IL-21–Induced MHC Class II+ NK Cells Promote the Expansion Of

Home , Naive T cell

IL-21−Induced MHC Class II+ NK Cells Promote the Expansion of Human Uncommitted CD4+ Central Memory T Cells in a Macrophage Migration Inhibitory This information is current as Factor−Dependent Manner of September 23, 2021. Romain Loyon, Emilie Picard, Olivier Mauvais, Lise Queiroz, Virginie Mougey, Jean-René Pallandre, Jeanne Galaine, Patricia Mercier-Letondal, Guillaume Kellerman,

Nathalie Chaput, John Wijdenes, Olivier Adotévi, Downloaded from Christophe Ferrand, Pedro Romero, Yann Godet and Christophe Borg J Immunol published online 27 May 2016 http://www.jimmunol.org/content/early/2016/05/27/jimmun

ol.1501147 http://www.jimmunol.org/

Supplementary http://www.jimmunol.org/content/suppl/2016/05/27/jimmunol.150114 Material 7.DCSupplemental

Why The JI? Submit online. by guest on September 23, 2021

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2016 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published May 27, 2016, doi:10.4049/jimmunol.1501147 The Journal of Immunology

IL-21–Induced MHC Class II+ NK Cells Promote the Expansion of Human Uncommitted CD4+ Central Memory T Cells in a Macrophage Migration Inhibitory Factor–Dependent Manner

Romain Loyon,*,1 Emilie Picard,*,1 Olivier Mauvais,† Lise Queiroz,*,‡ Virginie Mougey,*,x Jean-Rene´ Pallandre,*,{ Jeanne Galaine,* Patricia Mercier-Letondal,*,x Guillaume Kellerman,{ Nathalie Chaput,‖ John Wijdenes,* Olivier Adote´vi,*,‡,# Christophe Ferrand,*,x Pedro Romero,** Yann Godet,* and Christophe Borg*,‡,x,#

NK cells are critical for innate immunity–mediated protection. The main roles of NK cells rely on their cytotoxic functions or Downloaded from depend on the tuning of Th1 adaptive immunity by IFN-g. However, the precise influence of inflammatory cytokines on NK cell and CD4 T lymphocyte interactions was never investigated. In this study, we provide evidence that IL-21, a cytokine produced during chronic inflammation or infectious diseases, promotes the differentiation of a specific subset of NK cells coexpressing CD86 and HLA-DR and lacking NKp44. More importantly, IL-21–propagated HLA-DR+ NK cells produce macrophage migration inhibitory factor and provide costimulatory signaling during naive CD4+ T cell priming inducing the differentiation of uncommitted central memory T cells. Central memory T cells expanded in the presence of HLA-DR+ NK cells are CXCR3+CCR62 http://www.jimmunol.org/ CCR42CXCR52 and produce IL-2, as well as low levels of TNF-a. Costimulation of CD4+ T cells by HLA-DR+ NK cells prevents the acquisition of effector memory phenotype induced by IL-2. Moreover, we identified this population of NK HLA-DR+ macrophage migration inhibitory factor+ cells in inflammatory human appendix. Collectively, these results demonstrate a novel function for IL-21 in tuning NK and CD4+ T cell interactions promoting a specific expansion of central memory lymphocytes. The Journal of Immunology, 2016, 197: 000–000.

atural killer cells are innate lymphocytes contributing to modulation of innate or adaptive immunity through the production early defenses against various pathogens (1). The role of of several cytokines (1). A growing body of evidence has confirmed by guest on September 23, 2021 N NK cells in humans was first demonstrated in immu- the implication of NK cell–activating receptors in the prevention of nocompromised patients with high incidence of HSV infections human infectious disease. Indeed, the importance of NKG2D in (2–5). Established functions of NK cells include elimination of the control of EBV diseases in humans was recently highlighted in virus-infected or cancer cells by direct cytotoxic attack and the X-linked immunodeficiency with Mg2+ defect, EBV infection, and neoplasia (6). In contrast, the precise mechanisms underpinning the tuning of *INSERM Unit 1098, University of Franche-Comte´, 25000 Besançon, France; †Department of Head and Neck Surgery, University Hospital of Besançon, 25000 adaptive immunity by NK cells remain to be further explored. Among Besançon, France; ‡Clinical Investigation Center for Biotherapies, 25000 Besançon, NK cell–derived cytokines, IFN-g promotes dendritic cell (DC) France; xEtablissement Français du Sang, 25000 Besançon, France; {INSERM Unit ‖ maturation and IL-12 production (7–10). Martıń-Fontecha et al. (11) 1007, University of Paris Descartes, 75270 Paris, France; INSERM Unit 1015, Gustave Roussy Cancer Campus, 94805 Villejuif, France; #Department of Medical showed that DC-mediated NK cell activation participates in the Oncology, University Hospital of Besançon, 25000 Besançon, France; and generation of Th1. In these experiments, NK cell depletion pre- **Ludwig Center for Cancer Research of the University of Lausanne, Lausanne + 1066, Switzerland vented the Th1 polarization of CD4 T lymphocytes following DC 1 vaccination. Moreover, NK cell reconstitution from wild-type but R.L. and E.P. contributed equally to this work. not from IFN-g–deficient mice restored the ability of DC vaccina- ORCIDs: 0000-0001-8039-7893 (R.L.); 0000-0003-4146-3565 (O.M.); 0000-0002- + 2262-0572 (V.M.); 0000-0003-2506-7853 (N.C.); 0000-0003-0119-3919 (C.F.). tiontoprimeandexpandTh1CD4 T lymphocytes (11). The role of Received for publication May 20, 2015. Accepted for publication April 18, 2016. NK cells in innate and adaptive immunity interface was further evidenced in Leishmania major infections. Bajeńoff et al. (12) This work was supported by Le Conseil Regional de Franche-Comte´, La Ligue Contre le Cancer, the LabEx LipSTIC, and by the Inflammation–Cancer and Bio- showed that L. major parasites induce NK cells to produce IFN-g therapy Network of the University of Franche-Comte´. E.P. was supported by a Ph.D. andtoberecruitedinclosevicinitytoDCandCD4+ T lymphocytes ´ grant from Besançon City, and R.L. by the Region of Franche-Comte. shortly after disease initiation. NK cell activation during Leishmania Address correspondence and reprint requests to Prof. Christophe Borg, Medical infections requires myeloid DC (13), with NK cell–derived IFN-g Oncology Unit, University Hospital of Besançon, Boulevard Alexandre Fleming, 25000 Besançon, France. E-mail address: [email protected] skewing T cell differentiation toward Th1 immune responses (14). The online version of this article contains supplemental material. Indeed, when the role of NK cell–derived IFN-g in the regu- + Abbreviations used in this article: ChIP, chromatin immunoprecipitation; DC, den- lation of CD4 T lymphocytes during infection or vaccination was dritic cell; ILC, innate lymphoid cell; MHC-II, MHC class II; MIF, macrophage well characterized, the influence of inflammatory cytokines on migration inhibitory factor; qPCR, quantitative PCR; ROR, retinoic acid–related these interactions was never clearly assessed. Even though NK orphan receptor; T , central memory T cell; T , stem cell memory T cell. CM SCM cells have been observed in cutaneous lesions of common in- Copyright Ó 2016 by The American Association of Immunologists, Inc. 0022-1767/16/$30.00 flammatory diseases such as psoriasis (15), their precise role in

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1501147 + 2 IL-21–INDUCED MHC-II NK CELLS PROMOTE TCM EXPANSION inflammation is still elusive. Recently, genetic investigations of (1G1), and CXCR5 BV510 (RF8B2) from BD Biosciences, NKp80 PE NKp30 polymorphisms, an important regulator of DC/NK cell (5D12) from BioLegend, NKG2A PE (131411) from R&D Systems, and cross-talk (9), have suggested the involvement of human NK cells HLA-DR PE (BF-1) from Diaclone. Isotype Abs corresponding for each manufacturer were used as a control. in inflammation. In this study, genetically determined NKp30 iso- The following anti-human Abs were used for confocal microscopy: MIF forms (rs11575837 and rs2736191 variants), which are associated from BioLegend, CD3 from Dako, CD56 from Cell Signaling Technology, with reduced levels of NKp30 gene transcription and function (16), HLA-DR from Thermo Scientific, and CD3 and MIF from Abcam. Alexa were correlated to a decreased susceptibility to Sjo¨gren’s syndrome Fluor 488 goat anti-rat IgG (H+L), Cy3 goat anti-rabbit IgG (H+L), and Alexa Fluor 647 goat anti-mouse IgG (H+L) from Jackson ImmunoResearch (17). Additional evidence pointing out a possible role of NK cells in Laboratories were used as secondary Abs. DAPI was used for nuclear staining inflammatory diseases was reported in the experimental acute en- (Molecular Probes). cephalopathy model. In these studies, IL-21 injections during ex- Staining and flow cytometry perimental acute encephalopathy induction, using a myelin-derived peptide sensitization, increased experimental acute encephalopathy To prepare immunostaining, cells were counted and suspended in PBS 13 severity in a NK cell–dependent manner (18, 19). (Life Technologies) before labeling with appropriate Abs during 30 min in 3 g IL-21 is an important component of CD4+ T cell help and is dark at 4˚C. Then, cells were washed with PBS 1 . For intracellular IFN- detection, cells were stimulated for 6–18 h with or without 10 mmol/l required to control CMV chronic infections (20). Along with IL-2 peptide or PMA (50 ng/ml, Roche) and ionomycin (250 ng/ml, Roche), and IL-15 receptors, IL-21 receptor is associated with the common with 1 ml/ml GolgiPlug (BD Biosciences) before staining for flow cyto- g-chain subunit, the signaling of which is involved in NK ho- metric analysis. Surface staining was performed. Cells were then fixed meostasis as evidenced by the absence of circulating NK cells in and permeabilized using Cytofix/Cytoperm buffer (BD Biosciences) before intracytoplasmic staining. Samples were acquired on a FACSCanto II (BD SCID-X1 patients (21). However, whereas IL-2 and IL-15 signaling Biosciences) and analyzed with the FACSDiva software. Assessment of STAT3 Downloaded from relies mostly on STAT5, IL-21 signals mainly through STAT3 (22, phosphorylation was performed by culturing NK cells with IL-2 or IL-2 plus 23). IL-21 and its receptor IL-21R were discovered in 2000 and IL-21 in the presence of pharmacological inhibitors (STA21, 20 mM, Enzo Life shown to play a key role in inflammation (24). IL-21 is produced by Sciences; AG490, 20 mM, Enzo Life Sciences) or DMSO. NK cells were activated CD4 T cells and NKT cells. When the role of IL-21 on NK stained using BD Cytofix fixation buffer (BD Biosciences) and BD Phosflow perm buffer III (BD Biosciences) according to the manufacturer’s protocol. cell maturation and activation was well described, its precise influ- FACS gates set on unstimulated cells (medium alone or isotype controls) were ence on NK and CD4 T cell cross-talk was never clearly deciphered. applied in standard format across all samples and all conditions. http://www.jimmunol.org/ In this study, we characterized a population of human NK cells Patients and samples induced by IL-21 and expressing HLA-DR and CD86. Moreover, MHC class II+ (MHC-II) NK cells differentiated in the presence of Patients with the clinical diagnosis of appendicitis underwent appendec- IL-21 promote the expansion of uncommitted central memory CD4 tomy at the University Hospital of Besançon or Montbe´liard. Slides were obtained from paraffin-embedded appendix removed from patients who Tcells(TCM) in a macrophage migration inhibitory factor (MIF)– underwent surgery for an appendicitis. dependent manner. Finally, we have demonstrated the presence of MIF+ NK cells in inflammatory human appendix. Confocal microscopy Cells were spread onto poly-L-lysine–coated slides (Sigma-Aldrich, Lyon, Materials and Methods France) for 25 min at 37˚C. Cells were fixed and permeabilized with 0.2% by guest on September 23, 2021 Lymphocyte culture and isolation NaDodSO4. After 20 min of blocking in 20% FBS and washing, cells were stained with the appropriate Abs and secondary conjugated Abs. Stacks of Peripheral blood was obtained from healthy donors of the French blood confocal images were collected with an Olympus FV1000 laser scanning bank (Etablissement Français du Sang Bourgogne Franche-Comte´). In- confocal microscope and FluoView 10-ASW (v4.2) software. formed consents for functional tests and genetic analyses were obtained from healthy donors and approved by the local ethic committee. PBMC Proliferation assay were isolated by density centrifugation on Ficoll (GE Healthcare). Cells Cells were labeled with 5 mM CFSE (CellTrace CFSE cell proliferation kit, were maintained in RPMI 1640 (Life Technologies) supplemented with Molecular Probes) in PBS 13 (Life Technologies) during 15 min in the 10% human serum and 100 U/ml penicillin-streptomycin (Invitrogen). dark under agitation. Then, 2 vol FBS (Life Technologies) was added and NK cells were negatively selected from PBMC using the EasySep human cells were incubated again in the dark under agitation during 10 min. Fi- NK cell enrichment kit (Stemcell Technologies) according to the manufac- nally, cells were washed twice with PBS 13 before culture. Proliferation turer’s protocol, before culture during 7 d with IL-2 (100 U/ml, Novartis) or was assessed by flow cytometry. IL-2 plus IL-21 (50 ng/ml, Shenandoah Biotechnology) renewed at day 3 and day 5. The purity of NK cells was assessed by flow cytometry with anti- NK cell degranulation assay CD3 and anti-CD56 and ranged from 94 to 98%. CD4 T cells were negatively selected from PBMC using the EasySep NK cells were cocultured in the presence of target cells for 4 h at a 10:1 E:T human CD4+ T cell enrichment kit (Stemcell Technologies) or EasySep ratio with GolgiStop (BD Biosciences) according to the manufacturer’s human naive CD4 cell isolation kit (Stemcell Technologies) according to protocol. Degranulation of NK cells was monitored by flow cytometric the manufacturer’s instructions. The purity of CD4 T cells or naive CD4 analysis of CD107a (H4A3, BD Biosciences) expression. T cells was assessed by flow cytometry with, respectively, anti-CD3 and Assessment of cytokine production anti-CD4 or anti-CD3, anti-CD4, anti-CD45RA, and anti-CD62L Abs and ranged from 94 to 98%. Memory CD4 T cells were enriched by CD45RA Human IFN-g and IL-4 production was assessed in cell culture superna- depletion using CD45RA MicroBeads (Miltenyi Biotec) according to the tants using an ELISA assay (Diaclone) according to the manufacturer’s manufacturer’s protocol. protocol. In some experiments, a proteome array to assess NK cell secretome wasperformed.Forthispurpose,NKcellswereculturedduring5dwithIL-2, Abs IL-2 plus IL-21, or with medium only. At day 7, supernatants were collected The following anti-human Abs were used for flow cytometry: CD3 Pacific and used for human cytokine array panel A (R&D Systems) according to the Blue (UCHT1), CD4 PE-Cy7 (SK3), CD16 FITC/PE (3G8), CD40L manufacturer’s protocol. allophycocyanin (TRAP1), CD45RA PE (HI100), CD56 allophycocyanin/ CD4 T cell activation PE (B159), CD62L allophycocyanin (DREG-56), CD70 PE (Ki-24), CD80 FITC (L307.4), CD86 FITC (2331 FUN-1), ICOSL PE (2D3/B7-H2), CD4 T cell activation was achieved using cultured 5 mg/ml plate-bound NKp30 PE (p30-15), NKp44 PE (p44-8.1), NKp46 PE (9E2/NKp46), anti-CD3 (HIT3a, BD Biosciences) with or without 2.5 mg/ml soluble anti- NKG2D PE (1D11), DNAM-1 PE (DX11), 2B4 PE (2-69) , IFN-g CD28 (CD28.2, BD Biosciences). In some experiments, IL-21R–Fc (human allophycocyanin (B27), TNF-a PE (MAb11), IL-2 allophycocyanin (MQ1- IgG1) (R&D Systems) was added at 12.5 mg/ml in the culture medium to 17H12), pY705 pSTAT3 PE (4/P-STAT3), pS727 pSTAT3 PE (49/p-Stat3), neutralize IL-21, and cetuximab (human IgG1) was used as a control. Abs CXCR3 allophycocyanin (1C6/CXCR3), CCR6 BV510 (11A9), CCR4 PE were renewed at days 2, 4, and 6 before analysis at day 7. The Journal of Immunology 3

Th1 polarization was performed using rIL-12 (50 ng/ml, PeproTech) and NK cells recovered from lymph nodes or splenocytes (data not anti–IL-4 (MP4-25D2, 10 mg/ml, BD Biosciences). MIF was neutralized shown). MHC-II+ NK cells were activated lymphocytes because using the pharmacological inhibitor ISO-1 (50 mM, Focus Biomolecules) the expression of CD86 and HLA-DR appeared only in prolifer- and purified monoclonal Abs (10C3, 10 mg/ml, BioLegend). MIF receptor (CD74) was neutralized using purified anti-CD74 (M-B741, 25 mg/ml, ating cells after at least 3 d of culture (Fig. 1C). Moreover, IFN-g LSBio). Isotype Abs were used as a control for anti-MIF and anti-CD74, was not involved in CD86 and HLA-DR upregulation mediated by and DMSO was used as a control for ISO-1. Inhibitors were added in IL-21 because IFN-g–neutralizing Abs did not prevent their ex- cultures at day 0 and renewed at days 2 and 4. Transwell experiments were pression (data not shown). realized with 24-well Millicell cultures (Millipore). Considering that IL-21 is produced by activated lymphocytes, Real-time quantitative PCR we decided to investigate the capacity of CD4 T cells to promote Total RNA from T cells was extracted with an RNeasy mini kit (Qiagen). CD86 and HLA-DR expression on NK cells. Accordingly, CD4 Reverse transcription was performed using high-capacity RNA-to-cDNA T cells were purified and activated using anti-CD3 and anti-CD28 polymerase (Life Technologies) and was analyzed by the TaqMan in presence of autologous NK cells. A fusion protein encompassing method (Applied Biosystems) with the CFX-96 real-time PCR system, the human IL-21 receptor and an Fc fragment (IL-21R–Fc) was according to the manufacturers’ instructions (Bio-Rad). Expression of target RNA was normalized to human GAPDH. Primers used were IFN-g added to the cultures. After 7 d, the expression of CD86 and HLA- + + (Hs00989291_m1), IL-4 (Hs00174122_m1), GATA3 (Hs00231122_m1), DR on NK cells was monitored. A population of CD86 HLA-DR retinoic acid–related orphan receptor (ROR)c (Hs01076122_m1), Foxp3 NK cells appeared in T and NK lymphocyte cocultures where CD4 (Hs01085834_m1), and T-bet (Hs00203436_m1). T cells were activated by anti-CD3/CD28. IL-21R–Fc prevented Chromatin immunoprecipitation assay the induction of CD86+HLA-DR+ NK cells (Fig. 1D). Similar

results were achieved using PBMC activated with anti-CD3/CD28 Downloaded from Primary NK cells were cultured for 7 d with IL-2 and IL-2 plus IL-21. At day 7, NK cells were collected and chromatin immunoprecipitation (ChIP) (data not shown). was performed using the ChIP-IT PBMC (Active Motif, La Hulpe, Belgium). The next set of experiments addressed the distribution of NK Briefly, NK cells were fixed with 1% formaldehyde, pelleted, and resus- receptors in NK cell subsets propagated in the presence of IL-21. pended in lysis buffer. Chromatin was sonicated to two hundred 1000-bp CD86+HLA-DR+ NK cells expressed CD16, NKG2D, NKp46, fragments using a sonicator. NKp30, NKp80, as well as DNAM-1 (Fig. 1E). Moreover, NKG2A Immunoprecipitation was performed using anti-STAT3 (D3Z2G) (Cell + 2 Signaling Technology), or rabbit IgG (DA1E) (Cell Signaling Technology) expression was upregulated by IL-21 both on CD86 and CD86 http://www.jimmunol.org/ as control, and incubated overnight at 4˚C with rotation. Immune com- NK cells (Fig. 1E). Of note, CD86+HLA-DR+ NK cells never plexes were collected on protein G and eluted. Cross-linking was reversed expressed NKp44, including when IL-21 was combined with high with proteinase K by heating 30 min at 55˚C and then the temperature was doses of IL-2 (data not shown). increased to 80˚C during 2 h. DNA was purified with the a ChIP DNA purification kit (Active Motif). Quantitative PCR (qPCR) was performed Collectively, these results showed that IL-21 promotes the dif- on purified DNA and analyzed with CFX96 (Bio-Rad). ferentiation of an NK cell subset coexpressing CD86 and HLA-DR, Primers for IL-10 and MIF binding site in STAT3 gene promoter used for which could be identified in CD3/CD28-stimulated human PBMC. the ChIP assay were purchased from Qiagen, including EpiTect ChIP qPCR assay for human GAPDH, NM_002046.3 (+)1 kb; EpiTect ChIP qPCR assay IL-21 differentiates HLA-DR+ NK subset from CD56dimCD16+ for human MIF, NM_002415.1 (2)1 kb; and EpiTect ChIP qPCR assay for NK cells by guest on September 23, 2021 human IL-10, NM_000572.2 (2)8 kb. Data were normalized to the cycle threshold values for the gene encoding NK cells might be subdivided into two populations according GAPDH and are presented as enrichment relative to the results obtained by to CD56 and CD16 expression. The CD56brightCD16low NK cell immunoprecipitation with Ig. subset represents ∼5% of circulating NK cells, producing high levels of cytokines (mainly IFN-g, TNF-a, and GM-CSF) and Results displaying weak cytotoxic functions. Conversely, CD56lowCD16+ IL-21 promotes the differentiation of a subset of NK cells NK cells have effective cytotoxic functions (30). CD56bright NK expressing CD86 and HLA-DR cells are located in secondary lymphoid organs whereas CD56dim IL-21, along with IL-2, is a cytokine produced by activated T cells NK cells are preferentially observed in the peripheral immune and plays a key role in several inflammatory diseases (25–28). system (31). To identify whether CD86+HLA-DR+ NK cells arise Alternatively, IL-21 has the potential to directly activate NK cell from a specific lymphocyte subset, we decided to reproduce pre- proliferation and cytotoxic functions (29). Nevertheless, the previous experiments using NK cells sorted according to the level cise role of IL-21–activated NK cells in the regulation of CD4 of CD56 and CD16 expression. CD56brightCD162 NK cells and T cells was never investigated. We therefore decided to analyze CD56dimCD16+ NK cells derived from peripheral mononuclear whether IL-21 modulates the expression of CD4 T cell costimu- cells were cultured with IL-2 with and without IL-21 for 7 d. latory molecules on NK cells. For this purpose, NK cells were These experiments showed that the CD86+HLA-DR+ NK subset negatively purified from the PBMC of healthy volunteers and originates from CD56dimCD16+ NK cells. In contrast, CD86 and activated with IL-2 (100 UI/ml) with or without 50 ng/ml IL-21 HLA-DR were never observed when CD56brightCD162 NK cells during 7 d. Whereas 4.1BBL, OX40L, CD80, CD40L, ICOSL, and were exposed to IL-2 and IL-21 (Fig. 2).

CD70 were not detectable or were weakly detectable on activated + + NK cells, IL-21 promoted the expression of HLA-DR and CD86 IL-21–induced CD86 HLA-DR NK cells produce IFN-g but (Fig. 1A). displayed decreased cytotoxic functions The presence of CD86 and HLA-DR was clustered on a similar Previous results showed that IL-21–mediated CD86+HLA-DR+ NK cell subpopulation (Fig. 1B). Activation of NK by IL-2 alone NK cells expressed NK cell activating receptors. We therefore (either 100 or 1000 IU/ml) did not induce expression of MHC-II decided to assess the functions of this NK cell subpopulation. or CD86 (Supplemental Fig. 1). IL-15, another cytokine involved Cytotoxic functions of IL-2 or IL-2 plus IL-21–activated NK cells in gc-dependent NK cell homeostasis, was not able to trigger were assessed in cocultures with the K562 cell line. CD107a de- CD86 or HLA-DR expression (data not shown). Induction of granulation was decreased in CD86+HLA-DR+ NK cells com- CD86 and HLA-DR expression was variable among donors (range pared with IL-21–activated HLA-DR2 NK cells or NK cells from 8 to 60% of total CD32CD56+ cells after 7 d of culture). cultured in the presence of IL-2 (Fig. 3A). Furthermore, IL-21– Moreover, IL-21 did not modulate CD86 or MHC-II on murine activated HLA-DR+ NK cells were able to produce IFN-g,as + 4 IL-21–INDUCED MHC-II NK CELLS PROMOTE TCM EXPANSION Downloaded from http://www.jimmunol.org/ by guest on September 23, 2021

FIGURE 1. IL-21 induces a subpopulation of NK cells expressing both CD86 and HLA-DR. (A–C and E) NK cells negatively selected from human PBMC were cultured with IL-2 (100 U/ml) or IL-2 and IL-21 (50 ng/ml) for 7 d. (A) Expression of costimulation molecules was assessed by flow cytometry on CD32CD56+ lymphocytes. (B)CD3/CD56/CD86/HLA-DRphenotypingwasperformedonNKcellspropagatedduring7dinIL-2withorwithoutIL-21. CD86 and HLA-DR expression was assessed on CD32CD56+ lymphocytes by flow cytometry. Results showed a coexpression of CD86 and HLA-DR on an NK cell subset in the presence of IL-21. (C) NK cells were labeled by CFSE before culture with cytokines to investigate whether CD86 expression arises in activated NK cells. CD86 expression was studied by flow cytometry in CD32CD56+ lymphocytes labeled with CFSE. (D) NK cells that were cultured with autologous CD4 T cells stimulated by CD3/CD28 Abs (10:1 NK cells/CD4 T cells) and IL-21R–Fc (12.5 mg/ml) or rituximab were added in culture at days 0, 2, 4, and 6 then CD86 and HLA-DR expression was assessed by flow cytometry. (E) NK cell receptor expression was assessed by flow cytometry on CD32CD56+ lymphocytes. Data are representative of at least three independent experiments. The Journal of Immunology 5 Downloaded from

FIGURE 2. CD86+HLA-DR+ NK cells derive from CD56dim NK cells. CD32CD56brightCD162 NK cells and CD32CD56dimCD16+ NK cells

were isolated by flow cytometry cell sorting and cultured with IL-2 or IL-2 http://www.jimmunol.org/ and IL-21 for 7 d. CD86 expression was assessed by flow cytometry on CD32CD56+ lymphocytes. Data are representative of three independent experiments. shown by intracellular flow cytometry (Fig. 3B). The levels of IFN-g secreted by IL-21–activated HLA-DR NK cells were not significantly different in ELISA assay than those derived from NK activated with 1000 IU/ml IL-2 (data not depicted). Of note, by guest on September 23, 2021 CD86+HLA-DR+ NK cells were also able to produce GM-CSF, TNF-a, and IL-10 but not IL-13 and IL-5 (data not shown). These results suggest that IL-21 promotes the differentiation of CD86+ HLA-DR+ NK cells with potent ability to produce IFN-g but exhibiting reduced cytotoxic functions. Role of STAT3 in CD86+HLA-DR+ NK cell generation

Whereas IL-2 and IL-21 both signal through the gc subunit, IL-21 has an increased capacity to activate STAT3 (22). We therefore used Jak2 and STAT3 inhibitors to characterize the role of this FIGURE 3. CD86+HLA-DR+ NK cells can produce IFN-g but have signaling on the appearance of CD86+HLA-DR+ NK cells. Cul- reduced cytotoxic function. (A) NK cells activated by IL-2 or IL-2 plus tures of human NK cells were reproduced as described above with IL-21 were incubated with K562 cells (ratio 10:1) and degranulation was IL-2 with and without IL-21 during 5 d, and we first analyzed the assessed by flow cytometry by CD107a staining. (B) NK cells cultured level of STAT3 phosphorylation. STAT3 phosphorylation on ty- with IL-2 (100 U/ml) or IL-2 and IL-21 (50 ng/ml) for 7 d were stimulated rosine 705 or serine 727 was enhanced in the presence of IL-21 with PMA (50 ng/ml)/ionomycin (250 ng/ml) for 5 h. IFN-g expression (Fig. 4A). Then, the role of STAT3 on CD86+HLA-DR+ NK cell was assessed by intracellular flow cytometry. Data are representative of three independent experiments. differentiation was questioned using pharmacological inhibitors targeting STAT3 activation. STAT3 pharmacological inhibition + using either AG490 or STA21 prevented the ability of IL-21 to therefore next sought to investigate the influence of HLA-DR NK induce the CD86+HLA-DR+ NK cell subset (Fig. 4B), in line with cells differentiated in the presence of IL-21 on naive CD4 T cells. a direct role of STAT3 in CD86+HLA-DR+ NK cell subset ap- For this purpose, coculture experiments were reproduced using pearance in these cultures. naive CD4 T cells negatively selected from human healthy volunteers and autologous IL-2– or IL-21–propagated NK cells (ratio IL-21–mediated HLA-DR+ NK cells promote the differentiation - + 1:2). CD4 lymphocyte TCR activation was induced by immobi- of naive CD4 T cells into IFN-g IL-2 central lized agonist anti-CD3 mAbs, and Th1 polarization was used as a memory lymphocytes control of naive T lymphocyte activation. CD4 T cell proliferation In another set of experiments, we identified that HLA-DR+ NK was monitored by CFSE dilution assay. As shown in Fig. 5B, NK cells specifically expressed CD62L compared with IL-2–activated cells propagated in IL-21 were able to induce naive CD4 T cell NK cells or HLA-DR2 NK cells recovered from cultures with IL-21 proliferation, whereas naive CD4 T cells were not activated in (Fig. 5A). This observation raised the hypothesis that IL-21–driven the presence of NK cells cultured with IL-2 (Fig. 5B) or with HLA-DR+ NK cells could interact with naive CD4 T cells. We immobilized anti-CD3 mAb alone. The memory phenotype of + 6 IL-21–INDUCED MHC-II NK CELLS PROMOTE TCM EXPANSION

FIGURE 4. Expression of CD86 and HLA-DR on NK cells depend on IL-21–induced STAT3 phosphorylation. (A and B) Expression of CD86 and HLA-DR (top) and phosphorylation of pSTAT3 (pY705 and Downloaded from pS727) (bottom) were assessed by ﬂow cytometry on NK cells cultured 7 d with (A) IL-2 (100 U/ml) or IL-2/ IL-21 (50 ng/ml) and (B) IL-2/IL-21 with AG490 (20 mM), STA21 (20 mM), or DMSO. Data are representative of two independent experiments. http://www.jimmunol.org/ by guest on September 23, 2021

CD4 T cells costimulated by NK cells or differentiated with Th1 absence of Th1, Th2, Th17, or regulatory T cell polarization was polarizing conditions was investigated. Naive CD4 T cells acti- confirmed by RT-PCR because TCM generated in the presence of vated in Th1 conditions lost CD45RA expression and became CD86+HLA-DR+ NK cells lacked significant expression of the CD62L2 effector memory cells (Fig. 5C). Interestingly, IL-21– transcription factors T-bet, Gata3, RORgt, or Foxp3 (Supplemental activated NK cells prevented the loss of CD62L on memory CD4 Fig. 2). These experiments showed that CD4 T lymphocytes co- + T cells generated in these experiments. The presence of CD86 stimulated by IL-21–activated NK cells are uncommitted pre-TCM. HLA-DR+ NK cells sustained the expansion of CD45RA2CD62L+ STAT3-dependent production of MIF by IL-21–activated CD4 T cells corresponding to a central memory phenotype + + (Fig. 5C). Considering that IL-2 fosters the differentiation of naive CD86 HLA-DR NK cells is involved in pre-TCM expansion T cells toward CD62L2 effector memory lymphocytes (32), we We next sought to determine which molecular mechanism accounts + also tested the influence of this cytokine during CD4 cocultures for pre-TCM expansion during cultures of naive CD4 T cells and with NK cells. As expected, addition of IL-2 in CD4 and NK IL-21–activated NK cells. Previous experiments were first repro- cocultures increased proliferation of T lymphocytes. However, the duced using Transwells to determine whether cellular contacts are + + presence of CD86 HLA-DR NK cells enables CD4 T cells to re- needed for pre-TCM expansion. These results showed that pre-TCM tain a central memory phenotype (Fig. 5C). TCM inducedbyIL-21– expansion after coculture with IL-21–activated NK cells involved activated NK cells expressed CXCR3 but not CCR6, CCR4, or a soluble factor (Fig. 6A). Supernatants of IL-21–activated NK CXCR5 (Fig. 5D), suggesting a potential Th1 polarization (33). cells were analyzed using a protein array. An overexpression of However, the precise analysis of these TCM showed an uncom- MIF was observed in IL-21–driven NK cells (Fig. 6B). Further mitted profile. Indeed, whereas Th1 cells differentiated from naive analysis of IL-21–activated NK cells using confocal microscopy T cells produced TNF-a, IL-2, and IFN-g, CFSElow CD4 T cells revealed that MIF expression was restricted in HLA-DR+ NK cells propagated in the presence of CD86+HLA-DR+ NK cells secreted within these lymphocytes (Fig. 6C). Therefore, the precise role of only IL-2 and low levels of TNF-a (Fig. 5E). Additionally, these MIF in pre-TCM expansion was investigated using pharmacological CD4 T cells did not express IL-4 or IL-17 (data not show). The inhibitors (ISO-1) or MIF-neutralizing Abs. These experiments The Journal of Immunology 7 Downloaded from http://www.jimmunol.org/ by guest on September 23, 2021

FIGURE 5. Naive CD4 T lymphocyte costimulation and polarization by IL-21–activated NK cells. (A) NK cells were cultured with IL-2 (100 U/ml) or IL-2 and IL-21 (50 ng/ml) for 7 d and then CD86 and CD62L expression was assessed by flow cytometry. (B–E) Cultured NK cells were incubated with autologous CFSE-labeled naive CD4 T cells for 5–6 d on a plate-bound anti-CD3 (HIT3a, 5 mg/ml). Th1 polarization mixture was used as positive control. (B) Proliferation of naive CD4+ T lymphocytes activated in the presence autologous NK was assessed by CFSE dilution assay. (C) CD4 T cells memory phenotype was characterized by CD45RA and CD62L staining. (D) Chemokine receptor expression on CFSElow CD4 T cells after culture with NK cells cultured with IL-2 and IL-21 was assessed by flow cytometry. (E) Production of IFN-g, IL-2, and TNF-a was analyzed on CD3+CD4+CFSElow T cells by intracellular cytokine staining after overnight stimulation with PMA (50 ng/ml)/ionomycin (250 ng/ml). Data are representative of at least three independent experiments. provided evidence that MIF neutralization decreased the expan- (Fig. 6D, 6E). ChIP assays were then performed to determine + sion of pre-TCM in cocultures of CD4 T cells and NK cells whether STAT3, induced by IL-21 in NK cells, could directly (Fig. 6D, 6E). Similar results were achieved by using an antago- activate MIF production. These experiments showed that STAT3 nist mAb targeting CD74, a receptor of MIF on activated T cells proteins recovered from IL-21–cultured NK cells were able to + 8 IL-21–INDUCED MHC-II NK CELLS PROMOTE TCM EXPANSION bind MIF promoter (Fig. 6F). Collectively, these results demon- authors used CD56 mircobeads to positively isolate NK cells strated that IL-21 induces a subpopulation of NK cells expressing before culture with IL-2. In the present study, to determine the MHC-II, producing MIF, and endows them with the ability to origin of HLA-DR+ NK cells, we negatively isolated NK cells generate uncommitted pre-TCM lymphocytes. with a selection strategy that included anti-HLA-DR. Thus, our results pointed out that IL-21 can regulate a transcriptional pro- Identification of MIF+ NK cells in inflammatory human gram leading to the emergence of a specific population of NK appendix cells that express MHC-II molecules such as HLA-DR, but also IL-21 is produced during the inflammation phase, and thus we CIITA, cathepsins, RFX3, and HLA-DM (data not show). Fur- decided to investigate whether NK cells from human tissues might thermore, the expression of MHC-II–related molecules induced by express MIF. In a first set of experiments, we obtained 20 tonsils IL-21 is restricted to the HLA-DR+CD86+ NK cell subset. removed for a history of local infections. Tissues were mechan- The molecular mechanism and physiopathological context ically dissociated and lymphocytes recovered and immediately regulating MHC-II expression and Ag presentation functions in analyzed after surgery to assess HLA-DR and CD86 expression by human NK cells were never clearly investigated. In this context, flow cytometry. HLA-DR or CD86 was never observed on NK Kim and colleagues (38) confirmed that activated NK cells ex- derived from human tonsils in these experiments (data not shown). press increased levels of MHC-II. In addition, they also have We therefore hypothesized that tonsils from these children were shown that NK cells alone can present Ags to CD4 T cells in the recovered after inflammation resolution because most surgeries context of HSV infection after TLR activation. They suggested were performed .1 mo following tonsil infections. We then decided that this interaction could occur in herpetic lesions and also in to study another cohort of patients who were referred to our surgery draining lymph nodes (38). These observations support the hy- Downloaded from department for appendix inflammation and who underwent surgery pothesis that cytokines, such as IL-21, involved in chronic in- at the acute phase of their disease. Confocal microscopy was used to flammation or infections, could account for HLA-DR+ NK cell 2 + assess MIF expression in CD3 CD56 lymphocytes. These exper- differentiation (20). Thus, we decided to characterize the presence iments showed the presence of MIF in NK cells during inflamma- of HLA-DR+CD86+ NK cells in tonsils removed from 20 children tion in human appendix of 12 patients of the 20 samples analyzed with a history of infections. HLA-DR+ NK cells were not ob- 6 2 + (Fig. 7). In these patients, 53 12% of CD3 CD56 lymphocytes served in these patients (data not shown). We then hypothesized http://www.jimmunol.org/ expressed MIF. Thus, these experiments show that NK cells might that the expression of HLA-DR and CD86 on NK cells might be be a source of MIF production in humans. dependent on the presence of IL-21. Indeed, removal of IL-21 after 5 d of culture led to the progressive decrease of HLA-DR Discussion and CD86 expression (data not shown), suggesting that the ca- We have shown that IL-21 sustains the differentiation of a specific pacity of these NK cells to modulate CD4 T lymphocytes might subset of NK cells coexpressing CD86 and MHC-II and lacking require the presence of IL-21 in vivo. We therefore decided to NKp44. Therefore, IL-21 produced by activated CD4+ T cells not evaluate the presence of MIF and HLA-DR+ NK cells in inflamed only promotes the conventional cytotoxic functions of NK cells appendix where the Th17-like cytokine pattern was already described but also generates a specific subset of NK cells endowed with Ag (43). Confocal microscopy analysis of these samples showed the by guest on September 23, 2021 processing capacities and helper functions for adaptive immunity. presence of MIF+ NK cells (Fig. 7) and HLA-DR expression on NK Our results reveal that HLA-DR+CD86+ NK cells differentiated in cells (data not shown). the presence of IL-21, but not IL-2–activated NK cells, produce Direct interactions between human CD4 Th lymphocytes and MIF and specifically expand uncommitted TCM. NK cells have been unraveled by the pioneering study of Fehniger IL-2, IL-15, and IL-21 receptors share the common cytokine et al. (44). These authors have emphasized that endogenous g-chain subunit. Although IL-15 plays a critical role in the early T lymphocyte–derived IL-2 is involved in lymph node–resident differentiation of NK cells (34), both cytokines are involved in CD56bright NK cell activation through high-affinity IL-2 receptors. peripheral NK cell homeostasis. Even if IL-2 and IL-21 contribute IL-21 might represent another mechanism implicated in the reg- to the terminal maturation of NK cells and promote cytotoxic ulation of adaptive and innate lymphocyte regulation. Particularly, functions, our results confer to IL-21 a unique role in driving the we provided evidence that IL-21 produced by human lymphocytes differentiation of HLA-DR+CD86+ NK cells. It is well established acts on CD56dim NK cells to differentiate MHC-II+ NK cells pro- that resting NK cells do not express MHC-II RNA or proteins. ducing MIF. This process requires a long-term exposure to IL-21 However, NK cells isolated from MLRs were previously shown because we observed the appearance of MHC-II+ NK cells fol- to display membrane-associated MHC-II molecules (35). In this lowing at least 4 d of culture. Moreover, this phenomenon seems to study, we provide evidence that IL-21 neutralization, using IL-21– be tightly regulated in a STAT3-dependent manner. Accordingly, Fc protein, prevents the ability of activated lymphocytes to induce IL-21 might represent a regulatory mechanism tuning CD4 and NK MHC-II+ NK cells in these cocultures. lymphocyte interactions during inflammation or chronic infections, MHC-II molecules are critical for the presentation of peptides as suggested by the identification of HLA-DR+ NK cells during processed from extracellular soluble proteins to CD4 T cells. The HSV infection (38). constitutive expression of MHC-II is restricted to professional Conversely, NK cells were also demonstrated to modulate CD4 APC such as DC, B lymphocytes, or macrophages (36). The regula- T lymphocyte polarization. DC–NK cell cross-talk is one of the tion of MHC-II expression following exposition to inflammatory major components of these interactions. In vitro, mature DC stim- stimuli depends on the CIITA (36). However, although it was ulate NK cell proliferation, cytotoxic functions, and production of previously acknowledged that NK cell clones could process and IFN-g. Both NK cell–activating receptors and soluble mediators present some soluble Ags such as tetanus toxoid (37), the ex- such as IL-12, IL-15, IFN-a, or IL-18 contribute to DC-mediated pression of HLA-DR on NK cells was mostly considered as an NK cell activation (8, 10, 45). The dynamic network in which DC activation marker (38–42). As previously suggested by Evans interact closely with surrounding NK cells and CD4 was also et al. (42), the identification of HLA-DR+ NK cells after culture described in vivo during Leishmania or HSV infections (12, 46, with IL-2 might result from the expansion of rare HLA-DR+ cells 47). In vivo interactions of DC and NK cells promote Th1 po- initially present in peripheral blood NK cells. In that study, the larization of naive CD4+ T cells (11, 48). These studies indicated The Journal of Immunology 9 Downloaded from http://www.jimmunol.org/ by guest on September 23, 2021

FIGURE 6. MIF is involved in the costimulation of naive CD4 T lymphocytes by IL-21–activated NK cells. (A) NK and autologous naive CD4 T lymphocytes were puriﬁed from PBMC of healthy donors and cocultures as previously described in the presence or absence of a Transwell. NK cells cultured with IL-2 (100 U/ml) or IL-2 and IL-21 (50 ng/ml) (top chamber) were incubated with autologous CFSE-labeled naive CD4 T cells (bottom chamber) on a plate-bound anti-CD3 (HIT3a, 5 mg/ml) for 5–6 d. Then, proliferation of CD4 T lymphocytes was analyzed with CFSE dilution assay. Results are representative of three independent experiments. (B) Cytokine array was performed in two independent experiments using supernatants derived from NK cells cultured for 5 d with IL-2 or IL-2 and IL-21. (C) MIF and HLA-DR localization were characterized, in three independent experiments, by confocal microscopy (original magniﬁcation 340) on NK cells cultured with IL-2 or IL-2 and IL-21 at day 7. (D and E) NK cells were cultured with IL-2 (100 U/ml) or IL-2 and IL-21 (50 ng/ml) for 7 d before culture with autologous CFSE-labeled naive CD4 T cells during 5–6 d on a plate-bound anti-CD3 (HIT3a, 5 mg/ml) with anti-CD74 (M-B741), anti-MIF (10C3), or the MIF pharmacological inhibitor ISO-1. Then, proliferation was assessed with CFSE dilution assay (D). (E) Statistical analysis of cell proliferation with ISO-1 (n = 5), anti-MIF (n = 3), or anti-CD74 (n = 7) was realized with a paired t test. Data are presented as the means 6 SEM. *p , 0.05, **p , 0.01. (F) ChIP experiments were performed to investigate the (Figure legend continues) + 10 IL-21–INDUCED MHC-II NK CELLS PROMOTE TCM EXPANSION

FIGURE 7. MIF is expressed in NK cells in human inﬂamed appendix. Parafﬁn- embedded tissues from appendix removed for an appendicitis were collected. Con- focal microscopy assessment of MIF expression was performed in CD32CD56+ lymphocytes. Scale bar, 10 mm. Downloaded from

that the “helper” functions of NK for Th1 polarization mainly rely memory lymphocytes producing IL-2, low levels of TNF-a, and on IFN-g production. Nevertheless, the precise influence of NK no IFN-g. Expansion of TCM and production of IL-2 (data not cells on the fate of memory T cell differentiation was never clearly depicted) were shown to be dependent on soluble factors. MIF is a http://www.jimmunol.org/ characterized. proinflammatory cytokine with chemokine-like functions. After Recently, several distinct innate lymphoid cell (ILC) populations binding to CD74, MIF activates PI3K/Akt and ERK pathways, have been identified and play a critical role in the modulation of which both play an important role in cell survival and proliferation adaptive immune systems. However, the functions of ILC in the (56–58). MIF has already been identified to act as a mitogenic and presence of adaptive immunity and their potential to influence IL-2–activating factor for T cells (59). The role of MIF in the adaptive immune cell responses against cancers in inflammation central memory polarization of CD4 T cells had not been inves- context remain poorly understood. IL-12–activated NCR+ ILC3 tigated before. Accordingly, the expression of MIF was assessed

has been described to initiate antitumor immunity by upregulation in NK cells activated by IL-2 or IL-21. We could identify a se- by guest on September 23, 2021 of adhesion molecules in the tumor vasculature, which resulted in lective increase of MIF induced by IL-21 in HLA-DR+ but not more leukocyte invasion (49). Conversely, ILC3, which expresses HLA-DR2 NK cells. Moreover, we could identify the presence of CCR6 and produces IL-22, was shown to tumorigenic potential MIF+ NK cells in inflamed appendix removed from patients who (50, 51). Indeed, colon cancer is accompanied by accumulation of underwent surgery for appendicitis. Neutralization of MIF using a IL-17+IL-22+ ILC3, and its depletion in mice blocks the devel- pharmacological inhibitor, a neutralizing mAb, or blocking its opment of invasive colon cancer (52). More recently, both ILC2 receptor (CD74) prevented the proliferation of naive lymphocytes and ILC3 were also shown to modulate CD4 T cell responses mediated by HLA-DR+ NK cells. Nevertheless, addition of re- through Ag-peptide presentation by MHC-II (53, 54). Further- combinant MIF failed to expand naive CD4+ T cells when com- more, the microenvironment seems to be important to determine bined with anti-CD3 activation and IL-2–pretreated NK cells (data the modulation of CD4 T cells by MHC-II+ ILC3. Hepworth et al. not shown), suggesting that MIF is a necessary but not sufficient (53) have shown that RORgt+ ILC3 may process Ags and present component of the IL-21–induced HLA-DR+ NK costimulation. them on MHC-II molecules without the expression of costimula- Chemokine receptor expression identifies functional subpopu- tory molecules resulting in the negative regulation of CD4 T cell lations of memory T cells (60). In line with these studies (60, 61), responses. In contrast, von Burg et al. (55) have shown that IL-1b we observed that CD4+ T lymphocytes differentiated in the activates ILC3 and induces both the expression of costimulatory presence of IL-21–activated NK cells expressed CXCR3 as Th1 molecules and the upregulation of MHC-II molecules able to cells but lacked CXCR5, CCR6, and CCR4. Moreover, these promote T cell proliferation. lymphocytes produced IL-2 and a low amount of TNF-a, but no In the present study, we described novel conditions where IL-21 IFN-g. Of note, more recently, a new memory T cell subset in + directly promotes the differentiation of an HLA-DR NK subset humans has been described, the stem cell memory T cells (TSCM), + that is endowed with the ability to interact with CD4 Tcells.More which have superior proliferation and survival potential (62). TSCM importantly, the IL-21–induced HLA-DR+ NK cell subset might are characterized by the expression of Fas (CD95) and IL-2Rb play a particular role during naive T cell differentiation. IL-21 (CD122) and also retain the expression of CD45RA following drives STAT3 binding to the promoter of MIF. Our results showed TCR-mediated activation. CD4+ T cells costimulated by HLA-DR+ that MIF production by HLA-DR+ NK cells leads to the differenti- NK cells in our experiments lacked CD95 and CD45RA, ruling out + 2 + ation of naive CD4 T lymphocytes into CD45RA CD62L central a potential acquisition of the TSCM phenotype. Collectively, these

binding capacity of STAT3 on the putative STAT3 response elements in the MIF promoter and in the IL-10 promoter in NK cells cultured with IL-2 or IL-2 and IL-21. Results obtained with anti-STAT3 ChIP are presented in arbitrary units (AU) relative to those obtained with control Ig (isotype-matched control Ab). Data are representative of two independent experiments. The Journal of Immunology 11 results support that CD4+ T cells costimulated in the presence of 21. Cavazzana-Calvo, M., S. Hacein-Bey, G. de Saint Basile, F. Gross, E. Yvon, + P. Nusbaum, F. Selz, C. Hue, S. Certain, J. L. Casanova, et al. 2000. Gene therapy IL-21–induced HLA-DR NK cells acquire characteristics of un- of human severe combined immunodeficiency (SCID)-X1 disease. Science 288: committed TCM containing pre-Th1 lymphocytes and retaining 669–672. CD62L expression upon TCR-driven expansion even when IL-2 22. Lieberman, L. A., and G. C. Tsokos. 2010. The IL-2 defect in systemic lupus erythematosus disease has an expansive effect on host immunity. J. Biomed. was included in the cocultures. Thus, this NK cell subset displays Biotechnol. 2010: 740619. a particular capacity to modulate memory T cell differentiation in 23. Leonard, W. J., and R. Spolski. 2005. Interleukin-21: a modulator of lymphoid the presence of IL-21. This novel helper function of NK cells might proliferation, apoptosis and differentiation. Nat. Rev. Immunol. 5: 688–698. 24. Parrish-Novak, J., S. R. Dillon, A. Nelson, A. Hammond, C. Sprecher, be of interest to expand the pool of TCM during chronic infections. J. A. Gross, J. Johnston, K. Madden, W. Xu, J. West, et al. 2000. Interleukin 21 and its receptor are involved in NK cell expansion and regulation of lymphocyte function. Nature 408: 57–63. Acknowledgments 25. Meguro, A., K. Ozaki, I. Oh, K. Hatanaka, H. Matsu, R. Tatara, K. Sato, We are grateful to Dr. Sun who provided appendix tissues. W. J. Leonard, and K. Ozawa. 2010. IL-21 is critical for GVHD in a mouse model. Bone Marrow Transplant. 45: 723–729. 26. Hippen, K. L., C. Bucher, D. K. Schirm, A. M. Bearl, T. Brender, K. A. Mink, K. S. Waggie, R. Peffault de Latour, A. Janin, J. M. Curtsinger, et al. 2012. Disclosures Blocking IL-21 signaling ameliorates xenogeneic GVHD induced by The authors have no financial conflicts of interest. human lymphocytes. Blood 119: 619–628. 27. Herber, D., T. P. Brown, S. Liang, D. A. Young, M. Collins, and K. Dunussi- Joannopoulos. 2007. IL-21 has a pathogenic role in a lupus-prone mouse model References and its blockade with IL-21R.Fc reduces disease progression. J. Immunol. 178: 3822–3830. 1. Vivier, E., D. H. Raulet, A. Moretta, M. A. Caligiuri, L. Zitvogel, L. L. Lanier, 28. Young, D. A., M. Hegen, H. L. M. Ma, M. J. Whitters, L. M. Albert, L. Lowe, W. M. Yokoyama, and S. Ugolini. 2011. Innate or adaptive immunity? The Downloaded from M. Senices, P. W. Wu, B. Sibley, Y. Leathurby, et al. 2007. Blockade of the example of natural killer cells. Science 331: 44–49. interleukin-21/interleukin-21 receptor pathway ameliorates disease in animal 2. Biron, C. A., K. S. Byron, and J. L. Sullivan. 1989. Severe herpesvirus infections models of rheumatoid arthritis. Arthritis Rheum. 56: 1152–1163. in an adolescent without natural killer cells. N. Engl. J. Med. 320: 1731–1735. 29. Brady, J., Y. Hayakawa, M. J. Smyth, and S. L. Nutt. 2004. IL-21 induces the 3. Lee, S.-H., T. Miyagi, and C. A. Biron. 2007. Keeping NK cells in highly reg- functional maturation of murine NK cells. J. Immunol. 172: 2048–2058. ulated antiviral warfare. Trends Immunol. 28: 252–259. 30. Cooper, M. A., T. A. Fehniger, and M. A. Caligiuri. 2001. The biology of human 4. Orange, J. S., and C. A. Biron. 1996. An absolute and restricted requirement for natural killer-cell subsets. Trends Immunol. 22: 633–640. IL-12 in natural killer cell IFN-g production and antiviral defense. Studies of 31. Farag, S. S., and M. A. Caligiuri. 2006. Human natural killer cell development natural killer and T cell responses in contrasting viral infections. J. Immunol. and biology. Blood Rev. 20: 123–137. http://www.jimmunol.org/ 156: 1138–1142. 32. Gattinoni, L., and N. P. Restifo. 2013. Moving T memory stem cells to the clinic. 5. Nguyen, K. B., T. P. Salazar-Mather, M. Y. Dalod, J. B. Van Deusen, X. Q. Wei, Blood 121: 567–568. F. Y. Liew, M. A. Caligiuri, J. E. Durbin, and C. A. Biron. 2002. Coordinated and 33. Mahnke, Y. D., T. M. Brodie, F. Sallusto, M. Roederer, and E. Lugli. 2013. The distinct roles for IFN-ab, IL-12, and IL-15 regulation of NK cell responses to who’s who of T-cell differentiation: human memory T-cell subsets. Eur. J. viral infection. J. Immunol. 169: 4279–4287. Immunol. 43: 2797–2809. 6. Chaigne-Delalande, B., F.-Y. Li, G. M. O’Connor, M. J. Lukacs, P. Jiang, 34. Vosshenrich, C. A. J., T. Ranson, S. I. Samson, E. Corcuff, F. Colucci, L. Zheng, A. Shatzer, M. Biancalana, S. Pittaluga, H. F. Matthews, et al. 2013. Mg2+ regulates cytotoxic functions of NK and CD8 T cells in chronic EBV E. E. Rosmaraki, and J. P. Di Santo. 2005. Roles for common cytokine receptor infection through NKG2D. Science 341: 186–191. g-chain-dependent cytokines in the generation, differentiation, and maturation of 7. Walzer, T., M. Dalod, S. H. Robbins, L. Zitvogel, and E. Vivier. 2005. Natural- NK cell precursors and peripheral NK cells in vivo. J. Immunol. 174: 1213–1221. killer cells and dendritic cells: “l’union fait la force”. Blood 106: 2252–2258. 35. Phillips, J. H., A. M. Le, and L. L. Lanier. 1984. Natural killer cells activated in a 8. Gerosa, F., B. Baldani-Guerra, C. Nisii, V. Marchesini, G. Carra, and human mixed lymphocyte response culture identified by expression of Leu-11 by guest on September 23, 2021 G. Trinchieri. 2002. Reciprocal activating interaction between natural killer cells and class II histocompatibility antigens. J. Exp. Med. 159: 993–1008. and dendritic cells. J. Exp. Med. 195: 327–333. 36. Mach, B., V. Steimle, E. Martinez-Soria, and W. Reith. 1996. Regulation of 9. Ferlazzo, G., M. L. Tsang, L. Moretta, G. Melioli, R. M. Steinman, and C. Munz.€ MHC class II genes: lessons from a disease. Annu. Rev. Immunol. 14: 301–331. 2002. Human dendritic cells activate resting natural killer (NK) cells and are 37. Roncarolo, M. G., M. Bigler, J. B. Haanen, H. Yssel, R. Bacchetta, J. E. de Vries, recognized via the NKp30 receptor by activated NK cells. J. Exp. Med. 195: and H. Spits. 1991. Natural killer cell clones can efficiently process and present 343–351. protein antigens. J. Immunol. 147: 781–787. 10. Borg, C., A. Jalil, D. Laderach, K. Maruyama, H. Wakasugi, S. Charrier, 38. Kim, M., N. R. Osborne, W. Zeng, H. Donaghy, K. McKinnon, D. C. Jackson, B. Ryffel, A. Cambi, C. Figdor, W. Vainchenker, et al. 2004. NK cell activation and A. L. Cunningham. 2012. Herpes simplex virus antigens directly activate by dendritic cells (DCs) requires the formation of a synapse leading to IL-12 NK cells via TLR2, thus facilitating their presentation to CD4 T lymphocytes. polarization in DCs. Blood 104: 3267–3275. J. Immunol. 188: 4158–4170. ´ 39. Zingoni, A., T. Sornasse, B. G. Cocks, Y. Tanaka, A. Santoni, and L. L. Lanier. 11. Martın-Fontecha, A., L. L. Thomsen, S. Brett, C. Gerard, M. Lipp, + A. Lanzavecchia, and F. Sallusto. 2004. Induced recruitment of NK cells to 2004. Cross-talk between activated human NK cells and CD4 T cells via OX40- OX40 ligand interactions. J. Immunol. 173: 3716–3724. lymph nodes provides IFN-g for TH1 priming. Nat. Immunol. 5: 1260–1265. 12. Bajeńoff, M., B. Breart, A. Y. C. Huang, H. Qi, J. Cazareth, V. M. Braud, 40. Skak, K., K. S. Frederiksen, and D. Lundsgaard. 2008. Interleukin-21 activates R. N. Germain, and N. Glaichenhaus. 2006. Natural killer cell behavior in lymph human natural killer cells and modulates their surface receptor expression. Im- nodes revealed by static and real-time imaging. J. Exp. Med. 203: 619–631. munology 123: 575–583. 13. Schleicher,U.,J.Liese,I.Knippertz,C.Kurzmann,A.Hesse,A.Heit,J.A.A.Fischer, 41. Li, W., D. M. Janowicz, K. R. Fortney, B. P. Katz, and S. M. Spinola. 2009. S. Weiss, U. Kalinke, S. Kunz, and C. Bogdan. 2007. NK cell activation in visceral Mechanism of human natural killer cell activation by Haemophilus ducreyi. J. leishmaniasis requires TLR9, myeloid DCs, and IL-12, but is independent of plas- Infect. Dis. 200: 590–598. macytoid DCs. J. Exp. Med. 204: 893–906. 42. Evans, J. H., A. Horowitz, M. Mehrabi, E. L. Wise, J. E. Pease, E. M. Riley, and 14. Laouar, Y., F. S. Sutterwala, L. Gorelik, and R. A. Flavell. 2005. Transforming D. M. Davis. 2011. A distinct subset of human NK cells expressing HLA-DR growth factor-b controls T helper type 1 cell development through regulation of expand in response to IL-2 and can aid immune responses to BCG. Eur. J. natural killer cell interferon-g. Nat. Immunol. 6: 600–607. Immunol. 41: 1924–1933. 15. Cameron, A. L., B. Kirby, W. Fei, and C. E. M. Griffiths. 2002. Natural killer and 43. Rube´r, M., M. Andersson, B. F. Petersson, G. Olaison, R. E. Andersson, and natural killer-T cells in psoriasis. Arch. Dermatol. Res. 294: 363–369. C. Ekerfelt. 2010. Systemic Th17-like cytokine pattern in gangrenous appendi- 16. Delahaye, N. F., S. Rusakiewicz, I. Martins, C. Meńard, S. Roux, L. Lyonnet, citis but not in phlegmonous appendicitis. Surgery 147: 366–372. P. Paul, M. Sarabi, N. Chaput, M. Semeraro, et al. 2011. Alternatively spliced 44. Fehniger, T. A., M. A. Cooper, G. J. Nuovo, M. Cella, F. Facchetti, M. Colonna, bright NKp30 isoforms affect the prognosis of gastrointestinal stromal tumors. Nat. and M. A. Caligiuri. 2003. CD56 natural killer cells are present in human Med. 17: 700–707. lymph nodes and are activated by T cell-derived IL-2: a potential new link be- 17. Rusakiewicz, S., G. Nocturne, T. Lazure, M. Semeraro, C. Flament, S. Caillat- tween adaptive and innate immunity. Blood 101: 3052–3057. Zucman, D. Se`ne, N. Delahaye, E. Vivier, K. Chaba, et al. 2013. NCR3/NKp30 45. Ferlazzo, G., M. Pack, D. Thomas, C. Paludan, D. Schmid, T. Strowig, contributes to pathogenesis in primary Sjo¨gren’s syndrome. Sci. Transl. Med. 5: G. Bougras, W. A. Muller, L. Moretta, and C. Munz.€ 2004. Distinct roles of IL- 195ra96. 12 and IL-15 in human natural killer cell activation by dendritic cells from 18. Liu, R., Y. Bai, T. L. Vollmer, X.-F. Bai, Y. Jee, Y. Y. Tang, D. I. Campagnolo, secondary lymphoid organs. Proc. Natl. Acad. Sci. USA 101: 16606–16611. M. Collins, D. A. Young, A. La Cava, and F.-D. Shi. 2008. IL-21 receptor ex- 46. Kassim, S. H., N. K. Rajasagi, X. Zhao, R. Chervenak, and S. R. Jennings. 2006. pression determines the temporal phases of experimental autoimmune enceph- In vivo ablation of CD11c-positive dendritic cells increases susceptibility to alomyelitis. Exp. Neurol. 211: 14–24. herpes simplex virus type 1 infection and diminishes NK and T-cell responses. J. 19. Vollmer, T. L., R. Liu, M. Price, S. Rhodes, A. La Cava, and F.-D. Shi. 2005. Virol. 80: 3985–3993. Differential effects of IL-21 during initiation and progression of autoimmunity 47. Barr, D. P., G. T. Belz, P. C. Reading, M. Wojtasiak, P. G. Whitney, W. R. Heath, against neuroantigen. J. Immunol. 174: 2696–2701. F. R. Carbone, and A. G. Brooks. 2007. A role for plasmacytoid dendritic cells in 20. Elsaesser, H., K. Sauer, and D. G. Brooks. 2009. IL-21 is required to control the rapid IL-18-dependent activation of NK cells following HSV-1 infection. chronic viral infection. Science 324: 1569–1572. Eur. J. Immunol. 37: 1334–1342. + 12 IL-21–INDUCED MHC-II NK CELLS PROMOTE TCM EXPANSION

48. Mailliard, R. B., S. M. Alber, H. Shen, S. C. Watkins, J. M. Kirkwood, 56. Lue, H., A. Kapurniotu, G. Fingerle-Rowson, T. Roger, L. Leng, M. Thiele, R. B. Herberman, and P. Kalinski. 2005. IL-18-induced CD83+CCR7+ NK helper T. Calandra, R. Bucala, and J. Bernhagen. 2006. Rapid and transient activation of cells. J. Exp. Med. 202: 941–953. the ERK MAPK signalling pathway by macrophage migration inhibitory factor 49. Eisenring, M., J. vom Berg, G. Kristiansen, E. Saller, and B. Becher. 2010. IL-12 (MIF) and dependence on JAB1/CSN5 and Src kinase activity. Cell. Signal. 18: initiates tumor rejection via lymphoid tissue-inducer cells bearing the natural 688–703. cytotoxicity receptor NKp46. Nat. Immunol. 11: 1030–1038. 57. Lue, H., M. Thiele, J. Franz, E. Dahl, S. Speckgens, L. Leng, G. Fingerle- 50. Ji, Y., X. Yang, J. Li, Z. Lu, X. Li, J. Yu, and N. Li. 2014. IL-22 promotes the Rowson, R. Bucala, B. Luscher,€ and J. Bernhagen. 2007. Macrophage migration migration and invasion of gastric cancer cells via IL-22R1/AKT/MMP-9 sig- inhibitory factor (MIF) promotes cell survival by activation of the Akt pathway naling. Int. J. Clin. Exp. Pathol. 7: 3694–3703. and role for CSN5/JAB1 in the control of autocrine MIF activity. Oncogene 26: 51. Kryczek, I., Y. Lin, N. Nagarsheth, D. Peng, L. Zhao, E. Zhao, L. Vatan, 5046–5059. W. Szeliga, Y. Dou, S. Owens, et al. 2014. IL-22+CD4+ T cells promote colo- 58. Starlets, D., Y. Gore, I. Binsky, M. Haran, N. Harpaz, L. Shvidel, S. Becker- rectal cancer stemness via STAT3 transcription factor activation and induction of Herman, A. Berrebi, and I. Shachar. 2006. Cell-surface CD74 initiates a sig- the methyltransferase DOT1L. Immunity 40: 772–784. naling cascade leading to cell proliferation and survival. Blood 107: 4807–4816. 52. Kirchberger, S., D. J. Royston, O. Boulard, E. Thornton, F. Franchini, 59. Bacher, M., C. N. Metz, T. Calandra, K. Mayer, J. Chesney, M. Lohoff, R. L. Szabady, O. Harrison, and F. Powrie. 2013. Innate lymphoid cells sustain D. Gemsa, T. Donnelly, and R. Bucala. 1996. An essential regulatory role for colon cancer through production of interleukin-22 in a mouse model. J. Exp. macrophage migration inhibitory factor in T-cell activation. Proc. Natl. Acad. Med. 210: 917–931. Sci. USA 93: 7849–7854. 53. Hepworth, M. R., L. A. Monticelli, T. C. Fung, C. G. K. Ziegler, S. Grunberg, 60. Langenkamp, A., K. Nagata, K. Murphy, L. Wu, A. Lanzavecchia, and R. Sinha, A. R. Mantegazza, H.-L. Ma, A. Crawford, J. M. Angelosanto, et al. F. Sallusto. 2003. Kinetics and expression patterns of chemokine receptors in 2013. Innate lymphoid cells regulate CD4+ T-cell responses to intestinal com- human CD4+ T lymphocytes primed by myeloid or plasmacytoid dendritic cells. mensal bacteria. Nature 498: 113–117. Eur. J. Immunol. 33: 474–482. 54. Mirchandani, A. S., A.-G. Besnard, E. Yip, C. Scott, C. C. Bain, V. Cerovic, 61. Rivino, L., M. Messi, D. Jarrossay, A. Lanzavecchia, F. Sallusto, and J. Geginat. R. J. Salmond, and F. Y. Liew. 2014. Type 2 innate lymphoid cells drive CD4+ 2004. Chemokine receptor expression identiﬁes pre-T helper (Th)1, pre-Th2, and Th2 cell responses. J. Immunol. 192: 2442–2448. nonpolarized cells among human CD4+ central memory T cells. J. Exp. Med. 55. von Burg, N., S. Chappaz, A. Baerenwaldt, E. Horvath, S. Bose Dasgupta, 200: 725–735. D. Ashok, J. Pieters, F. Tacchini-Cottier, A. Rolink, H. Acha-Orbea, and 62. Gattinoni, L., E. Lugli, Y. Ji, Z. Pos, C. M. Paulos, M. F. Quigley, J. R. Almeida, Downloaded from D. Finke. 2014. Activated group 3 innate lymphoid cells promote T-cell- E. Gostick, Z. Yu, C. Carpenito, et al. 2011. A human memory T cell subset with mediated immune responses. Proc. Natl. Acad. Sci. USA 111: 12835–12840. stem cell-like properties. Nat. Med. 17: 1290–1297. http://www.jimmunol.org/ by guest on September 23, 2021