Hormone receptor 4 is required in muscles and distinct ovarian cell types to regulate specific steps of Drosophila oogenesis

Female reproduction is an energy-intensive process that is highly sensitive to changes in physiology (Drummond-Barbosa, 2019; Laws and Drummond-Barbosa, 2017). Transcription factors in the evolutionarily conserved nuclear receptor superfamily bind to and mediate the effects of a diverse set of circulating molecules, thus playing key roles in linking whole-body physiology to a wide range of conserved processes, including oogenesis. For example, estrogen-related receptors are required for germ cell development (Festuccia et al., 2017) and Liver receptor homolog 1 is required for steroid synthesis and ovulation (Zhang et al., 2013) in mammals, whereas multiple nuclear receptors are required in insects for egg production (Ables and Drummond-Barbosa, 2017; Xu et al., 2010). The complex mechanisms by which nuclear receptors mediate inter-organ communication to influence oogenesis, however, are only beginning to be uncovered.

As in mammals, female reproduction in Drosophila is regulated by hormones and other circulating factors (Ables and Drummond-Barbosa, 2017; Drummond-Barbosa, 2019), making it an ideal model system to investigate how nuclear receptors influence oogenesis. The Drosophila ovary is composed of 16-20 ovarioles, each consisting of an anterior germarium followed by developing egg chambers or follicles (Fig. 1A,B). Two-to-three germline stem cells (GSCs) reside in a specialized niche composed primarily of somatic cap cells at the anterior end of each germarium. Each GSC division generates a self-renewing GSC and a cystoblast that undergoes four synchronous divisions, forming a two-, four-, eight- and eventually 16-cell cyst comprising one oocyte and 15 nurse cells. A somatic population of escort cells closely interacts with early germline cysts as they divide mitotically, and somatic follicle cells subsequently surround the 16-cell germline cyst to form a new follicle that leaves the germarium and develops through 14 stages of oogenesis. As the follicle grows, follicle cells divide mitotically through stage 7, after which they undergo endoreplication. At stage 8, the oocyte begins vitellogenesis (or yolk uptake), and by stage 14 the mature oocyte is ready for ovulation and fertilization (Greenspan et al., 2015; Laws and Drummond-Barbosa, 2017).

The Drosophila genome encodes 18 nuclear receptors (representing all major subfamilies of vertebrate receptors) (King-Jones and Thummel, 2005; Mazaira et al., 2018), many of which have been implicated in regulating different aspects of oogenesis. For example, the steroid hormone ecdysone acts directly on GSCs through a heterodimer of Ecdysone receptor (EcR) and Ultraspiracle (Usp) to regulate their maintenance and proliferation (Ables and Drummond-Barbosa, 2010), whereas EcR functions in escort cells to support entry into meiosis (Morris and Spradling, 2012). E75 (downstream of EcR) promotes progression through vitellogenesis (Buszczak et al., 1999), and Ftz-f1 is required for ovulation (Knapp et al., 2020; Sun and Spradling, 2013). Additional studies show that nuclear receptors also act in peripheral tissues to indirectly influence oogenesis. EcR in the central nervous system modulates female feeding behavior to support egg production (Sieber and Spradling, 2015). We have recently shown that Seven up is required in adipocytes to influence GSC maintenance and early cyst survival, and in hepatocyte-like oenocytes to regulate vitellogenesis (Weaver and Drummond-Barbosa, 2019). These studies suggest that nuclear receptor activity is required in multiple peripheral tissues to indirectly influence distinct stages of oogenesis.

In this study, we investigate the role of the orphan nuclear receptor Hormone receptor 4 (Hr4) in regulating Drosophila oogenesis. Hr4 has been extensively studied in the context of larval development and is required to control the timing of ecdysone pulses and larval feeding behavior (King-Jones et al., 2005; Ou et al., 2011, 2016). The mammalian homolog of Hr4, germ cell nuclear factor (GCNF; also known as Nr6a1), has roles in embryonic stem cell differentiation, regulation of spermatogenesis and oogenesis (Akamatsu et al., 2009; Kavarthapu and Dufau, 2015; Sabour et al., 2014; Wang et al., 2016; Weikum et al., 2016). In mammalian adults, GCNF is predominantly expressed in germ cells (Katz et al., 1997; Lan et al., 2003a), and oocyte-specific knockout of GCNF results in aberrant steroidogenesis and reduced fertility (Zhao et al., 2007). However, previous studies have not addressed whether Hr4/GCNF might be required in multiple cell types/tissues to control oogenesis. Here, we show that Drosophila Hr4 functions in the germline and in multiple somatic cell types/tissues of the adult female to regulate distinct processes during oogenesis. Hr4 is required in the germline to maintain normal GSC numbers, and it is also required in cap cells for GSC proliferation. Interestingly, we uncovered a novel requirement for Hr4 in adult female muscles to help maintain GSCs in the niche and to promote follicle growth. These findings illustrate the complexity of nuclear receptor signaling in the control of oogenesis and underscore the importance of considering the potential requirement for any given nuclear receptor in multiple cell types/tissues when investigating its role in reproduction.

We previously described a ubiquitous somatic RNAi-based screen to identify nuclear receptors with novel roles in regulating adult Drosophila oogenesis, and identified a new role for the nuclear receptor encoded by svp (Weaver and Drummond-Barbosa, 2019). In that screen, we used the temperature-sensitive Gal4 inhibitor tub-Gal80^ts (McGuire et al., 2003) in combination with the tub-Gal4 driver (Nabel-Rosen et al., 2002) (tub-Gal4^ts; which is well expressed in all somatic cells but not the germline) to knock down nuclear receptors in adult females, and measured egg output (Weaver and Drummond-Barbosa, 2019). In addition to identifying svp (Weaver and Drummond-Barbosa, 2019), we also found that somatic knockdown of Hr4 in adult females (Fig. S1A,B) drastically reduced egg production compared with Luciferase (Luc) control RNAi or UAS-hairpin alone (Fig. 1C,D). These results show a requirement for Hr4 in the regulation of adult female oogenesis.

Physiological changes can affect many processes during adult oogenesis (Drummond-Barbosa, 2019). To investigate how ubiquitous somatic Hr4 RNAi knockdown impairs oogenesis, we first asked whether loss of Hr4 in somatic tissues affects GSC maintenance in the niche. GSCs were lost faster in somatic Hr4 RNAi compared with Luc RNAi control or UAS-hairpin alone females (Fig. 2A-C; Fig. S1C,D). The increased loss of GSCs was not due to changes in cap cell numbers, which remained similar between control and experimental females (Fig. 2D; Fig. S1E). We also examined whether Hr4 controls GSC proliferation based on the frequencies of EdU (a thymidine analog) incorporation and phospho-histone H3 (a mitotic marker) staining in GSCs (Fig. 3). The frequencies of both EdU-positive and phospho-histone H3-positive GSCs were significantly lower in females with ubiquitous somatic knockdown of Hr4 compared with Luc control (Fig. 3C,D). These results indicate that Hr4 is required in somatic tissues to promote GSC maintenance and proliferation.

To determine whether Hr4 is required in somatic cells for early germline cyst survival, we labeled dying cysts using ApopTag, a TUNEL-based assay for detection of DNA fragmentation, as previously described (Drummond-Barbosa and Spradling, 2001). Ubiquitous somatic knockdown of Hr4 in adult females significantly increased the percentage of germaria containing dying germline cysts in region 2 (which contains 16-cell germline cysts not yet completely enveloped by follicle cells) relative to Luc RNAi control (Fig. 4). The higher frequency of early cyst death (together with the faster loss and reduced proliferation rate of GSCs; see above) is consistent with our observations that the germaria of Hr4 somatic knockdown females appear shorter than those of controls (Fig. 2A,B). These results indicate that Hr4 is required in somatic cells to control survival of early GSC progeny.

Changes in physiology can also influence the growth of follicles and their later survival during vitellogenesis (Drummond-Barbosa, 2019; Laws and Drummond-Barbosa, 2017), prompting us to test whether Hr4 somatic knockdown affects these processes. Mitotically dividing follicle cells proliferate proportionately to follicle growth (Drummond-Barbosa and Spradling, 2001; LaFever and Drummond-Barbosa, 2005; Laws and Drummond-Barbosa, 2016; Maines et al., 2004); therefore, we asked whether follicle cell proliferation during stages 4-6 of oogenesis was altered in females with ubiquitous somatic Hr4 knockdown. The percentages of phospho-histone H3- and EdU-positive follicle cells were significantly decreased in ubiquitous somatic Hr4 knockdown females compared with Luc control RNAi (Fig. 5), indicating that Hr4 is required in somatic cells for normal rates of follicle growth. In addition, ubiquitous somatic Hr4 knockdown increased the percentage of ovarioles containing dying vitellogenic follicles (recognizable by the presence of pyknotic nuclei) relative to Luc control RNAi (Fig. 6). These results demonstrate that Hr4 is required in somatic tissues to promote follicle growth and survival through vitellogenesis.

To determine what specific tissues require Hr4 activity for oogenesis, we first examined the tissue distribution of Hr4 mRNA in adult Drosophila females using RT-PCR analysis (Fig. 7A). Hr4 transcripts were predominantly detected in the ovary and thorax (which is composed mostly of muscles), with considerably lower levels of expression in other tissues. To test whether Hr4 is required in adult muscles to influence oogenesis, we used tub-Gal80^ts in combination with the MHC-Gal4 driver (Schuster et al., 1996), which is exclusively expressed in muscles in adult females (Weaver and Drummond-Barbosa, 2019), to drive Hr4 hairpins for RNAi. Adult muscle-specific knockdown of Hr4 significantly increased the rate of GSC loss compared with Luc control knockdown (Fig. 7B; Fig. S2A). The loss of GSCs induced by Hr4 knockdown in muscles was not due to changes in niche E-Cadherin or Bone morphogenetic protein (BMP) signaling, which are required for GSC maintenance (Song and Xie, 2002; Xie and Spradling, 1998), as the levels of niche E-Cadherin and nuclear phosphorylated Mad (pMad, a BMP signaling reporter; Kai and Spradling, 2003) in GSCs were statistically similar in control and Hr4 knockdown females (Fig. S2B-E). In contrast to its role in GSC maintenance, muscle Hr4 does not regulate GSC proliferation or early germline cyst survival in the germarium (Fig. S3).

Adult female muscle-specific knockdown of Hr4 also delayed the growth of developing follicles. Specifically, we found that follicle cell proliferation rates were significantly reduced in muscle-specific Hr4 compared with Luc control RNAi females based on EdU incorporation and phospho-histone H3 staining (Fig. 7C,D). By contrast, knockdown of Hr4 in adult female muscles did not alter the survival of vitellogenic follicles relative to Luc control knockdown, as the percentage of ovarioles with degenerating follicles was similarly low in all females [Luc RNAi^JF01355 control, 0.67±0.67% (292 ovarioles); Hr4 RNAi^HM0560, 0.33±0.33% (300 ovarioles); Hr4 RNAi^HMJ21497, 0.33±0.33% (300 ovarioles); mean±s.e.m.]. Taken together, these results reveal novel roles for muscles in the control of GSC maintenance and follicle growth.

Hr4 is robustly expressed in the ovary (Fig. 7A), prompting us to examine a potential requirement for Hr4 in ovarian somatic cell populations (and in the germline; see below). To test whether Hr4 is required in the adult GSC niche, we knocked down Hr4 using tub-Gal80^ts with the hh-Gal4 driver (Jenett et al., 2012), which is predominantly expressed in the terminal filament, cap cells and escort cells (Weaver et al., 2020). Hr4 RNAi in the adult female niche significantly decreased GSC proliferation relative to control RNAi based on EdU incorporation and phospho-histone H3 staining (Fig. 8A,B). In addition, knockdown of Hr4 in the niche also slightly decreased GSC maintenance; however, the decrease in GSC number only reached significance for one of the Hr4 hairpin transgenes tested (Fig. S4A,B). These results suggest that Hr4 signaling in the niche is required to control GSC proliferation and might have a minor role in GSC maintenance.

We next asked whether survival of early germline cysts requires Hr4 function in closely associated escort cells. We used the escort cell driver c587-Gal4 (Weaver et al., 2020) in combination with the neuronal nSyb-Gal80 transgene (Harris et al., 2015) to repress c587-Gal4 in the brain, where it is significantly expressed (Weaver et al., 2020) and with tub-Gal80^ts (for adult-specific expression; see above) to drive Hr4 hairpins in adult escort cells (Fig. S4C). Hr4 knockdown in adult escort cells did not have a significant effect on the percentage of germaria with dying germline cysts relative to Luc control knockdown (Fig. S4C), suggesting that Hr4 is not required in these cells for survival of early germline cysts.

Although the original RNAi screen that identified Hr4 as a regulator of oogenesis targeted somatic cells (Weaver and Drummond-Barbosa, 2019), we also investigated whether Hr4 is required in the germline for oogenesis based on its relatively high levels of expression in the ovary (Fig. 7A). Consistent with that possibility, the mammalian Hr4 homolog, GCNF, is required in the germline for spermatogenesis and oogenesis (Sabour et al., 2014). We first asked whether Hr4 is required in the germline for GSC maintenance using three separate Hr4 lethal alleles, namely Hr4¹ (King-Jones et al., 2005), Hr4^W728X and Hr4^Q867X (Haelterman et al., 2014) (Fig. S1A), for genetic mosaic analysis (Fig. 8C-E). In ‘mock’ mosaic females, in which both GFP-negative and -positive cells are wild-type, we observed GSC loss events (recognizable by the presence of GFP-negative cystoblasts/cysts in the absence of the original GFP-negative GSC; Laws and Drummond-Barbosa, 2015) in ∼6% of germaria with a mosaic germline (Fig. 8E). By contrast, homozygous mutant GFP-negative GSCs in Hr4 mosaic germaria were lost at significantly higher frequencies (Fig. 8E), indicating that Hr4 is intrinsically required for GSC maintenance.

We also used genetic mosaics to test whether Hr4 is required in the germline to control the number of progeny produced by GSCs. We counted the average number of GFP-negative or -positive cystoblasts and cysts (i.e. GSC progeny) generated per GFP-negative or -positive GSC, respectively, and found that the numbers of homozygous Hr4 mutant GFP-negative and control GFP-positive GSC progeny were comparable (Fig. 8F). These results suggest that neither GSC proliferation nor early germline cyst survival, both of which control the number of GSC progeny observed, require Hr4 function in the germline. Hr4 was similarly not required for cyst growth or survival of vitellogenic follicles, as GFP-negative follicles developed at similar rates relative to GFP-positive control follicles (Fig. S5) and dying follicles were not observed in Hr4 genetic mosaic ovarioles (n=>50 mosaic ovarioles per genotype). Taken together, these results show that Hr4 is required in the germline for GSC maintenance, but not for GSC proliferation, early germline cyst survival, follicle growth or vitellogenesis.

Nuclear receptors are transcriptional regulators that bind a variety of circulating ligands to couple the function of organs throughout the body with the physiological environment (Mazaira et al., 2018). Although many studies have focused on how nuclear receptors cell autonomously regulate tissue function (Ables and Drummond-Barbosa, 2017; Drummond-Barbosa, 2019; Evans and Mangelsdorf, 2014; King-Jones and Thummel, 2005; Laws and Drummond-Barbosa, 2017), it is clear that nuclear receptors can indirectly regulate remote tissues, including the Drosophila ovary (Sieber and Spradling, 2015; Weaver and Drummond-Barbosa, 2019), presumably through secreted targets. This study adds to this growing body of literature highlighting how nuclear receptor activity in multiple cell types/tissues influences distinct stages of Drosophila oogenesis. In this study, we found that Hr4, which encodes the Drosophila GCNF homolog, is required in the germline for GSC maintenance and in the niche to regulate GSC proliferation. Notably, we uncovered a novel role for Hr4 in adult muscles in the control of both GSC maintenance and follicle growth. Our results add to the few known examples of skeletal muscle-derived factors controlling GSCs or oogenesis in any organism.

Drosophila Hr4 and its mammalian homolog, GCNF, are both required in the germline for female fertility, although the extent of similarity between the specific mechanisms involved remains to be determined. Oocyte-specific knockout of GCNF in mice causes reduced fertility owing to overactive oocyte-to-follicle cell signaling (via de-repression of GCNF direct targets, BMP15 and GDF9), which in turn leads to altered estrous cycle and steroid signaling (Lan et al., 2003b). By contrast, we did not find a germline requirement for Hr4 during follicle development, arguing against defects in paracrine signaling between the germline and surrounding follicle cells in Drosophila. Instead, we uncovered a cell autonomous requirement for Hr4 in GSC maintenance. In mammalian females, however, the timing of proliferation of germline precursors is different from that in Drosophila, in that germ cell precursors exist and proliferate only during fetal stages and not in adults (Lei and Spradling, 2013). Therefore, it remains unclear from previous studies targeting growing oocytes (Lan et al., 2003a) whether GCNF might also be required for maintenance of a germ cell precursor pool during female fetal development.

Here, we show a new role for muscles in controlling GSC numbers and follicle growth. It is unlikely that the increased GSC loss and reduced follicle growth observed when Hr4 is knocked down in adult muscles is due to muscle wasting, changes in feeding behavior or mating defects. Specifically, different types of stress often alter the survival of early germline cysts and/or the degeneration of vitellogenic follicles. For example, Wolbachia infection affects death of early germline cysts (Fast et al., 2011), whereas organ wasting due to tumors results in reduced insulin signaling and increased follicle death (Figueroa-Clarevega and Bilder, 2015; Kwon et al., 2015). Changes in diet (Drummond-Barbosa and Spradling, 2001), adipocyte-specific reduced amino acid uptake (Armstrong et al., 2014) or metabolic changes (Matsuoka et al., 2017), and altered insulin signaling in the germline (LaFever and Drummond-Barbosa, 2005) or in adipocytes (Armstrong and Drummond-Barbosa, 2018) influence survival of early germline cysts and/or vitellogenic follicles. Therefore, if nutrients were limiting (e.g. due to altered feeding behavior), survival at these two stages would be affected; similarly, mating problems would lead to reduced survival of vitellogenic follicles (Soller et al., 1997). Muscle-specific knockdown of Hr4, however, does not lead to these death phenotypes and instead specifically affects GSC number and follicle cell growth.

Muscles have recently emerged as an endocrine organ that helps maintain energy homeostasis through the secretion of myokines (muscle-derived cytokines) that act on various organs (Severinsen and Pedersen, 2020). For example, mammalian skeletal muscle has a secretome comprising over 100 different myokines (Pedersen and Febbraio, 2012) that are altered in response to exercise and stress (Hoffmann and Weigert, 2017). Skeletal muscle-derived myostatin is highly expressed in rheumatoid arthritis patients and negatively regulates bone formation (Dankbar et al., 2015). Muscle-derived follistatin-like 1 inhibits injury-induced vascular smooth muscle cell proliferation in mice (Miyabe et al., 2014). In adult Drosophila, the transcriptional factor FOXO functions in skeletal muscle to control lipid synthesis in adipocytes and gut via a multi-organ signaling cascade (initiated by muscle-derived Upd2 cytokine) under circadian control (Zhao and Karpac, 2017). In addition, muscle-derived BMP signaling modulates brain dopamine biosynthesis to modulate Drosophila feeding behavior (Robles-Murguia et al., 2020). Muscle-derived factors have also been reported to influence germ cell behavior in Caenorhabditiselegans. For example, loss of the skeletal muscle extracellular matrix protein encoded by him-4 (hemicentin) causes germline chromosome segregation defects due to cleavage furrow instability (Vogel and Hedgecock, 2001; Xu and Vogel, 2011). SWM-1, a secreted protein produced by body wall muscles, antagonizes the spermiogenesis activator TRY-5 within the gonad to regulate the timing of post-meiotic sperm activation (Chavez et al., 2018). Our findings reveal that Hr4 is required in adult Drosophila female muscles to promote GSC maintenance and follicle growth, indicating that as-yet-unidentified muscle-derived factors control Drosophila oogenesis, further underscoring the importance of investigating the endocrine role of muscles.

The diverse cellular requirements for Hr4 in regulating distinct processes during Drosophila oogenesis suggest that Hr4 acts via multiple targets. In accordance, several key targets have been identified for the mammalian homolog GCNF. For example, GCNF acts as a transcriptional repressor of BMP15 and GDF9 in oocytes (Lan et al., 2003b), of Oct4 (also known as Pou5f1) in male and female germ cells (Sabour et al., 2014), and of Oct4 and Nanog in embryonic stem cells and during early embryo development (Gu et al., 2005). Conversely, GCNF acts as a transcriptional activator of cyclin D1 during embryonic stem cell differentiation (Wang et al., 2014). Given that we uncovered a cell autonomous requirement for Hr4 during earlier germline stages (in GSCs as opposed to in later follicles), it is likely that targets will be different. In fact, although BMP signals are required for Drosophila GSC maintenance (Xie and Spradling, 1998), they are produced in the niche (and not in GSCs themselves) and Hr4 knockdown in the niche had only a minor and inconsistent effect on GSC numbers (see Fig. S4A,B). In the future it will be important to determine the factors (and relevant mechanisms) acting downstream of Hr4 in the early germline (to control GSC maintenance), in cap cells (to control GSC proliferation), in muscles (to control GSC maintenance and follicle growth) and in as-yet-unidentified tissues (to control the survival of early germline cysts and vitellogenic follicles) in order to understand how different subsets of Hr4 targets in different cell types/tissues orchestrate the diverse effects of Hr4 on oogenesis.

In summary, our findings show new roles for Hr4 in regulating multiple steps of oogenesis in adult Drosophila (Fig. 8G). We found that Hr4 activity is required in muscles as well as in different cell types in the ovary for the control of specific steps of oogenesis. As mentioned above, however, additional somatic tissues require Hr4 for oogenesis, as ubiquitous somatic Hr4 knockdown led to early germline cyst death and degeneration of vitellogenic follicles (which we did not observe in our cell type/tissue-specific manipulations). In the future, it will be important to determine the downstream targets of Hr4 in each of the tissues that require its function, and how these targets modulate different Hr4-dependent processes during oogenesis. It will be particularly interesting to identify the endocrine factors produced by muscles downstream of Hr4 and their mechanisms of action. Notably, secreted factors downstream of Hr4 in any one organ could be acting directly on the ovary or through one or more intermediate tissues/organs (and their respective endocrine signals) to influence oogenesis. Finally, GCNF/Hr4 are orphan nuclear receptors, and the identification of their ligands and how they are regulated to tie Hr4 function with the physiological state of the organism is also an important research direction.

Drosophila stocks were maintained on medium containing 58 g/ml molasses, 46.5 g/ml yellow cornmeal, 4.7 g/ml agar, 17.4 g/ml active dry yeast, 0.1% tegosept, and 7.25 mM propionic acid at 22-25°C. Medium was supplemented with wet yeast paste for all experiments, unless otherwise noted. Previously described Gal4 and Gal80 lines were used, including tub-Gal4 (Nabel-Rosen et al., 2002), MHC-Gal4 (Schuster et al., 1996), hh-Gal4^JF (Jenett et al., 2012; Weaver et al., 2020), c587-Gal4 (Hsu and Drummond-Barbosa, 2009; Zhu and Xie, 2003), tub-Gal80^ts (McGuire et al., 2003), nSyb-Gal80 (Harris et al., 2015), Hr4¹ (King-Jones et al., 2005), Hr4^W728X FRT19A (Haelterman et al., 2014) and Hr4^Q867X FRT19A (Haelterman et al., 2014). The UAS-GFP.nls line was obtained from the Bloomington Drosophila Stock Center (BDSC; bdsc.indiana.edu). UAS-Luc^JF01355 (Matsuoka et al., 2017), UAS-Hr4^HM05260 and UAS-Hr4^MHJ21497 (Ni et al., 2011) (Transgenic RNAi Project; fgr.hms.harvard.edu) were obtained from the BDSC. UAS-Hr4^GD1464 was obtained from the Vienna Drosophila Resource Center (VDRC; stockcenter.vdrc.at); however, it should be noted that this line was difficult to keep alive and is no longer available at the VDRC. Lines carrying multiple genetic elements were generated by standard crosses. Balancer chromosomes, FRT and FLP strains, and other genetic elements, are described in Flybase (www.flybase.org).

Egg production was measured by maintaining five females of appropriate genotype (with five y w males) in perforated plastic bottles capped with molasses/agar plates overlaid with yeast paste, which were changed twice daily, as previously described (Weaver and Drummond-Barbosa, 2019). The number of eggs laid per day was counted in five replicates per genotype and results were subjected to a paired Student's t-test.

Females containing the UAS-hairpin of interest (against Luc or Hr4) in combination with either y w; tub-Gal80^ts/cyo; tub-Gal4/TM6B (for adult ubiquitous somatic RNAi), y w; tub-Gal80^ts/cyo; MHC-Gal4/TM6B (for adult muscle-specific RNAi), y w; tub-Gal80^ts/cyo; hh-Gal4^JF/TM6B (for adult terminal filament and cap cell-specific RNAi) and y w nSyb-Gal80 c587-Gal4/FM7a; tub-Gal80^ts/TM6B (for adult escort cell-specific RNAi) were raised at 18°C to block Gal4 activity (and thus RNAi) during development. Zero- to 2-day-old females (with y w males) were kept at 18°C for 3 days and subsequently switched to 29°C for RNAi induction. UAS-luc^JF01355 was used as an RNAi control. As additional controls, females of similar genotypes but without Gal4/Gal80^ts were raised and maintained under similar conditions for analysis at 4, 10 and 15 days after switching to 29°C.

Ovaries and other tissues were dissected in Grace's Insect Medium (Bio Whittaker), fixed, and washed as previously described (Weaver and Drummond-Barbosa, 2019). Samples were blocked for at least 3 h at room temperature in 5% normal goat serum (NGS; Jackson ImmunoResearch) plus 5% bovine serum albumin (BSA, Sigma-Aldrich) in phosphate-buffered saline [PBS; 10 mM NaH₂PO₄/NaHPO₄, 175 mM NaCl (pH 7.4)] containing 0.1% Triton X-100 (PBST), and incubated overnight at 4°C in primary antibodies diluted in blocking solution as follows: mouse monoclonal anti-α-Spectrin [3A9, Developmental Studies Hybridoma Bank (DSHB), 1:25]; mouse monoclonal anti-Lamin C (LC28.26, DSHB, 1:100); chicken polyclonal anti-GFP (ab13970, Abcam, 1:1000); rat monoclonal anti-E-Cadherin (DCAD2, DSHB, 1:3); rat monoclonal anti-Vasa (anti-vasa, DSHB, 1:20); and rabbit polyclonal anti-phospho-Histone H3 (Ser10, Millipore Sigma, 1:200). Samples were washed in PBST and incubated for 2 h at room temperature in 1:400 Alexa Fluor 488- or 568-conjugated goat species-specific secondary antibodies (A11004, A10667, A11034, A21212, A11039, Invitrogen) in blocking solution. For staining of pMad, ovaries were dissected in Grace's Insect Medium and fixed for 15 min in 4% formaldehyde in Grace's Medium at room temperature. Samples were washed four times for 20 min each in PBS containing 0.3% Triton-X100. Samples were blocked in 1× PBS containing 4% NGS for 1 h and then incubated overnight at 4°C with rabbit polyclonal anti-SMAD (phospho S423+S425, Abcam, 1:50). Samples were washed four times in PBS containing 0.3% Triton-X100 and incubated for 2 h at room temperature in 1:200 Alexa Fluor 568-conjugated goat rabbit-specific secondary antibodies (A11034, Invitrogen) in blocking solution. Stained samples were washed, mounted in Vectashield containing 1.5 µg/ml 4′,6-diamidino-2-phenylindole (DAPI) (Vector Laboratories), and imaged using a Zeiss LSM700 confocal microscope.

Cap cells and GSCs were identified as previously described (Weaver and Drummond-Barbosa, 2019), and two-way ANOVA with interaction (GraphPad Prism) was used to calculate the statistical significance of any differences among genotypes in the rate of cap cell or GSC loss from at least three independent experiments, as previously described (Armstrong et al., 2014). Progression through vitellogenesis was assessed using DAPI staining, as previously described (Weaver and Drummond-Barbosa, 2019). Three independent experiments were performed and subjected to a paired Student's t-test for statistical analysis. For pMAD quantification, the densitometric mean of individual GSC nuclei was measured from optical sections containing the largest nuclear diameter (visualized by DAPI) using ImageJ (Armstrong et al., 2014). For E-Cadherin quantification, the total densitometric value from maximum intensity projections around the cap cells (identified using LamC staining) was measured with ImageJ (Weaver and Drummond-Barbosa, 2018). Three independent experiments were performed, and results subjected to a Student's paired two-tailed t-test.

For EdU incorporation analysis, intact dissected ovaries were incubated for 1 h at room temperature in 100 µM EdU (Molecular Probes) diluted in Grace's Medium, washed, and fixed as described above. Following primary antibody incubation, samples were subjected to the Click-iT reaction according to the manufacturer's instructions (Life Technologies) for 30 min at room temperature. For each genotype, we measured the fraction of EdU- or phospho-histone H3-positive GSCs or follicle cells (from stage 4-6 follicles) as a percentage of the total number of GSCs or follicle cells, respectively, analyzed, as previously described (Hsu et al., 2008; LaFever et al., 2010).

To detect dying germline cysts, the ApopTag Indirect In Situ Apoptosis Detection Kit (Millipore Sigma) was used according to the manufacturer's instructions as previously described (Weaver and Drummond-Barbosa, 2019). Briefly, fixed and teased ovaries were rinsed in equilibration buffer twice for 5 min each at room temperature. Samples were incubated in 100 µl TdT solution at 37°C for 1 h with mixing at 15-min intervals. Ovaries were washed three times in 1× PBS followed by incubation in anti-digoxigenin conjugate for 30 min at room temperature protected from light. Samples were washed four times in 1× PBS and processed for immunofluorescence as described above.

Females of genotype y w hs-FLP His2AvGFP FRT19A/Hr4* FRT19A were generated through standard crosses (Hr4* represents hypomorphic or wild-type alleles of Hr4). Zero- to 3-day-old females were maintained on standard food with dry yeast and heat shocked twice daily at 37°C for 3 days to induce flippase (FLP)/FLP recognition target (FRT)-mediated mitotic recombination, as previously described (Laws and Drummond-Barbosa, 2015). After the final heat shock, females were maintained on medium supplemented with wet yeast paste for 7 days before dissection. Hr4* homozygous clones were recognized based on their loss of GFP expression. To quantify GSC loss, we calculated the percentage of germaria containing GFP-negative cystoblasts and/or cysts that had lost the GFP-negative mother GSC, which we refer to as ‘GSC loss events’ (Laws and Drummond-Barbosa, 2015). Follicle growth was assessed by comparing the size of GFP-negative follicles relative to that of neighboring wild-type GFP-positive follicles. Three independent experiments were performed, and statistical significance was calculated using a paired Student's t-test.

Five whole females, five pairs of ovaries, 10 abdominal carcasses (including adipocytes and oenocytes, but without ovaries and gut), 10 guts, 10 thoraces or 10 heads were incubated in RNAlater (Ambion) for 30 min, then incubated in 250 µl lysis buffer from the RNAqueous-4PCR DNA-free RNA isolation for RT-PCR kit (Ambion), and RNA was extracted using a motorized pestle and following the manufacturer's instructions, as previously described (Weaver and Drummond-Barbosa, 2019). cDNA was synthesized from 500 ng of total RNA for each sample using Supercript II Reverse Transcriptase (Thermo Fisher Scientific) according to the manufacturer's instructions. For RT-PCR analysis, Hr4 and Rp49 transcripts from 500 ng cDNA were amplified using EconoTaq Plus Green 2X Master Mix (Lucigen) and samples were run on a 0.8% agarose gel. PowerUp SYBR Green Master Mix (Thermo Fisher Scientific) was used for RT-qPCR. The reactions for three independent biological replicates were performed in triplicate using LightCycler 96 (Roche). Amplification fluorescence threshold was determined by LightCycler 96 software, and ΔΔCT were calculated using Microsoft Excel. Rp49 transcript levels were used as a reference. Fold change of transcript levels was calculated using the equation 2^−ΔΔCt (Microsoft Excel). The primers used for all PCR reactions are listed in Table S1. Rp49 primers were used as a control.

We thank the DSHB for antibodies and the BDSC (National Institutes of Health, P400D018537), VDSC (www.vdrc.at; Dietzl et al., 2007), Hugo Bellen, Kirst King-Jones and Chris Potter for Drosophila stocks. We are thankful for Flybase (www.flybase.org), an essential Drosophila research resource.

The peer review history is available online at https://dev.biologists.org/lookup/doi/10.1242/dev.198663.reviewer-comments.pdf