Tumor-Inﬁltrating Myeloid Cells Activate...

Molecular and Cellular Pathobiology

Tumor-Infiltrating Myeloid Cells Activate Dll4/Notch/TGF-bSignaling to Drive Malignant Progression

Hidetaka Ohnuki1, Kan Jiang1, Dunrui Wang3, Ombretta Salvucci1, Hyeongil Kwak1, David S�anchez-Martín1,Dragan Maric2, and Giovanna Tosato1

AbstractMyeloid cells that orchestrate malignant progression in the tumor microenvironment offer targets for a

generalized strategy to attack solid tumors. Through an analysis of tumor microenvironments, we explored anexperimental model of lung cancer that uncovered a network of Dll4/Notch/TGF-b1 signals that links myeloidcells to cancer progression. Myeloid cells attracted to the tumor microenvironment by the tumor-derivedcytokines CCL2 andM-CSF expressed increased levels of the Notch ligandDll4, thereby activating Notch signalingin the tumor cells and amplifying tumor-intrinsic Notch activation. Heightened Dll4/Notch signaling in tumorcells magnified TGF-b–induced pSMAD2/3 signaling and was required to sustain TGF-b–induced tumor cellgrowth. Conversely, Notch blockade reduced TGF-b signaling and limited lung carcinoma tumor progression.Corroborating these findings, by interrogating RNAseq results from tumor and adjacent normal tissue in clinicalspecimens of human head and neck squamous carcinoma, we found evidence that TGF-b/Notch crosstalkcontributed to progression. In summary, the myeloid cell-carcinoma signaling network we describe uncoversnovelmechanistic links between the tumormicroenvironment and tumor growth, highlighting newopportunitiesto target tumors where this network is active. Cancer Res; 74(7); 1–12. �2014 AACR.

IntroductionThe tumor microenvironment is increasingly recognized

as an enabling contributor to tumor progression (1, 2), andstrategies that target the tumor microenvironment are effec-tive at reducing tumor growth, despite persistence of genet-ically modified tumor cells (3–5).The vasculature is a component of the tumor microenvi-

ronment that contributes to tumor growth through angiogen-esis. Inhibition of angiogenesis by neutralization of VEGF iseffective at reducing progression of certain tumors despitehaving little effect on most tumor cells (5). "Inflammatory"cells, particularly "M2-type" myeloid cells (2) and stromalfibroblast-like cells/cancer-associated fibroblastic cells (1) areother protumorigenic components of the cancer microenvi-ronment. Through integrin-mediated adhesion signaling and

other mechanisms myeloid cells promote cancer cell survival(6). Acting directly or through effector molecules, includingTGF-b, fibroblast growth factors (FGF), and epidermal growthfactors, cancer-associatedmyeloid andfibroblastic cells supplymitogenic signals to tumors (1). By secreting VEGF, basic FGF,platelet-derived growth factor, placental growth factor, andBv8, myeloid cells promote tumor angiogenesis (5,7). By pro-ducing various proteases, myeloid cells induce the releaseVEGF and other mitogenic factors sequestered in the extra-cellular matrix and disrupt tissue integrity by cleaving homo-typic and heterotypic cell adhesion molecules (8).

Pressing motivation to abrogate myeloid-derived protu-morigenic signals has produced some encouraging results.Blocking macrophage recruitment with antagonists of colo-ny-stimulating factor-1 receptor reduced mammary tumorprogression and increased mice survival (4). TGF-b–targeteddrugs are currently in clinical trials after showing encouraginganticancer activity in preclinical studies (9,10). Matrix metalloproteinases inhibitors, which showed promising antitumoractivities in mouse but not in human cancer trials, are beingre-evaluated in light of emerging new understanding (11).

Despite these advances, the complexities of cell compositionof tumormicroenvironments and tendency to adaptive changepresent current obstacles. To overcome some of these diffi-culties, we have queried the "simpler" tumor microenviron-ment of growth factor independence-1 (Gfi1)-null mice, whichlack mature granulocytes and harbor functionally impairedmyeloid cells (12,13) to identify principal mechanisms thatsustain reciprocal communications between tumor cells andcells of the tumor microenvironment. By focusing on Notchsignaling, a crucial regulator of cell fate decisions activated by

Authors' Affiliations: 1Laboratory of Cellular Oncology, Center for CancerResearch, National Cancer Institute, NIH; 2Department of IntramuralResearch National Institutes of Neurological Disorders and Stroke, NIH,Bethesda, Maryland; and 3W2Motif, LLC, San Diego, California

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

Current address for K. Jiang: Lymphocyte Cell Biology Section, NationalInstitute of Arthritis and Musculoskeletal and Skin Diseases, Bethesda,Maryland

Corresponding Author: Giovanna Tosato, Laboratory of Cellular Oncol-ogy, CCR, NCI, NIH, Building 37, Room 4124, 37 Convent Drive, Bethesda,MD 20892. Phone: 301-594-9596; Fax: 301-594-9585; E-mail:[email protected]

doi: 10.1158/0008-5472.CAN-13-3118

�2014 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org OF1

Research. on May 6, 2018. © 2014 American Association for Cancercancerres.aacrjournals.org Downloaded from

Published OnlineFirst February 11, 2014; DOI: 10.1158/0008-5472.CAN-13-3118

http://cancerres.aacrjournals.org/

cell-to-cell interaction between Notch ligand (Dll1, Dll3, Dll4,Jagged1, and Jagged2) and Notch receptor (Notch1–Notch4)–expressing cells (14–16), we now uncovered a novel networkof Dll4/Notch/TGF-b signals linking myeloid cells and cancercells that drives lung carcinoma tumor progression. Thisnetwork provides a mechanistic link between tumor-infiltrat-ing myeloid cells and tumor cells with opportunities forintervention.

Materials and MethodsCell culture and in vitro treatments

TheEL4, LLC1, andB16F10murine cell lines [AmericanTypeCulture Collection (ATCC); authentication confirmed by ATCCthrough depositor's analysis of representative cultures fromthe master seed stock] were propagated in the laboratory forfewer than 6 months in culture medium (RPMI or Dulbecco'sModified Eagle Medium with 1% BSA, 2 mmol/L L-glutamine,100 IU/mL penicillin, 100 mg/mL streptomycin, 50 mmol/L 2-ME, and 1 mmol/L sodium pyruvate); proliferation was mea-sured by 3H thymidine incorporation (17). Recombinanthuman TGF-b1 (R&D Systems), DAPT [N-[N-(3,5-difluorophe-nacetyl-L-alanyl)]-S-phenylglycine tert-butyl ester; Sigma-Aldrich], DBZ [(2S)-2-[2-(3,5-difluorophenyl)-acetylamino]-N-(5-methyl-6-oxo-6,7-dihydro-5H-dibenzo[b,d]azepin-7-yl)-propionamide; Millipore], and appropriate diluent controlswere added to culture; recombinant mouse His tag-Dll4(R&D Systems) and His control (Millipore) were immobi-lized (18 hours at 4�C) to culture vessels before cell addition.

Animal studiesAnimal experiments were approved by the NCI-Bethesda

Institutional Animal Care and Use Committee and conductedaccording to protocol. Gfi1þ/þ, Gfi1þ/�, and Gfi1�/� male andfemalemice (18)wereusedbetween 4 and 8weeks of age.Mousetumor cell lines were implanted (10 � 106) subcutaneously inthe left abdominal quadrant. DAPT (10mg/kg, i.p. 5 days/week)or diluent control treatmentwas initiated 24 hours after tumor–cell injection. Tumors were removed in toto after 12 and 16 days.Spleens and bone marrows were obtained from tumor-bearingand control mice.

Flow cytometric analysis and cell sortingSingle-cell suspensions from bone marrow, spleen, and

tumor tissues were incubated with mouse Fc block CD16/32antibody (2.4G2; BD Biosciences) for 20 minutes at 4�C in PBScontaining 2%BSA (PBS/BSA) to reduce nonspecific antibodybinding. After washing in PBS/BSA, cells were incubatedwith control Ig or fluorophore-conjugated antibodies in PBSwith 1%BSA and 2 mmol/L EDTA. Cell sorting and datacollection were performed on a FACSVantage SE or FACSAria(BD Biosciences); data analysis used Flowjo software. Detailson antibodies are found in Supplementary ExperimentalProcedures.

Immunohistochemistry and immunoblottingTissues were fixed with 2% or 4% paraformaldehyde over-

night or 4 hours at 4�C (19). Tissue immunostaining andquantification was performed as described previously (19).

Protein extracts prepared as described (19) were run through4% to 12% bis-Tris gels (Invitrogen) or 10% to 20% polyacryl-amide gels (Novex), transferred to protran BA83 celluloseni-trate membranes (Whatman) and stained with the primaryand secondary antibodies as detailed in Supplementary Exper-imental Procedures.

Bioinformatics and statistical analysisAll bioinformatic analyses were conducted on the publically

available gene expression data (normalized values from Illu-mina RNAseq version 2, level 3) fromTheCancer GenomeAtlas(TCGA; http://cancergenome.nih.gov/). The data were down-loaded from TCGA matrix and was evaluated by box plotanalysis and the Mann–Whitney U test using the R system(2.14.1) for statistical computation and graphics. In all otherexperiments, group differences were analyzed by using the 2-tailed Student t test with equal variance assumption and theFisher exact test (Microsoft Excel). P � 0.05 were consideredsignificant.

ResultsHost dependency of LLC1 carcinoma and EL4 T-celllymphoma progression

To explore contributions of the tumor microenvironmentto tumor progression, we utilized Gfi1-null mice that lackmature granulocytes and have functionally defective mono-cytes, while displaying a mostly intact lymphoid system(12,13,18). Gfi1-heterozygote mice are indistinguishablefrom wild type (12,13). By analysis of syngeneic subcutane-ous transplant systems, we evaluated tumor growth inducedby cell lines representative of T-cell lymphoma (EL4), lungcarcinoma (LLC1), and melanoma (B16F10; Fig. 1A–C andSupplementary Fig. S1). EL4 cells generated tumors thatgrew more aggressively (Fig. 1A and Supplementary Fig. S1)in Gfi1-null (KO) mice compared with Gfi1þ/þ (wild type,WT) or Gfi1þ/� heterozygous (Het) mice. By contrast, LLC1cells generated tumors that grew more aggressively (Fig. 1Band Supplementary Fig. S1) in Gfi1-WT/Het mice comparedwith Gfi1 KO. B16F10 cells generated tumors that grewsimilarly in Gfi1-WT/Het and KO mice (Fig. 1C and Supple-mentary Fig. S1). We concluded that EL4 and LLC1 tumorprogression is significantly affected by host factors.

We hypothesized that the Gfi1 and wild-type tumor micro-environment differed in EL4 and LLC1 tumors, but not inB16F10 tumors. Because neutrophils, a source of the pro-angiogenic Bv8 factor (7), are absent in Gfi1-null mice, weexamined tumor vascularization. We found that vasculariza-tion of EL4 and LLC1 tumors fromWT/Het and Gfi1-null micewas quantitatively and morphologically similar, as assessed byCD31 immunostaining (Supplementary Fig. S2A and S2B). Acomprehensive analysis of major cell types revealed a signif-icantly greater infiltration of CD11bþLy6CþLy6G� cells inLLC1 tumors from Gfi1-null mice compared with control,whereas this population was similarly represented in EL4tumors from Gf1-null and wild-type hosts, and was rare inB16F10 tumor tissues (Fig. 1D and E). By contrast, theCD11bþLy6CþLy6Gþ cells were significantly more abundantin EL4 tumors from wild type compared with Gfi1-null mice;

Ohnuki et al.

Cancer Res; 74(7) April 1, 2014 Cancer ResearchOF2




Figure 1. The Gfi1-null microenvironmentregulates tumor progression. A–C, tumorweightfrom control (WT Gfi1þ/þ or het Gfi1þ/�) andGfi1-null (KO Gfi1�/�) mice analyzed 12 to 15days postsubcutaneous injection of EL4, LLC1,and B16F10 tumor lines. Data are averages �SD from individual experiments, eachrepresentative of three experiments performed:A, EL4 tumors WT/Het, n ¼ 12; KO, n ¼ 10; B,LLC1 tumors WT/Het, n ¼ 15; KO, n ¼ 12; C,B16F10 tumors WT/Het, n ¼ 12; KO, n ¼ 10;P values from the Student t test. D–G,monocytes and granulocytes infiltrate tumorsfrom control and Gfi1-null mice. In the bargraphs (D and F), flow cytometry data areexpressed as average percentage of total cellsfrom tumor � SD; EL4, n ¼ 5; LLC1, n ¼ 6;B16F10, n ¼ 3. In the representative flowcytometry plots (E and G), the numerical valuesare expressed as percentages of total CD11bþ

leukocytes in the tumor; P values from theStudent t test. H and I, distribution of CD4þ andCD8þ lymphocytes in tumors. The data areexpressed as average percentage of total cellsfrom tumor � SD; EL4, n ¼ 5; LLC1, n ¼ 6;B16F10, n¼ 3; P values from the Student t test.J, frequency of tumor development in wild-typemice injected with EL4 cells alone or withsplenocytes unfractionated (wild-type or KO) ordepleted of Ly6Gþ cells (WT). Splenocytes werefrom EL4-bearing mice. EL4 alone, EL4 þ WTcells, EL4 þ KO cells, n ¼ 10; EL4þWT LyG�

cells, n ¼ 6; wild-type or KO cells alone, n ¼ 3.Data indicate the percentage of mice injectedthat developed tumors over a period of 14 to16 days; P values from the Fisher exact test. K,tumor weight in wild-type mice injected withLLC1 cell alone or with wild-type monocytessorted from bonemarrow of LLC1-bearing wild-typeor KOmice; n¼6/group.Data are averaged� SD; P values from the Student t test.

Dll4 Expression in Monocytes Drives Tumor Growth

www.aacrjournals.org Cancer Res; 74(7) April 1, 2014 OF3




this population was virtually absent in LLC1 and B16 tumortissues from wild-type and Gfi1-null hosts (Fig. 1F and G).CD4þ and CD8þ T lymphocytes (Fig. 1H and I) were similarlyrepresented in EL4, LLC1, and B16F10 tumor tissues fromwild-type and Gfi1-null mice. CD25þ T cells; CD4þFOXP3þ T cells,B220þ B (and other) cells; CD11cþB220þ dendritic cell popula-tions; SMAþ myofibroblasts; CD49b NK cells and CD11bþF4/80 macrophages were also similarly represented in EL4, LLC1,and B16F10 tumor tissues from wild-type and Gfi1-null mice(not shown).

Reflecting the Gfi1-null phenotype, spleens (SupplementaryFig. S2C and S2D) and bone marrows (Supplementary Fig. S2Dand S2E) from Gfi1-null mice (control and tumor-bearing)showed an increase of CD11bþLy6CþLy6G� cells and adecrease of CD11bþLy6CþLy6Gþ cells compared with wild-type mice. Spleens of EL4 tumor-bearing wild-type mice dis-played a significant increase of granulocytes compared withna€�ve spleens, and spleens of LLC1 tumor-bearing wild-typemice displayed a significant increase of monocytes comparedwith na€�ve spleens (Supplementary Fig. S2C). Given the pre-dominant differences in infiltrates distinguishing EL4 andLLC1 tumors growing in the Gfi1 and wild-type hosts andchanges developing in the spleens of tumor-bearing mice, wefocused on the role of CD11bþLy6CþLy6G� andCD11bþLy6CþLy6Gþ cells in these models.

Adoptive transfer experiments supported a tumor-inhibito-ry activity of wild-type granulocytes in the EL4 system, becausedepletion of Ly6Gþ cells reduced the antitumor activity ofwild-type splenocytes (Fig. 1J), and a tumor-promoting function ofmonocytes in the LLC1 tumor model, becauseCD11bþLy6CþLy6G� cells enhanced LLC1 tumor growthwhereas this cell population from Gfi1-null mice did not (Fig.1K).

LLC1 cells express CCL2/MCP1 mRNA (Fig. 2A; ref. 20)and protein present in LLC1 culture supernatant (2.8 ng/mL).Control and Gfi1-null CD11bþGr1þ cells similarly migrated torecombinant CCL2 (Fig. 2B), suggesting that LLC1-derivedCCL2 may recruit myeloid cells to the tumor. LLC1 cells alsoexpress M-CSF/CSF1 mRNA (Fig. 2A), but protein was unde-tected in culture supernatant. EL4 cells express GM-CSF/CSF2mRNA (Fig. 2A) and protein detected in the culture superna-tant at 21 pg/mL, suggesting that itmay recruit granulocytes toEL4 tumors (21).

Identification of myeloid cell-derived signals thatmodulate tumor cell growth

Results from adoptive transfer experiments suggested thatmyeloid cell types recruited by EL4 and LLC1 tumor cellsmightproduce paracrine signals that modulate cancer cell growth/survival. To identify such signals, we profiled gene expressionin EL4, LLC1, and B16F10 tumors from wild-type and Gfi1-deficient mice. To distinguish signals derived from the tumormicroenvironment from others derived from the tumor cells,we profiled in parallel EL4, LLC1, and B16F10 cell lines fromculture. Focusing on genes previously linked to modulation oftumor growth, we measured 57 mRNAs in 10 EL4 tumors(5 each from wild-type and Gfi1-null mice) and in 15 LLC1tumors (10 from wild type; 5 from Gfi1-null mice) from two to

four different experiments (Table 1 and Supplementary TableS1). From this pool, we identified 10 mRNAs expressed atsignificantly different levels in EL4 and/or LLC1 tumors arisingin wild type as opposed to Gfi1-null mice (Table 1). Extendinganalysis of these 10 mRNAs to B16F10 tumors, we found noexpression difference between tumors from wild-type andGfi1-null mice (Table 1).

All but 1 (Cxcr4) of the 10 candidate genes fulfilled thecriteria of being likely host induced in that expression washigher in theEL4 or LLC1 tumors comparedwith the tumor cellline. For the remainder, we looked for myeloid-derived genespreferentially expressed in the wild-type host that might belinked to suppression of EL4 and stimulation of LLC1 tumorgrowth. Tgif1 (encodes TGF-b1) and Tgif2 (encodes TGF-b2)fulfilled these criteria (Table 1). TGF-b1/2 is a monocyte andneutrophil product that can inhibit and stimulate cell growthdependent on context: it is a tumor suppressor in early tumordevelopment, but a tumor enhancer in more advanced tumors(3). We confirmed that TGF-b1 mRNA is expressed at higherlevels in the LLC1 tumor microenvironment of wild typecompared with Gfi1-null mice by sorting the CD11bþGr1þ

cells (Fig. 2C). Similarly, we confirmed that TGF-b2 mRNA isexpressed at significantly higher levels in the EL4 tumormicroenvironment of wild-type mice compared with Gfi1-nullmice (not shown). Na€�ve spleens from wild-type mice consti-tutively express higher levels of TGF-b1 (Fig. 2D) and TGF-b2(not shown) mRNA compared with Gfi1-null mice, and na€�veCD11bþGr1þ from wild-type bone marrow secrete higherlevels of TGF-b1 compared with Gfi1-null bone marrow, bothconstitutively and after activation with M-CSF/CSF1 or GM-CSF/CSF2 (Fig. 2E). We also found that levels of the TGF-bsignaling mediator pSMAD3 were higher in tumors arising inwild-typemice than inGfi1-nullmice (Fig. 2F andG), indicativeof greater TGF-b activity in vivo. We tested the effects of TGF-bon tumor cell growth. TGF-b1 significantly and dose-depen-dently reduced EL4 proliferation but enhanced LLC1 prolifer-ation (Fig. 2H). Increased LLC1 cell proliferation by TGF-b1is cell density dependent, suggesting a requirement for cell–cellinteraction (Fig. 2I); no such cell dose dependency wasobserved with EL4 cells (not shown).

To investigate this cell dose dependency of LLC1 prolifer-ation to TGF-b, we examined the potential involvement ofNotch signaling, because it is induced by cell–cell interactionand can establish cooperative crosstalk with TGF-b signalingin other systems (14–16). We found that the Notch ligand Dll4and the Notch signaling mediator Hey2 were expressed athigher levels in LLC1 tumors from wild type compared withGfi1-KO (Table 1). To identify the cell sources of the differen-tially expressed Dll4 and Hey2, we characterized Notch recep-tors/ligands expression in tumor cells and myeloid cells. LLC1cells express Notch1 and Notch4 receptors mRNAs at higherlevels than the EL4 and B16F10 cells (Fig. 3A); primary wild-typemonocytes sorted frombonemarrow express higher levelsof Dll4 mRNA (Notch1 and Notch4 ligand) than Gfi1-nullmonocytes (Fig. 3B). LLC1 cells also express endogenous Dll1mRNA (Notch1/2/3 ligand) andDll4mRNA (Fig. 3A).We sortedwild-type and Gfi1-null CD11bþLy6CþLy6G- cells from LLC1tumor cell suspensions tomeasure expression levels of Notch1,

Ohnuki et al.





Figure 2. TGF-b inhibits EL4 cell proliferation and increases LLC1 cell proliferation. A, relative cytokine mRNA expression in cell lines; results are fromqPCR (averages from duplicate measurements). B, mouse CCL2 induces migration of bone marrow CD11bþGr1þ cells. Data are averages � SD; n ¼ 3experiments. C, relative TGF-b1 expression in CD11bþGr1þ cells sorted from tumor cell suspensions. Data from qPCR are averages � SD; n ¼ 5 to 6tumors/group. D, relative TGF-b1 expression in naïve splenocytes; qPCR data aremeans�SD; n¼ 5. E, TGF-b in supernatant of bonemarrowCD11bþGr1þ

cells; M-CSF (20 ng/mL), GM-CSF (40 ng/mL); ELISA results are means � SD; triplicate cultures. F, representative images of pSMAD3-expressingcells in LLC1 tumor; nuclei are visualizedwith DAPI. G, quantification of pSMAD3-expressing cells in EL4 and LLC1 tumors. Data are averages�SDcells/field(�40); 5 fields/tumor; n ¼ 5 tumors/group. H, proliferation of EL4 and LLC1 cells to TGF-b1. Representative results (of 3–5 experiments) are means � SD(triplicate cultures). I, LLC1 proliferation to TGF-b is cell-density dependent. Representative results (of three experiments) are means � SD of (triplicatecultures). All P values are from the Student t test.






Notch4, Dll4, Hey1, and Hey2 (Supplementary Fig. S3A). LLC1tumor-infiltrating CD11bþLy6CþLy6G� cells from wild-typemice expressed more Dll4 than this population sorted fromGfi1-null mice (Fig. 3C), andmore than LLC1 cells from culture(Supplementary Fig. S3B). Notch1 and Notch4 were expressedin LLC1 tumor-infiltrating CD11bþLy6CþLy6G� cells fromwild-type mice and Gfi1-null mice (Supplementary Fig. S3B)at somewhat lower levels than found in LLC1 cells (Supple-mentary Fig. S3B). Because Hey1 and Hey2 mRNAs were at thelimit of detection in tumor-infiltrating CD11bþLy6CþLy6G�

cells from wild-type and Gfi1-null mice (not shown), weconcluded that Hey2 mRNA detected at higher levels in LLC1tumor tissues from wild type as opposed to KOmice was likelytumor-cell derived. Collectively, these results suggested amodel in which Dll4-expressing tumor-infiltrating wild-typemyeloid cells, stimulate Notch1 and Notch4 signaling in LLC1cells inducingHey2 expression (Fig. 3D and Supplementary Fig.S3C). Supporting this model, immobilized Dll4-his specificallyinduced Hey1 and Hey2 expression in LLC1 cells, but not theexpression of TGF-b (Fig. 3E). By contrast, Dll4 did not induceHey1 and Hey2 expression in EL4 cells, which express Notch1and Notch4 at considerably lower levels than LLC1 cells(Fig. 3E).

Crosstalk between Notch and TGF-b signaling regulatescarcinoma cell growth

Next, we examined whether Notch/Hey2 signaling in LLC1cells, attributable at least in part to activation by wild-typemyeloid cell-derived Dll4, modulates TGF-b signaling andfunction in LLC1 cells. The Notch signaling inhibitors DAPTand DBZ reduced TGF-b–induced LLC1 proliferation (Fig. 4A)but minimally affected TGF-b–induced repression of EL4proliferation (Fig. 4B). These results support a functionalrequirement for Notch signaling in sustaining TGF-b–inducedLLC1 growth stimulation.

DAPT and DBZ block early steps in the Notch signalingcascade by preventing the g-secretase-dependent proteolyticcleavage ofNotch intracellular domain. To investigate points ofpotential integration of Notch and TGF-b signaling, we focusedon SMADs phosphorylation induced by TGF-b binding to type Iand type II serine–threonine kinase receptors. SMAD2 andSMAD3 receptor–regulated SMADs are recognized by type ITGF-b receptors, which are expressed by EL4 and LLC1 cells.We found that TGF-b similarly induces the phosphorylation ofSMAD2 and SMAD3 in LLC1 andEL4 cells, despite the differentbiologic outcomes (Fig. 4C). However, the Notch inhibitorDAPT reduces this phosphorylation in LLC1, but not in EL4cells (Fig. 4C), indicating that TGF-b–induced SMADs phos-phorylation is dependent, in part, upon Notch signaling inLLC1, not EL4, cells. This crosstalk of TGF-b and Notchsignaling pathways in LLC1 cells is consistent with the previ-ously recognized binding of the active Notch intracellulardomain to SMAD2/3 (14, 15). Based on experiments showingthat active Notch promotes cMyc transcription (22), whereasTGF-b inhibits cell-cycle progression by transcriptional andnontranscriptional cMyc repression inmany cell types (23, 24),we examined cMyc expression. We found that cMyc levelsincrease in LLC1 cells after TGF-b activation and that theNotch inhibitor DAPT reduces this effect, whereas cMyc levelsare unaffected by TGF-b and/or DAPT in EL4 cells (Fig. 4C).This provides additional evidence for cooperative crosstalkbetween TGF-b and Notch signaling resulting in increasedcMyc expression in LLC1 cells. Strengthening this evidence, wefound that TGF-b induces cMyc mRNA expression in LLC1cells, which is reduced by DAPT, and that TGF-b promotesexpression of the Notch target gene Hey1 (and Hey2, notshown) while minimally affecting expression of SMAD2,SMAD3 (not shown), and TbRI (Fig. 4D). TGF-b did not changeexpression levels of Hey1 or cMyc in EL4 cells (Fig. 4D).

Because Hey2 mRNA levels are 8- to 11-fold higher in LLC1tumors from wild-type mice compared with LLC1 tumors

Table 1. Differentially expressed genes in tumors from Gfi1-null and control mice

Relative expressiona Relative expression Relative expression

Gene EL4 WT KO P valueb LLC1 WT KO P value B16 WT KO P value

Tgif1 1 0.5 (0.4) 0.9 (0.8) 0.08 1 5.1 (0.2) 3.1 (0.1) 0.02 1 2.1 (0.4) 2.5 (0.2) 0.2Tgif2 1 4.5 (1.5) 0.6 (0.2) 0.02 1 2.1 (0.2) 2.2 (0.1) 0.6 1 3.0 (0.5) 3.3 (0.2) 0.6Dll4 1 13.8 (10.6) 2.3 (2.6) 0.2 1 57.2 (1.9) 5.0 (1.2) 0.001 1 1.8 (0.7) 1.7 (1.1) 0.6Hey1 1 30.2 (21.0) 12.2 (6.9) 0.2 1 15.1 (0.7) 16.4 (1.0) 0.5 1 2.4 (0.3) 2.9 (0.2) 0.3Hey2 1 2.7 (1.0) 1.3 (0.8) 0.06 1 10.7 (0.4) 1.3 (1.0) 0.05 1 2.8 (2.7) 3.1 (1.7) 0.9S100a9 nd 3.6 (0.5) 0.9 (0.4) 0.003 nd 1.1 (0.1) 44.9 (20.9) 0.001 nd 2.4 (2.5) 0.3 (0.3) 0.08Lcn2 nd 2.7 (2.2) 0.5 (0.3) 0.2 nd 1.1 (0.1) 58.9 (2.6) 0.001 nd 2.7 (4.0) 1.5 (1.8) 0.2Il4 1 1.8 (0.4) 12.3 (1.5) 0.01 nd nd nd nd nd ndIl18 1 3.9 (0.2) 12.3 (0.3) 0.01 1 1.5 (1.3) 4.7 (1.7) 0.03 1 5.7 (0.2) 5.9 (0.4) 0.8Cxcr4 1 0.22 (0.3) 0.86 (0.3) 0.003 nd nd nd 1 5.7 (0.2) 5.9 (0.4) 0.8

Abbreviation: nd, not detected.aResults from qPCR analysis of RNAs from cell lines (EL4, LLC1, B16) and tumor tissues from WT/Het or Gfi1-null (KO) mice injectedwith one of these cell lines. Data from tumors inWT/Het and KOmice are expressed as relative mRNA levels (�SD) compared with theinjected cell line when expression was detected or compared with each other when not detected in the injected cell line.bP values were determined by the Student t test; n ¼ 4 to 10 tumor tissues per group; values in bold reflect significant differences.

Ohnuki et al.





from Gfi1-null mice and to LLC1 cells from culture (Table 1),we hypothesized that this might be because of a contribu-tion by wild-type tumor-infiltrating monocytes that expressDll4 at higher levels than Gfi1-null monocytes (Fig. 3C andSupplementary Fig. S3B). We therefore investigated whetherincreased Hey2 levels in LLC1 tumors are attributable toDll4-expressing wild-type monocytes infiltrating the tumor,and whether this increased Notch signaling accelerates LLC1tumor cell growth. To mimic the monocyte effect, we firstused immobilized Dll4 to modulate Notch signaling intensityin LLC1 cells seeded at low density to reduce cell–cellcontact expected to activate Notch through membrane-anchored Dll1 and Dll4 ligands. When activated with Dll4-

his, low-density LLC1 cells responded to TGF-b withincreased proliferation, which was absent from controlcultures lacking Dll4-his (Fig. 4E). DAPT and DBZ bothreduced this proliferation (Fig. 4E), providing evidence offunctional cooperation between TGF-b and Notch signalingin promoting LLC1 proliferation. We then tested wild-typemonocytes, which express Dll4. TGF-b enhanced the prolif-eration of low-density LLC1 cells cocultured with bonemarrow CD11bþLy6CþLy6G- cells from wild-type, but notGfi1-null mice. Hey1 mRNA levels were significantly (P <0.05) higher in cocultures of LLC1 cells with wild-typemonocytes compared with cocultures of LLC1 cells withKO monocytes, confirming that wild-type monocytes

Figure 3. Tumor-infiltrating myeloidcells express Dll4, which activatesNotch signaling in LLC1 tumor cellsthat express Notch1 andNotch4. AandB,Notch receptors and ligandsexpression in tumor cell lines (A)and in CD11bþLy6CþLy6G�/CD11bþLy6CþLy6Gþ populations(B). Data from qPCR are averagesfrom triplicates measurements(variation <12%); the sortedpopulations are pools from threemice per group. C, Dll4 expressionin CD11bþGr1þ cells infiltratingLLC1 tumors. Data are averages �SD; n ¼ 3 mice/group. D, a modelfor tumor/tumor-infiltratingmyeloid cell interaction. E, Hey1and Hey2 expression in tumor cellsactivated with immobilized Dll4-his; controls, uncoated and His-coated wells. Data from qPCR areaverages� SD; n¼ 4 to 5.P valuesare from the Student t test.






activate Notch in LLC1 cells better than KO monocytes.DAPT reduced this stimulatory effect of wild-type mono-cytes (Fig. 4F). Collectively, these results show that TGF-benhances LLC1 tumor cell growth in the presence of Notchsignaling from either LLC1 cell–cell contact and/or throughinteraction with myeloid cells expressing Dll4.

Contribution of Notch signaling to carcinomaprogression

Next, we investigated whether Notch signaling contributesto increased LLC1 tumor progression. Wild-type mice bear-ing LLC1 tumors showed a significant reduction in tumorprogression when treated with the Notch inhibitor DAPT,whereas KOmice did not (Fig. 5A and Supplementary Fig. S4).Treatment with DAPT reduced Hey1 and Hey2mRNA expres-sion in the tumor tissue, indicative of Notch signaling inhi-bition in vivo (Fig. 5B). Instead, mRNA levels of SMAD2,

Figure 4. Dll4/Notch and TGF-bsignaling cooperate to promoteLLC1 cell growth. A andB, LLC1 (A)and EL4 (B) proliferation to TGF-b1with or without the inhibitors DAPTor DBZ; representative of fiveexperiments. Data are cpmaverages � SD of triplicatecultures. C, modulation of SMADsphosphorylation and cMyc proteinlevels in tumor cells by TGF-b1withor without DAPT. Representativeresults; band intensity ratiosrelative to actin are shown.D, relative gene expression intumor cells cultured with TGF-b1 (2ng/mL/48 hour) alone or with DAPT(1mmol/L). Data are averages�SD;n ¼ 4 to 5 replicates; �, P < 0.01.E, representative data(four experiments) of TGF-b1–induced LLC1 cell proliferationwithor without DAPT (1 mmol/L) or DBZ(0.1mmol/L);wells coatedwithDll4-histidine or control histidine. Dataare cpmaverages�SDof triplicatecultures. �, P < 0.01. F,representative results (threeexperiments) of TGF-b1 (0–5ng/mL)–induced LLC1 cellproliferation with or withoutadditionofwild-typeorKOCD11bþ

Ly6CþLy6G� cells (ratio 10:1);DAPT 1 mmol/L. P values are fromthe Student t test.

Ohnuki et al.





SMAD3, and TGF-b were similar in untreated and DAPT-treated tumors (Fig. 5B). Although this was not predicted bythe results of short-term experiments in vitro, we found that

TGF-bR1mRNA levels were reduced in DAPT-treated tumorsand that cMyc mRNA levels were similar in DAPT-treatedmice compared with controls (Fig. 5B). This could beexplained by observations linking Notch and its downstreammediator RPB-jk to the regulation of TGF-b receptors expres-sion (25), and the complexities of Myc transcriptional reg-ulation by many signaling pathways besides Notch and TGF-b (26). Reduced tumor growth with DAPT treatment couldnot be attributed to reduced accumulation of protumori-genic myeloid cells, because similar infiltration ofCD11bþLy6Cþ cells was present in LLC1 tumors treated ornot treated with DAPT (not shown). Overall, these resultssupport a model in which wild-type myeloid cells recruitedto the tumor accelerate tumor progression by secretingmitogenic TGF-b and enhancing TGF-b signaling in thetumor cells via expression of the ligand Dll4, which activatesNotch1 and Notch4 (Fig. 5C). We envision that the contri-bution of myeloid cells is most critical at the invasive edge ofthe tumor where myeloid cells accumulate (27), providing anopportunity for myeloid–tumor cell interactions resulting inNotch activation. Tumor cells are mobile at the tumorinvasive edge and tumor cell density is reduced (28), likelylimiting Notch signaling from tumor-intrinsic cell–cell inter-actions (Fig. 5C).

Prompted by the current findings showing a role of tumor-infiltrating monocytes in enhancing TGF-b/Notch signalingand LLC1 tumor progression, we sought evidence for thislink in clinical samples. We queried TGCA, focusing on headand neck squamous cell carcinoma, which resembles sub-cutaneous LLC1 in having significant monocyte infiltrationcorrelating with increased tumor aggressiveness andreduced survival, increased expression CCL2, deregulatedTGF-b and Notch signaling components, and evidence ofactive TGF-b signaling (29–32). We also examined lungsquamous cell carcinoma based on the lung carcinomaderivation of LLC, albeit LLC1 is a cloned line adapted toculture. Remarkably, head and neck squamous cell carcino-mas display significantly greater expression of Dll4, Notch4,Hey1, TGF-b1, and TGF-bRI, compared with normal tissue,much alike LLC1 tumors in wild-type mice (Fig. 6A). Bycontrast, lung squamous cell carcinoma display significantlylower expression of Dll4, Notch4, TGF-b1, TGF-b2, TGF-bR1,and TGF-bR2 compared with normal tissue (Fig. 6A). Thisdiscordant pattern of gene expression is strengthened bycomparing ratios of tumor/normal control from pairedsamples of individual patients (Fig. 6B). The gene expressionsignature emerging from this analysis highlights the poten-tial for activation of TGF-b/TGF-bR1 and Dll4/Notch4/Hey1signaling in head and neck squamous cell carcinoma andcontribution to tumor progression. Additional studies areplanned to measure TGF-b/TGF-bR1 and Notch signalingactivity in head and neck squamous cell tumors and corre-late signaling levels with disease outcome.

DiscussionOne of the emerging challenges to the successful treatment

of cancer is the tumor microenvironment that variously

Figure 5. Notch signaling promotes LLC1 tumor progression. A, groups oflittermate mice were treated intraperitoneally 5 days per week (7 totalinjections) with the Notch inhibitor DAPT (10 mg/kg) or vehicle alone,beginning 24 hours after inoculation of LLC1 cells. Tumorswere removed16 days after LLC1 inoculation. Individual tumor weights and averageweight� SD. B, relative gene expression in LLC1 tumors after treatmentwith vehicle alone or DAPT. Data are from individual tumors and groupaverages � SD. C, model showing the contribution of myeloid cells toNotch signaling and TGF-b function in tumor cells. Where the tumor isdense, tumor intrinsic cell-to-cell interactions aremost critical inducers ofNotch signaling: tumor cells express the Notch ligands Dll1 and Dll4 andall Notch receptors. At the infiltrating edge of the tumor, where tumordensity is low, Dll4-expressing tumor-infiltrating myeloid cells activateNotch1 and/or Notch4 signaling in the tumor cells. Notch signalingenhances TGF-b signaling and growth stimulation in the tumor cells. AllPvalues are from the Student t test.






contributes to cancer progression through reciprocal commu-nication with the tumor cells. In addressing this challenge, ourwork unveils the Notch and TGF-b signaling pathways asfunctional partners in a network of cancer cells and tumor-infiltrating CD11bþLy6CþLy6G� cells that promotes cancercell growth and tumor progression. This new understandingoffers an opportunity for the targeting of tumors where TGF-band Notch signaling are linked protumorigenic partners. Headand neck squamous cell carcinomamay represent such settingas we find that this tumor generally displays increased expres-sion of TGF-b and its receptors, in conjunction with evidenceof Notch signaling activity.

TGF-b, a product of myeloid-lineage cells in many tumormicroenvironments, plays a well-recognized role in tumorprogression and resistance to treatment (33, 34). TGF-b hascytostatic and proapoptotic effects for normal cells and

premalignant lesions (33). In advanced cancer, TGF-b pro-motes tumor progression (33). Such functional change isattributed to inactivating mutations of TGF-bR2 and SMAD4preventing TGF-b tumor-suppressive signaling in gastroin-testinal and pancreatic tumors (33, 34). In many othercancer types, including the LLC1 tumor model in our study,TGF-b receptors and SMAD signaling are intact (33). Itremains a puzzle how TGF-b can exert contextually differentfunctions. Here we show that crosstalk between the Notchand TGF-b pathways is critical to TGF-b signaling andfunction as a tumor enhancer in the LLC1 model. BlockingNotch signaling significantly reduces SMADs phosphoryla-tion and cell growth induced by TGF-b. It also slowsLLC1 tumor progression, which is accelerated by TGF-b–producing tumor-infiltrating myeloid cells.

Figure 6. Analysis of geneexpression in human head andneck squamous cell carcinomasreveals similarities to geneexpression in mouse LLC1 tumors.A, gene expression in head andneck normal tissue (n ¼ 37) andsquamous carcinoma (n ¼ 303);and in lung normal tissue (n ¼ 50)and squamous carcinoma (n ¼369). C, control tissue; T, tumor.Results from RNAseq (TGCA) areexpressed as transcript expressionlevels; data distribution isdisplayed as box plots (the boxlimits first and third quartiles; bandinside the box is median value; thevertical dotted lines indicatevariability; outlier values are shownas dots). Statistical significance ofgroup differences from the Mann–Whitney U test. B, distinct patternsof gene expression in head andneck and lung squamous cellcarcinomas. Results depict ratiosof gene expression in tumor tissueand corresponding normal tissue inpaired patient samples with headand neck (black bars; n ¼ 35) orlung (gray bars; n ¼ 50) squamouscell carcinomas.

Ohnuki et al.





TGF-b and Notch signaling converge in regulating a numberof developmental processes (35). Notch signaling modulatesexpression of BMP family members (36) and TGF-b targetgenes (37). Conversely, TGF-b induces expression of the Notchligand Jagged1 and target gene Hey1 (38). Interestingly, Notchsignal transducers can physically interact with components ofthe TGF-b signaling pathway contributing more directly toTGF-b and BMP function (14–16, 39). The present work showsthat crosstalk between the TGF-b and Notch signaling path-ways converge on SMAD2/3 activation, which sustains TGF-btumor-promoting function.Our current findings further identify a previously unrecog-

nized role of tumor-infiltrating CD11bþLy6CþLy6G� cells asactivators of Notch signaling in tumor cells enabling paracrineTGF-b growth stimulation. CD11bþLy6CþLy6G�myeloid cellsconstitute a functionally heterogeneous population (2, 40),which includes tumor-promoting myeloid-derived suppressorcells (MDSC) with T-cell suppressive functions via expressionof inducible forms of nitric oxide and arginase, and in somecases TGF-b (1, 2, 41). MDSCs have been identified in manytumor types (41), including experimental EL4 and LLC1tumors. In this study, T-cell immunity is not identified a majorcontributor to tumor growth modulation, and the tumor-promotingCD11bþLy6CþLy6G� cells do not express high-levelNos2 and Arg1 typical of MDSCs. Rater, the cells resemblephenotypically tumor-derived M2-type macrophages (2),which express the Notch ligand Dll4, similar to monocytesexposed to proinflammatory signals (42).Recent reports and the current work show that Notch

activity is a contributor to progression of some cancers (43),albeit not all cancers (44–48). It is currently unclear whetherNotch-activating signals from the tumor microenvironmentcontribute to tumor progression. This was suggested in mye-loma through tumor–stromal cells interactions (49). Usinggenetic, biochemical, and functional approaches our resultsshow that Dll4-expressing myeloid cells induce Notch stimu-lation in LLC1 tumor cells, which allows paracrine TGF-bsignaling and growth in the tumor cells. This provides evidencefor yet another mechanism of tumor progression by tumor-infiltratingmyeloid cells. Our results argue that this function oftumor-infiltrating myeloid cells is most critical at the locally

invasive edge of the tumor where single cells migration is aprincipal mode of invasion (28), and there is opportunity formyeloid–tumor cells interaction (27, 50).

In conclusion, our results provide mechanistic insights intoTGF-b tumor promotion by linkingTGF-b andNotch signaling,and disclose an activating Dll4/Notch signaling axis linkingtumor-promotingmyeloid cells and tumor cells. This raises thepossibility of dual targeting of Notch and TGF-b signaling incancers where TGF-b plays Notch-dependent protumorigenicfunctions, as in head andneck squamous cell carcinoma, whichexpresses components of the TGF-b and Notch signalingpathways at abnormally high levels. It is fortunate that inhi-bitors of TGF-b and Notch signaling (43) have been developedand hold promise in clinical development.

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

Authors' ContributionsConception and design: K. Jiang, G. TosatoDevelopment of methodology: O. Salvucci, G. TosatoAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): H. Ohnuki, K. Jiang, H. Kwak, D. Maric, G. TosatoAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): H. Ohnuki, K. Jiang, D. Wang, D. S�anchez-Martín,D. Maric, G. TosatoWriting, review, and/or revision of the manuscript: O. Salvucci, G. TosatoAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): G. TosatoStudy supervision: H. Kwak, D. Maric, G. Tosato

AcknowledgmentsThe authors thank the animal care personnel; Dr. J. Zhou, I.-I. Chen, M. Di

Prima, G. Ghilardi, and S. Park for their help; Dr. D. Lowy for research support;Dr. M. Elkin and members of the LCO for sharing reagents and insightfuldiscussions.

Grant SupportThe work was supported by the Intramural Research Program of CCR/NCI/

NIH and NINDS/NIH; H. Ohnuki was supported in part by Grant-in-Aid forScientific Research (S2207), JSPS.

The costs of publication of this article were defrayed in part by the paymentof page charges. This article must therefore be hereby marked advertisementin accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received October 29, 2013; revised January 24, 2014; accepted February 6, 2014;published OnlineFirst February 11, 2014.

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

recruited to the tumor microenvironment. Cancer Cell 2012;21:309–22.

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

3. Acharyya S, Oskarsson T, Vanharanta S, Malladi S, Kim J, Morris PG,et al. A CXCL1 paracrine network links cancer chemoresistance andmetastasis. Cell 2012;150:165–78.

4. DeNardo DG, Brennan DJ, Rexhepaj E, Ruffell B, Shiao SL, MaddenSF, et al. Leukocyte complexity predicts breast cancer survival andfunctionally regulates response to chemotherapy. Cancer Discov2011;1:54–67.

5. Ferrara N, Kerbel RS. Angiogenesis as a therapeutic target. Nature.2005;438:967–74.

6. Chen Q, Zhang XH, Massague J. Macrophage binding to receptorVCAM-1 transmits survival signals in breast cancer cells that invadethe lungs. Cancer Cell 2011;20:538–49.

7. Shojaei F, Wu X, Zhong C, Yu L, Liang XH, Yao J, et al. Nature 2007;450:825–31.

8. Mohamed MM, Sloane BF. Cysteine cathepsins: multifunctionalenzymes in cancer. Nat Rev Cancer 2006;6:764–75.

9. Akhurst RJ, HataA. Targeting the TGF-b signalling pathway in disease.Nat Rev Drug Discov 2012;11:790–811.

10. Katz LH, Li Y, Chen JS, Munoz NM, Majumdar A, Chen J, et al.Targeting TGF-b signaling in cancer. Expert Opin Ther Targets 2013;17:743–60.

11. Overall CM, Kleifeld O. Tumour microenvironment - opinion: validatingmatrix metalloproteinases as drug targets and anti-targets for cancertherapy. Nat Rev Cancer 2006;6:227–39.






12. Hock H, Hamblen MJ, Rooke HM, Traver D, Bronson RT, Cameron S,et al. Intrinsic requirement for zinc finger transcription factor Gfi-1 inneutrophil differentiation. Immunity 2003;18:109–20.

13. Karsunky H, Zeng H, Schmidt T, Zevnik B, Kluge R, Schmid KW,et al. Inflammatory reactions and severe neutropenia in mice lack-ing the transcriptional repressor Gfi1. Nat Genet 2002;30:295–300.

14. Blokzijl A, Dahlqvist C,ReissmannE, FalkA,Moliner A, LendahlU, et al.Cross-talk between the Notch and TGF-b signaling pathways medi-ated by interaction of theNotch intracellular domainwith Smad3. JCellBiol 2003;163:723–8.

15. Itoh F, Itoh S, Goumans MJ, Valdimarsdottir G, Iso T, Dotto GP, et al.Synergy and antagonism between Notch and BMP receptor signalingpathways in endothelial cells. Embo J 2004;23:541–51.

16. Moya IM, Umans L, Maas E, Pereira PN, Beets K, Francis A, et al. Stalkcell phenotype depends on integration of Notch and Smad1/5 signal-ing cascades. Dev Cell 2012;22:501–14.

17. De La Luz Sierra M, Gasperini P, McCormick PJ, Zhu J, Tosato G.Transcription factor Gfi-1 induced by G-CSF is a negative regulator ofCXCR4 in myeloid cells. Blood 2007;110:2276–85.

18. Zhu J, Jankovic D, Grinberg A, Guo L, Paul WE. Gfi-1 plays animportant role in IL-2-mediated Th2 cell expansion. Proc Natl AcadSci U S A 2006;103:18214–9.

19. Segarra M, Ohnuki H, Maric D, Salvucci O, Hou X, Kumar A, et al.Semaphorin6A regulates angiogenesis bymodulatingVEGFsignaling.Blood 2012;120:4104–15.

20. Stathopoulos GT, Psallidas I, Moustaki A, Moschos C, Kollintza A,Karabela S, et al. A central role for tumor-derived monocyte chemoat-tractant protein-1 in malignant pleural effusion. J Natl Cancer Inst2008;100:1464–76.

21. Gomez-Cambronero J, Horn J, Paul CC, Baumann MA. Granulocyte-macrophage colony-stimulating factor is a chemoattractant cytokinefor human neutrophils: involvement of the ribosomal p70 S6 kinasesignaling pathway. J Immunol 2003;171:6846–55.

22. Satoh Y, Matsumura I, Tanaka H, Ezoe S, Sugahara H, Mizuki M, et al.Roles for c-Myc in self-renewal of hematopoietic stem cells. J BiolChem 2004;279:24986–93.

23. AlexandrowMG,MosesHL. Transforming growth factor beta 1 inhibitsmouse keratinocytes late in G1 independent of effects on gene tran-scription. Cancer Res 1995;55:3928–32.

24. Chen CR, Kang Y, Siegel PM, Massague J. E2F4/5 and p107 as Smadcofactors linking the TGF-b receptor to c-myc repression. Cell2002;110:19–32.

25. Timmerman LA, Grego-Bessa J, Raya A, Bertran E, Perez-PomaresJM, Diez J, et al. Notch promotes epithelial-mesenchymal transitionduring cardiac development and oncogenic transformation. GenesDev 2004;18:99–115.

26. Dang CV. MYC on the path to cancer. Cell 2012;149:22–35.27. Wels J, Kaplan RN, Rafii S, Lyden D. Migratory neighbors and distant

invaders: tumor-associated niche cells. Genes Dev 2008;22:559–74.28. Friedl P, Wolf K. Plasticity of cell migration: a multiscale tuning model.

J Cell Biol 2010;188:11–9.29. Leethanakul C, Patel V, Gillespie J, Pallente M, Ensley JF, Koontong-

kaew S, et al. Distinct pattern of expression of differentiation andgrowth-related genes in squamous cell carcinomas of the head andneck revealed by the use of laser capture microdissection and cDNAarrays. Oncogene 2000;19:3220–4.

30. Marcus B, Arenberg D, Lee J, Kleer C, Chepeha DB, Schmalbach CE,et al. Prognostic factors in oral cavity and oropharyngeal squamouscell carcinoma. Cancer 2004;101:2779–87.

31. Rothenberg SM, Ellisen LW. The molecular pathogenesis of head andneck squamous cell carcinoma. J Clin Invest 2012;122:1951–7.

32. Xie W, Aisner S, Baredes S, Sreepada G, Shah R, Reiss M. Alterationsof Smad expression and activation in defining 2 subtypes of humanhead and neck squamous cell carcinoma. Head Neck 2013;35:76–85.

33. Massague J. TGF-b signalling in context. Nat Rev Mol Cell Biol2012;13:616–30.

34. Yang L, Pang Y, Moses HL. TGF-b and immune cells: an importantregulatory axis in the tumor microenvironment and progression.Trends Immunol 2010;31:220–7.

35. Yang J,Weinberg RA. Epithelial-mesenchymal transition: at the cross-roads of development and tumor metastasis. Dev Cell 2008;14:818–29.

36. Cornell RA, Eisen JS.Notch in thepathway: the roles ofNotch signalingin neural crest development. Semin Cell Dev Biol 2005;16:663–72.

37. Sjolund J, BostromAK, LindgrenD,Manna S,Moustakas A, LjungbergB, et al. The notch and TGF-b signaling pathways contribute to theaggressiveness of clear cell renal cell carcinoma. PLoS One 2011;6:e23057.

38. Zavadil J, Cermak L, Soto-Nieves N, Bottinger EP. Integration of TGF-beta/Smad and Jagged1/Notch signalling in epithelial-to-mesenchy-mal transition. Embo J 2004;23:1155–65.

39. Dahlqvist C, Blokzijl A, Chapman G, Falk A, Dannaeus K, Ibanez CF,et al. Functional Notch signaling is required for BMP4-induced inhi-bition of myogenic differentiation. Development 2003;130:6089–99.

40. Qian BZ, Pollard JW. Macrophage diversity enhances tumor progres-sion and metastasis. Cell 2010;141:39–51.

41. Gabrilovich DI, Ostrand-Rosenberg S, Bronte V. Coordinated regula-tion of myeloid cells by tumours. Nat Rev Immunol 2012;12:253–68.

42. Fung E, Tang SM, Canner JP, Morishige K, Arboleda-Velasquez JF,Cardoso AA, et al. Delta-like 4 induces notch signaling in macro-phages: implications for inflammation. Circulation 2007;115:2948–56.

43. Espinoza I, Miele L. Notch inhibitors for cancer treatment. PharmacolTher 2013;139:95–110.

44. Huang Y, Lin L, Shanker A, Malhotra A, Yang L, Dikov MM, et al.Resuscitating cancer immunosurveillance: selective stimulation ofDLL1-Notch signaling in T cells rescues T-cell function and inhibitstumor growth. Cancer Res 2011;71:6122–31.

45. Sriuranpong V, BorgesMW, Ravi RK, Arnold DR, Nelkin BD, Baylin SB,et al. Notch signaling induces cell cycle arrest in small cell lung cancercells. Cancer Res 2001;61:3200–5.

46. Weng AP, Ferrando AA, Lee W, Morris JPt, Silverman LB, Sanchez-Irizarry C, et al. Activating mutations of NOTCH1 in human T cell acutelymphoblastic leukemia. Science 2004;306:269–71.

47. Uyttendaele H, Marazzi G, Wu G, Yan Q, Sassoon D, Kitajewski J.Notch4/int-3, a mammary proto-oncogene, is an endothelial cell-specific mammalian Notch gene. Development 1996;122:2251–9.

48. Xu K, Usary J, Kousis PC, Prat A, Wang DY, Adams JR, et al. Lunaticfringe deficiency cooperates with the Met/Caveolin gene amplicon toinduce basal-like breast cancer. Cancer Cell 2012;21:626–41.

49. Jundt F, Probsting KS, Anagnostopoulos I, Muehlinghaus G, Chatter-jee M, Mathas S, et al. Jagged1-induced Notch signaling drivesproliferation of multiple myeloma cells. Blood 2004;103:3511–5.

50. Ruffell B, Affara NI, Coussens LM. Differential macrophage program-ming in the tumormicroenvironment. Trends Immunol 2012;33:119–26.


Ohnuki et al.




Published OnlineFirst February 11, 2014.Cancer Res Hidetaka Ohnuki, Kan Jiang, Dunrui Wang, et al. Signaling to Drive Malignant Progression

βTumor-Infiltrating Myeloid Cells Activate Dll4/Notch/TGF-

Updated version

10.1158/0008-5472.CAN-13-3118doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2014/02/12/0008-5472.CAN-13-3118.DC1

Access the most recent supplemental material at:

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

Subscriptions

Reprints and

[email protected] at

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

Permissions

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

.http://cancerres.aacrjournals.org/content/early/2014/03/19/0008-5472.CAN-13-3118To 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-13-3118

http://cancerres.aacrjournals.org/content/suppl/2014/02/12/0008-5472.CAN-13-3118.DC1

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/early/2014/03/19/0008-5472.CAN-13-3118


Tumor-Inﬁltrating Myeloid Cells Activate...

Documents

Transcript of Tumor-Inﬁltrating Myeloid Cells Activate...