Immunosuppressive and Prometastatic Functions of Myeloid ... · of Myeloid-Derived Suppressive Cells Rely upon Education from Tumor-Associated B Cells Monica Bodogai1, Kanako Moritoh1,

Microenvironment and Immunology

Immunosuppressive and Prometastatic Functionsof Myeloid-Derived Suppressive Cells Rely uponEducation from Tumor-Associated B CellsMonica Bodogai1, Kanako Moritoh1, Catalina Lee-Chang1, Christine M. Hollander2,Cheryl A. Sherman-Baust1, Robert P.Wersto3, Yoshihiko Araki4, Ichiro Miyoshi5, Li Yang2,Giorgio Trinchieri6, and Arya Biragyn1

Abstract

Myeloid-derived suppressive cells (MDSC) have been reportedto promote metastasis, but the loss of cancer-induced B cells/Bregulatory cells (tBreg) can block metastasis despite MDSC expan-sion in cancer. Here, using multiple murine tumor models andhuman MDSC, we show that MDSC populations that expand incancer have only partially primed regulatory function and limitedprometastatic activity unless they are fully educated by tBregs.Cancer-induced tBregs directly activate the regulatory function ofboth the monocyte and granulocyte subpopulations of MDSC,

relying, in part, on TgfbR1/TgfbR2 signaling.MDSC fully educatedin this manner exhibit an increased production of reactive oxygenspecies and NO and more efficiently suppress CD4þ and CD8þ

T cells, thereby promoting tumor growth and metastasis. Thus,loss of tBregs or TgfbR deficiency in MDSC is sufficient to disabletheir suppressive function and to block metastasis. Overall, ourdata indicate that cancer-induced B cells/B regulatory cells areimportant regulatorsof the immunosuppressive andprometastaticfunctions of MDSC. Cancer Res; 75(17); 3456–65. �2015 AACR.

IntroductionThe success of metastasis often depends on the ability of dis-

seminating cancer cells to escape immune attack by using the helpof regulatory immune cells, a heterogeneous group of specializedcells of granulocytic, myeloid, and lymphoid origins with seem-ingly redundant functions (1). Among these, myeloid-derivedsuppressive cells (MDSC) are thought to be key inhibitors ofantitumor effector cells and, as such, an independent prognosticfactor of patient survival (2). As a groupof immature cells poised todifferentiate into granulocytes, dendritic cells, and macrophages,MDSC are subdivided into PMN-MDSC andMo-MDSC cells (1, 3)on the basis of expression of Ly6GþLy6CInt/Low CD11bþ andLy6CHighLy6G� CD11bþ in mice (4, 5) and CD14�CD11bþ

CD15þCD33þ and CD14þCD11bþHLA-DRLow/� in humans(2, 6, 7), respectively. By producing granulocyte macrophage

colony-stimulating factor (GM-CSF), VEGF, TGFb, IL-6, IL10, IL13,and PGE4, cancer not only drastically expands MDSC but alsoevokes their regulatory function (for reviews, see refs. 1, 8, 9) byinducing their production of reactive nitrogen and oxygen species(NO, ROS, H2O2, and peroxinitrite) through the IL4–Stat6-depen-dent expression of arginase 1 (Arg-1; ref. 10) and Stat1- and Stat3-induced expression of nitric oxide synthase (iNOS) and NADPHoxidase (NOX2; refs. 11, 12). The expansionofMDSC is oftenusedas a criterion of increased tumor burden and metastasis (1, 13).However, using tumor models where MDSC were reported to beessential, we failed to detect the primary importance of MDSC incancermetastasis. The loss of regulatory T cells (Treg) or B cells wassufficient to almost completely block the metastasis of the highlyaggressive 4T1 cancer in BALB/c mice, a human model of triple-negative breast cancer (14), and retard the growth of B16 mela-noma in C57BL/6 mice (15–18). In the 4T1 model, cancer pro-duces 5-lipoxygenase metabolites to convert B cells into a newsubset of regulatoryB cells, termed tumor-evoked regulatoryB cells(tBreg; refs. 17, 19), that induce FoxP3þ Tregs to inactivate theantimetastatic natural killer (NK) and CD8þ T cells (15, 17, 19).

Here, using two different murine models and experimentingwith human ex vivo–generated MDSC, we report that canceronly expands MDSC with partially activated regulatory func-tion. As a result, the MDSC cannot support metastasis orpromote tumor growth. We show that cancer uses B cells toevoke their full regulatory and thereby prometastatic function.Our modeling studies using specific TgfbR1 inhibitor and micewith TgfbR2 deficiency in myeloid cells suggest that cancer-induced B cells/tBregs evoke the full regulatory activity inMDSC via using, at least in part, the TgfbR1/TgfbR2 signalingaxis. These results further underscore B cells/tBregs as key tumormessengers and initiators of the chain of suppressive eventsneeded for metastasis.

1Immune Regulation Section, Laboratory of Molecular Biology andImmunology,National InstituteonAging,Baltimore,Maryland. 2TumorMicroenvironment Section, Laboratory of Cancer Biology and Genet-ics, National Cancer Institute, Bethesda, Maryland. 3Flow CytometryUnit, National Institute on Aging, Baltimore, Maryland. 4JuntendoUniversity Graduate School of Medicine, Chiba, Japan. 5Center forExperimental Animal Science, Nagoya City University GraduateSchool of Medicine, Nagoya, Japan. 6Cancer Immunobiology Section,Laboratory of Experimental Immunology, National Cancer Institute,Frederick, Maryland.

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

Corresponding Author: Arya Biragyn, National Institute on Aging, 251 BayviewBoulevard, Suite 100, Baltimore, MD 21224. Phone: 410-558-8680; Fax: 410-558-8386; E-mail: [email protected]

doi: 10.1158/0008-5472.CAN-14-3077

�2015 American Association for Cancer Research.

CancerResearch

Cancer Res; 75(17) September 1, 20153456

on April 19, 2020. © 2015 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 16, 2015; DOI: 10.1158/0008-5472.CAN-14-3077

http://cancerres.aacrjournals.org/

Materials and MethodsReagents, cells, and mice

TGFbRI (ALK5) inhibitor (SB431542) was purchased fromTocris Bioscience and catalase (1,000 U/mL) from Sigma Aldrich.L-NMMA and nor-NOHA (0.5 mmol/L) were from CaymanChemical. Nitrate and NO were detected with the Griess ReagentKit and DAF-FM diacetate, respectively, and ROS was detectedwith 1 mmol/L DHE (dihydroethidium) or DCFDA (20,70-dichlor-odihydrofluorescein diacetate) and was from Molecular Probesand used as described elsewhere (5). a-TGFb neutralizing anti-body (clone 1D11.16.8), a-mouse Gr1 (clone RB6-8C5), mouseIgG, and rat IgG2b were purchased from BioXcell.

The following flow cytometric antibodies and their isotypecontrols (from Biolegend and eBioscience, except otherwise spec-ified) were used: CD11b APC or FITC (M1/70), Gr1 PE or FITC(RB6-8C5), Ly6G Alexa Fluor700 or PerCP Cy 5.5 (1A8), Ly6CPacific blue or FITC (HK1.4), IL4Ra PE (I015F8), F4/80 PerCPCy5.5 or APC (BM8), CD40 PE Cy7 (3/23), CD115 PE (AFS98),CD80 brilliant violet 421 (16-10A1), CD83 brilliant violet 650(Michel-19), GrB FITC (GB11), and IFNg PE-Cy7 (XMG1.2).TgfbR antibodies were from R&D (TgfbR1, clone FAB5871A andTgfbR2, clone FAB532F).

For intracellular staining of phosphorylated Stat proteins, cellswere fixed with 2% paraformaldehyde in PBS for 10 minutes at37�C and resuspended in prechilled 90% methanol (in water).The cells were stained with anti-mouse CD11b Fitc, Ly6G PE-Cy7,Ly6C Pacific blue (Biolegend), and rabbit anti-mouse pStat1 orpStat3-Alexa Fluor 647 (Tyr705, Cell Signaling) at 1:200 dilution.For lineage-negative MDSC sorting, phycoerythrin (PE)-conju-gated CD19 (clone 6D5), CD3 (clone 145-2C11), B220 (cloneRA3-6B2), and CD49b (DX5) antibodies were used. Humansorting was performed using CD11b PE (clones M1/70) or CD14Fitc (clone M5E2), CD19 PE (HIB19), CD3 PE, and CD56 PEantibodies (Biolegend and eBioscience).

4T1 adenocarcinoma cells and B16F10 melanoma cells werefrom ATCC. 4T1.2 cells, a subset of 4T1 cells, were a gift fromDr. Robin L. Anderson (Peter McCallum Cancer Center, Mel-bourne, Australia). Nonmetastatic 4T1.2-PE cells were describedelsewhere (15). Female (8- to 15-week-old) BALB/c, C57BL/6mice, and JHT mice in C57BL/6 background (JHT; B6.129P2-Igh-Jtm1Cgn/J) were from the Jackson Laboratory; mMTmice in BALB/cbackground were a gift from Professor Dr. Thomas Blankenstein(Max-Delbrück-Center for Molecular Medicine, Berlin, Ger-many); C57BL/6 mice with spontaneous ovarian cancer due tomurine oviduct–specific glycoprotein promoter-driven expres-sion of simian virus 40 large T-antigen (20) were housed atNational Institute on Aging (Baltimore, MD) and TGFbRIImyeKO

mice at NCI (Bethesda, MD; ref. 21).

In vitro manipulationstBreg generation and T-cell suppression assays were performed

as previously described (15). Unless specified, control B cells weretreated with 100 ng/mL of recombinant mouse BAFF/BlyS (R&D)in cRPMI. tBregs were characterized as CD81hiCD25þCD20low4-1BBLlow cells gated within CD19þ cells using mAbs [CD19-APC-eFluor780, CD81-APC, CD25Pacific blue, CD20-PE, and 4-1BBL-PerCP Cy5.5 (Biolegend and eBioscience)]. Flow cytometric datawere collected on a FACS Canto II (BD) and analyzedwith FlowJosoftware (Tree Star, Inc.). Cell sorting was performed on FACSAria III (BD) or MoFlo XDP (Beckman Coulter, Inc.). Mo-MDSCwere isolated as Lin�, CD11bþ, Ly6G�, Ly6Chi, and PMN-MDSC

were isolated as Lin�, CD11bþ, Ly6Gþ, Ly6C low-int. Magneticsort puritywasmore than96%andflow cytometric sort puritywasmore than 98%. All magnetic sorts performed in these experi-ments were double sorts to increase purity. For themajority of theexperiments, PMN- and Mo-MDSC were cell-sorted (purity >98%) or isolated using the Miltenyi Biotech Myeloid-derivedSuppressor Cell Isolation Kit. MDSC isolated from peripheralblood or spleen yielded similar results. For MDSC education,Gr1þ cells were isolated from tumor-bearing mice by magneticpositive selection (Miltenyi Biotech) and mixed at experimentratios (10:1) with B cells (B-cont or tBregs) for 5 hours. Then,Gr1 cells were resorted by magnetic positive selection prior toassays, such as testing for in vitro T-cell suppression or adoptivetransfer experiments. When MDSC education was done with Btumor cells, negatively isolated MDSC (using Miltenyi BiotechMDSC isolation kit) were cocultured overnight with positivelyisolated CD19þ B cells and isolated from spleen or lymph node(yielded similar results) of na€�ve or 4T1 tumor–bearing BALB/cmice. After coculture, B cells were depleted and MDSC used inT-cell suppression assays or adoptive transfer experiments. Forsuppression assay, MDSC were cocultured at 1:2 to 1:16 ratioswith splenic CD3þ T cells, which were isolated with the mouseT-cell enrichment columns (R&D Systems), labeled with the cellproliferation dye eFluor450 (eBioscience), and stimulated withanti-CD3/28 beads (Life Technologies) for 4 to 5 days.Dilution ofeFluor450 in CD4þ and CD8þ T cells (stained with anti-mouseCD4-PerCPCy5.5, clone GK1.5, and CD8-FITC, clone 53-6.7,Biolegend) is considered proliferation. TGFb, GrB, and IFNg weredetected by intracellular staining of splenic B cells from tumor-bearing mice after 4-hour stimulation with phorbol 12-myristate13-acetate (PMA; 50 ng/mL and 500 ng/mL ionomycin, TocrisBioscience) and 10mmol/Lmonensin (eBioscience). To test TgfbRsignaling, we used 20 mmol/L SB431542 in the coculture of B cellswith MDSC. Alternatively, wild-type (WT) B cells were culturedwith MDSC from tumor-bearing TgfbR2 KO mice.

Human peripheral blood cell isolation and suppressivemyeloid cell assays

Humanperipheral bloodwas collected by theHealth ApheresisUnit and the Clinical Core Laboratory, the National Institute onAging, under Human Subject Protocol # 2003054 and TissueProcurement Protocol # 2003-071. Peripheral blood mononu-clear cells (PBMC) were isolated using Ficoll-Paque (GE Health-care) density gradient separation, and B, T, and NK cells weredepleted with PE-coupled antibodies to CD19, CD3, CD56, andanti-PE microbeads following the manufacturer's instructions(Miltenyi Biotech). To educate myeloid cells, PBMCs (1 � 106

cells/mL, with or without B cells) were cultured with conditionedmedia from human breast cancer cell line (MDA-MB-231, 50%,volume) and 20 ng/mL GM-CSF with or without 20 mmol/LSB431542 for 4 days. Then, CD11bþ or CD14þ cells were isolatedby positive selection (Miltenyi Biotech) and used in T-cell sup-pression assay.Healthydonor PBMCs (depletedof T andNKcells)were cocultured with CD19þ cells from peripheral blood ofpatients with B-CLL for 48 hours. Then, CD19þ cells were deplet-ed and CD14þ cells were positively isolated bymagnetic selectionand used in T-cell suppression assays.

In vivo manipulationsMouse experiments were performed in a pathogen-free envi-

ronment at the National Institute on Aging Animal Facility in

Cross-Talk between MDSCs and Bregs

www.aacrjournals.org Cancer Res; 75(17) September 1, 2015 3457




accordancewith the procedures outlined in theGuide for the Careand Use of Laboratory Animals (NIH Publication No. 86-23,1985). 4T1.2 cells and B16-F10 cells (1 � 105) were subcutane-ously injected into BALB/c mice or congenic mMT andTgfbR2myeKO or JHT and C57BL/6 mice, respectively. At days 32and 20 after 4T1 and B16-F10 cell injections, respectively, lungmetastasis and tumor weight were evaluated as described else-where (15). MDSC were isolated frommice after 25 to 30 and 18to 20 days after 4T1 and B16 tumor challenge, respectively. Micewith spontaneous ovarian cancer were 12 to 15 weeks old. Foradoptive transfer experiments, mice were intravenously injectedwith congenic 1.5 � 106 to 10.0 � 106 MDSC or tBregs at 1 to 3days after tumor challenge. For Fig. 2B, the mice were tumorinoculated at day 0, andB-cell adoptive transferwas performedonday 18 (the late B-cell transfer was done to allow sufficient MDSCaccumulation). MDSC were sorted 3 days after B-cell transfer toavoid the conversion of na€�ve B cells due to the tumor pressure.

a-mouse Gr1 depletion antibody was used at 100 mg/mouseintraperitoneal injections every fourth day. For the B-cell deple-tion experiments, mice were intraperitoneally injected witha-CD20 antibody (250 mg/mouse) on day 7 after tumor inocu-lation. Peripheral blood MDSC were assessed on day 28 aftertumor inoculation.

Statistical analysisThe results are presented as the mean � SEM. To assess signif-

icance between two groups, we used the Mann–Whitney test,whereas the Holm–Sidak method with a of 5.0% was used tocorrect for multiple comparison tests of tumor growth and titra-tion curve analysis (Prism 6, GraphPad Software, Inc.).

ResultsB-cell deficiency impairs the regulatory activity of MDSC

To understand the prometastatic role of MDSC, we subcuta-neously implanted 4T1 cancer cells in the mammary gland of WTmice and B-cell–deficient mMT mice that do not generate tBregs(16, 17). Tumor growth in the mammary gland was retarded(P < 0.005), and lung metastasis was blocked in mMT mice ascomparedwithWTmice (P¼ 0.028; Fig. 1A).However, bothmicecomparably and strongly expanded MDSC (CD11bþGr1þ;Supplementary Fig. 1A), such as Mo-MDSC (Ly6G�Ly6Cþ) andPMN-MDSC (Ly6GþLy6CInt/Low) in the peripheral blood as wellas in the secondary lymphoid organs, lungs, and tumors of bothmice (Fig. 1B). Because these results question the importance ofMDSC inmetastasis, we tested their role by either treating tumor-bearing WT mice with RB6-8C5 Ab that depletes Gr1þ myeloidcells (includingMDSC) or, conversely, by the adoptive transfer ofMDSC from tumor-bearingWTmice (WT-MDSC) into mMTmice.While the depletion of Gr1þ cells reduced metastasis in WT mice(P ¼ 0.004; Supplementary Fig. 1B), metastasis was restored inmMTmice transferred with WT-MDSC at a similar extent with themice injected with ex vivo–generatedmetastasis-supporting tBregs(tBregs; Fig. 1C). The transfer of MDSC from tumor-bearing mMTmice (mMT-MDSC; Fig. 1C) failed to affect metastasis in mMTmice, confirming the importance of B cells/tBregs in the prometa-static function of MDSC.

Next, we tested whether the loss of B cells/tBregs can impair theregulatory function of MDSC. To do this, we cocultured purifiedWT- and mMT-MDSC isolated from individual mice challengedwith 4T1 cancer at the same time (n ¼ 9, from here on, unless

specified, theMDSCwere only isolated from tumor-bearingmice)with na€�ve mouse T cells stimulated with anti-CD3/CD28 Abs for5 days. At the cell-to-cell comparisons, while WT-MDSC stronglyinhibited proliferation (Fig. 1D) and production of granzyme B(GrB; Fig. 1E) and IFNg (Fig. 1F) inCD8þT cells,mMT-MDSCweresignificantly less efficient (P < 0.001; Fig. 1D–F). To furtherconfirm this result, we also tested suppressive activity of individ-ual subsets of MDSC by coculturing sort-purified PMN-MDSCand Mo-MDSC with T cells stimulated with anti-CD3/CD28 Abs.Despite overall stronger activity of PMN-MDSC over Mo-MDSC,both mMT-MDSC subsets inhibited less efficiently CD4þ andCD8þ T cells as compared with their respective WT subsets(P < 0.001; Fig. 1G and Supplementary Fig. S1C). Because thesuppression involves NO and ROS, their production could beimpaired in mMT-MDSC. In support, specific inhibitors NOHA,L-NMMA, and particularly catalase (scavenges H2O2, dismutationproduct of ROS), almost completely abolished the ability ofMDSC to suppress T-cell proliferation (Supplementary Fig. S1Dand S1E). The relative expression levels of NOX2 and Arg geneswere significantly reduced in mMT-MDSC as compared with WT-MDSC (Supplementary Fig. S2A and S2B). In concordance, nitrateproduction (theNOreadout)wasonly detected inWTMo-MDSC,but not PMN-MDSC or mMT-MDSC, cultured with T cells(P ¼ 0.028 as compared with mMT-MDSC; Fig. 1H). Bothsubsets of mMT-MDSC also produced significantly less ROS(detected by staining for O2

� with DHE; Supplementary Fig.S2C) and H2O2 (Fig. 1I and Supplementary Fig. 2D, detected byDCFDA; ref. 22) as compared with WT-MDSC. For example, themedian fluorescence index (MFI) of DCFDAþMo-MDSC wasdecreased from 14,400 � 150 to 6,459 � 247 (P < 0.001) in thelungs and from 17,565� 1,153 to 12,109� 345 (P < 0.001; Fig.1I) in the tumor of mMT mice compared with WT mice,respectively. For DCFDAþ PMN-MDSC, it was reduced from14,569 � 64 to 6,082 � 41 (P < 0.001) and 29,616 � 251 to21,847 � 74 (P < 0.001; Fig. 1I), respectively. Moreover, ROSproduction in WT-MDSC of tumor-bearing WT mice was fur-ther increased (P < 0.03; Supplementary Fig. 2C) if the micewere treated with a-CD20 Ab that enriches for tBregs viadepleting B cells (17). Thus, given that mMT mice do notgenerate tBregs (17), these results suggest that the regulatoryfunction of MDSC requires tBregs.

tBregs activate the regulatory function of cancer-primed MDSCTo test this possibility, we transferred B cells from na€�ve

(B-cont) or tumor-bearing WT mice (B-tumor, which containtBregs; ref. 17) into mMT mice (n ¼ 5 per group; Fig. 2A). After3 days, peripheral blood MDSC were isolated and tested forsuppression of T-cell proliferation. mMT-MDSC from B-cont–transferred mice suppressed proliferation of CD8þ T cells slightlystronger than the ones from mock-treated mice (Fig. 2B), pre-sumably due to some in vivo conversion of B-cont cells into tBregs(16, 17). However, mMT-MDSC from mice transferred with B-tumor inhibited T-cell proliferation significantly stronger (P <0.001; Fig. 2B) and produced higher levels of ROS (P ¼ 0.029 ascompared with B-cont–transferred mouse Mo-MDSC and PMN-MDSC, respectively; Fig. 2C). Thus, these experiments reproducedat least 3 independent times indicate that the full regulatoryfunction of MDSC in mice requires tBregs. In concordance, thetransfer of ex vivo–generated tBregs also significantly increased thesuppressive activity of MDSC in mMT mice (P < 0.008; Fig. 2D),besides restoring lung metastasis (Fig. 1C and also refs. 16–18).

Bodogai et al.

Cancer Res; 75(17) September 1, 2015 Cancer Research3458




tBregs educate MDSC promoting cancer escape and metastasisTounderstand how tBregs activateMDSC,weperformed in vitro

experiments by coculturing freshly purified mMT-MDSC withna€�ve mouse B cells (B-cont), or with tBregs, or B cells fromtumor-bearing mice (B-tumor). After 5-hour incubation, the

MDSC were depleted of B cells and mixed with T cells stimulatedwith anti-CD3/CD28Ab.Unlike B-cont, the pretreatment ofmMT-MDSCwith B-tumor (Fig. 3A and B) or tBregs (Fig. 3C and D andSupplementary Fig. S3A) strongly (P ¼ 0.029) and in a dose-dependent manner inhibited T-cell proliferation and ROS. The

Figure 1.B-cell deficiency impairs the regulatory and prometastatic function of MDSC. Tumor size, lung metastasis (inset, A), and frequency of MDSC (B) wereevaluated in mMT and WT mice with subcutaneously injected 4T1.2 cancer cells (1 � 105 cells). To assess metastasis-inducing ability of tBregs and MDSC(from mice with 4 week old tumor), they (10 � 106 cells) were adoptively transferred in mMT mice (n ¼ 4–5 mice per group) 24 hours after tumorimplantation (C). MDSC (D–F) or MDSC subsets (G and H) were sorted from mMT and WT mice 4 weeks after tumor inoculation and tested for their abilityto suppress T cells stimulated with anti-CD3/CD28 Abs at indicated (x-axis, D–F) or at 1:8 (G) effector-to-target ratio. To test expression of GrB andIFNg , MDSC were intracellularly stained (E and F); and nitrate production (H) was assessed in conditioned media of cells used in G. The DCFDAstaining of MDSC subsets in indicated tissues from mMT and WT mice with 4 week tumor is shown in I. The y-axis shows tumor size (A) or number ofmetastatic foci (inset, A), percentage of total cells per indicated tissue (B), mM of nitrate (H) � SEM of 3 to 5 mice per group, experiments reproduced atleast three times. � , P < 0.05; �� , P < 0.01; �� , P < 0.001; �� , P < 0.0001. NS, not significant.






treatment also blocked production of key factors required forsuccessful antitumor activity of CD8þ T cells (17, 23), such as GrBand IFNg (Supplementary Fig. S3A–S3C). The B-tumor and tBreg-stimulated mMT-MDSC also expressed higher levels of NOX2,iNOS, and Arg1 genes (Supplementary Fig. S4A–S4G) andproduced more ROS (P < 0.03; Fig. 3B and D) than B-cont–treated MDSC. They also upregulated expression of IL10 andTGFb (Supplementary Fig. S4E and S4F) and other factorsassociated with regulatory MDSC, such as IL4Ra, CD80, CD83,CD40, and phosphorylated Stat1 and Stat3 (Fig. 3E and Sup-plementary Fig. S4H). The mMT-MDSC also upregulated surfaceexpression of TgfbR1 (P < 0.03) and TgfbR2 (P < 0.002) uponstimulation with tBregs (Supplementary Fig. S4I). When wetransferred these mMT-MDSC (pretreated with B-tumor/tBregs)into mMT mice, the mice succumbed to significant lung metas-tasis (P < 0.03; Fig. 3F). In contrast, B-cont–pretreated mMT-MDSC failed to restore metastasis in mMT mice. Thus, theseresults, which were independently confirmed in multipleexperiments, indicate that tBregs directly evoke the regulatoryand thereby prometastatic function of MDSC.

Murine and human cancer B cells can also educate MDSCSimilar "education" of MDSC appears to also occur in other

mouse backgrounds and tumor models, as the growth of a highlyaggressive B16 melanoma is also retarded in B-cell–deficient JHTmice compared with congenic C57BL/6mice (17–19, 24). To testthis possibility, we adoptively transferred JHT mice bearing B16melanoma with MDSC purified from WT C57BL/6 mice withorthotopic B16 melanoma or with spontaneous ovarian cancer(WT B16-MDSC and OC-MDSC, respectively; ref. 20). UnlikeMDSC from JHT mice with B16 melanoma (JHT B16-MDSC), thetransfer of WT B16-MDSC and OC-MDSC reversed the retardedtumor growth in JHTmice yielding significantly larger tumors (P <0.05; Fig. 4A). Moreover, WT B16-MDSC and OC-MDSC alsoproduced higher levels of ROS (P < 0.03; Fig. 4B) and nitric oxide(P < 0.03; Fig. 4C) and strongly inhibited CD4þ and CD8þ T-cellproliferation (P ¼ 0.03; Fig. 4D).

To see whether similar education can also occur in humanMDSC, healthy donor PBMC depleted of T andNK cells (but withor without CD19þ B-cell depletion) were treated with condi-tioned medium (CM) from MDA-MB-231 breast cancer cell line,

NaïveA

C

D

B

WT mice

Tumor-bearingWT mice

B cells

B cells(tBregs)

Unloaded control

0.00

Mo-

MD

SC

Mo-MDSC PMN-MDSC Mo-MDSC PMN-MDSCDHE

0

10

20

30

40

50

60

70

0 0 0

200 200

400400

600600

800

800

1,000

1,000

1,200

1,200

1,400

1,400

1,600

1,800

10

20

30

40

50

60

70

DH

E (

%)

Moc

k

B-c

ont

B-t

umor

Moc

k

B-c

ont

B-t

umor

Moc

k

B-c

ont

B-t

umor

Moc

k

B-c

ont

B-t

umor

DH

E (

MF

I)

PM

N-M

DS

C

0.1 59.2 59.2 64.5

53.8 54.2 62.8

Mock B-cont B-tumor

0Mock B-cont

CD4

0

Unt

reat

ed

αCD

3/C

D28

μMT MDSC

Moc

k

B-c

ont

tBre

g

10

20

30

40

50

60

Pro

lifer

atio

n (%

)

70

80

90

100CD8

B-tumor Mock B-cont B-tumor0

10 10

20

CD

4 pr

olife

ratio

n (%

)

CD

8 pr

olife

ratio

n (%

)

20

30

30

40

40

50NS

Adoptive

Tumor-bearing

Sort-isolate MDSC

and mix with T cells

μMT mice

transfer

Figure 2.The prometastatic activity of MDSC requires tBregs. Adoptive transfer of B cells in mMT mice modulates the T-cell suppressive potential of MDSC (B, D)and ROS production (C). Schema of the experiment (A) depicted in B–D. MDSC were isolated from mMT mice and cultured with T cells in Fig. 1G at 1:8 effector-to-target ratio (B, D) or stained with DHE (C). A representative figure (top) and summary results (as percentage and MFI, bottom, C) are from four mice pergroup experiment reproduced twice. The y-axis is the percentage of proliferated CD4þ (B) and CD8þ T cells (B, D) � SEM of 4 to 5 mice per groupexperiments reproduced at least three times. � , P < 0.05; �� , P < 0.01; �� , P < 0.001. NS, not significant.

Bodogai et al.





the procedure that induces the generation of tBreg-like cells. Then,myeloid cells were sort-purified and tested for suppression ofhuman CD8þ T cells stimulated with anti-CD3/CD28 Abs. Whilemyeloid cells strongly inhibited T-cell proliferation if pretreatedwith CM in the presence of B cells (P¼ 0.03, MDA), mock-treatedcells failed to do so (mock; Fig. 4E; and Supplementary Fig. S5Aand S5B). Importantly, if we treated PBMCsdepleted of B cells, thecells failed to inhibit T-cell proliferation (MDA-CD19; Fig. 4E andSupplementary Fig. S5A and S5B). To further confirm the linkbetweenB cells and the inductionof suppressive activity ofMDSC,we cocultured healthy donor peripheral blood myeloid cells(depleted of B, T, and NK cells) with B cells from patients withB-CLL, which we previously reported either did (PS #154) or didnot (PS #174) contain tBreg-like cells (17), for 2 days in thepresence of GM-CSF. Then, myeloid cells were reisolated andtested in T-cell suppression assays. Unlike mock or PS #174cocultured cells, myeloid cells treated with PS #154 (whichcontained tBreg-like cells) strongly inhibited proliferation of Tcells (P < 0.03; Fig. 4F and Supplementary Fig. S5C). Takentogether, these results indicate that both murine and human

cancer–exposed B cells/Bregs can evoke the regulatory functionof MDSC.

tBregs educate MDSC by triggering TGFb signalingGiven that murine and human tBregs convert metastasis-pro-

moting FoxP3þTregs by overexpressing TGFb (16, 18) and the factthat the TgfbR2 deficiency in myeloid cells abrogates 4T1 cancermetastasis (21), we tested whether TGFb can be involved in theeducation of MDSC. As murine tBregs and human tBreg-like cellsgenerated after treatmentwith CMofMDA-MB-231 cells (16, 18),the B16/OC-associated B cells expressed elevated levels of TGFbcompared with na€�ve B cells (Supplementary Fig. S6A). Thus, wetested the role of TGFb in this process by coculturing tBregs withmMT-MDSC in the presence of TGFbblocking or control Ab. Then,MDSC (after removal of B cells) were tested in T-cell suppressionassay. The TGFb neutralization during the education of MDSCsignificantly abrogated the ability of MDSC to inhibit T-cellproliferation (P < 0.03; Fig. 5A and Supplementary Fig. S3). Toconfirm this finding, we also cocultured mMT-MDSC with tBregsin the presence of a specific TgfbR1 inhibitor SB431542. The

0

01/4 1/6

MDSC:CD3 ratio

CD115

CD40

Isotype control

PM

N-

PM

N-

Mo-

MD

SC

Mo-

MD

SC

Mock B-cont tBreg

CD80 CD83

CD124 pSTAT1 pSTAT3

1/8 1/10

10

10203040

CD

4 pr

olife

ratio

n (%

)C

D4

prol

ifera

tion

(%)

5060

70

A

E

F

B C DMo-MDSCMo-MDSC

DHE

Unloadedcontrol

Mock

Mock

5,000

4,000NS

3,000D

HE

(M

FI)

2,000

1,000

0 0

500

1,000

1,500

2,000

B-cont

B-cont

B-tumor

B-tumor

PMN-MDSC

PMN-MDSC

DH

E (

MF

I)

Non

-act

T

T a

lone

Moc

k

Moc

k

B-c

ont

tBre

g

tBre

gs

CD4

0 0

0

μMT

MD

SC

B-c

ont-

μMT

MD

SC

tBre

g-μM

T M

DS

C

WT

MD

SC

WT

Mo-

MD

SC

WT

PM

N-M

DS

C

500

1,000

1,500

2,000

2,500

3,000

10

μMT-MDSC

μMT mice with tumor

μMT micewith tumor

Gr1+cells

Gr1+

cells

Culture withB-cont

or B-tumor or

tBregs

WT micewith tumor

20

30

40

Pro

lifer

atio

n (%

)

50

60

10

20

30

40

Met

asta

tic fo

ci (

#)

50

60

CD8Mock B-cont B-tumor

2030405060

70

Figure 3.tBregs directly activate cancer-expanded MDSC rendering them regulatory. MDSC from peripheral blood of mMT mice with 4T1 cancer were cultured with Bna€�ve/B-cont or B-tumor/tBregs. After 5- to 16-hour incubation (longer for B-tumor cells), the cells were stained with DHE (B, D) or for several surface andintracellular molecules (E). After B-cell depletion, the MDSC were used in T-cell suppression assays as in Fig. 1G (A, C) or adoptively transferred in tumor-bearingmMT mice to assess lung metastasis (F). For the education experiment, we used ex vivo–generated tBregs or B-tumor cells sort-purified from at least threeWTmice with tumor. B cells were mixed with MDSC isolated from three mMT tumor–bearing mice. Each experiment was reproduced at least three times. � , P < 0.05;�� , P < 0.01; �� , P < 0.001. NS, not significant.






inhibitor indeed abrogated expression of iNOS, NOX2, and Arg(Supplementary Fig. S4A–S4C) and ROS production in MDSC(P<0.03; Fig. 5B and Supplementary Fig. S6B). Importantly, upontransfer of the mMT-MDSC educated in the presence of SB431542(after depletion of B cells and several washes with PBS to removetraces of the inhibitor) into mMT mice with 4T1 cancer, theyyielded significantly fewermetastatic foci in the lungs than controltBreg-educated mMT-MDSC (P < 0.03; Fig. 5C). Note the fact thatthe residual metastasis after the transfer of TgfbR1-inactivatedmMT-MDSC could be due to the fact that SB431542 only partiallyblocked ROS production in PMN-MDSC (Fig. 5B and Supple-mentary Fig. S6B). To further verify the importance of TGFbsignaling, we also used MDSC from mice with TgfbR2 deficiencyin myeloid cells (TgfbR2 KO). Although ROS/NO was reduced inTgfbR2 KO MDSC of mice with B16 melanoma (Fig. 4B and C)and 4T1 cancer (Supplementary Fig. S6C) as in mMT-MDSC,tBregs failed to upregulate its expression as in mMT-MDSC (Sup-plementary Fig. S6C). Importantly, unlike mMT-MDSC, tBregsfailed to increase the suppressive activity of TgfbR2 KOMDSC on

CD4þ and CD8þ T cells (P < 0.03; Fig. 5D). Upon their transferinto mMT mice, they also supported the lung metastasis of 4T1cancer cells less efficiently (P < 0.03; Fig. 5E). In summary, theseresults confirmed in multiple independent experiments suggestthat tBregs educate MDSC, at least in part, by targeting theTgfbR1/2 axis. A similar mechanism appears to be used byhuman myeloid cells, as SB431542 also impaired their educa-tion with ex vivo–generated tBregs (Fig. 4E and SupplementaryFig. S5B) and tBreg-like cells of B-CLL (Fig. 4F and Supple-mentary Fig. S5C), significantly (P < 0.03) reducing their abilityto suppress T-cell proliferation.

DiscussionHere,we attempted tomechanistically reconcile the issue raised

in our previous studies questioning the importance of MDSC inmetastasis. We repeatedly noticed that metastasis was blocked ifB cells/tBregs were lost (16–19) despite the fact that cancerdrastically expanded MDSC, the key facilitators of cancer escape

400

350

300

250

200

150

100

50

05 8 10 13 16 19

Days after tumor challenge

Tum

or s

ize

(mm

2 )

J HT

-B16

Moc

k

MDSC

WT

B16

WT

OC

5

4

3

2

1

0

Tum

or w

eigh

t (g)

MOCKJHT B16-MDSCWT B16-MDSC

WT OC-MDSC

600

500

400

300

200

100

0

RO

S (

DH

E M

FI)

J HT

B16

WT

B16

WT

OC

J HT

B16

WT

B16

WT

OC

700

600

500

400

300

200

100

0

J HT

B16

WT

B16

WT

OC

WT

OC

WT

B16

J HT

B16

TβR

2 K

O B

16

TβR

2 K

O B

16

TβR

2 K

O B

16

TβR

2 K

O B

16

700

600

500

400

300

200

100

0

DA

F-F

M (

MF

I)

Mo-MDSC PMN-MDSC PMN-MDSCMo-MDSC

700

600

500

400

300

200

100

0

CD4 CD8100

90

80

70

60

50

40

30

20

100

% P

rolif

erat

ion

T a

lone

Moc

k

MD

A

MD

A +

SB

MD

A −

CD

19 Moc

k

Moc

k +

SB

PS

#154

PS

#154

+S

B

PS

#174

PS

#174

+S

B

Activated CD3

80

70

60

50

40

30

20

10

0

% P

rolif

erat

ion

% P

rolif

erat

ion

70

60

50

40

30

20

10

0aCD3/CD28

WT B16 MDSC

WT OC MDSC

JHT B16 MDSC −−− − − −

−−−− − −

+ + ++

++

E FD

AA B C

− +

Figure 4.Murine and human cancer B cells also educate MDSC. MDSC (5 � 106) purified from peripheral blood and spleen of B16 melanoma–bearing JHT mice or WTmice (B16-MDSC) or mice with spontaneous ovarian cancer (OC-MDSC) were adoptively transferred into JHT mice (n ¼ 4–5 mice per group) 24 hours afterB16 tumor challenge (A). Peripheral blood MDSC from WT, JHT, TgfbR2 KO mice with B16 melanoma (WT B16, JHT B16, and TbR2KO B16, respectively) or micewith ovarian cancer (WT OC) were tested for ROS and NO production by DHE and DAF-FM staining (B, C) or used in T-cell suppression assay as in Fig. 1G(D). Experiments in A and C and D were reproduced three and two times, respectively. PBMC (n ¼ 7, healthy human donors), depleted of CD19þ cells, wereincubated with conditioned media of MDA-MB-231 cancer cells with or without SB 431542 (20 mmol/L) in the presence of 20 ng/mL GM-CSF for 48 hours. CD11bþ orCD14þ cells were sorted and used in T-cell suppression assay (E). In a similar experiment, human PBMCs (n ¼ 3, healthy human donors), depleted of T cellsand NK cells, were incubated with B cells from patients with B-CLL. After 48 hours, CD14þ cells were sorted and used in T-cell suppression assays (F). � , P < 0.05;�� , P < 0.01; �� , P < 0.001.

Bodogai et al.





and metastasis (1, 13). By comparing MDSC in WT BALB/c andC57BL/6micewith their congenic B-cell–deficientmice (mMTandJHT) with highly aggressive tumors, such as 4T1 cancer and B16melanoma, we demonstrate that tumors only expand and primethe regulatory function of MDSC. As such, MDSC from tumor-bearing mMTand JHT less efficiently produced ROS/NO, inhibitedproliferation of CD4þ and CD8þ T cells, and promoted tumorgrowth and metastasis. However, these impaired functions werecompletely reversed if the MDSC were briefly stimulated with Bcells from WT mice with tumors (which contains tBregs andcancer-induced B cells) but not na€�ve mouse B cells. Similarly,using modeling studies, we demonstrate that human cancer-induced B cells and tBreg-like cells from patients with B-CLL arealso required in the activation of human MDSC. Thus, theseresults clearly indicate that the cancer-primedMDSC require helpfrom tBregs/cancer-associated B cells to empower them fully

regulatory and thereby prometastatic. Our data also uncouplethe regulatory activity of MDSC from their expansion in responseto cancer, suggesting that the expansion of MDSC per se is not acriterion of their regulatory and prometastatic activity. Althoughwe can readily detect activated MDSC in various sites, such asperipheral blood, spleen, the lungs, and tumor, the site of theirencounter with tBregs remains unknown.

Activated B cells and Bregs are known to upregulate and useTGFb, for example, in conversion of FoxP3þ T cells and modu-lation of macrophages (25, 26). Similarly, we recently reportedthat murine and human tBregs also convert FoxP3þ T cells byoverexpressing TGFb (16, 18). Here, using various complemen-tary in vitro and in vivo modeling studies, we demonstrate thattBregs and cancer-induced B cells also use the TGFb-TgfbR1/2 axisin the activation (education) of both Mo and PMN subsets ofcancer-expanded MDSC. First, the presence of TGFb neutralizing,

Mock B-cont tBregs tBregsmlgG

tBregsαTFGβ

μMT TGFβRII KO

CD4 CD8

NSNS

Pro

lifer

atio

n (%

)

NS

NS

1/6 1/8 1/10

18,000

16,000

14,000

12,000

10,000

8,000

DH

E (

MF

I)

6,000

4,000

2,000

0

4,500

4,000

3,500

3,000

2,500

2,000

1,500

1,000

500

0

NS

Mo-MDSC PMN-MDSC

Moc

k

Moc

k

Moc

k

Moc

k

Moc

k

Moc

ktB

regs

tBre

gs

tBre

gs

tBre

gs

SB SBPBSPBS μMT MDSCs

tBre

gs

B-c

ont

80

70

60

50

40

30

20

10

0

Met

asta

tic fo

ci (

#)

80

70

60

50

40

30

20

10

0

90

Met

asta

tic fo

ci

Moc

k

Moc

k

B-c

ont

tBre

gs

tBre

gs

Moc

k

Moc

k

Moc

k

tBre

g

tBre

g

B-c

ont

Moc

k

Alo

neN

on-a

ct

Non

-act

Alo

ne

MDSCMDSC

Act TAct T

80

70

60

50

40

30

20

10

0

90

Pro

lifer

atio

n (%

)

TG

FβR

II K

O

WT

μMT

TGFβ

TGFβ

5

4

1

2

3

Tumor

tBreg

tBregtBreg

tBreg

CD4T cell

CD8T cell

TGFbRI/I

D E F

CBA

tBre

gs+

SB

60

50

40

30

20

10

0

Figure 5.tBregs educate MDSC via the TGFb-TgfbR1/TgfbR2 axis. 4T1 cancer–bearing mouse mMT-MDSC (A–C) and TgfbR2 KO-MDSC (D and E) were stimulated in vitrowith B-cont and tBregs with anti-TGFb neutralizing antibody or mouse IgG (50 mg/mL; A) in the presence or absence of SB431542 (20 mmol/L, SB; B and C).Then, MDSC were tested for T-cell suppression as in Fig. 1G at indicated effector-to-target ratio (A, D) or for expression of ROS (B) or induced metastasesupon transfer into mMT mice (C, E). Prior to the transfer or use in suppression assays, MDSC were depleted of B cells. The y-axis shows DHE staining (MFI) inMo-MDSCandPMN-MDSC (B), percentage of T-cell proliferation (gated in CD4þ andCD8þT cells; A, D), and number ofmetastatic foci in the lungs ofmMTmice (C, E)� SEM in 4 to 5 mice per group experiments reproduced three times. F, schematic: tumor initiates an expansion of MDSC (1) and conversion of tBregs fromna€�ve B cells (2). Cancer-activated B cells/tBregs use TGFb (3) to induce TGFb receptors on MDSC (4) and to upregulate ROS and NO production in MDSC. As aresult, MDSC become fully suppressive for T cells and thereby prometastatic (5). � , P < 0.05; �� , P < 0.01; �� , P < 0.001. NS, not significant.






but not control, Ab was sufficient to inhibit the MDSC educationin vitro. Second, when we blocked TgfbR1 signaling with a specificinhibitor SB431542 during the education of mMT-MDSC withtBregs,we failed to detect the upregulationof ROSproduction andsuppression of T cells. As a result, these MDSC supported metas-tasis less efficiently in mMT mice upon their adoptive transfer.Similarly, the TgfbR1 inhibitor also blocked the education ofhuman MDSC by tBregs induced by CM of MDA-MB-231 cellsand tBreg-like cells of patients with B-CLL. We also confirmedthese results usingMDSC from tumor-bearing mice withmyeloidcells deficient in TgfbR2, which is required for the signaling ofTgfbR1 (27). Unlike mMT-MDSC, tBregs failed to educate MDSCwith TgfbR2 KO, as shown by the loss of upregulation of ROSproduction, T-cell inhibition, and ability to support metastasisupon transfer into mMT mice. Thus, these results unequivocallyindicate that cancer-induced B cells and tBregs render the fullregulatory function of MDSC by, at least, targeting their TgfbR1/TgfbR2 signaling axis. In support, others recently reported that4T1 cancer also fails to metastasize in BALB/c mice deficient inTgfbR2 in myeloid cells (21), suggesting that this is due to theinability of their MDSC to get education from tBregs.

Our ex vivo studies with human MDSC and 3 murine tumormodels tested so far, such asmicewith 4T1 cancer, B16melanoma,and spontaneous ovarian cancer, suggest that theMDSCeducationis a common feature of cancer-activated B cells/Bregs. As in 4T1cancer model (where tBregs induce MDSC), the loss of mature Bcells also impaired the regulatory and tumor-augmenting func-tions ofMDSC inmice with B16melanoma that does not generatetBregs. Moreover, unlike na€�ve mouse B cells that cannot educateMDSCnor promote tumor growth (17–19), B cells frommicewithB16melanoma and spontaneous ovarian cancer not only inducedregulatory activity of MDSC and reversed the retarded tumorgrowth in B-cell deficient mice but also upregulated TGFb as intBregs (18). Similarly, B cells were required for the ex vivo gener-ation of suppressive myeloid cells from the peripheral blood ofhealthy human donors, as their removal or the blockage of TgfbR1signaling abrogated the generation of suppressive myeloid cells.Finally, tBreg-like cells of patients with B-CLL, but not B cells frompatients without tBregs or from healthy donors, also induced thegeneration of suppressive myeloid cells using TgfbR1 signaling.

In summary, our data indicate that the generation of regulatoryMDSC is a 2-step process (Fig. 5F). While drastically expandingMDSC, probably using GM-CSF, IL1b, VEGF, TGFb, and otherfactors (1, 13, 28), cancer also activates their inducers, such astBregs inmice with 4T1 cancer and ovarian cancer (16, 17) or/andyet to be identified Bregs in mice with B16 melanoma. Na€�ve Bcells do not educateMDSC. Although the nature and induction ofcancer-induced B cells in mice with B16 melanoma is a focus of aseparate study, we recently reported that 4T1 cancer directlyinduces the generation of tBregs from B2 cells via production of5-lipoxygenase metabolites (19). The reasons behind the 2-stepactivation process for MDSC remain unclear. This can be a way tolimit the use of MDSC only at the sites of inflammation andmetastasis, such as the tumors and lungs, via triggering thefeedback loop induced by TGFb and ROS/NO, as shown for othercells (29–31) where the activation of TGFb and production of

ROS/H2O2/NO are mutually regulated. In concordance, mMT-MDSC became fully regulatory and prometastatic within a brief(<5 hours) encounter with tBregs. The education not onlyincreased a number of surface molecules linked with regulatoryMDSC (e.g., IL4Ra, TGFb, and IL10) but also upregulated surfaceexpression of TgfbRs presumably as a part of the feedback loop.Thus, the educationmay also bring additional factors that furtheraugment the differentiation and regulatory function ofMDSC, thefocus of a different study. On the other hand, it is tempting tospeculate that the early stage of cancer requires tBregs to supportmetastasis of 4T1 cancer via inducing metastasis-protectingFoxP3þTregs (17), as we detect tBregs at least 1 week before themassive expansion of MDSC. As such, the loss of Tregs mostlyblocks 4T1 cancer metastasis without affecting primary tumorgrowth (15, 17). At the later stage of cancer, tBregs appear to beneeded to counteract the induction of antitumormyeloid cells (8)by switching and enhancing their regulatory function.Overall, thedata presented here further underscore the importance of B cells incancer (24, 32–36) by adding for the first time the education ofMDSC to a growing list of protumorigenic functions, such as theinhibition of cytotoxic CD8þ T andNK cells (37), Treg conversion(38), andM1-to-M2macrophage polarization (39, 40). As such, Bcells have to be targeted to enhance antitumor immune responses.

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

Authors' ContributionsConception and design: M. Bodogai, A. BiragynDevelopment of methodology: M. Bodogai, C. Lee-ChangAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.):M. Bodogai, K. Moritoh, C.M. Hollander, R.P. Wersto,Y. Araki, I. Miyoshi, L. YangAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): M. Bodogai, A. BiragynWriting, review, and/or revision of the manuscript: M. Bodogai,R.P. Wersto, L. Yang, G. Trinchieri, A. BiragynAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): M. Bodogai, C.A. Sherman-Baust, R.P. WerstoStudy supervision: A. Biragyn

AcknowledgmentsThe authors thank Dr. Kathy Perdue (NIA/NIH) for the help with re-

derivation of mMT mice, Dr. Charles Hesdorffer (VA, Washington, DC) forB-CLL cells, Drs. Salman Tajuddin and Ilya Goldberg (NIA/NIH) for help withstatistical analysis, Ana Lustig and Dr. Nicole N. Hooten (NIA/NIH) for criticalreading this article.

Grant SupportThis research was supported by the Intramural Research Program of the

National Institute on Aging, NIH, and CRADA with Janssen Research Devel-opment program.

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 November 10, 2014; revised June 4, 2015; accepted June 13, 2015;published OnlineFirst July 16, 2015.

References1. Gabrilovich DI, Nagaraj S. Myeloid-derived suppressor cells as regulators

of the immune system. Nat Rev Immunol 2009;9:162–74.2. Gabitass RF, Annels NE, StockenDD, PandhaHA,MiddletonGW. Elevated

myeloid-derived suppressor cells in pancreatic, esophageal and gastric

Bodogai et al.





cancer are an independent prognostic factor and are associated withsignificant elevation of the Th2 cytokine interleukin-13. Cancer ImmunolImmunother 2011;60:1419–30.

3. Serafini P, De SC, Marigo I, Cingarlini S, Dolcetti L, Gallina G, et al.Derangement of immune responses by myeloid suppressor cells.Cancer Immunol Immunother 2004;53:64–72.

4. Movahedi K, Guilliams M, Van den Bossche J, Van den Bergh R, GysemansC, Beschin A, et al. Identification of discrete tumor-induced myeloid-derived suppressor cell subpopulations with distinct T cell-suppressiveactivity. Blood 2008;111:4233–44.

5. Youn JI, Nagaraj S, Collazo M, Gabrilovich DI. Subsets of myeloid-derivedsuppressor cells in tumor-bearing mice. J Immunol 2008;181:5791–802.

6. Mandruzzato S, Solito S, Falisi E, Francescato S, Chiarion-Sileni V,MocellinS, et al. IL4Ralphaþ myeloid-derived suppressor cell expansion in cancerpatients. J Immunol 2009;182:6562–8.

7. Liu CY, Wang YM, Wang CL, Feng PH, Ko HW, Liu YH, et al. Populationalterations of L-arginase- and inducible nitric oxide synthase-expressedCD11bþ/CD14(-)/CD15þ/CD33þmyeloid-derived suppressor cells andCD8þ T lymphocytes in patients with advanced-stage non-small cell lungcancer. J Cancer Res Clin Oncol 2010;136:35–45.

8. Kallberg E, Stenstrom M, Liberg D, Ivars F, Leanderson T.CD11bþLy6CþþLy6G- cells show distinct function in mice with chronicinflammation or tumor burden. BMC Immunol 2012;13:69.

9. Bronte V, Zanovello P. Regulation of immune responses by L-argininemetabolism. Nat Rev Immunol 2005;5:641–54.

10. Bronte V, Serafini P, De Santo C, Marigo I, Tosello V, Mazzoni A, et al. IL-4-induced arginase 1 suppresses alloreactive T cells in tumor-bearing mice.J Immunol 2003;170:270–8.

11. Gallina G, Dolcetti L, Serafini P, De Santo C, Marigo I, Colombo MP, et al.Tumors induce a subset of inflammatory monocytes with immunosup-pressive activity on CD8þ T cells. J Clin Invest 2006;116:2777–90.

12. Corzo CA, Cotter MJ, Cheng P, Cheng F, Kusmartsev S, Sotomayor E, et al.Mechanism regulating reactive oxygen species in tumor-induced myeloid-derived suppressor cells. J Immunol 2009;182:5693–701.

13. Bunt SK, Sinha P, Clements VK, Leips J, Ostrand-Rosenberg S. Inflamma-tion induces myeloid-derived suppressor cells that facilitate tumor pro-gression. J Immunol 2006;176:284–90.

14. Lelekakis M, Moseley JM, Martin TJ, Hards D, Williams E, Ho P, et al. Anovel orthotopic model of breast cancer metastasis to bone. Clin ExpMetastasis 1999;17:163–70.

15. Olkhanud PB, Baatar D, Bodogai M, Hakim F, Gress R, Anderson RL, et al.Breast cancer lung metastasis requires expression of chemokine receptorCCR4 and regulatory T cells. Cancer Res 2009;69:5996–6004.

16. Olkhanud PB, Damdinsuren B, Bodogai M, Gress RE, Sen R, Wejksza K,et al. Tumor-evoked regulatory B cells promote breast cancer metastasis byconverting resting CD4þ T cells to T-regulatory cells. Cancer Res 2011;71:3505–15.

17. BodogaiM, Lee Chang C,Wejksza K, Lai J, MerinoM,Wersto RP, et al. Anti-CD20 antibody promotes cancer escape via enrichment of tumor-evokedregulatory B cells expressing low levels of CD20 and CD137L. Cancer Res2013;73:2127–38.

18. Lee-Chang C, Bodogai M, Martin-Montalvo A, Wejksza K, Sanghvi M,Moaddel R, et al. Inhibition of breast cancer metastasis by resveratrol-mediated inactivation of tumor-evoked regulatory B cells. J Immunol2013;191:4141–51.

19. Wejksza K, Lee-Chang C, Bodogai M, Bonzo J, Gonzalez FJ, Lehrmann E,et al. Cancer-produced metabolites of 5-lipoxygenase induce tumor-evoked regulatory B cells via peroxisome proliferator-activated receptoralpha. J Immunol 2013;190:2575–84.

20. Sherman-Baust CA, Kuhn E, Valle BL, Shih IeM, Kurman RJ,Wang TL, et al.A genetically engineered ovarian cancer mouse model based on fallopiantube transformation mimics human high-grade serous carcinoma devel-opment. J Pathol 2014;233:228–37.

21. Pang Y, Gara SK, Achyut BR, Li Z, YanHH,DayCP, et al. TGF-beta signalingin myeloid cells is required for tumor metastasis. Cancer Discov 2013;3:936–51.

22. Cai H, Dikalov S, Griendling KK, Harrison DG. Detection of reactiveoxygen species and nitric oxide in vascular cells and tissues: comparisonof sensitivity and specificity. Methods Mol Med 2007;139:293–311.

23. Overwijk WW, Theoret MR, Finkelstein SE, Surman DR, de Jong LA, Vyth-Dreese FA, et al. Tumor regression and autoimmunity after reversal of afunctionally tolerant state of self-reactive CD8þ T cells. J Exp Med2003;198:569–80.

24. Qin Z, Richter G, Schuler T, Ibe S, Cao X, Blankenstein T. B cells inhibitinduction of T cell-dependent tumor immunity. Nat Med 1998;4:627–30.

25. Bao Y, Cao X. The immune potential and immunopathology of cytokine-producing B cell subsets: A comprehensive review. J Autoimmun 2014;55:10–23.

26. Reyes JL, Wang A, Fernando MR, Graepel R, Leung G, van Rooijen N, et al.Splenic B cells from Hymenolepis diminuta-infected mice amelioratecolitis independent of T cells and via cooperation with macrophages.J Immunol 2015;194:364–78.

27. Taylor AW. Review of the activation of TGF-beta in immunity. J Leukoc Biol2009;85:29–33.

28. DuPre SA, Redelman D, Hunter KW Jr The mouse mammary carcinoma4T1: characterization of the cellular landscape of primary tumours andmetastatic tumour foci. Int J Exp Pathol 2007;88:351–60.

29. Michaeloudes C, Sukkar MB, Khorasani NM, Bhavsar PK, Chung KF. TGF-beta regulates Nox4, MnSOD and catalase expression, and IL-6 release inairway smoothmuscle cells. Am J Physiol LungCellMol Physiol 2011;300:L295–304.

30. De Bleser PJ, Xu G, Rombouts K, Rogiers V, Geerts A. Glutathione levelsdiscriminate between oxidative stress and transforming growth factor-betasignaling in activated rat hepatic stellate cells. J Biol Chem 1999;274:33881–7.

31. AyacheN, BoumedieneK,Mathy-HartertM, Reginster JY,Henrotin Y, PujolJP. Expression of TGF-betas and their receptors is differentially modulatedby reactive oxygen species and nitric oxide in human articular chondro-cytes. Osteoarthritis Cartilage 2002;10:344–52.

32. DongHP, ElstrandMB,Holth A, Silins I, Berner A, Trope CG, et al. NK- andB-cell infiltration correlates with worse outcome in metastatic ovariancarcinoma. Am J Clin Pathol 2006;125:451–8.

33. Brodt P, Gordon J. Anti-tumor immunity in B lymphocyte-deprivedmice. I. Immunity to a chemically induced tumor. J Immunol 1978;121:359–62.

34. de Visser KE, Korets LV, Coussens LM.De novo carcinogenesis promoted bychronic inflammation is B lymphocyte dependent. Cancer Cell 2005;7:411–23.

35. Ammirante M, Luo JL, Grivennikov S, Nedospasov S, Karin M. B-cell-derived lymphotoxin promotes castration-resistant prostate cancer.Nature 2010;464:302–5.

36. Lindner S, Dahlke K, Sontheimer K, HagnM, Kaltenmeier C, Barth TF, et al.Interleukin 21-induced granzyme B-expressing B cells infiltrate tumors andregulate T cells. Cancer Res 2013;73:2468–79.

37. Inoue S, Leitner WW, Golding B, Scott D. Inhibitory effects of B cells onantitumor immunity. Cancer Res 2006;66:7741–7.

38. Townsend MJ, Monroe JG, Chan AC. B-cell targeted therapies in humanautoimmune diseases: an updated perspective. Immunol Rev 2010;237:264–83.

39. Wong SC, Puaux AL, Chittezhath M, Shalova I, Kajiji TS, Wang X, et al.Macrophage polarization to a unique phenotype driven by B cells.European J Immunol 2010;40:2296–307.

40. Affara NI, Ruffell B, Medler TR, Gunderson AJ, Johansson M, Bornstein S,et al. B cells regulate macrophage phenotype and response to chemother-apy in squamous carcinomas. Cancer Cell 2014;25:809–21.






2015;75:3456-3465. Published OnlineFirst July 16, 2015.Cancer Res Monica Bodogai, Kanako Moritoh, Catalina Lee-Chang, et al. Tumor-Associated B CellsMyeloid-Derived Suppressive Cells Rely upon Education from Immunosuppressive and Prometastatic Functions of

Updated version

10.1158/0008-5472.CAN-14-3077doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2015/07/17/0008-5472.CAN-14-3077.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/75/17/3456.full#ref-list-1

This article cites 40 articles, 18 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/75/17/3456.full#related-urls

This article has been cited by 7 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

[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://cancerres.aacrjournals.org/content/75/17/3456To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-14-3077

http://cancerres.aacrjournals.org/content/suppl/2015/07/17/0008-5472.CAN-14-3077.DC1

http://cancerres.aacrjournals.org/content/75/17/3456.full#ref-list-1

http://cancerres.aacrjournals.org/content/75/17/3456.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/75/17/3456


Documents

Immunosuppressive and Prometastatic Functions of Myeloid ... · of Myeloid-Derived Suppressive Cells Rely upon Education from Tumor-Associated B Cells Monica Bodogai1, Kanako Moritoh1,