Cell Reports

Article

Indian Hedgehog Controls Proliferationand Differentiation in Skin Tumorigenesisand Protects against Malignant ProgressionParisa Kakanj,1,5 Karen Reuter,1 Gilles Sequaris,1 Claudia Wodtke,1 Peter Schettina,1 Daniela Frances,1

Christos C. Zouboulis,2 Beate Lanske,3 and Catherin Niemann1,4,*1Center for Molecular Medicine Cologne (CMMC), University of Cologne, 50931 Cologne, Germany2Department of Dermatology, Venereology, Allergology and Immunology, Dessau Medical Center, 06847 Dessau, Germany3Department of Oral Medicine, Infection, and Immunity, Harvard School of Dental Medicine, Boston 02115, MA, USA4Center for Biochemistry, Medical Faculty, University of Cologne, 50931 Cologne, Germany5Present address: Max Planck Institute for Biology of Ageing, 50931 Cologne, Germany, and Institute for Genetics, University of Cologne,

50674 Cologne, Germany

*Correspondence: cnieman1@uni-koeln.dehttp://dx.doi.org/10.1016/j.celrep.2013.06.037

This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use,

distribution, and reproduction in any medium, provided the original author and source are credited.

SUMMARY

Mutations in the hedgehog pathway drive the forma-tion of tumors in many different organs, including thedevelopment of basal cell carcinoma in the skin.However, little is known about the role of epidermalIndian hedgehog (Ihh) in skin physiology. Usingmouse genetics, we identified overlapping anddistinct functions of Ihh in different models ofepidermal tumorigenesis. Epidermal deletion of Ihhresulted in increased formation of benign squamouspapilloma. Strikingly, Ihh-deficient mice showed anincrease in malignant squamous cell carcinoma anddeveloped lung and lymph node metastases. In asebaceous gland tumor model, Ihh deficiency in-hibited tumor cell differentiation. More mechanisti-cally, IHH stimulated cell proliferation by activatingthe transcription factor GLI2 in human keratinocytesand human tumors. Thus, our results uncover impor-tant functions for Ihh signaling in controlling prolifer-ation, differentiation, malignant progression, andmetastasis of epithelial cancer, establishing Ihhas a gatekeeper for controlling the grade of tumormalignancy.

INTRODUCTION

Morphogenesis and homeostasis of mammalian tissues rely on

stringent spatiotemporal control of hedgehog (Hh) signaling

activity (McMahon et al., 2003). Indian hedgehog (Ihh) consti-

tutes one of the three mammalian ligands of the Hh family and

plays an essential role in formation and maintenance of bone

and cartilage (Lanske et al., 1996; St-Jacques et al., 1999; Chung

et al., 2001; Kobayashi et al., 2005; Maeda et al., 2007). Ihh has

340 Cell Reports 4, 340–351, July 25, 2013 ª2013 The Authors

also been shown to function in T cell development, embryo

implantation, and enterocyte differentiation (Lee et al., 2006;

van den Brink, 2007; Outram et al., 2009; Kosinski et al., 2010).

Different mutations within the Ihh coding sequences have been

identified and are mainly associated with defects in the skeleton

(Anderson et al., 2012).

In general, Hh ligands bind to the receptor patched (Ptch),

resulting in a block of Ptch-mediated repression of the Hh sig-

naling receptor smoothened (Smo). Activation of Smo induces

a cascade of signaling events that trigger changes in gene

expression by Gli transcription factors (Gli1, Gli2, and Gli3).

Many of the transcriptional targets are also important mediators

of the Hh pathway, such as Gli1, Ptch1, and the Hh interacting

protein (Hhip) (Ingham et al., 2011).

Sonic hedgehog (Shh), the most widely studied ligand of the

Hh signaling pathway, is essential for hair follicle down-growth

during early stages of hair follicle development (St-Jacques

et al., 1998; Chiang et al., 1999). Epidermal Shh stimulates the

underlying dermal papillae (DP), a structure composed of

specialized dermal cells, to signal back to neighboring keratino-

cytes and to control their proliferation and differentiation (Woo

et al., 2012). Disrupting signaling of the downstream transcrip-

tion factor Gli2 results in hair follicle arrest during early morpho-

genesis (Mill et al., 2003). This block in development is rescued in

Gli2-deficient mice by epidermal Gli2 expression, suggesting

that Hh signaling in the skin epithelium is sufficient for normal

epidermal morphogenesis (Mill et al., 2003). In adult skin, Shh in-

hibition blocks hair follicle regeneration, demonstrating its role in

hair follicle renewal during epidermal homeostasis (Wang et al.,

2000; Gritli-Linde et al., 2007). Furthermore, Hh target genes

are expressed in proliferative cells during hair follicle regenera-

tion (Oro and Higgins, 2003). Additionally, Gli1 expressing stem

cells of the hair follicle bulge have been identified to respond

to Shh produced by neurons in the adult tissue (Brownell et al.,

2011). More mechanistically, Shh signals via Gli transcription

factors to activate Brg1 in bulge stem cells to initiate hair regen-

eration. Subsequently, Brg1 recruits nuclear factor kB (NF-kB) to

induce Shh to sustain proliferation in matrix cells, thereby

generating a positive feedback loop for hair follicle regeneration

(Xiong et al., 2013). Collectively, these data indicate that Hh

signaling plays crucial roles in both epidermal morphogenesis

and homeostasis.

It is well established that uncontrolled activation of Hh sig-

naling results in the development of cancer in different organs,

including brain, muscle, and skin (Pasca di Magliano and

Hebrok, 2003; Teglund and Toftgard, 2010; Ng and Curran,

2011). Therefore, to design effective therapies, it is important

to gain a detailed molecular understanding of Hh signaling and

identify activating mutations. Mutations in the Ptch1 receptor

are found in patients with basal cell nevus syndrome (BCNS,

also known as Gorlin-Goltz syndrome), which is characterized

by a high incidence of medulloblastoma and basal cell carci-

noma (BCC). Additionally, Ptch1 mutations have been identified

in a subset of sporadic BCC (Teglund and Toftgard, 2010).

Whether mutations in other Hh pathway components contribute

to Hh-driven skin cancer has not been sufficiently investigated.

Previously, we and others (Niemann et al., 2003; Takeda et al.,

2006; Lo Celso et al., 2008) detected expression of Ihh in human

sebaceous skin tumors and human sebocyte cell lines. However,

the functional relevance of Ihh signaling in skin physiology and

skin cancer is unknown.

The aim of this study was to unravel the role of Ihh in epidermal

development and tumorigenesis. We investigated the specific

function of epidermis-derived Ihh by deleting Ihh in mouse

epidermis and identified a critical gatekeeper function of Ihh

in skin cancer metastasis. Furthermore, we studied the role of

individual Gli transcription factors in different types of skin

tumors and unraveled an important role for Gli2 in mediating

the effects of Ihh. Our findings also point to nonoverlapping

signaling activities of Ptch1-dependent Hh signaling and Ihh in

skin cancer.

RESULTS

Ihh Controls Differentiation and Progenitor MarkerExpression in Sebaceous Skin TumorsTo investigate the function of Ihh in mammalian epidermis, we

generated mice with a conditional epidermal deletion of Ihh

(thereafter IhhEKO) by crossing Ihhfl/fl mice with keratin 14-Cre

animals (K14Cre) (Hafner et al., 2004; Razzaque et al., 2005).

K14Cre is active from embryonic day 13 onward in basal kerati-

nocytes, leading to Ihh deletion in all epithelial compartments,

including the interfollicular epidermis, hair follicles, and seba-

ceous glands (Figure 1A). In mouse epidermis, expression of

Ihh is closely associated with hair follicle morphogenesis and

its temporal regulation is similar to that of Shh and Ptch1 (Figures

S1A–S1C). Although the overall expression level of Ihh in

epidermis is low, the strongest expression of Ihh messenger

RNA (mRNA) is detected in the epidermis of adult mice (Fig-

ure S1C). In particular, Ihh mRNA expression was detected in

the Lrig1+ve (both integrina6high and integrina6low) sebaceous

gland progenitor compartment of adult mouse epidermis (Fig-

ure 1D; Jensen et al., 2009). Interestingly, a correlation between

Lrig1 and Ihh expression was also detected in sebaceous skin

tumor cells (Figures S1F and S1G).

As expected, the epidermis and primary keratinocytes of

newborn IhhEKO mice showed efficient Cre recombination (Fig-

ure 1B), and no Ihh expression was detected at postnatal day

2 (P2; Figure 1C) and P7 in IhhEKO mice (Figure S1E). IhhEKO

mice were viable and showed no macroscopic skin phenotype.

Histological analysis revealed that hair follicles and sebaceous

glands appeared normal in IhhEKO mice compared with control

littermates (Figures 1E and S1D). To examine whether Shh

expression is regulated to potentially compensate for Ihh loss

in mouse epidermis, we determined the expression levels of

Shh in P2 epidermis of IhhEKO and Ihhfl/fl control mice. No signif-

icant changes were detected in Shh or Ptch1 receptor expres-

sion in Ihh-deficient epidermis (Figures 1F and 1G), indicating

that either Ihh is dispensable for epidermal morphogenesis or

normal Shh expression is sufficient to take over Ihh functions in

IhhEKO mice.

The fact that Hh and the receptor Ptch1 are associated with

the sebaceous gland cell lineage is further supported by data

demonstrating that Ihh is a direct target of c-myc, an important

regulator of sebocyte lineage differentiation (Takeda et al., 2006;

Lo Celso et al., 2008). Furthermore, Hh activity was strongly

increased in cell lines established from individual sebaceous tu-

mors compared with normal primary keratinocytes (Figure S2E).

To investigate whether Ihh has a function in sebaceous tumor

formation, IhhEKO mice were crossed with K14DNLef1 trans-

genic mice (Niemann et al., 2002). In this mouse model, topical

treatment with a single subthreshold dose of the carcinogen

7,12-dimethylbenz[a]anthracene (DMBA) induces the formation

of sebaceous skin tumors with high frequency (Niemann et al.,

2007; Figure 2A). No Ihh expression was detected in sebaceous

tumors generated in IhhEKO mice (Figure 2G). In control mice,

expression of Ihh is rather low, indicating that only a subpopu-

lation of tumor cells produces this Hh ligand (Figure 2G). The

incidence of sebaceous tumors was not altered upon deletion

of Ihh from epidermis and the frequency of tumors was not

significantly reduced in IhhEKO mice (Figures 2B and 2C).

However, major differences were observed in tumor histology

and differentiation of tumor cells. K14DNLef1/Ihhfl/fl control

mice developed well-differentiated sebaceous adenoma typi-

cally consisting of multiple tumor lobules (sl), with each lobule

featuring a layer of basal germinative epithelial cells at the

periphery and mature sebocytes in the inner compartment of

the lobule (Figure 2D). In contrast, the lobular structure of the

IhhEKO tumors was significantly altered and the number of

mature sebocytes was dramatically reduced (Figure 2D, s).

Instead, tumors generated in IhhEKO mice frequently displayed

a reticular morphology that was not observed in tumors of con-

trol mice (Figure 2D). A detailed analysis of cell differentiation

within the tumors further strengthened this observation. Expres-

sion of the sebocyte marker stearoyl coenzyme A desaturase 1

(SCD1) was strongly reduced and the disturbed lobular archi-

tecture of sebaceous tumors was highlighted by the abnormal

distribution of the sebaceous duct marker keratin 6a (K6a) and

squamous differentiation marker keratin 10 (K10) in IhhEKO

mice (Figures 2D and 2E). These data demonstrate that Ihh

signaling is dispensable for initiating sebaceous tumors but

is crucial for governing the sebocyte differentiation program of

tumor cells.

Cell Reports 4, 340–351, July 25, 2013 ª2013 The Authors 341

Figure 1. Epidermal Morphogenesis and

Homeostasis in IhhEKO Mice

(A) Schematic depiction of different compartments

of mammalian epidermis. The K14 promoter (K14,

red) is active in basal keratinocytes of the inter-

follicular epidermis, hair follicle, and sebaceous

gland.

(B) PCR analysis of genomic DNA isolated for

floxed (1.2 kb) and deleted (400 bp) Ihh allele from

epidermis of Ihhfl/fl (�) and IhhEKO (+) mice (n = 3).

(C) qRT-PCR for Ihh mRNA expression in P2

mouse epidermis of IhhEKO (+) and Ihhfl/fl con-

trol mice (�). PCR results are presented as

mean ± SD.

(D) Ihh mRNA expression in Lrig1- and integrina6-

sorted (Itga6) keratinocytes from adult mice, as

determined by qRT-PCR (n = 2). PCR results are

shown as mean ± SD.

(E) Histology (hematoxylin and eosin [H&E]) and

immunostaining for expression of sebocyte dif-

ferentiation marker adipophilin (Adipo) in skin

sections of 7-week-old IhhEKO mice (n = 7) and

Ihhfl/fl control mice (n = 5). Scale bars, 50 mm.

(F and G) qRT-PCR for Ptch1 (F) and Shh (G)

mRNA expression in P2 mouse epidermis of

IhhEKO (+) and Ihhfl/fl control mice (�). PCR results

are shown as mean ± SD.

See also Figure S1.

Next, we wanted to know whether sebaceous gland and

squamous differentiation is also altered in the skin of IhhEKO

mice. Therefore, we compared the expression of marker mole-

cules in skin from P2 and adult IhhEKO and control mice.

Expression of both the sebaceous gland marker adipophilin

(Figures 1E and S2A) and SCD1 (Figure S2B) was not affected

upon epidermal Ihh deletion. In addition, the sebaceous

glands developed normally, as seen in epidermal whole mounts

of P2 and P9 IhhEKO mice (Figures S1D and S2B). The expres-

sion pattern for K10 and filaggrin, marker molecules for squa-

mous differentiation of keratinocytes, was also not disturbed

in IhhEKO P2 and adult skin (Figures S2C and S2D). We

conclude that abnormal differentiation is specific to pathophys-

iological conditions (e.g., sebaceous skin tumors in IhhEKO

mice), but is not detected during normal epidermal develop-

ment and homeostasis in adult skin upon epidermal Ihh

deletion.

342 Cell Reports 4, 340–351, July 25, 2013 ª2013 The Authors

To investigate whether the observed

alterations in tumor morphology and dif-

ferentiation could result from differences

in the distribution and number of hair folli-

cle stem and progenitor cells within the

tumors, we examined the expression of

the sebaceous lineage-associated pro-

genitor marker Plet1 (Nijhof et al., 2006;

Raymond et al., 2010). Remarkably, cells

found in reticular structures of IhhEKO

tumors were strongly positive for Plet1,

whereas only a few Plet1-positive cells

were found in control tumors (Figure 2D).

An increase in Plet1 was also detected in

protein lysates of tumors generated in IhhEKO mice compared

with sebaceous tumors in Ihhfl/fl mice (Figure 2F), suggesting

that loss of mature sebocytes could be a direct consequence of

the inhibition of precursor cell differentiation.

Ihh Deficiency Does Not Affect Stem Cell Functionduring Skin HomeostasisGiven the important function of Shh and Gli1 in hair follicle bulge

stem cells (Brownell et al., 2011; Xiong et al., 2013) and the regu-

lation of the progenitor marker Plet1 in sebaceous tumors devel-

oping in IhhEKO mice, we wanted to know whether normal stem

cell function is generally affected upon Ihh deletion from mouse

epidermis. To address this important issue, we analyzed distinct

stem cell and progenitor compartments of the pilosebaceous

unit in great detail. The bulge stem cell marker Sox9, keratin 15

(K15), and CD34 were not changed in IhhEKO mice (Figures 3A,

S3A, and S3B). Moreover, the number of CD34+/integrina6high

Figure 2. Ihh Deletion from Mouse

Epidermis Leads to Impaired Differentiation

of Sebaceous Skin Tumors

(A) Experimental layout for sebaceous tumor ex-

periments performed in K14DNLef1 mice.

(B and C) Incidence (B) and frequency (C) of

sebaceous tumors in K14DNLef1/IhhEKO (n = 19)

and K14DNLef1/Ihhfl/fl mice (n = 17). Two-way

ANOVA with repeated measures for ± SEM.

(D) Histology (H&E) and immunostaining for

expression of differentiation marker SCD1, K10,

and progenitor marker Plet1 (all in green) in tumors

of K14DNLef1/IhhEKO and K14DNLef1/Ihhfl/fl mice

(n = 25). sl, sebaceous lobules; s, sebocytes. Scale

bars, 100 mm.

(E and F) Western blot analysis for sebocyte dif-

ferentiation marker SCD1 and a-Tubulin (loading

control) (E), and Plet1 and GapDH (loading control)

(F) in lysates of sebaceous tumors formed in Ihhfl/fl

and IhhEKO mice.

(G) qRT-PCR for Ihh mRNA expression in seba-

ceous tumors generated in IhhEKO and Ihhfl/fl con-

trol mice. PCR results are shown as mean ± SD.

See also Figure S2.

stem cells, as well as the location and number of label-retaining

cells (LRCs) of the hair follicle bulge, was not altered in adult mice

upon Ihh deletion (Figures 3B and 3C). Next, we examined the

expression of Lrig1 and Plet1, markers for hair follicle stem and

progenitor cells that are associated with Ihh function in seba-

ceous tumors (Figures 1D, 2D, and 2F). Both markers showed

a normal expression pattern in P3 and 7-week-old IhhEKO mice

compared with littermate controls (Figures 3D and 3E). Thus,

Plet1 marker expression and localization are specifically altered

in the sebaceous tumor model, but not in normal skin in IhhEKO

mice.

Finally, we tested for the physiological function of epidermal

stem cells and the sebaceous gland. Measurements of transepi-

dermal water loss (TEWL) revealed that there was no inside-out

Cell Reports 4, 340–

barrier defect in IhhEKO newborn mice

(Figure S3C). In addition, we analyzed

the production of sebaceous gland lipids

and epidermal lipids of the stratum cor-

neum. These assays demonstrated that

lipid production was not affected and

that crucial functionsof theepidermal bar-

rier and the sebaceous glands were not

disturbed in IhhEKO mice (Figures S3D

and S3E). Thus, we conclude that Ihh

signaling regulates progenitor function in

sebaceous tumors, but is not crucial for

stem cell function and homeostasis of

the pilosebaceous unit.

IhhDeficiency Promotes SquamousSkin Tumor FormationBecause our results showed that Ihh

plays an important role in differentiation

of sebaceous tumor cells, we sought to

address whether this function also applies to epidermal cancer

in general. To that end, we induced squamous skin tumors in

IhhEKO mice using the two-stage chemical skin carcinogenesis

protocol. This protocol involves the topical application of

DMBA, leading to oncogenic Ras mutations, followed by repet-

itive treatment with the tumor promoter 12-O-tetradecanoyl-

phorbol-13-acetate (TPA), resulting in the proliferation of initiated

cells (Figure 4A; DiGiovanni, 1992). Typically, this leads to the

formation of benign papillomas that may eventually progress to

malignant SCCs. Interestingly, and in contrast to the sebaceous

tumor model, IhhEKO mice developed more papillomas and the

incidence of squamous tumors was increased upon epidermal

deletion of Ihh (Figures 4B and 4C). Initially, the papillomas

grewmore slowly in IhhEKOmice (Figure 4D). These data indicate

351, July 25, 2013 ª2013 The Authors 343

Figure 3. Stem Cell Function Is Not Affected in IhhEKO Mice

(A and B) Immunostaining for (A) the hair follicle bulge stem cell marker Sox9

(green) in a skin section, and (B) LRC (green) andK14 (red) in anepidermalwhole

mount from adult Ihhfl/fl (n = 5) and IhhEKO mice (n = 7; LRC: n = 3). B, bulge.

(C) Representative fluorescence-activated cell sorting plot and quantification

for CD34+/integrina6high (CD49f) keratinocytes from 7-week-old Ihhfl/fl and

IhhEKO mice (n = 2).

(D and E) Immunostaining for the progenitor marker molecules Lrig1 (D) and

Plet1 (E; both green) in skin sections of P3 (n = 5) and 7-week-old (n = 3) Ihhfl/fl

and IhhEKO mice. I, isthmus; JZ, junctional zone.

Scale bars, 50 mm (A and D), 100 mm (B), and 20 mm (E). See also Figure S3.

that Ihh suppresses the formation of squamous tumors, but once

tumor growth is initiated, it stimulates the proliferation of tumor

cells, resulting in increased tumor mass. In addition, squamous

differentiation of keratinocytes was also altered in the papillomas

of IhhEKO mice. Particularly, the expression of differentiation

marker K10 was strongly reduced, demonstrating a block in

squamous differentiation in the absence of epidermal Ihh (Fig-

ure 4F). In contrast, expression of the basal keratinocyte marker

K5 was not dramatically altered in different tumors generated in

IhhEKO mice (Figures S4A–S4C). Together, our results unravel a

novel function of Ihh in governing cellular differentiation pro-

grams in different tumor types, including sebaceous and squa-

mous skin tumors.

Increase in Malignant Tumor Progression andMetastasis in Ihh-Deficient MiceTo investigate whether Ihh is important for the progression of

benign papillomas to malignant SCCs, we monitored the mice

for up to 50 weeks following DMBA treatment. Surprisingly,

344 Cell Reports 4, 340–351, July 25, 2013 ª2013 The Authors

epidermal deletion of Ihh resulted in a strong increase in SCC

formation, with significantly more mice displaying malignant

tumors and an increased number of papillomas progressing to

SCC (Figures 4B, 4E, and 5A). SCC developed slightly later in

IhhEKO mice (25 weeks in Ihhfl/fl versus 28 weeks in IhhEKO after

DMBA treatment), but once the progression of benign papil-

lomas to malignant SCCs began, this process proceeded

much more rapidly in Ihh-deficient epidermis compared with

controls (Figure 4B). At the end of the two-stage chemical

carcinogenesis experiment, IhhEKO mice displayed significantly

fewer benign squamous papillomas and keratoacanthomas

than Ihhfl/fl control mice, whereas the number of SCCs was

strongly increased (Figure 4E). Expression analysis revealed

that Ihh was not detected in papillomas and SCCs formed in

IhhEKO mice. Papillomas and SCCs of control mice showed

variable but low expression of Ihh, indicating that only a subpop-

ulation of tumor cells produced the ligand (Figure 5G). Of note, no

upregulation of Shh RNA expression and no significant changes

in Ptch1 expression were detected in tumors of IhhEKO mice

(Figure S5).

Importantly, phenotypic differences were observed between

SCCs from IhhEKO and control mice. While SCCs from control

mice grew in defined epithelial tumor cell clusters that often

seemed to be separated from the surrounding stromal tissue,

SCCs of IhhEKO mice frequently exhibited a dispersed and inva-

sive pattern of tumor growth, and some SCCs even resembled

spindle cell carcinoma (SpCC; Figures 5B and 5C). Such aggres-

sive tumors were not formed in control mice. Staining for the

a-5 subunit of laminin revealed disruption of the basement

membrane between the stromal and epithelial cells of SpCC in

Ihh-deficient mice (Figure 5C). In agreement with the histological

features of aggressive growth and poor differentiation, mitotic

figures and abnormal nuclei were common to SCCs of IhhEKO

mice (inset, Figure 5B). Moreover, we observed suprabasal

integrina6 expression in SCCs from IhhEKO mice. In contrast,

integrina6 was localized to the basement membrane and

basal compartment of SCCs in control mice (Figure 5D). Given

that suprabasal expression of a6b4 integrins is a hallmark of

epidermal neoplastic progression, our data clearly demonstrate

an increase in malignancy upon epidermal Ihh deletion (Tennen-

baum et al., 1993; Van Waes et al., 1995).

Strikingly, our analysis of tumor-bearing mice disclosed the

presence of metastasis in Ihh-deficient mice but not in control

animals. Of note, mice in a C57BL/6 background are normally

extremely resistant to malignant conversion and formation of

metastasis. Metastases were identified in lymph nodes (Fig-

ure 5E) and in the lung of IhhEKO mice (Figure 5F). Abundant

immunofluorescent staining with antibodies against K14, a

marker for undifferentiated keratinocytes, confirmed the pres-

ence of epidermis-derived cells in metastases (Figures 5E and

5F). These results show that epidermal Ihh not only inhibits the

formation of papillomas but, importantly, also blocks the malig-

nant progression of benign squamous tumors and the develop-

ment of metastasis.

To investigate the mechanism of the increased metastatic

behavior, we examined the potential role of Ihh in keratinocyte

migration. For this purpose, we isolated primary keratinocytes

from IhhEKO and control mice, and analyzed the migration of

Figure 4. Loss of Ihh from Mouse Epi-

dermis Promotes Ras-Mediated Skin

Cancer and Alters Proliferation and Tumor

Differentiation

(A) Experimental layout for two-stage chemical-

induced carcinogenesis studies.

(B and C) Incidence (B) and frequency (C) of

squamous tumors developing in Ihhfl/fl (n = 19)

and IhhEKO (n = 20) mice. Treatments in control

groups (n = 14 per genotype) did not result in

papillomas.

(D) Quantification of tumor size for papillomas and

SCCs in Ihhfl/fl and IhhEKO mice (at weeks 41 and

50 after tumor initiation). Results are presented as

mean ± SEM. *p < 0.05; p = 0.057 (pap week 50)

and p = 0.580 (SCC week 50).

(E) Incidence of papilloma and keratoacanthoma

(Pap/KA) and SCC in Ihhfl/fl (n = 26) and IhhEKO (n =

28) mice at the end of one representative experi-

ment. Note that only tumors that could be clearly

categorized based on histological appearance are

presented.

(F) Histology (H&E) and immunostaining for keratin

10 (K10, green) in papillomas of IhhEKO and Ihhfl/fl

mice (n = 9). Scale bars, 100 mm.

See also Figure S4.

primary cells in vitro. Interestingly, Ihh-deficient keratinocytes

migrated significantly faster than control cells (Figure 5H).

Thus, our data indicate that Ihh blocks metastasis by inhibiting

the migration of keratinocytes, thereby preventing keratinocytes

from leaving the primary tumor.

Ihh Regulates Cell Proliferation in MammalianEpidermis and Different Types of Skin TumorsGiven the reduced size of papillomas developing in IhhEKO mice,

we investigated the involvement of Ihh in regulating proliferation.

To this end, we analyzed bromodeoxyuridine (BrdU) incorpora-

tion in epidermal tumors generated in the different mouse

models. Interestingly, the number of BrdU positive keratinocytes

was >2-fold reduced in papillomas developing in IhhEKO mice

compared with controls (Figure 6A). In addition, Ihh-deficient

mice showed less cell proliferation in SCC (Figure 6B) and a

strong reduction in proliferation sebaceous tumors generated

in the K14DNLef1 background (Figure 6C). Moreover, cell prolif-

Cell Reports 4, 340–

eration was also suppressed in primary

keratinocytes isolated from IhhEKO mice

and in hair follicles of newborn Ihh-

deficient mice compared with control

animals (Figures 6D and S6A). This

reduced proliferation rate was observed

until P3. At this time of development,

Shh mRNA expression normally is

detected in hair follicle keratinocytes,

and potentially could compensate for

this defect seen in Ihh-deficient epi-

dermis from P3 onward (St-Jacques

et al., 1998; Figure S1B). Taken together,

our results reveal that Ihh stimulates pro-

liferation during epidermal morphogenesis and promotes growth

in various types of skin tumors.

Ihh Controls Cell Proliferation by Activating GLI2SignalingOur data clearly demonstrate that Ihh plays an important role in

regulating tumor cell proliferation and malignant conversion. To

investigate the molecular mechanism of this effect in human

epidermal cells and human skin tumors, we analyzed the func-

tion of GLI transcription factors in this process. In general, the

transcriptional effectors responsible for mediating Ihh functions

are not well understood and most likely are dependent on the

cell type and the microenvironment (Joeng and Long, 2009).

First, we studied the functions of the individual GLI proteins

in vitro using the sebaceous gland cell line SZ95 (Zouboulis

et al., 1999; Niemann et al., 2003). The expression of key compo-

nents of the HH pathway was low in undifferentiated sebocytes

but strongly increased when proliferation or differentiation was

351, July 25, 2013 ª2013 The Authors 345

Figure 5. Epidermal Deletion of Ihh Pro-

motes Malignant Tumor Progression and

Tumor Metastasis

(A) Number of papillomas progressing to malig-

nant SCC in Ihhfl/fl (n = 19) and IhhEKOmice (n = 20).

Quantification of a representative experiment.

(B) Histology (H&E) of SCC and mitotic figures

(arrows) and abnormal nuclei (insets) of repre-

sentative tumors developing in Ihhfl/fl and IhhEKO

mice.

(C) Histology (H&E) and immunostaining for lami-

nin (green) and nuclei (magenta) of SpCC devel-

oping in IhhEKO mice. Note disruption of basement

membrane staining (arrows).

(D) Immunostaining for integrina6 (Itga6; green) in

SCC of Ihhfl/fl (n = 3) and IhhEKO mice (n = 8).

(E and F) Macroscopic view, histology (H&E), and

immunostaining for K14 (green) of metastases in

lymph nodes (E) and lung (F) observed in IhhEKO

and Ihhfl/fl control mice.

(G) qRT-PCR for Ihh mRNA expression in benign

papillomas and malignant SCCs generated in

IhhEKO and Ihhfl/fl control mice. PCR results are

presented as mean ± SD.

(H) Representative images of scratch assay (18 hr)

and quantification of the migration of primary

keratinocytes isolated from Ihhfl/fl and IhhEKO mice

(n = 3). Areas are presented as mean ± SD. *p <

0.05, ***p < 0.001.

Scale bars, 100 mm (B–D), 50 mm (E and F), and

20 mm (inset in B). See also Figure S5.

induced (Figure S6B). Importantly, IHH was the only HH ligand

that was highly expressed by mature SZ95 sebocytes, and

expression of SHH and DHH was not detected in sebaceous

gland cells (Figure 6B; data not shown). This observation further

strengthens our initial finding that Ihh signaling directs the seba-

ceous gland fate in tumors of K14DNLef1 mice (Figures 2D and

S6B). In addition, the fact that only IHH was expressed made

this cell line an ideal model for specific molecular dissection of

IHH signaling. In line with the expression data, HH activity was

strongly increased in mature sebocytes, as assessed by the

activity of a Gli reporter construct (Figure 6E). Remarkably,

GLI2 and GLI3, but not GLI1, were highly expressed in SZ95

cells, suggesting that these twomolecules govern sebocyte pro-

liferation and differentiation, and thatGLI1 is not a target gene of

HH activity in sebocytes (Figure S6B; data not shown).

Next, we studied the effect of individual GLI transcription fac-

tors on cell proliferation by overexpressing them in undifferenti-

ated sebocytes. Interestingly, GLI2 strongly increased BrdU

346 Cell Reports 4, 340–351, July 25, 2013 ª2013 The Authors

incorporation (Figure 6F), indicating that

GLI2 could be transmitting the effect

of IHH on proliferation in mammalian

epidermis and different types of skin

tumors. To test this hypothesis, we trans-

fected primary keratinocytes isolated

form IhhEKO and control mice with an

expression vector for Gli2. Indeed, Gli2

expression was sufficient to rescue the

proliferation defect in IhhEKO keratino-

cytes (Figure 6H), identifying this transcription factor as a crucial

mediator of Ihh function in cell proliferation.

To further confirm that GLI signaling indeed regulates kerati-

nocyte proliferation, we treated SZ95 sebocytes with Gant61,

a small-molecule inhibitor of the HH-GLI signaling axis that

interferes with DNA binding of GLI transcription factors

(Lauth et al., 2007). Treatment of SZ95 cells with Gant61 led

to a strong reduction of BrdU incorporation in a dose-depen-

dent manner (Figure 6G), showing that GLI2 is an important

regulator of proliferation in mammalian epidermis and skin

tumors. In addition, a similar result was also observed with

cyclopamine, an inhibitor of the Hh coreceptor Smo (Fig-

ure 6G). Remarkably, growth of sebaceous tumors generated

in K14DNLef1 mice was blocked following treatment with

cyclopamine (Figure S6C). Thus, these data indicate that block-

ing of Ihh and Gli2 signaling is a potential therapeutic strategy

for treating sebaceous gland cell hyperproliferation in mamma-

lian skin.

Figure 6. GLI2 Mediates Ihh-Driven Regula-

tion of Proliferation

(A–C) Immunostaining for BrdU (green), K14 (red),

and DAPI (blue), and quantification of BrdU+ve

keratinocytes in squamous papilloma (A; n = 6),

SCC (B; n = 3), and sebaceous skin tumors (C; n =

20) in Ihhfl/fl and IhhEKO mice. Quantification is

shown as mean ± SD. *p < 0.05, ***p < 0.001.

Scale bars: 100 mm (A and C) and 50 mm (B).

(D) Quantification of BrdU incorporation into pri-

mary keratinocytes isolated from newborn Ihhfl/fl

and IhhEKO mice (n = 3). Quantification is shown as

mean ± SD. ***p < 0.001.

(E) Gli reporter gene activity tested with Gli-BS-

luciferase (light gray; mutant Gli-binding sites as

control in dark gray) in undifferentiated und

differentiated SZ95 human sebocytes (n = 3). Fold

change is presented as mean ± SD.

(F) Quantification of BrdU incorporation into

human SZ95 sebocytes following transfection

with Gli expression constructs. Fold change over

mock-transfected cells is presented as mean ±

SD. **p < 0.01, ***p < 0.001.

(G) Quantification of BrdU-positive SZ95 sebo-

cytes following treatment with increasing con-

centrations of cyclopamine (5, 10, and 50 mM) and

GANT61 (10, 50, and 100 mM). Numbers are

presented as mean ± SD. *p < 0.05, **p < 0.01,

***p < 0.001.

(H) Quantification of BrdU incorporation into pri-

mary keratinocytes isolated from newborn Ihhfl/fl

and IhhEKO mice (n = 2) following transfection with

Gli expression constructs (n = 3). Fold change is

presented as mean ± SD.

See also Figure S6.

Finally, we tested whether GLI transcription factor expression

and IHH signaling are also altered in human sebaceous tumors.

To that end, we analyzed mRNA expression of GLI1,GLI2,GLI3,

and PTCH1 in human sebaceous tumors and normal skin by

quantitative RT-PCR (qRT-PCR). Interestingly, GLI1 was ex-

pressed in normal skin but was almost absent from tumor sam-

ples (Figure 7A), and mRNA levels for GLI3 were low in tumors,

which suggests that these two GLI proteins do not play a crucial

role in this particular type of tumor. In contrast,GLI2was strongly

expressed in differentiated and undifferentiated human seba-

Cell Reports 4, 340–

ceous tumors compared with human

SCCs, pointing to an important role of

this transcription factor in mediating the

IHH effect on proliferation in human seba-

ceous skin tumors (Figure 7B). The recep-

tor PTCH1 was strongly expressed in

human sebaceous tumors compared

with normal skin (Figure 7A). Given that

PTCH1 is also an HH target gene, this

result further strengthens our initial

observation that IHH signaling is impor-

tant for driving the proliferation and differ-

entiation of sebaceous tumors.

Taken together, our experimental data

clearly demonstrate a function for Ihh in

regulating proliferation and differentiation in different pathophys-

iological settings. We provide evidence that the transcription

factor Gli2 plays a central role in mediating these effects. Finally,

epidermal Ihh is essential for protecting against malignant pro-

gression and metastasis of epidermal tumors (Figure 7C).

DISCUSSION

Activation of the Hh signaling cascade has been linked to several

human cancers, and changes in Hh activity are commonly linked

351, July 25, 2013 ª2013 The Authors 347

Figure 7. Hh Signaling in Human Sebaceous

Tumors and SCCs

(A and B) qRT-PCR for expression of PTCH1,

GLI1, and GLI3 (A) and Gli2 (B) in human seba-

ceous tumors (A and B), human SCCs (B), and

normal human skin.

(C) Model for specific functions of Ihh in regulating

epidermal homeostasis and skin tumorigenesis.

to mutations in components of the pathway (Pasca di Magliano

and Hebrok, 2003; Scales and de Sauvage, 2009). In addition,

ligand-dependent cancers have been proposed to result from

an autocrine or paracrine signaling response. Here, we identified

in vivo functions for Ihh in mammalian skin tumors. Ihh stimulates

cell proliferation in different experimental settings and is required

for sebocyte differentiation in a mutant Lef1-induced sebaceous

tumor model. Most strikingly, Ihh blocks the progression of

benign papillomas to malignant SCCs and prevents metastasis

of SCCs in a Ras-driven skin tumor model. These results unravel

crucial functions for Ihh in epithelial cancer and clearly demon-

strate that Ihh is required beyond its function in morphogenesis

and normal tissue homeostasis.

Generally, abnormal activation of Hh signaling is strongly

associated with tumor initiation and maintenance of tumor

growth (Pasca di Magliano and Hebrok, 2003). Remarkably, we

observed that Ihh blocks tumor initiation, progression, and

metastasis of SCCs. This points to diverse functions of Hh

signaling in skin cancer, and strongly suggests that Shh and

Ihh play different roles in this process. In line with our results, a

similar function for Ihh in tumor cell differentiation and concom-

itant repression of progenitor cell proliferation was observed in

colon cancer (van den Brink et al., 2004; van den Brink, 2007;

Fu et al., 2010). In this model, Ihh expression strongly correlated

with the grade of malignancy, with the highest Ihh expression

levels observed in differentiated intestinal tumors. In contrast,

strong expression of Shh was detected in undifferentiated malig-

nancies (van den Brink, 2007; Saqui-Salces and Merchant,

2010), further supporting the notion that these two Hh ligands

play opposing roles. Notably, specific and nonoverlapping func-

tions for these Hh ligands have also been proposed to govern tis-

sue homeostasis in the intestine and skin, where expression of

both factors is restricted to a distinct compartment within the

tissue. In the intestine, Shh is expressed in crypts where it is

348 Cell Reports 4, 340–351, July 25, 2013 ª2013 The Authors

associated with proliferation, whereas

Ihh promotes differentiation in the villi

(van den Brink, 2007). Likewise, Shh reg-

ulates hair growth in the lower hair follicle

of the skin, whereas Ihh has been pro-

posed to affect differentiation in seba-

ceous gland cells (Niemann et al., 2003;

Lo Celso et al., 2008).

A central finding of our study is that

epidermal deletion of Ihh leads to

reduced proliferation during epidermal

development in newborn mice as well as

in both sebaceous and squamous skin

tumor models. This clearly points to a

general role of Ihh activity in keratinocyte proliferation. Surpris-

ingly, although keratinocyte proliferation was strongly sup-

pressed in IhhEKO during early postnatal stages of epidermal

development, hair follicle morphogenesis was not arrested as

seen in Shh�/� and Gli2i�/� mice (St-Jacques et al., 1999; Mill

et al., 2003). The morphology of DP appears normal in IhhEKO

mice, and despite a temporary decrease in cell-cycle progres-

sion, no impact on hair follicle morphogenesis is observed. Our

data therefore demonstrate that Ihh does not regulate DP func-

tion to the same extent as Shh (Woo et al., 2012). Alternatively,

Shh and potentially Dhh (produced by the DP cells) (Brownell

et al., 2011; Xiong et al., 2013) could compensate for the loss

of epidermal Ihh.

Remarkably, although keratinocyte proliferation is strongly

reduced, malignant progression, cell migration, and metastasis

are induced upon epidermal deletion of Ihh. Generally, the high

proliferation rate of tumor cells constitutes an important prereq-

uisite to allow for further accumulation of mutations leading to

expansion of more aggressive tumor cell clones. Our data

demonstrate that Ihh plays a dominant gatekeeper role in block-

ing the progression and expansion of malignant tumor cell pop-

ulations. Interestingly, human BCC, an HH-driven skin cancer,

rarely metastasizes. Our results raise the question as to whether

HH signaling plays a general role in suppressing the migration

of keratinocytes and metastasis of epidermal cancers despite

promoting their proliferation. This may involve a mechanism

whereby HH stimulates the proliferation of a subpopulation of

more differentiated and immobile tumor cells that are incapable

of malignant progression and metastatic spread. This idea is

further supported by our finding that the pool of Plet1-expressing

progenitor cells was increased in tumors upon epidermal Ihh

deletion.

Mechanistically, we observe that Ihh exerts its effect on

proliferation mainly by activating the transcription factor Gli2.

Importantly, Gli2 expression can rescue the proliferation defect

in IhhEKO keratinocytes. A strong increase in GLI2 and GLI3

expression is seen during proliferation and differentiation of

SZ95 human sebocytes, whereas GLI1 does not seem to play

a role in these cells. Because IHH is the only Hh ligand expressed

by SZ95 sebocyte cells and HH activity is strongly increased

upon differentiation of these human cells, we conclude that

IHH is important for balancing the proliferation and maturation

of epidermal cells during skin homeostasis and under patho-

physiological conditions. It was recently shown that the inhibition

of HH signaling at the level of GLI transcription factors sup-

presses the growth of neuroblastoma (Wickstrom et al., 2012),

further strengthening our conclusion that GLI2 is indeed a rele-

vant therapeutic target for other types of tumors.

Our results identify overlapping but also diverse functions of

Ihh in two different models for epidermal tumorigenesis.

Whereas epidermal deletion of Ihh does not influence tumor

development in mutant Lef1-induced sebaceous tumors, it pro-

motes the incidence, frequency, and malignant progression of

Ras-driven squamous skin cancer. This highlights the tremen-

dous impact the type of oncogenicmutation and cellular environ-

ment has on driving the development of a particular type of

tumor. Indeed, an important role for a crosstalk between Hh

and EGFR/Ras signaling has been shown for various tumor

models (Mangelberger et al., 2012). Furthermore, the cell of

origin could differ between sebaceous and squamous skin

cancers and subsequently lead to distinct Hh responses. Given

our finding that Ihh activity in skin tumor formation is context

dependent, our work emphasizes the need for a more detailed

analysis of HH components and activity in human cancer and

points to novel experimental strategies for designing personal-

ized therapies (Atwood et al., 2012).

Together, our results identify specific functions for Ihh in

different types of skin tumors. Ihh stimulates differentiation and

proliferation, and protects against malignant progression, migra-

tion, and metastasis of squamous skin cancer.

EXPERIMENTAL PROCEDURES

Experimental Mice and Tumor Tissues

To generate epidermis-specific deletion of Ihh (IhhEKO), mice carrying a condi-

tional Ihh allele (Razzaque et al., 2005) were crossed with K14Cre mice (Hafner

et al., 2004). For sebaceous tumor studies, IhhEKO mice were crossed with

K14DNLef1 transgenics (Niemann et al., 2002). All mouse strains had been

backcrossed and were maintained on a pure C57Bl/6 background. C57Bl/6

mice were used to determine expression of Hh pathway components during

different stages of epidermal development.

The TEWL of newborn mice was examined with the use of a Tewameter

(Courage+Khazaka Electronic) as described by Barel and Clarys (1995).

Mouse experiments were performed according to institutional guidelines

and the animal license given by the State Office of North Rhine-Westphalia,

Germany.

The institutional review board of the Medical Faculty of the University of

Cologne approved the use of human tumor tissues in this study. Written

informed consent was obtained from patients in accordance with the guide-

lines and policies of the review board.

Tumor Experiments

The formation of sebaceous tumors was induced in 9-week-old K14DNLef1/

IhhEKO (n = 19) and K14DNLef1/Ihhfl/fl mice (n = 17) as described previously

(Niemann et al., 2007). Briefly, mice of both genotypes were treated with a

single subthreshold dose of DMBA (100 nmol; Sigma-Aldrich). Alternatively,

mice received acetone vehicle as a treatment control (n = 9). Generally, seba-

ceous tumors developed 4–5 weeks following DMBA application, and scoring

of tumors was carried out once a week for up to 25 weeks after DMBA

treatment.

In one setting, sebaceous tumors generated in K14DNLef1 transgenic mice

were treated daily with cyclopamine (60 nmol; Sigma Aldrich) or control vehicle

for three consecutive days according to previous protocols (Silva-Vargas et al.,

2005) and the size of the tumors was measured every day.

To induce squamous papillomas and SCCs, 9-week-old IhhEKO (n = 20) and

Ihhfl/fl mice (n = 19) received a topical subthreshold dose of DMBA once

(100 nmol in acetone) followed by repetitive application of the tumor promoter

TPA (6 nmol in acetone; LC Laboratories). TPA treatment was done thrice per

week for up to 25 weeks. Mice of control groups (n = 14) were treated with

DMBA and acetone, acetone and TPA, or acetone alone. Scoring and mea-

surement of tumors were done once per week and mice were monitored for

up to 50 weeks following DMBA application. Tumor size was calculated by

measuring the length and width of tumors once per week. Both tumor studies

were done twice and gave similar results.

Histological Analysis and Immunofluorescent Stainings

The mice were sacrificed and tail and back skin, epidermal whole mounts, or

tumors were harvested according to standard protocols (Braun et al., 2003).

Tissue was analyzed from at least three mice per genotype and experiment.

For LRC studies, pups received four injections of BrdU (50 mg/kg body

weight; Sigma-Aldrich) at 12 hr intervals. To analyze the LRCs, mice were

sacrificed at 7 weeks of age and epidermal whole mounts were prepared

as previously described (Braun et al., 2003). To examine cell proliferation

in epidermis and tumors, mice were injected intraperitoneally with BrdU

(100 mg/kg body weight; Sigma-Aldrich) 1 hr before they were sacrificed.

Tissue was embedded into OCT (Sakura) or paraffin following fixation with

4% formaldehyde.

To quantify cell proliferation within SCCs, tumor areas that had a suprabasal

localization of BrdU-positive nuclei and were clearly distinct from benign

papilloma structures were counted.

The following antibodies were used to stain tissue sections: K14 (1:1,000;

Covance), K5 (1:1,000; Covance), K6a (1:600, Covance), K10 (1:250, Cova-

nce), K15 (1:250; Neomarkers), SCD1 (1:100; Santa Cruz), adipophilin

(1:1,000; Fitzgerald Industries), filaggrin (1:1,000; Covance), BrdU (1:10; BD),

CD34 (1:100; e-Bioscience), Sox9 (1:750, kindly provided by M. Wegner [Stolt

et al., 2003]), Lrig1 (1:100; R&D), Plet1 (1:5, kindly provided by A. Sonnenberg

[Raymond et al., 2010]), integrina6 (1:3,000; BD PharMingen), and laminin sub-

unit a-5 (1:500, kindly provided by L. Sorokin [Sorokin et al., 1997]). Secondary

antibodies were coupled to Alexa-488 and Alexa-594 (1:1,000; Invitrogen) and

applied together with propidium iodide or DAPI (Sigma-Aldrich) to counterstain

nuclei. Images were acquired with a Nikon eclips-E800, Olympus U-TB190, or

Leica DM 4000B and analyzed with ImageJ software.

Lipid Analysis

Lipids of the stratum corneum and the sebaceous glands were measured as

described previously (Reichelt et al., 2004; Fehrenschild et al., 2012).

Statistical Analysis

Statistical analysis of differences in tumor frequency was carried out using

two-way ANOVA with repeated measures. Data sets were analyzed for statis-

tical significance using a two-tailed unpaired Student’s t test. All p values

below 0.05 were considered significant. All statistical analyses were done

using Microsoft Exel, GraphPad Prism 5, and SPSS Version 7.

Additional details regarding the experimental procedures can be found in

the Extended Experimental Procedures and Tables S1 and S2.

SUPPLEMENTAL INFORMATION

Supplemental Information includes Extended Experimental Procedures, six

figures, and two tables and can be found with this article online at http://dx.

doi.org/10.1016/j.celrep.2013.06.037.

Cell Reports 4, 340–351, July 25, 2013 ª2013 The Authors 349

ACKNOWLEDGMENTS

We are grateful for excellent technical support from T.M. Brocks, V. Czichow-

ski, D. Fehrenschild, R. Hoppe, and A. Kraus (CMMC, University of Cologne),

and expert lipid analysis by S. Brodesser (CECAD and IMMI, University of Co-

logne). We thank R. Toftgard (Karolinska Institutet, Stockholm) for the Gli

expression plasmids, H. Sasaki (RIKEN, Kobe) for Gli reporter constructs, L.

Sorokin (University of Munster) for the laminin antibody, M. Wegner (University

of Erlangen) for the Sox9 antibody, C. Mauch and P. Zigrino (Z2, SFB 829) for

human tumor tissues, and S. Iden (CECAD, Cologne), D.M. Owens (Columbia

University), and S. Wickstrom (MPI, Cologne) for critically discussing the

manuscript. This work was supported by the German Research Foundation

(DFG NI 667/2-1, SFB 829). P.K. performed most of the experiments and

data analysis. K.R., G.S., C.W., P.S., and D.F. performed a significant amount

of the work. C.C.Z. provided a cell line, and B.L. generated and provided the

Ihhfl/fl mouse line. C.N. designed the experiments and wrote the manuscript.

Received: December 22, 2012

Revised: May 10, 2013

Accepted: June 24, 2013

Published: July 18, 2013

REFERENCES

Anderson, E., Peluso, S., Lettice, L.A., and Hill, R.E. (2012). Human limb abnor-

malities caused by disruption of hedgehog signaling. Trends Genet. 28,

364–373.

Atwood, S.X., Chang, A.L.S., and Oro, A.E. (2012). Hedgehog pathway inhibi-

tion and the race against tumor evolution. J. Cell Biol. 199, 193–197.

Barel, A.O., and Clarys, P. (1995). Study of the stratum corneum barrier func-

tion by transepidermal water loss measurements: comparison between two

commercial instruments: Evaporimeter and Tewameter. Skin Pharmacol. 8,

186–195.

Braun, K.M., Niemann, C., Jensen, U.B., Sundberg, J.P., Silva-Vargas, V., and

Watt, F.M. (2003). Manipulation of stem cell proliferation and lineage commit-

ment: visualisation of label-retaining cells in wholemounts ofmouse epidermis.

Development 130, 5241–5255.

Brownell, I., Guevara, E., Bai, C.B., Loomis, C.A., and Joyner, A.L. (2011).

Nerve-derived sonic hedgehog defines a niche for hair follicle stem cells

capable of becoming epidermal stem cells. Cell Stem Cell 8, 552–565.

Lo Celso, C., Berta, M.A., Braun, K.M., Frye, M., Lyle, S., Zouboulis, C.C., and

Watt, F.M. (2008). Characterization of bipotential epidermal progenitors

derived from human sebaceous gland: contrasting roles of c-Myc and beta-

catenin. Stem Cells 26, 1241–1252.

Chiang, C., Swan, R.Z., Grachtchouk, M., Bolinger, M., Litingtung, Y., Robert-

son, E.K., Cooper, M.K., Gaffield, W., Westphal, H., Beachy, P.A., and Dlu-

gosz, A.A. (1999). Essential role for Sonic hedgehog during hair follicle

morphogenesis. Dev. Biol. 205, 1–9.

Chung, U.I., Schipani, E., McMahon, A.P., and Kronenberg, H.M. (2001). Indian

hedgehog couples chondrogenesis to osteogenesis in endochondral bone

development. J. Clin. Invest. 107, 295–304.

DiGiovanni, J. (1992). Multistage carcinogenesis in mouse skin. Pharmacol.

Ther. 54, 63–128.

Fehrenschild, D., Galli, U., Breiden, B., Bloch, W., Schettina, P., Brodesser, S.,

Michels, C., Gunschmann, C., Sandhoff, K., Niessen, C.M., and Niemann, C.

(2012). TCF/Lef1-mediated control of lipid metabolism regulates skin barrier

function. J. Invest. Dermatol. 132, 337–345.

Fu, X., Yang, X., Li, J., Tian, X., Cai, J., and Zhang, Y. (2010). Opposite expres-

sion patterns of Sonic hedgehog and Indian hedgehog are associated with

aberrantmethylation status of their promoters in colorectal cancers. Pathology

42, 553–559.

Gritli-Linde, A., Hallberg, K., Harfe, B.D., Reyahi, A., Kannius-Janson, M., Nils-

son, J., Cobourne, M.T., Sharpe, P.T., McMahon, A.P., and Linde, A. (2007).

350 Cell Reports 4, 340–351, July 25, 2013 ª2013 The Authors

Abnormal hair development and apparent follicular transformation to mam-

mary gland in the absence of hedgehog signaling. Dev. Cell 12, 99–112.

Hafner, M., Wenk, J., Nenci, A., Pasparakis, M., Scharffetter-Kochanek, K.,

Smyth, N., Peters, T., Kess, D., Holtkotter, O., Shephard, P., et al. (2004).

Keratin 14 Cre transgenic mice authenticate keratin 14 as an oocyte-ex-

pressed protein. Genesis 38, 176–181.

Ingham, P.W., Nakano, Y., and Seger, C. (2011). Mechanisms and functions of

Hedgehog signalling across the metazoa. Nat. Rev. Genet. 12, 393–406.

Jensen, K.B., Collins, C.A., Nascimento, E., Tan, D.W., Frye, M., Itami, S., and

Watt, F.M. (2009). Lrig1 expression defines a distinct multipotent stem cell

population in mammalian epidermis. Cell Stem Cell 4, 427–439.

Joeng, K.S., and Long, F. (2009). The Gli2 transcriptional activator is a crucial

effector for Ihh signaling in osteoblast development and cartilage vasculariza-

tion. Development 136, 4177–4185.

Kobayashi, T., Soegiarto, D.W., Yang, Y., Lanske, B., Schipani, E., McMahon,

A.P., and Kronenberg, H.M. (2005). Indian hedgehog stimulates periarticular

chondrocyte differentiation to regulate growth plate length independently of

PTHrP. J. Clin. Invest. 115, 1734–1742.

Kosinski, C., Stange, D.E., Xu, C., Chan, A.S., Ho, C., Yuen, S.T., Mifflin, R.C.,

Powell, D.W., Clevers, H., Leung, S.Y., and Chen, X. (2010). Indian hedgehog

regulates intestinal stem cell fate through epithelial-mesenchymal interactions

during development. Gastroenterology 139, 893–903.

Lanske, B., Karaplis, A.C., Lee, K., Luz, A., Vortkamp, A., Pirro, A., Karperien,

M., Defize, L.H., Ho, C., Mulligan, R.C., et al. (1996). PTH/PTHrP receptor in

early development and Indian hedgehog-regulated bone growth. Science

273, 663–666.

Lauth, M., Bergstrom, A., Shimokawa, T., and Toftgard, R. (2007). Inhibition of

GLI-mediated transcription and tumor cell growth by small-molecule antago-

nists. Proc. Natl. Acad. Sci. USA 104, 8455–8460.

Lee, K., Jeong, J., Kwak, I., Yu, C.-T., Lanske, B., Soegiarto, D.W., Toftgard,

R., Tsai, M.-J., Tsai, S., Lydon, J.P., and DeMayo, F.J. (2006). Indian hedgehog

is a major mediator of progesterone signaling in the mouse uterus. Nat. Genet.

38, 1204–1209.

Maeda, Y., Nakamura, E., Nguyen, M.-T., Suva, L.J., Swain, F.L., Razzaque,

M.S., Mackem, S., and Lanske, B. (2007). Indian Hedgehog produced by post-

natal chondrocytes is essential for maintaining a growth plate and trabecular

bone. Proc. Natl. Acad. Sci. USA 104, 6382–6387.

Mangelberger, D., Kern, D., Loipetzberger, A., Eberl, M., and Aberger, F.

(2012). Cooperative Hedgehog-EGFR signaling. Front. Biosci. 17, 90–99.

McMahon, A.P., Ingham, P.W., and Tabin, C.J. (2003). Developmental

roles and clinical significance of hedgehog signaling. Curr. Top. Dev. Biol.

53, 1–114.

Mill, P., Mo, R., Fu, H., Grachtchouk, M., Kim, P.C.W., Dlugosz, A.A., and Hui,

C.C. (2003). Sonic hedgehog-dependent activation of Gli2 is essential for

embryonic hair follicle development. Genes Dev. 17, 282–294.

Ng, J.M.Y., and Curran, T. (2011). The Hedgehog’s tale: developing strategies

for targeting cancer. Nat. Rev. Cancer 11, 493–501.

Niemann, C., Owens, D.M., Hulsken, J., Birchmeier, W., andWatt, F.M. (2002).

Expression of DeltaNLef1 in mouse epidermis results in differentiation of

hair follicles into squamous epidermal cysts and formation of skin tumours.

Development 129, 95–109.

Niemann, C., Unden, A.B., Lyle, S., Zouboulis, ChC., Toftgard, R., and Watt,

F.M. (2003). Indian hedgehog and beta-catenin signaling: role in the seba-

ceous lineage of normal and neoplastic mammalian epidermis. Proc. Natl.

Acad. Sci. USA 100(Suppl 1), 11873–11880.

Niemann, C., Owens, D.M., Schettina, P., and Watt, F.M. (2007). Dual role of

inactivating Lef1 mutations in epidermis: tumor promotion and specification

of tumor type. Cancer Res. 67, 2916–2921.

Nijhof, J.G.W., Braun, K.M., Giangreco, A., van Pelt, C., Kawamoto, H., Boyd,

R.L., Willemze, R., Mullenders, L.H.F., Watt, F.M., de Gruijl, F.R., and van

Ewijk, W. (2006). The cell-surface marker MTS24 identifies a novel population

of follicular keratinocytes with characteristics of progenitor cells. Development

133, 3027–3037.

Oro, A.E., and Higgins, K. (2003). Hair cycle regulation of Hedgehog signal

reception. Dev. Biol. 255, 238–248.

Outram, S.V., Hager-Theodorides, A.L., Shah, D.K., Rowbotham, N.J., Drako-

poulou, E., Ross, S.E., Lanske, B., Dessens, J.T., and Crompton, T. (2009).

Indian hedgehog (Ihh) both promotes and restricts thymocyte differentiation.

Blood 113, 2217–2228.

Pasca di Magliano, M., and Hebrok, M. (2003). Hedgehog signalling in cancer

formation and maintenance. Nat. Rev. Cancer 3, 903–911.

Raymond, K., Richter, A., Kreft, M., Frijns, E., Janssen, H., Slijper, M., Praetzel-

Wunder, S., Langbein, L., and Sonnenberg, A. (2010). Expression of the orphan

protein Plet-1 during trichilemmal differentiation of anagen hair follicles.

J. Invest. Dermatol. 130, 1500–1513.

Razzaque, M.S., Soegiarto, D.W., Chang, D., Long, F., and Lanske, B. (2005).

Conditional deletion of Indian hedgehog from collagen type 2alpha1-express-

ing cells results in abnormal endochondral bone formation. J. Pathol. 207,

453–461.

Reichelt, J., Breiden, B., Sandhoff, K., and Magin, T.M. (2004). Loss of keratin

10 is accompanied by increased sebocyte proliferation and differentiation.

Eur. J. Cell Biol. 83, 747–759.

Saqui-Salces, M., and Merchant, J.L. (2010). Hedgehog signaling and gastro-

intestinal cancer. Biochim. Biophys. Acta 1803, 786–795.

Scales, S.J., and de Sauvage, F.J. (2009). Mechanisms of Hedgehog pathway

activation in cancer and implications for therapy. Trends Pharmacol. Sci. 30,

303–312.

Silva-Vargas, V., Lo Celso, C., Giangreco, A., Ofstad, T., Prowse, D.M., Braun,

K.M., and Watt, F.M. (2005). Beta-catenin and Hedgehog signal strength can

specify number and location of hair follicles in adult epidermis without recruit-

ment of bulge stem cells. Dev. Cell 9, 121–131.

Sorokin, L.M., Pausch, F., Durbeej, M., and Ekblom, P. (1997). Differential

expression of five laminin alpha (1–5) chains in developing and adult mouse

kidney. Dev. Dyn. 210, 446–462.

St-Jacques, B., Dassule, H.R., Karavanova, I., Botchkarev, V.A., Li, J., Danie-

lian, P.S., McMahon, J.A., Lewis, P.M., Paus, R., and McMahon, A.P. (1998).

Sonic hedgehog signaling is essential for hair development. Curr. Biol. 8,

1058–1068.

St-Jacques, B., Hammerschmidt, M., and McMahon, A.P. (1999). Indian

hedgehog signaling regulates proliferation and differentiation of chondrocytes

and is essential for bone formation. Genes Dev. 13, 2072–2086.

Stolt, C.C., Lommes, P., Sock, E., Chaboissier, M.-C., Schedl, A., andWegner,

M. (2003). The Sox9 transcription factor determines glial fate choice in the

developing spinal cord. Genes Dev. 17, 1677–1689.

Takeda, H., Lyle, S., Lazar, A.J.F., Zouboulis, C.C., Smyth, I., and Watt, F.M.

(2006). Human sebaceous tumors harbor inactivating mutations in LEF1.

Nat. Med. 12, 395–397.

Teglund, S., and Toftgard, R. (2010). Hedgehog beyond medulloblastoma and

basal cell carcinoma. Biochim. Biophys. Acta 1805, 181–208.

Tennenbaum, T., Weiner, A.K., Belanger, A.J., Glick, A.B., Hennings, H., and

Yuspa, S.H. (1993). The suprabasal expression of alpha 6 beta 4 integrin is

associated with a high risk for malignant progression in mouse skin carcino-

genesis. Cancer Res. 53, 4803–4810.

van den Brink, G.R. (2007). Hedgehog signaling in development and homeo-

stasis of the gastrointestinal tract. Physiol. Rev. 87, 1343–1375.

van den Brink, G.R., Bleuming, S.A., Hardwick, J.C.H., Schepman, B.L., Offer-

haus, G.J., Keller, J.J., Nielsen, C., Gaffield, W., van Deventer, S.J.H., Roberts,

D.J., and Peppelenbosch, M.P. (2004). Indian Hedgehog is an antagonist of

Wnt signaling in colonic epithelial cell differentiation. Nat. Genet. 36, 277–282.

VanWaes, C., Surh, D.M., Chen, Z., Kirby,M., Rhim, J.S., Brager, R., Sessions,

R.B., Poore, J., Wolf, G.T., and Carey, T.E. (1995). Increase in suprabasilar

integrin adhesion molecule expression in human epidermal neoplasms ac-

companies increased proliferation occurring with immortalization and tumor

progression. Cancer Res. 55, 5434–5444.

Wang, L.C., Liu, Z.Y., Gambardella, L., Delacour, A., Shapiro, R., Yang, J.,

Sizing, I., Rayhorn, P., Garber, E.A., Benjamin, C.D., et al. (2000). Regular

articles: conditional disruption of hedgehog signaling pathway defines its

critical role in hair development and regeneration. J. Invest. Dermatol. 114,

901–908.

Wickstrom, M., Dyberg, C., Shimokawa, T., Milosevic, J., Baryawno, N.,

Fuskevag, O.M., Larsson, R., Kogner, P., Zaphiropoulos, P.G., and Johnsen,

J.I. (2012). Targeting the hedgehog signal transduction pathway at the level

of GLI inhibits neuroblastoma cell growth in vitro and in vivo. Int. J. Cancer

132, 1516–1524.

Woo, W.-M., Zhen, H.H., and Oro, A.E. (2012). Shh maintains dermal papilla

identity and hair morphogenesis via a Noggin-Shh regulatory loop. Genes

Dev. 26, 1235–1246.

Xiong, Y., Li, W., Shang, C., Chen, R.M., Han, P., Yang, J., Stankunas, K., Wu,

B., Pan, M., Zhou, B., et al. (2013). Brg1 governs a positive feedback circuit in

the hair follicle for tissue regeneration and repair. Dev. Cell 25, 169–181.

Zouboulis, C.C., Seltmann, H., Neitzel, H., and Orfanos, C.E. (1999). Establish-

ment and characterization of an immortalized human sebaceous gland cell line

(SZ95). J. Invest. Dermatol. 113, 1011–1020.

Cell Reports 4, 340–351, July 25, 2013 ª2013 The Authors 351