Introduction

In animals, mitochondrial inheritance occurs exclusively through the mothers’ oocytes. Quality mitochondria not only improve offspring survival and health but are also important for maintaining oocyte competence. In Drosophila, the quality and dynamics of the mitochondria are regulated throughout oogenesis and heavily influenced by nutrition and age. During early oogenesis in the germarium, replication of mitochondrial DNA occurs, and mitochondria are fragmented and selected for by a process called mitophagy1,2,3,4. As oogenesis progresses, mitochondria enter the oocyte by moving along the fusome (an Actin-rich structure that connects the germline cells in a cyst) where they form the Balbiani body5. The mitochondria are then trafficked towards the oocyte poleplasm during stage 10 of oogenesis to ensure inheritance of mitochondria by the offspring primordial germ cells6. Towards late oogenesis, mitochondria also undergo active remodeling to enter respiratory quiescence for subsequent storage in the mature oocyte7,8. Starvation is known to induce mitochondrial elongation and clumping in the nurse cells, while aging causes mitochondrial fragmentation in germline stem cells (GSCs)9,10.

Mitochondrial clumping is a general sign of unhealthy mitochondria. Nutrient deprivation, induction of oxidative damage10,11 and mutations of genes that are involved in mitochondrial related functions, such as clueless (clu), will result in mitochondrial clumping12,13,14, suggesting that accumulation of mitochondrial clumps is a sign of mitochondrial damage.

Stable intronic sequence (sis)RNAs are intron-containing long noncoding (lnc)RNAs that are not rapidly degraded in cells15,16. First discovered in Xenopus troplicalis oocytes, sisRNAs were later found to be conserved from yeast to human17,18,19,20,21,22,23,24,25. Studies in Drosophila have revealed that sisRNAs function to regulate various processes during oogenesis and embryonic development20,21,22,26,27,28,29,30. Previous studies have focused on the roles of sisRNAs in regulating gene expression20,31, however, it is unknown whether sisRNAs exhibit other modes of actions such as regulating organelle dynamics.

It was previously reported that the abundance of several sisRNAs was upregulated in the ovaries during fasting (or starvation) in Drosophila26,27,28,32. Upregulation of 4 such sisRNAs (sisR-1 to sisR-4) appeared to inhibit the production of mature oocytes by blocking the oogenesis progression of stage 8/9 egg chambers, which is known to be highly sensitive to nutrient availability. The molecular and cellular mechanisms of how sisRNAs are regulated and how they modulate oogenesis during fasting are not known.

Here we show that sisR-1 is upregulated and localizes to mitochondrial clumps in Drosophila nurse cells during fasting. Knocking down sisR-1 or refeeding promotes the clearance of these fasting-induced mitochondrial clumps, thereby improving oocyte quality. The abundance of sisR-1 is regulated by nuclear DIP1 and cytoplasmic Clueless (Clu), and the expression level or localization of both are sensitive to nutrient availability. Mechanistically, sisR-1 inhibits defective mitochondrial clearance (or mitophagy) by suppressing Parkin (Park)-mediated polyubiquitination of VDAC1/Porin. Thus, our study reveals a nutrient-sensitive pathway consisting of DIP1, Clu and sisR-1 that modulates mitophagy during oogenesis in Drosophila.

Results

Fasting promotes sisR-1-mediated mitochondrial clumping

We first examined the effects of sisR-1 knockdown (hereafter sisR-1 RNAi) on the production of stage 14 oocytes (or mature oocytes) under fed and fasted conditions (Fig. 1A). Newly eclosed female flies were fed or fasted for 2 days before the ovaries were dissected to count the number of stage 14 oocytes present (see Methods). Under fed conditions, the stage 14 oocyte count was lower in sisR-1 RNAi flies (2 independent RNAi lines) than the controls, a phenotype that is likely explained by a reduction in GSCs in sisR-1 RNAi flies as reported previously (Fig. 1B)29. Consistent with what was observed previously28, we also observed an increase in the number of stage 14 oocytes produced under fasted conditions in sisR-1 RNAi flies (Fig. 1B), suggesting that sisR-1 inhibits the production of mature oocytes during fasting. We further verified the expression of sisR-1 in the different genotypes under both fed and fasted conditions by performing single molecule fluorescent in situ hybridization (smFISH) (Supplementary Fig. S1).

Fig. 1: sisR-1 regulates mitochondrial clumping during fasting.
figure 1

A Germline cells at different stages of oogenesis. G: germarium, S: stage. B Number of stage 14 oocytes per ovary in fed and fasted y w, sisR-1 RNAi line 1 and 2. Data are presented as mean values +/−SD. *p < 0.001. n = 30 ovaries. Two-tailed t-test. C Relative levels of sisR-1 normalized against actin5C in y w fed versus fasted oocytes. Data are presented as mean values +/−SD from three biological replicates. *p = 0.01. Two-tailed t-test. D Relative levels of sisR-1 normalized against actin5C in control sibling versus nos-Gal4 > TOR RNAi ovaries. Data are presented as mean values +/−SD from n = 3 biological replicates. *p < 0.01. Two-tailed t-test. E Confocal images of stage 8/9 egg chambers showing localization of sisR-1 (Green), ATP5α (Red) and DAPI (Blue) in Oregon R fed with yeast paste for 2 days and fasted for 1 day post eclosure. Boxed region is magnified. Arrowheads point to mitochondrial clumps that colocalize with sisR-1 in fasted females. Scale bar = 20 µm. F Percentages of cytoplasmic sisR-1 dots that colocalize with mitochondrial clumps. Data are presented as mean values +/−SD. ***p < 0.001. n = 5 (fed), 8 (fasted) slices. Two-tailed t-test. G Mitochondria isolation showing enrichment of sisR-1 in mitochondrial fractions. ATP5α and β-Tubulin were used as positive controls. N = 2 biological replicates. H Confocal images of stage 8/9 egg chambers showing ATP5α localization in fasted Oregon R, sisR-1 RNAi line 1 and 2. Arrowheads point to mitochondrial clumps in fasted Oregon R. Scale bar = 20 µm. I Percentages of total stage 8/9 egg chambers with mitochondrial clumps in Oregon R and sisR-1 RNAi line 1 and 2 (n = 100 egg chambers) *p < 0.001. Chi-square test, one-sided.  J Number of clumps per egg chamber present in stage 8/9 egg chamber that contain clumps. Data are presented as mean values +/−SD. *p < 0.01. n = 9 (Oregon R), 7 (RNAi #1), 9 (RNAi #2) egg chambers. Two-tailed t-test. K Proposed model on how fasting affects sisR-1 levels and causes mitochondrial clumping. Source data are provided as a Source Data file.

Also consistent with previous report28, sisR-1 was found to be upregulated in the ovaries of fasted and TOR RNAi flies (Fig. 1C,D). To understand the cellular mechanism by which sisR-1 regulates oogenesis during fasting, we examined the localization of sisR-1 by performing smFISH. Under fed conditions, sisR-1 levels remained low and localized predominantly in the nuclei of nurse cells (Fig. 1E). During fasting, an accumulation of sisR-1 and appearance of mitochondrial clumps could be seen in the nurse cell cytoplasm (Fig. 1E). These mitochondrial clumps were reminiscent of those reported in previous studies10,11,12,13,14 and localized perinuclearly with width ranging from ~2 µm to ~8 µm (Supplementary Fig. S2). Interestingly, sisR-1 transcripts were often seen localizing near and inside the mitochondrial clumps (Fig. 1E, F, arrowheads). To confirm the association of sisR-1 with mitochondria, we performed mitochondrial fractionation and found an increase in the abundance of sisR-1 transcripts in the mitochondria after fasting (Fig. 1G).

Because mitochondrial clumping is often an indication of unhealthy mitochondria10,11,12,13,14, we asked if sisR-1 is required for the production of these mitochondrial clumps during fasting. Indeed, in sisR-1 RNAi flies, the percentages of stage 8/9 egg chambers containing mitochondrial clumps were reduced (Fig. 1H,I, arrowheads). Furthermore, when we examined egg chambers that still contained some clumping, the overall number of clumps per egg chamber was also drastically reduced (Fig. 1J). Next, we asked if overexpression of sisR-1 alone is sufficient to induce clumping of mitochondria without fasting. In support of this idea, we observed that vasa-Gal4 driven overexpression of sisR-1 led to the formation of mitochondrial clumps in germline cells (Supplementary Fig. S3). Thus, the upregulation of sisR-1 is at least partially responsible for the formation of mitochondrial clumps during fasting (Fig. 1K).

Clueless represses sisR-1 abundance in the cytoplasm

Since sisR-1 expression is upregulated during fasting (Fig. 1C)28, we moved to investigate the mechanism that represses sisR-1 expression under fed conditions. During fasting, sisR-1 localizes to mitochondria and promotes clumping. We explored the possibility of a nutrient-sensitive mitochondria-associated RNA-binding protein as a regulator of sisR-1. Drosophila Clu is a conserved RNA-binding protein that localizes to the surface of mitochondria as discrete cytoplasmic bodies in nurse cells14,33. Under nutrient deprivation, Clu bodies disassemble, concomitant with the formation of mitochondrial clumps34. Similar clumps were also observed in fed clu mutants, suggesting that clu suppresses mitochondrial clumping12,14.

We hypothesized that during fasting, Clu bodies disassemble, leading to the accumulation of mitochondrial sisR-1, which causes clumping. In clu mutants, sisR-1 expression was upregulated as measured by RT-qPCR (Supplementary Fig. S4A). clu mutant egg chambers also exhibited cytoplasmic sisR-1 that associates with mitochondrial clumps (Supplementary Fig. S4B, arrowheads), which is a phenotype also seen in fasted wildtype ovaries (Fig. 1E).

Previous studies showed that clu is important to facilitate translation of mitochondrial proteins12. To circumvent this issue, we observed clu heterozygous mutants instead, which had normal translation12. Heterozygous clu mutants also showed an upregulation of sisR-1 expression and the colocalization of sisR-1 with mitochondrial clumps (Fig. 2A,B). To test whether sisR-1 mediates mitochondrial clumping in clu heterozygous mutants, we knocked down sisR-1 to see if it could rescue the clumping. Consistent with the model, we saw a reduction in the percentages of egg chambers containing mitochondrial clumps as well as the overall number of clumps (Fig. 2C-E).

Fig. 2: clueless regulates sisR-1 expression.
figure 2

A Relative expression of sisR-1 in Oregon R and clu[d08713] heterozygous stage 14 oocytes. Data are presented as mean values +/−SD from three biological replicates. *p < 0.05. Two-tailed t-test. B Confocal images of stage 8/9 egg chambers showing localization of sisR-1 (Green), ATP5α (Red) and DAPI (Blue) in fed Oregon R and clu[d08713] heterozygous. Arrowheads point to colocalization of sisR-1 and mitochondrial clumps. Scale bar = 10 µm. C Confocal images of stage 8/9 egg chambers showing localization of ATP5α in clu[d08713]/+, clu[d08713]/+;sisR-1 RNAi line 1 and 2. Arrowheads point to mitochondrial clumps. Scale bar = 20 µm. D Percentages of stage 8/9 egg chambers with mitochondrial clumps in clu[d08713]/+, clu[d08713]/+;sisR-1 RNAi line 1 and 2 (n = 100 egg chambers) *p < 0.001. Chi-square test, one-sided. E Number of mitochondrial clumps present in stage 8/9 egg chambers that contain clumps in clu[d08713]/+, clu[d08713]/+;sisR-1 RNAi line 1 and 2. Data are presented as mean values +/−SD. **p < 0.05. n = 54 (clu/+), 25 (clu/+;RNAi #1), 18 (clu/+;RNAi #2) egg chambers. Two-tailed t-test. F Enrichment of sisR-1, but not U85, in Clu-GFP immunoprecipitate compared to IgG control. G Relative levels of sisR-1 normalized to Rp49 after alpha-amanitin treatment in Oregon R and clu ovaries. Data are presented as mean values +/−SD from three biological replicates. *p < 0.05. Two-tailed t-test. H Model on the effects of fasting on Clu protein and sisR-1 in promoting mitochondrial clump formation. Source data are provided as a Source Data file.

We next investigated if Clu directly regulates the stability of sisR-1. First, by performing immunoprecipitation of endogenous GFP tagged Clu using a protein trap line, we found physical interaction between Clu and sisR-1 in the ovaries (Fig. 2F). As a negative control, Clu did not interact with U85 RNA (Fig. 2F). Next, alpha-amanitin assay showed that sisR-1 was more stable in clu mutant ovaries compared to controls (Fig. 2G). On the other hand, the stability of a positive control Arglu1 pre-mRNA was similar in both cases (Supplementary Fig. S4C). Taken together, our data suggest that Clu negatively regulates the levels of sisR-1 by promoting its decay.

DIP1 represses sisR-1 abundance in the nucleus

DIP1 is a nuclear RNA binding protein that binds and promotes sisR-1 degradation in Drosophila29,35. To investigate if DIP1 is also involved in the regulation of sisR-1 during fasting, we first examined the expression of DIP1 in ovaries of fed and fasted flies. RT-qPCR and western blotting revealed that although the levels of DIP1 mRNA were similar, DIP1 protein abundance was downregulated in ovaries of fasted flies (Fig. 3A,B). Intriguingly, fasting also abolished the nuclear localization of DIP1 protein in those nurse cells (Fig. 3C), suggesting that the downregulation of DIP1 may lead to upregulation of sisR-1 during fasting.

Fig. 3: DIP1 regulates sisR-1 expression during fasting.
figure 3

A Relative expression of DIP1 normalized to actin5C in fed and fasted Oregon R stage 14 oocytes. Error bars depict SD from two biological replicates. ns: not significant, p > 0.05. Two-tailed t-test. B Levels of DIP1 protein in ovaries of fed and fasted Oregon R flies. Graph shows the quantified relative levels of DIP1 normalized to Actin. *p < 0.05. Data are presented as mean values +/−SD from three biological replicates. Two-tailed t-test. C Confocal images of stage 8/9 egg chambers showing localization of DIP1 (Green) and DAPI (Blue) in fed and fasted Oregon R. Scale bar = 20 µm. D Confocal images of stage 8/9 egg chambers showing localization of sisR-1 (Green), ATP5α (Red) and DAPI (Blue) in fed Oregon R, DIP1−/−, clu−/− and DIP1−/−;clu−/−. Scale bar = 10 µm. E Percentages of nurse cell per stage 8/9 egg chamber with mitochondrial clumps in fed Oregon R, DIP1−/−, clu−/− and DIP1−/−;clu −/−. Data are presented as mean values +/−SD. ***p < 0.001. n = 9 (Oregon R), 7 (DIP1−/−), 4 (clu−/−) and 9 (DIP1−/−;clu −/−) egg chambers counted. Two-tailed t-test. F Relative expression of sisR-1 by smFISH in the nucleus and cytoplasm of stage 8/9 egg chambers in fed Oregon R, DIP1 −/−, clu−/− and DIP1−/−;clu −/−. Data are presented as mean values +/−SD. ***p < 0.001. n = 9 (N), 10 (C) for Oregon R, 10 (N) and 13 (C) for DIP1−/−, 8 (N) and 10 (C) for clu−/− and 6 (N) and 5 (C) for DIP1−/−;clu−/−. egg chambers. Two-tailed t-test. G Relative levels of sisR-1 in fed Oregon R, DIP1−/−, clu−/− and DIP1−/−;clu−/− oocytes. Data are presented as mean values +/−SD from three biological replicates. *p < 0.05 **p < 0.01. Two-tailed t-test. H Images of dissected ovary pairs of fed Oregon R, DIP1−/−, clu−/− and DIP1−/−;clu−/− taken at 1.5X zoom. I Proposed model on the effects of fasting on DIP1, Clu and sisR-1 in promoting mitochondrial clump formation. Source data are provided as a Source Data file.

When we looked at the localizations of sisR-1 and mitochondria in DIP1 mutants, we did not observe any obvious differences when compared to Oregon R controls (Fig. 3D, Supplementary Fig. S5). The only noticeable difference was the increase in nuclear sisR-1, which is consistent with DIP1 repressing sisR-1 in the nucleus as reported previously (Fig. 3D, F, Supplementary Fig. S5)29. We hypothesized that the lack of cytoplasmic upregulation of sisR-1 in DIP1 mutants could be due to functional Clu particles that promote the degradation of sisR-1 in the cytoplasm. In support of this idea, we observed a more severe upregulation of sisR-1 by smFISH and RT-qPCR in DIP1;clu double mutants, which was concomitant with clumping of mitochondria in the nurse cells (Fig. 3D–G, Supplementary Fig. S5). The size of these ovaries was also significantly smaller compared to single mutants (Fig. 3H), indicating that DIP1 and Clu act synergically in repressing sisR-1. During fasting, both the downregulation of nuclear DIP1 and disassembly of cytoplasmic Clu bodies lead to the upregulation of sisR-1 in the cytoplasm, which consequently causes the formation of mitochondrial clumps (Fig. 3I).

sisR-1 inhibits p62-mediated clearance of mitochondria during fasting

To investigate the downstream function of sisR-1 in regulating mitochondria during fasting, we examined p62 (SQSTM1 or Drosophila Ref(2)P), which is a receptor of autophagy36,37. p62 recognizes ubiquitinated cargos and targets them to autophagosomes for degradation36. During mitophagy, which is a specialized pathway of autophagy that targets mitochondria, p62 binds to ubiquitinated outer mitochondrial membrane proteins and selectively targets them for degradation (Fig. 4A)38. During this process, p62 is similarly degraded simultaneously.

Fig. 4: sisR-1 regulates p62-mediated mitophagy via polyubiquitination of VDAC1.
figure 4

A A drawing of p62-mediated mitophagy. B Confocal images showing localizations of sisR-1 (Green), p62 (Red), ATP5α (Magenta) and DAPI (Blue) in fasted Oregon R. n = 3 biological replicates. Scale bar = 10 µm. C Stacked bar graph showing the percentages of ATP5α-positive mitochondrial clumps that colocalize with sisR-1, p62 or both sisR-1 and p62. D Western blots showing levels of p62 in insoluble fractions extracted from ovaries of fed and fasted Oregon R and sisR-1 line 1 flies. E, F Graphs showing the relative levels of insoluble p62 in fed and fasted Oregon R and sisR-1 RNAi line 1, as shown in the blot in D. Data are presented as mean values +/−SD from four biological replicates. *p < 0.05. Two-tailed t-test. G Model on the effects of Park on ubiquitination of VDAC1 affecting p62-mediated mitophagy of defective mitochondria. H Western blots showing mono- and poly-ubiquitination of VDAC1 in fed and fasted Oregon R and fasted sisR-1 RNAi line 1 against control β-Tubulin. n = 2 biological replicates. I Chart showing the percentages of stage 8/9 egg chambers with mitochondrial clumps in fasted Oregon R, VDAC1-WT overexpression, VDAC1-K273R overexpression, VDAC1-poly-KR overexpression and Parkin overexpression ovaries. n = 30 (Oregon R), 37 (WT), 35 (K273R), 30 (poly-KR), 24 (Park OE). egg chambers counted. *p < 0.05. ns not significant. Chi-square test, one-sided. J Chart showing the average numbers of mitochondrial clumps present per stage 8/9 egg chamber that contain clumps in fasted Oregon R, VDAC1-WT overexpression, VDAC1-K273R overexpression, VDAC1-poly-KR overexpression and Parkin overexpression ovaries. Data are presented as mean values +/−SD. *p < 0.001. n = 30 (Oregon R), 37 (WT), 35 (K273R), 30 (poly-KR), 24 (Park OE). egg chambers counted. Two-tailed t-test. K Confocal images of ATP5α localization in fasted Oregon R, VDAC1-WT overexpression, VDAC1-K273R overexpression, VDAC1-poly-KR overexpression and Parkin overexpression ovaries Arrowheads point to mitochondrial clumps. n = 3 biological replicates. Scale bar = 20 µm. Source data are provided as a Source Data file.

When fed, p62 forms small cytoplasmic bodies in the nurse cell cytoplasm (Supplementary Fig. S6A). We examined the localization of p62 in relation to sisR-1 and mitochondrial clumps in ovaries of fasted flies (Fig. 4B, C). Unexpectedly, we found that majority (~70%) of the aggregates within mitochondrial clumps colocalized with only sisR-1, while only ~20% colocalized with only p62. The remaining minority fraction (~10%) colocalized with both sisR-1 and p62. The high mutual exclusivity of mitochondrial clumps colocalizing with either sisR-1 or p62 suggests the possibility that sisR-1 may suppress active localization of p62 to mitochondrial clumps.

To test if sisR-1 is involved in the regulation of p62-mediated mitophagy (or autophagy) during fasting, we monitored the levels of insoluble p62 in the ovaries when flies were fed or fasted39. The levels of insoluble p62 were significantly lower in sisR-1 RNAi ovaries when compared to Oregon R controls only under fasting conditions (Fig. 4D, E). To corroborate this finding, we observed that the number of p62 bodies was also lower in sisR-1 RNAi ovaries under fasting conditions (Supplementary Fig. S6B). These observations suggest that sisR-1 regulates p62 only during fasting. As expected, the levels of insoluble p62 decreased during fasting in Oregon R controls (Fig. 4D, F), which reflected the clearance of p62 via autophagy. Importantly, the degree of p62 downregulation was significantly greater in sisR-1 RNAi flies (Fig. 4F), supporting the idea that sisR-1 inhibits p62-mediated autophagy.

It was recently reported in mammalian cells of a vault noncoding RNA that inhibits autophagy by directly binding to p62 to inhibit its oligomerization40. To check if sisR-1 interacts with p62, we performed immunoprecipitation of GFP-p62 from a transgenic line expressing GFP-p62. GFP-p62 colocalized with endogenous p62 and exhibited similar dynamics during fasting (Supplementary Fig. S6C). Immunoprecipitation of GFP-p62 also pulled down endogenous p62 but not sisR-1 in both control and sisR-1 overexpression conditions (Supplementary Fig. S6D, E). Our data suggest that sisR-1 does not function to regulate p62 through direct binding.

sisR-1 inhibits poly-ubiquitination of VDAC1 during fasting

To identify the downstream target of sisR-1 in the regulation of mitophagy, we considered VDAC1 (or Drosophila Porin), which is an outer mitochondrial protein that has been shown to be a p62 target during mitophagy36. It was recently shown that the ubiquitination levels of VDAC1 can influence the cell decision to undergo apoptosis or mitophagy41. Mono-ubiquitination of VDAC1 inhibits apoptosis, whereas poly-ubiquitination promotes mitophagy via p62 (Fig. 4G). Western blotting of ovarian lysates showed that both mono- and poly-ubiquitination of VDAC1 were downregulated during fasting (Fig. 4H). Furthermore, this downregulation of VDAC1 ubiquitination was dependent on sisR-1 (Fig. 4H). These results are consistent with a role for sisR-1 in the downregulation of mitophagy during fasting.

Our results so far suggest that during fasting, upregulation of mitochondrial sisR-1 inhibits VDAC1 poly-ubiquitination and mitophagy, consequently leading to mitochondrial clumping. Since Drosophila Parkin (Park) facilitates mitophagy by promoting the poly-ubiquitination of VDAC141, we tested if germline overexpression of Park can suppress the mitochondrial clumping phenotypes during fasting. Indeed, overexpression of Park significantly reduced the number of mitochondrial clumps and the percentages of egg chambers containing clumps (Fig. 4I–K). Consistent with the model, germline overexpression of WT and K273R (mono-ubiquitination mutant), but not poly-KR (poly-ubiquitination mutant), forms of VDAC1 significantly suppressed the mitochondrial clumping phenotypes (Fig. 4I–K). Taken together, our experiments support the model that poly-ubiquitination of VDAC1 is required to promote mitophagy in the ovaries during fasting.

Clearance of fasting-induced mitochondrial clumps improves oocyte quality

Our results so far show that fasting leads to the upregulation of sisR-1, which inhibits the clearance of mitochondrial clumps by mitophagy. Next, we asked if the accumulation of these clumps during fasting has any impact on oocyte quality. First, we compared the hatch rates of fasted Oregon R and sisR-1 RNAi flies and observed a significant improvement from ~45% (Oregon R) to ~65% in sisR-1 RNAi flies, consistent with the model that a reduction of sisR-1-mediated mitochondrial clumps improves oocyte quality during fasting (Figs. 1H–J and 5A, B).

Fig. 5: Clearance of fasting-induced mitochondrial clumps improves oocyte quality.
figure 5

A Stage 8/9 egg chambers showing fasting-induced mitochondrial clumping persists in Oregon R but not in sisR-1 RNAi ovaries. B Hatch rates 2 days after egg laying in fasted Oregon R and sisR-1 RNAi. *p < 0.05. Chi-square test, one-sided. n = 183 eggs in 20 experiments (Oregon R) and 86 eggs in 7 experiments (RNAi). C Stage 8/9 egg chambers showing sisR-1 expression in initially fed and fasted Oregon R flies and after being subsequently fed. D Hatch rates of Oregon R females that underwent the indicated duration of feeding and fasting regimes. Data are presented as mean values +/−SD. *p < 0.05. ns not significant. Two-tailed t-test. n = 3 (16 h fed), 3 (16 h fasted), 6 (30 h fed), 8 (30 h fasted) biological replicates. E Confocal images of ATP5α localization in fed or fasted Oregon R ovaries at 16 or 30 h. Arrowheads point to mitochondrial clumps. Scale bar = 20 µm. F Percentages of nurse cell per stage 8/9 egg chamber with mitochondrial clumps in fed or fasted Oregon R ovaries at 16 or 30 h. Data are presented as mean values +/−SD. **p < 0.01, ***p < 0.001. n = 7 (16 h fed), 7 (16 h fasted), 3 (30 h fed), 6 (30 h fasted) egg chambers counted. Two-tailed t-test. G Confocal images of sisR-1 (Green) and DAPI (Blue) localizations in fed or fasted Oregon R ovaries at 16 or 30 h. Arrowheads point to mitochondrial clumps. Scale bar = 20 µm. H Levels of sisR-1 normalized to actin5C in fed or fasted Oregon R ovaries at 16 or 30 h. Data are presented as mean values +/−SD. n = 4 biological replicates. Two-tailed t-test. I Relative expression of sisR-1 by smFISH in the cytoplasm of stage 8/9 egg chambers in fed or fasted Oregon R ovaries at 16 or 30 h. Data are presented as mean values +/−SD. ***p < 0.001. n = 6 (16 h fed), 6 (16 h fasted), 8 (30 h fed), 9 (30 h fasted) egg chambers counted. Two-tailed t-test. Source data are provided as a Source Data file.

Since the expression of sisR-1 is dynamically modulated by food availability, we tested if clearance of mitochondrial clumps after fasting by refeeding had any impact on oocyte quality (Fig. 5C). Newly-eclosed Oregon R female flies were exposed to either fed or fasted conditions for 30 h before they were transferred to a nutrient-rich environment with males for 24 h. The flies were subsequently removed to determine the hatch rates (Fig. 5C). Interestingly, an initial fasting of 30 h significantly improved the hatch rates as compared to those that are fed for the same duration (Fig. 5D). Furthermore, the effect was only seen when flies were exposed to a longer duration of fasting (for 30 h), but not at 16 h (Fig. 5D). Different feeding durations did not alter the hatch rates (Fig. 5D), indicating that the effects were specific to fasting.

The impact of fasting duration on oocyte quality was further corroborated by the abundance of sisR-1 and mitochondrial clumps under these conditions. Higher amounts of sisR-1 and mitochondrial clumps were present in nurse cells fasted for 30 h than 16 h (Fig. 5E–I). However, no significant difference was observed under fed conditions (Fig. 5E–I). These observations suggest that fasting induces the accumulation of sisR-1, which in turn leads to mitochondrial clumping, and further clearance of these mitochondrial clumps by either genetic knockdown of sisR-1 or refeeding led to the production of better-quality oocytes.

sisR-1 response in other types of stresses

We also explored other forms of stresses to examine if they alter sisR-1 levels and mitochondrial clumping. Interestingly, protein deprivation alone was sufficient to induce sisR-1 expression and mitochondrial clumping (Fig. 6A–D). We also found that subjecting flies to a transient 1-h heat shock at 37 °C also led to an induction of sisR-1 and mitochondrial clumps (Fig. 6A–D). As a result, we further investigated if this form of transient heat stress could also improve oocyte quality, similar to what we observed with fasting. Indeed, flies subjected to a 1-h heat shock at 37 °C showed improved hatch rates as compared to those kept at 25 °C continuously (Fig. 6E).

Fig. 6: sisR-1 response in other types of stresses.
figure 6

A Confocal images of ATP5α localization in ovaries during aging, protein deprivation and after heat shock. Arrowheads point to mitochondrial clumps. Scale bar = 20 µm. B Confocal images of sisR-1 (Green) and DAPI (Blue) localizations in ovaries during aging, protein deprivation and after heat shock. Scale bar = 20 µm. C Relative expression of sisR-1 as measured by smFISH in the cytoplasm of stage 8/9 egg chambers during aging, protein deprivation and after heat shock. Data are presented as mean values +/−SD. *p < 0.05, **p < 0.001. n = 7 (day 4), 9 (day 12), 7 (fed), 8 (protein deprived), 7 (25 °C), 9 (37 °C 1 h) egg chambers counted. Two-tailed t-test. D Chart showing percentages of nurse cell per stage 8/9 egg chamber with mitochondrial clumps in ovaries during aging, protein deprivation and after heat shock. Data are presented as mean values +/−SD. ***p < 0.001. n = 3 (day 4), 6 (day 12), 3 (RNAi day 12), 3 (fed), 7 (protein deprived), 3 (25 °C), 6 (37 °C 1 h) egg chambers counted. Two-tailed t-test. E Hatch rates of control females kept at 25oC and with transient heat shock at 37 °C. Data are presented as mean values +/−SD. *p < 0.05. n = 10 biological replicates per group. Two-tailed t-test. F Hatch rates of females at day 2 and 12 in Oregon R and sisR-1 RNAi #1. Data are presented as mean values +/−SD. **p < 0.001. n = 10 biological replicates per group. Two-tailed t-test. G Number of eggs laid per female per day at day 2 and 12 in Oregon R and sisR-1 RNAi #1. Data are presented as mean values +/−SD. **p < 0.001. n = 10 biological replicates per group. Two-tailed t-test. Source data are provided as a Source Data file.

We next examined if sisR-1 and mitochondrial clumps had any impact on oogenesis during aging. Ovaries of young females (4-day old) displayed lower expression of sisR-1 and less mitochondrial clumps than that of older 14-day old flies (Fig. 6A–D). This was concomitant with a significant decrease in egg production from ~70 eggs/female/day in 4-day old flies to ~43 eggs/female/day in 14-day old flies (Fig. 6F). To investigate if the decrease in egg production during aging was caused by sisR-1 and mitochondrial clumping, we examined the number of eggs produced by sisR-1 RNAi flies. In sisR-1 RNAi flies, the decrease in egg production from day 4 to day 14 was not significant (Fig. 6F). The number of mitochondrial clumps was also significantly lower in 14-day old sisR-1 RNAi flies compared to control flies of the same age (Fig. 6A, D). Although we observed a decrease in egg production in Oregon R during aging, we did not observed any significant differences in hatch rates (Fig. 6F, G). On the other hand, in sisR-1 RNAi flies, a decrease in hatch rate but not egg production, was observed during aging (Fig. 6F, G). These results suggest that during aging, an increase in sisR-1 expression may induce mitochondrial clumping and activate a checkpoint that suppresses the production of poorer quality oocytes.

Discussion

Mitochondrial quality is controlled by mitophagy, a process which is activated by fasting. How fasting induces mitophagy has been extensively studied in somatic cells and less is known in germline cells37,38. In this study, we unexpectedly found that fasting inhibits the mitophagy-mediated clearance of damaged mitochondria in Drosophila ovaries. Mechanistically, fasting causes the downregulation of the DIP1-Clu pathway, leading to an up-regulation of the lncRNA or sisRNA sisR-1 which localizes to mitochondria. At the mitochondria, sisR-1 inhibits the poly-ubiquitination of the outer mitochondrial membrane protein VDAC1 to suppress p62-mediated clearance of the clumps (Fig. 7). This model is supported by a few lines of evidence presented in this study and by others. (1) clu RNAi induced formation of mitochondrial clumps, which could be rescued by overexpression of Park, thus placing clu genetically upstream of park14. (2) Both fasting and downregulation of clu led to upregulation of sisR-1, causing formation of mitochondrial clumps (Figs. 1 and 2). (3) Fasting-induced sisR-1 expression led to downregulation of polyubiquitination of VDAC1 (Fig. 4H). (4) Park promotes polyubiquitination of VDAC1, which is important for mitophagy to occur41. (5) Overexpression of Park and VDAC-WT, but not poly-KR (poly-ubiquitination mutant), suppressed mitochondrial clumping during fasting (Fig. 4I–K). Together, these observations provide genetic evidence that sisR-1 functions to suppress Park-mediated polyubiquitination of VDAC1 during fasting in the ovaries. In future, it will be interesting to characterize the molecular mechanism behind this process.

Fig. 7: Proposed working model.
figure 7

The role of sisR-1 in inhibiting Park-dependent ubiquitination of VDAC1 during fasting and how subsequent feeding leads to clearance of sisR-1 by Clu and activation of mitophagy.

In Drosophila ovaries, starvation triggers mitochondria to enter quiescence during stage 8/9 of oogenesis, a process similar to developmentally regulated mitochondria quiescence7,8. Interestingly, during this process, mitochondria are highly remodeled and are not cleared by mitophagy. At the same time, stage 8/9 is a highly nutrient-sensitive stage of oogenesis, where a shortage of nutrients causes short-term arrest to coordinate food intake and egg production42. One possible mechanism is that mitochondria are more sensitized to damage during nutritional stress, however sisR-1 suppresses the efficient clearance of damaged mitochondria, which serves as a signal to trigger pre-vitellogenesis arrest. The resumption of mitophagy upon refeeding eventually allows oogenesis to proceed and improves overall mitochondria and oocyte quality.

VDAC1 appears to be a central player in coordinating mitochondrial remodeling during both quiescence and mitophagy. During entry into quiescence, GSK3 phosphorylates VDAC1 to mediate proteasome recruitment8. During mitophagy, Park is recruited to damaged mitochondria and poly-ubiquitinates VDAC136,41. However, in germline cells, mitochondrial localized sisR-1 suppresses poly-ubiquitination of VDAC1. We speculate that sisR-1 may block the recruitment of Park to mitochondria, which is supported by a previous study reporting a role for clu in promoting Park localization to mitochondria14. Recently, vault RNA has been shown to bind to p62 and inhibit its dimerization/oligomerization, thereby directly suppressing autophagy40. We did not find any evidence that sisR-1 binds to p62 or regulates the dimerization of p62, suggesting that it functions as a novel regulator of mitophagy.

A previous study has shown that Clu acts to regulate mitochondrial dynamics by recruiting Drp1 to promote fission43. Our study provides a novel function of Clu in safeguarding mitochondria by promoting decay of mitochondrial localized sisR-1. We found that Clu binds to sisR-1 and negatively regulates its stability (Fig. 2). Localization of Clu to mitochondria is a highly dynamic event that depends on nutrient availability34. Interestingly, the mammalian homolog Cluh had been also implicated in mitochondria homeostasis44,45,46,47,48,49,50. We have also previously shown that a nuclear RNA binding protein DIP1 binds to sisR-1 and promotes it decay29. Here we further show that during fasting, DIP1 expression and localization are disrupted, contributing to a de-repression of sisR-1 in the nucleus. Taken together, our results show that under well-fed conditions, sisR-1 levels are actively controlled at the posttranscriptional manner by two RNA binding proteins in the nucleus (by DIP1) and cytoplasm (by Clu). During fasting, the expression and/or localization of DIP1 and Clu are disrupted, leading to greater stability of sisR-1. In the future, it would be important to decipher how the DIP1-Clu-sisR-1 pathway is controlled by nutrient availability and whether a similar pathway functions in mammalian germline cells to modulate oocyte mitochondrial quality control.

Methods

Fly strains

The following fly strains were used in this study: y w and Oregon R (used as control unless otherwise stated), vasa-Gal4/CyO51, TOR RNAi[HMS00904] (Bloomington #33951), sisR-1 RNAi #1 and #221, sisR-1 OE21, clu[d08713/CyO] and clu[CA06604] (gifts from Rachel Cox)12, DIP1 (Bloomington #15577), GFP-p62 (gift from Bhupendra Shravage)52, UAS-park41, UAS-porin/VDAC1 WT, UAS-porin/VDAC1 K273R (mono)41, UAS-porin/VDAC1 poly-KR (poly)41. To knockdown sisR-1, UASp-sisR-1 RNAi were crossed with daughterless (da)-Gal4 as described previously21. They were maintained in standard cornmeal medium unless otherwise stated. To prepare the cornmeal medium, 290.93 g cornmeal, 254.5 g dextrose, 118 g brewer’s yeast, 40 g agar and 150 ml 10% Nipagin were mixed with deionized water to prepare 5 litres of food.

Feeding protocol

To obtain RNA and protein from regularly fed female flies that are not used for the fasting assays, newly eclosed female flies were fed with yeast paste for 2 days before ovaries were dissected. For protein deprivation, the flies were placed in regular food vials without yeast paste.

Fasting and hatch rate protocol

Pupae about to eclose were placed in vials containing 1% agarose (dissolved in water) without yeast paste for 2 days before female flies were dissected for RNA, protein or immunostaining. For crossing, 2–7 day old fed males were added into the respective agarose vials without yeast paste and allowed to mate and lay eggs for 2 days. The flies were transferred to fresh vials daily. Eggs and hatched larvae were counted for 3 days after the flies were mated.

Fasting assay of oocytes

Newly eclosed flies were fasted or fed for 2 days before they were dissected, and the number of stage 14 oocytes scored. The average number of stage 14 oocytes were calculated in control flies and sisR-1 RNAi flies in both fed and fasted conditions. For each experiment, at least 32 ovaries from at least 16 flies were counted.

Heat shock and aging protocols

Newly eclosed flies were fed with yeast paste at 25 °C continuously for 4 days. For controls, the flies were transferred to food vials without yeast paste at 25 °C for 1 h. For heat shock, flies were transferred to food vials without yeast paste and placed in a 37 °C water bath or incubator for 1 h. After which, both groups of flies were dissected immediately for staining or smFISH experiments. For fertility assays, the flies were mated to 2–7 day old fed males in food vials with yeast paste and allowed to lay eggs for 1 day.

For aging assay, newly eclosed flies were fed with yeast paste at 25 °C and transferred into fresh food vials with yeast paste every 2-3 days. The flies were dissected or mated with males at day 4 or 12 for staining/smFISH experiments and fertility assays.

Alpha-amanitin treatment

Whole ovaries dissected from flies fatten with yeast paste were incubated in Grace’s medium containing 50 µg/mL of alpha-amanitin (Sigma) for the indicated time points before RNA was extracted.

Mitochondria fractionation

The fractionation was performed using Mitochondria Isolation Kit for Cultured Cells kit (Thermo Scientific) with slight modifications of the manufacturer’s protocol. About 60 pairs and 120 pairs of whole ovaries were dissected from the fed and fasted flies respectively and homogenized in Mitochondria Isolation Reagent A supplemented with 1× Halt Protease Inhibitor Cocktail (Thermo Scientific) using a motorized pellet homogenizer. Then, equal volume of Mitochondria Isolation Reagent C supplemented with 1× Halt Protease Inhibitor Cocktail was added to the lysed ovaries and centrifuged for 10 min at 700 × g, 4 °C. The supernatant was carefully transferred to a new tube and centrifuged again for 15 min at 3000 × g, 4 °C to enriched for mitochondria. The resulted supernatant was transferred to a new tube and labeled as the cytosol fraction while the resulted pellet is the mitochondria fraction. The pellet was washed twice using Mitochondria Isolation Reagent C supplemented with 1× Halt Protease Inhibitor Cocktail and spun down at 12,000 × g, 4 °C for 5 min. After removing all washing buffer, the pellet was resuspended with protein extraction buffer (50 mM Tris-HCl pH 7.5, 150mN NaCl, 5 mM MgCl2, 0.1% NP-40) supplemented with Protease Inhibitor Cocktail (Roche). The samples were kept in −80 °C until further experiments.

RNA extraction

RNA extraction was performed as described previously using TRIzol (Ambion) and the Direct-zol miniprep kit (Zymo Research)21,53. To detect maternally deposited sisRNA and mRNA, stage 14 oocytes were manually isolated to avoid contamination with pre-mRNA in transcriptionally active germline cells. For experiments involving detection of pre-mRNA and mRNA for gene expression, whole ovaries were used as they reflect the steady state levels of pre-mRNA and mRNA.

RT-PCR/qRT-PCR

RT-PCR/qRT-PCR were performed as described previously29,53. Transcript abundance was normalized against actin5C mRNA. RT was performed with M-MLV RT (Promega). PCR products were run on 1% agarose gel to visualize DNA. qRT-PCR was done using SYBR Fast qPCR kit master mix (2x) universal (Kapa Biosystems, USA) and on the Applied Biosystems 7900HT Fast Real-Time PCR system. Oligonucleotides used were sisR-1 Fw TCATGGAATCAGAAGCCCGT, sisR-1 Rv GGTTGTAAGCGTGGTGTCTC, DIP1 Fw TAATACGACTCACTATAGGGAGAAAGAAGTTGCGACAGAACCG, DIP1 Rv TAATACGACTCACTATAGGGAGACGAACAGCTTGTAGATGGCA, actin5C Fw TGCCCATCTACGAGGGTTAT, actin5C Rv AGTACTTGCGCTCTGGCGG, Rp49 Fw CCAAGGACTTCATCCGCCACC, Rp49 Rv GCGGGTGCGCTTGTTCGATCC.

Western blot

Western blotting was done as described previously29. Dissected ovaries were homogenized in protein extraction buffer (50 mM Tris-HCl pH 7.5, 150 mN NaCl, 5 mM MgCl2, 0.1% NP-40) supplemented with Protease Inhibitor Cocktail (Roche) and 2X sample buffer containing beta-mercaptoethanol were added to desired volume to load. Extraction of insoluble fractions was done as previously described54. Transfer of proteins was performed at 100 V for 1.5 h. Primary antibodies used were rabbit anti-DIP1 (1:5,000)29, rabbit anti-p62 (1:1,000, gift from Shoichiro Kurata and Tamaki Yano)55, mouse anti-porin (1:200, abcam 14734), mouse anti-lamin (1:1,000, DSHB ADL67.10-c), mouse anti-GFP (1:2,000, Invitrogen 3E6 A-11120), mouse anti-ATP5a (1:1000, Abcam 15H4C4), mouse anti-beta-Tubulin (1:1000, DSHB E7) and mouse anti-Actin (1:100, DSHB JLA20). Immunoblotting for Actin was done to determine equivalent loading. Detection was done using the ChemiDoc Touch Imaging System (BioRad) under non-saturating conditions and Odessey imaging system. Quantification of western blots was performed using ImageJ software as previously described22,28.

Immunoprecipitation

Immunoprecipitation was done as described previously20,35. Dissected ovaries were lysed in protein extraction buffer (50 mM Tris-HCl pH 7.5, 150 mN NaCl, 5 mM MgCl2, 0.1% NP-40) supplemented with Protease Inhibitor Cocktail (Roche). This was followed by blocking the protein lysates with protein A/G agarose beads (Merck Millipore). Antibodies (anti-GFP) were added and incubated overnight at 4 ˚C. For negative controls, no antibody was added. The next day, protein A/G beads were added and incubated for 2 h before washing in protein extraction buffer 3 times. Thereafter, protein and RNA were extracted using 2X sample buffer containing beta-mercaptoethanol and Direct-zol RNA miniprep kit (Zymo Research), respectively.

Immunostaining

Immunostaining was performed as described previously29,56. Briefly, ovaries were fixed in 16% PFA: Grace’s media (ratio 1:2) for 10–20 min before washing with PBS-T (0.2% Triton-X100) and blocked for 30 min in 5% normal goat serum. Ovaries were incubated in primary antibody mix overnight at room temperature. The next day, they were washed in PBS-T for three times and incubated in secondary antibody mix for 4 h at room temperature. DNA was stained using DAPI and ovaries were mounted in Vectashield and examined under the Leica SPEII microscope.

Primary antibodies used were as follows: mouse anti-ATP5A (1:500, Abcam 14748), rabbit anti-p62 (1:1000, gift from Shoichiro Kurata and Tamaki Yano)55, rabbit anti-DIP1 (1:300)29.

For Figs. 1I, 2D and 4I, the percentage of mitochondrial clumps were calculated by scoring the number of egg chambers positive for mitochondrial clumps over the total number of stage 8/9 egg chambers from 6 to 8 ovaries. For Figs. 3F, 5F and 6B, and Supplementary Fig. S3B, the percentage of mitochondrial clumps were calculated by scoring the number of nurse cells positive for mitochondrial clumps over the total nurse cells counted in an egg chamber from 6–8 ovaries. Both methods obtained comparable and similar results. Quantification of signal intensities was performed using identical confocal settings and Image J software as previously described28.

RNA single molecule Fluorescence in situ Hybridization (smFISH)

RNA fluorescence in situ hybridization (FISH) was performed using an RNAscope multiplex fluorescence kit (ACDbio, 320851). In brief, ovaries were fixed in 16% PFA: Grace’s media (ratio 1:2) for 10–20 min before washing with 1X PBS-T (0.2% Triton-X100) and blocked in 1% BSA + PBS-T at room temperature for 30 minutes. 1X target retrieval solution was added to the ovaries and boiled at 100 °C for 15 minutes. Ovaries were then rinse with 1% BSA + PBS-T and 100% methanol was added to the tube to dehydrate the sample. Ovaries was then washed carefully in 1% BSA + PBS-T. ovaries were treated with Protease III:1X PBS (ratio 1:5) for 5 min in 40 °C water bath and subsequently washed with probe diluent. The probe against sisR-1 (ACDbio, 417991) was added into the tube. After 2 h of incubation at 40 °C, hybridize-Amp probes #1–4 were added into the tubes sequentially to amplify the signal. All ovaries were counterstained with DAPI (Invitrogen, 62248) and mounted in Vectashield and examined under the Leica SPEII scanning microscope. Contrast and brightness were corrected using Image J software.

Statistics

For all experiments, the tests and the number of independent biological replicates were indicated in the figure legends or figures. P values and definitions of error bars were indicated in the legends. Sample sizes were not pre-determined prior to the experiments. T-tests were performed on samples that are normally distributed.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.