The adult Drosophila central brains are post-mitotic

(A) Left: representative confocal images of adult control (w1118) central brains ubiquitously expressing Fly-FUCCI cell-cycle reporters. EGFP::E2F1(1–230) (green) marks cells in G1/G0 phase, whereas mRFP::NLS-CycB (1–266) (red) marks cells in S/G2/M phases. Brains were counterstained with DAPI (blue) and the mitotic marker phosphorylated histone H3 Ser10 (pH3, grey). Upper panels show anterior views and lower panels show posterior views. Nearly all cells in the adult central brain were GFP-positive and lacked detectable RFP signals, indicating that they reside in G1/G0 phase. No pH3-positive mitotic cells were detected. A small number of cells near the antennal lobes exhibited weak RFP signals, suggesting rare G2-phase cells. Scale bars: 50 μm. Right: schematic representation of the Fly-FUCCI system used to visualise cell-cycle states based on cell-cycle-dependent degradation of E2F1 and CycB fragments. (B) Top: tSNE plot showing cell clusters identified from adult Drosophila brain scRNA-seq data (Davie et al., 2018). Cells are classified into 17 clusters, with different colours indicating neuronal, glial, and neuroblast-like populations. Bottom: dot plot showing expression of 112 selected cell-cycle regulator (CCR) genes across the 17 clusters. Colour intensity indicates average expression level, and dot size indicates the percentage of cells within each cluster expressing the gene. CCR genes are arranged according to broad functional categories. (C) Dot plot showing expression of the same selected CCR genes across 29 cell clusters identified from larval Drosophila brain scRNA-seq data (Avalos et al., 2019). Colour intensity and dot size are as described in (B). Most positive CCRs are enriched in larval neuroblast clusters but are minimally expressed in adult neuronal and glial clusters, consistent with the postmitotic state of the adult brain.

Neuroblast-specific Kr depletion and the KrIf-1 mutation cause mushroom body neuroblast retention and prolonged neurogenesis in adult brains.

(A) Representative confocal images of adult central brains after EdU incorporation assays. Brains from Kr wild-type controls (Kr+/Kr+), heterozygous KrIf-1 mutants (KrIf-1/Kr+), UAS control flies (insc>lacZ) and Kr-depleted flies (insc>KrIR#1) were stained with DAPI (blue) and EdU (red). No EdU-positive clones were detected in control brains, whereas insc>KrIR#1 and KrIf-1 mutant brains contained multiple EdU-positive clones, predominantly in the dorsoposterior region. Scale bars: 100 µm. (B) Quantification of EdU-positive clones in adult brains aged 4–6, 12–14 and 19–21 days after eclosion. Scatter dot plots show the number of EdU-positive clones per brain hemisphere. Thick and thin red bars indicate means and SDs, respectively, with values annotated above. n indicates the number of brain hemispheres analysed. Statistical significance was determined using the Mann-Whitney U test. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p < 0.0001; ns, not significant. (C) Representative confocal images of Kr wild-type control and KrIf-1 mutant adult brains expressing mCD8::GFP (green) under the NB-specific insc-Gal4 driver. Brains were stained with DAPI (blue), pH3 (red) and Mira (grey). Mitotic and interphase NBs, identified by Mira and GFP expression, were detected in the dorsoposterior region of KrIf-1 mutant brains but not in controls. Scale bars: 50 µm. (D) Representative confocal images of insc>KrIR#2 adult brains expressing mCD8::GFP (green), stained with DAPI (grey) and Mira (red). insc>GFP- and Mira-positive mitotic NBs exhibited crescent-like cortical Mira localisation, characteristic of dividing NBs, whereas surrounding smaller cells lacked NB marker expression. Scale bars: 20 μm. (E) Representative confocal images of insc>KrIR#2 adult brains expressing mCD8::GFP (green), stained for EdU (red) and the pan-neuronal marker Elav (grey). EdU-positive cells were observed among Elav-positive neuronal progeny, indicating continued neurogenesis in adult brains following Kr depletion. Scale bars: 20 µm. (F) Representative images of wild-type adult brain expressing mCD8::GFP (green) under OK107-Gal4 driver to visualise MB structures. Brains were stained with DAPI (blue). Posterior and anterior views are shown. The MB cell body region is located dorsoposteriorly, whereas MB lobes are visible in the anterior view. Scale bars: 100 µm. (G) Representative confocal images of KrIf-1 mutant adult brains expressing mCD8::GFP under mb247-Gal4 (green), stained with DAPI (blue), pH3 (red) and Mira (grey). Mitotic cells in the MB cell body region showed cortical Mira localisation, while surrounding smaller cells lacked NB markers, consistent with dividing MBNBs undergoing asymmetric division. Scale bars: 20 µm. (H) Representative confocal images of KrIf-1 adult brains expressing mCD8::GFP under OK107-Gal4 after EdU labelling. EdU incorporation was detected in Mira-positive MBNBs, which were weakly marked by OK107>GFP. Scale bars: 20 µm.

Kr is expressed in MBNBs and regulates their elimination during the pupal stage

(A) Quantification of MBNB number during pupal development in control, insc>KrIR and KrIf-1 mutant brains. MBNB number progressively decreased during pupal development in controls but was maintained in insc>KrIR and KrIf-1 mutant brains. Bars indicate medians and interquartile ranges. Numbers indicate the number of brain hemispheres analysed. Statistical significance was determined using a Kruskal–Wallis test followed by Dunn’s multiple comparisons test. (B) Representative confocal images and quantification of MBNB size during pupal development. MBNBs were identified by insc>GFP and DAPI staining. Magnified views show individual MBNBs, with yellow lines indicating the measured MBNB diameter. Scale bars: 20 µm. Quantification shows MBNB diameter in early and late pupal stages. In controls, MBNB size decreased from early to late pupal stages, whereas MBNB size was maintained upon Kr depletion. Numbers indicate the number of MBNBs analysed. Early and late pupal stages were defined based on developmental morphology, as described in Materials and methods. (C) Developmental stage-specific knockdown of Kr using insc-Gal4, pTub-Gal80ts and KrIR#2. Left: experimental scheme showing nine temperature-shift conditions. Flies maintained at 29°C (pale red boxes) induced Kr RNAi, whereas flies maintained at 19°C (pale blue boxes) suppressed RNAi through Gal80ts. n indicates the number of brain hemispheres analysed for EdU-positive clone counting. Right: quantification of EdU-positive clones in adult brains from each condition. Scatter dot plots show the number of EdU-positive clones per hemisphere. Greyed boxes and red bars indicate means and SDs, respectively, with values annotated above. Statistical significance was determined using pairwise Mann-Whitney U tests. (D) Representative confocal images of posterior adult brain regions from condition 2 and condition 5 in (C), stained for DAPI (blue), Mira (green) and EdU (red). Persistent EdU-positive clones were observed when Kr RNAi was induced during the pupal stage. Scale bars: 100 µm. (E) Representative confocal images of the MB cell body region in wild-type pupal brains at 54–60 h APF carrying the Kr::GFP(Bac) reporter. Brains were stained with rat anti-Kr antibody (Rt α-Kr, red), rabbit anti-Kr antibody (Rb α-Kr, cyan) and DAPI (grey). Kr::GFP is shown in green. MBNBs, identified by their position and large cell size within the MB cell body region, showed weak Kr signals detected by both antibodies and the Kr::GFP reporter. Scale bars: 20 µm. (F) Quantification of Kr signal intensity in individual MBNBs during pupal development. Scatter dot plots show normalised cytoplasmic and nuclear Kr signal intensities at different pupal stages. Thick and thin red bars indicate means and SDs, respectively. n indicates the number of brain hemispheres analysed per condition. Statistical significance was assessed separately for cytoplasmic and nuclear signals using Welch’s ANOVA followed by Dunnett T3 multiple comparisons test. No pairwise comparisons reached statistical significance after multiple-comparison correction. See Materials and methods for details of quantification. Statistical significance is indicated as follows: *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p < 0.0001; ns, not significant.

Kr promotes MBNB elimination by regulating the Imp–Syp transition.

(A) Quantification of Imp- and Syp-expressing areas within MB cell body regions in adult brains. Left: representative 3D reconstructions of dorsoposterior adult brain surfaces from the indicated genotypes, stained for Imp (red), Syp (green) and Eyeless (Ey; blue), which marks MB lineages. Scale bars: 50 µm. Right: scatter dot plots showing the ratios of Imp-positive and Syp-positive areas to the MB cell body surface region on the dorsoposterior brain surface, quantified per hemisphere. Thick and thin red bars indicate means and SDs, respectively. n indicates the number of hemispheres analysed per condition. Statistical significance was determined using pairwise Mann–Whitney U tests. (B) Quantification of Imp and Syp signal intensities in MBNBs during late pupal development. Left: representative confocal images of MBNBs in 72–96 h APF pupal brains from control and insc>KrIR#2 flies, stained for Imp (red), Syp (grey), DAPI (blue) and insc>GFP (green). Yellow arrowheads indicate MBNBs. Scale bars: 50 µm. Right: scatter dot plots showing normalised Imp and Syp fluorescence intensities in individual MBNBs. Thick and thin red bars indicate means and SDs, respectively. n indicates the number of MBNBs analysed. Statistical significance was determined using unpaired two-tailed t-tests; Welch’s correction was applied for Syp because variances were unequal. (C) Imp depletion suppresses MBNB persistence in Kr RNAi brains. Left: representative confocal images of adult brains from insc>KrIR#2, mCherryIR and insc>KrIR#2, ImpIR, stained for pH3 (red), insc>GFP (green), Mira (grey) and DAPI (blue). Mitotic MBNBs persisted in insc>KrIR#2, mCherryIR brains but were not detected in insc>KrIR#2, ImpIR brains. Scale bars: 100 µm. Right: scatter dot plots showing the number of mitotic MBNBs per brain. Thick and thin red bars indicate means and SDs, respectively. n indicates the number of hemispheres analysed. Statistical significance was determined using a Mann–Whitney U test. insc>ImpIR alone caused lethality, preventing adult brain analysis. Statistical significance is indicated as follows: *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p < 0.0001; ns, not significant.

Kr antagonises Kr-h1 and supports E93 expression to promote MBNB termination.

(A) Kr-h1 co-depletion suppresses MBNB persistence in Kr RNAi adult brains. Left: representative confocal images of insc>Kr-h1IR, insc>KrIR#2, mCherryIR and insc>KrIR#2, Kr-h1IR adult brains stained for EdU (red), insc>GFP (green), Mira (grey) and DAPI (blue). Kr-h1 RNAi alone did not prevent MBNB elimination, whereas the number of persistent EdU-positive MBNB clones observed in insc>KrIR#2, mCherryIR brains was significantly reduced when Kr-h1 was co-depleted with Kr. Scale bars: 100 µm. Right: quantification of EdU-positive MBNB clones per hemisphere in the indicated genotypes. Scatter dot plots show individual data points, with means and SDs represented by thick and thin red bars, respectively. n indicates the number of hemispheres analysed. Statistical significance was determined using a Kruskal–Wallis test followed by Dunn’s multiple comparisons test. (B) Kr depletion reduces E93 expression in MBNB progeny without detectably altering Kr-h1::GFP levels in MBNBs. Left: representative confocal images of 72–96 h APF pupal brains from control insc>Kr-h1::GFP and Kr-depleted insc>KrIR#2; Kr-h1::GFP flies, stained for E93 (red), Kr-h1::GFP (green), DAPI (grey) and Mira (blue). MBNBs were identified by Mira staining and cell morphology, and the surrounding progeny region was analysed separately. Scale bars: 50 µm. Right: quantification of normalised Kr-h1::GFP signal intensity in MBNBs, and normalised E93 signal intensities in MBNBs and surrounding progeny. Scatter dot plots show individual data points, with means and SDs represented by thick and thin red bars, respectively. n indicates the number of hemispheres analysed. Statistical significance was determined using unpaired two-tailed t-tests. (C-E) Kr-h1 overexpression promotes NB proliferation and blocks neuronal differentiation, leading to a tumour-like NB overgrowth phenotype. (C) Representative posterior views of adult brains overexpressing Kr-h1 in NBs (insc>Kr-h1ORF::FLAG), stained for DAPI (blue), insc>GFP (green), EdU (red) and Mira (grey), with magnified views of the MB cell body regions shown on the right. Most EdU-positive cells within enlarged clones co-expressed Mira and insc>GFP, indicating their NB-like status and a failure in differentiation. Scale bars: 100 µm. (D) Confocal images of insc>Kr-h1ORF::FLAG adult brains stained for DAPI (blue), insc>GFP (green), EdU (red) and Elav (grey). EdU-labelled clones in Kr-h1-overexpressing brains contained very few Elav-positive neurons, indicating impaired neuronal differentiation. Scale bars: 100 µm. (E) Confocal images of insc>Kr-h1ORF::FLAG adult brains stained for DAPI (grey), insc>GFP (green) and Imp (red). Scale bars: 20 µm. Statistical significance is indicated as follows: *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p < 0.0001; ns, not significant.

Kr coordinates MBNB termination and mushroom body development.

(A) Representative confocal images of anterior views of adult brains from the indicated genotypes, with MB lobe structures visualised by OK107>GFP (green). No overt structural abnormalities were observed in MB lobes following Kr depletion or in KrIf-1 mutant adult brains. Scale bars: 50 µm. (B) Representative images of MB α/β lobe morphology in control and insc>KrIR#2 adult brains, visualised by Fasciclin II (FasII) staining (red). Kr depletion led to thinner and more curved α/β lobes compared to controls. Scale bars: 50 μm. (C) Effects of Kr or Kr-h1 overexpression in MB lineages on MB development. Representative confocal images of anterior views of adult brains from OK107>mCherryIR control, OK107>KrIR#2, OK107>KrORF::V5 and OK107>Kr-h1ORF::FLAG flies, are shown, with MBs visualised by OK107>GFP (green) and FasII staining (red). Kr overexpression caused high lethality; surviving OK107>KrORF::V5 adult flies exhibited severe disorganisation of MB lobes, including pronounced reduction of α/β and α′/β′ lobes and extensive disorganisation of γ lobes. Kr-h1 overexpression also caused severe MB disorganisation, with persistent FasII-positive structures in the β and/or β′ lobe regions. Scale bars: 50 μm. (D) Proposed model for Kr function in coordinating MBNB termination. During larval development, high Imp and Kr-h1 levels maintain MBNB proliferation and early neuronal fate. As development progresses into the pupal stage, the intrinsic temporal programme shifts towards high Syp and E93 expression, promoting cell cycle exit, neuronal maturation and eventual MBNB elimination. Kr acts as an MBNB-specific coordinator that promotes this transition by suppressing the persistent Imp-dominated programme while facilitating Syp and E93 activation within the MBNB lineage. In parallel, Kr-h1 antagonises E93 as part of the ecdysone-responsive transcriptional programme. Solid regulatory arrows indicate relationships supported by data in this study, whereas dashed inhibitory arrows indicate regulatory interactions reported in other developmental contexts or inferred from the literature but not directly tested in MBNBs in this study. Activin/Babo signalling and ecdysone signalling provide extrinsic developmental cues that regulate the timing of this transition. Our data suggest that Kr does not primarily regulate E93 through detectable changes in Kr-h1 expression, but instead promotes the E93-dependent termination programme through a parallel pathway while simultaneously coordinating the Imp–Syp temporal transition. Loss of Kr disrupts these coordinated transitions, resulting in prolonged MBNB proliferation and continued neurogenesis in adult brains.

Expression of CCR genes across 84 cell clusters in the adult Drosophila brain.

Top: tSNE plot showing 84 cell clusters identified in the adult Drosophila brain scRNA-seq data (Davie et al., 2018). Bottom: dot plot showing expression levels of 112 selected CCR genes across the 84 clusters. Colour intensity indicates average expression level, and dot size indicates the percentage of cells within each cluster expressing the gene. Positive CCRs, including Cdk1, Cdk2, CycB and Polo, are largely absent from neuronal and glial clusters but are enriched in a small neuroblast-like population. In contrast, negative regulators such as fzr, Wee1 and rux remain broadly expressed across neuronal clusters, consistent with maintenance of the postmitotic state in the adult brain.

Forced expression of positive CCRs induces mitotic entry in postmitotic neurons.

(A) Overexpression of Dp-E2F1 and CycE-Cdk2 in dopaminergic neurons using TH-Gal4 (left panels) and in mushroom body neurons using mb247-Gal4 (right panels) induces mitotic entry in young adult flies, ∼2 days post-eclosion. Gene expression was induced by shifting flies to 29°C for 10 days. Brains were stained for DAPI (blue), GFP reporter expression (green) and pH3 (red). pH3-positive neurons were detected following E2F1-Dp and Cdk2-CycE overexpression, whereas no pH3-positive neurons were observed in controls. When gene induction was initiated in older flies, 10 days post-eclosion, no pH3-positive neurons were detected (n ≥ 10), suggesting that neurons become increasingly refractory to cell-cycle re-entry with age. Scale bars: 50 μm. (B) Overexpression of Dp-E2F1 and CycE-Cdk2 in dopaminergic neurons during late pupal development (∼2 days before eclosion) allowed some adult flies to emerge; however, these flies exhibited premature lethality, dying within 3–5 days after eclosion. Brains dissected 5 days after gene induction contained pH3-positive neurons (12 out of 27 brains examined). Some of these neurons also showed Dcp-1 staining (grey), indicating apoptosis. Overexpression of the same CCR combination in the mushroom body neurons using mb247-Gal4 caused lethality before adult eclosion. Scale bars: 50 μm. (C) Kaplan-Meier survival analysis of adult flies overexpressing Dp-E2F1 and CycE-Cdk2 in dopaminergic neurons. Experimental flies (TH>E2F-Dp, CycE-Cdk2, mCD8::GFP, gal80ts) and controls (TH>lacZ, mCD8::GFP, gal80ts) were shifted from 19°C to 29°C two days after eclosion to induce transgene expression. Flies overexpressing E2F1-Dp and Cdk2-CycE exhibited significantly reduced lifespan compared to controls (P < 0.0001, log-rank Mantel-Cox test). Pale blue and pale red lines represent three individual biological replicates, and dark blue and red lines represent mean survival curves. Error bars indicate SD.

Kr expression in MBNBs during development

(A, B) Representative confocal images of stage 12 wild-type embryonic central nervous system (CNS) stained with rabbit anti-Kr antibody (red), OK107>GFP (green), DAPI (blue), and Mira (grey). (A) Overview of the embryonic CNS and magnified views of the MB lineage region, with merged images and individual greyscale channels shown. Yellow asterisks indicate MBNBs identified by OK107>GFP and Mira co-expression. (B) Representative images of the MB lineage region. Yellow asterisks indicate MBNBs, and yellow arrows indicate a neuron within the MB lineage expressing Kr. Although Kr was broadly expressed in the embryonic CNS, it was not detected in MBNBs. Scale bars: 20 µm. (C-E) Kr expression in third-instar larval brains. (C) Confocal image of the MB cell body region stained for OK107>GFP (green), Mira (cyan), and Kr (red) using a rat anti-Kr antibody. The individual Kr channel is shown in grey. Yellow asterisks mark MBNBs, which exhibit weak cytoplasmic Kr signals. (D) Kr::GFP(Bac) reporter expression in the MB cell body region, showing Kr::GFP (green), DAPI (grey), and Mira (red). The individual GFP channel is shown in grey. Asterisks mark MBNBs. Strong Kr::GFP signals were detected in single cells adjacent to MBNBs, likely corresponding to GMCs or early-born neurons that may have inherited Kr::GFP from MBNBs. (E) Confocal image of the MB calyx region stained for OK107>GFP (green), Mira (cyan) and Kr (red), with the individual Kr channel shown in grey. Kr-expressing neurons near the MB calyx are indicated by arrowheads. Scale bars: 20 µm. (F) Kr antibody signal is reduced following Kr depletion. Left: representative confocal images of 24-48 h APF pupal brains from control and insc>KrIR#2 conditions, stained with rat anti-Kr antibody (red), DAPI (grey) and insc>GFP (green). Scale bar: 50μm. Right: quantification of normalised cytoplasmic Kr signal intensity. Scatter dot plot shows individual MBNB measurements. Thick and thin red bars indicate means and SDs, respectively. Statistical significance was determined using an unpaired t-test. See Materials and methods for details on quantification. (G, H) Representative confocal images of KrIf-1 mutant late pupal (G) and adult (H) brains stained with rat anti-Kr antibody (red) and DAPI (grey) using tyramide signal amplification (TSA). Yellow asterisks indicate MBNBs identified by their position and DAPI morphology. Scale bars: 100 µm. (I) Quantification of TSA-enhanced Kr signal intensities in individual MBNBs in KrIf-1mutant pupal brains, showing Kr misexpression in MBNBs during late pupal stages in KrIf-1 mutants. Statistical significance is indicated as follows: *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p < 0.0001; ns, not significant.

Imp expression persists in MBNBs in Kr RNAi and KrIf-1 mutant flies.

(A) Representative confocal images of dorsoposterior regions of wild-type, KrIf-1 mutant and insc>KrIR adult brains. EdU labelling (red) marks proliferating cells, Imp expression (green) was visualised using an Imp-specific antibody, and DAPI (blue) marks DNA. Compared to controls, Imp-expressing regions around the MB cell body region were expanded in KrIf-1 and insc>KrIR brains. Scale bars: 50 µm. (B) Representative confocal images of dorsoposterior MB cell body regions in wild-type, KrIf-1 and insc>KrIR adult brains. EdU labelling (red) marks proliferating cells, Syp expression (green) was visualised using Syp-specific antibodies, and DAPI (blue) marks DNA. These images correspond to the Syp analysis shown in Fig. 4A. Scale bars: 50 µm. (C) Validation of Imp depletion in Kr RNAi adult brains. Representative confocal images of control insc>mCherryIR, insc>KrIR#2, mCherryIR and insc>KrIR#2, ImpIR adult brains stained for Imp (red), insc>GFP (green) and DAPI (blue). Expression of Imp RNAi under insc-Gal4 efficiently depleted endogenous Imp proteins in brain tissue, including the MB lineage region. Scale bars: 50 µm.

E93 expression in the MB lineages is regulated by Kr-h1.

(A) Developmental expression profiles of temporal factors in the MBNB lineage. Expression levels (mean ± SD) of selected temporal regulators in MB lineages at 24 h ALH, 50 h ALH, 84 h ALH and 36 h APF are shown. Values were calculated from three biological replicates using the MB lineage-specific RNA-seq dataset reported by Liu et al. (2015). Imp, chinmo and Kr-h1 are highly expressed during larval stages and decline during pupal development, whereas Syp and E93 are strongly induced during the larval-to-pupal transition. Kr expression remains relatively low throughout development. (B) Kr-h1 regulates E93 expression in the MB lineage during late pupal development. Left: representative confocal images of 72–96 h APF pupal brains from control insc>mCherryIR, mCherryIR, insc>Kr-h1IR, mCherryIR and insc>Kr-h1ORF::FLAG, mCherryIR flies, stained for E93 (red) and insc>GFP (green). Dotted lines indicate insc>GFP-positive MB lineage regions. Scale bars: 50 µm. Right: scatter dot plots showing normalised E93 signal intensity in the MB lineage region, quantified per hemisphere. Thick and thin red bars indicate means and SDs, respectively. n indicates the number of hemispheres analysed per condition. Statistical significance was determined using pairwise Mann–Whitney U tests. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p < 0.0001; ns, not significant. (C) Representative confocal images showing Kr-h1::GFP(Bac) expression in MBNBs in third-instar larval brains. Kr-h1::GFP was enriched in the nuclei of MBNBs. Scale bars: 10 µm.