Abstract
Stem cells preferentially use glycolysis instead of oxidative phosphorylation and this metabolic rewiring plays an instructive role in their fate; however, the underlying molecular mechanisms remain largely unexplored. PIWI-interacting RNAs (piRNAs) and PIWI proteins have essential functions in a range of adult stem cells across species. Here, we show that piRNAs and the PIWI protein Aubergine (Aub) are instrumental in activating glycolysis in Drosophila female germline stem cells (GSCs). Higher glycolysis is required for GSC self-renewal and aub loss-of-function induces a metabolic switch in GSCs leading to their differentiation. Aub directly binds glycolytic mRNAs and Enolase mRNA regulation by Aub depends on its 5′UTR. Furthermore, mutations of a piRNA target site in Enolase 5′UTR lead to GSC loss. These data reveal an Aub/piRNA function in translational activation of glycolytic mRNAs in GSCs, and pinpoint a mechanism of regulation of metabolic reprogramming in stem cells based on small RNAs.
Similar content being viewed by others
Introduction
Energy metabolism has long been known as a homeostatic system that can adapt to cellular energy needs. However, a growing body of evidences have now directly implicated energy metabolism in cell fate. Metabolic remodeling takes place early during transitions between cell states, before the establishment of a particular cell fate, and has been shown to play an important role in stem cell biology and reprogramming. In particular, induction of pluripotency during cell reprogramming correlates with a shift from oxidative phosphorylation (oxphos) to glycolysis, and stimulation of glycolysis facilitates pluripotency1,2. Conversely, oxidative metabolism is activated during the differentiation of pluripotent stem cells, and differentiation can be suppressed through decreased mitochondrial respiration1,3,4. These findings underscore the importance of energetic pathways in pluripotency and stem cell differentiation. Increased glycolysis is a common feature across a number of stem cell populations, including hematopoietic and mesenchymal stem cells5,6,7. One reason for the glycolytic state of stem cells is to preserve their genome integrity by maintaining low oxidative stress through reduced oxphos8. A key question is thus to understand how energy metabolism is controlled in stem cells at the molecular level. Regulation occurs in part at the transcriptional level through hypoxia as various stem cell populations reside within niches with low oxygen tension. The transcription factors hypoxia-inducible factor (HIF)1α and 2β are stabilized under hypoxic conditions and associate with a constitutive HIF subunit to promote glycolysis and suppress mitochondrial biogenesis. This occurs through transcriptional activation of several glycolytic genes and repression of PGC-1α, a general regulator of genes encoding mitochondrial proteins9,10. Besides this transcriptional regulation, very little is known regarding the regulatory mechanisms underlying metabolic remodeling in stem cells, despite the importance of this remodeling in stem cell fate.
Piwi-interacting RNAs (piRNAs) are a class of small non-coding RNAs first identified in the germline of animal species, but widely present in somatic tissues as well11. piRNAs are loaded into specific Argonaute proteins, called PIWI proteins, and interact with their target mRNAs by complementarity. piRNAs and PIWI proteins are involved in repressing transposable elements, but they also have essential functions as regulators of cellular mRNAs via various mechanisms. These include mRNA cleavage through the endonuclease activity of PIWI proteins, deadenylation and decay, stabilization, and translational activation, thanks to the capacity of PIWI proteins to recruit different mRNA regulatory machineries12,13. PIWI proteins and piRNAs are required in various stem cell populations. Drosophila Piwi, the founding member of the PIWI family was identified based on its function in germline stem cell (GSC) asymmetric division14. Piwi is necessary for both GSCs and somatic niche cells for GSC self-renewal and differentiation15. The functions of PIWI proteins in GSC proliferation and maintenance are conserved from C. elegans to the mouse16,17. Recently, a role of Aubergine (Aub), another of the three Drosophila PIWI proteins was also demonstrated in GSC self-renewal and to a lesser extent differentiation18,19. In aub mutant, GSCs can differentiate into germline cysts, but they lose the capacity to self-renew and are rapidly lost. This role of Aub depends on its capacity to bind piRNAs and its function in gene regulation19. The proto-oncogene Cbl was identified as a relevant Aub mRNA target, and Aub acts as a negative regulator of Cbl mRNA for GSC self-renewal19. However, Aub positively regulates other mRNAs such as dunce, and was proposed to do so through activation of translation initiation18.
PIWI proteins are also required in stem cell populations outside the gonads. Drosophila Piwi has functions in the gut in establishing the intestinal stem cell population in pre-adult stages20, and in the maintenance and differentiation of intestinal stem cells in response to stress and during aging21. Strikingly, PIWI proteins are also markers of pluripotent stem cells in primitive species with high regenerative capacities, such as sponges, cnidarians and planarians22,23,24. Moreover, knock-down approaches have revealed their requirement for whole-body regeneration in several of these species15,23.
Our understanding of piRNA biogenesis has revealed tight links between energy metabolism and the piRNA pathway. piRNA biogenesis occurs in part at the mitochondrial outer membrane and the endonuclease (Drosophila Zucchini, mouse MitoPLD), as well as other factors involved in piRNA production are integral components of the mitochondrial outer membrane25,26,27,28. In addition, a genetic screen identified a role of subunits of the mitochondrial respiratory chain complexes in piRNA biogenesis29. Moreover, glycolytic enzymes were also reported to be involved in piRNA biogenesis30. Using Aub iCLIP (individual-nucleotide resolution UV crosslinking and immunoprecipitation) datasets from early embryos31 and cultured GSCs18, we identified a total of seven glycolytic mRNAs among mRNAs encoding the eleven glycolytic enzymes (Fig. 1a). This suggested that Aub could regulate glycolysis through direct interaction with glycolytic mRNAs. Here, we show that glycolysis is a critical energetic pathway in GSCs and that it is required for GSC self-renewal. Either reducing glycolysis by decreasing the levels of glycolytic enzymes, or increasing oxphos in GSCs promote GSC differentiation, preventing their maintenance. Strikingly, a switch in energy metabolism towards oxphos takes place in aub mutant GSCs. Aub-dependent increase of glycolytic enzyme levels in GSCs compared to differentiating cells, together with piRNA-dependent Aub binding to glycolytic mRNAs suggest a direct role of Aub in translational activation of glycolytic mRNAs in GSCs. This was confirmed using reporter transgenes containing the 5′UTR of Enolase (Eno) that were found to display Aub-dependent higher expression in GSCs. In addition, CRISPR-based deletions or mutations of a piRNA target site in Eno 5′UTR result in GSC loss. Finally, the strong GSC loss characteristic of aub mutants was reduced by increasing glycolysis. Together, these results reveal a key role of Aub and piRNAs in activating glycolysis in GSCs for their self-renewal. They also identify a regulatory mechanism of the metabolic remodeling in stem cells depending on piRNAs. Because piRNAs and PIWI proteins are present in various stem cell populations including in primitive regenerative species, this mode of regulation of stem cell metabolic reprogramming might be conserved throughout evolution.
a The ten steps of glycolysis. Glycolytic mRNAs found in GFP-Aub iCLIP datasets18,31 are in red. Aub binding to glycolytic mRNAs is statistically significant, p = 1.0e-5 using the Generalized hypergeometric test. b Scheme of the germarium. Germ cells are in green, except the oocyte that is in purple (GSC: germline stem cell; CB: cystoblast). The spectrosome and fusome are in red. The spectrosome acquires an elongated form during GSC division. Differentiation proceeds from anterior (left) to posterior (right). Regions 1, 2 and 3 contain mitotically active germ cells, differentiating 16 cell-cysts, and the newly formed egg chamber, respectively. c–e’ Confocal images of wild-type (wt) germaria immunostained with an anti-Ald (c, c’), anti-Eno (d, d’) or anti-PyK (e, e’) (green), anti-Hts (red) and DAPI. White arrows indicate GCSs and white arrowheads indicate the region containing germline cysts where glycolytic enzyme levels were quantified. f–h Quantification of Ald1 (f), Eno (g) and PyK (h) protein levels in GSCs and differentiating cyst cells using immunostaining experiments shown in c–e’. Fluorescence intensity was measured in arbitrary units using the ImageJ software. Horizontal bars represent the mean and standard deviations. ****p-value <0.0001 using the paired two-tailed Student’s t test. p = 1.84e−7 in (f); 5.99e−7 in (g); 8.76e−10 in (h). i–k’ Confocal images of mosaic germaria containing control (i, i’), Ald1EY13155 (j, j’) or EnoKG01162 (k, k’) clonal cells stained with anti-GFP (green) (lack of GFP indicates clonal cells), anti-Hts (red) and DAPI (blue), 14 days after clonal induction. The white dashed line (i, i’) indicates a control clonal GSC, and the yellow dashed line (j–k’) indicate mutant clonal differentiating cysts in contact with the niche. Heterozygous Ald1EY13155 females in which mitotic clones were induced are not viable after 17 days. l Relative percentages of germaria with at least a clonal GSC at 7, 14 or 21 days after clonal induction. The number of germaria analyzed is indicated. Data are presented as mean values with standard deviations. Quantifications were from two independent experiments, except for Enof07543 for which it was from one experiment. Scale bars: 10 μm. Source data are provided as a Source Data file.
Results
High glycolysis is required for GSC self-renewal
Although various populations of stem cells are known to preferentially use glycolysis as their main energetic pathway, this question has not been addressed in Drosophila ovarian GSCs, one of the best-described model of adult stem cells. The Drosophila ovary is composed of 16 to 18 ovarioles in which egg chambers develop towards the posterior. The anterior-most region of ovarioles is called the germarium and contains two to three GSCs localized anteriorly and in contact with the niche composed of somatic cells. GSCs are identified by a spherical organelle, the spectrosome localized anteriorly towards the niche (Fig. 1b). GSCs divide asymmetrically to produce a new GSC -the cell that remains in contact with the niche- by a process known as self-renewal, and a cystoblast -the cell that loses the contact with the niche- that starts differentiation. The cystoblast divides four consecutive times with incomplete cytokinesis to produce a germline cyst composed of sixteen interconnected germline cells containing a branched fusome derived from the original spectrosome (Fig. 1b).
To document the metabolic reprogramming in Drosophila GSCs, we first analyzed the expression patterns of three glycolytic enzymes, Aldolase 1 (Ald1), Eno and Pyruvate kinase (PyK) in germaria using immunostaining. We used available antibodies directed against mammalian proteins since these proteins are highly conserved. The three glycolytic enzymes were present in all cells in the germarium consistent with their fundamental role in energy metabolism, but their levels were higher in GSCs than in early differentiating germ cells (Fig. 1c–h). The specificity of the three antibodies was confirmed by immunostaining of germaria expressing the corresponding RNAi with the UAS/Gal4 system and the germline-specific driver nos-Gal4 (Supplementary Fig. 1a–g). PyK higher protein levels in GSCs compared to differentiating germ cells was validated using a P-GFP (Wee-P) insertion in the PyK locus producing a GFP-PyK fusion protein32 (Supplementary Fig. 1h, i). To address whether PyK higher protein levels in GSCs was underlied by a similar mRNA expression profile, we performed PyK single-molecule fluorescence in situ hybridization (smFISH) in germaria. We found that, in contrast to PyK protein, PyK mRNA was more abundant in differentiating cells than in GSCs (Supplementary Fig. 1j–n), suggesting that PyK protein accumulation in GSCs would depend on translational control. Consistent with higher expression of glycolytic enzymes in GSCs, previous studies have shown that GSCs do not express high levels of oxphos components and have reduced mitochondrial membrane potential33,34. Specifically, subunits of ATP synthase, the last complex of the mitochondrial electron transport chain were not found in GSCs, but were present in high amounts in differentiating cyst cells33. In addition, mitochondrial membrane potential as well as electron transport chain activity were very low in GSCs and the following dividing cyst cells and sharply increased in sixteen cell cysts34.
We next determined whether glycolysis plays an instructive role in GSC stemness. We addressed the requirement of high glycolysis for GSC self-renewal through reduction of glycolytic enzyme levels using clonal analysis with Eno and Ald1 mutants. The Ald1EY13155, EnoKG601162 35 and Enof07543 (Supplementary Fig. 1o) mutants are lethal but clonal mutant germ cells do survive, suggesting their hypomorphic nature30. Wild-type, Ald1- and Eno- clonal GSCs were generated with the FLP-mediated FRT recombination system36 and the presence of clonal GSCs, revealed by the loss of GFP expression, was quantified at three time points, i.e. 7, 14 and 21 days after clone induction. Wild-type clonal GSCs were maintained over time (Fig. 1i, i’, l), whereas a proportion of Ald1EY13155, EnoKG601162 or Enof07543 mutant GSCs were lost 14 and 21 days after clone induction, showing a defect in their self-renewal (Fig. 1j–k’, l). Importantly, immunostaining of these germaria with anti-Hts antibody that marks spectrosomes and fusomes showed that both Ald1 and Eno mutant differentiating cysts were present in the niche, indicating that mutant clonal GSCs were able to differentiate and were likely lost by differentiation (Fig. 1j, k). This was confirmed by recording cell death in clonal mutant GSCs using anti-cleaved Caspase 3 staining. The percentage of germaria with clonal GSCs positive for anti-cleaved Caspase 3 was very low and not higher in mutant than wild-type GSCs (Supplementary Fig. 1p-t), showing that Ald1 and Eno mutant GSCs did not undergo apoptosis, but were lost by differentiation.
To confirm the role of glycolytic genes in GSC self-renewal, we expressed available UAS-RNAi transgenes with the nos-Gal4 driver to reduce the levels of glycolytic enzymes in GSCs. UAS-RNAi transgene efficiency was validated using RT-qPCR (Supplementary Fig. 2a). Immunostaining of ovaries with anti-Hts antibody, 7, 14, and 21 days after eclosion, showed that downregulation in germ cells of the nine tested glycolytic mRNAs led to GSC loss quantified by the percentage of germaria with less than two GSCs (Supplementary Fig. 2b–l). This loss increased with time and differentiated cysts could be found in the niche, indicating that differentiation was not prevented. To reinforce the conclusion that GSC loss did not result from cell death, but rather their differentiation, we co-expressed the baculovirus anti-apoptotic protein P35 in germ cells37 together with UAS-RNAi against either Ald1 or Eno. The presence of P35 did not significantly reduce the GSC loss resulting from decreased levels of Ald1 or Eno, confirming that these GSCs did not undergo massive cell death (Supplementary Fig. 2m–p). Finally, we asked whether decreasing differentiation might affect the GSC loss resulting from glycolytic mRNA knock-downs in GSCs. Bag of marble (Bam) is the major differentiation factor in the GSC lineage38. We found that nos-Gal4 bamΔ86/+ heterozygous females showed a weak dominant phenotype of reduced differentiation visualized by the presence of mislocalized pre-cystoblasts with a spectrosome in germaria (Supplementary Fig. 2q). We thus addressed whether the presence of the bamΔ86/+ heterozygous mutation would reduce the GSC loss of Phosphofructokinase (Pfk) or Eno knock-downs. In both cases, the GSC loss in 7-day-old females was low and not significantly different to that of the nos-Gal4 bamΔ86/+ control, indicating a decrease of the GSC loss phenotype when differentiation was weakened (compare Supplementary Fig. 2t to Supplementary Fig. 2l for UAS-Pfk-RNAi and UAS-Eno-RNAi).
These results show that glycolytic enzymes accumulate at higher levels in GSCs than in differentiating cysts and that these higher levels of glycolytic enzymes are required for GSC self-renewal by preventing their differentiation.
Low Oxphos is required for GSC self-renewal
In addition to the increased levels of ATP synthase components in differentiating cysts, these components were shown to be required for division and differentiation of cyst cells, but not for GSC self-renewal33. ATP synthase is involved in the maturation of mitochondrial cristae in differentiating cyst cells, a prerequisite for oxphos33. To strengthen the conclusion of metabolic reprogramming towards glycolysis playing an important role in GSC homeostasis, we analyzed the effect of increasing oxphos in GSCs using two different approaches. First, we deregulated the mitochondrial pyruvate dehydrogenase (PDH) complex that catalyzes the first step of pyruvate metabolism for oxphos (Fig. 2a). The E1 enzyme of PDH is activated by pyruvate dehydrogenase phosphatase (PDP) and inhibited by pyruvate dehydrogenase kinase (PDK)39. We used the nos-Gal4 driver to express UAS-RNAi transgenes in germ cells and reduce the levels of Pdp or Pdk mRNAs (Supplementary Fig. 3a, b). Reducing PDP levels should reduce oxphos and this had no effect on GSC self-renewal, consistent with GSCs not heavily depending on oxphos (Fig. 2b, c, g). In contrast, reducing PDK levels that was expected to favor oxphos, led to GSC loss visualized by the presence of differentiating cysts in the niche, which increased with time (Fig. 2d, g). Next, we overexpressed general activators of mitochondrial mass and activity in germ cells, which is expected to increase oxphos. The coactivator PPAR coactivator 1 (PGC-1, Spargel in Drosophila) and the transcription factor nuclear respiratory factor-1 (NRF-1, Erect wing in Drosophila) coregulate a large set of genes involved in mitochondrial function40,41,42. spargel and erect wing (ewg) were overexpressed in germ cells using P-UAS constructs inserted into the genes and the nos-Gal4 driver (Supplementary Fig. 3c, d). Increase of mitochondrial mass following spargel and ewg overexpression was confirmed by visualizing mitochondria with MitoTracker and measuring the mitochondrial surface in GSCs (Supplementary Fig. 3e–h). Overexpression of both spargel and ewg led to GSC loss and the presence of differentiating cyst cells in the niche, consistent with the notion that increased oxphos prevented GSC self-renewal, but not differentiation (Fig. 2e–g). To confirm that the GSC loss resulted from differentiation rather than cell death, we co-expressed the P35 anti-apoptotic protein together with spargel or ewg overexpression and found that P35 did not reduce GSC loss (Supplementary Fig. 3i–l). In contrast, the presence of bamΔ86/+ heterozygous mutant strongly reduced the GSC loss resulting from spargel or ewg overexpression (Supplementary Fig. 3m–p; compare Fig. 2g to Supplementary Fig. 3p), consistent with GSC loss occurring through differentiation.
a Scheme of Pyruvate Dehydrogenase (PDH) complex regulation by Pyruvate Dehydrogenase Phosphatase (PDP) and Pyruvate Dehydrogenase Kinase (PDK). b–f Confocal images of germaria from 7 day old-females stained with anti-Vasa (green), anti-Hts (red) and DAPI (blue). The white dashed line indicates two GSCs in a control nos-Gal4/+ germarium (b) and the yellow dashed lines indicate two GSCs in a nos-Gal4/UAS-Pdp-RNAi germarium (c) and a single GSC or differentiating cysts in the niche in nos-Gal4/UAS-Pdk-RNAi (d), nos-Gal4/spargelEY05931 (e) and ewgEY05137/+; nos-Gal4/+ (f) germaria, respectively. g Quantification of germaria showing GSC loss (0-1 GSC) 7 and 14 days after eclosion. The number of scored germaria (n) is indicated. ****p-value <0.0001, ns, non-significant using the χ2 test. 7 days, p = 0.34 between wt and Pdp RNAi; 2.4e−6 between wt and Pdk RNAi; 1.0e−5 between wt and spargelEY05931; 1.0e-5 between wt and ewgEY05137. 14 days, p = 0.16 between wt and Pdp RNAi; 1.0e−6 between wt and Pdk RNAi; 1.0e−5 between wt and spargelEY05931; 1.0e−5 between wt and ewgEY05137. Scale bars: 10 μm. Source data are provided as a Source Data file.
We conclude that high oxphos levels induce differentiation and are not compatible with GSC maintenance. These results concur with a recent study showing that increasing mitochondrial membrane potential by interfering with mitochondrial dynamics results in GSC loss43.
Aubergine is required for metabolic reprogramming towards glycolysis in GSCs
Aub iCLIP assays revealed that glycolytic mRNAs interact with Aub. We, therefore, asked whether Aub could modulate energy metabolism by directly regulating glycolytic mRNAs. We first performed immunostaining of aub mutant ovaries with anti-Eno and anti-PyK antibodies to address whether the higher levels of glycolytic enzymes found in wild-type GSCs were maintained in aub mutant GSCs. Both Eno and PyK levels were similar in GSCs and differentiating cyst cells, in aub mutant germaria (Fig. 3a–e’), as quantified by the ratio of fluorescence intensity in GSCs to cyst cells that was significantly reduced in aub mutant compared to wild-type germaria (Fig. 3c, f). Thus, Aub increases glycolytic enzyme levels in GSCs, possibly through direct regulation of glycolytic mRNAs.
a–b’, d–e’ Confocal images of immunostaining of wt and aubQC42/HN2 germaria with anti-Eno (green) and DAPI (blue) (a–b’) or anti-PyK (green) and anti-Hts (red) (d–e’). White and yellow arrows indicate GCSs in wt and aub mutant germaria, respectively. White and yellow arrowheads indicate the region containing germline cysts where glycolytic enzyme levels were quantified, in wt and aub mutant germaria, respectively. c, f Quantification of Eno and PyK protein levels in wt and aubQC42/HN2 mutant GSCs and differentiating cyst cells using immunostaining experiments shown in (a–b’, d–e’). Fluorescence intensity was measured in arbitrary units using the ImageJ software. Ratios of fluorescence intensity of one GSC to one cyst cell per germarium were plotted. Horizontal bars represent the mean and standard deviations. ****p-value < 0.0001 using the unpaired two-tailed Student’s t test. p = 1.93e−5 in (c); 9.36e−11 in (f). g Schematic representation of the FRET Laconic sensor. The binding of lactate to the ligand-binding domain leads to a conformational change that increases the distance between the donor (mTFP) and the acceptor (Venus), resulting in reduced FRET efficiency. Thus, FRET Laconic sensor efficiency inversely correlates with lactate concentration. h, i FRET ratio images in control nos-Gal4/UASz-Laconic (h) and aub mutant, aubQC42/HN2; nos-Gal4/UASz-Laconic (i) anterior-most region of germaria. Control and aub mutant GSCs are indicated with white and yellow dashed lines, respectively. The rainbow colormap indicates the FRET efficiency levels. j Quantification of FRET Laconic sensor efficiency in control and aub mutant, aubQC42/HN2 and aubQC42/g1, GSCs from three day old-females based on acceptor photobleaching. Horizontal bars represent the mean and standard deviations. ***p-value < 0.001, ****p-value < 0.0001, using the unpaired two-tailed Student’s t test. p = 7.14e−5 between control and aubQC42/HN2 and 1.72e−4 between control and aubQC42/g1. Scale bars: 10 μm in (a–e’); 5 μm in (h, i). Source data are provided as a Source Data file.
To more directly assess the role of Aub in the regulation of energy metabolism in GSCs, we implemented the utilization of FRET metabolic sensors in GSCs. These genetically-encoded fluorescent probes sensitive to intracellular concentration of metabolites allow to quantify specific metabolites at single cell resolution. Pyruvate, the end-product of glycolysis is the input energy substrate for oxphos and can be converted into lactate under low oxphos. We used a lactate sensor called Laconic (an UASz version of the Laconic sensor44,45) expressed in germ cells with the nos-Gal4 driver. This Laconic sensor is built such that FRET efficiency decreases in the presence of lactate (Fig. 3g). FRET efficiency of the Laconic sensor was quantified based on acceptor photobleaching (see Methods). We first validated that the Venus acceptor alone did not produce donor mTFP-like fluorescence by photoconversion upon photobleaching46 (Supplementary Fig. 4a–b’). FRET Laconic sensor efficiency increased in aub mutant GSCs compared to wild-type revealing lower lactate levels (Fig. 3h–j). These reduced amounts of lactate indicated lower glycolysis and/or higher oxphos in aub mutant GSCs, consistent with the proposed role of Aub in activating glycolysis in GSCs.
We next addressed whether mitochondrial morphology and activity might be affected in aub mutant GSCs.Mitochondrial activity is intimately linked to their structure and organization. Globular fissed mitochondria with underdeveloped cristae are a hallmark of stem cells and induced pluripotent stem cells that mainly rely on glycolysis, whereas fused mitochondria with numerous cristae are present in differentiated cells allowing ATP production through increased oxphos activity8,47. Importantly, in Drosophila GSCs, mitochondria were also reported to be fissed with poorly developed cristae, and they became more fused with abundant cristae following differentiation33,43. We thus analyzed mitochondrial morphology in wild-type and aub mutant GSCs using transmission electron microscopy. As described previously, in wild-type GSCs, mitochondria tended to be globular with undeveloped cristae (Fig. 4a, a’). In aub mutant GSCs, mitochondria were more fused as shown by the increased number of longer mitochondria compared to their length in wild-type GSCs, and they had more developed cristae (Fig. 4b–d, Supplementary Fig. 4c, d). Therefore, in aub mutant GSCs, the mitochondrial network was more mature than in wild-type GSCs, resembling that in differentiating cells. We then analyzed the expression of ATP synthase α (ATPsynα) in aub mutant GSCs. In the wild type, the levels of ATPsynα are very low in GSCs and strongly increase in differentiating 8- and 16-cell cysts33. We recorded ATPsynα levels in aub mutant GSCs by inducing aub mutant clonal GSCs followed by immunostaining of ovaries with anti-ATPsynα antibody. aub mutant GSCs expressed significantly higher levels of ATPsynα than control heterozygous GSCs (Fig. 4e, f). To confirm this result, we quantified ATPsynα levels using immunostaining in wild-type and aub mutant GSCs identified with the GFP-Par1 marker to visualize spectrosomes48. ATPsynα levels were higher in aub mutant than in wild-type GSCs (Supplementary Fig. 4e–g). To investigate whether reduced glycolysis in GSCs could lead to indirect increased expression of oxphos genes, we quantified ATPsynα expression in Ald1 and Eno mutant GSCs using the clonal analysis. Similarly to aub mutant GSCs, the levels of ATPsynα were higher in Ald1 and Eno mutant GSCs than in control heterozygous GSCs, suggesting a metabolic rewiring upon lower glycolysis (Supplementary Fig. 5a–f).
a–b’ Transmission Electron Microscopy images of a wt (a) and an aubQC42/HN2 mutant (b) GSC. The white arrow in wt and yellow arrow in aub mutant GSCs point to mitochondria that are shown enlarged in (a’) and (b’). c Measurement of mitochondrial length in wt and aubQC42/HN2 mutant GSCs from the Transmission Electron micrographs. Percentages of mitochondria with a length above or below 0.25 μm are shown. The number of mitochondria (n) from at least ten GSCs per condition is indicated. *p-value <0.05 using the χ2 test. p = 0.025. d Quantification of mitochondria with and without cristae in wt and aubQC42/HN2 mutant GSCs from the Transmission Electron micrographs. Percentages of mitochondria with and without cristae are shown. The number of mitochondria (n) from at least ten GSCs per condition is indicated. ****p-value <0.0001 using the χ2 test. p = 1.0e−5. e, e’ Confocal images of a mosaic germarium containing an aubHN2 mutant clonal GSC stained with anti-GFP (green), anti-ATPsynα (red) and DAPI (blue). The non-clonal (GFP+) and aub mutant clonal (GFP-) GSCs are indicated by dashed white and yellow lines, respectively. f Quantification of ATPsynα protein levels in control non-clonal aubHN2/+ and clonal aubHN2 mutant GSCs using fluorescence intensity measured in arbitrary units using the ImageJ software. Horizontal bars represent the mean and standard deviations. *p-value < 0.05 using the paired two-tailed Student’s t test. p = 0.026. g, h Quantification of mitochondrial membrane potential: ratiometric images of wt and aubQC42/HN2 mutant germaria stained with TMRM to record mitochondrial membrane potential and Deep Red MitoTracker to record mitochondrial mass. The ratio of TMRM to MitoTracker intensity was calculated using ImageJ and a pseudo-colored image was generated using the rainbow RGB gradient. White and yellow arrows indicate wt and aub mutant GSCs, respectively. i Quantification of TMRM/MitoTracker ratios in wt and aubQC42/HN2 mutant GSCs using ImageJ. Horizontal bars represent the mean and standard deviations. **p-value < 0.01 using the two-tailed Mann–Whitney test. p = 0.0019. Scale bars: 1 μm in (a, b); 10 μm in (e, g, h). Source data are provided as a Source Data file. Panel (a’) and (b’) were created with BioRender.com released under a Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International license (https://creativecommons.org/licenses/by-nc-nd/4.0/deed.en).
Finally, we quantified mitochondrial membrane potential as a measurement of mitochondrial activity in wild-type and aub mutant GSCs, using tetramethylrhodamine methyl ester (TMRM) staining to record mitochondrial membrane potential and Deep Red MitoTracker to measure mitochondrial mass34,43. As reported previously, in wild-type germaria, mitochondrial membrane potential was very low in GSCs and dividing cyst cells, and sharply increased in 16 cell cysts (Fig. 4g, i)34. In contrast, in aub mutant germaria mitochondrial membrane potential was already high in GSCs (Fig. 4h, i).
Together these results show that Aub plays a key role in increasing glycolysis in GSCs. In the absence of Aub, expression of glycolytic enzymes is reduced in GSCs and this is accompanied by a switch in energy metabolism towards oxphos with higher expression of ATP synthase that might contribute to mitochondrial maturation.
Aubergine directly regulates glycolytic mRNAs
GFP-Aub iCLIP assays performed in cultured GSCs revealed five glycolytic mRNAs significantly bound by Aub18 (Supplementary Fig. 6a, Supplementary Table 1). Aub iCLIP sites were mostly found in 5′UTRs and 3′UTRs, with higher accumulation in 3′UTRs in most cases. To confirm Aub interaction with glycolytic mRNAs and address whether this interaction depends on Aub capacity to load piRNAs, we performed RNA immunoprecipitation (RIP) in ovaries. GFP-Aub and GFP-AubAA 31, an Aub double point mutant in the PAZ domain preventing Aub loading with piRNAs were expressed in GSCs and germ cells using UASp-based transgenes and the nos-Gal4 driver (Fig. 5a, b). GFP-Aub and GFP-AubAA immunoprecipitations were validated using western blots (Fig. 5c) and enrichment of glycolytic mRNAs in GFP-Aub and GFP-AubAA RIPs were quantified using RT-qPCR (Fig. 5d, e, Supplementary Fig. 6b, c). The seven glycolytic mRNAs identified in GFP-Aub iCLIP data sets from GSCs and embryos18,31 were enriched in GFP-Aub RIPs, but not in GFP-AubAA RIPs (Fig. 5d, e, Supplementary Fig. 6b, c), confirming Aub interaction with these mRNAs and demonstrating the requirement of Aub loading with piRNAs for this interaction.
a, b Confocal images of immunostaining of germaria expressing GFP-Aub (nos-Gal4/UASp-GFP-Aub) or GFP fused to AubAA that does not load piRNAs (UASp-GFP-AubAA/+; nos-Gal4/+) with anti-GFP (green), anti-Hts (red) and DAPI (blue). GFP-Aub is present in the nuage around nuclei, whereas GFP-AubAA is not. c Western blot showing immunoprecipitation (IP) of GFP-Aub and GFP-AubAA with anti-GFP in ovaries. Genotypes were as in (a, b). w1118 ovaries were used as negative control. Input corresponds to 1/8 of extract prior to immunoprecipitation. The blot was revealed with anti-GFP. Experiments in (a–c) were repeated twice with similar results. d, e Quantification of mRNAs using RT-qPCR in GFP-Aub (d, e) and GFP-AubAA (e) IPs. mRNA enrichment compared to mRNA levels in w1118 IPs set to 1 (horizontal bars). mRNA levels were normalized to U1 snRNA and to RNA levels in the input. nanos (nos) mRNA was used as positive control31. Mean of two biological replicates for Hex-A and Pfk mRNAs and three biological replicates for the other mRNAs in (d). Error bars represent standard deviation. One biological replicate in (e). f–i Genetic interaction between aub and either Ald1 or Eno for GSC self-renewal. Confocal images of immunostaining of aubHN2/+ (f), aubHN2/+; Def(Ald1)/+ (g) and aubHN2/EnoKG01162 (h) germaria with anti-Vasa (green), anti-Hts (red) and DAPI (blue). The white dashed line indicates two GSCs in a control aubHN2/+ germarium (f) and the yellow dashed lines indicate differentiating cysts in the niche in aubHN2/+; Def(Ald1)/+ (h) and aubHN2/EnoKG01162 (g) germaria. i Quantification of germaria showing GSC loss (0-1 GSC) in heterozygous and double heterozygous mutant females of the indicated genotypes, 7 days after eclosion. The number of scored germaria (n) is indicated. * p-value < 0.05, ** p-value < 0.01, ns, non-significant using the χ2 test, compared to the sum of GSC loss from the two corresponding single heterozygous. p = 0.0084 for aubHN2/+; Df(Ald1)/+; 0.021 for aubQC42/+; Df(Ald1)/+; 0.58 aubg1/+; Df(Ald1)/+; 0.31 for aubHN2/EnoKG01162; 0.0015 for aubQC42/EnoKG01162; 0.0067 for aubg1/EnoKG01162. Scale bars: 10 μm. Source data are provided as a Source Data file.
Using a genetic approach, we analyzed the defect in GSC self-renewal when both aub and either Ald1 or Eno gene dosage was reduced by half. The concomitant reduction of gene dosage of aub and its potential target Ald1 or Eno led to a synergistic effect in GSC loss analyzed by immunostaining, with higher GSC loss in double heterozygous mutants than the sum of GSC loss from both heterozygous mutants (Fig. 5f–i), in lines with a positive regulation of Ald1 and Eno by Aub.
Together, these results confirm the direct binding of Aub to glycolytic mRNAs and show the requirement of Aub loading for this binding. Furthermore, genetic data provide functional evidence that Aub positively regulates glycolytic mRNAs through this interaction.
Aubergine regulation of glycolytic mRNAs depends on piRNA targeting
We then addressed the potential of piRNAs to target glycolytic mRNAs bound by Aub in GSCs, in the vicinity of GFP-Aub iCLIPs. We used published piRNA libraries from cultured GSCs49 and analyzed potential piRNA targeting within ±60 nucleotides (nt) from significant GFP-Aub iCLIPs using various complementarities31 (see Methods). We then filtered targeting with the highest piRNA occurrences (total occurrence of overlapping piRNAs >500) and a low minimum free energy prediction of the piRNA-target RNA duplex (<−24 kcal/mol)50. Using these criteria, we identified four glycolytic mRNAs targeted by abundant GSC piRNAs in regions overlapping GFP-Aub iCLIPs (Supplementary Fig. 7a).
To functionally confirm the importance of glycolytic mRNA regulation by Aub and piRNAs for GSC maintenance, we focused on Eno. This gene produces different mRNAs but a single of them is highly expressed in GSCs (Fig. 6a, Supplementary Fig. 6a, Supplementary Table 1). This Eno mRNA is targeted in both its 5′UTR and 3′UTR by two piRNA populations produced from the Quasimodo transposable element (Fig. 6a, Supplementary Fig. 7a). We generated deletions of the piRNA target site in Eno 5′UTR, using the CRISPR-Cas9 approach. Two short deletions of 11 nt and 12 nt overlapping the target site of the piRNA seed region (nt 2-7) were obtained and named EnoΔpi11 and EnoΔpi12 (Fig. 6a). Both deletions removed a donor splice site and could prevent splicing in the 5′UTR. However, we found that splicing occurred between a new downstream donor splice site and the native acceptor site (Fig. 6a, Supplementary Fig. 7b). We analyzed GSC loss in EnoΔpi11 and EnoΔpi12 mutants either homozygous or in combination with another allele, Enof07543, using immunostaining 14 and 21 days after fly eclosion. GSC loss was significant with both deletion mutants at both time points, showing that Eno mRNA regulation by piRNAs is required for GSC self-renewal (Fig. 6b–d, f). To confirm that the GSC loss in EnoΔpi11 and EnoΔpi12 mutants resulted from preventing Eno 5’UTR targeting by Quasimodo piRNAs, we produced an Eno point mutant, Enopimut1 that bore seven point mutations in the piRNA target site to prevent piRNA targeting (Fig. 6a). This mutant displayed a wild-type splicing, confirmed by sequencing, and conserved the coding sequence of the longer Eno isoform that is expressed at low levels in GSCs (Fig. 6a, Supplementary Table 1). Similarly to the deletion mutants, Enopimut1 either homozygous or in combination with Enof07543 showed a GSC loss 14 and 21 days after eclosion (Fig. 6e, g). In addition, immunostaining with anti-Eno antibody revealed that the higher Eno levels in GSCs compared to cyst cells in wild-type germaria were reduced in Enopimut1/f07543 mutant germaria (Supplementary Fig. 8a–c).
a Scheme of the Eno gene; thick boxes: coding sequences; thin boxes: UTRs; lines: introns. GFP-Aub iCLIPs in cultured GSCs from three replicates are shown in red18. The arrowheads at the 5′ ends indicate mRNA isoforms expressed in GSCs. Open arrowhead: low expression (>10 and <100 Transcripts Per Million (TPM): 10 to 30 TPM for this transcript); green arrowhead: high expression (>100 TPM: 150 to 240 TPM for this transcript). The Eno sequence overlapping GFP-Aub iCLIPs in the 5′UTR and intron is shown. Nucleotides (nt) in red are bound by Aub in the iCLIP datasets18. The sequence and occurrences of GSC Quasimodo piRNAs potentially targeting Eno are indicated. Nt in blue are non-complementary to the Eno sequence. The sequences of EnoΔpi11, EnoΔpi12, and Enopimut1 are shown. Dashes and stars indicate deleted and modified nt, respectively. A part of the protein sequence of Eno long isoform is shown; the modified nt do not change the coding sequence. Boxed nt indicate the donor and new donor splice sites. b–e Confocal images of immunostaining of wt (b), EnoΔpi11 (c), EnoΔpi12 (d) and Enopimut1 (e) germaria with anti-Vasa (green), anti-Hts (red) and DAPI (blue). The white and yellow dashed lines indicate GSCs in wt (b) and mutant (c-e) germaria, respectively. f, g Quantification of germaria showing GSC loss (0–1 GSC) in wt and mutant females, 14 and 21 days after eclosion. The number of scored germaria (n) is indicated. *p-value < 0.05, **p-value <0.01, ***p-value <0.001, ****p-value <0.0001 using the χ2 test. In (f), 14 days, p = 2.0e−4 between wt and EnoΔpi11; 1.0e−4 between wt and EnoΔpi11/f07543; 0.0012 between wt and EnoΔpi12; 0.0017 between wt and EnoΔpi12/f07543. 21 days, p = 8.52e−5 between wt and EnoΔpi11; 6.99e−5 between wt and EnoΔpi11/f07543; 6.0e−4 between wt and EnoΔpi12; 0.0088 between wt and EnoΔpi12/f07543. In (g), 14 days, p = 0.0021 between wt and Enopimut1; 0.015 between wt and Enopimut1/f07543. 21 days, p = 0.004 between wt and Enopimut1; 0.047 between wt and Enopimut1/f07543. Scale bars: 10 μm. Source data are provided as a Source Data file.
We next generated Eno reporter transgenes containing nanos (nos) promoter to allow germline expression, followed by Eno 5′UTR, GFP coding sequence and either Eno 3′UTR or SV40 3′UTR (Fig. 7a). Strikingly, both transgenes recapitulate Eno protein expression profile in wild-type germaria, with higher GFP levels in GSCs than in differentiating cyst cells (Fig. 7b, d, f, g). Furthermore, this difference between GFP levels in GSCs and differentiating cells was significantly reduced in aub mutant germaria (Fig. 7c, e–g). These results show that Aub-dependent Eno mRNA activation depends mainly on its 5′UTR, strongly suggesting a translational activation.
a Schemes of Eno reporter transgenes; open boxes: nos promoter; orange boxes: Eno UTRs; green boxes: GFP coding sequence; blue box: SV40 3′UTR. b–e Confocal images of immunostaining of germaria expressing Eno reporter transgenes nosP-5′Eno-GFP-3′Eno (b, c) and nosP-5′Eno-GFP-3′SV40 (d, e), in wt (b, d) and aubQC42/g1 mutant context (c, e) with anti-GFP. White and yellow arrows point to GSCs in wt and mutant germaria, respectively. White and yellow arrowheads indicate the region containing germline cysts, where GFP was quantified in wt and mutant germaria, respectively. f, g Quantification of GFP levels in wt and aubQC42/g1 mutant GSCs and differentiating cyst cells using immunostaining experiments shown in (b–e). Fluorescence intensity was measured in arbitrary units using the ImageJ software. Ratios of fluorescence intensity of one GSC to one cyst cell per germarium were plotted. Horizontal bars represent the mean and standard deviations. ****p-value < 0.0001 using the unpaired two-tailed Student’s t test with Welch’s correction (p = 2.36e−5) in (f) and the two-tailed Mann–Whitney test (p = 2.34e−8) in (g). Scale bars: 10 μm. Source data are provided as a Source Data file.
Together, these data demonstrate that Eno mRNA activation by Aub in GSCs depends on piRNA targeting and is required for GSC self-renewal. They further show that Eno mRNA regulation by Aub involves its 5′UTR and likely occurs at the level of translation.
Aub function in GSC self-renewal involves its role in activating glycolysis
Aub was shown to be essential for GSC self-renewal through its role in the regulation of cellular mRNAs18,19. We find here that Aub activates the translation of glycolytic mRNAs in GSCs and that high glycolysis is required for GSC self-renewal. Thus, we asked whether activation of glycolysis contributes to Aub function in GSC self-renewal. We analyzed if aub mutant phenotype of GSC loss could be rescued to some extent by increasing glycolysis. Phosphofructokinase (Pfk) catalyzes an irreversible reaction in glycolysis that corresponds to the first commitment of glucose to the glycolytic pathway. Because Pfk is rate limiting, it plays a critical role in determining glycolytic flux and is considered as the pacemaker of glycolysis51. We overexpressed Pfk in germ cells using a Drosophila strain containing a P-UASp insertion upstream of the coding sequence, crossed with the nos-Gal4 germline driver (Supplementary Fig. 8d). GSC analysis using immunostaining 7 days and 14 days after fly eclosion in two different aub allelic combinations, revealed that GSC loss in aub mutant germaria was significantly reduced following germline overexpression of Pfk (Fig. 8a–d). In particular, the complete lack of germ cells (empty germaria) that became substantial with time due to GSC loss, sharply decreased when Pfk was overexpressed (Fig. 8b, d).
a–c Confocal images of immunostaining of aubQC42/HN2; nos-Gal4/+ (a, b) and aubQC42/HN2PfkEY23126/+; nos-Gal4/+ (c) germaria with anti-Vasa (green), anti-Hts (red) and DAPI (blue). The yellow dashed lines indicate the GSCs. An empty germarium, devoid of germ cells is shown in (b). d Quantification of germaria showing GSC loss (0–1 GSC) in mutant females of the indicated genotypes, 7 and 14 days after eclosion, and of empty germaria in mutant females 14 days after eclosion. The number of scored germaria (n) is indicated. **p-value < 0.01, ****p-value < 0.0001, ns, non-significant using the χ2 test. For GSC loss at 7 days, p = 1.0e−6 between aubQC42/HN2; nos-Gal4/+ and aubQC42/HN2 PfkEY23126/+; nos-Gal4/+; 0.0019 between aubQC42/g1; nos-Gal4/+ and aubQC42/g1 PfkEY23126/+; nos-Gal4/+. For GSC loss at 14 days, p = 6.56e−5 between aubQC42/HN2; nos-Gal4/+ and aubQC42/HN2 PfkEY23126/+; nos-Gal4/+; 0.08 between aubQC42/g1; nos-Gal4/+ and aubQC42/g1 PfkEY23126/+; nos-Gal4/+. For empty gemaria, p = 8.72e−6 between aubQC42/HN2; nos-Gal4/+ and aubQC42/HN2 PfkEY23126/+; nos-Gal4/+; 1.37e-8 between aubQC42/g1; nos-Gal4/+ and aubQC42/g1 PfkEY23126/+; nos-Gal4/+. e Model of Aub and piRNA function in metabolic remodeling in GSCs. Aub guided by piRNAs binds glycolytic mRNAs in GSCs leading to their translational activation and higher levels of glycolytic enzymes. Translational activation might involve Aub interaction with the translation initiation factors PABP and eIF3 as is the case in the embryo54. Aub-dependent activation of glycolytic mRNAs leads to high glycolysis in GSCs that is required for their self-renewal. During cyst cell differentiation, mitochondrial maturation takes place leading to an increased number of fused mitochondria equipped with cristae. This allows a gradual metabolic rewiring towards oxphos during the process of germ cell differentiation. Other factors (question mark) are expected to cooperate with Aub to restrict translational activation to GSCs. Scale bars: 10 μm. Source data are provided as a Source Data file. Panel e was created with BioRender.com released under a Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International license (https://creativecommons.org/licenses/by-nc-nd/4.0/deed.en).
These data provide functional evidence that Aub-dependent activation of glycolysis is a key contribution to Aub function in GSC self-renewal.
Discussion
Over the recent years, it has become clear that, in addition to its bioenergetic functions, metabolism plays a role in the regulation of cellular programs, and in particular in the control of stem cell fate52. Various populations of stem cells depend on glycolysis to generate ATP and this specific metabolism actively contributes to stemness maintenance, whereas ATP production through mitochondrial respiration is directly associated to differentiation6. The molecular mechanisms of this metabolic rewiring remain poorly understood. Transcriptional regulation by HIF transcription factors are known to be involved. The stabilization of these transcription factors under hypoxic conditions in stem cell niches, leads to both transcriptional activation of glycolytic genes and reduction of mitochondrial biogenesis9,10. Another level of regulation in human pluripotent stem cells is based on increased expression of uncoupling protein 2 (UCP2) that shunts pyruvate from mitochondria, thus favoring glycolysis1,53. However, the molecular bases of this regulation remain unknown. A key information from this study is to identify an unanticipated level of control of energy metabolism in stem cells based on mRNA regulation by piRNAs and PIWI proteins. We show that Aub targets glycolytic mRNAs and this leads to elevated levels of glycolytic enzymes and higher glycolysis in GSCs. Aub loading with piRNAs is necessary for Aub interaction with glycolytic mRNAs and mutating a potential piRNA target site in Eno mRNA 5′UTR results in defective GSC self-renewal. Moreover, Eno reporter transgenes containing Eno UTRs with an unrelated promoter recapitulate Aub-dependent Eno mRNA regulation in GSCs. These results demonstrate a control of metabolic remodeling in stem cells at the posttranscriptional level involving Aub and piRNAs (Fig. 8e).
The piRNA target site in Eno mRNA shown to be functional for Eno regulation in GSCs is located in the 5′UTR, 23 nt upstream of the translation initiation codon. Furthermore, an Eno reporter transgene containing only Eno 5′UTR, followed by exogenous coding sequence and 3′UTR, reproduces Aub-dependent higher protein levels in GSCs. These data point to a function of Aub and piRNAs in translational activation of glycolytic mRNAs in GSCs (Fig. 8e). A role for Aub and piRNAs in translational activation has been demonstrated for nos mRNA in the Drosophila embryo54. It depends on direct protein interactions between Aub and the translation initiation factors poly(A) binding protein (PABP) and eIF3 subunits. Importantly, a similar role in translational activation has also been described for Miwi, the mouse homolog of Aub, and also involves direct Miwi interactions with PABP1 and subunits of eIF355, showing that this positive function of PIWI proteins in mRNA regulation is conserved12.
The function of mitochondria has been analyzed in the Drosophila ovarian GSC lineage, although the energetic pathways used at the different steps of GSC differentiation remain uncharacterized. Components of ATP synthase (complex V), the last complex of the mitochondrial electron transport chain, are expressed in the germarium in dividing cyst cells and required for their division and differentiation33,37. However, ATP synthase at these stages is involved in mitochondrial maturation, i.e. cristae formation, but not for oxphos. Indeed, ATP synthase requirement in differentiation of early cyst cells was shown to depend on its role in mitochondrial remodeling that prevents ER stress, premature meiosis and cell death37. The formation of mitochondrial cristae increases the surface of the inner mitochondrial membrane and would accommodate the presence of more electron transport chain complexes, favoring a switch towards ATP production through oxphos56,57. However, components of complexes II, III and IV of the electron transport chain are not required for early cyst cell differentiation, indicating that early differentiating cells do not strongly depend on oxphos37. The switch towards oxphos would be more gradual, and indeed mutants targeting complexes III or IV stop oogenesis after full differentiation of the egg chamber at stages 8/9. These data are in lines with the upregulation of electron transport chain complexes at the transcription level in differentiated 16 cell-cysts and with the actual rise of oxphos levels in these cells, measured through mitochondrial membrane potential34.
Here, we identify the energetic pathway used in GSCs. We find that GSCs rely heavily on glycolysis to generate ATP, since reducing the levels of glycolytic enzymes leads to GSC loss. However, differentiation is not prevented in GSCs with decreased glycolysis indicating that glycolysis is not the main energetic pathway in early differentiating cyst cells. These data raise the question of which energetic pathway is used by these early cyst cells. Based on mRNA expression profiling, it was previously shown that GSCs and their immediate progeny express high levels of arginine kinase suggesting that early differentiating cyst cells might employ a phosphagen system involving arginine-phosphate and arginine kinase to use energy provided from neighboring cells58. This specific metabolism could reflect the fact that selection of mitochondria for the next generation occurs in these early differentiating cyst cells59. Therefore, the use of a different energetic pathway might favor this key selection process without affecting cell fate.
Consistent with the importance of glycolysis to maintain GSC stemness, we show that increasing oxphos in GSC leads to their differentiation. Similarly, interfering with mitochondrial fusion or fission in GSCs was shown to induce their differentiation by increasing oxphos43.
Importantly, in aub mutant GSCs, the reduced levels of glycolytic enzymes and decreased lactate accumulation are accompanied by a metabolic rewiring towards oxphos. Mitochondrial maturation occurs precociously in these mutant GSCs with a large proportion of mitochondria containing cristae. Consistent with the role of ATP synthase in mitochondrial maturation, we also find increased levels of ATPsynα in aub mutant GSCs. Moreover, these mitochondria are active in generating ATP as measured by inner membrane potential. This metabolic switch in aub mutant GSCs could be indirect and reflect a metabolic adaptation to reduced glycolysis. In lines with this hypothesis, we found increased levels of ATPsynα in GSCs mutant for the Ald1 and Eno glycolytic enzymes. However, Aub does bind a large set of mRNAs18,31 and might also play a more general role in the metabolic control in the GSC lineage through the regulation of other metabolic mRNAs.
Our functional data reveal that Aub function in activating glycolysis substantially contributes to its role in GSC self-renewal, since increasing the levels of Phosphofructokinase, the rate limiting enzyme in glycolysis, significantly reduces the aub mutant phenotype of GSC loss. PIWI proteins are expressed in different populations of adult stem cells and were shown to be required for stem cell homeostasis. However, in most cases, the molecular mechanisms underlying this PIWI function remain uncharacterized. Since both a role of PIWI proteins in stemness and a metabolic reprogramming towards glycolysis are conserved features of stem cells, an intriguing possibility would be that the metabolic regulation by piRNAs and PIWI proteins might also be conserved in stem cells across species. In planarians, the PIWI protein SMEDWI-3 interacts with a large set of mRNAs, leading to either their decay or binding without decay60. This indicates a role of SMEDWI-3 in mRNA regulation that might extend to metabolic mRNAs.
Aub has been previously described to regulate target mRNAs directly involved in stem cell homeostasis. Aub represses Cbl mRNA in GSCs leading to reduced protein levels, a regulation required for GSC self-renewal19. In addition, Aub iCLIP in cultured GSCs identified several mRNAs encoding self-renewal and differentiation factors, among which dunce self-renewal mRNA was shown to be positively regulated by Aub in GSCs, whereas bam differentiation mRNA was activated by Aub in early differentiating cyst cells18. Therefore, Aub would have a dual function in regulating mRNAs involved in both metabolism and stemness thus coupling metabolism and development to secure stem cell fate. Similarly, coregulation of both metabolism and pluripotency genes has also been reported at the transcription level through direct binding of HIF-1 to their promoters47.
The importance of metabolism in piRNA biogenesis is intriguing. Both oxphos components and glycolytic enzymes play a role in piRNA production29,30. Here we show that metabolic rewiring in stem cells is regulated by Aub and piRNAs, thus revealing a crosstalk between energy metabolism and the piRNA pathway. Understanding how this crosstalk impacts on piRNA biology constitutes a major challenge for future studies.
Methods
Drosophila stocks
All Drosophila stocks used in this study are listed in Supplementary Table 2. Drosophila were raised at 25 °C on standard medium. The same numbers of flies were used for control and experimental crosses and kept to a maximum of 12 females and 6 males per vial. Crosses were transferred to fresh vials every three days. All crosses involving nos-Gal4 were performed using nos-Gal4 females.
DNA constructs and transgenic fly lines
To produce UASz-Laconic sensor, the pUASt-Laconic plasmid (a gift from T. Préat44) was digested with EcoRI and XbaI. The resulting fragment containing the Laconic sensor coding sequence was purified from agarose gel and cloned into the pBSII-SK vector digested with EcoRI and XbaI. This construct was then digested with XhoI and the resulting fragment was purified and cloned into the pUASz1.0 vector61 digested with XhoI. The resulting plasmid was validated by sequencing. Transgenic lines containing UASz-Laconic were generated using PhiC31-based integration into the attP-9A[VK00027] landing site on chromosome III, by BestGene Inc. (California). nosP-5′Eno-GFP-3’Eno and nosP-5′Eno-GFP-3′SV40 plasmids were generated by assembly of PCR fragments into the pNOSPE_MCP_eGFP vector62 digested by NotI and BamHI using the NEBuilder HiFi DNA Assembly Cloning Kit (NEB). pNOSPE_MCP_eGFP contains the nanos promoter, the mini-white gene and an attB site for PhiC31-mediated transformation. The primers used to generate these constructs are listed in Supplementary Table 3. The PCR fragments were amplified with Q5 High-Fidelity DNA Polymerase or Phusion® High-Fidelity DNA Polymerase and the resulting plasmids were validated by sequencing. Transgenic lines were generated using PhiC31-based integration into the attP2 landing site on chromosome III, by BestGene Inc. (California).
CRISPR/Cas9 editing
EnoΔpi11 and EnoΔpi12 were generated as follows. The donor plasmid pBS-Eno-Mutpi was generated by assembly of PCR fragments into pBS-SK+ digested by EcoRI and KpnI using the NEBuilder HiFi DNA Assembly Cloning Kit (NEB) and primers designed for amplifying ≈1 kb long homology arms from w1118 genomic sequence. This donor plasmid was designed to introduce 7 point mutations in the Quasimodo piRNA target site in Eno 5′UTR. The guide RNA producing plasmid pDCC6-Eno-gRNA was generated as follows. pDCC6 vector63 was digested by BbsI and ligated with two annealed primers corresponding to sgRNA in Eno 5′UTR. The primers used to generate these constructs are listed in Supplementary Table 3. PCR fragments were amplified with Q5 High-Fidelity DNA Polymerase or Phusion® High-Fidelity DNA Polymerase and the resulting plasmids were validated by sequencing. The donor plasmid and pCCD6-Eno-gRNA that also produces the Cas9 enzyme were co-injected into the w1118 stock at the Madrid Drosophila Transgenesis Facility (Centro de Biología Molecular Severo Ochoa). Injected flies were individually crossed to produce independent lines and these lines were screened by PCR using a primer corresponding to the mutant sequence in Eno 5′UTR. A CRISPR edited line containing this mutant sequence was not recovered, but two lines containing short deletions in this region were identified (EnoΔpi11 and EnoΔpi12). Eno sequence overlapping the donor plasmid was validated in these lines. The Enopimut1 allele was obtained as follows. A new donor plasmid pBS-Eno-Mutpi-DsRed was generated by inserting the scarless-DsRed cassette (PBac[3xP3-DsRed]) into an TTAA site in the right homology arm of pBS-Eno-Mutpi using the NEBuilder HiFi DNA Assembly Cloning Kit (NEB). pBS-Eno-Mutpi-DsRed was co-injected with pDCC6-Eno-gRNA into the Act5c-Cas9 DNAlig4(169) stock (BL #58492) by BestGene Inc. (California), and the CRISPR mutants were selected on the basis of DsRed fluorescent eyes by BestGene Inc. The DsRed cassette was removed by crosses with the PBac Transposase source stock (BL #8285). The mutant sequence was validated by sequencing the Eno region overlapping the donor plasmid.
Production of mitotic clones in ovaries
The Flipase under a heat shock promoter was used to induce mitotic clones in the germline of adult females. Clone induction was performed in adults to analyze to role of Ald1 and Eno in GSCs during adult stages. Ald1EY13155, EnoKG01162, Enof07543 and aubHN2 mutant germline clones were induced in 3 day-old females by two 37 °C heat shocks of 1 h spaced by an 8 h-recovery period at 25 °C, during three consecutive days. The flies were then maintained at 25 °C until ovarian dissection, either 3, 7, 14 or 21 days following the last heat shock.
Immunostaining
Ovaries were dissected in PBS followed by a fixation in 4% paraformaldehyde in PBT (PBS supplemented with 0.1% Tween 20) for 20 min at room temperature in rotation. Ovaries were then rinsed for 10 min in PBT, blocked with 10% bovine serum albumin (BSA) in PBT for 1 h and incubated with primary antibodies in PBT supplemented with 1% BSA overnight at 4 °C in rotation. The following primary antibodies were used at the indicated dilutions: mouse anti-Hts (1B1 clone, DSHB, RRID: AB528070) 1/50, rabbit anti-GFP (Invitrogen, Cat# A6455) 1/500, mouse anti-GFP (Invitrogen, Cat# A11120) 1/50, rabbit anti-Vasa (Santa Cruz Biotechnology, Cat# sc-30210) 1/200, rat anti-Vasa (DSHB, RRID: AB760351) 1/50, rabbit anti-cleaved Caspase 3 (Cell Signaling, Cat# 9661) 1/300, rabbit anti-PyK (Sigma, Cat# SAB4200105) 1/200, rabbit anti-Aldolase (Cell Signalling, Cat# 3188S) 1/100, mouse anti-Eno (Santa Cruz Biotechnology, Cat# sc-271384) 1/100, and mouse anti-Atp5A (Santa Cruz Biotechnology, Cat# sc-136178) 1/100. After incubation with primary antibodies, the ovaries were washed three times for 30 min in PBT-1% BSA and then incubated with fluorescent secondary antibodies diluted in PBT-0.1% BSA for 4 h at room temperature in rotation. After rising two times for 10 min in PBT, the ovaries were incubated in 0.1 μg/mL DAPI in PBT to stain DNA. They were then mounted in Vectashield mounting medium. Images were acquired using a Confocal Leica SP8 and analyzed with the ImageJ software. For immunostaining with anti-Atp5A, the fixation was with 5% formaldehyde in PBS for 25 min at room temperature in rotation. Ovaries were then washed with PBS followed by a permeabilization step with 1% Triton X-100 in PBS for 2 h, and blocking in PBS-1% BSA for 1 h.
Quantification of fluorescence intensity
Confocal images of germaria were analyzed using ImageJ. The cytoplasm of one GSC and one cyst cell in a same plane were delineated using the Freehand selections tool. The GFP fluorescence intensity was measured for each cell and this quantification was replicated three times. The mean of these three measurements was then calculated. The ratio: intensity in GSC/intensity in cyst cell was calculated by dividing the mean intensity value obtained for the GSC by the one obtained for the cyst cell in the same plane. The graphs were produced using the GraphPad Prism software.
Single-molecule FISH and quantification
For smFISH, ovaries were dissected in PBT and fixed in 4% paraformaldehyde for 1 h at room temperature. After two quick washes in PBS, ovaries were incubated overnight at 4 °C in 70% ethanol. Ovaries were incubated in Wash Buffer (10% Formamide and 2X SSC) for 20 min. Wash buffer was removed, replaced by pre-warmed Hybridization Buffer (10% Formamide, 2X SSC, 100 mg/ml dextran sulfate and Stellaris smFISH probes at 125 nM) and incubated overnight at 37 °C. Ovaries were washed thrice in pre-warmed Wash Buffer (DAPI was added at 1/1000 on the second wash) for 20 min, then incubated for 5 min in 2X SSC before addition of liquid anti-fade mounting medium (Abberior). The sequence of PyK smFISH probes are listed in Supplementary Table 4. For quantification of PyK mRNA foci/surface, confocal images of germaria were analyzed using the ImageJ software. From a maximum projection of 4 planes (z step of 0.3 µm), the cytoplasm of one GSC and two cyst cells were delineated using the Freehand selections tool and defined as ROIs. smFISH foci in each ROI were defined and counted using FIJI plugin ComDet v.0.5.5 (https://github.com/UU-cellbiology/ComDet). The number of foci counted was divided by the cytoplasm area to obtain the number of mRNA foci/surface. The graphs were produced using the GradPad Prism software.
Transmission electron microscopy
The ovaries were dissected in PBS pH 7.4 and immediately fixed in 2.5% Glutaraldehyde in PHEM buffer (60 mM PIPES, 25 mM HEPES, 10 mM EGTA, 4 mM MgSO4·7 H20) at room temperature for 1 h, and then overnight at 4 °C. Ovaries were post-fixed with 1% osmium tetroxide for 1 h at 4 °C and then en bloc stained with 1% uranyl acetate in double-distilled H2O at 4 °C for 1 h. Dehydration series were carried with out with successive ethanol baths at 30%, 50% and 70% at 4 °C, 85% at room temperature. To preserve mitochondrial crista structure, dehydration steps were limited to 5 min each. Ovaries were processed in a standard manner and embedded in Epoxy resin. 700 nm semi-thin sections were stained with 0.1% toluidine blue to evaluate the area of interest. 70 nm ultrathin sections were cut, mounted on formvar coated slotted copper grids and stained with 1.5% uranyl acetate in 70% ethanol and lead citrate. Stained grids were examined under a Tecnai G2 20 S-TWIN Transmission Electron Microscope.
Quantification of Laconic sensor FRET efficiency
Ovaries were rinsed twice in 75% ethanol and once in Schneider’s Insect Medium (Sigma) in a glass block for disinfection. They were then dissected in clean Schneider’s Insect Medium using forceps. The epithelial sheath was removed from several ovarioles using needles during no more than 20 min and ovarioles were then transferred to a drop of 20 μL of 10S oil on a high-resolution microscope slide. Confocal images and FRET efficiencies were acquired in vivo using a Confocal Leica SP8 microscope. Quantification of FRET efficiency was based on mTFP fluorescence before and after Venus photobleaching. The microscope settings were the following: for mTFP, excitation was at 458 nm and the detection window between 460 and 510 nm; for Venus, excitation was at 514 nm and the detection window between 560 and 620 nm. The bleaching was performed at 514 nm for 2 min. FRET efficiency was calculated as follows: (mTFP fluorescence intensity after Venus photobleaching)—(mTFP fluorescence intensity before Venus photobleaching)/(mTFP fluorescence intensity after Venus photobleaching). Fluorescence intensity was measured using ImageJ. Similar quantifications were obtained using the LEICA FRET-AB application.
Quantification of mitochondrial surface
Ovaries were dissected in Schneider’s Insect Medium (Sigma). They were stained with Red CMX Ros MitoTracker (Invitrogen M7512) at 500 nM for 30 min at room temperature with rotation. After three washes in PBS, ovaries were fixed in 4% paraformaldehyde in PBS for 20 min at room temperature. After two washes of 10 min in PBT, the ovaries were incubated in 0.1 μg/mL DAPI in PBT for 10 min at room temperature. The ovaries were washed three times for 10 min in PBS and mounted in Vectashield mounting medium. Images were acquired using a Leica SP8 Confocal microscope and analyzed using the ImageJ software. The cytoplasm of GSCs was delineated using the Freehand selections tool. The MitoTracker fluorescent signal over the threshold of 10% identified mitochondria and was defined as ROI; its total surface was calculated in µm2 using the Measure tool. The graphs were produced using the GradPad Prism software.
Quantification of mitochondrial membrane potential
Ovaries were dissected in Schneider’s Insect Medium (Sigma). They were stained with Deep Red FM MitoTracker (ThermoFisher Scientific) at 500 nM during 15 min at room temperature with rotation. TMRM (Biotium) was then added at 100 nM to the previous mix for 15 min at room temperature. After three washes in PBS, ovarioles were mounted in PBS on glass slides and immediately imaged using a Leica SP8 confocal microscope. Acquired images were analyzed using ImageJ. Ratiometric images were generated as follows. After inverting the colors for each channel and changing them to gray, the images were converted to 32-bit. Ratiometric images were then obtained by dividing the TMRM channel by the MitoTracker channel, using the Image Calculator function of ImageJ. The color of the resulting image was changed for Rainbow RGB. A scale bar and calibration bar were added. Quantification of mitochondrial membrane potential was performed as follows. The cytoplasm of one or two GSCs per germarium was delineated using the Freehand selections tool. The fluorescence intensity of the two channels, TMRM and MitoTracker, were measured and the value obtained for the TMRM channel was divided by the value obtained for the MitoTracker channel.
RNA-immunoprecipitations and RT-qPCR
For RNA-immunoprecipitation, ovaries from young females (1 to 3 days old) were dissected in PBS and kept at −80 °C. 100 pairs of ovaries per genotype were crushed in 600 μL DXB-150 (25 mM HEPES pH 6.8, 250 mM sucrose, 1 mM MgCl2, 1 mM DTT, 150 mM NaCl, 0.1% Triton X-100) containing cOmpleteTM EDTA-free Protease Inhibitor Cocktail (Roche) and RNase Inhibitor (0.25 U/μL; Promega). 25 μL of GFP-Trap beads (ProteinTech) were incubated with the extracts for 3 h on a wheel at 4 °C. The beads were then washed seven times with DXB-150 for 10 min at room temperature. The beads were mixed with 100 μL of DXB-150. 10 μL of beads were used to confirm immunoprecipitation by western blots; western blots were revealed using rabbit anti-GFP antibody (Invitrogen, Cat# A6455) at 1/1000. RNA was prepared with the remaining 90 μL of beads using TRIzol (Invitrogen), followed by DNA removal with TURBO DNA-free (Ambion). For RT-qPCR without immunoprecipitation, 10 pairs of ovaries were directly crushed in TRIzol and RNA was prepared as described above. 500 ng of total RNA was used for reverse transcription. qPCR was performed using Universal SYBR Green Master Mix and the LightCycler® 480 Instrument (Roche). Primers used for RT–qPCR are listed in Supplementary Table 5.
Bioinformatics and statistics
Quantification of glycolytic mRNAs using RNA-seq
We used published datasets (GSE119862, https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE119862)49 in which we selected RNA-seq performed in cultured GSCs and nos-Gal4/UASp-tkvM1 ovaries, which we analyzed using Salmon64 (https://combine-lab.github.io/salmon). Transcript quantification was done using --validateMappings option with the raw data on Drosophila melanogaster genome (FlyBase release 6.28). Each RNA-seq dataset was analyzed independently and glycolytic mRNAs were filtered using thresholds of minimum expression of 10 and 100 Transcripts Per Million (TPM) in all replicates.
Construction of piRNA library from small RNA-seq in cultured GSCs
The small RNA-seq datasets in cultured GSCs (GSE119862, https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE119862)49 were analyzed using a simplified homemade version of the pipeline piPipes65 (https://github.com/bowhan/piPipes). The raw data were first trimmed with Cutadapt66 (https://github.com/marcelm/cutadapt) as follow -a TGGAATTCTCGGGTGCCAAGGAACTCCAGTCACnnnnnnATCTCGTATGCCGTCTTCTGCTTG -m 18 -M 30 --discard-untrimmed. The alignment on the Drosophila melanogaster genome (FlyBase release 6.28) was performed using bowtie67 (https://bowtie-bio.sourceforge.net/index.shtml) with the configuration -v 0 --a --best --strata. All mapped reads were next annotated in the following order to rRNAs, snoRNAs, tRNAs, snRNAs, small mtRNAs and miRNAs using bowtie with the same parameters as above, except for rRNAs where only the best alignment was reported (-m 1). At each annotation step, unmapped reads were conserved (--un option) and used as input against the next small RNA species. All non-coding RNA references were obtained from diverse sources: NCBI, FlyBase (http://flybase.org) and RNAcentral (https://rnacentral.org). The remaining reads were then filtered by size, selecting only the 23-29 nt, and the pool was cleaned by collapsing identical sequences and attributing a unique identifier to each, along with the number of times a read had been retrieved in the whole pool. We got a library of 4,370,902 sequences of putative piRNAs. The library was then annotated by independent mappings against transposable elements and piRNA clusters (from piPipes) and Drosophila transcriptome (from FlyBase), with the following bowtie’s options (transposons: -v 3 -a --best --strata; piRNA clusters: -v 0 -m 1; transcriptome: -v 0 -m 1 --norc). Using a handmade script, annotations were next merged with sequence headers alongside the unique identifier. Unannotated piRNAs were kept and their annotation field complemented with a « 0 ».
piRNA targeting close to Aub binding sites in glycolytic mRNAs
Aub cross-link sites in cultured GSCs were obtained from published iCLIP experiments (GSE96751, https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE96751)18 and iCLIP genomic coordinates were converted to the dm6 release using the online converter of FlyBase. Using our piRNA library, bowtie was configured with different settings to identify potential target sites on transcripts. Nine targeting conditions were defined as follows: 0 to 3 mismatches within the piRNA sequence using option -v set up from 0 to 3; targeting of a 5′ seed of 16 nt without mismatch, 18 nt with 0 to 2 mismatches, or 20 nt with 1 mismatch using options -l and -n. For all targeting conditions, the first nt of the piRNA was excluded from the alignment using the −5 option. Each targeting was done on the whole Drosophila genome. All valid alignments were reported and when a seed was defined, the -e option was set to an arbitrary value of 2000, disabling the quality values. Using the BEDtools Suite68 (https://bedtools.readthedocs.io/en/latest/content/bedtools-suite.html), the alignment files were next converted into the bed format and piRNA coordinates were overlapped with protein coding genes coordinates obtained from FlyBase. Only antisense hits were reported using the –wa -wb -S options of intersectBED. Finally, antisense piRNAs closed to Aub cross-link sites were conserved using windowBED with a bin size of 60 nt in each direction of a cross-link site (-r 60 –l 60). Targeting results were then visualized and studied on the Integrative Genomics Viewer69 (https://software.broadinstitute.org/software/igv).
Calculation of MFE scores
We evaluated each piRNA-mRNA duplexes by calculating the Minimum Free Energy (MFE) using the RNAduplex tool from the ViennaRNA package70 (https://www.tbi.univie.ac.at/RNA/RNAduplex.1.html).
Statistics and reproducibility
Statistical tests were performed using GraphPad Prism. For each figure, the tests are indicated in the figure legend. The exact values of the tests are indicated the figure legends and/or in the Source Data file. Each experiment was repeated two to four times and quantifications were from two to four independent experiments, except those in Fig. 5i that were performed one to two times; a similar phenotype was reproduced with different alleles of the same gene in this experiment.
Ethics and inclusion statement
We support inclusive, diverse and ethical conduct in research.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
All data are available in the main text or Supplementary Information. Uncropped versions of gels are in Supplementary Fig. 9. Source data are provided with this paper.
References
Folmes, C. D. et al. Somatic oxidative bioenergetics transitions into pluripotency-dependent glycolysis to facilitate nuclear reprogramming. Cell Metab. 14, 264–271 (2011).
Zhu, S. et al. Reprogramming of human primary somatic cells by OCT4 and chemical compounds. Cell Stem Cell 7, 651–655 (2010).
Prigione, A., Fauler, B., Lurz, R., Lehrach, H. & Adjaye, J. The senescence-related mitochondrial/oxidative stress pathway is repressed in human induced pluripotent stem cells. Stem Cells 28, 721–733 (2010).
Yanes, O. et al. Metabolic oxidation regulates embryonic stem cell differentiation. Nat. Chem. Biol. 6, 411–417 (2010).
Chen, C. T., Shih, Y. R., Kuo, T. K., Lee, O. K. & Wei, Y. H. Coordinated changes of mitochondrial biogenesis and antioxidant enzymes during osteogenic differentiation of human mesenchymal stem cells. Stem Cells 26, 960–968 (2008).
Folmes, C. D. & Terzic, A. Energy metabolism in the acquisition and maintenance of stemness. Semin. Cell Dev. Biol. 52, 68–75 (2016).
Simsek, T. et al. The distinct metabolic profile of hematopoietic stem cells reflects their location in a hypoxic niche. Cell Stem Cell 7, 380–390 (2010).
Lisowski, P., Kannan, P., Mlody, B. & Prigione, A. Mitochondria and the dynamic control of stem cell homeostasis. EMBO Rep. 19 (2018).
Kreipke, E. R., Wang, Y., Miklas, J. W., Mathieu, J. & Ruohola-Baker, H. Metabolic remodeling in early development and cardiomyocyte maturation. Semin. Cell Dev. Biol. 52, 84–92 (2016).
Mahato, B. et al. Regulation of mitochondrial function and cellular energy metabolism by protein kinase C-lambda/iota: a novel mode of balancing pluripotency. Stem Cells 32, 2880–2892 (2014).
Ozata, D. M., Gainetdinov, I., Zoch, A., O’Carroll, D. & Zamore, P. D. PIWI-interacting RNAs: small RNAs with big functions. Nat. Rev. Genet. 20, 89–108 (2019).
Ramat, A. & Simonelig, M. Functions of PIWI proteins in gene regulation: new arrows added to the piRNA Quiver. Trends Genet.: TIG 37, 188–200 (2021).
Wang, X., Ramat, A., Simonelig, M. & Liu, M. F. Emerging roles and functional mechanisms of PIWI-interacting RNAs. Nat. Rev. Mol. Cell Biol. 24, 123–141 (2023).
Lin, H. & Spradling, A. C. A novel group of pumilio mutations affects the asymmetric division of germline stem cells in the Drosophila ovary. Development 124, 2463–2476 (1997).
Rojas-Rios, P. & Simonelig, M. piRNAs and PIWI proteins: regulators of gene expression in development and stem cells. Development 145, dev161786 (2018).
Cox, D. N. et al. A novel class of evolutionarily conserved genes defined by piwi are essential for stem cell self-renewal. Genes Dev. 12, 3715–3727 (1998).
Unhavaithaya, Y. et al. MILI, a PIWI-interacting RNA-binding protein, is required for germ line stem cell self-renewal and appears to positively regulate translation. J. Biol. Chem. 284, 6507–6519 (2009).
Ma, X. et al. Aubergine controls germline stem cell self-renewal and progeny differentiation via distinct mechanisms. Dev. Cell 41, 157–169.e155 (2017).
Rojas-Rios, P., Chartier, A., Pierson, S. & Simonelig, M. Aubergine and piRNAs promote germline stem cell self-renewal by repressing the proto-oncogene Cbl. EMBO J. 36, 3194–3211 (2017).
Tang, X., Liu, N., Qi, H. & Lin, H. Piwi maintains homeostasis in the Drosophila adult intestine. Stem Cell Rep. 18, 503–518 (2023).
Sousa-Victor, P. et al. Piwi is required to limit exhaustion of aging somatic stem cells. Cell Rep. 20, 2527–2537 (2017).
Juliano, C., Wang, J. & Lin, H. Uniting germline and stem cells: the function of Piwi proteins and the piRNA pathway in diverse organisms. Annu. Rev. Genet. 45, 447–469 (2011).
Kim, I. V., Riedelbauch, S. & Kuhn, C. D. The piRNA pathway in planarian flatworms: new model, new insights. Biol. Chem. 401, 1123–1141 (2020).
Ross, R. J., Weiner, M. M. & Lin, H. PIWI proteins and PIWI-interacting RNAs in the soma. Nature 505, 353–359 (2014).
Ge, D. T. et al. The RNA-Binding ATPase, Armitage, Couples piRNA Amplification in Nuage to Phased piRNA Production on Mitochondria. Mol. Cell 74, 982–995.e986 (2019).
Huang, H. et al. piRNA-associated germline nuage formation and spermatogenesis require MitoPLD profusogenic mitochondrial-surface lipid signaling. Dev. Cell 20, 376–387 (2011).
Munafo, M. et al. Daedalus and Gasz recruit Armitage to mitochondria, bringing piRNA precursors to the biogenesis machinery. Genes Dev. 33, 844–856 (2019).
Watanabe, T. et al. MITOPLD is a mitochondrial protein essential for nuage formation and piRNA biogenesis in the mouse germline. Dev. Cell 20, 364–375 (2011).
Handler, D. et al. The genetic makeup of the Drosophila piRNA pathway. Mol. cell 50, 762–777 (2013).
Gao, M. et al. Glycolytic enzymes localize to ribonucleoprotein granules in Drosophila germ cells, bind Tudor and protect from transposable elements. EMBO Rep. 16, 379–386 (2015).
Barckmann, B. et al. Aubergine iCLIP reveals piRNA-dependent decay of mRNAs involved in germ cell development in the early embryo. Cell Rep. 12, 1205–1216 (2015).
Clyne, P. J., Brotman, J. S., Sweeney, S. T. & Davis, G. Green fluorescent protein tagging Drosophila proteins at their native genomic loci with small P elements. Genetics 165, 1433–1441 (2003).
Teixeira, F. K. et al. ATP synthase promotes germ cell differentiation independent of oxidative phosphorylation. Nat. Cell Biol. 17, 689–696 (2015).
Wang, Z.H., Liu, Y., Chaitankar, V., Pirooznia, M. & Xu, H. Electron transport chain biogenesis activated by a JNK-insulin-Myc relay primes mitochondrial inheritance in Drosophila. eLife 8 (2019).
Bellen, H. J. et al. The BDGP gene disruption project: single transposon insertions associated with 40% of Drosophila genes. Genetics 167, 761–781 (2004).
Golic, K. G. & Lindquist, S. The FLP recombinase of yeast catalyzes site-specific recombination in the Drosophila genome. Cell 59, 499–509 (1989).
Monteiro, V. L., Safavian, D., Vasudevan, D. & Hurd, T. R. Mitochondrial remodelling is essential for female germ cell differentiation and survival. PLoS Genet. 19, e1010610 (2023).
Ohlstein, B. & McKearin, D. Ectopic expression of the Drosophila Bam protein eliminates oogenic germline stem cells. Development 124, 3651–3662 (1997).
Peters, S. J. Regulation of PDH activity and isoform expression: diet and exercise. Biochem Soc. Trans. 31, 1274–1280 (2003).
Herzig, R. P., Andersson, U. & Scarpulla, R. C. Dynein light chain interacts with NRF-1 and EWG, structurally and functionally related transcription factors from humans and drosophila. J. Cell Sci. 113, 4263–4273 (2000).
Scarpulla, R. C. Transcriptional paradigms in mammalian mitochondrial biogenesis and function. Physiol. Rev. 88, 611–638 (2008).
Tiefenbock, S. K., Baltzer, C., Egli, N. A. & Frei, C. The Drosophila PGC-1 homologue Spargel coordinates mitochondrial activity to insulin signalling. EMBO J. 29, 171–183 (2010).
Garcez, M., Branco-Santos, J., Gracio, P. C. & Homem, C. C. F. Mitochondrial dynamics in the Drosophila ovary regulates germ stem cell number, cell fate, and female fertility. Front Cell Dev. Biol. 8, 596819 (2020).
Placais, P. Y. et al. Upregulated energy metabolism in the Drosophila mushroom body is the trigger for long-term memory. Nat. Commun. 8, 15510 (2017).
San Martin, A. et al. A genetically encoded FRET lactate sensor and its use to detect the Warburg effect in single cancer cells. PLoS ONE 8, e57712 (2013).
Valentin, G. et al. Photoconversion of YFP into a CFP-like species during acceptor photobleaching FRET experiments. Nat. Methods 2, 801 (2005).
Wanet, A., Arnould, T., Najimi, M. & Renard, P. Connecting mitochondria, metabolism, and stem cell fate. Stem Cells Dev. 24, 1957–1971 (2015).
Villa-Fombuena, G., Lobo-Pecellin, M., Marin-Menguiano, M., Rojas-Rios, P. & Gonzalez-Reyes, A. Live imaging of the Drosophila ovarian niche shows spectrosome and centrosome dynamics during asymmetric germline stem cell division. Development 148 (2021).
Story, B. et al. Defining the expression of piRNA and transposable elements in Drosophila ovarian germline stem cells and somatic support cells. Life Sci. Alliance 2 (2019).
Halbach, R. et al. A satellite repeat-derived piRNA controls embryonic development of Aedes. Nature 580, 274–277 (2020).
Mor, I., Cheung, E. C. & Vousden, K. H. Control of glycolysis through regulation of PFK1: old friends and recent additions. Cold Spring Harb. Symp. Quant. Biol. 76, 211–216 (2011).
Chandel, N. S., Jasper, H., Ho, T. T. & Passegue, E. Metabolic regulation of stem cell function in tissue homeostasis and organismal ageing. Nat. Cell Biol. 18, 823–832 (2016).
Zhang, J. et al. UCP2 regulates energy metabolism and differentiation potential of human pluripotent stem cells. EMBO J. 30, 4860–4873 (2011).
Ramat, A. et al. The PIWI protein Aubergine recruits eIF3 to activate translation in the germ plasm. Cell Res. https://doi.org/10.1038/s41422-020-0294-9 (2020).
Dai, P. et al. A translation-activating function of MIWI/piRNA during mouse spermiogenesis. Cell 179, 1566–1581.e1516 (2019).
Toth, A. et al. Kinetic coupling of the respiratory chain with ATP synthase, but not proton gradients, drives ATP production in cristae membranes. Proc. Natl Acad. Sci. USA 117, 2412–2421 (2020).
Zick, M., Rabl, R. & Reichert, A. S. Cristae formation-linking ultrastructure and function of mitochondria. Biochim. Biophys. Acta 1793, 5–19 (2009).
Kai, T., Williams, D. & Spradling, A. C. The expression profile of purified Drosophila germline stem cells. Dev. Biol. 283, 486–502 (2005).
Lieber, T., Jeedigunta, S. P., Palozzi, J. M., Lehmann, R. & Hurd, T. R. Mitochondrial fragmentation drives selective removal of deleterious mtDNA in the germline. Nature 570, 380–384 (2019).
Kim, I. V. et al. Planarians recruit piRNAs for mRNA turnover in adult stem cells. Genes Dev. 33, 1575–1590 (2019).
DeLuca, S. Z. & Spradling, A. C. Efficient expression of genes in the Drosophila germline using a UAS promoter free of interference by Hsp70 piRNAs. Genetics 209, 381–387 (2018).
Dufourt, J. et al. Imaging translation dynamics in live embryos reveals spatial heterogeneities. Science 372, 840–844 (2021).
Gokcezade, J., Sienski, G. & Duchek, P. Efficient CRISPR/Cas9 plasmids for rapid and versatile genome editing in Drosophila. G3 4, 2279–2282 (2014).
Patro, R., Duggal, G., Love, M. I., Irizarry, R. A. & Kingsford, C. Salmon provides fast and bias-aware quantification of transcript expression. Nat. Methods 14, 417–419 (2017).
Han, B. W., Wang, W., Zamore, P. D. & Weng, Z. piPipes: a set of pipelines for piRNA and transposon analysis via small RNA-seq, RNA-seq, degradome- and CAGE-seq, ChIP-seq and genomic DNA sequencing. Bioinformatics 31, 593–595 (2015).
Martin, M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet. J. 17, 10–12 (2011).
Langmead, B., Trapnell, C., Pop, M. & Salzberg, S. L. Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 10, R25 (2009).
Quinlan, A. R. & Hall, I. M. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics 26, 841–842 (2010).
Robinson, J. T. et al. Integrative genomics viewer. Nat. Biotechnol. 29, 24–26 (2011).
Lorenz, R. et al. ViennaRNA Package 2.0. Algorithms Mol. Biol. 6, 26 (2011).
Acknowledgements
The authors are grateful to A. Arkov, T. Kai, T.R. Hurd and the Bloomington Drosophila Stock Center for providing Drosophila stocks. We thank P.Y. Plaçais, T. Préat and J.R. Huynh for their gifts of plasmids and R. Lopez-Alemany for her gift of antibodies. We thank M. Jault for participating in initial experiments shown in Fig. 2 and P.P. Marie for participating in experiments shown in Figs. 6 and 7, and in Supplementary Fig. 1 and 5. This work was supported by the CNRS-University of Montpellier UMR9002, ANR (ANR-15-CE12-0019-01, ANR-19-CE12-0031, ANR-21-CE12-0035-01), FRM (Equipe FRM EQU202303016322) and MSDAvenir. P.P. Marie held a salary from FRM. PRR held a salary from CNRS and VI PPIT-University of Seville. CE held a salary from ANR. JC held a salary from ANR and LabUM Montpellier, France 2030. CG held a salary from ANR and MSDAvenir; and AR held a salary from ANR, Fondation ARC and CNRS.
Author information
Authors and Affiliations
Contributions
P.R.R., A.C., C.E., C.G., A.B., and A.R. performed experiments and analyzed the data. J.C. performed bioinformatic analyses. M.S. conceptualized and supervised the study, and acquired funding; M.S. wrote the manuscript with inputs from all authors. All authors edited the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks the anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Source data
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Rojas-Ríos, P., Chartier, A., Enjolras, C. et al. piRNAs are regulators of metabolic reprogramming in stem cells. Nat Commun 15, 8405 (2024). https://doi.org/10.1038/s41467-024-52709-4
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467-024-52709-4