The Coordinated Actions of TIM-3 on Cancer and Myeloid ...Research Article The Coordinated Actions...

Research Article

The Coordinated Actions of TIM-3 on Cancer andMyeloid Cells in the Regulation of Tumorigenicityand Clinical Prognosis in Clear Cell Renal CellCarcinomasYoshihiro Komohara1, Tomoko Morita2, Dorcas A. Annan2, Hasita Horlad1, Koji Ohnishi1,Sohsuke Yamada3,Toshiyuki Nakayama3, Shohei Kitada4, Shinya Suzu5, Ichiro Kinoshita6,Hirotoshi Dosaka-Akita6, Koichi Akashi7, Motohiro Takeya1, and Masahisa Jinushi2,8

Abstract

Clear cell renal cell carcinoma (ccRCC) is one ofmost commoncancers in urogenital organs. Although recent experimental andclinical studies have shown the immunogenic properties of ccRCCas illustrated by the clinical sensitivities to various immunothera-pies, the detailed immunoregulatory machineries governing thetumorigenicity of human ccRCC remain largely obscure. In thisstudy, we demonstrated the clinical significance and functionalrelevance of T-cell immunoglobulin andmucin domain-contain-ing molecule-3 (TIM-3) expressed on tumor cells and myeloidcells in patients with ccRCC. TIM-3 expression was detected oncancer cells and CD204þ tumor-associated macrophages (TAM),and higher expression level of TIM-3 was positively correlatedwith shorter progression-free survival (PFS) in patients withccRCC. We found that TIM-3 expression was detected on a large

number of tumors, and there was significant correlation betweenan increased number of TAMs and high expression level of TIM-3in patients with ccRCC. Furthermore, TIM-3 rendered RCC cellswith the ability to induce resistance to sunitinib and mTORinhibitors, the standard regimen for patients with ccRCC, as wellas stem cell activities. TIM-3 expression was induced on CD14þ

monocytes upon long-term stimulation with RCC cells, andTIM-3–expressing myeloid cells play a critical role in augment-ing tumorigenic activities of TIM-3-negative RCC cells. Moreimportantly, treatment with anti–TIM-3 mAb suppressed itstumorigenic effects in in vitro and in vivo settings. These findingsindicate the coordinated action of TIM-3 in cancer cells and inmyeloid cells regulates the tumorigenicity of human RCC.Cancer Immunol Res; 3(9); 999–1007. �2015 AACR.

IntroductionKidney cancer is the fifteenth most common cancer in the

world, and the global incidence rate is 4 cases per 100,000persons. The incidence rate is significantly higher in the NorthAmerican, Australian, and European regions (1). Clear cell renalcell carcinoma (ccRCC) is the most common histologic type inkidney cancers. Overall median progression-free survival (PFS) is

12 months in ccRCC and 17 months in non–clear cell RCCs (2).Althoughnearly 80%of ccRCC cases are considered to be cured byresection, the median overall survival (OS) for patients withmetastatic RCC is less than 3 years (3). Therefore, suitablemarkersfor predicting outcomes are necessary to guide clinical therapeuticmanagement.

T-cell immunoglobulin (Ig) and mucin domain-containingmolecule-3 (TIM-3), also known as hepatitis A virus cellularreceptor 2, is widely expressed on immune cells, such asmonocytes/macrophages, dendritic cells, natural killer cells,and T cells (4, 5). Signaling via TIM-3 is generally involved inthe regulation of immune responses via negatively regulatingT-helper type 1–cell viability and interferon secretion (3, 6–8). Recently, TIM-3 expression has also been found on mel-anoma, liver cancer, and lung cancer cells. In these cancers,higher TIM-3 expression correlated with poor clinical prog-nosis (9–11).

In this study, we demonstrated that, in patients with ccRCC,TIM-3 is frequently expressed on tumor tissues and higherTIM-3 expression levels are significantly associated withshorter PFS. Furthermore, TIM-3 on tumor cells and myeloidcells coordinately contributed to the tumorigenic activities ofRCCs. These findings suggest that TIM-3 may serve as a usefulbiomarker for predicting prognosis and a potential therapeutictarget for improving therapeutic responses in patients withccRCC.

1Department of Cell Pathology, Graduate School of Medical Sciences,Kumamoto University, Kumamoto, Japan. 2Research Center for Infec-tion-Associated Cancer, Institute for Genetic Medicine, Hokkaido Uni-versity, Sapporo, Japan. 3Department of Pathology and Cell Biology,School of Medicine, University of Occupational and EnvironmentalHealth, Kitakyusyu, Japan. 4Department of Urology, School of Med-icine, University of Occupational and Environmental Health, Kitakyu-syu, Japan. 5Center for AIDS Research, Kumamoto University, Kuma-moto, Japan. 6Department of Medical Oncology, Hokkaido UniversitySchool of Medicine, Sapporo, Japan. 7Department of Medicine andBiosystemic Sciences, Kyushu University Graduate School of Medi-cine, Fukuoka, Japan. 8Institute for Advanced Medical Research, KeioUniversity Graduate School of Medicine, Tokyo, Japan.

Corresponding Author: Masahisa Jinushi, Institute for Advanced MedicalResearch, Keio University Graduate School of Medicine, 35 Shinanomachi,Shinjuku-ku, Tokyo, 160-8582, Japan. Phone: 81-3-5363-3778; Fax: 81-3-5362-9259; E-mail: [email protected]

doi: 10.1158/2326-6066.CIR-14-0156

�2015 American Association for Cancer Research.

CancerImmunologyResearch

www.aacrjournals.org 999

on February 18, 2020. © 2015 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from

Published OnlineFirst March 17, 2015; DOI: 10.1158/2326-6066.CIR-14-0156

http://cancerimmunolres.aacrjournals.org/

Materials and MethodsTissue samples

In total, 91 paraffin-embedded tissue samples (not tissuearray), derived from patients diagnosed with ccRCC, who hadundergone curative surgery between 1998 and 2008 at UniversityHospital of Occupational and Environmental Health, and whosesamples were pathologically confirmed as ccRCC, were selectedfor this study (12). Cases with massive necrosis were not selected.Data for PFS and cancer-specific OS were obtained from themedical records of these patients. All samples were obtained withwritten informed consent from patients in accordance with pro-tocols approved by each university review board. Tissue sampleswere fixed in 10% neutral buffered formalin and were embeddedin paraffin per routine methods.

ImmunohistochemistryA goat-polyclonal antibody against TIM-3 was purchased from

LifeSpan BioSciences. Mouse monoclonal anti-CD204 antibody(SRA-E5; Transgenic) was used for detecting tumor-associatedmacrophages (TAM; ref. 13). For TIM-3 immunostaining, CanGet Signal (Toyobo) was used as an antibody diluent. Secondaryantibodies were purchased from Nichirei, and reactions werevisualized using a diaminobenzidine substrate system (Nichirei).Two investigators, who were blinded to any information aboutthe samples, evaluated the infiltration of CD204þ cells and TIM-3þ cancer cells. CD204þ cells in six randomly selected areas werecounted by these two investigators, and the averages of the resultswere used.

Cell culturesCD14þmonocytes were isolated fromperipheral bloodmono-

nuclear cells (PBMC) obtained from healthy volunteers usingCD14 microbeads (Miltenyi Biotec). The blood mononuclearcells from patients or healthy donors were obtained with writteninformed consent from healthy volunteers in accordance withprotocols approved by each university review board. Prior tococulture experiments, monocytes were labeled with PKH26fluorescence (Sigma). Three human RCC cell lines (ACHN,786-O, and Caki-1) were obtained from the American TypeCulture Collection (ATCC). CD14þ monocytes and human RCCcells were cocultured in serum-free media supplemented withhuman EGF and b-FGF on the ultra-low attachment plates. In thiscondition, the tumor cell growth kinetics was low while the cellviability was preserved, and the ratio of tumor cells and mono-cytes remained constant during the experimental periods. Myco-plasma tests were performed using a PCR detection kit (Takara BioInc.).

Construction of human TIM-3 gene plasmidThe human TIM-3 mRNA (accession number NM_032782.4)

was isolated from TIM-3þ tumor-infiltrating monocytes usingprimers for full-length amplification (Fw, 50-TCAGATCTCGAG-CTCATGTTTTCACATCTTCCCTTTGACTG-30; Rv, 50-CGGTGGA-TCCCGTGGCATTGCAAAGCGACAACC-30), and inserted intopPRIME-dsRed plasmid. The plasmids were transfected intohuman ccRCC cells with Lipofectamine 3000 for 48 hours accord-ing to themanufacturer's instructions (Life Technologies). TIM-3–expressing cells were selected with G418 (400 mg/mL) for 14 to 21days. After selection, the mRNA levels of human TIM-3 were

approximately 250 times higher in the transfected cells (Fig.3C). The empty pPRIME-dsRed plasmid was used as a control.

Apoptosis assayACHN cells were transfected with human TIM-3 or control

gene-introduced plasmids, and cells were then treated with suni-tinib (50 nmol/L) or rapamycin (50 nmol/L) for 48 hours.Caspase-3 activity in the tumor cell lysates was quantified witha colorimetric assay kit according to the manufacturer's instruc-tions (Invitrogen). In all indicated experiments, monoclonalantibodies (mAb) recognizing human TIM-3 (Clone F38–2E2:10 mg/mL) were used as described previously (14).

Sphere formation assayACHN cells were transfected with human TIM-3 or control

genes, or stimulated with TIM-3þ or TIM-3� macrophages for 7days. The cells were then cultured in ultra-low attachment culturedishes (Corning) inDMEM/F-12 serum-freemedium supplemen-ted with 20 ng/mL epithelial growth factor and 10 ng/mL basicfibroblast growth factor-2 (PeproTech). Cell digestions with tryp-sin and cell passageswere performed every 3days, and the size andnumber of spheres were counted under the microscope.

Quantitative RT-PCRmRNA was isolated from TIM-3þ or TIM-3� ACHN cells, and

the genes associated with cancer stem cell properties (Twist1,Snail, and Pou5F) were quantified by real-time PCR using SYBRGreen Gene Expression Assays (Applied Biosystems).

In vivo tumor initiation assayNSG mice were purchased from The Jackson Laboratory. All

experiments were conducted under a protocol approved by theanimal care committees of Hokkaido University (Sapporo,Japan). For analysis of ACHN tumor-initiating activities, ACHNcells were injected s.c. into NSGmice (ranged from 1� 104 to 1�105 cells/mouse), and tumor growth was measured once a weekthrough the entire experimental period.

Flow cytometryFor in vitro analysis, untreated ACHN cells or those stimulated

with PKH26-labeled TIM-3þ or TIM-3� monocytes were stainedwith anti-BrdUrd mAbs or anti–TIM-3 mAbs (BioLegend).BrdUrd labeling was performed using the FITC BrdUrd Flow Kitaccording to the manufacturer's instructions (BD Biosciences).For in vivo assays, EpCAMþ tumor cells or CD68þ tumor-infil-tratingmacrophageswere analyzed byflow cytometry usingmAbsspecifically recognizing humanEpCAM,CD68,CD163, andMHCclass II (BioLegend). Cell acquisition and analysis were performedwith a FACS-Canto (Becton Dickinson).

Myeloid cell–mediated in vivo tumorigenesisFor in vivo tumor–myeloid cell interaction analysis, ACHN cells

were injected s.c. into NSGmice (1� 105/mouse) in conjunctionwith i.v. administration of CD14þ macrophages obtained fromcancer patients (1 � 106/mouse). The blood mononuclear cellsfrom patients were obtained with written informed consent inaccordance with protocols approved by each university reviewboard. Recombinant M-CSF proteins (500 mg/mL) were alsoadministered i.p. to support the survival of transferred humanmonocytes in vivo, as shown by a previous report (15). Additional

Komohara et al.

Cancer Immunol Res; 3(9) September 2015 Cancer Immunology Research1000




micewere treatedwith i.p. injections of control Ig or the anti–TIM-3 mAb. Tumor growth was measured on the indicated days, andthe number of human macrophages in the tumors was evaluatedfor each mouse.

Statistical analysisStatistical analyseswere carried out using JMP10 (SAS Institute)

and StatMate III (ATOMS). The Kruskal–Wallis test, the Kaplan–Meier method, and the Cox hazard test were used to analyze theclinical course associations. The Student t test was used for two-group comparisons in in vitro and in vivo studies, and data areexpressed as means � SD. A value of P < 0.05 was consideredstatistically significant.

ResultsHigher expression of TIM-3 in cancer cells is associated withshorter PFS

Immunostaining of ccRCC tissues revealed TIM-3 expressionon both immune and cancer cells (Fig. 1A). TIM-3 was mainlyexpressed on cell surface membranes in cancer cells, and TIM-3–positive cancer cells were detected in 63 of 91 cases. In noncancertissues, TIM-3 was detected in a subset of immune cells (Fig. 1A)and renal tubules (Fig. 1B). As shown in Fig. 1A, the stainingintensity of TIM-3 varied from case to case, and therefore thestaining intensities for cancer cells were classified into threegroups (score 0, negative or weak; score 1, intermediate; score

2, strong). The correlations between TIM-3 expression and clin-icopathologic factors were then analyzed. As a result of thisanalysis, we found that the presence of TIM-3–positive cancercells was preferentially detected in cases with higher clinicalT stage and nuclear grade (Table 1), and was significantly asso-ciated with shorter PFS, but not with OS (Fig. 1C and D). Thepatients with strong (score 2) TIM-3 expression in cancercells showed the shortest PFS (Fig. 1C). Statistical analysis wasalsoperformed inpatientswith lower clinical stage (T1) and lowernuclear grade (grade 1 and 2), and the presence of TIM-3–positivecancer cells was significantly associated with shorter PFS in bothgroups (Fig. 1E).

Increased TAM infiltration in cases with higher TIM-3expression

Because we had previously shown that CD204þ TAMs areintimately involved in cancer cell activation (16), we hypothe-sized that TAM-derived factors induce TIM-3 expression in cancercells. To test this hypothesis, serial sections were stained usinganti-CD204 antibody, and the correlation of CD204 stainingwithTIM-3 expression was analyzed. The number of CD204þ TAMswas increased in TIM-3–expressing cases (Fig. 2A), and the num-ber of TAMswas significantly correlatedwith the staining intensityof TIM-3 (Fig. 2B). The number of TAMs was associated with aworse OS; however, the correlation with PFS was not statisticallysignificant upon analysis (Fig. 2C and Table 2). Although aquantitative evaluation of TIM-3þ TAMs was too difficult to

Figure 1.Immunohistochemical determination of TIM-3 expression in ccRCC. A, the intensity of TIM-3 immunohistochemical staining was classified into three groups(score 0, 1, and 2). TIM-3þ immune cells are indicated as arrowheads. B, TIM-3 expression was weakly detected in renal tubules in noncancer kidney tissue. TheKaplan–Meier analysis of PFS (C) and OS (D) were performed to investigate the correlations between TIM-3 expression and clinical prognosis. E, the Kaplan–Meieranalysis was performed in patients with lower stage (T1) or lower nuclear grade (grade 1 and 2).

Tumor and Myeloid Cell TIM-3 in RCC

www.aacrjournals.org Cancer Immunol Res; 3(9) September 2015 1001




perform in tissue sections of patients with high expression level(score 3) of TIM-3, we observed that some TAMs also expressedTIM-3 in tissue sections of no or low expression level (score 0 and1)of TIM-3 (Fig. 2D). TIM-3þTAMswere counted, and the averagepercentage of TIM-3þ TAMs among CD204þ TAMs was found tobe 15.7%. Then, patients were divided into two groups, andstatistical analysis demonstrated that patients with higher num-ber of CD204þ TAMs and higher percentage of TIM-3þ TAMs hadshorter PFS (Fig. 2E).

TIM-3 on RCC cells contributes to anticancer drug resistanceand tumorigenic activities

To determine the effect of tumor-derived TIM-3 in thismodel, we transfected human TIM-3 gene into cells of thehuman RCC line ACHN, because no TIM-3 expression wasdetected on any of the human RCC cell lines used in our study(data not shown). Although TIM-3 did not significantly influ-ence the growth of ACHN cells in vitro (data not shown), TIM-3–expressing ACHN cells manifested increased resistance tosunitinib and the mTOR inhibitor rapamycin (Fig. 3A). Suni-tinib and mTOR inhibitors have been approved as standardanticancer regimens against human RCC (17, 18). Indeed, TIM-3þ ACHN cells have superior self-renewal activities comparedwith control ACHN cells, as shown by increased sphere-form-ing activities (Fig. 3B). We next examined whether TIM-3confers cancer-stem cell characteristics to RCC cells. The mRNAlevels of genes that have been reported to be associated withrenal cancer-stem cell properties, such as Twist1, Snail, andPau5F1 (19), were compatible between TIM-3þ and TM-3�

ACHN cells (Fig. 3D). However, tumor formation was observedwhen TIM-3þ ACHN cells were s.c. inoculated into NSG mice atthe smaller numbers (104 cells per mouse), whereas TIM-3�

ACHN tumor cells needed larger numbers (105 cells per mouse)to initiate tumor formation (Fig. 3E). These results suggest thatTIM-3 promotes tumorigenicity of RCC cells by inducing can-cer-stem cell properties.

Log-rank: P = 0.025Wilcoxon: P = 0.006

TIM-3 score 1,2 (n = 79)

Months

>=PF

S

TIM-3 CD204

Log-rank: P = 0.024Wilcoxon: P = 0.028

Log-rank: P > 0.05Wilcoxon: P > 0.05

0 20 40 60 80 100 120

1.0

0.8

0.6

0.4

0.2

0.0

1.0

0.8

0.6

0.4

0.2

0.0

1.0

0.8

0.6

0.4

0.2

0.0

CD204 high(n = 46)

CD204 low(n = 45)

CD204 high(n = 46)

CD204 low(n = 45)

P = 0.047

The

num

ber

of C

D20

4+

mac

roph

ages

/mm

2 600

500

400

300

200

100

050 μm

50 μm

A B

D E

C

Pt. 8 (TIM-3 score 0)

CD204 CD204

TIM-3 → Score 0 Score 1 Score 2

Pt. 4 (TIM-3 score 2)

. CD204 high (TIM-3+ TAM 15%) (n = 20)

. CD204 high (TIM-3+ TAM <15%)

. CD204 low (n = 59)

PF

S

OS

Months0 20 40 60 80 100 120

0 20 40 60 80 100 120

Months

Figure 2.Immunohistochemical determinationof CD204 expression. A,immunostaining of CD204 wasperformed to detect TAMs. B, thecorrelation between the number ofTAMs and TIM-3 expression wasanalyzed using the Kruskal–Wallis test.C, the Kaplan–Meier analyses of PFSand OS were performed. D, TIM-3expression was observed on TAMs. E,the Kaplan–Meier analysis of PFS wasperformed to investigate thecorrelations between TIM-3þ

TAMs and clinical prognosis.

Table 1. TIM-3 expression in cancer cells and clinicopathologic parameters

TIM-3 expression CD204/mm2

n Score 0 Score 1 Score 2 P Mean P

Age, y<60 38 9 24 5 >0.05 392 >0.05�60 53 18 28 7 384

GenderM 59 17 37 5 >0.05 399 >0.05F 32 10 15 7 364

T classificationT1 42 20 19 3 <0.001 389 >0.05T2�T4 49 7 33 9 387

Nuclear gradeG1, G2 74 26 42 6 <0.001 373 >0.05G3, G4 17 1 10 6 454

NOTE: Correlation with TIM-3 was tested using a cumulative x2 test. Correlationwith the number of CD204þ cells was tested using the Mann–Whitney U test.Underline, statistically significant.

Komohara et al.





TIM-3þ myeloid cells contribute to anticancer drug resistanceand tumorigenic activities of RCC cells

Recent studies have revealed that TIM-3 on myeloid cellsmediates immunoregulatory functions, leading to impairedantitumor immunosurveillance (20). Because we show thatTIM-3 was also expressed on TAMs in ccRCC in this study

(Fig. 2C), we next investigated the contribution of TIM-3þ

TAMs in coculture experiments and in a mouse model. ACHNcells and PKH26-labeled monocytes were directly coculturedfor 2 weeks, and TIM-3 expression was examined by FACS(Fig. 4A). After coculture, TIM-3 expression was observed onCD14þ monocytes, although TIM-3 was not detected on the

0

100

200

300

400

500

600

700

DMSO Sunitinib Rapamycin

TIM3–

TIM3+

A

Act

ive

casp

ase-

3 (n

g/m

L)

B

*

*

D

C

*

ns

Fol

d in

duct

ion

(/β-

actin

)

0

0.5

1

1.5

2

2.5

Snail

0

0.5

1

1.5

2

2.5

Pou5F1

Control

TIM-3

Contro

l

TIM-3

Contro

l

TIM-3

Contro

l

TIM-3

TIM-3

–

TIM-3

–

TIM-3

+

TIM-3

+

Contro

l

TIM-3

0

0.5

1

1.5

2

2.5Twist1

nsns

TIM-3– ACHN

TIM-3+ ACHN

0.0

5.0

10.0

15.0

20.0

25.0

0 1 2 3 4 5 6

Tum

or a

rea

(mm

2 )

Week

1×104 cells

0.0

10.0

20.0

30.0

40.0

50.0

0 1 2 3 4 5 6

Tum

or a

rea

(mm

2 )

Week

1×105 cells

0

50

100

150

200

250

TIM-3

TIM-3– ACHN

TIM-3+ ACHN

ACHNTIM-3– monocyte

ACHNTIM-3+ monocyte

ACHN

10 μm

0

40

80

120

160

200

Sph

ere

form

ing

cells

(/1

03)

*

Figure 3.Increased tumorigenic activities of TIM-3–expressing RCC cells. A, ACHN cells transfected with human TIM-3 or control gene plasmid were treated with 50 nmol/Lof sunitinib or rapamycin for 6 hours and then assayed for cell death by measuring active caspase-3. Similar results were observed in three experiments. B, bulkACHN cells, TIM-3þ or TIM-3� ACHN cells, or those stimulated with TIM-3þ or TIM-3� CD14þ monocytes isolated from patients' PBMCs were then cultured withthree passages in ultra-lowattachment plates. The size (left) and numbers (right) of formed spheres generated per 100,000 cellsweredetermined. C,mRNAexpressionof TIM-3, Twist1, Snail, and Pou5F in TIM-3þ or TIM-3� ACHN cells were quantified by RT-PCR. D, TIM-3þ or TIM-3� ACHN cells (104

–105 per mouse) were inoculated s.c.into NSG mice, and tumor growth was measured. Similar results were obtained in three independent experiments. � , P < 0.05. ns, not statistically significant.

Table 2. Univariate Cox regression analysis of PFS and OS

PFS OSPatients (n) HR (95% CI) P HR (95% CI) P

Age, <60 vs. 60 38/53 1.3 (0.6–2.9) 0.48 0.9 (0.4–2.7) 0.99Gender, M vs. F 59/32 1.3 (0.6–3.0) 0.49 2.3 (0.2–0.7) 0.28Stage, T1 vs. T2þ3þ4 42/49 3.5 (1.6–7.8) 0.001 2 (0.8–5.4) 0.16Nuclear grade, G1þ2 vs. G3þ4 74/17 5.5 (2.5–11.8) <0.001 3.9 (1.5–9.9) 0.007TIM-3, score 0 vs. score 1þ2 27/64 6.1 (1.8–37.7) 0.002 3.7 (0.7–68) 0.12CD204, <300 vs. 300 45/46 1.6 (0.8–3.6) 0.18 3.1 (1.1–11.0) 0.029

Abbreviations: CI, confidence interval; HR, hazard ratio. Underline, statistically significant.






human RCC cell lines examined in this study (ACHN, Caki-1,MAMIYA, and 786-0; Fig. 4B; and data not shown). Next, weisolated TIM-3þ and TIM-3� monocytes from tumor–monocytemixtures to evaluate the significance of TIM-3 in cell–cellinteractions. We investigated whether TIM-3 on monocytesinfluences the self-renewal properties of RCC cells, and foundthat the sphere-forming capacity of ACHN cells was augmentedby coculture with TIM-3þ monocytes (Fig. 3B). Next, TIM-3þ orTIM-3� monocytes were cocultured with ACHN cells for 72hours. TIM-3þ monocytes rendered ACHN cells with the abilityto promote proliferation and resist apoptotic cell death medi-ated by sunitinib and rapamycin compared with ACHN cellsalone or those cocultured cells with TIM-3� monocytes (Fig. 4Cand D). Moreover, the treatment with anti–TIM-3 mAb resultedin the educed proliferative and chemoresistant phenotypes ofACHN cells specifically mediated by TIM-3þ monocytes (Fig.4C and D). Together, these findings highlight the significantinvolvement of TIM-3 expression on myeloid cells in humanRCC tumorigenicity.

TIM-3 serves as a therapeutic target for suppressing RCC growthin a myeloid cell–dependent manner

Finally, we evaluated the impact of TIM-3 on in vivo tumori-genicity. For this purpose, ACHN cells were injected s.c. into NSGmice in the presence of control Ig or anti-humanTIM-3mAb (F38-2E2), and in vivo tumor formation was evaluated at the indicatedtimes. To examine the involvement of humanmyeloid cells in theregulation of TIM-3–mediated antitumor responses, NSG micewere treated with human M-CSF (500 mg/kg/d) twice per week toimprove human myeloid cell reconstitution. During this proce-dure, adoptive transfer of human-derived monocytes (1 � 106

/mouse) was performed via i.v. injection (Fig. 5A). Althoughcotransfer of monocytes had little impact on the tumor growthcompared with nontransfer controls, tumor growth was signifi-cantly suppressed by treatment with anti–TIM-3 mAb in themonocyte-transferred groups (Fig. 5B). We detected the expres-sion of human TIM-3 in tumor-infiltrating CD68þ myeloid cells,but not in RCC cells (Fig. 5C and D). TIM-3þ cells are observed intumor tissues in the monocyte-transferred groups, but not in the

Act

ive

casp

ase-

3 (n

g/m

L)

B

D

0.0

99.8

0.0

0.1

0.0

99.8

0.0

0.1

0.1

82.1

0.2

17.6

1.1

80.2

2.4

16.3

PKH26 TIM-3

Tumor + monocytes

Tumor

1st week 2nd weekAMonocytes PKH26-labeled monocytes

Evaluate TIM-3 expression by flow cytometry

ACHN cells(PKH26–)

Monocytes(PKH26+)

C

0

100

200

300

400

500

600

Control αTIM-3 Control αTIM-3 Control αTIM-3

DMSO

Rapamycin

Sunitinib

TIM-3– monocyte TIM-3+ monocyte

0

5

10

15

20

25

30

TIM-3+ monocyteTIM-3– monocyte(–)

ControlαTIM-3

ns ns

*

Brd

Urd

+ A

CH

N c

ells

(%

)

**

Figure 4.Increased tumorigenic activities of RCC cells stimulated with TIM-3–expressing myeloid cells. A, schema of the experimental design. B, PKH26-labeled CD14þ

monocytes isolated from PBMCs of patients (5� 105) were cocultured with unlabeled ACHN cells (1� 105) for 14 days. TIM-3 expression onmonocytes (PKH26þ) orACHN cells (PKH26�) was evaluated by flow cytometry. C, ACHN cells were cultured with TIM-3þ or TIM-3� monocytes, and treated with anti–TIM-3 mAb orcontrol Ig (10 mg/mL) for 7 days. Cell proliferation was then analyzed by quantifying BrdUrd uptake with flow cytometry. D, ACHN cells cocultured with CD14þ

monocytes as described above were treated with 50 nmol/L of sunitinib or rapamycin for 6 hours and then assayed for cell death by measuring activecaspase-3. Similar results were obtained in three independent experiments. � , P < 0.05. ns, not statistically significant.

Komohara et al.





control groups (Fig. 5D). The numbers of TIM-3þ myeloid cellswere not changed by treatment with anti–TIM-3 mAb (Fig. 5D).Interestingly, treatment with anti–TIM-3mAb decreased the infil-tration of CD163-expressing M2 macrophages in the tumortissues of the monocyte-transferred group, suggesting that TIM-3may support thedifferentiation of protumormyeloid cells in thetumor microenvironments (Fig. 5E). Overall, these findings pro-vide clear evidence that TIM-3–mediated regulation of cancercell–myeloid cell interactions serves as a critical pathway support-ing RCC tumorigenicity.

DiscussionAccumulating evidence has revealed that TIM-3 expressed on

tumor cells and tumor-associated immune cells has diversetumorigenic activities mediated through recognition of multipleligands, such as galectin-9, high-mobility group box 1 (HMGB1),and phosphatidylserine (6, 7, 21). Galectin-9 recognizes TIM-3expressed on exhausted CD8þ T lymphocytes and triggers apo-

ptosis by antagonizing human leukocyte antigen B (HLA-B)–associated transcript 3 (Bat3)–mediated survival signals. HMGB1binds TIM-3 on dendritic cells and suppresses innate immunesignals mediated by nucleic acid pattern recognition receptors(PRR; refs. 7, 22). Furthermore, TIM-3 is detected on varioustumor cells and tumor-initiating cells and contributes to tumor-initiating and tumor-promoting activities (23, 24). Thus, TIM-3has pleiotropic functions that influence multiple immunologicand biologic properties of various types of cells.

Importantly, TIM-3 expressed on tumor cells and tumor-infil-trating immune cells may serve as a key sentinel linking impairedtumor immunosurveillance with amplified tumorigenicity in thetumor microenvironment. Although the ligand recognized spe-cifically by tumor cells, including RCC cells, remains to beidentified, recent studies suggest that TIM-3 activates the NF-kBpathway via phosphorylation of cytoplasmic tyrosine kinasemotifs in murine B16 melanoma cells (25). Moreover, NF-kBfunctions as a critical hub governingmultiplemodes of oncogenicprocesses, including antiapoptotic pathways, inflammatory

0

50

100

150

200

250

A

B

C

M-CSF

Isotype or

αTIM-3 Ab

Transfer of human CD14+ monocytes

0

10

20

30

40

50

Tum

or v

olum

e (m

m3 )

*

NSG mice

E

NS

50 μm

Control abMonocytes⇒Control ab

TIM

-3+ c

ells

/mm

2

ns

Monocytes⇒αTIM-3

2.62

TIM-3

CD

68

SSC

CD

68

F

ND ND

0

2

4

6

8

10

Total (CD68+) M2 (CD68+CD163+)

Control abαTIM-3ns

*

TAM

num

ber

(%)

D

M 3 AA

Contro

l ab

Mon

ocyte

s

⇒Con

trol a

b

Mon

ocyte

s

⇒TIM

-3 ab

αTIM-3

ab

Contro

l ab

Mon

ocyte

s⇒Con

trol a

b

Mon

ocyte

s⇒αTIM

-3

αTIM-3

Figure 5.The significance of TIM-3–expressing myeloid cells in an animal model. A, schema of the in vivo experiment. B, NSG mice (n ¼ 5 per group) were intraperitoneallytreated with M-CSF at 500 mg/kg twice per week. CD14þ monocytes isolated from patients' PBMCs were transferred i.v. into NSG mice along with s.c.injections of ACHN tumor cells. The tumor volume in each mouse treated as described above was measured 4 weeks after tumor challenge. � , P < 0.05. C,the frequencies (%) of CD68þ populations within tumor. D, the frequencies (%) of TIM-3þ populations within CD68þ macrophages. E, TIM-3þ cells wereobserved by immunohistochemistry. F, the frequencies (%) of CD68þ total and CD68þCD163þ M2 macrophage-infiltrating tumor tissues were evaluated byflow cytometry. Similar results were obtained in four independent experiments. � , P < 0.05. ns, not statistically significant.






carcinogenesis, and the acquisition of cancer-stem cell properties(26, 27). Thus, it is of interest to evaluate whether the TIM-3–NF-kB axis may be a key pathway that coordinately stimulatesintrinsic oncogenic signals and immune-mediated carcinogenicpathways, such as shown in this study, in which TIM-3 mediatestumor- and myeloid cell–mediated regulation of RCC tumori-genesis. It is noteworthy that the same antibody exerts antitumoreffects by coordinately targeting TIM-3 on cancer cells and mye-loid cells, raising the possibility that inhibition of TIM-3 maycreate a cancer microenvironment that antagonizes tumorigenic-ity and stimulates endogenous cancer immunosurveillance.

In addition, we found higher TIM-3 expression on RCC tumorsclosely correlated with increased TAMs infiltration, raising thepossibility that TIM-3 expression on tumor cells might be regu-lated by cell–cell interactions with TAMs. We have previouslydemonstrated that macrophages induce signal transducer andactivator of transcription 3 (Stat3) signal activation in RCC cellsby direct cell–cell contact. Furthermore, the engagement of mem-brane-type M-CSF expressed on cancer cells with CD115 onmacrophages contributes to this cell–cell interaction (16). More-over, TAMs support tumorigenic activities and trigger resistance toanticancer drugs by inducing cancer-stem cell properties in tumorcells (28–30). However, we did not observe myeloid cell–medi-ated induction of TIM-3 on RCC cell lines in vitro, although RCCtumors express TIM-3 in the vicinity of macrophages in clinicalsamples. A previous report suggested that various mediatorspreferentially produced from the tumor microenvironment, suchas IL10, VEGF-A, and arginase-I, were responsible for inducingTIM-3 expression on myeloid cells (7), though the downstreamsignals critical for transcriptional and translational regulation ofTIM-3 remain obscure. Thus, it is likely that various mediatorsderived from complex networks of the tumor microenvironmentare required; those from in vitro cultured myeloid cells may beinsufficient for inducing TIM-3 expression on RCC cells. It willtherefore be important to elaborate the detailed mechanismswhereby tumor–myeloid cell interactions affect TIM-3 expressionon a variety of tumor cells, including RCC cells.

The adaptive transfer of human CD14þ monocytes had littleimpact on in vivo tumor growth, whereas CD14þ monocytesincreased tumorigenic activities of in vitro–cultured RCC cell lines.Because complexity of the tumor microenvironment serves as acritical factor to determine the direction ofmyeloid cell–mediatedregulation of tumorigenicity, it is likely that the tumor microen-vironment influences the differentiation of transferred humanmonocytes toward subsets with protumor or antitumor proper-ties. We demonstrated that treatment with anti–TIM-3 mAbsuppressed tumor growth when human monocytes were adop-tively transferred, and this observation might indicate that block-

ing of TIM-3 would induce antitumor properties in transferredhuman monocytes. Thus, TIM-3 may be a critical regulator ofmyeloid cell plasticity in the tumor microenvironment.

In conclusion, we demonstrate herein the significant involve-ment of TIM-3 on cancer cells and myeloid cells in the regulationof RCC tumorigenesis. High expression of TIM-3 on cancer tissueswith increased infiltration of TAMs serves as useful biomarker topredict poor clinical prognosis for patients with ccRCC. TIM-3 oncancer cells is involved in the resistance to sunitinib and small-molecule inhibitors of mTOR. TIM-3 on myeloid cells interactswith cancer cells to increase tumorigenicity and trigger resistanceto anticancer drugs. The essential function of TIM-3on cancer cellsand myeloid cells underscores the inhibition of TIM-3 as asuitable strategy with which to rewire tumorigenic microenviron-ments and enhance the clinical efficacy of standard anticancerregimens in patients with RCC.

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

Authors' ContributionsConception and design: Y. Komohara, K. Akashi, M. JinushiDevelopment of methodology: Y. KomoharaAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): Y. Komohara, T. Morita, H. Horlad, S. Yamada,S. Kitada, I. Kinoshita, H. Dosaka-Akita, K. AkashiAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): Y. Komohara, T. Morita, D.A. Annan, K. Ohnishi,S. Yamada, K. Akashi, M. JinushiWriting, review, and/or revision of themanuscript: Y. Komohara, I. Kinoshita,K. Akashi, M. JinushiAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): Y. Komohara, T. Nakayama, S. SuzuStudy supervision: T. Nakayama, M. Takeya, M. Jinushi

AcknowledgmentsThe authors thank Muhammad Baghdadi, Akihiro Yoneda, Emi Kiyota,

Osamu Nakamura, Takenobu Nakagawa, and Yui Hayashida for their technicalassistance.

Grant SupportThis work was supported by grants from the Ministry of Education, Culture,

Sports, Science and Technology (MEXT), and the Ministry of Health, Labor andWelfare of Japan.

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

Received August 24, 2014; revised February 1, 2015; acceptedMarch 4, 2015;published OnlineFirst March 17, 2015.

References1. Patel AR, Prasad SM, Shih YC, Eggener SE. The association of the human

development index with global kidney cancer incidence and mortality.J Urol 2012;187:1978–83.

2. Russell CM, Espiritu PN, Kassouf W, Schwaab T, Buethe DD, Dhilon J,et al. Surgical outcomes in the management of isolated nodal recur-rences: a multicenter, international retrospective cohort. J Urol 2014;192:350–6.

3. Motzer RJ, Hutson TE, Tomczak P,MichaelsonMD, Bukowski RM,OudardS, et al. Overall survival and updated results for sunitinib compared withinterferon alfa in patientswithmetastatic renal cell carcinoma. J ClinOncol2009;27:3584–90.

4. Monney L, Sabatos CA, Gaglia JL, Ryu A,WaldnerH, Chernova T, et al. Th1-specific cell surface protein Tim-3 regulates macrophage activation andseverity of an autoimmune disease. Nature 2002;415:536–41.

5. Sakuishi K, Jayaraman P, Behar SM, Anderson AC, Kuchroo VK. EmergingTim-3 functions in antimicrobial and tumor immunity. Trends Immunol2011;32:345–9.

6. Zhu C, Anderson AC, Schubart A, Xiong H, Imitola J, Khoury SJ, et al. TheTim-3 ligand galectin-9 negatively regulates T helper type 1 immunity.Nat Immunol 2005;6:1245–52.

7. Chiba S, BaghdadiM, Akiba H, YoshiyamaH, Kinoshita I, Dosaka-Akita H,et al. Tumor-infiltrating DCs suppress nucleic acid-mediated innate


Komohara et al.




immune responses through interactions between the receptor TIM-3 andthe alarmin HMGB1. Nat Immunol 2012;13:832–42.

8. Jinushi M. Regulatory mechanisms of nucleic acid-mediated innateimmune responses in the tumor microenvironment. Oncoimmunology2012;1:1632–4.

9. Wiener Z, Kohalmi B, Pocza P, Jeager J, Tolgyesi G, Toth S, et al. TIM-3 isexpressed in melanoma cells and is upregulated in TGF-beta stimulatedmast cells. J Invest Dermatol 2007;127:906–14.

10. Li H, Wu K, Tao K, Chen L, Zheng Q, Lu X, et al. Tim-3/galectin-9 signalingpathway mediates T-cell dysfunction and predicts poor prognosis inpatients with hepatitis B virus-associated hepatocellular carcinoma. Hepa-tology 2012;56:1342–51.

11. Zhuang X, Zhang X, Xia X, ZhangC, Liang X,Gao L, et al. Ectopic expressionof TIM-3 in lung cancers: a potential independent prognostic factor forpatients with NSCLC. Am J Clin Pathol 2012;137:978–85.

12. Kitada S, Yamada S, Kuma A, Ouchi S, Tasaki T, Nabeshima A, et al.Polypeptide N-acetylgalactosaminyl transferase 3 independently predictshigh-grade tumours and poor prognosis in patients with renal cell carci-nomas. Br J Cancer 2013;109:472–81.

13. Komohara Y, Ohnishi K, Kuratsu J, Takeya M. Possible involvement of theM2 anti-inflammatory macrophage phenotype in growth of human glio-mas. J Pathol 2008;216:15–24.

14. Fourcade J, Sun Z, BenallaouaM, Guillaume P, Luescher IF, Sander C, et al.Upregulation of Tim-3 and PD-1 expression is associated with tumorantigen-specific CD8þ T cell dysfunction in melanoma patients. J Exp Med2010;207:2175–86.

15. Sakurai T, Misawa R, Tanaka-Douzono M, Hagesita H, Nomaguchi K,Kuhara T, et al. Effect of coadministration of M-CSF and IFN-a on NK1.1þ

cells in mice. J Interferon Cytokine Res 2002;22:701–8.16. Komohara Y, Hasita H, Ohnishi K, Fujiwara Y, Suzu S, Eto M, et al.

Macrophage infiltration and its prognostic relevance in clear cell renal cellcarcinoma. Cancer Sci 2011;102:1424–31.

17. Motzer RJ, Hutson TE, Tomczak P, MichaelsonMD, Bukowski RM, Rixe O,et al. Sunitinib versus interferon alfa in metastatic renal-cell carcinoma.N Engl J Med 2007;356:115–24.

18. Hudes G, Carducci M, Tomczak P, Dutcher J, Figlin R, Kapoor A, et al.Temsirolimus, interferon alfa, or both for advanced renal-cell carcinoma.N Engl J Med 2007;356:2271–81.

19. Nishizawa S, Hirohashi Y, Torigoe T, Takahashi A, Tamura Y, Mori T, et al.HSDNAJB8 controls tumor-initiating ability in renal cancer stem-like cells.Cancer Res 2012;72:2844–54.

20. Tang D, Lotze MT. Tumor immunity times out: TIM-3 and HMGB1. NatImmunol 2012;13:808–10.

21. FreemanGJ, Casasnovas JM, UmetsuDT,DeKruyff RH. TIM genes: a familyof cell surface phosphatidylserine receptors that regulate innate andadaptive immunity. Immunol Rev. 2010;235:172–89.

22. Rangachari M, Zhu C, Sakuishi K, Xiao S, Karman J, Chen A, et al. Bat3promotes T cell responses and autoimmunity by repressing Tim-3–medi-ated cell death and exhaustion. Nat Med 2012;18:1394–400.

23. Kikushige Y, Shima T, Takayanagi S, Urata S, Miyamoto T, Iwasaki H, et al.TIM-3 is a promising target to selectively kill acute myeloid leukemia stemcells. Cell Stem Cell 2010;7:708–17.

24. Jan M, Chao MP, Cha AC, Alizadeh AA, Gentles AJ, Weissman IL, et al.Prospective separation of normal and leukemic stem cells basedon differential expression of TIM3, a human acute myeloid leuke-mia stem cell marker. Proc Natl Acad Sci U S A 2011;108:5009–14.

25. Wu FH, Yuan Y, Li D, Lei Z, Song CW, Liu YY, et al. Endothelial cell-expressed Tim-3 facilitates metastasis of melanoma cells by activating theNF-kappaB pathway. Oncol Rep 2010;24:693–9.

26. Ben-Neriah Y, Karin M. Inflammation meets cancer, with NF-kB as thematchmaker. Nat Immunol 2011;12:715–23.

27. Schwitalla S, Fingerle AA, Cammareri P, Nebelsiek T, G€oktuna SI, ZieglerPK, et al. Intestinal tumorigenesis initiated by dedifferentiation and acqui-sition of stem-cell-like properties. Cell 2013;152:25–38.

28. Jinushi M, Chiba S, Yoshiyama H, Masutomi K, Kinoshita I, Dosaka-AkitaH, et al. Tumor-associated macrophages regulate tumorigenicity andanticancer drug responses of cancer stem/initiating cells. Proc Natl AcadSci U S A 2011;108:12425–30.

29. Iliopoulos D, Hirsch HA, Wang G, Struhl K. Inducible formation ofbreast cancer stem cells and their dynamic equilibrium with non-stemcancer cells via IL6 secretion. Proc Natl Acad Sci U S A 2011;108:1397–402.

30. Lu H, Clauser KR, Tam WL, Frose J, Ye X, Eaton EN, et al. A breast cancerstem cell niche supported by juxtacrine signaling from monocytes andmacrophages. Nat Cell Biol 2014;16:1105–17.






2015;3:999-1007. Published OnlineFirst March 17, 2015.Cancer Immunol Res Yoshihiro Komohara, Tomoko Morita, Dorcas A. Annan, et al. Cell Renal Cell Carcinomasthe Regulation of Tumorigenicity and Clinical Prognosis in Clear The Coordinated Actions of TIM-3 on Cancer and Myeloid Cells in

Updated version

10.1158/2326-6066.CIR-14-0156doi:

Access the most recent version of this article at:

Cited articles

http://cancerimmunolres.aacrjournals.org/content/3/9/999.full#ref-list-1

This article cites 30 articles, 6 of which you can access for free at:

Citing articles

http://cancerimmunolres.aacrjournals.org/content/3/9/999.full#related-urls

This article has been cited by 3 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

Permissions

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

.http://cancerimmunolres.aacrjournals.org/content/3/9/999To request permission to re-use all or part of this article, use this link



http://cancerimmunolres.aacrjournals.org/lookup/doi/10.1158/2326-6066.CIR-14-0156

http://cancerimmunolres.aacrjournals.org/content/3/9/999.full#ref-list-1

http://cancerimmunolres.aacrjournals.org/content/3/9/999.full#related-urls

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

mailto:[email protected]

http://cancerimmunolres.aacrjournals.org/content/3/9/999


The Coordinated Actions of TIM-3 on Cancer and Myeloid ...Research Article The Coordinated Actions...

Documents

Transcript of The Coordinated Actions of TIM-3 on Cancer and Myeloid ...Research Article The Coordinated Actions...