Downregulation of NHP2 promotes proper cyst formation in Drosophila ovary

Abstract

In Drosophila ovary, germline stem cells (GSCs) divide to produce two daughter cells. One daughter is maintained as a GSC, whereas the other initiates cyst formation, a process involving four synchronous mitotic divisions that form 2‐, 4‐, 8‐, and 16‐cell cysts. In this study, we found that reduction in the level of NHP2, a component of the H/ACA small nucleolar ribonucleoprotein complex that catalyzes rRNA pseudouridylation, promotes progression to 8‐cell cysts. NHP2 protein was concentrated in the nucleoli of germline cells during cyst formation. NHP2 expression, as well as the nucleolar size, abruptly decreased during progression from 2‐cell to 4‐cell cysts. Reduction in NHP2 activity in the germline caused accumulation of 4‐ and 8‐cell cysts and decreased the number of single cells. In addition, NHP2 knockdown impaired the transition to 16‐cell cysts. Furthermore, a tumorous phenotype caused by Sex‐lethal (Sxl) knockdown, which is characterized by accumulation of single and two‐cell cysts, was partially rescued by NHP2 knockdown. When Sxl and NHP2 activities were concomitantly repressed, the numbers of four‐ and eight‐cell cysts were increased. In addition, Sxl protein physically interacted with NHP2 mRNA in ovaries. Thus, it is reasonable to conclude that Sxl represses NHP2 activity at the post‐transcriptional level to promote proper cyst formation. Because NHP2 knockdown did not affect global protein synthesis in the germarium, we speculate that changes in NHP2‐dependent pseudouridylation, which is involved in translation of specific mRNAs, must be intact in order to promote proper cyst formation.

1 INTRODUCTION

In Drosophila, the adult ovary consists of approximately 20 elongated tubes, called ovarioles. At the distal tip of each ovariole is a germarium containing two or three germline stem cells (GSCs; Figure 1a;Lin, 1997; Xie & Spradling, 1998). GSCs asymmetrically divide to produce two daughter cells (Lin, 1997). The daughter cell adjacent to niche cells called cap cells is maintained as a GSC, whereas the other daughter cell (cystoblast) is detached from the niche, initiates differentiation, and undergoes cyst formation, a process involving four synchronous mitotic divisions with incomplete cytokinesis, to form two‐, four‐, eight‐, and 16‐cell cysts interconnected by fusomes (de Cuevas, Lilly, & Spradling, 1997; Lin, Yue, & Spradling, 1994; Robinson, Cant, & Cooley, 1994; Figure 1a). Among the 16 cells in a cyst, one differentiates into oocyte, and the other 15 become nurse cells that support oocyte maturation (Figure 1a).

Figure 1

Expression of NHP2 mRNA in germarium. (a) Schematic diagram of Drosophila germarium. Germline stem cells (GSCs)s (light green) adhere to cap cells (orange) located at the distal tip of the germarium. In region 1 (R1), a GSC divides to produce two daughter cells: one remains as a GSC, whereas the other (cystoblast) differentiates. In this article, the GSC and cystoblasts are referred to as single cells. Cystoblast undergoes four synchronous mitotic division with incomplete cytokinesis (cyst formation) to form two‐, four‐, eight‐, and 16‐cell cysts interconnected by the fusome (red), a derivative of the spectrosome (red) in the single cell. In region 2 (R2), among the 16 cells, one differentiates into an oocyte (blue) and remaining 15 cells become nurse cells. Oocyte and nurse cells are encapsulated by a monolayer of follicle cells (light brown). (b) NHP2 encodes two mRNA variants, NHP2‐RA and NHP2‐RB. Exons (boxes), introns (straight lines), protein‐coding regions (black boxes), Sxl‐binding motif (red), and the region detected by the RNA probe (green line) are shown. (c–c”) nos‐Gal4 germarium in situ hybridized with an RNA probe for NHP2 (green in c and c') and immunostained for Vasa, a marker of germline cells (magenta in c and c”). White brackets indicate R1 and R2. Scale bar: 10 µm. (d, d', e and e') nos‐Gal4 germarium stained for NHP2 mRNA (gray in d and e); Hts, a marker of spectrosomes and fusomes; FasIII, a marker of a subset of somatic cells including follicle cells (green in d' and e'); and Vasa (magenta in d' and e'). Single cells (arrowheads in d and d'), two‐cell cysts (arrows in d and d'), four‐cell cysts (arrowheads in e and e'), and eight‐cell cysts (arrows in e and e') are indicated by white dotted circles. Scale bars: 10 µm

A great deal of effort has been devoted to elucidating the mechanisms regulating GSC maintenance and self‐renewal by short‐range signals from the niche. GSC daughter cells that escape from the niche signal undergo differentiation and cyst formation (Figure 1a;Kirilly & Xie, 2007; Spradling, Fuller, Braun, & Yoshida, 2011). Upon differentiation, the daughter cell expresses bag of marbles (bam) gene, which is both necessary and sufficient for its differentiation (McKearin & Ohlstein, 1995; McKearin & Spradling, 1990; Ohlstein & McKearin, 1997). This is supported by the fact that germline cells lacking bam function fail to progress through cyst formation, resulting in over‐proliferation of GSC‐like single cells (McKearin & Ohlstein, 1995; McKearin & Spradling, 1990). By contrast, ectopic expression of bam in GSCs prevents these cells from maintaining stem cell identity (Ohlstein & McKearin, 1997). Transcriptional silencing of bam in GSCs is induced by the niche signal mediated by Decapentaplegic (Dpp), a member of the bone morphogenetic protein (BMP) ligand family (Chen & McKearin, 2003a). Once GSCs receive Dpp ligand from the niche, Mothers against dpp (Mad) protein is phosphorylated in GSCs, which in turn represses bam transcription by binding to a bam silencer element (Song et al. 2004).

However, the mechanism that regulates cyst formation and differentiation of GSC daughters remains to be fully elucidated. Recent studies have highlighted the roles of ribosome biogenesis and protein synthesis in control of the transition between GSC maintenance and differentiation (Sanchez et al. 2016; Xi, Doan, Liu, & Xie, 2005; Zhang, Shalaby, & Buszczak, 2014). For example, under‐developed (udd) encodes a component of the RNA polymerase I regulatory complex, which is involved in rRNA transcription in nucleolus. Disruption of udd results in a reduction in Mad protein expression, which in turn prevents maintenance of GSCs (Zhang et al. 2014). This is compatible with the observation that GSCs retain a large nucleolus, and its size changes dynamically during cyst formation (Sanchez et al. 2016; Zhang et al. 2014). These data lead us to speculate that ribosome biogenesis occurring in the nucleolus may affect cyst formation and differentiation, in addition to GSC maintenance.

To address this issue, we focused on NHP2, a Drosophila homolog of yeast Nhp2, which is widely conserved among eukaryotes (FlyBase: http://flybase.org). The yeast Nhp2 protein is a component of the H/ACA small nucleolar ribonucleoprotein complex (H/ACA snoRNP), which consists of one unique H/ACA RNA and four common core proteins, Nhp2, Gar1, Nop10, and Cbf5 (Henras et al. 1998; Watkins & Bohnsack, 2012). H/ACA snoRNPs catalyze site‐specific rRNA pseudouridylation guided by unique base‐pairing between the guide sequence of H/ACA RNAs and substrate rRNA (Bousquet‐Antonelli, Henry, Gélugne, Caizergues‐Ferrer, & Kiss, 1997; Ganot, Bortolin, & Kiss, 1997; Ge & Yu, 2013; Ni, Tien, & Fournier, 1997). In this study, we examined the role of NHP2 in cyst formation and differentiation of GSC daughters during early oogenesis.

2 MATERIALS AND METHODS

2.1 Flies

Flies were maintained on standard Drosophila medium at 25°C. The germline‐specific Gal4 driver nanos‐Gal4::VP16 (nos‐Gal4) (Van Doren, Williamson, & Lehmann, 1998) was used to express the following UAS constructs: UAS‐Sxl^RNAi (34393) and UAS‐NHP2^RNAi (51784) from the Bloomington Drosophila Stock Center (BDSC), and UASp‐Flag‐GFP‐Sxl and UASp‐Flag‐GFP (Ota et al. 2017). Snf¹⁴⁸ (7398) was provided by BDSC. NHP2‐EGFP and UASp‐NHP2 constructs were generated in this study (see following sections).

2.2 EGFP knock‐in mediated by CRISPR/Cas9

The clustered regularly interspaced short palindromic repeat (CRISPR)/CRISPR‐associated endonuclease (Cas9) system was used to generate an EGFP knock‐in allele of the NHP2 locus by expressing an EGFP knock‐in construct, Cas9, and guide RNA (gRNA;Gokcezade, Sienski, & Duchek, 2014). Using CRISPR Optimal Target Finder, the EGFP knock‐in site was determined to be just upstream of the stop codon of the NHP2 protein‐coding region. The gRNA sequence encompassing the knock‐in site was amplified using primer pair NHP2‐EGFP‐sgRNA‐Fw/NHP2‐EGFP‐sgRNA‐Rv (Supporting Information Table S1), and cloned into BbsI‐digested pDCC6; this vector expresses gRNA and Cas9 under the control of the U6‐2 and hsp70Bb promoters, respectively (Gokcezade et al. 2014).

The EGFP knock‐in construct contains the EGFP‐coding sequence flanked by the 1‐kb regions upstream and downstream of the NHP2 stop codon. The EGFP coding sequence was amplified from pEGFP‐N1 (Clontech) using primer pair pEGFP‐N1‐Fw/pEGFP‐N1‐Rv (Supporting Information Table S1). The genomic regions upstream and downstream of the NHP2 stop codon were amplified from genomic DNA of y w flies using primer pairs NHP2‐upstream‐1 kb‐Fw/NHP2‐upstream‐1 kb‐Rv and NHP2‐downstream‐1 kb‐Fw/NHP2‐downstream‐1 kb‐Rv (Supporting Information Table S1), respectively. Using Gibson Assembly (New England BioLabs), the upstream, EGFP‐coding, and downstream regions were assembled in pBluescript II SK(+) amplified using primer pair pBlue‐Fw/pBlue‐Rv (Supporting Information Table S1). Approximately 0.1 nl of solution containing the NHP2 knock‐in construct and pDCC6 plasmid containing target sequences for NHP2 (150 ng/μl each in distilled water) were injected into y w embryos. Progeny carrying the EGFP insertion at the NHP2 target site was selected by PCR and used to establish stock lines.

2.3 Transgene

For construction of UASp‐NHP2, the open reading flame (ORF) of NHP2 was amplified from DNA clone RH66170 from the Drosophila Genomics Resource Center. Primer pair “NHP2‐RA‐Fw/NHP2‐RA‐Rv” (Supporting Information Table S1) was used to amplify the NHP2 ORF. The NHP2 ORF was then cloned into the KpnI/NotI digested pUASp vector (Rørth, 1998). Germline transformation was performed using y w embryos as recipients. w⁺ transformants were crossed to w; Sp/CyO; PrDr/TM3 females to establish a homozygous stock of UASp‐NHP2.

2.4 UV‐crosslinked RNA immunoprecipitation (CLIP) and RT‐qPCR

To perform CLIP analysis, ovaries were dissected from nos‐Gal4, nos‐Gal4 UASp‐EGFP, and nos‐Gal4 UASp‐Sxl‐EGFP flies. Twenty ovaries were homogenized in 10 μl of ice‐cold buffer‐A [8.3 mM KCl, 1.7 mM NaCl, 1 mM Tris/HCl (pH 7.5), Complete Protease Inhibitor Cocktail (Roche), and 100 U/ml RNasin Plus RNase Inhibitor (Promega)] and centrifuged at 20,000 g for 10 min at 4°C. Five microliters of supernatant was mixed with 5 μl of binding buffer [BB: 200 mM KCl, 200 ng/μl yeast tRNA, 1 mg/ml heparin, 20 mM HEPES (pH 7.2), 6 mM MgCl₂, 2 mM DTT and 10% glycerol], irradiated using a CL‐1000 Ultraviolet Crosslinker (UVP) at 17200 μJ/cm², and diluted to a final volume of 110 μl with PBST (PBS containing 0.1% TritonX‐100 and 0.1% Tween20). RNAs were extracted using ISOGEN (NIPPON GENE) from 10 μl of the solution (representing 10% input). The remaining 100 μl of the solution was incubated with anti‐GFP antibody (Thermo Fisher Scientific) in the presence of protein A–Sepharose beads (GE Healthcare) for 1 hr at 4°C. After incubation, beads were precipitated and rinsed three times each with PBST and Pro‐K buffer [0.01 M Tris‐HCl (pH 7.8), 0.01 M EDTA, 0.5% SDS], and then incubated in the presence of 2 mg/ml protease K in Pro‐K buffer for 20 min at 37°C. RNAs were extracted from the immunoprecipitate using ISOGEN. cDNAs were synthesized using SuperScript III (Thermo Fisher Scientific). NHP2, nos, Tub84D, and rp49 were amplified using primer pairs NHP2‐qPCR‐Fw/NHP2‐qPCR‐Rv, nos‐qPCR‐Fw/nos‐qPCR‐Rv, Tub84D‐qPCR‐Fw/Tub84D‐qPCR‐Rv, and rp49‐qPCR‐Fw/rp49‐qPCR‐Rv (Supporting Information Table S2), respectively. Quantification was performed on a Light Cycler 480 system (Roche) with the QuantiTect SYBR Green RT‐PCR Kit (Qiagen). Thermal cycling conditions were as follows: 15 min at 95°C, followed by 55 cycles of 15 s at 94°C, 30 s at 60°C, and 20 s at 72°C. Data were analyzed using the Light Cycler Software (Roche) and Microsoft Excel (Microsoft). Three independent experiments were carried out using different batches of flies.

The amounts of NHP2‐RA, NHP2‐RB (RA/RB), and NHP2‐RB (RB) transcripts were determined using external standards identical to the target sequences. Standard curves for NHP2‐RA/RB and NHP2‐RB were generated using a dilution series of SpeI‐digested pGEM‐T easy vector (Promega) containing a sequence common to NHP2‐RA and RB (RA/RB) or an NHP2‐RB‐specific (RB) cDNA fragment. Target sequences were amplified from cDNA synthesized from bam mutant ovaries using primer pairs NHP2‐RARB‐abs‐qPCR‐Fw/NHP2‐RARB‐abs‐qPCR‐Rv and NHP2‐RB‐abs‐qPCR‐Fw/NHP2‐RB‐abs‐qPCR‐Rv (Supporting Information Table S2). Quantification was performed using the Light Cycler 480 system and the QuantiTect SYBR Green RT‐PCR Kit. Data were analyzed using the Light Cycler Software and Microsoft Excel. Absolute transcript levels were calculated by comparing the Cp values of RA/RB and RB transcripts with standard curves. Three independent experiments were carried out using different batches of flies.

2.5 Immunostaining

Immunofluorescence staining of adult ovaries was performed as described (Hayashi, Hayashi, & Kobayashi, 2004). Ovaries were dissected from adult flies 5–10 days after eclosion. The following primary antibodies were used at the indicated dilutions: mouse anti‐Fibrillarin (1:500; Abcam), rabbit anti‐GFP (1:500; Thermo Fisher Scientific), rabbit anti‐Sxl (1:1000, a gift from Dr. Hiroshi Sakamoto), rabbit anti‐Bam (1:1000), rabbit anti‐active Caspase3 (1:1000; Abcam), chick anti‐Vasa (1:500), mouse anti‐Hts [1:10, 1B1; Developmental Studies Hybridoma Bank (DSHB)], and mouse anti‐FasIII (1:10, 7G10 anti‐FasciclinIII; DSHB). For detection of primary antibodies, the following secondary antibodies were used: Alexa Fluor 488–conjugated goat anti‐rabbit (1:500; Molecular Probe), Alexa Fluor 546–conjugated goat anti‐mouse (1:500; Molecular Probe), and Alexa Fluor 633–conjugated goat anti‐chick (1:500; Molecular Probe). Samples were mounted in ProLong^™ Diamond Antifade Mountant (Thermo Fisher Scientific) and observed by confocal microscopy SP5 (Leica Microsystems).

2.6 In situ hybridization

Whole‐mount in situ hybridization of adult gonads was performed as described (Hayashi et al. 2004). Total RNA was isolated from ovaries using ISOGEN, and cDNA was synthesized using SuperScript III. cDNA corresponding to NHP2 was PCR‐amplified using primer pair NHP2‐FISH‐Fw/NHP2‐FISH‐Rv (Supporting Information Table S1). The resultant cDNA amplicon was cloned into pGEM‐T Easy (Promega). Digoxigenin (DIG)‐labeled RNA probes were synthesized with T7 or SP6 RNA polymerase (Roche) using the fragment amplified from the plasmid using the T7 and SP6 primers. Signal was detected using horseradish peroxidase–conjugated anti‐DIG antibody (Roche) and amplified using the TSA Biotin System and streptavidin–fluorescein (FITC, PerkinElmer). In situ hybridization combined with immunostaining was performed as described (Hayashi et al. 2004). Ovaries were stained with chick anti‐Vasa, mouse anti‐Hts, and mouse anti‐FasIII antibodies. Secondary antibodies were Alexa Fluor 546–conjugated goat anti‐rabbit and Alexa Fluor 633–conjugated goat anti‐chick.

2.7 OPP staining

OPP staining was performed using the Click‐iT Plus OPP Alexa Fluor 488 Protein Synthesis Assay Kit (Molecular Probes). Ovaries were dissected in Grace's insect medium (Thermo Fisher Scientific) from adult flies 5–10 days after eclosion. The ovaries were incubated in 20 μM Click‐iT OPP reagent in Grace's insect medium for 30 min at room temperature. After incubation, ovaries were fixed in 4% paraformaldehyde in PBS for 20 min, and then washed with PBS containing 0.1% Tween‐20 and 0.1% Triton X‐100 (PBT) three times for 20 min each. For the Click‐iT reaction, ovaries were incubated in Click‐iT reaction cocktail in the dark at room temperature. After 30 min, ovaries were washed with Click‐iT reaction rinse buffer. Ovaries were then washed with PBT three times for 20 min each and immunostained as described above.

3 RESULTS AND DISCUSSION

3.1 Expression of NHP2 mRNA and its protein product during cyst formation

The NHP2 locus encodes two mRNA isoforms, which contain identical protein‐coding regions (Figure 1b; FlyBase: http://flybase.org). To examine the expression of NHP2 mRNA during cyst formation, we performed fluorescence in situ hybridization (FISH) of ovaries using a probe capable of detecting both mRNA isoforms (Figure 1b; green line). The signal for NHP2 mRNA was almost evenly distributed throughout the cytoplasm of the germline during cyst formation from single cells to 8‐cell cysts in region 1 (R1), and was increased in disc‐shaped 16‐cell cysts surrounded by follicle cells in region 2 (R2) (Figure 1c–c'', d, d', e and e'). Next, we examined expression of NHP2 protein during cyst formation. Because we were unable to raise an antibody against NHP2 protein, we inserted a tag sequence encoding an enhanced green fluorescent protein (EGFP) into the C‐terminus of the NHP2 coding region (NHP2‐EGFP) (Gokcezade et al. 2014), enabling us to detect NHP2 protein using an anti‐GFP antibody. We found that NHP2‐EGFP accumulated in nuclei of germline cells during cyst formation (Figure 2a–a”). In their nuclei, NHP2‐EGFP was enriched in a region stained by an antibody against Fibrillarin, a component of the nucleolus (Figure 2b–b”, c–c” and d–d”), indicating that NHP2‐EGFP was concentrated in the nucleoli of germline cells during cyst formation. This is consistent with the fact that the yeast ortholog, Nhp2, is a component of the H/ACA‐box small nucleolar RNAs and protein complex (H/ACA snoRNPs), which guides pseudouridylation of rRNA in the nucleolus (Henras et al. 1998; Kierzek et al. 2014; Maiorano, Brimage, Leroy, & Kearsey, 1999; Pogacic, Dragon, & Filipowicz, 2000).

Figure 2

Expression of NHP2 protein in germarium. (a–a”) NHP2‐EGFP germarium immunostained for EGFP (green in a and a') and Fibrillarin, a marker of nucleoli (magenta in a and a”). Scale bar: 10 µm. (b–b”, c–c” and d–d”) Distal part of R1 (b–b”), proximal part of R1 (c–c”), and R2 (d–d”) in NHP2‐EGFP germaria immunostained for Vasa (upper panels), EGFP (middle panels), and Fibrillarin (lower panels). White arrowheads indicate nucleoli in the germline nuclei. Scale bars: 5 µm. (e–j and e'–j') NHP2‐EGFP germaria immunostained for Vasa (green in e–j), Hts and FasIII (magenta in e–j), and EGFP (gray in e'–j'). Single cells (e and e'), two‐cell cysts (f and f'), four‐cell cysts (g and g'), eight‐cell cysts (h and h'), newly formed 16‐cell cysts (i and i'), and disc‐shaped 16‐cell cysts (j and j') are shown. White arrowheads indicate NHP2‐EGFP foci in germline nuclei. Scale bars: 5 µm

Changes in NHP2‐EGFP expression and nucleolar morphology were evident in the germline during cyst formation (Figure 2b–d, b'–d' and b”–d” ;Sanchez et al. 2016). Single cells and two‐, four‐, eight‐, and 16‐cell cysts can be distinguished by counting the number of germline cells interconnected by branched fusomes (Figure 1a). In single cells and two‐cell cysts in R1 of the germarium, we observed large nucleoli marked by NHP2‐EGFP (Figure 2e, e', f and f'). When these germline cells progressed to four‐ and eight‐cell cysts in R1, NHP2‐EGFP staining was significantly reduced (Figure 2g, g', h and h'), and nucleolar size was decreased (Figure 2c”). Although NHP2‐EGFP staining remained at a low level in newly formed 16‐cell cysts (Figure 2i and i'), in disc‐shaped 16‐cell cysts in R2, staining increased (Figure 2j and j'), and the nucleoli were expanded (Figure 2d”). These observations demonstrate that abrupt downregulation of NHP2 expression occurs during progression from two‐cell to four‐cell cysts.

3.2 Germline‐specific knockdown (GS‐KD) of NHP2 promotes progression to eight‐cell cysts

NHP2 GS‐KD causes a defect in egg‐chamber formation (Sanchez et al. 2016). NHP2 function is required in GSCs for complete cytokinesis, and reduction in the level of this protein results in formation of interconnected undifferentiated cells, or stem cysts (Sanchez et al. 2016). Consistent with this, NHP2 GS‐KD impaired 16‐cell cyst formation (Figure 3a, a', b, b' and c). Moreover, the numbers of single cells were drastically reduced in NHP2 GS‐KD germarium, and the numbers of four‐ and eight‐cell cysts were elevated, in comparison with nos‐Gal4 (Figure 3c). This suggests that NHP2 is required to maintain single cells, and that NHP2 downregulation initiates progression to four‐ and eight‐cell cysts. Next, we asked whether accumulation of four‐ and eight‐cell cysts in the NHP2 GS‐KD germarium results from progression of cyst formation, or merely from disruption of cytokinesis between undifferentiated germline stem cells (GSCs) (single cells). Knockdowns of translation initiation factors and ribosomal assembly factors cause formation of abnormal cysts (stem cysts) containing Bam‐negative undifferentiated germline cells connected with each other by branched fusomes in the germaria (Sanchez et al. 2016). Furthermore, stem cysts exhibit nucleolar hypertrophy, a characteristic of single cells and two‐cell cysts (Sanchez et al. 2016).

Figure 3

Defects in cyst formation in NHP2 GS‐KD germarium. (a, a', b and b') nos‐Gal4 (a and a') and NHP2 GS‐KD germarium (b and b') immunostained for Hts and FasIII (green) and Vasa (magenta in a and b). Scale bars: 10 µm. (c) Average numbers of single cells and two‐, four‐, eight‐, and 16‐cell cysts per a germarium from nos‐Gal4 (open bars) and NHP2 GS‐KD females (hatched bars) are indicated. Error bars indicate standard error. Significance of the difference relative to nos‐Gal4 was calculated by Student's t‐test (*p < 0.05). The number of germaria examined is shown in parentheses. (□, nos‐Gal4; , NHP2 GS‐KD)

Similar to four‐ and eight‐cell cysts in normal germaria, we found that four‐ and eight‐cell cysts in NHP2 GS‐KD germaria expressed high levels of Bam protein (Bam‐positive four‐ and eight‐cell cysts; Figure 4a–a” and b–b”). The number of Bam‐positive eight‐cell cysts was significantly higher in NHP2 GS‐KD than in nos‐Gal4 germaria (Table 1), although there was no significant difference in the number of Bam‐positive four‐ and eight‐cell cysts in NHP2 GS‐KD germaria did not exhibit nucleolar hypertrophy (Figure 4c–c” and d–d”). These observations suggest that NHP2 GS‐KD promotes proper differentiation to eight‐cell cysts, which retain proper cell identity as evidenced by high levels of Bam expression and smaller nucleolar size. However, differentiation into 16‐cell cysts appeared to be arrested in NHP2 GS‐KD germaria (Figure 3c), and some of these cysts may have eventually been eliminated by apoptosis (Supporting Information Figure S1a, a', b, b' and c).

Figure 4

Expression of Bam in 4‐ and 8‐cell cyst of NHP2 GS‐KD germarium. (a–a” and b–b”) nos‐Gal4 (a–a”) and NHP2 GS‐KD germarium (b–b”) immunostained for Bam (green in a, a', b and b'), Hts and FasIII (gray in a” and a”), and Vasa (magenta in A and B). White dotted circles indicate four‐cell cysts (arrowheads) and eight‐cell cysts (arrows). Yellow arrowheads indicate Bam‐negative four‐cell cyst. Scale bars: 10 µm. (c–c” and d–d”) nos‐Gal4 (c–c”) and NHP2 GS‐KD germarium (d–d”) immunostained for Bam (green in A and B), Fibrillarin (magenta in c, c', d and d'), and Vasa (gray in c” and d”). Bam‐positive cysts are indicated by white dotted circles. In d–d”, the germline cells above the Bam‐positive cyst expressed low levels of Bam and showed nucleolar hypertrophy. These cysts are regarded as “Bam‐negative cysts” (see text), and are classified as “stem cysts” based on their weak Bam expression and nucleolar morphology. Scale bars: 10 µm

Table 1. Number of Bam‐positive cysts in nos‐Gal4 and NHP2 GS‐KD germaria

Ovariesa ^aNHP2 GS‐KD ovaries were obtained from the females derived from nos‐Gal4/nos‐Gal4 females mated with UAS‐NHP2^RNAi/UAS‐NHP2^RNAi males. Ovaries were dissected from adults 5–10 days after eclosion, and stained for Bam, Hts, FasIII, and Vasa. Average numbers (±SE) of Bam‐positive and ‐negative cysts (Total) and Bam‐positive cysts (Bam^+ cysts) per germarium were calculated. A total of 16 nos‐Gal4 germaria and 19 NHP2 GS‐KD germaria were examined. Bam‐positive cysts were defined as those in which Bam staining was more intense than in the germline cells adjacent to the niche. *p > 0.05, **p < 0.05: Significance of the difference relative to the nos‐Gal4 was calculated using Student's t‐test.	Four‐cell cysts/germarium		Eight‐cell cysts/germarium
	Total	Bam^+ cysts	Total	Bam^+ cysts
nos‐Gal4	0.88 ± 0.20	0.88 ± 0.20	0.75 ± 0.17	0.75 ± 0.17
NHP2 GS‐KD	2.11 ± 0.38	0.79 ± 0.32* ^aNHP2 GS‐KD ovaries were obtained from the females derived from nos‐Gal4/nos‐Gal4 females mated with UAS‐NHP2^RNAi/UAS‐NHP2^RNAi males. Ovaries were dissected from adults 5–10 days after eclosion, and stained for Bam, Hts, FasIII, and Vasa. Average numbers (±SE) of Bam‐positive and ‐negative cysts (Total) and Bam‐positive cysts (Bam^+ cysts) per germarium were calculated. A total of 16 nos‐Gal4 germaria and 19 NHP2 GS‐KD germaria were examined. Bam‐positive cysts were defined as those in which Bam staining was more intense than in the germline cells adjacent to the niche. *p > 0.05, **p < 0.05: Significance of the difference relative to the nos‐Gal4 was calculated using Student's t‐test.	2.79 ± 0.27	1.79 ± 0.27** ^aNHP2 GS‐KD ovaries were obtained from the females derived from nos‐Gal4/nos‐Gal4 females mated with UAS‐NHP2^RNAi/UAS‐NHP2^RNAi males. Ovaries were dissected from adults 5–10 days after eclosion, and stained for Bam, Hts, FasIII, and Vasa. Average numbers (±SE) of Bam‐positive and ‐negative cysts (Total) and Bam‐positive cysts (Bam^+ cysts) per germarium were calculated. A total of 16 nos‐Gal4 germaria and 19 NHP2 GS‐KD germaria were examined. Bam‐positive cysts were defined as those in which Bam staining was more intense than in the germline cells adjacent to the niche. *p > 0.05, **p < 0.05: Significance of the difference relative to the nos‐Gal4 was calculated using Student's t‐test.

Notes

• ^aNHP2 GS‐KD ovaries were obtained from the females derived from nos‐Gal4/nos‐Gal4 females mated with UAS‐NHP2^RNAi/UAS‐NHP2^RNAi males. Ovaries were dissected from adults 5–10 days after eclosion, and stained for Bam, Hts, FasIII, and Vasa. Average numbers (±SE) of Bam‐positive and ‐negative cysts (Total) and Bam‐positive cysts (Bam⁺ cysts) per germarium were calculated. A total of 16 nos‐Gal4 germaria and 19 NHP2 GS‐KD germaria were examined. Bam‐positive cysts were defined as those in which Bam staining was more intense than in the germline cells adjacent to the niche. *p > 0.05, **p < 0.05: Significance of the difference relative to the nos‐Gal4 was calculated using Student's t‐test.

On the other hand, four‐ and eight‐cell cysts with undetectable or low levels of Bam expression (Bam‐negative four‐ and eight‐cell cysts) were frequently observed in the proximity of the distal tips of germaria, where niche cells reside. These Bam‐negative cysts were regarded as “stem cysts” (Figure 4b–b'', d–d'', Table 1). This observation suggests that stem cysts were also formed in NHP2 GS‐KD germaria. Thus, proper progression of cyst formation, along with abnormal stem‐cyst formation, results in the accumulation of four‐ and eight‐cell cysts in NHP2 GS‐KD germaria.

3.3 NHP2 GS‐KD partially rescues the defect caused by Sxl GS‐KD in cyst formation

The data presented above indicate that downregulation of NHP2 facilitates progression to eight‐cell cysts. This is compatible with the observation that expression of NHP2 protein is decreased in four‐ and eight‐cell cysts compared with single and two‐cell cysts (Figure 2e'–h'). We speculate that downregulation of the NHP2 protein promotes progression from two‐ to eight‐cell cysts through four‐cell cysts. Because NHP2 mRNA was almost evenly detected throughout cyst formation until 8‐cell cysts (Figure 1d, d', e and e'), it is reasonable to speculate that NHP2 protein production is repressed at a post‐transcriptional level in four‐ and eight‐cell cysts. The strongest candidate for the gene responsible for downregulating NHP2 during cyst formation is Sex lethal (Sxl). Reduction in Sxl activity causes a tumorous phenotype, which is characterized by accumulation of single cells and two‐cell cysts in germaria (Figure 5a, a' and c;Steinmann‐Zwicky, Schmid, & Nöthiger, 1989). Sxl GS‐KD caused a reduction in the numbers of four‐, eight‐, and 16‐cell cysts (Figure 5a, a' and c). The defects caused by Sxl GS‐KD were partially rescued by introduction of NHP2 GS‐KD (Figure 5b, b' and c). When Sxl and NHP2 activities were concomitantly repressed (NHP2‐Sxl GS‐KD), the tumorous phenotype was suppressed, and the numbers of four‐ and eight‐cell cysts were increased (Figure 5b, b' and c).

Figure 5

Defects in cyst formation in NHP2‐Sxl GS‐KD germarium. (a, a', b and b') Sxl GS‐KD (a and a') and NHP2‐Sxl GS‐KD germarium (b and b') immunostained for Hts and FasIII (green) and Vasa (magenta in a and b). Arrows and arrowheads indicate four‐ and eight‐cell cysts, respectively. Scale bars: 10 µm. (c) Average numbers of single cells and two‐, four‐, eight‐, and 16‐cell cysts per a germarium from Sxl GS‐KD (open bars) and NHP2‐Sxl GS‐KD females (hatched bars) are indicated. Error bars indicate standard error. Significance of the difference related to Sxl GS‐KD was calculated by Student's t‐test (*p < 0.05). The number of germaria examined is shown in parentheses. (d–d” and e–e”) Sxl GS‐KD (d–d”) and NHP2‐Sxl GS‐KD germarium (e–e”) immunostained for Bam (green in D, D', E and E'), Vasa (magenta in D and E), and Hts and FasIII (gray in d” and e”). Scale bars: 10 µm. (□, Sxl GS‐KD; , NHP2‐Sxl GS‐KD)

We next asked whether these four‐ and eight‐cell cysts properly express high levels of Bam protein, as is the case for normal cyst formation. In Sxl GS‐KD germaria, the accumulating single cells and two‐cell cysts exhibited modest levels of Bam expression (Figure 5d–d” ;Chau, Kulnane, & Salz, 2009). Because Bam expression is undetectable or low in these cells in normal germaria (Figure 4a–a'';Chen & McKearin, 2003b; McKearin & Ohlstein, 1995; McKearin & Spradling, 1990), single cells and two‐cell cysts may partially adopt four‐ and eight‐cell cyst identity in Sxl GS‐KD. However, we found that four‐ and eight‐cell cysts expressed Bam in NHP2‐Sxl GS‐KD at a much higher level than in single cells and two‐cell cysts in Sxl GS‐KD (Figure 5d–d” and e–e”). There was a statistically significant increase in the number of Bam‐positive eight‐cell cysts in NHP2‐Sxl GS‐KD germaria in comparison with those of Sxl GS‐KD, but no significant difference between NHP2‐Sxl and Sxl GS‐KD in the number of Bam‐positive four‐cell cysts (Table 2). These data suggest that single cells and two‐cell cysts in Sxl GS‐KD can progress properly to eight‐cell cysts by introduction of NHP2 GS‐KD. Based on these findings, it is reasonable to conclude that Sxl represses NHP2 activity to promote proper cyst formation. However, 16‐cell cysts were rarely observed in either NHP2 GS‐KD or NHP2‐Sxl GS‐KD germaria (Figures 3c and 5c). This suggests that NHP2 activity is also required to promote progression from eight‐ to 16‐cell cysts.

Table 2. Number of Bam‐positive cysts in Sxl GS‐KD and NHP2‐Sxl GS‐KD germaria

Ovariesa ^aSxl GS‐KD ovaries were obtained from females derived from nos‐Gal4/nos‐Gal4 females mated with UAS‐SxlRNAi/UAS‐SxlRNAi males. NHP2‐Sxl GS‐KD ovaries were obtained from females derived from nos‐Gal4/nos‐Gal4 females mated with UAS‐NHP2RNAi/UAS‐NHP2RNAi; UAS‐SxlRNAi/UAS‐SxlRNAi males. Ovaries were dissected from adults 5–10 days after eclosion, and stained for Bam, Hts, FasIII, and Vasa. Average numbers (±SE) of Bam‐positive and ‐negative cysts (Total) and Bam‐positive cysts (Bam+ cysts) per germarium were calculated. A total of 19 Sxl GS‐KD germaria and 11 NHP2‐Sxl GS‐KD germaria were examined. Bam‐positive cysts were defined as those in which Bam staining was more intense than in the germline cells adjacent to the niche. *p > 0.05, **p < 0.05: Significance of the difference relative to the Sxl GS‐KD was calculated using Student's t‐test.	Four‐cell cysts/germarium		Eight‐cell cysts/germarium
	Total	Bam^+ cysts	Total	Bam^+ cysts
Sxl GS‐KD	1.50 ± 0.20	0.70 ± 0.20	1.30 ± 0.17	0.10 ± 0.17
NHP2‐Sxl GS‐KD	4.64 ± 0.38	1.36 ± 0.32* ^aSxl GS‐KD ovaries were obtained from females derived from nos‐Gal4/nos‐Gal4 females mated with UAS‐SxlRNAi/UAS‐SxlRNAi males. NHP2‐Sxl GS‐KD ovaries were obtained from females derived from nos‐Gal4/nos‐Gal4 females mated with UAS‐NHP2RNAi/UAS‐NHP2RNAi; UAS‐SxlRNAi/UAS‐SxlRNAi males. Ovaries were dissected from adults 5–10 days after eclosion, and stained for Bam, Hts, FasIII, and Vasa. Average numbers (±SE) of Bam‐positive and ‐negative cysts (Total) and Bam‐positive cysts (Bam+ cysts) per germarium were calculated. A total of 19 Sxl GS‐KD germaria and 11 NHP2‐Sxl GS‐KD germaria were examined. Bam‐positive cysts were defined as those in which Bam staining was more intense than in the germline cells adjacent to the niche. *p > 0.05, **p < 0.05: Significance of the difference relative to the Sxl GS‐KD was calculated using Student's t‐test.	4.45 ± 0.27	2.36 ± 0.27** ^aSxl GS‐KD ovaries were obtained from females derived from nos‐Gal4/nos‐Gal4 females mated with UAS‐SxlRNAi/UAS‐SxlRNAi males. NHP2‐Sxl GS‐KD ovaries were obtained from females derived from nos‐Gal4/nos‐Gal4 females mated with UAS‐NHP2RNAi/UAS‐NHP2RNAi; UAS‐SxlRNAi/UAS‐SxlRNAi males. Ovaries were dissected from adults 5–10 days after eclosion, and stained for Bam, Hts, FasIII, and Vasa. Average numbers (±SE) of Bam‐positive and ‐negative cysts (Total) and Bam‐positive cysts (Bam+ cysts) per germarium were calculated. A total of 19 Sxl GS‐KD germaria and 11 NHP2‐Sxl GS‐KD germaria were examined. Bam‐positive cysts were defined as those in which Bam staining was more intense than in the germline cells adjacent to the niche. *p > 0.05, **p < 0.05: Significance of the difference relative to the Sxl GS‐KD was calculated using Student's t‐test.

• ^aSxl GS‐KD ovaries were obtained from females derived from nos‐Gal4/nos‐Gal4 females mated with UAS‐SxlRNAi/UAS‐SxlRNAi males. NHP2‐Sxl GS‐KD ovaries were obtained from females derived from nos‐Gal4/nos‐Gal4 females mated with UAS‐NHP2RNAi/UAS‐NHP2RNAi; UAS‐SxlRNAi/UAS‐SxlRNAi males. Ovaries were dissected from adults 5–10 days after eclosion, and stained for Bam, Hts, FasIII, and Vasa. Average numbers (±SE) of Bam‐positive and ‐negative cysts (Total) and Bam‐positive cysts (Bam+ cysts) per germarium were calculated. A total of 19 Sxl GS‐KD germaria and 11 NHP2‐Sxl GS‐KD germaria were examined. Bam‐positive cysts were defined as those in which Bam staining was more intense than in the germline cells adjacent to the niche. *p > 0.05, **p < 0.05: Significance of the difference relative to the Sxl GS‐KD was calculated using Student's t‐test.

3.4 Sxl protein binds to NHP2 mRNA during cyst formation

Sxl encodes an RNA‐binding protein that regulates sex‐biased translation and splicing of target mRNAs (Penalva & Sánchez, 2003). In the soma, Sxl is expressed in embryos in a female‐specific manner (Bopp, Bell, Cline, & Schedl, 1991; Keyes, Cline, & Schedl, 1992). Sxl protein regulates alternative splicing of transformer (tra) RNA to produce functional Tra protein only in female soma. Furthermore, Sxl protein binds to 5′ and 3′ untranslated regions (UTRs) of male‐specific lethal 2 (msl2) RNA in female soma to repress its translation (Bashaw & Baker, 1997; Beckmann, Grskovic, Gebauer, & Hentze, 2005; Kelley, Wang, Bell, & Kuroda, 1997; Lucchesi & Kuroda, 2015). Considering the molecular function of Sxl protein in the soma, we speculate that Sxl binds to NHP2 mRNA and represses its translation in four‐ and eight‐cell cysts in germaria.

NHP2 mRNA co‐immunoprecipitates with Sxl‐enhanced green fluorescent protein (EGFP) fusion protein in primordial germ cells and contains a Sxl‐binding motif (Figure 1b;Ota et al. 2017). The Sxl‐binding motif is located at the 3′ UTR of NHP2‐RB, but not in NHP2‐RA 3′ UTR (Figure 1b). We observed NHP2‐RB expression in single cells and two‐cell cysts in bam mutant ovaries (Figure 6a), but we could not determine whether NHP2‐RB was expressed in four‐ and eight‐cell cysts in normal germaria because the RB‐specific probe (116 bp) is too short to detect NHP2‐RB expression by in situ hybridization.

Figure 6

Binding of Sxl protein to NHP2 mRNA in oogenesis. (a) Expression level of both NHP2‐RA and NHP2‐RB mRNA (RA/RB, light gray) and NHP2‐RB mRNA (RB, dark gray) in bam mutant ovaries. To detect mRNA levels, qRT‐PCR was performed on cDNA derived from bam mutant ovaries. Average expression levels are presented in arbitrary units. Error bars indicate standard error of three biological replicates. (b) UV‐crosslinked RNA immunoprecipitation (CLIP) was performed using extracts obtained from adult ovaries from nos‐Gal4, nos‐Gal4 UASp‐EGFP, and nos‐Gal4 UASp‐EGFP‐Sxl females (Supporting Information Table S3). RT‐qPCR assay to detect NHP2‐RB, nos (a positive control), Tub84D, and rp49 (negative controls) mRNA were performed using 10% of total extract (10% input) and samples immunoprecipitated with anti‐GFP antibody. The amount of each RNA in the immunoprecipitated samples from nos‐Gal4, nos‐Gal4 UASp‐EGFP (EGFP IP), and nos‐Gal4 UASp‐Sxl‐EGFP (Sxl‐EGFP IP) ovaries was normalized against the corresponding amount of RNA in the 10% input, and the average enrichment (log₂ fold change) of each RNA in EGFP IP (open bars) and Sxl‐EGFP IP (hatched bars) in comparison with nos‐Gal4 immunoprecipitate is shown. Error bars indicate standard error of three biological replicates. Significance of the difference relative to EGFP IP was calculated by Student's t‐test (*p < 0.05). (□, EGFP IP; , Sxl‐EGFP IP)

To determine whether Sxl protein binds to NHP2‐RB 3′ UTR during cyst formation, we performed UV‐crosslinked RNA immunoprecipitation (CLIP). Extracts prepared from ovaries expressing Sxl‐EGFP and EGFP in the germline were immunoprecipitated with an anti‐GFP antibody, and NHP2‐RB 3′ UTR was detected by RT‐qPCR using a primer set amplifying the 3′ UTR region encompassing the Sxl‐binding motif. As a positive control, we examined the precipitation of nos mRNA, whose 3′ UTR contains a Sxl‐binding motif and binds Sxl protein (Chau, Kulnane, & Salz, 2012). NHP2‐RB and nos 3′ UTRs were more abundant in immunoprecipitate from Sxl‐EGFP‐expressing ovaries (Sxl‐EGFP IP) than from EGFP‐expressing ovaries (EGFP IP) (Figure 6b). By contrast, Tub84D and rp49, which contain no Sxl‐binding motif in their 3′ UTRs, were not enriched in Sxl‐EGFP IP in comparison with EGFP IP (Figure 6b). These results strongly suggest that Sxl protein binds to the NHP2‐RB 3′ UTR in the ovarian germline.

Sxl protein accumulates in GSCs (Chau et al. 2009), but its expression during cyst formation remains unclear. We found that Sxl protein was highly expressed in single cells and two‐ and four‐cell cysts (Supporting Information Figure S2a–a” and b–b”), but its expression was gradually reduced in eight‐cell cysts (Supporting Information Figure S2c–c”) and 16‐cell cysts (Supporting Information Figure S2a–a”, b–b” and c–c”). Thus, Sxl was expressed in four‐ and eight‐cell cysts, in which expression of NHP2 protein was reduced. By contrast, in single cells and two‐cell cysts, both Sxl and NHP2 were highly expressed. Therefore, we speculate that Sxl requires a corepressor to suppress NHP2 production in 4‐ and 8‐cell cysts, and that its expression is restricted in these cysts. Sxl acts with a corepressor to suppress translation of msl‐2 mRNA in the soma (Duncan, Strein, & Hentze, 2009; Duncan et al. 2006).

In Figure 6a, we show that NHP2‐RB constitutes approximately 30% of the total expression level of NHP2 mRNA in bam mutant ovaries. Although we cannot exclude the possibility that NHP2‐RB is a major variant of NHP2 mRNA in four‐cell and eight‐cell cysts, we speculate that Sxl‐dependent translational repression of the NHP2‐RB variant causes a reduction in NHP2 protein expression in four‐ and eight‐cell cysts. Our data suggest that this downregulation of NHP2 protein facilitates progression from two‐cell cysts to four‐ and eight‐cell cysts. If this is the case, one would expect that germline‐specific overexpression of NHP2 (NHP2 GS‐OE) causes a tumorous phenotype characterized by the accumulation of single cells and two‐cell cysts in the germarium, such as that observed in Sxl GS‐KD. Indeed, NHP2 GS‐OE caused a statistically significant but modest increase in the number of single cells and two‐cell cysts in the germarium (Supporting Information Figure S3). However, four‐ to 16‐cell cysts were still observed in the NHP2 GS‐OE germarium (Supporting Information Figure S3). These data suggest that NHP2 GS‐OE is insufficient to cause the tumorous phenotype. This can be explained by the idea that NHP2 requires co‐factors to execute its function in cyst formation. NHP2 co‐factor(s) may be active in single cells and two‐cell cysts, but may be inactivated by Sxl in four‐ and eight‐cell cysts. Thus, NHP2 GS‐KD is able to eliminate NHP2 function in single cells and two‐cell cysts, but NHP2 GS‐OE is unable to upregulate NHP2 function in four‐ and eight‐cell cysts.

3.5 Proper regulation of pseudouridylation is required for progression of cyst formation

The observations described above support the idea that progression to four‐ and eight‐cell cysts requires downregulation of NHP2 expression, and that NHP2 is suppressed at the translational level in four‐ and eight‐cell cysts by Sxl function.

The genes required for GSC differentiation are repressed by nanos (nos). Once the daughter cells of asymmetric GSC divisions detach from the niche, Sxl becomes active and represses Nos expression at a translational level, collaborating with Benign gonial cell neoplasia (Bgcn), Mei‐P26, and Bam proteins; expression of all of these factors is repressed by Dpp signaling from the niche (Chau et al. 2012; Chen & McKearin, 2003b; Li, Minor, Park, McKearin, & Maines, 2009; Li et al. 2013; Song et al. 2004). Thus, Sxl enables the switch from the GSC to differentiating state by suppressing nos translation (Slaidina & Lehmann, 2014). By analogy, we speculate that Sxl plays an additional role in progression of cyst formation by suppressing NHP2 production. Of course, we cannot rule out the alternative possibility that accumulation of Bam‐positive four‐ and eight‐cell cysts in NHP2 GS‐KD and NHP2‐Sxl GS‐KD germaria resulted from de‐repression of Bam in the stem cysts, which are separated from the niche. To address this issue, further work is necessary to examine whether four‐ and eight‐cell cysts have gene expression profiles similar to those of normally developing cysts.

It remains to be determined how NHP2 modulates cyst formation. Knockdowns of genes involved in ribosome assembly, including H/ACA snoRNP complex, impair GSC differentiation and cause formation of stem cysts. Moreover, knockdowns of genes encoding ribosomal subunits and rRNA transcription induce a GSC loss phenotype (Sanchez et al. 2016). Furthermore, four‐ and eight‐cell cysts accumulate in bam mutant germaria when mutations are introduced in either pelota (which encodes a translational release factor–like protein) or under‐developed (which encodes an RNA polymerase I cofactor) (Xi et al. 2005; Zhang et al. 2014). Thus, it is reasonable to assume that modulation of translational activity is essential for GSC differentiation and cyst formation. However, we found that global protein synthesis, monitored by O‐propargyl‐puromycin and L‐homopropargylglycine Click‐iT staining (OPP staining) (Liu, Xu, Stoleru, & Salic, 2012), did not decrease in four‐ and eight‐cell cysts, in which NHP2 expression was downregulated (Supporting Information Figure S4a–a” ;Sanchez et al. 2016), and NHP2 GS‐KD did not affect OPP staining in germarium (Supporting Information Figure S4b–b”). Thus, we speculate that NHP2‐dependent pseudouridylation, rather than global translation activity, must be precisely regulated to promote proper cyst formation.

Alteration in pseudouridylation of rRNA causes a defect in translation of specific mRNAs containing IRES (internal ribosome entry site) in yeast, mouse, and human (Gilbert, 2011; Jack et al. 2011; Xue & Barna, 2012; Yoon et al. 2006). Furthermore, in Xenopus embryos and human embryonic stem cells, a complex containing NHP2 and RNA Polymerase I induces neural crest specification by altering translation of specific mRNAs without affecting global translation activity (Werner et al. 2015). Therefore, it is possible that reduction in rRNA pseudouridylation alters translation of specific mRNAs, which in turn promotes progression of cyst formation. Further studies are needed to identify the mRNAs whose translation is altered by NHP2 GS‐KD, and to examine their roles in the progression of cyst formation.

ACKNOWLEDGMENTS

We thank Dr. Hiroshi Sakamoto for providing us with anti‐Sxl antibody. We also thank the Bloomington Drosophila Stock Center for providing us with fly stocks, Drosophila Genomic Resource Center for a cDNA clone, and the Developmental Studies Hybridoma Bank for antibodies. This work was supported in part by Grants‐in‐Aid for Scientific Research from the Japan Society for the Promotion of Science (JSPS) (KAKENHI Grant Numbers: 16J00758, 25114002, and 24247011). We thank members of the Kobayashi laboratory for valuable discussions.

AUTHOR CONTRIBUTIONS

S. M. and S. K. designed the experiments; S. M. and R. O. performed the experiments; and S. M. and S. K. wrote the paper. All authors reviewed the manuscript.