Cell Reports Article Identification of Lgr5-Independent Spheroid-Generating Progenitors of the Mouse...

Please cite this article in press as: Mustata et al., Identification of Lgr5-Independent Spheroid-Generating Progenitors of the Mouse Fetal IntestinalEpithelium, Cell Reports (2013), http://dx.doi.org/10.1016/j.celrep.2013.09.005

Cell Reports

Article

Identification of Lgr5-IndependentSpheroid-Generating Progenitorsof the Mouse Fetal Intestinal EpitheliumRoxana C. Mustata,1,5 Gabriela Vasile,1,4 Valeria Fernandez-Vallone,1,4 Sandra Strollo,1 Anne Lefort,1 Frederick Libert,1

Daniel Monteyne,2 David Perez-Morga,2,3 Gilbert Vassart,1,* and Marie-Isabelle Garcia1,*1Institut de Recherche Interdisciplinaire en Biologie Humaine etMoleculaire (IRIBHM), Faculty of Medicine, Universite Libre de Bruxelles ULB,

Route de Lennik 808, 1070 Brussels, Belgium2Laboratory of Molecular Parasitology, IBMM, Universite Libre de Bruxelles, 12 rue des Professeurs Jeener et Brachet, 6041 Gosselies,

Belgium3Center for Microscopy and Molecular Imaging (CMMI), Universite Libre de Bruxelles, 8 rue Adrienne Bolland, B-6041 Gosselies, Belgium4These authors contributed equally to this work5Present address: Wellcome Trust and Medical Research Council, Cambridge Stem Cell Institute, Tennis Court Road, Cambridge,CB2 1QR, UK

*Correspondence: [email protected] (G.V.), [email protected] (M.-I.G.)

http://dx.doi.org/10.1016/j.celrep.2013.09.005

This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial-No Derivative WorksLicense, which permits non-commercial use, distribution, and reproduction in any medium, provided the original author and source are

credited.

SUMMARY

Immortal spheroids were generated from fetalmouse intestine using the culture system initiallydeveloped to culture organoids from adult intestinalepithelium. Spheroid proportion progressively de-creases from fetal to postnatal period, with a corre-sponding increase in production of organoids. Likeorganoids, spheroids show Wnt-dependent indefi-nite self-renewing properties but display a poorlydifferentiated phenotype reminiscent of incom-pletely caudalized progenitors. The spheroid tran-scriptome is strikingly different from that of adultintestinal stem cells, with minimal overlap of Wnttarget gene expression. The receptor LGR4, butnot LGR5, is essential for their growth. Trop2/Tacstd2 and Cnx43/Gja1, two markers highly en-riched in spheroids, are expressed throughout theembryonic-day-14 intestinal epithelium. Comparisonof in utero and neonatal lineage tracing using Cnx43-CreER and Lgr5-CreERT2 mice identified spheroid-generating cells as developmental progenitorsinvolved in generation of the prenatal intestinalepithelium. Ex vivo, spheroid cells have the potentialto differentiate into organoids, qualifying as a fetaltype of intestinal stem cell.

INTRODUCTION

The adult intestinal epithelium is one of the most rapidly self-re-

newing tissues in adult mammals. The steady-statemaintenance

and self-repairing ability of this tissue are ensured by a hierarchy

of stem cells present in the crypts of Lieberkhun (Barker et al.,

2012). Crypt base columnar cells (CBCs) are rapidly dividing

stem cells, expressing the specific marker Lgr5, responsible

for the constant production of transit-amplifying (TA) cells, while

simultaneously maintaining their own population in steady state

(Barker et al., 2007). Upon leaving the crypts, TA cells differen-

tiate into postmitotic enterocytes, Goblet cells, enteroendocrine

cells, and Tuft cells, which populate the intestinal villi before

being shed at their tip. Other rarely dividing adult intestinal

stem cells have been described, located just above the Paneth

cells in the ‘‘+4 position.’’ These ‘‘label-retaining cells’’ are char-

acterized by the expression of several marker genes, including

Bmi1, Hopx, Tert, and Lrig1 (Sangiorgi and Capecchi, 2008;

Yan et al., 2012; Takeda et al., 2011; Montgomery et al., 2011;

Wong et al., 2012). The possibility of interconversion between

slow and rapidly cycling LGR5+ intestinal stem cells has recently

been demonstrated in healing processes after tissue injury or in

ex vivo intestinal organoids (Tian et al., 2011; Takeda et al., 2011;

Roth et al., 2012, Buczacki et al., 2013).

Embryonic development of the murine intestine has been well

characterized morphologically, and recent studies have pro-

vided important information regarding the inductive cues and

transcription factors implicated in the differentiation of the gut

(Walton et al., 2012; Kim et al., 2007; Verzi et al., 2011; Noah

and Shroyer, 2013; Spence et al., 2011). Despite these pro-

gresses, there is still limited understanding about the origin of

the complex adult stem cell pool during development. Fetal pro-

genitors are believed to originate from the region between the

newly formed villi, around embryonic day 15–16 (E15–E16). We

previously showed that Lgr5-expressing cells are detected in

the ileal epithelium of E15 embryos and then become exclusively

localized to the intervillus region in the late fetal intestine, before

being restricted to CBCs in the crypts (Garcia et al., 2009). Earlier

studies with chimeric mice have suggested the existence of mul-

tiple progenitors in intervillus regions preceding crypt formation

Cell Reports 5, 1–12, October 31, 2013 ª2013 The Authors 1

mailto:[email protected]

mailto:[email protected]


Figure 1. Ex Vivo Culture of Fetal Small Intestine

Generates Mixed Populations of Spheroids and

Organoids

(A and B) Ex vivo culture of embryonic E16.5 small

intestine (A) and replating of selected spheroids and

organoids from a primary culture at day 7 (B) were per-

formed as reported in the scheme. Selected fields were

followed and photographed at the times indicated.

(C) Quantification of the percentage of spheroids and

organoids in cultured small intestine obtained at dif-

ferent embryonic and postnatal stages (mean ± SEM).

For each time point, the number of embryos/mice and

total counted elements is as follows: E14 (3, 67), E15

(7, 2476), E16 (5, 982), E17 (9, 1288), E18 (10, 923), P1

(10, 492), P4 (4, 2150), P5 (5, 2682), P15 (5, 383).

Scale bars represent 200 mm. See also Figure S1.


(Wong et al., 2002), but their relation with adult stem cells is still

unclear.

In the present study, we took advantage of the ex vivo culture

system of intestinal epithelium, successfully used to characterize

the adult CBCs (Sato et al., 2009), to culture fetal intestine at

different developmental stages. Self-renewing progenitor cells

were identified that could be cultured indefinitely as undifferenti-

ated hollow spheroids. Spheroid transcriptome differed mark-

edly from that of intestinal organoids, with low or absent

expression of intestinal differentiation genes and CBC markers,

and upregulation of several genes among which the Trop2/

Tacstd2 and Gja1/Cx43/Cnx43 genes (hereafter referred to as

Trop2 and Cnx43, respectively). Ex vivo, spheroid cells demon-

strated their ability to generate minigut-forming adult intestinal

stem cells. Finally, results from lineage-tracing experiments

2 Cell Reports 5, 1–12, October 31, 2013 ª2013 The Authors

showed that the intestinal epithelium in mice

is generated in two successive waves relying

on different kinds of progenitors: a transient,

fetal wave relies on Cnx43-positive cells,

whereas the postnatal epithelium is generated

from Lgr5-positive precursors of CBCs.

RESULTS

Ex Vivo Culture of Fetal Small IntestineIdentifies a Population of ImmortalSpheroidsThe ex vivo culture system described by Sato

et al. (2009) allows indefinite propagation of

organoid structures containing all differenti-

ated cell types present in normal intestinal

epithelium when adult intestinal crypts are

used as starting material. When E16.5 intes-

tine was used, as described in Figure 1A, we

observed that, in addition to adult-type mini-

gut organoids, a proportion of hollow spheres

(hereafter referred to as ‘‘spheroids’’) were

generated (Figure 1A). Upon serial replating,

both structures ‘‘bred true’’: i.e., spheroids

generated spheroids and organoids gener-

ated organoids (currently for >50 generations

or 10 months) (Figure 1B). Among the supplements present in

the ex vivo culture medium, spheroids required Rspondin1 for

growth and Noggin for efficient replating, whereas EGF did not

appear essential (Figure S1). Of note, fetal spheroids were grown

in total absence of Wnt ligand supplementation, making them

different from the spheroid-like structures generated from

normal adult tissue-derived organoids in presence of Wnt3a

(Sato et al., 2011).

The relative proportion of spheroids and organoids generated

from cultured small intestine was studied at different develop-

mental stages. Whereas E14–E15 intestines generated almost

exclusively spheroids, their proportion progressively decreased

during late fetal development, representing 60% at E16 and

less than 4% at postnatal day 5 (P5). Intestinal crypts har-

vested at P15 generated exclusively organoids (Figure 1C).

Figure 2. Spheroids Are Composed of a Polarized Epithelium of Intestinal Origin

(A) Frozen sections of spheroids and organoids were immunostained for E-cadherin, ZO-1, Villin, and CD44.

(B) Scanning electron microscopy of spheroids showed a monostratified epithelium. Higher magnification of the rectangle is shown on the right image of

the panel.

(C) Transmission electron microscopy of spheroids and organoids; Lu indicates lumens; insets show a higher magnification of the microvilli at the apical

membrane.

(D) Spheroid and organoid sections were stained with EdU and TUNEL for proliferative and apoptotic cells, respectively.

Scale bars represent 20 mm (A and D), 5 mm (C), and 100 and 10 mm (B, left and right panels, respectively). See also Figure S2.


This observation demonstrated that the relative abundance of

spheroid-generating cells is clearly linked to the developmental

stage, these cells being predominant in the period preceding

villogenesis.

Spheroids Are Made of a Proliferating PolarizedEpithelium of Intestinal OriginBasolateral labeling of E-cadherin and apical staining of the

tight-junction marker ZO-1 in both spheroids and organoids

demonstrated the epithelial nature and polarized state of the

two structures (Figure 2A). Expression of villin on the luminal sur-

face of the spheroids provided evidence for their intestinal origin

(Figure 2A). Electron microscopy further showed that spheroids

were formed by a monostratified layer of cells with a high degree

of interdigitations and exhibiting microvilli sparser and shorter

than in organoids (Figures 2B and 2C). With regard to cell prolif-

eration, the overall rate was similar in both kinds of structures

when computed as percentage of living cells, but, contrary to

organoids showing cycling cells restricted to crypt-like protru-

sions only, spheroids displayed proliferating cells all over their

surface (Figures 2D and S2A). Apoptotic cells were rarely de-

tected in the epithelium and lumen of spheroids, whereas the

lumen of organoids appeared full of dead cells (Figure 2D).

Spheroid Transcriptome Is Radically Different from thatof Both Organoids and CBCsWe compared global gene expression of spheroid/organoid

pairs obtained from four different embryos coming from three

different litters (two at E16, one at E18, and one at P0) (see

Experimental Procedures). Significance analysis of microarrays

(SAM) followed by further selection of transcripts with a 2-fold

up- or downregulation and a q value <0.054 allowed identif-

ication of 317 upregulated and 179 downregulated genes

(Table S1). Among the most strongly downregulated genes


Figure 3. Spheroid and Organoid Transcriptomes Are Different

(A) List of the 33 genes most up- or downregulated in spheroids versus organoids. Data were generated from four independent pairs of spheroid/organoids

samples.

(B) qRT-PCR analysis of transcripts from spheroids and organoids. Six pairs of spheroids/organoids were used. Bars represent mean ± SEM.

(C) GSEA analysis of microarray data versus CBC signature of Munoz et al. (2012).

(D) GSEA analysis of microarray data versus Cdx1KO/Cdx2KO list of upregulated genes (left panel) or downregulated genes (right panel). NES, normalized

enrichment score.

See also Figures S3 and S4 and Tables S1, S2, and S3.


were differentiation markers corresponding to the four main in-

testinal cell types (Figure 3A and Table S1). Loss of differentiation

markers was confirmed by quantitative RT-PCR (qRT-PCR) and

immunofluorescence (Figures 3B and 4A).

Interestingly, compared to organoids, spheroids also showed

low expression levels of several adult intestinal stem cell markers

(Lgr5/Gpr49, Smoc2, Axin2, Cdx1) (Figure 3B and Table S1).

Gene set enrichment analyses (GSEA) comparing our microarray

results with the set of CBC-enriched genes (Munoz et al., 2012)


confirmed the downregulation of the CBC signature in spher-

oids (Figure 3C). These data were validated by qRT-PCR and

extended to additional adult stem cell markers: whereas expres-

sion of Tert and Olfm4 were also downregulated in spheroids,

Bmi1 and Ascl2 were expressed at similar levels in spheroids

and organoids (Figure 3B). The uncoupling between Lgr5 and

Ascl2 gene expression was unexpected as both genes are

markers for adult CBCs (van der Flier et al., 2009). In accordance

with qRT-PCR data, expression of Lgr5 and the Wnt reporter

Figure 4. Compared Expression of Differentia-

tion, Stem/Progenitor Markers, and the Wnt

Stimulatory Tone in Paired Spheroids and

Organoids

(A) Immunofluorescence detection of Paneth (lysozyme

staining) and enteroendocrine cells (serotonin) in orga-

noids and spheroids. The arrow and arrowhead point to

an enteroendocrine and a Paneth cell, respectively.

(B) X-gal staining ofLgr5LacZ/+ or Lgr4LacZ/+ orAxin2LacZ/+

spheroids and organoids. Quantification of the relative

proportion of Xgal-positive (pos) and Xgal-negative (neg)

elements among Axin2LacZ/+ spheroids and organoids.

Fifty spheroids and 250 organoids were counted per

sample (n = 4 independent samples coming from

cultured P0 intestine stained at day 4 after the second

replating). Morphological classification of the counted

elements is detailed on the right side.

(C) Inhibition of porcupine activity affects growth of

spheroids. Spheroids were grown for 4 days in the

presence of IWP2 1 mM supplemented or not with re-

combinant Wnt3a at final concentration of 100 ng/ml. A

mean of 20 elements were measured per condition and

sample (n = 4/5 independent experiments performed on

three different samples). Spheroid size is graphed rela-

tive to control conditions (mean ± SEM).

Significance was computed from paired t test: Ctrl

versus IWP2, p = 0.0055; IWP2 versus IWP2 + Wnt3a,

*p = 0.027. Scale bars, 20 (A) and 200 (B and C) mm.

See also Figures S2–S4.


marker Axin2 appeared strong in the protrusions of organoids

containing adult stem cells but undetectable or very low in the

paired spheroids (Figure 4B). Indeed, 99.9% ± 0.1% of the orga-

noids and 3.8% ± 2.6% (mean ± SEM) of the spheroids were

Xgal positive in Axin2LacZ/+ samples, suggesting different Wnt

stimulatory tones in both kinds of elements (Figures 4B and

S2B). Noteworthy, the Lgr4 receptor was highly and similarly ex-

pressed in both spheroids and organoids when analyzed by

qRT-PCR and Xgal staining on Lgr4LacZ/+ samples (Figures 3B

and 4B), whereas Lgr6 was undetectable in any in vivo or

ex vivo intestinal sample (Figure S3A). The role of the Lgr4 and

Lgr5 receptors in self-maintenance of spheroids was investi-

gated. As demonstrated earlier for P0 intestine (Mustata et al.,

Cell Reports 5

2011), Lgr4 was essential for growth of both

spheroids and organoids as no explants could

be maintained from the small intestine of Lgr4-

deficient embryos (n = 11 embryos analyzed

between E16 and birth) (Figure S3B). The role

of Lgr5 and Lgr5-expressing cells in spheroids

was addressed using Lgr5-deficient mice

(Morita et al., 2004) and the Lgr5-DTR mouse

line (Tian et al., 2011), respectively. Contrary

to Lgr4, stable spheroids were readily cultured

from Lgr5 KO embryos (Figure S3C). Similarly,

specifically killing Lgr5-expressing cells by

administration of diphteria toxin to explants

cultured from heterozygous Lgr5-DTR em-

bryos was without effect on spheroid growth.

In contrast, the toxin severely affected initial

formation of protrusions in the paired organo-

ids (Figure S3D). Together with the low level of Lgr5 expression in

spheroids, these data indicate minor if any contribution of the

Lgr5 receptor to spheroid maintenance.

Among the most upregulated genes in spheroids versus

organoids (Figure 3A), several have been associated with

stem/progenitor cells (Cnx43, Trop2, and Ly6a/Sca1) (Todorova

et al., 2008; Goldstein et al., 2008; Holmes et al., 2007) or

reported to be expressed in malignant tumors (Spp1, Trop2,

and Clu) (Cao et al., 2012; Trerotola et al., 2013; Rizzi and Bet-

tuzzi, 2010). Also, some of the upregulated genes are reportedly

involved in tissue regeneration and/or development (Ctgf, Trop2,

Clu, and Vsig1) (Gunasekaran et al., 2012; Goldstein et al., 2008;

Lee et al., 2011; Oidovsambuu et al., 2011) (Table S1). The

, 1–12, October 31, 2013 ª2013 The Authors 5

Figure 5. Tissue Expression of Spheroid

Markers Ex Vivo and during Fetal Intestinal

Development

(A) Immunofluorescence images of Cnx43 and Trop2

coexpression in spheroids and organoids. Individual

channels for visualizing Cnx43 and Trop2 expression

in spheroids are depicted below the corresponding

merged image.

(B) Immunohistochemical detection of Cnx43 and

Trop2 in duodenal sections at E14, E16.5, and P0.

Arrow shows cells expressing high Trop2 at E16.5,

empty arrowheads evidence epithelial-expressing

Cnx43 at E16.5 and arrowheads points to mesen-

chymal expression of Cnx43.

Scale bars represent 20 mm (A and B). See also

Figure S5.


preferential expression of Cnx43 and Trop2 in spheroids was

confirmed by qRT-PCR and immunofluorescence on matched

pairs of spheroids/organoids (Figures 3B and 5A).

Before reaching a mature state with multiple crypt-like protru-

sions, organoids display a transient spheroid-like appearance

(Sato et al., 2009) (Figure S4A). The possibility that organoids

would pass through a less differentiated state similar to that of

the spheroids described here was ruled out by qRT-PCR anal-

ysis. Indeed, at their spheroid-like stage, organoids already ex-

hibited low Trop2, Cnx43, and Ccnd1 transcript levels, with

high expression of the differentiation markers (Crypt4, Muc2,

Chga, and Si) (Figure S4A). Of note, the gene expression profiles

of spheroids and organoids were stable, with no substantial

changes of their ratio after ten replatings (Figure S4B).

Given the striking morphological similarity of fetal spheroids

with those reported from cultured adult Apcmin adenomas, we

comparedourmicroarraydata to the short list of 38genesupregu-

lated in adenomas and their related organotypic cultures (Farrall

et al., 2012). Twelve genes appeared commonly upregulated in

both kinds of spheroid cultures, with the Trop2marker ranking in


the top list, pointing to b-catenin-dependent

activation of partially overlappinggenetic pro-

grams in the two cases (Table S2). Coherent

with this view, spheroid growth was inhibited

in a dose-dependent manner by the porcu-

pine inhibitor IWP2 and partially restored by

Wnt3a addition to the culture medium (Fig-

ures 4CandS4C). This suggests dependence

of spheroids on autocrine production of Wnt

ligand for their survival. Unexpectedly, only

two genes (Trex2 and Foxq1) out of the 80

making the intestinal Wnt/TCF signature

reported by Van der Flier et al. (2007) were

found to be upregulated in fetal spheroids.

Togetherwith the lowexpressionof additional

Wnt target genes in fetal spheroids (Lgr5

and Axin2), this suggests that Wnt would act

differently, depending on the epigenetic

status of the Apcmin and fetal spheroids.

From these observations, we concluded

that spheroids are made of poorly differenti-

ated intestinal cells with progenitor/stem cell characteristics

different from those of adult CBCs.

Spheroids Cells Correspond to Incompletely CaudalizedProgenitorsFurther inspection of the list of genes upregulated in spheroids

pointed to gastric and esophageal genes such asGkn1,2,3, Invo-

lucrin, Vsig1, and Krt4 (Figure 3A and Table S1). Together with

the strong downregulation of the caudal type homeobox 1

gene Cdx1 observed in spheroids (Figure 3B), this prompted

us to compare byGSEA ourmicroarray results to the list of genes

up- and downregulated in the intestine of Cdx1/Cdx2 double

knockouts, which are known to lose intestinal differentiation

and acquire expression of gastric markers (Verzi et al., 2011). A

clear positive relation was observed between the spheroid

versus organoids and the Cdx1/Cdx2 knockout intestine data

sets (Figure 3D), suggesting a lack of caudal differentiation in

spheroid cells. Surprisingly, this phenotype is observed despite

low but detectable expression of Cdx2 in spheroids (Figure 3B),

suggesting that the reported redundancy of Cdx genes in the


intestine (Beck and Stringer, 2010) is not effective in the context

of spheroids.

Trop2 and Cnx43 Are Expressed in the Fetal Epitheliumof the Mouse IntestineThe Trop2 molecule is highly expressed in several types of

tumors, including colorectal cancer, but its expression pattern

in the developing intestine has not yet been studied (Ohmachi

et al., 2006). As reported in Figures 5B and S5A, Trop2 was ex-

pressed at high levels in all epithelial proliferating progenitors of

the E14 duodenum. At E15.5, strong membrane staining was

observed in cells present at the tip of the newly formed villi

with lower but still detectable signal, in the intervillus zone (Fig-

ure S5A). Between E16.5 and birth, these strongly labeled cells

progressively disappeared, having likely been shed from the villi

(Figure 5B). Similar staining was observed in the ileum (Fig-

ure S5B). When sorted from fetal intestine and cultured

ex vivo, the epithelial Trop2+ cells generated spheroids, demon-

strating a direct filiation between in vivo Trop2+ cells and ex vivo

spheroids (Figure S5C). Of note, several transcripts found to

be enriched in spheroids displayed lower or absent expression

in Trop2+ cells sorted from the E15 intestine, indicating an

effect of the ex vivo culture conditions on spheroid phenotype

(Table S3).

Intestinal expression was also studied for the gap junction pro-

tein Cnx43 (second position in themicroarray list). At E14, Cnx43

was expressed in the lateral intercellular epithelial membranes

and at the basal pole of epithelial cells at the mesenchymal inter-

face (Figure 5B). At E16.5, Cnx43 was found in a more restricted

pattern in the epithelium, being almost exclusively detected in

cells of the intervillus region and showing a characteristic punc-

tuate staining (Figures 5B and S5D). In addition, Cnx43 was also

detected in rare mesenchymal cells, many of them located un-

derneath the epithelium of the intervillus region and within the

mesenchymal bud of nascent villi. At P0, Cnx43 was almost

exclusively expressed in mesenchymal cells surrounding the

epithelium, often organized into clusters (Figures 5B and S5D).

Cnx43-Expressing Cells Contribute to Prenatal VillusFormationWe used lineage tracing to explore the respective contribution of

Cnx43+ and Lgr5+ cells to generation of the fetal and postnatal

intestine. Cnx43-Cre/Rosa26R and Lgr5-Cre/Rosa26R embryos

were pulsedwith tamoxifen at the onset of villogenesis (E15), and

the labeling patterns were assessed at several time points pre-

and postnatally. One day post pulse (dpp), Cnx43-Cre/Rosa26R

embryos displayed numerous labeled cells in newly formed villi

and intervillus regions, mainly as single cells but also sometimes

as small groups of two to three cells, whereas 3 dpp the number

of labeled cells had decreased (Figure 6A). At a later time, most

epithelial positive cells disappeared from the intestine, with

labeling confined to villus extremities and only sparse intervillus

regions around birth (6 dpp, Figure S6A) and very rare ribbons of

crypt/villus units later (2 weeks pp, Figure 6A). In contrast, in E15-

pulsed Lgr5-Cre/Rosa26R embryos, only very rare labeled cells

were observed after 1 and 3 dpp, which persisted postnatally

(2 weeks pp), suggesting a minor contribution of Lgr5+ cells to

prenatal villus formation (Figure 6A). Low lineage tracing due to

potential variegation of the Lgr5-Cre knockin allele was unlikely

considering the number of embryos analyzed giving similar re-

sults (see Experimental Procedures). When a long pulse was per-

formed during the neonatal period via the lactating female (from

P5 to P8), Cnx43-dependent recombination yielded only mesen-

chymal labeling and virtually no epithelial labeling, in agreement

with the observed loss of Cnx43 epithelial expression around

birth (Figure 6B). In contrast, Lgr5-Cre/Rosa26R intestine dis-

played abundant labeling of crypt/villus units, compatible with

the tracing of immediate CBC precursors (Figures 6B and

S6B). Quantification of lineage tracing confirmed these observa-

tions (Figure 6C). Together, these results suggest that Cnx43+

cells function as progenitors of prenatal villi, their offspring being

lost by shedding in the lumen as Trop2+ cells (Figure S5B). Only a

small proportion of them, likely those gaining expression of Lgr5

in the forming intervillus region, may function as progenitors of

adult crypt-villus units. In agreement with this view, the rare post-

natal ribbons of cells generated from E15 Cnx43-positive pro-

genitors express Lgr5 in the crypts (Figure S6C).

Fetal Spheroids Can Differentiate into OrganoidsThe ability of spheroid Cnx43+ cells to potentially convert into

adult Lgr5+ CBCs was investigated in the ex vivo culture system.

After 7–8 days of culture, despite the stability of the phenotype, a

small proportion of the grown elements generated dark spheres,

with intraluminal accumulation of dead cells, and some of these

emitted protrusions similar to the crypt-like domains of organo-

ids (�1.4%) (Figures 7A and 7D). As expected for miniguts, the

newly formed organoids demonstrated strong expression of

Lgr5 restricted to the protrusions and exhibited expression of dif-

ferentiation markers from the absorptive and secretory lineages

(Figure 7C). Interestingly, the proportion of spheroids displaying

morphological differentiation toward an organoid-like phenotype

increased to 11.5% when spheroids were grown in a medium

containing the gamma secretase inhibitor DAPT (1 mM) (Figures

7B and 7D). As observed in the case of spontaneous differentia-

tion, after replating in normal medium, the organoid-like ele-

ments generated in the presence of DAPT gave rise to stable

organoids. These results indicate that fetal spheroids have the

potential to generate adult-type CBCs.

DISCUSSION

Using the minigut culture system, we have isolated from fetal in-

testine self-renewing cells that generate immortal epithelial

‘‘spheroids.’’ Spheroids display genetic commitment to intesti-

nal differentiation but express low levels of intestinal markers in

comparison to organoids and their gene expression profile dif-

fers radically from that of adult CBC stem cells. Several lines of

evidences suggest that spheroid cells represent a ‘‘frozen state’’

of progenitors found in the epithelium before the onset of villo-

genesis in vivo (i.e., around E14). A first strong argument in favor

of this hypothesis is given by the inverse relation between the

proportion of spheroids obtained from fetal tissue explants and

the developmental stage: whereas E14–E15 tissue generates

close to 100% spheroids, the ratio progressively decreases dur-

ing late fetal development, approaching zerowhen crypts start to

form (P5 onward). Second, we have shown that all epithelial cells


Figure 6. Lineage-Tracing Experiments

Performed on Cnx43-CreER/Rosa26R and

Lgr5-CreERT2/Rosa26R Embryos and

Neonatal Mice

(A) E15 embryos pulsed with tamoxifen were

sacrificed 1 day, 3 days, and 2 weeks post pulse

(pp). Arrows point to single recombined Lgr5-

positive cells.

(B) P5-old mice were tamoxifen-pulsed for 4

consecutive days (viamaternal milk) and sacrificed

at P10 (2 days pp). Full arrowheads point to re-

combined Cnx43-positive cells localized in the

mesenchyme. Immunohistochemical detection of

b-gal or YFP-positive cells (see Experimental

Procedures).

(C) Quantification of lineage-tracing experiments

reported as the number of positive clones counted

per 200 villus-intervillus/crypt units (mean ± SEM)

in E15-pulsed embryos harvested after 3 days or

2 weeks and in P5-P8-pulsed mice harvested at

P10. The numbers of embryos/mice used for

quantification were as follows: three and five for

Cnx43-Cre and Lgr5-Cre lines at E15 + 3 days pp

and two for both lines at E15 + 2 weeks pp; two

and three for Cnx43-Cre and Lgr5-Cre lines at

P5-P8 + 2 days pp, respectively. A mean of 200

villus-intervillus/crypt units were analyzed for each

sample.

Scale bars represent 200 and 20 mm (A, left and

right panels, respectively) and 50 mm (B). See also

Figure S6.


of the E14 intestine express Trop2 and Cnx43, two among the

most upregulated genes identified in spheroids by microarray.

In addition, Trop2+ cells sorted from fetal intestine generate

similar spheroids structures when cultured ex vivo. Finally, in

agreement with spheroids being generated from intestinal cells

preceding villogenesis, the global gene expression of intestinal

spheroids was found to display similarities with that of

double Cdx1/Cdx2 knockout intestine, a pattern consistent

with an intestinal dedifferentiation-like phenotype characterized


by expression of anterior foregut markers

(Verzi et al., 2011). The overall transcrip-

tomes of stomach and intestine epithelia

are very similar at E14.5 (Li et al., 2009),

with only the latter displaying striking

changes between E14.5 and E16.5, after

emergence of villi. This process was

coined ‘‘intestinalization’’ by Gumucio

and colleagues (Li et al., 2009) and

corresponds to the induction of the differ-

entiation markers of the three intestinal

lineages present at birth. Of note, despite

sharing similar morphology with embry-

onic spheroids described here, the cystic

vesicles generated by ex vivo culture of

Cdx2 knockout intestinal crypts appeared

different because these do not survive

replating (Stringer et al., 2012). We

conclude that our spheroids likely corre-

spond to immortalization of progenitors in a state just preceding

the ‘‘intestinalization’’ process.

The status of spheroids regarding Wnt signaling displays con-

tradictory features. On the one hand, several arguments point to

a role for activation of a b-catenin-dependent transcription pro-

gram in fetal spheroids; genes upregulated in spheroids show

significant overlap with those of Apcmin adenomas (Farrall

et al., 2012); spheroid growth and maintenance require the pres-

ence of the Wnt coactivator Rspondin, one of its receptor Lgr4,

Figure 7. Spheroids Can Generate Intestinal Organoids in Culture

(A and B) Spheroids were cultured in control medium (A) or in presence of DAPT 1 mM (B). Arrows point to crypt-like structures.

(C) X-gal staining of organoid-like structure having spontaneously differentiated from culture of Lgr5LacZ/WT spheroids shows Lgr5+ cells in crypt-like structures

(left panel). Newly formed organoid-like structures express high levels of cell lineage differentiation markers as compared to the surrounding spheroids from the

same well (right panel).

(D) Quantification of spheroid’s conversion to organoids after 7 days of culture.

Data represent the mean ± SEM of four independent experiments. *p < 0.016. Scale bars represent 200 mm.


and autocrineWnt production. On the other hand, only two out of

80 genes making the intestinal Wnt signature (Van der Flier et al.,

2007) are upregulated in spheroids, andAxin2 and Lgr5, two pro-

totypical Wnt targets highly expressed in CBCs, are virtually not

expressed in spheroids. This suggests that activation of the Wnt

pathway causes upregulation of different sets of genes in Apcmin

adenomas, CBCs, and fetal spheroid cells possibly as a result of

their different epigenetic status, or crosstalk with different con-

current regulatory cascades. Of interest, the dramatic effect of

Lgr4 deficiency on self-renewing ability of spheroids ex vivo

compared with the mild intestinal phenotype in vivo (Mustata

et al., 2011) parallels the situation described for Lgr4 deficiency

in organoids versus postnatal crypts (Mustata et al., 2011). The

absence of detectable levels of the Lgr6 paralog in the intestine

suggests that fetal extraepithelial signals, rather than potential

in vivo redundancy of Lgr receptors, compensate for the default

of Lgr4 function.

Expression of Trop2 and Cnx43 is clearly associated with the

normal development of prenatal intestinal epithelium in vivo. The

Trop2 molecule, highly expressed in the trophoblast and several

organs during development, marks subpopulations of adult

prostatic stem cells with regeneration capabilities and is upregu-

lated in a wide series of malignant tumors including colorectal

cancer (Yagel et al., 1994; Tsukahara et al., 2011; McDougall

et al., 2011, Goldstein et al., 2008, Trerotola et al., 2013). This

justified its qualification as a possible therapeutic target (Cubas

et al., 2009). Recently, it has been proposed that Trop2 enhances

stem-like properties of tumor tissue via b-catenin-dependent

mechanisms (Stoyanova et al., 2012).

Connexins, and, in particular, Cnx43, are responsible for the

establishment of cell-to-cell communication, involving transfer

of small soluble molecules (for a review, see Kar et al., 2012).

Expression of Cnx43 in progenitors has been reported in human

fetal bulge stem cells of the hair follicle (Arita et al., 2004) and in

the embryonic brain where it appears crucial to prevent prema-

ture neuronal differentiation (Santiago et al., 2010). Whether

Trop2 and Cnx43 are simply markers of fetal spheroids or

contribute functionally to their stem cell phenotype will need to

be addressed in future studies.

Although spheroid cells and progenitors sorted from fetal in-

testinal epithelium share expression of a series of genes, a few

transcripts showing high expression in spheroids (e.g., loricrin,

Sca1) were barely detected in the E15 intestinal epithelium.

This suggests that the ex vivo culture conditions, although



allowing survival of progenitors as spheroids, cause some distor-

tion of gene expression profile. A similar observation has been

reported when culturing adult normal and Apcmin adenoma tis-

sues (Farrall et al., 2012). Nonetheless, the strikingly different

gene expression profiles displayed by fetal spheroids and orga-

noids generated from the same intestine, when exposed to the

same culture conditions, highlight the different nature of the

self-renewing cells from which they originate. The fundamental

difference between CBCs and spheroid-generating cells is

further attested by their different Lgr5 status. Whereas CBCs

are currently defined as Lgr5-positive cells, fetal spheroids ex-

press low level of this gene and experiments with Lgr5-DTR

embryos demonstrate that spheroids do thrive in the absence

of Lgr5-expressing cells.

The progressive decrease in the number of Cnx43+ cells in

fetal intestinal epithelium during the E14–E18 period parallels

the spheroid-generating ability of intestinal explants and fits

with the lineage-tracing experiments of Cnx43+ cells performed

in utero and during the neonatal period. Together, these obser-

vations suggest that Trop2/Cnx43+ cells act as transient stem

cells responsible for the generation of fetal intestine in an envi-

ronment characterized by low Wnt and high Bmp stimulatory

tones prevailing at this period of intestinal development (Karls-

son et al., 2000; Li et al., 2009; Kim et al., 2007; Korinek et al.,

1998). In contrast, the majority of Lgr5+ cells are generated later

as precursors of adult CBCs. Ex vivo experiments showing the

capacity of Cnx43+ cells to convert to organoids suggest that a

fraction of these cells would be the precursors of Lgr5+ cells pre-

sent in the intervillus region during late gestation (Garcia et al.,

2009), whereas the vast majority of them are lost at the tip of

prenatal villi. Only very rare E15 Cnx43+ cells contribute directly

to the postnatal epithelium, being likely those that already ex-

pressed, or gained expression, of Lgr5 in a higher Wnt environ-

ment (Li et al., 2009). The boosting effect on spontaneous

conversion of spheroids to organoids observed ex vivo in the

presence of DAPT suggests that Notch pathway contributes to

maintenance of spheroid progenitor cells in an undifferentiated

state, as it contributes to proliferation of adult intestinal stem

cells (Noah and Shroyer, 2013).

Reminiscent of a switch from Cnx43+ progenitors of the early

intestinal epithelium to Lgr5+ stem cells, establishment of the

definitive intestinal epithelium adapted to digestion of adult

food type is known to be a two-step process in batrachians

(for a review, see Ishizuya-Oka and Hasebe, 2013). During meta-

morphosis of Xenopus laevis, the intestinal epithelium of the

tadpole is totally replaced by a novel, definitive epithelium gener-

ated from rare Lgr5+ cells. Similarly, our results suggest that the

intestinal epithelium in mammals is generated in two waves

relying on different kinds of stem cells: a transient, fetal wave

relies on Cnx43-positive progenitors, whereas the postnatal

epithelium is generated from Lgr5-positive precursors of CBCs.

Altogether, our data establish a relationship between transient

progenitors responsible for generation of fetal intestinal epithe-

lium and immortal spheroid-generating cells having the capa-

bility to ‘‘differentiate’’ into CBCs. This is of particular interest

in the recently documented context of interconversion of the

various adult intestinal stem cell types, in situations of epithelial

regeneration (Takeda et al., 2011; Tian et al., 2011; Parry et al.,


2013; Yan et al., 2012; Roth et al., 2012; Munoz et al., 2012;

Montgomery et al., 2011). Considering that most spheroid cells

are mitotically active and display an early, poorly differentiated

intestinal phenotype, grafting of spheroids cells could be partic-

ularly efficient in regeneration of injured gut epithelium.

EXPERIMENTAL PROCEDURES

Animal Experiments and Tissue Processing

Animal procedures complied with the guidelines of the EU and were approved

by the Local Ethical Committee. Mice strains were CD1 (Charles River Labora-

tories), Lgr5/LacZ-NeoR knockin (Morita et al., 2004), Lgr5-DTR knockin (Tian

et al., 2011), Lgr4/Gpr48^Gt (Leighton et al., 2001), Cnx43-KI-Cre-ER(T)

(EMMA), Rosa26R-LacZ, Rosa26R-YFP, Lgr5-Cre-ERT2, and Axin2-lacZ

(Jax mice). The day the vaginal plug was observed was considered as E0.5.

For lineage-tracing experiments, tamoxifen (Sigma-Aldrich) was dissolved

in a sunflower oil (Sigma-Aldrich)/ethanol mixture (9:1) at 10 mg/ml and used

in all experiments at a dose of 0.1 mg/g of body weight. For in utero induction,

pregnant mothers were injected intraperitoneally at E15. When required,

cesarean sections were performed for delivery, and newborn mice were

nursed by adoptive lactating females. For neonatal induction, lactating

mothers were injected intraperitoneally once a day for 4 consecutive days

(from P5 to P8). TheRosa26R-LacZ background was used in all Lgr5-CreERT2

experiments. For the experiments using Cnx43-CreERT, the Rosa26R-LacZ

background was used for in utero pulse + 1 and 6 dpp, and the Rosa26R-

YFP context was used in the embryonic pulse + 3 dpp and + 2 weeks pp as

well as in neonatal pulse + 2 dpp. The number of embryos analyzed for each

time point were as follows: lineage tracing with Cnx43-Cre (n = 4, 4, 3, 2,

and 2, for E15 + 1 dpp, 3 dpp, 6 dpp, 2 weeks pp and postnatal pulse, respec-

tively); lineage tracing with Lgr5-Cre (n = 10, 7, 2, and 3, for E15 + 1 dpp, 3 dpp,

2 weeks pp and postnatal pulse, respectively).

Tissue processing, histological protocols, and immunofluorescence/histo-

chemistry experiments were carried out as previously described (Garcia

et al., 2009). The primary antibodies used for staining were mouse anti-E-

cadherin, rat anti-CD44, mouse antibromodeoxyuridine, all from BD Biosci-

ences, goat anti-Villin (Santa Cruz Biotechnology), mouse antiserotonin and

rabbit antilysozyme from Dako, goat anti-Trop2 (R&D Systems), rabbit anti-

Cnx43 (Cell Signaling), rabbit anti-ZO-1 (Invitrogen), and chicken anti-b-gal

and anti-YFP (Abcam). EdU staining (Invitrogen) and TUNEL assays (Roche)

were performed according to the manufacturer’s instructions. Samples were

visualized with Zeiss Axioplan2 (immunohistochemistry) or Zeiss Observer

Z1 microscope (immunofluorescence).

Ex Vivo Culture

Small intestinal tissue was dissociated, and epithelial samples were cultured

as previously described (Mustata et al., 2011). Specifically, the culture medium

was composed of Advanced-DMEM/F12 medium supplemented with 2 mM

L-glutamine, gentamycin, penicillin-streptomycin cocktail, and 2% fetal

bovine serum (Gibco). The only growth factors added to the culture medium

were at a final concentration of 50 ng/ml EGF (PeproTech), 100 ng/ml Noggin

(PeproTech), and 500 ng/ml R-spondin1 (R&D Systems). Culture medium was

changed every other day, and, after 5–7 days in culture, spheroids and organo-

ids pairs were harvested, mechanically dissociated, and replated in fresh

Matrigel.

Diphteria toxin, IWP2, or DAPT compounds (all from Sigma-Aldrich) were

added together with fresh medium. In experiments with Trop2+-sorted cells,

1 mM JAG-1 (Anaspec) was added to the Matrigel, and the culture medium

was supplemented with 10 mM Y-27632 (Sigma). For Xgal staining experi-

ments, ex vivo cultures were prefixed for 15 min at room temperature before

proceeding to staining as described (Garcia et al., 2009). Pictures were

acquired with a Moticam Pro camera connected to Motic AE31 microscope

or with a Leica DFC 420C camera using the Leica Application Suite V3.8 soft-

ware. For electron microscopy studies, spheroids and organoids cultured in

Matrigel were layered onto a nitrocellulose filter, and samples were processed

as described in Supplemental Experimental Procedures. Fluorescence-acti-

vated cell sorting (FACS) is detailed in Supplemental Experimental Procedures.


Microarray Experiments

Two-channel microarray experiments were performed from spheroid/orga-

noid pairs isolated each from a given embryo. Specifically, spheroid/organoid

pairs were obtained as reported in Figure 1A. Following initial seeding of small

intestine from a given embryo/mouse (at E16, E18, or P0), spheroids and orga-

noids were selectively picked up for each animal and replated for three pas-

sages to reach sample homogeneity. Hybridization was performed on the

four independent pairs with dye-swap, on Mouse ReadyArray MM1100 slides

(38,467 70-mer probes; MI-Microarrays), as described (Garcia et al., 2009).

SAM analysis was performed using default parameters. A list of 1,982

upregulated and 1,276 downregulated genes was obtained. Unknown and

duplicate genes were removed and genes modulated more than 2-fold with

q value <0.054 were kept. The resulting short list of 317 upregulated and

179 downregulated genes is provided in Table S1.

Quantitative Real-Time PCR

Quantitative real-time PCR was performed on total RNA as reported (Garcia

et al., 2009). Expression levels were normalized to that of the house keeping

genes (HPRT, RPL13, and TBP). Each sample was run in duplicate. Primer

sequences were previously reported (Mustata et al., 2011) or are listed in

Supplemental Experimental Procedures.

Statistical Evaluation

Statistical analyses were performed with GraphPad Prism. All experimental

data are expressed as mean ± SEM. The significance of differences between

groups was determined by unpaired nonparametric (Mann-Whitney) test or

paired t test analysis.

ACCESSION NUMBERS

Microarray data sets were deposited in the Gene Expression Omnibus under

accession number GSE49803.

SUPPLEMENTAL INFORMATION

Supplemental Information includes Supplemental Experimental Procedures,

six figures, and three tables and can be found with this article online at

http://dx.doi.org/10.1016/j.celrep.2013.09.005.

AUTHOR CONTRIBUTIONS

R.C.M., G. Vasile, V.F.-V., S.S., and M.-I.G. performed the majority of the ex-

periments. A.F. and F.L. performed the microarray experiments and made the

related statistical analyses. D.M. and D.P.-M. performed and analyzed the

electron microscopy experiments. R.C.M., G.Vassart, V.F.-V., G. Vasile, and

M.-I.G. conceived the experiments and analyzed the results. R.C.M., G. Vas-

sart, and M.-I.G. wrote the paper.

ACKNOWLEDGMENTS

We are grateful to William C. Skarnes, Genentech, and Hans Clevers for

providing us with Lgr4/Gpr48^Gt, Lgr5-DTR, and Lgr5-LacZ-NeoR knockin

mice, respectively. We thank Christine Dubois for assistance in cell sorting ex-

periments and Cedric Blanpain, David Communi, and Pierre Vanderhaeghen

for critical reading of the manuscript. This work was supported by the Interuni-

versity Attraction Poles Programme-Belgian State-Belgian Science Policy

(6/14), the Fonds de la Recherche Scientifique Medicale of Belgium, the

Walloon Region (program ‘‘Cibles’’), and the not-for-profit Association

Recherche Biomedicale et Diagnostic. The CMMI is supported by the Euro-

pean Regional Development Fund and the Walloon Region.

Received: May 22, 2013

Revised: July 16, 2013

Accepted: September 4, 2013

Published: October 17, 2013

REFERENCES

Arita, K., Akiyama, M., Tsuji, Y., McMillan, J.R., Eady, R.A., and Shimizu, H.

(2004). Gap junction development in the human fetal hair follicle and bulge

region. Br. J. Dermatol. 150, 429–434.

Barker, N., van Es, J.H., Kuipers, J., Kujala, P., van den Born, M., Cozijnsen,

M., Haegebarth, A., Korving, J., Begthel, H., Peters, P.J., and Clevers, H.

(2007). Identification of stem cells in small intestine and colon by marker

gene Lgr5. Nature 449, 1003–1007.

Barker, N., van Oudenaarden, A., and Clevers, H. (2012). Identifying the stem

cell of the intestinal crypt: strategies and pitfalls. Cell Stem Cell 11, 452–460.

Beck, F., and Stringer, E.J. (2010). The role of Cdx genes in the gut and in axial

development. Biochem. Soc. Trans. 38, 353–357.

Buczacki, S.J., Zecchini, H.I., Nicholson, A.M., Russell, R., Vermeulen, L.,

Kemp, R., and Winton, D.J. (2013). Intestinal label-retaining cells are secretory

precursors expressing Lgr5. Nature 495, 65–69.

Cao, D.X., Li, Z.J., Jiang, X.O., Lum, Y.L., Khin, E., Lee, N.P., Wu, G.H., and

Luk, J.M. (2012). Osteopontin as potential biomarker and therapeutic target

in gastric and liver cancers. World J. Gastroenterol. 18, 3923–3930.

Cubas, R., Li, M., Chen, C., and Yao, Q. (2009). Trop2: a possible therapeutic

target for late stage epithelial carcinomas. Biochim. Biophys. Acta 1796,

309–314.

Farrall, A.L., Riemer, P., Leushacke, M., Sreekumar, A., Grimm, C., Herrmann,

B.G., and Morkel, M. (2012). Wnt and BMP signals control intestinal adenoma

cell fates. Int. J. Cancer 131, 2242–2252.

Garcia, M.I., Ghiani, M., Lefort, A., Libert, F., Strollo, S., and Vassart, G. (2009).

LGR5 deficiency deregulates Wnt signaling and leads to precocious Paneth

cell differentiation in the fetal intestine. Dev. Biol. 331, 58–67.

Goldstein, A.S., Lawson, D.A., Cheng, D., Sun, W., Garraway, I.P., and Witte,

O.N. (2008). Trop2 identifies a subpopulation of murine and human prostate

basal cells with stem cell characteristics. Proc. Natl. Acad. Sci. USA 105,

20882–20887.

Gunasekaran, U., Hudgens, C.W., Wright, B.T., Maulis, M.F., and Gannon, M.

(2012). Differential regulation of embryonic and adult b cell replication. Cell

Cycle 11, 2431–2442.

Holmes, C., Khan, T.S., Owen, C., Ciliberti, N., Grynpas, M.D., and Stanford,

W.L. (2007). Longitudinal analysis of mesenchymal progenitors and bone qual-

ity in the stem cell antigen-1-null osteoporotic mouse. J. Bone Miner. Res. 22,

1373–1386.

Ishizuya-Oka, A., and Hasebe, T. (2013). Establishment of intestinal stem cell

niche during amphibian metamorphosis. Curr. Top. Dev. Biol. 103, 305–327.

Kar, R., Batra, N., Riquelme, M.A., and Jiang, J.X. (2012). Biological role of

connexin intercellular channels and hemichannels. Arch. Biochem. Biophys.

524, 2–15.

Karlsson, L., Lindahl, P., Heath, J.K., and Betsholtz, C. (2000). Abnormal

gastrointestinal development in PDGF-A and PDGFR-(alpha) deficient mice

implicates a novel mesenchymal structure with putative instructive properties

in villus morphogenesis. Development 127, 3457–3466.

Kim, B.M., Mao, J., Taketo, M.M., and Shivdasani, R.A. (2007). Phases of

canonical Wnt signaling during the development of mouse intestinal epithe-

lium. Gastroenterology 133, 529–538.

Korinek, V., Barker, N., Moerer, P., van Donselaar, E., Huls, G., Peters, P.J.,

and Clevers, H. (1998). Depletion of epithelial stem-cell compartments in the

small intestine of mice lacking Tcf-4. Nat. Genet. 19, 379–383.

Lee, S., Hong, S.W., Min, B.H., Shim, Y.J., Lee, K.U., Lee, I.K., Bendayan, M.,

Aronow, B.J., and Park, I.S. (2011). Essential role of clusterin in pancreas

regeneration. Dev. Dyn. 240, 605–615.

Leighton, P.A., Mitchell, K.J., Goodrich, L.V., Lu, X., Pinson, K., Scherz, P.,

Skarnes, W.C., and Tessier-Lavigne, M. (2001). Defining brain wiring patterns

and mechanisms through gene trapping in mice. Nature 410, 174–179.

Li, X., Udager, A.M., Hu, C., Qiao, X.T., Richards, N., andGumucio, D.L. (2009).

Dynamic patterning at the pylorus: formation of an epithelial intestine-stomach

boundary in late fetal life. Dev. Dyn. 238, 3205–3217.




McDougall, A.R., Hooper, S.B., Zahra, V.A., Sozo, F., Lo, C.Y., Cole, T.J.,

Doran, T., and Wallace, M.J. (2011). The oncogene Trop2 regulates fetal

lung cell proliferation. Am. J. Physiol. Lung Cell. Mol. Physiol. 301, L478–L489.

Montgomery, R.K., Carlone, D.L., Richmond, C.A., Farilla, L., Kranendonk,

M.E., Henderson, D.E., Baffour-Awuah, N.Y., Ambruzs, D.M., Fogli, L.K.,

Algra, S., and Breault, D.T. (2011). Mouse telomerase reverse transcriptase

(mTert) expression marks slowly cycling intestinal stem cells. Proc. Natl.

Acad. Sci. USA 108, 179–184.

Morita, H., Mazerbourg, S., Bouley, D.M., Luo, C.W., Kawamura, K., Kuwa-

bara, Y., Baribault, H., Tian, H., and Hsueh, A.J. (2004). Neonatal lethality of

LGR5 null mice is associated with ankyloglossia and gastrointestinal disten-

sion. Mol. Cell. Biol. 24, 9736–9743.

Munoz, J., Stange, D.E., Schepers, A.G., van de Wetering, M., Koo, B.K.,

Itzkovitz, S., Volckmann, R., Kung, K.S., Koster, J., Radulescu, S., et al.

(2012). The Lgr5 intestinal stem cell signature: robust expression of proposed

quiescent ‘‘+4’’ cell markers. EMBO J. 31, 3079–3091.

Mustata, R.C., Van Loy, T., Lefort, A., Libert, F., Strollo, S., Vassart, G., and

Garcia, M.I. (2011). Lgr4 is required for Paneth cell differentiation and mainte-

nance of intestinal stem cells ex vivo. EMBO Rep. 12, 558–564.

Noah, T.K., and Shroyer, N.F. (2013). Notch in the intestine: regulation of

homeostasis and pathogenesis. Annu. Rev. Physiol. 75, 263–288.

Ohmachi, T., Tanaka, F., Mimori, K., Inoue, H., Yanaga, K., andMori, M. (2006).

Clinical significance of TROP2 expression in colorectal cancer. Clin. Cancer

Res. 12, 3057–3063.

Oidovsambuu, O., Nyamsuren, G., Liu, S., Goring, W., Engel, W., and Adham,

I.M. (2011). Adhesion protein VSIG1 is required for the proper differentiation of

glandular gastric epithelia. PLoS ONE 6, e25908.

Parry, L., Young, M., El Marjou, F., and Clarke, A.R. (2013). Evidence for a

crucial role of paneth cells in mediating the intestinal response to injury.

Stem Cells 31, 776–785.

Rizzi, F., and Bettuzzi, S. (2010). The clusterin paradigm in prostate and breast

carcinogenesis. Endocr. Relat. Cancer 17, R1–R17.

Roth, S., Franken, P., Sacchetti, A., Kremer, A., Anderson, K., Sansom, O., and

Fodde, R. (2012). Paneth cells in intestinal homeostasis and tissue injury. PLoS

One 7, e38965.

Sangiorgi, E., andCapecchi, M.R. (2008). Bmi1 is expressed in vivo in intestinal

stem cells. Nat. Genet. 40, 915–920.

Santiago, M.F., Alcami, P., Striedinger, K.M., Spray, D.C., and Scemes, E.

(2010). The carboxyl-terminal domain of connexin43 is a negative modulator

of neuronal differentiation. J. Biol. Chem. 285, 11836–11845.

Sato, T., Vries, R.G., Snippert, H.J., van de Wetering, M., Barker, N., Stange,

D.E., van Es, J.H., Abo, A., Kujala, P., Peters, P.J., and Clevers, H. (2009).

Single Lgr5 stem cells build crypt-villus structures in vitro without a mesen-

chymal niche. Nature 459, 262–265.

Sato, T., van Es, J.H., Snippert, H.J., Stange, D.E., Vries, R.G., van den Born,

M., Barker, N., Shroyer, N.F., van de Wetering, M., and Clevers, H. (2011).

Paneth cells constitute the niche for Lgr5 stem cells in intestinal crypts. Nature

469, 415–418.

Spence, J.R., Lauf, R., and Shroyer, N.F. (2011). Vertebrate intestinal endo-

derm development. Dev. Dyn. 240, 501–520.


Stoyanova, T., Goldstein, A.S., Cai, H., Drake, J.M., Huang, J., andWitte, O.N.

(2012). Regulated proteolysis of Trop2 drives epithelial hyperplasia and stem

cell self-renewal via b-catenin signaling. Genes Dev. 26, 2271–2285.

Stringer, E.J., Duluc, I., Saandi, T., Davidson, I., Bialecka, M., Sato, T., Barker,

N., Clevers, H., Pritchard, C.A., Winton, D.J., et al. (2012). Cdx2 determines the

fate of postnatal intestinal endoderm. Development 139, 465–474.

Takeda, N., Jain, R., LeBoeuf, M.R., Wang, Q., Lu, M.M., and Epstein, J.A.

(2011). Interconversion between intestinal stem cell populations in distinct

niches. Science 334, 1420–1424.

Tian, H., Biehs, B., Warming, S., Leong, K.G., Rangell, L., Klein, O.D., and de

Sauvage, F.J. (2011). A reserve stem cell population in small intestine renders

Lgr5-positive cells dispensable. Nature 478, 255–259.

Todorova, M.G., Soria, B., and Quesada, I. (2008). Gap junctional intercellular

communication is required to maintain embryonic stem cells in a non-differen-

tiated and proliferative state. J. Cell. Physiol. 214, 354–362.

Trerotola, M., Cantanelli, P., Guerra, E., Tripaldi, R., Aloisi, A.L., Bonasera, V.,

Lattanzio, R., de Lange, R., Weidle, U.H., Piantelli, M., and Alberti, S. (2013).

Upregulation of Trop-2 quantitatively stimulates human cancer growth.

Oncogene 32, 222–233.

Tsukahara, Y., Tanaka, M., and Miyajima, A. (2011). TROP2 expressed in the

trunk of the ureteric duct regulates branching morphogenesis during kidney

development. PLoS ONE 6, e28607.

Van der Flier, L.G., Sabates-Bellver, J., Oving, I., Haegebarth, A., De Palo, M.,

Anti, M., Van Gijn, M.E., Suijkerbuijk, S., Van de Wetering, M., Marra, G., and

Clevers, H. (2007). The Intestinal Wnt/TCF Signature. Gastroenterology 132,

628–632.

van der Flier, L.G., van Gijn, M.E., Hatzis, P., Kujala, P., Haegebarth, A.,

Stange, D.E., Begthel, H., van den Born, M., Guryev, V., Oving, I., et al.

(2009). Transcription factor achaete scute-like 2 controls intestinal stem cell

fate. Cell 136, 903–912.

Verzi, M.P., Shin, H., Ho, L.L., Liu, X.S., and Shivdasani, R.A. (2011). Essential

and redundant functions of caudal family proteins in activating adult intestinal

genes. Mol. Cell. Biol. 31, 2026–2039.

Walton, K.D., Kolterud, A., Czerwinski, M.J., Bell, M.J., Prakash, A.,

Kushwaha, J., Grosse, A.S., Schnell, S., and Gumucio, D.L. (2012).

Hedgehog-responsive mesenchymal clusters direct patterning and emer-

gence of intestinal villi. Proc. Natl. Acad. Sci. USA 109, 15817–15822.

Wong,M.H., Huelsken, J., Birchmeier, W., andGordon, J.I. (2002). Selection of

multipotent stem cells during morphogenesis of small intestinal crypts of

Lieberkuhn is perturbed by stimulation of Lef-1/beta-catenin signaling.

J. Biol. Chem. 277, 15843–15850.

Wong, V.W., Stange, D.E., Page, M.E., Buczacki, S., Wabik, A., Itami, S., van

de Wetering, M., Poulsom, R., Wright, N.A., Trotter, M.W., et al. (2012). Lrig1

controls intestinal stem-cell homeostasis by negative regulation of ErbB sig-

nalling. Nat. Cell Biol. 14, 401–408.

Yagel, S., Shpan, P., Dushnik, M., Livni, N., and Shimonovitz, S. (1994).

Trophoblasts circulating in maternal blood as candidates for prenatal genetic

evaluation. Hum. Reprod. 9, 1184–1189.

Yan, K.S., Chia, L.A., Li, X., Ootani, A., Su, J., Lee, J.Y., Su, N., Luo, Y., Heils-

horn, S.C., Amieva, M.R., et al. (2012). The intestinal stem cell markers Bmi1

and Lgr5 identify two functionally distinct populations. Proc. Natl. Acad. Sci.

USA 109, 466–471.

Supplemental Experimental Procedures

Electron microscopy protocol: Samples were washed with PBS and fixed with ice-cold glutaraldehyde

2% (Sigma). For TEM, samples were post-fixed in OsO4 (2%) in 0.1M cacodylate buffer (pH 7.2),

serially dehydrated in increasing ethanol concentrations, embedded in Agar 100 resin (Agar Scientific

Ltd, UK) and left to polymerize at 60°C for 2 days. Ultrathin sections (50-70 nm thick) were produced

with a Leica EM UC6 ultra-microtome, collected on formvar-carbon-coated copper grids and stained with

uranyl acetate and lead citrate by standard procedures. Observations were made on a Tecnai 10 TEM

(FEI) and images were captured with a Veleta CCD camera and processed with SIS iTEM (Olympus).

For SEM, samples were fixed over night at 4°C in glutaraldehyde 2.5%, 0.1M cacodylate buffer (pH 7.2),

and post-fixed in OsO4 (2%) in the same buffer. After serial dehydration, samples were dried at critical

point and coated with platinum by standard procedures. Observations were made in a Tecnai FEG ESEM

QUANTA 200 (FEI) and images processed by SIS iTEM (Olympus) software.

Flow cytometric analysis and cell sorting (FACS)

Small intestine of E15 embryos (pool of 12 embryos from a given litter per experiment, n=2 independent

experiments), or spheroids cultured ex vivo for 14 passages collected from Matrigel, were completely

dissociated with the Stem Pro accutase cell dissociation reagent (Thermo electron) and passed through a

40-μm nylon cell strainer (Greiner). Cells were Fc-blocked for 15 min on ice with anti-CD16/CD32 (BD

Biosciences) before incubation with fluorochrome-conjugated anti-Trop2-APC antibody, or the relevant

isotype control (R&D systems) in PBS-1% BSA for 30 min on ice. Live/Dead reagent (Invitrogen) was

used to exclude dead cells from the analysis. Cell sorting was done by using the Facs Aria I (BD

Biosciences). Sorted cells were either cultured ex vivo for cell plating efficiency measurements or

collected for RNA analysis.

List of primers used in RT-PCR and qRT-PCR.

Hprt fw: 5’-GCTACTGTAATGATCAGTCAACGGG, Hprt Rev: 5’-AAGCTTGCAACCTTAACCATTTTG; rlp13 Fw: 5’- CCCGTGGCGATTGTGAA, rlp13 Rev: 5’-TCATTGTCCTTCTGTGCAGGTT; tbp Fw:5’-TGTACCGCAGCTTCAAAATATTGTAT, tbp Rev: 5’-AAATCAACGCAGTTGTCCGTG; cdx1 Fw: 5’-GCGGTGGCAGCGGTAAGACC, cdx1 Rev: 5’-AGCTCGGACTTGCGCCGGAT; cdx2 Fw: 5’-CTGCTCTGGGTCCCTCGCCA, cdx2 Rev: 5’-CTGCGGAGCCAGGTTCAGGC; cnx43 Fw: 5’-TGGGGGAAAGGCGTGAGGGA, cnx43 Rev: 5’-ACCCATGTCTGGGCACCTCTCTT; ctgf Fw: 5’-TGACCCCTGCGACCCACACA, ctgf Rev: 5’-CAGGGTGCACCATCTTTGGCAGT; lgr6 Fw: 5’-CTACGCTGCAGCCGGTGAGCTG, lgr6 Rev: 5’-CAAAGTGGCTCCCTCTGCCTTCAGC; lor Fw: 5’-TCCCTGGTGCTTCAGGGTAAC, lor Rev: 5’-CTTTCCACAACCCACAGGAGG; olfm4 Fw: 5’-CAGCTGCCTGGTTGCCTCCG, olfm4 Rev: 5’-

GGCAGGTCCCATGGCTGTCC; sca1 Fw: 5’-GAAAGAGCTCAGGGACTGGAGTGTT, Sca1 Rev: 5’-TTAGGAGGGCAGATGGGTAAGCAA; smoc2 Fw:5’-GTTCGCACACCGGATCTTC, smoc2 Rev: 5’-TTGATCGACTCTCAAGAACGTG; spp1 Fw: 5’-TCGGAGGAAACCAGCCAAGGACT, spp1 Rev: 5’-AAGCTTCTTCTCCTCTGAGCTGCCA; tert Fw: 5’-TGAGAGCAGCAGCAGCCTGT, tert Rev: 5’-TAGGCTGGAGCCCTGGGGGA; trop2 Fw: 5’-GAACGCGTCGCAGAAGGGC, trop2 Rev: 5’-CGGCGGCCCATGAACAGTGA

Figure S1. Growth requirements for ex vivo cultures of paired spheroids and organoids, related to Figure

1. Upon replating, spheroids and organoids were cultured either in complete medium supplemented with

Rspondin1, Noggin and EGF or in absence of one of these growth factors. Selected fields were followed

over the culture period for each culture condition. Scale bar 200µm.

Figure S2. Comparison of proliferation and Wnt stimulatory tones in spheroids and organoids, related to

Figures 2 and 4. (A) Quantification of proliferative cells in spheroids vs. organoids (mean ± SEM).

Samples were labeled with BrdU for 1.5 hours, fixed and processed for staining. (B) Low magnification

view of Xgal staining performed on ex vivo cultured heterozygous Axin2LacZ/+ spheroids and organoids.

Scale bar: 1 mm.

Figure S3. Role of Lgr receptors in the growth of intestinal spheroids, related to Figures 3 and 4. (A)

Lgr6 is not expressed at detectable levels in intestinal samples. RT-PCR performed on spheroid (S) or

organoid (O) RNA samples and E15, E17 or adult small intestine RNA, using primers specific for Lgr4,

Lgr6 and HPRT. A E18 skin RNA sample was used as positive control for Lgr6 expression. RT (-)

corresponds to PCR of the E17 sample processed without reverse transcription. (B) Intestinal extracts

from E17 Lgr4 wild-type (WT, n=10) and Lgr4 knockout (KO, n=11) embryos were seeded and cell

growth was followed over the time. None of the cultured Lgr4KO samples generated productive spheroid

or organoid structures. (C) Intestinal extracts from E18 Lgr5 WT (n=3) and Lgr5 KO (n=3) embryos were

seeded and growth was followed over the time. The elements grown from the first seeding could be

efficiently replated (#3=passage 3). (D) Spheroid and organoid pairs from E15 or E17 WT (n=4) or

heterozygous (HE, n=4) Lgr5-DTR embryos were cultured in absence or presence of diphteria toxin (DT,

20 ng/ml final concentration). Medium was replenished every other day. DT treatment did not affect HE

spheroids but interfered with protrusion emergence in organoids (see inset). Scale bars 200 µm.

Figure S4. Gene expression and Wnt signaling of fetal spheroids differs from that of early spheroid-like

stage of organoids, related to Figures 3 and 4. (A) qPCR analysis of spheroids and organoids at day 2 and

day 5 of culture. * represent values ≤ 0.03. Images represent typical cultures of spheroids and organoids at

day 2. (B) qPCR analysis of spheroids and organoids at passage 5 and 15. * represent values ≤ 0.01. (C)

Porcupine inhibition with the IWP2 compound affects fetal spheroid growth in a dose-dependent way.

Spheroids were grown for 4 days in presence of IWP2 at the indicated concentrations. A mean of 20

elements were measured per condition and sample (n=2/4 independent experiments performed on 3

different samples). Spheroid size is graphed relative to control conditions (mean ± SEM). Significance

was computed from paired t-test: *p <0,035. Scale bars: 200 µm.

Figure S5. Expression of spheroid markers in the fetal intestine, related to Figure 5. (A) Most of the

Trop2-hi cells are BrdU+ cells in the epithelium of E14 intestine. In the E15.5 tissue, only few Trop2-hi

cells at the tip of the nascent villi are proliferative cells. (B) E16.5 ileal tissue shows Trop2-hi cells

shedding off in the lumen. Cnx43 is also expressed in the ileal epithelium at this stage. (C) FACS-sorted

Trop2 positive cells from E15 embryos (blue dots in the left panel) generated spheroids when cultured ex

vivo (central panel) while Trop2 negative cells did not. Quantification of spheroid forming efficiency of

both populations is shown on the right panel (mean ± SEM of 2 independent experiments.). (D) Higher

magnifications for immunohistochemical detection of Cnx43 in duodenal sections at E16.5. Empty arrow-

heads show punctuated staining observed in epithelial cells localized in the intervillus region. Scale bars

20 µm (A, B, D) and 200 µm (C).

Figure S6. Lineage tracing experiment performed on Cnx43-CreER and Lgr5-CreERT2/Rosa26R

embryos, related to Figure 6. (A) Cnx43-CreER/Rosa26-LacZ embryos pulsed with tamoxifen at E15

were sacrificed 6 days post-pulse. Tissues were immunostained with anti-βgal antibody to detect

recombined Cnx43 positive cells. (B) Lgr5-CreERT2/Rosa26LacZ P5-old mice were tamoxifen-pulsed

for 4 consecutive days (via maternal milk) and sacrificed at P10 (2 days pp). Low, medium and high

magnification images of a representative field are shown. Positive clones are evidenced by circles. (C)

Cnx43-CreER/Rosa26-YFP/Lgr5-LacZ embryo pulsed at E15 and sacrificed 2 weeks post-pulse. Serial

sections of 5 µm show that crypt-villus units generated from Cnx43 positive cells contain Lgr5 expressing

cells in the bottom of the crypt. Tissues were immunostained with anti-GFP and anti-βgal antibodies to

detect recombined Cnx43 positive and Lgr5 expressing cells, respectively. Scale bars 20 µm (A, B right

panel, C) and 200 µm (B left and central panels).

Table S1. Short List of 317 Upregulated and 179 Downregulated Genes in Spheroids versus Organoids,

Related to Figure 3 and Experimental Procedures

Provided as a separate Excel file.

Table S2. Comparison of the Fetal Spheroid versus Organoids Microarray List with Genes Upregulated

in Apcmin Adenomas and the Corresponding Organotypic Cultures, Related to Figure 3

Up > 2fold in Apc

min adenoma vs normal tissue

Genes in column 1 and up > 2 fold in adenoma spheroids vs organoids

Genes in column 1 and up > 2 fold in fetal spheroids vs organoids (present study)

1 Cfi Cfi

2 Lcn2 Lcn2

3 Krt23 Krt23 Krt23

4 Ctse Ctse

5 Foxq1 Foxq1 Foxq1

6 Scara3 Scara3

7 Tnfrsf11b Tnfrsf11b

8 S100a9 S100a9

9 Avil Avil

10 Stra6 Stra6

11 Onecut2 Onecut2

12 Cxcl1 Cxcl1

13 Tacstd2 Tacstd2 Tacstd2=Trop2

14 Prr18 Prr18

15 Rbp1 Rbp1

6 Phlda1 Phlda1

17 Tnfrsf12a Tnfrsf12a

18 Rgs17 Rgs17 Rgs17

19 Mmp14 Mmp14

20 Akr1b8 Akr1b8 Akr1b8

21 Il33 Il33 Il33

22 Tns4 Tns4 Tns4

23 Tnfrsf19 Tnfrsf19

24 Slc2a1 Slc2a1

25 Egln3 Egln3

26 Anxa3 Anxa3 Anxa3

27 Bmp4 Bmp4

28 Rbms3 Rbms3

29 H6pd H6pd

30 Cgnl1 Cgnl1 Cgnl1

31 Pls3 Pls3

32 Rbms1 Rbms1 Rbms1

33 Krt7 Krt7 Krt7

34 Bace1 Bace1

35 Pdlim4 Pdlim4 Pdlim4

36 Tspan12 Tspan12

37 Ptprb Ptprb

38 Cldn2 Cldn2

39 1190003m12rik

40 Wif1

41 S100a8

42 Otop2

43 Etv5

44 Cyp2c55

45 Fxyd3 Fxyd3

46 Cdkn1c

47 Asah3l

48 Ccdc3

49 Igfbp4

50 Mt2

51 Loc100047651

52 Rnf43

53 Sox4

54 Htra1

55 Sidt1

56 Gstm5

57 Paip1

58 2210407c18rik

59 Smoc1

60 Mt1

61 Fndc3b

62 Mafg

63 1110012d08rik

64 Nedd9

65 Mir16

66 Fgl1

67 Tmc7

68 Prkcbp1

69 Loc100043257

70 Nfib

71 Fvt1

72 Mycl1

The genes upregulated in both fetal (present study) and adenoma spheroids (Farrall et al., 2012) are

overlined in pale yellow. The single gene upregulated only in fetal spheroids is overlined in pale green.

Spheroids (CT) Sorted Trop2+ cells (CT)

SPP1 19.8 ± 0.0 26.4 ± 0.1 CNX43 22.0 ± 0.1 25.2 ± 0.5 TROP2 22.1± 0.5 24.7 ± 0.0 CTGF 20.0 ± 0.1 24.5 ± 0.1 SCA1 18.8 ± 0.1 29.5 ± 0.3 LOR 21.5 ± 0.0 undetermined

Table 3. Compared Expression of Trop2+ Cells Isolated In Vivo and Ex Vivo, Related to Figure 3

qPCR analysis of Trop2+ cells sorted from cultured spheroids or E15 intestinal tissue. Values represent

CTs (mean ± range of duplicate determinations).

Cell Reports Article Identification of Lgr5-Independent Spheroid-Generating Progenitors of the Mouse...

Documents

Transcript of Cell Reports Article Identification of Lgr5-Independent Spheroid-Generating Progenitors of the Mouse...