Rbp4-Gal4, a germline driver that activates in meiosis, reveals functions for VCP in spermatid development

ABSTRACT Valosin-containing protein (VCP) is a versatile and ubiquitously expressed AAA+ ATPase that regulates multiple stages of Drosophila spermatogenesis. While VCP has documented roles in mitotic spermatogonia and meiotic spermatocytes, it is also highly expressed in post-meiotic spermatids, suggesting potential late-stage developmental functions as well. However, tools to assess late-stage activities of pleiotropic spermatogenesis genes such as VCP are lacking. Available germline-specific Gal4 drivers activate in stem cells or spermatogonia; consequently, knocking down VCP using one of these drivers disrupts or blocks early germ-cell development, precluding analysis of VCP in later stages. A Gal4 driver that activates later in development, such as at the meiotic spermatocyte stage, may permit functional analyses of VCP and other factors in post-meiotic stages. Here, we describe a germline-specific Gal4 driver, Rbp4-Gal4, which drives transgene expression beginning in the early spermatocyte stage. We find that Rbp4-Gal4-driven knockdown of VCP causes defects in spermatid chromatin condensation and individualization without affecting earlier developmental stages. Interestingly, the defect in chromatin condensation appears linked to errors in the histone-to-protamine transition, a key event in spermatid development. Overall, our study reveals roles for VCP in spermatid development and establishes a powerful tool to dissect the functions of pleiotropic spermatogenesis genes.


Introduction
Pleiotropy is the ability of a gene to serve multiple functions, sometimes even in the same biological process. One such example is Valosin-containing protein (VCP, also known as TER94), a AAA+ ATPase that regulates diverse biological processes [1], including multiple stages of Drosophila spermatogenesis. During the initial stages of Drosophila spermatogenesis, germline stem cells (GSCs) give rise to mitotic spermatogonia, which then differentiate into meiotic spermatocytes (Figure 1a). When VCP is genetically knocked down specifically at the start of spermatogenesis, mitotic spermatogonia hyper-proliferate [2], suggesting that VCP regulates developmental timing early in spermatogenesis. Recently, we discovered that VCP performs an additional, essential function during meiosis; VCP downregulates a repressive histone modification, monoubiquitinated H2A (H2Aub), to promote spermatocyte differentiation, and experimentally blocking this activity causes an early meiotic arrest [3]. Thus, VCP is a pleiotropic spermatogenesis gene that executes separable functions at distinct developmental timepoints and inhibition of VCP function in the testis can yield different phenotypes depending on the precise stage of inhibition.
While previous findings indicate at least two separable functions of VCP during spermatogenesis, it remains possible that VCP may exert even more activities during this process, perhaps even at later stages. Following meiotic divisions, spermatocytes differentiate into spermatids (Figure 1a). During the spermatid stage, germ cells elongate and histones are replaced by arginine-rich proteins known as protamines, driving hyper-condensation of chromatin to ultimately facilitate the formation of mature sperm [4][5][6]. Interestingly, a large-scale mutagenesis analysis [7] identified a male-sterile allele of VCP (FBal0325644) that is reported to cause defects at the spermatid stage (personal communication to Flybase; FBrf0235035). Yet, this mutation is present in all cell types, and it is unclear whether the associated defects are due to VCP dysfunction in the soma or germline. Potentially, tools to control VCP activity in a germline-and stage-specific manner could help to clarify the extent to which VCP regulates previously unlinked events in spermatogenesis, including at the spermatid stage.
Gene function in Drosophila spermatogenesis is commonly manipulated using the Gal4/UAS system [8,9], a powerful genetic toolkit that permits spatiotemporal control of transgene expression [10,11]. Currently, there are three main Gal4 drivers used to study gene function in the male germline: Vasa-Gal4, Nanos-Gal4 and Bam-Gal4 [8]. Vasa-Gal4 becomes active in GSCs and maintains activity throughout the male germline ( Figure 1a). Nanos-Gal4 is also active in GSCs but exhibits minimal activity outside of GSCs (Figure 1a). Bam-Gal4 becomes active in spermatogonia but also supports transgene expression later in development, such as in spermatocytes and postmeiotic round spermatids (Figure 1a). A germlinespecific Gal4 driver that consistently and robustly activates in meiotic-stage germ cells (i.e. following the completion of mitotic stages) has, to our knowledge, not been reported. Researchers have previously attempted to control transgene expression in meiotic stages of spermatogenesis using the β2-tubulin promoter, which activates late in the spermatocyte stage [9]. However, the generation of functional Gal4 lines under control of the β2-tubulin promoter has been unsuccessful, possibly because the promoter activates too far along in development to drive sufficient Gal4 expression [9]. The lack of a driver specific to postspermatogonial stages of Drosophila spermatogenesis has hindered the study of pleiotropic spermatogenesis genes at later timepoints in germ-cell development. In the case of VCP, RNAi-mediated knockdown using one of the three aforementioned Gal4 drivers causes a developmental abnormality or arrest early in germ-cell development [2,3], abrogating functional analysis later in spermatogenesis. Accordingly, a germline-specific Gal4 driver that activates developmentally later than Bam-Gal4 could be of use in deciphering later-stage functions of VCP and possibly other pleiotropic spermatogenesis genes.
In this study, we designed a Gal4 driver under the control of the promoter of Rbp4, a germlinespecific gene that is strongly expressed following the completion of mitosis (Fig. S1) [12][13][14]. We find that Rbp4-Gal4 is inactive until germ cells enter the spermatocyte stage; thus, it can be used to effectively drive transgene expression in germ cells that have progressed past the spermatogonial stage of spermatogenesis. When used to drive VCP-RNAi, Rbp4-Gal4 reliably knocks down VCP expression in both spermatocytes and spermatids. Interestingly, VCP knockdown driven by Rbp4-Gal4 reveals a previously unknown, germline-specific requirement for VCP in spermatid chromatin condensation and individualization. Intriguingly, our data also suggest that VCP promotes the histone-to-protamine transition in developing spermatids. Overall, our study establishes a tool that enables the analysis of later-stage functions of pleiotropic spermatogenesis genes, using VCP as a proof-of-principle.

Fly husbandry and strains
Flies were maintained on standard cornmeal/agar food at 25°C, unless otherwise noted. For RNAi experiments, flies were incubated on standard cornmeal/agar food supplemented with yeast paste at 29°C for 7 d prior to dissection and imaging to boost Gal4 activity and gene knockdown.

Identification of candidate promoters
To identify an appropriate spermatocyte-specific promoter, we mined previously published gene expression datasets to identify candidate genes with spermatocyte-specific gene expression profiles. First, we used the dataset from [14] to identify candidate genes that were lowly expressed in FLY spermatogonia but highly expressed throughout the spermatocyte stage. This was done by dividing RPKM expression values at the early spermatocyte stage by expression values at the 16-cell spermatogonia stage (EC/S16). We filtered out genes with an EC/S16 less than 10, RPKM above 50 at any spermatogonial stage, and any gene symbol beginning with 'CG' or 'CR' (unannotated genes), which resulted in a list of approximately 5000 genes. We then cross-referenced the selected genes with genes that were specifically mentioned to be expressed in spermatocytes in an independent RNAseq study [13] and other genes we were familiar with from previous studies, such as hsc70-2 [16], mst35Bb [5] and htrA2 [17]. To this point, we had identified 22 candidate genes that are highly expressed in spermatocytes and lowly expressed in spermatogonia. Finally, we cross-referenced this list to FlyAtlas [18,19] to select against genes that were expressed in somatic tissues, which resulted in a final list of 15 genes. Following this analysis, we sorted genes from highest to lowest EC value. The top gene was Rbp4 (Fig. S1).

Immunostaining, microscopy and image processing
Testes were dissected in 1X phosphate-buffered saline (PBS; 137 mM NaCl, 2.7 mM KCl, 10 mM Na 2 HPO 4 , 1.8 mM KH 2 PO 4 ) and then immediately fixed in 4% paraformaldehyde diluted in PBS. Testes were washed three times in PBT (1X PBS, 0.1% Tween-20), then incubated in blocking buffer (3% BSA in 1X PBS) for at least 1 h at room temperature. Testes were incubated with the primary antibody diluted in blocking buffer containing 2% Triton X-100 overnight at 4°C. The next day, testes were washed five times with PBT prior to applying the secondary antibody. Testes were incubated with the secondary antibody for at least 3 h at room temperature in the dark. After the secondary antibody solution was removed, testes were washed five times with PBT. 1 µM Hoechst 33342 was incubated in the first wash to stain DNA. Testes were mounted in Vectashield antifade mounting medium prior to imaging.
Images were acquired using an inverted Leica SP8 confocal microscope, equipped with a 40× objective (NA 1.30) and a white-light laser. Images were processed using Leica LAS X software, and quantifications were performed using Fiji (NIH) on 8-bit images.

VCP-GFP quantification
We used Fiji (NIH) to quantify VCP-GFP intensity in the nucleus and cytosol of germ cells at various stages of spermatogenesis (see below for stage identification information). Briefly, we outlined the nucleus and a small region within the cytosol in five cells per stage per testis and measured the mean fluorescence intensity. This was done for 10 testes total for each condition. VCP-GFP was imaged using a 488 nm excitation laser and a Hybrid detector set at 500-550 nm. The excitation laser was set at 2.5% for all testes imaged with gain left at its default level, such that GFP intensity could be reliably compared between genotypes. For presentation purposes, the mean intensity of the control group was normalized to one.

Fertility assay
Single male flies were placed in a vial with 2-3 wild type (w 1118 ) virgin females shortly after eclosion. Males were transferred to a fresh vial with new wild type virgin females 2-3 d later. The presence or absence of progeny was scored after each mating. Males that failed to produce progeny in both matings were scored as infertile. Males that were able to produce progeny in at least one mating were scored as fertile. Flies were kept at 25°C on standard cornmeal/agar food for all matings.

Phalloidin staining
Phalloidin staining was performed by fixing testes in 4% paraformaldehyde immediately following dissection. Testes were then washed three times in PBT (1X PBS, 0.1% Tween-20). After washing, testes were incubated in 1:400 phalloidin (Invitrogen A12380) for 1 h at room temperature in the dark. Following staining, testes were washed three times in PBT (1X PBS, 0.1% Tween-20). 1 µM Hoechst 33342 was incubated in the first wash to stain DNA. Testes were mounted in Vectashield antifade mounting medium prior to imaging.

Germ-cell stage identification
Germ-cell staging was performed primarily based on chromatin morphology. Spermatocytes were identified based on chromatin features described in [21]. Spermatids were identified based on chromatin and mitochondrial features described in [4].

Statistical analyses
Information on sample size and statistics is provided in figure legends where applicable. Data normality was tested via the D'Agostino-Pearson normality test in combination with Q-Q plots prior to performing follow-up statistical analyses using GraphPad Prism software. Statistical tests used to determine significance are indicated in figure legends. A student's unpaired t-test was used when unpaired data for two groups were normally distributed and standard deviation was equal between groups. Welch's unpaired t-test was used when unpaired data for two groups were normally distributed but standard deviation was not equal. The Mann-Whitney U-test was used when unpaired data for two groups were not normally distributed.

VCP is expressed ubiquitously throughout the male Drosophila germline
VCP serves critical functions in the development of both spermatogonia [2] and spermatocytes [3]. However, whether VCP is also expressed and performs important, germline-specific functions in later stages of spermatogenesis, such as the spermatid stage, has not been explicitly explored. Using a VCP-GFP fly line, where a GFP tag was inserted at the C-terminus of the endogenous VCP locus via CRISPR-Cas9 [15], we noticed that VCP is expressed throughout the male germline ( Figure 1b). As previously described [3], VCP is cytosolic in spermatogonia (Figure 1c) but enters the nucleus as cells transition into the spermatocyte stage (Figure 1c). Notably, we also observed bright VCP-GFP in the nucleus of round spermatids and canoe-stage spermatids (Figure 1c). This suggests that VCP is subject to developmental regulation at multiple stages, and it may serve some function in spermatid nuclei.

Rbp4-Gal4 drives transgene expression in germ cells after mitotic stages of spermatogenesis but not earlier
Given this localization pattern, we aimed to develop a Gal4/UAS strain that would allow us to probe possible VCP functions in the spermatid stage. As noted previously, the three major germline-specific Gal4 drivers (Vasa-Gal4, Nanos-Gal4 and Bam-Gal4) initiate transgene expression at the GSC and/ or spermatogonia stages [8]. Because a germlinespecific Gal4 driver capable of driving transgene expression starting at the spermatocyte stage, but not at earlier stages, currently does not exist, we sought to generate a germline-specific Gal4 driver that activates in the early spermatocyte stage, as this could permit the study of gene function during later stages of spermatogenesis. We reasoned that a Gal4 driver of this nature must meet the following criteria: 1) the driver must be under control of a promoter that is inactive or minimally active in spermatogonia; 2) the promoter must be highly active in spermatocytes, particularly at early stages; and 3) the promoter should be minimally active in somatic cells to prevent potential embryonic lethality and unintended physiological effects. To identify the gene promoters that fit these criteria, we examined Drosophila transcriptomic datasets that distinguish genes based on germline versus somatic expression [18,19] and expression at different germ-cell stages [13,14]. This analysis yielded several genes, but we prioritized genes that had been experimentally shown to be expressed and to function in the male germline ( Fig. S1; see Methods). We found that Rbp4 fit our criteria the best; not only is Rbp4 lowly expressed in spermatogonia (Fig. S1) [13,14] and minimally expressed outside of the germline [18,19] but, among the candidates, it showed the highest early-spermatocyte expression (Fig. S1) [14]. We thus cloned the promoter of Rbp4 upstream of the Gal4 coding sequence (Figure 2a) and generated Rbp4-Gal4 transgenic flies.
We tested the specificity of transgene expression under Rbp4-Gal4 control by driving nuclearlocalized GFP (GFP nls ) expression. For comparison, we also examined GFP nls expression under the control of Bam-Gal4. Rbp4-Gal4-driven GFP nls signal was absent from mitotic spermatogonia but clearly visible in spermatocytes ( Figure 2b). In contrast, Bam-Gal4-driven GFP nls signal was present in late spermatogonia stages, as well as in spermatocytes (Figure 2c). Thus, Rbp4-Gal4 activates developmentally later than Bam-Gal4. Importantly, Rbp4-Gal4-driven GFP nls signal was undetectable in somatic cyst-cell nuclei (Figure 2b, outlines), indicating that Rbp4-Gal4 activity is restricted in the testis to the germline. To further verify that Rbp4-Gal4 is not active in spermatogonia, we knocked down bag of marbles (bam), which is required for the spermatogonia-to -spermatocyte transition [22]. Whereas Bam-Gal4driven knockdown of bam blocked the spermatogonia-to-spermatocyte transition, as indicated by an absence of spermatocytes (Figure 2d, left panel), Rbp4-Gal4-driven knockdown of bam did not affect developmental progression to the spermatocyte stage (Figure 2d, right panel). Collectively, these data indicate that Rbp4-Gal4 activity is restricted in the testis to germ cells and becomes active at the spermatocyte stage.

Rbp4-Gal4 drives efficient knockdown of endogenous VCP in spermatocytes and spermatids
We next assessed the usefulness of this tool in manipulating gene expression in postspermatogonial stages of male germline development. We recently found that VCP knockdown using the Bam-Gal4 driver arrested cells at the spermatocyte stage, when VCP normally redistributes from the cytosol to the nucleus (Figure 1c) [3]. Because Bam-Gal4 activates in spermatogonia, whereas Rbp4-Gal4 activates in the spermatocyte stage, we reasoned that VCP knockdown using this new driver may produce a different phenotype; the later-stage knockdown may not activate in time to produce the initial degree of knockdown required to block spermatocyte differentiation, but it might reveal later defects.
As a first step to evaluate VCP-RNAi using Rbp4-Gal4, we quantified endogenous VCP-GFP fluorescence intensities at different stages of spermatogenesis. As expected, VCP-GFP fluorescence intensity was not affected in spermatogonia of VCP-RNAi testes compared to controls (Figure 3a-c). In contrast, VCP-GFP signal was substantially reduced by Rbp4-Gal4-driven VCP-RNAi at later stages of male germ-cell development (Figure 3d-l). Moreover, unlike testes with VCP-RNAi driven by Bam-Gal4, testes with VCP-RNAi driven by Rbp4-Gal4 produced post-meiotic round spermatids (Figure 3j), consistent with a later timepoint of gene knockdown and inhibition. These data indicate that Rbp4-Gal4 is indeed capable of knocking down VCP expression in spermatocytes and round spermatids, and that Rbp4-Gal4-driven knockdown of VCP permits development to the spermatid stage, providing an opportunity to experimentally assess potential VCP functions in later stages of spermatogenesis.
Given the apparent defect in mature sperm formation in Rbp4-Gal4 VCP-RNAi testes, we hypothesized that VCP may support maturation of elongated spermatids and/or spermatid individualization, the terminal differentiation event of spermatogenesis [4]. To test the first possibility, we analysed chromatin condensation, which is tightly linked to spermatid maturation [4]. If VCP were to promote spermatid maturation, one would expect to observe incompletely condensed spermatid chromatin in VCP-RNAi testes. Indeed, we found that elongated spermatid chromatin failed to completely condense upon VCP knockdown (Figure 4d). This phenotypic abnormality was accompanied by a significant reduction in the number of mature elongated spermatids (Figure 4e). Thus, we concluded that one germline-specific function of VCP during late spermatogenesis is to promote elongated spermatid chromatin condensation.
We next tested whether VCP might additionally regulate spermatid individualization. During the individualization process, filamentous actin (F-actin)-based investment cones form near the nucleus of each of the 64 elongated spermatids within a cyst at the basal end of the testis. Collectively, these investment cones make up a complex known as the individualization complex (IC) [4]. Once ICs have fully formed, ICs synchronously translocate down spermatid tails in a basal to apical direction to remove excess cytoplasm and organelles [4]. To test if VCP promotes proper spermatid individualization, we labelled ICs with an F-actin dye, phalloidin, in control and VCP-RNAi testes. Because we had observed a reduction in the number of mature spermatid bundles (Figure 4e), which would not form ICs, and would thus generate confounding results, we normalized the number of ICs to the total number of mature spermatid bundles present. Consistent with a defect in spermatid individualization, we observed a significant reduction in the number of ICs per mature spermatid bundle at the basal end of the testis in VCP-RNAi testes compared to controls (Figure 4f,). We likewise observed a similar reduction in the number of non-basal ICs (i.e. those that had begun movement to the apical end) per mature spermatid bundle in VCP-RNAi testes compared to controls (Figure 4h). Notably, of the few non-basal ICs we did see in VCP-RNAi testes, we found that they often moved asynchronously (Figure 4i). Overall, these data indicate that VCP plays an additional role in the formation and synchronous movement of ICs to support proper spermatid individualization.

VCP promotes the histone-to-protamine transition
As spermatids mature, histones are replaced by protamines to drive hyper-condensation of chromatin [5,6]. Interestingly, mono-ubiquitinated H2A (H2Aub), which is downregulated in spermatocytes by VCP [3], is cleared from round spermatid nuclei just prior to the histone-to-protamine transition [6]. Because nuclei fail to properly condense in VCP-RNAi testes, we were curious whether H2Aub dynamics in spermatids may also be affected by knockdown of VCP. In control testes, we observed that H2Aub was present in round spermatid nuclei (Figure 5a) and absent from canoe-stage spermatids (Figure 5b  previously demonstrated [6]. In VCP-RNAi testes, H2Aub was present in round spermatids at comparable levels to controls (Figure 5a). However, H2Aub was retained in canoe-stage spermatids of VCP-RNAi testes (Figure 5b).
Because clearance of H2Aub from spermatid nuclei has been postulated to be a critical step for the histone-to-protamine transition in flies [6], we hypothesized that protamine expression and/or incorporation may be affected by the absence of VCP. We tested this hypothesis by imaging a GFPtagged protamine (protB-GFP) that was expressed from its endogenous promoter. While we did observe protB-GFP expression in some VCP-RNAi canoe-stage spermatids (Figure 5c), we noted that protB-GFP was not uniformly expressed in all spermatid nuclei; strikingly, protB-GFP signal was present in some nuclei, but absent from adjacent nuclei within the same cyst (Figure 5c, arrows). Collectively, our data suggest that VCP affects the robustness of the histone-toprotamine transition, possibly by controlling H2Aub levels, and this function likely in turn supports chromatin condensation and spermatid maturation.

Discussion
Previous studies have used the β2-tubulin promoter to express transgenes in the late stages of spermatogenesis [23][24][25][26][27][28][29][30][31]. However, attempts to generate a late-stage Gal4 driver under the control of the β2-tubulin promoter have not been fruitful [9], likely because the promoter becomes active too late in the spermatocyte stage to produce a sufficient amount of Gal4 to drive transgene expression. In this study, we have overcome this obstacle by generating Rbp4-Gal4, which activates early in the spermatocyte stage. As a proof-ofprinciple, we used Rbp4-Gal4 to successfully knock down VCP in late spermatogenesis, providing information on late-stage functions of VCP in spermatogenesis.
In theory, the Rbp4-Gal4 driver could also be applied to study the function of virtually any other gene that regulates the development of spermatids, or maybe even late-stage spermatocytes. Recent single-cell RNA sequencing studies in Drosophila [13,19,32,33] and a stage-specific Drosophila spermatogenesis proteomics study [34] have provided robust datasets that may suggest important functions for particular genes in the late stages of spermatogenesis. However, some of these genes may also serve important functions in the early stages of spermatogenesis. In the past, this would have posed an issue in studying the function of these genes in later stages, as had been the case for VCP [2,3]. As demonstrated in our study, Rbp4-Gal4 permits the investigation of potential pleiotropic gene functions in the spermatogenesis program, which provides an opportunity to advance this field of study. Notably, because UAS-RNAi lines are readily available from various stock centres (BDSC, VDRC, Kyoto, etc.), the Rbp4-Gal4 driver can now be combined with gene knockdown experiments and even screens to rapidly identify genes that function in the late stages of spermatogenesis.
Consistent with this driver activating at the spermatocyte stage, Rbp4-Gal4 sufficiently knocked down VCP in spermatocytes and spermatids ( Figure 3). Although VCP-RNAi using Rbp4-Gal4 decreased VCP levels in immature and mature spermatocytes, it did not lead to a developmental arrest at the spermatocyte stage, as had VCP-RNAi when driven by Bam-Gal4 [3]. One important point to note is that Rbp4-Gal4 becomes active developmentally later than Bam-Gal4, which activates in spermatogonia (Figure 2b, c). It could be that VCP knockdown must be initiated at the spermatogonia-tospermatocyte transition to cause an arrest at the spermatocyte stage, and that Rbp4-Gal4, while active in spermatocytes, activates too late to produce this arrest. These small differences in timing in fact enabled us to probe VCP function later in spermatogenesis using Rbp4-Gal4, which was previously not possible using available Gal4 drivers.
Currently, there are some open questions regarding how VCP molecularly supports spermatid development. We previously found that the knockdown of VCP using Bam-Gal4 decreases the expression of mst35Bb [3], which encodes protB [5]. While we detected protB in some spermatid nuclei upon VCP knockdown using Rbp4-Gal4 (Figure 5c), protB was evidently not incorporated in all spermatid nuclei following this