The PAF1 complex cell autonomously promotes oogenesis in Caenorhabditis elegans

Abstract The RNA polymerase II‐associated factor 1 complex (PAF1C) is a protein complex that consists of LEO1, RTF1, PAF1, CDC73, and CTR9, and has been shown to be involved in RNA polymerase II‐mediated transcriptional and chromatin regulation. Although it has been shown to regulate a variety of biological processes, the precise role of the PAF1C during germ line development has not been clarified. In this study, we found that reduction in the function of the PAF1C components, LEO‐1, RTFO‐1, PAFO‐1, CDC‐73, and CTR‐9, in Caenorhabditis elegans affects oogenesis. Defects in oogenesis were also confirmed using an oocyte maturation marker, OMA‐1::GFP. While four to five OMA‐1::GFP‐positive oocytes were observed in wild‐type animals, their numbers were significantly decreased in pafo‐1 mutant and leo‐1(RNAi), pafo‐1(RNAi), and cdc‐73(RNAi) animals. Expression of a functional PAFO‐1::mCherry transgene in the germline significantly rescued the oogenesis‐defective phenotype of the pafo‐1 mutants, suggesting that expression of the PAF1C in germ cells is required for oogenesis. Notably, overexpression of OMA‐1::GFP partially rescued the oogenesis defect in the pafo‐1 mutants. Based on our findings, we propose that the PAF1C promotes oogenesis in a cell‐autonomous manner by positively regulating the expression of genes involved in oocyte maturation.

F I G U R E 1 PAF1C is essential for oogenesis. (a-m) Germ cell development of wild-type (a), leo-1(gk1081) (b, c), rtfo-1(tm5670) (d), pafo-1(tm13347) (e), pafo-1(tm13347);tjIs280[pafo-1p::pafo-1::mCherry::pafo-1 3 0 -UTR] (f), control(RNAi) (g), leo-1(RNAi) (h, i), rtfo-1(RNAi) (j), pafo-1(RNAi) (k), cdc-73(RNAi) (l), and ctr-9(RNAi) (m) in hermaphrodite day 1 adult posterior gonads. (n, o) Quantitative analysis of the RNAi efficiency of leo-1(RNAi) (n) and pafo-1(RNAi) (o). P-values are indicated for Student's t-test in comparison with the control(RNAi). ***p < .005. The error bars represent ±SD. See Figures S1 and S2 for more information about the fluorescent images. (p) Percentages of oogenesis defects found in 1 day-adult wild type, mutants, transgenic rescued, control(RNAi), and RNAi-knockdown animals of each PAF1C component. P-values are indicated for Fisher's exact test in comparison with WT or control(RNAi). ***p < .005. Error bars represent ±SD. In all the panels, the anterior region of the gonad was to the left, and the dorsal region was at the top of the image. The posterior gonads are shown. The orange dotted lines mark the gonad boundaries and the green dotted lines mark the oocyte boundaries. Scale bar (white), 50 μm the recruitment and activation of RNA polymerase II (Pol II) to the transcriptional target is required. In addition, chromatin remodeling affects DNA accessibility during transcription through epigenetic modification of nucleosomes. The RNA polymerase II-associated factor 1 (PAF1) complex, or PAF1C, is a highly conserved protein complex in eukaryotes, which is involved in multiple aspects of Pol II-mediated transcriptional regulation, including transcriptional elongation, 3 0 -end processing, and epigenetic modification. Moreover, the PAF1C is involved in the posttranscriptional step of gene expression and translational regulation via its interaction with the regulatory sequences of mRNAs (Francette et al., 2021;Jaehning, 2010).
The development of the nematode, Caenorhabditis elegans, is highly reproducible, which makes it a reliable model organism for analyzing the regulatory mechanism of development. The hermaphrodite gonad of this nematode temporally produces sperm at the late larval stage, which are stored in the spermatheca, and subsequently produces oocytes during the adult stage. During this process, spatiotemporal regulation of gene expression, cell proliferation, cell differentiation, cell shape change, cell growth, and meiotic progression occurs (Arur, 2017;Huelgas-Morales & Greenstein, 2018;Kim et al., 2013). However, the mechanism of oogenesis has not been fully elucidated.
In this study, we found that all the PAF1C components are involved in promoting oogenesis, and that the expression of OMA-1, a CCCH-type zinc finger protein involved in oocyte maturation, is promoted by the PAF1C in a cell-autonomous manner.

| The PAF1C is essential for oogenesis
To analyze whether the PAF1C is involved in germ cell development in C. elegans, we observed germ cell development of the posterior gonads in day 1 adults of the PAF1C mutants by DIC microscopy (Figure 1). Compared with the wild-type animals, oogenesis was insufficient in the F I G U R E 2 PAF1C is dispensable for spermatogenesis. (a-l) Differential interference contrast (DIC) (a, c, e, g, i, and k) and fluorescence (b, d, f, h, j, and l) images of wild type (a, b), pafo-1 (tm13347) (c, d), control(RNAi) (e, f), leo-1(RNAi) (g, h), pafo-1 (RNAi) (i, j), and cdc-73(RNAi) (k, l) day 1 adult animals with tjIs57 [pie-1p::mCherry::H2B::pie-1 3 0 -UTR]. In all the panels, the anterior region of the gonad is to the left, and the dorsal region is at the top of the image. The posterior gonads are shown. The orange dotted lines mark the gonad boundaries, the green dotted lines mark the oocyte boundaries and the blue dotted lines surround the spermlike small cells. Scale bar (white), 50 μm PAF1C mutants, leo-1(gk1081), rtfo-1(tm5670), and pafo-1 (tm13347) mutants, although the penetrance of leo-1 (gk1081) was lower than that of the other mutants ( Figure 1a-1e and 1p). An integrated transgene, tjIs280 [pafo-1::mCherry], expressed by the pafo-1 regulatory region (Figure 4a,h), rescued the oogenesis defect of the pafo-1 (tm13347) deletion mutant (Figure 1f,p). Similar to the observations of deletion mutants, although the penetrance of the oogenesis defect of the leo-1(RNAi) was relatively low, all the five RNAi-knockdown animals of the PAF1C components exhibited oogenesis defects (Figure 1g-m,p). Next, we analyzed the expression patterns of LEO-1 and PAFO-1 by using animals expressing the tjIs308[leo-1p:: GFP::leo-1] and tjIs280[pafo-1p::pafo-1::mCherry] transgenes by authentic promoters. The results indicated that both of the GFP::LEO-1 and PAFO-1::mCherry were ubiquitously expressed within the germ cells ( Figures S1 and S2). The efficiency of both leo-1(RNAi) and pafo-1(RNAi) was over 90% (Figure 1n,o; Figures S1 and S2), as measured by the fluorescent signal of GFP::LEO-1 and PAFO-1::mCherry, respectively, at the distal region of the gonad. These results suggest that the PAF1C is essential for oogenesis, and that the contribution of LEO-1 to the PAF1C function may be the lowest among the PAF1C components.

| The PAF1C is involved in the expression of OMA-1 in oocytes
Next, we investigated how the PAF1C regulates oogenesis. To visualize matured oocytes, we used an oocyte maturation marker, OMA-1::GFP, which was derived from the oma-1 regulatory region ( Figure 3a). In the day 1 adult stage of control(RNAi) animals, 4.8 OMA-1::GFP-positive cells were arranged linearly in the ventral region of each gonad on an average (N = 50, Figure 3b,c,j). In contrast, the average number of OMA-1::GFP-positive cells were 1.4, 0.10, and 0.36 in leo-1(RNAi), pafo-1(RNAi), and cdc-73(RNAi) animals, respectively (N = 50, Figure 3d-j). These results suggest that the PAF1C positively regulates the expression of OMA-1.
2.5 | Overexpression of OMA-1::GFP Partially rescues the oogenesis defect in the pafo-1(tm13347) mutant We tested whether the reduction in the expression of OMA-1 is the major cause of oogenesis defects in the PAF1C mutants. When OMA-1::GFP was overexpressed in the pafo-1(tm13347) mutant, the oogenesis defect was partially rescued ( Figure 5). Therefore, a possible role of the PAF1C in the germline is to promote oogenesis by positively regulating the expression of oma-1.

| DISCUSSION
The PAF1C is a highly conserved protein complex that consists of five conserved components, namely LEO1, RTF1, PAF1, CDC73, and CTR9. Although it has been shown to be required in diverse biological processes, its contribution to germ cell development has not yet been explored. In this study, we performed functional analysis of the PAF1C in the germ cell development of C. elegans and demonstrated its requirement for the promotion of oogenesis and expression of OMA-1 during oogenesis.
Although PAFO-1 is ubiquitously expressed in germ cells, OMA-1 is expressed exclusively in maturing oocytes (Detwiler et al., 2001). This distribution implies that PAFO-1 could act as a regulator of OMA-1 expression during the oocyte maturation process. Consistent with this hypothesis, the OMA-1 expression level was found to be significantly decreased in pafo-1 mutants, suggesting that the PAF1C may function as a positive regulator of OMA-1 expression.
Although the PAF1C components, LEO-1 and PAFO-1, are expressed ubiquitously, including in germ cells (Kubota et al., 2014), the PAF1C is required only for oogenesis but not for spermatogenesis. Because the PAF1C is not required for sperm formation, it is unlikely that the oogenesis-defective phenotype is causative of earlier defects in germ cell development. These results suggest that PAF1C regulates a specific set of genes that are required for oogenesis.
The number of OMA-1::GFP-positive maturing oocytes was decreased in leo-1(RNAi), pafo-1(RNAi), and cdc-73 (RNAi) animals and in pafo-1 deletion mutants. Therefore, one possible role of the PAF1C is the promotion of OMA-1 expression. We also show that the overexpression of OMA-1::GFP partially rescued the oogenesis defect in the pafo-1 mutant. Therefore, the PAF1C may promote oogenesis by positively regulating the specific set of downstream oocyte maturation regulators, including OMA-1 ( Figure 6). Our data also suggest that the germ cell expression of PAF1C is sufficient for the formation of fully functional oocytes and F I G U R E 6 Models for the PAF1Cdependent regulation of oogenesis. (a) The PAF1C and RNA polymerase II (RNA pol II) are recruited to the regulatory region of target genes, and directly or indirectly promote the expression of the oocyte maturation regulators. (b) During oocyte maturation, the PAF1C promotes oocyte maturation regulators, which then promotes oogenesis in the ventral region of the gonad that the maternal contribution of the PAF1C is sufficient for embryonic development, larval development, and larval-adult transition.
It has been shown that oma-1 and oma-2 act redundantly to promote the later part of oocyte maturation to complete oocyte maturation (Detwiler et al., 2001;Spike et al., 2014;Tsukamoto et al., 2017). In contrast, the expansion defect of oocytes in the PAF1C-depleted animals and mutants occurred in the early part of the oocyte maturation process. Although the overexpression of OMA-1::GFP partially rescued the oogenesis defect of the pafo-1 mutant, the phenotypic similarity of the oogenesis defect was not observed in the pafo-1(tm13347) mutant and the oma-1(RNAi);oma-2(RNAi) animals. These results indicate that the phenotypic severity of PAF1C-depleted animals was stronger than that of OMA-1/2-double depleted animals. Therefore, it is pertinent to discuss as to why the overexpression of OMA-1 rescues the failure of oocyte expansion phenotype of the pafo-1 mutant. A possible explanation is that although the major function of PAF1C is to promote the expression of OMA-1/2, the PAF1C may also regulate other targets that are involved in the promotion of the expansion process of oocytes in parallel with the OMA-1/2 functions. Further analysis is required to confirm this hypothesis.
In this study, we found that oogenesis defects were less severe in leo-1(RNAi) animals among the animals with RNAi-knockdown of the five components of the PAF1C. Similar to our observations, RTF1, PAF1, CDC73, CTR9, but not LEO1, were reported to be required for the specification of an appropriate number of cardiomyocytes and for elongation of the heart tube in zebrafish (Langenbacher et al., 2011). Taken together, these observations suggest that in the specific context of the differentiation process, among the five PAF1C components, the requirement for LEO-1 is less critical, and this difference is conserved in vertebrates and invertebrates. At present, there are several possible avenues for exploring this phenomenon. In yeast, Ctr9, Cdc73, and Rtf1, but not Leo1, were shown to require Paf1 at normal levels, and loss of Cdc73 resulted in a lower abundance of Rtf1 (Mueller et al., 2004;Squazzo et al., 2002). Therefore, it is expected that the other four components may achieve only a part of the PAF1C function in the absence of LEO-1.
There are several possible mechanisms to explain why the leo-1(gk1081) mutant exhibited less severe defects in germ cell development. Recently, crystal structure analysis revealed that the reconstituted PAF1C quaternary Ctr9/Paf1/Cdc73/RTf1 complex may act as the core module of the yeast PAF1C in the absence of the Leo1 subunit (Chen et al., 2022). The mutant LEO-1 (gk1081) protein was shown to have lost more than two-thirds of the carboxyl-terminal region (Kubota et al., 2014), which include the PAF1-binding domain of the Homo sapiens LEO1 protein (Chu et al., 2013). Therefore, the mutant LEO-1(gk1081) protein might be excluded from the LEO-1-containing complex. Because Leo1 appears to mediate the interaction between the PAF1C and elongating Pol II through its RNA-binding activity in yeast (Dermody & Buratowski, 2010), in parallel to the LEO-1-dependent function of the PAF1C, an unknown LEO-1-independent PAF1C function may also be involved in the PAF1Cmediated regulation of the oogenesis in C. elegans. Therefore, it is conceivable that the leo-1(gk1081) mutant achieves the LEO-1-independent regulation of C. elegans oogenesis in consequence of physical interaction among the components of the quaternary complex.
Although all the PAF1C components are required for its function, each component has a specific role in regulating the expression of gene encoding cell differentiation regulators, possibly by affecting the formation of the protein complex, specific protein-protein interactions, and protein-DNA/RNA interactions. Further studies are required to determine how PAF1C regulates tissuespecific development in multicellular organisms.

| Caenorhabditis elegans strains
Caenorhabditis elegans strains used in this study were derived from the wild-type (WT) Bristol strain N2 (Brenner, 1974). Worms were incubated at 20 C, except those that were fed RNAi bacteria and were maintained at 22 C.
The leo-1 locus encodes a predicted polypeptide of 430 amino acids (aa), and the gk1081 allele (isolated by the C. elegans Gene Knockout Consortium) deleted 627 bp that would result in a C-terminally truncated protein of 137 aa (intrinsic 132 aa with an extra 5 aa) (Kubota et al., 2014). The rtfo-1 locus encodes a predicted polypeptide of 613 aa, and the tm5670 allele (isolated by the National Bioresource Project Japan) deleted 361 bp that would result in a C-terminally truncated protein product of 349 aa (intrinsic 314 aa with an extra 35 aa). The pafo-1 locus encodes a predicted polypeptide of 425 aa, and the tm13347 allele (isolated by the National Bioresource Project, Japan) deleted 83 bp that would result in a C-terminally truncated protein product of 288 aa (intrinsic 285 aa with an extra 3 aa). The Cterminus of Rtf1 in yeast has been shown to be required for its efficient anchoring to the PAF1C; rtfo-1(tm5670) is expected to be a complex formation-defective mutant (Warner et al., 2007). As mentioned above, leo-1(gk1081) is not a protein-null mutant. The leo-1(gk1081) mutant has been shown to produce reduced amounts of Cterminally truncated proteins in a previous study (Kubota et al., 2014). At present, we have not checked whether or not rtfo-1(tm5670) and pafo-1(tm13347) are null mutants.
The strains used in this work are listed in Table S1.

| Plasmid construction
The plasmids used in this study are listed in Table S2. A miniMos backbone vector, denoted as pYK13, was constructed by inserting a 376 bp fragment containing a multicloning site into a StuI site in pCFJ910. To construct the targeting vectors, the following fragments were amplified and individually subcloned into pYK13 at the NotI and AscI sites. To construct transgenes that expressed GFP-fusion proteins from putative endogenous 5 0 -cis regulatory regions of oma-1, the genomic fragments, oma-1:: GFP derived from the oma-1 regulatory region (2860 bp), the coding region, and the oma-1 3 0 -UTR (2935 bp), were PCR amplified and then fused with GFP. For germ cell-specific expression experiments, the pafo-1 genome was subcloned into a carboxyl-terminal mCherry-fusion protein expression vector, which has cis regulatory regions of pie-1, pYK229, a modified vector derived from pYK13.

| Feeding RNAi
The worms were fed on RNAi-feeding plates as previously described (Kamath et al., 2001). Full-length leo-1, rtfo-1, pafo-1, and cdc-73 cDNAs and 1000 bp ctr-9 (1st-1000th coding region) cDNA were isolated from a C. elegans cDNA library and inserted into the feeding RNAi vector, L4440. An L4440 vector lacking an insert was used as a control(RNAi). After confirming that each inserted sequence was correct, the feeding vectors were individually transformed into Escherichia coli HT115 (DE3) samples, which were then seeded separately onto plates of nematode growth medium agar containing Luria-Bertani medium and 50 μg/ml ampicillin, and incubated overnight at 37 C. Thereafter, each culture was seeded onto a 60 mm feeding agar plate containing 50 μg/ml ampicillin and 1 mM isopropyl β-D-1-thiogalactopyranoside and incubated at 23 C for 2 days. L4-stage worms were transferred to a feeding plate and cultured at 22 C. Phenotypes of F1 worms were determined at the day 1 adult stage.

| Microscopy
Fluorescence and differential interference contrast; DIC microscopy procedures were performed using an Olympus BX63 microscope with an ORCA-Spark camera (Hamamatsu Photonics) and UPlanSApo X60 water NA 1.20 or UPlanXApo X40 NA 0.95 objective lens (Olympus). The microscope system was controlled using the cellSens Dimension software (Olympus). Images were processed using the ImageJ (NIH) or Adobe Photoshop 2021 software.
4.6 | Analyses of the efficiency of RNAi knockdown of leo-1 and pafo-1 Fluorescence signals of GFP::LEO-1 and PAFO-1:: mCherry from images of the RNAi-treated animals at the perinuclear region of the germ cells of the distal gonad obtained under the fluorescence/DIC microscope, fluoresce signals were measured using Image J software (National Institutes of Health; see Figure 1n,o). A rectangular area (8 μm Â 40 μm) was used to analyze the GFP/mCherry signals. The relative expression levels of GFP::LEO-1 or PAFO-1::mCherry were then calculated and compared between control(RNAi) and leo-1(RNAi) or pafo-1(RNAi) animals, respectively.