squamous cell tumors recruit gd t cells producing either ......e-mail: [email protected]...

Research Article

Squamous Cell Tumors Recruit gd T CellsProducing either IL17 or IFNg Depending on theTumor StageElena Lo Presti1,2, Francesca Toia3, Sebastiano Oieni3, Simona Buccheri4,Alice Turdo5, Laura Rosa Mangiapane5, Giuseppina Campisi6, Valentina Caputo7,Matilde Todaro1,8, Giorgio Stassi5, Adriana Cordova3, Francesco Moschella3,Gaetana Rinaldi9, Serena Meraviglia1,2, and Francesco Dieli1,2

Abstract

The identification of reciprocal interactions between tumor-infiltrating immune cells and the microenviroment may help usunderstand mechanisms of tumor growth inhibition or pro-gression. We have assessed the frequencies of tumor-infiltratingand circulating gd T cells and regulatory T cells (Treg) from 47patients with squamous cell carcinoma (SCC), to determine ifthey correlated with progression or survival. Vd1 T cells infil-trated SSC tissue to a greater extent than normal skin, but SCCpatients and healthy subjects had similar amounts circulating.However, Vd2 T cells were present at higher frequencies incirculation than in the tissue of either cancer patients or healthydonors. Tregs were decreased in the peripheral blood of SCCpatients, but were significantly increased in the tumor com-partment of these patients. Tumor-infiltrating gd T cells pref-

erentially showed an effector memory phenotype and madeeither IL17 or IFNg depending on the tumor stage, whereascirculating gd T cells of SCC patients preferentially made IFNg .Different cell types in the tumor microenvironment producedchemokines that could recruit circulating gd T cells to the tumorsite and other cytokines that could reprogram gd T cells toproduce IL17. These findings suggest the possibility that gd Tcells in SCC are recruited from the periphery and their featuresare then affected by the tumor microenvironment. Elevatedfrequencies of infiltrating Vd2 T cells and Tregs differentlycorrelated with early and advanced tumor stages, respectively.Our results provide insights into the functions of tumor-infil-trating gd T cells and define potential tools for tumor immu-notherapy. Cancer Immunol Res; 5(5); 397–407. �2017 AACR.

IntroductionEpithelial surfaces are selective barriers and, together with

immune system, defend from tissue-damaging radiation, bacte-rial and viral invasion, toxins and mutagens. Skin cancers[melanoma, basal cell carcinoma (BCC) and squamous cellcarcinoma (SCC)] are the most frequent cancer types in white

populations (1), with SCC having an 18 to 20 times higherincidence than malignant melanoma (2, 3).

Tumor-infiltrating leukocytes have been found in differentsolid tumors (4) and the extent of their infiltration has beenoften associated with improved prognosis (5). A minor popula-tion of T cells, named gd T cells, that is about 5%ofCD3þ T cells inthe peripheral blood, is present in other anatomic sites, such as theintestine or the skin (6). These invariant gd T cells bind to differentligands, depending on which V genes are used. Those expressingthe Vd2 gene paired with the Vg9 chain account for 50% to >90%of the gd T-cell population in the peripheral blood of healthysubjects and recognize phosphoantigens (PAg) in associationwith butyrophilin 3A1 (BTN3A1; ref. 7), either exogenous PAgsderived from themicrobial nonmevalonate isoprenoid byosynth-esis pathway or endogenous PAgs derived from the mevalonatepathway in human cells, which is highly active upon infection ortumor transformation (8, 9). In contrast, gd T cells that expressVd1 gene paired with different Vg elements, abound in theintestine or the skin (10), recognize the MHC class I–relatedmolecules MICA, MICB, and ULBPs expressed on epithelial cellsupon oxidative stress and heat shock and onmany hematopoieticand epithelial tumoral cells (8, 11).

gd T cells are among the TILs in many types of cancer, but it isunclear if they correlate positively or not with tumor growth orhave any prognostic value. For instance, Vd1-expressing and IL17-producing gd TILs are negative prognosticators in colorectal andbreast cancer (12–14), whereas the presence of the Vd2 subsetamong TILs positively correlates with early-stage melanoma (15)

1Central Laboratory of Advanced Diagnosis and Biomedical Research (CLADI-BIOR), University of Palermo, Palermo, Italy. 2Department of Biopathology andMedical Biotechnologies (DIBIMED), University of Palermo, Palermo, Italy.3Department of Surgical, Oncological and Oral Sciences, Plastic and Recon-structive Surgery, University of Palermo, Palermo, Italy. 4Department for theTreatment and Study of Abdominal Diseases and Transplantation, Mediterra-nean Institute for Transplantation and Advanced Specialized Therapies(ISMETT), Palermo, Italy. 5Department of Surgical and Oncological and OralSciences, Cellular and Molecular Pathophysiology Laboratory, University ofPalermo, Palermo, Italy. 6Department of Surgical, Oncological andOral Sciences,University of Palermo, Palermo, Italy. 7Department ofDermatology, University ofPalermo, Palermo, Italy. 8Department of DIBIMIS, University of Palermo,Palermo, Italy. 9Department of Surgical, Oncological and Oral Sciences, MedicalOncology, University of Palermo, Palermo, Italy.

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

CorrespondingAuthor:SerenaMeraviglia, University of Palermo, Via del Vespro129, 22, Palermo 90144, Italy. Phone: 39-338-3222559; Fax: 39-091-6555901;E-mail: [email protected]

doi: 10.1158/2326-6066.CIR-16-0348

�2017 American Association for Cancer Research.

CancerImmunologyResearch

www.aacrjournals.org 397

on April 26, 2021. © 2017 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from

Published OnlineFirst March 28, 2017; DOI: 10.1158/2326-6066.CIR-16-0348

http://crossmark.crossref.org/dialog/?doi=10.1158/2326-6066.CIR-16-0348&domain=pdf&date_stamp=2017-4-21

http://cancerimmunolres.aacrjournals.org/

and patient's survival. Taken together, these data show a dual rolefor gd T cells in the tumor microenvironment that probablydepends on the specific gd T-cell subset/function recruited to thetumor site.

In contrast, regulatory T cells (Treg) inhibit T-cell activation andimpair antitumor immune responses. Several human cancers,including SCC, have elevated frequencies of Tregs in peripheralblood and within tumors, which are associated with poor out-comes (16–18). In addition, human skin also contains a residentpopulation of Tregs that account for 5% to 10% of the total skin-resident T cells (17, 19, 20). A reciprocal cross-regulation has beenfound between gd T cells and Tregs during mycobacterial infec-tion: Tregs inhibit gd T-cell proliferation and IFNg production (21,22), but, on the other hand, Vg9Vd2 T cells antagonize expansionof Tregs and subsequent suppression of T-cell responses (23).However, the relationship between Tregs and gd T cells in SCCandother cancers has not been studied thoroughly.

In this article, we have analyzed the frequency and thephenotype of gd T cells (both Vd1 and Vd2 T cells) and Tregsamong TILs and in the peripheral blood of patients with SCC todetermine if a specific subset of gd T cells contributes to thepathogenesis of SCC.

Materials and MethodsCharacteristics of sample cohort

In this study, we enrolled 47 patients with histologicallyconfirmed diagnosis of cutaneous SCC treated between 2012 and2015 at the Plastic Surgery Unit of the University of Palermo. Thecohort was composed by 9 females and 38males. Themedian agewas 71 years for male and 74 for female (range, 45–96). Allpatients were staged according to the international staging systemfor skin cancer as described in Supplementary Table S1. Patientswere followed for a minimum of 1 year and data on disease-related mortality and relapse rates were collected.

Healthy subjects were volunteers in good and stable clinicalcondition and had laboratory parameters in the physiologicrange; median age was 74 years (range, 45–96).

Blood was drawn before the surgical excision. The study wasapproved by the Ethical Committee of the University Hospital,Palermo, where the patients were recruited. The study was per-formed in accordance with the principles of the Helsinki decla-ration and those of the "Good Clinical Practices," and all indi-viduals gave written informed consent to participate.

Isolation of circulating and tumor-infiltrating immune cellsand flow cytometry analysis

Peripheral bloodmononuclear cells (PBMCs)were obtainedbydensity gradient centrifugation using Ficoll–Hypaque (PharmaciaBiotech). Tissue specimes were obtained from 47 differentpatients undergoing surgical procedures for tumor skin. Tissuewas obtained fresh and transported to the laboratory for proces-sing. Tissue wasminced into small pieces followed digestion withCollagenase type IV and DNAse (Sigma) for 2 hours at 37�C 5%CO2. After digestion, the extracted cells were washed twice inincomplete medium (RPMI 1640, Gibco). Both PBMCs andtumor-infiltrating cells were stained for live/dead discriminationusing Invitrogen Live/Dead fixable violet dead cell stain kit(Invitrogen). Fc receptor blocking was performed with humanimmunoglobulin (Sigma, 3 mg/mL final concentration) followedby surface staining with different fluorochrome-conjugated anti-

bodies to study the composition of the different subpopulations.The fluorescein isothiocyanate (FITC)-, phycoerythrin (PE)-, PE-Cy5-, allophycocyanin (APC)-, phycoerythrin-Cy7 (PECy7)-,allophycocyanin-Cy7 (APC-Cy7)-conjugated monoclonal anti-bodies (mAb) used to characterize the entire population werethe following: anti-CD3 (Cat 45-0037 eBioscience, Cat 300412and Cat 300420 BioLegend), anti-CD45 (Cat 560274 BD Biosci-ence), anti-pan gd TCR (Cat 555717 BD), anti-Vd1 (CatPG196007 Thermo Fisher), anti-Vd2 (Cat 331408 BioLegend),anti-CD27 (Cat 17-0279 eBioscience), anti-CD45RA (Cat 25-0458 eBioscience), anti-CD14 (Cat 325632 BioLegend), anti-CD19 (Cat 302228 BioLegend). Anti-CXCR1 (Cat 555939),anti-CXCR2 (Cat 551127), anti-CXCR3 (Cat 560831), anti-CXCR4 (Cat 560669), anti-CCR2 (Cat 561744), anti-CCR3 (Cat564189), and anti-CCR5 (Cat 555992) were all purchased fromBD Biosciences. To evaluate Tregs, we used anti-CD3 (Cat 45-0037 eBioscience), anti-CD4 (Cat 130-080-501 Miltenyi Biotec),anti-CD45 (Cat 560274 BD Biosciences), anti-CD25 (Cat 130-101-426 Miltenyi Biotec), and FoxP3 (Cat 130-093-013 MiltenyiBiotec). Cells were fixed and permibilized according to themanufacture's instructions (Miltenyi Biotec GmbH, Treg detec-tion kit). Expression of surface markers was determined by flowcytometry on a FACSCanto II Flow Cytometer with the use ofthe FlowJo software (BD Biosciences). The gating strategyinvolved progressively measuring total cells, viable cells only,and lymphomonocytes and specific cell types. For every sample,100,000 nucleated cells were acquired and values are expressedas a percentage of viable lymphomonocytes, as gated by for-ward and side scatter.

Cytokine productionThe cytokine production capacity of Vd1 and Vd2 T cells

was determined after stimulation with PMA (BD Biosciences,20 ng/mL, final concentration) and ionomycin (BD Biosciences,1 mmol/L, final concentration) for 4 hours at a cell concentrationof 106/mL. After 1-hour stimulation,Monensin (Sigma-Aldrich, 2mmol/L, final concentration) was added, and the cells wereincubated at 37�C in 5% CO2 for remaining time. Staining ofintracellular cytokines was performed by incubation of fixedpermeabilized cells with FITC- or PE-labelled anti-IFNg (Cat554700 and 554701 BD Biosciences) and APC-labelled anti-IL17A mAbs (Cat 506916 BioLegend). After two more washes inPBS containing 1% FCS, the cells were analyzed by FACS as abovedescribed. Viable lymphocytes were gated by forward and sidescatter, and analysis was performed on 100,000 acquired eventsfor each sample by using FlowJo and the following gating strategyto detect lymphocytes: FSC/SSC, live cells, single cells, double-positive CD3, and TCR Vd1 and Vd2 cells.

Negative control (background) values were not subtracted, asthemedian backgrounds for isotype-matchedmAbswas 0.0028%(range, 0.000%–0.0063%). Samples were considered positiveif the number of cells was equal to or greater than 0.01% andat least 10 clustered events were apparent. This empiric cutoffvalue was predicted to be >90% different from background, at ana of 0.05 (24).

CAF, CSC, and SDAC conditioned medium; cytokine andmigration analyses

Primary cancer-associated fibroblasts (CAF), skin squamouscancer stem cells (CSC), and sphere-derived adherent cells(SDAC) were obtained from surgical resection of skin SCC

Lo Presti et al.

Cancer Immunol Res; 5(5) May 2017 Cancer Immunology Research398




subjected to mechanical and enzymatic digestion with collage-nase (0.6 mg/mL, Gibco) and hyaluronidase (10 mg/mL, Sigma).Cell suspension was cultured in 10% fetal bovine serum (FBS)Dulbecco's Modified Eagle's Medium (DMEM) in adhesionflasks, in order to obtain CAFs or in low-adhesion conditionsand in serum-free medium supplemented with EGF and b-FGF,which allows the selective growth of skin squamous CSCs. Skinsquamous CSCs were cultured in 10% FBS DMEM in adherentconditions to obtain SDACs. Cells were washed twice in PBS andincubated in their fresh CSCmedium for 48 hours. CSCsmediumwas deprived of EGF and b-FGF. The medium was then collectedand used for luminex assay.

Forty-eight cytokines [IL1a, IL1b, IL1R antagonist, IL2, IL2Ra,IL3, IL4, IL5, IL6, IL7, IL9, IL10, IL12, IL12 (p40), IL13, IL15, IL16,IL17, IL18, TNF-a, TNF-b, IFNa2, IFNg , G-CSF, GM-CSF, M-CSF,FGF-b, VEGF, PDGF, MIF, MIG, HGF, LIF, b-NGF, SCF, SCGF-b,SDF-1a, TRAIL, eotaxin, IP-10, IL8, MIP-1a, MIP-1b, MCP-1,RANTES, CTACK, GRO-a, and MCP-3] were analyzed in CAF,CSC, and SDAC-conditioned medium obtained from SCC byxMAPmultiplex technology on the Luminex platform (Luminex),using Bio-Rad reagents (Bio-Plex Pro Human Cytokine 27-plexAssay #M500KCAF0Y and Bio-Plex Pro Human Cytokine 21-plexAssay #MF0005KMII, Bio-Rad) acquired and analyzed with theBioplex Manager Software (Bio-Rad). Responses were scoredpositive if the value was 2-fold over the negative control. Briefly,50 mL bead solution (containing assay buffer and 5,000 beads)was added to the appropriate wells in a 96-well Millipore filterplate (Millipore). Fifty microliters assay buffer was added to eachbackground well: 50 mL diluted standard serum pool, diluted 2-fold from 1:25 to 1:3,200 to each standard well and 50 mL dilutedpositive serum control, diluted 1:25 to each positive control well.

Fifty microliters sample, diluted to 1:25 and 1:400, respectively,was added to each sample well. Standard and positive controlswere diluted in assay buffer, and samples were diluted in assaybuffer with 10% sample blocking buffer. After 30 minutes ofincubation at room temperature on a plate shaker and twowashes, 50 mL biotinylated detection Ab, diluted 1:1000 in assaybuffer with a final Ab concentration of 1.0 mg/mL, was added toeach well. After further 30 minutes of incubation at room tem-perature on a plate shaker and two washes, 50 mL diluted strepta-vidin–R-PE, diluted 1:250 in assay buffer, was added to each well.After further 30 minutes of incubation at room temperature on aplate shaker and two washes, the samples were analyzed on theLuminex machinery (Luminex).

TGFb was measured by ELISA according to the manufacturer'sinstructions (R&D Systems).

gd T cells migration was evaluated using a 24-well, Transwellplate (8.0-mm pore size; Corning); in brief, gd T cells werewashed once with RPMI1640 medium, cell count readjusted(105 cells/mL) in T-cell medium, and an aliquot (100 mL) of T-cell suspension was placed in the top chamber of the Transwell.The bottom chamber of the Transwell plate received condi-tioned medium obtained from SSC tissue upon culture in theabsence of FCS at the indicated concentration prior to theaddition of T cells in the top chamber. As a negative control,fresh medium without FCS was added to the lower chamber.After 4 hours of incubation at 37�C in a 5% CO2 atmosphere,the top chamber was removed, and the gd T cells that hadmigrated into the bottom chamber were stained by the use of amodified Giemsa staining (DiffQuik). Cells in 3 or 4 randomlychosen fields (�400 magnification) were counted. For eachexperiment, 3 replicates were performed.

Figure 1.

Frequency of infiltrating and circulating gd T cells expressing either Vd1 or Vd2TCR d chains inHDs andSCCpatients.A,Flow cytometry plot of representative primarydata to define gd T cells expressing either Vd1 or Vd2 TCR d chains in tissue and peripheral blood of HD subjects and SCC patients. Viable lymphocytes weregated by forward and side scatter, and the analysiswas performed on 100,000 acquired events by using FlowJo. The following gating strategywas used to detect gdT lymphocytes: FSC/SSC, live cells, single cells, CD3/Vd1 or CD3/Vd2 double-positive T cells. B, Histogram of cumulative data from 47 patients. Error barsindicate SEM (�� , P < 0.01; � , P < 0.05). Data are from three independent experiments.

Analysis of SCC Infiltrating gd T Cells

www.aacrjournals.org Cancer Immunol Res; 5(5) May 2017 399




Statistical analysisStatistical analysis was performed using GraphPad software

(Prism Software). Data from different groups were comparedwith the nonparametric two-tailed Mann–Whitney test andStudent t test.

A two-tailed nonparametric Spearman test was used to com-pare the effect of cancer stage on the recovery rate and percentageof gd T cells. Differences between group of patients with aprobability of <0.05 were regarded as significant. All P values aretwo-tailed, and all statistical analyses were performed using theGraphPad software.

SPADE analysis. Spanning-tree progression analysis of density-normalized events (SPADE; ref. 25) clustering algorithm on theCytobank.org platform was done to visualize single cells, amongthe live CD45þ lymphocytes from six subjects. The nodes of thetree reproduce clusters of cells that are similar in marker expres-sion. SPADE uses the size and color of each node to signify thenumber of cells and median marker expression, respectively.

Resultsgd T cells were increased in TILs of SCC patients

In order to study the prevalence of gd T cells among TILs in SCC,a single-cell suspension of tumor tissue or normal skin fromhealthy donors was prepared by enzyme digestion and Vd1 orVd2 cell composition assessed by immunophenotyping. Theresults were compared with the frequencies of gd T cells inperipheral blood mononuclear cells (PBMC) isolated from thesame subjects. Figure 1A describes representative data of infiltrat-

ing and circulating gd T cells expressing Vd1 or Vd2 TCR d chainsin SCC patients and in healthy donors (HD).

The gd T cells expressing Vd1 or Vd2 were slightly increased inPBMCs of SCC patients compared with HDs, but only Vd2differences were statistically significant (Fig. 1B). The frequencyof Vd1 T cells was significantly elevated in tumor compartments,but not in the peripheral blood of SCC patients, and the relativepercentage of Vd1 T cells was significantly higher in TILs of SCCpatients compared with that of normal skin obtained from HDs.

Similarly, the percentage of Vd2 T cells was faint and oftenundetectable in normal skin of HDs, but significantly increased intissue of SCC patients. Finally, and as expected, Vd2 T cells weresignificantly elevated in the peripheral blood of SCC patientscompared with healthy donor peripheral blood. Altogether, theseresults indicate that gd T cells are an important component of TILsin SCC patients.

Phenotypic analysis of gd T cells in SCC patientsHuman gd T cells include those with na€�ve or central-memory

phenotypes (Tna€�ve, CD45RAþCD27þ; TCM, CD45RA�CD27þ)that are abundant in secondary lymphoid organs, without imme-diate effector function, and those with effector-memory (TEM,CD45RA�CD27�) and terminally differentiated (TEMRA,CD45RAþCD27�) phenotypes thatmigrate to inflammatory siteswhere they exert effector functions as cytokine production andcytotoxic activity (26). We studied the memory subset distribu-tion of circulating and tumor-infiltrating gd T cells, assessing theexpression of CD27 and CD45RA on Vd1 and Vd2 cells. Analysisshowed that the majority of Vd1 and Vd2 T cells in TILs of SCC

Figure 2.

Memory phenotype of Vd1 and Vd2T cells in TILs and PBMCs of SCC patients. Phenotypical analysis of (A) infiltrating and (B) circulating Vd1 and Vd2 T cells from SCCpatients and HDs upon staining with mAbs to CD45RA and CD27, after gating on CD3þVd1þ or CD3þVd2þT cells. Data are from three independent experiments.C, Flow cytometry panels of a representative experiment, using the gating strategy described in the legend to Fig. 1. Isotype-matched mAbs were usedas controls.

Lo Presti et al.





patients had a TEM (50%) phenotype, whereas the vastmajority ofVd1 and Vd2 T cells in normal skin from HDs had a TCM (80%)phenotype (Fig. 2A).

To understand if the predominance of tumor-infiltrating Vd1and Vd2 T cells with a TEM phenotype was due to the tumormicroenvironment or simply reflected an overall bias in SCCpatients, we studied the phenotype of gd T cells in the PBMCsofHDsandSCCpatients, and compared themwith thephenotypeof the infiltrating gd T cells. Vd1 and Vd2 T cells obtained fromPBMCs of HDs showed a predominant TCM phenotype, but Vd1and Vd2 T cells obtained from PBMCs of SCC patients showed ahomogeneous distribution of all memory subsets, with a slightand nonsignificant predominance of the TEMRA phenotype (Fig.2B). In any event, both Vd1 and Vd2 gd T cells with a TEM

phenotype were poorly represented among circulating gd T cellsof HDs and SCC patients (usually less than 10%), while being thepredominant population (>50%) in the tumor-infiltrating gd Tcells.

Thus, the gd T cells committed to effector activities at the tumorsite are phenotypically different with respect to circulating and/orresident gd T cells. A typical FACS analysis of one representativesample illustrates the different phenotypes of Vd1 and Vd2 gd Tcells in each tested group (Fig. 2C).

Cytokine production by gd T cells in SCC patientsIn order to study cytokine production by tumor-infiltrating and

circulating gd T cells in SCC patients, PBMCs were isolated fromSCC patients(n ¼ 28) and HDs (n ¼ 10). Similarly, single-cell

Figure 3.

Cytokine production of gd T cells in SCC patients. A, Representative gating strategy to define Vd1 and Vd2T cells making IL17 or IFNg among TILs or in PBMCs of SCCpatients. Cells were stimulated in vitro as described in Materials and Methods and were stained with mAbs to IFNg and IL17 after gating on the Vd1þ and Vd2þ T-cellpopulations. B, Histograms of cumulative data from 45 SSC patients. Error bars indicate SEM (�� , P < 0.01; � , P < 0.05). Each experiment was repeated threetimes independently.C,CD45þ single cellswere used to generate the SPADE tree. Infiltrating cells were grouped in two different populations, CD3– and CD3þ (blackouter circles). The color patterns indicate that the nodes contained within the dark black boundary of CD3þ subset are Vd1 or Vd2 cells. The distributionof the major populations is shown for one representative sample. The branching tree is based on the number of cells included in each node and the legendindicates the range of cell per node according to relative median fluorescence intensity.






suspension of tumor tissue (n ¼ 45) or normal skin from HDs(n ¼ 10) was prepared. Cells were stimulated with PMA andIonomycin in the presence of Brefeldin A followed byimmunophenotyping. Figure 3A describes representative gatingstrategy to define Vd1 and Vd2 gd T cells making IL17 or IFNg .Vd1T cells from the PBMCs of both HDs and SCC patients, andnormal skin of HD, preferentially produced IFNg , whereas IL17production was very low, if any (Fig. 3B). Conversely, Vd1Tcells from TILs of SCC patients preferentially produced IL17,whereas IFNg production was decreased compared with that ofVd1T cells from the PBMCs of both HDs and SCC patients, andnormal skin of HDs.

Vd2 T cells fromPBMCs ofHDs and SCCpatients preferentiallyproduced IFNg , but not IL17; the intratumoral Vd2 T cells fromSCCpatients behaved likewise. Productionof IL17was low inVd2T cells from TILs of SCC patients, but significantly higher com-pared with that of Vd2 T cells from the PBMCs of both HDs andSCC patients. The yield of primary Vd2T cells isolated from thenormal skin from HDs was always too low to allow analysis ofcytokine production. Altogether, these results indicated that gd Tcells (bothVd1 andVd2),making IL17 (IL17-producing gd T cells)were increased at the tumor site, but not in the circulation, of SCCpatients.

To better define the IL17 or IFNg profiles of tumor-infiltratinggd T cells of SCC patients andHD, we used the SPADE (Spanning-Tree Progression Analysis of Density-normalized Events) algo-rithm, which distinguishes cell subsets by clustering, based onsurface antigen expression denoted by a colour gradient (Fig. 3C).IL17 had homogeneous maps with highest intensity of Vd1expression, whereas IFNg had diverse maps compared with thatof IL17 and the highest intensity of Vd2 surfacemarker expression.

Possible gd T cell recruitment and polarization in the SCCtumor microenvironment

The increased frequency of gd T cells observed in the tumorenvironment of SCC patients suggested that they would bemigrating to the tumor from the circulation. To test this hypoth-esis, we analyzed expression of chemokine receptors on circulat-ing gd T cells of HDs and SCC patients. The gd T cells had elevatedexpression of a variety of chemokine receptors, among whichwere CXCR1, CXCR2, CXCR3, CXCR4, CCR2, CCR3, and CCR5(Fig. 4A). Of note, CCR5 was preferentially expressed on Vd2, butnot onVd1T cells, whereas Vd1T cells expressedmoreCXCR1 thanVd2T cells. Other chemokine receptors were equally expressed byVd1 and Vd2 T cells. Next, we evaluated chemokine production inthe context of the tumor microenvironment. To this end, we

250

200

150

100

50

0Che

mok

ine

rece

ptor

(MFI

)

CXCR1

CXCR2

CXCR3

CXCR4CCR2

CCR3CCR5

Vδ1Vδ2

HD SKIN

IL6

IL3

IL16IL1aIL18IL1bIL2IL4IL7IL9IL10IL13IL15IL17

IL1ra

IL5

IL2Ra

TNFaTNFbIFNa2LIF

IFNg

IL12p40IL12(p70)

CSC CAF SDAC

MIF

TRAILTGFb

HD SKIN CSC CAF SDAC

FGF-bG-CSF

M-CSF

HGF

b-NGFPDGF-bb

SCF

SCGF-b

GM-CSFVEGF

CXCL1CXCL8CXCL9CXCL10CXCL12CCL2CCL3CCL4CCL5CCL7CCL11CCL27

HD Skin CSC CAF SDAC

Chemokines

Cytokines/growth Factors

Fold change(Log

10 )

+3

–3

Fold change(Log

10 )

+3

–3

10080604020

0

Mig

rate

d ce

lls(%

inpu

t ± S

D)

w/o FCS w/o FCS

RPMI SCC CM

A

B

C

D

Figure 4.

gd T-cell recruitment at the tumor microenvironment of SCC patients. A, Analysis of chemokine receptor expression by circulating Vd1 and Vd2 T cells. Values areexpressed as MFI; error bars indicate SEM. Data are from three independent experiments. B and D, Luminex analysis of chemokines (B) and cytokines/growth factors (D) production byCSCs, CAFs, andSDACs obtained fromSCCpatients, and cultured in vitro as described inMaterials andMethods. The concentrationsof each molecule is evaluated according to a color intensity scale. C, gd T-cell migration in response of SCC conditioned medium was evaluated as described inMaterials and Methods, data are expressed as a percentage of migrated cells among input. Error bars indicate SEM (� , P < 0.05). Each experiment wasrepeated five times independently.

Lo Presti et al.





cultured allogeneic CAFs, SDACs, andCSCs from tumor tissues ofSCC patients (n ¼ 5), and chemokine concentrations were mea-sured in culture supernatant collected after 48 hours; as a control,we also analyzed fibroblasts isolated for the skin of HDs. CSCsproduced large amounts of CXCL1, CXCL8, and CCL2, andintermediate concentrations of CXCL9, CXCL10, CXCL12, CCL5,CCL7, and CCL11 (Fig. 4B). Conversely, both CAFs and SDACshad a more limited chemokine-producing capacity and secretedhigh amounts of CXCL1 and CXCL8, and intermediate concen-trations ofCXCL12 andCCL2. Thus, several cell types in the tumormicroenvironment of SCC patients produce chemokines with thepotential to recruit circulating gd T cells that express the relevantchemokine receptors.

In order to investigate gd T-cell migration to the tumor micro-environment, we performed a migration assay in the presence ofbulk SCC supernatant. The gd T cells were attracted by SCC-conditioned medium compared with medium, indicating thatchemokines present in the tumor microenvironment may causerecruitment of gd T cells at the tumor site (Fig. 4C).

We next addressed the possibility that the tumor microenvi-ronment of SCC patients also contained cytokines capable of

polarizing the differentiation of IL17-producing gd T cells: in fact,we previously demonstrated that gd T cells may be induced toproduce IL17 by combinations of IL6, TGFb, IL1b, and IL23 (27).CSCs, CAFs, and SDACs produced significantly elevated concen-trations of IL6, CSCs produced intemediate amounts of IL23 andTGFb, and less IL1b; and CAFs produced much TGFb, and littleIL1b and IL23 (Fig. 4D). Finally, SDACs produced little TGFb,IL1b, and IL23. Altogether, these results show that cytokines withthe potential to polarize IL17-producing gd T cells are producedby different cell types in the tumor environment of SCC patients.

Correlation between cytokine production, clinical stage of SSC,and clinical outcome

To investigate the clinical significance of gd T cells, frequenciesof these cells in TILs were analyzed according to the clinical stageand the clinical outcome of SCC patients. Because no referencevalue for these populationswas available, patientswere divided intwo groups based onmean percent values of respective cells (datanot shown and Fig. 5).

We observed no associations between the percentage of intra-tumoral Vd1 and Vd2 T cells and any of the well-established

Figure 5.

Correlation between cytokine production, clinical stage of SSC, and clinical outcome. A, Cumulative data of the frequencies of Vd1 and Vd2 T cells producingIL17 or IFNg in SCC patients at early or late clinical stage (I–II vs. III–IV) and in HD subjects. B, Box plot of the percentages of infiltrating Vd1 and Vd2 T cells inthree different groups: HD, SSC patients at stage I/II, and SSC patients at stage III/IV. C, Cumulative data of the frequencies of IL17- or IFNg-producing SSCinfiltrating Vd1 and Vd2 T cells in recurrence/nonrecurrence, dead/live, and no lymph node invasion/lymph node invasion patients. In all experiments, data arereported as mean of percentage of positive cells � SEM (� , P < 0.05).






prognostic factors for SCC (clinical stage, recurrence/not recur-rence, lymph node invasion, and overall survival; data notshown).

However, a correlation was found between the capability of gdT cells in TILs to produce IL17 or IFNg , and the clinical stage ofSCC patients (Fig. 5A). Significantly more IL17-producing Vd1and Vd2 T cells were found in SCC patients with advanced disease(stages III and IV), compared with patients with early disease(stages I and II). In contrast, the frequencies of IFNg-producingVd1 and Vd2 T cells were higher in SCC patients at stages I and II,but significantly decreased in patients with advanced disease(stages III and IV). As a control, percentages of gd T cells amongTILs of SCC patients did not change with disease stage (Fig. 5B).This clearly indicated that the immune response was skewedtoward IL17-producing cells in SCC patients.

We also assessedwhether any correlation existed between IL17-or IFNg-producing intratumoral Vd1 and Vd2 T cells and clini-copathologic staging and follow-up data of SCC patients (Fig.5C). The frequency of IL17-producing Vd1 and Vd2 T cells washigher in patients with high relapse, lymph node metastasis, andmortality rates, whereas the frequence of IFNg-producing Vd2 T

cells was higher in patients with favorable outcome (no relapse,no lymph nodes invasion, and alive).

CD4þCD25þFoxp3þ Tregs are increased in TILs of SCC patientsThe frequency of Tregs in PBMCs of SCC patients and HDs

was analyzed by surface staining for CD4 and CD25, followedby intracellular staining for Foxp3. We assessed infiltrating andcirculating Tregs either in SCC patients or in HDs (Fig. 6A).Tregs were slightly decreased in PBMCs of SCC patients com-pared with HDs, but differences did not attain statistical sig-nificance. However, the percentage of Tregs in TILs was signif-icantly higher than in PBMCs of SCC patients and HDs. Thepercentage of Tregs was significantly higher in TILs from SCCpatients compared with the normal skin of HDs (Fig. 6B).Therefore, Tregs were decreased in the peripheral blood of SCCpatients, but were significantly increased in the tumor com-partment of these patients.

The frequency of infiltrating Tregs was significantly higher inSCC patients with advanced disease (stages III and IV) who hadlymph node metastasis and a high relapse rate, and Treg fre-quency correlate with overall survival (Fig. 6C and 6D).

Figure 6.

Frequency of infiltrating and circulating Tregs in HDs and SCC patients.A, Flow cytometry plot of representative primary data to define Tregs in tissue and peripheralblood of HD subjects and SCC patients. Analysis was performed as described in the legend to Fig. 1. The following gating strategy was used to detect Tregs:FSC/SSC, live cells, single cells, CD3/CD4 double-positive T cells, and CD25þFoxP3þ. Isotype-matchedmAbswere used as controls.B,Histogram of cumulative datafrom 47 patients. Error bars indicate SEM (�� , P < 0.01; � , P < 0.05). Data are from three independent experiments. C, Cumulative data of the frequencyof Tregs in SCC patients at early or late clinical stage (I–II vs. III–IV) and in HD subjects. Error bars indicate SEM (�, P < 0.05). D, Cumulative data of the frequency ofTregs in SCC patients related with relapse rate and overall survival; error bars indicate SEM (�, P < 0.05).

Lo Presti et al.





Correlations between Tregs, Vd1 and Vd2 T cells, clinical stage,and outcome

A significant inverse correlation was found between Vd2 T cellsand Tregs at early clinical stages (I and II) of SCC, which flipped toa statistically significant direct correlation in stages III and IV.Conversely, no correlation was found between Tregs and Vd1 Tcells at different tumor stages (data not shown). This analysissuggests that a reciprocal relationshipmay exist betweenTregs andVd2 T cells, which is largely influenced by the specific tumormicroenvironment.

Related to this, we further evaluated the relationship betweenVd2 T cells and Tregs through ROC curve analysis of the Normal-izedEuclidianDistance (NED)of infiltrating Tregs andVd2T cells,calculated according to the following formula:

2

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi%Treg

%Treg �mean%Tregð Þ� �2

� %Vd2%Vd2�mean%Vd2ð Þ

� �2s

NEDwill have low values when normalized Tregs and Vd2 varyin similar ranges of values, and higher values when normalizedTreg and Vd2 vary in different ranges of values.

The area under the ROC curve (AUC) obtained comparing theNED of percentage of infiltrating Tregs and Vd2 T cells at early (Iand II) versus advanced (III and IV) stage of SCC was 0.89 [95%confidence interval (CI), 0.768–1.04] with P ¼ 0.0016 (Fig. 7).According to the Youden Index, using a cutoff value ofNED<2.33,chosen for scoring purposes to maximize the sum of sensitivityand specificity, we found that the sensitivity was 78.57% andspecificitywas 100%. Therefore, basedon theNEDvalueswith theselected cutoff, 78.57% of subjects at the advanced stage of thetumorwere correctly identifiedwith a positive test, whereas 100%of subjects at the early stage of the tumor were correctly identifiedwith a negative test. The ROC curve analysis further confirms thatthemutual relationships between infiltrating Tregs andVd2 T cellsin early and advanced stage of SCC are different.

DiscussionThe infiltration of tumors by certain subsets of T lymphocytes

has largely been linked to favorable outcome in different types of

cancer (4, 5), whereas other leukocyte subsets, such as Tregs andM2 macrophages, can confer a poor prognosis (28, 29).

gd T cells have unique features thatmake themgood candidatesfor effective antitumor immunotherapy (30), because they are notMHC restricted anddo not rely on costimulatory signals. gd T cellsdisplay potent cytotoxic and antitumor activities in vitro (31–35)and in xenograftmodels in vivo (36, 37), leading to the explorationof their therapeutic potential.

Several studies have shown that gd T cells are present amongtumor-infiltrating lymphocytes (TIL) obtained from differenttypes of cancers, but their clinical relevance is still obscure,because some studies showed a correlation of gd T cells andtumor remission,whereas others showed a correlationwith tumorprogression. Other papers found no correlation of these cells withany prognostic feature.

However, an analysis of expression signatures from �18,000human tumors with overall survival outcomes across 39 malig-nancies identified tumor-infiltrating gd T cells as the most signif-icant favorable cancer-wide prognostic signature (38), includingalso 74 SSC patients (head and neck carcinoma). These disparatefindings strongly suggest that gd T cells in the tumor microenvi-ronment may play substantially different functions, dependingon the specific gd T-cell subset/function recruited to the tumorsite. Furthermore, the net positive or negative biologic effect of gdT cellsmay depend on the histologic tumor type and on the tumorsite, reflecting microenvironmental differences.

Results reported here show that Vd1 T-cell numbers weresignificantly elevated in the tumor compartment compared withperipheral blood in SCC patients and normal skin obtained fromHDs. Similarly, the percentage of Vd2 T cells was low and oftenundetectable in normal skin fromHDs, but significantly increasedin TILs of SCC patients. Tregs were significantly increased in thetumor compartment, but were decreased in the peripheral bloodof SCC patients.

Cells with a TEM phenotype were the predominant populationamong tumor-infiltrating Vd1 andVd2 T cells, highly suggestive ofcells that are home to sites of inflammation and that displayimmediate effector functions such as cytokine production.Accordingly, tumor-infiltrating Vd1 and Vd2 T cells preferentiallyproduced IL17, but not IFNg , whereas circulating Vd1 and Vd2 Tcells had a reciprocal cytokine production pattern. The capabilityof gd T cells in TILs to produce IL17 or IFNg correlated with theclinical stage of SCC patients; IL17-producing gd T cells weresignificantly more plentiful in patients with advanced disease(stages III and IV), whereas SCC patients at stages I and II hadmore IFNg-producing gd T cells.

Five different studies, three in mouse models (12, 39, 40) andtwo in human cancer (13, 14), show that IL17-producing T cellsare key mediators of tumor-associated immunosuppression andpromote tumor progression by several different mechanisms.Typically, murine gd T cells are an innate source of IL17, apotential they acquire in the neonatal thymus independentlyupon encountering their specific antigen (41–44). Conversely,human gd T cells default toward type 1 cytokine production andpredominantly produce IFNg uponactivation.However, gd T cellscan divert from this typical Th1-like phenotype upon appropriateculture condition and polarize to different cytokine-producingsubsets; for example, the addition of IL1b, IL6, IL23, and TGFbtogether with TCR triggering promotes expression of the tran-scription factor RORC and polarization to IL17 gd T cells (27). Inother words, we reasoned that at early stages of tumor growth,

Figure 7.

Correlation between Tregs andVd2þT cells amongTILs and clinical stage of SCCpatients. The graph shows ROC curve (AUC) obtained comparing the NEDof percentage of infiltrating Tregs and Vd2þ T cells at early versus advancedstages of SCC patients.






IFNg-producing gd T cells either expand locally (Vd1) or arerecruited to the tumor site from the peripheral blood (Vd2) andmaypossibly exert antitumor activity; however, with aprogressingtumor, cytokines present in the tumor microenvironment mightreprogram gd T cells to produce IL17, which instead promotestumor growth. In agreementwith this hypothesis, CSCs andpartlySDACs and CAFs produced large amounts of CXCL1, CXCL8, andCCL2, and intermediate concentrations of CXCL9, CXCL10,CXCL12, CCL5, CCL7, and CCL11, and circulating gd cellsexpressed counter-receptors for many of these chemokines, sug-gesting that gd cellsmaymigrate from the circulation to the tumorsite. IL6, IL1b, IL23, and TGFb were also produced by CSCs,SDACs, and CAFs, thus making the environment conducive fordifferentiation of IL17-producing gd T cells.

However, concerning IL23, wewould like to point out that onlyp40 was tested, and p40 is also a constituent of IL12. Consistentwith this, IL12 p70 was highly produced by any cell type in theSCC environment, and this balance between IL12 and IL23mightbe therefore critical in order to determine whether the environ-ment is oriented toward IFNg polarization (in the case of IL12) orIL17 (in the case of IL23). Concerning IL12, CSCs and SDACsproduced high amounts either of p40 (which is commonbetweenIL12 and IL23) and p70, conditioning the polarization towardIFNg-producing cells. Additional studies are thus required toclarify the role of IL12/IL23 produced in the tumor microenvi-ronment as a major determinant of gd T-cell polarization.

We have no direct evidence supporting our contention thatcytokines produced by tumor cells polarize gd T cells toward IL17production. We speculate this may occur on grounds that (i) gd Tcells (both Vd1 and Vd2) that produce IL17 are more highlyrepresented in tumor tissue compared with peripheral blood ofthe same patients, indicating that this IL17 phenotype is largelyshaped by the tumor microenvironment, (ii) tumor cells producecytokines that are known to mediate differentiation of IL17-producing gd cells in humans, and (iii) in human colorectalcancer, DC-derived IL23 was proposed to polarize gd T cellstoward IL17 production (13). Therefore, it is conceivable that gdT cells are recruited into the tumor from draining lymph nodes orperipheral blood and then undergo functional changes inresponse to the local microenvironment, although we cannotrule out the possibility that tissue-resident gd T cells are function-ally affected locally.

We observed a correlation between cytokine-producing intra-tumoral Vd1 and Vd2 T cells and clinicopathologic staging andfollow-up data on SCC patients. Moreover, a significant inversecorrelation was found between Vd2 T cells and Tregs at earlystages, which turned to a significant direct correlation at latestages. Moreover, the Vd2/Treg ratio was prognostically accurate,suggesting that a reciprocal relationship may exist between Tregsand Vd2 T cells in vivo, which is largely influenced by the specifictumor microenvironment.

In conclusions, results here reported show that gd T cells(both Vd1 and Vd2) infiltrate the tumor tissue in SCC patientsand their phenotype and functions are largely influenced by thetumor microenvironment, thus differentially affecting patients'prognoses.

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

Authors' ContributionsConception and design: M. Todaro, G. Stassi, A. Cordova, F. Moschella,S. Meraviglia, F. DieliDevelopment of methodology: E. Lo Presti, A. Turdo, L.R. MangiapaneAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): E. Lo Presti, F. Toia, S. Oieni, S. Buccheri, G. Campisi,V. Caputo, G. RinaldiAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): E. Lo Presti, F. Toia, S. BuccheriWriting, review, and/or revision of the manuscript: E. Lo Presti, S. Meraviglia,F. DieliAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): S. MeravigliaStudy supervision:M.Todaro,G. Stassi, A. Cordova, F.Moschella, S.Meraviglia,F. Dieli

AcknowledgmentsThe authors would like to thank Jean Jacques Fourni�e (CRCT, UMR1037,

Inserm-Univ. Toulouse III Paul Sabatier-ERL5294 CNRS, Toulouse, France) forreading the manuscript.

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

Received December 1, 2016; revised February 6, 2017; accepted March 23,2017; published OnlineFirst March 28, 2017.

References1. Chen J, Ruczinski I, Jorgensen TJ, Yenokyan G, Yao Y, Alani R, et al.

Nonmelanoma skin cancer and risk for subsequent malignancy. J NatlCancer Inst 2008;100:1215–22.

2. Staples M, Marks R, Giles G. Trends in the incidence of nonme-lanocytic skin cancer (SCC) treated in Australia 1985–1995: areprimary prevention programs starting to have an effect? Int J Cancer1998;78:144–8.

3. Holme SA, Malinovszky K, Roberts DL. Changing trends in nonmela-noma skin cancer in South Wales, 1988–98. Br J Dermatol 2000;143:1224–29.

4. GoodenMJ, de BockGH, LeffersN,Daemen T,NijmanHW. The prognosticinfluence of tumour-infiltrating lymphocytes in cancer: a systematic reviewwith meta-analysis. Br J Cancer 2011;105:93–103.

5. Sasada T, Suekane S. Variation of tumor-infiltrating lymphocytes in humancancers: controversy on clinical significance. Immunotherapy 2011;3:1235–51.

6. HaydayAC. gdT cells:a right time anda right place for a conserved thirdwayof protection. AnnuRevImmunol 2000;18:975–1026.

7. Vavassori S, Kumar A,WanGS, RamanjaneyuluGS,CavallariM, ElDaker S,et al. Butyrophilin 3A1 binds phosphorylated antigens and stimulateshuman gd T cells. Nat Immunol 2013;14:908–16.

8. Tanaka Y, Morita CT, Tanaka Y, Nieves E, Brenner MB, Bloom BR. Naturaland synthetic non-peptide antigens recognized by human gd T cells. Nature1995;75:155–8.

9. Eberl M, HintzM, Reichenberg A, Kollas AK,Wiesner J, JomaaH.Microbialisoprenoid biosynthesis and human gd T cell activation. FEBS Lett2003;544:4–10.

10. Bonneville M, O'Brien RL, Born WK. gd T cell effector functions: a blend ofinnate programming and acquired plasticity. Nat Rev Immunol2010;10:467–78.

11. Poggi A, VenturinoC, Catellani S, ClavioM,MiglinoM,GobbiM, et al. Vd1T lymphocytes from B-CLL patients recognize ULBP3 expressed on leuke-mic B cells and up-regulated by trans-retinoic acid. Cancer Res 2004;64:9172–9.

12. Wakita D, Sumida K, Iwakura Y, Nishikawa H, Ohkuri T, Chamoto K, et al.Tumor-infiltrating IL-17-producing gammadelta T cells support the

Lo Presti et al.





progression of tumor by promoting angiogenesis. Eur J Immunol2010;40:1927–37.

13. Wu P, Wu D, Ni C, Ye J, Chen W, Hu G, et al. gdT17 cells promote theaccumulation and expansion of myeloid-derived suppressor cells inhuman colorectal cancer. Immunity 2014;40:785–800.

14. MaC, ZhangQ, Ye J,Wang F, Zhang Y,Wevers E, et al. Tumor-infiltrating gdT lymphocytes predict clinical outcome in humanbreast cancer. J Immunol2012;189:5029–36.

15. Cordova A, Toia F, LaMendola C, Orlando V,Meraviglia S, Rinaldi G, et al.Characterization of human gd T lymphocytes infiltrating primary malig-nant melanomas. PLoS One 2012;7:e49878.

16. Strauss L, Bergmann C, Gooding W, Johnson JT, Whiteside TL. Thefrequency and suppressor function of CD4þCD25highFoxp3þ T cells inthe circulation of patients with squamous cell carcinoma of the head andneck. Clin Cancer Res 2007;13:6301–11.

17. Clark RA, Huang SJ, Murphy GF, Mollet IG, Hijnen D,MuthukuruM, et al.Human squamous cell carcinomas evade the immune response by down-regulation of vascular E-selectin and recruitment of regulatory T cells. J ExpMed 2008;205:2221–34.

18. Gasparoto TH, de Souza Malaspina TS, Benevides L, de Melo EJJr, CostaMR, Damante JH, et al. Patients with oral squamous cell carcinoma arecharacterized by increased frequency of suppressive regulatory T cells in theblood and tumor microenvironment. Cancer Immunol Immunother2010;59:819–28.

19. Clark RA, Kupper TS. IL-15 and dermal fibroblasts induce proliferationof natural regulatory T cells isolated from human skin. Blood 2007;109:194–202.

20. Clark RA. Skin-resident T cells: the ups and downs of on site immunity.J Invest Dermatol 2010;130:362–70.

21. Kunzmann V, Kimmel B, Herrmann T, Einsele H,WilhelmM. Inhibition ofphosphoantigen-mediated gd T-cell proliferation by CD4þCD25þFoxP3þ

regulatory T cells. Immunology 2009;126:256–67.22. Li L, Wu CY. CD4þCD25þ Treg cells inhibit human memory gd T cells to

produce IFNg in response to M. tuberculosis antigen ESAT-6. Blood2008;111:5629–36.

23. GongG, Shao L,Wang Y, ChenCY, HuangD, Yao S, et al. Phosphoantigen-activated Vg9Vd2 T cells antagonizeIL-2-induced CD4þCD25þFoxp3þ Tregulatory cells in mycobacterial infection. Blood 2009;113:837–45.

24. Motulsky H. Choosing an appropriate sample size. Intuitive Biostatistics.1995;199–200. Oxford Univ. Press, New York.

25. Qiu P, Simonds EF, Bendall SC, Gibbs KDJr, Bruggner RV, LindermanMD,et al. Extracting a cellular hierarchy from high-dimensional cytometry datawith SPADE. Nat Biotechnol 2011;29:886–91.

26. Dieli F, Poccia F, Lipp M, Sireci G, Caccamo N, Di Sano C, et al. Differ-entiation of effector/memory Vd2 T cells and migratory routes in lymphnodes or inflammatory sites. J Exp Med 2003;198:391–7.

27. Caccamo N, La Mendola C, Orlando V, Meraviglia S, Todaro M, Stassi G,et al. Differentiation, phenotype, and function of interleukin-17-produc-ing human Vg9Vd2 T cells. Blood 2011;118:129–38.

28. Quezada SA, Peggs KS, Simpson TR, Allison JP. Shifting the equilibrium incancer immunoediting: from tumor tolerance to eradication. Immunol Rev2011;241:104–18.

29. Mahmoud SM, Lee AH, Paish EC, Macmillan RD, Ellis IO, Green AR.Tumour-infiltrating macrophages and clinical outcome in breast cancer.J Clin Pathol 2012;65:159–63.

30. Buccheri S, Guggino G, Caccamo N, Li Donni P, Dieli F. Efficacy and safetyof gd T cell-based tumor immunotherapy: a meta-analysis. J Biol RegulHomeost Agents 2014;28:81–90.

31. Kunzmann V, Bauer E, Feurle J, Weissinger F, Tony HP, Wilhelm M.Stimulation of gd T cells by aminobisphosphonates and inductionof anti plasma cell activity in multiple myeloma. Blood 2000;96:384–92.

32. Corvaisier M, Moreau-Aubry A, Diez E, Bennouna J, Mosnier JF, Scotet E,et al. Vg9Vd2 T cell response to colon carcinoma cells. J Immunol2005;175:5481–8.

33. Belmant C, Decise D, Fournie JJ. Phosphoantigens and aminobispho-sphonates: new leads targeting gd T lymphocytes for cancer immunother-apy. Drug Discov Today Ther Strateg 2006;3:17–23.

34. Liu Z, Guo BL, Gehrs BC, Nan L, Lopez RD. Ex vivo expanded humanVg9Vd2 gdþ T cells mediate innate antitumor activity against humanprostate cancer cells in vitro. J Urol 2005;173:1552–6.

35. Viey E, Fromont G, Escudier B, Morel Y, Da Rocha S, Chouaib S, et al.Phosphostim-activated gd T cells kill autologous metastatic renal cellcarcinoma. J Immunol 2005;174:1338–47.

36. Kabelitz D, Wesch D, Pitters E, Z€oller M. Characterization of tumorreactivity of human Vg9Vd2 gd T cells in vitro and in SCID mice in vivo.J Immunol 2004;173:6767–76.

37. D'Asaro M, La Mendola C, Di Liberto D, Orlando V, Todaro M, Spina M,et al. Vg9Vd2 T lymphocytes efficiently recognize and kill zoledronate-sensitized, imatinib-sensitive, and imatinib-resistant chronicmyelogenousleukemia cells. J Immunol 2010;184:3260–8.

38. Gentles AJ, Newman AM, Liu CL, Bratman SV, Feng W, Kim D, et al. Theprognostic landscape of genes and infiltrating immune cells across humancancers. Nat Med 2015;21:938–45.

39. Ma S, Cheng Q, Cai Y, Gong H, Wu Y, Yu X, et al. IL-17A produced by gd Tcells promotes tumor growth in hepatocellular carcinoma. Cancer Res2014;74:1969–82.

40. Rei M, Goncalves-Sousa N, Lanca T, Thompson RG, Mensurado S,Balkwill FR, et al. Murine CD27(�) Vg6(þ) gd T cells producing IL-17A promote ovarian cancer growth via mobilization of protumorsmall peritoneal macrophages. Proc Natl Acad Sci U S A 2014;111:e3562–70.

41. Serre K, Silva-Santos B.Molecularmechanisms of differentiation ofmurinepro-inflammatory gd T cell subsets. Front Immunol 2013;5:431.

42. Schmolka N, Wencker M, Hayday AC, Silva-Santos B. Epigeneticand transcriptional regulation of gd T cell differentiation: program-ming cells for responses in time and space. Semin Immunol2015;27:19–25.

43. Schmolka N, Serre K, Grosso AR, Rei M, Pennington DJ, Gomes AQ,et al. Epigenetic and transcriptional signatures of stable versus plasticdifferentiation of proinflammatory gd T cell subsets. Nat Immunol2013;14:1093–100.

44. Silva-Santos B, Serre K, Norell H. gd T cells in cancer. Nat Rev Immunol2015;15:683–91.






2017;5:397-407. Published OnlineFirst March 28, 2017.Cancer Immunol Res Elena Lo Presti, Francesca Toia, Sebastiano Oieni, et al.

Depending on the Tumor StageγIFN T Cells Producing either IL17 orδγSquamous Cell Tumors Recruit

Updated version

10.1158/2326-6066.CIR-16-0348doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerimmunolres.aacrjournals.org/content/suppl/2017/03/28/2326-6066.CIR-16-0348.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerimmunolres.aacrjournals.org/content/5/5/397.full#ref-list-1

This article cites 43 articles, 16 of which you can access for free at:

Citing articles

http://cancerimmunolres.aacrjournals.org/content/5/5/397.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/5/5/397To 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-16-0348

http://cancerimmunolres.aacrjournals.org/content/suppl/2017/03/28/2326-6066.CIR-16-0348.DC1

http://cancerimmunolres.aacrjournals.org/content/5/5/397.full#ref-list-1

http://cancerimmunolres.aacrjournals.org/content/5/5/397.full#related-urls

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

mailto:[email protected]

http://cancerimmunolres.aacrjournals.org/content/5/5/397


squamous cell tumors recruit gd t cells producing either ......e-mail: [email protected]...

Documents