TGFbProgramsCentralMemoryDifferentiationin Ex Vivo ... · (stem cell memory: Tscm; central memory:...

Research Article

TGFb Programs Central Memory Differentiation inEx Vivo--Stimulated Human T CellsAmina Dahmani1, Val�erie Janelle1, C�edric Carli1, Manon Richaud1,Caroline Lamarche1,2,3, Myriam Khalili1, Mathieu Goupil1, Ksenia Bezverbnaya4,Jonathan L. Bramson4, and Jean-S�ebastien Delisle1,5,6

Abstract

The adoptive transfer of ex vivo--expanded T cells is apromising approach to treat several malignancies. Severallines of evidence support that the infusion of T cells withearly memory features, capable of expanding and persistingafter transfer, are associated with better outcomes. Wereport herein that exposure to exogenous TGFb duringhuman T-cell stimulation ex vivo leads to the accumulationof early/central memory (Tcm) cells. Exposure to TGFbsuppressed the expression of BLIMP-1, a key orchestratorof effector T-cell differentiation, and led to the upregula-tion of the memory-associated transcription factor ID3.Accordingly, this was associated with an early memorytranscriptional signature in both CD4þ and CD8þ T-cellsubsets. The T cells stimulated in the presence of TGFb

expanded normally, and displayed polyfunctional featuresand no suppressive activity. The adoptive transfer ofex vivo--stimulated T cells into immunodeficient mice con-firmed that TGFb-conditioned cells had an enhancedcapacity to persist and mediate xenogeneic graft-versus-host disease, as predicted by their early T-cell memoryphenotype. Chimeric antigen receptor--expressing T cellsgenerated in the presence of exogenous TGFb were cytotoxicand more effective at controlling tumor growth in immu-nodeficient animals. This work unveils a new role for TGFbin memory T-cell differentiation and indicates that TGFbsignaling may be harnessed to program Tcm differentiationin the context of ex vivo T-cell stimulation for adoptiveimmunotherapy in humans.

IntroductionThe integration of stimulatory signals during T-cell activation

programs the differentiation of effector and memory T cells.According to the progressive differentiation model, T cells differ-entiate depending on the nature and strength of activation signalsfollowing a one-way linear path from na€�ve to early memory(stem cell memory: Tscm; central memory: Tcm), effector mem-ory (Tem), and finally terminally differentiated effector T cells(Teff; ref. 1). Hence, the gradual acquisition of effector featuresresulting from "strong" activation signals is associated with adecreased potential for long-term memory T-cell generation andpersistence. Although challenged by evidence supporting thepossibility to revert from effector to long-lived memory T cells

(2), a consensus in T-cell--adoptive immunotherapy is to useex vivo--expanded "early memory" Tcm and Tscm capable ofproliferating and persisting in vivo after transfer (3--6). Thus, theuse of various cytokine combinations, the alteration of metabolicpathways, and the modulation of signaling cascades involved inT-cell memory or effector fate determination are widely pursuedto confer Tcm or Tscm characteristics to ex vivo--manipulatedT cells for therapy (3--8).

The cytokine TGFb has pleiotropic effects in the hematopoieticsystem (9, 10). Although primarily known for its immunoregu-latory and antiproliferative properties, TGFb orchestrates bothregulatory T-cell (Treg) and inflammatory subset differentiationdepending on the presence of additional signals (11--14). Thispleiotropy is further exemplified by the contrasting prosurvivaleffects of TGFb on na€�ve and memory T cells, and the pro-apoptotic and functional inhibitory effects on differentiatedTeff (15--17). The role of TGFb in memory T-cell differentia-tion remains incompletely understood, but given the potentialof this cytokine to mitigate T-cell activation signals (18, 19), onemay expect that TGFb exposure during T-cell activation mayfavor early memory differentiation.

We show herein that exogenous TGFb exposure duringhuman T-cell stimulation ex vivo favored Tcm differentiation.In the presence of TGFb, the transcriptional regulator of Teffdifferentiation, BLIMP-1, was suppressed, and ID3, a masterregulator of T-cell memory differentiation, was induced, cor-relating with an early T-cell memory transcriptional signature(20, 21). The T cells generated in TGFb-supplemented condi-tions expanded normally and displayed increased polyfunc-tional cytokine secretion relative to unexposed T cells inkeeping with early memory differentiation (1). The adoptive

1Centre de Recherche de l'Hôpital Maisonneuve-Rosemont (CRHMR), Montreal,Quebec, Canada. 2Department of Surgery, University of British Columbia,Vancouver, British Columbia, Canada. 3BCChildren'sHospital Research Institute,Vancouver, British Columbia, Canada. 4Department of Pathology and MolecularMedicine, McMaster University, Hamilton, Ontario, Canada. 5D�epartement deM�edecine, Universit�e de Montr�eal, Montreal, Quebec, Canada. 6Hematology-Oncology Division, Hôpital Maisonneuve-Rosemont, Montreal, Quebec, Canada.

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

Corresponding Author: Jean-S�ebastien Delisle, Hôpital Maisonneuve-Rosemont-CIUSSS-EMTL, 5415 Boulevard de l' Assomption, Montr�eal, QCH1T 2M4, Canada. Phone: 514-252-3400, ext. 6381; Fax: 514-252-3430;E-mail: [email protected]

Cancer Immunol Res 2019;7:1426--39

doi: 10.1158/2326-6066.CIR-18-0691

�2019 American Association for Cancer Research.

CancerImmunologyResearch

Cancer Immunol Res; 7(9) September 20191426

on June 18, 2021. © 2019 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from

Published OnlineFirst July 15, 2019; DOI: 10.1158/2326-6066.CIR-18-0691

http://crossmark.crossref.org/dialog/?doi=10.1158/2326-6066.CIR-18-0691&domain=pdf&date_stamp=2019-8-13http://crossmark.crossref.org/dialog/?doi=10.1158/2326-6066.CIR-18-0691&domain=pdf&date_stamp=2019-8-13http://cancerimmunolres.aacrjournals.org/

transfer of activated T cells in immunodeficient mice revealedthat TGFb exposure in culture conferred an enhanced capacityto expand, persist, and mediate xenogeneic graft-versus-hostdisease (GVHD), as previously reported for cells with Tscmfeatures (3). Similarly, chimeric antigen receptor (CAR)--modifiedT cells transduced and expanded in TGFb-supplementedculture were more effective at controlling tumor growthin vivo. Hence, the TGFb pathway can be used to program earlymemory differentiation in human T cells and has, therefore,immediate relevance for the field of adoptive immunotherapy.

Materials and MethodsEx vivo T-cell cultures, proliferation, and apoptosis assays

Peripheral bloodmononuclear cells (PBMC) were obtained byvenipuncture after informed consent from a total of 39 healthyvolunteers in accordance with institutional policies. A total of1 � 105 T cells (enriched using the Human T cell EnrichmentKit, StemCell Technologies) were stimulated in 96-well U-bottomculture plates with plate-bound anti-CD3e (5 mg/mL; UCHT1; BDBiosciences,) and soluble anti-CD28 (1 mg/mL; CD28.2; BDBiosciences) in T-cell media (Advanced RPMI1640, Gibco;10% human serum and 1� L-glutamine, Sigma-Aldrich) withPenicillin (100 U/mL)--Streptomycin (100 mg/mL) Solution(Sigma-Aldrich) at 37�Cand5%CO2.Recombinant humanTGFb(active form) was used at 5 ng/mL (Feldan). IL7 (10 ng/mL), IL15(5 ng/mL; Miltenyi Biotec), IL2 (100 U/mL; Stem Cell Techno-logies), and the type I and II TGFb receptor kinase inhibitorGW788388 (2.5 mmol/L; Selleckchem; ref. 22) were used as indi-cated. At days 3, 7, and 11, half of the media was replacedwith fresh media alone or fresh media containing TGFb orGW788388.

For pathogen-specific T-cell line generation, 1 � 106 PBMCsfrom healthy volunteers were cocultured in 24-well flat-bottomculture plates with autologous irradiated (40 Gy) mature den-dritic cells (DC) obtained as described in ref. 23 by monocyteisolation (plastic adherence method), culture in IL4- andGM-CSF--supplemented media for 7 days, and maturation for48 hours following the addition of IL6, IL1b, TNFa, PGE2, andIFNg . Mature DCs were loaded with a peptide library (1 mg/mL)consisting of 15-mers, overlapping by 11 amino acids and cov-ering the entire Epstein--Barr nuclear antigen-1 (EBNA1) proteinsequence (JPT peptides). Antigen-loaded DCs were coculturedwith PBMC at a 1:10 ratio (stimulator:responder) in T-cell mediasupplemented with IL7 (10 ng/mL), IL15 (5 ng/mL), and TGFb(5 ng/mL) when indicated. Restimulation of T cells with antigen-loaded DCs was performed weekly, and half-media changes wereperformed twice a week.

To test the specific reactivity of EBNA1-specific T-cell lines,ELISpot assays were performed at the end of the culture (28 days)using the Human IFNg ELISpot PLUS Kit (HRP) ELISpot Assays(Mabtech, Inc.), and corresponding spot-forming cells and activ-ity per 105 cells were counted using a vSpot Reader Spectrum(AID) according to the manufacturer's instructions. Viability andapoptosis were evaluated by double staining of 105 cells withAnnexin V (AV; BD Biosciences) and propidium iodide (PI;2.5 mg/mL; Invitrogen) after 7 and 14 days of culture with anti-CD3/CD28 stimulation (performed as described above). Stainingfor AV was performed in AV binding buffer (10 mmol/L HEPESpH 7.4, 140 mmol/L NaCl, 2.5 mmol/L CaCl2) for 15 minutes atroom temperature. PI was added to the cell suspension right

before flow cytometry assessment of the percentage of viable(AV�PI�), apoptotic (AVþ/PI�), and dead (AVþPIþ, AV�/PIþ)cells. Proliferation and viability datawere acquired using an LSR IIFlow Cytometer (BD Biosciences) and analyzed with DIVAVersion 8.7 Software (BD Biosciences). For proliferation assays,105 T cells were labeled with 1 mmol/L CellTrace Violet (CTV;Invitrogen) prior to the cultures with anti-CD3/CD28 as describ-ed previously (24) and according to the manufacturer's instruc-tions. Dye dilution was assessed by flow cytometry as above.

ImmunophenotypingAntibodies targeting the following antigens (name of clone in

parenthesis) were purchased from BD Biosciences: CD3 (SK7),CD4 (RPA-T4), CD8 (RPA-T8), CD45RA (5H9), CD45RO(UCHL1), CD95 (DX2), CCR7 (150503), CD62L (DREG-56),CD27 (M-T271), CD28 (CD28.2), CD127 (HIL-7F-M21),IL2 (MQ1-17H12), IFNg (4S.B3), TNFa (MAb11), and FOXP3(PCH101). The anti-CD271 (ME204) antibody was purchasedfrom BioLegend. Cell surface staining was performed by incub-ating up to 106 cells with the antibodies in PBS supplementedwith 2% FBS in the dark at 4�C for 30 minutes. When required,intracellular staining was performed in the dark at 4�C for30 minutes after staining for cell surface antigens, followedby permeabilization using the Cytofix/Cytoperm Buffer (BDBiosciences) as per the manufacturer's instructions. For func-tional analysis, 105 T cells were restimulated with phorbol12-myristate 13-acetate (PMA; 50 ng/mL), ionomycin(500 ng/mL; Sigma-Aldrich), and anti-CD28 and anti-CD49d(1 mg/mL; BD Biosciences) in the presence of brefeldin A at7.5 mg/mL (Sigma-Aldrich) for 4 hours prior to cell surface andintracellular staining. All data were acquired on a LSR II flowcytometer, and sorting was performed using a FACS Aria IIISorter (BD Biosciences). Data were analyzed with DIVA Version8.7 Software (BD Biosciences), Flow Logic Software (InivaiTechnologies), or FlowJo V10 Software (Tree Star).

Western blottingAn equal number of cells (5 � 105) per condition were

used. Protein extraction was done in Laemmli Lysis buffer andresolved on 10% SDS-PAGE gels and transferred to polyvinyli-dene difluoride membranes. Anti-phospho SMAD3 (EP823Y;1:2,000; Abcam), anti-SMAD3 (EP568Y; 1:2,500; Abcam), anti-ID3 (D16D10; 1:700; Cell Signaling Technology), anti--BLIMP-1(C14A4; 1:700; Cell Signaling Technology), anti--b-actin (Ab-5;C4/actin; BD Biosciences), and Goat Anti-Rabbit IgG (H þ L)-HRP--conjugated secondary antibody (1:3,000; Bio-Rad, catalogno. 1706515) were used. The Amersham ECLTMWestern BlottingDetection Reagent (GE Healthcare) was used for revelation.

Gene expression studies and Treg-specific demethylationregion analysis

T-cell subset qPCR (RT-PCR) analyses were performed asdescribed previously (7). The RNA from sorted cells (FACSARIAIII, BD Biosciences) was extracted using the RNeasy MicroKit (Qiagen) according to the manufacturer's instructions. AfterDNase treatment (Ambion, Life Technologies), 0.5 mg of RNAwasreverse transcribed with randomprimers using the High-CapacitycDNA Reverse Transcription Kit (Life Technologies). ID3 andPRDM1 gene expression was determined by qRT-PCR, in dupli-cate, using TaqMan Gene Expression Assays (Thermo FisherScientific, catalog no. 4331182; ID3 hs00171409, PRDM1

Early Memory T-cell Differentiation by TGFb

www.aacrjournals.org Cancer Immunol Res; 7(9) September 2019 1427



http://cancerimmunolres.aacrjournals.org/

hs00153357), and TaqMan Fast Advanced Master Mix (LifeTechnologies) performed according to themanufacturer's instruc-tions. The Viia7 qPCR Instrument (Life Technologies) was usedto detect amplification. Relative expression (RQ ¼ 2�DDCt) wascalculated using the Expression Suite Software (Life Techno-logies), and normalization was done using GAPDH (ThermoFisher Scientific, catalog no. 4331182; GAPDH hs03929097).

For next-generation sequencing, all reagents and devices werefromThermo Fisher Scientific. Sampleswere prepared using 35ngtRNA that was reverse transcribed using the SuperScriptVILOcDNA Synthesis Kit and amplified (Ion Ampliseq TranscriptomeHuman Gene Expression Core Panel targeting 20812 RefSeqgenes). Libraries were generated and barcoded using Ion XpressBarcode Adapters 1-16 kit, quantified with the Ion LibraryQuantitation Kit, and loaded on the Ion Chef instrument (IonPI Hi-Q Chef Kit) for template preparation and loaded on IonP1 v3 chips. Each sample was sequenced and aligned to thehuman reference genome (hg19) using the Torrent Suite softwarev5.0.5. Raw read counts were generated by the ampliSeqRNAplugins. The data were analyzed using Bioconductor packages(http://www.bioconductor.org/) and R statistical language(www.r-project.org). Differentially expressed gene analysis wasperformed using the DESeq2 package version 1.6.3 with raw readcounts from AmpliSeq. The data can be found in the GeneExpression Omnibus (https://www.ncbi.nlm.nih.gov/geo, acces-sion number GSE87508). The T-cell--specific demethylationregion (TSDR) analysis for FOXP3 promoter methylation statuswas performed on male T cells as described previously (25).Briefly, after DNA bisulfite conversion (EZ DNA Methylation-Direct Kit, Zymo Research) and PCR (Human FOXP3 Kit, EpigenDX), pyrosequencing was performed on a Biotage PyroMarkQ96 MD Pyrosequencer (Qiagen) and analyzed with the PyroQ-CpG software (Biotage).

Cytokine quantification and T-cell--suppression assayCytokine quantification in 25 mL of undiluted culture super-

natants harvested at day 7 of anti-CD3/CD28 stimulation (asdescribed above) was performed as described previously (7)using the Human Th1/Th2/Th17 Magnetic 8-Plex Panel (LifeTechnologies). Data were acquired on a MagPix Instrument(Bio-Rad).

T-cell--suppression assays used anti-CD3/CD28--stimulatedor sorted peripheral blood Treg cells (CD4þCD25highCD127�;suppressors) added to a mixed lymphoid reaction whereCTV-labeled responder T cells (autologous to cultured cells)were mixed with allogeneic irradiated PBMCs (stimulator) ata 1:1:8 ratio (stimulator:responder:suppressor). After 6 days,the proliferation of responder T cells was analyzed by flowcytometry.

In vivo models of GVHD and adoptive immunotherapy inNOD/SCID/IL2Rgnull mice

For GVHD induction, 8- to 12-week-old NOD/SCID/IL2Rgnull

(NSG) mice (Jackson Laboratory) received total body irradiation(250 cGy) 1 day prior to intravenous injection of 0.5� 106 T cellsthat had been stimulated for 7 days with anti-CD3/CD28 in thepresence or absence of TGFb (5 ng/mL). Human recombinantrhIL15 (1 mg, 2,000 U; Miltenyi Biotec) was administered intra-peritoneally every 2 to 3 days for 3 weeks after transplant.Monitoring of circulating human cells was performed weekly bysubmandibular bleeding, cell surface staining, andflowcytometry

as described above. When indicated, cellular suspensions fromspleens were obtained by mechanical dissociation of the spleenperformed in RPMI media (Gibco).

For evaluation of antitumor responses, second-generationB-cell maturation antigen (BCMA)--specific CAR (described inref. 26) were transduced by lentivirus on day 2 of CD3/CD28stimulation (TransAct, Miltenyi Biotec) in the presence of IL7(10 ng/mL) and IL15 (5 ng/mL)� TGFb (5 ng/mL) and expandedfor 7 days prior to adoptive transfer or cytotoxicity assays.To evaluate the specific cytotoxicity of BCMA-CART cells treatedor not with TGFb, human BCMA-expressing KMS11 cells wereused. The KMS-11 cells were provided by Jonathan Bramson(McMaster University, Hamilton, Ontario, Canada) and wereoriginally obtained from the ATCC and Kelvin Lee (Roswell ParkComprehensive Cancer Institute, Buffalo, NY). Jurkat cells(ATCC) were likewise provided by Jonathan Bramson (McMasterUniversity, Hamilton, Ontario, Canada) and used as non--BCMA-expressing negative controls in our in vitro assays. The cell lineswere not further authenticated. The cell lines were engineeredwith a transgene for enhanced firefly luciferase expression andallowed for puromycin selection (27). Both cell lines weretested for Mycoplasma (PlasmoTest Mycoplasma Detection Kitfrom InvivoGen) and were used after 7 to 13 passages. Celllines were cultured in RPMI media (Gibco) supplemented with10% FBS, 2 mmol/L L-glutamine (Sigma-Aldrich), penicillin(100 U/mL)--streptomycin (100 mg/mL; Sigma-Aldrich), andpuromycin (8 mg/mL; InvivoGen) at 37�C and 5% CO2.

For cell cytotoxicity assays, KMS-11 and Jurkat cells werelabeled with either CTV or CellTrace Yellow (CTY; Invitrogen LifeTechnologies) and plated at equal numbers (2.5 � 104) beforecoculture with BCMA-CART cells in a U-bottom 96-well plate(Sarstedt) at indicated effector:target ratio for 16 hours at 37 �C.After coculture, tumor cell viability was determined by flowcytometry using Flow Count Beads (Beckman Coulter). Tumorcell viability was calculated as: (100 � (Target cell alive/Targetcell alone) � 100).

To assess the in vivo antitumor response of TGFb-treatedCART cells, a total of 2 � 105 CART cells (approximately 2 �106 total T cells) were injected into NSG mice bearing enhancedfirefly luciferase--expressing KMS-11 cells. A total of 106 KMS-11cells were injected intravenously 7 days prior to adoptive transfer.To delay the occurrence of GVHD, recipient mice did not receiveirradiation. Bioluminescent imaging was performed weekly10 minutes after the intraperitoneal injection of 150 mg/kgof fresh sterile D-Luciferin Solution (PerkinElmer). Dorsal andventral views were obtained using an IVIS Spectrum (IVIS100IVIS Lumina System, Caliper LifeSciences). Images were analyz-ed using Living Image Software version 4.5 for Windows(PerkinElmer). Peripheral blood sampling and flow cytometryassessments were performed weekly (as described above) tomonitor CART cell persistence, and recipients were sacrificed atday 35 following tumor inoculation. Mice were maintained in aspecific pathogen-free environment. The study was approved inaccordance with the Canadian Council on Animal Care guidelinesat Hôpital Maisonneuve-Rosemont (Montreal, Quebec, Canada;protocol #2016-FE-002).

Statistical analysisUnless otherwise specified, all statistical analyses were paired

to best assess the impact of test conditions for every donor andperformed using the Wilcoxon signed-rank test or Student t test

Dahmani et al.

Cancer Immunol Res; 7(9) September 2019 Cancer Immunology Research1428



http://www.bioconductor.org/www.r-project.orghttps://www.ncbi.nlm.nih.gov/geohttp://cancerimmunolres.aacrjournals.org/

depending on data distribution (assessed by the Shapiro--Wilktest) using the IBM SPSS statistics 21 software or R statisticallanguage (www.r-project.org). P < 0.05 was considered significant.

ResultsExposure to exogenous TGFb during T-cell activation leads toTcm accumulation

Total T cells from healthy donor volunteers were stimulatedwith plate-bound anti-CD3 and soluble anti-CD28 in the pre-sence or absence of exogenous TGFb. The expression of CD45ROand CD62L on both CD4þ and CD8þ T cells was used to assessfor central memory (Tcm; CD45ROþCD62Lþ) and effectormem-ory (Tem; CD45ROþCD62L�) differentiation. T cells were stim-ulated and incubated in parallel with the TGFb receptor I and IIkinase inhibitor GW788388 (22) given previous reports showingthat autocrine TGFb can have important effects on T-cell activa-tion and differentiation (Fig. 1A--C; refs. 12, 28). The addition ofTGFb rapidly induced, whereas GW788388 suppressed, the phos-phorylation of the canonical TGFb mediator SMAD3 (Supple-mentary Fig. S1). After 7 days in the culture, nearly all cells couldbe identified as Tcm or Tem (Fig. 1A and B). The percentage ofTcm was significantly increased, and reciprocally, the percentageof Tem was decreased upon exposure to exogenous TGFb forbothCD4þT andCD8þT cells at day7. The samepatternpersistedafter 14 days in culture, reaching statistical significance in CD4þ

T cells. Likewise, CCR7, the other widely used Tcm marker, wasexpressed by a significantly greater percentage of CD4þ andCD8þ

T cells after TGFb exposure at both day 7 and day 14 (Fig. 1A andC). Conversely, the inhibition of autocrine TGFb signaling withGW788388 favored Tem over Tcm accumulation. The impactof TGFb supplementation or signaling blockade on the expres-sion of other T-cell memory markers (CD27, CD28, and CD127)was also performed (Supplementary Fig. S2). A statisticallysignificant difference was noted for the proportion of CD27-expressing T cells (lower in the TGFb condition), and a slightlylower proportion of GW388788-treated T cells expressed CD127and CD28 (CD8þ T cells only).

We next assessed whether TGFb exposure could favor Tcmmarker expression on different T-cell populations. To this end,we sorted na€�ve (CD45RAþCD45RO�CD62LþCCR7þCD95�),bulk memory (CD45ROþCD45RA�), as well as Tem (CD45ROþ

CD45RA�CD62L�CCR7�) and Teff (CD45RO�CD45RAþ

CD62L�CCR7�) CD4þ and CD8þ T cells. For both na€�ve CD4þ

and CD8þ T cells, TGFb exposure during stimulation yieldeda higher percentage of Tcm relative to the control condition(Fig. 1D and E). The same observation was made following thestimulation of sorted CD4þ and CD8þ bulk memory T-cellpopulations (Fig. 1D and E). Autocrine TGFb signaling blockadehad no effect. Sorted Tem mostly kept their phenotype afterstimulation irrespective of TGFb exposure, whereas sorted andstimulated Teff mainly died or reverted to a Tem phenotype(CD8þ T cells from 2/4 donors) in both TGFb-supplementedand control conditions (Supplementary Fig. S3). Globally, thesedata implied that TGFb exposure favored Tcm differentiation inhuman T-cell cultures through effects on na€�ve and preexistingearly memory T cells.

We then assessed whether exogenous TGFb exposure wouldlikewise confer Tcm features to T cells expanded using natural-ly occurring antigens. To this end, we expanded EBNA1-specificT cells ex vivo (29). Four weekly T-cell stimulations with auto-

logous DCs loaded with peptide libraries were performed inthe presence of cytokines previously shown to favor earlymemoryT-cell differentiation (IL7 and IL15; ref. 3). SupplementationwithTGFb during the first 2 weeks increased the proportion of CD4þ

and CD8þ T cells expressing CCR7 at day 14 by roughly 20%(Supplementary Fig. S4). Despite repeated antigen stimulationsin the absence of exogenous TGFb beyond day 14, CCR7þ T cellsand CD62L-expressing CD4þ T cells were found in higher pro-portions in T-cell lines previously exposed to TGFb at day 28. Asopposed to T cells stimulated with anti-CD3/CD28, antigen-stimulated T cells in the presence of TGFb expressed CD27 in ahigher proportion relative to unexposed T cells, but no effectswere found forCD127 andCD28expression. Finally, the additionof TGFbdid not limit the generationof antigen-reactive T cells.Weconclude that early TGFb exposure during T-cell stimulation witheither anti-CD3/CD28 or antigenic peptides globally favors Tcmmarker expression.

TGFb confers an early memory gene expression signature toex vivo--stimulated T cells

It was previously found that TGFb suppresses BLIMP-1(encoded by PRDM1), a central regulator of Teff differentiation(30, 31). The transcriptional repressor BLIMP-1 inhibits memorydifferentiation, notably through TBX21 induction (32) and therepression of ID3, a key memory-associated transcription fac-tor (21). Given the limited effect of autocrine TGFb signalinginhibition in our system, we focused on the impact of TGFbexposure on the expression of BLIMP-1 and ID3 in anti-CD3/CD28--stimulated T cells. Early after stimulation (72 hours),qPCR revealed that the TGFb-exposed cells expressed lowerPRDM1 (coding for BLIMP-1) and tended to have higher ID3transcript levels relative to cells unexposed to TGFb (Fig. 2A andB). This translated to significant differences in protein levelsfor both BLIMP-1 and ID3 (Fig. 2C and D). RNA sequencingperformed on sorted CD4þ and CD8þ T cells after 7 days ofculture linked this early pattern of BLIMP-1 and ID3 expressionwith the differential expression of several effector and memorygenes between the two experimental conditions (Fig. 2E; refs. 1,5, 6, 8, 20, 33, 34). Several transcripts associated with effectordifferentiation were downregulated in TGFb-exposed cells relativeto the reference condition. These included granzymes (GZM), FASligand (FASLG), STAT4, SOCS3, ID2, IFNG, and the mastertranscription factor of Teff differentiation transcripts PRDM1and TBX21. Conversely, and consistent with Tcm differentiation,TGFb exposure led to higher expression of the memory-associatedtranscription factors transcripts ID3, FOXP1, SOX4, and FOXO1.Globally, these results showed that during T-cell activation, TGFbmitigated BLIMP-1 and associated effector gene expression, andincreased ID3expression,whichwas associatedwith theexpressionof T-cell memory--associated transcripts.

TGFb does not limit T-cell expansion and favorspolyfunctionality

After establishing that TGFb favored Tcm differentiation at thephenotypic and gene expression levels, we sought to determinewhether TGFb impacted T-cell growth and function. BecauseTGFb is known to restrict cellular proliferation and promoteapoptosis of Teff (14), we next assessed whether the effects ofTGFb on differentiation marker expression were biased by arestriction in cellular expansion (35). Cell counts and flow cyto-metry analysis performed at days 3, 7, and 14 revealed that TGFb





www.r-project.orghttp://cancerimmunolres.aacrjournals.org/

A

CD45RO

CD

62L

ØTG

FβG

W78

8388

B

CD45RO

CD4

CC

R7

C

D

E

ØTGFβGW788388

0

20

40

60

80

100

Day 7

CD45RO-

CD62L+CD45RO-

CD62L-Tcm Tem

Day 14

CD45RO-

CD62L+CD45RO-

CD62L-Tcm Tem

***

***

** *

*

*

%of

CD

4+

TC

ells

0

20

40

60

80

100

Day 7

CD45RO-

CD62L+CD45RO-

CD62L-Tcm Tem

Day 14

CD45RO-

CD62L+CD45RO-

CD62L-Tcm Tem

**

**

***

**

%of

CD

8+T

Cel

ls

CD4+

0

20

40

60

80

100

Day 7 Day 14

***

***

%of

CC

R7

+

CD8+

0

20

40

60

80

100

Day 7 Day 14

***

****

%of

CC

R7

+

From naïve CD4+ T cells

0

20

40

60

80

100

*

**

%of

CD

4+

TC

ells

CD45RO -

CD62L+CD45RO -

CD62L-Tcm Tem

From naïve CD8+ T cells

0

20

40

60

80

100

*

%of

CD

8+

TC

ells

CD45RO -

CD62L+CD45RO-

CD62L-Tcm Tem

From memory CD4+ T cells

0

20

40

60

80

100

CD45RO-

CD62L+CD45RO -

CD62L-Tcm Tem

**

%of

CD

4+

TC

ells

From memory CD8+ T cells

0

20

40

60

80

100

CD45RO-

CD62L+CD45RO-

CD62L-Tcm Tem

*

*

*%of

CD

8+T

Cel

ls

Figure 1.

Exposure to exogenous TGFb favors Tcmmarker expression in activated human T cells. A, Representative staining of CD4þ T-cell differentiation on day 7 ofculture based on the expression of CD45RO and CD62L or CCR7markers. B, Percentages (data points for each donor and average represented by histograms) ofT-cell subpopulations measured on CD4þ or CD8þ T cells on days 7 and 14 of culture: Tcm, CD62LþCD45ROþ; Tem, CD62L�CD45ROþ. C, Percentage of CCR7-expressing CD4þ and CD8þ T cells on days 7 and 14 of culture. ø, no added TGFb; square, TGFb supplementation; triangle, GW788388 addition. Seven to 10different donors, eight independent experiments. D and E, Tcm and Tem profiles of sorted na€�ve (D) and bulk memory (E) CD4þ and CD8þ T cells after 7 days ofanti-CD3/CD28 stimulation and TGFb signaling modulation (6 donors from three independent experiments). Error bars, SEM. All comparisons are with thereference condition (� , P < 0.05; �� , P < 0.01; pairedWilcoxon signed-rank test).

Dahmani et al.





did not significantly affect cell accumulation or the proportion ofCD4þ andCD8þT cells (Fig. 3A andB).Dye dilutionproliferationassays and AV/PI staining showed no significant differences in

proliferation or apoptosis rates between experimental groups(Fig. 3C--E). Likewise, the addition of TGFb did not compromiseT-cell expansion relative to other commonly used cytokines for

C D

E

BA

Fol d

chan

ge ( r

elat

ive

to

refe

renc

e co

nditi

on)

ID3PRDM1

**

Rel

ativ

e de

nsity

***

+ +-+-TGFβ

αCD3/CD28

ID3 (15 kDa)

72 h 7 days

+ +-+-

**

BLIMP1 (100 kDa)

b-Actin (42 kDa)

TGFβ

ø

Fold

cha

nge

(rel

ativ

e to

re

fere

nce

cond

ition

)

0123

00.5

11.5

--ID3 BLIMP1

CD4+

Log 2

fold

cha

nge

(rel

ativ

e to

refe

renc

e co

nditi

on)

CD8+

Effector-associated genesMemory-associated genes

ID3

IL7R

IL2RB

CCR7

CD27

FOXO1

FOXP1

SOX4

KLF2

FASLG

PRDM1

TBX21

ID2

GZMA

GZMB

GZMH

GZMK

GZMM

IFNG

IL2RG

ZEB2

SOCS3

STAT4

*

*

* * * *

**

**

** ** **

** ** **

****

**

**

***

***

*** ******

***

******

***************

******

***

***

P =

0.0

52

P =

0.0

56

P =

0.0

53

P =

0.0

53

3

2

1

0

−1

−2

−3

2

1

0

−1

−2

−3

2.0

1.5

1.0

0.5

0.0

Figure 2.

TGFb confers an early-memory gene expression signature to ex vivo--stimulated T cells. PRDM1 (A) and ID3 (B) transcript levels by qPCR performed at 72 hoursof culture (4 donors, two independent experiments). Data are expressed as fold change expression in TGFb-exposed cells relative to the reference condition (noadded TGFb, ø) set at 1. One representativeWestern blot analysis performed on stimulated T cells at 72 hours (h) and 7 days (C) and compiled densitometricanalyses of ID3 and BLIMP1 protein levels at 72 hours in T cells activated in the absence or presence of TGFb (3 different donors, reference condition set at 1; D).E, Gene expression analyses in sorted CD4þ and CD8þ T cells after 7 days of culture in the presence or absence of TGFb (4 donors). Data represented asdifferential expression (log2) of indicated transcripts in TGFb-exposed T cells relative to the reference condition (no exogenous TGFb, arbitrarily set at 0).Histograms represent means and error bars represent SEM variations. Transcripts included were at least differentially expressed 1.5-fold and statisticallysignificant between the two conditions in either CD4þ or CD8þ T cells as calculated on normalized data (� , P < 0.05; �� , P < 0.01; �� , P < 0.001; DESeq2 package).






humanT-cell stimulation/expansion (IL2, IL7, and IL15) at days 7and 14 of culture (Supplementary Fig. S5), and the effects ofTGFb on Tcm proportions were preserved when combinedwith these cytokines. Thus, the impact of TGFb on Tcm differ-entiationwas largely independent of T-cell expansion, survival, orthe presence of other cytokines.

We next aimed to determine whether TGFb exposure duringT-cell stimulation impacted functionality, as TGFb can inhibit thesecretion of IFNg , TNFa, and IL2 (33, 36). However, Tcm areknown to secrete all three cytokines, with a skewing towardmonofunctional IFNg secretion upon further effector differenti-ation (1). We assessed the impact of TGFb exposure during

Figure 3.

A TGFb-mediated increase in Tcm is independentof T-cell expansion and apoptosis. Cell counts at3, 7, and 14 days of culture, initiated with 1� 105T cells (A), and proportion of CD4þ and CD8þ Tcells at 7 and 14 days of culture (10 different donorsfrom eight independent experiments) followinganti-CD3/CD28 stimulation in the presence orabsence (ø) of TGFb (B). Representative CTVdilution plots on CD4þ T cells (C) and compileddata from 3 donors showing equivalent proportionof proliferating cells across conditions for bothCD4þ and CD8þ T cells (D). E, T-cell viabilityassessed by flow cytometry using AV and PIstaining at days 7 and 14 of culture. Dead andapoptotic cells include AV�PIþ, AVþPIþ, andAVþPI� T cells. Live cells include AV�PI� T cells(3--4 different donors from two independentexperiments). Histograms and horizontal barsrepresent means, and error bars represent SEMvariations. A pairedWilcoxon signed-rank test wasperformed to compare the two experimentalconditions, and no statistically significantdifferences were found.

Dahmani et al.





Figure 4.

TGFb exposure favors polyfunctional cytokine secretion by ex vivo--stimulated T cells. A, Representative staining of intracellular IL2, TNFa, and IFNg in T cellsstimulated with PMA/ionomycin (PMA/IONO) or unstimulated (øPMA/IONO) after 7 of 14 days in culture. Numbers in the dot plots indicate the percentage ofexpressing cells. B, Ratio of cytokine-expressing CD4þ and CD8þ T cells across experimental conditions at days 7 and 14 of culture (normalized to the referencecondition—no added TGFb (ø) in 4 different donors (over two independent experiments). Representative gating for the isolation of polyfunctional T cells (C) andcompiled results in 4 donors (over two independent experiments; D). Error bars, SEM variations (� , P < 0.05; paired Student t test).






T-cell stimulation on cytokine production by intracellularflow cytometry after briefly exposing days 7 and 14 T cells toPMA/ionomycin (Fig. 4A). To correct for interdonor variability,the percentage of cytokine-producing cells was arbitrarily set at1 in the reference condition (T cells stimulated in the absenceof TGFb; Fig. 4A and B). The percentage of IFNg-, TNFa-, andIL2-producing T cells was not significantly different betweenTGFb-exposed and unexposed T cells at day 7. However, at day14, a higher fraction of TGFb-exposed CD8þ T cells were pro-ducing TNFa and IL2, but a lower percentage produced IFNgrelative to control. When we assessed for polyfunctionality, ahigher percentage of TGFb-exposed CD4þ and CD8þ T cellsexpressed all three cytokines at day 14 (Fig. 4C and D). Takentogether, our results showed that TGFb led to Tcm-associatedcytokine secretion.

TGFb exposure during T-cell activation does not generateTregs

TGFb has pleiotropic effects on T-cell differentiation, whichmay interfere with the potency of T-cell immunotherapies, notablythrough the induction of Tregs (37). Despite our data show-ing Th1/Tc1 cytokine secretion following activation in the pre-sence of TGFb, as well as previous work showing that TGFbalone does not induce the Treg fate in human T cells (38, 39),we sought to rule out Treg skewing following T-cell activation inthe presence of TGFb. Our gene expression data confirmed thatFOXP3, along with the other Treg gene transcripts, IKZF2 (codingfor Helios) and IKZF4 (coding for Eos), were upregulated in T cellsexposed to TGFb relative to the control condition (Fig. 5A).Although, FOXP3 is expressed in both activated conventionalT cells and Tregs in humans (40), it was imperative to show thatTGFb did not induce suppressive cells. A majority of anti-CD3/CD28--stimulated T cells expressed FOXP3 at the protein level after7 days of culture, followed by a decline at day 14, in both controland TGFb-supplemented cultures (Fig. 5B). TGFb exposure alone(or with IL2) also did not lead to demethylation of the FOXP3promoter, which is a hallmark of stable Treg differentiation(Fig. 5C). Suppression assays showed that the addition of acti-vated T cells from both control and TGFb-supplemented culturesat day 7 did not suppress but rather enhanced the proliferativeresponse of autologous T cells (labeled) mixed with irradiatedallogeneic targets. This contrasted with conditions where eitherunstimulated T cells or sorted Tregs (CD4þCD25þCD127�) fromthe same donors were added to the culture (Fig. 5D and E). IL10was undetectable in culture supernatants, further arguing againstTreg skewing (Fig. 5F). Finally, the cytokine arrays also failed todemonstrate significant or differential IL17 or IL9 productionin TGFb-supplemented cultures, respectively, indicating that theTGFb-dependent Th17 and Th9 T-cell subsets are not expandedin our cultures (14). Our results confirmed that the addition ofTGFb during the stimulation of T cells does not induce alter-native differentiation schemes in our system.

TGFb exposure confers an enhanced capacity to cause GVHDEarly memory T cells have a greater capacity to expand,

persist, and retain functionality following injection in immu-nodeficient mice (3, 4). To assess whether the observed Tcmbias conferred by TGFb in vitro would translate into functionalearly memory T-cell features in vivo, T cells (5 � 105) wereharvested at day 7 of anti-CD3/CD28 stimulation in theabsence or presence of TGFb and injected intravenously into

NSG mice. Polyclonal early memory T cells are predicted tocause severe xenogeneic GVHD as opposed to their moredifferentiated counterparts (3). Weekly assessments of peri-pheral blood and spleen assessments 4 weeks after adoptivetransfer showed that TGFb-conditioned T cells persisted athigher frequencies (Fig. 6A--C). This translated to weight loss(Fig. 6D) and signs of GVHD (ruffled fur, prostration) in thegroup that received TGFb-exposed cells, but not in the controlgroup, confirming that TGFb-exposed cells had retained highfunctional xeno-reactive capacity. Hence, TGFb exposure dur-ing the early stages of T-cell activation programmed T-cellpersistence, differentiation, and function up to at least 1 monthafter infusion, as expected for early memory cells (3, 41). Thesedata corroborated the in vitro phenotypic, functional, and geneexpression characteristics of ex vivo--stimulated human T cells,linking TGFb to early T-cell memory differentiation.

Ex vivo CART cell--exposure to TGFb prior to infusionimproves tumor control in vivo

On the basis of our GVHD data, where a brief exposure toTGFb during T-cell stimulation in vitro led to different outcomesseveral weeks after adoptive transfer, we evaluated whetherthe Tcm-promoting properties of TGFb could be leveragedtherapeutically. To this end, we used a myeloma xenograftmodel, where BCMA-expressing KMS-11 cells geneticallyengineered to express luciferase were injected intravenouslyinto NSG mice 1 week prior to the adoptive transfer of humanT cells expressing a BCMA-targeting CAR (BCMA-CART). Weused a combination of anti-CD3/CD28 stimulation, IL7, IL15,and TGFb (or not, in the reference condition) for transductionbased on previous data highlighting the potential of IL7/IL15combination to generate early memory T cells in this setting(42). Exogenous TGFb enhanced the proportions of Tcm gen-erated in the context of combined IL7 and IL15 supplementa-tion, and did not affect the BCMA-CAR transduction rates(Supplementary Fig. S6). After transduction and a 7-day expan-sion, cytotoxicity was assessed in vitro using BCMA-expressingKMS-11 cells and Jurkat cells as BCMA-negative controls(Fig. 7A). Differentially labeled targets and control cells werecoincubated for 16 hours with BCMA-CART effectors at differ-ent ratios. The BCMA-specific killing of target cells was robustand equivalent whether the effectors had been exposed to TGFbor not during transduction and expansion (Fig. 7B), therebyconfirming that a short exposure to TGFb did not impede theacquisition of effector functions. Adoptive transfer of BCMA-CARTs in KMS-11 tumor-bearing NSGs revealed that BCMA-CARTs exposed to TGFb in vitro prior to transfer outcompetedtheir unexposed counterparts in terms of tumor control oreradication (Fig. 7C and D; Supplementary Fig. S7 for dorsalviews). Weekly peripheral blood assessments starting at day 14after injection showed that TGFb-exposed BCMA-CARTs werealways found in greater numbers relative to unexposed BCMA-CARTs (statistically significant at day 14 after transfer; Fig. 7E).These data further support that ex vivo TGFb exposure pro-grammed T-cell fates associated with high functionality in thecontext of adoptive immunotherapy.

DiscussionIn the hematopoietic system, TGFb is a quiescence factor that

promotes stem cell as well as memory T-cell maintenance and

Dahmani et al.





na€�ve T-cell survival (10, 15, 37). TGFb has been shown todampen signaling downstream of the T-cell receptor (18, 43) andCD28 (44).We found that TGFb increased the expression of early-memory markers in activated human T cells in multiple culturesettings and improved polyfunctional cytokine production with-

out altering T-cell expansion or leading to Treg generation. Ourdata showed that a brief TGFb exposure could program adurable effect on T cells in vivo, as predicted for early memoryT cells. In the two models used, TGFb-exposed T cells ex vivo hada significant impact on so-called "hard" endpoints such as

FOXP3

IKZF2 (

HELIO

S)

IKZF4 (

EOS)

0.0

0.5

1.0

1.5

2.0

2.5

*** ***

***

Log 2

fold

cha

nge

(rel

ativ

e to

unt

reat

ed c

ontro

l)A

CD4

FOXP3

IKZF2 (

HELIO

S)

IKZF4 (

EOS)

0.0

0.5

1.0

1.5

2.0 ******

Log 2

fold

cha

nge

(rel

ativ

e to

unt

reat

ed c

ontro

l) CD8

UnstimulatedØTGFβ

FOXP3

Cou

nt

CD4 CD8B

% F

OXP

3 E

xpre

ssio

n

CD4 CD8100

80604020

0Day 7 Day 14 Day 7 Day 14

C

pTreg

+TGF

β (D1

4)

+TGF

β (D7

)

+TGF

β+IL2

+TGF

β+IL

20

20

40

60

80

100Sorted CD4+ Total T cells

SortedpTreg

%of

Met

hyla

tion

CTV Dilution

Unstimulated aCD3/CD28TRegø TGFβ

FSC

-A

ED

% P

rolif

erat

ing

cells

re

lativ

e to

add

ed T

reg

cond

ition

TReg Unstimulated Ø TGFβ0

2

4

6

8 **

F

pg/m

L

IL2 IL4 IL5 IL9 IL10 IL13 IL17 IFNγ

ØTGFβ

TGFβØ

Unstimulated2,000

1,500

1,000

500

0

Figure 5.

TGFb exposure during T-cell activation does not induce Treg differentiation. A,Differential FOXP3, IKZF2 (coding for HELIOS), and IKZF4 (coding for EOS)expression (log2) in sorted CD4

þ and CD8þ T cells after 7 days of culture in the presence or absence of TGFb (4 donors from two independent experiments) andrepresented as differential expression of transcript levels in TGFb-exposed T cells relative to the reference condition (no exogenous TGFb) arbitrarily set at 0.B, Representative FOXP3 expression in CD4þ and CD8þ T cells using intracellular flow cytometry and compiled data at days 7 and 14 on both CD4þ and CD8þ

FOXP3-expressing T cells (4 donors from two independent experiments). C, Regulatory TSDR pyrosequencing analysis results following analysis of the nine CpGsites within the TSDR region located from -2330 to -2263 base pairs upstream of the transcriptional start codon (ATG) of FOXP3 on peripheral blood Tregs(CD4þCD25highCD127�, pTreg) sorted from CD4þ T cells after 7 days of anti-CD3/CD28 stimulation in the presence of TGFb (5 ng/mL) and/or IL2 (100 U/mL)and total T cells from days 7 and 14 of culture (4 donors from two independent experiments).D, Representative dot plots showing CTV dilution of responderT cells 6 days following stimulation with irradiated allogeneic PBMCs. Autologous unlabeled sorted pTregs, autologous unstimulated T cells, or day 7--stimulatedT cells in the absence (ø) or presence of TGFbwere added to the culture on the first day. E,Mean percentage of proliferating cells relative to the pTreg condition(set at 1) from 3 different donors and three independent experiments at a 8:1 "suppressor":responder ratio. F, Cytokines in the supernatants of unstimulated oranti-CD3/CD28--stimulated T-cell culture in the presence or absence of TGFb, harvested at day 7 of culture (4 donors over two independent experiments). Errorbars, SEM variations (� , P < 0.05, pairedWilcoxon signed-rank test; �� , P < 0.001, DESeq2 package).






death from GVHD or antitumor responses. Globally, theseresults support that the gain in Tcm brought by TGFb exposureis sufficient to mediate clinically meaningful effects.

Our results are in line with other reports establishing a role forTGFb in early memory T-cell differentiation and maintenance inmice (15, 45). In the field of human adoptive immunotherapy, aprevious study done with tumor-infiltrating lymphocytes (TIL)provides insights that overlap with some of our conclusions(46). In this context, exogenous TGFb added to a rapid expan-sion procedure (REP) using feeder cells, anti-CD3 stimulation(OKT3), and IL2; enhanced functional antigen-specific CD8þ

T cells; and prevented terminal CD8þ T-cell effector differentia-tion without leading to Treg expansion. Given that TILs areantigen-experienced T cells presenting evidence of exhaustionprior to expansion (47), which contrasts with the steady-stateperipheral T cells that we used, this previous work supports thatthe memory-promoting effects of TGFb may be extended toseveral adoptive immunotherapy approaches. However, ourresults contrast with another study primarily done in mice thatreported that endogenous TGFb signaling blockade in CD8þ Tcells favored the expansion of Tcm cells relative to other sub-types (28). This is the reason why we tested paracrine TGFbsignaling blockade initially, which in contrast to the previousstudy, had opposite and marginal Tem-promoting effects. Suchdiscrepancy may pertain to differences in the inhibitor used and/or experimental conditions. We used enriched T cells (vs. wholePBMCs) and combined anti-CD3/CD28 stimulation (vs. anti-CD3 alone), notably to reproduce the type of stimulation used inthe CAR field. Our results following repeated antigenic pepti-de--loaded DC stimulations or culture using cytokine combina-tions corroborate the Tcm-promoting effects of TGFb in several

experimental conditions. Nevertheless, our results may not applyto adoptive immunotherapy models in mice, and optimizationsmay be required (concentration, timing of exposure, combina-tion with other cytokines) before TGFb supplementation is usedin clinical scale ex vivo T-cell manufacturing.

Our gene expression studies, along with the suppression ofBLIMP-1 and overexpression of ID3 following TGFb exposure,are consistent with previous data contrasting effector versusmemory signatures. These results linking TGFb to the molecularunderpinnings of early memory differentiation in human T cellsextend our understanding of TGFb and T-cell biology but raiseseveral questions. Future studies will be required to determinewhether TGFb mostly acts indirectly by mitigating T-cell activa-tion signals or whether a more direct relationship exists betweencanonical or noncanonical TGFb signaling and BLIMP-1 sup-pression (18, 19, 30). We identified that two memory markerscan be downregulated by TGFb (modestly for CD27 and morerobustly for KLF2). Although this had no functional consequen-ces in our assays, it suggests that TGFb may be explored toconfer specific properties to T cells. Notably, the previouslydescribed direct effect of TGFb on KLF2 expression during tis-sue-resident memory T-cell differentiation may be harnessed forimmunotherapy (14, 48).

There are significant implications of our findings for the field ofT-cell immunotherapy. The programming of earlymemory T cellsis desirable to ensure robust in vivo expansion, generation ofTeff, and self-renewal as early memory T cells leading to persis-tence of the infused cells (41). Likewise, the avoidance of Treggeneration is paramount to the success of T-cell therapies. Ourresults showed that the supplementation of TGFb alongwith anti-CD3/CD28 stimulation could achieve this balance. Our results

A

C

hCD

8

* * ***

% o

f Cel

ls

***

Peripheralblood

SpleenhCD4

**

*

B4

3

16

20

110

100

90

Wei

ght v

aria

tion

(% fr

om p

retra

nsfe

r)

7 14 21 28Days posttransfer

0

10

20

30

4050 TGFβ

øTGFβøhCD4

hCD

8Day 7 Day 14 Day 21 Day 28

48

14

35

11

5

16

12

5

47

15

28

21

27

24

27

32TGFβ

Ø

D

TGFβ

Ø

Figure 6.

TGFb exposure confers an enhancedcapacity to expand, persist, and mediatexenoreactivity after adoptive transfer inimmunodeficient mice. Representative dotplots of mouse peripheral blood at varioustime points (A) and in the spleen (B) at day28 after adoptive transfer of T cellsstimulated with anti-CD3/CD28 for 7 daysin the presence (TGFb) or absence (ø) ofTGFb. Numbers in dot plots indicate thepercentage of events in the human CD4(hCD4) and human CD8 (hCD8) gates.C,Mean percentage of human T cells in theperipheral blood and spleen at day 28 afteradoptive transfer (8 mice/group, 3different donors over three independentexperiments). D, Variation in mouseweight over time after adoptive transfer.Error bars, SEM variations (� , P < 0.05;�� , P < 0.001; unpaired Student t test).

Dahmani et al.





Figure 7.

BCMA-CARTs exposed to TGFb ex vivo are superior to control tumor growth in vivo. A, Representative staining showing specific lysis of BCMA-expressingtargets. Numbers on dot plots refer to percentages of cells in the gates. B, Compilation of four independent experiments with different donors showingequivalent lytic potential of TGFb-exposed and unexposed BCMA-CARTs at multiple target:effector ratios. C, Serial bioluminescence imaging (ventral view) ofluciferase-expressing KMS-11 cells starting 7 days after intravenous injection; on the day of adoptive transfer (day 0). Total of 7 to 8mice/group from twoindependent experiments receiving BCMA-CARTs previously exposed or unexposed to TGFb. Mice inoculated with tumor cells but that did not receive BCMA-CARTs were used as controls of tumor growth (no CART cells). Compiled ventral and dorsal luciferase output at day 35 between the two experimental groups (D)and count of BCMA-CARTs previously exposed or not (ø) to TGFb (as assessed by CD271 expression) gated on human CD3þ cells and acquired using constantacquisition time from 100 mL of peripheral blood at indicated days after adoptive transfer (E). Error bars, SEM variations (� , P < 0.05; �� , P < 0.001; unpairedStudent t test).






also support that TGFboperates according to theprogressive T-celldifferentiation model, whereby ex vivo--expanded early memoryT cells originate from a pool of T cells that have not reached a Temor Teff differentiation state. In addition to the targeted T-cellsubsets preferentially leading to Tcm accumulation followingTGFb exposure, the activation of TGFb signaling needs to becarefully timed. Although TGFb can be used to program earlymemory T cells during T-cell activation, it could also impedeTeff function and survival within infected or neoplastic micro-environments (16, 17, 33, 49, 50). Our data support the notionthat the TGFb-dependent acquisition of Tcm features duringT-cell stimulation does not prevent the generation of potentT-cell responses in vitro or after transfer in NSG mice. A briefactivation of the TGFb pathway in T cells in vitro prior to adoptivetransfer could harness the promemory properties of this mastercytokine without compromising the therapeutic potential of theinfused T cells.

Disclosure of Potential Conflicts of InterestJ.-S. Delisle reports receiving a commercial research grant from and has

ownership interest (including stocks and patents) in SpecificiT Pharma. Nopotential conflicts of interest were disclosed by the other authors.

Authors' ContributionsConception and design: A. Dahmani, J.-S. DelisleDevelopment of methodology: A. Dahmani, K. BezverbnayaAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): A. Dahmani, C. Carli, C. Lamarche, M. Khalili,M. Goupil, J.L. Bramson

Analysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): A. Dahmani, V. Janelle, C. Carli, J.-S. DelisleWriting, review, and/or revision of the manuscript: A. Dahmani, V. Janelle,C. Carli, C. Lamarche, K. Bezverbnaya, J.L. Bramson, J.-S. DelisleAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): A. Dahmani, V. Janelle, C. Carli, M. RichaudStudy supervision: A. Dahmani, J.-S. Delisle

AcknowledgmentsThe authors are grateful to the blood donors, the animal care facility

personnel, Denis-Claude Roy and Vibuthi Dave for access toNSGmice, MartineDupuis for expert flow cytometry support, Manuel Buscarlet and LambertBusque for next-generation sequencing, S�ebastien Lemieux for advice regardingthe bioinformatics analysis, Jana Gillies and Megan Levings for FOXP3 pro-moter methylation studies, as well as Nathalie Labrecque, Claude Perreault,Sylvie Lesage, and Heather Melichar for helpful discussions and revision of themanuscript. J.-S. Delisle is a Fonds de recherche du Qu�ebec -- Sant�e (FRQS)scholar, aCole FoundationEarlyCareer TransitionAward laureate and aTh�eCellnetwork member, as well as a Canadian Donation and Transplant ResearchProgram member. This work was supported by the National Science andEngineering Research Council (NSERC) of Canada through a Discovery Grant(418607-2012-RGPIN) and by a Leukemia/Lymphoma Society of Canada(LLSC) operating grant (#202379), both held by J.-S. Delisle.

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

Received September 30, 2018; revisedMarch 27, 2019; accepted July 9, 2019;published first July 15, 2019.

References1. Gattinoni L, Klebanoff CA, Restifo NP. Paths to stemness: building the

ultimate antitumour T cell. Nat Rev Cancer 2012;12:671--84.2. Youngblood B, Hale JS, Kissick HT, Ahn E, Xu X, Wieland A, et al. Effector

CD8 T cells dedifferentiate into long-livedmemory cells. Nature 2017;552:404--9.

3. Cieri N, Camisa B, Cocchiarella F, ForcatoM,Oliveira G, Provasi E, et al. IL-7 and IL-15 instruct the generation of human memory stem T cells fromnaive precursors. Blood 2013;121:573--84.

4. van der Waart AB, van de Weem NM, Maas F, Kramer CS, Kester MG,Falkenburg JH, et al. Inhibition of Akt signaling promotes the generation ofsuperior tumor-reactive T cells for adoptive immunotherapy. Blood 2014;124:3490--500.

5. SukumarM, Liu J, Ji Y, SubramanianM,Crompton JG, YuZ, et al. Inhibitingglycolytic metabolism enhances CD8þ T cell memory and antitumorfunction. J Clin Invest 2013;123:4479--88.

6. Pilipow K, Scamardella E, Puccio S, Gautam S, De Paoli F, MazzaEM, et al. Antioxidant metabolism regulates CD8þ T memorystem cell formation and antitumor immunity. JCI Insight 2018;3.pii: 122299.

7. Orio J, Carli C, Janelle V, Giroux M, Taillefer J, Goupil M, et al.Early exposure to interleukin-21 limits rapidly generated anti-Epstein-Barr virus T-cell line differentiation. Cytotherapy 2015;17:496--508.

8. Gattinoni L, Zhong XS, Palmer DC, Ji Y, Hinrichs CS, Yu Z, et al. Wntsignaling arrests effector T cell differentiation and generates CD8þmemory stem cells. Nat Med 2009;15:808--13.

9. Carli C, Giroux M, Delisle JS. Roles of transforming growth factor-beta ingraft-versus-host and graft-versus-tumor effects. Biol Blood Marrow Trans-plant 2012;18:1329--40.

10. Blank U, Karlsson S. TGF-beta signaling in the control of hematopoieticstem cells. Blood 2015;125:3542--50.

11. Nakatsukasa H, Zhang D, Maruyama T, Chen H, Cui K, Ishikawa M, et al.The DNA-binding inhibitor Id3 regulates IL-9 production in CD4(þ)T cells. Nat Immunol 2015;16:1077--84.

12. Li MO, Wan YY, Flavell RA. T cell-produced transforming growth factor-beta1 controls T cell tolerance and regulates Th1- and Th17-cell differen-tiation. Immunity 2007;26:579--91.

13. Bettelli E, Carrier Y, Gao W, Korn T, Strom TB, Oukka M, et al. Reciprocaldevelopmental pathways for the generation of pathogenic effector TH17and regulatory T cells. Nature 2006;441:235--8.

14. Dahmani A, Delisle JS. TGF-beta in T cell biology: implications forcancer immunotherapy. Cancers 2018;10. pii: E194.

15. Ma C, Zhang N. Transforming growth factor-beta signaling is constantlyshaping memory T-cell population. Proc Natl Acad Sci U S A 2015;112:11013--7.

16. Filippi CM, Juedes AE, Oldham JE, Ling E, Togher L, Peng Y, et al.Transforming growth factor-beta suppresses the activation of CD8þT-cells when naive but promotes their survival and function once antigenexperienced: a two-faced impact on autoimmunity. Diabetes 2008;57:2684--92.

17. Sanjabi S, Mosaheb MM, Flavell RA. Opposing effects of TGF-beta andIL-15 cytokines control the number of short-lived effector CD8þ T cells.Immunity 2009;31:131--44.

18. ChenCH, Seguin-DevauxC., BurkeNA,Oriss TB,Watkins SC, ClipstoneN,et al. Transforming growth factor beta blocks Tec kinase phosphorylation,Ca2þ influx, and NFATc translocation causing inhibition of T cell differ-entiation. J Exp Med 2003;197:1689--99.

19. Delisle JS, Giroux M, Boucher G, Landry JR, Hardy MP, Lemieux S, et al.The TGF-beta-Smad3 pathway inhibits CD28-dependent cell growthand proliferation of CD4 T cells. Genes Immun 2013;14:115--26.

20. Yang CY, Best JA, Knell J, Yang E, Sheridan AD, Jesionek AK, et al. Thetranscriptional regulators Id2 and Id3 control the formation of distinctmemory CD8þ T cell subsets. Nat Immunol 2011;12:1221--9.

21. Ji Y, Pos Z, RaoM, Klebanoff CA, Yu Z, Sukumar M, et al. Repression of theDNA-binding inhibitor Id3 by Blimp-1 limits the formation of memoryCD8þ T cells. Nat Immunol 2011;12:1230--7.

22. PetersenM, ThorikayM,DeckersM, vanDintherM,Grygielko ET, GellibertF, et al. Oral administration of GW788388, an inhibitor of TGF-beta type I

Dahmani et al.





and II receptor kinases, decreases renal fibrosis. Kidney Int 2008;73:705--15.

23. Janelle V, Carli C, Taillefer J, Orio J, Delisle JS. Defining novel parametersfor the optimal priming and expansion of minor histocompatibilityantigen-specific T cells in culture. J Translat Med 2015;13:123.

24. Gudmundsdottir H, Wells AD, Turka LA. Dynamics and requirements ofT cell clonal expansion in vivo at the single-cell level: effector functionis linked to proliferative capacity. J Immunol 1999;162:5212--23.

25. Hoeppli RE, MacDonald KN, Leclair P, Fung VCW, Mojibian M, Gillies J,et al. Tailoring the homing capacity of human Tregs for directed migrationto sites of Th1-inflammation or intestinal regions. Am J Transplant 2019;19:62--76.

26. Helsen CW, Hammill JA, Lau VWC, Mwawasi KA, Afsahi A, Bezverb-naya K, et al. The chimeric TAC receptor co-opts the T cell receptoryielding robust anti-tumor activity without toxicity. Nat Commun2018;9:3049.

27. Rabinovich BA, Ye Y, Etto T, Chen JQ, Levitsky HI, Overwijk WW, et al.Visualizing fewer than 10mouse T cells with an enhanced firefly luciferasein immunocompetent mouse models of cancer. Proc Natl Acad Sci U S A2008;105:14342--6.

28. Takai S, Schlom J, Tucker J, Tsang KY, Greiner JW. Inhibition of TGF-beta1signaling promotes centralmemory T cell differentiation. J Immunol 2013;191:2299--307.

29. Icheva V, Kayser S, Wolff D, Tuve S, Kyzirakos C, Bethge W, et al.Adoptive transfer of epstein-barr virus (EBV) nuclear antigen 1-specifict cells as treatment for EBV reactivation and lymphoproliferativedisorders after allogeneic stem-cell transplantation. J Clin Oncol2013;31:39--48.

30. Lin R, Chen L, Chen G, Hu C, Jiang S, Sevilla J, et al. Targeting miR-23a inCD8þ cytotoxic T lymphocytes prevents tumor-dependent immunosup-pression. J Clin Invest 2014;124:5352--67.

31. Benevides L, Costa RS, Tavares LA, RussoM,Martins GA, da Silva LLP, et al.Blimp-1 controls Th9 cell development, IL-9 production and allergicinflammation. J Allergy Clin Immunol 2018;143:1119--30.

32. Cimmino L, Martins GA, Liao J, Magnusdottir E, Grunig G, Perez RK, et al.Blimp-1 attenuates Th1 differentiation by repression of ifng, tbx21,bcl6 gene expression. J Immunol 2008;181:2338--47.

33. Thomas DA, Massague J. TGF-beta directly targets cytotoxic T cellfunctions during tumor evasion of immune surveillance. Cancer Cell2005;8:369--80.

34. Kim MV, Ouyang W, Liao W, Zhang MQ, Li MO. The transcription factorFoxo1 controls central-memory CD8þ T cell responses to infection.Immunity 2013;39:286--97.

35. Crompton JG, Sukumar M, Restifo NP. Uncoupling T-cell expansion fromeffector differentiation in cell-based immunotherapy. Immunol Rev 2014;257:264--76.

36. Das L, Levine AD. TGF-beta inhibits IL-2 production and promotescell cycle arrest in TCR-activated effector/memory T cells in the presenceof sustained TCR signal transduction. J Immunol 2008;180:1490--8.

37. Li MO, Flavell RA. TGF-beta: a master of all T cell trades. Cell 2008;134:392--404.

38. Tran DQ, Ramsey H, Shevach EM. Induction of FOXP3 expression in naivehuman CD4þFOXP3 T cells by T-cell receptor stimulation is transforminggrowth factor-beta dependent but does not confer a regulatory phenotype.Blood 2007;110:2983--90.

39. Schmidt A, Eriksson M, Shang MM, Weyd H, Tegner J. Comparativeanalysis of protocols to induce human CD4þFoxp3þ regulatory T cellsby combinations of IL-2, TGF-beta, retinoic acid, rapamycin and butyrate.PLoS One 2016;11:e0148474.

40. McMurchy AN, Gillies J, Gizzi MC, RibaM, Garcia-Manteiga JM, Cittaro D,et al. A novel function for FOXP3 in humans: intrinsic regulation ofconventional T cells. Blood 2013;121:1265--75.

41. Graef P, Buchholz VR, Stemberger C, Flossdorf M, Henkel L, SchiemannM, et al. Serial transfer of single-cell-derived immunocompetencereveals stemness of CD8(þ) central memory T cells. Immunity 2014;41:116--26.

42. Xu Y, Zhang M, Ramos CA, Durett A, Liu E, Dakhova O, et al. Closelyrelated T-memory stem cells correlate with in vivo expansion of CAR.CD19-T cells and are preserved by IL-7 and IL-15. Blood 2014;123:3750--9.

43. Boussiotis VA, Chen ZM, Zeller JC, Murphy WJ, Berezovskaya A, Narula S,et al. Altered T-cell receptor þ CD28-mediated signaling and blocked cellcycle progression in interleukin 10 and transforming growth factor-beta-treated alloreactive T cells that do not induce graft-versus-host disease.Blood 2001;97:565--71.

44. Delisle JS, GirouxM, Boucher G, Landry JR,HardyMP, Lemieux S, et al. TheTGF-beta-Smad3 pathway inhibits CD28-dependent cell growth and pro-liferation of CD4 T cells. Genes Immun 2013;14:115--26.

45. Ishigame H, Mosaheb MM, Sanjabi S, Flavell RA. Truncated form ofTGF-betaRII, but not its absence, induces memory CD8þ T cell expan-sion and lymphoproliferative disorder in mice. J Immunol 2013;190:6340--50.

46. Liu S, Etto T, Rodriguez-Cruz T, Li Y, Wu C, Fulbright OJ, et al. TGF-beta1induces preferential rapid expansion and persistence of tumor antigen-specific CD8þ T cells for adoptive immunotherapy. J Immunother 2010;33:371--81.

47. Gros A, Robbins PF, YaoX, Li YF, Turcotte S, Tran E, et al. PD-1 identifies thepatient-specific CD8(þ) tumor-reactive repertoire infiltrating humantumors. J Clin Invest 2014;124:2246--59.

48. Skon CN, Lee JY, erson KG, Masopust D, Hogquist KA, Jameson SC.Transcriptional downregulation of S1pr1 is required for the establishmentof resident memory CD8þ T cells. Nat Immunol 2013;14:1285--93.

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

50. Bollard CM, Rossig C, Calonge MJ, Huls MH,Wagner HJ, Massague J, et al.Adapting a transforming growth factor beta-related tumor protectionstrategy to enhance antitumor immunity. Blood 2002;99:3179--87.






2019;7:1426-1439. Published OnlineFirst July 15, 2019.Cancer Immunol Res Amina Dahmani, Valérie Janelle, Cédric Carli, et al. Stimulated Human T Cells

−Ex Vivo Programs Central Memory Differentiation in βTGF

Updated version

10.1158/2326-6066.CIR-18-0691doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerimmunolres.aacrjournals.org/content/suppl/2019/07/13/2326-6066.CIR-18-0691.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerimmunolres.aacrjournals.org/content/7/9/1426.full#ref-list-1

This article cites 50 articles, 18 of which you can access for free 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/7/9/1426To 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-18-0691http://cancerimmunolres.aacrjournals.org/content/suppl/2019/07/13/2326-6066.CIR-18-0691.DC1http://cancerimmunolres.aacrjournals.org/content/7/9/1426.full#ref-list-1http://cancerimmunolres.aacrjournals.org/cgi/alertsmailto:[email protected]://cancerimmunolres.aacrjournals.org/content/7/9/1426http://cancerimmunolres.aacrjournals.org/

/ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /CropGrayImages false /GrayImageMinResolution 200 /GrayImageMinResolutionPolicy /Warning /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /CropMonoImages false /MonoImageMinResolution 600 /MonoImageMinResolutionPolicy /Warning /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 900 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False

/CreateJDFFile false /Description > /Namespace [ (Adobe) (Common) (1.0) ] /OtherNamespaces [ > /FormElements false /GenerateStructure false /IncludeBookmarks false /IncludeHyperlinks false /IncludeInteractive false /IncludeLayers false /IncludeProfiles false /MarksOffset 18 /MarksWeight 0.250000 /MultimediaHandling /UseObjectSettings /Namespace [ (Adobe) (CreativeSuite) (2.0) ] /PDFXOutputIntentProfileSelector /NA /PageMarksFile /RomanDefault /PreserveEditing true /UntaggedCMYKHandling /LeaveUntagged /UntaggedRGBHandling /LeaveUntagged /UseDocumentBleed false >> > ]>> setdistillerparams> setpagedevice

TGFbProgramsCentralMemoryDifferentiationin Ex Vivo ... · (stem cell memory: Tscm; central memory:...

Documents

Transcript of TGFbProgramsCentralMemoryDifferentiationin Ex Vivo ... · (stem cell memory: Tscm; central memory:...