+1443 776-2705 panelessays@gmail.com

Article

Regulation of Drosophila intestinal stem cell
maintenance and differentiation by the
transcription factor Escargot
Mariano A Loza-Coll1,2, Tony D Southall3,†, Sharsti L Sandall1, Andrea H Brand3 & D Leanne Jones1,2,4,*

Abstract

Tissue stem cells divide to self-renew and generate differentiated
cells to maintain homeostasis. Although influenced by both intrin-
sic and extrinsic factors, the genetic mechanisms coordinating the
decision between self-renewal and initiation of differentiation
remain poorly understood. The escargot (esg) gene encodes a tran-
scription factor that is expressed in stem cells in multiple tissues
in Drosophila melanogaster, including intestinal stem cells (ISCs).
Here, we demonstrate that Esg plays a pivotal role in intestinal
homeostasis, maintaining the stem cell pool while influencing fate
decisions through modulation of Notch activity. Loss of esg
induced ISC differentiation, a decline in Notch activity in daughter
enteroblasts (EB), and an increase in differentiated enteroendo-
crine (EE) cells. Amun, an inhibitor of Notch in other systems, was
identified as a target of Esg in the intestine. Decreased expression
of esg resulted in upregulation of Amun, while downregulation of
Amun rescued the ectopic EE cell phenotype resulting from loss of
esg. Thus, our findings provide a framework for further compara-
tive studies addressing the conserved roles of Snail factors in coor-
dinating self-renewal and differentiation of stem cells across
tissues and species.

Keywords Amun; Drosophila; Escargot; Notch; stem cell

Subject Categories Development & Differentiation; Stem Cells; Transcription

DOI 10.15252/embj.201489050 | Received 21 May 2014 | Revised 20 October

2014 | Accepted 27 October 2014 | Published online 28 November 2014

The EMBO Journal (2014) 33: 2983–2996

See also: J Korzelius et al (December 2014)

Introduction

During homeostasis, tissue stem cells maintain the stem cell popula-

tion through self-renewal and give rise to differentiating progeny to

replace cells lost to normal turnover of the tissue (Biteau et al,

2011; Simons & Clevers, 2011; Wang & Jones, 2011). In response to

acute and chronic stress (infection, wounding, aging, metabolic

challenges), tissue stem cells can undergo dynamic waves of

symmetric self-renewing or differentiating divisions to quickly prop-

agate and replace damaged tissue (Morrison & Kimble, 2006; Egger

et al, 2010; O’Brien et al, 2011; Piccin & Morshead, 2011; Simons &

Clevers, 2011). While significant progress has been made in the

identification, isolation and manipulation of tissue stem cells in

organisms ranging from plants to vertebrates (Amatruda & Zon,

1999; Gentile et al, 2011; Takashima et al, 2013), our understanding

of conserved mechanisms that regulate the choice between self-

renewal and the onset of differentiation in vivo is lacking. In this

regard, model organisms such as Caenorhabditis elegans and

Drosophila melanogaster have been instrumental for the character-

ization of basic regulatory mechanisms in stem cells, such as the

role of asymmetric divisions (Yamashita et al, 2003; Wu et al, 2008;

Egger et al, 2010; Inaba & Yamashita, 2012) and the interaction

between stem cells and their niche (Wong et al, 2005; Spradling

et al, 2008; Losick et al, 2011; Resende & Jones, 2012).

The identification and characterization of stem cells in the

posterior midgut of adult flies (Micchelli & Perrimon, 2006; Ohlstein

& Spradling, 2006) has revealed numerous regulatory mechanisms

conserved between flies and vertebrates (Biteau et al, 2011; Jiang &

Edgar, 2012). The Drosophila midgut epithelium is composed of

intestinal stem cells (ISCs), enteroblasts (EBs), secretory enteroen-

docrine (EE) cells and absorptive enterocytes (ECs) (Fig 1A).

Through cell division, ISCs self-renew to maintain the ISC pool and

generate progenitor cells, which adopt either an EE or an EC fate. In

addition, ISCs can divide symmetrically to generate either two

daughter ISCs or two cells that will differentiate (O’Brien et al,

2011; Goulas et al, 2012; de Navascues et al, 2012). Indeed, it has

been proposed that intestinal homeostasis in flies is maintained

through a neutral drift between all possible mitotic outcomes

(de Navascues et al, 2012).

ISCs express the ligand Delta (Dl), which activates Notch (N)

signaling in adjacent EBs to promote differentiation and influence

cell fate decisions: a weak N signal specifies EE fate, whereas

1 Laboratory of Genetics, The Salk Institute for Biological Studies, La Jolla, CA, USA
2 Department of Molecular, Cell, and Developmental Biology, University of California Los Angeles, Los Angeles, CA, USA
3 The Gurdon Institute, University of Cambridge, Cambridge, UK
4 Eli and Edythe Broad Center of Regenerative Medicine and Stem Cell Research, University of California Los Angeles, Los Angeles, CA, USA

*Corresponding author. Tel: +1 310 206 7066; E-mail: [email protected]
†Present address: Department of Life Sciences, Imperial College London, London, UK

ª 2014 The Authors The EMBO Journal Vol 33 | No 24 | 2014 2983

stronger N signaling generates ECs (Micchelli & Perrimon, 2006;

Ohlstein & Spradling, 2006, 2007). Accordingly, strong loss-of-function

mutations in the N pathway cause an accumulation of ISC-like cells,

due to lack of EB differentiation, whereas weaker loss-of-function

mutations of Notch generate clusters of ISC-like cells and EEs, due

to a combination of impaired EB differentiation and a bias toward

the EE fate. In contrast, ectopic activation of N in ISCs results in

precocious differentiation, with a bias toward the EC fate (Micchelli

& Perrimon, 2006; Ohlstein & Spradling, 2006, 2007). While the

regulation of the ISC lineage by the Notch pathway and its down-

stream effectors has been well established previously (Micchelli &

Perrimon, 2006; Ohlstein & Spradling, 2006; Bardin et al, 2010;

Perdigoto et al, 2011), little is known about upstream mechanisms

that control the levels of Notch activity in this system.

The expression of esg reporter transgenes has been used to mark

ISCs and EBs since their initial characterization (Micchelli & Perri-

mon, 2006). Subsequently, the restricted expression of endogenous

esg mRNA in ISC/EB nests was confirmed by fluorescence in situ

hybridization in combination with immunofluorescence staining

(FISH/IF) (Fig 1B; Toledano et al, 2012). To date, however, whether

esg plays any specific role in the regulation of ISCs remains

unknown.

Esg is a member of the Snail family of transcription factors that

act primarily through competition with transcriptional activators for

access to a consensus-binding site, the E-box, within the promoter

region of target genes (Hemavathy et al, 2000; Nieto, 2002; Barrallo-

Gimeno & Nieto, 2005). The Snail factors were first characterized in

Drosophila and are conserved from mollusks to humans (Nieto,

2002). In addition to expression in ISCs, Esg is expressed in germ-

line stem cells (GSCs) and cyst stem cells (CySCs) of the testis (Kiger

et al, 2000; Voog et al, 2014) and, during development, in neural

stem cells and imaginal disks (Hayashi et al, 1993; Ashraf et al,

1999; Cai et al, 2001). Moreover, our previous work demonstrated

that Esg is required for the maintenance of CySCs and hub cells, a

critical component of the stem cell niche in the adult testis (Voog

et al, 2014).

Given the restricted expression of esg in ISC/EBs of the Drosoph-

ila intestine, we sought to characterize the function of Esg in these

cells. Here, we demonstrate that Esg is required for maintenance of

ISCs and an important regulator of Notch signaling within EBs.

Furthermore, DNA binding analysis by DamID identified Amun, a

previously characterized negative regulator of Notch signaling

(Abdelilah-Seyfried et al, 2000; Shalaby et al, 2009), as a putative

target of Esg in the gut. Accordingly, Esg knockdown in ISC/EBs

caused an upregulation in Amun expression in these cells. Further-

more, abrogating the increase in Amun rescued the reduction in

Notch activity and accumulation of EE cells caused by loss of Esg.

Based on our data, we conclude that Esg positively modulates Notch

signaling within EBs through repression of Amun, influencing the

decision between EC and EE fates. Therefore, we propose that Esg

plays a pivotal role in intestinal homeostasis, simultaneously

promoting stem cell maintenance and regulating the differentiation

of EBs.

Results

Loss of Escargot function in ISC/EBs leads to loss of ISCs and a
bias toward the enteroendocrine cell fate

Most esg alleles result in lethality during development when homo-

zygous; however, the shutoff (shof) allele of esg is a homozygous

viable mutation in the esg locus, which permits investigation of

adult phenotypes (Voog et al, 2014). FISH/IF analysis revealed that

esgshof homozygotes express normal levels of esg mRNA in ISC/EBs

(Supplementary Fig S1A), and intestines from these flies appeared

normal. Therefore, in order to probe the role of Esg in the intestine,

FRT-mediated recombination was used to generate ISCs homozy-

gous mutant for a null allele of esg, esgG66 (Whiteley et al, 1992; Lee

& Luo, 1999; Voog et al, 2014). In this experiment, a heat shock-

induced recombination generates esg mutant cells that become

permanently labeled by expression of GFP. Progeny derived from

marked ISCs are similarly marked, permitting characterization of

cells derived from esg mutant ISCs (or that of corresponding wild-

type counterparts, in control animals). Clones of esgG66 mutant cells

did not appear grossly different from wild-type clones at early time

▸Figure 1. Loss of Escargot induces ISC loss and a bias toward the enteroendocrine cell fate.A Drosophila posterior midgut epithelium. Schematic representation of cell types and their lineage relationships (see text for details; Micchelli & Perrimon, 2006;
Ohlstein & Spradling, 2006). Intestinal stem cells (ISCs) and enteroblasts (EBs) can be usually identified by an esg-GFP reporter, or expression of GFP under control
of ISC/EB-specific drivers. ISCs express Delta (Dl, which results in a characteristic punctate staining of ISCs), which activates Notch in EBs (revealed by b-GAL
staining of flies that carry a Su(H)-lacZ reporter of Notch activity (Bray & Furriols, 2001)). Enteroendocrine cells (EE) are identified by nuclear Prospero (Pros)
staining, whereas enterocytes (ECs) can be distinguished based on their large polyploidy nuclei (as revealed by DAPI staining of DNA).

B esg mRNA is restricted to ISC/EB cells. FISH/IF staining for esg mRNA (red, gray) and GFP protein (green) in midguts of 3- to 5-day-old adults carrying an ISC/EB-
specific reporter (“esg > GFP” = esg-Gal4, UAS-GFP).

C Clonal analysis of esg mutant ISCs. Representative images of wild-type (control) and esg mutant (esgG66) MARCM clones stained as indicated with DAPI, Pros and
GFP, 4 or 10 days after heat shock (dphs). esgG66 mutant clones are smaller and contain EE cells more frequently than controls (arrows).

D, E Loss of esg results in loss of ISCs and an increase in EE cells. A CellProfiler analysis of images as those in (C) confirmed that esgG66 mutant clones are significantly
enriched for EE cells (D) and have significantly less cells (E) than their control counterparts (***P < 0.001 and **P < 0.01, Kruskal–Wallis/Dunn test).

F, G Phenotypes induced by RNAi-mediated depletion of esg in ISC/EBs. (F) RNAi-mediated knockdown of Esg in ISC/EBs caused an accumulation of EE cells and a
noticeable change in the morphology and size of some ISC/EBs (arrows in bottom panel; e.g. compare the large GFP+ cell identified by the rightmost arrow to its
two smaller neighbors). Midguts from adults of the indicated genotypes were stained with DAPI (all nuclei), GFP (ISC/EB) and Pros (EE cells) following a 6-day
incubation at 25°C on 10 lg/ml RU486 or EtOH-containing food (as indicated). (G) Images as those in (F) were processed with CellProfiler to quantify the relative
proportion of EE cells in the indicated genotypes/treatments (see Materials and Methods for details). Each data point is an average proportion calculated from four
independent images per midgut, and the bars are the geometric mean � SEM of those averages. *** denotes a significant enrichment of EE cells following Esg
knockdown in ISC/EBs compared to either control group (P < 0.001, Kruskal–Wallis/Dunn test).

Data information: Scale bars = 20 lm.

The EMBO Journal Vol 33 | No 24 | 2014 ª 2014 The Authors

The EMBO Journal Regulation of intestinal stem cells by Escargot Mariano A Loza-Coll et al

2984

A B

C

F

D

E

G

Figure 1.

ª 2014 The Authors The EMBO Journal Vol 33 | No 24 | 2014

Mariano A Loza-Coll et al Regulation of intestinal stem cells by Escargot The EMBO Journal

2985

points (Fig 1C, 4 dphs); however, quantification of Prospero-

expressing (Pros+) cells within esgG66 clones revealed a significant

enrichment of EE cells (Fig 1D). At later time points, esgG66 clones

were significantly smaller than control clones (Fig 1C and E,

10 dphs) and remained significantly enriched for EE cells (Fig 1D).

We used CellProfiler (Carpenter et al, 2006; Kamentsky et al,

2011) to automatically classify and quantify cells within GFP+

esgG66 and control clones (Supplementary Fig S1B and Supplemen-

tary Table S1; see Materials and Methods for details). Our analysis

showed a higher prevalence of multicellular esgG66 clones that

contained only differentiated cells, consistent with a role for Esg in

ISC maintenance (polyploid ECs, EE cells or combinations thereof,

examples are shown in Supplementary Fig S1C, insets iv and v).

The proportion of esgG66 that did not contain ISCs or EBs was

approximately double that of wild-type counterparts, both at 4 and

10 dphs (Supplementary Fig S1D). In addition, esgG66 clones lacking

ISC/EBs had a significantly larger proportion of EE cells compared

to controls (Supplementary Fig S1E). Of note, the frequency of wild-

type clones that lost the ISC at 4 dphs (12.5%) is in close agreement

with previously reported rates of symmetric/differentiating ISC divi-

sions (O’Brien et al, 2011; Goulas et al, 2012; de Navascues et al,

2012). Particularly evident among esgG66 GFP+ clones were

instances of Pros+ doublets (Supplementary Fig S1C, inset v), which

were only rarely observed in control clones (3.13% of esgG66 clones

with more than one cell vs. 0.26% of control clones, Supplementary

Table S1).

To confirm these findings, we used the GAL4-UAS system to

induce RNAi-mediated knockdown of Esg expression. Specifically,

a UAS-esgRNAi construct (referred to as esgRNAi hereafter) was

expressed under control of a drug-inducible GAL4 driver, 5961GS,

whose expression pattern recapitulates esg expression in the diges-

tive tract (Biteau et al, 2010; Mathur et al, 2010). Esg knockdown

caused an alteration in the morphology of ISC/EBs, some of which

appeared larger and had an overall morphology reminiscent of ECs

(arrows in Fig 1F). In addition, a striking accumulation of Pros+

EE cells was observed (Fig 1F and G), with the total proportion

doubling after 6 days of esgRNAi induction. Similar results were

obtained with an independent UAS-esgRNAi line (Supplementary Fig

S1F) and with another inducible and considerably stronger ISC/

EB-specific GAL4 driver (esg-Gal4, tub-Gal80ts, referred to as

“esgts” hereafter) (Supplementary Fig S3C and D). Taken together,

these data indicate that loss of esg function in ISC/EBs leads to a

progressive loss of ISCs and a shift in differentiation toward the EE

lineage.

To determine in which cell type Esg is required to influence cell

fates, we directed the expression of UAS-esgRNAi to ISCs or EBs.

Restricted expression of UAS-esgRNAi and a UAS-2xYFP reporter to

ISCs (Wang et al, 2014) revealed morphological changes in YFP+

cells consistent with the induction of ISC differentiation (Fig 2A,

Supplementary Fig S2A). In addition, a significant accumulation of

EE cells (Fig 2C) was also observed. Depletion of Esg exclusively in

EBs, with a temperature-sensitive version of a Su(H)-Gal4 driver

(referred to as Su(H)-Gal4ts, Supplementary Fig S2B) (Zeng et al,

2010), also led to an alteration in the size and morphology of EBs,

including some cells that resembled polyploid ECs (Fig 2B, arrows).

Interestingly, there was an even more striking and significant

enrichment of EE cells (Fig 2B and D), indicating that the loss Esg

function in EBs is sufficient to accelerate their differentiation and

bias their fate toward the EE lineage.

Loss of Escargot function in ISC/EBs correlates with reduced
Notch activity in EBs

Numerous reports have established a connection between activation

of the Notch signaling pathway and cell fate decisions in the intes-

tine (Micchelli & Perrimon, 2006; Ohlstein & Spradling, 2006, 2007;

Maeda et al, 2008; Bardin et al, 2010; Perdigoto et al, 2011). Notch

is a transmembrane receptor that, upon binding to its ligands Delta

or Serrate, undergoes a series of proteolytic cleavages leading to the

cytoplasmic release of its intracellular domain (NICD or Nintra). Nintra

translocates into the nucleus and binds Suppressor of Hairless,

Su(H), a co-factor through which Notch activates the expression of

its target genes (Koch et al, 2013). Targeted depletion of Esg in ISC/

EBs caused a noticeable reduction in the expression of a Su(H)-lacZ

reporter of Notch activation in EBs (Supplementary Fig S3A),

including those adjacent to ISCs that continued to express Delta

(Fig 3A). These data indicate that loss of Esg leads to a decrease in

N signaling between ISCs and EBs. In support of these observations,

a modest induction of esgRNAi expression in flies heterozygous for a

null allele of Notch (N81K1) noticeably enhanced the accumulation

of EE cells, which was not obvious in either the esgRNAi or the

N81K1/+ background (Fig 3B). In 17% of midguts examined

(n = 18), supernumerary esg-GFP+ diploid cells were observed,

along with a striking accumulation of EE cells, reminiscent of the

characteristic phenotypes caused by Notch loss-of-function muta-

tions (Supplementary Fig S3B).

If the accumulation of EE cells upon Esg knockdown was due to

a reduction in Notch activity within EBs, then ectopic activation of

▸Figure 2. Loss of Esg function causes an accumulation of EE cells and an altered EB morphology.A Phenotypes caused by depletion of esg in ISCs. Midguts of the indicated genotype were stained with DAPI (all nuclei), YFP (primarily ISCs) and Pros (EE cells) following
6 days of incubation at the indicated temperatures (control = outcross to w1118 flies).

B Phenotypes caused by depletion of esg in EBs. Midguts of the indicated genotypes were stained with DAPI (all nuclei), GFP (EBs) and Pros (EE cells) following 6 days
of incubation at 29°C to allow the EB-restricted expression of UAS-esgRNAi or a UAS-GFPnls (control). Notice the stronger nuclear GFP staining in control samples due to
GFPnls expression. Arrows point to examples of EBs with a wider cytoplasm, dimmer GFP staining and seemingly polyploid nuclei following Esg knockdown.

C EE cell accumulation induced by depletion of esg in ISCs. CellProfiler was used as in Fig 1G to quantify the relative proportions of EE cells in midguts in images as
those in (A). *** and * denote a statistically significant difference in the relative proportion of EE cells in pairwise post-test comparisons indicated by the
corresponding bars (P < 0.001 and P < 0.05 Kruskal–Wallis/Dunn test).

D EE cell accumulation induced by depletion of esg in EBs. CellProfiler was used as in (C). *** denotes a statistically significant accumulation of EE cells following
EB-specific Esg knockdown, as compared to all other samples (P < 0.001, Kruskal–Wallis/Dunn test).

Data information: Scale bars = 20 lm.

The EMBO Journal Vol 33 | No 24 | 2014 ª 2014 The Authors

The EMBO Journal Regulation of intestinal stem cells by Escargot Mariano A Loza-Coll et al

2986

A

B

C D

Figure 2.

ª 2014 The Authors The EMBO Journal Vol 33 | No 24 | 2014

Mariano A Loza-Coll et al Regulation of intestinal stem cells by Escargot The EMBO Journal

2987

the Notch pathway through forced expression of a constitutively

active Notch construct should preserve the normal balance between

EEs and ECs. Expression of UAS-Nintra in ISCs and EBs with the

inducible 5961GS driver resulted in strong activation of the Su(H)-

lacZ reporter in both ISC and EBs (Fig 3C). Furthermore, as

predicted, co-expression of Nintra and esgRNAi in ISC/EBs resulted in

a suppression of ectopic EE cells (Fig 3C and D). Similar results

were obtained by expressing both transgenes with the esgts driver

(Supplementary Fig S3C and D). Restricted expression of Nintra and

esgRNAi to EBs, using the Su(H)-Gal4ts driver, resulted in lethality,

even at the restrictive temperature (18°C), likely due to leaky

expression in other tissues during development.

Interestingly, expression of Nintra alone in ISC/EBs led to an

expected trend toward EC enrichment, which became highly

significant when Nintra was expressed in combination with esgRNAi

(Supplementary Fig S3E). Similarly, there was a significant accumu-

lation of ECs following expression of esgRNAi alone, likely a conse-

quence of ISC depletion due to loss of Esg. To address whether the

proposed accumulation of EE cells reflected the non-specific, relative

accumulation of ECs and EEs as a result of ISC/EB loss, we calcu-

lated the relative proportion of EEs to ECs (Supplementary Fig S3F).

However, this analysis confirmed the enrichment of EE cells, relative

to ECs. Furthermore, the co-expression of Nintra and esgRNAi fully

reversed this trend, as predicted by the model that activated Notch

rescues the bias toward the EE fate in favor of EC differentiation. In

summary, reduced activation of the Su(H)-lacZ reporter following

esg knockdown (Fig 3A and Supplementary Fig S3A), together with

the observation that Nintra is epistatic to esgRNAi (Fig 3C and D and

Supplementary Fig S3D), strongly suggests that Notch functions

downstream of Esg to regulate cell fate decisions in the intestine.

Amun is a candidate Esg target in ISCs/EBs

In order to identify mediators of Esg function in ISC/EBs, we used

DamID to conduct a genome-wide in vivo mapping of Esg binding to

DNA (van Steensel & Henikoff, 2000; Southall & Brand, 2009). A

fusion of Esg and the bacterial DNA methylase Dam was expressed

in flies, which led to the methylation of DNA surrounding Esg bind-

ing sites in vivo. Genomic DNA was then extracted from midguts,

and the methylated regions were labeled and hybridized to whole-

genome tiling arrays (see the Supplementary Materials and Methods

for further details). Putative targets of Esg in the intestine were then

identified by proximity to the methylated regions identified via

DamID, or “Esg binding regions” (EBRs) (Supplementary Table S2).

We then used FlyMine to cross-reference genetic interactors of

Notch with a list of in vivo DamID targets of Esg (Supplementary

Table S2). From this approach, we identified Amun as a putative

candidate (Fig 4A). Amun is a nuclear protein with a predicted DNA

glycosylation domain, which has been shown to suppress Delta

gain-of-function phenotypes when overexpressed in the eye and the

wing (Shalaby et al, 2009) and to cause the formation of extra

scutellar bristles when overexpressed in wing imaginal disks

(Abdelilah-Seyfried et al, 2000). Both phenotypes are consistent with

Amun acting as an inhibitor of Notch signaling. Accordingly, when we

overexpressed Amun in ISC/EBs using esg-Gal4ts, we observed a down-

regulation in Su(H)-lacZ expression (Fig 4B) and a corresponding

enrichment of EE cells (Fig 4C). We confirmed these observations using

the 5961GS driver, which led to more subtle, but reproducible, pheno-

types (Supplementary Fig S4).

To further explore the relationship between Esg and Amun, we

tested the hypothesis that Esg knockdown would lead to upregula-

tion (de-repression) of Amun. Analysis of Amun mRNA expression

in whole intestines by reverse transcriptase (RT)–qPCR showed a

modest but statistically significant upregulation in Amun following

Esg knockdown (Supplementary Fig S5A). Because such slight

upregulation could be a consequence of ISC/EBs representing only a

very small proportion of the intestinal biomass, RT–qPCR measure-

ments were repeated in ISC/EBs purified by FACS sorting from

dissociated intestines (Dutta et al, 2013) (Supplementary Fig S5B).

This complementary approach confirmed the upregulation in Amun

mRNA levels following Esg knockdown (Fig 4D). We also assayed

Amun expression in posterior midguts via fluorescent in situ hybrid-

ization/immunofluorescence (FISH/IF) staining (Toledano et al,

2012). Although Amun mRNA was undetectable in ISC/EBs of

control guts using this method, some expression could be detected

within esg-GFP+ progenitor cells following RNAi-mediated depletion

of esg (Supplementary Fig S5C).

If Amun de-repression mediates the downregulation of Notch

signaling in EBs caused by Esg depletion, then the co-expression of

an AmunRNAi construct along with esgRNAi should rescue the bias in

cell fates toward the EE cell lineage. We tested this prediction by co-

expressing esgRNAi and AmunRNAi in EBs using the Su(H)-Gal4ts

▸Figure 3. Loss of Esg function in ISC/EBs leads to decreased Notch activity in EBs.A Esg knockdown in ISC/EBs causes reduced Notch reporter activity in EBs. Use of a Notch (N) activity reporter (Su(H)-lacZ; Bray & Furriols, 2001) reveals a noticeable
reduction in N signaling within EBs, despite of seemingly unaltered Delta expression in ISCs. Midguts of the indicated genotypes were stained with DAPI (all
nuclei), Delta (ISCs) and b-GAL (EBs) following 4 days of incubation in 10 lg RU486/ml or ethanol-containing food as indicated.

B Epistasis analysis between Esg (esgRNAi) and Notch (N81K1). Midguts of the indicated genotypes were stained with DAPI (all nuclei), GFP (esg+ cells) and Pros (EE
cells) following 6 days of incubation in 5 lg/ml RU or ethanol, as indicated. Notice that a lower concentration of RU486, resulting in a more moderate induction of
esgRNAi expression, did not produce the larger and dimmer GFP+ nuclei or EE cell accumulation typically observed with higher doses of RU486. Approximately 83%
(15/18) of the N81K1/+; esgRNAi guts showed a noticeable accumulation of EE cells in an otherwise normal-looking epithelium (right column, middle panel), whereas
the remaining 17% showed a drastic expansion of diploid, esg+ and EE cells (right column, lower panel), reminiscent of stronger Notch loss-of-function mutations
(compare with Supplementary Fig S3B).

C, D Constitutively active Notch signaling rescues the EE cell enrichment phenotype caused by Esg knockdown. (C) An esgRNAi construct and a constitutively active
Notch construct (Nintra) were expressed alone or in combination in ISC/EBs using the 5961GS driver. Midguts of the indicated genotypes were stained for GFP,
b-GAL (EBs) and Pros (EE cells) following 7 days of incubation in ethanol or 25 lg/ml RU486 as indicated. (D) Images as those in (C) were processed with
CellProfiler to quantify the relative proportion of EE cells. Midguts overexpressing the esgRNAi construct alone were the only sample that showed a significant EE
enrichment relative to the corresponding EtOH control (***P < 0.001, one-way ANOVA/Bonferroni test).

Data information: Scale bars = 20 lm.

The EMBO Journal Vol 33 | No 24 | 2014 ª 2014 The Authors

The EMBO Journal Regulation of intestinal stem cells by Escargot Mariano A Loza-Coll et al

2988

A

C

D

B

Figure 3.

ª 2014 The Authors The EMBO Journal Vol 33 | No 24 | 2014

Mariano A Loza-Coll et al Regulation of intestinal stem cells by Escargot The EMBO Journal

2989

A

B

D

C

Figure 4. Amun is a candidate target of Esg.

A Esg DamID profile surrounding the Amun locus. Data are displayed as custom UCSC Genome Browser tracks. Green/red bars represent the average log2 (intensity
ratio) between the Esg:Dam and Dam-control samples, mapped to the genomic regions � 10 kb from Amun (see Supplementary Materials and Methods for further
details). The yellow shading highlights an Esg-bound region (EBR), which includes a consensus Esg E-box ([G/A]CAGGTG; Fuse et al, 1994).

B, C Amun overexpression in ISC/EBs resembles a reduction in N signalling. (B) Midguts of the indicated genotype were stained with DAPI (nuclei), GFP (ISC/EB) and
b-GAL (N activation). (C) CellProfiler quantification of the relative proportion of EE cells in midguts from (B). A similar result was obtained using the milder 5961GS

driver (Supplementary Fig S4). ***P < 0.001, Mann–Whitney U-test. Scale bars = 20 lm.
D Amun mRNA upregulation following Esg knockdown in ISC/EBs. qPCR measurements of relative transcript abundances for the indicated genes from esg-GFP+ and

esg-GFP� cells isolated by FACS sorting from esg-GFP, 5961GS > UAS-esgRNAi flies incubated on ethanol or RU486 (25 lg/ml for 3 days) to induce RNAi expression.
Shown are means (� SEM) of efficiency-corrected relative quantities for each primer set, normalized to the corresponding GFP+/EtOH sample and to RpL32 levels
(used as reference). * denotes a significant reduction in esg and an increase in Amun transcript levels, …