Syndecan is required for long-term Drosophila intestinal stem cell maintenance

Sdc, a previously uncharacterised player in the adult Drosophila intestine, is expressed throughout the intestinal epithelium, in all cell types (Figs 1B, 1C and S1). To test whether Sdc contributes to intestinal cell production, we knocked down Sdc by RNAi expression in progenitor cells (Fig 1A) using the esg^ts F/O (“escargot flip out”) system [15] (Methods). This system also drives heritable GFP expression in progenitor cells and their progeny, allowing assessment of progenitor maintenance and new cell production. Expression of two out of three independent RNAi lines impaired new cell production as evidenced by limited appearance of newly differentiated large GFP^+ve cells in the tissue (Figs 1D-F, S2A and S2B) and resulted in a progressive loss of small GFP^+ve progenitor cells (Figs 1D and S2A–S2C).

To identify in which cell type(s) of the posterior midgut Sdc is required, we performed RNAi knockdown with cell type-specific GAL4 drivers (Fig 1A and Methods). We used Sdc RNAi 3 (henceforth referred to as ‘Sdc RNAi’ for simplicity) as this line yielded highly effective knockdown (S2A’ Fig) and has also been validated and used in other work [16,17]. We employed w¹¹¹⁸ as the control (referred to as ‘Control’) to match the genetic background in which Sdc RNAi 3 was originally constructed [18], as this is known to impact intestinal stem cell dynamics during aging [19]. We monitored the proportion of each cell type to assess whether Sdc is required for its maintenance. In addition, since failure in new cell production is accompanied by low epithelial cell density (Fig 1F), we measured cell density as a proxy for new cell production. Sdc knockdown in ISCs resulted in a progressive reduction in ISC proportion (Fig 1G and 1H). This ISC loss was associated with reduced cell density (Fig 1I), presumably due to a failure in new cell production (Fig 1E). ISC loss was also observed using two further Sdc RNAi lines (S2D Fig). Fly survival was not compromised (S3 Fig), consistent with other reports using alternative modes of ISC elimination [20,21].

Knockdown in other cell types did not perturb their proportion (Fig 1J), except for EEs which became more abundant. Based on known roles of Sdc in other Drosophila tissues [22–25], this could be due to Sdc modulating Slit/Robo signalling involved in EE specification [26,27]. Nevertheless, across all knockdowns in cell types other than ISCs, cell density was comparable to control guts (Fig 1K), suggesting Sdc is dispensable in other epithelial cell types. We then tested whether Sdc overexpression in ISCs would induce an opposite effect on ISC proportion or cell density, but it did not (Fig 1H and 1I). Altogether, we have identified Sdc as a previously unrecognised but important factor required for stem cell maintenance in the adult Drosophila intestine.

Sdc-depleted intestinal stem cells are partially lost through apoptosis and display aberrant cell shapes

ISCs could be eliminated from the tissue by differentiation or cell death. Differentiation is ruled out by the reduction in differentiated cells upon expression of Sdc RNAi with the esg^ts F/O lineage tracing system (Figs 1D, 1E and S2; compare to control guts where production of large GFP^+ve ECs is clear). To test whether ISC loss occurs by apoptosis, a common form of programmed cell death, we expressed the inhibitor of apoptosis, DIAP1 [28], in these cells. DIAP1 expression partially rescued (~50%) ISC loss caused by Sdc depletion and did not affect ISC numbers in control guts (Fig 2A and 2B). Thus, some ISCs depleted of Sdc are eliminated by apoptosis.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Sdc-depleted ISCs adopt abnormal cell shapes.

(A) Surface views of midguts expressing +/− DIAP1 and +/− Sdc RNAi in ISCs. DNA stain (blue) marks the nuclei of all cells; YFP (green) marks ISCs. (B) ISC proportion. n = number of guts, from three replicates. (C) Surface views of ISCs. DNA stain (blue/white) marks nuclei; YFP (green) marks ISCs. Images are from either 14 or 28 days expression with esgSu(H). (D) ISC cytoplasmic area; (E) ISC aspect ratio and (F) ISC form factor. n = number of ISCs analysed, from ≥27 guts, from three replicates. (G) Time lapse of control or Sdc RNAi-expressing ISCs. YFP (black) marks ISCs. Right panels show colour-coded snapshots of the cell outline during the time lapse. ns: not significant, ****: p < 0.0001.

https://doi.org/10.1371/journal.pgen.1011586.g002

We then sought to determine the cellular changes induced by loss of Sdc, which could cause ISC elimination. Observation of labelled Sdc-depleted ISCs at high magnification revealed striking changes in cell shape. Drosophila ISCs are normally small and triangular shaped [6,29,30] (Fig 2Ci-ii). In contrast, Sdc-depleted ISCs showed a variety of morphological defects including larger size, cytoplasmic protrusions, and occasional blebs (Fig 2Ciii-vii). Quantification of ISC area and shape revealed that, compared to controls, ISCs expressing Sdc RNAi were larger (Fig 2D), more elongated (Fig 2E) and more convoluted (Fig 2F). ISC-specific DIAP1 expression, with or without concomitant Sdc depletion, also resulted in larger cells (S4A Fig). In all genotypes, larger cells were more convoluted (S4B-E Fig), with a stronger effect seen when Sdc RNAi was also expressed (S4C and S4E Fig). Focussing on these larger (>50 μm²) cells, we observed similar deformations when Sdc RNAi was expressed both with and without DIAP1 (S4F and S4G Fig), suggesting that these cell shape changes are not primarily caused by apoptosis.

Abnormal cell shape and size might reflect changes in the cortical cytoskeleton and cellular dynamics. We therefore turned to live imaging of fluorescently labelled ISCs in ex vivo guts to compare dynamic cellular behaviour over approximately eight hours (>400 ISCs imaged across >14 guts for each genotype). Whilst most control ISCs remained static (Fig 2Gi), Sdc RNAi-expressing ISCs displayed more dynamic protrusions and shape changes (Fig 2Gii). In addition, rare Sdc RNAi-expressing ISCs appeared to attempt but fail in cell division (Fig 2Giii), a behaviour that we have never observed in controls. We concluded that Sdc modulates stem cell shape and that its loss can contribute to changes in cellular morphology compatible with defective divisions.

Sdc is dispensable for intestinal stem cell abscission but prevents nuclear lamina invaginations and DNA damage

How could Sdc interfere with stem cell morphology and maintenance? Sdc-4 links the abscission machinery to the plasma membrane in mammalian cultured cells, with Sdc-4 depletion delaying abscission or more rarely causing abscission failure, generating binucleate cells [31]. To explore whether Sdc depletion from proliferating Drosophila ISCs impairs their resolution into two separate daughter cells and causes them to become binucleate prior to their elimination, we immunostained guts with Lamin B to label the nuclear compartment and α-catenin to label the cell perimeter (Fig 3A). To confirm that our approach could detect binucleate cells, we used RNAi to knockdown the cytokinesis-promoting Aurora B [32] (Fig 3A). We did not detect any binucleate stem cells upon Sdc knockdown (n>100 ISCs examined across >20 guts) (Fig 3A), indicating that Sdc is not required for ISC abscission.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Sdc prevents nuclear lamina invaginations and DNA damage.

(A) Control ISC, and ISC expressing Aurora B RNAi or Sdc RNAi. DNA stain (white) marks nuclei; YFP (green) marks ISCs; anti-α-catenin (magenta) marks cell junctions; anti-Lamin B (white) marks nuclear lamina. Yellow arrowheads indicate nuclear lamina invaginations. (B) The Drosophila LINC complex. (C) ISC nuclear lamina invaginations (Lamin B pixels located in the nuclear interior), (D) ISC nuclear eccentricity, and (E) ISC nuclear form factor. n = number ISCs analysed, from a minimum of 17 guts per genotype, from a minimum of two replicates. (F) Surface view images of midguts. DNA stain (blue) marks nuclei; YFP (green) marks ISCs; anti-γH2Av (white) marks DNA damage. Yellow arrowheads indicate ISCs with DNA damage. (G) Proportion of ISCs, per gut, with DNA damage. n = number of guts, from a minimum of two replicates. >100 ISCs analysed per genotype per timepoint. (H) Relationship between ISC nuclear lamina invaginations and DNA damage. Each dot represents an individual ISC, coloured orange if ISC has DNA damage and grey if ISC does not. ISCs are defined as having nuclear lamina invaginations if there are ≥2000 Lamin B pixels in the nuclear interior. n = number ISCs analysed, from ≥22 guts per genotype, from four replicates. Example ISCs, with and without nuclear lamina invaginations and DNA damage. DNA stain (white) marks nuclei; YFP (green) marks ISCs; anti-Lamin B (white) marks the nuclear lamina; anti-γH2Av (white) marks DNA damage. (I) Model representing the role of Sdc in Drosophila ISCs. ns: not significant, **: p < 0.01, ***: p < 0.001, ****: p < 0.0001.

https://doi.org/10.1371/journal.pgen.1011586.g003

We noted, however, that unlike control ISC nuclei which appeared approximately spherical, with a smooth, round nuclear lamina (Fig 3A), the nuclei of Sdc-depleted ISCs presented frequent nuclear lamina invaginations (Fig 3A and 3C) and aberrant nuclear shapes, with more elongated and lobed nuclei (Fig 3D and 3E). These phenotypes were also seen with other Sdc RNAi lines and when DIAP1 was co-expressed with Sdc RNAi (S5A Fig), indicating that they were neither an off-target effect of the RNAi nor a secondary effect of apoptotic cell death.

We reasoned that Sdc depletion could impair a relay from the plasma membrane to the nucleus, which would in turn alter nuclear properties. The Linker of Nucleoskeleton and Cytoskeleton (LINC) complex is a conserved multiprotein complex that connects the cytoskeleton with the nucleoskeleton, allowing transmission of mechanical force from the cell surface to the nucleus, regulating nuclear deformation, positioning and functions [33,34]. The Drosophila LINC complex consists of two KASH proteins, Klarsicht (Klar) and Msp300, and two SUN proteins, Klaroid (Koi) and the testes-specific SPAG4 (Fig 3B) [33]. We hypothesised that Sdc might act via the LINC complex and thus tested whether individual knockdowns of Klar, Koi and Msp300 in ISCs recapitulated the lamina invaginations and nuclear shape changes seen upon Sdc depletion. The knockdown efficacy of all RNAi lines used here has previously been validated by qPCR and/or immunostaining [35–38]. Only Msp300 knockdown resulted in increased lamina invaginations compared to control nuclei, but not to the same extent as Sdc knockdown (Figs 3C and S5A). Knockdown of Klar or Koi produced only modest nuclear elongation (Fig 3D) and nuclear lobing was unaffected (Fig 3E).

Disruptions to the nuclear lamina are often associated with DNA damage [39,40]. Immunolabelling of DNA double strand breaks with γH2Av [41] revealed striking DNA damage acquisition in ISCs upon Sdc knockdown but not in control ISCs (Fig 3F and 3G). Increased DNA damage was also seen with a second Sdc RNAi line, albeit at a lower frequency (S5B Fig). These results suggest that Sdc affects a mechanism protecting the genome. Knockdown of individual or pairs of LINC complex components from ISCs (to exclude potential redundancy [33,34]) did not recapitulate the level of DNA damage seen upon Sdc depletion (Figs 3F, 3G and S5C), suggesting that Sdc function is unlikely to be fully accounted for by LINC complex proteins. It should be noted that γH2Av staining in the Msp300 RNAi condition is predominantly found in ECs, not ISCs (Fig 3F). How Msp300 loss in ISCs causes DNA damage in neighbouring cells remains to be determined, however, it is clear that this does not recapitulate the high levels of ISC-specific DNA damage seen upon Sdc RNAi knockdown (Fig 3F and 3G). Notably, DNA damage was not a secondary consequence of apoptosis (S5D Fig), and thus may instead contribute to ISC elimination.

We carefully examined the frequency of both nuclear lamina invaginations and DNA damage in a large number of ISCs (Fig 3H). DNA damage was found more frequently in Sdc-depleted ISCs with lamina invaginations compared to those without (Fig 3H), demonstrating a relationship between acquisition of DNA damage and disruption of the nuclear lamina.

Altogether, our results suggest that Sdc modulates nuclear shape and the nuclear lamina, in a manner that is not dependent on single members of the LINC complex. In the absence of Sdc, nuclear aberrations arise early, and the acquisition of DNA damage could conceivably trigger ISC elimination (Fig 3I).

Sdc controls nuclear envelope remodelling of neural stem cells but is dispensable from female germline stem cells

Next, to test whether Sdc plays similar roles in other stem cell types, we examined female germline stem cells (fGSCs) and larval neural stem cells (neuroblasts).

In the female Drosophila germline (S6A Fig), Sdc depletion from fGSCs did not cause germline loss (S6A and S6B Fig), suggesting that Sdc is dispensable for fGSC survival. Moreover, we did not detect any evidence of germline stem cysts, which arise upon defective fGSC abscission [42] (S6A and S6B Fig), indicating that Sdc is not required for abscission in these cells. In addition, we did not detect any γH2Av staining in fGSCs (S6C Fig), indicating that Sdc is not required for genome protection in these cells. Overall, presumably due to low levels of expression compared to the surrounding somatic cells (S6D Fig), our results suggest Sdc is dispensable in fGSCs, supporting previous data [43].

In the larval Drosophila brain, which is highly amenable to live imaging [44,45], neuroblasts undergo rapid asymmetric cell division, generating a large self-renewing neuroblast with a large nucleus and a smaller differentiating ganglion mother cell (GMC) with a small nucleus [46] (Fig 4A). Wild type neuroblasts undergo semi-closed mitosis, whereby the nuclear envelope persists throughout mitosis, dependent on the maintenance of a supporting nuclear lamina [47]. Some reservoirs of nuclear membrane are also present in the cytoplasm of the renewing neuroblast and contribute to differential nuclear growth of the daughter cells [47] (Fig 4A and 4B).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Sdc knockdown in neural stem cells alters nuclear envelope remodelling and asymmetric cell division.

(A) Drosophila neuroblast division. (B-C & D-E) Klaroid::GFP (white/green) marks inner nuclear membrane; Cherry::Jupiter (magenta) marks mitotic spindle. Neuroblast outlined with white dashed line. Time stamp in minutes and seconds, with anaphase onset defined as time 0. (B) Time lapse of a control mitotic neuroblast. Representative time lapse from 36 neuroblasts from 4 independent experiments (20 brains total). (C) Example images of mitotic neuroblasts expressing Sdc RNAi with mitotic nuclear envelope phenotypes. Representative images from 45 neuroblasts from two independent experiments (7 brains total). (C’) Prevalence of mitotic nuclear envelope phenotypes. n numbers same as B&C. (D) Time lapse of a mitotic neuroblast expressing Sdc RNAi, with a large rupture of the nuclear envelope (yellow arrowheads). Yellow dashed line marks cleavage furrow. (E) Time lapse of a mitotic neuroblast expressing Sdc RNAi, with complete dispersion of the nuclear envelope. Arrowheads mark abnormal nuclear membrane aggregates in the cytoplasm of the renewing neuroblast. (F) Neuroblast nuclear form factor, (G) Neuroblast nuclear size and (H) Neuroblast cell size, measured during interphase. n = number of neuroblasts, from 13 brains from 3 independent experiments (control) or 7 brains from 2 independent experiments (Sdc RNAi). (I) Nuclear size ratio between the renewed neuroblast and newly formed GMC, measured at telophase. Neuroblast nuclear envelope phenotype indicated for each data point. n = number of neuroblast/GMC pairs measured, from 13 brains from 3 independent experiments (control) or 7 brains from 2 independent experiments (Sdc RNAi). ***: p < 0.001, ****: p < 0.0001.

https://doi.org/10.1371/journal.pgen.1011586.g004

After confirming the presence of Sdc in neuroblasts (S7 Fig), we examined whether its depletion affected cell division and nuclear properties. Sdc-depleted neuroblasts exhibited prolonged cell division (Fig 4B, 4D and 4E) and had a variety of cell division, nuclear envelope and nuclear shape defects. Approximately 50% of mitotic neuroblasts expressing Sdc RNAi displayed abnormal mitotic nuclear envelope remodelling (Fig 4C-C’) including ruptures (Fig 4C and 4D) and dispersion (Fig 4C and 4E) of the nuclear envelope, which appeared as early as metaphase. Outside of mitosis, Sdc-depleted neuroblasts had aberrant nuclear shape and size (Fig 4F and 4G) and abnormal cell size (Fig 4H).

Furthermore, Sdc depletion resulted in abnormal ratio in nuclear size between the renewed neuroblast and the newly formed GMC (Fig 4I). This was sometimes caused by mispositioning of the cleavage furrow and abnormal nuclear membrane partitioning between the daughter cells, resulting in abnormally large GMC nuclei (Fig 4D), and in other cases was caused by impaired nuclear growth, resulting in abnormally small neuroblast nuclei (Fig 4E). These defects in nuclear size ratio were predominantly associated with divisions where the nuclear envelope was either ruptured or completely dispersed (Fig 4I), suggesting that mitotic nuclear envelope remodelling defects induced by Sdc knockdown might have a causative effect on nuclear size ratio. These defects did not elicit detectable levels of DNA damage (S8A-C Fig). Furthermore, brain lobe sizes from larvae with Sdc-depleted neuroblasts were comparable to control brains (S8D Fig), suggesting that Sdc depletion does not affect neuroblast ability to generate differentiated progeny.

Thus, in at least two different somatic stem cell types, in the adult gut and developing brain, the transmembrane protein Sdc modulates stem cell nuclear properties, with tissue-specific impacts on cell division and tissue renewal.

Fly strains

All Drosophila stocks were maintained at 18°C or amplified at 25°C on standard medium (cornmeal, yeast, glucose, agar, water, Nipagin food medium). Fly strains are referenced in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Fly stocks used in this work.

https://doi.org/10.1371/journal.pgen.1011586.t001

We used FlyBase (release FB2023_06) to find information on phenotypes/function/stocks/gene expression etc. [82].

The fly line expressing Sdc-Scarlet-2xHA was produced by Genetivision. The following guide RNAs (gRNA) were selected for CRISPR/Cas9 mediated double strand breaks, for targeting immediately after the final codon of the syndecan gene: TATCTCAGGCGTAGAACTCG CGG and GTGTGCGTATGACTGGACGA AGG. The donor template for homology directed repair was designed to contain a four-serine linker, Scarlet and a HA-GG-HA tag.

Experimental conditions for gut analyses

All analyses were performed on mated females.

For experiments involving GAL80^ts, flies were reared at 18°C throughout embryonic, larval and pupal development, with crosses flipped every 3–4 days. Flies were collected every 3–4 days after hatching and transferred to 29°C, (which was defined as day 1) and transferred to new food every two days.

For the survival experiment, crosses of 15 esgSu(H) virgins and 6 w¹¹¹⁸ or Sdc RNAi males were set up at 18°C and flipped every 3–4 days. Upon hatching, adult flies were collected and transferred to 29°C for two days to allow mating. On day 2, flies were anesthetised for minimal time to retrieve female flies of the correct genotypes, which were transferred into vials of standard medium at a density of 12 females per vial. All vials were placed horizontally at 29°C, with vial position randomised to control for any variation in temperature or humidity in the incubator. Survival was recorded daily, and flies were transferred to fresh food every 2 days until the end of the experiment, when the final fly died.

esg^ts F/O (“escargot flip out”) experiments [15] were performed as in [58]. At the non-permissive temperature (18°C), GAL80^ts which is expressed under the ubiquitous tubulin promoter, inhibits GAL4 preventing expression of UAS-GFP, UAS-RNAi and UAS-flp-recombinase. Upon hatching, adult flies are transferred to the permissive temperature (29°C). This inactivates GAL80^ts, allowing the intestinal progenitor cell-specific esg-GAL4 to drive expression of UAS-transgenes. The flp-recombinase excises a STOP codon between the ubiquitous actin promotor and GAL4, resulting in permanent and heritable GAL4 expression, and thus UAS-transgene expression, regardless of cell type. Thus, all cells which arise from progenitors after the temperature shift will express GFP, providing a visual readout for new cell production.

Gut dissection and immunostaining

Midguts were dissected in 1X PBS (Oxoid, BR0014G) and fixed for 20 minutes at room temperature in fresh 4% methanol-free paraformaldehyde (Polysciences, 18814-10) diluted in PBS, 0.025% Triton X-100 (Sigma Aldrich, X100). Samples were given three 5-minute washes in 0.25% PBST and permeabilised for 30 minutes in 1% PBST. Samples were then incubated for 30 minutes at room temperature in blocking buffer (0.1% PBST, 0.1% BSA (Sigma, A2153-10G)), followed by incubation overnight at 4°C with primary antibodies diluted in blocking buffer. Samples were next given three 20-minute washes in 0.25% PBST and incubated for 2 hours at room temperature in secondary antibodies diluted in 0.25% PBST. Finally, guts were given three 20-minute washes in 0.25% PBST. Guts were mounted in Vectashield (Vector Laboratories, H-1000) on a glass slide (VWR International, SuperFrost 1.0mm, ISC 8037/1) with coverslip (Menzel-Gläser, 22x50mm#1, 12342118). The following primary antibodies were used: chicken anti-GFP, 1/1000 (Abcam, ab13970); rat anti-α-catenin, 1/20 (DSHB, D-CAT-1-s); rabbit anti-HA, 1/500 (Cell Signaling, 3724T); mouse anti-Sdc, 1/200 (Abmart 11053-1-2/C87); rabbit anti-γH2Av, 1/700 (Rockland, 600-401-914); mouse anti-LaminDm0 (Lamin B), 1/10 (DSHB, ADL84.12). Alexa-488-, 555-, and 647-conjugated secondary goat antibodies (Molecular Probes) were used. F-actin was stained with Phalloidin (Molecular Probes, A12380 or A22287). Nuclei were stained with DAPI (Molecular Probes, D1306) or Hoechst (Molecular Probes, H1399).

Ovary dissection and immunostaining

Ovaries were dissected and collected in 1X PBS and then fixed for 25 minutes in 4% formaldehyde in 0.3 PBSTX (0.3% Triton-X in 1X PBS) at room temperature. Samples were given three 15-minute washes in 0.3% PBSTX and then incubated for one hour in blocking buffer (0.2 ug/μl BSA in 0.3% PBSTX), followed by incubation overnight at 4°C with primary antibodies and Phalloidin diluted in blocking buffer. Samples were then given three 15-minute washes in 0.3% PBSTX and incubated for 2 hours at room temperature in secondary antibodies diluted in blocking buffer. Samples were then given three 15-minute washes in 0.3% PBSTX, including Hoechst (33342) DNA stain in the first wash. Samples were mounted in Vectashield media (Vector Laboratories, H-1000). The following primary antibodies were used: mouse anti-γH2Av, 1/200 (DSHB, UNC93-5.2.1); mouse anti-α-Spectrin, 1/200 (DSHB, 3A9).

smFISH probes against Sdc were designed with the Stellaris Probe Designer (Biosearch Technologies), as described in S1 Table, and labelled with ATTO 633 as described in [84]. Briefly, after fixation and washes as above, ovaries were incubated in wash buffer (2X saline sodium citrate (SSC), 10% deionised formamide in nuclease-free water) for 10 minutes at room temperature. smFISH probes, primary antibodies and Phalloidin (Alexa Fluor 488 Phalloidin, ThermoFisher Scientific) were added to hybridisation buffer (2X SSC, 10% deionised formamide, 20mM vanadyl ribonucleoside complex, 0.1 mg/ml BSA, competitor (1:50 dilution of 5 mg/ml E.coli tRNA and 5 mg/ml salmon sperm ssDNA) in nuclease-free water), and ovaries were incubated in this buffer at 37°C overnight in the dark. Samples were given three 15-minute washes and incubated with secondary antibodies in wash buffer for 2 hours at room temperature. Finally, samples were washed in wash buffer, with the addition of Hoechst in one wash step, and mounted using Vectashield media (Vector Laboratories).

Larval brain dissection and immunostaining

Seventy-two to one hundred forty-four-hour larvae were dissected in Schneider’s insect medium (Sigma-Aldrich, S0146) and brains were fixed for 20 minutes in 4% paraformaldehyde in PEM (100mM PIPES pH 6.9, 1mM EGTA and 1mM MgSO4). After fixing, the brains were washed with PBSBT (1 × PBS (pH7,4), 0.1% Triton X-100 and 1% BSA) then blocked with PBSBT for one hour. Primary antibody dilution was prepared in PBSBT and brains were incubated for 48 hours at 4°C. Brains were given four 30-minute washes with PBSBT, then incubated with secondary antibodies diluted in PBSBT at 4°C overnight. The next day, brains were given four 20-minute washes with PBST (1x PBS, 0.1% Triton X-100) and kept in Vectashield antifade mounting medium with DAPI (Vector laboratories, H1200), at 4°C. The following primary antibodies were used: chicken anti-GFP, 1/1000 (Abcam, Ab13970); mouse anti-Prospero, 1/100 (DHSB, MR1A); rat anti-Elav, 1/100 (DHSB, 7E8A10); rat anti-Deadpan, 1/50 (Abcam, Ab195172); rabbit anti-Asense, 1/500 (gift from Yuh Nung Jan); rabbit anti-γH2Av, 1/500 (Rockland. 600-401-914).

RNA quantification

For each biological replicate (n = 3), total RNA was extracted from five third instar larvae using RNeasy Plus mini kit (Qiagen) according to manufacturer’s instructions. cDNA synthesis was performed using Superscript III reverse transcriptase (Invitrogen). For each experiment, transcript levels were measured relative to Ribosomal Protein 49 (Rp49) expression, in technical triplicates using the iTaq Universal SYBR green supermix (Biorad) and the primer pairs shown below.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Image acquisition and processing

Confocal images of fixed samples were acquired on a Leica TCS SP8 advanced confocal microscope with laser gain and power consistent within experiments. Gut overview images were acquired on a Leica M165FC microscope with Leica DFC3000G camera or EVOS M7000 (ThermoFisher Scientific). All images were processed and analysed with Fiji [85]. Imaris 3 64 7.5.2 was used to analyse neuroblast data presented in Fig 4. Figures were compiled in Adobe Illustrator.

Unless otherwise specified, all images show the R4/5 posterior midgut region. Midgut surface views are z-projections through half the gut depth. Images of single ISCs for nuclear lamina and shape qualifications are single z sections.

Live imaging of larval neuroblasts

Live imaging experiments were performed on intact brains. Larvae expressing nuclear membrane and spindle marker (Klaroid::GFP and UAS-Cherry::Jupiter) together with UAS-Dicer, UAS-Sdc RNAi (VDRC 13322) and worniu-GAL4 were dissected seventy-two or ninety-six hours after egg laying, respectively, in imaging medium (Schneider’s insect medium mixed with 10% fetal bovine serum (FBS) (Sigma, F7524), 2% PenStrepNeo (Sigma, P4083), 0.02 mg/mL insulin (Sigma, 11070), 20mM L-glutamine (Sigma, G8540), 0.04 mg/mL L-glutathione (Sigma, G4251) and 5 mg/mL 20-hydroxyecdysone (Sigma, H5142)) warmed up to room temperature before use. Brains were then transferred onto IbiTreat micro-slide 15 well 3D (Ibidi, 81506) and imaged with a confocal spinning disc. Neuroblasts were imaged in the brain lobes, and we did not distinguish between type I and II neuroblasts. All images presented in Fig 4 are single z sections. Anaphase onset is defined as t=0.

Live imaging and analysis of adult intestinal stem cells

Guts expressing YFP +/− UAS Sdc RNAi in ISCs (with esgSu(H) system) were carefully dissected 7 to 16 days after induction of GAL4 expression at 29°C in Shield and Sang M3 medium (Sigma, S3652), keeping the crop, Malpighian tubules and ovaries attached. Guts were mounted in imaging medium (Shields and Shang M3 medium), supplemented with 2% Fetal Bovine Serum, 0.5% penicillin-streptomycin (Invitrogen, 15140–122) and methylcellulose (Sigma, M0387-100G) 2.5% wt/vol, to stabilise the gut in the chamber [86] between a concanavalin-A coated coverslip and an oxygen permeable membrane and left to settle for 5 min. All guts were live imaged for 10–15 hours with a 2-minute interval between each scan comprising 20–25 sections of 1μm, on a 40X air objective using the EVOS M7000 (ThermoFisher Scientific). Image files were processed in FiJi, using Bleach correction and Stackreg plugins. Regions of interest (25 × 25μm) were cropped, upscaled to 512 × 512 pixels and reduced to 264 images (528 min). Cell shapes at specified timepoints were drawn and superposed using Adobe Illustrator.

Image analysis and quantifications

Statistical analyses

Data were plotted with GraphPad Prism 10 for Mac OS X. In all graphs, red line represents median. For pairwise comparisons, statistical significance was calculated using a Mann–Whitney test and p values <0.05 were considered statistically significant. Categorical data in Fig 3H were compared using the Chi square test. Categorical data in S2A and S2B Fig were compared using the Fisher Exact test due to the low expected frequency in some categories. Survival was assessed using the Log-rank test. Correlation in S4 Fig was tested with a Spearman nonparametric test. Statistics compare to age-matched control, unless otherwise indicated.