Priority Report

Epithelial-to-Mesenchymal Transition Contributesto Immunosuppression in Breast CarcinomasAnushka Dongre1,2, Mohammad Rashidian1, Ferenc Reinhardt1,2, Aaron Bagnato1,Zuzana Keckesova1,2, Hidde L. Ploegh1,3, and Robert A.Weinberg1,2,3

Abstract

The epithelial-to-mesenchymal transition (EMT) is a cellbiological program that confers mesenchymal traits on carci-noma cells and drives their metastatic dissemination. It isunclear, however, whether the activation of EMT in carcinomacells can change their susceptibility to immune attack. Wedemonstrate here that mammary tumor cells arising frommore epithelial carcinoma cell lines expressed high levels ofMHC-I, low levels of PD-L1, and contained within their stromaCD8þ T cells and M1 (antitumor) macrophages. In contrast,tumors arising from more mesenchymal carcinoma cell linesexhibiting EMT markers expressed low levels of MHC-I, high

levels of PD-L1, and contained within their stroma regulatoryT cells, M2 (protumor) macrophages, and exhausted CD8þ Tcells. Moreover, the more mesenchymal carcinoma cells withina tumor retained the ability to protect their more epithelialcounterparts from immune attack. Finally, epithelial tumorswere more susceptible to elimination by immunotherapythan corresponding mesenchymal tumors. Our resultsidentify immune cells and immunomodulatory markers thatcan be potentially targeted to enhance the susceptibility ofimmunosuppressive tumors to various therapeutic regimens.Cancer Res; 77(15); 3982–9. �2017 AACR.

IntroductionThe microenvironment of breast carcinomas is enriched with

immunosuppressive cells that interact with carcinoma cells via avariety of heterotypic signaling mechanisms (1, 2). However, thefactors that contribute to immunosuppression in the tumormicroenvironment are poorly defined. The epithelial-to-mesen-chymal transition (EMT) is a cell biological program that drivesthe metastatic dissemination of carcinomas including breastcancers. After activation of an EMT, carcinoma cells begin tosecrete a distinct cohort of cytokines, such as TGFb (3). TGFb isrequired for the formation of immunosuppressive T regulatorycells (Treg) and reduces the cytotoxicity of natural killer (NK) cells(4, 5). It remains unclear, however, whether immunosuppressionoccurs as a direct consequence of the EMT program or developsthrough the mediation of additional, still-uncharacterized signal-ing channels. Moreover, carcinoma cells that have undergone anEMT are more refractory to various chemotherapeutic regimens(6). However, whether carcinoma cells residing in the epithelialversusmesenchymal statemount distinct responses to checkpointblockade therapy is unknown. For these reasons, we undertook tounderstand the dynamics of themicroenvironmental interactions

of breast cancers cells that remain in a largely epithelial state orhave acquired mesenchymal characteristics following activationof an EMT program.

Materials and MethodsMice

C57BL/6 mice were obtained from The Jackson laboratory. Acolony of NOD/SCID mice was generated and maintained in-house. All animals used for experiments were 6- to 8-week-oldfemale mice. Animals were maintained in compliance withthe guidelines and protocols approved by the Animal Care andUse Committees at the Massachusetts Institute of Technology(Cambridge, MA).

Cell lines and cell culturePyMT tumor cell lines from which epithelial (pB-2) and

mesenchymal (pB-3) versions were derived were kind giftsfrom the Harold L. Moses laboratory and were established andmaintained as described previously (7). Cells were cultured inDMEM/F12 medium containing 5% adult bovine serum with1� penicillin–streptomycin and 1� nonessential amino acidsfor the duration of this study. MCF7 and T47D cell lines wereobtained from ATCC and established and maintained aspreviously described (8) in DMEM supplemented with 10%FBS and 1� penicillin–streptomycin. Cells tested negativefor mycoplasma (2014) and were not authenticated sincefirst acquisition. All cell lines were obtained between 2009and 2012 and used within 2 to 3 years since first acquisition.Cells were stably transfected with GFP, RAS, or doxycycline-inducible EMT-inducing transcription factors (EMT-TF) asdescribed previously (8, 9). All cell lines containing doxycy-cline-inducible expression vectors were treated with 1.5 mg/mL(PyMT) or 1 mg/mL (MCF7RAS and T47DRAS) doxycyclinehyclate (Sigma Aldrich).

1Whitehead Institute for Biomedical Research, Cambridge, Massachusetts. 2MITLudwig Center for Molecular Oncology, Cambridge, Massachusetts. 3Depart-ment of Biology, Massachusetts Institute of Technology, Cambridge,Massachusetts.

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

Corresponding Author: Robert. A. Weinberg, Whitehead Institute for Biomed-ical Research, 9 Cambridge Center, WHTH-367, Cambridge, MA 02142. Phone:617-258-5159; Fax: 617-258-5213; E-mail: weinberg@wi.mit.edu

doi: 10.1158/0008-5472.CAN-16-3292

�2017 American Association for Cancer Research.

CancerResearch

Cancer Res; 77(15) August 1, 20173982

on August 31, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst April 20, 2017; DOI: 10.1158/0008-5472.CAN-16-3292

In vivo models and tumor dissociationFor orthotopic tumor transplantations, sorted cell popula-

tions, EpCAMHI (epithelial PyMT cell line-pB-2), EpCAMLO

(mesenchymal PyMT cell line-pB-3), EpCAMHISnail-YFPLO

(Snail-lo), and EpCAMLOSnail-YFPHI (Snail-hi), were resus-pended in 30 mL media containing 20% Matrigel. Cells(1 � 106) were implanted into the mammary fat pads ofC57BL/6 or NOD/SCID mice. Animals were sacrificed oncetumors reached 2 cm in size. Tumors were excised and dividedinto two parts. One part was digested and used for flowcytometry analysis, and the other part was fixed in formalinfor tissue sections. For tumor digestions, tumors were mincedwith a razor blade and digested in RPMI containing 2 mg/mLcollagenase and 100 U/mL hyaluronidase (Roche) in a rotatorat 37�C for 1 hour. Dissociated tumors were washed twice inPBS and filtered through a 70- and 40-mm cell strainer toobtain single-cell suspensions. For immunotherapy experi-ments, mice bearing tumors arising from various cell lineswere treated with anti-CTLA4, 200 mg, clone 9H10, every 3 daysfor 20 days.

Flow cytometryDissociated tumors were resuspended in wash buffer (PBS

containing 0.1%BSA) and stained for surfacemarkers usingCD45PECy7 (30F-11; Affymetrix), CD45 FITC (30F-11; Affymetrix),CD4 PE (RM4-5; Affymetrix), CD8a APC (53-6.7; Affymetrix),CD25 PercpCy5.5 (pc61.5; Affymetrix), CD44 PercpCy5.5 (IM7;Affymetrix), PD-1 FITC (J43; Affymetrix), CTLA4 PE (UC10-4B9;Affymetrix), CD107a PercpCy5.5 (1D4B; Affymetrix), CD11BPercpCy5.5 (M1/70; Affymetrix), F480 PECY7 (BM8; Affymetrix),LY6C E450 (HK1.4; Affymetrix), LY6G APC (1A8; Affymetrix),CD206 APC (MR6F3; Affymetrix), CD3 PercpCy5.5 (17A2; Affy-metrix), NK1.1 PE (PK136; Affymetrix), MHC-I (H-2Kb) PE (AF6-88.5.5.3; Affymetrix), CD274BV605 (10F.9G2; BioLegend),HLA-ABC APC (W6/32; Affymetrix), and PDL1 FITC (MIH1; BDBiosciences). CD8þ T cells were sorted from digested tumorsamples and cocultured with the respective PyMT cells (1 �106 cells/mL) for 5 hours in the presence of Monensin GolgiStop(BD Biosciences). Intracellular cytokine staining was performedusing the Intracellular Fixation and Permeabilization Buffer Set(Affymetrix). Intracellular staining for FOXP3 was performedusing the FOXP3/Transcription Factor Staining Buffer Set (Affy-metrix) using FOXP3 Alexa488 (FJK-16s; Affymetrix), NOS2 FITC(CXNFT; Affymetrix), IL12 PE (C15.6; BioLegend), and IFNgPECY7 (XMG1.2; Affymetrix). Flow cytometry data were acquiredon a BD LSRFortessa, and data were analyzed using the FlowJo(TreeStar) software.

Western blot analysisWhole-cell lysates were made in RIPA buffer (150 mmol/L

NaCl, 1% IgeCal-CA 360, 0.1% SDS, 50 mmol/L Tris, pH 8.0,0.5% sodium deoxycholate) and resolved on a gradient gel.Protein was transferred on a nitrocellulose membrane andblocked in 5% milk powder and 0.2% Tween-20 in PBS. Mem-branes were probed overnight with primary antibody, washed,and incubated with horseradish peroxidase–labeled secondaryantibody, and developed using ECL substrate (Thermo FisherScientific). Primary antibodies used were E-cadherin, vimentin,Zeb1, Snail, GAPDH, b-tubulin, b2-microglobulin (b2M; CellSignaling Technology), Fibronectin (BD Biosciences), and Twist(Abcam).

Immunofluorescence stainingTumors were fixed in 10% neutral-buffered formalin for 12 to

24 hours and transferred to 70%ethanol, followed by embeddingand sectioning. Tumor sectionswerewashed twice inHistoclear II,followed by onewash each in 100%, 95%, 75%ethanol, PBS, and1� wash buffer (Dako). Antigen retrieval was done in 1� TargetRetrieval Solution, pH 6.1 (Dako) in a microwave. Sectionswere blocked in PBS containing 0.3% Triton X-100 and 1%normal donkey serum (Jackson ImmunoResearch Laboratories)for 1 hour at room temperature. Sections were incubated withprimary antibody at 4�Covernight. Sections were washed twice in1�wash buffer, followed by incubation with secondary antibody(Biotium) for 2 hours. Sections were washed three times with 1�wash buffer and incubatedwithDAPI for 10minutes, followed by1 wash in PBS. Sections were mounted using ProLong GoldAntifade Reagent (Invitrogen). Tumor cell lines were fixed in2.5% neutral-buffered formalin on ice for 15 minutes, followedby three washes in PBSþ. Cells were fixed in Triton X-100 for 3minutes and blocked in PBS� containing 3% normal donkeyserum. Cells were incubated with primary antibody at 4�C over-night. Cells were washed three times with PBS�, followed byincubation with secondary antibody for 2 hours. Cells werewashed three times with PBS� and incubated with DAPI for 10minutes, followed by 1 wash in PBS. Cells were mounted usingProLong Gold Antifade Reagent. Primary antibodies used were E-cadherin (BDBiosciences), vimentin (Cell Signaling Technology),Zeb1 (Santa Cruz Biotechnology), CD3 (AbCam), CD8 (Abbio-tec), F480 (Thermo Fisher Scientific), and Arginase1 (Santa CruzBiotechnology). Images were acquired using the Zeiss confocalmicroscope and analyzed using the Zen software.

Statistical analysisAll data are represented asmean� SEM. Statistical analysis was

performed using the GraphPad Prism software. P values werecalculated using an unpaired two-tailed Student t test (�, P < 0.05;��, P < 0.01; ���, P < 0.001; ns, not significant).

ResultsTo determine whether the immune responses to epithelial and

mesenchymal carcinoma cells were fundamentally different, weused cell lines that were established from tumors arising in thetransgenic MMTV-PyMTmouse model of breast adenocarcinomadevelopment (7). Some of these cell lines were more epithelialand expressed high levels of E-cadherin and EpCAM, whereasothers were more mesenchymal and in contrast expressed vimen-tin, low levels of EpCAM, and significant levels of the Zeb1,Twistm and Snail EMT-TFs (Fig. 1A; Supplementary Fig. S1Aand S1B).

To determine whether tumors formed from the epithelial andmesenchymal cell lines were differentially susceptible to immuneattack, we implanted these cells orthotopically into syngeneicimmunocompetent hosts. Tumors arising from the more epithe-lial cell line were well-differentiated adenocarcinomas andexpressed E-cadherin, whereas tumors arising from the moremesenchymal cells were poorly differentiated, exhibited a sarco-matoid histomorphology, and expressed instead the mesenchy-mal marker vimentin (Fig. 1A).

Most strikingly, the stroma of primary tumors arising from themore mesenchymal cell line, relative to those spawned by themore epithelial carcinoma cells, contained significantly higher

Immunosuppression in Breast Carcinomas

www.aacrjournals.org Cancer Res; 77(15) August 1, 2017 3983

on August 31, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst April 20, 2017; DOI: 10.1158/0008-5472.CAN-16-3292

E-Cadherin

N-Cadherin

Twist

Vimen�n

Zeb1

Epi Mes

Snail

β-Tub

A

B

pB-2

pB-3

CD8+ T cells

010203040

% o

CD

45

***

0

50

100

150

CD107a Gzymb Perforin IFNγ

***

***

***

*

% o

CD

8+

01020304050

PD1+CTLA-4+

***%

o C

D8+

020406080 ***

% o

f CD

4

Tregs T-eff-

01020304050 ***

% o

f CD

4

Macrophages

0

50

100

150 ***

% o

f CD

45

020406080 ***

% o

f Mac

s

CD206

020406080

100

iNOS IL12

******

% o

f Mac

s

M1 cytokinesC

D

Epi

F480 Arg1 DAPIMes

CD3 F480 DAPI

Epi Mes

H&E Ecad PanCk DAPI Vim PanCk DAPI

Epi

Mes

Epi Mes

CD11

B

iNOS

38.6% 13.9%

CD11

B

IL12

Epi Mes

42.5% 2.04%

EpiMes

CD8 DAPIEpi Mes

Figure 1.

Immunosuppressive cells in tumors arising from mesenchymal (Mes) cell lines. A, Western blot analysis for EMT markers expressed by cell lines. Hematoxylinand eosin (H&E) and immunofluorescent staining (B and D) of tumor sections for the indicated markers (N ¼ 5). Scale bar, 100 mm. B and C, Digestedtumors were analyzed for the indicated immune populations by flow cytometry after gating on CD45þ cells. Epi, epithelial. Data represent three independentexperiments. � , P < 0.05; ��� , P < 0.001.

Dongre et al.

Cancer Res; 77(15) August 1, 2017 Cancer Research3984

on August 31, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst April 20, 2017; DOI: 10.1158/0008-5472.CAN-16-3292

numbers of CD25þ FOXP3þ Tregs and lower numbers of CD25þ

FOXP3� effector CD4þ T cells (Fig. 1B; Supplementary Fig. S1D).Conversely, tumors arising from the epithelial cell line containedhigher numbers of CD8þ cytotoxic T cells instead (Fig. 1B;Supplementary Fig. S1C). Furthermore, a higher percentage ofCD8þ T cells present in the stroma of epithelial tumors expressedCD107a, perforin, granzyme B, and IFNg compared with thosepresent in mesenchymal tumors, indicative of effector function(Fig. 1B). Finally, the fewCD8þ T cells present in the stroma of themore mesenchymal tumors coexpressed markers associated withfunctional exhaustion, PD-1 and CTLA4 (Fig. 1B).

Tumors arising from the more mesenchymal cell lines alsocontained a significantly higher number of macrophages thatexpressed the mannose receptor CD206 and stained positive forArginase1, indicating a protumor M2 phenotype, relative to themacrophages present in tumors arising from the more epithe-lial cells, which by contrast did not express either of these twomarkers but instead stained positively for iNOS and IL12,markers of M1, antitumor macrophages (Fig. 1C; Supplemen-tary Fig. S1E and S1F).

Immunofluorescence analysis of tumor sections revealed thatCD3þ andCD8þT cells had infiltrated the tumors arising from theepithelial cell line, while themacrophages remained largely at theperiphery. This contrasted with tumors arising from the mesen-chymal cell line, which were infiltrated bymacrophages (Fig. 1D).We did not, however, observe any significant differences in thenumbers of monocytic or granulocytic MDSCs infiltrating thesetwo classes of tumors (Supplementary Fig. S1G and S1H). Theseresults indicated that, in contrast to the stroma assembled by themore epithelial tumors, the stroma of tumors arising from themore mesenchymal cells was largely immunosuppressive.

Using MMTV-PyMT reporter mice that bear knock-in IRES-YFPconstructs reporting the activity of the endogenous Snail and Slugpromoters, we recently demonstrated that, when expressed atphysiologic levels, Snail is associated with a CSC-like phenotype,whereas Slug choreographs the normal mammary epithelial stemcell state (7). To determine whether the Snail-expressing carcino-ma cells would also be immunosuppressive, we sorted cells thatdiffered in their expression of the Snail-YFP reporter and thusendogenous Snail protein, from tumors arising in MMTV-PyMT-Snail-YFPmice using YFP expression in conjunctionwith EpCAM,as described previously (Supplementary Fig. S2A; ref. 7). Snail-hi-YFP cells were mesenchymal and did not express any epithelialmarkers, but expressed insteadN-cadherin, vimentin, fibronectin,and the Zeb1 EMT-TF. In contrast, Snail-lo-YFP cells were lessmesenchymal and expressed high levels of both EpCAM and E-cadherin and low levels of Zeb-1 (Fig. 2A; Supplementary Fig.S2B). This approach allowed us to compare the immune responsewith related cell lines, which were established from the sametumor and differed in their levels of Snail expression.

Tumors arising from orthotopic implantation of the Snail-locells were heterogeneous and contained some sectors of E-cad-herin–expressing cells. In contrast, Snail-hi tumors were poorlydifferentiated and lacked E-cadherin expression (Fig. 2A).Although we did not observe any difference in the numbers andfunction of CD8þ T cells (Fig. 2B; Supplementary Fig. S2C), Snail-hi tumors exhibited a significantly higher percentage of infiltratingTregs (Fig. 2B; Supplementary Fig. S2D) and an increase inCD206-expressingM2macrophages (Fig. 2C; Supplementary Fig.S2E and S2F), relative to Snail-lo tumors. Thesemacrophages alsoexpressed Arginase1 in contrast to Snail-lo tumors, which lacked

Arginase1-expressing macrophages (Fig. 2C). Furthermore, thesemacrophages had infiltrated Snail-hi tumors, which was in sharpcontrast to Snail-lo tumors,where themacrophageswere confinedlargely at the tumor periphery (Fig. 2D). Conversely, Snail-lotumors were infiltrated by CD3þ and CD8þ T cells (Fig. 2D). Inaddition, adenocarcinomas arising from another epithelial cellline established from Slug reporter mice (Supplementary Fig. S3Aand S3B; ref. 7) were also infiltrated by significantly highernumbers of CD8þ T cells and significantly fewer numbers of Tregsand M2 macrophages relative to Snail-hi cells (SupplementaryFig. S3C–S3H).Once again, these observations demonstrated thatmore mesenchymal Snail-hi cells were capable of inducing atumor microenvironment rich in immunosuppressive Tregs andM2 macrophages in contrast to the less mesenchymal Snail-locells.

We also observed that the more mesenchymal cell lines, whichclearly recruited immunosuppressive cells to the tumor-associat-ed stroma, showed a striking reduction in the levels of surfaceMHC-I compared with themore epithelial cell lines (Supplemen-tary Fig. S4A–S4D). Furthermore, tumors arising from the moremesenchymal carcinoma cell lines in both immunocompetentand immunodeficient mice exhibited significantly fewer MHC-Iþ

cells compared with those arising from the epithelial cell lines. Incontrast, we found that tumors arising from the more mesenchy-mal carcinoma cell lines contained instead a higher percentage ofPD-L1þ cells (Supplementary Fig. S4E–S4H). Of note, activationof an EMT in epithelial cell lines by forced expression of variousdoxycycline-inducible EMT-TFs (Supplementary Fig. S5A–S5C)resulted in a decrease in protein levels of MHC-I expression,which was also accompanied by decreased expression of b2M(Supplementary Fig. S5D, S5F, and S5H) and increased expres-sion of PD-L1 (Supplementary Fig. S5E and S5G). Thus, theimmunoevasive and immunosuppressive maneuvers assembledby the carcinoma cells resulted directly from the cell-autonomousactions of the EMT program.

Carcinomas of the type studied here are comprised of hetero-geneous mixtures of both epithelial and mesenchymal cells.However, it was unclear whether the minority subpopulationsof immunosuppressive mesenchymal carcinoma cells are able toprotect the entire tumor, including its epithelial cells, fromimmune attack or whether their immunosuppressive effects werelimited to their immediate vicinity. To address this question, weinjected mixtures of unlabeled epithelial cells with two differentGFP-labeled, mesenchymal cell lines described earlier.

We observed that a decrease in the proportion of epithelial cells(10:1, 4:1, and 1:1 ratio of E:M cells) was accompanied by aconcomitant decrease in the number of infiltrating CD8þ T cells(Fig. 3A; Supplementary Fig. S6A and S6C). CD8þ T cells presentin these various comixed tumors did not show any reduction inthe expression of markers associated with effector function com-pared with those present in tumors arising exclusively fromepithelial cells. Nonetheless, the CD8þ T cells that were presentin all three types of comixed tumors coexpressed PD1 and CTLA4,indicative of partial exhaustion, relative to CD8þ T cells present intumors arising only from epithelial cells (Fig. 3A; SupplementaryFig. S6A). In sharp contrast, a decrease in the proportion ofepithelial cells in all comixed tumors was accompanied by anincrease in the number of suppressive Tregs (CD25þ FOXP3þ; Fig.3A; Supplementary Fig. S6A and S6C) and a decrease in thenumber of CD4þ effector T cells (CD25þ FOXP3� and CD44þ

FOXP3�; Fig. 3B; Supplementary Fig. S6B).

Immunosuppression in Breast Carcinomas

www.aacrjournals.org Cancer Res; 77(15) August 1, 2017 3985

on August 31, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst April 20, 2017; DOI: 10.1158/0008-5472.CAN-16-3292

Similarly, macrophages present in all types of comixed tumorsexpressed protumor M2 markers CD206 and Arginase1 and lowlevels of the iNOS antitumor,M1marker, relative tomacrophagespresent in tumors arising from the epithelial cells alone (Fig. 3B;Supplementary Fig. S6B and S6C). Hence, even when the mes-enchymal cells were present as small minority subpopulations(�10%), they were nonetheless able to influence the numbers ofsuppressive Tregs andM2macrophages in the tumors as a whole.

Both immunofluorescent and hematoxylin and eosin stainingof tumors arising from the mixtures of cells injected in differentratios revealed that the epithelial and mesenchymal cells becamesegregated to distinct sectors of the tumor and thus did notphysically intermingle (Fig. 3C, top; Supplementary Fig. S6D).In addition, mesenchymal sectors within the comixed tumors

continued to express features associated with the EMT, such asvimentin and nuclear Zeb1, in contrast to their epithelial counter-parts, which expressed E-cadherin instead (Fig. 3C, middle andbottom; Supplementary Fig. S6D).

This clear segregation of the two types of tumor cells to distinctsectors of tumors allowed us to examine the topological relation-ships between recruited immune cells and readily identifiableepithelial and mesenchymal sectors of the tumors. As we found,CD8þ T cells were found near the epithelial sectors of mixedtumors. Conversely, numerous Arginase1þ F480þ M2 macro-phages were interspersed only among the GFPþ, mesenchymalcells of the tumor, while only a few scattered macrophages werefound at the periphery of epithelial sectors within the same tumor(Fig. 3D, top; Supplementary Fig. S6E). Most importantly, the

05

101520

ns

% o

f CD

45+

CD8+ T cells

0

10

20

30

PD1+CTLA-4+

ns

% o

f CD

8

010203040 ***

% o

f CD

4

Tregs

020406080

ns

% o

f Cel

ls

Macrophages

020406080 ***

% o

Mac

s

CD206

SnailLO

SnailHI

A

B

C

D

SnailLO

E-Cadherin

Zeb1

Snail

Fibronec�n

SnailHI

Vimen�n

Twist

β-Tubulin

020406080

Gzymb IFNGCD107a

% o

f CD

8

nsns

ns

CD3 F480GFPDAPI

Interior

Periphery

Snail HISnail LO

Snail LO Snail HI

F480 Arg1 DAPI

Snail HISnail LO

Snail HI

Snail LO

H&E Ecad PanCk DAPI Vim PanCk DAPI

CD8DAPI

Snail HISnail LO

Figure 2.

Immunosuppressive cells in tumorsarising from Snailþ cells. A, Westernblot analysis for EMT markers.Hematoxylin and eosin (H&E) andimmunofluorescent staining (C and D)of tumor sections for the indicatedmarkers (N ¼ 5). Scale bar, 100 mm. Band C, Digested tumors were analyzedfor the indicated immune populationsby flow cytometry after gating onCD45þ cells. ns, not significant. Datarepresent three independentexperiments. ���, P < 0.001.

Dongre et al.

Cancer Res; 77(15) August 1, 2017 Cancer Research3986

on August 31, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst April 20, 2017; DOI: 10.1158/0008-5472.CAN-16-3292

010203040

% o

f CD

45 ** * *****

CD8+ T cells

010203040

PD1+CTLA4+

% o

f CD

8 ****

******

Tregs

010203040

% o

f CD

4

*****

**

*

T-eff

020406080

100

CD25 CD44

% o

f CD

4

***

*****

*****

0

20

40

60

% o

f Cel

ls

Macrophages

020406080

% o

f Mac

s *****

CD206

020406080

100

% o

f Mac

s

***

iNOS

A

B

C

D

Epithelial

(SnailMes HI)

: MesEpi (9:1): MesEpi (4:1): MesEpi (1:1)

020406080

100

CD107a GzymbCD44

*

***% o

f CD

8 **

***

ns

ns

MixedEpithelial Mesenchymal

H&E

E-cadVimGFPDAPI

PanCkZeb1GFP

DAPI

ns

E-cadCD8GFP DAPI

Epi-sector

Mes-sector

Mixed tumorsMixed tumors

E-cadF480GFP

DAPI

F480Arg1GFP

DAPI

Epi-sectorComplete sec�on Mes-sector

Figure 3.

Ability of mesenchymal cells to protect epithelial cells from immune attack. A and B, Digested tumors were analyzed for the indicated immune populationsby flow cytometry after gating on CD45þ cells. ns, not significant. C, Hematoxylin and eosin (H&E) and immunofluorescent staining (C and D) of tumorsections for the indicated markers (N ¼ 5). Scale bar, 100 mm. Data represent three independent experiments. � , P < 0.05; �� , P < 0.01; ��� , P < 0.001.

Immunosuppression in Breast Carcinomas

www.aacrjournals.org Cancer Res; 77(15) August 1, 2017 3987

on August 31, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst April 20, 2017; DOI: 10.1158/0008-5472.CAN-16-3292

latter macrophages did not stain positive for Arginase1 (Fig. 3D,middle), indicating that the juxtaposition of macrophages withmesenchymal but not epithelial cells within a tumor was likely tobe crucial for regulating their immunosuppressive function. Thesedata suggest that the proportion ofmesenchymal cells dictates thenumbers and types of suppressive immune cells recruited to thetumor.

Having observed substantial differences in T cells infiltratingthe primary tumors arising from implanted epithelial versusmesenchymal carcinoma cells, we asked whether these tumorswould differ in their susceptibility to checkpoint blockade therapy(10, 11). Although epithelial tumor–bearing mice showed acomplete response to anti-CTLA4 immunotherapy,mesenchymaltumor–bearing mice were refractory to treatment (Fig. 4A and B).Interestingly, tumors arising from1:1 or 9:1mixtures of epithelialandmesenchymal cells, respectively, also did not respond to anti-CTLA4 treatment (Fig. 4C). In addition, anti-CTLA4 treatment didnot alter the numbers or EMT features of epithelial and mesen-chymal cells within mixed tumors (Supplementary Fig. S7A andS7B). Hence, these data suggest that the immunosuppressivetumor microenvironment of mesenchymal tumors made themresistant to elimination via checkpoint blockade therapy (12, 13).Moreover, the presence of even a small minority of mesenchymalcarcinoma cells resulted in the loss of responsiveness of tumors tocheckpoint blockade therapy.

DiscussionSeveral studies have demonstrated an association between the

EMT and immunosuppression in other cancer types followingforced expression of EMT-TFs (12). However, high-level expres-sion of EMT-TFs elicits functional changes in carcinoma cells thatdo not recapitulate the changes occurring when these EMT-TFs are

expressed under cell physiologic control (7). This explainswhyweundertook the currently described work to determine whether anassociation between the EMT and immunosuppression doesindeed occur when EMT-TFs are expressed under physiologiccontrol using cell lines that were established directly from tumorsarising in MMTV-PyMT mice.

Mesenchymal cell lines and the tumors arising from themexpressed markedly lower levels of cell surface MHC-I comparedwith epithelial cell lines and their corresponding tumors, raisingthe question of whether such MHC-I–low expression levelsrender tumor cells vulnerable to NK-mediated killing (14).Indeed, we did observe a higher percentage of NK cells infiltratingMHC-I–low tumors arising from mesenchymal cell lines relativeto MHC-I–high tumors arising from epithelial cell lines (Supple-mentary Figs. S1E and S2E). Importantly, however, these tumorswere not eliminated despite the presence of such NK cells. Weattribute this outcome to the actions of other immunosuppressivesubsets, such as Tregs, which were present abundantly in mesen-chymal tumors and are known to dampen NK function (4, 15).

The presence of cytotoxic lymphocytes in MHC-I–expressingepithelial tumors is likely to make them more susceptible toelimination via checkpoint blockade therapy. Indeed, patientswith tumor-infiltrating lymphocytes generally have a relativelygood prognosis and are thought to generate favorable responsesto checkpoint blockade treatment regimens (10, 16). In contrast,the strongly immunosuppressive stroma of mesenchymal tumorsmay render them refractory to such therapy (13).

Our studies demonstrate that epithelial and mesenchymalcarcinomas differ in their susceptibility to immune attack.Although we have yet to determine the precise molecularmechanism(s) by which EMT-induced immunosuppressiontakes place, we suggest that this may likely occur by severalmechanisms. Among these are dysregulated expression of

A B

C

Epithelial tumors

00 10 20 30

Time (days)

Time (days)

50

Tum

or si

ze (m

m2 ) 100

0

50

Tum

or si

ze (m

m2 )

100

00 5 10 15 20 25

Time (days)0 5 10 15 20 25

50

Tum

or si

ze (m

m2 ) 100

ControlAnti-CTLA4

tumorsMesenchymal

Mixed tumors

Mesenchymal

Epithelial1:1 9:1

E:M

E:M

Figure 4.

Response of epithelial andmesenchymal tumors toimmunotherapy. A–C, Average tumorgrowth of mice that did or did notreceive anti-CTLA4 therapy. Datarepresent three independentexperiments.

Dongre et al.

Cancer Res; 77(15) August 1, 2017 Cancer Research3988

on August 31, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst April 20, 2017; DOI: 10.1158/0008-5472.CAN-16-3292

immunomodulatory markers associated with immune evasionand alterations in cytokine and chemokine secretion patterns,which may act as the first step that triggers a cascade of eventsthat ultimately culminate in immunosuppression. All of theseimmunoevasive maneuvers utilized by carcinoma cells couldact in concerted fashions to (i) recruit specific immunosup-pressive cells to the tumor microenvironment and/or (ii)reprogram preexisting cells to adopt an immunosuppressivephenotype.

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

Authors' ContributionsConception and design: A. DongreDevelopment of methodology: A. Dongre, Z. KeckesovaAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): A. Dongre, M. Rashidian, F. Reinhardt, A. BagnatoAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): A. Dongre, M. Rashidian, H.L. PloeghWriting, review, and/or revision of the manuscript: A. Dongre, M. Rashidian,H.L. Ploegh, R.A. WeinbergStudy supervision: H.L. Ploegh, R.A. Weinberg

AcknowledgmentsWe thank Richard A. Goldsby and Barbara A. Osborne for critical reading of

themanuscript.We thankBrian Bierie andXin Ye for cells and all othermembersof the Weinberg Laboratory for helpful discussion. We thank the Flow Cyto-metry Core facility at theWhitehead Institute for assistance with cell sorting andflow cytometry analysis and the Keck Imaging facility at theWhitehead Institutefor assistance withmicroscopy.We thank Roderick Bronson for histopathologicanalysis of tumor sections and the Histology Facility at the Koch Institute/MITfor tissue sectioning.

Grant SupportR.A. Weinberg received support from the NIH (P01 CA080111), Breast

Cancer Research Foundation, Samuel Waxman Cancer Research Founda-tion, and Ludwig Center for Molecular Oncology. H.L. Ploegh receivedsupport from NIH (R01-AI087879-06, R01-GM100518-04) and the Lust-garten Foundation. A. Dongre was supported by the Ludwig Fund forCancer Research. M. Rashidian was supported by the Cancer ResearchInstitute.

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received December 8, 2016; revised February 28, 2017; accepted April 13,2017; published OnlineFirst April 20, 2017.

References1. Hanahan D, Coussens LM. Accessories to the crime: functions of cells

recruited to the tumor microenvironment. Cancer Cell 2012;21:309–22.2. Kerkar SP, Restifo NP. Cellular constituents of immune escape within the

tumor microenvironment. Cancer Res 2012;72:3125–30.3. Scheel C, Eaton EN, Li SH, Chaffer CL, Reinhardt F, Kah KJ, et al. Paracrine

and autocrine signals induce and maintain mesenchymal and stem cellstates in the breast. Cell 2011;145:926–40.

4. Crane CA, Han SJ, Barry JJ, Ahn BJ, Lanier LL, Parsa AT. TGF-beta down-regulates the activating receptor NKG2D on NK cells and CD8þ T cells inglioma patients. Neuro Oncol 2010;12:7–13.

5. Viel S, Marcais A, Guimaraes FS, Loftus R, Rabilloud J, Grau M, et al. TGF-beta inhibits the activation and functions of NK cells by repressing themTOR pathway. Sci Signal 2016;9:ra19.

6. Singh A, Settleman J. EMT, cancer stem cells and drug resistance:an emerging axis of evil in the war on cancer. Oncogene 2010;29:4741–51.

7. Ye X, TamWL, Shibue T, Kaygusuz Y, Reinhardt F, Eaton EN, et al. DistinctEMT programs control normal mammary stem cells and tumour-initiatingcells. Nature 2015;525:256–60.

8. Guo W, Keckesova Z, Donaher JL, Shibue T, Tischler V, Reinhardt F, et al.Slug and Sox9 cooperatively determine the mammary stem cell state. Cell2012;148:1015–28.

9. De Cock JM, Shibue T, Dongre A, Keckesova Z, Reinhardt F, Weinberg RA.Inflammation triggers Zeb1-dependent escape from tumor latency. CancerRes 2016;76:6778–84.

10. Pardoll DM. The blockade of immune checkpoints in cancer immuno-therapy. Nat Rev Cancer 2012;12:252–64.

11. Sharma P, Allison JP. The future of immune checkpoint therapy. Science2015;348:56–61.

12. Kudo-Saito C, Shirako H, Takeuchi T, Kawakami Y. Cancer metastasis isaccelerated through immunosuppression during snail-induced EMT ofcancer cells. Cancer Cell 2009;15:195–206.

13. De Henau O, Rausch M, Winkler D, Campesato LF, Liu C, Cymerman DH,et al. Overcoming resistance to checkpoint blockade therapy by targetingPI3Kgamma in myeloid cells. Nature 2016;539:443–7.

14. Waldhauer I, Steinle A. NK cells and cancer immunosurveillance. Onco-gene 2008;27:5932–43.

15. Rook AH, Kehrl JH, Wakefield LM, Roberts AB, Sporn MB, Burlington DB,et al. Effects of transforming growth factor beta on the functions of naturalkiller cells: depressed cytolytic activity and blunting of interferon respon-siveness. J Immunol 1986;136:3916–20.

16. Callahan MK, Wolchok JD, Allison JP. Anti-CTLA-4 antibody therapy:immunemonitoring during clinical development of a novel immunother-apy. Semin Oncol 2010;37:473–84.

www.aacrjournals.org Cancer Res; 77(15) August 1, 2017 3989

Immunosuppression in Breast Carcinomas

on August 31, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst April 20, 2017; DOI: 10.1158/0008-5472.CAN-16-3292

2017;77:3982-3989. Published OnlineFirst April 20, 2017.Cancer Res   Anushka Dongre, Mohammad Rashidian, Ferenc Reinhardt, et al.   Immunosuppression in Breast CarcinomasEpithelial-to-Mesenchymal Transition Contributes to

  Updated version

  10.1158/0008-5472.CAN-16-3292doi:

Access the most recent version of this article at:

  Material

Supplementary

  http://cancerres.aacrjournals.org/content/suppl/2017/04/20/0008-5472.CAN-16-3292.DC1

Access the most recent supplemental material at:

   

   

  Cited articles

  http://cancerres.aacrjournals.org/content/77/15/3982.full#ref-list-1

This article cites 16 articles, 4 of which you can access for free at:

  Citing articles

  http://cancerres.aacrjournals.org/content/77/15/3982.full#related-urls

This article has been cited by 34 HighWire-hosted articles. Access the articles at:

   

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

  Subscriptions

Reprints and

  .pubs@aacr.org

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://cancerres.aacrjournals.org/content/77/15/3982To request permission to re-use all or part of this article, use this link

on August 31, 2021. © 2017 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst April 20, 2017; DOI: 10.1158/0008-5472.CAN-16-3292