Autophagy is dispensable in germline stem cells but is required in the cap cells for their maintenance in the Drosophila ovarian niche

Summary

Autophagy is a cytoprotective mechanism responsible for the maintenance and long-term survival of various cell types, including stem cells. However, its role in the germline stem cell (GSC) niche remains unexplored. We demonstrate that autophagy flux in female Drosophila GSCs is low and dependent on the core autophagy gene, Atg5. However, the maintenance of Atg5^−/− GSCs within the GSC niche was unaffected even under nutrient stress. In contrast, disruption of autophagy within the cap cells (niche cells) leads to the loss of both cap cells and GSCs during aging. Further, reduced autophagy in cap cells severely impairs the crucial GSC self-renewal signal mediated by BMP-pMad emanating from the cap cells at the onset of midlife. Autophagy was essential for the long-term survival of cap cells. Our study reveals a differential role for autophagy, which is dispensable in GSCs but necessary in niche cells, where it supports signaling and survival to maintain GSCs.

Graphical abstract

Highlights

• •Autophagy is intrinsically low in GSCs and dispensable for maintenance under nutrient stress
• •Cap cell autophagy is essential for sustaining niche integrity and GSCs with age
• •Atg5-mediated autophagy preserves cap cell viability and BMP signaling during aging

Shravage and Nilangekar demonstrate that autophagy exerts a compartment-specific function within the Drosophila germline stem cell niche. Autophagy is intrinsically low and nonessential in GSCs but is required in cap cells, where Atg5-mediated autophagy sustains cellular viability and BMP signaling during aging. Autophagy impairment induces cap cell apoptosis and compromises GSC maintenance, highlighting its niche-specific role.

Introduction

Macroautophagy (autophagy hereafter) is a catabolic process that maintains cellular homeostasis by the sequestration and lysosome-mediated degradation of cytoplasmic material that is toxic or superfluous. Autophagy occurs constitutively at a basal level and is upregulated under stress conditions, including nutrient deprivation. Autophagy is an important factor that influences the aging and longevity of stem cells. The stem cell niche maintains the “stemness” of stem cells by activating pro-stemness cellular pathways and repressing differentiation. Stem cell niches replace dysfunctional cells within the tissues and organs but themselves age. However, the role of autophagy in the maintenance of stem cell niches throughout the lifespan is poorly understood.

Multiple Atg (autophagy-related) proteins form distinct complexes that regulate steps of autophagy (Klionsky et al., 2021). The Atg12-Atg5-Atg16 complex is essential for lipidation of Atg8 (LC3), which is crucial for the formation of mature autophagosomes. Knockdown (KD) of Atg5 blocks canonical autophagy (Chang et al., 2013; Klionsky et al., 2021; Rusten et al., 2004). Atg5^−/− mice do not survive postnatally, while germline-specific knockout (KO) of Atg5 disrupts sperm development and fertility in mice (Huang et al., 2021). Conditional KO of Atg5 in murine neural stem cells showed impaired autophagy as observed by reduced LC3 lipidation, accumulation of p62, and dysfunctional mitochondria (Wang et al., 2016). Additionally, mutations in ATG5 were found to cause ataxia in humans (Kim et al., 2016). Importantly, the ubiquitous overexpression of Atg5 in mice was sufficient to enhance autophagy levels and extend lifespan (Pyo et al., 2013). In Drosophila, basal autophagy is disrupted in both Atg5 KD and in Atg5 null mutants (Kim et al., 2016; Scott et al., 2004). Given its essentiality for canonical autophagy, Atg5 is an ideal candidate for investigating autophagy and its role in the GSC niche.

Autophagy is crucial in adult stem cells for their self-renewal, differentiation, metabolism, and aging (Adelipour et al., 2022; Chen et al., 2018; Shravage and Turksen, 2023). In aged muscle stem cells and hematopoietic stem cells (HSCs), autophagy maintains quiescence and regenerative function (García-Prat et al., 2016; Ho et al., 2017). Autophagy is impaired in both aged murine and human muscle stem cells. Loss of autophagy drives stem cell senescence, a phenotype that can also be induced in young stem cells upon its disruption. Defective autophagy compromises proteostasis, perturbs mitochondrial quality control, and increases oxidative stress, resulting in reduced stem cell function and number. Notably, restoring autophagy reverses senescence and rejuvenates regenerative capacity. Interestingly, about one-third of aged HSCs maintain high autophagy, low metabolism, and robust regenerative potential, underscoring autophagy as a key determinant of stem cell fitness during aging (Ho et al., 2017).

It remains unclear whether autophagy supports niche cell maintenance and, in turn, indirectly affects GSCs. Our earlier work showed that Atg1 maintains GSCs by regulating mitochondrial dynamics, independent of canonical autophagy (Ayachit and Shravage, 2023). In Drosophila, loss of autophagy in male cyst stem cells (CySCs), a key niche component, disrupts their maintenance and causes GSC loss (Sênos Demarco et al., 2020). Previous studies have shown that age-dependent niche cell loss and reduced niche stem cell signaling factors lead to impaired stem cell maintenance, renewal, and differentiation in various stem cell niches, including murine spermatogonial stem cells (GSCs) (Dulken et al., 2019; Enwere et al., 2004; Ryu et al., 2006; Schüler et al., 2021; Silva-Vargas et al., 2016). Age-related niche deterioration ultimately causes stem cell loss (reviewed in Brunet et al., 2022). Currently, the molecular role of autophagy in adult stem cell niche cells remains unknown.

Several Drosophila studies highlight the importance of autophagy in development, cell death, nutrient stress, and aging (reviewed in McPhee and Baehrecke, 2009; Mulakkal et al., 2014). The female Drosophila GSC niche has been an excellent model for studying stem cell biology, stem cell-niche interaction, and GSC-niche aging (Ishibashi et al., 2020; Xie and Spradling, 2000). The female GSC niche is located at the anterior tip of the germarium, the anterior region of the ovariole that contains developing egg chambers. It comprises the terminal filament, the cap, and the escort cells that constitute the niche architecture (Xie, 2013). Cap cells, the primary niche cells, secrete BMP ligands decapentaplegic (Dpp) and glass bottom boat (Gbb), essential for GSC self-renewal and anchor GSCs through E-cadherin-based adherens junctions (Xie, 2013). The GSC niche typically contains six to eight cap cells and two to three GSCs. BMP-pMad signaling maintains GSCs in an undifferentiated state. Previous studies report reduced BMP/Dpp signaling and E-cadherin-mediated adhesion in aged germaria, leading to loss of GSCs and cap cell function (Pan et al., 2007). Niche deterioration, along with intrinsic stem cell aging, serves as an extrinsic driver of GSC aging (Boyle et al., 2007; Pan et al., 2007; Wallenfang et al., 2006).

In this study, we demonstrate the role of autophagy within the GSC niche. Our data show that Atg5 is required for basal autophagy in both GSCs and cap cells; it is essential for the survival of the cap cells but not GSCs. We found that a lack of autophagy in cap cells leads to the progressive reduction of their numbers, subsequently leading to GSC loss during aging. We further demonstrate that the lack of autophagy in the cap cells affects niche-GSC signaling required for GSC maintenance and leads to increased cap cell death.

Results

Autophagy in GSCs is low and dispensable for GSC maintenance

We previously developed a germline-specific autophagy reporter, nosP-mCherry-Atg8a, expressing mCherry-tagged Atg8a under the nanos promoter (Nilangekar et al., 2019). Nutrient limitation strongly induces autophagy in the germarium (Barth et al., 2011; Hou et al., 2008; Nezis et al., 2009; Nilangekar et al., 2019). However, mCherry-Atg8a puncta in GSCs remain few and unchanged under starvation, or even pharmacological modulation of autophagy using rapamycin or chloroquine (CQ) compared to differentiating cysts in regions 2a and2b. (Nilangekar et al., 2019). We quantified autophagy flux in GSCs using nosP-mCherry-Atg8a flies, where mCherry-Atg8a marks autophagosomes and autolysosomes. After CQ treatment and cathepsin L immunostaining (which marks autolysosomes and lysosomes), the number of both autophagosomes and autolysosomes remained unchanged (Figure 1B). On average, one punctum per GSC was positive for both markers, indicating the presence of an autolysosome (Figures 1A and 1B). In contrast, CQ treatment increased the number of mCherry-Atg8a puncta in differentiating cells, suggesting increased autophagy (Nilangekar et al., 2019). These data confirm that basal autophagy and autophagy flux in GSCs are low.

Figure 1.

Autophagy in GSCs is low and dispensable for maintenance

(A) Representative image showing autophagic vesicles in GSCs. The GSC is marked by a dashed outline, and the cap cells are marked by asterisks. Puncta are indicated by colored arrowheads; mCherry-Atg8a/autophagosome (red), cathepsin L/lysosome (green), and colocalized punctum/autophagosome (yellow). Scale bar, 5 μm.

(B) Interleaved scatterplot showing the number of autophagic vesicles in control GSCs and CQ-treated GSCs. The blue line represents the average, and the error bars represent the SD. n = 20 GSCs per treatment.

(C) Representative image showing Ref(2)P in mosaic GSCs. Atg5-null (−/−) GSC is marked by a white dashed outline (RFP–), heterozygous (+/−) GSC is marked by a red dashed outline (RFP+), and cap cells are marked by asterisks. Scale bar, 10 μm. Interleaved scatterplot showing the number (D) and size (E) of Ref(2)P puncta in Atg5 mutant GSCs. The blue line represents the average, and the error bars represent the SD. Two independent experiments were performed. For the data presented, n = 19 control GSCs, which were heterozygous (+/−) or wild type (+/+) and n = 20 Atg5 mutant GSCs. (F) Representative image showing mosaic GSC clones. Atg5 null GSC (indicated as Atg5^−/−) with no visible RFP and neighboring RFP-positive heterozygous GSC (indicated as Atg5^+/−). Scale bar, 10 μm.

(G) Line graph showing the maintenance of Atg5-null GSCs compared to control GSC clones. Error bars represent SD. The data presented are aggregated from three independent biological replicates. n = 105 ± 6 (I), 80 ± 2 (II), and 70 ± 3 (III) germaria per genotype per time point for the three replicates (I, II, and III) (except III Atg5^5cc5 2 days n = 54).

(H) Line graph showing the retention of Atg5-null GSCs under complete starvation compared to control GSC clones. The data presented are aggregated from two independent biological replicates. n = 105 ± 3 germaria per genotype per condition for each replicate (I and II) (except I control starved n = 92, II Atg5^5cc5 3 days n = 87). ^∗∗∗∗p < 0.0001.

Additionally, we assessed autophagy flux under genetically induced conditions, where overexpression of core autophagy genes such as Atg1, Atg8a, or Atg5 has been shown to enhance autophagy activity (Bjedov et al., 2020; Pyo et al., 2013; Simonsen et al., 2008). Overexpression of Atg5 in the germ line using nosGal4VP16 > UASp-eGFP-Atg5 resulted in strong EGFP fluorescence in germline cells, including GSCs, confirming effective expression (Figure S1B). Further, we quantified autophagy flux in the presence or absence of CQ, in GSCs to determine whether Atg5 overexpression enhances autophagy. Despite Atg5 overexpression, autophagy flux in GSCs remained comparable to that of controls (Figures S1A and 1B).

As recommended by Klionsky et al. (2021), autophagy was evaluated using two independent markers (Klionsky et al., 2021). The autophagy receptor Ref(2)P (Refractory to Sigma P), a Drosophila homolog of p62/SQSTM1, was used to assess autophagic activity. Ref(2)P binds and delivers cargo to autophagosomes and is degraded along with it; hence its levels are inversely correlated with autophagy. Inhibition of autophagic degradation using agents such as CQ causes Ref(2)P accumulation, and its turnover in the presence or absence of such inhibitors serves as a reliable measure of autophagy flux (Klionsky et al., 2021). We used an anti-Ref(2)P antibody to detect Ref(2)P aggregates in germline cells from untreated and CQ-treated flies. Germaria from CQ-treated flies showed an increased number of Ref(2)P aggregates (Figure S1D); however, the number of Ref(2)P puncta within GSCs remained comparable between control and CQ-treated flies (Figure S1C). Furthermore, Atg5 overexpression did not alter the number of Ref(2)P puncta in GSCs, even after CQ treatment (Figure S1C). These results indicate that autophagy flux in GSCs is intrinsically low and is not enhanced by Atg5 overexpression.

We tested whether Atg5 is required for basal autophagy in GSCs by depleting Atg5 mRNA in the germline using nosGal4VP16-driven Atg5IR expression (Nilangekar et al., 2019). Atg5 KD significantly increased the number and size of Ref(2)P puncta in GSCs (Figures S1E–S1G). The null allele Atg5^5cc5, which lacks over 85% of the coding region including the start site, confirmed the requirement of Atg5 for basal autophagy (Kim et al., 2016). Using the FLP-FRT system, we generated Atg5^−/− GSC clones. Ref(2)P puncta were significantly increased in number and size in (Atg5^−/−) GSCs as compared to heterozygous (Atg5^−/+) or wild-type (Atg5^+/+) GSCs (Figures 1C–1E). In summary, These findings confirm that Atg5 is essential for Ref(2)P clearance, indicating that canonical autophagy operates in GSCs.

Our data show that Atg5 is required for basal autophagy in GSCs, consistent with reports linking autophagy to stem cell maintenance (Chen et al., 2018; García-Prat et al., 2016; Ho et al., 2017; B. V. Shravage and Turksen, 2023). We performed a GSC retention assay using Atg5 null mutants, comparing the persistence of Atg5^−/− GSC clones with control clones (Atg5^+/+) generated via the FLP-FRT system (isoFRT19A; RFP^−/− isogenized wild-type chromosome containing FRT19A) (Figures 1F and 1G). Surprisingly, Atg5-deficient and control GSCs exhibited comparable retention for up to 19 days, indicating that Atg5-dependent autophagy is dispensable for GSC maintenance (Figure 1G). Atg5 is required for autophagy in GSCs but not for their maintenance, indicating that autophagy is dispensable for GSC maintenance.

Autophagy supports survival under nutrient stress (Klionsky et al., 2021). To assess its role, flies with control and Atg5 mutant GSC clones were subjected to 4 days of complete starvation and compared with fed counterparts. Two-way ANOVA revealed comparable GSC numbers across fed and starved, control and mutant, and pre- and post-treatment groups (Figure 1H), indicating that starvation does not impair the maintenance of Atg5/autophagy-deficient GSCs.

Autophagy is required extrinsically in the cap cells for GSC maintenance

In the Drosophila male GSC niche, Demarco et al. reported that autophagy gene KD in CySCs significantly reduced GSC number (Sênos Demarco et al., 2020). We next aimed to establish and validate an “autophagy-defective niche” model. We targeted cap cells, the primary niche, using UAS-Gal4-driven RNAi to knock down Atg1, Atg13, Atg9, Atg5, Atg12, Atg16, and Atg8a, driven specifically by HhGal4 or Bab1Gal4. Autophagy disruption was verified by Ref(2)P aggregation, which accumulated in all Atg-RNAi cap cells except Atg12 and Atg13 KDs (Figures 2A–2D and S1H–S1K). Further, to assess autophagosome formation, we used the 3xmCherry-Atg8a reporter, which expresses mCherry-Atg8a under the endogenous Atg8a promoter. Atg5 KD cap cells showed markedly fewer mCherry-Atg8a puncta than controls (Figures 2B and 2E). Taken together, the autophagy-defective niche model was established and validated for Atg1, Atg9, Atg5, Atg16, and Atg8a.

Figure 2.

Autophagy in Cap cells is necessary for their and GSC maintenance

(A and A′) Representative image showing Ref(2)P in cap cells upon Atg5 KD.

(B and B′) Representative image showing the autophagosomes in Atg5 RNAi cap cells. The region of cap cells is marked by dashed outlines. Scale bars, 5 μm. Interleaved scatterplots showing the number (C) and size (D) of Ref(2)P puncta in cap cells with Atg5 KD. n = 14 ± 1 cap cell planes/area per genotype.

(E) Interleaved scatterplot showing the number of 3xmCherry-Atg8a puncta (autophagosomes) in Atg5 KD cap cells. n = 30 cap cell planes/area per genotype. Blue line represents the average, and error bars represent SD.

(F–L) (F and F′) Representative image showing the number of cap cells. Cap cells are marked by asterisks. Scale bars, 5 μm. Interleaved scatterplots showing the number of cap cells in old niches with Atg5 RNAi (G), Atg1 RNAi (H), and Atg9 RNAi (I) in the cap cells. Blue line represents the average, and error bars represent SD. Bar graphs showing the number of GSCs in old niches with Atg5 RNAi (J), Atg1 RNAi (K), and Atg9 RNAi (L) in the cap cells. Error bars represent SD. Sample sizes are the number of germaria as indicated in the graphs. ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001.

Previous work established that cap cells are critical for GSC maintenance. We examined whether autophagy in cap cells is required for non-autonomous GSC maintenance. To test this, we used autophagy-defective niche models with Atg1, Atg9, or Atg5 KD and quantified GSC retention in 7-week-old germaria. Notably, GSCs were significantly reduced in these niches compared with age-matched controls (Figures 2J–2L). Hence, autophagy in cap cells is required for GSC maintenance during aging. Previous studies have shown that the number of GSCs correlates with the number of cap cells (≈1 GSC per 2.5 cap cells) (Ward et al., 2006; Xie and Spradling, 2000). As aged autophagy-defective niches contained fewer GSCs than the controls, we quantified the numbers of cap cells in these niches. Notably, autophagy-defective niches contained significantly fewer cap cells than age-matched controls (Figures 2F–2I). Because targeting multiple Atgs of different complexes yielded identical outcomes, the observed accelerated loss of GSC niche cells during aging is attributable to disrupted autophagy in cap cells. Thus, autophagy is required to sustain cap cells, which are essential for preserving GSCs.

As HhGal4 and Bab1Gal4 are active during development and gonad morphogenesis, we examined whether the observed effects occurred prior to the adult ovary stage. To assess this, cap cells and GSCs were quantified at 12 days post-eclosion, and in a separate, higher-temporal resolution experiment, as early as 7 days post-eclosion. The average cap cell and GSC numbers across experimental conditions were comparable to controls at early ages (Figures S2A, S2C, and S2D). These data indicate that the cap cell-to-GSC ratio was similar in controls and Atg5 KD niches, confirming that GSC-niche morphogenesis was unaffected during development. Thus, the accelerated loss of niche cells resulted from the absence of autophagy during aging.

Lack of autophagy affects cap cell function and causes cap cell death during aging

In female Drosophila, GSC self-renewal depends on secreted and membrane-bound cues. Cap cells secrete BMP ligands Dpp and Gbb, which drive GSC self-renewal and division. Loss of BMP signaling activates differentiation programs, causing GSC depletion (Xie and Spradling, 1998). Phosphorylated Mad (pMad), the key BMP/Dpp signaling effector, serves as a reliable marker of cap cell-GSC signaling strength. We quantified pMad levels in GSCs from autophagy-deficient niches across ages and observed a marked reduction from day 20 onward compared with controls (Figures 3A–3F). The same phenotype was reproduced using the cap cell-specific driver Bab1Gal4 (Figure S2E). Thus, autophagy impairment in cap cells disrupts BMP-pMad signaling in the GSC niche during mid to late aging.

Figure 3.

Lack of autophagy in cap cells affects their function and causes cell death during aging

(A–E) Interleaved scatterplots showing pMad intensity in GSCs of the mentioned genotypes at the five mentioned time points. Error bars represent SD. n = 20 ± 2 GSCs per genotype per time point.

(F and F′) Representative image showing pMad in GSCs at 30 days. Scale bars, 5 μm.

(G–G‴) Representative images showing a TUNEL-positive cap cell in the GSC niche. Scale bar, 5 μm.

(H) Line graph showing the number of cap cells across age. Error bars represent the SEM.

(I) Column graph showing percent germaria with TUNEL-positive cap cells across age in the mentioned genotypes (^∗∗p < 0.01, Z score = −2.76). n = 80 germaria per genotype per time point for (H) and (I). ^∗∗p < 0.01, ^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001.

In autophagy-defective niches, cap cells were lost rapidly with age (Figures 3H, S2A and S2C). We tested whether cap cell loss resulted from apoptosis using the TUNEL assay, which labels double-stranded DNA breaks characteristic of programmed cell death. TUNEL-positive cap cells were quantified in control and autophagy-defective niches across ages. Cap cell death was absent across all experimental conditions until 20 days. By 40 days, TUNEL-positive cap cells appeared in both control and autophagy-defective niches. At 50 days, however, Atg5 KD niches showed a significantly higher proportion of dying cap cells than age-matched controls (Figures 3G–3I and S2B). These results suggest that loss of autophagy triggers apoptosis in aged cap cells, indicating that autophagy acts as a survival factor during aging.

Discussion

Our study reveals a differential role for autophagy in the Drosophila ovarian GSC niche. We show that autophagy is not required intrinsically in GSCs but is essential in niche cells, where it acts extrinsically to maintain GSCs. While Atg5 mediated Ref(2)P clearance in GSCs, Atg5-null GSCs were maintained despite complete loss of autophagy, even during starvation. Atg5-mediated autophagy was required for Ref(2)P clearance within the cap cells and, in contrast to its role in GSCs, was found to be critical for the survival of the cap cells. The autophagy-defective cap cells are compromised in function and thereby unable to maintain GSCs during aging. Disrupted autophagy drives GSC-niche decline, marked by cap cell apoptosis in old age and reduced niche-GSC BMP signaling from mid-life onward.

Atg5 is crucial for the self-renewal and differentiation of diverse adult stem cell types (Chen et al., 2018). Autophagy is widely recognized as essential for stem cell maintenance; thus, we expected it to be required for GSC maintenance. But surprisingly, in female Drosophila GSCs, autophagy is dispensable. Previously, we suggested that GSCs may be protected from starvation-induced autophagy (Nilangekar et al., 2019). This study used a comprehensive strategy to measure basal autophagy flux in GSCs. Autophagic vesicles were quantified across the full GSC volume, and both inhibition (CQ) and induction (Atg5 overexpression) conditions were applied according to guidelines (Klionsky et al., 2021). Multiple markers, including nosP-mCherry-Atg8a, cathepsin-L, and Ref(2)P, enabled rigorous assessment (Klionsky et al., 2021). We demonstrated disruption of autophagy in Atg5 mutant GSCs. Zhao et al. similarly showed that autophagy is low in GSCs and that their maintenance is unaffected in Atg6 and Fip200 mutants (Zhao et al., 2018). Demarco et al. also found that KD of multiple autophagy genes (Atg1, Atg5, Atg6, Atg7, and Atg8a/8b) in male GSCs did not alter their numbers at 10 days post-eclosion (Sênos Demarco et al., 2020). Thus, autophagy is dispensable in both male and female GSCs.

Nutrition shapes Drosophila oogenesis. Drummond-Barbosa and Spradling reported that GSC numbers are unaffected by diet, whereas their proliferation and differentiation into eggs are strongly influenced (Drummond-Barbosa and Spradling, 2001). Our analyses recapitulate that poor nutrition does not affect GSC numbers, even under conditions of complete starvation. Importantly, even Atg5/autophagy-deficient GSCs were resistant to GSC loss under severe nutrient stress, indicating their resilience, which may be crucial in the wild. This contrasts with the fact that starvation induces autophagy and autophagy-mediated cell death in the differentiated cells in the germarium (Barth et al., 2011; Hou et al., 2008; Nezis et al., 2009; Nilangekar et al., 2019). Taken together, this indicates differential regulation and strict repression of autophagy in the GSCs.

Autophagy in GSCs appears to be under tight control, since eGFP-Atg5 overexpression induced puncta in differentiating cells but not in GSCs (Figure S1B). eGFP-Atg5 puncta localize to phagophore assembly sites, isolation membranes, and immature autophagosomes; therefore, this observation also corroborates that autophagy in GSCs is low. Alternatively, this could be due to a lack of other core autophagy proteins required to form these intermediates. By contrast, murine Atg5 overexpression and neuronal Atg8a overexpression in Drosophila have each been shown to elevate autophagy (Pyo et al., 2013; Simonsen et al., 2008). Thus, autophagy in GSCs may be tightly and differently regulated compared to the differentiating germline cells.

The regulation of autophagy in GSCs is intriguing, as adjacent cysts exhibit moderate to high levels. Zhao and colleagues speculated that low autophagy in the GSCs may be explained by activation of BMP signaling (Zhao et al., 2018). Consistent with this view, Varga and colleagues demonstrated that BMP signaling suppresses autophagy in male Drosophila GSCs (Varga et al., 2022). As BMP-pMad signaling sustains female GSCs, our data imply that the self-renewal machinery actively inhibits autophagy in these cells. Elevated autophagy in cysts suggests that differentiation factors drive its upregulation, a view supported by Varga et al. (2022). Further, defective mitochondria in the female germ line undergo clearance via Atg1-dependent mitophagy (Ayachit and Shravage, 2023; Lieber et al., 2019; Palozzi et al., 2022). Insulin/InR-mTOR signaling, known to regulate GSC growth and division, may also control autophagy, either independently or via BMP signaling.

Although autophagy contributes to stem cell maintenance, a few studies demonstrate that Atg5 deletion did not affect stem cell maintenance. In mice, Atg12 and Atg5 conditional KO HSCs persisted in normal numbers and Atg5 loss did not impair their maintenance (Ho et al., 2017). In case of murine neural stem cells, deletion of Atg5, Atg16L1, Atg7, or FIP200 all showed defective autophagy and accumulation of mitochondria, but only FIP200 deletion affected neural stem cell maintenance (Wang et al., 2016). In mice, Atg5 deletion did not compromise the maintenance of HSCs or neural stem cells, raising the yet unresolved question of why Atg5 is indispensable for autophagy in these cells but dispensable for their maintenance.

According to Demarco et al., autophagy supports male GSC maintenance non-autonomously via CySCs in the niche (Sênos Demarco et al., 2020). Similarly, in females, GSC maintenance depends on the niche, specifically the cap cells. The dichotomy between soma and germ line makes the ovarian tissue interesting. Autophagy is regulated differently in these cells and has distinct functions. For instance, Barth et al. demonstrated that Atg1 and Atg13 are required in follicle cells but not in germ cells for egg development (Barth et al., 2011). We observe a similar phenomenon in the GSC niche; autophagy is required in the somatic cells, i.e., cap cells, and is dispensable in GSCs. Another contrast between germ line and soma in the germarium arises from GSCs and follicle stem cells. Our findings, together with prior studies, indicate that autophagy is dispensable in GSCs, whereas follicle stem cells of somatic origin are strongly influenced by it. In these cells, autophagy increases with age and contributes to their loss (Singh et al., 2018). In summary, autophagy has distinct functions in the female Drosophila germ line and GSCs.

Cap cells are terminally differentiated, and Pan et al. reported their numbers remain largely stable throughout the fly lifespan (Pan et al., 2007). Significant cap cell loss does not occur until 7 weeks; even so, the loss is less than 20%. Autophagy supports the longevity of terminally differentiated post-mitotic cells such as neurons (Simonsen et al., 2008). In our study, despite autophagy disruption, cap cell death was only evident from 40 days onward, possibly due to incomplete Atg5 KD by RNAi, although severe Ref(2)P accumulation was observed. Apoptosis and autophagy both contribute to oogenesis cell death in a cell type-, stage-, and context-dependent manner (Hou et al., 2008; Nezis et al., 2009; Pritchett et al., 2009). Autophagy acts upstream of apoptosis in early oogenesis (Nezis et al., 2009). Importantly, while cap cell death began at 40 days, functional deterioration was evident by 20 days, suggesting loss of homeostasis. We did not detect niche disruption during early life, possibly reflecting age-dependent physiological changes in Drosophila that emerge post-midlife (Aparicio et al., 2019; Rana et al., 2017). Alternate stress-mitigating pathways may also operate at earlier stages and in younger adults.

Although there is a significant and thorough understanding of the role of autophagy in stem cells as an intrinsic factor, there are very few reports connecting autophagy with the stem cell niche. Recent work demonstrated that sensory nerves activate autophagy in mesenchymal stem cells via fibroblast growth factor 1, a pathway critical for their maintenance (Pei et al., 2023). Emerging evidence, summarized in a recent review, points to systemic autophagy as a protective mechanism for vascular niche homeostasis (Dergilev et al., 2024).

Our finding that GSC maintenance is autophagy independent is baffling, especially considering the nature of these cells. The first level of distinction is that they are stem cells, and uniquely among the type of stem cells, they are GSCs. This is the first study that describes a novel role of autophagy in niche cells for stem cell maintenance. This study exemplifies a model to study the effect of autophagy-defective niches on the stem cells that they harbor. Atg5 is conserved in metazoans and plays an important role in autophagosome formation and closure. Thus, these findings could have critical implications for understanding the role of autophagy in niche-regulated stem cells, including in diseases such as cancer. Our findings add to the understanding of the fundamental role of autophagy in stem cell maintenance and where it has a non-cell-autonomous effect.

Limitations of the study

The influence of cap cells on GSC maintenance via autophagy likely involves pMad (BMP) signaling and cell survival, although the precise mechanism remains unresolved. Our findings suggest that loss of autophagy in cap cells disrupts homeostasis, which may lead to dysregulated transcription/translation and altered BMP ligand trafficking, resulting in reduced pMad levels. Similar effects may extend to adhesion factors such as E-cadherin. Beyond BMP, multiple pathways—insulin, Jak-Stat, Notch, and Hedgehog—also function in cap cells to support GSCs (Xie, 2013), raising the need to assess their regulation under autophagy loss. Variability in autophagy-dependent phenotypes may stem from incomplete Atg5 KD due to the UAS-Gal4 system, which could be addressed by mosaic analysis, although generating clones in adult cap cells is technically challenging. Alternatively, Atg mutant clones could be induced during niche specification. Another caveat lies in Gal4 drivers: Bab1-Gal4 and Hh-Gal4 mark the niche but are not cap cell specific, also targeting terminal filament cells. While these cells are important during niche formation, their role in the adult ovary is largely redundant, as cap cells provide key factors such as Piwi, Yb, and Dpp (Xie, 2013). Thus, cap cells remain the indispensable niche component, and taken together, our results indicate that autophagy-deficient niche cells compromise GSC maintenance primarily through cap cell dysfunction.

Methods

Genotypes of all fly stocks

+; +; + (OregonR) |w¹¹¹⁸; +; + |w; +; nosGal4VP16 (BL-4937) |w; nosP-mCherry-Atg8a; + |w; nosP-mCherry-Atg8a/CyO; nosGal4VP16/TM6b, Hu |yw; +; UASp-eGFP-drAtg5 (derived from BL-59848) |ysc^∗ v; Atg1 RNAi; + (BL-44034) |ysc^∗ v; +; Atg5 RNAi (BL-34899) |ysc^∗ v; Atg8a RNAi; + (BL-58309) |ysc^∗ v; +; Atg9 RNAi (BL-34901) |ysc^∗ v; Atg13 RNAi; + (BL-40861) |ysc^∗ v; +; Atg16 RNAi (BL-34358) |w; If/CyO; HhGal4, UAS-GFP/TM6B, Tb |w; +; HhGal4, UAS-GFP/TM6B, Tb |w; +; Bab1Gal4/TM6B (BL-6803) |yw; 3xmCherry-Atg8a; + |w; 3xmCherry-Atg8a/CyO; HhGal4, UAS-GFP/TM6B |y w⁺ Atg5^5cc5 FRT19A/FM7i; +; + |y w iso FRT19A; +; + |w hsFLP12, Ubi-RFP FRT19A; +; + (BL-31418) |y w hsFLP12, His2Av.GFP FRT19A/FM7a; +; + (BL-32045) |.

Aging

Large crosses were set to collect a large number of age-synchronized progeny of the desired genotype, which were collected within 24–48 h of eclosion. Until dissection the flies were housed under standard conditions as 15 females and 7–10 males in a vial, which were flipped every 3 to 4 days to fresh food vials supplemented with dry yeast pellets. The vials were kept horizontal throughout the duration in order to avoid death of flies due to sticking onto the food surface. Subsets of the collected and aged flies were dissected at each of the time points.

Immunostaining and TUNEL staining

The immunostaining procedure was followed as published previously (Nilangekar et al., 2019). In Situ Cell Death Detection Kit, TMR red (Roche, 12156792910) was used for TUNEL staining with a modified procedure. The details of immunostaining and TUNEL staining are elaborated in the supplementary text.

Confocal microscopy

The imaging was performed on the Leica SP8 confocal microscope or the confocal mode of the Zeiss LSM900 Airyscan microscope. Detailed settings and parameters are elaborated in the supplementary text.

Statistical analysis

For all comparative analyses, Student’s t test assuming unequal variance was used. To statistically compare the effect of fed versus starved between control versus Atg5 mutant GSC clones, two-way ANOVA was used. For statistical comparison of the proportion of TUNEL positive cap cells out of total cap cells among different genotypes, Z test for two proportions was used. Microsoft Excel was used to store, arrange, and analyze data. GraphPad Prism was used for plotting all the graphs.

Resource availability

Lead contact

Further information and requests for resources should be directed to and will be fulfilled by the lead contact, Bhupendra V Shravage (bvshravage@aripune.org, bhupendra.shravage@gmail.com).

Materials availability

No new materials or reagents were generated in the study.

Data and code availability

All datasets supporting the findings of this study are available from the corresponding author upon reasonable request.

Acknowledgments

We thank Dr. Gábor Juhász for providing the Atg5 mutant and 3xmCherry-Atg8a fly stocks, Dr. Manish Jaiswal for providing the FRT19A stocks, and Dr. Richa Rikhy and the IISER-Pune fly facility for providing many essential fly stocks. We thank Agharkar Research Institute, Pune, India, and the Developmental Biology fraternity for support and confocal facility access. This work was supported by DST-SERB grant number ECR/2015/000239, BT/RLF/Re-entry/58/2013, and BT/PR12718/MED/31/298/2015 to B.V.S. K.S.N. was supported by ICMR-SRF 2020-6879/CMB-BMS and was a registered Ph.D. student affiliated with the Department of Biotechnology, Savitribai Phule Pune University, Pune, India (registration no. 175, PGS/4204). B.V.S. is affiliated to Savitribai Phule Pune University, Pune, India, and is recognized by SPPU as a PhD guide in biotechnology and zoology.

Author contributions

K.S.N. and B.V.S. conceived the project and designed the experiments. K.S.N. conducted all experiments and analysis. K.S.N. and B.V.S. wrote the manuscript.

Declaration of interests

The authors declare no conflicts of interest.

Published: November 13, 2025