Notch signaling regulates neural stem cell quiescence entry and exit in Drosophila

Most stem cells in adult tissues remain in a poised, non-proliferative state known as quiescence. Quiescence, typically thought of as G0 arrest, is associated with reduced transcription, translation and ribosome biogenesis and, more recently, with increased lysosomal activity and fatty acid utilization (Kalucka et al., 2018; Coller, 2019; Kobayashi et al., 2019). In adults, stem cells reactivate from quiescence to maintain tissue homeostasis and for repair, whereas in development, quiescence is ‘pre-programmed’ and required to ensure that sufficient dietary nutrients or other key factors are available to fuel cell divisions needed to support continued growth (Cheung and Rando, 2013; Cavallucci et al., 2016; Kalamakis et al., 2019; Cho et al., 2019; Urbán et al., 2019). Although stem cell entry and exit from quiescence is important in development and adulthood, mechanisms and the cell signaling pathways that regulate quiescence entry and exit are incompletely understood (Li et al., 2017; Mohammad et al., 2019; Sueda and Kageyama, 2020).

In Drosophila, neural stem cells, known as neuroblasts (NBs), enter and exit quiescence at defined developmental times (Ito and Hotta, 1992; Truman and Bate, 1988; Britton and Edgar, 1998; Tsuji et al., 2008; Chell and Brand, 2010; Sousa-Nunes et al., 2011; Sipe and Siegrist, 2017; Yuan et al., 2020). In the central brain (CB), all NBs, except for a small subset, enter quiescence at the end of embryogenesis, coincident with declining maternal nutrient stores, and then reactivate after newborn larvae consume their first complete meal (Britton and Edgar, 1998; Chell and Brand, 2010; Sousa-Nunes et al., 2011; Sipe and Siegrist, 2017; Yuan et al., 2020). In response to animal feeding, PI3-kinase becomes active in NBs, their cortex glial niche and trachea, leading to coordinated increases in cell growth, endoreplication of glia and trachea, and resumption of NB divisions (Chell and Brand, 2010; Sousa-Nunes et al., 2011; Sipe and Siegrist, 2017; Spéder and Brand, 2018; Yuan et al., 2020). Concurrent with nutrient-dependent PI3-kinase activation, Yorkie, a transcriptional co-activator negatively regulated by Hippo signaling, localizes to the NB nucleus to promote growth (Ding et al., 2016). When PI3-kinase activity is reduced NB reactivation is delayed and, when Hippo activity is reduced, NBs reactivate prematurely (Chell and Brand, 2010; Sousa-Nunes et al., 2011; Ding et al., 2016; Yuan et al., 2020). Although dietary nutrients and growth signaling are key extrinsic cues required for NB reactivation, it remains unclear how extrinsic cues coordinate with NB intrinsic cues to control quiescence.

A low-level pulse of the homeodomain transcription factor Prospero (Pros) triggers quiescence, and timing of the Pros pulse is regulated by NB intrinsic temporal factors (Lai and Doe, 2014). Pros is usually partitioned asymmetrically to ganglion mother cells (GMC) after NBs divide, where it functions to promote GMC cell cycle exit (Hirata et al., 1995; Choksi et al., 2006; Colonques et al., 2011). In nubbin/pdm2 mutants, the NB Pros pulse and quiescence occur prematurely, whereas in castor (cas) mutants, the NB Pros pulse and quiescence are delayed (Tsuji et al., 2008; Lai and Doe, 2014). Nubbin/Pdm2 and Cas are both temporal factors and are expressed sequentially in embryonic NBs. Dacapo (dap), a cyclin-dependent kinase inhibitor, and tribbles (trbl), a protein kinase that inhibits both insulin signaling activity and String (stg), also are reported to regulate NB quiescence (Colonques et al., 2011; Otsuki and Brand, 2018, 2019). But how cell cycle gene regulation is coupled with temporal factor expression and Pros is not yet understood. Nor is it clear whether other factors are required.

From a large-scale RNAi screen aimed at identifying genes that regulate neural stem cell proliferation decisions during development, we identified Notch, an evolutionarily conserved transmembrane receptor that functions in cell-cell communication. In Drosophila, there is one Notch receptor and two Notch ligands, Delta and Serrate (Rebay et al., 1991; Fiuza and Arias, 2007; Kopan and Ilagan, 2009). Notch is activated after the Notch receptor binds its ligand expressed on neighboring cells. After ligand binding, Notch is proteolytically cleaved, first by Kuzbanian (Kuz), an ADAM metalloprotease, and then by γ-secretase, resulting in a liberated, cleaved and active form of Notch, known as Notch ICD (Notch intracellular domain). Notch ICD translocates to the nucleus to regulate gene expression through interactions with Suppressor of Hairless [Su(H)] and the transcriptional co-activator Mastermind (Fortini and Artavanis-Tsakonas, 1994; Pan and Rubin, 1997; De Strooper et al., 1999; Mumm et al., 2000; Kopan and Ilagan, 2009). Context-dependent Notch signaling can determine cellular proliferation status by regulating expression of cell cycle inhibitors such as Dap and cell cycle activators such as Cyclin E, String and E2f (Sriuranpong et al., 2001; Deng et al., 2001; Noseda et al., 2004; Herranz et al., 2008; Ishikawa et al., 2008; Bivik et al., 2016). Here, we show that Notch pathway activity increases in NBs during embryogenesis and that high Notch triggers quiescence entry. As a consequence of quiescence and cell cycle exit, Notch signaling becomes attenuated and low Notch is required for NBs to exit quiescence in response to dietary nutrient cues.

All neuroblasts in the central brain, except for a small subset, stop proliferating at two defined times in development. First, during the embryonic-to-larval transition when most CB NBs enter quiescence (Fig. 1A) (Tsuji et al., 2008; Lai and Doe, 2014; Otsuki and Brand, 2018). Second, within 24 h after pupal formation, when most CB NBs either terminally differentiate or undergo cell death (Ito and Hotta, 1992; Truman and Bate, 1988; Maurange et al., 2008; Siegrist et al., 2010; Homem et al., 2014; Yang et al., 2017). From a large-scale RNAi screen aimed at identifying genes regulating termination of neurogenesis in the CB during pupal stages, we identified Notch and Notch pathway components (Pahl et al., 2019). To better understand how neurogenesis terminates, we asked whether Notch signaling is also required for CB NB quiescence during the embryonic-to-larval transition.

At 0 h after larval hatching (ALH), the majority of CB NBs are small and quiescent, except for a subset, which include the four mushroom body (MB) NBs located on the dorsal surface (designated 1-4 in all confocal images) and the ventro-lateral (VL) NB (designated with an asterisk) (Fig. 1A; Fig. S1A-C) (Ito and Hotta, 1992; Truman and Bate, 1988; Britton and Edgar, 1998; Sousa-Nunes et al., 2011; Lin et al., 2013; Ding et al., 2016; Yuan et al., 2020). At this time, the MB and VL NBs are larger than quiescent CB NBs and are actively dividing based on expression of the S-phase indicator pcna:GFP, incorporation of the thymidine analogue EdU and their generation of EdU-positive progeny (Fig. S1A-C).

To determine whether Notch signaling regulates CB NB quiescence, we used GAL4/UAS to express UAS-RNAi transgenes targeted to Notch pathway components in NBs (worGAL4). In NotchRNAi animals, in addition to the MB+VL NBs, we observed on average one, occasionally two, Deadpan (Dpn)-positive CB NBs that incorporated EdU after 2 h of animal feeding, as well as more CB NBs expressing pcna:GFP compared with controls (Fig. 1B,C,G; Fig. S1D-F). We used a second RNAi line targeted against a different Notch exon and observed a similar phenotype (Fig. 1D,G). Unlike MB+VL NBs, the ectopic EdU positive CB NBs in NotchRNAi animals were small, similar in size to other quiescent CB NBs (Fig. 1H). Next, we assayed kuzRNAi, neuralizedRNAi and Su(H)RNAi animals and, again, found small ectopically proliferating EdU-positive CB NBs (Fig. S1G-J). To further substantiate whether Notch signaling regulates quiescence, we assayed the loss-of-function mutant allele, kuz^e29-4. In kuz^e29-4 homozygous animals, a modest, but significant, increase in the number of EdU-positive CB NBs was found compared with NotchRNAi animals (Fig. 1E,G). The number of Dpn-positive CB NBs was the same in kuz^e29-4 mutants as in controls (108.1±3.027 per brain hemisphere, n=6 animals; mean±s.e.m.), consistent with previous reports (Ulvklo et al., 2012; Bivik et al., 2016). Next, we used DeltaGAL4 to increase expression levels of UAS-RNAi transgenes. Likewise, we found an increase in the number of EdU-positive CB NBs compared with worGAL4,UAS-NotchRNAi animals (Fig. 1F,G). To confirm that DeltaGAL4 increased NotchRNAi expression levels resulting in better knockdown of Notch signaling, we assayed expression of the Notch activity reporter E(spl)mγ-GFP (Almeida and Bray, 2005). Compared with controls, E(spl)mγ-GFP reporter expression was reduced in CB NBs in worGAL4,UAS-NotchRNAi animals and even further reduced in DeltaGAL4,UAS-NotchRNAi (Fig. S1K-M). We conclude that Notch pathway activity positively regulates CB NB quiescence.

Next, we assayed earlier developmental time points to determine whether Notch signaling promotes quiescence entry or is required to maintain the quiescent state. In controls at embryonic stage 16, ∼30 CB NBs on average (excluding MB+VL NBs) incorporated EdU after 1 h treatment and, at stage 17, ∼3 CB NBs (Fig. 2A,D,F,H). At both stages, EdU-positive CB NBs were larger than EdU-negative CB NBs, yet the size difference was less at stage 17 (Fig. 2E). This suggests that CB NBs, as a population, complete their final reductive divisions between embryonic stages 16 and 17. In Notch RNAi animals at embryonic stages 16 and 17, we found more CB NBs that were proliferating based on EdU incorporation and pcna:GFP expression (Fig. 2B,D,G,H; Fig. S2A-C). We assayed total CB NB number in each of the brain hemispheres as well as CB NB size and found no differences in Notch RNAi animals compared with controls (Fig. 2I; Fig. S2D). This suggests that Notch signaling regulates CB NB entry into quiescence and when Notch pathway activity is reduced quiescence is delayed. To further test this possibility, we expressed a constitutively active form of the Notch receptor (Notch^ΔECD) in CB NBs (Vaccari et al., 2008). In Notch^ΔECD animals at embryonic stage 16, we found a reduction in the number of proliferating CB NBs compared with controls (Fig. 2C,D). We conclude that Notch pathway activity positively regulates timing of quiescence entry.

Next, we asked whether Notch signaling promotes quiescence by inhibiting CB NB cell cycle progression or by regulating temporal patterning. First, we used fluorescent ubiquitination-based cell cycle indicator (FUCCI) to determine cell cycle states of quiescent CB NBs (Zielke et al., 2014; Otsuki and Brand, 2018). In control animals at the freshly hatched larval stage (0 h ALH), more than half of quiescent CB NBs were G0/G1-arrested and the rest were G2-arrested (Fig. 3A,B,D). In contrast, in Notch RNAi animals, less than half were G0/G1-arrested and the rest were G2-arrested or actively dividing (Fig. 3C,D). This suggests that Notch signaling inhibits CB NB cell cycle progression. Next, we co-expressed a number of UAS-transgenes to manipulate CB NB cell cycle and/or growth in NotchRNAi animals, including UAS-dp110, the catalytic subunit of PI3-kinase, UAS-hippoRNAi, UAS-trblRNAi and UAS-dacapoRNAi. These transgenes were selected because they all affect genes and/or cell signaling pathways implicated previously in quiescence, entry and exit, and they all regulate aspects of NB cell cycle and growth (Leevers et al., 1996; Ding et al., 2016; Otsuki and Brand, 2018, 2019). We found a significant increase in the number of EdU-positive NBs in NotchRNAi, dacapoRNAi animals compared with animals expressing either NotchRNAi or dacapoRNAi alone (Fig. 3F,I,J). No differences were detected following co-expression of other transgenes (Fig. 3E-J). Next, to distinguish whether Notch promotes quiescence by regulating temporal patterning, we assayed expression of the late temporal factor, Cas. We found no difference in the number of Cas-positive CB NBs in control animals compared with NotchRNAi animals, consistent with previous reports (Fig. S2E; Ulvklo et al., 2012; Chang et al., 2013). Together, these results support that Notch signaling promotes quiescence by regulating activity of the cell cycle exit gene, dap, a known Notch target.

Next, to better understand how Notch controls quiescence, we used E(spl)mγ-GFP to assay Notch pathway activity (Almeida and Bray, 2005). At embryonic stage 16, E(spl)mγ-GFP reporter expression was detected in 90% of CB NBs, whereas at stage 17, 10% of CB NBs, and at 0 h ALH, 6% of CB NBs, retained expression of E(spl)mγ-GFP (Fig. 4A-C) (Zacharioudaki et al., 2012). Next, we assayed E(spl)mγ-GFP expression in MB+VL NBs, which divide continuously during the embryonic-to-larval transition. We found E(spl)mγ-GFP reporter expression in MB+VL NBs at these stages (Fig. 4D). This suggests that Notch signaling is active in proliferating, but not in quiescent, CB NBs. To test this possibility, we assayed E(spl)mγ-GFP reporter expression in cas mutants. Cas positively regulates quiescence in ventral nerve cord NBs by inhibiting expression of the early temporal factor Pdm (Kambadur et al., 1998; Tsuji et al., 2008). Consistent with previous reports (Tsuji et al., 2008), we found a significant number of EdU-positive CB NBs in brains of cas homozygous mutants compared with controls (Fig. 4E,G), the majority of which expressed E(spl)mγ-GFP (Fig. 4F,H). Next, we fed freshly hatched pcna:GFP larvae a sucrose-only diet. After 24 h of animal feeding, MB NBs continued dividing and expressed E(spl)mγ-GFP (Fig. 4I). In contrast, the VL NB stopped and E(spl)mγ-GFP reporter expression was not detected (Fig. 4I). Lastly, we blocked CB NB mitosis and division by knocking down the cdc25 phosphatase, string, in NBs. At embryonic stage 16, we found an equal number of EdU-positive CB NBs compared with controls, but a strong reduction in the number of EdU-positive CB NB progeny, consistent with G2-arrest (Fig. S3A,B). Compared with controls, stringRNAi animals had a strong reduction in CB NBs expressing the E(spl)mγ-GFP reporter (Fig. S3C-E). We conclude that Notch signaling is active in proliferating NBs, but not in NBs that are in quiescence or those that have exited the cell cycle.

To better understand how Notch activity is regulated during the embryonic-to-larval transition and to understand how Notch pathway activity becomes attenuated, we assayed expression and localization of the Notch receptor and the Notch ligands, Delta and Serrate. We assayed expression in both MB+VL NBs, which maintain Notch activity and continue dividing, and in the other CB NBs, which stop dividing and have low to no Notch activity. At embryonic stage 16 and in freshly hatched larvae at 0 h ALH, Notch was expressed in CB NBs, including MB+VL NBs (Fig. 5A; Fig. S4A). At embryonic stage 16, Delta was also expressed in CB NBs, including MB+VL NBs, but by 0 h ALH Delta was reduced in quiescent CB NBs compared with MB+VL NBs (Fig. 5B,C; Fig. S4B). This is similar to E(spl)mγ-GFP reporter expression (Fig. 4C), suggesting that decreases in Delta could account for Notch pathway attenuation. Next, we fed freshly hatched larvae (0 h ALH) a complete diet to reactivate quiescent CB NBs. Reactivated CB NBs expressed both Notch and Delta and had active Notch signaling based on E(spl)mγ-GFP reporter expression (Fig. S4C-H). In addition, we observed Delta in newborn GMCs, adjacent to their CB NB mothers (Fig. 5B; Fig. S4F, yellow bracket). Serrate was not detected at any of these time points (Fig. S4I,J). Together, we conclude that Notch signaling is required for quiescence, but that Notch is not active in quiescent CB NBs. Moreover, proliferating CB NBs, but not quiescent, generate Delta-expressing GMCs.

Expression of Delta in GMCs adjacent to proliferating NBs suggests that GMC-localized Delta transactivates Notch in NBs. To further examine this possibility, we fed freshly hatched animals a complete diet for 12 h to visualize the first CB NB S-phase and cell division after quiescence (Fig. 5D,E). At 12 h after feeding before S-phase entry, Delta levels were increased in CB NBs compared with CB NBs at 0 h ALH (Fig. 5D). Delta was localized symmetrically in mitotic NBs and, after division, Delta-expressing GMCs were found adjacent to their NB mothers (Fig. 5D,F). Coinciding with the production of Delta-expressing GMCs, CB NBs expressed E(spl)mγ-GFP (Fig. 5E) (Zacharioudaki et al., 2012). We conclude that NBs generate their own ligand-expressing daughters for Notch pathway transactivation and because CB NBs exit cell cycle, Notch activity becomes attenuated.

To determine whether inactivation of Notch signaling is important for quiescence or is simply a consequence of NB cell cycle exit and quiescence entry, we expressed UAS-numbRNAi in NBs to activate Notch signaling. Numb inhibits Notch signaling and, in numbRNAi animals, we found E(spl)mγ-GFP reporter expression in quiescent CB NBs at 0 h ALH, consistent with Notch pathway activation (Fig. 6A,B). Next, we fed numbRNAi animals a complete diet. After 24 h of feeding, most CB NBs in control animals had reactivated from quiescence based on their increased size, expression of pcna:GFP, incorporation of EdU and generation of EdU-positive progeny (Fig. 6C,E) (Sousa-Nunes et al., 2011; Lin et al., 2013; Sipe and Siegrist, 2017; Yuan et al., 2020). In contrast, in numbRNAi animals, few CB NBs incorporated EdU after 24 h of animal feeding, yet E(spl)mγ-GFP reporter expression was still maintained (Fig. 6D-G). We conclude that inactivation of Notch signaling is required for CB NBs to reactivate in response to dietary nutrients.

Although Notch is active in CB NBs throughout most of embryogenesis, Notch induces quiescence only at late stages. This suggests that Notch activity or Notch targets change over time. To assay Notch activity, we examined the subcellular localization of Notch ICD. We found a modest, but significant, increase in nuclear localized Notch ICD in stage 16 CB NBs compared with stage 10 (Fig. 7A), consistent with other studies (Ulvklo et al., 2012). Next, we asked whether temporal patterning restricts Notch function to late stages. In nubbin/pdm2 mutant wing discs, Notch target genes are ectopically expressed, and in nubbin/pdm2 mutants, ventral cord NBs enter quiescence prematurely (Neumann and Cohen, 1998; Tsuji et al., 2008). Because of technical constraints, we were unable to reduce Notch pathway components in nubbin/pdm2 mutants. However, we did assay quiescence in cas mutants following constitutive activation of Notch signaling. Compared with cas mutants alone, we found no difference in the number of EdU-positive CB NBs (Fig. 7B). These results suggest that temporal factor patterning provides Notch signaling competence to induce CB NB cell cycle exit and quiescence, either by regulating Notch activity levels or by regulating Notch targets.

Here, we report that Notch signaling regulates quiescence, entry and exit in Drosophila CB NBs (model Fig. 7C). Increasing Notch pathway activity induces CB NBs to exit cell cycle via a Dap-dependent mechanism. Dap, a cyclin-dependent kinase inhibitor and CIP/KIP family member, is a known Notch target gene as is Cyclin E, String (Cdc25) and E2F (Sriuranpong et al., 2001; Deng et al., 2001; Noseda et al., 2004; Herranz et al., 2008; Ishikawa et al., 2008; Bivik et al., 2016; Shang et al., 2016). Whether Notch regulates other cell cycle genes required for CB NB exit remains unknown. Once CB NBs stop dividing, Notch pathway activity becomes attenuated. Low to no Notch activity is required for CB NBs to exit quiescence in response to dietary nutrient cues. Thus, levels of CB NB Notch activity regulate both the entry and exit from quiescence. High Notch is required for entry, whereas low Notch is required for exit. Intestinal stem cells (ISCs) also experience a period of low to no Notch activity during mid-pupal stages and, if Notch is ectopically induced at this time, ISCs terminally differentiate into secretory enteroendocrine cells (Guo and Ohlstein, 2015). Whether any of the quiescent CB NBs with ectopic Notch terminally differentiate remains unknown.

Although Notch is required for quiescence, most CB NBs do stop dividing, albeit late. This suggests that other genes or signaling pathways are required or that residual Notch activity is sufficient to induce CB NB quiescence. Unfortunately, Notch null mutants are embryonic lethal (Lehmann et al., 1983; Parody and Muskavitch, 1993; Brennan et al., 1997; Leonardi et al., 2011). In addition, we have shown that Notch is sufficient to induce quiescence. Although quiescence occurs prematurely, it is still restricted to late embryonic stages. Restriction of Notch function (cell cycle exit) is likely due to CB NB intrinsic temporal programs. Temporal programs that likely vary across CB NB lineages, as is the case in the ventral nerve cord. Whether temporal programs regulate levels of Notch pathway activity or provide additional factors needed for CB NB cell cycle exit remains unanswered.

In mammals, Notch signaling is more complicated because of gene duplication. Yet, recently, Notch signaling has been shown to regulate neural stem cell quiescence. In Notch2 conditional knockout mice, more neural stem cells are actively dividing in the hippocampus and subventricular zone brain regions in adult animals compared with controls (Engler et al., 2018; Zhang et al., 2019). This results in premature depletion of the neural stem pool and reduced neurogenesis in older mice (Engler et al., 2018; Zhang et al., 2019). This is similar to what we report here and, in the future, it will be interesting to determine which Notch ligands are required and whether neural stem cells in mammals use their newborn daughters for pathway activation.

Fly stocks used in this study and their source are listed in Table S1.

Embryos were collected for 0-2 h or 0-4 h after egg laying (AEL) and aged for 13-22 h at 25°C. For experiments using tub-Gal80ts (temperature sensitive), embryos and larvae were kept at 29°C. For larval staging, animals were collected immediately after hatching and transferred to either standard Bloomington fly food or 20% sucrose-only solution for the desired amount of time. For embryonic staging, embryos were aged at 25°C for 13-15 h AEL to get stage 16 embryos and for 16-20 h AEL to get stage 17 embryos. For genotypes (kuz^e29-4 and cas²⁴) where embryos failed to hatch, embryos were dechorionated and dissected (Lee et al., 2009) at 24 h AEL to get larvae.

For EdU incorporation assays, Bloomington fly food or Schneider's insect media was supplemented with 0.2 mM EdU. For larval EdU incorporation experiments, freshly hatched larvae were transferred to EdU food for 2 h at 25°C. For embryonic EdU incorporation experiments, embryos were dechorionated and dissected (Lee et al., 2009). Dissected embryonic tissue was transferred to Schneider's insect media supplemented with EdU for 60 min at room temperature. EdU incorporation was detected using the Click-iT EdU Proliferation Kit for Imaging and Alexa Fluor 647 dye (Thermo Fisher Scientific, C10340) as described previously (Sipe and Siegrist, 2017; Yuan et al., 2020).

Embryonic and larval brains were dissected as described previously (Sipe and Siegrist, 2017; Yuan et al., 2020). In brief, dissected tissues were fixed in 4% EM-grade formaldehyde in PEM buffer for 20 min and rinsed in 1× PBS with 0.1% Triton X-100 (PBT). Tissues were blocked overnight at 4°C in 10% normal goat serum in PBT followed by antibody staining. Primary antibodies used include chicken anti-GFP (1:500, Abcam, ab13970), rat anti-Deadpan (Dpn, 1:100, Abcam, ab195173), mouse anti-Notch ICD (1:250; Developmental Studies Hybridoma Bank, C17.9C6), rabbit anti-dsRed (1:1000; Clontech, 632496) and rabbit anti-Scribble (Scrib, 1:1000, gift from Chris Q. Doe, University of Oregon, OR, USA). To detect primary antibodies, the following Alexa-Fluor conjugated secondary antibodies were used: goat anti-chicken Alexa 488 (1:300; A32931), goat anti-mouse Alexa 488 (1:300, A11001), goat anti-rat Alexa 555 (1:300, A48263), goat anti-rabbit Alexa 555 (1:300, A21428) and goat anti-rabbit Alexa 633 (1:300, A21071) (Thermo Fisher Scientific).

Images encompassing the entire brain hemispheres were acquired using a Leica SP8 laser scanning confocal microscope equipped with a 63×/1.4 NA oil-immersion objective and analyzed using Fiji software. All images were processed using Fiji and Adobe Photoshop and figures were assembled using Adobe Illustrator. NBs were identified based on Dpn expression and superficial location. The Fiji ‘cell counter’ plugin was used to count and track the number of EdU- and/or GFP-expressing Dpn-positive NBs. NB size was calculated by averaging the lengths of two perpendicular lines through the center of the NB in Fiji.

Quantification of fluorescence was performed in Fiji. Cytoplasmic and nuclear Notch ICD levels were quantified as follows. NBs labeled with Scribble and nuclei labeled with Dpn were manually traced and the average Notch ICD fluorescence intensity measured in the whole cell and the nucleus using Fiji. Notch ICD nuclear fluorescence intensity was determined as a ratio of nuclear Notch ICD fluorescence intensity to whole cell Notch ICD fluorescence intensity.

All data is represented as mean±s.e.m. and statistical significance was determined using unpaired two-tailed Student's t-tests or ANOVAs in Prism 9.

We thank Cheng-Yu Lee for providing fly stocks. We thank the Bloomington Drosophila Stock Center, Harvard TRiP and the Developmental Studies Hybridoma Bank for providing flies and antibody reagents. We especially thank Siegrist Lab members, Md Ausrafuggaman Nahid, and Karsten Siller for providing comments on the manuscript.

The peer review history is available online at https://journals.biologists.com/dev/article-lookup/doi/10.1242/dev.200275.