Introduction

Asymmetric cell division (ACD) is a universal mechanism for generating cellular diversity during development as well as for regulating tissue homeostasis in the adult (1). Stem and progenitor cells undergo ACD to simultaneously give rise to a self- renewing stem/progenitor cell and to a daughter cell that is committed to enter a differentiation program. Remarkably, over the past years it has become apparent that undermining this critical process can lead to tumor-like overgrowth (2). This link between failures in the process of ACD and tumorigenesis was first established using the Drosophila neural stem cells, called neuroblasts (NBs), as a model system (3). In this work, authors showed how Drosophila larval mutant brains for ACD regulatory genes were able to induce the formation of massive tumoral masses after weeks of being transplanted into the abdomen of wild-type fly hosts. Intriguingly, some genes firstly identified in Drosophila as tumor suppressor genes, such as discs large 1 (dlg1), lethal (2) giant larvae (l(2)gl) and brain tumor (brat) (4), were later uncovered as ACD regulators (5–11), further supporting the connection between compromised ACD and tumorigenesis.

Tumor initiating cells or cancer stem cells (CSCs) were originally identified in acute myeloid leukemia (AML) (12,13) but, to date, several studies have also revealed the existence of CSCs in multiple solid tumors (14). These CSCs represent a very small but highly malignant population within the tumor, able to proliferate without control by symmetrical, self-renewal divisions, as well as to provide differentiated progeny by ACD that contribute to the heterogeneity of the tumor. These properties, along with their quiescent state, increase in metabolic activity and high capacity of DNA repair, make CSCs extremely resistant to radio- and chemotherapy and responsible for malignant relapse (14–18). Thus, understanding the unique nature of these cells and their landmarks is crucial to target CSCs and hence to completely destroy the tumor. An intriguing possibility is that an unbalanced ACD process in CSCs, favoring symmetric renewal divisions, might contribute to the transition from a chronic (low grade) to an acute (high grade) phase in the cancer progression (19,20).

Glioblastoma (GBM), the most aggressive and lethal brain tumor with very poor prognosis, is among the solid tumors in which the presence of CSCs, called glioma or glioblastoma stem cells (GSCs), has been proven. As other CSCs, GSCs are responsible for the formation, maintenance, resistance to conventional therapy and consequent relapse of GBM (21–25). Several GSC biomarkers, such as CD133 and Nestin, and signaling pathways, such as Notch and Hedgehog, have been identified to help isolate and target these crucial cells (14,26,27). Despite of that, there are still many unknowns regarding the nature, properties and origin of CSCs in general and of GSCs in particular. Here, we analyze the consequences of compromising ACD in the GSCs within GBM neurosphere cell cultures. Specifically, we show that RAP2A, the human homolog of Rap2l, which encodes a Drosophila small GTPase and novel ACD regulator, displays low expression levels in GBM, and that restitution of RAP2A in GBM neurosphere cultures increases the ACD and decreases the expression of stem cell markers in the GSCs. Our results support that failures in ACD increase GSC proliferation promoting their spreading and that ACD amendment might contribute to reduce the expansion of GBM.

Results

RAP2A is highly downregulated in GBM patients

As a first approach to analyze whether ACD might be impaired in GBM, we analyzed the expression levels of human homologs of 21 Drosophila ACD regulators in a GBM microarray (28) (Fig. 1A). We also investigated which of those genes were more similar in expression pattern to the others by performing a Gene Distance Matrix (Fig. 1B). The human genes TRIM2 and RAP2A showed the lowest levels in the GBM patient samples analyzed. TRIM2 belongs to the same family that TRIM3 and TRIM32, all of them related to the Drosophila ACD regulator gene brat. TRIM3, the closest homolog of brat, has been previously found to be expressed at low levels in GBM (29). Hence, we concentrated on RAP2A, the second human gene consistently found at very low levels in all the patient samples (Fig. 1A). RAP2A is also known to have a tumor suppressor role in the context of glioma migration and invasion (30,31). First, to further support the microarray data expression levels in a bigger sample size, we performed in silico analyses taking advantage of established datasets of Glioblastoma IDH wild-type, such as “The Cancer Genome Atlas” (TCGA) and “Gravendeel cohort”. We confirmed that low RAP2A expression levels in human GBM are associated with a poor prognosis (Fig. 1C). Hence, we focused on the human gene RAP2A to analyze its potential role in ACD and its relevance in GBM.

Figure 1.

Drosophila RAP2A homolog Rap2l regulates ACD

We previously showed that the Drosophila gene Rap1 regulates the ACD process (32). The closest human homolog of Drosophila Rap1 is RAP1A, while human RAP2A is more similar to Drosophila Rap2l, which has not been previously analyzed in the context of ACD. Hence, we first wanted to investigate whether Rap2l was, like Rap1, implicated in ACD regulation. With that aim, we started analyzing the ACD of neural stem cells (neuroblasts, NBs) and intermediate neural progenitors (INPs) within Drosophila larval brain type II NB lineages (NBII) (Fig. 2A) (8,33,34). In NBII lineages, there is one NB that expresses the transcription factor Deadpan (Dpn) and several INPs, which express both Dpn and the transcription factor Asense (Ase) (Fig. 2A). We first observed the presence of ectopic NBs (eNBs; Dpn⁺ Ase^-) after downregulating Rap2l in NBII lineages, while in control lineages only one NB was usually found (Fig. 2B). This result suggested that the process of ACD was impaired within those Rap2l mutant NBII lineages. To more directly support that, we analyzed the localization of key ACD regulators in metaphase

Figure 2.

NBs and INPs of NBII lineages, such as the apical proteins aPKC and Canoe (Cno), and the cell-fate determinant Numb (35–37). In control metaphase NBs and INPs, those ACD regulators form crescents at the apical (aPKC and Cno) or basal (Numb) poles of the dividing cell (Fig. 2C). The apical localization of aPKC was not significantly altered after downregulating Rap2l in NBII lineages; however, the localization of Cno and Numb did fail (Fig. 2C). All these results strongly suggested that Rap2l, as Rap1, was an ACD regulator. Hence, we came back to the human homolog of Drosophila Rap2l, RAP2A, to determine its potential relevance in the context of the ACD of GSCs.

RAP2A expression in GBM neurosphere cultures reduces the stem cell population

After confirming by in silico analyses that low RAP2A expression levels in GBM patients were associated to a poor prognosis, we decided to use established GBM cell lines to investigate in neurosphere cell cultures whether the downregulation of RAP2A was affecting the properties of GSCs. We tested six different GBM cell lines finding similar mRNA levels of RAP2A in all of them, and significantly lower levels than in control Astros (Fig. 3A). We focused on the GBM cell line called GB5, which grew well in neurosphere cell culture conditions, for further analyses. Given that stem cell markers, such as CD133, SOX2 and NESTIN are landmarks of GSCs, we started analyzing their expression after restoring RAP2A in the neurosphere culture (Fig. 3B). Immunofluorescences of each of those markers revealed a significant reduction in the protein intensity in the RAP2A-expressing neurospheres (GB5-RAP2A) compared to the control neurospheres (GB5) (Fig. 3C). Likewise, RT-PCRs and Westerns blots also showed a significant decrease in the mRNA and protein expression levels, respectively, of those stem cell markers (Fig. 4). Hence, RAP2A seems to be contributing to decrease the stem cell population within GBM neurosphere cultures.

Figure 3.

Figure 4.

RAP2A expression in GBM neurosphere cultures decreases cell proliferation and sphere size

Next, we wondered whether the reduction in the stem cell population after expressing RAP2A in neurosphere cultures was reflected in the level of cell proliferation and, consequently, in the size of the neurospheres. To investigate this, we first analyzed the expression of the proliferation marker and key tool in cancer diagnostics Ki-67, highly expressed in cycling cells but absent in quiescent (G0) cells (38,39). We detected a significant decrease in the number of proliferating cells per neurosphere in those cultures expressing RAP2A (Fig. 5A). Moreover, the average area size of these GB5- RAP2A neurospheres was also significantly reduced compared to control GB5 neurospheres (Fig. 5B). Thus, RAP2A promotes a decrease in cell proliferation.

Figure 5.

RAP2A expression in GBM neurosphere cultures fosters ACD in GSCs

Given the reduction in cell proliferation and in the stem cell population after adding RAP2A to the control GB5 neurosphere cultures, we wanted to investigate whether RAP2A was favoring an asymmetric mode of cell division in these cells. Thus, we decided to follow over time the GB5-RAP2A and the control GB5 neurospheres, analyzing the number of cells per neurosphere cluster. We reasoned that an even number of cells per cluster would be indicative of symmetric cell divisions, while an odd number of cells would point to an asymmetric mode of cell division. We observed a significant increase of cell clusters with an odd number of cells in the GB5-RAP2A neurospheres (n= 124 cell clusters analyzed) compared to control GB5 neurospheres (n=153 cell clusters) (Fig. 6A). This result was very suggestive of an increase in the number of ACDs in GB5-RAP2A neurosphere cells. To further support this result, we decided to analyze the distribution of the cell fate determinant NUMB, a key conserved ACD regulator whose asymmetric distribution in the mother and progeny ultimately promotes the repression of stem cell self-renewal in the daughter cell in which is segregated (40–44). We observed a significant increase in the number of GB5- RAP2A dividing cells that showed an asymmetric localization of NUMB in the progeny, compared to control GB5 dividing cells (Fig. 6B). Thus, all these results strongly supported a function of RAP2A promoting ACD in the GSCs.

Figure 6.

Discussion

ACD is an evolutionary conserved mechanism used by stem and progenitor cells to generate cell diversity during development and to regulate tissue homeostasis in the adult. CSCs, present in many human tumors, can divide asymmetrically to generate intratumoral heterogeneity or symmetrically to expand the tumor by self-renewal. Over the past 15 years, it has been suggested that dysregulation of the balance between symmetric and asymmetric cell divisions in CSCs, favoring symmetric divisions, can trigger tumor progression in different types of cancer (19,20,45) including mammary tumors (43,46), glioblastoma (29), oligodendrogliomas (47,48), colorectal cancer (49,50) and hepatocellular carcinoma (50). Thus, it is of great relevance to get a deeper insight into the network of regulators that control ACD, as well as the mechanisms by which they operate in this key process.

Drosophila neural stem cells or NBs have been used as a paradigm for many decades to study ACD (51). NBs divide asymmetrically to give rise to another self- renewing NB and a daughter cell that will start a differentiation process. Over all these years, a complex network of ACD regulators that tightly modulate this process has been characterized. For example, the so-called cell-fate determinants, including the Notch inhibitor Numb, accumulate asymmetrically at the basal pole of mitotic NBs and are exclusively segregated to one daughter cell, promoting in this cell a differentiation process. The asymmetric distribution of cell-fate determinants in the NB is, in turn, regulated by an intricate group of proteins asymmetrically located at the apical pole of mitotic NBs, generically known as the “apical complex”. This apical complex includes kinases (i.e. aPKC), small GTPases (i.e. Cdc42, Rap1) and Par proteins (i.e. Par-6, Par3), among others (51).

Given the potential relevance of ACD in CSCs, we decided to take advantage of all the knowledge accumulated in Drosophila about the network of modulators that control asymmetric NB division. As a first approach, we aimed to analyze whether the levels of human homologs of known Drosophila ACD regulators were altered in human tumors. Specifically, we centered on human GBM, as the presence of CSCs (GSCs) have been shown in this tumor. The microarray we interrogated with GBM patient samples had some limitations. For example, not all the human genes homologs of the Drosophila ACD regulators were present (i.e. the human homologs of the determinant Numb). Likewise, we only tested seven different GBM patient samples. Nevertheless, the output from this analysis was enough to determine that most of the human genes tested in the array presented altered levels of expression. We selected for further analyses RAP2A, one of the human genes that showed the lowest levels of expression, compared to the control samples. However, it would be interesting to analyze in the future the potential consequences that altered levels of expression of the other human homologs in the array can have in the behavior of the GSCs. In silico analyses, taking advantage of the existence of established datasets, such as the TCGA, can help to more robustly assess, in a bigger sample size, the relevance of those human genes expression levels in GBM progression, as we observed for the gene RAP2A.

We have previously shown that Drosophila Rap1 acts as a novel NB ACD regulator in a complex with the apical regulators Canoe (Cno), aPKC and Par-6 (32). Here we have shown that Drosophila Rap2l also regulates NB ACD by ensuring the correct localization of the ACD modulators Cno and Numb. We also have demonstrated that RAP2A, the human homolog of Drosophila Rap2l, behaves as an ACD regulator in GBM neurosphere cultures, and its restitution to these GBM cultures, in which is present at low levels, targets the stemness of GSCs increasing the number of ACDs. Thus, this study supports the relevance of ACD in CSCs of human tumors to refrain the expansion of the tumor. Other studies, however, claim that ACD should be targeted in human tumors as it promotes the intratumoral heterogeneity that hampers the complete tumor loss after chemotherapy (52–54). More studies should be carried out with other human genes homologs of ACD regulators to further confirm the results of this study. Likewise, analyses in vivo (i.e. in mouse xenografts) would be also required to reinforce our conclusions. This would be very relevant in order to consider ACD restitution in CSCs of human tumors, what has been called “differentiation therapy” (55), as a potential alternative therapeutic treatment.

Materials and Methods

Drosophila husbandry, strains and genetics

The following fly stocks were used: wor-Gal4 ase-Gal80 (56); UAS-CD8::GFP (Bloomington Drosophila Stock Center -BDSC- #5137); UAS-CD2 (BDSC #1373); UAS- Rap2l^KKRNAi (Vienna Drosophila Resource Center #107745); cno::GFP (57). All the fly stocks were raised and kept in 18°C or 25°C incubators. Experimental temperatures for the assays were maintained using 25°C or 29°C incubators. The Gal-4 x UAS-crosses were carried out at 29°C (the first 2 days they were kept at 25°C, and then transferred to a new tube and left at 29°C) until third instar larvae developed.

Drosophila histology, immunofluorescence and microscopy

Larval brains were dissected out in PBS and fixed with 4% PFA in PBT (PBS and Triton X- 100 0.1%) for 20 min at room temperature with gentle rocking. Fixed brains were washed 3 times for 15 min with PBT (PBS and Triton X-100 0.3%) and then incubated in PBT-BSA for at least 1h before incubation with the corresponding primary antibody/antibodies. The following primary antibodies were used in this study: guinea pig anti-Dpn (1:2,000) (57), rabbit anti-Ase (1:100) (57), rabbit anti PKCζ (1:100; Santa Cruz Biotechnology, sc-216), goat anti-Numb (1:200; Santa Cruz Biotechnology, sc- 23579), mouse anti-PH3 (1:2,000; Millipore, 05-806), mouse anti-γTub (1:200; Sigma- Aldrich, T5326) and mouse anti-CD2 (1:50; BioRad). Secondary antibodies conjugated to fluorescent dyes (Invitrogen) were used at 1:400 dilution. Samples were mounted in VECTASHIELD antifade mounting medium for fluorescence (Vector Laboratories, H- 1000). Fluorescent images were captured using a Super-resolution Inverted Confocal Microscope Zeiss LSM 880-Airyscan Elyra PS.1. Images were analyzed using the image processing package FIJI from ImageJ and assembled using Adobe Photoshop CS6.

GBM microarray

The cohort of patients consisted of seven GBM biopsies obtained from the Valencian Network of Biobanks. The procedures for obtaining tissue samples were developed in accordance with national ethical and legal standards, and following the guidelines established in the Declaration of Helsinki, and approved of Clinical Research Ethics Committee of the Hospital General Universitario de Elche, Spain. All patients had been diagnosed with GBM grade IV, and all of them exhibited an IDH1 plus IDH2 wild-type genotype. Normal human brain RNA (frontal lobe; pool from 5 donors) was purchased from BioChain and used as a control for gene expression experiments. A total of 12,516 genes were used to generate a custom Agilent Two-Color 8 x 15 K Agilent gene expression DNA microarray, which were deposited in the NCBI Gene Expression Omnibus (GEO) Database under accession number GSE182697 (28).

Bioinformatics analysis

The level of expression of human homologs of Drosophila ACD regulators was assessed in seven different GBM patients versus control samples of the custom-made microarray (GSE182697) (28). Statistical analysis of microarray data was done in Multiexperiment Viewer (MeV) (58). Hierarchical clustering analyses were done optimizing gene and sample leaf order and using Pearson Correlation for the distance metric selection. Average linkage clustering was used as a linkage method. For the Gene Distance Matrix (GDM), the distance, inverse of similarity, between two genes was calculated using a distance metric as indicated in MeV.

Human samples and cell lines

Glioma samples were obtained with patients’ written informed consent, in accordance with the Declaration of Helsinki, and approved by the Ethics Committee at Hospital 12 de Octubre, Madrid, Spain (CEI 14/023). The primary glioblastoma cell lines were provided by the Biobank at “Hospital 12 de Octubre”. Fresh tissue samples were enzymatically dissociated using Accumax (Millipore) and cultured in stem cell medium (Neurobasal, Invitrogen), supplemented with N2 (1:100, Invitrogen), GlutaMAX (1:100, Invitrogen), penicillin-streptomycin (1:100, Lonza), 0.4% heparin (Sigma-Aldrich), 40 ng/ml EGF and 20 ng/ml bFGF2 (Peprotech). The GB cell lines used in this study were: GB4, GB5, GB7, GB18, GB19, 12o89, Astros (59,60) and HEK293T. The cell lines HEK293T and Astros were maintained in Dulbeccós modified Eagle Medium supplemented with 10% fetal bovine serum 2 mM L-glutamine, 0.1% penicillin (100 U/ml) and streptomycin (100 μg/ml) and Non-Essential Aminoacids (NEA 1x).

GBM-derived neurosphere cell cultures

For stem cell culture of GBM-derived neuroespheres we used DMEM:F12 with FGF-2 (20 ng/ml), EGF-2 (20 ng/ml), and N2 supplements, 0.1% penicillin (100 U/ml) and streptomycin (100 μg/ml), as previously described (61,62). The tumor spheres were dissociated to single cells using Accumax, and 1x10⁵ dissociated cells/ml were plated in a Flask and cultured for 6-7 days.

In Silico Analysis

Data from The Cancer Genome Atlas (TCGA) and the Chinese Glioma Genome Atlas (CGGA) for GBM cohorts were accessed through UCSC Xena-Browser (https://xenabrowser.net), and Gliovis (http://gliovis.bioinfo.cnio.es) for extraction of overall survival data, gene expression levels, and distribution of genetic alterations. Kaplan-Meier survival curves were generated by stratifying samples into low- and high- expression groups for each gene. Survival differences between these groups were assessed using the log-rank test.

DNA Constructs and Lentiviral Production

Lentiviral vectors were used to produce cells overexpressing RAP2A (pLJM1RAP2A #Plasmid 19311, Addgene) or eGFP (pLJMEGFP #Plasmid 19319, Addgene) as infection control. Infected cells were selected with Puromycin. To obtain the lentivirus, HEK293T cells were transiently co-transfected with 5 μg of the appropriate lentiviral plasmid (pLJEGFP or pLJRRap2A, respectively), 5 μg of the packaging plasmid pCMVdR8.74 (#Plasmid 22036, Addgene), and 5 μg of VERSUSV-G envelope protein plasmid pMD2G (#Plasmid 12259, Addgene), using Lipofectamine Plus reagent (Invitrogen, Carisbad, CA, USA). Lentiviral supernatants were collected after 48 h of HEK293T transfection.

Neurosphere Immunofluorescences and Imaging

For immunostaining, floating cultured neurospheres were incubated on Matrigel coated micro slide glass for 2 hours at 37°C. After having been washed, the attached neurospheres were fixed in 4% paraformaldehyde for 20 min. Cells were blocked for 1 h in 1% FBS and 0.1% Triton X-100 in PBS; then, they were incubated for 2 hours with the primary antibody: mouse anti-CD133 (HB#7, Developmental Studies Hybridoma Bank), mouse anti-Ki67 (3E6, Developmental Studies Hybridoma Bank), rabbit anti- Sox2 (N1C3, GTX101507 GeneTex), mouse anti-Nestin (10c2, sc-23927, Santa Cruz Biotechnology) and mouse anti-Numb (48, sc-136554, Santa Cruz Biotechnology). The secondary antibodies used were Alexa555 goat anti-rabbit (A32732, Invitrogen) or Alexa555 goat anti-mouse (A32727, Invitrogen). Phalloidin 633 (Sigma-Aldrich, 68825 Phalloidin-Atto 633) was added and incubated for 20 min in PBS; Vectashield Mounting medium with DAPI (Linaris, H-1200) was added before mounting. Fluorescent images were captured using a Super-resolution Inverted Confocal Microscope Zeiss LSM 880- Airyscan Elyra PS.1. Images were analyzed using the image processing package FIJI from ImageJ and assembled using Adobe Photoshop CS6 program.

Immunofluorescent quantification

Fluorescence intensity measurements were carried out using ImageJ software package. In all cases, images of neuroespheres transfected with GFP (GB5, control) or RAP2A (GB5-RAP2A) lentivirus were analyzed:

CD133: Using the “line tool”, a line was drawn on the channel with the CD133 signal on three different pictures per genotype. Fluorescence was measured across the line sections with the “Plot Profile Tool”, and reported as a subset of images showing the signal intensity plot. For quantification, only the 100 highest gray value points, were taken.

Sox: After adjusting the threshold on the channel with the Sox2 signal and eliminating particles smaller than 5 microns, the fluorescence intensity was measured as the mean gray value (sum of all pixels with gray values inside the selection, divided by the number of pixels). In Figure 3 is shown one example out of three quantifications achieved with at least 20 cells per genotype.

Nestin: The regions of interest (ROIs) were set with the “hand drawing tool” using the Phalloidin staining to follow the cell periphery. At least 20 cells (ROIs) from three different pictures were selected per condition. The background fluorescence was subtracted from the mean gray value fluorescence for each well, and the fluorescence intensity was normalized by the nuclear DAPI signal coming from the same cell (ROI). Plot in Figure 3C represents the mean gray value that is the sum of all pixels with gray values inside the selection divided by the number of pixels.

RNA isolation and qRT-PCR assay

Total RNA was extracted from 7 days old neurospheres using TRI reagent (AM9738, Invitrogen) following manufacturer indications and quantified using a nanodrop (ND- 1000, Thermo Scientific). 2ug of RNA was treated with DNaseI (EN0521, Thermo Scientific) and reverse transcribed with SuperScriptTM III Reverse Transcriptase (1808044, Invitrogen). Quantitative real time PCR (qRT-PCR) was performed using Power SYBR Green PCR Master Mix (PN4367218, Applied Biosystems), following established protocols with 60°C for annealing/extension and 40 cycles of amplification on a QuantStudioTM 3 apparatus (applied Biosystems). Act5 primers were used for normalization, and relative expression was calculated using the comparative ΔΔCt. The average of at least 3 independent experiments is shown in Figure 4.

Neurosphere Western Blots

Protein extracts were prepared by re-suspending cell pellets in a lysis buffer (50-mM Tris, pH 7.5, 300-mM NaCl, 0.5% SDS, and 1% Triton X-100) and incubating the cells for 10 min at 95°C. The lysed extracts were centrifuged at 13,000 g for 10 min at room temperature, and the protein concentration was determined using a commercially available colorimetric assay (BCA Protein Assay Kit). Approximately 25 µg of protein was resolved by 10% or 12,5% SDS-PAGE, and they were then transferred to a PVDF membrane (Hybond-ECL, Amersham Biosciences). The membranes were blocked for 1 h at room temperature in TBS-T (10-mM Tris–HCl, pH 7.5, 100-mM NaCl, and 0.1% Tween-20) with 5% skimmed milk, and then incubated overnight at 4°C with the corresponding primary antibody diluted in TBS-T with 2,5% skimmed milk. After washing 3 times with TBS-T, the membranes were incubated for 2 h at room temperature with their corresponding secondary antibody, diluted in TBS-T with 2,5% skimmed milk. The proteins were visible by enhanced chemiluminescence with ECL (Pierce) using Amersham imager 680, and the signal was quantified by Fiji-ImageJ software.

Proliferative index (Ki67 analysis)

The number of Ki67+ cells with respect to the total number of nuclei labelled with DAPI within a given neurosphere were counted to calculate the Proliferative Index (PI), which was expressed as the % of Ki67+ cells over total DAPI+ cells. More than 20 neurospheres coming from different pictures were analyzed.

Sphere size measurement

Glioblastoma cells (GB5 cell line) were infected by control lentivirus (LV-GFP) or by lentivirus directing expression of RAP2A (LV-RAP2A). The tumor spheres were Accumax-dissociated to single cells, during three consecutive passages. Seven days after the third plating in a 6-well plate, spheres were captured using a 10x objective (Carl Zeiss Axio Imager A1 microscope). 20 random pictures of each genotype were taken and the images were analyzed by measuring the area of all the spheres present, using the “freeform drawing tool” in the ImageJ software package. The data presented show the average area measured in 3 independent experiments.

Pair assay and Numb segregation analysis

Dissociated cells were plated at low density on Matrigel-coated glass slides. Cultures were incubated for 21 h at 37°C with 5% CO2, fixed and immunolabeled for Numb, as well as for DAPI to identify nuclei. First, isolated groups of cells (most probably coming from the same single progenitor) were identified and the number of cells in each cluster was annotated. For each identified DAPI-labeled pair, the localization of Numb was visualized and scored from at least 100 cell pairs. Images in Figure 6 were acquired in an Inverted Microscope Leica Thunder Imager.

Statistical analysis

The data were first analyzed using the Shapiro-Wilk test to determine whether the sample followed a normal distribution. Parametric T-test or a nonparametric two- tailed Mann Whitney U test for those that did not follow a normal distribution were used to compare statistical differences between two different groups. Mean and standard deviation values were calculated by standard methods. A Chi-square test was used to analyze the data in Figure 2C. To assume statistical significance, p-values were determined below 0.05. Sample size (n) and the p-value are indicated in the figure or figure legend; * p < 0.05, ** p < 0.01, *** p < 0.001, ns: not significant (p > 0.05).

Acknowledgements

We thank the Bloomington Drosophila Stock Center at the University of Indiana and the Developmental Studies Hybridoma Bank at the University of Iowa for kindly providing fly strains and reagents. R.G. was funded by “Instituto de Salud Carlos III (ISCIII)” through the project “Miguel Servet Contracts” (CP21/00116), PI22/01171 and co-funded by the European Union. M.S. was supported by the Spanish grant from the Instituto de Salud Carlos III (ISCIII) PI22/00824. A.C. was financed by the Spanish grants from the Ministry of Science, Innovation and Universities (PGC2018- 097093-B-100), the Ministry of Science and Innovation (PID2021-123196NB-100), the Generalitat Valenciana Prometeo/2021/052 grant and by FEDER (European Regional Development Fund). The “Instituto de Neurociencias” in Alicante is a “Severo Ochoa Center of Excellence in R&D”. Our group belongs to “Conexión Cáncer- CSIC”.

Additional information

Author contributions

M.F. participated in the design of the experiments, conducted most of them and analyzed the data; R.G. developed the GB lines and performed the in silico analyses; V.M.B. and M.S. carried out the GB microarray and bioinformatics analysis; A.C. conceived the study and wrote the manuscript.