CalnexinImpairstheAntitumorImmunityofCD4þ and CD8 T CellsGRP78, calreticulin, calnexin, GRP94, and ERP57, are a large family of proteins that have been discovered to have many importantroles

Research Article

Calnexin Impairs the Antitumor Immunity of CD4þ

and CD8þ T CellsYichen Chen1, Da Ma1, Xi Wang1, Juan Fang1, Xiangqi Liu1, Jingjing Song1,Xinye Li1, Xianyue Ren1, Qiusheng Li1, Qunxing Li1, Shuqiong Wen1, Liqun Luo2,Juan Xia1, Jun Cui3, Gucheng Zeng2, Lieping Chen4, Bin Cheng1, and Zhi Wang1

Abstract

Elucidation of the mechanisms of T-cell–mediated anti-tumor responses will provide information for the rationaldesign and development of cancer immunotherapies. Here,we found that calnexin, an endoplasmic reticulum (ER)chaperone protein, is significantly upregulated in oral squa-mous cell carcinoma (OSCC). Upregulation of its membra-nous expression on OSCC cells is associated with inhibitedT-cell infiltration in tumor tissues and correlates with poorsurvival of patients with OSCC. We found that calnexininhibits the proliferation of CD4þ and CD8þ T cells isolatedfrom the whole blood of healthy donors and patients withOSCC and inhibits the secretion of IFNg , TNFa, and IL2from these cells. Furthermore, in a melanoma model,

knockdown of calnexin enhanced the infiltration and effec-tor functions of T cells in the tumor microenvironmentand conferred better control of tumor growth, whereastreatment with a recombinant calnexin protein impairedthe infiltration and effector functions of T cells and pro-moted tumor growth. We also found that calnexinenhanced the expression of PD-1 on CD4þ and CD8þ Tcells by restraining the DNA methylation status of a CpGisland in the PD-1 promoter. Thus, this work uncovers amechanism by which T-cell antitumor responses are regu-lated by calnexin in tumor cells and suggests that calnexinmight serve as a potential target for the improvement ofantitumor immunotherapy.

IntroductionAlthough T-cell immunity plays a critical role in mediating

antitumor immunity, the molecular mechanisms underlyingimpaired antitumor T-cell immunity are not fully understood.Immune checkpoint blockade with mAbs directed against theinhibitory immune receptors CTLA-4, PD-1, and PD-L1 hasemerged as a successful treatment approach that has shownstriking antitumor activity in a variety of cancer types (1–3).Currently, there are more than 10 CTLA4, PD-1, and PD-L1antibodies in different stages of clinical trial in tumors. Despitethe durable response rates observed with cancer immunothera-pies, themajority of patients do not benefit from the treatment. A"hot" tumormicroenvironment, typified by an increased numberof CD8þ CTLs and PD-L1–positive cells, has been identified to be

a reliable predictive biomarker of response to immune checkpointblockade (4–6). Some new strategies have been developed toconvert immunologically "cold" tumors into "hot" tumors. Exam-ples of these strategies includemetabolic reprogrammingof T cellsor modulation of the gut microbiome (7, 8). Still anotherapproach needs to be developed based on novel insights intoT-cell responses and immune systems.

Endoplasmic reticulum (ER) chaperones, including BiP/GRP78, calreticulin, calnexin, GRP94, and ERP57, are a largefamily of proteins that have been discovered to have manyimportant roles in maintaining ER homeostasis and contributingto cancer cell survival and progression. The correlation betweenER chaperone expression and tumorigenesis has been extensivelystudied in various cancers, and most reports have indicated thatthese proteins promote the proliferation, migration, and attach-ment of cancer cells (9–13). The ER chaperone calnexin, which isan ER-specific type I transmembrane protein, regulates the foldingand quality control of newly synthesized glycoproteins (14).Calnexin can escape from the ER and be transported to the plasmamembrane or released into the extracellular space by interactingwith glycoproteins (15–17) or via its phosphorylation at twoserine residues (Ser554/564) by protein kinase CK2 (18). Somestudies have found that extracellular calnexin could be involved ininnate and adaptive immunity in non-mammals (19–21), but theimpact of surface or secreted calnexin on the human immunesystem has not been reported. The results from a lung cancerpatient cohort provided evidence that calnexin may be a sero-diagnostic marker for lung cancer (15).

In this study, we found that calnexin inhibits the infiltrationand effector functions of T cells and enhances the expression ofPD-1 on T cells. Membrane expression of calnexin was positivelycorrelated with poor prognosis of patients with oral squamous

1Guanghua School of Stomatology, Guangdong Provincial Key Laboratory ofStomatology, Stomatological Hospital, Sun Yat-Sen University, Guangzhou, P.R.China. 2Zhongshan School of Medicine, Sun Yat-sen University, Guangzhou, P.R.China. 3Key Laboratory of Gene Engineering of the Ministry of Education, StateKey Laboratory of Biocontrol, School of Life Sciences, Sun Yat-Sen University,Guangzhou, P.R. China. 4Department of Immunobiology and Yale Comprehen-sive Cancer Center, Yale University, New Haven, Connecticut.

Note: Supplementary data for this article are available at Cancer ImmunologyResearch Online (http://cancerimmunolres.aacrjournals.org/).

Y. Chen, D. Ma, and X. Wang contributed equally to this article.

Corresponding Authors: Zhi Wang, Sun Yat-Sen University, No. 56, Lingyuan-west Road, Guangzhou 510055, Guangdong, China. Phone: 8620-8733-0591;Fax: 8620-8382-2807; E-mail: [email protected]; and Bin Cheng,[email protected]

doi: 10.1158/2326-6066.CIR-18-0124

�2018 American Association for Cancer Research.

CancerImmunologyResearch

www.aacrjournals.org 123

on April 3, 2020. © 2019 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from

Published OnlineFirst November 6, 2018; DOI: 10.1158/2326-6066.CIR-18-0124

http://crossmark.crossref.org/dialog/?doi=10.1158/2326-6066.CIR-18-0124&domain=pdf&date_stamp=2018-12-19

http://cancerimmunolres.aacrjournals.org/

cell carcinoma (OSCC). Our work thus uncovers a mechanism bywhich antitumor responses are regulated by calnexin expressed intumor cells and suggests that calnexin may serve as a target forantitumor immunotherapy. Elucidationof thesemechanismswillreveal clues as to the next steps that need to be taken to potentiallyovercome resistance to immunotherapy.

Materials and MethodsPatients and tissue samples

For Western blot analyses, 8 pairs of primary OSCC samplesand corresponding normal oral epithelial tissues were obtainedduring surgeries at the Hospital of Stomatology, Sun Yat-senUniversity (Guangzhou, Guangdong, P.R. China). For real-timePCR analyses, 33 pairs of primary OSCC samples and corres-ponding normal oral epithelial tissues were obtained duringsurgeries at the Hospital of Stomatology, Sun Yat-sen University(Guangzhou, Guangdong, P.R. China). For IHC and immuno-fluorescence analyses, the expression of calnexin was investigatedusing a tissue microarray (TMA) containing samples from 357patients with primary OSCC who were treated at the Hospital ofStomatology, Sun Yat-sen University (Guangzhou, Guangdong,P.R. China), between January 2007 and January 2009. All speci-mens were confirmed histologically by hematoxylin and eosinstaining, and tumor tissue was present in more than 80% of thespecimens. The follow-up interval was calculated from the date ofsurgery to the date of death or last clinical evaluation. The detailedinformation of these patients is described in Supplementary TableS2. This study protocol was approved by the Institutional ReviewBoard of the Hospital of Stomatology, Sun Yat-sen University(Guangzhou, Guangdong, P.R. China) and was conducted inagreement with the Helsinki Declaration, and written informedconsent was obtained from all study participants.

Cell lines and reagentsThe HSC-3 cell line was purchased from the cell bank of the

Japanese Collection of Research Bioresource (JCRB). SCC15,SCC25, CAL27, B16F10, and MB49 cells were purchased fromthe ATCC. The normal keratinocyte cell line (NOK-SI) was kindlyprovided by J. Silvio Gutkind (UCSD, San Diego, CA). Cell linesused in these experiments were passaged a maximum of fourtimes before the experiments. Cells were tested for Mycoplasmacontamination and identified by short tandem repeat.

MiceC57BL/6 mice were purchased from the experimental animal

center of Sun Yat-sen University (Guangzhou, Guangdong, P.R.China). NOD-Prkdcem26Cd52Il2rgem26Cd22/Nju (NCG)micewere purchased from Nanjing Biomedical Research Institute ofNanjing University (Nanjing, China). All experiments wereapproved by the Institutional Animal Care and Use Committeeof Sun Yat-sen University (Guangzhou, Guangdong, P.R. China)and performed following local rules.

Flow cytometrySingle-cell preparations were stained with antibodies pur-

chased from eBioscience, BD Biosciences, and BioLegend. Iso-type-matched control mAbs were used. Intracellular staining wasdone using a Foxp3/Transcription factor staining kit or Intracel-lular fix & Perm Set according to the manufacturer's instructions(BD Biosciences). Briefly, cells were stimulated with 50 ng/mL

PMA(Sigma-Aldrich) and5mg/mL ionomycin (Sigma-Aldrich) inthe presence of GolgiPlug (BD Biosciences). After 4 hours, cellswere stained for dead cells using a FVD dye (eBioscience) andsurface markers then fixed prior to intracellular staining forcytokines. Data were analyzed using a FACSVerse flow cytometer(BD Biosciences) and using FlowJo software (Tree Star). Thefollowing antibodies were used for flow cytometry: anti-mouseCD4 (cloneGK1.5); anti-mouseCD8a (clone 53-6.7); anti-mouseCD3 (clone 17A2); anti-mouse/rat Foxp3 (clone FJK-16s);anti-mouse CD45.2 (clone 104); anti-mouse ki-67 (clone16A8); anti-mouse TNFa (clone MP6-XT22); and anti-mouseIFNg (clone MOB-47).

Xenograft assays in immunodeficient miceCNX-knockdown (sh-CNX) cells or control cells (6� 106)were

injected subcutaneously into rightflankofNCGmice, and1�107

human peripheral blood mononuclear cells (PBMCs) wereinjected via the tail vein after tumor implantation. The animalsweremonitored for tumor formation every 2 days and euthanized3 weeks later. Tumor length (L) and width (W) were measured atthe end of the experiment, and tumor volume was calculated bythe formula (L � W2)/2. Serum cytokines were analyzed at theindicated timepoints, andhumanCD3þT cellswere counted aftertumor dissociation.

Tumor experimentB16F10 tumor cells were retrovirally transduced with sh-CNX

or a control and selected with puromycin (3 mg/mL). For tumorvaccination, na€�ve C57BL/6 mice were immunized with 1 � 106

irradiated B16F10 (1 � 104 rad) cells that were inoculatedsubcutaneously into the left flank. On day 14, the vaccinatedmice were challenged with live transduced tumor cells that wereinoculated subcutaneously into the right flank. A CNX-Ig fusionprotein or Flag-Ig (200 mg) was injected intraperitoneally (i.p.)into eachmouse once aweek. Tumor growthwasmonitored every2 days. The mice were euthanized when the tumor size reached15 mm diameter.

Isolation of tumor-infiltrating leukocytes from tissuesTumor-infiltrating leukocytes (TILs) from the xenograft

tumors were prepared according to the protocol describedpreviously (22, 23). Briefly, tumors were dissected and homog-enized using a GentleMACS dissociator (Miltenyi Biotec),digested with 0.05% collagenase IV (Sigma-Aldrich), 0.002%DNase I (Roche) at 37�C for 1 hour prior to centrifugation onPercoll density gradient (40%–80%), and the TILs were washedand resuspended in RPMI.

Retroviral constructs and transduction of OSCC cell linesHSC3 cells were transfected with CNX shRNA or empty vector

(Genechem) using Polybrene. The cells were trypsinized andreplated in 0.5 mg/mL puromycin 48 hours after transfection.Two months later, the puromycin-resistant stable line wasestablished andmaintained inmediumwith 1 mg/mLpuromycin.The transfected cells were incubated for 24 hours and harvestedfor real-time PCR and Western blot analysis.

In vitro antigen-specific T-cell response assayOSCC tumor cells were isolated from fresh specimen; single-

cell suspensions were obtained as described above. HumanPBMCs from 6 healthy donors and 8 patients with OSCC were

Chen et al.

Cancer Immunol Res; 7(1) January 2019 Cancer Immunology Research124




density-enriched by Ficoll (TBD). Tumor antigens were preparedas described previously (24). Briefly, 2 � 107 tumor cells weresubjected to four freeze (liquid nitrogen) and thaw (37�C waterbath) cycles to obtain a crude lysate as tumor antigen. Afterremoval of large particles by centrifugation and sterilization byfiltering (0.22 mm), the protein concentration in the supernatantwas measured (Coomassie blue protein assay kit, Thermo Scien-tific) and aliquots stored at �80�C until use. A total of 1 � 105

PBMCswere stimulated for 48 hours in the presence or absence oftumor lysate from the same patient and 1 mg/mL PHA-M (SigmaAldrich). GolgiPlug was added for 5 hours to the cells in culture.After 5 hours, cells were stained for viability using a FVD dye(eBioscience), and surface markers were then fixed prior to intra-cellular staining for cytokines using Transcription Factor Buffer Setand Fixation/Permeabilization Solution Kit (BD Biosciences) asper themanufacturer's instructions (25). The culture supernatantswere collected for examination by a human Th1/Th2 cytometricbead assay (BD Biosciences) to quantitatively measure IL2, IL4,IL5, IL10, TNFa, and IFNg protein levels. Briefly, test samples(50 mL) and PE detection antibody were incubated with capturebead reagent for 3 hours in the dark at room temperature. Allunbound antibodies were washed (1.0 mL wash buffer) andresuspended in 300 mL before acquisition on FACSVerse FlowCytometer (BD Biosciences).

IHC and immunofluorescenceTissues were deparaffinized and rehydrated prior to antigen

retrieval in citrate buffer. Tissues were stained with anti-calnexin(C5C9, Cell Signaling Technology), anti-CD3 (17A2,eBioscience), anti-CD4 (OKT4, eBioscience), anti-CD8 (HIT8A,eBioscience).Horseradish peroxidase stainingwas visualizedwith3-39 diaminobenzidine (Gene Company), and fluorescencestaining was visualized with Alexa Fluor–conjugated secondaryantibody (Thermo Scientific). Slides were counterstained, cleared,and mounted.

Pathology assessmentEach TMA staining result was confirmed by a tissue slide

from the same patient. The slides were scanned using an AperioScan Scope AT Turbo for digital image analysis. The imageswere blindly reviewed and scored by two certified anatomicpathologists. Calnexin expression was defined as cytoplasmicor membranous based on its immunoreactivity. As previouslydescribed (26–28), cytoplasmic calnexin staining in tumor cellswas evaluated using the staining–intensity distribution (SID)score. The staining intensity of positive tumor cells was cate-gorized into three grades by comparing the intensity with thatof internal controls: 0, negative staining; 1 (weak), lighter thanskeletal muscle; 2 (moderate), equal to skeletal muscle; and 3(strong), more intense than skeletal muscle. The distribution ofpositive tumor cells was graded as follows: 0, no stained cells; 1,<25% stained cells; 2, 25%–50% stained cells; and 3, >50%stained cells. After multiplying the distribution score by theintensity score in eight different high-powered image fields, theaverage of the eight fields was the SID score for the sample.Cytoplasmic calnexin expression was categorized into low andhigh expression groups using the median of SID scores of thetotal patients. Although upon analysis, calnexin membranousimmunoreactivity correlated with overall survival, we definedextensive staining as positive and no staining or sporadicstaining as negative.

In vitro plate-bound T-cell activation assayHuman CD3þ T cells were isolated from PBMCs of healthy

donor using pan T-cell isolation kits (Stem Cell Inc.). Then,labeling was performed by incubating 106 cells/mL at 37�C for15 minutes with 5 mmol/L CFSE in PBS. Carboxyfluoresceindiacetate succinimidyl ester (CFSE) was quenched by addingtwice the volume of complete media, followed by three washesin complete media before stimulation. 96-well flat-bottom plateswere coated with 1 mg/mL anti-CD3 (clone OKT3) at 4�C over-night. The wells were washed three times with PBS to removeunbound antibody and coated with another 5 mg/mL (ratio 1:5)calnexin-Ig or control-Ig protein in PBS at 4�C overnight.Wells were washed three times with PBS before adding cells.Replicate cultures (1 � 105 cells per well) were maintainedin complete RPMI1640 medium supplemented with 10%FBS, 10 mmol/L HEPES, 50 mmol/L b-mercaptoethanol, andpenicillin/streptomycin/L-glutamine. The cultures were analyzedfor CFSE profiles according to a time course as indicated.

Real-time PCRTotal RNAwas derived from cultured cells with Rnaiso Reagent

(Takara). Quantitative real-time PCR was done using SYBR GreenIDye (Roche) andaccording to theprotocol of LightCycle 480kits(Roche). The primer sequences of CNX were 50-CATGATGGA-CATGATGATGACAC-30 (forward) and 50-CTAGAGGCTTGGTGTATAC-30. Results were normalized to the expression ofGAPDH(forward, 50-AACTTTGGCATTG TGGAAGG-30; reverse, 50-ACA-CATTGGGGGTAGGAACA-30).

Western blot analysisCells and tissues were lysed with radioimmunoprecipitation

assay (RIPA) buffer containing proteases inhibitor cocktail(Sigma Aldrich) and ultrasonication. Protein quantification wasperformed using BCA Protein Assay Reagent (Thermo Fisher Sci-entific), and 45mg protein per sample was loaded into SDS-PAGEand sequentially immunoblotted with calnexin mAb (C5C9, CellSignaling Technology). The proteins were using GAPDH antibo-dies (D16H11, Cell Signaling Technology) as loading controls.

CTL killing assayTumor antigen–specific CD8þ human T-cell clones were gener-

ated fromPBMCs froma healthy donor by in vitro stimulation usingdendritic cells loaded with corresponding peptide epitopes (irradi-ated HSC3 cells were used as tumor antigens). CNX-overexpressingHSC3 cells or control cells were labeled with CFSE and coculturedwith CTLs at an effector-to-target ratio (E/T) of 5:1 and 10:1 for 4hours. Then, 0.1 mg of DAPI was added to each sample, and thesamples were immediately analyzed by flow cytometry. CTL killing(%) ¼ CFSEþDAPIþ cells/total CFSEþ cells � 100%.

Statistical analysisBaseline characteristics were described by mean and SD for

continuous variables or described by numbers and percentages forcategorical variables. To compare the baseline characteristicsbetween different groups, Student t test was used for continuousvariables, whereas x2 tests were used for categorical variables.Overall survival was calculated and described by Kaplan–Meiermethod. The difference of survival curves was tested by log-ranktest. Univariate and multivariate Cox proportional models wereused to analyze the associations between baseline characteristicsandoverall survival, and theHRswith95%confidence interval (CI)

Calnexin Impairs T-cell Antitumor Immunity

www.aacrjournals.org Cancer Immunol Res; 7(1) January 2019 125




were calculated. All statistical analyses were performed byGraphPad Prism 7.0 and Stata/MP 14.0. All tests were two-sided,and a P value of less than 0.05 was considered significant.

ResultsUpregulation of membranous calnexin is correlated withreduced T-cell infiltration

iTRAQ-coupled 2D LC-MS/MS technique was used to studythe protein expression patterns of OSCC tumor tissue and

control normal tissue. In total, 6 pairs of tissue lysates wereanalyzed. When the protein patterns of the primary tumor andits corresponding normal tissue were compared, multiple pro-teins were found to be differentially expressed. SupplementaryTable S1 shows 43 proteins that are significantly upregulated(>2-fold). Calnexin exhibited a higher expression in cancertissues (2.7-fold elevation) when compared with the corre-sponding normal tissues. We further confirmed its expressionby qRT-PCR (Fig. 1A), Western blotting (Fig. 1B), and IHCstaining (Fig. 1C) in OSCC cells; paired cancer and adjacent

A B

C

CNX

GAPDH

CNX

GAPDH

NOK Cal27 SCC15 SCC25HSC3

N Ca N Ca N Ca N Ca N Ca N Ca N Ca N Ca1 2 3 4 5 6 7 8

Normal tissueOSCC40X40X

4X

E

D

NOKCAL2

7HSC3

SCC15

SCC250

5

10

15

20

25

Rel

ativ

em

RN

Aex

pres

sion

****

*****

*

102 103 104 105 106 107

3

4

5

6

7

2 3 4 5 6 7

SS

C

FSC

SS

C

PI

CN

X

EpCAM2 3 4 5 6 7

2

3

4

5

6

7

CD

4/8

CD3

F

Cance

r

Normal

0

2

4

6

8

10

CN

X/G

AP

DH

Rat

io

P= 0.0121

OSCC WGADAPI CNX Merge

Melanoma WGADAPI CNX Merge

CNX+ Epithelial cells (x107/g)

0 2 4 60

1

2

3 R = −0.5080P = 0.0222

CNX+ Epithelial cells (x107/g)

CD

4+ Cel

ls (x

107 /g

)

CD

8+ Cel

ls (x

107 /g

)

0 2 4 60

1

2

3

4

5 R= −0.5197P= 0.0189

0

2.0 M

4.0 M

6.0 M

8.0 M

0 2.0 M 4.0 M 6.0 M 8.0 M

0

2.0 M

4.0 M

6.0 M

8.0 M

10 10 10 10 10 10

10 10 10 10 10 10

10

10

10

10

10

10

10

10

10

10

10

0

10

10

10

10

10

Figure 1.

Calnexin expression is upregulated in OSCC, and its membranous expression is correlated with reduced infiltration of T cells in OSCC tissues. A, qRT-PCRanalyses of calnexin (CNX) expression in tumor and adjacent normal tissues derived from the same OSCC patient (n ¼ 33; left); qRT-PCR analyses of CNXexpression in OSCC cell lines (right). NOK cells served as control cells. B, Western blotting analysis of CNX expression in the cell lines (top) and pairedtissues (bottom). N, normal tissue; Ca, cancer (n ¼ 8). C, IHC analysis of CNX expression in tumor and adjacent normal tissues derived from a typical patientwith progressive OSCC. D, Representative immunofluorescence microscopic images of CNX in OSCC tissue and melanoma tissue. In addition to be primarilyexpressed in the cytosol of OSCC cells, a significant fraction of calnexin colocalizedwithWheatgerm agglutinin (WGA) in the cell membrane of tumor cells (scale bar,50 mm). E, The gating strategy of CNXþ epithelial cell and T cells in single-cell suspension of fresh surgical specimen. F, The number of CNXþ epithelial cellsamong every gram of OSCC tissue, was correlated with CD4þ and CD8þ T cells in corresponding single-cell suspensions, n ¼ 20. Pearson correlationcoefficient was used. Bar graph, mean � SEM. � , P < 0.05; ��, P < 0.001; �� , P < 0.0001. One representative experiment of three is depicted.

Chen et al.





noncancerous tissues were derived from the same patients withOSCC. We found that calnexin was significantly upregulatedin OSCC tissues.

In addition, immunofluorescence analysis suggested thatalthough most of the calnexin protein expression was observedin the cytosol of OSCC cells, a considerable fraction ofcalnexin colocalized with WGA in the cell membrane of tumorcells. Membrane localization of calnexin has been found inseveral tumors, including OSCC and melanoma (Fig. 1D),suggesting that calnexin may exert its biological or immuno-logic regulation functions as either a secreted or membrane-bound form. Because T cells play a critical role in mediatingantitumor immunity, and impaired T-cell infiltration is pos-itively correlated with poorer prognosis of tumor patients(29), we first determined whether there was any correlationbetween membranous expression of calnexin and T-cell infil-tration in tumor tissues. Cell surface calnexin expression wasdetermined by flow cytometry. We found that higher calnexinmembranous expression was negatively correlated with thenumbers of infiltrated CD4þ and CD8þ T cells in OSCC tumortissues (Fig. 1E and F).

Upregulation of membranous calnexin correlated with poorclinical prognosis

Given that calnexin membranous expression was upregu-lated in OSCC tissues, we next determined whether calnexinexpression in the cytosol and cell membrane was correlatedwith tumor prognosis. To address this, IHC staining of calnexinwas performed to assess the expression of calnexin in samplesfrom 357 patients with primary OSCC. Representative imagesof the intensity stages are shown in Fig. 2A. Overall, cyto-plasmic calnexin expression was categorized into low (178 ofthe 357 tumor samples, 49.86%) and high (179 of the 357tumor samples, 50.14%) expression groups using the cutoffpoint 5.04 based upon the median of SID scores of the totalpatients. In addition, 71 of the 357 tumor samples (19.89%)showed apparent calnexin expression at the plasma membrane,whereas 286 of the 357 tumor samples (80.11%) showednegative staining of calnexin at the plasma membrane (Sup-plementary Table S2). Representative images are shownin Fig. 2A. The expression of calnexin was assessed for associ-ation with a number of clinicopathologic variables (Supple-mentary Table S2).

Kaplan-Meier survival curves show, overall patient survivalwas not significantly different when compared between lowand high cytoplasmic expression of calnexin (P ¼ 0.405),whereas patients with positive calnexin membranous expres-sion had a significantly reduced overall survival than patientswith negative expression [3-year OS: 47.89% (35.93%–

58.88%) vs. 66.43% (60.64%–71.58%), P ¼ 0.016; Fig. 2B].In the univariate analysis, calnexin membranous expression(P ¼ 0.018) along with nodal stage (P < 0.001), clinical TNMstage (P ¼ 0.030), and radiotherapy (P ¼ 0.042) were signif-icantly associated with overall survival. Adjusted for nodal stageand radiotherapy, patients with positive membranous expres-sion of calnexin was significantly associated with reducedoverall survival, compared with patients with negative expres-sion (HR, 1.59; 95% CI, 1.10–2.30; P ¼ 0.013; Fig. 2C). Therewas no significant association between cytoplasmic expressionof calnexin and overall survival among patients with OSCC(Supplementary Tables S3 and S4).

Calnexin inhibits T-cell proliferation and antitumor effectorfunctions

Because upregulation of calnexin in OSCC tumor tissues wascorrelated with reduced infiltration of CD4þ andCD8þ T cells, wehypothesized that calnexin impairs the antitumor immunity ofeffector T cells. A calnexin-Ig fusion protein (CNX-Ig) was gener-ated to examine the regulatory roles of calnexin in T-cellresponses. Indeed, we found that when immobilized on a micro-plate, calnexin-Ig, but not control-Ig, suppressed the proliferationof bulk purified CD4þ and CD8þ T cells in response to anti-CD3stimulation (Fig. 3A) and inhibited the production of effectormolecules such as IFNg , TNFa, and IL2 (Fig. 3B). Furthermore,calnexin inhibited the antitumor cytolytic functions of CD8þ

T cells against HSC3 tumor cells (Fig. 3C). These data collectivelysuggested that calnexin inhibited the proliferation and antitumoreffector functions of CD4þ and CD8þ T cells. Although thereceptor for calnexin is unknown, we speculated that the engage-ment of calnexin-R on T cells suppresses T-cell receptor (TCR)signaling. To test this hypothesis, proximal TCR signaling eventswere examined using calnexin-Ig. LAT is a proximal signalingadaptor that is phosphorylated upon TCR stimulation and formsa complex with multiple signaling molecules, including SH2domain containing a leukocyte protein of 76 kDa (SLP76) andphospholipase C (PLC)-g1 (30). Immobilized calnexin-Ig sub-stantially reduced the amount of SLP76 recruited to the CD3complex, as well as its phosphorylation. When total cell lysateswere examined, the phosphorylation of several downstreamsignaling molecules, such as Akt and Erk1/2 was also impaired(Fig. 3D).

In addition to demonstrating the inhibitory effect of calnexinon peripheral blood T cells from healthy donors, we also evalu-ated its role in circulating blood T cells from patients with OSCC.Tumor lysates from the same patients were used as tumor anti-gens. We found that PBMCs fromOSCC patients cocultured withcalnexin-Ig showed inhibitory effects on the proliferation ofCD8þ T cells (Fig. 4A) and reduced the number of functionalCD8þ T cells producing IFNg by nearly half (Fig. 4B). Thesefindings were confirmed by our cytometric bead assay (CBA)results, which showed decreased production of IFNg and TNFabut increased production of IL10 by T cells (Fig. 4C). Thesechanges were more significant in the tumor antigen–experiencedcells. Because an increase in IL10 production by T cells wasobserved, the induction of Treg was examined. However, calnexincould not promote Treg production in an antigen-specificmanner(Supplementary Fig. S1). These data collectively suggested thatcalnexin inhibited the proliferation and antitumor effector func-tions of CD8þ T cells in patients with OSCC in an antigen-dependent manner.

Calnexin promotesOSCC tumor growth in ahumanizedmousemodel

Because calnexin inhibits the proliferation and effector func-tions of CD4þ and CD8þ T cells, we next determined whethercalnexin-mediated impairment of antitumor T-cell responsescontributes to tumor growth. Because OSCC tumor cells do notgrowwell in wild-typemice, a humanizedmousemodel was used(31, 32). HSC3 tumor cells expressing shRNA-CNX (sh-CNX) orshRNA-control (sh-NEG)were inoculated intoNCGmice, and themice were engrafted with human PBMCs after tumor implanta-tion. Themice were euthanized before experiencing weight loss, asymptom of graft-versus-host disease (GVHD) that occurs in this






humanized mouse model (Fig. 5A). In this model, the tumorgrowth in the sh-CNX group was lower than that in the sh-NEGgroup (P¼ 0.047; Fig. 5B).We also detected increased frequenciesof multifunctional CD3þ T cells producing IFNg in the sh-CNXgroup comparedwith the sh-NEGgroupuponPBMCengraftment(Fig. 5C and D). We also examined control mice without PBMC

injection in HSC3 tumor model and found that in contrast to theresults fromhumanizedmice, calnexin silencing promoted tumorgrowth in the immunodeficient mice (Fig. 5E and F), indicatingthat calnexin might have another tumor-intrinsic role that isindependent of its function on T cells. These data indicated thatcalnexin suppressed antitumor immunity and promoted OSCC

A

CNX Intensity (weak) CNX Intensity (moderate) CNX Intensity (strong)

B

CNX Cytoplasmic staining

CNX Plasma membrane staining

Positive Negative

C

0 12 24 36 48 60 720

50

100

Months to last follow-up

Ove

rall

surv

ival

(%)

NO. at risk179178

153154

123122

114110

10592

8360

33

High expressionLow expression

High expression

Low expression

Log-rank test: P = 0.405

0 12 24 36 48 60 720

50

100

Months to last follow-up

Ove

rall

surv

ival

(%)

Negative

NO. at risk28671

25057

20540

19034

16334

11825

60

NegativePositive

Positive

Log-rank test: P= 0.016

0 1 2 3

Multivariable risk factor

CNX membranous expression(Positive vs. Negative)

Nodal stage(N1-N3 vs. N0)Radiotherapy(Yes vs. No)

1.59 (1.1−2.3) 0.013

1.88 (1.36−2.59) <0.001

0.56 (0.36−0.88) 0.012

HR (95% CI) p

CNX Cytoplasmic expression CNX Membrane expression

Figure 2.

Upregulation of membranous calnexin is correlated with poorer overall survival rates of patients with OSCC. A, The calnexin (CNX) cytoplasmic expressionwas determinedon the basis of the staining intensity andproportion, and then divided into high and lowexpression groups using the cutoff point 5.04based upon themedian of SID scores of the total patients. The calnexin plasma membrane expression was determined, and the samples were divided into positive and negativeexpression groups. Representative images showing strong-, moderate- and weak-intensity cytoplasmic staining and positive/negative membranous staining intumor tissue samples derived from patients with OSCC. B,Analysis of the associations between the cytoplasmic andmembranous expression of calnexin and overallsurvival among 357 patients with OSCC. Kaplan–Meier survival curves showed that patients with positive membranous expression of calnexin had a significantlyreduced overall survival than patients with negative expression (P¼0.016). Calnexin cytoplasmic expressionwas not significantly associated with overall survival (P¼ 0.405). The P value was determined by log-rank test. C, Adjusted multivariable risk factor cohort of overall survival. Calnexin membranous expression wassignificantly associated with overall survival (P ¼ 0.013).

Chen et al.





tumor growth via inhibiting the proliferation and effectorfunctions of CD4þ and CD8þ T cells.

Calnexin deficiency promotes antitumor immunity andcontrols tumor growth

We next developed a mouse melanoma model to determinewhether calnexin-mediated impairment of T cells contributes totumor growth. We injected mice subcutaneously with B16F10

cells expressing shRNA targeting calnexin (sh-CNX) or controlshRNA (sh-NEG) and monitored tumor growth. To generateprotective immunity, na€�ve mice were vaccinated with irradiatedB16F10 tumor cells in advance (Fig. 6A). We found that knock-down of calnexin inmelanoma tumor cells significantly inhibitedmelanoma growth inmice, whereas administration of calnexin-Igenhanced melanoma growth (Fig. 6B). Furthermore, knockdownof calnexin in melanoma tumor cells increased the infiltration of

CD8 α-CD3α-CD3+CNX-Igα-CD3+Flag-IgControl

CFSE

0

20

40

60

80

100

CD

4+C

FSE-

low

(%)

α-CD3

CNX-Ig

Flag-Ig − −

−

−

− − −

+ ++

+

+

****

*

****

0

20

40

60

80

100C

D8+

CFS

E-lo

w(%

)****

N.S.

**** ****

α-CD3

CNX-Ig

− −

−

−

− − −

+ ++

+

+

Flag-Ig

CD4

24 48 720

100

200

300

400

500

Time (h)

TNFα

(pg/

mL)

CNX-IgFlag-Ig

N.S. * **

C

CNX-Ig

p-SLP76

SLP76

PLCγ-1

p-PLCγ-1

AKT

p-AKT

Erk1/2

p-Erk1/2

β-Actin

Flag-IgD

A B

0 25.5

74.50

10 10 10 10 10 10

0

1

2

3

4

50 5.68

94.30

0 1 2 3 4 5

0

1

2

3

4

50 7.91

92.10

10 10 10 10 10 10

10

10

10

10

10

10

0 38.8

61.20

100

101

102

103

104

105

100

101

102

103

104

105

0 14.3

85.70

100

101

102

103

104

105

100

101

102

103

104

105

VectorCNX

Effector : Target = 5 : 1 Effector : Target = 0 : 1

DA

PI

CFSE

5:1 10:1

0

10

20

30

40

50

CTL

Killi

ng(%

)

CNXVector

*** ***

Effector : Target = 10 : 1

DA

PI

CFSE

24 48 720

200

400

600

800

Time (h)

IL2

(pg/

mL)

** **

24 48 720

500

1,000

1,500

Time (h)

IFN

γ(p

g/m

L)

* * *

0

1

2

3

4

5

0 1 2 3 4 5 0 1 2 3 4 5

10

10

10

10

10

10

10

10

10

10

10

10

10 10 10 10 10 10

Figure 3.

Calnexin inhibited the proliferation of CD4þ/CD8þ T cells and the cytotoxicity of CD8þ T cells. Fresh PBMCs were isolated from 6 healthy donors. A,CFSE-labeled, bulk-purified pan T cells were stimulated by plate-bound anti-CD3 together with coabsorbed calnexin-Ig (CNX-Ig) or control-Ig (Flag-Ig) protein.Top, representative CFSE dilution profiles. Bottom, the percentage of CFSE-low cells was quantified. B, Culture supernatants in A were collected at theindicated times. The concentrations of IL2, IFNg , and TNFa were analyzed by ELISA. C, Tumor antigen–specific CD8þ human T-cell clones (CTL) weregenerated from PBMCs of a healthy donor by in vitro stimulation using dendritic cells loaded with irradiated HSC3 cells. Calnexin-overexpressing HSC3 cells(CNX) or control cells (Vector) were labeled with CFSE and cocultured with CTLs at an effector-to-target ratio (E/T) of 5:1 and 10:1 for 4 hours. D,Engagement of calnexin during TCR activation maximally suppresses proximal adaptor signaling. Na€�ve pan-T cells purified from human PBMCs wereincubated on ice for 30 minutes with anti-CD3/CD28 and CNX-Ig or Flag-Ig. Then, cell lysates were prepared, and the phosphorylation status of SLP76,PLC-g1, AKT, and Erk1/2 was examined by immunoblotting. Bar graph is shown as themean� SEM (n¼ 6); N.S., not significant. � , P < 0.05; �� , P < 0.01; �� , P < 0.001;�� , P < 0.0001. One representative experiment of three is depicted.






CD3þ, CD4þ, andCD8þ T cells inmelanoma tumors (Fig. 6C andD) and enhanced the expression of Ki67 in CD4þ and CD8þ

T cells (Fig. 6E). In addition, treatment with calnexin-Ig inhibitedthis infiltration in melanoma tumors (Fig. 6C and D) and theexpressionofKi67 in these T cells (Fig. 6E).Moreover, knockdownof calnexin in melanoma tumor cells enhanced the expressionof the antitumor effector molecules IFNg and TNFa by CD4þ

and CD8þ T cells in melanoma tumors, and this effect wassignificantly reversed by treatment with calnexin-Ig (Fig. 6F).

No differences in Tregs and MDSC frequencies among TILs werefound (Supplementary Fig. S2). There were no significantdifferences in the proliferation and effector functions of CD4þ

and CD8þ T cells located in the spleen, lymph nodes andPBMCs between the groups (Supplementary Fig. S3). To con-firm that there is no intrinsic enhancement of tumor growth inthe absence of T-cell–mediated antitumor immunity, tumorswere inoculated in T-cell–deficient nude mice. As shown inSupplementary Fig. S4, administration of calnexin-Ig no longer

C

B

IFN

γ

CD8CNX-IgFlag-Ig

Tumor antigenMedium

A

0

10

20

30

40

Ki67

+am

ong

CD

8(%

) *

Tumor antigen

Flag-Ig

CNX-Ig ++

+ +

+ +−−

−−

−−

N.S.

0

100

200

300

400

500

TNFα

(pg/

mL)

***

Tumor antigenFlag-Ig

CNX-Ig ++

+ ++ +−−

−−

−−

0

50

100

150

IFN

γ (p

g/m

L)


CNX-Ig ++

+ ++ +--

--

--

*N.S.

0

50

100

150

200

IL10

(pg/

mL)

*N.S.


CNX-Ig ++

+ ++ +−−

−−

−−

CNX-IgFlag-Ig

0

10

20

30

40

IFN

γ+ am

ong

CD

8(%

)

****N.S.


CNX-Ig ++

+ ++ +−−

−−

−−

CNX-IgFlag-Ig

Tumor antigenMedium

CNX-IgFlag-Ig

37.2

0 102

103

104

105

0

102

103

104

105

13.4

0 102

103

104

105

0

102

103

104

105

17.7

0 102

103

104

105

0

102

103

104

105

21.9

0 102

103

104

105

0

102

103

104

105

Ki6

7

CD8

Figure 4.

Calnexin inhibited tumor antigen–specific Ki67 and IFNg expression on CD8þ T cells during progressive OSCC. Fresh PBMCs isolated from 8 patients withprogressive OSCC were restimulated by tumor lysate in the presence of calnexin (CNX)-Ig or Flag-Ig for 48 hours. The PBMCs were then subjected to flowcytometric analyses of Ki67 and IFNg expression, and the culture supernatants were subjected to cytometric bead assay (CBA) analysis of Th1/Th2 cytokineconcentrations. A, Representative flow cytometric data and bar graph show the percentages of Ki67þCD8þ T cells in the presence of OSCC tumor lysatewith calnexin-Ig or Flag-Ig. B, Representative flow cytometric data and bar graph data show the percentages of IFNgþCD8þ T cells in the presence ofOSCC tumor lysate with calnexin-Ig or Flag-Ig. C, Concentrations of IFNg , TNFa, and IL10 in the culture supernatants. Bar graph, mean � SEM (n ¼ 8).N.S., not significant; � , P < 0.05; �� , P < 0.01; �� , P < 0.0001. One representative experiment of two is depicted.

Chen et al.





enhanced melanoma growth upon T-cell deficiency. Calnexin-silenced B16F10 tumors grew more rapidly than controltumors, which is consistent with our previous observation(Fig. 5E and F) that tumor-intrinsic calnexin itself suppressedtumor growth. Together, these data indicated that calnexindeficiency promotes antitumor immunity and controls tumorgrowth in a T-cell–dependent manner.

Calnexin enhances the expression of PD-1 by repressing PD-1promoter methylation

Given that T-cell surface receptors such as TIGIT, CTLA-4, PD-1,and LAG-3play critical roles in inhibiting T-cell responses,wenextdetermined whether upregulation of calnexin might enhance theexpressionof thesemolecules and therefore induce impairment ofthe proliferation and effector functions of CD4þ and CD8þ T cellsin tumors. To address this, we analyzed the expression of TIGIT,CTLA-4, PD-1H, PD-1, and LAG-3 on CD4þ and CD8þ T cellsderived from melanoma tumor samples. We found that knock-downof calnexin inmelanoma tumor cells significantly decreased

the expression of PD-1, but not TIGIT, CTLA-4, PD-1H, or LAG-3,in CD4þ and CD8þ T cells derived from melanoma tumorsamples (Fig. 7A). In contrast, calnexin-Ig treatment partlyreversed the decrease in PD-1 expression on CD4þ and CD8þ

T cells conferred by knockdown of calnexin in melanoma tumors(Fig. 7A). Similar results were found in an MB49 tumor modelwith calnexin-Ig treatment (Fig. 7B). In addition, calnexin-Igenhanced the expression of PD-1 on CD8þ T cells in PBMCsderived from patients with progressive OSCC, and this enhance-ment was more significant in tumor antigen–experienced T cells,as shown in Fig. 7C. Thus, these data suggested that calnexinenhanced the expression of PD-1 on CD4þ and CD8þ T cells inOSCC in an antigen-dependent manner.

We then determined the mechanism by which calnexinenhanced the expression of PD-1 on T cells in tumors. BecausePD-1 promoter CpG islandmethylation status plays a central rolein mediating PD-1 expression (33, 34), we analyzed the methyl-ation of this region using bisulfite sequencing in T cells fromOSCC patients' PBMCs (Fig. 7D). In contrast to control-Ig, T cells

sh-CNX

sh-NEG

Day1

6×106 sh-CNX or

sh-NEG HSC3 s.c.

Day22

Tumor isolated

FE

C

NCG mice

Day7

Day14

0

10

20

30

40

50

Hum

an-IF

Nγ

(pg/

mL) sh-CNX

sh-NEG

N.S. *

sh-C

NX

sh-N

EG

0

200

400

600

Hum

an-C

D3+

Tce

llin

each

mg

oftu

mor *

0 5 10 15 20 250

200

400

600

800

Days

Tum

orvo

lum

e(m

m3 )

Tum

orvo

lum

e(m

m3 )

sh-CNX

sh-NEG

*

sh-CNX

sh-NEG

0 5 10 15 20 250

200

400

600

800

Days

sh-CNX

sh-NEG**

Day1

1×107 human-PBMC i.v. +

6×106 sh-CNX or sh-NEG HSC3 s.c.

Day7 Day14

Day21

Blood withdrawal

Tumor isolated

NCG mice

BA

D

Figure 5.

Calnexin promotes OSCC tumor growth in humanized NCG mice. A, Schematic diagram showed the experiment protocol used to determine the role ofcalnexin (CNX) in OSCC tumor growth in immune-integrity environment. On day 1, mouse was transplanted with HSC3 cells transduced with sh-CNX or sh-NEG,1 � 107 human PBMCs were injected intraperitoneally. On days 7 and 14, peripheral blood samples were taken. B, Representative in situ images of OSCCtumors and tumor volume kinetics were measured and calculated using the following formula: V ¼ L � W2/2. C and D, The bar graph shows the increasedfrequencies of CD3þ T cells and functional T cells producing IFNg in the sh-calnexin group compared with the sh-NEG group after PBMC engraftment. E,Schematic diagram shows the experiment protocol used to determine the role of calnexin in OSCC tumor growth in immune-deficient environment.HSC3 cells transduced with shRNA targeting calnexin (sh-CNX) or control shRNA (sh-NEG) were injected subcutaneously at indicated time. F,Representative in situ images of OSCC tumors in NCG mice and tumor volume kinetics were measured and calculated using the formula described inB. Bar graph, mean � SEM (n ¼ 5); �, P < 0.05. One representative experiment of two is depicted.






Figure 6.

Calnexin expressed in tumor cells inhibits the antitumor protective immunity of CD4þ and CD8þ T cells in a mouse melanoma model. A, Schematic diagramshows the protocol used to determine the effect of calnexin (CNX) on melanoma tumor growth in mice. B, Kinetics of tumor volumes in mice as indicated (n¼ 4–5).C, Flow cytometry analysis of the number of CD4þ and CD8þ T cells infiltrated in tumors derived from mice with the indicated treatments. Note thatknockdown of calnexin (CNX) significantly increased the number of infiltrated CD4þ or CD8þ T cells in tumors. However, treatment with calnexin-Igdecreased the number of tumor-infiltrated CD4þ or CD8þ T cells (n ¼ 4–5). D, IHC analysis of tumors derived from mice with indicated treatments suggestedthat significantly larger numbers of infiltrated CD3þ T cells in tumors were observed in the calnexin-deficient group. E, Flow cytometry analysis of the Ki67expression on tumor infiltrated CD4þ and CD8þ T cells. F, Representative flow cytometric analysis of expression of IFNg and TNFa in CD4þ or CD8þ T cellsisolated from tumors derived from mice with the indicated treatments; note that knockdown of calnexin in melanoma tumor cells enhanced expression ofthe antitumor effector molecules IFNg and TNFa produced by CD4þ and CD8þ T cells, and this enhancement of effector functions was significantly reversedby treatment with the calnexin-Ig. Bar graph, mean � SEM (n ¼ 4–5); � , P < 0.05; �� , P < 0.01; �� , P < 0.001; �� , P < 0.0001; N.S., not significant.One representative experiment of two is depicted.

Chen et al.





Figure 7.

Calnexin promotes the expression of PD-1 on CD4þ/CD8þ T cells in tumor by restraining the DNAmethylation status of a CpG region in the PD-1 (PDCD1) promoter.A, Representative flow cytometric analysis and dot plot data show the expression of PD-1, TIGIT, CTLA-4, PD-1H, and LAG-3 on CD4þ or CD8þ T cellsisolated from mice in Fig. 6. The data showed that knockdown of calnexin (CNX) in B16F10 tumor cells significantly reduced the expression of PD-1, but notTIGIT, CTLA-4, PD-1H, or LAG-3, on CD4þ or CD8þ T cells (n ¼ 4–5). B, Representative flow cytometric analysis and dot plot data from an MB49 tumor modelshowed that treatmentwith calnexin-Ig, but not Flag-Igprotein, significantly enhanced the expressionof PD-1 onCD4þandCD8þTcells (n¼4–5).One representativeexperiment of two is depicted. C, Representative flow cytometric data and bar graph data showed that calnexin-Ig increased the expression of PD-1 among CD8þ

T cells in the presence of OSCC tumor antigen (n ¼ 8). D, Schematic of the CpG island and bisulfite pyrosequencing region in the PDCD1 promoter. TSS,transcription start site; red letters, CG sites for bisulfite pyrosequencing. Bisulfite pyrosequencing was used to detect the methylation of PD-1 promoter CpG island.E, The average methylation in the calnexin-Ig and Flag-Ig group was calculated. Note that recombinant calnexin-Ig significantly suppressed PD-1 promoterCpG island methylation in T cells. The data are representative of three independent experiments. Bar graph is shown as mean � SEM; � , P < 0.05;�� , P < 0.01; �� , P < 0.001; N.S., not significant.






treated with calnexin-Ig significantly repressed themethylation ofthe PD-1 promoter CpG island (Fig. 7E).

DiscussionIn this study, we discovered that the ER chaperone protein

calnexin was highly upregulated in OSCC tumor tissues andmultiple tumors. Upregulation of membranous calnexin waspositively correlated with poor prognosis of patients with OSCC.We found that calnexin played a central role in inhibiting theinfiltration and effector functions of T cells and promoting theexpression of PD-1 on CD4þ and CD8þ T cells in tumors, whichtherefore enhanced tumor growth, demonstrating the potential ofcalnexin as a new antitumor immunotherapy target.

Calnexin has been reported to play a role in the folding andquality control of newly synthesized glycoproteins (14, 35). Awide variety of important cellular and viral glycoproteins areknown substrates of calnexin, including HIV gp120 and gp160,class I MHC heavy chain, and TCR subunits (36, 37). Althoughseveral reports have shown that calnexin expression may beassociated with the progression of breast cancer, lung cancer, andcolorectal cancer,most previous studies of calnexin focusedon therelationship between the expression of calnexin and clinicaloutcome (38–40). Whether calnexin regulates the T-cell responseduring tumor development is unknown. Here, we first identifiedthat upregulation of calnexin in tumor cells could inhibit theinfiltration of T cells in tumors and the proliferation and effectorfunctions of CD4þ and CD8þ T cells. As increasing evidence hassuggested that the infiltration and effector functions of T cells intumors are critical for antitumor immunity, this finding thereforereveals a mechanism responsible for poor survival of tumorpatients.

A finding of this study is the establishment of an immunologiclink between calnexin and PD-1 on T cells. We found thatknockdown of calnexin in melanoma tumor cells significantlydecreased the expression of PD-1. In addition, calnexin-Ig treat-ment partly reversed the decrease of PD-1 expression on T cells.Calnexin-Ig also enhanced the expression of PD-1 on T cells inPBMCs derived from patients with progressive OSCC by inhibit-ing the PD-1 promoter CpG island methylation. PD-1 is upregu-lated on activated T cells. The binding of PD-1 and PD-L1 inducesT-cell anergy and cell death (5, 41). This study provides evidencesuggesting that PD-1 expressiononT cells canbe influencedby theexpression of calnexins. The detailed mechanism by which cal-nexin mediates PD-1 expression during tumor developmentshould be investigated in future studies. There are substantialefforts underway to identify reliable predictive biomarkers ofresponse and resistance to immune checkpoint blockade, includ-ing total tumor mutational load (42, 43), as well as markers of aneffective immune infiltrate within a tumor signifying a "hot" or"cold" tumor microenvironment (44). This study provides apotential target for the improvement of responses to anti–PD-1immunotherapy. Because the density or distribution of T cells andPD-1/PD-L1 axis activation could affect the differential responsesto checkpoint blockade, the effect of calnexinon the enhancement

of antitumor responses during PD-1 blockade should be deter-mined in further studies.

Although we found that calnexin expressed in tumor cellslimited the infiltration and effector functions of CD4þ and CD8þ

T cells in tumors and therefore promoted tumor cell growth, thespecific receptor expressed on CD4þ and CD8þ T cells thatinteracts with calnexin remains unknown. In addition, the obser-vations that only membranous calnexin expressed in tumor cellswas associated with poorer survival of patients with OSCC indi-cate that direct contactwithPBMCs is required for calnexin to exertits regulatory function. Identification of the receptor that interactswith calnexin expressed in tumor cells will allow us to betterunderstand the mechanism by which calnexin impairs the infil-tration and effector functions of CD4þ and CD8þ T cells in thetumor microenvironment. Although low concentrations of cal-nexin were isolated from lung cancer patients' peripheral bloodserum (15), the interaction between calnexin and T cells mayprimarily occur in the tumor site. Thus, the interaction betweencalnexin and T cells in the circulation and lymphoid tissuemay benot sufficient for inhibition of the proliferation and effectorfunctions of T cells. Previous studies from other groups suggestthat calnexin could be transported to the plasma membrane tointeract with glycoproteins such as clonotype-independent CD3complexes (16, 45). Further studies are required to determinewhat protein interacts with calnexin.

Disclosure of Potential Conflicts of InterestL. Chen is a scientific founder of and has ownership interest in NextCure and

TAYU Biotech, reports receiving commercial research funding from NextCure,and isa consultant/advisoryboardmember forPfizer, Vcanbio, andGenomiCare.No potential conflicts of interest were disclosed by the other authors.

Authors' ContributionsConception and design: Y. Chen, J. Cui, G. Zeng, L. Chen, B. Cheng, Z. WangDevelopment of methodology: D. Ma, X. Wang, J. SongAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): Y. Chen, D. Ma, X. Wang, J. Fang, Q. Li, Q. Li, S. WenAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): Y. Chen, D. Ma, X. Li, X. Ren, L. ChenWriting, review, and/or revision of the manuscript: J. Cui, G. Zeng, L. Chen,Z. WangAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): X. Wang, X. Liu, L. Luo, J. XiaStudy supervision: B. Cheng, Z. Wang

AcknowledgmentsWe thank Luisa A.DiPietro (College ofDentistry, UIC) for the suggestions on

this manuscript. This project was supported by the National Natural ScienceFoundations of China (grant nos. 81772896, 81472524, 81630025, 81602383,and 81602384).

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

ReceivedMarch 7, 2018; revised August 6, 2018; accepted November 2, 2018;published first November 6, 2018.

References1. Okazaki T, Chikuma S, Iwai Y, Fagarasan S,Honjo T. A rheostat for immune

responses: the unique properties of PD-1 and their advantages for clinicalapplication. Nat Immunol 2013;14:1212–8.

2. Zou W, Wolchok JD, Chen L. PD-L1 (B7-H1) and PD-1 pathway blockadefor cancer therapy: Mechanisms, response biomarkers, and combinations.Sci Transl Med 2016;8:328rv4.


Chen et al.




3. Schadendorf D, Hodi FS, Robert C, Weber JS, Margolin K, Hamid O, et al.Pooled analysis of long-term survival data from phase II and phase III trialsof ipilimumab in unresectable or metastatic melanoma. J Clin Oncol2015;33:1889–94.

4. Taube JM, Klein A, Brahmer JR, Xu H, Pan X, Kim JH, et al. Association ofPD-1, PD-1 ligands, and other features of the tumor immune microenvi-ronment with response to anti-PD-1 therapy. Clin Cancer Res 2014;20:5064–74.

5. Tumeh PC, Harview CL, Yearley JH, Shintaku IP, Taylor EJ, Robert L, et al.PD-1 blockade induces responses by inhibiting adaptive immune resis-tance. Nature 2014;515:568–71.

6. Sharma P, Hu-Lieskovan S, Wargo JA, Ribas A. Primary, adaptive, andacquired resistance to cancer immunotherapy. Cell 2017;168:707–23.

7. Buck MD, O'Sullivan D, Klein Geltink RI, Curtis JD, Chang CH, Sanin DE,et al. Mitochondrial dynamics controls T cell fate through metabolicprogramming. Cell 2016;166:63–76.

8. Sivan A, Corrales L, Hubert N, Williams JB, Aquino-Michaels K, Earley ZM,et al. Commensal Bifidobacterium promotes antitumor immunity andfacilitates anti-PD-L1 efficacy. Science 2015;350:1084–9.

9. WiersmaVR,MichalakM, Abdullah TM, Bremer E, Eggleton P.Mechanismsof translocation of ER chaperones to the cell surface and immunomodu-latory roles in cancer and autoimmunity. Front Oncol 2015;5:7.

10. Gutierrez T, Simmen T. Endoplasmic reticulum chaperones and oxidor-eductases: critical regulators of tumor cell survival and immunorecogni-tion. Front Oncol 2014;4:291.

11. Graner MW, Lillehei KO, Katsanis E. Endoplasmic reticulum chaperonesand their roles in the immunogenicity of cancer vaccines. Front Oncol2014;4:379.

12. Dejeans N, Glorieux C, Guenin S, Beck R, Sid B, Rousseau R, et al.Overexpression of GRP94 in breast cancer cells resistant to oxidativestress promotes high levels of cancer cell proliferation and migration:implications for tumor recurrence. Free Radic Biol Med 2012;52:993–1002.

13. Zamanian M, Veerakumarasivam A, Abdullah S, Rosli R. Calreticulin andcancer. Pathol Oncol Res 2013;19:149–54.

14. Schrag JD, Bergeron JJ, Li Y, Borisova S, Hahn M, Thomas DY, et al.The structure of calnexin, an ER chaperone involved in quality control ofprotein folding. Mol Cell 2001;8:633–44.

15. Kobayashi M, Nagashio R, Jiang SX, Saito K, Tsuchiya B, Ryuge S, et al.Calnexin is a novel sero-diagnostic marker for lung cancer. Lung Cancer2015;90:342–5.

16. Wiest DL, Burgess WH, McKean D, Kearse KP, Singer A. The molecularchaperone calnexin is expressed on the surface of immature thymocytes inassociation with clonotype-independent CD3 complexes. EMBO J1995;14:3425–33.

17. Okazaki Y, Ohno H, Takase K, Ochiai T, Saito T. Cell surface expression ofcalnexin, a molecular chaperone in the endoplasmic reticulum. J BiolChem 2000;275:35751–8.

18. Myhill N, Lynes EM,Nanji JA, Blagoveshchenskaya AD, Fei H, Simmen KC,et al. The subcellular distribution of calnexin is mediated by PACS-2.Mol Biol Cell 2008;19:2777–88.

19. Huang Y, Hui K, Jin M, Yin S, Wang W, Ren Q. Two endoplasmicreticulum proteins (calnexin and calreticulin) are involved in innateimmunity in Chinese mitten crab (Eriocheir sinensis). Sci Rep 2016;6:27578.

20. Wuthrich M, Brandhorst TT, Sullivan TD, Filutowicz H, Sterkel A, StewartD, et al. Calnexin induces expansion of antigen-specific CD4(þ) T cells thatconfer immunity to fungal ascomycetes via conserved epitopes. Cell HostMicrobe 2015;17:452–65.

21. Zhang Q, Wang XQ, Jiang HS, Jia WM, Zhao XF, Wang JX. Calnexinfunctions in antibacterial immunity of Marsupenaeus japonicus.Dev Comp Immunol 2014;46:356–63.

22. Lemos H, Mohamed E, Huang L, Ou R, Pacholczyk G, Arbab AS, et al.STING promotes the growth of tumors characterized by low antigenicityvia IDO activation. Cancer Res 2016;76:2076–81.

23. Khan MW, Keshavarzian A, Gounaris E, Melson JE, Cheon EC, Blatner NR,et al. PI3K/AKT signaling is essential for communication between tissue-infiltrating mast cells, macrophages, and epithelial cells in colitis-inducedcancer. Clin Cancer Res 2013;19:2342–54.

24. Bachleitner-Hofmann T, Strohschneider M, Krieger P, Sachet M, Dubsky P,Hayden H, et al. Heat shock treatment of tumor lysate-pulsed dendriticcells enhances their capacity to elicit antitumor T cell responses againstmedullary thyroid carcinoma. J Clin Endocrinol Metab 2006;91:4571–7.

25. Singhal S, Bhojnagarwala PS, O'Brien S, Moon EK, Garfall AL, Rao AS, et al.Origin and role of a subset of tumor-associated neutrophils with antigen-presenting cell features in early-stage human lung cancer. Cancer Cell2016;30:120–35.

26. Harris MR, Yu YY, Kindle CS, Hansen TH, Solheim JC. Calreticulin andcalnexin interact with different protein and glycan determinants during theassembly of MHC class I. J Immunol 1998;160:5404–9.

27. WilliamsDB. Beyond lectins: the calnexin/calreticulin chaperone systemofthe endoplasmic reticulum. J Cell Sci 2006;119:615–23.

28. Lewis MJ, Pelham HR. A human homologue of the yeast HDEL receptor.Nature 1990;348:162–3.

29. Fridman WH, Galon J, Pages F, Tartour E, Sautes-Fridman C, Kroemer G.Prognostic and predictive impact of intra- and peritumoral immuneinfiltrates. Cancer Res 2011;71:5601–5.

30. Zhang W, Sloan-Lancaster J, Kitchen J, Trible RP, Samelson LE. LAT:the ZAP-70 tyrosine kinase substrate that links T cell receptor to cellularactivation. Cell 1998;92:83–92.

31. Pearson T, Greiner DL, Shultz LD. Creation of "humanized" mice to studyhuman immunity. Curr Protoc Immunol 2008;Chapter 15:Unit 15 21.

32. Brehm MA, Cuthbert A, Yang C, Miller DM, DiIorio P, Laning J, et al.Parameters for establishing humanized mouse models to study humanimmunity: analysis of humanhematopoietic stem cell engraftment in threeimmunodeficient strains of mice bearing the IL2rgamma(null) mutation.Clin Immunol 2010;135:84–98.

33. Youngblood B, Oestreich KJ, Ha SJ, Duraiswamy J, Akondy RS, West EE,et al. Chronic virus infection enforces demethylation of the locus thatencodes PD-1 in antigen-specific CD8(þ) T cells. Immunity 2011;35:400–12.

34. Ahn E, Youngblood B, Lee J, Lee J, Sarkar S, AhmedR.Demethylation of thePD-1 promoter is imprinted during the effector phase of CD8 T cellexhaustion. J Virol 2016;90:8934–46.

35. LeachMR, Cohen-DoyleMF, ThomasDY,WilliamsDB. Localization of thelectin, ERp57 binding, and polypeptide binding sites of calnexin andcalreticulin. J Biol Chem 2002;277:29686–97.

36. Jennelle L, Hunegnaw R, Dubrovsky L, Pushkarsky T, Fitzgerald ML,SviridovD, et al. HIV-1 proteinNef inhibits activity of ATP-binding cassettetransporter A1 by targeting endoplasmic reticulum chaperone calnexin.J Biol Chem 2014;289:28870–84.

37. Thammavongsa V, Mancino L, Raghavan M. Polypeptide substrate recog-nition by calnexin requires specific conformations of the calnexin protein.J Biol Chem 2005;280:33497–505.

38. Liu Y, Komohara Y, Domenick N, Ohno M, Ikeura M, Hamilton RL, et al.Expression of antigen processing and presentingmolecules in brainmetas-tasis of breast cancer. Cancer Immunol Immunother 2012;61:789–801.

39. Okayama A, Miyagi Y, Oshita F, Nishi M, Nakamura Y, Nagashima Y, et al.Proteomic analysis of proteins related to prognosis of lung adenocarcino-ma. J Proteome Res 2014;13:4686–94.

40. Yeates LC, Powis G. The expression of the molecular chaperone calnexin isdecreased in cancer cells grown as colonies compared to monolayer.Biochem Biophys Res Commun 1997;238:66–70.

41. Chen DS, Mellman I. Oncology meets immunology: the cancer-immunitycycle. Immunity 2013;39:1–10.

42. Roszik J, Haydu LE, Hess KR, Oba J, Joon AY, Siroy AE, et al. Novelalgorithmic approach predicts tumor mutation load and correlates withimmunotherapy clinical outcomes using a defined genemutation set. BMCMed 2016;14:168.

43. Chan TA, Wolchok JD, Snyder A. Genetic basis for clinical response toCTLA-4 blockade in melanoma. N Engl J Med 2015;373:1984.

44. van der Woude LL, Gorris MAJ, Halilovic A, Figdor CG, de Vries IJM.Migrating into the tumor: a roadmap for T cells. Trends Cancer 2017;3:797–808.

45. van Leeuwen JE, Kearse KP. Calnexin associates exclusively with individualCD3 delta and T cell antigen receptor (TCR) alpha proteins containingincompletely trimmed glycans that are not assembled into multisubunitTCR complexes. J Biol Chem 1996;271:9660–5.






2019;7:123-135. Published OnlineFirst November 6, 2018.Cancer Immunol Res Yichen Chen, Da Ma, Xi Wang, et al. Cells

T+ and CD8+Calnexin Impairs the Antitumor Immunity of CD4

Updated version

10.1158/2326-6066.CIR-18-0124doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerimmunolres.aacrjournals.org/content/suppl/2018/11/06/2326-6066.CIR-18-0124.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerimmunolres.aacrjournals.org/content/7/1/123.full#ref-list-1

This article cites 44 articles, 15 of which you can access for free at:

Citing articles

http://cancerimmunolres.aacrjournals.org/content/7/1/123.full#related-urls

This article has been cited by 1 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/7/1/123To 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-18-0124

http://cancerimmunolres.aacrjournals.org/content/suppl/2018/11/06/2326-6066.CIR-18-0124.DC1

http://cancerimmunolres.aacrjournals.org/content/7/1/123.full#ref-list-1

http://cancerimmunolres.aacrjournals.org/content/7/1/123.full#related-urls

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

mailto:[email protected]

http://cancerimmunolres.aacrjournals.org/content/7/1/123


Documents

CalnexinImpairstheAntitumorImmunityofCD4þ and CD8 T CellsGRP78, calreticulin, calnexin, GRP94, and ERP57, are a large family of proteins that have been discovered to have many importantroles