Src kinase function controls progenitor cell pools during regeneration and tumor onset in the Drosophila intestine


Src non-receptor kinases have been implicated in events late in tumor progression. Here, we study the role of Src kinases in the Drosophila intestinal stem cell (ISC) lineage, during tissue homeostasis and tumor onset. The adult Drosophila intestine contains only two progenitor cell types, division-capable ISCs and their daughters, postmitotic enteroblasts (EBs). We found that Drosophila Src42a and Src64b were required for optimal regenerative ISC division. Conversely, activation of Src42a, Src64b or another non-receptor kinase, Ack, promoted division of quiescent ISCs by coordinately stimulating G1/S and G2/M cell cycle phase progression. Prolonged Src kinase activation caused tissue overgrowth owing to cytokine receptor-independent Stat92E activation. This was not due to increased symmetric division of ISCs, but involved accumulation of weakly specified Notch+ but division-capable EB-like cells. Src activation triggered expression of a mitogenic module consisting of String/Cdc25 and Cyclin E that was sufficient to elicit division not only of ISCs but also of EBs. A small pool of similarly division-capable transit-amplifying Notch+ EBs was also identified in the wild type. Expansion of intermediate cell types that do not robustly manifest their transit-amplifying potential in the wild type may also contribute to regenerative growth and tumor development in other tissues in other organisms.


Numerous studies have implicated elevated Src levels in cancer, including colorectal cancer.1,2 Despite evidence for deregulation of Src kinases at all stages of tumor progression,3 c-Src has been found to function mostly late in cancer progression, such as during invasion and metastasis.4 Therefore, whether and how Src kinases affect the earliest stages of tumorous cell cycle and growth control is interesting to clarify.2,3 In this context, the Drosophila system has proven to be useful for mapping dedicated Src functions in cell proliferation control underlying tumor growth.5, 6, 7, 8, 9, 10, 11 Mammals contain nine Src family kinases, which show a degree of functional redundancy,12 whereas Drosophila has only two c-Src homologs, Src42a and Src64b.7

Here, we study the role of Src kinases in cell cycle control in the adult Drosophila intestinal stem cell (ISC) lineage,13,14 which is a promising model for tumor onset in regeneratively growing tissues. The adult midgut grows homeostatically and regenerates by stem cell activation following tissue damage.15, 16, 17, 18, 19 Control of stem cell function during tissue feedback control in the fly midgut has emerged as a useful model to understand tumor onset.20, 21, 22 We use this system to study the two Src kinases in pathophysiological context during the very initial phases of tumor onset.

Mammalian Src kinase activation has been shown to be important for platelet-derived growth factor-induced Myc activation and hence for G1–S progression,23 as well as for an elusive step controlling G2–M cell cycle progression.24 Mammalian Src kinases are physiologically activated by mitogenic Pdgf, Egf and Csf25 or integrin signaling, and are believed to function upstream of cyclin E/A-cyclin-dependent kinase 2 (CycE/A-Cdk2) activation in these cases.26,27 Src kinases can also activate Janus kinase/signal transducers and activators of transcription (Jak-Stat) signaling in mammals and flies.6,28,29 This is particularly interesting, as Egfr and Stat signaling control ISC division rates and differentiation of ISC progeny in the Drosophila intestine.16,17,19,30,31

The contribution of pathological Src function to tumorous tissue overgrowth is more complex and has been studied in diverse contexts. Src kinases can induce context- and dosage-dependent stimulation of proliferation or apoptosis in different systems,32 whereas low levels of Src activity can be antiapoptotic.9 When cell death is blocked11 or Src is coexpressed with other oncogenes like Ras in Drosophila imaginal discs, malignant tissue overgrowth occurs.9 Also, in Drosophila imaginal discs, Src64b uses JNK signaling and stabilizes F-actin fibers to stimulate the transcription factor Yorkie, an effector of the Hippo signaling pathway, to control proliferation.10,11 As Ras, Yki and JNK signaling are all known to regulate proliferation of adult Drosophila ISCs, we set out to investigate how, downstream of Src kinase function, core cell cycle regulation might influence ISC behavior during the onset of tumor growth in the adult fly intestine (midgut).

In the intestine, differentiated absorptive cells are short-lived and, when damaged or exhausted, delaminate from the intestinal epithelium to be replaced by the progeny of ISCs.15,18,19,33 Drosophila ISCs typically divide asymmetrically to self-renew and produce an enteroblast (EB) daughter cell. EB fate determination has been suggested to be controlled, at least in part, by integrin-dependent asymmetric segregation of the PAR3/6/aPKC (atypical protein kinase C) complex to the EB, which may contribute to Notch activation in EBs by a yet undefined mechanism.34 In contrast to the mammalian intestine, which has many transit-amplifying stem cell progeny, Drosophila’s EBs typically differentiate directly without intervening transit-amplifying (TA) cell divisions. This is due to rapid Notch activation in EBs by Delta signaling from ISCs, which promotes EB-to-enterocyte (EC) differentiatation, typically without intervening divisions.13,14,22,35 As terminal cell differentiation is often coupled to permanent exit from the cell cycle into G1, we reasoned that not only cell cycle regulation in ISCs but also control of exit from the mitotic cycle and, possibly, exit from pluripotency in EBs might be important for tumor onset. Indeed, both unrestrained self-renewal of ISCs and dedifferentiation of differentiated cells like EBs may lead to unrestrained tumorigenic proliferation.36

Mitotic cell cycle progression is due to oscillatory activation of cyclin/Cdk complexes, which is driven by oscillations in cell cycle gene expression and proteolysis.37 E2F transcription factors regulate expression not only of cyclins, Cdks and the the Cdc25-type phosphatase, String (Stg), but also of many other genes required for DNA replication and mitosis.38 In Drosophila, G1 CycE–Cdk2 activity is essential for the S-phase entry. For G2/M progression, the CycA/Cdk1 and CycB/Cdk1 complexes must become activated by the Stg, which removes inhibitory phosphates from Cdk1. For mitotic exit and G1 re-entry, the anaphase-promoting complex (APC/C), an E3 ubiquitin ligase, is required for the degradation of mitotic proteins. Cdh1/Fizzy-related (Fzr) activates the APC/C from the end of mitosis through the G1 phase and, thereby, suppresses CDK activity.39 A comprehensive study of cell cycle properties and their regulation in Drosophila’s ISC lineage has not yet been reported.

In Drosophila imaginal discs, differentiation-associated mitotic exit involves concurrent inhibition of G1 cyclin/Cdk activity and repression of E2F activity.40 Thus, after exit, activating CycE alone is not sufficient to trigger the S-phase progression. This is due to ongoing E2F1 degradation,41 as well as due to depletion of E2F1-dependent target genes and independent genes critical for the S-phase completion.42 In this situation, also E2F1 activation alone is not sufficient to bypass cell cycle exit, as APC/C activity rises during differentiation and inhibits accumulation of mitotic cyclins and Stg.42 How mitotic exit is executed in EBs is still unknown. At least during differentiation to ECs, EBs obviously engage in mitotic-to-endocycle (ME) switching, and are expected to retain E2F, CycE and APC/C activities, but also the ME switch in EBs remains to be mechanistically explored.

We show here that Src42a and Src64b are required for the normal division capacity of Drosophila intestinal progenitor cells, and that Src42a, Src64b and another non-receptor kinase, the activated Cdc42-associated kinase (Ack),43, 44, 45, 46 cause progenitor overproliferation in the intestine via core cell cycle activation. Interestingly, we find that these kinases drive overgrowth not only by accelerating ISC cell cycle progression but also by expanding the pool of a previously unidentified cell type that is division capable like ISCs, but weakly positive for reception of Notch signaling, as in postmitotic committed EBs. A similar weakly committed TA cell type was also found in the wild type, suggesting more plasticity in Drosophila’s intestinal progenitor cell pool than previously believed.


Src levels control progenitor division capacity

To identify possible regulators of mitotic division capacity in intestinal progenitors, we conditionally overexpressed several oncogenes by escargot-Gal4-mediated transactivation in esg+ progenitor cells and measured proliferation rates. We found that the expression of a constitutively activated Src42a allele (Src42a.CA), or of wild-type Src64b, triggered mitosis in esg+ cells (Figure 1a and Supplementary Figure S1A) and increased the ratio of esg+ progenitor cells to the total number of intestinal cells (Figure 1a′). We then used fluorescence-activated cell sorting (FACS) analysis to determine how quiescent wild-type progenitors were induced to divide at the level of cell cycle control (Figure 1b). Wild-type esg+ progenitor cells showed low S-phase levels, consistent with low or absent tissue replacement in healthy animals. Infection with the enteric bacterial pathogen Pseudomonas entomophila (P.e.) triggered G1/S progression and enhanced endoreplicative S-phase progression in postmitotic esg+ progenitor cells (Figure 1b). Similarly, Src42a.CA or Src64b expression in uninfected esg+ progenitors stimulated G1/S as well as G2/M progression and endoreplicative S-phase progression (Figure 1c). The latter was consistent with increased nuclear sizes of esg+ cells compared with wild-type controls (Supplementary Figures S1A and C). As Src64b expression conferred lower proliferative rates than Src42a.CA, we tested whether cell cycle stimulation in the ISC lineage was Src kinase dosage-sensitive or whether there was a difference between Src64b and Src42a functionality in the gut. We expressed different transgenic alleles of Src64b or Src42a, whose strength in promoting overproliferation or cell death had been previously reported to range from moderate to strong in eye imaginal disc cells.7 Overexpression of moderate Src42a alleles in esg+ progenitors accelerated cell division rates more robustly than strong or constitutively active Src64b alleles (Supplementary Figure S1B), indicating a more fundamental difference between Src42a and Src64b. Additionally, expression of Src64b in esg+ cells increased Stg mRNA levels less than did Src42a.CA or the Ack non-receptor kinase (Supplementary Figure S1B′). On the other hand, Src64b expression resulted in higher levels of rhomboid mRNA, an Egfr signaling factor that has previously47 been associated with gut stress and tissue damage (Supplementary Figure S1B′). To define the underlying cause for these differences, we tested whether Src64b promoted cell death or elimination. As we could not detect activated caspase-3 in the Drosophila intestine (data not shown), we used an alternative assay to measure effects on cell elimination. We found that the expression of Src64b in differentiated ECs (using the Myo1a-Gal4 driver) causes a strong reduction in gut epithelial size (Supplementary Figure S1B′′). As epithelial shrinkage has been linked to anoikis of ECs,48 we suggest that Src64b induces significant rates of EC delamination from the intestine. Thus, Src64b-mediated cell loss may effectively counteract its proproliferative function in progenitor cells. This interpretation was supported by FACS analysis: we found that after tissue dissociation, the population of Src64b-expressing esg+ cells included 10.3% propidium iodide-positive dead esgGFP+ cells, as opposed to 2.99% dead cells in esg+ Src42a.CA-expressing cells, and only 1.34% in wild-type controls (data not shown).

Figure 1

Src42a.CA and Src64b control progenitor division capacity and are required for regeneration. (a) Stimulation of mitotic rates in Drosophila midguts following overexpression of constitutively activated Src42a (Src42a.CA) or wild-type Src64b in esg+ progenitors for 4 or 7 days. Numbers of mitotic histone H3-S10-phosphorylated nuclei (pH3+) per midgut were quantified (a); n=10 midguts each. (a′) Src42a.CA or Src64b expression for 7 days amplifies the ratio of esg+ progenitor cells to the total number of intestinal cell pool 2.2-fold from 24.3% (s.d. ±10.6) in wild-type to 52.4% (s.d. ±11.8) following Src42a expression and to 55.6% (s.d. ±14.9) following Src64b expression. Numbers of total cells and esg+ cells per region of interest (ROI=100 mm2) are plotted; n=5 ROIs each. (b) Cell cycle profiles of quiescent (black trace) and mitotically activated wild-type esg+ progenitor cells after 1 day of P.e. infection (red trace). FACS analysis of DNA content of esg+ cells (x axis). Black and red horizontal dotted lines indicate S-phase levels in uninfected and P.e.-infected wild-type esg+ progenitors, respectively. (c) FACS analysis of DNA content following Src42a.CA or Src64b expression in esg+ progenitors for 7 days. Percentages of cells in G1, S+G2 or endoreplicated states (E, >4C) are indicated. Red horizontal dotted lines indicate S-phase levels in Src-expressing esg+ progenitors. (d and e) Depletion of Src42a or of Src64b by transgenic RNAi (IR) in esg+ cells for 7 days (d) but not depletion of Src kinases in ECs (e) impairs regenerative ISC division capacity upon P.e. infection; n=7–11 (d), n=5–8 (e) midguts. (f) Combined double RNAi of Src42a and Src64b for 7 days impairs subsequent regenerative ISC divison upon P.e. infection for 24 h; n=12–17 midguts. (g) Combined double RNAi of Src42a and Src64b for 7 days impairs accumulation of esgGFP+ progenitors upon P.e. infection for 24 h. (h) Schematic model for esg+ progenitor cell pool sized control by Drosophila Src kinases. Src kinases are required for the division of esg+ adult intestinal progenitor cells during tissue regeneration (left arrow, wild-type Src levels). Permanently increased Src42a or Src64b levels lead to hyperplastic amplification of proliferative esg+ progenitor cells (right arrow). Reducing Src42a and Src64b levels impairs ISC divison and endoreplication of EBs. Correspondingly, inferred Src42a or Src64b levels are indicated on top (from quiescent wild type (left) to pathologically enriched (right)).

Next, we tested the endogenous function of Src kinases by depleting Src42a or Src64b using transgenic RNA interference (RNAi) expressed in esg+ progenitor cells. Src42a or Src64b kinase depletion in healthy, noninfected esg+ progenitors did not cause loss of stem cell nests within 1 week, but did impair ISC division capacity during the acute tissue regeneration that results from oral P.e. infection (Figure 1d) (nonparametric two-tailed Mann–Whitney test: P=0.0014 (Src42a RNAi), P=0.0008 (Src64b RNAi)). Inhibition of regenerative ISC divisions was not observed when Src64b or Src42a were depleted from differentiated enterocytes during infection (Figure 1e). Knocking down Src42a and Src64b simultaneously in esg+ cells for 1 week suppressed, but did not completely abolish regenerative divisions following 1 day of P.e. infection (P=0.0038) (Figure 1f). These infected Src42a+Src64b-codepleted intestines displayed a reduction in the number of esg+-labeled progenitors (Figure 1g). To investigate whether Src42a and Src64b have functions not only during damage-induced acute regeneration, we knocked down Src42a and Src64b simultaneously for 2 weeks in healthy, noninfected animals (Supplementary Figure S2A). After 2 weeks at 29 °C, the restrictive temperature for Gal80ts-controlled transgene expression, numbers of esg+ cells were increased in the wild type, consistent with earlier reports of ageing-associated stress and stress-mediated proliferation.18,49 The combined Src42a+Src64b knockdown blocked this ageing/stress-induced amplification of the esg+ cell pool, but did not deplete all esg+ cells or Delta+ ISCs cells or induce premature differentiation of ISCs to ECs (Supplementary Figure S2A). Next, we used FACS analysis to test whether loss of ISC division capacity was due to cell cycle arrest. Src42a-depleted esg+ cells displayed impaired G1/S progression of diploid as well as of endoreplicative cells during tissue regeneration (Supplementary Figure S2B) and increased numbers of early endoreplicating esg+ cells (4–8C) at the expense of diploid (2C) cells (Supplementary Figure S2B). Permanent clonal marking with the esgts Gal4 flipout (esgFO) TARGET system19 showed, however, that Src42a-depleted cells did not prematurely differentiate to form mature endoreplicated ECs (Supplementary Figure S2C). Moreover, as Src42a-depleted ISCs were also still Dl+ and negative for two different Notch activity reporters, and thus undifferentiated (Supplementary Figures S2C and C′), our data indicate that Src42a depletion causes a cell cycle arrest and not loss of stemness. We also measured Src42a and Src64b mRNA abundance in FACS-sorted esg+ progenitors or Dl+ ISCs or Gbe-Su(H)+ EBs in wild-type conditions or after P.e. infection. We did not see a specific upregulation of Src kinases transcripts after P.e. infection (Supplementary Table 1) and also did not observe strong upregulation of levels of active phosphorylated Src42a (pSrc42aY40050)50 after P.e. infection (Supplementary Figure S2D). From these results, we conclude that intestinal esg+ progenitor cell proliferation requires Src kinase function and is positively determined by Src kinase dosage (model Figure 1h).

How do Src kinases control progenitor cell proliferation? Src may directly control cell cycle regulators, indirectly use receptor tyrosine kinsase-RAS-mitogen-activated protein kinase (MAPK) signaling,23,27,51,52 or crossactivate other proproliferative pathways. For example, mammalian Src kinases have been implicated in direct activation of STAT3 by phosphorylation53 as well as indirectly in growth factor receptor kinase-mediated STAT3/5 activation.54 In contrast, in the Drosophila intestine esg+ progenitor growth can trigger unpaired3 (Upd3) cytokine secretion in adjacent enterocytes,55 and Upd3 ligands noncell autonomously feedback and activate the single Jak-Stat receptor in flies, Domeless, in esg+ progenitors to convey cell proliferation.19 For this reason, we hypothesized that Src kinase-mediated Stat92E activation might, atypically, depend also on Domeless. Therefore, we tested whether Drosophila Src kinase-driven esg+ progenitor overproliferation depended on Egfr, Erk, Stat92E or Domeless.

Epistasis tests indicated that Src42a as well as Src64b required Egfr signaling and Stat92E, the transcriptional effector of Jak-Stat signaling, to stimulate esg+ cell division (Figures 2a and a′). However, Src-dependent esg+ progenitor cell overproliferation was only partially blocked by Mkp3, an ERK inhibitor,56 and was not blocked by dominant-negative domeless receptor mutants57 that lack the cytoplasmic signaling domain, domeΔCyt (P=0.2222) or domeΔTMCyt (P=0.5185) (Figures 2a and b). We deduce that Src can induce proliferation via Jak-Stat signaling, and that this may proceed by crossactivation via Egfr as in mammals, or in parallel to MAPK signaling, at the level of the Jak kinase, or Stat92E itself.

Figure 2

Cell cycle control of intestinal progenitor pool size. (a and b) Src kinase-dependent progenitor overproliferation depends on Egfr, MAPK (Erk) and Stat92E. Stat92E RNAi (a) or also Egfr RNAi (a′) block mitotic rates induced by (MAE)Src64b (a) or wild-type Src64b (a′) overexpression in esg+ cells. Expression of dominant-negative domeless (unpaired (upd) receptor) alleles lacking the cytoplasmic, or both the cytoplasmic and transmembrane domains (a), or of Mkp3 (a′) do not inhibit, or only partially impair Src64b-dependent overprolifertion. (b) Src42a.CA-dependent esg+ cell overproliferation also depends on Stat92E, and Src42a43–4-dependent overproliferation depends on Egfr receptor function. A heterozyogous Egfr mutant (egfrCO) does not block Src42a433-dependent overproliferation (b′). n=16–17 (a, Stat92E RNAi), n=5–9 (a, Dome alleles), n=6–10 (a′), n=7–10 (b) and n=8–11 midguts (b′). (c) Increased Jak-Stat signaling following Src42A.CA activation using esgts or esgFO-dependent expression for 7 days. Fold increase of mRNA abundance of the transcriptional Stat92E target gene socs36e by RNA deep sequencing compared with esgts or esgFO controls respectively. (d and d′) Coexpression of Stg and CycE is sufficient to elicit division of Delta+ ISCs (d) and of Gbe-Su(H)+ Notch+ EBs (d′). Also, coexpression of E2f1, Dp, CycE and Cdk2 induces mitoses; n=7–10 midguts (d and d′). (e) Cluster formation of Gbe-Su(H)+ EBs and mitotic figures in Notch+ EBs following Stg/CycE coexpression. (e′) Zoom-in of mitoses occurring in two Gbe-Su(H)+ EBs. (f) Visualization of Gbe-Su(H)-LacZ (Notch reporter) levels following Stg+CycE coexpression in esg+ cells. Asymmetric fate determination between ISC (white arrow) and the Notch+ ISC daughter cell (control, top panel). Stg+cycE coexpression causes esg+ clusters to form. Only one Notch-negative cell persists (ISC, white arrow), and ISC daughter cells are Notch+. The latest newborn EB shows very faint lacZ levels (dashed arrow).

To demonstrate whether Src kinases indeed activate Stat signaling or also MAPK-ERK signaling, or both, for stimulating ISC division, we measured reporters for pathway activity. The phosphorylated active form of extracellular signal-regulated kinase ERK (pERK) is an established marker for ERK MAPK signaling activity, and pERK staining in the wild-type intestinal epithelium is observed in undifferentiated progenitors and is absent in differentiated esg polyploid enterocytes (Supplementary Figure S1C). esgFO Gal4-mediated clonal expression of Src42a.CA or of Src64b in esg+ progenitors and their progeny amplified the number of progenitor cells, and many but not all of these accumulating cells were pERK+ (Supplementary Figure S1C). Next, focusing on these esgFO-labeled cells, we quantified nuclear pERK levels per cell. Despite the amplification of the numbers of pERK+ cells, the mean anti-pERK fluorescence intensity levels per individual esg+ cell was not increased, but slightly decreased following Src42a.CA or Src64b expression (Supplementary Figure S3A). This indicated that Src kinase-dependent ERK signaling is essential for proliferation, but might not be the causative determinant for esg+ progenitor cell overproliferation. To quantify Stat signaling strength, we quantified mRNA levels of Socs36E, a direct transcriptional target gene of Stat92E. Socs36E mRNA abundance was increased after Src42a.CA expression (Figure 2c). Next, we inspected Stat92E protein levels using an Stat92E-GFP BAC reporter,29 where Stat92E is expressed from its endogenous promoter. In controls, Stat92E is expressed in esg+ progenitors, but is absent in differentiated esg polyploid enterocytes (Supplementary Figure S3A′). Following Src42a.CA expression in esg+ cells, the number of progenitor cells was increased, and many of these accumulating cells expressed Stat92E. Closer inspection revealed that both surplus esg+ diploid progenitors as well as surplus esg+ endoreplicating cells were Stat92E+ (Supplementary Figure S3A′). Taken together, we conclude that Src kinase function is likely to be generally required for ISC proliferation under ageing- as well as pathogen-induced tissue turnover. Additionally, our data indicate a stimulatory role of Src kinases for the Jak-Stat signaling pathway to promote cell cycle progression in esg+ progenitors. Activation of Stat92E appears necessary for Src to induce mitosis in quiescent esg+ progenitors.

A core cell cycle control module regulates intestinal progenitor pool size

Based on Src kinases-dependent esg+ cell overproliferation, we reasoned that downstream factors controlling esg+ progenitor pool size might be identified as common targets in mRNA deep sequencing expression profiles. Esg+ progenitors have also been shown to overproliferate following transgenic Notch inhibition,13,14,22 as well after Ack kinase expression (Supplementary Figure S1B′). We found 236 transcripts to be commonly upregulated (cutoff twofold induction) after Src42a.CA and Ack activation, including 30 functional G1/S and G2/M cell cycle regulatory factors defined previously by genome-wide screens in different Drosophila systems38,58 (Supplementary Table 2). Twenty-nine of these factors were also induced following esgts-mediated Notch depletion. We identified the E2F1 target and developmentally regulated G2/M regulator Stg, the fly Cdc25 phosphatase,59 as the top enriched cell cycle regulator after esgFO-mediated Src42a.CA as well as Ack expression.

Based on the identified candidate regulators, we set out to identify functionally the minimally required set of core cell cycle regulators that were sufficient to stimulate division of quiescent ISCs, or to delay mitotic cell cycle exit in wild-type EBs and elicit mitotic division. For this, we expressed different transgenic combinations of cyclins, Cdks and the cell cycle regulatory E2f1/Dp transcription factor in ISCs or, separately, in EBs. We found that a combination of E2f1/Dp+CycE/Cdk2 coexpression induced high levels of cell division (mitosis) in ISCs (Figure 2d) (P=0.0004) and EBs (Figure 2d′ and Supplementary Figure S3B) (P=0.0051). As E2f1 can activate the Stg promoter by direct binding in cultured Drosophila cells and imaginal discs,38 we reasoned that E2f1 might also drive genes required for cell cycle progression in intestinal progenitors, including the G2/M cell cycle activating Stg phosphatase. Consistent with this notion, we found that the combination of dl-Gal4-mediated Stg and CycE coexpression in ISCs was sufficient to induce sustained cell division (Figure 2d) (P=0.0005). Surprisingly, Stg+CycE expression also triggered division of Notch+ (Gbe-Su(H)+) EBs (P=0.0005) (Figure 2d′). Stg alone (P=0.4408), or CycE alone (P=0.8171), was not sufficient to drive prolonged overproliferation in EBs (Figure 2d′). To address whether E2F1 activity alone was also capable of overcoming G1 and G2 arrests in esg+ progenitors, we pulse-expressed E2F1/Dp in ISCs or in EBs for one day (Supplementary Figure S3B). E2F1/Dp were capable of inducing the division of ISCs to 10.2 (±7.7) mitoses per midgut (P=0.0004) and less robustly in EBs (3.6 (±3.81) mitoses per midgut (P=0.0286)) compared with 0.22 (±0.44) mitoses per midgut in control animals expressing GFP in Gbe-Su(H)+ cells. Compared with Stg+CycE expression in ISCs or EBs, E2F1/Dp in ISCs or EBs exerted only 59.0% or 47.8% of the proproliferative effects, respectively.

Stg+CycE coexpression thus represents a core proproliferative module causing prolonged esg+ proliferation and can cause EB clusters to form (Figures 2e and e′). The finding that Stg+CycE could induce division in EBs was surprising, because only E2f1/Dp+CycE/Cdk2, but not Stg+CyE, had previously been shown to be capable of bypassing cell cycle exit in differentiating cells in other Drosophila organs.40 As conditional expression of E2f1/Dp+CycE/Cdk2 for one day induced higher division rates than Stg+CycE (Supplementary Figure S3B), and as Stg is a described E2F1 target,38,60,61 E2F1 targets other than Stg must be limiting, at least initially, in quiescent ISCs and EBs.

Progenitor pool regulation via cell cycle control

Athough mutants in cell cycle regulators can impair the asymmetric cell division machinery,62 activation of the core cell cycle regulators studied here are not known to exhibit such effects.62 However, they have not yet been tested in adult fly stem cell lineages in this role. Hence, we tested whether Stg or Stg and CycE coexpression might affect ISC stem cell and EB identities. Cell cycle activation might ‘neutrally’ duplicate an existing cell state (e.g. ISC division creating two ISCs, or EB division creating two EBs), without directly controlling fate determination. Alternatively, Stg or Stg+CycE-mediated cell cycle activation might more directly impair differentiation and thereby confer stem cell self-renewal, as cell cycle dynamics can impact differentiation dynamics in vertebrate embryonic stem cells.63,64 To test these possibilities, we clonally coexpressed Stg and CycE (Figure 2f) or, separately, E2F1/Dp (Supplementary Figure S3C′). Despite increased division rate and cluster formation, only one truly Notch-negative cell (the ISC) was maintained in each clone, while most other labeled ISC progeny showed Notch activation as assessed using the Gbe-Su(H)-LacZ transcriptional reporter (Figure 2f and Supplementary Figure S3C′). Nevertheless, the newborn EBs showed only very low LacZ levels (Figure 2f and Supplementary Figure S3C′) and in parts of the midgut cell clusters with faint Notch reporter activity arose (Supplementary Figure S3C). In addition, we occasionally found small cell clusters without detectable Notch activity (Supplementary Figure S3C). We investigated this phenomenon and observed that, especially in middle midgut regions (e.g. R365) that normally have low endogenous tissue turnover and little Delta expression, progenitor cell clusters expressing Stg+CycE contain initially only immature and undifferentiated diploid cells (Figures 3a and b). Dynamic analysis revealed that Delta, an established ISC stem cell marker, was weakly detectable in these progenitor clusters, while known differentiation features were absent (Pdm1, Pros and p4E-BP, a dTOR target under Notch control in EBs66) (Figures 3a and b and Supplementary Figure S3C). In addition, prolonged, permanent esgFO-mediated expression of Stg alone was able to produce cell clusters that did not show acquisition of mature EC identity as judged from lack of endoreplication (Figure 3c). FACS analysis revealed that this phenomenon was reflected by accumulation of progenitors in G1 and suppression of endoreplication onset (Figure 3d). As these clusters grew, over time differentiation was eventually observed in these Stg-expressing cell clusters, indicative of a transient delay and not absolute block in EC maturation (Figure 3e). Interestingly, Stg-expressing Dl+N cells (bona fide ISCs, white arrows) were found to coexist next to Dl+N+ cells (likely recently born or immature preEBs, red arrows) and DlN+ cells (mature EBs, cyan arrow) within single progenitor clusters. At later time points, endoreplicated Pdm1+ ECs also developed in the Stg-expressing progenitor cells clusters (data not shown). Owing to eventual differentiation, we think that although cell cycle activation accumulated young diploid and only weakly Notch+ progenitors, this was likely not due to a switch to symmetric ISC amplification. The presence of Dl+N+ cells also indicates the unexpected existence of transitory cell states intermediate between ISCs and EBs. Moreover, this indicates that Delta is possibly not a completely accurate stem cell marker, and poses the question whether these or similarly weakly specified EB-like cells divide in addition to ISCs. For technical reasons and owing to the low levels of Delta staining in the initial clusters (Figures 3b and c), we were not able to quantify convincingly cells with triple costaining for LacZ and Delta and pH3. In principle, such a test could address whether both Dl+N+ EBs, and also the presumably more mature DlN+ EBs, are capable of division.

Figure 3

Amplification of weakly specified intestinal progenitors. (a) esgFO>-mediated Stg+CycE coexpression for 4 days causes transient amplification of Prospero- and Pdm1-negative undifferentiated progenitors. Prospero marks mature differentiated EEs and Pdm1 marks mature differentiated ECs (arrows). Single EEs and ECs inside the GFP+ clone likely represent cells already determined before onset of transgene expression. (b) Lower p4E-BP labeling defines the Delta+ ISC (white arrow, bottom panel), whereas EBs show relatively higher p4E-BP levels (cyan arrows). Shown is a wild-type ISC nest (yellow dashed line) containing one ISC and two EBs (top panels). esgFO-mediated Stg expression for 1 week is sufficient to cause amplification of undifferentiated progenitors in a stem cell nest (yellow outline). Overproliferating undifferentiated Stg-expressing esg+ progenitors are weakly Delta+ and show reduced p4E-BP levels compared with neighboring differentiated cells in the tissue (bottom panels). (c) Wild-type GFP-marked ISC progeny differentiated to GFP-marked endoreplicated ECs within 14 days of clone growth, whereas Stg expression transiently impairs EC endoreplication (dashed arrow indicating diploid immature cells). (d) Inhibition of onset of endoreplication after 14 days of Stg expression. esgFO-mediated Stg-expressing ISCs and their progeny accumulate with 2C DNA content, whereas wild-type ISC progeny develops endoreplicating cells during normal differentiation to ECs. FACS analysis. (e) Stg expression does not permanently inhibit progenitor differentiation. After 14 days, differentiation onset can be detected in some clones, indicated by the activation of p4E-BP labeling in diploid cells. Two representative bona fide Dl+ ISCs (white arrows) and one mature EBs (cyan arrow) are indicated. Note three Dl+p4E-BP+ double-positive EB-like states (red arrows) that are not typically seen in the ISC lineage during baseline tissue homeostasis.

Next, we investigated whether this novel intermediate progenitor cell type was amplified during Src42a-, Src64b- or Ack-driven progenitor overproliferation. To address whether Src kinase- or Ack-driven amplification of Dl+ cell was due to stimulation of symmetric ISC division, we tested whether the esg+ cell pool remained permanently increased following pulses and subsequent deactivation of ectopic Src kinases or Ack expression. If ISCs were amplified by stimulation of symmetric division, pool size (ISC number) should remain increased. Conversely, if transient Src kinases or Ack pulses caused intermediate progenitors (EB-like cells) to increase, progenitor cell amplification should be reversible, as non-stem cells should be depleted by differentiation. Therefore, we conditionally expressed Src42a.CA, Src64b or Ack in progenitors for 1 week, resulting in strong esg+ progenitor amplification (Figure 1a′). We subsequently extinguished further transgene expression for up to 18 days (Figure 4). By 18 days at the restrictive temperature for Src42a.CA expression, a near-wild-type epithelium with distinct small, separated esg+ marked stem cell nests developed. In the case of Src42a.CA expression, this resulted in only a slight elevation in esg+ cell numbers (P=0.0200), and in the case of Src64b, there was no significant difference from controls (P=0.2683) (Figure 4a). Total cell numbers were not increased in either case (Src42a.CA (P=0.7857), Src64b (0.1429); Figure 4a′). Secondary re-expression of Src42a.CA for 10 h in esg+ progenitors that had previously overproliferated caused only a weak increase in division rates as compared with Src42a.CA expression in naive esg+ progenitors (Figure 4a′′). Secondary Src64b re-expression effected slightly decreased division rates compared with initial transgene expression in naive esg+ progenitors (Figure 4a′′). Taken together, these results further indicate that the final pool of esg+ progenitors after an earlier Src kinase expression pulse was not larger or much more proliferative than the initial esg+ progenitor pool ruling out that Src expression amplified the number of permanently self-renewing ISCs.

Figure 4

Src- and Ack-driven progenitor cell amplification is reversible. (a and b) Transient esg+ progenitor overproliferation upon Src42a.CA, Src64b (a) or Ack expression (b) in esg+ progenitors is reversible. (a) Numbers of esg+ progenitors in regions of interest (ROI) in the posterior midgut. Transgenic Src kinases expression in esg+ cells was conditionally induced for 7 days, and repressed for the next 18 days. The terminal expression pulse for 10 h served to induce efficient GFP labelling to allow reliable quantification of esgGFP+ cell numbers. ROI=100 μm2; n=5 ROIs. See Figures 1a and a′ for numbers of pH3+, esg+ and total cells per ROI after initial Src kinase expression for 7 days. (a′) Total numbers of all cell types in ROI. (aa′) ROI=100 μm2. (a′′) Following transient induction and subsequent repression of transgenic Src kinases expression, the mitotic division rate in midguts is not significantly increased. (b) Representative images revealing esg+ progenitor pool amplification upon conditional transgenic Ack expression for 7 days, and progenitor pool retraction after inhibition of further Ack expression for the next 18 days. After expansion and retraction, progenitors could be stimulated to divide and amplify a second time when Ack was reinduced for 3 days (rightmost panel). (c) Schematic representation of Ack activity in (b). Red=wild-type Ack levels, orange=weak Ack overexpression levels and red=strong Ack overexpression levels. Ack levels in (c) explain the experiment above (b).

Similarly, pulsed Ack expression amplified esg+ progenitor numbers that covered the epithelial surface, but these excess esg+ cells receded to normal numbers (Figure 4b) after ectopic Ack expression was silenced (Figure 4c). As with Src42a and Src64b, these esg+ cells we competent to resume division following a secondary pulse of Ack expression (Figure 4b, rightmost panel). We infer that transient Drosophila Ack or Src activation does not permanently transform progenitor cells or exhaust their proliferative potential, and that the surplus esg+ cells generated in these experiments were likely not autonomously self-renewing ISCs but, rather, intermediate EB-like cells. Consistent with this interpretation, mRNA expression profiling indicated that genes that were differentially induced after esg-mediated Ack overexpression were more frequently EB-specific than ISC-specific (Supplementary Figure S4A). In summary, our data indicate that while Src-expressing, Ack-expressing or Stg+CycE-expressing esg+ cells that are amplified are mitotically active and certainly include hyperproliferating ISCs, many cells in this amplified esg+ cell pool resembled an intermediate EB-like cell.

To investigate effects on ICS lineage asymmetry, we assayed the symmetry of the PAR3/6/aPKC complex in dividing ISCs. We found that a Pon (Partner of Numb)::GFP protein reporter, which reports Par-complex-dependent cell polarity,67 was asymmetrically partitioned in 80.8% of dividing wild-type ISCs in metaphase, and in 80.0% of dividing wild-type ISCs in telophase (Figure 5a). The fraction of asymmetric divisions was not decreased following Ack expression: 85.3% of Ack-expressing ISCs in metaphase and 82.6% in telophase showed asymmetric Pon-GFP distribution (Figure 5a′). To resolve the apparent incongruence that many Dl+ ISC-like cells accumulated after Ack activation, but that Dl+ ISCs nevertheless divided mostly asymmetrically, we stained the Ack-expressing progenitors for phospho-4E-BP, a dTOR target that is under Notch control in EBs, and normally mirrors Gbe-Su(H)-lacZ expression.66 Phospho-4E-BP levels were decreased in wild-type diploid basally located Dl+ ISCs compared with Dl differentiating and more apically located ISC progeny (Figures 3b and 5b), corroborating a previous report.66 We found, however, that after Ack expression, not all Dl+ cells showed low phospho-4E-BP levels, as would have been expected for true ISCs (Figure 5b′, white arrows). This suggested inhomogeneity in the Dl+ cell pool and the unexpected onset of differentiation in some Dl+ cells (Figure 5b′, red arrows).

Figure 5

Diminution of esg+ progenitor pool size by cell differentiation and delamination. (a) Wild-type ISCs divide primarily asymmetrically. Asymmetric cortical Pon-GFP protein localization in metaphase wild-type ISCs (green arrow marks the pole of the future ISC (basally localized), white arrowhead marks the future EB (more apically localized)). We determined that Pon-GPF was segregated to the future ISC and not the future EB. This was determined, because the divisions of ISCs are oriented in the epithelium, as has been described also previously.22 Therefore, the future Pon-GFP retaining ISC nucleus is positioned more basally and closer to the basal membrane than the nucleus of the daughter EB. (a′) Ack-expressing ISCs divide also primarily asymmetrically. Ack was expressed for 2 days. Progressive mitotic phases are indicated from left to right; n=34 (Ack expressing ISCs in metaphase), n=23 (Ack expressing ISCs in telophase), 16/73 (21.9%) of all observed mitoses could not be classified as either symmetric or asymmetric owing to the lack of Pon-GFP signal or a dotted cytoplasmic signal. n=26 (wild-type metaphases), n=20 (wild-type telophases), 9/55 (16.4%) or all wild-type mitoses could not be classified. (a′′) Scheme describing asymmetric segregation of Pon-GFP (green) to the basal cortical pole of the dividing ISC. Visceral muscles in gray, ISC in green and EB in orange. (a′′′) Quantification of symmetrical and asymmetrical mode of ISC division based on cortical Pon-GFP distribution in metaphase or telophase. (b) In the wild-type, Delta+ ISCs display lower abundance of p4E-BP levels (white arrow) than sourrounding differentiated cell progeny (cyan arrows). (b′) Ack expression as well as integrin αPS3+4 coinhibition or aPKC inhibition for 7 days leads to the accumulation of weakly specified Delta+/ p4E-BP+ cells (red arrows). Bona fide ISCs are Delta+/p4E-BP (white arrows). (c) Scheme describing high activity of bantam microRNA in ISC (marked by destablization of the bantam microRNA GFP reporter) and low bantam activity in adjacent EB (marked by stabilization of GFP microRNA GFP reporter) (top panel). ISC in light green and EB in dark green. Bottom panels: Ack expression does not only accumulate bona fide ISCs (Delta+ bantam sensor negative, white arrows) but also leads to the accumulation of Delta+ cells with EB-type low bantam microRNA activity (high bantam sensor levels, red arrows). See Supplementary Figure S3B for bantam sensor levels in wild-type ISCs and EBs. Ack was induced for 7 days. (d) Ack and integrin αPS3+4 RNAi hairpin coexpression in esg+ cells for 7 days causes formation of surplus diploid esg+ progenitors. When further transgenic Ack expression was subsequently extinguished for 10 days, these esg+ cells differentiated to ECs (note esgGFP+ labelling in polyploid big EC nuclei originating from diploid esgGFP+ progenitors) and were lost by delamination from the apical epithelial surface facing the intestinal lumen (arrows). As judged by confocal optical sectioning, nuclei of these polyploid esg+ cells were positioned at apical sites in the pseudostratified intestinal epithelium facing the intestinal lumen, and not at basal sides, where wild-type diploid progenitors typically reside (scheme in left panel). The position of the optical section taken for the immunostaining is shown. Transient Ack expression amplifies esg+ progenitors (basally localized). After extinguishing transgenic Ack expression, these surplus esg+ progenitors differentiate by endoreplication (bigger nuclei) and apical delamination (arrow, dashed outline of delaminating cells).

Hence, we tested the idea that not all Dl+ cells were true stem cells. Loss of integrin-mediated Par polarity has been shown to cause symmetric stem cell division.34 We found that following coinhibition of integrins αPS3 and 4, or following aPKC inhibition, indeed esg+ cells with increased Delta levels were produced. However, not all of these were Delta+ p4E-BP diploid ISCs (Figure 5b′). Instead, differentiating Dl+ p4E-BP+ cells were observed (Figure 5b′, red arrows). These data indicate that inhibition of integrins and aPKC amplifies a Delta+ Notch+ cell type, and hence not always necessarily duplicates stem cells.

In searching for another assay for the existence of weakly specified Dl+ EBs, we found using a transgenic bantam microRNA reporter that the bantam microRNA68 has higher activity in ISCs and lower activity in neighboring wild-type EBs (Supplementary Figure S4B and Figure 5c, schematic presentation). Ack expression caused not only amplification of bona fide Dl+ ISCs with low bantam sensor levels (high bantam activity) (Figure 5c and Supplementary Figure S4B) but also amplification of Dl+ cells with low bantam activity and high GFP sensor levels. This again indicated incipient differentiation in some Dl+ cells. In further tests, we coexpressed Ack and RNAi’s directed against integrins αPS3 and αPS4 as a possible way to obtain amplify bona fide ISCs. However, this did not cause ISC amplification. Instead, after pulsed transgene expression had ceased, esg+ cells eventually differentiated, as inferred from the onset of endoreplication, and delaminated from the epithelium in an esg+ state (Figure 5d). In summary, like the Src kinases or cell cycle activators, Ack expression, integrin inhibition or mutation of Par polarity stimulated ISC division, and also caused accumulation of weakly specified intermediate Dl+ cells that were neither self-renewing ISCs nor mature EBs. In all these cases, we suggest that these cells exist in a stabilized immature EB state.

A small pool of immature EBs divides mitotically in the wild-type Drosophila intestine. APC/C regulates the ME switch in EBs

Are these findings relevant to wild-type midgut homeostasis? Although previous reports have indicated that the Drosophila midgut does not contain TA cells and that ISCs are the sole dividing cell type,13,14 we reinvestigated whether EBs are division capable. We found that the population of mitotically dividing intestinal cells contained a small subset (4.5%) of Gbe-Su(H)+ EBs (Figure 6a). No dividing EBs were found in quiescent noninfected intestines. Unfortunately, we were not able to determine whether dividing Notch+ EBs were positive or negative for endocytic Delta foci, leaving open the question whether these dividing EBs were possibly newborn, less mature ISC progeny. To address this question by an alternative assay, we assayed the cell cycle states of wild-type EBs and ISCs by FACS analysis. We isolated Dl-Gal4-GFP+ or Gbe-Su(H)-Gal4-GFP+69 cells from dissociated adult intestines (Figure 6b). Dl+ ISCs resided primarily in G2 (4C) (79.9% (s.d. ±9.8)), less in G1 (2C) (17.36% (s.d. ±9.18)), and also infrequently in endoreplicating phases (>4C) (2.09% (s.d. ±1.75)). Pulsed GFP expression for 1 day revealed that Gbe-Su(H)+ EBs were primarily endoreplicated (>4C, 63.13% (s.d. ±9.14)), and less in the G2 phase (30% (s.d. ±8.23)) or in G1 (3.9% (s.d. ±3.8)). Thus, in contrast to G0-type quiescence of many vertebrate stem cells, wild-type ISCs in healthy midguts rest predominantly in G2, possibly poised for fast division during future tissue regeneration. On the other hand, while endoreplication is usually incompatible with mitotic division,70 a significant percentage of wild-type EBs were in cell cycle states (G1+G2), theoretically compatible with future mitotic division. A small proportion of EBs (3.8%) resided in G1, which hypothetically could be the source of the EBs that divided during regeneration (4.5%).

Figure 6

Some EBs have mitotic division capacity. (a) A small pool of wild-type EBs displays mitotic division capacity upon oral enteric P.e. infection for 1 day; 4.5% of all regenerative divisions (n=448 mitotic nuclei from 10 midguts) occur in Notch+ Gbe-Su(H)+ EBs. (b) Quiescent wild-type Delta (Dl)+ stem cells are diploid and reside primarily in the G2 phase. EBs are mostly in G2 and endoreplicative phases, but less in the G1 phase. FACS analysis of DNA content (x axis) of wild-type Dl+ ISCs (red) and Notch (GbeSu(H)+) EBs (blue). Arrows indicate differences between ISCs and EBs. E, DNA endoreplication; n=5 biological FACS replicates (dl-Gal4) and n=3 biological FACS replicates (Gbe-Su(H)-Gal4) using 10–12 midguts for each replicate. (c) Stg phosphatase expression is not inhibited in young differentiating ISC daughter cells. Stg expression reported by a GFP protein trap and seen in all esg+ progenitor cells residing within a stem cell nest (white arrows). Note that following bacterial oral infection with P.e. for 18 h, the number of progenitor cells per stem cell nest increases, all of which are Stg+ (bottom panels). (d) Conditional Fzr overexpression for 7 days causes increased endoreplication in esg+ progenitors causing formation of big nuclei with increased size. Compared with the nuclear size of esg+ cells in wild type (arrows). (e) Conditional Rca1 overexpression inhibits the ME switch in the EB->EC lineage owing to a G2 cell cycle arrest (blue trace). (f) Fzr expression increases the DNA amount per cell in endoreplicating esg+ cells, whereas Rca1 expression impairs endoreplication capacity in esg+ cells compared with wild-type esg+ progenitors. DNA amount was determined from single confocal sections of all esgGFP+ cells present in random ROIs in five posterior midguts as a product of mean 4′,6-diamidino-2-phenylindole (DAPI) fluorescence intensity and nuclear area (y axis, log scale). Each data point represents one nucleus (x axis). n=221 (wild type), n=47 (fzr) and n=61 (rca1 nuclei). Fzr and Rca1 were expressed for 7 days. Scheme below visualizing ME switch during maturation of the esg+ EB cell type. Endoreplication onset after mitotic cell cycle exit in mature EBs. (g) Fzr overexpression for 7 days rescues intestines from tumorous overproliferation following Notch inhibition; n=7–8 midguts. (h) Mitotic and endoreplicative cell cycle states are not fully mutually exclusive in the ISC lineage. Mitotic chromosome condensation can be observed not only in diploid cells (8 condensed chromosomes) following conditional Ack or Src64b expression in esg+ progenitors for 7 days but also in endoreplicated nuclei (containing >8 chromosomes) following Src64b expression or Fzr expression under regenerative conditions.

What could the molecular basis for the observed division potential of some EBs be? Using protein-trap reporter lines, 42,71 we found that wild-type EBs expressed the mitotic activators Stg and CycB3 (Figure 6c and Supplementary Figure S5A). These and other mitotic regulators are usually degraded by APCFzr and further transcriptionally inhibited by repressive E2F2 upon differentiation-associated mitotic cell cycle exit in other cell types.70,72 Thus, we tested the idea that execution of mitotic cell cycle exit during ME switching in mature EBs was controlled by APC/C levels. When we expressed Fzr, a coactivator for APCFzr/Cdh1, we observed strongly increased EC endoreplication (Figures 6d and f and Supplementary Figure S5C). Conversely, overexpression of Rca1, an APCFzr/Cdh1 inhibitor,72 imposed a G2 cell cycle block and impaired endocycling and maturation of EBs to Pdm1+ ECs (Figures 6e and f and Supplementary Figure S5B). Moreover, Fzr overexpression impaired the tumor-like esg+ progenitor amplification in Notch mutants (Figure 6g and Supplementary Figures S5C and C′). This supported the important role of ME switching in progenitor pool size control. Finally, we observed histone-H3-S10 phosphorylation, a mitotic marker, not only in diploid wild-type and Ack-expressing cells, as expected, but observed that Fzr and also Src42a.CA expression led to pH3 staining in clearly endoreplicated nuclei (Figure 6h). This observation corroborates the existence of an EB stage or EB subtype that has not fully exited the mitotic cycle, while already partially engaging in the endocycle. Collectively, these results show that APC/C activity influences the ME transition during EB maturation. These results also suggest that APC/C activity is unexpectedly low in EBs, a condition that might explain why mitotic and endoreplicative features are not yet fully disentangled in immature EBs and consequently why these EBs might enter mitosis.


This work reveals that exit from the mitotic cell cycle is dynamically regulated in the ISC lineage. Besides the previously described and unrelated symmetric ISC amplification,34,73,74 dynamic progenitor pool amplification may accelerate the regenerative response to physiological stress by providing more differentiation-capable cells. Our study defines Src kinases as upstream regulators of Stat signaling. Thereby, unrestrained Src activity over-rides quiescence in ISCs as well as mitotic exit in ISC progeny during the earliest steps of hyperplastic progenitor overproliferation.

Src kinases function during the earliest steps of esg+ progenitor cell overproliferation

Our data indicate that Drosophila Src kinases activate Stat signaling to drive esg+ progenitor cell proliferation. This might occur at the level of the Jak kinase or Stat92E, or proceed via a function of Src in Egfr-mediated Stat92E activation, as in mammals.54,75 Hypothetically, Src kinases might also function in the Egfr pathway and derepress Jak-Stat signaling by stimulating bantam microRNA-dependent repression of the Jak-Stat repressor Soc36E.76,77 It is thus also possible that in the absence of Src kinases, Stat92E signaling might not be sufficiently active to sustain ISC proliferation. This is consistent with reports that for Stat92E activation, Src42a and Src64b are required to concentrate Stat92E in the apical membrane of embryonic ectodermal cells, in close proximity to the domeless receptor.29 JAK kinase-independent STAT3/5 activation by c-SRC or Abl kinases has also been reported in mammals, and STAT3 is additionally required for SRC-dependent transformation.78, 79, 80 Direct Stat92E targets have not yet been described in the Drosophila ISC lineage. Whether Drosophila Src kinase signaling in the intestine also engages other transcriptional regulators like the Hippo pathway effector Yki10,81 or the Egfr effector Pointed82 will have to be investigated. This is interesting because these factors can act as direct transcriptional activators of cell cycle regulators like CycE81 and Stg.82 While Src42a and Src64b are required for ISC proliferation, the Src kinases may not be essential for ISC viability. This could, however, also be attributed to partial function of the RNAi’s we used to deplete them. Hence, an essential function for cell survival in the gut remains to be formally addressed using conditional genetic knockouts. Nevertheless, we propose that the duality of proliferative and proapoptotic functions inherent to the Src kinase family has been co-opted for control of tissue homeostasis in the gut. Allowing dynamic tuning in a limited bandwidth, Src kinases may serve as one rheostat for esg+ progenitor pool size control.

Underlying this function, we show that while Ack or Src activation triggered early events in tumor-like proliferation, transient expression of these kinases did not permanently transform ISCs or EBs. Given that Ack expression triggers mitosis mostly in diploid cells, and that Ack also causes significant accumulation of bona fide ISCs (Delta+ p4E-BP or Delta+ bantam sensor GFP), Ack likely maintains esg+ progenitors in a more naive state than the Src kinases. While asymmetric segregation of Notch signaling to EBs defines the ISC state, in contrast, Yki, Egfr, Stat92E and Src kinases signaling are prominent in ISCs, but their signaling intensities in EBs are not weaker than in ISCs, at least on cytological levels. As bantam is more active in ISCs than EBs, and since it has been shown to confer slight proliferative advantages in the midgut83 and, in other contexts, can be activated by Yki or Egfr83, 84, 85 or Wnt signaling while being inhibited by Notch,86 bantam+ progenitors might be more proliferative than bantam progenitors. Bantam is, however, likely not solely causative for selective division activity of self-renewing ISCs as opposed to EBs, because it is also not required for overproliferation of Ras-driven or Notch-mutant tumors in the intestine.83

Irrespective of what actively determines the self-renewing ISC state, our data indicate that during physiological progenitor division, such as during regeneration or cell cycle activation by Src family kinases, intermediate stem cell daughters divide. This cell type may be undetermined and favoured in the course of tumor initiation.

In contrast to Drosophila, although human c-SRC has also only weak transforming activity,87 a transient pulse of human c-SRC expression is sufficient to transform permanently immortalized human breast epithelial cells to mammospheres that contain cancer stem cells.88 In this case, the c-SRC pulse established an autocrine, self-maintaining proproliferative inflammatory interleukin-6/Jak-Stat signaling circuit88 via reciprocal interleukin-6/STAT3 and nuclear factor-κB activation by microRNAs.89 In the Drosophila ISC lineage, Upd3/Jak-Stat signaling acts between cells rather than autocrinely: enterocytes express the Upd3 cytokine, whereas Jak-Stat activation occurs in Upd3-negative stem cells. Therefore, tissue turnover and replacement of spent Upd3-secreting ECs by healthy newborn Upd3-negative ECs would interrupt a self-maintaining proproliferative circuit.

Mitotic cell cycle exit during differentiation of stem cell daughters

Differentiation onset in Drosophila wing and eye cells is causal for cell cycle exit and establishes parallel, independent restraints on E2F1 and cyclin–Cdk activities.40 Thereby, neither CycE nor E2F1 nor combined CycE and Stg activation could bypass cell cycle exit.40 In contrast, in undifferentiated progenitor cells, CycE–Cdk2 activity could increase E2F1 activity via phosphorylation of retinoblastoma proteins (Rbf, Rbf2), and E2F1 could induce CycE and Cdk2, effectively establishing a positive feedforward loop.40 As CycE and Stg coexpression activated cell divisions in EBs in the Drosophila intestine, EBs might be considered undifferentiated in terms of their cell cycle state: the E2F1-CycE feedforward loop90 seems to be intact. How is cell division possible in a cell type like an EB, which has already received the Notch signal, which is thought to lock-in differentiation and cell cycle exit?22,91 This may be because G2/M regulators are not depleted from younger EBs. Thus, EBs are committed and not differentiated, and cell cycle exit has not been completely locked-in (Figure 6c and Supplementary Figure S5A) as compared with other postmitotic cells types such as differentiated imaginal disc cells.42 In these differentiated cells, depletion of mitotic factors is often due to APC/C92-mediated degradation, which reduces G2–M cyclin–Cdks and Stg.42,92 Our observations indicate that APC/C activity is low in esg+ cells including EBs, opening a short window of opportunity for mitotic division. Indeed, these cells rest predominantly in G2, a phase of low APC/C activity. Alternatively, it may be that the lock-in of Notch-dependent transcriptional programs conferring differentiation takes some time, creating a temporal window during which young EBs are susceptible to mitotic stimuli.

We also describe that core mitotic cell cycle activation does not impair the Notch activation process per se in young EBs. Nevertheless, subsequent EB→EC maturation was functionally impaired. As decreasing Cdk1 activity has been shown to be essential for ME progression70,93 and as Stg can activate Cdk1,94 it is conceivable that Stg activation blocked endoreplication onset. Thereby, a delay in acquisition of terminal differentiated cell fate would be a secondary consequence of delays in endocycle initiation. In such a model, we suggest that endoreplication onset may serve as a timer for maturation of EBs into ECs.

Rudimentary TA cells in the Drosophila midgut

Although the fly midgut has been thought not to contain TA cells, the mammalian intestine contains a well-described TA cell population. In mice and humans, intestinal adenomas are thought to be able to initiate from bona fide stem cells, and from TA cells, and even from differentiated postmitotic epithelial cells following dedifferentiation.95 In healthy flies, in agreement with previous studies, we did not detect divisions of EBs, the committed progenitors likely to be most similar to mammalian TA cells. However, we did observe mitotic divisions in a small pool of such Notch+ EBs when the intestine was induced to regenerate, following infection (Figures 6a and b). The identification of this TA cell population defines Drosophila as tumor model relevant to the mammalian intestine, and sets aside a seemingly fundamental difference between fly and mammalian intestinal homeostasis.

Importantly, previous clonal MARCM (mosaic analysis with a repressible cell marker) analysis during baseline tissue homeostasis13,14 did not reveal such a cell population. We speculate that the pool of dividing Notch+ intermediate progenitors was too small or transient to be identified by MARCM analysis. Corroborating our findings, earlier reports do, however, show that following enteric infection or DSS-induced tissue damage15 both the number of Delta+ cells (defined as ISCs) and the number of Notch+ Gbe-Su(H)-lacZ EBs increases. Of note, one study reported that a fraction of dividing cells (15%) were not Delta+,19 and thus not likely to be ISCs. When we overexpressed either the Src or Ack non-receptor tyrosine kinases, we observed a transient, reversible expansion of a population of diploid, mitotic, weakly Delta+, weakly Notch+ cells. Based on their similar characteristics, we suggest that these cells are functionally similar to the division-capable EBs that divide during active regeneration in the wildtype. Moreover, congruent to our observations, increased neoplastic risk in aged flies has been ascribed to accumulation of atypical Delta Notch double-positive progenitor cells.18 Thus, cells with TA characteristics can be found in many conditions in the fly gut. This is clearly a dynamic cell population, however, and it nearly disappears during periods of low gut epithelial turnover.

Our study suggests that Src family kinase and Ack-dependent tissue overgrowth increases the number of esg+ cells. These derive from accelerated ISC division but include division-capable specified but immature EB-like cells—the TA-like cells discussed above—that expand in number during the course of enhanced ISC division. Similarly, in the central nervous system of the fly, brain tumors created by mutating brat or numb arise preferentially in a small TA lineage of the PAN neuroblasts,96 specifically from a weakly specified immature progenitor cell type (Ase).

In summary, the expansion of intermediate weakly specified progenitor cells in the Drosophila intestine may constitute a novel mechanism of tumor onset. The expansion of these TA-like cells appears to occur independently of stem cell expansion or dedifferentiation of mature gut epithelial cells. We suggest that similar mechanisms may apply during regeneration and tumorigenesis in other stem cell-supported tissues in humans.

Materials and methods

Temperature shift experiments

Females were used for all experiments. Genetic crosses using the Gal4 TARGET system (esgts, Myo1ats, Dlts, Gbe-Su(H)ts and esgFO) were set up and raised at 18 °C until adulthood without a controlled day–night cycle. One- to two-week-old animals were shifted to 29 °C for the indicated amount of time. For MARCM clones, animals were heat shocked for 45 min 1 week after eclosion.

Bacterial infection

Gut infections were performed by oral infection with P.e. in 5% sucrose/phosphate-buffered saline (PBS) on Whatman filter paper for 18 h at 29 °C except for Figures 1f and g and Supplementary Figure S2D where P.e. was provided with yeast paste.

FACS analysis

Ten to twelve midguts were dissected in PBS (w/o Mg2+, Ca2+), and cut before dissociation in PBS containing 5 mg/ml collagenase (Sigma-Aldrich, Schnelldorf, Germany; Blend type H) at 29 °C on a shaker (600 r.p.m.) for 1 h. Then, 1/10 vol 10 × trypsin were added for further dissociation for 1 h/29 °C aided by manual trituration. Before analysis, cell suspensions were filtered (40 μm cell strainer). Dead cells were excluded from analysis by propidium iodide staining. Samples were analyzed on a FACS Canto II using the DIVA software (BD Biosciences, Heidelberg, Germany). DNA profiles were generated using the FlowJo Software (Miltenyi Biotec, Bergish Gladbach, Germany).


Female adult flies were dissected in PBS. The gastrointestinal tract was removed and fixed in PBS with 4% paraformaldehyde for 30 min. Samples were washed in 0.1% Triton X-100 (PBST), permeabilized for 30 min in 0.3% PBST and preblocked for 1 h in 10% normal goat serum before incubation with primary antibody overnight at 4 °C. Samples were incubated with secondary antibody for 3–4 h at room temperature. Finally, samples were washed in PBST and mounted in mounting medium containing 4′,6-diamidino-2-phenylindole (Vectashield; Vector Laboratories, LINARIS, Dossenheim, Germany). EdU labeling was performed using the Click-it chemistry (Invitrogen, Darmstadt, Germany) after ex vivo labeling for 1 h in 10 μM EdU/Ringer’s solution.

Fluorescence images were acquired on a Leica TCS SP5 II confocal microscope (Leica Mikrosysteme, Mannheim, Germany). Images were processed using Rasband WS, ImageJ (National Institutes of Health, Bethesda, MD, USA), Fiji97 and Adobe photoshop (Adobe Systems, München, Germany).

Determination of DNA amount

To quantify the nuclear DNA content of individual nuclei, the mean intensity of Hoechst fluorescence was measured manually with the Fiji software (ImageJ) from a single confocal section at the widest diameter of a given nucleus as described.98

Quantification and statistical analysis

For determining the numbers of pH3+ cells per midgut, total cells or esg+ cells per region of interest (ROI), data from 5–10 samples were quantified per condition (see figure legends and text for exact numbers). For representative immunostainings without quantification, >10 adult posterior midguts were analyzed, derived as progeny from mass crosses of 10 females and 5 male parents. As we notice that proliferation rate of ISCs in wild-type animals is not always normally distributed,99 data range was presented as boxplot, and statistical significance was calculated by nonparametric two-tailed Whitney–Mann testing at 95% confidence intervals. Data presentation uses Bioconductor R’s default boxplot function showing median, interquartile range and outliers with whiskers representing minimum and maximum. Terminally sick animals (after prolonged tumor induction or P.e. infection) that were unable to move were excluded from the study. No specific statistical method was used to choose sample size to determine adequate power to detect a prespecified effect size. No method of randomization into experimental groups was used. Investigators were not blinded when assessing the outcome of animal studies.

For technical details of RNA sequencing and for annotation of antisera and Drosophila lines used in the study see Supplementary Information.


  1. 1

    Bolen JB, Veillette A, Schwartz AM, DeSeau V, Rosen N . Activation of pp60c-src protein kinase activity in human colon carcinoma. Proc Natl Acad Sci USA 1987; 84: 2251–2255.

    CAS  Google Scholar 

  2. 2

    Yeatman TJ . A renaissance for SRC. Nat Rev Cancer 2004; 4: 470–480.

    CAS  Google Scholar 

  3. 3

    Cartwright CA, Meisler AI, Eckhart W . Activation of the pp60c-src protein kinase is an early event in colonic carcinogenesis. Proc Natl Acad Sci USA 1990; 87: 558–562.

    CAS  Google Scholar 

  4. 4

    Brunton VG, Ozanne BW, Paraskeva C, Frame MC . A role for epidermal growth factor receptor, c-Src and focal adhesion kinase in an in vitro model for the progression of colon cancer. Oncogene 1997; 14: 283–293.

    CAS  Google Scholar 

  5. 5

    Stewart RA, Li DM, Huang H, Xu T . A genetic screen for modifiers of the lats tumor suppressor gene identifies C-terminal Src kinase as a regulator of cell proliferation in Drosophila. Oncogene 2003; 22: 6436–6444.

    CAS  Google Scholar 

  6. 6

    Read RD, Bach EA, Cagan RL . Drosophila C-terminal Src kinase negatively regulates organ growth and cell proliferation through inhibition of the Src, Jun N-terminal kinase, and STAT pathways. Mol Cell Biol 2004; 24: 6676–6689.

    CAS  Google Scholar 

  7. 7

    Pedraza LG, Stewart RA, Li DM, Xu T . Drosophila Src-family kinases function with Csk to regulate cell proliferation and apoptosis. Oncogene 2004; 23: 4754–4762.

    CAS  Google Scholar 

  8. 8

    Vidal M, Larson DE, Cagan RL . Csk-deficient boundary cells are eliminated from normal Drosophila epithelia by exclusion, migration, and apoptosis. Dev Cell 2006; 10: 33–44.

    CAS  Google Scholar 

  9. 9

    Vidal M, Warner S, Read R, Cagan RL . Differing Src signaling levels have distinct outcomes in Drosophila. Cancer Res 2007; 67: 10278–10285.

    CAS  Google Scholar 

  10. 10

    Enomoto M, Igaki T . Src controls tumorigenesis via JNK-dependent regulation of the Hippo pathway in Drosophila. EMBO Rep 2013; 14: 65–72.

    CAS  Google Scholar 

  11. 11

    Fernandez BG, Jezowska B, Janody F . Drosophila actin-capping protein limits JNK activation by the Src proto-oncogene. Oncogene 2013; 33: 2027–39.

    Google Scholar 

  12. 12

    Klinghoffer RA, Sachsenmaier C, Cooper JA, Soriano P . Src family kinases are required for integrin but not PDGFR signal transduction. EMBO J 1999; 18: 2459–2471.

    CAS  Google Scholar 

  13. 13

    Ohlstein B, Spradling A . The adult Drosophila posterior midgut is maintained by pluripotent stem cells. Nature 2006; 439: 470–474.

    CAS  Google Scholar 

  14. 14

    Micchelli CA, Perrimon N . Evidence that stem cells reside in the adult Drosophila midgut epithelium. Nature 2006; 439: 475–479.

    CAS  Google Scholar 

  15. 15

    Amcheslavsky A, Jiang J, Ip YT . Tissue damage-induced intestinal stem cell division in Drosophila. Cell Stem Cell 2009; 4: 49–61.

    CAS  Google Scholar 

  16. 16

    Buchon N, Broderick NA, Chakrabarti S, Lemaitre B . Invasive and indigenous microbiota impact intestinal stem cell activity through multiple pathways in Drosophila. Genes Dev 2009; 23: 2333–2344.

    CAS  Google Scholar 

  17. 17

    Cronin SJ, Nehme NT, Limmer S, Liegeois S, Pospisilik JA, Schramek D et al. Genome-wide RNAi screen identifies genes involved in intestinal pathogenic bacterial infection. Science 2009; 325: 340–343.

    CAS  Google Scholar 

  18. 18

    Biteau B, Hochmuth CE, Jasper H . JNK activity in somatic stem cells causes loss of tissue homeostasis in the aging Drosophila gut. Cell Stem Cell 2008; 3: 442–455.

    CAS  Google Scholar 

  19. 19

    Jiang H, Patel PH, Kohlmaier A, Grenley MO, McEwen DG, Edgar BA . Cytokine/Jak/Stat signaling mediates regeneration and homeostasis in the Drosophila midgut. Cell 2009; 137: 1343–1355.

    Google Scholar 

  20. 20

    Lee WC, Beebe K, Sudmeier L, Micchelli CA . Adenomatous polyposis coli regulates Drosophila intestinal stem cell proliferation. Development 2009; 136: 2255–2264.

    CAS  Google Scholar 

  21. 21

    Apidianakis Y, Pitsouli C, Perrimon N, Rahme L . Synergy between bacterial infection and genetic predisposition in intestinal dysplasia. Proc Natl Acad Sci USA 2009; 106: 20883–20888.

    CAS  Google Scholar 

  22. 22

    Ohlstein B, Spradling A . Multipotent Drosophila intestinal stem cells specify daughter cell fates by differential notch signaling. Science 2007; 315: 988–992.

    CAS  Google Scholar 

  23. 23

    Barone MV, Courtneidge SA . Myc but not Fos rescue of PDGF signalling block caused by kinase-inactive Src. Nature 1995; 378: 509–512.

    CAS  Google Scholar 

  24. 24

    Roche S, Fumagalli S, Courtneidge SA . Requirement for Src family protein tyrosine kinases in G2 for fibroblast cell division. Science 1995; 269: 1567–1569.

    CAS  Google Scholar 

  25. 25

    Abram CL, Courtneidge SA . Src family tyrosine kinases and growth factor signaling. Exp Cell Res 2000; 254: 1–13.

    CAS  Google Scholar 

  26. 26

    Prathapam T, Tegen S, Oskarsson T, Trumpp A, Martin GS . Activated Src abrogates the Myc requirement for the G0/G1 transition but not for the G1/S transition. Proc Natl Acad Sci USA 2006; 103: 2695–2700.

    CAS  Google Scholar 

  27. 27

    Chu I, Sun J, Arnaout A, Kahn H, Hanna W, Narod S et al. P27 phosphorylation by Src regulates inhibition of cyclin E–Cdk2. Cell 2007; 128: 281–294.

    CAS  Google Scholar 

  28. 28

    Ishizawar R, Parsons SJ . c-Src and cooperating partners in human cancer. Cancer Cell 2004; 6: 209–214.

    CAS  Google Scholar 

  29. 29

    Sotillos S, Krahn M, Espinosa-Vazquez JM, Hombria JC . Src kinases mediate the interaction of the apical determinant Bazooka/PAR3 with STAT92E and increase signalling efficiency in Drosophila ectodermal cells. Development 2013; 140: 1507–1516.

    CAS  Google Scholar 

  30. 30

    Beebe K, Lee WC, Micchelli CA . JAK/STAT signaling coordinates stem cell proliferation and multilineage differentiation in the Drosophila intestinal stem cell lineage. Dev Biol 2010; 338: 28–37.

    CAS  Google Scholar 

  31. 31

    Jiang H, Grenley MO, Bravo MJ, Blumhagen RZ, Edgar BA . EGFR/Ras/MAPK signaling mediates adult midgut epithelial homeostasis and regeneration in Drosophila. Cell Stem Cell 2011; 8: 84–95.

    CAS  Google Scholar 

  32. 32

    Thomas SM, Brugge JS . Cellular functions regulated by Src family kinases. Annu Rev Cell Dev Biol 1997; 13: 513–609.

    CAS  Google Scholar 

  33. 33

    Buchon N, Broderick NA, Poidevin M, Pradervand S, Lemaitre B . Drosophila intestinal response to bacterial infection: activation of host defense and stem cell proliferation. Cell Host Microbe 2009; 5: 200–211.

    CAS  Google Scholar 

  34. 34

    Goulas S, Conder R, Knoblich JA . The Par complex and integrins direct asymmetric cell division in adult intestinal stem cells. Cell Stem Cell 2012; 11: 529–540.

    CAS  Google Scholar 

  35. 35

    Perdigoto CN, Schweisguth F, Bardin AJ . Distinct levels of Notch activity for commitment and terminal differentiation of stem cells in the adult fly intestine. Development 2011; 138: 4585–4595.

    CAS  Google Scholar 

  36. 36

    Knoblich JA . Asymmetric cell division: recent developments and their implications for tumour biology. Nat Rev Mol Cell Biol 2010; 11: 849–860.

    CAS  Google Scholar 

  37. 37

    Nasmyth K . A prize for proliferation. Cell 2001; 107: 689–701.

    CAS  Google Scholar 

  38. 38

    Dimova DK, Stevaux O, Frolov MV, Dyson NJ . Cell cycle-dependent and cell cycle-independent control of transcription by the Drosophila E2F/RB pathway. Genes Dev 2003; 17: 2308–2320.

    CAS  Google Scholar 

  39. 39

    Pesin JA, Orr-Weaver TL . Regulation of APC/C activators in mitosis and meiosis. Annu Rev Cell Dev Biol 2008; 24: 475–499.

    CAS  Google Scholar 

  40. 40

    Buttitta LA, Katzaroff AJ, Perez CL, de la Cruz A, Edgar BA . A double-assurance mechanism controls cell cycle exit upon terminal differentiation in Drosophila. Dev Cell 2007; 12: 631–643.

    CAS  Google Scholar 

  41. 41

    Shibutani ST, de la Cruz AF, Tran V, Turbyfill WJ 3rd, Reis T, Edgar BA et al. Intrinsic negative cell cycle regulation provided by PIP box- and Cul4Cdt2-mediated destruction of E2f1 during S phase. Dev Cell 2008; 15: 890–900.

    CAS  Google Scholar 

  42. 42

    Buttitta LA, Katzaroff AJ, Edgar BA . A robust cell cycle control mechanism limits E2F-induced proliferation of terminally differentiated cells in vivo. J Cell Biol 2010; 189: 981–996.

    CAS  Google Scholar 

  43. 43

    Prieto-Echague V, Miller WT . Regulation of Ack-family nonreceptor tyrosine kinases. J Signal Transd 2011; 2011: 742372.

    Google Scholar 

  44. 44

    Sem KP, Zahedi B, Tan I, Deak M, Lim L, Harden N . ACK family tyrosine kinase activity is a component of Dcdc42 signaling during dorsal closure in Drosophila melanogaster. Mol Cell Biol 2002; 22: 3685–3697.

    CAS  Google Scholar 

  45. 45

    Schoenherr JA, Drennan JM, Martinez JS, Chikka MR, Hall MC, Chang HC et al. Drosophila activated Cdc42 kinase has an anti-apoptotic function. PLoS Genet 2012; 8: e1002725.

    CAS  Google Scholar 

  46. 46

    Abdallah AM, Zhou X, Kim C, Shah KK, Hogden C, Schoenherr JA et al. Activated Cdc42 kinase regulates Dock localization in male germ cells during Drosophila spermatogenesis. Dev Biol 2013; 378: 141–153.

    CAS  Google Scholar 

  47. 47

    Jiang H, Edgar BA . EGFR signaling regulates the proliferation of Drosophila adult midgut progenitors. Development 2009; 136: 483–493.

    CAS  Google Scholar 

  48. 48

    Buchon N, Broderick NA, Kuraishi T, Lemaitre B . Drosophila EGFR pathway coordinates stem cell proliferation and gut remodeling following infection. BMC Biol 2010; 8: 152.

    CAS  Google Scholar 

  49. 49

    Choi NH, Kim JG, Yang DJ, Kim YS, Yoo MA . Age-related changes in Drosophila midgut are associated with PVF2, a PDGF/VEGF-like growth factor. Aging Cell 2008; 7: 318–334.

    CAS  Google Scholar 

  50. 50

    Shindo M, Wada H, Kaido M, Tateno M, Aigaki T, Tsuda L et al. Dual function of Src in the maintenance of adherens junctions during tracheal epithelial morphogenesis. Development 2008; 135: 1355–1364.

    CAS  Google Scholar 

  51. 51

    Maa MC, Leu TH, McCarley DJ, Schatzman RC, Parsons SJ . Potentiation of epidermal growth factor receptor-mediated oncogenesis by c-Src: implications for the etiology of multiple human cancers. Proc Natl Acad Sci USA 1995; 92: 6981–6985.

    CAS  Google Scholar 

  52. 52

    Smith MR, DeGudicibus SJ, Stacey DW . Requirement for c-ras proteins during viral oncogene transformation. Nature 1986; 320: 540–543.

    CAS  Google Scholar 

  53. 53

    Cao X, Tay A, Guy GR, Tan YH . Activation and association of Stat3 with Src in v-Src-transformed cell lines. Mol Cell Biol 1996; 16: 1595–1603.

    CAS  Google Scholar 

  54. 54

    Olayioye MA, Beuvink I, Horsch K, Daly JM, Hynes NE . ErbB receptor-induced activation of stat transcription factors is mediated by Src tyrosine kinases. J Biol Chem 1999; 274: 17209–17218.

    CAS  Google Scholar 

  55. 55

    Cordero JB, Stefanatos RK, Myant K, Vidal M, Sansom OJ . Non-autonomous crosstalk between the Jak/Stat and Egfr pathways mediates Apc1-driven intestinal stem cell hyperplasia in the Drosophila adult midgut. Development 2012; 139: 4524–4535.

    CAS  Google Scholar 

  56. 56

    Kim M, Cha GH, Kim S, Lee JH, Park J, Koh H et al. MKP-3 has essential roles as a negative regulator of the Ras/mitogen-activated protein kinase pathway during Drosophila development. Mol Cell Biol 2004; 24: 573–583.

    CAS  Google Scholar 

  57. 57

    Brown S, Hu N, Hombria JC . Identification of the first invertebrate interleukin JAK/STAT receptor, the Drosophila gene domeless. Curr Biol 2001; 11: 1700–1705.

    CAS  Google Scholar 

  58. 58

    Guest ST, Yu J, Liu D, Hines JA, Kashat MA, Finley RL Jr . A protein network-guided screen for cell cycle regulators in Drosophila. BMC Syst Biol 2011; 5: 65.

    CAS  Google Scholar 

  59. 59

    Edgar BA, Lehman DA, O'Farrell PH . Transcriptional regulation of string (cdc25): a link between developmental programming and the cell cycle. Development 1994; 120: 3131–3143.

    CAS  Google Scholar 

  60. 60

    Neufeld TP, de la Cruz AF, Johnston LA, Edgar BA . Coordination of growth and cell division in the Drosophila wing. Cell 1998; 93: 1183–1193.

    CAS  Google Scholar 

  61. 61

    Reis T, Edgar BA . Negative regulation of dE2F1 by cyclin-dependent kinases controls cell cycle timing. Cell 2004; 117: 253–264.

    CAS  Google Scholar 

  62. 62

    Prokopenko SN, Chia W . When timing is everything: role of cell cycle regulation in asymmetric division. Semin Cell Dev Biol 2005; 16: 423–437.

    CAS  Google Scholar 

  63. 63

    Orford KW, Scadden DT . Deconstructing stem cell self-renewal: genetic insights into cell-cycle regulation. Nat Rev Genet 2008; 9: 115–128.

    CAS  Google Scholar 

  64. 64

    Kohlmaier A, Edgar BA . Proliferative control in Drosophila stem cells. Curr Opin Cell Biol 2008; 20: 699–706.

    CAS  Google Scholar 

  65. 65

    Buchon N, Osman D, David FP, Fang HY, Boquete JP, Deplancke B et al. Morphological and molecular characterization of adult midgut compartmentalization in Drosophila. Cell Rep 2013; 3: 1725–1738.

    CAS  Google Scholar 

  66. 66

    Kapuria S, Karpac J, Biteau B, Hwangbo D, Jasper H . Notch-mediated suppression of TSC2 expression regulates cell differentiation in the Drosophila intestinal stem cell lineage. PLoS Genet 2012; 8: e1003045.

    CAS  Google Scholar 

  67. 67

    Wang H, Ouyang Y, Somers WG, Chia W, Lu B . Polo inhibits progenitor self-renewal and regulates Numb asymmetry by phosphorylating Pon. Nature 2007; 449: 96–100.

    CAS  Google Scholar 

  68. 68

    Brennecke J, Hipfner DR, Stark A, Russell RB, Cohen SM . Bantam encodes a developmentally regulated microRNA that controls cell proliferation and regulates the proapoptotic gene hid in Drosophila. Cell 2003; 113: 25–36.

    CAS  Google Scholar 

  69. 69

    Zeng X, Chauhan C, Hou SX . Characterization of midgut stem cell- and enteroblast-specific Gal4 lines in Drosophila. Genesis 2010; 48: 607–611.

    CAS  Google Scholar 

  70. 70

    Zielke N, Edgar BA, DePamphilis ML . Endoreplication. Cold Spring Harb Perspect Biol 2013; 5: a012948.

    Google Scholar 

  71. 71

    Chau J, Kulnane LS, Salz HK . Sex-lethal facilitates the transition from germline stem cell to committed daughter cell in the Drosophila ovary. Genetics 2009; 182: 121–132.

    CAS  Google Scholar 

  72. 72

    Zielke N, Querings S, Rottig C, Lehner C, Sprenger F . The anaphase-promoting complex/cyclosome (APC/C) is required for rereplication control in endoreplication cycles. Genes Dev 2008; 22: 1690–1703.

    CAS  Google Scholar 

  73. 73

    de Navascues J, Perdigoto CN, Bian Y, Schneider MH, Bardin AJ, Martinez-Arias A et al. Drosophila midgut homeostasis involves neutral competition between symmetrically dividing intestinal stem cells. EMBO J 2012; 31: 2473–2485.

    CAS  Google Scholar 

  74. 74

    O'Brien LE, Soliman SS, Li X, Bilder D . Altered modes of stem cell division drive adaptive intestinal growth. Cell 2011; 147: 603–614.

    CAS  Google Scholar 

  75. 75

    Ren Z, Schaefer TS . ErbB-2 activates Stat3 alpha in a Src- and JAK2-dependent manner. J Biol Chem 2002; 277: 38486–38493.

    CAS  Google Scholar 

  76. 76

    Quesnelle KM, Boehm AL, Grandis JR . STAT-mediated EGFR signaling in cancer. J Cell Biochem 2007; 102: 311–319.

    CAS  Google Scholar 

  77. 77

    Herranz H, Hong X, Hung NT, Voorhoeve PM, Cohen SM . Oncogenic cooperation between SOCS family proteins and EGFR identified using a Drosophila epithelial transformation model. Genes Dev 2012; 26: 1602–1611.

    CAS  Google Scholar 

  78. 78

    Silva CM . Role of STATs as downstream signal transducers in Src family kinase-mediated tumorigenesis. Oncogene 2004; 23: 8017–8023.

    CAS  Google Scholar 

  79. 79

    Turkson J, Bowman T, Garcia R, Caldenhoven E, De Groot RP, Jove R . Stat3 activation by Src induces specific gene regulation and is required for cell transformation. Mol Cell Biol 1998; 18: 2545–2552.

    CAS  Google Scholar 

  80. 80

    Zhong Z, Wen Z, Darnell JE Jr . Stat3: a STAT family member activated by tyrosine phosphorylation in response to epidermal growth factor and interleukin-6. Science 1994; 264: 95–98.

    CAS  Google Scholar 

  81. 81

    Huang J, Wu S, Barrera J, Matthews K, Pan D . The Hippo signaling pathway coordinately regulates cell proliferation and apoptosis by inactivating Yorkie, the Drosophila homolog of YAP. Cell 2005; 122: 421–434.

    CAS  Google Scholar 

  82. 82

    Baonza A, Murawsky CM, Travers AA, Freeman M . Pointed and Tramtrack69 establish an EGFR-dependent transcriptional switch to regulate mitosis. Nat Cell Biol 2002; 4: 976–980.

    CAS  Google Scholar 

  83. 83

    Huang H, Li J, Hu L, Ge L, Ji H, Zhao Y et al. Bantam is essential for Drosophila intestinal stem cell proliferation in response to Hippo signaling. Dev Biol 2014; 385: 211–219.

    CAS  Google Scholar 

  84. 84

    Ren F, Shi Q, Chen Y, Jiang A, Ip YT, Jiang H et al. Drosophila Myc integrates multiple signaling pathways to regulate intestinal stem cell proliferation during midgut regeneration. Cell Res 2013; 23: 1133–1146.

    CAS  Google Scholar 

  85. 85

    Herranz H, Hong X, Cohen SM . Mutual repression by bantam miRNA and Capicua links the EGFR/MAPK and Hippo pathways in growth control. Curr Biol 2012; 22: 651–657.

    CAS  Google Scholar 

  86. 86

    Herranz H, Perez L, Martin FA, Milan M . A wingless and Notch double-repression mechanism regulates G1–S transition in the Drosophila wing. EMBO J 2008; 27: 1633–1645.

    CAS  Google Scholar 

  87. 87

    Irby RB, Yeatman TJ . Role of Src expression and activation in human cancer. Oncogene 2000; 19: 5636–5642.

    CAS  Google Scholar 

  88. 88

    Iliopoulos D, Hirsch HA, Struhl K . An epigenetic switch involving NF-kappaB, Lin28, Let-7 MicroRNA, and IL6 links inflammation to cell transformation. Cell 2009; 139: 693–706.

    CAS  Google Scholar 

  89. 89

    Iliopoulos D, Jaeger SA, Hirsch HA, Bulyk ML, Struhl K . STAT3 activation of miR-21 and miR-181b-1 via PTEN and CYLD are part of the epigenetic switch linking inflammation to cancer. Mol cell 2010; 39: 493–506.

    CAS  Google Scholar 

  90. 90

    Buttitta LA, Edgar BA . Mechanisms controlling cell cycle exit upon terminal differentiation. Curr Opin Cell Biol 2007; 19: 697–704.

    CAS  Google Scholar 

  91. 91

    Liu W, Singh SR, Hou SX . JAK-STAT is restrained by Notch to control cell proliferation of the Drosophila intestinal stem cells. J Cell Biochem 2010; 109: 992–999.

    CAS  Google Scholar 

  92. 92

    Schaeffer V, Althauser C, Shcherbata HR, Deng WM, Ruohola-Baker H . Notch-dependent Fizzy-related/Hec1/Cdh1 expression is required for the mitotic-to-endocycle transition in Drosophila follicle cells. Curr Biol 2004; 14: 630–636.

    CAS  Google Scholar 

  93. 93

    Shcherbata HR, Althauser C, Findley SD, Ruohola-Baker H . The mitotic-to-endocycle switch in Drosophila follicle cells is executed by Notch-dependent regulation of G1/S, G2/M and M/G1 cell-cycle transitions. Development 2004; 131: 3169–3181.

    CAS  Google Scholar 

  94. 94

    Weiss A, Herzig A, Jacobs H, Lehner CF . Continuous cyclin E expression inhibits progression through endoreduplication cycles in Drosophila. Curr Biol 1998; 8: 239–242.

    CAS  Google Scholar 

  95. 95

    Schwitalla S, Fingerle AA, Cammareri P, Nebelsiek T, Goktuna SI, Ziegler PK et al. Intestinal tumorigenesis initiated by dedifferentiation and acquisition of stem-cell-like properties. Cell 2013; 152: 25–38.

    CAS  Google Scholar 

  96. 96

    Bowman SK, Rolland V, Betschinger J, Kinsey KA, Emery G, Knoblich JA . The tumor suppressors Brat and Numb regulate transit-amplifying neuroblast lineages in Drosophila. Dev Cell 2008; 14: 535–546.

    CAS  Google Scholar 

  97. 97

    Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T et al. Fiji: an open-source platform for biological-image analysis. Nat Methods 2012; 9: 676–682.

    CAS  Google Scholar 

  98. 98

    Zielke N, Kim KJ, Tran V, Shibutani ST, Bravo MJ, Nagarajan S et al. Control of Drosophila endocycles by E2F and CRL4(CDT2). Nature 2011; 480: 123–127.

    CAS  Google Scholar 

  99. 99

    Shaw RL, Kohlmaier A, Polesello C, Veelken C, Edgar BA, Tapon N . The Hippo pathway regulates intestinal stem cell proliferation during Drosophila adult midgut regeneration. Development 2010; 137: 4147–4158.

    CAS  Google Scholar 

Download references


We thank David Ibbserson and the Deep Sequencing Core Facility of CellNetworks University of Heidelberg and the Fred Hutchinson Cancer Research Center Seattle for RNA and library preparations and sequencing. We thank Nicholas Harden, Yuh Nung Jan, Marco Milàn, Steven Hou, Sarah Bray, Sol Sotillos and the Yale FlyTrap consortium (USA) for gifts of fly stocks. We are grateful to Sylvia Kreger for experimental support and Monika Langlotz (ZMBH) for help with FACS. We thank Juanita Reetz for critically reading of the manuscript. Work in BAE’s laboratory was funded by ERC Grant 268515, NIH Grant R01 GM51186 and DFG Grant SFB873. AK was supported by a Human Frontiers in Science Program Long-Term postdoctoral fellowship (LT00316/2008-L).

Author information



Corresponding author

Correspondence to B A Edgar.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Supplementary Information accompanies this paper on the Oncogene website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Kohlmaier, A., Fassnacht, C., Jin, Y. et al. Src kinase function controls progenitor cell pools during regeneration and tumor onset in the Drosophila intestine. Oncogene 34, 2371–2384 (2015).

Download citation

Further reading