughout the postembryonic germline development. This conclusion was further supported by the following observations: First, the expression level of pab-1 mRNA increased as the number of germ cells, which was scored as the number of PGL-1-positive nuclei in the gonad, increased during the larval development in wild-type N2 ( Figure 1H). Second, the level of PAB-1 protein, which was detected by our original rabbit polyclonal anti-PAB-1 antibody (see Materials and Methods), was drastically decreased in the glp-1(q231) mutant, which contained few germ cells, compared to wild-type N2 ( Figure 1I). pab-1 mRNA level was also significantly decreased in the glp-1(q231) mutant ( Figure S2). Given that PAB-1 was still weakly detected in glp-1(q231) mutant, PAB-1 is likely enriched in but not specific to the germline. We also observed that after pab-1 RNAi treatment, P granules were diffused ( Figure 1G1-G3). This observation suggests that PAB-1 affects P granule integrity or assembly. The effectiveness of pab-1 RNAi depletion in rrf-1(pk1417) mutant was confirmed by western blot analysis of the rrf-1(pk1417) mutant total worm proteins with or without pab-1 RNAi treatment using anti-PAB-1 antibody, and the specificity of the anti-PAB-1 antibody was demonstrated using pab-1 mutant alleles, bn116 and bn119 ( Figure 1I).

PAB-1 Localizes to P Granules and the Cytoplasm in the germline and Embryos
To determine the localization of PAB-1, embryos and adult gonads of wild-type N2 were immunostained with anti-PAB-1 ( Figure 2). PAB-1 localized to P granules and it was also dispersed in the cytoplasm in embryos (Figure 2A-D) and extruded adult gonads ( Figure 2E-H). Localization of PAB-1 to P granules was demonstrated by co-immunostaining with OIC1D4 [38], a monoclonal antibody that recognizes P granules ( Figure 2B, F), along with Hoechst 33342 nuclear staining ( Figure 2C, G). PAB-1 granules that were not positive by OIC1D4 immunostaining were also observed in some somatic blastomeres in the embryos ( Figure 2D, arrows). This expression pattern was also observed for some P-body components [15], suggesting that PAB-1 is a component of Pbodies.

PAB-1 affects the localization of CGH-1 and CAR-1
Having observed that PAB-1 colocalized with CAR-1 and CGH-1, to further explore the interaction among them, we examined the localization of each protein in the absence of other P-body components by immunostaining ( Figure 4). PAB-1 localized to P granules around the germ nuclei, and it was also evenly dispersed in the rachis in adult gonads of wildtype N2 ( Figure 4A). However in cgh-1(ok492) mutants, the localization of PAB-1 was significantly affected ( Figure 4A). PAB-1 was aggregated as patched structures in the rachis of cgh-1(ok492) gonads. The PAB-1 aggregates colocalized with CAR-1 in the cgh-1(ok492) gonads (data not shown). In contrast to cgh-1(ok492) mutant, localization of PAB-1 was not affected in car-1(tm1753) gonads ( Figure 4A). These results suggest that the localization of PAB-1 is influenced by CGH-1 but not by CAR-1. Altered localization of PAB-1 was also observed in embryos after cgh-1 RNAi ( Figure 4D). PAB-1 foci were concentrated mainly in the P 2 blastomere in the control (mock RNAi) 4-cell embryo, whereas in the cgh-1(RNAi) 4-cell embryo, they were dispersed to all the blastomeres ( Figure  4D).

Depletion of pab-1 decreased the mRNA levels of germline-enriched genes
To determine whether PAB-1 plays a role in mRNA stability in the germline, we measured mRNA levels of representative germline-enriched genes by quantitative real-time RT-PCR. rec-8 and glp-1, which are strongly expressed in mitotic germ cells, and rme-2 and msp-152, which are specifically expressed during oogenesis and spermatogenesis, respectively, were examined with or without pab-1 RNAi, and their relative mRNA levels were shown after normalization to act-1 ( Figure 5). act-1 was used as an internal control because its mRNA level was not significantly changed with or without pab-1 RNAi (data not shown). The effectiveness of pab-1 RNAi depletion was confirmed by simultaneously measuring the mRNA level of pab-1 after pab-1 RNAi in each set of experiments ( Figure 5). pab-1 RNAi was treated for 24 hours to the L4-stage rrf-1(pk1417) hermaphrodites, which already contained well proliferated mitotic germ cells. Therefore, numbers of germ cells were not significantly decreased by this RNAi treatment. Further, the RNAi treated worms were recovered to seeded NGM plates and cultured for 2 more days before harvesting them as adult worms. We found that after pab-1 RNAi PAB-1 Regulates mRNA Metabolism PLOS ONE | www.plosone.org treatment, rec-8 and glp-1 mRNA levels were reduced to 55% and 41%, respectively, of the mock RNAi control levels ( Figure  5A). rme-2 mRNA level was reduced to 75% of the mock RNAi control level after pab-1 RNAi in feminized fem-1(lf) worms, which produce only oocytes ( Figure 5B). msp-152 mRNA level was reduced to 20% of the mock RNAi control level after pab-1 RNAi in masculinized fem-3(gf) worms, which produce only sperm ( Figure 5C). These results indicate that RNAi depletion of pab-1 at the L4 stage significantly decreased the mRNA levels of multiple germline-enriched genes without significantly affecting the numbers of germ cells that expressed these genes. This finding was further verified by measuring the protein levels of REC-8, GLP-1, and RME-2 with or without pab-1 RNAi at the L4 stage either by digitally quantifying immunostaining signals ( Figure 6A, B, D, E, G, H) or by western blot analysis ( Figure 6C, F, I). The results of immunostaining signal quantification revealed that protein levels of REC-8, GLP-1, and RME-2 were reduced to 52%, 57%, and 14%, respectively, of the mock RNAi control levels after pab-1 RNAi ( Figure 6B, E, H). The results of western blot analysis showed that protein levels of REC-8, GLP-1, and RME-2 were reduced to 73%, 85%, and 44%, respectively, of the mock RNAi control levels after pab-1 RNAi ( Figure 6C, F, I). These results confirmed that protein levels of germlineenriched genes were reduced along with their mRNA levels after pab-1 RNAi depletion.

PAB-1 is required for the maintenance of germlineenriched mRNAs
CGH-1 is required for the accumulation of a subset of germline-enriched mRNAs in C. elegans [33,41,42]. PAB-1 and CGH-1 mutually affected the other's localization, and depletion of pab-1 decreased the mRNA levels of several germlineenriched genes in our study. Therefore, we examined the  (Figure 7), as previously reported [33]. In contrast, the mRNA level of msp-152 was increased 1.5-fold in the cgh-1 mutant compared to wild-type N2 (Figure 7). We then investigated whether RNAi depletion of pab-1 can reduce the high level of msp-152 mRNA in the cgh-1 mutant. The mRNA level of msp-152 was indeed reduced to 2.1% of the mock RNAi control level after pab-1 RNAi in the cgh-1 mutant ( Figure  7). On the other hand, the mRNA level of rme-2 was not significantly changed after pab-1 RNAi in the cgh-1 mutant, possibly because the mRNA level of rme-2 was already reduced in the cgh-1 mutant, and pab-1 RNAi treatment, therefore, could not further reduce the mRNA level of rme-2 in the cgh-1 mutant. These results suggest the following. First, PAB-1 is required for the accumulation or stabilization of a subset of germline-enriched mRNAs including msp-152. Second, CGH-1 may function not only towards the

Discussion
Previous studies carried out by us as well as by others showed that pab-1 mutants exhibit a specific defect in germline stem cell proliferation [21][22][23]. In this study, we demonstrated that pab-1 RNAi treatment at each larval stage resulted in the arrest of germline development at the following stage ( Figure  1). These results indicate that pab-1 has essential functions not only for germline stem cell proliferation but also for later stages of germline development. These functions, masked in the studies of pab-1 mutants with proliferation defects, have been revealed for the first time in this study by a series of stagespecific pab-1 RNAi treatments.
There are 2 cytoplasmic PABPs, PAB-1 and PAB-2, in C. elegans. The pab-2 gene is located on the X chromosome; thus, germline specific inactivation of the X chromosome prevents its expression in the germline [23]. On the other hand, PAB-1 protein and pab-1 mRNA was weakly detected in the glp-1(q231) mutant, which contains few germ cells (Figure 1), indicating that PAB-1 is germline-enriched but is also expressed weakly in the soma. Due to the soma-limited pab-2 activity, pab-1 mutants or pab-1 RNAi would show germlinespecific defects without causing any somatic defects. Nevertheless, the pab-1(ok1656) deletion null mutant showed larval arrest (data not shown). This result indicates that depletion of pab-1 activity can cause somatic defects under certain conditions. To suppress possible somatic defects, we used rrf-1 mutants for pab-1 RNAi analysis in this study. In rrf-1 mutants, RNAi is effective only in the germline and limited somatic tissues [34,43].
PAB-1 was shown to colocalize with CGH-1 and CAR-1, which are RNA-binding proteins and components of P-bodies ( Figure 3). P-bodies are cytoplasmic aggregates consisting of RNAs and proteins, which include translational repressors and components of the mRNA decay machinery [13,14]. Among the components of P-bodies, a DEAD-box RNA helicase, Dhh1/ Me31B/RCK/p54, is involved in the translational repression and decapping of mRNAs, and CGH-1 is a C. elegans ortholog of Dhh1 [19,44,45]. CGH-1 promotes the stability of numerous maternal mRNAs in some cases [33,41,42]. On the other hand, CAR-1, an ortholog of Scd6/trailer hitch/RAP55 containing Smlike and FDF domains represses translation of target mRNAs [46][47][48]. Colocalization of PAB-1 with CGH-1 and CAR-1 suggests functional interactions among them. Our finding that PAB-1, CGH-1, and CAR-1 mutually affect each other's localization further supports this view (Figure 4). In previous reports, PAB-1 and CAR-1 were found in CGH-1 immunoprecipitates [18,41]. However, CGH-1 was absent after RNase treatment in CAR-1 immunoprecipitates [20]. These observations suggest that interactions among PAB-1, CGH-1, and CAR-1 are RNA-mediated. In pab-1 depleted conditions, CGH-1 and CAR-1 formed aggregates in the rachis ( Figure 4B-C). To test the possibility that generation of these aggregates is a consequence of general translational arrest, rather than because they require PAB-1 for their proper localization, we observed CGH-1 and CAR-1 localization after RNAi treatment of ifg-1, which encodes the sole C. elegans ortholog of eIF4G, an essential component of translation initiation complex ( Figure   PAB -1(q224), a germline proliferation defective mutant, was used as a negative control. α-Tubulin was used as a loading control. (F) The antibody against the intracellular region of GLP-1 recognized 160 kDa and 95 kDa protein bands as indicated [27]. P values were calculated by Student's t-test. *p < 0.05.   S3G-I) [49]. After ifg-1RNAi treatment, CGH-1 and CAR-1 also formed minor aggregates, but the degree of accumulation of CGH-1 and CAR-1 in the rachis was not so prominent compared to that in pab-1(RNAi) gonads ( Figure S3D-F). From these observations we consider that aggregation of CGH-1 and CAR-1 after pab-1 depletion is not a simple consequence of a translational arrest.
RNAi depletion of pab-1 decreased the mRNA levels of rec-8, glp-1, rme-2, and msp-152, 4 representative germline genes that function at different developmental stages ( Figure  5). In addition, their protein levels were also reduced after pab-1 RNAi (Figure 6). These results indicate that PAB-1 is required to maintain the mRNA and protein levels of multiple germline genes. This may be achieved by promoting mRNA stability or stimulating translational initiation, given that PABP is actively involved in these processes in yeast and mammals [50,51]. PAB-1 was enriched in the polysome fractions of total worm extract, suggesting that PAB-1 is actively involved in translation [33]. Among the 4 germline genes, the protein level of RME-2 was more significantly reduced ( Figure 6H, I) than the mRNA level ( Figure 5B) after pab-1 RNAi compared to the untreated controls. This supports the view that PAB-1 regulates gene expression not only at the mRNA level but also at the translational level.
In the cgh-1 mutant, the mRNA level of msp-152 was increased, whereas that of rme-2 was decreased ( Figure 7). However, the increased msp-152 mRNA level in the cgh-1 mutant was significantly reduced upon pab-1 RNAi treatment as in wild-type N2 (Figure 7). This result indicates that PAB-1 functions consistently as a positive regulator for the stabilization or accumulation of mRNAs ( Figure 8). Our study suggests that PAB-1, CGH-1, and CAR-1 regulate the mRNA levels of germline genes by associating and by functionally cooperating for the proper germline development.

dsRNA Interference (RNAi)
RNAi was performed by the soaking method with minor modifications [53]. The DNA template for pab-1 dsRNA synthesis was amplified from N2 cDNA by PCR with a T7 promoter sequence using the following primers: 5′-GTAATACGACTCACTATAGGGCG-AAATGAACGTCGCT-3′ and 5′-GTAATACGACTCACTATAGGGCTTGCTTCTGAGCG-G-3′. In vitro transcription and purification of dsRNA were performed as described previously [54]. To examine the effects of pab-1 depletion at each developmental stage, synchronized worms were soaked in dsRNA solution at each stage for 24 hours at 20°C, transferred to NGM plates seeded with OP 50, and allowed to develop to adults. Their germline development was then examined. Each set of experiment was repeated three times using 100 worms. To measure mRNA and protein levels after RNAi, L4 hermaphrodites (48 hours after synchronized L1) were treated with dsRNA for 24 hours. The treated worms were transferred to seeded plates and allowed to recover for 2 days. To measure mRNA levels after RNAi in cgh-1(ok492), young adult worms were treated. For cgh-1 and ifg-1 RNAi, dsRNA was transcribed in vitro from amplified cDNA templates flanked by T7 promoter sequences. The cDNA template was PCR-amplified from the yk85e1 and yk450c12 clone, respectively, using the primer pair 5′-GCGTAATACGACTCACTATAGGGAACAAAAGCTGGAGCT-3 ′ and 5′-GTAATACGACTCACTATAGGGC-3′. The yk cDNA clones were generously provided by Y. Kohara (National Institute of Genetics, Japan).
The relative mRNA expression level of each gene was averaged from triplicate measurements of 3 independent experiments and normalized to that of act-1, indicated as the control value as 1.

Western blot analysis
For western blot analysis, up to 150 worms of each strain were collected in 20 µL sample buffer (10% glycerol + 60 mM Tris-HCl, pH 6.8, + 4% SDS + 0.05% bromophenol blue + 5% 2-mercaptoethanol) and then boiled at 100°C for 8 minutes. The protein samples were centrifuged at 13000 rpm for 10 minutes, and loaded onto 10% SDS-PAGE gels to detect PAB-1 and 7.5% SDS-PAGE gels to detect GLP-1, REC-8, RME-2, and MSPs and then transferred to PROTRAN membranes (Whatman) with a current of 60 mA for 75 minutes. Antibody dilutions were as follows: rabbit anti-PAB-1 (1:1000), rabbit anti-GLP-1 (1:5), rabbit anti-REC-1 (1:1000), rabbit anti-RME-2 (1:500), mouse 4A5 (1:10), mouse anti-α-tubulin (1:100; Sigma). After the addition of primary antibody, the membrane was incubated overnight at 4°C. Immunoreactive protein bands were detected with either HRP-conjugated goat anti-rabbit IgG (1:10,000; Santa Cruz Biotechnology) or HRP-conjugated donkey anti-mouse IgG (1:1000; Jackson ImmunoResearch). Blots were visualized with an ECL Plus kit (Amersham) and analyzed using LAS-3000 (Fuji Film). The relative abundance of GLP-1, REC-8, and RME-2 after pab-1 RNAi was compared with controls, and the intensity of the bands was calculated by densitometry. α-Tubulin was used as a reference standard. Figure S1. pab-1 RNAi at the L1 stage causes developmental arrest of germ cells before entering meiosis. Extruded gonads of rrf-1(pk1417) worms with or without pab-1 RNAi treatment were co-immunostained with PAB-1 Regulates mRNA Metabolism PLOS ONE | www.plosone.org anti-HIM-3 (A, D, G), a meiotic marker, and OIC1D4 (B, E, H), a monoclonal antibody specifically recognizing P granules, along with TO-PRO-3 nuclear staining (C, F, I). pab-1 RNAi was administered either at L1 (D-F) or L2 (G-I) stage for 24 hours. Germ cells were observed after RNAi-treated worms were recovered and grown to the adult stage. A control adult gonad arm with mock RNAi treatment (A-C) is also shown. Asterisk indicates the distal end of each gonad. Bars, 10 µm. (TIFF) Figure S2. pab-1 mRNA is decreased in the germline proliferation defective mutant. The expression levels of pab-1 mRNA measured by quantitative real-time RT-PCR in wild-type N2 and glp-1(q231) mutant are shown. The average values from 3 independent experiments were normalized to that of act-1, and the relative expression levels are shown with the N2 value taken as 1.