Casein kinase 2 maintains the self-renewal of neural stem cells via prospero phosphorylation in Drosophila

Abstract

Most neural stem cells (NSCs) maintain self-renewal to ensure proper development of central nervous system (CNS) until a specific time and undergo terminal differentiation after neurogenesis. However, the regulation of the NSC maintenance during brain development is not well understood. Our study reveals that casein kinase 2 (CK2) is essential for maintaining NB self-renewal in late larval stages in Drosophila. Deficiency of CK2 results in increased nuclear localization of the differentiation promoting factor Prospero (Pros), and consequently a significant reduction in NB number. Mechanistically, Pros is phosphorylated by CK2 and sequestered in the cytoplasm, which prevents premature differentiation of the NB in late larval stages. While in pupal stages, the developmentally declined expression of CK2 allows NBs to undergo terminal differentiation. Overall, our study uncovers a critical role of CK2 as a time switch on NB transition from self-renewal to differentiation, highlighting a novel mechanism of the temporal regulation of neurogenesis by post-translational modifications (PTMs).

Supplementary Information

The online version contains supplementary material available at 10.1186/s13578-026-01532-z.

Introduction

In mammals, neural stem cells (NSCs) maintain a stable population within a specific time window while also producing differentiated progenies [1]. The NSC pool expands rapidly in early embryonic stage and remains relatively stable during neurogenesis [2, 3]. Then the population of NSCs progressively declines, with only a limited number persisting in restricted brain regions [4]. The temporal regulation of NSC maintenance is critical for proper formation and homeostasis of the larval central nervous system (CNS) [5]. Disrupting the temporal regulation of NSC maintenance, whether through premature depletion or prolonged persistence of NSCs, may result in neurodevelopmental disorders due to aberrant neuronal population sizes [6–8]. Therefore, identifying the key regulators and elucidating mechanisms governing the time window of NSC maintenance is pivotal to deciphering the cellular basis of nervous system development.

In Drosophila, the NSCs (neuroblasts, NBs) are also maintained in specific time window [9, 10]. The NBs in the central brain (CB) and the ventral nerve cord (VNC) detach from the neuroectoderm in the embryonic stage, which can be characterized by biomarkers of Miranda (Mira), Asense (Ase), and Deadpan (Dpn) [11, 12]. In the CB, around 90 NBs are maintained from embryonic to early pupal stages, after which they undergo progressive elimination and are completely absent before adulthood [10, 13, 14]. The dynamic pattern of NB number during neurogenesis resembles that of mammalian NSCs, which makes NB an ideal model to study the temporal regulation on NSC maintenance [15–17]. For instance, the steroid hormone ecdysone, acting through the Mediator complex, initiates a metabolic shift to oxidative phosphorylation, which uncouples the cell cycle from growth and drives a reduction in NB size, ultimately leading to cell-cycle exit [18]. Although several factors influencing NB maintenance have been identified, the underlying temporal factors associated with the time window of NB maintenance remains enigmatic.

Prospero (Pros)/Prox1, an evolutionarily conserved homeobox transcription factor, acts as a critical differentiation promoting factor of NSCs during neurogenesis, which makes it also important for NB maintenance [19–21]. During larval stages, Pros is expressed in NBs, but remains in the cytoplasm through Mira-mediated cortical anchoring, so that its differentiation-promoting transcriptional activity is effectively suppressed [22, 23]. In contrast, during pupal stages, Pros is translocated into the NB nucleus to promote terminal differentiation [19, 24]. Therefore, the subcellular localization of Pros to either the nucleus or the cytoplasm is critical to NB fate determination in different stages, and the nuclear Pros can be regarded as a differentiation marker of NBs [25]. Previous studies have shown that post-translational modification (PTM) factors, for instance, phosphatase PP4 excludes Pros and Mira from the nucleus during interphase/prophase [26, 27]. These factors collectively regulate NB self-renewal and differentiation via manipulating Pros localization. The subcellular localization of Pros in NBs during larval or pupal stages is also regulated by temporal factors. For instance, Castor, one of NB-specific temporal series, interacts with Hedgehog signaling pathway to promote Pros nuclear localization to trigger cell cycle exit in pupal stages [28]. However, the detailed regulatory network controlling Pros localization across developmental stages is not fully understood.

Casein kinase 2 (CK2) is a conserved serine/threonine kinase, which is a heterotetramer composed of two catalytic subunits (CK2α) and two regulatory subunits (CK2β) [29, 30]. As CK2 is constitutively active and ubiquitously distributed in eukaryotes, its protein substrates make up a substantial proportion of the phosphoproteome and implicated in a major, if not all developmental processes, consistent with the fact that homozygous deletion of CK2 is early embryonic lethal [31, 32]. CK2 is widely expressed in the CNS, including NSCs, and is essential for brain development [33, 34]. Notably, variants in human CK2 leading to its functional dysregulation are correlated with neurodevelopmental disorders, including seizures and Okur–Chung syndrome [35–37]. Previous researches have demonstrated that CK2 is involved in cell cycle and survival pathways in leukemia stem cells and is required for human adipogenic stem cell differentiation, indicating that CK2 may be involved in stem cell maintenance [38, 39]. Altogether, this evidence prompts us to investigate whether CK2 functions in NSCs to regulate neurodevelopment.

In this study, we identify CK2 as a developmentally programmed temporal regulator that is essential for NB maintenance in late larval stages in Drosophila. Mechanistically, CK2 phosphorylates Pros to localize it in the cytoplasm, thereby preventing the premature nuclear entry of Pros and differentiation initiation in NBs. These findings indicate CK2 acts as a time switch via Pros phosphorylation that maintains NSC self-renewal during neurogenesis.

Materials and methods

Fly genetics

Fly strains were bred on yeast-containing medium at a constant of temperature 25℃ in 12 h light/dark cycle unless otherwise stated. UAS-CK2α-RNAi (NIG 17520R1) and UAS-CK2β-RNAi (NIG 15224R1) were used for most experiments except Fig. 1—figure supplement 1 H (VDRC: 330507, VDRC:32377). UAS-CK2α (FlyORF F001005). UAS-CK2β (FlyORF F001372). UAS-P35 (BDSC: 5072). UAS-dcp1-RNAi (BDSC: 38315). UAS-mCherry-RNAi (BDSC: 35785). UAS-lacZ (BDSC: 3955). UAS-pros-RNAi (THU2372). UAS-3XFLAG-pros.S (BDSC: 32245). Insc-gal4, tubulin-gal80; UAS-Dicer2 (generated in this work). Insc-gal4, UAS-Dicer2; UAS- CD8::GFP (generated in this work). UAS-Pros^A and UAS-Pros^D (generated in this work).

Fig. 1.

CK2 is required for maintaining NB number and proliferation. A Third-instar larval brains expressing the indicated RNAi transgenes were driven by insc-gal4. Dpn, NB nuclei, red; PH3, proliferative marker, green; DAPI, nuclei, blue. Scale bars: 50 μm. B–C Quantification of NB number(B) and proliferation rate(C) from genotypes in (A). D Rescue of NB number and proliferation when simultaneously overexpressing wild-type CK2α and CK2α RNAi (or CK2β and CK2β RNAi) in NB. Dpn, red; PH3, green; DAPI, blue. Scale bars: 50 μm. E–F Quantification of NB number (E) and proliferation rate (F) in (D). B, n = 13; C, n = 10; E and F, n = 13; Statistical results were presented as means ± SEM, p values were performed by one-way ANOVA with a Bonferroni test; ****p < 0.0001

CK2α and CK2β antibody production

For the generation of antibodies against Drosophila CK2α and CK2β, we generated N-terminal His-tagged full-length constructs using the Pet30a vector via the restriction enzyme sites NdeI and HindIII. The accuracy of the final expression construct was confirmed by restriction enzyme digestion and sequencing. The construct was then transformed into the BL21(DE3) expression strain. The protein was induced for expression using IPTG and subsequently purified by affinity chromatography using Ni-IDA resin (Nanjing DetaiBio). Following primers were used:

CK2α N-term: CACCATGACACTTCCTAGTG,

CK2α C-term: TTATTGCTGATTATTGGGAT;

CK2β N-term: CACCATGAGCAGCTCCGAGGAAG,

CK2β C-term: TTAGTTTTTCGCTCGTAGTGG.

DNA and plasmids

pJFRC28-10XUAS-IVS-GFP-p10 vector (Addgene Plasmid #36431) which has 10xUAS, the attB site and miniwhite marker was used as the backbone. For each construct, a NotI/XbaI fragment containing either the dv-pros-pmA or dv-pros-pmD coding sequence was synthesized in vitro and cloned into the corresponding restriction sites of the linearized pJFRC28 vector, replacing the GFP coding sequence, to generate pUAS-dv-pros-pmA and pUAS-dv-pros-pmD plasmids respectively.When the injected P0 embryos grew into adults, they were crossed with TM2/TM6B. F1 flies were screened for candidates that carried mini white marker in eyes (orange eye). PCR was performed using primers for validation of 10UAS-pros-pmA integration. Then, the flies were balanced with TM6B. Primers for validation, HSP-F: AAGTAACCAGCAACCAAGTA; dv-pros-pmA-5R: ATGCTGTTCGCCGCTACTGGCG, dv-pros-pmD-R: GTGCTGTTCTCCGCTGGAGGCA).

Immunohistochemistry

Drosophila larvae were dissected in phosphate-buffered saline (PBS), and the larval brains were fixed in 4% EM-grade formaldehyde in PBST for 40 min. After washing thrice with 0.3% PBST (10 min each), brain samples were blocked with 3% BSA in 0.3% PBST for 1 h at room temperature. Blocked brain samples were incubated with primary antibodies diluted in 3% BSA overnight at 4 °C. Following this, they were again washed thrice with 0.3% PBST (10 min each) and incubated with secondary antibodies diluted in 0.3% PBT for 1.5 h, and after washing four times with 0.3% PBT (10 min each), larval brains were mounted onto Vector shield (Vector Laboratory) for Confocal microscopy.

Primary antibodies: Guinea pig anti-Dpn (1/1000, gift from Juergen A. Knoblich), rabbit anti-PH3 (1/100, CST, 9701), rat anti-Mira (1/1000, abcam, ab197788), rabbit anti-GFP (1/2000, Thermo Fisher Scientific, A10260), rabbit anti-Dcp1 (1/100, CST, 9578), mouse anti-Pros (1/20, DSHB, MR1A), rabbit anti-PAX6 (1/200, proteintech, 12323-1-AP), mouse anti-Ki67 (1/1600, CST, 9449),rabbit anti-CK2α (1/1000, this study), rabbit anti-CK2β (1/1000, this study).

The secondary antibodies used were conjugated with goat anti-rat Alexa 555 (1/100, A21434), goat anti-guinea pig Alexa 555 (1/100, A21435), donkey anti-rabbit Alexa 488 (1/100, A21206), donkey anti-rat Alexa 488 (1/100, A21208), donkey anti-mouse Alexa 488 (1/100, A21202), goat anti-rat Alexa 488 (1/100, A21094), donkey anti-mouse Alexa 647 (1/100, A31571). All immunostaining images were captured using an LSM700 or an LSM900 (Zeiss) confocal microscope.

Protein extraction and Western blots

Brains were dissected in PBS, and then homogenized in the RIPA buffer with protease phosphatase inhibitors (Roche). Homogenate was centrifuged at 12,000 rpm for 15 min at 4℃ and boiled with SDS for 10 min. The protein was separated on 4–12% SDS-PAGE and transferred to PVDF membrane. The membranes were blocked in 5% milk and incubated overnight at 4℃ with primary antibodies. Following extensive washing, the membranes were incubated with appropriate HRP-conjugated secondary goat antibodies in PBST containing 5% milk for 1.5 h at room temperature. The blots were imaged via the Tanon 5200 imaging system. The primary antibodies used for western blotting analyses in this study were as follows: rabbit anti-CK2α (1/1000, this study), rabbit anti- CK2β (1/1000, this study), mouse anti-tubulin (1:5000, Sigma‒Aldrich), anti-Pros (1:2000, DSHB, MR1A).

Co-Immunoprecipitation (Co-IP) assay

Drosophila larval brains were dissected at certain time points, and were collected and lysed in NP-40 lysis buffer along with protease inhibitors (Roche) and phosphatase inhibitors (Roche) for 30 min on a rotor at 4 °C. Protein A/G magnetic beads (Thermo Fisher Scientific, 8803) were conjugated with an anti- CK2α (1:1000, generated in this work) or control IgG at room temperature for 1 h. The antibody-bound beads were then incubated with total protein extracts at 4 °C for 2–4 h with rotation. After washing with NP-40 lysis buffer, the immunoprecipitated complexes were eluted by boiling in SDS-PAGE loading buffer and analyzed by western blotting. Blots were probed with mouse anti-Pros (1:2000, DSHB, MR1A). The secondary antibodies used were conjugated with HRP (Horseradish Peroxidase).

Nuclear-Cytoplasmic fractionation of third-instar larval brain

100 larval brains were dissected in ice-cold 0.7% NaCl and homogenized in 150µL ice-cold Cytoplasmic Extraction Buffer (CytoEB1X, CytoEB2X: 30 mM Hepes-KOH pH 7.6, 20 mM KCl, 10 mM MgCl2, 0.2 mM EDTA pH 8.0, 20% Glycerol) supplemented with protease and phosphatase inhibitors (the experimental group used for the dephosphorylation assay did not receive phosphatase inhibitors). The homogenate was subjected to a series of centrifugations: first at 600 × g for 1 min to remove debris, then the supernatant was sequentially centrifuged at 5000 × g and 14,000 × g for 10 min each at 4 °C to obtain the cytoplasmic fraction. The pellet was washed sequentially with 500µL WASH150 (CytEB 1X, 150 mM Sucrose) and WASH250 (CytEB 1X, 250 mM Sucrose) buffers at 4 °C and 5000 × g for 10 min to remove residual cytoplasmic contaminants. Nuclear proteins were extracted by resuspending the pellet in Nuclear Extraction Buffer (NEB, 350 mM Sucrose, 15 mM Hepes-KOH pH 7.6, 385 mM KCl, 5 mM MgCl2, 0.1 mM EDTA pH 8.0, 0.05% Tween 20, 10% Glycerol), followed by vigorous vortexing and centrifugation to obtain the Nuclear Fraction (NF). All steps were performed on ice to minimize protein degradation.

Dephosphorylation assay

The cytoplasmic fraction was divided into two aliquots for dephosphorylation assays. One aliquot was incubated with 1× FastAP™ reaction buffer only (control), while the other was treated with 10 U of FastAP™ Thermosensitive Alkaline Phosphatase (Thermo Fisher Scientific, EF0651) at 37 °C for 1 h to allow for dephosphorylation. After incubation, 1X SDS loading buffer was added to the samples, which were subsequently boiled at 95 °C for 10 min to denature the proteins.

Temperature shift

GAL80^ts was introduced into this system to control the function of GAL4 according to the temperature for determining the temporal window during which CK2 functions to maintain NBs. The eggs of tubulin-GAL80^ts, insc-gal4 crossing with UAS- CK2α RNAi or UAS- CK2β RNAi were collected, and cultured at 18℃ (permissive temperature for GAL80^ts) before hatching, then transferred to 29℃ restrictive temperature for GAL80^ts) after 24 h ALH, or vice versa until third-instar larval stage. Then brains were observed and analyzed.

Quantifications and statistical analysis

For quantification of NBs, Dpn- or Mira-positive NSC of CB or the thoracic VNC at the indicated stage were counted. Mitotic index is the number of PH3-positive cells among Dpn or Mira-positive cells. For quantifications of fluorescence or western blot intensity, NIH ImageJ was used to measure.

Statistical analysis was performed using GraphPad Prism 10. P values were calculated by performing two-tailed, unpaired Students’ T-test and one-way ANOVA with a Bonferroni test. In all graphs, ∗ indicates p < 0.05, ∗∗indicates p < 0.01, ∗∗∗indicates p < 0.001, ∗∗∗∗indicates p < 0.0001, ns indicates p > 0.05. Cell population numbers were quantified and represented as mean ± SEM and/or percentage of phenotype. Each experiment has at least three independent replications.

Protein fluorescence quantification

For all comparative analyses of protein fluorescence intensity, control and experimental larvae were processed in parallel under identical conditions for dissection, fixation, and immunostaining. Confocal images were acquired using the same settings. Fluorescence quantification was performed using ImageJ. The mean gray value of the protein to be assessed was measured by selecting a region of interest (ROI) outlining the cell on the original confocal image. The average fluorescence intensity of the selected area was measured, and the mean background fluorescence, determined from at least three adjacent cell-free regions, was subtracted. The total protein intensity was calculated as the product of the cell area and the background-corrected mean intensity.

Human brain organoid maintenance and organoid generation

H9 human embryonic stem cells (H9 ESCs; Shanghai Huzhen Industrial Co., Ltd.) were maintained in Essential 8 (E8) medium (Gibco, A1517001) on Matrigel (Corning)-coated plates under standard culture conditions. On day 0, H9 ESCs were dissociated into single cells using Accutase (STEMCELL Technologies). The cell suspension was resuspended in E8 medium supplemented with 50 µM ROCK inhibitor (Y-27632, Tocris Bioscience) and seeded into a 96-well U-bottom ultra-low attachment polystyrene plate (Corning, 7007) to form embryoid bodies (EBs). On day 3, EBs with normal morphology were transferred to ectoderm induction medium, which consisted of Essential 6 (E6) medium (Gibco, A1516401) supplemented with 2 mM GlutaMAX (Gibco), 0.1 mM Non-Essential Amino Acids (NEAA, Gibco), 0.1 mM β-mercaptoethanol (Gibco), 3 µM endo-IWR-1 (Tocris Bioscience), 0.1 µM LDN-193,189 (STEMgent), and 10 µM SB-431,542 (Tocris Bioscience). On day 7, the medium was replaced with neural induction medium, composed of DMEM/F-12 (Gibco) supplemented with 1:100 N-2 Supplement (Gibco), 2 mM GlutaMAX, 0.1 mM NEAA, 55 µM β-mercaptoethanol, and 1 µg/mL heparin (STEMCELL Technologies). On day 10, the developing brain organoids were embedded in Matrigel droplets and transferred to maturation medium for long-term culture. The maturation medium was Neurobasal Medium (Gibco) supplemented with 1:50 B-27 Supplement (Gibco), 2 mM GlutaMAX, 0.1 mM NEAA, 0.55 µM β-mercaptoethanol, 5 µg/mL heparin, 1% (v/v) Matrigel, 10 ng/mL brain-derived neurotrophic factor (BDNF, PeproTech), 10 ng/mL glial cell line-derived neurotrophic factor (GDNF, PeproTech), and 1 µM cyclic AMP (cAMP, Sigma-Aldrich). The medium was changed every 3–4 days until the organoids were harvested for downstream applications.

CK2 inhibitor treatment of human brain organoid

Organoids were treated with the 10 nM CK2 inhibitor CX-4945 (Adooq Bioscience, A11060) from day 14 onward. The CX-4945 was discontinued on day 28, and human brain organoid samples were harvested on day 35. Vehicle control groups received an equivalent volume of DMSO, which did not exceed 0.1% of the total medium volume and was confirmed to have no observable effect on organoid development or morphology.

Results

CK2 is required for maintaining NB number and proliferation

Previous studies have suggested a potential role for CK2 in stem cell maintenance [34, 38, 39], however, its role as a temporal regulator was unclear. To address this, we employed the NB-specific driver insc-GAL4 to knock down CK2α or CK2β and examine whether the genetic knockdown could lead to premature NB loss due to the shortened time window of NB maintenance in larval stages [40, 41]. Specifically, knocking down CK2α or CK2β, which are the catalytic subunit and the regulatory subunit encoding genes of CK2 respectively, significantly reduced the number of NBs at the third-instar larval stage (Fig. 1A–B). To assess the impact of CK2 on NB proliferation, we employed the mitotic marker phospho-Histone H3 (PH3), revealing a substantial decrease in the proliferation rate of NBs upon CK2 knockdown (Fig. 1A and C). These results indicated that CK2 is required for maintaining NBs and their proliferation in larval stages. These phenotypes were consistent in both central brain (CB) and ventral nerve cord (VNC) regions (Fig. S1A-C), so we only focused on NB phenotypes in the CB for convenience in this research.

To evaluate the efficiency of RNAi-mediated knockdown for CK2α or CK2β, we performed western blot (WB) analysis of larval brain protein extract using polyclonal antibodies against CK2α or CK2β, which further confirmed that both subunits were downregulated in knockdown of either subunit (Fig. S1D–G). To exclude the possibility of off-target effects of RNAi, we first confirmed the phenotype with multiple independent RNAi lines targeting CK2α or CK2β, and observed consistent phenotypes (Fig. S1H–J). Hereafter, UAS-CK2α-RNAi (NIG 17520R1) and UAS-CK2β-RNAi (NIG 15224R1) were mainly used for subsequent studies. Next, we used insc-GAL4 to overexpress full-length CK2α or CK2β in the context of CK2α or CK2β knockdown, respectively, which resulted in a significant increase in the number and proliferation rate of NBs compared to CK2 knockdown alone (Fig. 1D–E). With these results, we conclude that CK2 is necessary for NB maintenance in larval stages.

CK2 maintains NBs in late larval stages

To precisely determine the time window of CK2 on maintaining NBs, we conducted a quantitative analysis of NB numbers and proliferation rate across different developmental stages following CK2 knockdown. At the first (24 h after larval hatching [ALH]) and second (48 h ALH) instar larval stages, the NB number and proliferative capacity in CK2α or CK2β knockdown conditions were comparable to those of the control, but dropped significantly to approximately half of the control at the third-instar larval stage (72 h ALH) (Fig. 2A–C and S2A–C). These findings reveal that the NB loss phenotype of CK2 knockdown appears in late larval stages. To further confirm the time window of CK2 activity on NB maintenance, we used the temperature-inducible NB driver (insc-GAL4, tub-Gal80^ts) to achieve time-specific knockdowns of CK2. The Gal80^ts protein undergoes a conformational change at the restrictive temperature (29 °C), which prevents it from binding and inhibiting GAL4, thereby permitting RNAi expression. At the permissive temperature (18 °C), Gal80^ts binds GAL4 and blocks its activity. This temporal control allowed us to induce CK2α or CK2β knockdown at defined stages. Animals were transferred to the permissive temperature (18℃) from the embryonic stage to the first-instar larval stage, and then shifted to a restrictive temperature (29℃) during later stages, or vice versa until we dissected them at the third-instar larval stage. Knockdown of CK2 after the first-instar larval stage resulted in a significantly decreased number and proliferation rate of NBs (Fig. 2D–F), whereas CK2 knockdown before the first-instar larval stage did not cause any noticeable abnormalities (Fig. 2D–F).

Fig. 2.

CK2 maintains NBs in late larval stages. A Larval brains expressing the indicated RNAi transgenes were driven by insc-gal4 at different developmental stages (48, 72, 96 h ALH). Dpn, red; PH3, green; DAPI, blue. Scale bars: 50 μm. B–C Quantification of NB number (B) and proliferation rate (C) from genotypes in (A). D The temperature shift of animals with CK2α or CK2β knockdown from the permissive temperature (18℃) to the restrictive temperature (29℃). Dpn, red; PH3, green; DAPI, blue. Scale bars: 50 μm. E–F Quantification of NB number and proliferation rate in (D). G 96 h ALH brains expressing the indicated RNAi transgenes during 24–48 h ALH were driven by insc-Gal4, tub-Gal80^ts. Dpn, red; PH3, green; DAPI, blue. Scale bars: 50 μm. H–I Quantification of NB number (H) and proliferation rate (I) from genotypes in (G). J 96 h ALH brains expressing the indicated RNAi transgenes during 48–72 h ALH were driven by insc-Gal4, tub-Gal80^ts. Dpn, red; PH3, green; DAPI, blue. Scale bars: 50 μm. K–L Quantification of NB number (K) and proliferation rate (L) from genotypes in (J). M 72 h ALH brains expressing the indicated RNAi transgenes during 48–72 h ALH were driven by insc-Gal4, tub-Gal80^ts. Dpn, red; PH3, green; DAPI, blue. Scale bars: 50 μm. N–O Quantification of NB number (N) and proliferation rate (O) from genotypes in (M). B and C, n = 10; E, n = 11,13; F, n = 11, H and I, n = 11; K, L, N and O, n = 9. Statistical results were presented as means ± SEM, p values were performed by one-way ANOVA with a Bonferroni test; ****p < 0.0001, ns, not significant

To precisely determine when CK2 becomes essential, we then performed a more refined temporal analysis. When CK2α or CK2β was knocked down specifically during 24–48 h ALH, NB number and proliferative capacity remained comparable to controls at 96 h ALH (Fig. 2G–I). In contrast, knockdown restricted to the 48–72 h ALH interval resulted in a significant reduction in both parameters by 96 h ALH (Fig. 2J–L) and as early as 72 h ALH (Fig. 2M–O). These results collectively demonstrate that CK2 begins to exert a discernible impact on NB maintenance after the second-instar stage, with significant abnormalities first becoming apparent by 72 h ALH.

Together, these complementary results indicate that CK2 acts as an essential regulator that ensures NB maintenance in late larval stages.

CK2 expression in NBs is dynamic during brain development

To investigate why the regulation of CK2 on NBs occurs in late larval stages, we examined its expression across developmental stages (24 h ALH to 12 h after pupa formation [APF]) via immunostaining of brains with polyclonal antibodies. Our results showed that both CK2α and CK2β protein levels progressively increasing from early larval stages (24 h ALH) to peak at the third-instar larval stages (72 and 96 h ALH), followed by markedly declining during pupation (Fig. 3A–C), demonstrating a dynamic expression pattern of CK2 in NBs throughout development. This dynamic expression profile revealed by western blot analysis was consistent with the results obtained from immunofluorescence staining (Fig. S3A–C). The maximal expression of CK2 at the third instar-larval stage could explain the observed loss of NBs following CK2 knockdown in late larval stages but no substantial alterations of NB number in early larval stages. Together, these data indicate that the dynamic expression pattern of CK2 determines the time window of NB maintenance.

Fig. 3.

The expression of CK2 is dynamic in NBs A CK2α or CK2β staining at different developmental stages (24, 48, 72, 96 h ALH, 0, 12 h APF). CK2α, black; Mira, red; CK2β, black. The dashed circles point to NB region. Scale bars: 2 μm. B–C) Quantification of (A), CK2α (B), or CK2β (C) intensity in NB. B, n = 15; C, n = 16. Statistical results were presented as means ± SEM

CK2 is both necessary and sufficient for NB maintenance

The low abundance of CK2 during the early larval stage (24 h ALH), a period when NBs are undergoing reactivation from quiescence [42–44], led us to investigate whether CK2 influences the reactivation process itself. To test this, we overexpressed CK2α or CK2β in NBs from the embryonic stage and assessed key reactivation parameters. At both 9 and 24 h ALH, we found that NB number and proliferation rate were comparable to controls (Fig. S4A–F). Furthermore, NB diameter at 9 h ALH was also unchanged (Fig. S4G–H). Together with the absence of a phenotype upon CK2 knockdown at 24 h ALH (Fig. S2 A-C), this finding demonstrates that CK2 is neither necessary nor sufficient for NB reactivation.

During pupal stages, NBs undergo progressive elimination and terminal differentiation [18], so the downregulation of CK2 during pupal stages prompted us to investigate whether its expression decline could trigger NB elimination. To verify this hypothesis, we examined the phenotype of CK2 overexpression in NBs. During larval stages, CK2 overexpression did not significantly alter the number or proliferation rate of NBs (Fig. 4A–C). Then, we performed this phenotypic analysis of NBs at 24 h APF, a time point when most NBs have already disappeared in the WT. Overexpression of CK2 significantly prolonged the NB maintenance (Fig. 4D-E). Additionally, these ectopic NBs with CK2 overexpression remained mitotically active (Fig. 4D and F). This result indicates that ectopic CK2 expression in pupal stages could prolong NB maintenance. Therefore, we conclude that CK2 is both necessary and sufficient for NB maintenance based on results of the genetic knockdown and overexpression. These results suggest that the dynamically expressed CK2 functions as a time switch to control NB maintenance during larval and pupal stages.

Fig. 4.

CK2 determines that NBs are maintained in larval stages and disappeared in pupal stages A Brain lobes in control (LacZ) and UAS-CK2α or CK2β driven by insc-gal4 in 96 h ALH. Mira, red; PH3, green; DAPI, blue. Scale bars: 50 μm. B–C Quantification of NB number (B) and proliferation rate (C) from genotypes in (A). D Brain lobes in control (LacZ) and UAS-CK2α or CK2β driven by insc-gal4 in 24 h APF. Mira, red; PH3, green; DAPI, blue. Scale bars: 50 μm. (E–F) Quantification of NB number(B) and proliferation rate (C) from genotypes in (D). B and C, n = 10; E and F, n = 9. Statistical results were presented as means ± SEM, p values were performed by one-way ANOVA with a Bonferroni test; ns, not significant; ****p < 0.0001

The NB loss by CK2 knockdown results from premature differentiation

As defective origin, apoptosis and premature differentiation are the main causes of NB disappearance [13, 45], we next investigated which one led to the NB loss in CK2 knockdown. Previous results have shown that CK2 knockdown did not lead to a significant change in NB numbers in early larval stages compared to the control (Fig. 2A–B and S3A–F), suggesting that the number of NBs generated since their origin is not affected during early larval stages. To test the possibility of apoptosis, we next examined the apoptotic marker Caspase-1 (Dcp-1) in NBs at 72 h ALH. However, no significant increase of the ratio of the NBs positive for Dcp1 following CK2 knockdown was observed (Fig. S3A–B), suggesting that NBs did not undergo apoptosis after CK2 knockdown. Moreover, genetic validation via either baculoviral P35-mediated caspase inhibition or Dcp1 RNAi failed to rescue NB loss phenotype caused by CK2 knockdown (Fig. S3C–D). Based on these results, we conclude that the NB loss associated with CK2 knockdown is independent of apoptosis.

After excluding the possibilities of abnormal NB origin and apoptosis, we investigated whether CK2 knockdown induces NB premature differentiation. Previous studies have shown that the nuclear Pros is the marker of terminal differentiation which precedes NB elimination [13, 25, 46]. Therefore, we first quantified the ratio of NBs with nuclear Pros in control and CK2 knockdown, respectively. NBs with nuclear Pros were hardly detected in the wild-type third-instar larvae, but increased significantly in CK2 knockdown (Fig. 5A–B). To determine whether the decreased NB number upon CK2 knockdown was caused by increased Pros level in the nucleus, we used the insc-GAL4 driving UAS-pros-RNAi to prevent the NB loss, and found that the NB number was dramatically increased in the context of CK2 knockdown (Fig. 5C). These results indicate that CK2 deficiency induces the nuclear accumulation of Pros in NBs, leading to premature transition from self-renewal to differentiation.

Fig. 5.

CK2 knockdown causes the accumulation of Pros in the nucleus of NBs. A Pros staining of knocking down CK2α or CK2β in NB. Dpn, red; Pros, green. Scale bars: 2 μm. B Quantification of the ratio of NBs with nuclear Pros in (A). C Knockdown of pros in NBs rescues the NB number. Dpn, red; PH3, green. Scale bars: 50 μm. D Pros staining of knocking down CK2α or CK2β with marked by GFP and Dpn. GFP, cell membrane of NB, green; Dpn, red; Pros, blue. The solid circle point NB membrane, and the dashed circle point NB nucleus. Scale bars: 2 μm. E–G Quantification of normalized Pros intensity in whole NB (E), in NB cytoplasm (F) and in NB nucleus (G) in (D). H Quantification of normalized nucleo-cytoplasmic ratio of Pros intensity in (D). B, n = 10; E, n = 12, 24, 29; F, n = 17, 19, 24; G, n = 12, 20, 20; H, n = 12, 17, 17. Statistical results were presented as means ± SEM, p values were performed by one-way ANOVA with a Bonferroni test; **p < 0.01, ***p < 0.001, ****p < 0.0001

To examine whether the Pros accumulation in NB nucleus is due to its higher expression, we assessed the total intensity of Pros in NBs. To our surprise, we found that the Pros fluorescence intensity was significantly reduced in NBs after CK2 knockdown (Fig. 5E), indicating that the increased nuclear Pros did not result from the higher expression level. Therefore, we next examined the subcellular localization of Pros and quantified its levels in the cytoplasm and nuclei, respectively. In NBs of CK2 knockdown, the Pros fluorescence intensity was decreased in the cytoplasm (Fig. 5F), but considerably increased in nuclei (Fig. 5G), and the nucleocytoplasmic ratio was also significantly increased (Fig. 5H), which indicated that CK2 knockdown specifically facilitates the localization of Pros into the nucleus from cytoplasm in late larval stages. Altogether, we conclude that the loss of NBs induced by CK2 knockdown is due to the mislocalization of Pros to the nucleus, which triggers premature differentiation of NBs in late larval stages.

Pros is sequestered outside NB nuclei through phosphorylation by CK2

As previous researches have reported that Pros is a phosphorylated protein [47, 48]_, we hypothesized that CK2 phosphorylates Pros to inhibit its accumulation in the nucleus, which prevents NB differentiation. To examine this possibility, we first performed co-immunoprecipitation (co-IP) experiments and found that Pros was able to bind to CK2 (Fig. 6A). Then, we tried to confirm that the phosphorylated form of Pros exists in the cytoplasm of NBs. Given that in the larval brain, Pros is predominantly localized in the cytoplasm of NBs and in the nucleus of differentiated cells, the cytoplasmic Pros in the brain is primarily from NBs. Therefore, we performed nuclear-cytoplasmic fractionation on protein extracts from third instar larval brains. We treated the cytoplasmic Pros with alkaline phosphatase for dephosphorylation, while the control group was left untreated. We performed the WB analysis of Pros in these two groups and observed the higher molecular mass band from control, whereas the lower molecular mass band from the extract treated with alkaline phosphatase (Fig. 6B). Thus, we conclude that phosphorylated forms of Pros exist in the cytoplasm of NBs.

Fig. 6.

CK2 regulates NB maintenance by phosphorylation of Pros. A Coimmunoprecipitation (co-IP) of Pros and CK2α in the wild-type third-instar larval brain. B Western blotting analysis of cytoplasmic Pros from the wild-type third-instar larval brain treated with alkaline phosphatase (AP). C Overexpression of Pros^A during the 60–72 h ALH. Dpn, red; Pros, green. Scale bars: 20 μm, 2 μm. D–E Quantification of NB number (D) and the ratio of NBs with nuclear Pros (E) in (C). F Overexpression of Pros^D during the 48–72 h ALH. Dpn, red; Pros, green. Scale bars: 20 μm, 2 μm. G–H Quantification of NB number (G) and the ratio of NBs with nuclear Pros (H) in (F). D, E, G and H, n = 10. Statistical results were presented as means ± SEM, p value was performed by unpaired two-sided Student’s t test; ****p < 0.0001

To determine whether CK2 maintains NBs through Pros phosphorylation, we first tried to identify the potential Pros phosphorylation site(s) modulated by CK2. We used NetPhos, a computational tool designed for the prediction of phosphorylation sites in eukaryotic proteins using ensembles of neural networks, and found 17 evolutionarily conserved serine/threonine residues as potential targets of CK2 (Table S1). To address whether the phosphorylation status of the 17 sites was indeed related to Pros function, we mutated all 17 serine/threonine residues to alanine to mimic the non-phosphorylated form of Pros (Pros^A), or to aspartic acid to mimic the phosphorylated form of Pros (Pros^D). Overexpression of the wild-type Pros in NBs from the embryonic stage led to early larval death, thus we restricted the ectopic expression of Pros to late larval stages. We found that overexpression of wild-type Pros during the 60–72 h ALH period did not result in phenotypic changes in the NBs, whereas overexpression of Pros^A caused a significant decrease in NB number (Fig. 6C–D). Conversely, wild-type Pros overexpression during the 48–72 h ALH period led to a decrease in NB number (Fig. 6C and F), while Pros^D expression did not cause any change in NB number (Fig. 6F and G). To explore whether the decrease in NB number induced by Pros^A overexpression was attributed to its subcellular localization in the nuclei, we quantified the nuclear Pros level and found it significantly higher than wild-type Pros (Fig. 6C and E). In contrast, overexpression of Pros^D did not promote nuclear accumulation. (Fig. 6F and G). These findings support the hypothesis that CK2-mediated phosphorylation of Pros at identified residues enhances cytoplasmic retention to maintain NB self-renewal.

CK2 maintains the self-renewal of NSCs in human brain organoid

CK2 is a conserved protein kinase from insects to vertebrates that has been previously implicated in the development of the human CNS [49–51]. Therefore, we tried to investigate its function on NSCs during neurogenesis using a human-derived brain organoid model.

During early brain organoid development, there is an abundant population of PAX6-positive NSCs. This transcription factor suppresses the premature expression of neuronal differentiation genes, thereby ensuring that NSCs retain their self-renewal potential within an appropriate developmental time window [52–55]. We exposed brain organoids to 10nM CK2 specific inhibitor CX-4945 from day 12 after organoid expansion. We found that CX-4945 treatment led to disorganized neural rosette structures (Fig. 7A). In the remaining, discernible rosette regions, the PAX6 fluorescence intensity was significantly reduced (Fig. 7A–B), indicating that inhibition of CK2 expression in human brain organoids diminishes the self-renewal capacity of NSCs. Additionally, co-staining for the neural progenitor marker PAX6 and the proliferation marker Ki67 revealed a significant decrease in the percentage of PAX6 + cells that were Ki67+ (Fig. 7A and C), demonstrating a reduction in the proportion of proliferating neural progenitors. All these results demonstrate that inhibition of CK2 in human brain organoids reduces both the self-renewal and proliferative capacities of NSCs, resembling the conserved function found in Drosophila.

Fig. 7.

CK2 maintains the self-renewal of NSCs in human brain organoid. A Treated human brain organoids with 10 nM CK2-specific inhibitor XC-4945 from 12 d, fixed at 30 d. PAX6, NSC, green; Ki67, proliferative marker, red. The annular region between the solid and dashed circles points neural rosettes. Scale bars: 20 μm. B Quantification of PAX6 intensity in per neural rosettes in (A). C Quantification of proliferation rate of PAX6⁺ neural stem cell in per neural rosettes in (A). D A schematic model depicting CK2 regulates NBs self-renewal through phosphorylation of Pros. B, n = 10, 9; C, n = 10, 12. Statistical results were presented as means ± SEM, p value was performed by unpaired two-sided Student’s t test; ***p < 0.001

Discussion

The appropriate time window of NSC self-renewal is fundamental to proper neurodevelopment and homeostasis, yet the underlying molecular mechanisms remain incompletely understood. In this study, we revealed that CK2 exhibits a dynamic expression during neurogenesis. CK2 knockdown in late larval stages shortens the time window of NB self-renewal while CK2 overexpression extends it, which demonstrates CK2 is both necessary and sufficient for the NB maintenance. Therefore, the dynamic expression pattern of CK2 is required for the proper time window of NB maintenance. Here we revealed that CK2 acts as a temporal regulator that precisely switches NB from self-renewal in larval stages to differentiation in pupal stages.

Our results demonstrate that CK2 is essential for the proper maintenance of NBs in Drosophila. Larval NBs are classified into type I and type II based on their division mode. Type I NBs asymmetrically divide to self-renew and produce smaller ganglion mother cells (GMCs), which further divide to generate neurons and/or glial cells [56–59]. Type II NBs produce intermediate neural progenitor cells (INPs) that undergo multiple rounds of asymmetric division to generate GMCs [60–63], resembling the radial glial cells in the outer subventricular zone of the mammalian cerebral cortex [64, 65]. Although the mechanisms for maintaining the number of type I and II NBs are partially different [66, 67], we knock down CK2α or CK2β in type II NBs and found similar NB loss and decreased proliferation rates. This result indicates that CK2 plays a consistent role in the maintenance of both NB types, which further illustrates the conservation of CK2 function and suggests its important role in human brain development.

Our results indicate that CK2 maintains NBs self-renewal by phosphorylating Pros, thereby sequestering Pros in the cytoplasm and preventing premature nuclear localization. This phosphorylation-dependent mechanism occurs in a stage-specific manner: CK2 expression peaks during late larval stages to sustain NB self-renewal, while its decline during pupation permits Pros nuclear entry and terminal differentiation. Notably, the relatively low expression of CK2 during early larval stages coincides with the period of NB quiescence, when maintaining a proliferative pool is unnecessary [42, 43, 68, 69]. This observation is entirely consistent with our result that CK2 knockdown does not affect NB maintenance at early larval stages. Thus, high CK2 levels in late larvae maintain NBs, while its pupal downregulation serves as a switch that permits differentiation.

NetPhos, a neural network-based algorithm integrating multi-algorithm fusion strategies for enhanced precision, identified 17 evolutionarily conserved serine/threonine residues as potential CK2 phosphorylation targets. Next we generate a non-phosphorylated form Pros^A and a constitutively phosphorylated form Pros^D, to investigate the effect of Pros phosphorylation status on NB maintenance. Comparisons between wild-type and mutated forms of Pros overexpression indicate that these potential phosphorylation sites of Pros by CK2 are involved in the Pros localization and consequently in NB self-renewal or differentiation. However, there is still a lack of direct evidence coupling these predicted sites on Pros and the CK2-mediated phosphorylation. Furthermore, as a kinase with broad substrate specificity, CK2 may regulate NB maintenance through additional candidate pathways involving known targets like Dpn and Mbm [70, 71]. Future studies are required to elucidate the specific CK2 phosphorylation sites of Pros and their functional roles in modulating Pros subcellular localization, as well as to dissect the potential collaborative or parallel functions of these multiple substrates within the NB regulatory network.

This study highlights CK2 as a critical regulator to maintain NSC self-renewal. Consistent with findings in Drosophila, we have also demonstrated that CK2 is necessary for the self-renewal of NSCs in human organoids, underlying its critical and conserved role in human brain development. To unravel the detailed regulation of CK2 on NSC self-renewal will not only improve our understanding of NSC fate determination, but also potentially contribute to the development of therapeutic strategies for neurodevelopmental disorders in clinic.

Supplementary Information

Below is the link to the electronic supplementary material.

Funding

National Natural Science Foundation of China, 52033002, Su Wang. STI2030-Major Projects, 2021ZD0204000, Su Wang. Research Personnel Cultivation Programme of Zhongda Hospital Southeast University, CZXM-GSP-RC-152, Kun Yang. Zhongda Hospital Affiliated to Southeast University, Jiangsu Province High-Level Hospital Pairing Assistance Construction Funds, zdlyg05, Kun Yang. Medical Science and Technology Development Foundation of Nanjing, ZKX22041, Kun Yang. The funders had no role in study design, data collection, and interpretation, or the decision to submit the work for publication.

Data availability

All data and materials are available upon request from the corresponding authors Su Wang (wangsu@seu.edu.cn), Menglong Rui (ruimenglong@seu.edu.cn) or Kun Yang (yungkun@126.com).

Declarations

Ethics approval and consent to participate

Not applicable.

Competing interests

The authors declare that they have no conflict of interest.