Corrections

CELL BIOLOGYCorrection for “Role of p63 and the Notch pathway in cochleadevelopment and sensorineural deafness,” by Alessandro Terrinoni,Valeria Serra, Ernesto Bruno, Andreas Strasser, Elizabeth Valente,Elsa R. Flores, Hans van Bokhoven, Xin Lu, Richard A. Knight,and Gerry Melino, which appeared in issue 18, April 30, 2013, ofProc Natl Acad Sci USA (110:7300–7305; first published April 15,2013; 10.1073/pnas.1214498110).The authors note that “Thanks to an alert reader, we no-

ticed that in Fig. 3D, the control ChIP for MDM2 errone-

ously duplicated a p53-RE-III panel from an earlier paper(1). We thank the reader for bringing this issue to our at-tention, and we deeply apologize to the scientific community forthe error.”The corrected figure and its corrected legend appear below.

1. Tucci P, et al. (2012) Loss of p63 and its microRNA-205 target results in enhancedcell migration and metastasis in prostate cancer. Proc Natl Acad Sci USA 109(38):15312–15317.

www.pnas.org/cgi/doi/10.1073/pnas.1324224111

200 1000 2000 2200 2400 2600 2800 3000 3200 3400

RE PromoterA

Input HA IgG

MDM2

Hes5

SaOs-2 TAp63αD

Hes5 gene

Input HA IgGSaOs-2 TAp63α

F

Doxy

HA

Tub

SaOs-2 HA-TAp63αG

Atoh1

p21

H

HA

Actin

p21

C

E

3000 4200 7200 Enhancer A Enhancer B

RE II 3’Promoter RE I 5’

Atoh1 mRNA transcript

0

1

2

3

4

5

6

7 Luciferase activity

Fold

ove

r con

trol

pcDNA-H

A

TAp6

3a W

t

G530V

I537T

Q634X

R279H

R280C

R304W

S272N

Q566fs

X94

SAM DBD

TID

B Luciferase activity

SAM DBD

TID

35

30

25

20

15

10

5

0

Fold

over

contr

ol

40

45

pcDNA-H

A

TAp6

3a W

t

G530V

I537T

Q634X

R279H

R280C

R304W

S272N

Q566fs

X94

Q536L

- +

DoxySaOs-2 HA-TAp63α

- +

Fig. 3. p63 drives Hes5 and Atoh1 promoters. (A) The Hes5 gene structure shows the presence of a putative p53/p63 RE localized at −988, −966 from the TSS,in the promoter sequence. All promoters analyses were performed by MathInspector professional release 8.0.5, March 2011; Matrix Family Library Version 8.3,October 2010. (B) SaOs2 cells were transiently cotransfected with expression constructs for WT TAp63α-HA or TAp63α-HA mutant (G530V, I537T, Q536L,R280C, R304W, S272N, R279H, Q566fsX94, and Q634X) plus the hHes5-luc reporter vector. There was an increase in luciferase activity in cells transfected withTAp63α-HA WT, more pronounced with the TAp63α-Q634X-HA mutant, but not in cells transduced with the vectors encoding mutants for the DBD (mean ±SD, n = 3). (C) Western blot analysis was performed to verify TAp63a protein expression. (D) ChIP analysis shows the binding of TAp63α to the p53/p63-RE inthe Hes5 promoter. In the Lower panel, the ChIP on the MDM2 promoter used as positive control. (E) The Atoh1 gene structure shows the presence of two p53RE: RE-I5′ localized at −1,682, −1,660 from the TSS; and RE-II3′ in the Enhancer-A. (F) ChIP analysis of the Atoh1 p53-Res. TAp63α binds only to the p53-REAtoh1 enhancer sequence; the MDM2 promoter was used as a positive control. (G) Western blot showing TAp63α protein expression. (H) Luciferase activitywas increased by TAp63α-HA and TAp63α Q634X plasmids. SaOs2 cells were transiently cotransfected with an hAtoh1-luc expression vector, TAp63α-HA, andTAp63α-HA mutants (G530V, I537T, R280C, R304W, S272N, R279H, Q566fsX94, and Q634X) (mean ± SD, n = 3).

2854–2855 | PNAS | February 18, 2014 | vol. 111 | no. 7 www.pnas.org

CELL BIOLOGYCorrection for “Loss of p63 and its microRNA-205 target resultsin enhanced cell migration and metastasis in prostate cancer,” byPaola Tucci, Massimiliano Agostini, Francesca Grespi, Elke K.Markert, Alessandro Terrinoni, Karen H. Vousden, Patricia A. J.Muller, Volker Dötsch, Sebastian Kehrloesser, Berna S. Sayan,Giuseppe Giaccone, Scott W. Lowe, Nozomi Takahashi, PeterVandenabeele, Richard A. Knight, Arnold J. Levine, and Gerry

Melino, which appeared in issue 38, September 18, 2012, ofProc Natl Acad Sci USA (109:15312–15317; first publishedSeptember 4, 2012; 10.1073/pnas.1110977109).The authors note that “The y-axis label of Fig. 5C is incorrect.

Instead of ‘Biochemical Recurrence,’ it should read ‘Biochemical-Free Recurrence.’ We apologize to readers for the erroneous la-beling.” The corrected figure and its corrected legend appear below.

www.pnas.org/cgi/doi/10.1073/pnas.1324223111

Ctrl Scr miR-205 Np630

20

40

60

Num

ber o

f met

asta

sis

*** ***

TIME (months)

D

TIME (months)

C

PR

OB

AB

ILIT

Y

(Bio

chem

ical

-Fre

e R

ecur

renc

e, c

enso

red)

Np63/miR205 function Np63/miR205 intermediate Np63/miR205 loss

PR

OB

AB

ILIT

Y (O

vera

ll Fo

llow

-up

time)

A

B E

Fig. 5. In human prostate cancer, loss of ΔNp63 and miR-205 associates with invasive phenotype and poor clinical outcome. Tumor and normal prostatesamples clustered into three groups reflect the activity of the ΔNp63–miR-205 complex. Groups were indicative of metastatic and invasive behavior as well asclinical prognosis. (C) Time to biochemical-free recurrence for the three groups. (D) Clinical follow-up in the cohort was recorded over 5 y; all data arecensored for survival. Kaplan–Meier analysis with suppressed censoring shows a significant trend within this follow-up time toward poor survival in theΔNp63–miR-205 loss group. (A) miR-205/ΔNp63 relationship. Average expression of ΔNp63 was calculated from ΔNp63-specific probes and compared with miR-205 expression. Metastatic tumor samples are indicated in red diamonds, primary tumors in black squares, and normal samples in blue circles. Pearsoncorrelation was calculated together with the significance of the correlation. Notable is the clean separation between metastatic and normal samples. (B)Heatmap illustrating miR-205 and ΔNp63 expression in prostate cancer. Data show (i) clustering of samples by their ΔNp63/miR-205 expression, Middle bars;(ii) association between normal and ΔNp63/miR-205 expression as well as metastasis and ΔNp63/miR-205 loss, Top bars; and (iii) association between the EMTsignature and the expression of the ΔNp63/miR-205 axis, Bottom bars. Middle bars (clustering): Samples were clustered by their miR-205/ΔNp63 expression(red indicates significant overexpression and blue, significant underexpression of the gene/miR, true color: P ≤ 1.0E-05). This determined one group of sampleswith an active miR-205/ΔNp63 axis (“function”), one group with a clear loss of expression (“loss”), and one intermediate group (“intermediate”) as indicatedin the clustering bar. Top bars (primary, metastasis, normal): The group exhibiting loss of miR-205/ΔNp63 expression was enriched in metastatic samples,whereas the miR-205/ΔNp63 function group was enriched in normal tissue samples (P < 1.0E-06; compare also Fig. S8 A–C). Bottom bars: The miR-205/ΔNp63-loss group was also associated with an EMT transcriptional profile (P < 0.01, compare also Fig. S8D). Signature scores (Sarrio bars) for experimentally derivedsignatures according to Sarrio et al. (27) characterizing EMT are shown. Red indicates significant positive, and blue indicates significant negative association ofa sample with a signature. P value for enrichment of the cluster loss with the signature EMT_up is 0.0014. Full statistics are in Fig. S8D. (E) Lung metastasis innude mice. A total of 1.5 × 106 scrambled control–PC3–Tet–On (12 mice), ΔNp63α–PC3–Tet–On (10 mice), or miR-205–PC3–Tet–On (10 mice) cells were injectedthrough the tail vein of BALB/c nude male mice. ΔNp63α and miR-205 expression was induced with doxycycline through their drinking water. Animals werekilled after 3 wk and total number of lung metastases was counted using a stereomicroscope. ***P < 0.001. See also Fig. S9.

PNAS | February 18, 2014 | vol. 111 | no. 7 | 2855

CORR

ECTIONS

Role of p63 and the Notch pathway in cochleadevelopment and sensorineural deafnessAlessandro Terrinonia,1,2, Valeria Serraa,1, Ernesto Brunob, Andreas Strasserc,d, Elizabeth Valentec,d, Elsa R. Florese,Hans van Bokhovenf, Xin Lug, Richard A. Knighth, and Gerry Melinoa,h,2

aBiochemistry Laboratory Istituto Dermopatico Dell’Immacolata, c/o Department of Experimental Medicine and Surgery, University of Rome “Tor Vergata,”00133 Rome, Italy; bDepartment of Clinical Sciences and Translational Medicine, University of Rome “Tor Vergata,” 00133 Rome, Italy; cThe Walter and ElizaHall Institute of Medical Research, Parkville, VIC 3052, Australia; dDepartment of Medical Biology, Melbourne University, Parkville, VIC 3052, Australia;eDivision of Basic Science Research, Department of Biochemistry and Molecular Biology, University of Texas M. D. Anderson Cancer Center, Houston, TX 77030;fDepartment of Human Genetics, Nijmegen Centre for Molecular Life Sciences, Radboud University Nijmegen Medical Centre, 6500 HB, Nijmegen, TheNetherlands; gNuffield Department of Clinical Medicine, Ludwig Institute for Cancer Research, University of Oxford, Oxford OX3 7DQ, United Kingdom; andhToxicology Unit, Medical Research Council, Leicester University, Leicester LE1 9HN, United Kingdom

Edited by Michael Karin, University of California, San Diego School of Medicine, La Jolla, CA, and approved March 20, 2013 (received for reviewAugust 22, 2012)

The ectodermal dysplasias are a group of inherited autosomaldominant syndromes associated with heterozygous mutations inthe Tumor Protein p63 (TRP63) gene. Here we show that, in additionto their epidermal pathology, a proportion of these patients havedistinct levels of deafness. Accordingly, p63 null mouse embryosshowmarked cochlea abnormalities, and the transactivating isoformof p63 (TAp63) protein is normally found in the organ of Corti. TAp63transactivates hairy and enhancer of split 5 (Hes5) and atonal homo-log 1 (Atoh1), components of the Notch pathway, known to be in-volved in cochlear neuroepithelial development. Strikingly, p63 nullmice show morphological defects of the organ of Corti, with super-numerary hair cells, as also reported for Hes5 null mice. This pheno-type is related to loss of a differentiation property of TAp63 and notto loss of its proapoptotic function, because cochleas in mice lackingthe critical Bcl-2 homology domain (BH-3) inducers of p53- and p63-mediated apoptosis—Puma, Noxa, or both—are normal. Collectively,these data demonstrate that TAp63, acting via the Notch pathway, iscrucial for the development of the organ of Corti, providing a molec-ular explanation for the sensorineural deafness in ectodermal dys-plasia patients with TRP63 mutations.

epidermis | cell death | p53 family

Tumor protein p63 (TRP63) is the most ancient member of thep53 family of transcription factors (1) and acts as a keymolecule

in embryonic development. The structural organization of p63 (2, 3)is similar to p53, containing a transactivation domain (TA), DNAbinding domain (DBD), and an oligomerization domain. The ex-pression of p63 is regulated by two distinct promoters giving rise toproteins that either contain (TAp63) or do not contain (ΔNp63) theN-terminal TA domain, which is critical for the transcriptional in-duction of particular subsets of p63 target genes (2). Both p63isoforms give rise to differentially spliced mRNAs and proteins,with at least six p63 isoforms (α, β, γ) recognized (1). p63α alsocontains a sterile α motif (SAM) and a carboxy terminal-inhibitorydomain (TID) (4), both absent in p53 (5). The TID binds andinhibits the N-terminal TA domain by masking important N-ter-minal residues. The importance of this regulatory domain is evidentfrom the identification of mutations in the TAp63α-C terminus inhuman patients (6).The ectodermal dysplasia (ED) syndromes are a large and het-

erogeneous group of inherited human diseases that are character-ized by developmental abnormalities of ectodermally derivedstructures. A large subset of autosomal dominant ED syndromesare caused by heterozygous mutations in the TRP63 gene (7). Theprototypic Ectrodactyly, Ectodermal dysplasia, and Cleft lip/palate(EEC) syndrome (Online Mendelian Inheritance in Man: OMIM604292) has highly variable expression and penetrance. Clinically,EEC patients show ectodermal dysplasia affecting skin, hair, nailsand teeth, and facial clefts, as well as frequent lacrimal duct

abnormalities, urogenital problems, facial dysmorphism, and hearingloss. Nucleotide sequence analyses have provided evidence fora striking genotype–phenotype correlation with mutations in in-dividual domains of p63 in ED patients (8, 9) The SAM domainseems to be particularly important for skin development, whereasthe DBD and TID are crucial for limb development. In general,mutations in EEC patients show substitutions in the DBD thatimpair p63’s ability to bind to specific target sequences in DNA.Mice lacking both copies of the Trp63 gene are born lacking

limbs, skin, and skin appendages, such as hair shafts, follicles, andsebaceous glands, and consequently die from dehydration shortlyafter birth (10, 11). ΔNp63 is the predominant isoform in the basallayer of the epidermis (12, 13) and is crucial for the maintenance ofepithelial stem cells (10, 11). Accordingly, reconstitution withΔNp63 at least partially rescues the epidermal defects seen in p63−/−

mice (12). In EEC patients, ΔNp63 mutant proteins accumulate,and these proteins generally show reduced transcriptional activityon skin-specific promoters (compared with TAp63 proteins), al-though in other ED syndromes ΔNp63 mutant proteins are tran-scriptionally more active than the WT protein.TAp63 also exerts critical functions in the development and

function of the heart (14) and oocytes (15, 16). The latter suggestsa role for this gene in female infertility (17). Only limited infor-mation is available for the role of p63 in development and main-tenance of other tissues/organs. Although the relationship betweenΔNp63 mutations and ED is consistent with the epidermal defectsseen in the p63−/− mice, the role of mutant p63 isoforms in theconductive and sensorineural deafness seen in some patients (8, 18)is unclear. Because a deaf EEC patient presented with abnormal-ities in themorphology of the cochlea (18), we have investigated therole of p63 in the development of the inner ear and, in particular,the organ of Corti. Our studies, using gene expression analysis andmorphological examination of gene-targeted mice, demonstratethat TAp63, through activation of the Notch signaling pathway, isimportant for normal development of the cochlea.

ResultsDeafness in ED Patients and Cochlea Morphology in p63-Null Mice. Toevaluate the incidence of deafness among ED patients with p63

Author contributions: A.T. and G.M. designed research; A.T., V.S., E.B., and H.v.B. per-formed research; A.S., E.V., E.R.F., and X.L. contributed new reagents/analytic tools; A.T.,A.S., H.v.B., R.A.K., and G.M. analyzed data; and A.T., R.A.K., and G.M. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.1A.T. and V.S. contributed equally to this work.2To whom correspondence may be addressed. E-mail: alessandro.terrinoni@uniroma2.itor melino@uniroma2.it.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1214498110/-/DCSupplemental.

7300–7305 | PNAS | April 30, 2013 | vol. 110 | no. 18 www.pnas.org/cgi/doi/10.1073/pnas.1214498110

mutations, we analyzed a cohort of 112 patients: 17%of these wereaffected by deafness, albeit to different extents. Table 1 summa-rizes the relative incidence of deafness for each ED subgroup.Additional analysis on two new patients revealed by audiometrictests a mild hearing loss, both conductive and sensorineural (Fig.S1 A and B). Because the p63-deficient mice phenocopy to a sub-stantial extent the symptoms observed in ED patients, we in-vestigated whether these mice show cochlear abnormalities,consistent with the human deafness phenotype. To this end, wecompared p63−/− and WT embryos from embryonic day (E)14.5and E18.5 (adult p63−/− mice could not be investigated, becausethe pups die a few hours after birth). H&E staining of sectionsfrom embryos showed no structural abnormalities in the cochlea ofp63−/− E14.5 embryos (Fig. S2 A and B). However, morphologicaldefects were clearly detected in the cochlear duct (scala media),

scala tympani, and scala vestibuli (Fig. 1 A and B) in E18.5p63−/− embryos.Development of the cochlea during embryogenesis occurs from

prosensory cells at E14.5, which leads to the formation of bothsupporting cells as well as inner (IHC) and outer hair cells (OHC).At the molecular level, Sex determining region Y-box 2 (Sox2) isexpressed in supporting cells, whereas Myo-VIIa is found in IHCand OHC. In p63−/− E14.5 embryos, we identified Sox2 (Fig. S2 Cand F), Myosin-VIIa (Fig. S2 I and K), and Connexin 26 (Cx26),a cell-specific marker of the stria vascularis (Fig. S2 D and G) ingerminative prosensory cells and in the organ of Kolliker.WhereasSox2 and Myo-VIIa appeared to be normally expressed at thisstage, Cx26 showed a slightly altered distribution compared withE14.5 WT embryos (Fig. S2 D and G).At E18.5, Sox2 expression was detected in supporting cells, at

the base of the organ of Corti (Fig. S3 A and D) in both WT and

Table 1. Incidence of deafness in EEC-like patients

p63-associated syndrome Patients affected, % (n) Trp63 mutations

EEC syndrome 7 (11/86) DBD: R204, S272, R279, R304, D312, and ins 3′ end.RHS syndrome 20 (2/10) TID: S541 and frameshift.AEC syndrome 38 (6/16) SAM: L514, C522, G530, I537, and frameshiftTotal EEC-Like 17 (19/112)

AEC, Ankyloblepharon Ectodermal dysplasia Cleft lip/palate; RHS, Rapp-Hodgkin Syndrome.

18.5 E WT 18.5 E p63 -/-

TAp63α

ΔNp63α

Actin

M Cochlea Skin

18.5 E WT

MyoVIIaCx26

18.5 E p63 -/- 18.5 E TAp63 -/-

2,5

2

1,5

1

Wt-LWt-w TAp63-/- -L TAp63-/- -w

Coch

lea

dim

ensio

n m

m

0

2

4

6

8

10

Mill

isec

onds

8K

Hz

AEPs

Wt TAp63-/-

10

30

50

70

3/6 4/6 4/6 6/ 11 3/5 6/11

TAp63-/- mice

% o

f Sup

ernu

mer

ary

I/OHC

Sup. cells ratio

A B C

FED

G H I

Fig. 1. Morphological and functional defects in cochlea of p63-deficient mice. (A and B) H&E-stained sections of murine WT and p63−/− E18.5 embryos. Thep63−/− embryo shows an abnormal structure of the scalae tympani and media compared with the corresponding WT samples. (Scale bars, 350 μm.) (C) TAp63αand ΔNp63α expression in cochlea and epidermis of WT embryos. Semiquantitative PCR analysis shows much higher expression of TAp63α in cochlear tissuecompared with the expression of ΔNp63α. In contrast, expression of only ΔNp63α was seen in the skin; expression was normalized to the expression of actin.(D–F) Immunofluorescence staining of cochlear sections of mouse embryos (E18.5): WT (D), p63−/− (E), and TAp63−/− (F). MyoVIIa (green), Cx26 (red), and DAPInuclei (blue) staining. (Scale bars, 20 μm.) E and F show supernumerary inner and outer hair cells in both p63−/− and TAp63−/− embryos. The staining for Cx26protein suggests that the p63−/− embryos lack intercellular junctions compared with the WT embryos in which connections were correctly formed; the sameeffect was observed in TAp63−/− (F). (G) Auditory evoked potential analysis shows a variable delay of the fifth wave cerebral cortex potentials in response to8-KHz stimulation. (H) Comparative study of the width (w) and length (L) of the mouse Cochlea by CT images (Fig. S4). (I) Histological analysis of at least threesections (base, middle, apex) from each cochlea TAp63−/− mice revealed the presence of supernumerary IHC/OHC cells. The supernumerary cells were presentat least in 50% of section (n= 6 mice analyzed for each group), ratio shows defective sections over total sections analyzed in TAp63−/− mice.

Terrinoni et al. PNAS | April 30, 2013 | vol. 110 | no. 18 | 7301

CELL

BIOLO

GY

p63−/− mice. Cx26 was expressed in WT mice, as previously re-ported (19), in supporting and stria vascularis cells, with modestoverexpression seen in the p63−/− embryos (Fig. S3B andE), whereit was abnormally localized below the stria. The distribution ofsupporting cells was similar inWT and p63−/−mice. Staining of haircells withMyo-VIIa inE18.5 embryos showed an abnormal numberof hair cells in p63−/−mice (Fig. 1D andE): fourOHCand two IHCwere identified in p63−/− mice instead of the regular three OHCand one IHC seen in the WT. Moreover, both IHC and OHC inp63−/− embryos lacked the characteristic cilia, which are clearlyvisible in the sections fromWT embryos. The abnormal number ofhair cells in p63−/− mice suggests that p63 transcriptional targetgenes are essential for the normal development of the cochlea.

Selective Involvement of TAp63 Isoform in Cochlea Development.Whereas ΔΝp63 is normally expressed in epithelia, TAp63 wasreported to be restricted to oocytes and cardiomyocytes (14–16).To investigate the importance of each of these p63 isoforms in thecochlea, we performed RT-PCR on surgical extract from wholecochlea from E18.5 WT mice. Interestingly, this analysis revealeda strong expression of TAp63 in the cochlea (Fig. 1C), as well as inoocytes (15), whereas, as reported, ΔΝp63 was the predominantisoform expressed in the skin.According to this result, we analyzed the cochlear morphology

of selective TAp63−/− mice. As shown in Fig. 1F, the TAp63−/−

mouse shows the same abnormalities of the full p63−/−. Also inthis case, in fact, supernumerary OHC and IHC are evident; theTAp63−/−mouse cochlea show a higher number ofHC (Fig. S2O).

TAp63−/− Mice Have Sensorineural Deafness. Auditory evoked po-tential analysis performed on a cohort of WT and TAp63−/− miceshowed an altered response to 8-KHz stimulation, with a variabledelay of cerebral cortex potentials, especially on the fifth wave (Fig.1G). Furthermore, computerised axial tomography (CT-scan)analysis of mice cochleae demonstrated that this organ is lightlysmaller in size in knockout compared with the WT mice (Fig. 1H;sample images in Fig. S4). The analysis of IHC/OHC number, incochleae from TAp63−/− andWTmice, using at least three cochleasections located near the apex, middle, and base of the organ, didnot reveal defects inWTmice, whereas all of the TAp63mice show50% to approximately 70% sections with supernumerary IHC/OHC cells (Fig. 1I). This suggests that TAp63 ablation (or total p63ablation) could have different penetrance, supporting the differentdegrees of hearing loss.

p63 Regulation of the Cochlea Does Not Involve a Cell DeathMechanism. TAp63 can trigger apoptosis, and this is dependenton the direct transcriptional induction of the proapoptotic BH3-only proteins Puma and Noxa (20). To test the possibility that theadditional IHC andOHCpresent in p63−/−mice could result fromthe loss of p63-mediated apoptosis (21, 22), we examined co-chleae from single Puma−/−, Noxa−/−, or double Puma−/−Noxa−/−

knockout mice. Figure 2 A–C shows that, in all these mice, thenumbers of hair cells are normal. Therefore, the increased IHCand OHC cells in the p63−/− mice are unlikely to result froma failure of p63-mediated apoptosis, but alternative mechanismsmust be involved.

TAp63 Regulates the Notch Pathway via Hairy and enhancer of split 5(Hes5) and Atonal homolog 1 (Atoh1) in Cochlea Development. Toidentify the molecular mechanisms leading to the morphologicaldefects in the cochlea of p63−/− embryos, we performed a geneexpression microarray analysis using SaOs-2 cells transfected withTAp63α or ΔNp63α. In particular, we searched for genes knownto be involved in cochlear neuroepithelial development. A com-parative gene function map provided evidence for the in-volvement of the Notch pathway, which is known to be critical forcochlea development (23–25) (Fig. S5 A and B). Indeed, Hes1,

Hes7, Hes5, Atoh1, prospero homeobox 1 (Prox1), Jagged 1(Jag1), Jag2, Connexin 26 (Cx26), Notch1 and Notch3 have allbeen validated as differentially expressed genes by RT–quanti-tative PCR (RT-qPCR) in this Tet-On inducible system as well asin cells transiently transfected with expression constructs forTAp63α (Fig. 2D andE). Remarkably, Prox1 and Atoh1 were up-regulated ∼fivefold, and Hes5 by nearly 35-fold in response toTAp63α overexpression (Fig. 2 D).This suggests that TAp63 selectively transactivated genes may

be important in cochlear neuroepithelial development, and wefocused specifically on Hes5, Atoh1 (Math1), and Prox1. We firstidentified in silico p53-like responsive elements in their promoters(Mathinspector software; Genomatix). The analysis of the Hes5−2,000 bp (from transcription start site, TSS) promoter regionshowed the presence of a putative p63 responsive element (RE)located at −1,098/1,066 bp upstream of the TSS (Fig. 3A). Thisentire region was therefore cloned upstream of luciferase ORFand the resulting construct transfected into SaOs-2 cells. Cotrans-fection of WT TAp63 significantly enhanced luciferase reporteractivity, whereas DBD TAp63α mutants (R280C, R279H, S272N,R304W) were inactive, SAM domain mutants (G530V, I537T,Q536L) showed a significant loss of function, and a TAp63αN-terminal domain mutant (Q634X) showed a gain of functionactivity (Fig. 3B) (26, 27). Confirming the in silico detection of ap63 binding site within the Hes5 promoter, ChIP assays on Saos-2Tet-On expressing TAp63α-HA showed binding of TAp63α tothis site (Fig. 3D, Lower); the MDM2 promoter was used as apositive control (Fig. 3D, Upper; Fig. 3C for protein control).Atoh1 regulatory sequences (28) have characteristic canonical

promoter sequences that are located at the 5′ end of the gene andtwo downstream sequences, called Enhancer-A and Enhancer-B,which are located∼3 kb from the stop codon. These two sequencesare separated by ∼400 bp and span ∼500 bp each. Their strongconservation between human and mouse highlights their impor-tance in gene expression regulation. Indeed, these enhancer se-quences are responsible for timing and tissue specificity of geneexpression during embryogenesis (28). The in silico analysis iden-tified p63RE within the promoter and Enhancer-A (Fig. 3E); the

18.5 E Puma-Noxa -/-18.5 E Puma -/-18.5 E Noxa -/-

A B CMyoVIIa

Cx26

35

10

5

0

TAp63α

Hes1Hes

7Hey

1Ja

g1Ja

g2 Dll3Hey

LHes

5Sox

2Prox

1Atoh

1Gli2

Cx26

p27

Notch1

Notch3

Hey2R

elat

ive

Exp

ress

ion

HA

Tub

p21

D EDoxy

SaOs-2 HA-TAp63α

- +

Fig. 2. Mice lacking Puma, Noxa, or both BH3-only proteins have normalhair cells in the cochlea; TAp63 induces Notch-related genes. Immunofluo-rescence staining of cochlear sections from E18.5 noxa−/− (A), puma−/− (B),and puma−/−;noxa−/− (C) mouse embryos stained with MyoVIIa (green), Cx26(red), and DAPI nuclei (blue). (Scale bars, 20 μm.) (D) Prox1, Hairy and en-hancer of split 5 (Hes5), and Atonal homolog 1 (Atoh1) mRNA levels areregulated by TAp63α. Transcriptional activity (RT-qPCR) of TAp63α in doxy-cycline-inducible (24 h) TAp63α cell lines. Tet-on cells lacking p63-inducibleconstructs were used as a control, and β-actin was used as an internalstandard for 2̂ ΔΔCt calculations. (E) Western blot of Tet-on induced SaOs2cells, expressing HA-tagged TAp63α. p21 expression was analyzed as a posi-tive control, and expression of tubulin was used as a loading control.

7302 | www.pnas.org/cgi/doi/10.1073/pnas.1214498110 Terrinoni et al.

p63RE in the Enhancer-A bound TAp63 by ChIP (Fig. 3F; Fig. 3Gfor protein control) and responded to TAp63 in a Luciferase assay,though not to p63 mutants (Fig. 3H). These results support thehypothesis that this region is important for the expression of thisp63 target gene in vivo (28). In contrast, analysis of the Prox1putative promoter did not reveal the presence of any responsiveelement.To test the possibility that TAp63α drives Hes5 transcription,

we treated H1299 cells with LBH589, a histone deacetylase in-hibitor, to induce TAp63 (29) (Fig. 4A). LBH589 induced TAp63expression by ∼15-fold, compared with ΔNp63, and resulted in theinduction of both Hes5 (∼15-fold) and p21 (∼40-fold), the latterknown to be a p63 target gene. To confirm the relationship be-tween TAp63 and Hes5, Atoh1, and Prox1, we extracted RNAfrom cochleae of p63−/− E18.5 embryos and found that the levelsof Hes5 mRNA were considerably (∼10-fold) lower than in WTmice (Fig. 4B); a reduction is also shown for Atoh1 and Prox1,thereby providing a direct link between p63 and these genes invivo. Furthermore, the protein analysis, using cochlear extracts,revealed Hes5 and Atoh1 depletion in p63−/− E18.5 samples (Fig.4C), as well as Atoh1 induction in TAp63 transfected cells (Fig.

4D). Studies in transfected TAp63 cells showed a significant Hes5induction (Fig. S6D), in which a mechanism of nuclear relocali-zation is also evident (Fig. S6 A–C).Interestingly, the supernumerary hair cell phenotype observed

in our p63−/−mice (Fig. 1) is also seen inASPP2−/−mice (Fig. S6E)a regulator of p63 function (30), and is highly reminiscent of thatfound in Hes5−/− mice (31). This is consistent with the notion thatthe lack of TAp63-mediated regulation of the Hes5 promoter withconsequent reduction of Notch expression is responsible for thisdevelopmental defect.

DiscussionNotch signaling is critical for the development of many tissues andorgans. This pathway is activated by ligand–receptor interactions,resulting in transcriptional activation of target genes, includingHes1 and Hes5 (23, 24, 32). Specifically, the Notch pathway isinvolved in neuroepithelial differentiation of the inner ear, in theorgan of Corti (25, 33, 34). Here, the auditory sensory epithelium isthe specialized region of the cochlea that transduces sound. Thesensory neuroepithelia are characterized by amosaic of supportingcells and hair cells, both IHC and OHC, which are arranged into

200 1000 2000 2200 2400 2600 2800 3000 3200 3400

RE PromoterA

Input HA IgG

MDM2

Hes5

SaOs-2 TAp63αD

Hes5 gene

Input HA IgG

MDM2

Atoh1 Enhancer 3’

SaOs-2 TAp63α

Atoh1 Promoter 5’

FDoxy

HA

Tub

SaOs-2 HA-TAp63αG

Atoh1

p21

H

HA

Actin

p21

C

E

3000 4200 7200 Enhancer A Enhancer B

RE II 3’Promoter RE I 5’

Atoh1 mRNA transcript

0

1

2

3

4

5

6

7 Luciferase activity

Fold

ove

r con

trol

pcDNA-H

A

TAp6

3a W

t

G530V

I537T

Q634X

R279H

R280C

R304W

S272N

Q566fs

X94

SAM DBD

TID

B Luciferase activity

SAM DBD

TID

35

30

25

20

15

10

5

0

Fold

over

contr

ol

40

45

pcDNA-H

A

TAp6

3a W

t

G530V

I537T

Q634X

R279H

R280C

R304W

S272N

Q566fs

X94

Q536L

- +

DoxySaOs-2 HA-TAp63α

- +

Fig. 3. p63 drives Hes5 and Atoh1 promoters. (A) The Hes5 gene structure shows the presence of a putative p53/p63 RE localized at −988, −966 from the TSS,in the promoter sequence. All promoters analyses were performed by MathInspector professional release 8.0.5, March 2011; Matrix Family Library Version 8.3,October 2010. (B) SaOs2 cells were transiently cotransfected with expression constructs for WT TAp63α-HA or TAp63α-HA mutant (G530V, I537T, Q536L,R280C, R304W, S272N, R279H, Q566fsX94, and Q634X) plus the hHes5-luc reporter vector. There was an increase in luciferase activity in cells transfected withTAp63α-HA WT, more pronounced with the TAp63α-Q634X-HA mutant, but not in cells transduced with the vectors encoding mutants for the DBD (mean ±SD, n = 3). (C) Western blot analysis was performed to verify TAp63a protein expression. (D) ChIP analysis shows the binding of TAp63α to the putative p53/p63-RE. ChIP on the MDM2 promoter was used as positive control. (E) The Atoh1 gene structure shows the presence of two p53 RE: RE-I5′ localized at −1,682,−1,660 from the TSS; and RE-II3′ in the Enhancer-A. (F) ChIP analysis of the Atoh1 p53-Res. TAp63α binds only to the p53-RE Atoh1 enhancer sequence; theMDM2 promoter was used as a positive control. (G) Western blot showing TAp63α protein expression. (H) Luciferase activity was increased by TAp63α-HA andTAp63α Q634X plasmids. SaOs2 cells were transiently cotransfected with an hAtoh1-luc expression vector, TAp63α-HA, and TAp63α-HA mutants (G530V,I537T, R280C, R304W, S272N, R279H, Q566fsX94, and Q634X) (mean ± SD, n = 3).

Terrinoni et al. PNAS | April 30, 2013 | vol. 110 | no. 18 | 7303

CELL

BIOLO

GY

highly ordered rows. Several components of the Notch signalingpathway are known to be expressed in the developing organ ofCorti. Importantly, interference with their expression or mutationof the corresponding genes leads to an abnormal increase in haircells (25, 35), highlighting a role for Notch activation in theregulation of progenitors that develop into hair cells. Previousstudies have established the importance of the basic helix–loop–helix (bHLH) Notch target genes Hes1, Hes5, and Math1 in thedeveloping ear. Math1-deficient mice die shortly after birth (36)with complete disruption of hair cell development (37). Hes1 andHes5 were shown to negatively regulate the differentiation ofinner ear hair cells and, accordingly, Hes5-deficient mice havesupernumerary OHC (31).We have previously demonstrated that p63 can regulate com-

ponents of theNotch pathway by direct transcriptional induction ofJag1 and Jag2 during epidermal differentiation (21), further im-plicating a link between p63 and Notch in epidermal development(38). Here we identify Hes5 and Atoh1 as two Notch-relatedtranscriptional targets of TAp63 that are involved in the differen-tiation of the organ of Corti. Hes5 is involved in the formation ofthe hair cells in the organ of Corti (31), and here we demonstratethat it is primarily regulated by TAp63. TAp63 also regulates an-other bHLH transcription factor, Atoh1, which also plays an im-portant role in the differentiation of hair cells. Atoh1 is expressed ina transient population of actively proliferating progenitor cells (39),and Atoh1-deficient mice are characterized by the absence of au-ditory sensory hair cells (37). We show that TAp63 regulates Atoh1via binding to the enhancer A, a region known to be critical fortime- and tissue-specific expression of this gene. Themorphologicalcomparison of the inner ear of p63−/− and WT mouse embryosalso confirmed the involvement of p63 in the late stages of the

formation of the organ of Corti. Mice lacking Puma and/or Noxa,the BH3-only proteins that are transcriptionally induced byTAp63 and essential for its ability to trigger apoptosis, had normalcochlea and organs of Corti. This demonstrated that loss ofa nonapoptotic mechanism activated by TAp63 must be the causeof the defects in ear development in the p63−/− embryos. Indeed,H&E staining of E18.5 embryos showed disorganization of thecochlea, especially the scala medium and timpani, and a minorthickening of the stria vascularis, with supernumerary IHC andOHC, as well as an abnormal distribution of Cx26 (Fig. 1 D–F).This inhibition of hair progenitor cell by Notch signaling istherefore essential for correct morphological and functional de-velopment of the inner ear. Accordingly, we propose a modelwhereby TAp63 regulates the formation of the organ of Corti bytranscriptional induction of Hes5/Atoh1 (Fig. S7A), thereby reg-ulating this process of lateral inhibition. Indeed, bothHes5−/− andp63−/− mice develop supernumerary hair cells (Fig. S7B) (31) asa result of abnormal differentiation of supporting cells into haircells, because the block by Hes5 has been removed by the ablationof p63.To date, a developmental role for TAp63 has only been clearly

documented in oocytes, where it promotes apoptosis upon DNAdamage (15, 20), or in the heart (14). Here we show that p63 is alsoimportant for the maturation of the cochlea; this activity of p63requires transcriptional regulation of cell differentiation and notthe induction of its predominant apoptosis inducing target genes.Auditory analysis in two new patients demonstrated a sensorineu-ral component in hearing loss previously not reported. Furtherstudies performed on TAp63−/−mice, demonstrated a variable butconsistent sensorineural hearing loss, and histological examinationconfirmed the presence of supernumerary IHC/OHC, in allTAp63−/− mice cochleae. These data imply that TAp63 mutationslead to modification of sensorineural epithelium by affecting theintegrity of the Hes5/Atoh1 pathway and contribute with differentdegrees to the hearing loss seen in a subset of patients with ED.

Materials and MethodsMice. p63 knockoutmice (11) andmice lacking Puma, Noxa (40), or both Pumaand Noxa (41) and TAp63 (42), have been described. Mice lacking p63 weregenerated electroporating JI Es cells, which were microinjected into blasto-cysts from C57BL/6 mice. Heterozygous mice have been crossed with C57BL/6for more than 10 generations. Mice lacking Puma or Noxa were generated ona C57BL/6 background, using C57BL/6-derived ES cells. TAp63 knockout micewere in a C57BL/6 background. Animal experiment have been performed inthe Department of Experimental Medicine and Surgery, and approved by theDepartment Board, according Italian law 116/92, September 29, 2011.

Cell Culture, Transfection, and Plasmid. SaOs-2 cells with doxycycline-inducibleexpression of HA-TAp63a and HA-DNp63a were generated and treated aspreviously described (43). The luciferase reporter plasmid (hHes5-luc) wasmodified from the pGL3-Basic vector (Promega) by inserting the putativehHes5 promoter sequence containing the p53RE between Sac1 and Xho1sites. Similarly, the amplified hAtoh1 enhancer sequence was inserted inpGL3 using NheI sites.

Western Blot, Immunostaining, and Confocal Microscopy. See refs. 12 and 44for embryo preparation and confocal analysis. Subsequently, sections wereincubated for 2 h with the following primary antibodies: anti-p63 (H129 sc-8344 and H137 sc-8243; Santa Cruz), anti-MyoVIIa (25-6790, Proteus), anti-Sox2 (AB5603; Millipore), and anti-Cx26 (13-8100; Invitrogen). See ref. 26 forWestern blot preparation. All membranes were incubated for 2 h with thefollowing antibodies: anti-Hes5 (ab25374; Abcam), anti-Math1 or Atoh1(ab27667; Abcam;), anti-Cx26 (13-8100; Invitrogen), anti-p63 (Y4A3 p3362;Sigma-Aldrich), and anti-Parp (sa250; Biomol).

Real-Time qPCR, Semiquantitative RT-PCR, and ChIP. See ref. 44 for PCR analysisand Table S1 for specific PCR primer sequences. SaOs TAp63α-inducible cells(5 × 106) were cross-linked for 10 min in a solution containing 1% formalde-hyde, and ChIP assays were performed using a MAGnify ChIP system (Invi-trogen), according to the manufacturer’s instructions. Luciferase reporterassays and immunoblot analysis were performed as described previously (44).

p21Hes5

40

20

0

10

TA/ΔNp63α

CtrLBH589

Rele

tive

Expr

essio

n A

0

2

4

6

8

10

600Hes5

TAp63α

Cx26

WT

B

50Re

letiv

e Ex

pres

sion

Atoh1Prox1

Parp

H1299DpcDNA TAp63α

30

p63-/-

Atoh1

Cx26

Hes5

CochleaWT

C

Atoh1

p63

TAp63 -/-

Fig. 4. p63 drives Hes5 expression in the cochlea. (A) Treatment of H1299cells with LBH589 for 24 h induced the expression of TAp63, resulting in thetranscriptional induction of Hes5, as determined by RT-qPCR analysis (mean ±SD, n = 3). (B) Transcriptional analysis on cochlea tissue biopsies from WTand p63−/− mouse embryos (E18.5). Expression levels of Hes5 proved to bedependent on TAp63, and Hes5 mRNA levels were ∼10 times lower in co-chlea of p63−/− embryos compared with those from WT embryos; addition-ally, Prox1 and Atoh1 expression were lower in p63−/− embryos. (C) Proteinexpression of Hes5 and Atoh1 in WT and TAp63−/− embryos. C shows a lowerexpression of Hes5 in TAp63−/− sample; Cx26 protein was used as control. (D)Up-regulation of Atoh1 in TAp63 transfected H1299 cells. Parp protein wasanalyzed as loading control for nuclear protein extract.

7304 | www.pnas.org/cgi/doi/10.1073/pnas.1214498110 Terrinoni et al.

Auditory Evoked Potential.Micewereanesthetizedby i.m. injectionofketamine(100 mg/kg) and xylazine (10 mg/kg) and placed in a dark, electrically shieldedroom. Stainless steel electrodeswere inserted s.c. into the vertex (positive pole),oneretroauricularregion(negativepole),andtheopposite retroauricular region(ground)ofeachmouse.Acoustic stimuli, consistingof toneburstsat frequenciesof 8 kHz,were delivered to eachmouse. AmplaidMK12 (Amplifon) instrumentswas used to perform the analysis and elaborate data.

ACKNOWLEDGMENTS. We thank Dr. Eleonora Candi for constructive com-ments, Mr. M. Cook for help with harvesting embryos, and Drs. A. Villunger,C. L. Scott, and J. M. Adams for gifts of mice. This work was supported by theMedical Research Council, United Kingdom; MIUR/PRIN (20078P7T3K_001)/

FIRB (RBIP06LCA9_0023, RBIP06LCA9_0C), AIRC (2008-2010_33-08) (#5471)(2011-IG11955), Associazione Italiana Ricerca sul Cancro (AIRC) 5xmille(#9979), Ministero dell’Istruzione e Ricerca Scientifica (MIUR)/Progetti diRicerca di Interesse Nazionale (PRIN) 2008MRLSNZ_004, and Telethon GrantGGPO9133 (to G.M.); National Health and Medical Research Council of Aus-tralia Program Grant 461221, Australia Fellowship; and Leukemia and Lym-phoma Society Specialized Center of Research (SCOR) Grant 7413. Researchdescribed in this article was also supported in part by Min. Salute (Ricercaoncologica 26/07) Istituto Dermopatico dell’Immacolata (RF06 c.73, RF07c.57, RF08 c.15, RF07 c.57) (to G.M.) and Independent Research InstitutesInfrastructure Support Scheme (IRISS) grants through the Australian Gov-ernment and the Victorian State Government Operational InfrastructureSupport (OIS).

1. Melino G, Lu X, Gasco M, Crook T, Knight RA (2003) Functional regulation of p73 andp63: Development and cancer. Trends Biochem Sci 28(12):663–670.

2. Yang A, et al. (1998) p63, a p53 homolog at 3q27-29, encodes multiple products withtransactivating, death-inducing, and dominant-negative activities. Mol Cell 2(3):305–316.

3. Osada M, et al. (1998) Cloning and functional analysis of human p51, which struc-turally and functionally resembles p53. Nat Med 4(7):839–843.

4. Ghioni P, et al. (2002) Complex transcriptional effects of p63 isoforms: Identificationof novel activation and repression domains. Mol Cell Biol 22(24):8659–8668.

5. Thanos CD, Bowie JU (1999) p53 Family members p63 and p73 are SAM domain-containing proteins. Protein Sci 8(8):1708–1710.

6. Celli J, et al. (1999) Heterozygous germline mutations in the p53 homolog p63 are thecause of EEC syndrome. Cell 99(2):143–153.

7. van Bokhoven H, et al. (2001) p63 Gene mutations in eec syndrome, limb-mammarysyndrome, and isolated split hand-split foot malformation suggest a genotype-phe-notype correlation. Am J Hum Genet 69(3):481–492.

8. Rinne T, Hamel B, van Bokhoven H, Brunner HG (2006) Pattern of p63 mutations andtheir phenotypes—update. Am J Med Genet A 140(13):1396–1406.

9. van Bokhoven H, Brunner HG (2002) Splitting p63. Am J Hum Genet 71(1):1–13.10. Mills AA, et al. (1999) p63 is a p53 homologue required for limb and epidermal

morphogenesis. Nature 398(6729):708–713.11. Yang A, et al. (1999) p63 is essential for regenerative proliferation in limb, cranio-

facial and epithelial development. Nature 398(6729):714–718.12. Candi E, et al. (2006) Differential roles of p63 isoforms in epidermal development:

Selective genetic complementation in p63 null mice. Cell Death Differ 13(6):1037–1047.

13. Carroll DK, et al. (2006) p63 regulates an adhesion programme and cell survival inepithelial cells. Nat Cell Biol 8(6):551–561.

14. Rouleau M, et al. (2011) TAp63 is important for cardiac differentiation of embryonicstem cells and heart development. Stem Cells 29(11):1672–1683.

15. Suh EK, et al. (2006) p63 protects the female germ line during meiotic arrest. Nature444(7119):624–628.

16. Gonfloni S, et al. (2009) Inhibition of the c-Abl-TAp63 pathway protects mouse oo-cytes from chemotherapy-induced death. Nat Med 15(10):1179–1185.

17. Levine AJ, Tomasini R, McKeon FD, Mak TW, Melino G (2011) The p53 family:Guardians of maternal reproduction. Nat Rev Mol Cell Biol 12(4):259–265.

18. Haberlandt E, et al. (2001) Split hand/split foot malformation associated with senso-rineural deafness, inner and middle ear malformation, hypodontia, congenital ver-tical talus, and deletion of eight microsatellite markers in 7q21.1-q21.3. J Med Genet38(6):405–409.

19. White TW, Paul DL (1999) Genetic diseases and gene knockouts reveal diverse con-nexin functions. Annu Rev Physiol 61:283–310.

20. Kerr JB, et al. (2012) The primordial follicle reserve is not renewed after chemical orγ-irradiation mediated depletion. Reproduction 143(4):469–476.

21. Candi E, et al. (2007) DeltaNp63 regulates thymic development through enhancedexpression of FgfR2 and Jag2. Proc Natl Acad Sci USA 104(29):11999–12004.

22. Flores ER, et al. (2002) p63 and p73 are required for p53-dependent apoptosis inresponse to DNA damage. Nature 416(6880):560–564.

23. Jarriault S, et al. (1995) Signalling downstream of activated mammalian Notch. Na-ture 377(6547):355–358.

24. Kageyama R, Ohtsuka T (1999) The Notch-Hes pathway in mammalian neural de-velopment. Cell Res 9(3):179–188.

25. Lanford PJ, et al. (1999) Notch signalling pathway mediates hair cell development inmammalian cochlea. Nat Genet 21(3):289–292.

26. Browne G, et al. (2011) Differential altered stability and transcriptional activity ofΔNp63 mutants in distinct ectodermal dysplasias. J Cell Sci 124(Pt 13):2200–2207.

27. Serra V, et al. (2011) Functional characterization of a novel TP63 mutation in a familywith overlapping features of Rapp-Hodgkin/AEC/ADULT syndromes. Am J Med GenetA 155A(12):3104–3109.

28. Helms AW, Abney AL, Ben-Arie N, Zoghbi HY, Johnson JE (2000) Autoregulation andmultiple enhancers control Math1 expression in the developing nervous system. De-velopment 127(6):1185–1196.

29. Sayan BS, et al. (2010-1751) Induction of TAp63 by histone deacetylase inhibitors.Biochem Biophys Res Commun 391(4):1748–1751.

30. Notari M, et al. (2011) Inhibitor of apoptosis-stimulating protein of p53 (iASPP) pre-vents senescence and is required for epithelial stratification. Proc Natl Acad Sci USA108(40):16645–16650.

31. Zine A, et al. (2001) Hes1 and Hes5 activities are required for the normal developmentof the hair cells in the mammalian inner ear. J Neurosci 21(13):4712–4720.

32. Nishimura M, et al. (1998) Structure, chromosomal locus, and promoter of mouseHes2 gene, a homologue of Drosophila hairy and Enhancer of split. Genomics 49(1):69–75.

33. Kelley MW (2007) Cellular commitment and differentiation in the organ of Corti. Int JDev Biol 51(6-7):571–583.

34. Pan W, Jin Y, Stanger B, Kiernan AE (2010) Notch signaling is required for the gen-eration of hair cells and supporting cells in the mammalian inner ear. Proc Natl AcadSci USA 107(36):15798–15803.

35. Zine A, Van De Water TR, de Ribaupierre F (2000) Notch signaling regulates thepattern of auditory hair cell differentiation in mammals. Development 127(15):3373–3383.

36. Ben-Arie N, et al. (1997) Math1 is essential for genesis of cerebellar granule neurons.Nature 390(6656):169–172.

37. Bermingham NA, et al. (1999) Math1: An essential gene for the generation of innerear hair cells. Science 284(5421):1837–1841.

38. Okuyama R, Tagami H, Aiba S (2008) Notch signaling: Its role in epidermal homeo-stasis and in the pathogenesis of skin diseases. J Dermatol Sci 49(3):187–194.

39. Helms AW, Johnson JE (1998) Progenitors of dorsal commissural interneurons aredefined by MATH1 expression. Development 125(5):919–928.

40. Villunger A, et al. (2003) p53- and drug-induced apoptotic responses mediated byBH3-only proteins puma and noxa. Science 302(5647):1036–1038.

41. Michalak EM, Villunger A, Adams JM, Strasser A (2008) In several cell types tumoursuppressor p53 induces apoptosis largely via Puma but Noxa can contribute. CellDeath Differ 15(6):1019–1029.

42. Su X, et al. (2009) TAp63 prevents premature aging by promoting adult stem cellmaintenance. Cell Stem Cell 5(1):64–75.

43. Gressner O, et al. (2005) TAp63alpha induces apoptosis by activating signaling viadeath receptors and mitochondria. EMBO J 24(13):2458–2471.

44. Candi E, et al. (2006) p63 is upstream of IKK alpha in epidermal development. J Cell Sci119(Pt 22):4617–4622.

Terrinoni et al. PNAS | April 30, 2013 | vol. 110 | no. 18 | 7305

CELL

BIOLO

GY