Electron transport chain complex I and mitochondrial fusion regulate ROS for differentiation in Drosophila neural stem cells Authors: Rahul Kumar Verma, Atharva Bhingare, Dnyanesh Dubal, Richa Rikhy Published: 2026 ABSTRACT SummaryMitochondrial fusion and electron transport chain complex I are each essential for differentiation in Drosophila neuroblasts, but the mechanism by which they interact to mediate differentiation is unknown. We found that complex I subunit depletion did not affect type II neuroblast numbers but reduced their proliferation and decreased their lineage cells. Complex I depletion decreased the mitochondrial membrane potential and cristae numbers, increased fragmentation and ROS, and inhibited Notch signaling in lineage cells. Similarly, antioxidant enzyme depletion increased ROS and reduced lineage cells. Both complex I and antioxidant proteins promoted the G1/S transition and nuclear cyclin E levels. Additional mitochondrial fusion via Drp1 mutants restored ROS levels, proliferation, and differentiation defects in complex I and antioxidant protein-depleted neuroblasts. Overexpression of antioxidant proteins and an increase in Notch signaling alleviated ROS and the complex I depletion-driven defect in neuroblast proliferation and differentiation. Complex I and mitochondrial fusion together restrict ROS to support neuroblast proliferation and differentiation. Introduction Mitochondrial dynamics and activity regulate stem cell differentiation. In mammals, pluripotent stem cells predominantly rely on glycolysis for energy production. They contain fragmented or globular mitochondria, with poorly organized cristae and perinuclear localization. These fragmented mitochondria have low electron transport chain (ETC) activity, decreased mitochondrial membrane potential (MMP), and decreased ATP production. During differentiation, mitochondria transition to oxidative phosphorylation (OXPHOS), facilitated by fused or tubular mitochondrial architecture with well-developed cristae. Fused mitochondria have high OXPHOS, MMP, and ATP production (Liu et al., 2013; Schieke et al., 2008; Son et al., 2013; Varum et al., 2009). An equilibrium between the fusion and fission events determines the mitochondrial architecture and, consequently, their activity. The family of large Dynamin GTPases orchestrates these events. Mitofusin (Mfn) in mammals, or Mitochondrial assembly regulatory factor (Marf) in Drosophila, drives outer mitochondrial membrane fusion, followed by inner mitochondrial membrane fusion through Optic atrophy 1 (Opa1). Dynamin-related protein 1 (Drp1) is essential for mitochondrial fragmentation. Mitochondrial fusion and ETC activity regulate proliferation and differentiation in Drosophila neural stem cells, also known as neuroblasts. Mitochondrial fusion driven by Opa1 is essential for proliferation and differentiation in the central brain neuroblasts (Dubal et al., 2022). An increase in Notch signaling leads to the excessive proliferation of neuroblasts, which also requires mitochondrial activity and fusion (Bowman et al., 2008; Dubal et al., 2022; Zhang et al., 2018). Although mitochondrial fusion and activity are crucial for neuroblast proliferation and differentiation, the mechanisms by which they interact to produce differentiation outcomes remain largely unexplored. In this study, we investigate the interaction between mitochondrial fusion and ETC complex I during the proliferation and differentiation of neuroblasts in the central brain of Drosophila. The central brain in the third instar larva contains type I and type II neuroblasts. The type I neuroblasts are approximately 90 in number, and the type II neuroblasts are 8 in number in each brain lobe. Notch signaling is activated in both type I and II neuroblasts, and it maintains fused mitochondrial morphology (Dubal et al., 2022; Rossi and Desplan, 2017). Type I neuroblasts express transcription factors Deadpan (Dpn) and Asense (Ase). They divide asymmetrically to give rise to one ganglion mother cell (GMC) each, which further divides symmetrically to give two neurons or glia. The type II neuroblast lineages contain transit-amplifying cells, similar to those found in mammalian systems. Type II neuroblasts express Dpn and an Ets transcription factor, pointed P1 (pnt) (Zhu et al., 2011). They undergo asymmetric division to produce immature intermediate neural progenitors (INPs) (Bello et al., 2008). Immature INPs undergo transcriptional changes to form mature INPs (mINPs), which express Dpn and Ase. mINPs divide asymmetrically 5–6 times, to form GMCs. GMCs in both type I and II lineages express Prospero (Pros) and Ase. GMCs eventually divide symmetrically to produce neurons or glia (Figure 1A). High nuclear Pros and Elav are expressed in neurons (Kelsom and Lu, 2012; Knoblich, 2010).Figure 1Complex I depletion reduces neuroblasts proliferation in the Drosophila third instar larval brain(A) Type I neuroblasts in the third instar larval Drosophila brain divide asymmetrically to self-renew and produce one GMC, which divides once to give two neurons or glia. Type II neuroblasts divide asymmetrically to self-renew and generate one immature INP. Mature INPs divide 4–5 times asymmetrically to self-renew and generate GMCs, which then divide symmetrically to form neurons or glia. Red-Dpn, Yellow-Ase, Blue-Low cytoplasmic Pros, Purple-High nuclear Pros. OL, optic lobe; VNC, ventral nerve cord; NB, neuroblast.(B) Representative images for mitotically dividing neuroblasts (yellow arrowheads) in one brain lobe of indicated genotypes. wor-Gal4 driven UAS-mCD8-GFP marks the plasma membrane of neuroblasts and their lineage (green). Miranda (red), and pH3 marks mitotically dividing cells (gray).(C) Neuroblast number quantification per brain lobe of genotypes: mCherryi (6 brains), ND-51i (3), ND-39i (4), and ND-19i (4).(D) Percent mitotic neuroblast number quantification per brain lobe of wor-Gal4 driven mCherryi (n = 607 neuroblasts, 6 brains), ND-51i (301, 3), ND-39i (473, 5), and ND-19i (605, 6).(E) Representative images of mitotically dividing type II neuroblasts (yellow arrowheads) in larval brain of the genotypes: pnt-Gal4, mCD8-GFP driven neuroblasts (blue arrowheads) and its lineage (green), and pH3 marks mitotic cells (gray).(F) Type II neuroblast number quantification per brain lobe of genotypes: mCherryi (20 brains), ND-51i (20), ND-39i (20), and ND-19i (20).(G) Percent mitotic type II neuroblasts numbers quantification per brain lobe expressing pnt-Gal4 driven mCherryi (n = 144 neuroblasts, 9 brains), ND-51i (176, 11), ND-39i (192, 12), and ND-19i (176, 11).Scale bars, 50 μm (B), 10 μm (E). Graphs show mean ± sd. Statistical analysis: Mann-Whitney test. ns = non-significant, ∗- p < 0.05 and ∗∗- p < 0.01.Also see Figure S1. Complex I depletion reduces neuroblasts proliferation in the Drosophila third instar larval brain (A) Type I neuroblasts in the third instar larval Drosophila brain divide asymmetrically to self-renew and produce one GMC, which divides once to give two neurons or glia. Type II neuroblasts divide asymmetrically to self-renew and generate one immature INP. Mature INPs divide 4–5 times asymmetrically to self-renew and generate GMCs, which then divide symmetrically to form neurons or glia. Red-Dpn, Yellow-Ase, Blue-Low cytoplasmic Pros, Purple-High nuclear Pros. OL, optic lobe; VNC, ventral nerve cord; NB, neuroblast. (B) Representative images for mitotically dividing neuroblasts (yellow arrowheads) in one brain lobe of indicated genotypes. wor-Gal4 driven UAS-mCD8-GFP marks the plasma membrane of neuroblasts and their lineage (green). Miranda (red), and pH3 marks mitotically dividing cells (gray). (C) Neuroblast number quantification per brain lobe of genotypes: mCherryi (6 brains), ND-51i (3), ND-39i (4), and ND-19i (4). (D) Percent mitotic neuroblast number quantification per brain lobe of wor-Gal4 driven mCherryi (n = 607 neuroblasts, 6 brains), ND-51i (301, 3), ND-39i (473, 5), and ND-19i (605, 6). (E) Representative images of mitotically dividing type II neuroblasts (yellow arrowheads) in larval brain of the genotypes: pnt-Gal4, mCD8-GFP driven neuroblasts (blue arrowheads) and its lineage (green), and pH3 marks mitotic cells (gray). (F) Type II neuroblast number quantification per brain lobe of genotypes: mCherryi (20 brains), ND-51i (20), ND-39i (20), and ND-19i (20). (G) Percent mitotic type II neuroblasts numbers quantification per brain lobe expressing pnt-Gal4 driven mCherryi (n = 144 neuroblasts, 9 brains), ND-51i (176, 11), ND-39i (192, 12), and ND-19i (176, 11). Scale bars, 50 μm (B), 10 μm (E). Graphs show mean ± sd. Statistical analysis: Mann-Whitney test. ns = non-significant, ∗- p < 0.05 and ∗∗- p < 0.01. Also see Figure S1. ETC complex I, known as NADH dehydrogenase or NADH:ubiquinone oxidoreductase, is a multi-subunit enzyme with a 1000 kDa molecular weight that catalyzes the transfer of electrons from NADH to ubiquinone. Complex I can generate ROS during electron transfer in the ETC. It is an L-shaped structure, consisting of one hydrophobic arm embedded in the inner mitochondrial membrane and one hydrophilic arm present at the periphery, facing the matrix (Agip et al., 2023; Garcia et al., 2017; Guarani et al., 2014). The Drosophila complex I is composed of 43 subunits, a catalytic core of 14 evolutionarily conserved subunits alongside 29 accessory subunits (Agip et al., 2023; Garcia et al., 2017). Although these accessory subunits are not directly involved in the catalytic process, their perturbation can disrupt the assembly of the entire complex, leading to elevated ROS generation (Berger et al., 2008; Stroud et al., 2016). Electrons from complex I are transferred to complex III and then to IV, where molecular oxygen is reduced to water. This electron flow is coupled to the pumping of protons from the mitochondrial matrix to the intermembrane space by complex I, II, and IV, establishing an electrochemical gradient across the inner mitochondrial membrane, which manifests as the MMP. Complex V (ATP Synthase) utilizes the MMP to synthesize ATP from ADP. In this study, we found that the depletion of complex I subunits led to decreased proliferation in the central brain neuroblasts. There was a decrease in proliferation and differentiation in the type II neuroblast lineage, characterized by a reduction in the numbers of mINPs, GMCs, and neurons. Notably, the downregulation of complex I led to fragmented mitochondria, decreased cristae numbers, a low MMP, increased ROS generation, and decreased Notch signaling. These elevated ROS levels led to a decrease in differentiation by disrupting the transition from the G1 to S phase in the cell cycle. Inducing mitochondrial fusion or increasing ROS scavengers or increasing Notch signaling in complex I-depleted type II neuroblasts reversed the ROS levels and restored the differentiation defects in the type II neuroblast lineage. Results ETC complex I depletion leads to a decrease in the proliferation and differentiation of neuroblasts To evaluate the function of ETC complex I in neuroblast proliferation and differentiation, we used RNAi to deplete individual subunits in type I and II neuroblasts. We combined worniu-Gal4 (wor-Gal4), mCD8-GFP, and pointed P1-Gal4 (pnt-Gal4), mCD8-GFP with shRNA lines targeting complex I subunits (Figure S1A). mCD8-GFP marked the plasma membrane in the shRNA-expressing cells. RNAi depletion targeted core hydrophilic arm subunits, ND-49 (mammalian NDUFS2) and ND-51 (NDUFV1), and ND-19 (NDUFA8), ND-39 (NDUFA9), ND-42 (NDUFA10), and accessory hydrophobic arm subunits ND-PDSW (NDUFB10). When expressed with wor-Gal4 at 29°C, we observed pupal lethality on the depletion of ND-PDSW, ND-49, and ND-42, with the few flies that emerged from pupae being sluggish in behavior. Depletion of ND-51, ND-39, and ND-19 gave a more severe phenotype of complete pupal lethality (Figure S1A). Depletion of complex I subunits specifically in type II neuroblasts using pnt-Gal4 did not affect the development of F1 flies (Homem et al., 2014). Flies emerged normally out of pupae but were less active and sluggish and died within 15 days after eclosion (Figure S1A). The shRNA lines against ND-51 (ND-51i), ND-39 (ND-39i), and ND-19 (ND-19i) were selected for further analysis of the role of complex I in neuroblast proliferation and differentiation in the larval stage. To assess the impact of ND-51, ND-39, and ND-19 depletion on central brain neuroblasts, we crossed shRNA lines with wor-Gal4 and immunostained with Miranda and phospho-Histone 3 (pH3), a marker for dividing cells. Neuroblast numbers remained unchanged between the controls and brains depleted of ND-51, ND-39, and ND-19 (Figures 1B and 1C), but the proportion of mitotically active neuroblasts reduced significantly within each brain lobe (Figures 1B and 1D). Knockdown of ND-51, ND-39, and ND-19 in type II neuroblasts using the pnt-Gal4 driver also yielded unchanged neuroblast numbers but fewer mitotically dividing neuroblasts (Figures 1E–1G). Live imaging of third instar larval brains expressing ND-51i in type II neuroblasts showed decreased division. Control neuroblasts divided at least once during live imaging in 40 min, but ND-51-depleted neuroblasts did not show a division in the same duration (Figure S1B). Collectively, these observations show that complex I is a crucial regulator of neuroblast proliferation. To evaluate the impact of complex I depletion on differentiation in the type II neuroblast lineage, we quantified the numbers of mINPs, GMCs, and neurons by immunostaining for specific markers. The number of mINPs (Ase+, Dpn+) decreased significantly in ND-51i, ND-39i, and ND-19i expressing lineages as compared to mCherryi controls (Figures 2A and 2B). Depletion of ND-51, ND-39, and ND-19 resulted in a decreased number of dividing mINPs (Dpn+) labeled with pH3 within the lineage compared to controls (Figures 2D and 2E). This reduction in mINP proliferation corresponded to a decrease in GMC (Ase+ Dpn-) numbers (Figures 2A and 2C), subsequently leading to a diminished production of neurons (high nuclear Pros) (Figures 2F and 2G). We ruled out apoptosis as a cause for the decrease in lineage cells by immunostaining for cleaved Caspase 3. No discernible difference was found between control and complex I-depleted neuroblast lineages (Figure S2A). An independent shRNA targeting ND-51i (BL29534) confirmed the specificity of the RNAi, showing a similar and significant decrease in mINPs, GMCs, and neurons in the pnt-Gal4 lineage (Figures S2B-S2F).Figure 2Complex I depletion in type II neuroblasts decreases lineage cells(A) Representative images of type II neuroblast lineages of the mentioned genotypes. mINPs are Dpn+ (gray). GMCs are Dpn-/Ase+ (red) (yellow arrowheads).(B) mINP number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 28 type II neuroblast lineages, 8 brains), ND-51i (12, 3), ND-39i (20, 11), and ND-19i (14, 9).(C) GMC number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 24 type II neuroblast lineages, 11 brains), ND-51i (17, 9), ND-39i (9, 5), and ND-19i (13, 9).(D) Representative images showing proliferating mINPs in the type II neuroblast lineages of the mentioned genotypes. Dpn in the lineage marks mINPs (red). pH3-labeled cells mark mitotically dividing cells (gray).(E) Mitotic mINP number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 16 type II neuroblast lineages, 5 brains), ND-51i (7, 5), ND-39i (18, 5), and ND-19i (18, 5).(F) Representative images of high nuclear Pros positive neurons in type II neuroblast lineage (red).(G) Neuron number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 19 type II neuroblast lineages, 5 brains), ND-51i (15, 3), ND-39i (5, 3), and ND-19i (20, 4).Scale bars, 10 μm. Graphs show mean ± sd. Statistical analysis: Mann-Whitney test. ∗- p < 0.05, ∗∗- p < 0.01, and ∗∗∗- p < 0.001.Also see Figure S2. Complex I depletion in type II neuroblasts decreases lineage cells (A) Representative images of type II neuroblast lineages of the mentioned genotypes. mINPs are Dpn+ (gray). GMCs are Dpn-/Ase+ (red) (yellow arrowheads). (B) mINP number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 28 type II neuroblast lineages, 8 brains), ND-51i (12, 3), ND-39i (20, 11), and ND-19i (14, 9). (C) GMC number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 24 type II neuroblast lineages, 11 brains), ND-51i (17, 9), ND-39i (9, 5), and ND-19i (13, 9). (D) Representative images showing proliferating mINPs in the type II neuroblast lineages of the mentioned genotypes. Dpn in the lineage marks mINPs (red). pH3-labeled cells mark mitotically dividing cells (gray). (E) Mitotic mINP number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 16 type II neuroblast lineages, 5 brains), ND-51i (7, 5), ND-39i (18, 5), and ND-19i (18, 5). (F) Representative images of high nuclear Pros positive neurons in type II neuroblast lineage (red). (G) Neuron number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 19 type II neuroblast lineages, 5 brains), ND-51i (15, 3), ND-39i (5, 3), and ND-19i (20, 4). Scale bars, 10 μm. Graphs show mean ± sd. Statistical analysis: Mann-Whitney test. ∗- p < 0.05, ∗∗- p < 0.01, and ∗∗∗- p < 0.001. Also see Figure S2. In summary, we find that the reduction in type II neuroblast and mINP proliferation resulting from complex I depletion led to a decrease in the proliferation of transit-amplifying cells and differentiated cells. Drp1 mutant expression in complex I-deficient type II neuroblasts alleviates their proliferation and differentiation defect Depletion of ETC complex I and complex V in type I neuroblasts causes mitochondrial fragmentation (van den Ameele and Brand, 2019). Similarly, Opa1 depletion causes mitochondrial fragmentation and proliferation and differentiation defects in Drosophila type II neuroblasts (Dubal et al., 2022) and intestinal stem cells (Deng et al., 2018). Drp1 activity is essential for driving mitochondrial fragmentation in Opa1-depleted cells (Chowdhary et al., 2017; Deng et al., 2018; Trevisan et al., 2018). Additionally, concurrent depletion of Drp1 with Opa1 restores mitochondrial fusion and alleviates the differentiation defects, underscoring the necessity of a fused mitochondrial morphology for proper stem cell differentiation (Deng et al., 2018; Dubal et al., 2022). This precedent led us to investigate mitochondrial morphology in complex I-depleted type II neuroblasts and the effect of increasing mitochondrial fusion on differentiation. We assessed mitochondrial morphology in ND-51-depleted type II neuroblasts and muscle cells by crossing the shRNA lines with pnt-Gal4 and mhc-Gal4, respectively. ND-51 depletion caused expected mitochondrial fragmentation in type II neuroblasts and also showed relatively smaller mitochondria in muscle cells (Figure 3A). To test if Drp1 depletion could reverse the mitochondrial morphology defects, we co-expressed the dominant-negative form of Drp1 (Drp1SD) with complex I RNAi (Drp1SD;ND-51i) and used Drp1SD;mCherryi as a Gal4 dilution control, crossing them to mito-GFP;pnt-Gal4 and mhc-Gal4. Mitochondria in Drp1SD;mCherryi neuroblasts were clustered and localized to one side of the cell as compared to mCherryi alone. Drp1SD;mCherryi expression in muscle cells led to larger and interconnected mitochondria as compared to mCherryi alone. Drp1SD;ND-51i neuroblasts contained a mixed population of clustered mitochondria localized to one side of the cell and fragmented mitochondria. Muscle cells expressing Drp1SD;ND-51i showed larger mitochondria than ND-51i alone (Figure 3A).Figure 3Drp1SD expression in complex I-depleted neuroblasts shows mitochondrial fusion and alleviates the proliferation and differentiation defects(A) Representative images of mito-GFP in type II neuroblasts and ATPB immunostaining in third instar larval muscle of the indicated genotypes. ND-51i induces mitochondrial fragmentation in neuroblasts and muscle. Drp1SD;mCherryi neuroblasts show clustered mitochondria localized to one side (yellow arrowhead), Drp1SD;ND-51i neuroblasts exhibit clustered (yellow arrowhead) and fewer punctuated mitochondria (blue arrowhead).(B) Representative images of mitotic pH3-positive neuroblasts (yellow arrowheads) in the larval brain lobe of indicated genotypes. wor-Gal4, UAS-mCD8-GFP marks neuroblasts and their lineage (green). Miranda (red), pH3 (gray).(C) Quantification of percent mitotic neuroblasts per brain lobe of third-instar larvae, expressing wor-Gal4 driven mCherryi (n = 295 neuroblasts, 3 brains), Drp1SD; mCherryi (476, 5), ND-51i (267, 3), Drp1SD; ND-51i (488, 5).(D) Representative images show pH3-positive type II neuroblasts (yellow arrowheads) in the larval brain of the indicated genotypes. pnt-Gal4 driven mCD8-GFP marks neuroblasts (blue arrowheads) and their lineage (green), and pH3 (gray).(E) Percent pH3 positive type II neuroblast number quantification per brain lobe of genotype, mCherryi (n = 144 neuroblasts, 9 brains), Drp1SD; mCherryi (112, 7), ND-51i (160, 10), Drp1SD; ND-51i (200, 13).(F) Representative images of immunostained mINPs and GMCs in the type II neuroblast lineage of the indicated genotypes. mINPs are Dpn+ (gray). GMCs are Dpn-/Ase+ (red) (yellow arrowheads).(G) mINP number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 28 type II neuroblast lineages, 11 brains), Drp1SD; mCherryi (20, 7), ND-51i (12, 9), Drp1SD; ND-51i (12, 6).(H) GMC number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 12 type II neuroblast lineages, 3 brains), Drp1SD; mCherryi (18, 5), ND-51i (14, 3), Drp1SD; ND-51i (15, 5).(I) Representative images of neurons with high nuclear Pros (red) in type II neuroblast lineages of indicated genotypes.(J) Neuron number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 10 type II neuroblast lineages, 5 brains), Drp1SD; mCherryi (13, 5), ND-51i (8, 5), Drp1SD; ND-51i (10, 3).Scale bars, 5 μm (A), 50 μm (B), 10 μm (D, F, and I). Graphs show mean ± sd. Statistical analysis: Mann-Whitney test. ns = not significant, ∗- p < 0.05, and ∗∗∗- p < 0.001.Also see Figures S2 and S3. Drp1SD expression in complex I-depleted neuroblasts shows mitochondrial fusion and alleviates the proliferation and differentiation defects (A) Representative images of mito-GFP in type II neuroblasts and ATPB immunostaining in third instar larval muscle of the indicated genotypes. ND-51i induces mitochondrial fragmentation in neuroblasts and muscle. Drp1SD;mCherryi neuroblasts show clustered mitochondria localized to one side (yellow arrowhead), Drp1SD;ND-51i neuroblasts exhibit clustered (yellow arrowhead) and fewer punctuated mitochondria (blue arrowhead). (B) Representative images of mitotic pH3-positive neuroblasts (yellow arrowheads) in the larval brain lobe of indicated genotypes. wor-Gal4, UAS-mCD8-GFP marks neuroblasts and their lineage (green). Miranda (red), pH3 (gray). (C) Quantification of percent mitotic neuroblasts per brain lobe of third-instar larvae, expressing wor-Gal4 driven mCherryi (n = 295 neuroblasts, 3 brains), Drp1SD; mCherryi (476, 5), ND-51i (267, 3), Drp1SD; ND-51i (488, 5). (D) Representative images show pH3-positive type II neuroblasts (yellow arrowheads) in the larval brain of the indicated genotypes. pnt-Gal4 driven mCD8-GFP marks neuroblasts (blue arrowheads) and their lineage (green), and pH3 (gray). (E) Percent pH3 positive type II neuroblast number quantification per brain lobe of genotype, mCherryi (n = 144 neuroblasts, 9 brains), Drp1SD; mCherryi (112, 7), ND-51i (160, 10), Drp1SD; ND-51i (200, 13). (F) Representative images of immunostained mINPs and GMCs in the type II neuroblast lineage of the indicated genotypes. mINPs are Dpn+ (gray). GMCs are Dpn-/Ase+ (red) (yellow arrowheads). (G) mINP number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 28 type II neuroblast lineages, 11 brains), Drp1SD; mCherryi (20, 7), ND-51i (12, 9), Drp1SD; ND-51i (12, 6). (H) GMC number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 12 type II neuroblast lineages, 3 brains), Drp1SD; mCherryi (18, 5), ND-51i (14, 3), Drp1SD; ND-51i (15, 5). (I) Representative images of neurons with high nuclear Pros (red) in type II neuroblast lineages of indicated genotypes. (J) Neuron number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 10 type II neuroblast lineages, 5 brains), Drp1SD; mCherryi (13, 5), ND-51i (8, 5), Drp1SD; ND-51i (10, 3). Scale bars, 5 μm (A), 50 μm (B), 10 μm (D, F, and I). Graphs show mean ± sd. Statistical analysis: Mann-Whitney test. ns = not significant, ∗- p < 0.05, and ∗∗∗- p < 0.001. Also see Figures S2 and S3. We assessed the proliferation of Drp1SD;ND-51i expressing neuroblasts using wor-Gal4 and pnt-Gal4. In each brain lobe, the proportion of mitotically dividing neuroblasts significantly increased in Drp1SD;ND-51i compared to ND-51i alone (Figures 3B–3E). Next, we analyzed how Drp1SD affected differentiation in the ND-51i expressing type II neuroblast lineage. Consistent with the suppression of the proliferation defects in type II neuroblasts (Figures 3D and 3E), the number of mINPs (Figures 3F and 3G) and mitotically dividing mINPs (Figures S3A and S3B) increased in Drp1SD;ND-51i compared to ND-51i alone. This increased division of mINPs in Drp1SD;ND-51i type II neuroblast lineage resulted in the rescue of the number of GMCs (Figures 3F and 3H), and the partial rescue of the number of neurons (Figures 3I and 3J). To confirm that the rescue of differentiation defects upon Drp1SD expression was not due to Gal4 dilution, we expressed UAS-mito-GFP with ND-51i and found that the numbers of mINPs, GMCs, and neurons were comparable to ND-51i alone (Figures S2B–S2F). In summary, mitochondrial fusion suppresses the complex I depletion-induced differentiation defects in the type II neuroblast lineage. Drp1 mutants suppress the cell cycle defects in complex I knockdown type II neuroblasts To determine the cellular mechanism by which mitochondrial fusion restores differentiation in complex I-depleted type II neuroblast lineages, we examined cell cycle progression, as mitochondrial activity and dynamics are known regulators of G1/S phase transition. Depletion of complex I in Drosophila eye discs causes a delay in the G1/S transition and a decrease in cyclin E, mediated by high ROS levels and JNK-pathway activity (Mandal et al., 2010; Owusu-Ansah et al., 2008). Drp1 depletion also affects cyclin E levels in Drosophila follicle cells (Mitra et al., 2012; Parker et al., 2015). We assessed the impact of complex I depletion on G1-S progression and cyclin E accumulation in the nucleus of type II neuroblasts. We used EdU (5-ethynyl-2′-deoxyuridine), a thymidine analog, to mark S-phase cells. ND-51i expressing type II neuroblast lineages showed a decreased number of EdU-positive cells in each lineage and decreased percentage of mINPs positive for EdU compared to controls (Figures 4A–4C). This decrease was restored in Drp1SD;ND-51i lineages. Nuclear cyclin E accumulation was reduced in ND-51i expressing neuroblasts compared to controls, and this was rescued in Drp1SD;ND-51i expressing neuroblasts (Figures 4D and 4E). Therefore, complex I disruption induces G1/S phase transition defects by decreasing nuclear cyclin E in type II neuroblasts.Figure 4Cell cycle defects in complex I-depleted type II neuroblasts are rescued by Drp1SD expression(A) Representative images showing EdU-positive cells (red) and mINPs (gray) in type II neuroblast lineages of the mentioned genotypes.(B) EdU positive cell number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 19 type II neuroblast lineages, 6 brains), Drp1SD;mCherryi (13, 5), ND-51i (23, 6), Drp1SD;ND-51i (20, 4).(C) Percent EdU positive mINPs (yellow arrowheads) quantification in type II neuroblast lineages of genotypes: mCherryi (n = 12 type II neuroblast lineages, 6 brains), Drp1SD;mCherryi (11, 5), ND-51i (15, 6), Drp1SD;ND-51i (13, 4).(D) Representative images of cyclin E immunostaining (red) in type II neuroblasts (encircled by white dashed line) of indicated genotypes.(E) Cyclin E nuclear to cytoplasmic intensity quantification in type II neuroblasts of the genotypes: mCherryi (n = 62 type II neuroblasts, 9 brains), Drp1SD;mCherryi (28, 3), ND-51i (20, 3), Drp1SD;ND-51i (19, 3).Scale bars, 10 μm. Graphs show mean ± sd. Statistical analysis: Mann-Whitney test. ns = not significant, ∗∗- p < 0.01 and ∗∗∗- p < 0.001. Cell cycle defects in complex I-depleted type II neuroblasts are rescued by Drp1SD expression (A) Representative images showing EdU-positive cells (red) and mINPs (gray) in type II neuroblast lineages of the mentioned genotypes. (B) EdU positive cell number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 19 type II neuroblast lineages, 6 brains), Drp1SD;mCherryi (13, 5), ND-51i (23, 6), Drp1SD;ND-51i (20, 4). (C) Percent EdU positive mINPs (yellow arrowheads) quantification in type II neuroblast lineages of genotypes: mCherryi (n = 12 type II neuroblast lineages, 6 brains), Drp1SD;mCherryi (11, 5), ND-51i (15, 6), Drp1SD;ND-51i (13, 4). (D) Representative images of cyclin E immunostaining (red) in type II neuroblasts (encircled by white dashed line) of indicated genotypes. (E) Cyclin E nuclear to cytoplasmic intensity quantification in type II neuroblasts of the genotypes: mCherryi (n = 62 type II neuroblasts, 9 brains), Drp1SD;mCherryi (28, 3), ND-51i (20, 3), Drp1SD;ND-51i (19, 3). Scale bars, 10 μm. Graphs show mean ± sd. Statistical analysis: Mann-Whitney test. ns = not significant, ∗∗- p < 0.01 and ∗∗∗- p < 0.001. Mitochondrial fragmentation and activity defects in complex I-depleted type II neuroblasts are rescued by Drp1SD mutant expression ETC disruption causes mitochondrial fragmentation, decreased activity, and altered metabolites in neuroblasts (Bonnay et al., 2020; Yao et al., 2019). To evaluate the changes in mitochondrial activity and morphology in complex I-depleted type II neuroblasts, we measured the changes in MMP, cristae organization, and ROS generation to assess mitochondrial functionality. We used the potentiometric dye, tetra-methyl-rhodamine methyl ester (TMRM), to assess MMP in living third-instar larval brains. TMRM fluorescence was measured in type II neuroblasts and expressed as a ratio to neighboring control type I neuroblasts. We observed a significant reduction in MMP in ND-51i and ND-19i expressing type II neuroblasts (Figures 5A and 5B). Mitochondrial fusion by Drp1SD expression did not alter the MMP, and co-expression of Drp1SD with ND-51i and ND-19i in type II neuroblasts restored the MMP.Figure 5Complex I depletion decreases MMP and cristae and increases ROS in type II neuroblasts, and Drp1SD expression alleviates these defects(A) Representative images of TMRM fluorescence in type II neuroblasts (red, circled by white dashed line). Control type I neuroblasts are shown by yellow arrowheads.(B) TMRM fluorescence quantification in type II neuroblasts, normalized to the neighboring type I neuroblast of genotypes: mCherryi (n = 19 type II neuroblasts, 5 brains), Drp1SD;mCherryi (16, 5), ND-51i (17, 9), Drp1SD;ND-51i (28, 5), ND-19i (34, 6), Drp1SD;ND-19i (22, 4).(C) Representative electron micrographs of neuroblasts of the genotypes: mCherryi, Drp1SD;mCherryi, ND-51i and Drp1SD;ND-51i.(D) Mitochondrial area quantification in genotypes: mCherryi (n = 76 mitochondria, 10 neuroblasts, 3 brains), Drp1SD;mCherryi (66, 8, 3), ND-51i (247, 12, 3), Drp1SD;ND-51i (138, 16, 3).(E) Cristae number quantification in genotypes: mCherryi (n = 39 mitochondria, 10 neuroblasts, 3 brains), Drp1SD;mCherryi (20, 8, 3), ND-51i (85, 12, 3), Drp1SD;ND-51i (60, 16, 3).(F) Representative images showing DHE fluorescence (red, encircled by a white dashed line). Control type I neuroblasts are marked by yellow arrowheads.(G) DHE fluorescence quantification in type II neuroblast normalized to surrounding type I neuroblast of the genotypes: mCherryi (n = 9 type II neuroblasts, 5 brains), Drp1SD;mCherryi (11, 5), ND-51i (32, 7), Drp1SD;ND-51i (30, 5), ND-19i (10, 3), Drp1SD;ND-19i (26, 5).Scale bars, 10 μm (A and F), 2 μm (C), and 0.5 μm (C, zoomed-in view) . Graphs show mean ± sd. Statistical analysis: Mann-Whitney test. ns = not significant, ∗∗- p < 0.01 and ∗∗∗- p < 0.001. Also see Figure S4. Complex I depletion decreases MMP and cristae and increases ROS in type II neuroblasts, and Drp1SD expression alleviates these defects (A) Representative images of TMRM fluorescence in type II neuroblasts (red, circled by white dashed line). Control type I neuroblasts are shown by yellow arrowheads. (B) TMRM fluorescence quantification in type II neuroblasts, normalized to the neighboring type I neuroblast of genotypes: mCherryi (n = 19 type II neuroblasts, 5 brains), Drp1SD;mCherryi (16, 5), ND-51i (17, 9), Drp1SD;ND-51i (28, 5), ND-19i (34, 6), Drp1SD;ND-19i (22, 4). (C) Representative electron micrographs of neuroblasts of the genotypes: mCherryi, Drp1SD;mCherryi, ND-51i and Drp1SD;ND-51i. (D) Mitochondrial area quantification in genotypes: mCherryi (n = 76 mitochondria, 10 neuroblasts, 3 brains), Drp1SD;mCherryi (66, 8, 3), ND-51i (247, 12, 3), Drp1SD;ND-51i (138, 16, 3). (E) Cristae number quantification in genotypes: mCherryi (n = 39 mitochondria, 10 neuroblasts, 3 brains), Drp1SD;mCherryi (20, 8, 3), ND-51i (85, 12, 3), Drp1SD;ND-51i (60, 16, 3). (F) Representative images showing DHE fluorescence (red, encircled by a white dashed line). Control type I neuroblasts are marked by yellow arrowheads. (G) DHE fluorescence quantification in type II neuroblast normalized to surrounding type I neuroblast of the genotypes: mCherryi (n = 9 type II neuroblasts, 5 brains), Drp1SD;mCherryi (11, 5), ND-51i (32, 7), Drp1SD;ND-51i (30, 5), ND-19i (10, 3), Drp1SD;ND-19i (26, 5). Scale bars, 10 μm (A and F), 2 μm (C), and 0.5 μm (C, zoomed-in view) . Graphs show mean ± sd. Statistical analysis: Mann-Whitney test. ns = not significant, ∗∗- p < 0.01 and ∗∗∗- p < 0.001. Also see Figure S4. Fused mitochondria support well-organized lamellar cristae, which expand the inner membrane surface area, concentrating ETC components for efficient OXPHOS and higher MMP (Gilkerson et al., 2003; Ikon and Ryan, 2017; Song et al., 2013; Vogel et al., 2006). Cristae disruption may explain the reduced MMP upon complex I depletion. We examined cristae organization using transmission electron microscopy on brains expressing ND-51i and mCherryi with wor-Gal4. Control neuroblasts have relatively elongated mitochondria with cristae-rich morphology. ND-51i neuroblasts possess small, fragmented mitochondria with less cross-sectional area. The ND-51i neuroblasts have fewer and poorly organized cristae than controls (Figures 5C–5E). In Drp1SD;ND-51i expressing type II neuroblasts, mitochondrial morphology and cristae number were similar to Drp1SD;mCherryi controls (Figures 5C–5E). Thus, MMP rescue by mitochondrial fusion coincides with restored mitochondrial morphology and cristae architecture. Since ETC inhibition induces cellular ATP stress and increased phosphorylated AMP-activated protein kinase (pAMPK) in Drosophila embryos (Chowdhary et al., 2017) and pAMPK increases after 2-deoxyglucose incubation in larval brains (Dubal et al., 2022), we checked if complex I depletion decreased ATP. We assessed ATP levels in ND-51i type II neuroblasts using the BioTracker ATP-Red Live Cell Dye. The ATP-Red dye fluorescence was found to be comparable between ND-51i and control type II neuroblasts (Figures S4A and S4B). Additionally, pAMPK in ND-51i type II neuroblasts was similar to controls, suggesting that ATP stress was not seen in these cells (Figures S4C and S4D). Alterations in mitochondrial morphology and ETC change ROS levels in the mitochondria and the cytoplasm (Deng et al., 2018; Ježek et al., 2018; Sharma et al., 2011). Complex I of the ETC generates ROS, and ROS are recognized as crucial signaling molecules in cellular processes, including stem cell maintenance and differentiation (Deng et al., 2018; Juan et al., 2021; Khacho et al., 2016; Murphy, 2009; Owusu-Ansah and Banerjee, 2009). Superoxides generated in the mitochondria can diffuse into the cytoplasm as H2O2 and also be transported outside via the voltage-gated anion channels (VDAC) in the outer membrane (Han et al., 2003). We assessed the ROS levels in type II neuroblasts using the dihydroethidium (DHE) fluorescence as compared to neighboring control type I neuroblasts. ND-51 and ND-19 depletion in type II neuroblasts resulted in a significant elevation in ROS levels as compared to controls, and the ROS levels were restored in both Drp1SD;ND-51i and Drp1SD;ND-19i combinations (Figures 5F and 5G). In summary, Drp1SD-mediated mitochondrial fusion in complex I-depleted neuroblasts restored the mitochondrial cristae architecture and mitochondrial activity defects as measured by MMP and ROS levels. Cell cycle progression and differentiation in type II neuroblasts are regulated by ROS Given the increased ROS in complex I-depleted type II neuroblasts, we tested if ROS regulates differentiation by manipulating ROS levels via RNAi knockdown of the scavenger enzymes: cytoplasmic SOD1 and mitochondrial SOD2. Depleting SOD enzymes in type II neuroblasts and larval muscle showed mitochondrial fragmentation (Figure S5A). Further, MMP was unchanged in SOD1i type II neuroblasts and decreased slightly in SOD2i (Figures S5B and S5C). SOD1 and SOD2 downregulation in type II neuroblasts resulted in elevated ROS levels (Figures S5D and S5E) and decreased number of mINPs, GMCs, and neurons in the lineages as compared to controls (Figures 6A–6E). SOD1 and SOD2 downregulation also led to a decreased number of EdU positive cells per lineage, decreased percentage of EdU positive mINPs (Figures 6F–6H), and decreased nuclear cyclin E localization in type II neuroblasts (Figures 6I and 6J). These results confirm that enhanced ROS generation impedes nuclear cyclin E, causing proliferation and differentiation defects; thus, ROS mediates the reduction in lineage cells caused by complex I depletion.Figure 6ROS regulates differentiation and cell cycle progression in type II neuroblast lineages, and SOD2 overexpression suppresses ND-51i-driven differentiation defects(A) Representative image show mINPs and GMCs of the mentioned genotypes. mINPs are Dpn+ (gray). GMCs are Dpn-/Ase+ (red) (yellow arrowheads).(B) mINP number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 16 type II neuroblast lineages, 3 brains), SOD1i (15, 4), SOD2i (17, 4), SOD2OE (12, 3), ND-51i (15, 3), and SOD2OE;ND-51i (17, 5).(C) GMC number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 15 type II neuroblast lineages, 3 brains), SOD1i (15, 4), SOD2i (17, 4), SOD2OE (19, 3), ND-51i (17, 3), and SOD2OE;ND-51i (18, 5).(D) Representative images of neurons with high nuclear Pros (red) in type II neuroblast lineages of indicated genotypes.(E) Neuron number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 14 type II neuroblast lineages, 3 brains), SOD1i (12, 6), SOD2i (15, 4), SOD2OE (15, 4), ND-51i (15, 3), and SOD2OE;ND-51i (13, 4).(F) Representative images of EdU positive cells (red) and mINPs (gray) in type II neuroblast lineages of the mentioned genotypes.(G) EdU positive cell number quantification per type II neuroblast lineages of genotypes: mCherryi (n = 15 type II neuroblast lineages, 5 brains), SOD1i (15, 5), and SOD2i (15, 3).(H) Percent EdU positive mINP quantification (yellow arrowheads) in type II neuroblast lineages of genotypes: mCherryi (n = 12 type II neuroblast lineages, 5 brains), SOD1i (15, 5), and SOD2i (15, 3).(I) Representative images of cyclin E immunostaining (red) in type II neuroblasts of the indicated genotypes. Type II neuroblasts: circled with a white dashed line.(J) Cyclin E nuclear to the cytoplasmic intensity quantification of type II neuroblast of genotypes: mCherryi (n = 64 type II neuroblasts, 9 brains), SOD1i (12, 4), and SOD2i (24, 4).Scale bars, 10 μm. Graphs show mean ± sd. Statistical analysis: Mann-Whitney test. ns = not significant, ∗∗∗- p < 0.001. Also see Figures S5, S6 and S7. ROS regulates differentiation and cell cycle progression in type II neuroblast lineages, and SOD2 overexpression suppresses ND-51i-driven differentiation defects (A) Representative image show mINPs and GMCs of the mentioned genotypes. mINPs are Dpn+ (gray). GMCs are Dpn-/Ase+ (red) (yellow arrowheads). (B) mINP number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 16 type II neuroblast lineages, 3 brains), SOD1i (15, 4), SOD2i (17, 4), SOD2OE (12, 3), ND-51i (15, 3), and SOD2OE;ND-51i (17, 5). (C) GMC number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 15 type II neuroblast lineages, 3 brains), SOD1i (15, 4), SOD2i (17, 4), SOD2OE (19, 3), ND-51i (17, 3), and SOD2OE;ND-51i (18, 5). (D) Representative images of neurons with high nuclear Pros (red) in type II neuroblast lineages of indicated genotypes. (E) Neuron number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 14 type II neuroblast lineages, 3 brains), SOD1i (12, 6), SOD2i (15, 4), SOD2OE (15, 4), ND-51i (15, 3), and SOD2OE;ND-51i (13, 4). (F) Representative images of EdU positive cells (red) and mINPs (gray) in type II neuroblast lineages of the mentioned genotypes. (G) EdU positive cell number quantification per type II neuroblast lineages of genotypes: mCherryi (n = 15 type II neuroblast lineages, 5 brains), SOD1i (15, 5), and SOD2i (15, 3). (H) Percent EdU positive mINP quantification (yellow arrowheads) in type II neuroblast lineages of genotypes: mCherryi (n = 12 type II neuroblast lineages, 5 brains), SOD1i (15, 5), and SOD2i (15, 3). (I) Representative images of cyclin E immunostaining (red) in type II neuroblasts of the indicated genotypes. Type II neuroblasts: circled with a white dashed line. (J) Cyclin E nuclear to the cytoplasmic intensity quantification of type II neuroblast of genotypes: mCherryi (n = 64 type II neuroblasts, 9 brains), SOD1i (12, 4), and SOD2i (24, 4). Scale bars, 10 μm. Graphs show mean ± sd. Statistical analysis: Mann-Whitney test. ns = not significant, ∗∗∗- p < 0.001. Also see Figures S5, S6 and S7. Since high ROS in complex I or superoxide scavenging enzyme depletion led to the loss of differentiation in the type II neuroblasts, we overexpressed SOD2 in complex I-depleted type II neuroblasts to test for the suppression of differentiation defects. SOD2 overexpression in type II neuroblasts did not change ROS levels (Figures S5F and S5G) and the number of mINPs, GMCs, and neurons as compared to controls (Figures 6A–6E). But overexpression of SOD2 in ND-51 depleted type II neuroblasts (SOD2OE;ND-51i) decreased ROS levels compared to ND-51i alone (Figures S5F and S5G). Further, the SOD2OE;ND-51i restored the numbers of mINPs, GMCs, and neurons in each type II neuroblast lineage as compared to ND-51i alone (Figures 6A–6E). Rescue of differentiation defects in SOD2OE;ND-51i was consistent with the rescue of cell cycle defects in the lineage. The number of EdU-positive cells and the proportion of EdU-positive mINPs were higher than those of the ND-51i alone (Figures S6A–S6C). Also, nuclear cyclin E levels in SOD2OE;ND-51i type II neuroblasts were similar to those observed in controls (Figures S6D and S6E). SOD2OE;ND-51i expressing type II neuroblasts showed a loss of MMP as compared to controls (Figures S5B and S5C) and similar to ND51i (Figures 5A and 5B), suggesting that ROS is the downstream mediator of proliferation and differentiation defects on complex I depletion. Since mitochondrial fusion by expressing Drp1 mutants in complex I downregulated type II neuroblast rescued differentiation defects (Figure 3) by lowering ROS levels (Figures 5F and 5G), we expressed Drp1SG in SOD2i expressing type II neuroblasts. We observed that SOD2i-mediated differentiation defects were rescued; the number of mINPs, GMCs, and neurons was increased compared to SOD2i alone and was similar to controls (Figures S7A–S7E). We further combined Drp1SD with SOD2i and complex I RNAi (ND-39i) to determine whether mitochondrial fusion could rescue differentiation defects despite further elevation of ROS in neuroblasts. We found that Drp1SD/ND-39i;SOD2i expressing type II neuroblast showed rescue of differentiation defects similar to Drp1SD with SOD2i (Figures S7F and S7G) and Drp1SD with complex I depletion (Figure 3). It is possible that depletion of SOD2 in the Drp1SD and complex I depletion background did not increase ROS appreciably to give rise to defects in differentiation seen on complex I and SOD2 depletion. It is also possible that mitochondrial fusion rescues defects in differentiation on complex I depletion by suppressing levels of unidentified metabolites, which need to be characterized in the future. In summary, our findings show a crucial role for maintenance of appropriate ROS levels by mitochondrial fusion and complex I activity in regulating differentiation in the type II neuroblast lineage. Complex I depletion reduces notch signaling, and notch overexpression rescues complex I depletion driven lineage defects JNK and Notch pathways have been implicated in coordinating mitochondrial function with stem cell differentiation (Deng et al., 2018; Dubal et al., 2022; Khacho et al., 2016). We checked if their activation is affected in complex I-depleted type II neuroblast lineages. Mitochondrial fragmentation elevates ROS, activates JNK, and impairs differentiation in Drosophila intestinal cells (Deng et al., 2018). Since complex I depletion in type II neuroblasts showed mitochondrial fragmentation and increased ROS levels (Figures 3 and 5), we determined if JNK signaling was activated by immunostaining for activated phosphorylated-c-Jun (p-c-Jun) (Acharjee et al., 2025). p-c-Jun levels in ND-51i were comparable with mCherryi controls (Figures S8A and S8B), suggesting that JNK signaling activity was not enhanced upon complex I impairment. We increased JNK activity by the overexpression of Hemipterous (HepOE) (Worley et al., 2021) and reduced it by Jun-related antigen knockdown (Jrai) (Ray et al., 2017) or expression of dominant-negative Basket (BskDN) (Pérez et al., 2017) (Figures S8C and S8D) to further ascertain type II neuroblast lineage defects. Neither activation nor inhibition of JNK signaling significantly altered the mINP numbers within the lineage (Figures S8E and S8F). This suggests JNK signaling is not a major factor in complex I depletion-induced differentiation defects. We next investigated whether Notch signaling defects caused complex I depletion-driven differentiation loss. Notch signaling is a key determinant of neuroblast number and lineage progression in the Drosophila third instar larval brain (Bowman et al., 2008; Li et al., 2016; Wang et al., 2006). It is active in type II neuroblasts, downregulated in immature INPs, and reactivated in mINPs (Rossi and Desplan, 2017). Loss of Notch signaling results in a reduction of type II neuroblasts and their transformation into type I-like neuroblasts, accompanied by the absence of Dpn-positive mINPs. Conversely, hyperactivation of Notch signaling increases the number of neuroblasts through the dedifferentiation of immature INPs (Bowman et al., 2008; Weng et al., 2010). Notch signaling regulates mitochondrial morphology and differentiation in type II neuroblasts, where Notch depletion induces mitochondrial fragmentation (Dubal et al., 2022). Given that complex I depletion also causes mitochondrial fragmentation in type II neuroblasts, we examined whether it affects Notch pathway activity. We monitored Notch signaling using the E(spl)mγ-GFP reporter in complex I-depleted (ND-51i) type II neuroblasts and their lineages. E(spl)mγ-GFP expression was detected in type II neuroblasts and mINPs (Figures 7A–7C). While reporter levels in ND-51i neuroblasts were comparable to controls, the number of E(spl)mγ-GFP-positive mINPs was reduced, indicating impaired activation of Notch signaling at later stages of the lineage. We further examined Notch receptor distribution using an antibody to the Notch intracellular domain (NICD) in control and ND-51-depleted type II neuroblasts. NICD distribution was comparable in controls and ND51i-expressing type II neuroblasts. However, mINPs within the ND-51i lineage displayed a marked loss of NICD at the cell membrane (Figures 7D–7F). This showed that Notch signaling was defective in complex I-depleted type II neuroblast lineages.Figure 7Complex I depletion reduces Notch signaling in mINPs and Notch overexpression rescues complex I depletion-driven differentiation defects in type II neuroblast lineages(A) Representative images of type II neuroblast lineages of mentioned genotypes, showing E(spl)mγ-GFP (green), Dpn+ mINPs (red, small dashed circle), type II neuroblast (large white dashed circle), mINPs positive for E(spl)mγ-GFP (small white circle), and mINPs negative for E(spl)mγ-GFP (small yellow circle).(B) E(spl)mγ-GFP intensity quantification in type II neuroblasts in genotypes: mCherryi (n = 10 type II neuroblasts, 3 brains) and ND-51i (5, 5).(C) Percentage of mINPs negative for E(spl)mγ-GFP of genotypes: mCherryi (n = 16 type II neuroblast lineages, 3 brains), and ND-51i (14, 5).(D) Representative images show NICD (red) and Dpn positive mINPs (magenta) staining in type II neuroblast and its lineages (green) in the mentioned genotypes. Yellow arrowheads mark cells in the lineage that show loss of Notch receptors.(E) NICD fluorescence quantification in type II neuroblasts in genotypes: mCherryi (n = 23 type II neuroblasts, 5 brains), and ND-51i (30, 5).(F) Quantification of percentage mINPs with loss of NICD on the membrane of genotypes: mCherryi (n = 18 type II linages, 5 brains), and ND-51i (14, 5).(G) Representative images of larval brain lobes (white dashed circle) showing type II neuroblasts and their lineages (green) in the mentioned genotypes.(H) Neuroblast number quantification per brain lobe of genotypes: mCherryi (20 brain lobes, 10 brains), ND-51i (20, 10), NotchFL(4, 3), and ND-51i;NotchFL (6, 3).(I) Representative images of differentiated progenies in the type II neuroblast lineage of the mentioned genotypes. mINPs are Dpn+ (gray). GMCs are Dpn-/Ase+ (red) (yellow arrowheads).(J) mINP number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 28 type II neuroblast lineages, 5 brains), NotchFL (9, 3), ND-51i (12, 3), and ND-51i;NotchFL (8, 3).(K) GMC number quantification in type II neuroblast lineages of genotypes, mCherryi (n = 24 type II neuroblast lineages, 5 brains), NotchFL (9, 3), ND-51i (17, 3), and ND-51i;NotchFL (9, 3).(L) Representative images of neurons with high nuclear Pros (red) in type II neuroblast lineages of the indicated genotypes.(M) Neuron number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 14 type II neuroblast lineages, 3 brains), NotchFL (10, 3), ND-51i (15, 3), and ND-51i;NotchFL (10, 3).(N) Schematic summary shows that complex I depletion impairs neuroblast proliferation and differentiation due to mitochondrial fragmentation and increased ROS, leading to cell cycle defects. Mitochondrial fusion, SOD2, and Notch overexpression in complex I-depleted neuroblasts leads to the suppression of differentiation defects.Scale bars, 50 μm (G), 10 μm (A, D, I, and L). Graphs show mean ± sd. Statistical analysis: Mann-Whitney test. ns = not significant, ∗∗- p < 0.001 and ∗∗∗- p < 0.001. Also see Figures S8 and S9. Complex I depletion reduces Notch signaling in mINPs and Notch overexpression rescues complex I depletion-driven differentiation defects in type II neuroblast lineages (A) Representative images of type II neuroblast lineages of mentioned genotypes, showing E(spl)mγ-GFP (green), Dpn+ mINPs (red, small dashed circle), type II neuroblast (large white dashed circle), mINPs positive for E(spl)mγ-GFP (small white circle), and mINPs negative for E(spl)mγ-GFP (small yellow circle). (B) E(spl)mγ-GFP intensity quantification in type II neuroblasts in genotypes: mCherryi (n = 10 type II neuroblasts, 3 brains) and ND-51i (5, 5). (C) Percentage of mINPs negative for E(spl)mγ-GFP of genotypes: mCherryi (n = 16 type II neuroblast lineages, 3 brains), and ND-51i (14, 5). (D) Representative images show NICD (red) and Dpn positive mINPs (magenta) staining in type II neuroblast and its lineages (green) in the mentioned genotypes. Yellow arrowheads mark cells in the lineage that show loss of Notch receptors. (E) NICD fluorescence quantification in type II neuroblasts in genotypes: mCherryi (n = 23 type II neuroblasts, 5 brains), and ND-51i (30, 5). (F) Quantification of percentage mINPs with loss of NICD on the membrane of genotypes: mCherryi (n = 18 type II linages, 5 brains), and ND-51i (14, 5). (G) Representative images of larval brain lobes (white dashed circle) showing type II neuroblasts and their lineages (green) in the mentioned genotypes. (H) Neuroblast number quantification per brain lobe of genotypes: mCherryi (20 brain lobes, 10 brains), ND-51i (20, 10), NotchFL(4, 3), and ND-51i;NotchFL (6, 3). (I) Representative images of differentiated progenies in the type II neuroblast lineage of the mentioned genotypes. mINPs are Dpn+ (gray). GMCs are Dpn-/Ase+ (red) (yellow arrowheads). (J) mINP number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 28 type II neuroblast lineages, 5 brains), NotchFL (9, 3), ND-51i (12, 3), and ND-51i;NotchFL (8, 3). (K) GMC number quantification in type II neuroblast lineages of genotypes, mCherryi (n = 24 type II neuroblast lineages, 5 brains), NotchFL (9, 3), ND-51i (17, 3), and ND-51i;NotchFL (9, 3). (L) Representative images of neurons with high nuclear Pros (red) in type II neuroblast lineages of the indicated genotypes. (M) Neuron number quantification in type II neuroblast lineages of genotypes: mCherryi (n = 14 type II neuroblast lineages, 3 brains), NotchFL (10, 3), ND-51i (15, 3), and ND-51i;NotchFL (10, 3). (N) Schematic summary shows that complex I depletion impairs neuroblast proliferation and differentiation due to mitochondrial fragmentation and increased ROS, leading to cell cycle defects. Mitochondrial fusion, SOD2, and Notch overexpression in complex I-depleted neuroblasts leads to the suppression of differentiation defects. Scale bars, 50 μm (G), 10 μm (A, D, I, and L). Graphs show mean ± sd. Statistical analysis: Mann-Whitney test. ns = not significant, ∗∗- p < 0.001 and ∗∗∗- p < 0.001. Also see Figures S8 and S9. To test whether restoring Notch signaling could compensate for complex I depletion, we overexpressed the full-length Notch (NotchFL) receptor. As previously reported, NotchFL overexpression increased the number of neuroblasts (Dubal et al., 2022). However, knockdown of ND-51 with NotchFL overexpression reduced the number of neuroblasts compared with NotchFL alone (Figures 7G and 7H). Given the functional interaction between mitochondrial dynamics and Notch signaling, we next analyzed mitochondrial morphology and activity. NotchFL expression resulted in mitochondrial clustering on one side of the neuroblast (Figures S9A) but did not significantly alter ROS levels compared with mCherryi controls (Figures S9D and S9E). The overexpression of NotchFL in ND-51i depleted type II neuroblasts, however, restored mitochondrial morphology and rescued MMP (Figures S9A–S9C), suggesting that Notch signaling can enhance mitochondrial function on complex I impairment. We next examined whether enhancing Notch activity could rescue the differentiation defects caused by complex I depletion. NotchFL overexpression in ND-51i expressing type II neuroblasts restored the numbers of mINPs, GMCs, and neurons within the lineage (Figures 7I–7M) as compared to ND-51i alone, indicating that the activation of Notch signaling alleviates the lineage defects induced by complex I depletion. Together, these results show that complex I activity supports Notch signaling during lineage progression and that increasing Notch activity can compensate for defects in mitochondrial morphology and activity along with differentiation in type II neuroblast lineages. In summary, the elevation of ROS levels in type II neuroblasts, induced by either complex I or SOD1 or SOD2 depletion, leads to mitochondrial fragmentation and differentiation defects. However, restoring ROS levels by overexpressing scavenging enzymes or fusing mitochondria by either depleting Drp1 or increasing Notch signaling rescues the differentiation defects (Figure 7N). Together, these results demonstrate that a fused mitochondrial morphology is essential for maintaining ROS homeostasis and Notch signaling, which enable cell-cycle progression and differentiation of Drosophila type II neuroblasts (Figure 7N). Discussion We show that optimal complex I activity and mitochondrial fusion limit ROS to regulate the cell cycle, differentiation, and Notch signaling in the Drosophila type II neuroblast lineage. We discuss our results in the context of the following: 1] the mechanism by which there is a change in ROS by complex I depletion and how mitochondrial fusion impacts this change and 2] the mechanism of control of mitochondrial activity and morphology in vivo in stem cell differentiation. During electron transport through ETC complexes, 0.2–2% of electrons escape ETC and interact with oxygen to form superoxides (Cadenas and Davies, 2000). Complex I and complex III are the main sites of ROS production in mitochondria (Brand, 2010). Dysfunctionality of complex I or its inhibition by rotenone leads to inefficient transfer of electrons, an increase in their concentration, and formation of ROS by reaction with oxygen in the mitochondria (Sharma et al., 2009). We observed an increase in ROS on complex I depletion by expressing RNAi against its subunits ND-51 or mammalian NDUFV1 and ND-19 or mammalian NDUFA8. ND-51 belongs to the core hydrophilic matrix, catalyzes the oxidation of NADH to NAD+, and is part of ∼815kDa complex I assembly intermediate. Downregulation of ND-51 and ND-19 in Drosophila muscle decreases complex I assembly (Garcia et al., 2017). Thus, disrupting complex I formation in type II neuroblasts on ND-51 and ND-19 knockdown may cause complex I to lose electrons and produce ROS. Reverse electron transfer (RET) may be another possibility for the generation of increased ROS. RET catalyzes the reduction of matrix NAD+ with electrons provided by CoQH2. The electrons from complex I and II are transferred to CoQH2; some are further moved to complex III (FET, forward electron transfer), and others flow back to complex I (RET). Excessive ROS is generated during RET compared to FET, presumably due to the dissociation of FMNH2 from complex I and the reaction with matrix O2 (Ten and Galkin, 2019). FMN and the CoQ reduction sites of complex I are centers for ROS production during RET (Lambert and Brand, 2004; Murphy, 2016). In complex I-deficient neuroblasts, mitochondrial fusion might reduce ROS generation by facilitating ETC supercomplex formation in the cristae. These supercomplexes increase electron transport efficiency between ETC complexes by optimizing the distance between them (Berndtsson et al., 2020; Kohler et al., 2023). Hydrogen peroxide, a small and electrically neutral molecule, diffuses through the inner and outer membranes of mitochondria to reach the cytoplasm (Shadel and Horvath, 2015). Superoxides, on the other hand, may pass to the cytoplasm through mitochondrial channels. Increased ROS accumulation beyond a threshold level in mitochondria induces the opening of mitochondrial channels, mitochondrial permeability transition (MPT) pore (Haworth and Hunter, 1979), and inner membrane anion channel (IMAC) (Garlid and Beavis, 1986), which releases ROS to the cytoplasm, accompanied by MMP dissipation (Zorov et al., 2000). ROS accumulation in the cytosol promotes the opening of MPT and IMAC in other mitochondria, called ROS-induced ROS release (RIRR). Therefore, RIRR might be another reason for MMP loss on complex I depletion and mitochondrial fusion rescuing MMP by inhibiting RIRR through decreasing ROS levels. Decrease in complex I subunits leads to mitochondrial fragmentation. This is likely to occur due to the alteration of mitochondrial fusion and fission proteins or the impact of ROS on mitochondrial fusion and fission proteins. Complex I inhibition by rotenone induces mitochondrial fission in a Drp1-dependent manner in neurons in the Parkinson’s disease model (Barsoum et al., 2006). Decreasing mitochondrial fission by expressing a dominant-negative Drp1 or overexpressing Mfn led to the inhibition of apoptosis of these neurons. Increase in ROS also increases the expression of Drp1 and activates Drp1 fission by increasing phosphorylation at the 616 residue in mouse hippocampal neurons (Chang et al., 2023; Ježek et al., 2018). ROS can regulate mitochondrial dynamics by modulating fusion and fission proteins at the transcriptional level and through post-translational modifications. High cellular ROS activates MAP kinases such as ERK, which then phosphorylate and activate Drp1, promoting fission (Brillo et al., 2021). In hepatocellular carcinoma, elevated ROS generation facilitates survival by activating the AKT pathway, which then activates NF-κB. This pathway leads to the expression of proteins involved in mitochondrial fission, such as Drp1, and suppresses fusion proteins MFN1 (Huang et al., 2016). Multiple signaling pathways coordinate mitochondrial function with stem cell fate. In this study, we analyzed the role of Notch and JNK signaling in type II neuroblast differentiation in the context of ETC complex I function. We show that Notch signaling regulates mitochondrial fusion and enhances differentiation in complex I-depleted type II neuroblast lineages. An opposing effect has been observed in mice, where less Notch signaling maintains fragmented mitochondria and promotes intestinal stem cells and neural stem cells differentiation (Khacho et al., 2016; Ludikhuize et al., 2020). Fused mitochondria promote the differentiation of Drosophila intestinal stem cells by maintaining low ROS and suppressing the JNK pathway. Inducing mitochondrial fragmentation elevated ROS levels and JNK activity, and inhibited differentiation (Deng et al., 2018). However, JNK signaling was not affected in complex I-depleted type II neuroblast lineages. Signaling pathways could therefore regulate mitochondrial dynamics to affect specific differentiation outcomes (Verma et al., 2026). In conclusion, our study determines that complex I activity is essential for regulating stem cell differentiation by regulating ROS levels and mitochondrial dynamics. We propose that optimal complex I activity and mitochondrial fusion maintain redox balance, thereby supporting Notch signaling and differentiation in the Drosophila type II neuroblast lineage. It is possible that a change in unidentified metabolites and pathways leads to defects in differentiation on complex I depletion. Future analyses will help evaluate the mechanisms by which mitochondrial fusion alleviates defects in other metabolites in differentiation due to ETC deficiencies. In conclusion, our study reveals a mechanistic connection between mitochondrial activity and developmental signaling pathways, offering insight into how mitochondrial morphology and activity can influence stem cell proliferation and differentiation during neurogenesis in vivo. Resource availability Lead contact Requests for further information and resources should be directed to and will be fulfilled by the lead contact, richa@iiserpune.ac.in. Materials availability The materials are available on request from the lab and can also be bought from stock centers. Data and code availability All the data are included as a part of the manuscript. Acknowledgments We thank the RR lab for continuous feedback on this study. We thank the IISER Pune Drosophila and microscopy facility for help with stocks and microscopy. We thank Flybase and Bloomington Stock Center for help with fly stocks. We thank Jurgen Knoblich (IMB, Vienna, Austria) for stocks. We thank Yuh-Nung Jan, UCSF, USA, for the Asense antibody. We thank Mohit Prasad (IISER, Kolkata, India) for the p-c-Jun antibody. This study is supported by the 10.13039/501100009053Wellcome Trust DBT India Alliance grant number IA/S/22/1/506232. RKV thanks the 10.13039/501100001332CSIR, India, for his graduate fellowship. AB thanks DST INSPIRE for his undergraduate fellowship. Author contributions R.K.V. participated in conceptualization, investigation, methodology, visualization, manuscript writing, and review and editing. A.B. and D.D. participated in the investigation, methodology, visualization, and manuscript review and editing. R.R. participated in funding acquisition, conceptualization, investigation, methodology, visualization, manuscript writing, and review and editing. Declaration of interests The authors declare no competing interests. STAR★Methods Key resources table REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesChicken polyclonal anti-GFP (1:1000)Thermo Fisher ScientificCat# A10262; RRID: AB_2534023Rabbit polyclonal anti-Asense (1:10000)Gift from Yuh-Nung Jan, UCSF, USAN/ARat monoclonal anti-Deadpan (1:150)AbcamCat# Ab195173; RRID: AB_2687586Rabbit polyclonal anti-Deadpan (1:800)Gift from Yuh-Nung Jan, UCSF, USAN/AMouse monoclonal anti-Prospero (1:20)DSHBCat# MR1A; RRID: AB_528440Rabbit polyclonal Phospho-Histone H3 (Ser10) (1:500)Thermo Fisher ScientificCat# PA5-17869; RRID: AB_10984484Mouse Monoclonal anti- ATP Synthase Subunit Beta (1:200)AbcamCat# ab14730; RRID: AB_301438Rabbit polyclonal anti-cyclin E (d-300) (1:200)Santa Cruz BiotechnologyCat# sc-33748; RRID: AB_638863Rabbit Monoclonal anti-AMPK-alpha, phospho (Thr172) (1:200)Cell Signaling TechnologyCat# 2535; RRID: AB_331250Rat monoclonal anti-Miranda (1:400)AbcamCat# ab197788; RRID: AB_2936368Rabbit polyclonal anti-Cleaved Caspase-3 (Asp175) (1:100)Cell Signaling TechnologyCat# 9661; RRID: AB_2341188Mouse monoclonal anti-p-c-Jun (KM-1) (1:50)Santa Cruz BiotechnologyCat# sc-822; RRID: AB_627262Mouse monoclonal anti-Notch, intracellular domain (1:10)DSHBCat# c17.9c6; RRID: AB_528410Alexa Fluor 488 Goat Anti-Chicken IgG (H + L) (1:1000)Molecular ProbesCat# A-11039; RRID: AB_142924Alexa Fluor 568 Goat Anti-Mouse IgG (H + L) (1:1000)Molecular ProbesCat# A-11004; RRID: AB_2534072Alexa Fluor 647 Goat Anti-Mouse IgG (H + L) (1:1000)Molecular ProbesCat# A-21235; RRID: AB_2535804Alexa Fluor 647 Goat Anti-Rabbit IgG (H + L) (1:1000)Molecular ProbesCat# A-21245; RRID: AB_2535813Alexa Fluor 488 Goat Anti-Rabbit IgG (H + L) (1:1000)Molecular ProbesCat# A-11008; RRID: AB_143165Alexa Fluor 568 Goat Anti-Rabbit IgG (H + L) (1:1000)Molecular ProbesCat# A-11011; RRID: AB_143157Alexa Fluor 647 Goat Anti-Rat IgG (H + L) (1:1000)Molecular ProbesCat# A-21247; RRID: AB_141778Alexa Fluor 568 Goat Anti-Rat IgG (H + L) (1:1000)Molecular ProbesCat# A-11077; RRID: AB_141874Alexa Fluor 488 Goat Anti-Rat IgG (H + L) (1:1000)Molecular ProbesCat# A-11006; RRID: AB_141373Chemicals, peptides, and recombinant proteinsHoechst DNA dye (10μg/mL)Thermo Fisher ScientificCat# H3570Tetramethylrhodamine (TMRM) (100nM)Thermo Fisher ScientificCat# T668Dihydroethidium (DHE) (30μM)Thermo Fisher ScientificCat# D1168BioTracker ATP-Red Live Cell Dye (30μM)Merk MilliporeCat# SCT045Schneider’s Drosophila mediaThermo Fisher ScientificCat# 21720024ParaformaldehydeSigma-AldrichCat# 158127Triton X-100Sigma-AldrichCat# T8787Bovine serum albuminHimediaCat# MB083poly-L-lysineSigma-AldrichCat# P8920D(+) LimoneneLobachemieCat# 0441600500SlowFade™ Gold Antifade MountantThermo Fisher ScientificCat# S36937InsulinSigma-AldrichCat# I0516Fetal bovine serumGibcoCat# 10270-106GlutaraldehydeSigma-AldrichCat# G5882Osmium tetroxideSigma-AldrichCat# 201030Uranyl acetatePubChemCat# CID 21226249Critical commercial assayClick-iT™ EdU Cell Proliferation KitThermo Fisher ScientificCat# C10340Experimental models: Organisms/strainsD. melanogaster; wor-Gal4,UAS-mCD8-GFPJurgen Knoblich, IMP, Vienna, AustriaN/AD. melanogaster; pnt-Gal4, UAS-mCD8-GFPJurgen Knoblich, IMP, Vienna, AustriaN/AD. melanogaster; mhc-Gal4Bloomington Drosophila Stock center (BDSC)RRID: BDSC_38464D.melanogaster;UAS-ND-51-RNAi-TRiP.HMS01590BDSCRRID: BDSC_36701D.melanogaster; UAS-ND-51-RNAi-TRiP.HM05213BDSCRRID: BDSC_29534D.melanogaster; UAS-ND-39-RNAi-TRiP.HMC03662BDSCRRID: BDSC_52922D.melanogaster; UAS-ND-19-RNAi-TRiP.GLC01699BDSCRRID: BDSC_50577D.melanogaster; UAS-ND-PDSW-RNAi-TRiP.JF03271BDSCRRID: BDSC_29592D.melanogaster; UAS-ND-49-RNAi-TRiP.HMC04814BDSCRRID: BDSC_57499D.melanogaster; UAS-ND-39-RNAi-TRiP.HMS00852BDSCRRID: BDSC_34526D.melanogaster; Drp1S193D (Drp1SD)(Chowdhary et al., 2017)N/AD.melanogaster; UAS-mCherry-RNAiBDSCRRID: BDSC_35785D.melanogaster; UAS-mito-HA-GFP.APBDSCRRID: BDSC_8442D.melanogaster; UAS-SOD1-RNAi-TRiP.HMS0129BDSCRRID: BDSC_34616D.melanogaster; UAS-SOD2-RNAi-TRiP.JF01989BDSCRRID: BDSC_25969D.melanogaster; UAS-SOD2.MBDSCRRID: BDSC_24494D.melanogaster; E(spl)mγ-GFP(Almeida and Bray, 2005)N/AD.melanogaster; UAS-hep.BBDSCRRID: BDSC_9308D.melanogaster; UAS-bsk.DNBDSCRRID: BDSC_6409D.melanogaster; UAS-Jra-RNAi-TRiP.JF01184BDSCRRID: BDSC_31595D.melanogaster; UAS-Drp1SD/CyOGFP;UAS-mCherry-RNAi(BL35785)/TM6BThis paperN/AD.melanogaster; UAS-Drp1SD/CyOGFP;UAS-ND-51-RNAi(BL36701)/TM6BThis paperN/AD.melanogaster; UAS-Drp1SD/CyOGFP;UAS-ND-19-RNAi (BL50577)/TM6BThis paperN/AD.melanogaster; UAS-Drp1SD/CyOGFP;pnt-Gal4,UAS-CD8-GFP/TM6BThis paperN/AD.melanogaster; UAS-mito-HA-GFP/CyOGFP;pnt-Gal4/TM6BThis paperN/AD.melanogaster; UAS-mito-HA-GFP/CyOGFP;UAS-ND-51-RNAi (BL36701)/TM6BThis paperN/AD.melanogaster; UAS-SOD2/CyOGFP;UAS-ND-51-RNAi(BL36701)/TM6BThis paperN/AD.melanogaster; UAS-Drp1SG/FM7a;UAS-SOD2-RNAi(BL25969)/TM3,SerGFPThis paperN/AD.melanogaster; UAS-Drp1SD/CyOGFP;UAS-SOD2-RNAi(BL25969)/TM6BThis paperN/AD.melanogaster; UAS-E(spl)mγ-GFP/CyOGFP;pnt-Gal4/TM6BThis paperN/ASoftware and algorithmsImageJ/FijiFijihttp://fiji.sc/IllustratorAdobehttps://www.adobe.com/products/photoshop.htmlPrismGraphPadhttp://www.graphpad.com/OtherLabTEK chamberThermo Fisher ScientificCat# 155380 REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesChicken polyclonal anti-GFP (1:1000)Thermo Fisher ScientificCat# A10262; RRID: AB_2534023Rabbit polyclonal anti-Asense (1:10000)Gift from Yuh-Nung Jan, UCSF, USAN/ARat monoclonal anti-Deadpan (1:150)AbcamCat# Ab195173; RRID: AB_2687586Rabbit polyclonal anti-Deadpan (1:800)Gift from Yuh-Nung Jan, UCSF, USAN/AMouse monoclonal anti-Prospero (1:20)DSHBCat# MR1A; RRID: AB_528440Rabbit polyclonal Phospho-Histone H3 (Ser10) (1:500)Thermo Fisher ScientificCat# PA5-17869; RRID: AB_10984484Mouse Monoclonal anti- ATP Synthase Subunit Beta (1:200)AbcamCat# ab14730; RRID: AB_301438Rabbit polyclonal anti-cyclin E (d-300) (1:200)Santa Cruz BiotechnologyCat# sc-33748; RRID: AB_638863Rabbit Monoclonal anti-AMPK-alpha, phospho (Thr172) (1:200)Cell Signaling TechnologyCat# 2535; RRID: AB_331250Rat monoclonal anti-Miranda (1:400)AbcamCat# ab197788; RRID: AB_2936368Rabbit polyclonal anti-Cleaved Caspase-3 (Asp175) (1:100)Cell Signaling TechnologyCat# 9661; RRID: AB_2341188Mouse monoclonal anti-p-c-Jun (KM-1) (1:50)Santa Cruz BiotechnologyCat# sc-822; RRID: AB_627262Mouse monoclonal anti-Notch, intracellular domain (1:10)DSHBCat# c17.9c6; RRID: AB_528410Alexa Fluor 488 Goat Anti-Chicken IgG (H + L) (1:1000)Molecular ProbesCat# A-11039; RRID: AB_142924Alexa Fluor 568 Goat Anti-Mouse IgG (H + L) (1:1000)Molecular ProbesCat# A-11004; RRID: AB_2534072Alexa Fluor 647 Goat Anti-Mouse IgG (H + L) (1:1000)Molecular ProbesCat# A-21235; RRID: AB_2535804Alexa Fluor 647 Goat Anti-Rabbit IgG (H + L) (1:1000)Molecular ProbesCat# A-21245; RRID: AB_2535813Alexa Fluor 488 Goat Anti-Rabbit IgG (H + L) (1:1000)Molecular ProbesCat# A-11008; RRID: AB_143165Alexa Fluor 568 Goat Anti-Rabbit IgG (H + L) (1:1000)Molecular ProbesCat# A-11011; RRID: AB_143157Alexa Fluor 647 Goat Anti-Rat IgG (H + L) (1:1000)Molecular ProbesCat# A-21247; RRID: AB_141778Alexa Fluor 568 Goat Anti-Rat IgG (H + L) (1:1000)Molecular ProbesCat# A-11077; RRID: AB_141874Alexa Fluor 488 Goat Anti-Rat IgG (H + L) (1:1000)Molecular ProbesCat# A-11006; RRID: AB_141373Chemicals, peptides, and recombinant proteinsHoechst DNA dye (10μg/mL)Thermo Fisher ScientificCat# H3570Tetramethylrhodamine (TMRM) (100nM)Thermo Fisher ScientificCat# T668Dihydroethidium (DHE) (30μM)Thermo Fisher ScientificCat# D1168BioTracker ATP-Red Live Cell Dye (30μM)Merk MilliporeCat# SCT045Schneider’s Drosophila mediaThermo Fisher ScientificCat# 21720024ParaformaldehydeSigma-AldrichCat# 158127Triton X-100Sigma-AldrichCat# T8787Bovine serum albuminHimediaCat# MB083poly-L-lysineSigma-AldrichCat# P8920D(+) LimoneneLobachemieCat# 0441600500SlowFade™ Gold Antifade MountantThermo Fisher ScientificCat# S36937InsulinSigma-AldrichCat# I0516Fetal bovine serumGibcoCat# 10270-106GlutaraldehydeSigma-AldrichCat# G5882Osmium tetroxideSigma-AldrichCat# 201030Uranyl acetatePubChemCat# CID 21226249Critical commercial assayClick-iT™ EdU Cell Proliferation KitThermo Fisher ScientificCat# C10340Experimental models: Organisms/strainsD. melanogaster; wor-Gal4,UAS-mCD8-GFPJurgen Knoblich, IMP, Vienna, AustriaN/AD. melanogaster; pnt-Gal4, UAS-mCD8-GFPJurgen Knoblich, IMP, Vienna, AustriaN/AD. melanogaster; mhc-Gal4Bloomington Drosophila Stock center (BDSC)RRID: BDSC_38464D.melanogaster;UAS-ND-51-RNAi-TRiP.HMS01590BDSCRRID: BDSC_36701D.melanogaster; UAS-ND-51-RNAi-TRiP.HM05213BDSCRRID: BDSC_29534D.melanogaster; UAS-ND-39-RNAi-TRiP.HMC03662BDSCRRID: BDSC_52922D.melanogaster; UAS-ND-19-RNAi-TRiP.GLC01699BDSCRRID: BDSC_50577D.melanogaster; UAS-ND-PDSW-RNAi-TRiP.JF03271BDSCRRID: BDSC_29592D.melanogaster; UAS-ND-49-RNAi-TRiP.HMC04814BDSCRRID: BDSC_57499D.melanogaster; UAS-ND-39-RNAi-TRiP.HMS00852BDSCRRID: BDSC_34526D.melanogaster; Drp1S193D (Drp1SD)(Chowdhary et al., 2017)N/AD.melanogaster; UAS-mCherry-RNAiBDSCRRID: BDSC_35785D.melanogaster; UAS-mito-HA-GFP.APBDSCRRID: BDSC_8442D.melanogaster; UAS-SOD1-RNAi-TRiP.HMS0129BDSCRRID: BDSC_34616D.melanogaster; UAS-SOD2-RNAi-TRiP.JF01989BDSCRRID: BDSC_25969D.melanogaster; UAS-SOD2.MBDSCRRID: BDSC_24494D.melanogaster; E(spl)mγ-GFP(Almeida and Bray, 2005)N/AD.melanogaster; UAS-hep.BBDSCRRID: BDSC_9308D.melanogaster; UAS-bsk.DNBDSCRRID: BDSC_6409D.melanogaster; UAS-Jra-RNAi-TRiP.JF01184BDSCRRID: BDSC_31595D.melanogaster; UAS-Drp1SD/CyOGFP;UAS-mCherry-RNAi(BL35785)/TM6BThis paperN/AD.melanogaster; UAS-Drp1SD/CyOGFP;UAS-ND-51-RNAi(BL36701)/TM6BThis paperN/AD.melanogaster; UAS-Drp1SD/CyOGFP;UAS-ND-19-RNAi (BL50577)/TM6BThis paperN/AD.melanogaster; UAS-Drp1SD/CyOGFP;pnt-Gal4,UAS-CD8-GFP/TM6BThis paperN/AD.melanogaster; UAS-mito-HA-GFP/CyOGFP;pnt-Gal4/TM6BThis paperN/AD.melanogaster; UAS-mito-HA-GFP/CyOGFP;UAS-ND-51-RNAi (BL36701)/TM6BThis paperN/AD.melanogaster; UAS-SOD2/CyOGFP;UAS-ND-51-RNAi(BL36701)/TM6BThis paperN/AD.melanogaster; UAS-Drp1SG/FM7a;UAS-SOD2-RNAi(BL25969)/TM3,SerGFPThis paperN/AD.melanogaster; UAS-Drp1SD/CyOGFP;UAS-SOD2-RNAi(BL25969)/TM6BThis paperN/AD.melanogaster; UAS-E(spl)mγ-GFP/CyOGFP;pnt-Gal4/TM6BThis paperN/ASoftware and algorithmsImageJ/FijiFijihttp://fiji.sc/IllustratorAdobehttps://www.adobe.com/products/photoshop.htmlPrismGraphPadhttp://www.graphpad.com/OtherLabTEK chamberThermo Fisher ScientificCat# 155380 Experimental model and study participant details Drosophila stocks and maintenance The flies were maintained on a standard cornmeal-agar medium at 25°C. The fly lines used were listed in key resources table. Flybase (www.flybase.org) was used for obtaining information on fly stocks (https://doi.org/10.1093/genetics/iyad211 and https://doi.org/10.1007/978-1-0716-2541-5_1). Method details Genetic crosses 10 female and 5–6 male flies were added to polystyrene vials containing cornmeal agar media at 29°C. After 48 h, the flies were transferred to fresh vials. Third-instar larvae emerged after 5–6 days and were used for further analysis. The wor-Gal4, UAS-mCD8-GFP was used to express transgenes in all third-instar larval neuroblasts and pnt-Gal4, UAS-mCD8-GFP was used to express in type II neuroblasts. The combinations of RNAi and Drp1SD stocks were made using standard genetic crosses using balancers. The following strategy was used to generate complex I (ND-39i) and SOD2 RNAi neuroblasts expressing Drp1SD. ND-39i males were crossed with pnt-Gal4 females to generate ND-39i/+; pnt-Gal4/+ progeny, as the ND-39i/CyO-GFP; pnt-Gal4, CD8-GFP stock was unstable. ND-39i/+; pnt-Gal4/+ males were subsequently crossed with Drp1SD;SOD2i virgin females. Third instar larval brains were dissected, and GFP-positive brains (expressing pnt-Gal4, mCD8-GFP) were selected for analysis. Non-CyO-GFP larvae were chosen to ensure the presence of Drp1SD in all samples. Brains carrying ND-39i in combination with Drp1SD were identified based on lineage morphology. ND-39-depleted lineages are characteristically smaller with ruffled cellular boundaries compared to controls. Immunofluorescence Wandering third-instar larvae were selected to dissect the brains or cuticle muscle in Schneider’s media (Invitrogen-Thermofisher, 21720024), followed by 4% paraformaldehyde fixation for 25 min followed by washing in 0.1% PBST (1× PBS, 0.1% Triton X-100) for 30 min at room temperature (RT) and blocked with 1% BSA in 0.1% PBST for 1 h at RT. The tissues were then incubated with primary antibodies (diluted in 1% BSA) overnight at 4°C on a shaker. They were washed thrice (20min, 10min, 10min) with 0.1% PBST and incubated with fluorescently coupled secondary antibodies (diluted in 0.1% PBST) for 1 h at RT followed by 20 min washing with 0.1% PBST. The tissues were stained with the DNA dye, Hoechst for 10 min, washed for 10 min with 0.1% PBST, and then mounted in SlowFade Gold Antifade Mountant. The antibodies used were listed in key resources table. ROS detection Third instar larval brains were dissected in Schneider’s media, and each of the brain lobes was carefully cut at the outer margin of the optic lobe to facilitate the penetration of dihydroethidium (DHE) dye into brain tissue. The brains were then incubated in 30 μM DHE for 15 min at RT and washed with Schneider’s media for 10 min before being mounted in a poly-L-lysine coated LabTEK chamber, immersed in Schneider’s media. A coverslip was placed over the brain very gently to avoid movement. Imaging was performed immediately using the Zeiss LSM710 confocal system with a 63×/1.4 NA Oil objective and excitation/emission of 561nm/605nm. TMRM staining Larval brains were dissected out in Schneider’s media and treated with 1:150 D(+) Limonene (diluted in Schneider’s media) for 4 min. Brains were then immersed in 100 nM Tetramethylrhodamine methyl ester (TMRM) dye for 30 min followed by washing in Schneider’s media for 5 min and mounted in a poly-L-lysine coated LabTEK chamber, immersed in Schneider’s media. Imaging was performed immediately as described above. EdU uptake assay EdU assay was performed by using a Click-iTTM EdU Alexa FluorTM 647 imaging kit. Third-instar larvae were dissected in Schneider’s medium and immediately treated with 10μM EdU (5-ethynyl-2′-deoxyuridine) for 15 min at RT. EdU incorporation reaction in DNA was stopped by fixation in 4% PFA for 25 min followed by washing in 0.1% PBST for 30 min at RT and blocked with 1% BSA in 0.1% PBST for 1 h at RT. The brains were then incubated with the required primary antibodies overnight (∼14 h) at 4°C on a shaker. The next day, brains were washed for 20 min with 0.1% PBST and incubated with appropriate secondary antibodies for 1 h at RT, followed by 20 min of washing with 0.1% PBST. The Click-iT reaction cocktail was prepared according to the manufacturer’s instructions. The brains were incubated in it for 30 min at RT, followed by 10 min of washing with 0.1% PBST. The brains were treated with 10 μg/mL Hoechst in 0.1%PBST for 10 min and then washed for 5 min with 0.1% PBST before being mounted in Slow-Fade Gold. Imaging was performed using a Zeiss LSM710 with a 63×/1.4NA oil objective. ATP dye uptake assay Third instar larval brains were dissected in Schneider’s media, treated with 1:150 D (+) Limonene (diluted in Scheiders media) for 4 min and immersed in 30μM ATP-red dye for 30 min. They were washed with Schneider’s media for 5 min and mounted in a poly-L-lysine coated LabTEK chamber, immersed in Schneider’s media. Imaging was done immediately using the Zeiss LSM710 with 63×/1.4 NA Oil objective using 561 nm laser. Microscopy Imaging of fixed samples Fixed samples were imaged at room temperature using a Zeiss LSM710 or LSM780 inverted confocal microscope at the IISER Pune microscopy facility. A Plan-Apochromat 40× (1.4 NA) or 63× (1.4 NA) oil objective was used for image acquisition. Images were captured at 1024 × 1024 pixels with an averaging of 2 and an acquisition speed of 8 using Zen2010 software. Fluorescence intensity was maintained between 0 and 255 range on a 8-bit scale. During confocal microscopy, Hoechst, was excited using a 405 nm Diode laser, Argon laser (488 nm) was used for Alexa Fluor 488, DPSS laser at 561 nm was used to image Alexa Fluor 568 and Alexa Alexa Fluor 647 signals were acquired using a 633 nm HeNe laser. Representative images in Figures depict type II lineage, chosen to display the maximum number of cells within that lineage. Live imaging of type II neuroblast For live imaging of type II neuroblasts, third instar larval brains were dissected at room temperature in Schneider’s Drosophila media. An inverted ZMP LSM710 confocal microscope was used to scan the brains on a 35 mm dish in Schneider’s media supplemented with 10% FBS and 10 mM insulin (I0516, Sigma) at 3-min intervals for three hours at 25°C. Transmission electron microscopy Third instar larval brains were dissected in Schneider’s Drosophila media at room temperature. Brains were immediately fixed in 3% glutaraldehyde (Sigma, G5882) for 2h at 4°C. Washed twice with 0.1 M Sodium cacodylate buffer, pH 7.4, for 10 min each. Then, they were fixed in 1% Osmium tetroxide (OsO4) for 1 h at 4૦C. Brains were en-block stained with 2% uranyl acetate and dehydrated for ten minutes each in an increasing alcohol series. Araldite resins A and B were used to infiltrate the tissues, and Araldite resin B was used to embed the tissues and polymerize them for 48 h at 60૦C. Ultra-microtome (LeicaUC7) was used for ultrathin brain slices of 70 nm were cut and collected on copper slot grids. Lead citrate was used to contrast the sections. The transmission electron microscope, JEM 1400 Plus, JEOL (Japan), at 120 kV, was used to take the images. Neuroblasts were identified due to their distinct large size in the central brain. Quantification and Statistical analysis Mitotic index quantification Neuroblasts in the central brain hemisphere were quantified using Miranda as a cortical marker marked with mCD8-GFP 0.05, ∗-p < 0.05, ∗∗-p < 0.01, and, ∗∗∗-p < 0.001.