Impaired epithelial differentiationQ:1; 2 of inducedpluripotent stem cells from EEC patients isQ:3; 4 rescuedby APR-246/PRIMA-1MET

Q:5

Ruby Shalom-Feuersteina,b,1, Laura Serrorb,1, Edith Aberdama,b, Franz-Josef Müllerc, Hans van Bokhovend,Klas G. Wimane, Huiqing Zhoud, Daniel Aberdama,b,1,2, and Isabelle Petita,b,1Q:6

aInstitut National de la Santé et de laQ:7 Recherche Médicale U898, University of Nice, France; bBruce Rappaport Faculty of Medicine, INSERTECH, Technion,; 8

Haifa, Israel; cZentrum für Integrative Psychiatrie, Kiel, Germany; dDepartment of Human Genetics, Nijmegen Centre for Molecular Life Sciences, RadboudUniversity Nijmegen Medical Centrum, Nijmegen, The Netherlands; and eDepartment of Oncology-Pathology, Karolinska Institutet, Cancer Center Karolinska,SE-171 76 Stockholm, Sweden

Edited by Tak W. Mak, The Campbell Family Institute for Breast Cancer Research, Ontario Cancer Institute at Princess Margaret Hospital, University HealthNetwork, Toronto, ON, Canada, and approved May 8, 2012 (received for review February 7, 2012)

Ectodermal dysplasia is a group of congenital syndromes affectinga variety of ectodermal derivatives. Among them, ectrodactyly,ectodermal dysplasia, and cleft lip/palate (EEC) syndrome is causedby single pointmutations in the p63 gene,Q:9 which controls epidermaldevelopment and homeostasis. Phenotypic defects of the EEC syn-drome include skin defects and limbal stem-cell deficiency. In thisstudy,we designed a uniqueQ:10 cellularmodel that recapitulatedmajorembryonic defects related to EEC. Fibroblasts from healthy donorsand EEC patients carrying two different point mutations in the DNAbinding domain of p63 were reprogrammed into induced pluripo-tent stem cell (iPSC) lines. EEC-iPSC from both patients showed earlyectodermal commitment into K18+ cells but failed to further differ-entiate into K14+ cells (epidermis/limbus) or K3/K12+ cells (cornealepithelium). APR-246 (PRIMA-1MET), a small compound that restoresfunctionality of mutant p53 in human tumor cells, could revert cor-neal epithelial lineage commitment and reinstate a normal p63-re-lated signaling pathway. This study illustrates the relevance of iPSCfor p63 relateddisorders andpaves theway for future therapyof EEC.

corneaQ:11

Ectodermal dysplasia are rare syndromes characterized by ab-normal development of the skin and ectodermal derivatives,

like teeth, hair, cornea, and nails. Among these syndromes, someare related to mutations on the transcription factor p63 (TP63)and represent a group of autosomal dominant ectodermal dys-plasia associated with orofacial clefting and limb abnormalities.The severe phenotype of p63-null mice highlighted the major roleof p63 in embryonic development, and particularly in the devel-opment of ectodermal lineages (1, 2). Five syndromes in whichp63mutations have been detected include ectrodactyly, ectodermaldysplasia, and cleft lip/palate syndrome (EEC), ankyblepharon,ectodermal dysplasia, and cleft lip/palate syndrome (AEC), limbmammary syndrome, acro-dermato-ungual-lacrimal-tooth syn-drome, and Rapp-Hodgkin syndrome. EEC mutations are clus-tered in the DNA-binding domain and AEC mutations are foundin the sterile α-motif or transactivation inhibitory domain (3).Although there are some clinical similarities and overlap betweenthe syndromes, the specific location of p63 mutations in the dif-ferent domains of the gene shows a strong genotype-phenotypecorrelation, and thus different molecular mechanisms behind thevarious p63-associated syndromes. Clinical and penetrance vari-ability are observed for the same mutation, suggesting that thenumber of p63 patients could be underestimated (3, 4). More-over, variability of phenotype among syndromes could be a resultof specific functional consequences of a single mutation. For ex-ample, two hot-spot mutants, R304W and R204W, both locatedin the DNA binding domain of the p63 gene, represent EECsyndrome but, based on their analogy with p53 mutations, may actdifferently because R304W interferes with DNA binding and

R204W with global protein structure/stability of p63, influencingthe transcription of target genes differently (5, 6).In addition to skin defects, EEC patients suffer from visual

morbidity with progressive limbal stem-cell deficiency that leadsto severe visual impairments and blindness (7, 8). Therefore,modeling of these diseases is essential to identify abnormalitiesin molecular processes involving p63, their effects on cell growthand skin development, and for drug screening. In vitro cellularmodels of rare skin and corneal diseases are obtained by the useof patient-derived primary epidermal cells. Given that p63 isa master regulator of embryonic steps of epithelial development,cellular models that could recapitulate the main steps of skin andcorneal epithelial development in vitro are necessary. The re-cently discovered capacity of human somatic cells to be relativelyeasily reprogrammed into embryonic stem cell-like pluripotentstem cells (iPSC) offers numerous perspectives in therapies byproviding patient-specific differentiated cells on demand andnovel cellular models for specific pathologies. iPSC technologyprovide pluripotent stem cells carrying genetic characteristics ofpatients. These cells have the remarkable ability to recapitulatein vitro the main steps of human embryonic development, andthey provide urgently needed tools to generate patient-specific,organotypic disease models. These cellular models may be usedfor the discovery of novel drugs both in a flexible and highlyspecific manner, because they facilitate high-throughput com-pound screening and toxicity assays. Finally, unlike patient pri-mary cells, iPSC derived from patients’ cells provide researcherswith cells with unlimited proliferation capacity. The clinicalpenetrance of the p63 gene is highly variable, apparently be-cause of other genetic and epigenetic factors (9, 10). Therefore,animal models in which a single EEC mutation is inserted byknock-in may not reproduce the human pathology. Here, wederived iPSC lines from healthy control and EEC patients andevaluated their ability to differentiate into epidermal and cor-neal epithelial cells. Our study demonstrated that they displayedimpaired epithelial commitment that could be partially rescuedby a small therapeutic compound.

Author contributions Q:12; 13: R.S.-F., D.A., and I.P. designed research; R.S.-F., L.S., E.A., F.-J.M., andI.P. performed research; H.v.B., K.G.W., and H.Z. contributed new reagents/analytic tools;R.S.-F., L.S., F.-J.M., and I.P. analyzed data; and R.S.-F., D.A., and I.P. wrote the paper.

Conflict of interest statement: K.G.W. is cofounder, shareholder, and member of theboard of Aprea AB, a company that develops p53-based cancer therapy including thecompound APR-246.

This article is a PNAS Direct Q:14Submission.1R.S.-F., L.S., D.A., and I.P. contributed equally to this work.2To whom correspondence should be addressed. E-mail: daniel.aberdam@inserm.fr.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1201753109/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1201753109 PNAS Early Edition | 1 of 5

CELL

BIOLO

GY

123

456

789

10111213

141516

171819

20212223

242526

272829

303132

33343536

373839

404142

43444546

474849

505152

535455

56575859

606162

636465

666768

697071

72737475

767778

798081

82

8485

868788

9091

929394

95969798

99100101

102103104

105106107108

109110111

112113114

115116117

118119120121

122123124

ResultsDerivation of iPSC Lines from WT and EEC Fibroblasts. iPSC lineswere obtained by lentiviral infection of primary dermal fibro-blasts isolated from one healthy individual and two EEC patientscarrying single point mutations R304W or R204W in the p63gene. These two mutations located in the DNA binding domainare among the five hotspots accounting for 90% of EEC. Severalclones with typical iPSC morphology (Fig. S1A) wereQ:15 expandedand two to three lines for each donor were used for subsequentcharacterization and experiments (WT or +/+, R204W/+ andR304W/+). Pluripotency of iPSC was confirmed by the expres-sion of various markers such as octamer-binding transcriptionfactor 4 (OCT4), TRA-1-80, and alkaline phosphatase by stain-ing (Fig. S1A) and OCT4, sex-determining region Y box-2(SOX2), DNA methyltransferase 3b (DNMT3b), and NANOGby quantitative PCR (qRT-PCRQ:16 ) (Fig. S1B). We have previouslydeveloped a purely data-driven approach, termed PluriTest, to-ward empirically defining the human pluripotent state, becausethe gold standard germ-line transmission is impossible for humancells (11). Briefly, the PluriTest data model was derived via in-terrogation of large-scale datasets of genome-wide somatic andpluripotent expression profiles and can be used to rapidly andconfidently assess the pluripotency of human cells through bio-informatic analysis of microarray data from new stem-cell prep-arations without the sacrifice of laboratory animals for Teratomaassays (11). All iPSC lines displayed high pluripotency scores,similarly to human embryonic stem cells (hESC) and other fi-broblast-derived iPSC (Fig. S1C). Finally, iPSCWT and iPSCEEC

lines were able to form embryoid bodies in suspension and differ-entiate into cell types belonging to the three germ layers upon ad-hesion (Fig. S1D). This finding was confirmed by qRT-PCR analysison gene expression specific for neuroectodermal (NCAM1), en-dodermal (AFP), and mesodermal (CD31) fates. In addition, EEC-iPSC were able to differentiate into trophoectoderm, as shown byCDX2 expression after 6 d with bone morphogenetic protein-4(BMP-4) (Fig. S1E). Taken together, these data confirmed thatwe obtained pluripotent iPSC lines and we next explored theirpotential to model molecular characteristics of EEC syndrome.

Impaired Epidermal Differentiation of EEC-iPSC Lines. EEC patientssuffer from impaired skin development. To define whether EEC-iPSC lines could mimic these defects, we optimized protocols todifferentiate human iPSC into epidermal cells. Epidermal com-mitment of embryonic stem cells can be induced by BMP-4, whichinhibits neural differentiation and promotes epidermal fate ofneuroectodermal cells (12–14). Prolonged treatment with BMP-4and ascorbic acid combined with keratinocyte growth conditionshave been shown to promote efficiently hESC differentiation intomature keratinocytes (14). However, applying this protocol to iPSClines appeared inefficient because most of the cells differentiate ina heterogenous manner, with few ectodermal-like cells and withlarge cystic-like structures (Fig. S2A). We found that extraembry-onic cells expressing CDX2 and CGHaQ:17 are produced from iPSC inresponse to a high dose of BMP-4 (Fig. S2B), suggesting that iPSC,contrary to hESC, do not efficiently differentiate into ectodermallineage because they failed to first spontaneously commit into neu-roectoderm (15). Because inhibition of the TGF-β/nodal pathwaypromotes lost of pluripotency and neuroectodermal commitment(16), we tested the effect of a TGF-β inhibitor, SB431542, on theepidermal differentiation of iPSC. Ectodermal differentiation ofiPSC was dramatically improved in presence of SB431542. Mor-phologically, the colonies lost their pluripotent aspect faster andacquired a homogenous ectodermal phenotype within 7 d (Fig.S2A). Pluripotency markers (Oct4 and Dnmt3b) and trophoec-toderm markers (CDX2 and HCGa) were reduced but the ker-atinocyte marker K14 was increased (Fig. S2B). Differentiation ofiPSC treated with BMP-4, AA, and SB431542 for 30 d led to

mature keratinocytes (Fig. S2Ci), as demonstrated by expressionof K14 and K10 (Fig. S2C ii and iii), signs of spontaneous strati-fication (Fig. S2C iv and v), and ability to form typical keratinocytecolonies upon splitting (Fig. S2Cvi). This process demontrates thatTGF-β inhibition significantly improved the production of epi-dermal cells from iPSC lines.iPSCEEC were subjected to epidermal commitment according

to the above protocol. In response to SB431542 and BMP-4during the first 10 d, iPSCEEC initiated ectodermal commitment,as illustrated by typical ectodermal cell morphology similarly toiPSCWT cells (Fig. S2A). However, although iPSCWT cells un-derwent epidermal transition from days 13–15 for the productionof keratinocyte-like areas proliferating during the next 10 d,iPSCEEC continued to display the same ectodermal morphology(Fig. S3A). Only a small number of keratinocyte-like cells couldbe noticed occasionally. Immunofluorescence staining (Fig. 1A)and quantification by flow cytometry (Fig. S3B) confirmed thatboth iPSCR304W/+ and iPSCR204W/+ could produce K18+ ecto-dermal progenitors to a similar extent as iPSCWT (42.5% and46%, respectively). However, iPSCR304W/+ failed to further dif-ferentiate into K14+ epidermal cells (2.2%), compared withiPSCWT (27.5%) (Fig. 1 and Fig. S2B). Similar results have beenobtained for iPSCR204W/+ (Fig. 1B). Of note, expression of p63was up-regulated to the same extent during epidermal commit-ment of both WT and mutated iPSC (Fig. S3C).

Impaired Corneal Differentiation of EEC-iPSC Is Rescued by APR-246/PRIMA-1MET. EEC patients suffer from visual morbidity because ofimpaired cornea associated with limbal stem-cell deficiency (5,6). iPSC lines were induced to corneal fate using a slight modi-fication of a protocol designed by Lako and colleagues for hESC(17). In brief, iPSC lines were seeded on collagen IV in thepresence of medium conditioned by human corneal fibroblasts(COF) and treated with BMP-4 between days 0 and 3. As illus-trated by real-time qRT-PCR analysis, human iPSC lines un-derwent sequential differentiation into ectodermal precursors(K18+/Pax6+) at day 4, markers of corneal progenitors (K14+/p63+/pax6+) appeared at day 8, and markers of terminally dif-ferentiated corneal epithelial (Pax-6+/K3+/K12+) cells wereexpressed at day 14 (Fig. 2A). Remarkably, at day 14, most of thecells became corneal epithelial cells, as detected by immunoflu-orescence staining (Fig. 2B) and FACS analysis (Fig. 2C). p63 isa putative marker of corneal stem cells, which are located in thelimbus, a defined region at the corneal periphery (18). We nextchallenged iPSCEEC for their ability to undergo proper cornealepithelial commitment compared with the iPSCWT. Similarproduction of ectodermal progenitors (K18+/E-cadherin+) was

BADay 30Day 10

+/+

R30

4W/+

0

10

20

30

40

50

+/+

**Pe

rcen

tage

of K

14+

cells

K18/dapi

K18/dapi K14/dapi

K14/dapi

*

Fig. 1. Impaired epidermal differentiation of EEC-iPSC. iPSC+/+, iPSCR204W/+,and iPSCR304W/+ cells were subjected to epidermal differentiation protocol inpresence of BMP-4 and SB431542, and immunofluorescence for K18 and K14after 15 d of differentiation (A) or analyzed by flow cytometry analysis forK18 and K14 after 25 d of differentiation (B). The data are an average of thepercentage of K14+ cells ± SE from two independent iPSC clones. *P < 0.001.Q:21

2 of 5 | www.pnas.org/cgi/doi/10.1073/pnas.1201753109 Shalom-Feuerstein et al.

125126127

128129130

131132133

134135136137

138139140

141142143

144145146147

148149150

151152153

154155156

157158159160

161162163

164165166

167168169170

171172173

174175176

177178179

180181182183

184185186

187188189

190191192

193194195

196197198199

200201202

203204205

206207208209

210211212

213214215

216217218

219220221222

223224225

226227228

229230231232

233234235

236237238

239240241

242243244245

246247248

observed at day 10 for iPSCWT, iPSCR204W/+, and iPSCR304W/+

(Fig. 2D). However, K14 and K3 staining revealed the inability ofiPSCEEC to undergo further commitment for the production oflimbal cells and corneal cells, respectively, compared with iPSCWT.In parallel, expression of several p63-dependent genes was evalu-ated on day 10 of commitment by qRT-PCR (Fig. 2E). These genesinclude genes that are enhanced at embryonic day (E) 14.5 only inthe presence of p63 (GJA1,GJB6, KRTDAP, and KRT14) (19) andp63-target genes, of which deregulation during epithelial de-velopment is associated with ectodermal dysplasia [DLX5, DLX6,and CDH3 (P-Cadherin)] (20, 21). Interestingly, most of thesegenes were significantly less expressed in mutated cells comparedwith control cells (Fig. 2E).Because corneal epithelial commitment of iPSC appeared

more efficient and much faster than epidermal fate, we employedthis system for testing whether the small compound APR-246/PRIMA-1MET, that was recently shown to restore p63-inducedapoptosis in human tumor cells (22), could rescue corneal epi-thelial commitment of iPSCEEC. Thus, iPSCWT, iPSCR204W/+,and iPSCR304W/+ cells were treated with APR-246 (20 μM) fromday 3 to day 14 and corneal commitment was monitored at day14. APR-246/PRIMA-1MET had no significant effect on differ-entiation of iPSCWT cells (Fig. S4). Notably, a partial restorationin expression of epithelial markers by iPSCEEC was observed inthe presence of APR-246 (Fig. 3A). The average expression ofthese factors by iPSCEEC was ∼15% compared with iPSCWT.However, a significant increase was demonstrated in the pres-ence of APR-246, to 43 ± 12% in iPSCR304W/+ and to 24 ± 15%

and iPSCR204W/+, compared with iPSCWT (Fig. 3B). Similarly,the expression of K14 protein was significantly elevated by APR-246 treatment, as shown by flow cytometry analysis (Fig. 3C).APR-246/PRIMA-1MET had no significant effect on cell pro-liferation (Fig. S5), excluding the possibility that this effect wasa result of enhanced cell proliferation of iPSCEEC cells. Of note,the rescue effect was systematically stronger for the mutantR304W. We thus concluded that patient iPSC cells are able torecapitulate in vitro EEC major molecular defects, which couldbe potentially restored by APR-246 treatment.

DiscussionCongenital diseases, like ectodermal dysplasia syndromes, lackexperimental models that could recapitulate embryonic eventsin vitro. Attempts have been made to express exogenously mutatedforms of p63 in mouse embryonic stem cells that provided inter-esting insights on molecular signaling pathways (23). However, thisapproach does not reproduce pathophysiological conditions ofequimolarity between mutated and wild-type alleles. Here, wereprogrammed fibroblasts isolated from EEC patients into plurip-otent iPSC lines and stimulated them to undergo in vitro majorsteps of epidermal and corneal-epithelial development. iPSCEEC

recapitulated impaired epithelial commitment, demonstrating theutility of this model for congenital skin or corneal diseases. Weobserved that p63-mutated pluripotent cells failed to make thetransition from K8/K18 ectodermal precursors to K14 epidermalcells, confirming the key role of p63 in controlling the early

K18

no AbDay 7Day 14

K14

Fluorescence

A

1

10

100

1000

K18

pax6 K14 p63 K3

K12

Rel

ativ

e e

xpre

ssio

n

K18

pax6

K14 p 63 K3

K12

K18

pax 6 K14 p6

3K

3K

12

Day 4 Day 8 Day 14B

K14

K3

p63

pax6

C

no AbDay 7Day 14

D

E

K14 KRTDAP GJB6 GJA1 CDH3 DLX5 DLX61

10R

elat

ive

exp

ress

ion

100

1000+/+ R304W/+ R204W/+

+/+ R304W/+ R204W/+

K18 K5 K18 K5 K18 K5

E-Cad E-CadE-Cad

ΔNp63 K14 ΔNp63 K14 ΔNp63K14

Fig. 2. Impaired corneal epithelial commitment of iPSC lines. iPSC+/+ were seeded on collagen IV-coated dishes in corneal fibroblast-conditioned medium thatwas supplemented with BMP-4 for the first 3 d of differentiation. Cells were harvested at the indicated time points and subjected to real-time PCR analysis ofectodermal markers (pax6 and K18), corneal epithelial progenitor markers (K14 and p63), and markers of terminally differentiated corneal-epithelial cells (K3and K12) (A). Cells were collected at day 14 of differentiation and subjected to coimmunofluorescent staining of p63 and K14 or pax6 and K3 (B). Flowcytometry analysis of iPSC+/+ that were harvested at days 7 and 14 of differentiation and stained with K18 or K14 antibodies is shown in C. iPSC+/+ and iPSCEEC

lines were differentiated into corneal epithelial fate for 10 d (E). Immunostaining followed by fluorescent microscopy was performed for determining theexpression of the indicated proteins (D), and real-time PCR analysis was performed for determining the relative expression of the indicated epithelialtranscripts (E).Q:22

Shalom-Feuerstein et al. PNAS Early Edition | 3 of 5

CELL

BIOLO

GY

249250251

252253254

255256257

258259260261

262263264

265266267

268269270271

272273274

275276277

278279280

281282283284

285286287

288289290

291292293294

295296297

298299300

301302303

304305306307

308309310

311312313

314315316

317318319

320321322323

324325326

327328329

330

338339

340341342

343344345346

347348349

350351352

353354355356

357358359

360361362

363364365

366367368369

370371372

embryonic epidermal switch, as already suggested by the hESCmodel (23, 24) and by p63-deficient mice (19).BMP-4–based protocols have been established to obtain kera-

tinocytes from pluripotent cells (12, 14, 25) and very recently, Itohet al. used BMP-4 combined with retinoic acid for a limited time togenerate keratinocytes from iPSC from dystrophic epidermolysisbullosa patients reaching 30–40% of K14+ cells that could beenriched only after passaging or cell sorting (26). Despite possibleenrichment and satisfactory 3D potential (14, 26), improvedprotocols for highly homogenous production of keratinocytesfrom pluripotent cells are still needed to consider cell therapy. Weobserved that BMP-4–induced epidermal commitment of iPSCwas significantly low compared with what was reported with hESC(14), apparently because of limited spontaneous neuroectodermcommitment that led to the generation of BMP-4–induced tropho-ectoderm. We found that TGF-β inhibition efficiently promotedneuroectoderm engagement and that concomitant addition ofBMP-4 increased the production of K14+ cells. These findingsmay be useful for optimizing the production of keratinocytes fromhuman iPSC for future cell therapy.In contrast, corneal epithelial commitment appeared much

more efficient and homogenous. For that reason, the effect ofAPR-246 was tested on corneal fate. The small molecule APR-246, also known as PRIMA-1MET (p53-dependent reactivationand induction of massive apoptosis) can restore p53-inducedapoptosis in several types of cancer cells (27). APR-246 wasrecently tested in a phase I/II clinical trial. Furthermore, APR-246 has been shown to also target mutant forms of p73 and p63,an effect that is presumably caused by highly homologous struc-tural elements among the three p53 family member proteins(22). Our study isQ:18 unique in showing that APR-246 can partiallyrestore the molecular circuitry downstream of p63, which wasaffected in embryonic lineage commitment of EEC patient-de-rived cells. Four genes that were previously shown to be elevatedin WT mice but not in p63-null mice at E14.5 (GJA1, GJ6B,KRT14, KRTDAP) (19) were rescued by APR-246. Moreover,the level of expression of three p63-target genes (DLX5, DLX6,

and CDH3/P-Cadherin) that are known to be associated withEEC syndrome (20, 21) was significantly enhanced by APR-246.Interestingly, we found that APR-246 acted differently on theR204W and R304W mutants. Whereas both mutations arefound within the DNA-binding domain, R304W is located in theZn-binding pocket that is thought to affect the direct binding ofp63 to the DNA phosphate moiety, but R204W possibly causesmore drastic conformational changes and misfolding of the p63protein (5). As a matter of fact, Rökaeus et al., have also re-ported differential effects of APR-246 on the TAp63γR204Wand TAp63γR304W mutants: APR-246 caused cell death in thepresence of TAp63γR304W but induced mainly growth arrest incells with TAp63γR204W (22). It would be interesting to in-vestigate whether there is a difference between the binding ofAPR-246 to the two mutated p63 forms by in vitro assays. Onecould hypothesize that a putative binding site for APR-246 isless exposed in the p63R204W protein comparing to p63R304W,and that would explain the milder restoration of p63R204W ac-tivity, or that R204W-induced conformational changes are lesssensitive to APR-246 binding or APR-246-mediated refolding.Animal models are useful to confirm therapeutic function be-fore translating to patients. However, knock-in mice for EECmutations are not yet available and may not reproduce faithfullythe human pathology. Therefore, our study paves the way for fu-ture therapy of p63-related diseases that could be developed rel-atively fast, as APR-246 has already been tested in clinical trials inpatients with hematological malignancies or prostate cancer.

Materials and MethodsDerivation of iPSC. Dermal fibroblasts were isolated and amplified from skinbiopsy obtained after informed consent from one healthy individual and twoEEC patients (carrying p63R304W or p63R204W). Fibroblasts were reprog-rammed using a lentival polycistronic cassette expressing OCT4, SOX2, KLF4,and C-MYC or OCT4, SOX2, and KLF4 only (28), as previously described (29).

Pluripotency Characterization. Pluripotency of iPSC was determined by im-munofluorescence, real-time qRT-PCR and PluriTest analysis as described indetail in SI Materials and Methods and in ref. 29.

0.2

0.4

0.6

0.8

1.0

Krt

DapK14

GJB

6G

JA1

PCad

Dlx

5D

lx6

+/+ (-)

R304W/+(-)

R304W/+(+)

R204W/+(-)

R204W/+(+)

1.2

Rel

ativ

e ex

pres

sion

A

% o

f K14

+ ce

lls

10

20

30

C

Ave

rage

d ex

pres

sion

of

epi

thel

ial m

arke

rs

(% o

f iPS

CW

T )

B40

-APR-246:

R204W/+R304W/+- ++

+/+

20

4060

80

100120

-

* **

*

*

K14

GJB

6G

JA1

PCad

Dlx

5D

lx6

Krt

Dap K14

GJB

6G

JA1

PCad

Dlx

5D

lx6

Krt

Dap K14

GJB

6G

JA1

PCad

Dlx

5D

lx6

Krt

Dap K14

GJB

6G

JA1

PCad

Dlx

5D

lx6

Krt

Dap

-APR-246:

R204W/+R304W/+- ++

+/+-

Fig. 3. Rescue of corneal differentiation of EEC-iPSC by APR-246. The indicated iPSC lines were differentiated into corneal epithelial cells for 14 d in thepresence (+) or absence (−) of APR-246 treatment (as detailed in Materials and Methods). (A) Real-time PCR analysis showing the relative expression of theindicated epithelial transcripts. (B) Average of gene-expression data presented in A. Results show data obtained from two independent experiments ± SD. (C)Flow cytometry analysis for K14 expression. Results show average data obtained from three independent experiments ± SD. Asterisks indicate for statisticalsignificance (*P < 0.01; **P < 0.05) between APR-246 treated samples compared with control (untreated cells).

4 of 5 | www.pnas.org/cgi/doi/10.1073/pnas.1201753109 Shalom-Feuerstein et al.

373374375

376377378

379380381

382383384385

386387388

389390391

392393394395

396397398

399400401

402403404

405406407408

409410411

412413414

415416417418

419420421

422423424

425426427

428429430431

432433434

435436437

438439440

441442443

444445446447

448449450

451452453

454455456457

458459460

461462463

464465466

467468469470

471472473

474475476

477478479480

481482483

484485486

487488489

490491492493

494495496

In Vitro Differentiation Protocols. iPSC were differentiated in vitro as de-scribed in detail in SI Materials and Methods.

Epidermal Differentiation. iPSC were mechanically detached in small clumps(two to four colonies/six-well) and seeded on mitomycin-treated 3T3-G2feeders in medium [DMEM/F12 supplemented withQ:19 15% knockout serumreplacement (Invitrogen), 1 mM Glutamine, 100 μM β-Mercaptoethanol,100 μM nonessential amino acids]. Two days later, medium was switched toGreen medium (29). BMP-4 (25 ng/mL; Peprotech) and ascorbic acid (0.3 mM;Sigma), and SB431542 (10 μM; Tocris) were added from day 0 for up to 30 d.

Corneal Epithelial Differentiation. COFs were grown in fibroblast medium[DMEM(Invitrogen) andFCS (Invitrogen] andarrestedbymitomycin treatment(8 μg/mL; Sigma) for 3 h followed by washing with medium and incubationovernight. The next day, COFs were incubated with Green medium (detailedabove) and conditioned media was collected every day for up to 10 d and wasstored at −20 °C. To induce corneal epithelial differentiation, iPSC were

seeded on collagen IV-coated dishes (0.5 mg/mL; Sigma) in the presence ofconditionedmedia supplementedwith 0.5 nMBMP-4 for thefirst 3 d. APR-246treatment (20 μM) was initiated at day 3 to minimize toxic effects.

qRT-PCR, Immunofluorescence, and Flow Cytometry. Detailed procedures aregiven in SI Materials and Methods.

ACKNOWLEDGMENTS. This work has been partially supported by theEuropean Union 6th Framework Program within the EPISTEM IntegratedProject Grant LSHB-CT-2005-019067 (to D.A.); Agence Nationale de RechercheANR-08-GENOPAT-024-03 and ANR-erare2-SkinDev (to D.A.); Israel Ministryof Science and Technology Grant MOST3-6494 (to D.A.); “Poste Vert” fellow-ship of Institut National de la Santé et de la Recherche Médicale, “Chateau-briand” fellowship of the Embassy of France in Israel and EuropeanMolecularBiology Organization short-term fellowship (to R.S.-F.); a PhD fellowship fromthe Israeli ministry of integration (to L.S.); and an Q:20Else-Kröner Fresenius Stif-tung fellowship (to F.-J.M.).

1. Mills AA, et al. (1999) p63 is a p53 homologue required for limb and epidermalmorphogenesis. Nature 398:708–713.

2. Yang A, et al. (1999) p63 is essential for regenerative proliferation in limb, cranio-facial and epithelial development. Nature 398:714–718.

3. Rinne T, Brunner HG, van Bokhoven H (2007) p63-associated disorders. Cell Cycle 6:262–268.

4. Itin PH (2009) Rationale and background as basis for a new classification of the ec-todermal dysplasias. Am J Med Genet A 149A:1973–1976.

5. Celli J, et al. (1999) Heterozygous germline mutations in the p53 homolog p63 are thecause of EEC syndrome. Cell 99:143–153.

6. Browne G, et al. (2011) Differential altered stability and transcriptional activity ofΔNp63 mutants in distinct ectodermal dysplasias. J Cell Sci 124:2200–2207.

7. Di Iorio E, et al. (2012) Limbal stem cell deficiency and ocular phenotype in ec-trodactyly-ectodermal dysplasia-clefting syndrome caused by p63 mutations. Oph-thalmology 119(1):74–83.

8. McNab AA, Potts MJ, Welham RA (1989) The EEC syndrome and its ocular manifes-tations. Br J Ophthalmol 73:261–264.

9. Vanbokhoven H, Melino G, Candi E, Declercq W (2011) p63, a story of mice and men.J Invest Dermatol 131:1196–1207.

10. Rinne T, Hamel B, van Bokhoven H, Brunner HG (2006) Pattern of p63 mutations andtheir phenotypes—Update. Am J Med Genet A 140:1396–1406.

11. Williams R, Schuldt B, Müller FJ (2011) A guide to stem cell identification: Progressand challenges in system-wide predictive testing with complex biomarkers. Bioessays33:880–890.

12. Aberdam E, et al. (2008) A pure population of ectodermal cells derived from humanembryonic stem cells. Stem Cells 26:440–444.

13. Metallo CM, Ji L, de Pablo JJ, Palecek SP (2008) Retinoic acid and bone morphogeneticprotein signaling synergize to efficiently direct epithelial differentiation of humanembryonic stem cells. Stem Cells 26:372–380.

14. Guenou H, et al. (2009) Human embryonic stem-cell derivatives for full reconstructionof the pluristratified epidermis: A preclinical study. Lancet 374:1745–1753.

15. Xu RH, et al. (2002) BMP4 initiates human embryonic stem cell differentiation totrophoblast. Nat Biotechnol 20:1261–1264.

16. Smith JR, et al. (2008) Inhibition of Activin/Nodal signaling promotes specification ofhuman embryonic stem cells into neuroectoderm. Dev Biol 313:107–117.

17. Ahmad S, et al. (2007) Differentiation of human embryonic stem cells into cornealepithelial-like cells by in vitro replication of the corneal epithelial stem cell niche.Stem Cells 25:1145–1155.

18. Pellegrini G, et al. (2001) p63 identifies keratinocyte stem cells. Proc Natl Acad Sci USA98:3156–3161.

19. Shalom-Feuerstein R, et al. (2011) ΔNp63 is an ectodermal gatekeeper of epidermalmorphogenesis. Cell Death Differ 18:887–896.

20. Lo Iacono N, et al. (2008) Regulation of Dlx5 and Dlx6 gene expression by p63 is in-volved in EEC and SHFM congenital limb defects. Development 135:1377–1388.

21. Shimomura Y, Wajid M, Shapiro L, Christiano AM (2008) P-cadherin is a p63 targetgene with a crucial role in the developing human limb bud and hair follicle. De-velopment 135:743–753.

22. Rökaeus N, et al. (2010) PRIMA-1(MET)/APR-246 targets mutant forms of p53 familymembers p63 and p73. Oncogene 29:6442–6451.

23. Rostagno P, et al. (2010) Embryonic stem cells as an ectodermal cellular model ofhuman p63-related dysplasia syndromes. Biochem Biophys Res Commun 395:131–135.

24. Medawar A, et al. (2008) DeltaNp63 is essential for epidermal commitment of em-bryonic stem cells. PLoS ONE 3:e3441.

25. Coraux C, et al. (2003) Reconstituted skin from murine embryonic stem cells. Curr Biol13:849–853.

26. Itoh M, Kiuru M, Cairo MS, Christiano AM (2011) Generation of keratinocytes fromnormal and recessive dystrophic epidermolysis bullosa-induced pluripotent stem cells.Proc Natl Acad Sci USA 108:8797–8802.

27. Bykov VJ, et al. (2002) Restoration of the tumor suppressor function to mutant p53 bya low-molecular-weight compound. Nat Med 8:282–288.

28. Somers A, et al. (2010) Generation of transgene-free lung disease-specific humaninduced pluripotent stem cells using a single excisable lentiviral stem cell cassette.Stem Cells 28:1728–1740.

29. Petit I, et al. (2012) Induced pluripotent stem cells from hair follicles as a cellularmodel for neurodevelopmental disorders. Stem Cell Res (Amst) 8:134–140.

Shalom-Feuerstein et al. PNAS Early Edition | 5 of 5

CELL

BIOLO

GY

497498499

500501502

503504505

506507508509

510511512

513514515

516517518519

520521522

523524525

526527528

529530531532

533534535

536537538

539540541542

543544545

546547548

549550551

552553554555

556557558

559560561

562563564

565566567

568569570571

572573574

575576577

578

586587

588589590

591592593594

595596597

598599600

601602603604

605606607

608609610

611612613

614615616617

618619620

Q: 1_Please contact PNAS_Specialist.djs@sheridan.com if you have questions about the editorial

changes, this list of queries, or the figures in your article. Please include your manuscript number in

the subject line of all e-mail correspondence; your manuscript number is 201201753.

Q: 2_Please (i) review the author affiliation and footnote symbols carefully, (ii) check the order of the

author names, and (iii) check the spelling of all author names, initials, and affiliations. Please check

with your coauthors about how they want their names and affiliations to appear. To confirm that the

author and affiliation lines are correct, add the comment “OK” next to the author line. This is your

final opportunity to correct any errors prior to publication. Misspelled names or missing initials will

affect an author’s searchability. Once a manuscript publishes online, any corrections (if approved)

will require publishing an erratum; there is a processing fee for approved erratum.

Q: 3_Please review and confirm your approval of the short title: APR-246 rescues differentiation of EEC-

iPSC. If you wish to make further changes, please adhere to the 50-character limit.

Q: 4_PNAS requires all nonstandard abbreviations in article titles to be spelled out. As such, “APR-246,”

“EEC” and “PRIMA-1” are nonstandard and will need to be spelled out. If this is not possible,

please provide a modifier for this term that identifies what the term is (for example, ‘protein kinase

_____’).

Q: 5_If your article contains links to Web sites (other than the SI links for your article), please verify that

the links are valid and will direct readers to the proper Web page.

Q: 6_Author names may have been edited to match those provided during article submission; please check

carefully and note your approval in the margin. (Your article cannot be published until your approval

has been received.)

Q: 7_Please provide departmental affiliations for affiliations a, b, and c, and postal codes for affiliations a-

d.

Q: 8_Please expand ISERTECH in affiliation b.

Q: 9_PNAS italicizes the names of genes and alleles. Please check throughout the manuscript and correct

as necessary. If, by “XXX gene,” you mean “the gene that encodes protein XXX,” then italic type is

not necessary. (Note: If all instances of a gene/allele name should be changed, please make only

one correction and simply indicate that it should be made throughout the paper.)

Q: 10_PNAS does not allow statements of novelty or priority. As such, the word "novel" has been changed

to "unique" in the sentence beginning "In this study . . ." in the Abstract. Please approve edit.

Q: 11_PNAS allows up to five key terms that (i) do not repeat terms present in the TITLE ORABSTRACT

(which are searchable online) and (ii) do not include nonstandard abbreviations. As such, all terms

except "cornea" have been removed. Because we prefer to avoid publishing a single key term,

please add at least one more term (and specify your preferred order in which the terms should

appear) or delete the key terms entirely.

AUTHOR QUERIES

AUTHOR PLEASE ANSWER ALL QUERIES 1

Q: 12_Please review the information in the author contribution footnote carefully. Please make sure that

the information is correct and that the correct author initials are listed. Note that the order of author

initials matches the order of the author line per journal style. Youmay add contributions to the list in

the footnote; however, funding should not be an author’s only contribution to the work.

Q: 13_You will receive a notification from the PNAS eBill system in 1-2 days. Each corresponding author

is required to log in to the system and provide payment information for applicable publication

charges (purchase order number or credit card information) upon receipt of the notification. You

will have the opportunity to order reprints through the eBill system if desired, as well. Failure to log

in and provide the required information may result in publication delays.

Q: 14_Reminder: You have chosen not to pay an additional $1300 (or $975 if your institution has a site

license) for the PNAS Open Access option.

Q: 15_Please verify that all supporting information (SI) citations are correct. Note, however, that the

hyperlinks for SI citations will not work until the article is published online. In addition, SI that is

not composed in the main SI PDF (appendices, datasets, movies, and “Other Supporting

Information Files”) have not been changed from your originally submitted file and so are not

included in this set of proofs. The proofs for any composed portion of your SI are included in this

proof as subsequent pages following the last page of the main text. If you did not receive the

proofs for your SI, please contact PNAS_Specialist.djs@Sheridan.com.

Q: 16_quantitative PCR, quantitative RT-PCR, qPCR, and qRT-PCR appear to be used interchangibly

throughout the text. They have been changed to qRT-PCR for consistency throughout, with "real-

time" being added as applicable. Please check and amend as necessary.

Q: 17_Please expand CGHa in the sentence beginning "We found that . . ." Should this be CGHa?

Q: 18_PNAS does not allow statements of novelty or priority. As such, the phrase "Our study shows for the

first time . . ." has been changed to "Our study is unique in showing . . ." Please approve edit.

Q: 19_Throughout, for concentrations .1%, please state basis (eg, vol/vol, vol/wt, etc).

Q: 20_PNAS does not allow acronyms in the Acknowledgments. Please expand EPISTEM in this section,

if applicable, and check the section as amended (see specifically funding to D.A.).

Q: 21_Please provide a value for the scale bar in Fig. 1A.

Q: 22_Please provide the magnification used in Fig. 2 B and D.

AUTHOR QUERIES

AUTHOR PLEASE ANSWER ALL QUERIES 2

Supporting InformationShalom-Feuerstein et al. 10.1073/pnas.1201753109SI Materials and MethodsPluripotency Characterization. Expression of pluripotency markerswere determined by immunofluorescence using octamer-bindingtranscription factor 4 (OCT4) antibodies (Stemgent) and TRA-1-81 antibodies (Santa Cruz Biotechnology) and by real-time quan-titative PCR (qRT-PCR), as described below. Alkaline phospha-tase activity was revealed after 4% paraformaldehyde fixation byincubation with 0.01% Naphtol (Sigma) and 0.5 mg/mL Fast Blue(Sigma) in 100 mM Tris-HCL and 20 mM MgCl2.For whole-genome microarray and PluriTest analysis, we fol-

lowed the same procedures as previously described (1, 2). RNAQ:1

was isolated from two biological replicates per cell line (1 × 106

cells per sample) with the mirVana RNA isolation kit (Ambion).Illumina HT12v3 microarrays were hybridized following the man-ufacturers instructions and as previously described (1). The re-sulting raw data were processed with the PluriTest algorithm (1) fortesting pluripotent features in induced pluripotent stem cell(iPSC) lines. Genetic integrity was evaluated by G-banded kar-yotype analysis.

In Vitro Differentiation Protocols. iPSC colonies were cultured insuspension in Petri dishes in differentiation medium (DMEMsupplemented with 10% FBS, 1 mM Glutamine,100 μM nones-sential amino acids, 100 μM β-mercaptoethanol) to allow theformation of . Embryoid bodies were plated on day 7 on gelatin-coated plates and further cultured until day 20. Directed tro-phoectoderm differentiation was performed by plating iPSC onmatrigel in iPSC medium previously conditioned 24 h on mouseembryonic fibroblasts. The following day, medium was switchedto serum-free medium [DMEM/F12 supplemented with 1 mMglutamine, 100 μM nonessential amino acids, 1× N2 and 1× B27

(Invitrogen)] with bone morphogenetic protein-4 (BMP-4) (25ng/mL). Cells were collected 6 d later.

qRT-PCR. RNA was extracted using Aurum kit and cDNA wassynthetized from 1 μg RNA using the iscript kit (Bio-Rad). qRT-PCR was performed in duplicate using SYBR Green (Bioline)and specific primers (sequences available upon request). Eachreaction contained 12.5 μL SYBR-Green PCR Master Mix, 5 μLcDNA and 5 μL primer mix (0.5 μM), adjusted to 25 μL reactionvolume. The value of each reaction was normalized to GAPDHand the relative expression of each transcript was calculated as afold-change relative to control sample of undifferentiated cells.

Immunofluorescence.Differentiated cells thatwere grownon cover-slips were fixed in cold methanol for 20 min. Blocking was for 20min [2.5% BSA (Sigma), 2.5% naive Donkey serum (JacksonLaboratories)], primary and secondary antibodies were applied for45min followed by anti-fadeDAPImounting-gel fixation (Sigma).The following primary antibodies were used: rabbit anti-K14(1:200) (Covance), rabbit anti-K5 (1:200) (Covance), mouse anti-K18 (1:200) [Chemicon (Millipore)], rabbit anti-pax6 (1:100)(Chemicon), mouse anti-K3(1:100) (Millipore), mouse anti-p63(1:100) (Santa Cruz), and mouse anti–E-Cadherin (1:100) (R&DSystems).

Flow Cytometry. Cells were fixed in 2% paraformaldehyde for 20min at room temperature. After washing with PBS, cells wereincubated with 0.5% BSA, 0.5% saponin, and donkey serumfor 30 min. Primary antibodies for K14 (Millipore) and K18(Chemicon) were added for 45 min. Acquisition was performedon FACSCalibur using CellQuest software (BD Biosciences).

1. Rinne T, Hamel B, van Bokhoven H, Brunner HG (2006) Pattern of p63 mutations andtheir phenotypes—Update. Am J Med Genet A 140:1396–1406.

2. Petit I, et al. (2012) Induced pluripotent stem cells from hair follicles as a cellular modelfor neurodevelopmental disorders. Stem Cell Res (Amst) 8:134–140.

Shalom-Feuerstein et al. www.pnas.org/cgi/content/short/1201753109 1 of 4

123

456

789

10111213

141516

171819

20212223

242526

272829

303132

33343536

373839

404142

43444546

474849

505152

535455

56575859

606162

636465

666768

697071

72737475

767778

798081

82838485

868788

899091

929394

95969798

99100101

102103104

105106107108

109110111

112113114

115116117

118119120121

122123124

1

10

100

1000

10000

+/+ R304W/+ R204W/+

Rela

�ve

expr

essi

on

E

C D

1

10

100

1000

+/+

Rel

ativ

e ex

pres

sion

AFP CD31 NCAM1

0.0001

0.001

0.01

0.1

1

10

Oct4 TRA-1-80 APphaseA B

ES +/+

R30

4W/+

R20

4W/+

+/+

Oct4 Sox2

Dnmt3bNanog

Rel

ativ

e ex

pres

sion

Fibro

Fig. S1. Characterization of iPSC from EEC patients. (A) Morphology of iPSC+/+, iPSCR204W/+, and iPSCR304W/+ cells and expression of the pluripotent markersOct4, Tra-1-80, and alkaline phosphatase (AP). Immunofluorescence staining included DAPI staining (blue). (Scale bars, 100 mm.) (B) qRT-PCR for the plurip-otent markers Oct4, sex-determining region Y box-2 (Sox2), Nanog, and DNA methyltransferase 3b (Dnmt3b). hES and fibroblasts were included as positive andnegative controls, respectively. (C) Pluripotent transcriptional profile measured in PluriTest assay. PluriTest results are plotted in density distribution forpreviously referenced pluripotent cells (red cloud) and somatic cells (blue cloud). (D) In vitro differentiation of iPSC lines in embryoid bodies. iPSC were culturedin suspension for 7 d then plated on gelatin. At day 20 cells were analyzed by qRT-PCR for the germ layer markers AFP (endoderm), CD31 (mesoderm), andNCAM1 (neuroectoderm). Results show relative expression compared with undifferentiated iPSC. (E) Directed in vitro differentiation of iPSC toward troph-oectoderm. iPSC+/+, iPSCR204W/+, and iPSCR304W/+ cells were seeded on matrigel and cultured in presence of BMP-4 for 6 d. Expression of the trophoectodermmarker CDX2 was evaluated by qRT-PCR and results were expressed relative to CDX2 expression in undifferentiated iPSC. Results show average ± SE from twoindependent iPSC clones.

Shalom-Feuerstein et al. www.pnas.org/cgi/content/short/1201753109 2 of 4

125126127

128129130

131132133

134135136137

138139140

141142143

144145146147

148149150

151152153

154155156

157158159160

161162163

164165166

167168169170

171172173

174175176

177178179

180181182183

184185186

187188189

190191192

193194195

196197198199

200201202

203204205

206207208209

210211212

213214215

216217218

219220221222

223224225

226227228

229230231232

233234235

236237238

239240241

242243244245

246247248

C

0.001

0.01

0.1

1

10

100

1000

10000

Oct4 K14

A

0

20

40

60

80

100

CDX2 CGHa

BMP4 BMP4+SB431542

B

Rela

�ve

expr

essi

on

Rela

�ve

expr

essi

on

BMP4

BMP4+SB431542

i ii iii

iv v vi

K14/dapi K10/dapi

Fig. S2. Epidermal differentiation of iPSC by BMP-4 and TGF-β inhibition. iPSC+/+ were subjected to epidermal differentiation in presence of BMP-4 alone orwith BMP-4 and the TGF-β inhibitor SB431542. (A) Phase pictures at day 7. (B) qRT-PCR at day 10 for Oct4, K14, and the extraembryonic markers CDX2 andCGHa. (C) After 30 d of differentiation in presence of BMP-4 and SB431542, cells acquire a keratinocyte-like morphology (i) and highly express K14 (ii). Somemature K10+ cells are present (iii) and spontaneous stratification can be observed (iv and v) (Dashed lines outline stratified area in iv shows focus on thestratified keratinocytes, in v shows focus on the rest of the keratinocytes). Picked cells form typical keratinocyte colonies (vi). (Scare bars, 100 μm.)

Day 14Day 10Day 7

+/+

R204

W/+

R304

W/+

K14/p63

K14/p63

K14/p63

K14/p63

K14/p63

K14/p63

K14/p63

K14/p63

K14/p63

A

+/+

R30

4W/+

K14

K18

FSC

2.2%27.5%

42.5% 46%+/+ R304W/+B

C

Fig. S3. Impaired epidermal differentiation of EEC-iPSC. iPSC+/+, iPSCR204W/+, and iPSCR304W/+ cells were subjected to epidermal differentiation protocol inpresence of BMP-4 and SB431542. (A) Phase pictures of iPSC+/+ and iPSCR204W/+ after 10 and 30 d of differentiation. (B) Flow cytometry analysis of iPSC+/+ andiPSCEEC after 25 d of differentiation. The data show one representative experiment. (C) Immunofluorescence staining for K14 and p63 during epidermaldifferentiation of iPSC+/+ and iPSCEEC. (Scale bars, 100 μm.)

Shalom-Feuerstein et al. www.pnas.org/cgi/content/short/1201753109 3 of 4

249250251

252253254

255256257

258259260261

262263264

265266267

268269270271

272273274

275276277

278279280

281282283284

285286287

288289290

291292293294

295296297

298299300

301302303

304305306307

308309310

311312313

314315316

317318319

320321322323

324325326

327328329

330331332333

334335336

337338339

340341342

343344345346

347348349

350351352

353354355356

357358359

360361362

363364365

366367368369

370371372

0.2

0.4

0.6

0.8

1.0

K14

GJB

6

GJA

1

PCad

Dlx

5

Dlx

6

1.2

Rel

ativ

e ex

pres

sion

1.4

Krt

Dap

ControlPRIMA

Fig. S4. APR-246 effect on corneal epithelial differentiation of iPSC+/+cells. iPSC+/+ were differentiated into corneal epithelial cells for 10 d in the presence(PRIMA) or absence (Control) of APR-246 treatment, as detailed in Materials and Methods. Real-time PCR analysis is showing the relative expression of theindicated epithelial transcripts. Results show average data obtained from three independent experiments ± SD.

0

2

4

6

Control

PRIMA

+/+ R304W/+ R204W/+

B

Perc

enta

ge o

f ki

67-p

ositi

ve c

ells

0

5000

10000

15000

20000

25000

30000

35000

Control

PRIMA

+/+ R304W/+ R204W/+ +/+ R304W/+ R204W/+

Day 7 Day 14

A

Num

ber

of c

ells

per

cm

2

Fig. S5. APR-246 effect on proliferation during corneal epithelial differentiation of EEC-iPSC cells. The indicated iPSC lines were differentiated into cornealepithelial fate in the presence (PRIMA) or absence (Control) of APR-246 treatment, as detailed in Materials and Methods. Cells were at counted at day 7 andday 14 of differentiation. The averaged number of cells per square centimeter is shown in A. (B) Immunofluorescent staining of ki67 was performed at day 7 ofdifferentiation followed by quantification of ki67+ cells. Data represents the percentage of ki67+ cells in the whole population (determined by DAPI staining).Five different fields were randomly pictured and counted for each sample. Results show average data obtained from two independent experiments ± SD.

Shalom-Feuerstein et al. www.pnas.org/cgi/content/short/1201753109 4 of 4

373374375

376377378

379380381

382383384385

386387388

389390391

392393394395

396397398

399400401

402403404

405406407408

409410411

412413414

415416417418

419420421

422423424

425426427

428429430431

432433434

435436437

438439440

441442443

444445446447

448449450

451452453

454455456457

458459460

461462463

464465466

467468469470

471472473

474475476

477478479480

481482483

484485486

487488489

490491492493

494495496

Q: 1_PNAS requires that any references cited in the SI text generate a separate reference list. As such, refs.

10 and 29 have been renumbered as SI refs. 1 and 2, respectively, and a separate reference list has

been generated for the SI text. Please check.

AUTHOR QUERIES

AUTHOR PLEASE ANSWER ALL QUERIES