carcinogen-induced skin tumor development requires...

Research Article

Carcinogen-Induced Skin Tumor Development RequiresLeukocytic Expression of the Transcription Factor Runx3

Omri Bauer, Shay Hantisteanu, Joseph Lotem, and Yoram Groner

AbstractCarcinogen-induced skin tumorigenesis depends heavily on proinflammatory tumor-promoting pro-

cesses. Here, we show that leukocytic Runx3 expression is central to the two-stage DMBA/TPA-induced skin

tumorigenesis. Runx3-null mice were highly resistant to this process and concomitant ablation of Runx3 in

dendritic and T cells fully recapitulated this resistance. Mechanistically, this resistance was associated with a

shift in the skin cytokine milieu toward a tumor nonpermissive microenvironment. Specifically, leukocytic

Runx3 loss substantially increased the antitumorigenic cytokine thymic stromal lymphopoietin (TSLP) and

profoundly decreased two protumorigenic cytokines, interleukin-17a and osteopontin. Therefore, inflam-

mation-mediated tumor promotion requires leukocytic Runx3 expression, as its loss creates a unique

cytokine composition that polarizes the tumormicroenvironment to a potent antitumorigenic state.Cancer

Prev Res; 7(9); 913–26. �2014 AACR.

IntroductionThe two-stage (DMBA/TPA) skin tumorigenesis assay has

been instrumental to our understanding of carcinogenesisas a multistep process (1, 2). Like in other epithelial cancermodels, tumors induced by this assay evolve through pri-mary genetic and epigenetic changes that are propagatedthrough interactions with the tumor microenvironment(3). Tumor-promotingmechanisms are only partly defined,involving chronic recruitment of inflammatory cells, fibro-blasts, and endothelial cells that reside within the tumorstroma and contribute to its survival and proliferation(4, 5). Although inflammatory cells in the tumor microen-vironment can eliminate initiated cells and suppress tumorgrowth (6), there is increasing recognition of their tumor-promoting roles (3, 7).The tumor microenvironment cytokine/chemokine

milieu is a major determinant of cancer-related inflamma-tion. It coordinates communication between transformedcells and supporting stroma, thereby affecting severalaspects of tumor promotion and progression, includingimmune cell recruitment and function, tumor cell senes-cence, proliferation and survival,matrix remodeling, angio-genesis, and tumor invasion and metastasis (8, 9). Thus,tumor bed cytokines can polarize host immunity to either

inflammatory-promotive or -protective responses, leadingto one of two opposite outcomes: tumor progression orregression (10, 11). Therefore, there is a great clinicalpromise in modulating the tumor microenvironment cyto-kine repertoire so that immune response is shifted toward atumor-protective mode (8, 12).

The runt-related transcription factor Runx3 is highlyexpressed in murine skin leukocytes and mesenchyme,including hair follicle dermal papillae, but not in epithelia(13, 14). In lymphocytes, Runx3 is involved in Cd4 repres-sion in differentiating CD4þ/CD8þ thymocytes (15, 16), inregulation of cytokine gene expression in CD4þ Th1 lym-phocytes (17), and in growth/terminal differentiation ofmature CD8þ T cells (15). Runx3 is also mandatory for theperinatal homing of dendritic epidermal T cells (DETC) tothe epidermis, as reflected by their complete loss in Runx3-null mice (18). In dendritic cells (DC), Runx3 is part of thetumor growth factor b (TGFb) signaling cascade (19, 20).When lost, the development of TGFb-dependent epidermalLangerhans cells (LC) is abrogated, whereas dermal DCs(dDC) seem unaffected (18, 19). Runx3 germline inactiva-tion enhanced lung DC maturation, resulting in hyper-mature/-active alveolar DCs (19, 20) and reprogrammedgene expression of CD11bþEsamhigh splenic DCs (21).Despite these profound immune anomalies, in vivo studiesof wound closure, contact hypersensitivity, and epidermalapoptosis were unaltered in skins of Runx3-null mice (18).Runx3 is also highly expressed in respiratory and gastroin-testinal tract leukocytes, but is absent from their epithelia(13, 22, 23). Significantly, Runx3-null mice spontaneouslydevelop airway and colon inflammation (19, 20, 22), attest-ing toRunx3 involvement in leukocyte homeostasis in thesetissues. These earlier findings prompted us to investigate thepotential role of Runx3 in inflammation-mediated tumorpromotion in the skin.

Department of Molecular Genetics, The Weizmann Institute of Science,Rehovot, Israel.

Note:Supplementary data for this article are available atCancer PreventionResearch Online (http://cancerprevres.aacrjournals.org/).

Corresponding Author: Yoram Groner, Department of Molecular Genet-ics, The Weizmann Institute of Science, 25 Herzel Street, Rehovot 76100,Israel. Phone: 972-8-9343972; Fax: 972-8-9344108; E-mail:[email protected]

doi: 10.1158/1940-6207.CAPR-14-0098-T

�2014 American Association for Cancer Research.

CancerPreventionResearch

www.aacrjournals.org 913

Research. on May 26, 2019. © 2014 American Association for Cancercancerpreventionresearch.aacrjournals.org Downloaded from

Published OnlineFirst June 24, 2014; DOI: 10.1158/1940-6207.CAPR-14-0098-T

http://cancerpreventionresearch.aacrjournals.org/

Materials and MethodsMice

The mice (2–6months old) used in the study include thepreviously described strains: Runx3-null (ICR background;ref. 24); loxP-flanked Runx3 (Runx3lox/lox) and Runx3fl/fl/Pgk-cre (25); and Runx3P1-AFP/P2-GFP (23) and Runx3fl/fl/CD11c-cre (21). Transgenic Lck- and K14-cre mice wereobtained from The Jackson Laboratory. Transgenic hemi-zygous TG.ACmice (FVBbackground)werepurchased fromTaconic and crossed with Runx3-null and wild-type (WT)mice to yield Runx3-null/TG.AC and WT/TG.AC controlmice (using FVB/ICR F1 only). Control mice, that is, WTfor Runx3-null and Runx3þ/þ for all other transgenic mice(unless stated otherwise), were gender-matching littermatesfrom heterozygote breeding. Specific ablation of Runx3in dDCs (DCRunx3�/� mice) or in basal keratinocytes(Runx3fl/fl/K14-cre mice) was verified by PCR onDNA fromCD11chigh FACS-sorted dermal or DETC/LC–depleted epi-dermal cell suspensions, respectively, isolated using theArcturus PicoPure DNA Extraction Kit (Applied Biosys-tems). Cell-specific Runx3 ablation in sorted splenicCD8þ thymocytes from Runx3fl/fl/Lck-cre mice was validat-ed by western blotting, using an in-house polyclonal anti-Runx3 antibody (13). Mice husbandry details are given inSupplementary Materials and Methods.

Immunohistochemistry and histologyRunx3 immunohistochemistry (IHC) was performed on

formalin-fixed paraffin-embedded sections, as describedpreviously (16), using an in-house polyclonal anti-Runx3antibody (1:1,000; ref. 13). Further information aboutassessments of epidermal hyperplasia, keratinocyte prolif-eration, and apoptosis is included in Supplementary Mate-rials and Methods.

Flow cytometryEpidermal or dermal single-cell suspensions were gener-

ated from cohorts of mice (n � 3), as described previously(26). Further details are included in Supplementary Materi-als and Methods.

Chemical carcinogenesis assayThe DMBA/TPA skin tumorigenesis assay was conducted

as described previously (27), including several modifica-tions, as detailed in Supplementary Materials andMethods.

Epidermal sheet preparation, immunostaining, and LCquantification

Epidermal sheets were prepared from adult mouse ears asdescribed previously (19), including modificationsdescribed in Supplementary Materials and Methods.

Cytokine analysisNa€�ve and DMBA/TPA–treated (single/6 biweekly appli-

cations, respectively) whole-skin biopsies of Runx3-nulland control WT mice (pooled; n ¼ 3 mice/genotype) wereweighed, snap-frozen, and stored (�80�C) for further pro-cessing. Tissue protein lysate (R&D Systems/ARY006) and

RNA (Qiagen/RNeasyMicroarray Tissue Kit) were obtainedaccording to the manufacturers’ instructions and analyzedusing the Proteome Profiler/Mouse Cytokine Array Dot-Blot Kit (R&D Systems), the RT2 Profiler PCR Array System/Mouse Cytokines and Chemokines (SABiosciences/PAMM-150Z) and qRT-PCR, as detailed in Supplementary Materi-als and Methods.

Statistical analysisData were presented as mean� standard deviation (SD).

Datasets were compared using the Student two-tailed t test.Statistical significance was considered at P < 0.05.

ResultsRunx3 expression in mouse skin

To evaluate Runx3 expression in mouse skin, we per-formed IHC on normal and neoplastic skin sections. In linewith previous studies (13, 14), we found that Runx3 expres-sion was confined to large scattered cells of the interfolli-cular epidermis and to distinct hair follicle cells (Fig. 1A).However, Runx3 was not detected in either embryonic orpostnatal (normal or neoplastic) epidermal keratinocytes(Fig. 1A). We noted intense staining of dispersed morpho-logically heterogeneous cells in normal and pathologicdermis of adult mice (Fig. 1A), and therefore performedflow cytometry on single-cell suspensions from dermis andepidermis of Runx3/GFP knockin mice (23). Analysisrevealed intense expression of Runx3/GFP in resident Tcells and CD11bþ DCs (Fig. 1B), two well-establishedRunx3-expressing cell lineages (16, 19, 21, 23). Similarepidermal analysis identified Runx3 expression in subsetsof steady-state LCs and DETCs (Fig. 1C), correspondingwith their complete absence in Runx3-null mice (18, 19).Together, these findings demonstrated Runx3 expression inskin leukocytes and lack of its expression in normal orneoplastic epidermal keratinocytes.

Significant reduction in CD11bþ DCs in na€�ve Runx3-null dermis

Steady-state murine dermis contains two distinct DCsubtypes—"classical" dermal (d) (CD11bþ) and langerinþ

(CD11blow/neg) DCs—that share theCD11c pan-DCmarker(28). The aE-b7 integrin CD103, found on DC subsets invarious organs, including lung (29) and colon (30), furtherspecifies langerinþ dDCs. To evaluate changes in the pro-portion of dDC subsets in Runx3-null mice, we performedflow cytometry analysis on ear skin-isolated cells. Na€�veRunx3-null skin displayed a significant decrease inCD11bþ dDCs (30.3% � 11.6%), compared with WT con-trols (62.6%� 10.2%; P < 0.01; Fig. 1D, left). Furthermore,while the proportion ofCD11blow/negCD103þ dDCsdidnotchange significantly, approximately five times increase in auniquepopulationofCD11bþ/CD103þ dDCswas recordedin Runx3-null mice (27.3% � 1.6% vs. 4.9%� 4.2% inWTcontrols;P < 0.01; Fig. 1D, left andmiddle). DCmaturation/activation state analyses indicated comparable MHCII andT-cell costimulatory molecule CD86 levels in Runx3-null

Bauer et al.

Cancer Prev Res; 7(9) September 2014 Cancer Prevention Research914




A

B

C

Naï

ve s

kin

Naïve dermis

Pap

illom

a

WT

IHC, ×200

50 μm 50 μm

2 mm

IHC, ×400

IHC, ×50H&E, ×10

DD

E

E

HF

pK

HFSC

GFP

FS

C

CD

11c

0 102 103 104 105

0.62

1.28

0

0

102

102

103

103

104

104

105

105 0 102 103 104 105

1.31

0

50K

100K

150K

200K

250K

FS

C

SS

C

0

50K

100K

150K

200K

250K

CD3e0 102 103 104 105

0.47

GFP CD103

CD

11c

0.0845

85.40

0

102

102

103

103

104

104

00

102

2

101

101

103

103

104

104

105

105

GFP

WT Runx3-null

Runx3-nullWT

CD11b+/CD103+ MHC II

WT

% o

f Tot

al d

DC

s

Cou

nts

CD11b

CD11b0 102 103 104 105

1.83

0

0.5

1

1.5

2

2.5

Naïve epidermis

D

0 102 103 104 105

MHCII

CD

3e

0.049

0

0

102

102

103

103

104

104

105

105 0 102 103 104 105

GFP γδ TCR

0

2

4

6

8

0 102101 103 104 0 102101 103 104

15.2 28.9

23.50

5

10

15

20

25

30

0

20

40

60

80

100

CD86

0 102101 103 104

Runx3-null

100 μm200 μm

IHC, ×100

pK

111110

5.0 2.6

65.327.0

CD

103

32.4

Figure 1. Runx3 expression in naïve and papillomatotic skin leukocytes and changes in immunocytes in Runx3-null skin. A, top, IHC showing intense Runx3expression in scattered nonkeratinocytic cells of naïveWT skin. Note stained nuclei in large ovoid cells in basal epidermis (red arrows), in distinct hair follicle cellsand in dispersed heterogeneous dermal cells (black arrows), but not in keratinocytes. Dorsal skin, 2-month-old WT mouse, �200 (left); enlarged micrographof boxed area, �400 (right); E, epidermis; D, dermis; SC, subcutis (hypodermis); HF, hair follicle. Bottom, a section through a cauliflower-shaped benignpapilloma showing outwardly proliferating keratinocytes. H&E,�10 (left); Runx3 IHCon a papilloma showingmarked infiltration of Runx3-expressing leukocytesinto the tumor stroma (black arrows) in�50and�100micrographsof boxed area. Note lackofRunx3 in neoplastic keratinocytes (red arrows);pK, papillomatotickeratinocytes. Runx3 expression in CD3ehigh T cells and CD11bþ DCs from naïve WT dermis (B) and in LCs and DETCs (C) from epidermis. Shown areflowcytometry analyses of cells fromRunx3P1-AFP/P2-GFP knockinmice, depictingGFP-expressingCD3ehigh, CD103þ/CD11c� T cells, and theCD11cþ/CD11bþ

dDCsubsets (B) andCD11cþ/MHCIIþLCsandCD3ehigh/gdTCRþDETCs (C).D,flowcytometryanalysisofWTandRunx3-nulldDCsubsets innaïvedermis.Notemarked reduction in CD11bþ dDCs and CD11bþ/CD103þ dDC emergence (�5� more prevalent) in Runx3-null mice (values are mean � SD; n ¼ 4mice/genotype; P < 0.01). Right, comparative MHCII and CD86 levels in dDCs. Results in B–D represent one of three experiments with similar findings.

Leukocytic Runx3 Is Crucial for Skin Tumorigenesis

www.aacrjournals.org Cancer Prev Res; 7(9) September 2014 915




andWT dDCs (Fig. 1D, right), suggesting that in contrast tolung and splenic Runx3-null DCs (19), na€�ve Runx3-nulldDCs are neither hypermature nor hyperactivated.

Runx3 loss markedly attenuates DMBA/TPA–inducedpapilloma outgrowth

Despite the above-mentioned anomalies in major com-ponents ofRunx3-null cutaneous immunity, noquantifiableskin deficit was noted in a series of in vivo experimentsaddressing wound closure, contact hypersensitivity, andepidermal apoptosis (18). To investigate Runx3 role in skintumorigenesis, we compared tumor development in Runx3-null and WT mice, using the two-stage tumorigenesis pro-tocol (1). Our results revealed a striking resistance of Runx3-null mice to DMBA/TPA tumorigenesis (Fig. 2A and B),reflected by (i) an order-of-magnitude lower mean papillo-ma yield (multiplicity) in Runx3-null vs. WT control mice(peakof 1.1�1.8 vs. 12.1�6.9papillomas/mouse at17and20weeks after DMBA initiation, respectively; P < 0.01); (ii) adoubled period tofirst papilloma appearance (92 vs. 46 daysfrom DMBA initiation, respectively); and (iii) a decreasedrate of papilloma appearance in Runx3-null mice (0% vs.90% of mice bearing �1 papilloma at 8 weeks after DMBAinitiation, respectively). Furthermore, tumors that did formgrew significantly slower in the absence of Runx3, indicatingthat Runx3 is involved in papilloma formation and growth.Similar results were obtained when Runx3fl/fl/Pgk-cre mice,Runx3-null functional analogs, were subjected to DMBA/TPA treatment (Fig. 2C). Compared with WT, these miceyielded reducedmultiplicity (peak of 0.6� 0.8 vs. 12.6� 8.3papillomas/mouse at 20 weeks after DMBA initiation;P < 0.01), a significantly lower papilloma rate (0% vs.100% and 43% vs. 100% of mice bearing �1 papilloma at10 and 20 weeks after DMBA initiation, respectively) and anincreased latency period (99 compared with 39 days). Col-lectively, these data showed that Runx3-null mice exhibit aprofound resistance to DMBA/TPA–induced tumorigenesis.

Although our analyses failed to detect Runx3 in murinekeratinocytes, we repeatedly observed its presence in hairfollicle cells. Given that keratinocyte stem cells, residing inthe hair follicle bulge and basal interfollicular epidermis,are the primary cellular targets of the DMBA-induced ini-tiation phase of carcinogenesis (2), we addressed whetherlack of Runx3 in these cells contributed to the observedDMBA/TPA resistance. Using Runx3fl/fl/K14-cre mice thatlackRunx3 in thehair follicle and epidermal basal layer cells(Supplementary Fig. S1), we found similar papilloma yield,rate, and kinetics in these mice and their WT littermates(Fig. 2D), indicating that lack of Runx3 in epidermal ker-atinocytes was not involved in the observed tumor resis-tance of Runx3-null mice.

Runx3 loss mildly increases 30-methylcholanthrene–induced carcinogenesis

To evaluate Runx3 function in other neoplastic settings,we assessed fibrosarcoma induction by 30-methylcholan-threne (30-MCA), a complete carcinogen that mediates itsaction solitarily (31). Although 30-MCA carcinogenicity

does not involve a substantial inflammatory component(32–34), adaptive immunity plays a role in preventing late-induced sarcomas in this model (35). Mice were injectedsubcutaneously with 30-MCA (50 or 100 mg) and inspectedtwice-weekly for tumor development (Fig. 2E). In contrastto their resistance to DMBA/TPA, Runx3-null mice dis-played a mildly increased susceptibility, compared withWT littermates (Fig. 2F). This was reflected by a higher finitetumors/injected site in Runx3-null versus WT (22.7% vs.14.3% and 50.0% vs. 22.7%, in the low- and high-dosegroups, respectively) and an earlier appearance of firsttumor (106 vs. 141 days and 89 vs. 105 days in the low-and high-dose groups, respectively). Once established,however, sarcoma growth behavior was similar betweenRunx3-null andWTmice (Fig. 2F). Thus, Runx3 loss resultsin a mildly increased susceptibility to 30-MCA–inducedtumorigenicity, conceivably due to Runx3-null impairedimmunosurveillance-mediated immune response againstthe malignant sarcoma cells (6, 35). Taken together, these30-MCA results, the opposed DMBA/TPA findings, andRunx3 skin expression pattern indicated that while leu-kocytic Runx3 is required for DMBA/TPA inflammation-dependent tumorigenesis, it is nonessential for inflamma-tion-independent chemical carcinogenesis.

Altered TPA-mediated tumor promotion in Runx3-nullmice

To address whether papillomagenesis suppression inRunx3-null mice stemmed from a failure in DMBA-medi-ated initiation or in TPA-induced tumor promotion, wecrossed Runx3-null mice onto the TG.AC strain, whichconstitutively expresses an activated Ha-ras gene (36).TG.AC mice are considered "pre-initiated," because theydevelop papillomas following topical TPA application,without prior DMBA initiation (36). Using a biweeklyTPA promotion schedule, WT/TG.AC mice developed largepapillomas, as of 4 weeks after the first TPA application(Fig. 3A), consistent with their increased sensitivity (36). Incontrast, Runx3-null/TG.AC mice only occasionally devel-oped small papillomas that often regressed spontaneously(Fig. 3A). Thus, even in "pre-initiated" mice with a mutant/activatedHa-ras, Runx3 loss suppressed TPA-induced papil-lomagenesis, demonstrating that tumor promotion wasaffected by Runx3 loss.

Runx3 ablation decreases TPA-induced epidermalproliferation

To determine whether Runx3 absence affects keratinocyteproliferation, we assessed changes in TPA-induced epider-mal hyperplasia. The number of nucleated cell layers andthickness of untreated Runx3-null and WT epidermis werenot significantly different (Fig. 3B, top left). Although TPAapplication induced epidermal hyperplasia in both geno-types, the number of nucleated cell layers in Runx3-nullmicewas lower as comparedwithWTmice (Fig. 3B, top left).Specifically, following 2 or 10 TPA applications, Runx3-nullepidermal thickness has increased approximately 1.5-foldless than in WT mice (Fig. 3B, bottom left). These data

Bauer et al.





Mean papilloma/mouse

A

B

C

D

% Mice with papillomas

0

0

5

10

15

20 WT Runx3−/−

WT Runx3fi/fi/Pgk-cre

0

20

40

60

80

100

0

20

40

60

80

100

WT Runx3−/−

10

20

30

0

20

40

60

80

100

50 10 1515 2020

Weeks after DMBA application

0

5

10

15

50 10

Tumor incidence, % of injected sites

100 μg/site 50 μg/site

E

F

0

25

50

75

100

0 100 200 300

WT Runx3-null

0

25

50

75

100

0 100 200 300 Days after 3-MCA injection

E

D

M

FS×63 5 mm

×400

WTRunx3fl/fl/K14-cre

Figure 2. Runx3-null mice are resistantto DMBA/TPA tumorigenesis butsusceptible to 30-MCA–induced skintumors. A, representative mice showingmultiple papillomas on a WT versus asingle papilloma on a Runx3-null mouseat 20 weeks after DMBA initiation. B andC, delayed onset, reduced yield(multiplicity), and decreased rate ofpapillomas in DMBA/TPA–treatedRunx3-null (B) and Runx3fl/fl/Pgk-cremice (n ¼ 7, each), compared with WT(n ¼ 10) and control (n ¼ 9; bearing �1WT allele) littermate mice, respectively.D, similar papilloma yield, rate, andkinetics in DMBA/TPA–treatedRunx3fl/fl/K14-cre mice (n ¼ 10) andWT littermate mice (n ¼ 16). E, macroappearance of an autochthonousfibrosarcoma developing in a WTmouse lower back (left, circled). Asection through this tumor (right) showsits subcutaneous localization (H&E,�63) and the proliferating cancerousfibroblasts (inset, H&E �400; bar,50 mm); E, epidermis; D, dermis; M,hypodermal muscle; FS, fibrosarcoma.F, tumor formation kinetics andfrequency following 30-MCA injection.Shown are cumulative tumornumbers/injection sites (%) over time.Left, low dose (50 mg; 11 Runx3-nulland 14 WT mice). Right, high dose(100 mg; 10Runx3-null and 11WTmice).Values in B–D are mean � SD.






indicated that while Runx3 loss had no effect on na€�veepidermis, its ablation significantly attenuated TPA-inducedepidermal hyperplasia.

We then used bromodeoxyuridine (BrdUrd) IHC to eval-uate TPA-induced changes in keratinocyte proliferation.Whereas na€�ve Runx3-null mice exhibited BrdUrd labelingthat was similar to WTmice (1%–2% of stratum basale cells),TPA-treated skins of Runx3-null mice showed a significantlylower percentage of stratum basale BrdUrdþ cells (11.4% �8.2%) as compared with WT littermate mice (29.0% �4.5%); P < 0.01 (Fig. 3B, middle). These findings suggestedthat Runx3 loss abrogated TPA-induced keratinocyte prolif-eration. To determine whether enhanced keratinocyte apo-ptosis also contributed to Runx3-null mice resistance to

DMBA/TPA-induced papillomagenesis,we used the terminaldeoxynucleotidyltransferase–mediated dUTP nick end label-ing (TUNEL) assay to measure in vivo apoptosis. Analysisshowed that TPA-treated skins fromWTandRunx3-nullmicehad comparable numbers of epidermal TUNEL-positivenuclei (Fig. 3B, right top and bottom), indicating thatRunx3-nullmice lowerpapilloma incidence largely stemmedfrom diminished TPA-induced keratinocyte proliferation.

Runx3-null mice chronically inflamed skin showsaltered leukocyte recruitment

Toprobe into the cellularmicroenvironment that opposestumor promotion in Runx3-null mice, we quantified thecellular infiltrate in DMBA-initiated skin persistently treated

A

B

Figure 3. Decreased TPA-induced tumor promotion in Runx3-null mice. A, Runx3-null/TG.AC mice are resistant to TPA-induced papillomas, compared withWT/TG.AC mice. Left, representative mice at 15 weeks post-TPA showing multiple papillomas in a WT/TG.AC mouse versus null in a Runx3-null/TG.ACmouse. TPA was applied twice-weekly and papilloma multiplicity (middle) and rate (right) were monitored over a 15-week period. B, decreased TPA-inducedepidermal proliferation in Runx3-null mice. Top left, H&E-stained dorsal skin sections of approximately 5-month-old male mice (n ¼ 2, each) untreatedor after two (10 mg) TPA treatments (scale bars, 50 mm; E, epidermis; D, dermis). Bottom left, quantitative epidermal thickness analysis (n � 10 fields;�, P < 0.001). Middle, in vivo BrdUrd labeling of skins from the above naïve and twice-TPA–treated mice. Black arrows, BrdUrd-positive cells in interfollicularstratum basale of WT (top) and Runx3-null (middle) epidermis. Red arrows, scattered BrdUrdþ cells in the dermis (scale bars, 50 mm). Bottom, quantitativeanalysis of BrdUrdþ cells (n � 10 fields; �, P < 0.001). Right, epidermal cell apoptosis rate analysis in Runx3-null and WT littermate mice. TUNEL stainingin skin sections of TPA-treated mice. White arrows, TUNELþ cells (scale bars, 50 mm; E, epidermis; D, dermis). Bottom, quantitative analysis ofapoptotic (TUNELþ) cells (n � 10 fields at �200 magnification). Histogram values in B are mean � SD.

Bauer et al.





with TPA (8–10 biweekly applications) to establish chronicdermatitis. Flowcytometry revealedcomparableproportionsof several leukocyte populations, including dDC subsets,neutrophils, macrophages, and myeloid-derived suppressor

cells (MDSC) in Runx3-null and WT chronically inflameddermis (Supplementary Fig. S2A). However, a significantdifference was found in the dermal T-cell compartment(Fig. 4A). Despite similar overall numbers of CD3eþ T cells

DMBA/TPA-treated Runx3-null, dermis Dermal T-cell subsets

% o

f T c

ells

A

B

94.9

0.002

0.272

0

0

102

102

103

103

104

104

105

105

γδγδ T cells αβ T cells0

0102

103

104

105

50K 100K 150K 200K 250K

SSC

CD

3e

0102

103

104

105

CD

3e

CD

11c

SSC CD11b

CD

11c

0

0102

103

104

105

50K 100K 150K 200K 250K

0 50K 100K 150K 200K 250K

SSC

DCs (all) CD11b+ DCs

CD

3eC

D3e

αβ

0.586

0

0

102

102

103

103

104

104

105

105

0

0

102

102

103

103

104

104

105

105γδ

αβ

3.02

8.62

68.4

0

0

102

103

104

105

50K 100K 150K 200K 250K

SSC

DMBA/TPA-treated WT control, dermis

DMBA/TPA-treated Runx3-null, epidermis

Decreased DC numbers in thechronically inflamed epidermis

of Runx3-null mice

DMBA/TPA-treated WT control, epidermis

0102

103

104

105

1.12

CD

11c

SSC CD11b

CD

11c

0 50K 100K 150K 200K 250K

6.05

0

0

102

102

103

103

104

104

105

105

9.03

0102

103

104

105

0

20

40

60

80

100

4.16

0 50K 100K 150K 200K 250KSSC

Runx3-null

*

WT

*

0

2

4

6

8

10 Runx3-nullWT

*

% o

f Epi

derm

al c

ells

20.8

59.1

0

0

102

102

103

103

104

104

105

105γδ

αβ

Figure 4. Leukocyte distribution in Runx3-null DMBA/TPA–treated dermis. A, altered proportions of dermal gdþ and abþ T-cell populations in DMBA/TPA–treated Runx3-null, compared with WT littermate mice. Flow cytometry analysis of cells isolated from Runx3-null (top) or WT littermate (bottom) dermis.Top right, a histogram summarizing dermal T-cell subset proportions (�, P 0.02). B, reduced proportion of epidermal DCs in DMBA/TPA–treatedRunx3-null mice. Flow cytometry analysis of cells isolated from Runx3-mull (top) or WT littermate (bottom) epidermis. Right, a histogram summarizing theproportion of epidermal DC populations (�, P ¼ 0.03). Histogram values in A and B are mean � SD (n ¼ 3 mice/genotype).






in these two genotypes, wenoted a 2-fold decrease in the gdþ

T-cell subset (26.0% � 5.1% in WT vs. 12.0% � 2.5% inRunx3-null mice; P < 0.01) accompanied by a mild increasein the proportion of Runx3-null abþ T cells (69.9%� 4.9%in Runx3-null vs. 60.4% � 2.8% in WT mice; P ¼ 0.02;Fig. 4A). Furthermore, a CD3ehigh gdþ T-cell subset, whichcomprised 1% to 4% of inflamed WT dermal T cells, wasabsent from Runx3-null dermis (Fig. 4A). Similar analysesof the DMBA/TPA–treated epidermal compartment cor-responded with these dermal findings (SupplementaryFig. S2B). In the epidermis, the DC proportion decreased4-fold in Runx3-null mice (1.4% � 0.4% in Runx3-null vs.6.0% � 3.0% in WT mice; P ¼ 0.03; Fig. 4B). This decreasewas probably due to a decline in the CD11bþ DC subset(1.2% � 0.6% in Runx3-null vs. 3.5% � 2.3% in controlmice; P ¼ 0.07). Collectively, these data demonstrate thatRunx3 loss affected recruitment of T and DC lineage cellsubsets to the TPA-treated skin, attesting to the crucial role ofRunx3 in shaping the inflamed skin microenvironment.

DC- or T cell–specific Runx3 ablation partiallyphenocopy Runx3-null resistance to DMBA/TPA skintumorigenesis

Because Runx3 expression was confined to skin leuko-cytes, we addressed whether specific elimination of Runx3in DCs would endow resistance to DMBA/TPA–mediatedtumorigenesis. Upon DMBA/TPA treatment, Runx3fl/fl/CD11c-cre (DCRunx3�/�) mice (Fig. 5A, left) developed few-er papillomas/mouse than Runx3þ/þ littermates (11.9 �11.4 vs. 29.5 � 11.6 peak papillomas/mouse at 20 weeksafter DMBA initiation, respectively; P < 0.01; Fig. 5A, mid-dle). However, the percentage of DCRunx3�/� mice with �1papilloma did not differ significantly from controls (63%vs. 81% and 88% vs. 100%ofmice at 9.5 and 20weeks afterDMBA initiation, respectively; Fig. 5A, right). Furthermore,tumor latency period was 45 days in both DCRunx3�/� andRunx3þ/þ groups (Fig. 5A). These results suggested thatalthough Runx3-expressing DCs were essential, they werenot sufficient to furnish Runx3-null full tumor resistance.

Modi and colleagues (37) showed that absence of epi-dermal LCs confers complete resistance to DMBA/TPA–induced tumorigenesis. Therefore, we evaluated theinvolvement of LCs in Runx3-null tumor resistance byimmunostaining of epidermal sheets from na€�veDCRunx3�/� ear skin and flow cytometry (Fig. 5B). Althoughmorphologically sound MHCIIþ LCs were present in na€�veDCRunx3�/� epidermis (Fig. 5B and Supplementary Fig. S3),their numbers were significantly lower, compared withRunx3þ/þmice (Fig. 5B, right top and bottom). This findingindicates that while DCRunx3�/� mice tumor resistanceoccurred en face low numbers of epidermal LCs, additionaldermal components contributed to Runx3-null miceresistance.

We then evaluated the involvement of T cells inRunx3-null mice tumor resistance. Treatment of Runx3fl/fl/Lck-cre (TRunx3�/�) mice (Fig. 6A, left) with DMBA/TPArevealed a papilloma yield that was lower, compared withlittermate Runx3þ/þ mice (peak of 4.4 � 5.5 vs. 9.5 � 7.8

papillomas/mouse at 20 weeks after DMBA initiation; P ¼0.03; Fig. 6A,middle). Papilloma rate in TRunx3�/�mice wasalso lower, compared with Runx3þ/þ control mice (17% vs.50% and 83% vs. 96% of mice bearing �1 papilloma at10 and20weeks afterDMBA initiation, respectively; Fig. 6A,right). However, like in DCRunx3�/� mice, a similar timespan to first papilloma was observed in TRunx3�/� andlittermate Runx3þ/þ mice (43 days from DMBA initiation).These results show that Runx3 ablation in T cells con-ferred partial resistance to DMBA/TPA–mediated skintumorigenesis.

Given that skin of germline-inactivated Runx3-null micelacks DETCs (18), which are known to protect againstDMBA/TPA–induced neoplasms (27), skins of TRunx3�/�

mice were examined. Using whole-mount epidermal sheetimmunostaining, we readily detected morphologicallymature DETCs in na€�ve epidermis of adult TRunx3�/� mice(Fig. 6B and Supplementary Fig. S3). This finding indicatedthat Lck-cre–mediated Runx3 deletion in early CD4�/CD8�

T cells did not affect the generation/survival of epidermalDETCs. This result further implied that TRunx3�/� micepartial resistance to DMBA/TPA–induced tumorigenicitywas mediated by other T-cell species.

Together, the above results show that specific Runx3ablation in either the DC or T-cell compartments confersresistance to DMBA/TPA–induced skin tumorigenesis ontothese mice. However, the extent of resistance in both geno-types, based on the relative finite tumor yield, rate, andlatency period, was not as robust as in the two germlineRunx3-deficient mouse strains, Runx3-null and Runx3fl/fl/Pgk-cre. These findings implied that a cooperative activitybetween Runx3-deficient DCs and T cells is required to fullyrecapitulate observed resistance in Runx3-null mice.

Combined Runx3 deletion in DCs and T cells fullyrecapitulated Runx3-null mice resistance to skintumorigenesis

Mice concurrently lacking Runx3 in DCs and T cellswere generated by cross-breeding Runx3fl/fl/CD11c-cre andRunx3fl/fl/Lck-cre mice (DCRunx3�/�/TRunx3�/�). These miceand their littermate controls were subjected to the DMBA/TPA protocol. Significantly, DCRunx3�/�/TRunx3�/� miceconferred an extended resistance, similar to that noted inthe two Runx3-deficient mouse strains (Runx3-null andRunx3fl/fl/Pgk-cre). Accordingly, by 20 weeks of promotion,mean tumormultiplicity was 8.3-fold lower in DCRunx3�/�/TRunx3�/�mice, comparedwith control group (peak of 2.5�1.9 vs. 20.1 � 15.5 papillomas/mouse, respectively; P ¼0.015; Fig. 6C). Furthermore, the delay period to firstpapilloma was doubled in DCRunx3�/�/TRunx3�/� mice,compared with controls (77 vs. 39 days from DMBAinitiation; Fig. 6C), and papilloma rate was significantlylower (0% vs. 40% and 25% vs. 93% of mice bearing �1papilloma at 10 and 15 weeks after DMBA initiation,respectively; Fig. 6C). Collectively, these findings indicatedthat Runx3 cell-autonomous function in DCs and thymo-cytes is required for DMBA/TPA–mediated tumorigenesis,and thus Runx3 loss in both these cell types causes the

Bauer et al.





pronounced tumor resistance. But what is the underlyingmechanism?

Cytokine composition in Runx3-null na€�ve and TPA-treated skin generates a robust antitumormicroenvironmentThe fact that communication between tumor cells and

their supporting stroma is largely regulated by inflamma-

tory cytokine/chemokine networks raised the possibilitythat an altered cytokine/chemokine milieu in Runx3-nullskin participates in conferring the observed tumor resis-tance. To address this possibility, we used qRT-PCR andproteomic dot-blot assays to comparatively measure theexpression levels of multiple cytokines and chemokines.Using manual qRT-PCR and the RT2 Profiler PCR Array,which simultaneously monitors the expression of 84

A

B

% Mice with papillomasMean papilloma/mouse


DCRunx3 −/−WT

M

DCRunx3 −/− WT

LC quantification

105104103102

102

103

104

105

0

0

0

20

40

60

80

100

0 10 20 30

600

400

Ave

rage

num

ber

of L

Cs

FS

C-H

CD

11c

102

103

104

105

0

CD

11c

FSC-A FSC-AMHC II

200

0

400 10 20 30 400

10

20

30

40 WT DCRunx3−/−DC

Runx3−/−

Runx3lox/l

ox contro

l

×400 ×400

20 μm 20 μm

WT DCRunx3−/−

*

10.0

0 50K 100K 150K 200K 250K0

50K

100K

150K

200K

250K

FS

C-H

0

50K

100K

150K

200K

250K

13.1

0 50K 100K 150K 200K 250K

0.277

1051041031020

MHC II

0.867

1,000 bp

508

268

500 bp

100 bp

bp

Figure 5. DCRunx3�/� mice resistance to DMBA/TPA–induced tumorigenesis. A, DCRunx3�/� mice characterization. Left, PCR showing specific deletion of thefloxed Runx3 alleles (268-bp amplicon) in sorted dDCs from a DCRunx3�/� mouse (PCR details in Supplementary Fig. S1). Right, reduced yield (multiplicity)and extended time to first papilloma in DCRunx3�/� (n ¼ 8), compared with WT (n ¼ 21) mice (P < 0.01), and comparable papilloma rates in DCRunx3�/�

andcontrolmice. B, detection ofmorphologicallymature LCs inDCRunx3�/� epidermis. Left,MHCII-FITC immunostained epidermal sheets. Right, quantitativeanalysis of epidermal LCs (n ¼ 4 mice/genotype, �5 fields/mouse at �400 magnification). �, P < 0.01; scale bars, 20 mm. Bottom, flow cytometryanalysis of LCs (CD11cþ/MHCIIþ) isolated from DCRunx3�/� (left) or control littermate (right) epidermal cell suspensions. Results represent one of fourDCRunx3�/� or littermate control mice with similar findings. Values in A and B are mean � SD.






cytokines/chemokines (Supplementary Tables S1 and S2),we analyzedWTor Runx3-null skin before and afterDMBA/TPA treatment. A 15-fold increase in Tslp (thymic stromallymphopoietin) expression was noted in na€�ve Runx3-nullskin (Fig. 7A). Tslp is a master cytokine that confers adominant and lasting CD4þ/CD8þ T cell–mediated pro-tection against skin tumorigenesis (26, 38). In addition, wealso observed intense upregulation of Cd70 and Cxcl5,indicating the existence of immunologically active leuko-cytes (39) in na€�ve Runx3-null skin. On the other hand,expression of Spp1, encoding the protumorigenic cytokineosteopontin (4), was 6-fold lower in na€�ve Runx3-null,

compared withWT skin (Fig. 7A).We then used a ProteomeProfiler Array that concurrently detects the levels of 40cytokines/chemokines (Supplementary Table S3) to assayprotein extracts fromDMBA/TPA-treatedWT or Runx3-nullskins (Fig. 7B). This analysis revealed a�2.5-fold increase inseveral chemokines, including Ccl17, Ccl1, and Ccl12, aswell as cytokines such as colony-stimulating factor (Csf) 2and 3, tumor necrosis factor a (Tnfa), and interleukin-6(IL6) in Runx3-null skin. These proteomic results werefurther validated by qRT-PCR (Fig. 7B). Notably, a strikingdecrease in IL17a was also evident in DMBA/TPA-treatedRunx3-null skin (Fig. 7C). IL17a is a signature Th17

A

B C

CD8+ W

T

CD8+ /Lc

k-cre

Runx1Runx3

Emerin

25

3750

75

kDa


Mean papilloma/mouse % Mice with papillomas

WT TRunx3−/−

0

20

40

60

80

100

0 5 10 15 20 0

5

10

15

20

0 5 10 15 20

TRunx3-/-×100

WT×100

×630

×630

0

10

20

30

40

0 5 10 15 20


Mean papilloma/mouse % Mice with papillomas

WT DCRunx3 −/−/TRunx3 −/−

0

20

40

60

80

100

0 5 10 15 20

Figure 6. Runx3 combined deletion in DCs and thymocytes fully recapitulates Runx3-null tumor resistance. A, Runx3 ablation in CD8þ T cells. Western blotanalysis (left) of sorted splenic CD8þ T cells from WT and TRunx3�/� mice. Note the Runx1 expression in cells from both genotypes. Right, decreasedtumor multiplicity and delayed onset in TRunx3�/� (n ¼ 12), compared with WT littermate mice (n ¼ 28). B, detection of morphologically sound DETCs innaïve epidermis of TRunx3�/� and WT littermate mice. Anti-CD3e-FITC immunostained epidermal sheets (scale bars, 100 and 20 mm in �100 and �630,respectively). See also Supplementary Fig. S3. C, DCRunx3�/�/TRunx3�/� mice exhibit resistance to DMBA/TPA tumorigenesis, similar to that noted inRunx3-null mice. Reduced yield and delayed onset of papillomas in DCRunx3�/�/TRunx3�/�mice (n¼ 9), compared with controls (n¼ 21). Values in A and C aremean � SD.

Bauer et al.





cytokine, the deletion of which has been correlated withmarked resistance to the DMBA/TPA protocol (40). Similarto na€�ve skin, Spp1 levels in DMBA/TPA–treated Runx3-nullskin were lower, compared with WT (Fig. 7B). Collectively,these expression experiments show that leukocytic Runx3loss generates a unique assembly of cytokines that is asso-ciated with a tumor-refractory microenvironment in theskin.

DiscussionWe have used mouse genetic models to demonstrate

the pivotal role of Runx3-expressing skin leukocytes inshaping the tumor microenvironment, thereby promot-ing the DMBA-initiated keratinocyte to a mature tumor.Specifically, we show that Runx3 cell-autonomous func-tion in the innate (DC) and adaptive (T-cell) immunecompartments hubs the induction of a protumorigeniccytokine microenvironment in the DMBA/TPA model ofskin tumorigenesis. Accordingly, subjecting mice lackingRunx3 in DCs and T cells to DMBA/TPA generates an

inflammatory tumor nonpermissive stromal environmentthat strongly resists papillomagenesis. Our analysis alsorevealed approximately 2-fold decrease in the proportionof "classical" CD11bþ dDCs and a concomitant approx-imately 5-fold increase in CD11bþ/CD103þ dDCs inRunx3-null mice na€�ve dermis. Although an increase inCD11bþ/CD103þ DCs was noted in inflamed murine gutlamina propria (30), as well as in steady-state mesentericlymph nodes (28) and lungs (41), their in vivo functionsremain largely obscure. Given that Runx3 was correlatedwith both positive and negative regulation of Cd103,during CD8þ T-cell differentiation (18, 42), it is temptingto speculate that its loss in "classical" CD11bþ dDCs leadsto Cd103 de-repression and the subsequent emergence ofCD11bþ/CD103þDCs, observed in Runx3-null mice. Thefinding that Runx3 skin expression was restricted toimmunocytes and was not detected in na€�ve or DMBA/TPA–treated keratinocytes excludes the possibility of acell-autonomous oncogenic Runx3 function in themurine DMBA/TPA tumor target cell itself. This conclu-sion is further supported by the Runx3fl/fl/K14-cre mice

BADifferentially expressed cytokines

in Runx3-null skin (qRT-PCR)Differentially expressed cytokines

in Runx3-null skin

IL17a relativequantification (densitometry)

IL17a relativeexpression (qRT-PCR)

DMBA/TPA-treated skinNaïve skin

−7

Ccl12Dot-blot densitometryqRT-PCR

Runx3-null WT

Ccl1

Ccl17

Tnf-α

Csf3

Csf2

IL-6

Ref signal

Tslp Cd70 Cxcl5

Fol

d ch

ange

(co

mpa

red

with

WT

)

Fol

d ch

ange

(co

mpa

red

with

WT

)

Spp1

−4

−1

2

5

8

11

14

17

DMBA/TPA-treated skin

−8

−6

−4

−2

Ccl

12

Ccl

17

Csf

3

Csf

2

IL-6

Spp

1

Tnf

-α

Ccl

10

2

4

6

8

10 59.2

C

Runx3-null WT

IL17a

Ref point

0.E+00

WT

Runx3-null

2.E+05

4.E+05

Raw

inte

grea

ted

dens

ity (

pixe

ls)

6.E+05

8.E+05

Runx3-nullWT

% o

f WT

0

20

40

60

80

100

Figure 7. Runx3-nullmice skin cytokine analysis. A, qRT-PCRanalysis of Runx3-null mice naïve skin RNAusing theRT2Profiler PCRArray and/ormanual qRT-PCR. Results show four highly significant differentially expressed cytokines of 84 analyzed genes (Supplementary Tables S1 and S2). Primers used inTslp analysis are depicted in Supplementary Materials and Methods. B, semiquantitative proteomic dot-blot panel (left), densitometry and RT2 Profiler qRT-PCR (right) analyses of differentially expressed cytokines in DMBA/TPA–treated skins of Runx3-null and WT littermate mice (Supplementary Table S3).Ccl17 and IL6 signals were undetected in Runx3-null mice. C, semiquantitative proteomic dot-blot panel (left), densitometry (middle), and qRT-PCR (right)analyses of IL17a expression in DMBA/TPA–treated skin of Runx3-null and WT littermate mice. Analyzed RNA or protein in A–C were purified fromskins of 3 mice (each genotype) and pooled. Dot-blot reference (Ref) signals in B and C are membrane positive controls. Data in A–C represent three and fiveindependent experiments with similar findings, respectively.






findings, demonstrating that epidermal keratinocyteswere not involved in Runx3-null mice tumor resistance.

Othermousemodels displaying resistance toDMBA/TPAhave been described, including mice deficient in Tnfa,cycloxygenase-2, and IL12 (43–45). Although the magni-tude of tumor resistance in these models was probablyaffected by genetic background and promotion regimendifferences, none has—as far as we know—resulted inresistance as dramatic as that reported here for Runx3-deficient mice. Moreover, although DMBA/TPA–inducedpapillomas persisted >100 days after TPA cessation, withapproximately 7% progressing to malignant carcinoma inWT mice, the low overall incidence of these benign papil-lomas in Runx3-null mice resulted in null carcinomas inthis genotype. Therefore, our study did not address thepossibility that lack of leukocytic Runx3 could also inhibitpapilloma-carcinoma transition.

In contrast to Runx3-null mice resistance to DMBA/TPAtumorigenicity, these mice were slightly more susceptiblethan control mice to complete 30-MCA–induced carcino-genesis. Several possible explanations were considered inpreviously observed differential susceptibility to DMBA/TPA vs. 30-MCA treatments (33, 34): (i) the obvious differ-ences in tumor characteristics (proliferating keratinocytes/DMBA vs. malignant fibroblasts/30-MCA); (ii) the tumorlocation (epidermal vs. subcutaneous, respectively); and(iii) the stromal microenvironment [persistent inflamma-tion (46) vs. inflammation-independent "foreign-body"granuloma formation, respectively (33, 34)]. Hence, theobservation that Runx3-null mice are refractory to DMBA/TPA–induced tumorigenesis, but are susceptible to inflam-mation-independent 30-MCA carcinogenesis, underlinesRunx3 involvement in inflammation-dependent skintumorigenesis.

LC-deficient mice are almost completely resistant toDMBA/TPA tumorigenesis (47). When LCs are absent,metabolic activation of chemical carcinogens (e.g., DMBA)that induce tumor initiation was impaired (37). Moreover,although DETC-depleted mice manifest higher susceptibil-ity to DMBA/TPA tumorigenesis by themselves (27), com-bined depletion of DETCs and LCs retained the robust LC-deficient tumor-resistant phenotype (47). Significant differ-ences exist between our observation that Runx3 loss confersmarked resistance to DMBA/TPA tumorigenesis and thetumor-resistant phenotype of LC-deficient mice. TheTRunx3�/� mice, which have epidermal LCs, are approxi-mately three times more resistant than WT mice to theDMBA/TPA assay. More convincingly, Runx3-null/TG.ACmice are refractory to DMBA/TPA papillomagenesis, a factthat provides strong evidence that Runx3-null resistance isexerted at the post-initiation phase. Thus, although wecannot rule out the possibility that LC decrease or absencecontributes to the DMBA/TPA resistance, by yet anothernoninitiation mechanism, we conclude that the uniqueinflammatory milieu drives Runx3-null mice robust tumorresistance.

Studies of cytokine function in tumorigenesis are com-plicated by their pleiotropic effects and redundancy and

by the ways in which the overall cytokine environmentmodulates individual cytokine impact (48). Despite thesedifficulties, loss- and gain-of-function experiments haveyielded important insights into the complex cytokines–tumor associations. Remarkably, each of the highly dys-regulated cytokines in na€�ve and DMBA/TPA–treatedRunx3-null skin—Tslp, IL17a, and osteopontin—hasbeen implicated in decreased susceptibility to DMBA/TPA-induced tumorigenicity (4, 38, 40). Interestingly,ChIP-seq analysis in CD8þ T and natural killer (NK) cellsrevealed Runx3 binding to both IL17a and Spp1 genes(49), suggesting that Runx3 might be involved in regu-lating these genes. In summary, loss of Runx3 in DCs andT cells affects tumor and/or inflammatory cell cytokineproduction, thereby generating a unique antitumormicroenvironment that resists skin tumorigenesis.

Tumor biology can no longer be understood withoutconsidering the tumor microenvironment contributions totumorigenicity (3). Various observations, from human epi-demiology to gene-modifiedmice, have highlighted inflam-mation as a predisposing cause of cancer and it is nowconsidered an enabling characteristic of cancer (3). Currentestimates imply that a quarter of cancers are associated withchronic inflammation (of infective or sterile origin; ref. 5).Moreover, inflammatory cells and mediators are presentwithin the microenvironment of most tumors, includingthose that are not epidemiologically related to inflamma-tion (e.g., breast cancer; ref. 50). Our findings indicate thatRunx3-null mice provide a model for dissecting the criticalevents affecting clonal progression of initiated cells to amature neoplasm. Under the two-stage DMBA/TPA regi-men, loss of leukocytic Runx3 induces a nonpermissivemicroenvironment that potently opposes the growth ofinitiated cells. Whether the DMBA/TPA protocol faithfullymimics human skin tumorigenesis is uncertain. Yet, wheth-er targeting Runx3 in leukocytes blocks the expansion ofinitiated epithelial cells in other malignant settings is animportant issue to address in the future, implying a signif-icant preventive impact on cancer occurrence in the humanclinical arena.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: O. Bauer, Y. GronerDevelopment of methodology: O. Bauer, Y. GronerAcquisitionofdata (provided animals, acquired andmanagedpatients,provided facilities, etc.): O. Bauer, S. Hantisteanu, J. LotemAnalysis and interpretation of data (e.g., statistical analysis, biosta-tistics, computational analysis): O. Bauer, J. Lotem, Y. GronerWriting, review, and/or revision of the manuscript: O. Bauer, J. Lotem,Y. GronerAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): O. BauerStudy supervision: Y. Groner

AcknowledgmentsThe authors thank Avraham Keter, Zamir Kokhavi, Rafi Saka, Arie Gum-

berof, and Ofir Higfa for help in animal husbandry; Ori Brenner forhistopathologic analyses and helpful advice; Steffen Jung and Uri Gat for

Bauer et al.





helpful discussions; and Edna Schechtman and Yisrael Parmet for assistancein statistical analysis.

Grant SupportThis study was supported by a grant from Israel Science Foundation (ISF)

individual grants to Y. Groner

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received March 21, 2014; revised June 10, 2014; accepted June 12, 2014;published OnlineFirst June 24, 2014.

References1. Kemp CJ. Multistep skin cancer in mice as a model to study the

evolution of cancer cells. Semin Cancer Biol 2005;15:460–73.2. Abel EL, Angel JM, Kiguchi K, DiGiovanni J. Multi-stage chemical

carcinogenesis in mouse skin: fundamentals and applications. NatProtoc 2009;4:1350–62.

3. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation.Cell 2011;144:646–74.

4. Hsieh YH, Juliana MM, Hicks PH, Feng G, Elmets C, Liaw L, et al.Papilloma development is delayed in osteopontin-null mice: implicat-ing an antiapoptosis role for osteopontin. Cancer Res 2006;66:7119–27.

5. Mantovani A, Allavena P, Sica A, Balkwill F. Cancer-related inflamma-tion. Nature 2008;454:436–44.

6. Shankaran V, Ikeda H, Bruce AT, White JM, Swanson PE, Old LJ, et al.IFNgamma and lymphocytes prevent primary tumour developmentand shape tumour immunogenicity. Nature 2001;410:1107–11.

7. Hanahan D, Coussens LM. Accessories to the crime: functions of cellsrecruited to the tumor microenvironment. Cancer Cell 2012;21:309–22.

8. Dinarello CA. The paradox of pro-inflammatory cytokines in cancer.Cancer Metastasis Rev 2006;25:307–13.

9. Balkwill FR, Mantovani A. Cancer-related inflammation: commonthemes and therapeutic opportunities. Semin Cancer Biol 2012;22:33–40.

10. Dranoff G. Cytokines in cancer pathogenesis and cancer therapy. NatRev Cancer 2004;4:11–22.

11. Lin WW, Karin M. A cytokine-mediated link between innate immunity,inflammation, and cancer. J Clin Invest 2007;117:1175–83.

12. Vaughan S, Coward JI, Bast RC Jr, Berchuck A, Berek JS, Brenton JD,et al. Rethinking ovarian cancer: recommendations for improvingoutcomes. Nat Rev Cancer 2011;11:719–25.

13. Levanon D, Brenner O, Negreanu V, Bettoun D, Woolf E, Eilam R, et al.Spatial and temporal expression pattern of Runx3 (Aml2) and Runx1(Aml1) indicates non-redundant functions during mouse embryogen-esis. Mech Dev 2001;109:413–7.

14. Raveh E, Cohen S, Levanon D, Groner Y, Gat U. Runx3 is involved inhair shape determination. Dev Dyn 2005;233:1478–87.

15. Taniuchi I, Osato M, Egawa T, Sunshine MJ, Bae SC, Komori T, et al.Differential requirements for Runx proteins in CD4 repression andepigenetic silencing during T lymphocyte development. Cell2002;111:621–33.

16. Woolf E, XiaoC, FainaruO, LotemJ, RosenD,Negreanu V, et al. Runx3and Runx1 are required for CD8 T cell development during thymopoi-esis. Proc Natl Acad Sci U S A 2003;100:7731–6.

17. Djuretic IM, Levanon D, Negreanu V, Groner Y, Rao A, Ansel KM.Transcription factors T-bet and Runx3 cooperate to activate Ifngand silence Il4 in T helper type 1 cells. Nat Immunol 2007;8:145–53.

18. Woolf E, Brenner O, Goldenberg D, Levanon D, Groner Y. Runx3regulates dendritic epidermal T cell development. Dev Biol 2007;303:703–14.

19. Fainaru O, Woolf E, Lotem J, Yarmus M, Brenner O, Goldenberg D,et al. Runx3 regulates mouse TGF-beta–mediated dendritic cell func-tion and its absence results in airway inflammation. EMBO J 2004;23:969–79.

20. Fainaru O, Shseyov D, Hantisteanu S, Groner Y. Accelerated chemo-kine receptor 7-mediated dendritic cell migration in Runx3 knockoutmice and the spontaneous development of asthma-like disease. ProcNatl Acad Sci U S A 2005;102:10598–603.

21. Dicken J, Mildner A, Leshkowitz D, Touw IP, Hantisteanu S, Jung S,et al. Transcriptional reprogramming of CD11b(þ)Esam(hi) dendriticcell identity and function by loss of Runx3. PLoS ONE 2013;8:e77490.

22. Brenner O, Levanon D, Negreanu V, Golubkov O, Fainaru O, Woolf E,et al. Loss of Runx3 function in leukocytes is associated with spon-taneouslydevelopedcolitis andgastricmucosal hyperplasia. ProcNatlAcad Sci U S A 2004;101:16016–21.

23. LevanonD,BernsteinY,NegreanuV, BoneKR,Pozner A, EilamR, et al.Absence of Runx3 expression in normal gastrointestinal epitheliumcalls into question its tumour suppressor function. EMBO Mol Med2011;3:593–604.

24. Levanon D, Bettoun D, Harris-Cerruti C, Woolf E, Negreanu V, EilamR,et al. The Runx3 transcription factor regulates development andsurvival of TrkC dorsal root ganglia neurons. EMBO J 2002;21:3454–63.

25. Levanon D, Groner Y. Runx3-deficient mouse strains circa 2008:resemblance and dissimilarity. Blood Cells Mol Dis 2009;43:1–5.

26. Di Piazza M, Nowell CS, Koch U, Durham AD, Radtke F. Loss ofcutaneous TSLP-dependent immune responses skews the balance ofinflammation from tumor protective to tumor promoting. Cancer Cell2012;22:479–93.

27. GirardiM,OppenheimDE,SteeleCR, Lewis JM,GlusacE, Filler R, et al.Regulation of cutaneous malignancy by gammadelta T cells. Science2001;294:605–9.

28. Heath WR, Carbone FR. Dendritic cell subsets in primary and sec-ondary T cell responses at body surfaces. Nat Immunol 2009;10:1237–44.

29. Bursch LS, Wang L, Igyarto B, Kissenpfennig A, Malissen B, KaplanDH, et al. Identification of a novel population of Langerinþ dendriticcells. J Exp Med 2007;204:3147–56.

30. Varol C, Vallon-Eberhard A, Elinav E, Aychek T, Shapira Y, Luche H,et al. Intestinal lamina propria dendritic cell subsets have differentorigin and functions. Immunity 2009;31:502–12.

31. Tanooka H, Tanaka K, Arimoto H. Dose response and growth rates ofsubcutaneous tumors inducedwith 3-methylcholanthrene inmice andtiming of tumor origin. Cancer Res 1982;42:4740–3.

32. Qin Z, Kim HJ, Hemme J, Blankenstein T. Inhibition of methylcholan-threne-induced carcinogenesis by an interferon gamma receptor–dependent foreign body reaction. J Exp Med 2002;195:1479–90.

33. Blankenstein T, Qin Z. Chemical carcinogens as foreign bodies andsome pitfalls regarding cancer immune surveillance. Adv Cancer Res2003;90:179–207.

34. Swann JB, Vesely MD, Silva A, Sharkey J, Akira S, Schreiber RD, et al.Demonstration of inflammation-induced cancer and cancer immunoe-diting during primary tumorigenesis. Proc Natl Acad Sci U S A2008;105:652–6.

35. Koebel CM, Vermi W, Swann JB, Zerafa N, Rodig SJ, Old LJ, et al.Adaptive immunity maintains occult cancer in an equilibrium state.Nature 2007;450:903–7.

36. Leder A, Kuo A, Cardiff RD, Sinn E, Leder P. v-Ha-ras transgeneabrogates the initiation step in mouse skin tumorigenesis: effects ofphorbol esters and retinoic acid. Proc Natl Acad Sci U S A 1990;87:9178–82.

37. Modi BG, Neustadter J, Binda E, Lewis J, Filler RB, Roberts SJ, et al.Langerhans cells facilitate epithelial DNA damage and squamous cellcarcinoma. Science 2012;335:104–8.

38. Demehri S, Turkoz A, Manivasagam S, Yockey LJ, Turkoz M, KopanR. Elevated epidermal thymic stromal lymphopoietin levels






establish an antitumor environment in the skin. Cancer Cell 2012;22:494–505.

39. Coquet JM, Middendorp S, van der Horst G, Kind J, Veraar EA, Xiao Y,et al. The CD27 and CD70 costimulatory pathway inhibits effectorfunction of T helper 17 cells and attenuates associated autoimmunity.Immunity 2013;38:53–65.

40. Wang L, Yi T, Zhang W, Pardoll DM, Yu H. IL-17 enhances tumordevelopment in carcinogen-induced skin cancer. Cancer Res 2010;70:10112–20.

41. EdelsonBT,KCW, JuangR,KohyamaM,Benoit LA, KlekotkaPA, et al.Peripheral CD103þ dendritic cells form a unified subset developmen-tally related to CD8alphaþ conventional dendritic cells. J Exp Med2010;207:823–36.

42. Yarmus M, Woolf E, Bernstein Y, Fainaru O, Negreanu V, Levanon D,et al. Groucho/transducin-like Enhancer-of-split (TLE)-dependent and-independent transcriptional regulation by Runx3. Proc Natl Acad SciU S A 2006;103:7384–9.

43. Moore RJ, Owens DM, Stamp G, Arnott C, Burke F, East N, et al. Micedeficient in tumor necrosis factor-alpha are resistant to skin carcino-genesis. Nat Med 1999;5:828–31.

44. Tiano HF, Loftin CD, Akunda J, Lee CA, Spalding J, Sessoms A, et al.Deficiency of either cyclooxygenase (COX)-1 or COX-2 alters epider-

mal differentiation and reducesmouse skin tumorigenesis. CancerRes2002;62:3395–401.

45. Sharma SD, Meeran SM, Katiyar N, Tisdale GB, Yusuf N, Xu H, et al.IL-12 deficiency suppresses 12-O-tetradecanoylphorbol-13-acetate-induced skin tumor development in 7,12-dimethylbenz(a)anthracene-initiated mouse skin through inhibition of inflammation. Carcinogen-esis 2009;30:1970–7.

46. Alford JG, Stanley PL, Todderud G, Tramposch KM. Temporal infil-tration of leukocyte subsets into mouse skin inflamed with phorbolester. Agents Actions 1992;37:260–7.

47. Strid J, Roberts SJ, Filler RB, Lewis JM, Kwong BY, Schpero W, et al.Acute upregulation of an NKG2D ligand promotes rapid reorganizationof a local immune compartment with pleiotropic effects on carcino-genesis. Nat Immunol 2008;9:146–54.

48. Paul WE. Pleiotropy and redundancy: T cell–derived lymphokines inthe immune response. Cell 1989;57:521–4.

49. Lotem J, LevanonD, Negreanu V, Leshkowitz D, Friedlander G, GronerY. Runx3-mediated transcriptional program in cytotoxic lymphocytes.PLoS ONE 2013;8:e80467.

50. Colotta F, AllavenaP,SicaA,GarlandaC,Mantovani A.Cancer-relatedinflammation, the seventh hallmark of cancer: links to genetic insta-bility. Carcinogenesis 2009;30:1073–81.


Bauer et al.




2014;7:913-926. Published OnlineFirst June 24, 2014.Cancer Prev Res Omri Bauer, Shay Hantisteanu, Joseph Lotem, et al. Expression of the Transcription Factor Runx3Carcinogen-Induced Skin Tumor Development Requires Leukocytic

Updated version

10.1158/1940-6207.CAPR-14-0098-Tdoi:

Access the most recent version of this article at:

Material

Supplementary

DC1

http://cancerpreventionresearch.aacrjournals.org/content/suppl/2014/06/28/1940-6207.CAPR-14-0098-T.Access the most recent supplemental material at:

Cited articles

http://cancerpreventionresearch.aacrjournals.org/content/7/9/913.full#ref-list-1

This article cites 50 articles, 18 of which you can access for free at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerpreventionresearch.aacrjournals.org/content/7/9/913To request permission to re-use all or part of this article, use this link



http://cancerpreventionresearch.aacrjournals.org/lookup/doi/10.1158/1940-6207.CAPR-14-0098-T

http://cancerpreventionresearch.aacrjournals.org/content/suppl/2014/06/28/1940-6207.CAPR-14-0098-T.DC1

http://cancerpreventionresearch.aacrjournals.org/content/suppl/2014/06/28/1940-6207.CAPR-14-0098-T.DC1

http://cancerpreventionresearch.aacrjournals.org/content/7/9/913.full#ref-list-1

http://cancerpreventionresearch.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerpreventionresearch.aacrjournals.org/content/7/9/913


carcinogen-induced skin tumor development requires...

Documents