The Interactions of Multiple Control NK Cell Maturation Jason Brady, Sebastian Carotta, Rebecca P. L. Thong, Christopher J. Chan, Yoshihiro Hayakawa, Mark J. Smyth This information is current as and Stephen L. Nutt of September 24, 2021. J Immunol 2010; 185:6679-6688; Prepublished online 25 October 2010; doi: 10.4049/jimmunol.0903354

http://www.jimmunol.org/content/185/11/6679 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2010/10/25/jimmunol.090335 Material 4.DC1 http://www.jimmunol.org/ References This article cites 63 articles, 33 of which you can access for free at: http://www.jimmunol.org/content/185/11/6679.full#ref-list-1

Why The JI? Submit online.

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

by guest on September 24, 2021 • 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 All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

The Interactions of Multiple Cytokines Control NK Cell Maturation

Jason Brady,*,1 Sebastian Carotta,* Rebecca P. L. Thong,* Christopher J. Chan,† Yoshihiro Hayakawa,†,2 Mark J. Smyth,† and Stephen L. Nutt*

Although NK cells are well known for their cytotoxic functions, they also produce an array of immunoregulatory cytokines and . During an immune response, NK cells are exposed to complex combinations of cytokines that influence their differ- entiation and function. In this study, we have examined the phenotypic and functional consequences of exposing mouse NK cells to IL-4, IL-12, IL-15, IL-18, and IL-21 and found that although all factors induced signs of maturation, characterized by decreased proliferation and IFN-g secretion, distinct combinations induced unique secretion profiles. In contrast, the immunosup- pressive factors IL-10 and TGF-b had little direct effect on NK cell effector functions. Sustained IL-18 signals resulted in IL-13 and GM-CSF production, whereas IL-12 and IL-21 induced IL-10 and TNF-a. Surprisingly, with the exception of IL-21, all Downloaded from cytokines suppressed cytotoxic function of NK cells at the expense of endogenous cytokine production suggesting that “helper- type” NK cells were generated. The cytokine signals also profoundly altered the cell surface phenotype of the NK cells—a striking example being the downregulation of the activating receptor NKG2D by IL-4 that resulted in decreased NKG2D-dependent killing. IL-4 exposure also modulated NKG2D expression in vivo suggesting it is functionally important during immune responses. This study highlights the plasticity of NK cell differentiation and suggests that the relative abundance of cytokines at sites of inflammation will lead to diverse outcomes in terms of NK cell phenotype and interaction with the immune system. The Journal http://www.jimmunol.org/ of Immunology, 2010, 185: 6679–6688.

atural killer cells are components of the innate immune the blood and lungs. In vitro functional assays indicate that the system that play a protective role against some tumors CD27+ NK cells are more potent effector cells as measured by N and viral through their cytolytic and immu- cytotoxicity and cytokine secretion (6). Recently, CD94 has been nomodulatory capabilities. In contrast with B and T lymphocytes proposed as an alternative marker of NK cell heterogeneity; where there is abundant evidence for multiple sublineages and however, the relationship between the subsets defined by CD27 functional states, the nature and consequences of heterogeneity in and CD94 is at present unclear (8). the NK cell lineage are only now emerging. Although it is well Human NK cells can be divided into CD56bright cells capable of by guest on September 24, 2021 established that a predominant pathway of murine NK cell de- enhanced proliferation and cytokine production and CD56dim cells velopment occurs in the bone marrow, a thymic pathway char- that display more potent cytotoxic function (9). Some data suggest acterized by a dependence on IL-7 and the transcription factor that the CD56bright NK cells are the equivalent of the thymus- Gata3 has been revealed (1, 2). Phenotypically distinct NK cell dependent population in the mouse, whereas others have pro- populations are also found in the uterus and gastrointestinal tract posed that the CD56bright population derives from hematopoietic (3–5). We have shown that maturing NK cells can be divided on progenitors in the lymph node (10–12). The factors that regulate the basis of CD27 and/or KLRG1 expression to distinguish dis- NK cell heterogeneity are poorly characterized, although by tinct subsets (6, 7). The CD27+ NK cells predominate in the bone analogy to other lymphocyte lineages, it is likely that extrinsic marrow and spleen and are selectively recruited into lymph nodes factors such as cytokines play a role. upon immunization, whereas the CD272 subset is found mainly in Mouse NK cell progenitors express IL-2/-15Rb-chain (CD122) in the absence of NK1.1 (13). Thereafter, NK1.1 expression is *The Walter and Eliza Hall Institute of Medical Research, Parkville; and †Cancer Immu- maintained throughout the lineage and is followed by the rapid nology Program, Sir Donald and Lady Trescowthick Laboratories, Peter MacCallum acquisition of the NKG2-CD94 and Ly49 receptor families (8, 14). Cancer Centre, East Melbourne, Victoria, Australia Besides IL-15, which is continually required for cell survival (15, 1Current address: Cancer Immunology Program, Sir Donald and Lady Trescowthick 16), little is known about the requirements of other cytokines Laboratories, Peter MacCallum Cancer Centre, East Melbourne, Victoria, Australia. in cellular development as mice deficient in one or multiples of 2 Current address: Graduate School of Pharmaceutical Sciences, The University of IL-2, IL-4, IL-7, IL-12, IL-18, and IL-21 all have relatively nor- Tokyo, Tokyo, Japan. mal numbers of peripheral NK cells (17–19). Received for publication October 13, 2009. Accepted for publication September 21, 2010. In contrast with their minor role in NK cell development, This work was supported by the National Health and Medical Research Council of cytokines such as IL-12, IL-4, IL-18, IL-21, and TGF-b have been Australia and by the Pfizer Australia Research Fellowship (to S.L.N.). shown to have pronounced effects on human and mouse NK cell Address correspondence and reprint requests to Dr. Stephen L. Nutt, The Walter and function (18, 20–27). IL-12 and IL-18 are produced by activated Eliza Hall Institute of Medical Research, 1G Royal Parade, Parkville, Victoria 3050, macrophages and dendritic cells and have an important function Australia. E-mail address: [email protected] in promoting Th1 responses and IFN-g production and tumor The online version of this article contains supplemental material. clearance by NK cells (18). IL-21 is a pleiotropic cytokine made Abbreviations used in this paper: PI, propidium iodide; TRAIL, TNF-related apoptosis- by activated CD4+ T cells (28) and NKT cells (29) that enhances inducing ligand. maturation from human multipotent progenitors and activates Copyright Ó 2010 by The American Association of Immunologists, Inc. 0022-1767/10/$16.00 peripheral NK cells (28). IL-21 is a potent inducer of IL-10 www.jimmunol.org/cgi/doi/10.4049/jimmunol.0903354 6680 NK CELL DIFFERENTIATION IN VITRO and IFN-g and promotes cytotoxicity against a broad spectrum of 31G2 biotin for detection, the TNF-a ELISA used TN3-19.12 for capture targets in vitro (22) and in vivo (30, 31). and rabbit polyclonal anti-mouse TNF biotin for detection, and the IL-4 As the NK cell differentiation pathways elicited by these in- ELISA used BVD4-1D11 for capture and BVD6-24G2.3 biotin for de- tection. Recombinant cytokine standards were purchased from R&D Sys- flammatory cytokines have many parallels to the Th1/Th2 dif- tems or PeproTech (Rocky Hill, NJ). ferentiation of CD4+ T cells, a number of studies have suggested that the NK cell cytokine responses can be categorized into NK1 Western blot analysis and NK2 type cells producing Th1 (IFN-g) or Th2 (IL-5, IL-13) Western blot analysis was performed as previously described (36). cytokines, respectively (24, 32–34). However, the diversity of cy- Hydrodynamic delivery tokines capable of supporting the functional maturation of NK 2 2 cells suggested to us that this model does not fully explain the Rag1 / mice were injected with 20 mg pOrf-Empty or pOrf–IL-4 hy- breadth of NK cell functions observed. In this study, we have drodynamic vectors in saline (37). The vectors were injected into the mice intravenously in a volume less than 10% of their body weight with systematically examined the phenotypic and functional conse- a maximum of 2 ml injected over a period of 3–5 s. NK cells and serum quences of exposing mouse NK cells to inflammatory cytokines IL-4 were analyzed 3 d postinjection. and demonstrate that the nature of these cytokine signals pro- RT-PCR foundly affected the cell surface phenotype and function of the resulting NK cells. Total RNA from NK cells was subjected to semiquantitative or quantitative RT-PCR as previously described (38). For semiquantitative RT-PCR, cDNA was normalized using Hprt1 expression, and 5-fold serial dilutions of the Materials and Methods cDNA were prepared. Hprt and Tbx21 primers were previously described Downloaded from Mice (39). Primers were: Gata3:59-CTACCGGGTTCGGATGTAAG-39,59-TG- CTAGACATCTTCCGGTTTC-39; Ets1:59-CAGCGGGTCCCCTCCTA- 2 2 Rag1 / mice were maintained on a C57/BL6 background. Mice at 5–10 TGA-39,59-CGTTGCCCGCCGTCTTG-39; Elf4:59-GGGGAAGGCAC- wk of age were used in all experiments. GGAAGAGAAA-39,59-GAAGTCGAGGCCTGGGAAGAGTC-39; Irf1: 59-GATGAGACCCTGGCTAGAGATG-39,59-TCTCTCTTTCCTCTGGT- NK cell isolation and in vitro culture TCCTG-39; Irf2:59-GAGTCTACCGGATGCTGCCCTTAT-39,59-GTCT- NK cells were enriched from the spleen of Rag12/2 mice using DX5- CCCCTCTGCGTTCTCTTCAT-39; Id2:59-ACCAGAGACCTGGACAG- conjugated microbeads (purity typically 70–95% CD49b+ cells; Miltenyi AAC-39,59-AAGCTGAGAAGGGAATTCAG-39. http://www.jimmunol.org/ Biotech, Auburn, CA). CD49b+NK1.1+TCRb2 NK cells were isolated from C57/BL6 spleens by flow cytometric sorting. NK cells were cultured for up to 7 d in IMDM supplemented with 10% FCS, 2 mM L-glutamine, Results and 50 mM 2-ME using the following recombinant cytokine concentrations NK cell maturation is associated with decreased proliferative (R&D Systems, Minneapolis, MN): human IL-15 50 ng/ml and TGF-b 5 potential in vitro ng/ml, mouse IL-21 100 ng/ml, IL-12 2 ng/ml, IL-18 50 ng/ml, IL-10 10 ng/ml, and IL-4 10 ng/ml. The numbers of viable NK cells were enu- To investigate the comparative roles of the cytokines IL-4, merated using trypan blue exclusion and annexin V staining/propidium IL-12, IL-18, and IL-21 in NK cell differentiation, we isolated iodide (PI) staining. Cell morphology was examined by cytospin and + 2/2 2 2 mature DX5 NK cells from the spleens of Rag1 mice using staining with May–Grunwald–Giemsa solutions. Pax5 / pro-B cells were ¨ 2/2 by guest on September 24, 2021 differentiated into NK cells as described (35) with the exception that IL-4 magnetic beads. The use of Rag1 mice removed T and NKT was added to some cultures. cells from the cultures, which may themselves produce regulatory cytokines in response to the exogenous factors. The NK cells were Cytotoxicity assays then cultured in the presence of optimal concentrations of cyto- The cytotoxicity activity of NK cells was assessed using RMA, RMA-S, and kines for up to 7 d. Culture of NK cells in IL-15 alone resulted in RMA-Rae1b tumor cell lines as we have previously described (22). Ex- pronounced proliferation and increased cumulative cell number, periments were performed at least three times in triplicate. whereas, as we have previously reported (22), IL-21 antagonized Abs and flow cytometry the proliferative effect of IL-15 by ∼10-fold (Fig. 1A). The ad- Abs to NK1.1 (PK136), CD11b (M1/70), CD16 (2.4G2), Ly49A (YE132), dition of the other cytokines involved in NK cell maturation, IL-4, Ly49C/I (SW5E6), Ly49D (4E5/E1), Ly49G2 (4D11), IFN-g (HB170 and IL-12, or IL-18, either alone or with IL-21, similarly antagonized XMG1.2), IL-10 (JE5S-2A5.1), GM-CSF (MP1-22E9 and MP1-31G6), the IL-15 proliferative effect (Fig. 1A). Significantly reduced cell and IL-4 (BVD4-1D11 and BVD6-24G2.3) were purified from hybrid- numbers were also observed when wild-type NK cells were cul- oma supernatants on -G Sepharose columns (Amersham Pharmacia tured in the same cytokines over a 7-d time course (Fig. 1B). Biotech, Arlington Heights, IL) and conjugated to allophycocyanin (Pro- Zyme, San Leandro, CA) or biotin (Pierce, Rockford, IL) when required. These cultures showed that statistically significant reductions in Abs to TNF-a (TN3-19.12), CD94 (18d3), CXCR4 (2B11), NKG2A/C/E NK cell numbers were observed for all NK cell maturation (20d5), Ly49F (HBF-719), KLRG1 (2F1), CD122 (TM-b1), CD25 (7D4), cytokines as early as day 5. A similar decrease in cell yield re- CD244 (2B4), CD43 (S7), c-Kit (2B8), IL-10 (SXC-1), CD49b (HMa2), sulted from the addition of TGF-b to the cultures, whereas IL- TCRb (H57-597), TNF-related apoptosis-inducing ligand (TRAIL), and TNF-a were purchased from BD Pharmingen (San Diego, CA). Anti- 10, another immunosuppressive cytokine, had no impact on cell perforin (OMAK-D), DNAM1 (10E5), NKp46 (29A1.4), and NKG2D number. To ascertain the mechanism by which the cytokines re- (CX5) were purchased from eBioscience (San Diego, CA). Anti–b-actin– duced the cell yields, cell viability and proliferative assays were HRP was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Goat performed. Although CFSE labeling studies demonstrated that cells anti-rat IgG HRP was purchased from Amersham Pharmacia Biotech. For in all conditions were proliferating vigorously (data not shown), flow cytometry, appropriate Abs were used to stain single-cell suspensions in PBS containing 2% FCS. Biotinylated Abs were revealed by PE-streptavidin analysis of the apoptotic rates on days 4 and 5 of the cultures, using (Southern Biotechnology Associates, Birmingham, AL). Cells were ana- a combination of annexin V and PI staining, revealed a pronounced lyzed using an LSR flow cytometer (BD Biosciences, Mountain View, upregulation in the frequency of apoptotic cells with the addition of CA), with dead cells excluded by PI staining. all cytokines except IL-15 (Fig. 1C). Cytokine production assays Besides the effects that the various cytokines had on cell survival, changes in phenotypes indicative of differentiation were observed, NK cells were cultured in appropriate cytokines for 6 d, washed, and 1 3 105 cells were seeded in fresh cytokine for 24 h after which supernatants including increased cell size. Cytocentrifuge and histological were harvested. IFN-g and IL-10 were assayed by ELISA as previously analysis revealed that IL-12, IL-18, and IL-21 all induced dramatic reported (22). The GM-CSF ELISA used MP1-22E9 for capture and MP1- morphological changes in the cultures, including increased granularity The Journal of Immunology 6681

FIGURE 1. Cytokine regulation of NK cell proliferation in vitro. A,NKcellswere isolated from Rag12/2 spleens using DX5+ magnetic beads and 1 3 105 cells cultured for 7 d in the indicated cytokines. Total viable cell number was determined by manual counting and trypan blue exclusion at day 7. Data are the mean 6 SD of five to eight independent experiments. *p , 0.001 compared with cells cultured in IL-15 alone using a paired t test. B, FACS-sorted CD49b+NK1.1+TCRb2 NK cells (0.85 3 105/well) isolated from C57/ BL6 spleens were cultured in the indicated cytokines for 7 d. Total viable cell number was determined on days 3, 5, and 7. Data are the mean 6 SD of triplicate wells from a repre- sentative of two experiments. *p , 0.05 compared with cells cultured in IL-15 alone using a paired t test. C, Proportion of viable and apoptotic NK cells cultured as in A were determined after 5 d using annexin V stain- Downloaded from ing and PI exclusion. Live cells were annexin V2PI2, early apoptotic cells were annexin V+ PI2, and dead cells were annexin V+PI+. D, May–Gru¨nwald–Giemsa staining of day 7 cultures (original magnification 3100). Data in C and D are representative of at least two http://www.jimmunol.org/ experiments. and in many cases large cytoplasmic vacuoles (Fig. 1D). Taken and others have reported that IL-21 further boosts this cytolytic together, these data suggest that the reduced cell numbers ob- capacity against an array of targets (17, 22, 41). To test for syn- served in the cultures exposed to IL-4, IL-12, IL-18, or IL-21 re- ergistic effects of cytokine stimulation, we have tested lytic ac- sulted from increased cell death at later points in the time course, tivity of day 7 cultures against the relatively resistant RMA and potentially through the differentiation into effector NK cells. sensitive RMA-S (MHC-I2) targets. Whereas IL-15 and IL-15/IL- 21 showed the expected killing capacity against these lines, ad- Distinct cytokine secretion profiles are induced by different dition of IL-12, IL-18, IL-4, but not IL-10 or TGF-b, significantly by guest on September 24, 2021 culture conditions impaired their lytic function (Fig. 3A, Supplemental Fig. 1). This To address potential functions of cytokine-differentiated NK cells, inhibitory effect was most pronounced for the IL-15/IL-4, IL-15/ we have tested the secretion of a selection of known NK cell IL-12, and IL-15/IL-18 combinations that exhibited almost no products after 7 d culture in the various cytokine combinations. specific lysis. In keeping with the cytokine secretion data, these NK cells propagated in IL-15 alone produced baseline levels inhibitory effects were dominant over the prolytic IL-21 function. of IFN-g, MIP-1a (CCL3), and MIP-1b (CCL4; Fig. 2). These This inhibition of killing function was not the result of impaired factors were all further induced in vitro in the presence of IL-12, expression of perforin as Western blotting revealed high amounts IL-21, and/or IL-18. As previously noted, the combination of of perforin in all conditions (Fig. 3B) and did not correlate with an IL-12/IL-18 showed pronounced synergy in eliciting cytokine ability to degranulate, as assessed by cell surface CD107 staining secretion (40). Interestingly, GM-CSF, IL-13, IL-10, and TNF-a (data not shown) or expression of the death domain showed much more stringent cytokine requirements. GM-CSF TRAIL or (Fig. 4 and data not shown). and IL-13 were almost exclusive signatures of IL-18 stimulation, Modulation of the NK cell surface phenotype by whereas IL-10 was induced by IL-12 or IL-21 but completely cytokine-driven differentiation repressed by IL-18 signaling. TNF-a in contrast showed an in- termediate phenotype, being induced strongly by IL-12 and IL-21 We have previously reported that culturing NK cells in the pres- but only mildly repressed by IL-18. Finally IL-4, in conjunction ence of IL-21 modulates the expression of a number of impor- with IL-15, had little positive effect on cytokine secretion but did tant NK cell surface receptors (22). We have therefore examined the 2/2 inhibit IL-21 function. In support of these cytokine profiles being NK cell receptor expression for Rag1 and wild-type NK cells a consequence of cellular differentiation, assessment of cytokine cultured as in Fig. 1 using a panel of 25 markers (Figs. 4, 5, Sup- production after 1 d in culture revealed a similar profile of IFN-g plemental Figs. 2, 3, Supplemental Tables I, II). These markers secretion; however, IL-10 was not detectable. This observation represented an array of molecules that can be used to define the agrees with a previous report that found that NK cells secrete maturation state of NK cells in vivo, as well as their activation and IL-10 only after prolonged engagement with activating cyto- ability to respond to activating and inhibitory signals. As one of our kines (25). These data demonstrate that distinct cytokine signa- aims was to define the maturation states induced by various cyto- tures can be observed in vitro and that hierarchies of activity can kine combinations, this analysis was also expected to provide flow be readily discerned. cytometric signatures for particular NK cell states that can be further tested in vivo. Impaired cytotoxic function mediated by IL-4, IL-12, and IL-18 The pan-NK cell marker NK1.1 was significantly downregulated NK cells cultured in IL-15 or IL-2 are able to lyse efficiently an in some conditions, whereas other markers, such as CD122 (IL-2/ array of target cells suggesting that they are already fully active. We -15b), NKp46 (CD335), and CD244 (2B4), were uniformly posi- 6682 NK CELL DIFFERENTIATION IN VITRO

FIGURE 2. Cytokine production from cultured NK cells. Rag12/2 splenic NK cells were cultured for 6 d in the indicated con- ditions. Viable cells (1 3 105) were washed and reseeded in fresh cytokine for 24 h, after Downloaded from which supernatants were harvested and ex- amined for cytokine production using an ELISA. Data are the mean 6 SEM of trip- licate wells from a representative of at least three experiments. Il13 expression levels were determined by quantitative real-time PCR http://www.jimmunol.org/ and are shown relative to Hprt. by guest on September 24, 2021

tive and confirmed that these cells were NK cells (Fig. 5, Sup- IL-15 or IL-15/IL-12. Notably, IL-21 had no obvious effect in the plemental Table I). The NK cell maturation markers CD11b presence of IL-18, demonstrating again a hierarchy in the activity (Mac1), CD43, and c-Kit (CD117) were also downregulated in of these cytokines. Most surprisingly, culture in IL-4 or IL-12/IL- vitro and could not be used to determine NK cell differentiation 18 markedly downregulated this receptor complex as well as the status, although each again showed specific patterns, as the com- unrelated NKG2D molecule (Fig. 4). Culture of NK cells in the bination IL-12/IL-18 maintained CD43 expression whereas CD11b presence of TGF-b resulted in some modest suppression of the expression was silenced in the presence of IL-18. expression of several NK cell markers, including NK1.1, NKG2D, More decisively, a number of receptors that modulate NK cell and CD25, whereas IL-10 had little effect on receptor expression function or reflect activation status were strongly affected by (Supplemental Fig. 2, Supplemental Table I). particular culture conditions. KLRG1, a receptor we have recently Finally, we examined the expression of a panel of NK cell found to be expressed specifically on a subset of the most mature receptors encoded by the Ly49 cluster (Supplemental Fig. 2, NK cell in vivo (7, 42), was significantly induced by combinations Supplemental Table I). Expression of the Ly49 family was gen- of IL-12 and IL-18, suggesting that a similar maturation process erally decreased in all conditions except IL-15 6 IL-21. The only occurred in vitro (Fig. 4, Supplemental Table I). A similar profile exception was Ly49G2, which was maintained in all conditions. was observed for CD25 (IL-2Ra), whereas CD16, a receptor that The modulation of Ly49 receptors by cytokine treatment has re- is modulated on human NK cell subsets, was specifically repressed cently also been reported for Ly49E, which is induced by IL-15 or by IL-12/IL-18. One striking example was the expression of the IL-2 (43). homing receptor, CXCR4, which was expressed on splenic NK It was possible that many of the changes in cell surface Ag cells ex vivo but was strongly downregulated in all conditions expression observed were a consequence of the differential out- except IL-12/IL-18 (Fig. 4). growth of a subset of NK cells with a distinct phenotype. To ex- The analysis of NK cell receptors was also informative. The clude this scenario, we have performed experiments where NK CD94/NKG2 heterodimer was induced by IL-21 in the presence of cells were first activated and cultured in IL-15 for 4 d and then The Journal of Immunology 6683 Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021 FIGURE 3. Cytolytic activity of cultured NK cells. A, Rag12/2 splenic NK cells were cultured in the indicated cytokines for 7 d and subjected to FIGURE 4. Cytokine modulation of NK cell receptor expression. 2 2 a standard 51Cr release assay. Data are the mean of triplicate wells 6 SEM Rag1 / splenic NK cells were cultured for 7 d in the indicated conditions and are representative of at least three experiments. B, NK cells cultured and analyzed for the NK cell receptor expression by flow cytometry. 2 2 as in A were analyzed for perforin expression by Western blotting. b-actin Freshly isolated Rag1 / NK cells (gated as NK1.1+CD49b+) were used as was analyzed as a loading control. The p values compare the indicated a comparison (ex vivo). Data are representative of at least 10 experiments. culture condition with cells cultured in IL-15 alone: *p , 0.05; **p , 0.01. treated NK cells were relatively inefficient killers despite main- further differentiated in the indicated cytokines for a further 3 d. taining NKG2D expression, indicating that the inhibition of lytic Fig. 6A shows a representative experiment for the downregulation function occurred through a different mechanism. of NKG2D. Comparable results were obtained for the other NK To investigate whether IL-4 has the capacity to downregulate cell receptors assayed in Figs. 4 and 5, suggesting that the pheno- NKG2D expression in vivo, we have used the hydrodynamic gene typic diversity observed was elicited from a common starting pool therapy to express transiently mouse IL-4 (Fig. 7A,7B). Three days 2/2 of mature NK cells (data not shown). after the introduction of the IL-4 expression vector into Rag1 mice, ELISA could readily detect IL-4 in the serum (Fig. 7B). Downregulation of NKG2D by IL-4 Importantly, the NK cells from mice that received the pOrf–IL-4 The downregulation of the critical activating receptor, NKG2D, construct showed a clear downregulation of several NK cell re- by IL-4 and the combination, IL-12/IL-18, was surprising as only ceptors, including NK1.1, CD49b, CD94/NKG2, and NKG2D (Fig. TGF-b had previously been shown to decrease mouse and human 7A, Supplemental Fig. 4), demonstrating that IL-4 downregu- NKG2D activity (26, 44, 45) (Supplemental Fig. 2), although lated specific NK cell receptors including NKG2D in vivo. IL-21 (46), IL-12, and type I IFNs (27) are reported to decrease To address if IL-4 had any function during earlier stages of human NKG2D. To assess the functional consequences of the NK cell development, we have taken advantage of the NK cell NKG2D downregulation, we tested the ability of cultured NK differentiation capacity of Pax52/2 pro-B cells (35). Pax52/2 pro- cells to lyse RMA-Rae1b cells that express an activating NKG2D B cells are maintained on OP9 stromal cells in the presence of ligand and are predominantly killed by this pathway. In line with IL-7; however, transient exposure to notch signaling via coculture the decreased NKG2D expression, NK cells cultivated in the with OP9 cells expressing the notch ligand D-like 1 (OP9-DL1) presence of IL-4 or IL-12/IL-18 did not kill Rae1b+ cells (Fig. results in NK cell commitment (step 1) (35). These Pax52/2-de- 6B). In contrast, the combination IL-15/IL-18, IL-15/TGF-b, and rived NK cells can then be further matured by culture on OP9 cells IL-15/IL-10 maintained NKG2D expression and cytotoxic activ- in the presence of IL-15 (step 2), resulting in NK1.1+CD49b+ ity, whereas IL-15/IL-21 strongly promoted killing. IL-15/IL-12– NKG2D+CD94+ NK cells (Fig. 7C,7D). Introduction of IL-4 into 6684 NK CELL DIFFERENTIATION IN VITRO Downloaded from http://www.jimmunol.org/ FIGURE 6. Downregulation of NKG2D expression by IL-4 in vitro. A, Rag12/2 splenic NK cell cultures were cultured for 7 d in the indicated cytokines (upper panel) or for 3 d in IL-15 and then transferred into dif- ferentiation conditions for 4 d as indicated (lower panel). B, NK cells were isolated as in A and cultured for 7 d in the indicated cytokines and sub- jected to a standard 51Cr release assay using RMA-Rae1b+ (NKG2D li- gand+) target cells. Data are the mean of triplicate wells 6 SEM and are representative of at least three experiments. The p values compare the indicated culture conditions: *p , 0.05; **p , 0.01. by guest on September 24, 2021 pressed in all conditions, whereas Gata3 and Tbx21 (encoding FIGURE 5. Cytokine modulation of NK cell maturation marker ex- T-bet) showed some heterogeneity. Gata3, whose expression is pression. Rag12/2 or wild-type (DNAM1, NKp46) splenic NK cells were required for thymic and liver NK cell differentiation (10, 50), was cultured for 7 d in the indicated conditions and analyzed for differentiation significantly downregulated in the presence of IL-12, IL-21, and markers by flow cytometry. Freshly isolated Rag12/2 NK cells (gated as IL-18. Tbx21 in contrast was significantly decreased only in the NK1.1+CD49b+) were used as a comparison (ex vivo). Data are repre- presence of IL-15/IL-18 and strongly upregulated in IL-12/IL-18 sentative of 4–10 experiments. conditions that also induce the T-bet target gene Ifng. These data suggest that despite the strong phenotypic and functional differ- ences between the distinct cytokine-differentiated NK cells, the these cultures during step 1 resulted in lower NKG2D and reduced NK cell transcription factor profile was remarkably stable and not cell yields (Fig. 7C,7D), whereas addition of IL-4 only during the predominant cause of the distinct effector profiles. step 2 produced a phenotype reminiscent of the in vivo IL-4– treated mice; that is, decreased NK1.1, CD49b, NKG2D, and Discussion CD94. These data indicated that IL-4 had a strong effect on pro- The development and peripheral survival of mouse NK cells ab- liferation and gene expression in NK cells from an early stage in solutely depends on the cytokine IL-15 (13, 15). IL-15 can also differentiation. promote the survival and proliferation of NK cells in vitro; a function that can also be provided by high concentrations of IL-2. Transcriptional regulation of NK cell differentiation in vitro A number of other cytokines have been shown to influence It is well established that the distinct cytokine profiles produced NK cell maturation and effector function, including IL-12, IL-18, from in vitro-differentiated CD4+ T cells are the result of the activity IL-4, IL-21, and TGF-b. Importantly, none of these factors have of a number of lineage-promoting transcription factors, most no- been shown to be essential for NK cell development or homeo- tably, T-bet in Th1 cells (47), Gata3 in Th2 cells (48), and RAR- stasis in vivo (17–19). In this study, we have used primary NK related orphan receptor g t in Th17 cells (49). To examine if the cell cultures to address the interactions between these NK cell distinct NK cell fates we have observed in vitro are produced by modulating cytokines in controlling NK cell differentiation and differential transcription factor expression, we analyzed NK cells effector function. We found that distinct cytokine combinations cultivated in various cytokine conditions for the expression of a showed striking phenotypic, cellular, and functional consequences panel of factors that are essential for NK cell differentiation in vivo. in vitro and in the case of IL-4 in vivo. These findings suggest that Transcription factors were examined, first by semiquantitative the cytokine milieu present in situ during the early stages of an RT-PCR and then by real-time PCR (Fig. 8). Several factors, in- immune response are likely to have strong influences on the cluding Elf4 (Mef), Ets1, Irf1, Id2, and Irf2, were similarly ex- functional properties of the resulting NK cells. The Journal of Immunology 6685 Downloaded from http://www.jimmunol.org/

FIGURE 7. Downregulation of NKG2D expression by IL-4 in vivo and during early NK cell differentiation. A, Analysis of NKG2D expression in NK cells from the spleen of Rag12/2 injected with either pOrf or pOrf–IL-4. Rag12/2 mice were injected with PBS (control), empty vector (pOrf), or an IL-4– expressing vector (pOrf–IL-4) using the hydrodynamic gene transfer. At day 3, spleens were analyzed for NK cells (CD49b+NK1.1+). Note the lower expression of both markers after IL-4 treatment. Gated CD49b+NK1.1+ NK cells were analyzed for the indicated markers, including NKG2D. Values within the histograms are the mean fluorescent index 6 SD of four to six mice from each group. The p values compare the mice treated with pOrf–IL-4 and the control pOrf mice for the indicated marker using a paired t test. B, Serum IL-4 levels were determined after 3 d by ELISA. Data are the mean of triplicate 6 2/2 wells SEM from four to six mice. C and D, In vitro differentiation of NK cells from Pax5 pro-B cells in the presence or absence of IL-4. To induce by guest on September 24, 2021 NK cell differentiation, Pax52/2 pro-B cells were cultured for 7 d in the presence of IL-7 and OP9 stromal cells that express the notch ligand D-like 1 (OP9- DL1) before being transferred onto parental OP9 cells in the presence of IL-15 for a further 7 d. Parallel cultures were carried out in the presence of IL-4. Cultures were assayed for cell number (C) and expression of the indicated NK cell receptors (D). Data in C are the mean 6 SEM. C and D are repre- sentative of two identical experiments.

We have previously reported that culture of mature NK cells in resulted in increased production of a range of effector cytokines. IL-15 and IL-21 promoted the differentiation of mouse NK cells One group of cytokines including IFN-g,TNF-a, MIP-1a,and into effector cells that were strongly cytotoxic and secreted IFN-g MIP-1b were induced in all conditions and represent generic acti- and IL-10 (22). We have now addressed the role of IL-21 in vation responses, although in all cases further activation could be combination with other cytokines, including IL-12, IL-18 (18, 23, induced by specific stimuli such as IL-12/IL-18. Other NK cell 25, 40), and, to a lesser extent, IL-4 and TGF-b (23, 24, 26, 27, 43, products were however specifically produced by distinct cell pop- 51, 52), which have been reported to influence NK cell functions. ulations, with IL-13 and GM-CSF being a signature of IL-18 stim- As we predominately focused on understanding the cellular dif- ulation, whereas IL-10 was induced by IL-21 and IL-12. Diagnostic ferentiation process, we have not assayed the short-term effects of hierarchies of cytokine secretion were also apparent where IL-4 cytokine activation that may have distinct functions compared and IL-18 acted to inhibit IL-21 function. Surprisingly, of all the with that in longer-term cultures (53). A striking example of the cytokine combinations, only IL-21 promoted cytolytic function, difference between short- and long-term cultures is IL-10 ex- whereas IL-12, IL-18, and IL-4 actively suppressed target killing. pression. Whereas Il10 mRNA is rapidly induced in activated The mechanism for this reduced killing activity remains to be de- NK cells, IL-10 is not secreted until much later in an immune termined as the cytokine-inhibited NK cells retained high perforin response (54). expression and underwent some target-dependent degranulation. All culture conditions, except IL-15 alone, decreased cell pro- Moreover, the expression of other cytotoxic effector molecules such liferation and increased the proportion of cells undergoing apoptosis. as TRAIL and Fas ligand did not correlate with lytic potential. Taken This finding, along with the marked changes in cell size, morphology, together, these data suggest that, with the exception of IL-21, the in- and receptor expression, suggests that the predominant consequence flammatory cytokines favored the differentiation of NK cells with a of exposure to these inflammatory cytokines is to induce the NK cell “helper-type” capacity to produce immunoregulatory cytokines. Simi- lineage terminal differentiation program. Of the immunosuppres- lar helper NK cells have been proposed for human NK cells (23, 25). sive cytokines, IL-10 had very little effect on NK cell phenotype and These pronounced functional differences between the differ- function, whereas TGF-b very strongly antiproliferative but left entiated NK cells also correlated with dramatically altered cell NK cell effector function largely intact. In agreement with this surface profiles and repertoires of activating and inhibitory NK concept, exposure of NK cells to either IL-4, IL-12, IL-18, or IL-21 receptors. Of particular interest, many pan-NK cell lineage markers 6686 NK CELL DIFFERENTIATION IN VITRO

ligand–dependent cytotoxic function. The repressive effect of IL-4 occurred not only in vitro but also in developing NK cells in vivo. Previous studies have reported that TGF-b, IL-12, and type I IFN suppress NKG2D expression and activity in human NK cells (26, 27), and clinical studies have shown that reduced NKG2D expression is associated with high risk of progression of myelo- dysplastic syndrome to acute myeloid (59). The sup- pressive effects of IL-4 and IL-12/IL-18 on NKG2D in vitro do not appear to be mediated through autocrine secretion as TGF-b levels were low and unchanged in all NK cell culture conditions tested, and exogenous TGF-b had only a modest effect on cell surface NKG2D. IL-21 was also reported to decrease NKG2D on human NK cells (46); however, we observed no effects of IL-21 on either NKG2D expression level or cytotoxic function in mouse cells, suggesting that some species-specific differences may also occur in the function of IL-21 in the NK cell lineage. There is increasing evidence for diversity of the NK cell lineage. This includes anatomically distinct developmental and differenti-

ation sites, such as thymus, uterus, and gut, as well as the generation Downloaded from of functionally distinct subsets (1–3, 9). Our data describe potential cytokine signals by which NK cell subsets can be produced in vitro and, potentially, in vivo. It is well established that distinct cytokine profiles are produced from in vitro-differentiated CD4+ T cells as a result of the activity of a number of lineage-promoting tran-

scription factors. These factors function to directly regulate the http://www.jimmunol.org/ differentiation of these lineages as well as the diagnostic effector profiles, IFN-g in the case of T-bet and IL-4 for Gata3 (47, 48). Based on the production of Th1/2 type cytokines, a similar NK1 and NK2 model has been proposed for human and mouse NK cells (24, 33). The studies presented here highlight the diversity of NK cell helper and effector functions and indicate that NK cells do not fit into a clear NK1/NK2 model. Indeed, in the past few years it has become apparent that CD4+ responses are much more di- verse and flexible than the Th1/Th2 paradigm allows as other sub- by guest on September 24, 2021 lineages such as Th17 cells (regulated by RAR-related orphan recep- tor g t), regulatory T cells (controlled by Foxp3), and most recently FIGURE 8. Transcription factor expression in cultured NK cells. Gene 2/2 T follicular helper cells (Bcl6) have emerged (60). This diversity expression analysis of Rag1 splenic NK cells that were cultured in the is further complicated by the existence of Th1/Th2 type programs indicated cytokines for 7 d. A, Fivefold serial dilutions of the cDNA were within the Foxp3 expressing regulatory T cell lineage (61, 62). subjected to semiquantitative RT-PCR analysis for the indicated . B, Quantitative real-time PCR analysis for Tbx21 and Gata3. The quantity of Although our data are compatible with a similar diversity of input cDNA was normalized to Hprt. Data are the mean 6 SEM of trip- NK cells, our analysis of the expression of several transcriptional licate measurements. A and B are representative of two identical experi- regulators implicated in NK cell maturation found an overall very ments. The p values compare the gene expression from the indicated similar expression profile from these otherwise distinct NK cells. culture condition with cells cultured in IL-15 alone: *p , 0.05. This included identical expression of Elf4, Ets1, Irf1, Irf2, and Id2, all known to be important for NK cell differentiation (63). Small were downregulated. Although previous studies have shown that changes were observed for Gata3 (decreased in IL-12 and IL-18) IL-21 (22) or IL-12/IL-18 (55) exposure decreased NK1.1 and that and Tbx21 (increased in IL-12/IL-18); however, these changes CD49b expression was lost upon culture in IL-2 (56), few studies were modest compared with the strongly polarized gene expres- have systematically examined NK cell gene expression in re- sion in similarly cultured CD4+ T cells (32). Although these sponse to differentiation signals. The downregulation of many findings may simply represent the fact that the crucial polarizing markers commonly associated with the NK cell lineage, such as transcription factor(s) has yet to be found, it may also reflect a NK1.1 and CD49b, raise the possibility that prolonged exposure to greater plasticity of NK cells that possess a much more limited inflammatory cytokines may make NK cells difficult to identify capacity to clonally expand in response to Ag compared with using standard FACS conditions in situations such as or T cells. This plasticity thus enables the NK cells to respond rap- autoimmunity. Indeed, IL-21 is known to repress NK1.1 expres- idly to the distinctive cytokines produced as a reaction to partic- sion in vivo (22), and here we show a similar effect of IL-4 ex- ular challenges in an immunologically appropriate manner. posure. A similar loss of NK1.1 expression occurs in activated NKT cells (57). These studies suggest that the inflammatory en- Acknowledgments vironment and the exact cytokine milieu will profoundly affect We thank A. Kallies and D. Andrews for critical comments on the manu- NK cell plasticity and function. script, K. Elder for animal husbandry, and J.C. Zuniga-Pflucker (University The regulation of NKG2D was particularly striking. Under of Toronto, Toronto, Canada) for the OP9 and OP9-DL1 cell lines. resting conditions, NKG2D is expressed on all mouse NK cells (58). However, we found that NKG2D expression can be repressed Disclosures by exposure to IL-4 or IL-12/IL-18 resulting in reduced NKG2D- The authors have no financial conflicts of interest. The Journal of Immunology 6687

References 27. Muntasell, A., G. Magri, D. Pende, A. Angulo, and M. Lo´pez-Botet. 2010. In- hibition of NKG2D expression in NK cells by cytokines secreted in response to 1. Huntington, N. D., C. A. Vosshenrich, and J. P. Di Santo. 2007. Developmental human cytomegalovirus infection. Blood 115: 5170–5179. Nat. pathways that generate natural-killer-cell diversity in mice and humans. 28. Spolski, R., and W. J. Leonard. 2008. Interleukin-21: basic biology and impli- Rev. Immunol. 7: 703–714. cations for cancer and autoimmunity. Annu. Rev. Immunol. 26: 57–79. 2. Hayakawa, Y., N. D. Huntington, S. L. Nutt, and M. J. Smyth. 2006. Functional 29. Coquet, J. M., K. Kyparissoudis, D. G. Pellicci, G. Besra, S. P. Berzins, subsets of mouse natural killer cells. Immunol. Rev. 214: 47–55. M. J. Smyth, and D. I. Godfrey. 2007. IL-21 is produced by NKT cells and 3. Colonna, M. 2009. Interleukin-22-producing natural killer cells and lymphoid modulates NKT cell activation and cytokine production. J. Immunol. 178: 2827– tissue inducer-like cells in mucosal immunity. Immunity 31: 15–23. 2834. 4. Satoh-Takayama, N., C. A. Vosshenrich, S. Lesjean-Pottier, S. Sawa, M. Lochner, 30. Smyth, M. J., M. E. Wallace, S. L. Nutt, H. Yagita, D. I. Godfrey, and F. Rattis, J. J. Mention, K. Thiam, N. Cerf-Bensussan, O. Mandelboim, et al. 2008. Y. Hayakawa. 2005. Sequential activation of NKT cells and NK cells provides Microbial flora drives interleukin 22 production in intestinal NKp46+ cells that effective innate immunotherapy of cancer. J. Exp. Med. 201: 1973–1985. provide innate mucosal immune defense. Immunity 29: 958–970. 31. Takaki, R., Y. Hayakawa, A. Nelson, P. V. Sivakumar, S. Hughes, M. J. Smyth, 5. Trowsdale, J., and A. Moffett. 2008. NK receptor interactions with MHC class I and L. L. Lanier. 2005. IL-21 enhances tumor rejection through a NKG2D- molecules in pregnancy. Semin. Immunol. 20: 317–320. dependent mechanism. J. Immunol. 175: 2167–2173. 6. Hayakawa, Y., and M. J. Smyth. 2006. CD27 dissects mature NK cells into two 32. Katsumoto, T., M. Kimura, M. Yamashita, H. Hosokawa, K. Hashimoto, subsets with distinct responsiveness and migratory capacity. J. Immunol. 176: A. Hasegawa, M. Omori, T. Miyamoto, M. Taniguchi, and T. Nakayama. 2004. 1517–1524. STAT6-dependent differentiation and production of IL-5 and IL-13 in murine 7. Huntington, N. D., H. Tabarias, K. Fairfax, J. Brady, Y. Hayakawa, M. A. Degli- NK2 cells. J. Immunol. 173: 4967–4975. Esposti, M. J. Smyth, D. M. Tarlinton, and S. L. Nutt. 2007. NK cell maturation 33. Kimura, M. Y., and T. Nakayama. 2005. Differentiation of NK1 and NK2 cells. and peripheral homeostasis is associated with KLRG1 up-regulation. J. Immunol. Crit. Rev. Immunol. 25: 361–374. 178: 4764–4770. 34. Loza, M. J., L. Zamai, L. Azzoni, E. Rosati, and B. Perussia. 2002. Expression of 8. Yu, J., M. Wei, H. Mao, J. Zhang, T. Hughes, T. Mitsui, I. K. Park, C. Hwang, type 1 (interferon gamma) and type 2 (interleukin-13, interleukin-5) cytokines at S. Liu, G. Marcucci, et al. 2009. CD94 defines phenotypically and functionally distinct stages of differentiation from progenitor cells. Blood distinct mouse NK cell subsets. J. Immunol. 183: 4968–4974.

99: 1273–1281. Downloaded from 9. Caligiuri, M. A. 2008. Human natural killer cells. Blood 112: 461–469. 35. Carotta, S., J. Brady, L. Wu, and S. L. Nutt. 2006. Transient Notch signaling 10. Vosshenrich, C. A., M. E. Garcı´a-Ojeda, S. I. Samson-Ville´ger, V. Pasqualetto, induces NK cell potential in Pax5-deficient pro-B cells. Eur. J. Immunol. 36: L. Enault, O. Richard-Le Goff, E. Corcuff, D. Guy-Grand, B. Rocha, A. Cumano, 3294–3304. et al. 2006. A thymic pathway of mouse natural killer cell development charac- 36. Smyth, M. J., R. W. Johnstone, E. Cretney, N. M. Haynes, J. D. Sedgwick, terized by expression of GATA-3 and CD127. Nat. Immunol. 7: 1217–1224. H. Korner, L. D. Poulton, and A. G. Baxter. 1999. Multiple deficiencies underlie 11. Freud, A. G., A. Yokohama, B. Becknell, M. T. Lee, H. C. Mao, A. K. Ferketich, NK cell inactivity in lymphotoxin-alpha gene-targeted mice. J. Immunol. 163: and M. A. Caligiuri. 2006. Evidence for discrete stages of human natural 1350–1353. killer cell differentiation in vivo. J. Exp. Med. 203: 1033–1043. 37. Song, Y. K., F. Liu, G. Zhang, and D. Liu. 2002. Hydrodynamics-based trans- 12. Freud, A. G., B. Becknell, S. Roychowdhury, H. C. Mao, A. K. Ferketich, fection: simple and efficient method for introducing and expressing transgenes in http://www.jimmunol.org/ G. J. Nuovo, T. L. Hughes, T. B. Marburger, J. Sung, R. A. Baiocchi, et al. 2005. animals by intravenous injection of DNA. Methods Enzymol. 346: 92–105. A human CD34(+) subset resides in lymph nodes and differentiates into 38. Kallies, A., A. Xin, G. T. Belz, and S. L. Nutt. 2009. Blimp-1 transcription factor CD56bright natural killer cells. Immunity 22: 295–304. is required for the differentiation of effector CD8(+) T cells and memory 13. Rosmaraki, E. E., I. Douagi, C. Roth, F. Colucci, A. Cumano, and J. P. Di Santo. responses. Immunity 31: 283–295. 2001. Identification of committed NK cell progenitors in adult murine bone 39. Kallies, A., E. D. Hawkins, G. T. Belz, D. Metcalf, M. Hommel, L. M. Corcoran, marrow. Eur. J. Immunol. 31: 1900–1909. P. D. Hodgkin, and S. L. Nutt. 2006. Transcriptional repressor Blimp-1 is es- 14. Kim, S., K. Iizuka, H. S. Kang, A. Dokun, A. R. French, S. Greco, and sential for T cell homeostasis and self-tolerance. Nat. Immunol. 7: 466–474. W. M. Yokoyama. 2002. In vivo developmental stages in murine natural 40. Lauwerys, B. R., N. Garot, J. C. Renauld, and F. A. Houssiau. 2000. Cytokine killer cell maturation. Nat. Immunol. 3: 523–528. production and killer activity of NK/T-NK cells derived with IL-2, IL-15, or the 15. Kennedy, M. K., M. Glaccum, S. N. Brown, E. A. Butz, J. L. Viney, M. Embers, combination of IL-12 and IL-18. J. Immunol. 165: 1847–1853. N. Matsuki, K. Charrier, L. Sedger, C. R. Willis, et al. 2000. Reversible defects

41. Parrish-Novak, J., D. C. Foster, R. D. Holly, and C. H. Clegg. 2002. Interleukin- by guest on September 24, 2021 in natural killer and memory CD8 T cell lineages in -deficient 21 and the IL-21 receptor: novel effectors of NK and T cell responses. J. Leukoc. mice. J. Exp. Med. 191: 771–780. Biol. 72: 856–863. 16. Huntington, N. D., H. Puthalakath, P. Gunn, E. Naik, E. M. Michalak, 42. Robbins, S. H., K. B. Nguyen, N. Takahashi, T. Mikayama, C. A. Biron, and M. J. Smyth, H. Tabarias, M. A. Degli-Esposti, G. Dewson, S. N. Willis, et al. L. Brossay. 2002. Cutting edge: inhibitory functions of the killer cell lectin-like 2007. Interleukin 15-mediated survival of natural killer cells is determined by receptor G1 molecule during the activation of mouse NK cells. J. Immunol. 168: interactions among Bim, Noxa and Mcl-1. Nat. Immunol. 8: 856–863. 2585–2589. 17. Kasaian, M. T., M. J. Whitters, L. L. Carter, L. D. Lowe, J. M. Jussif, B. Deng, 43. Gays, F., K. Martin, R. Kenefeck, J. G. Aust, and C. G. Brooks. 2005. Multiple K. A. Johnson, J. S. Witek, M. Senices, R. F. Konz, et al. 2002. IL-21 limits cytokines regulate the NK gene complex-encoded receptor repertoire of mature NK cell responses and promotes antigen-specific T cell activation: a mediator of NK cells and T cells. J. Immunol. 175: 2938–2947. the transition from innate to adaptive immunity. Immunity 16: 559–569. 44. Ghiringhelli, F., C. Me´nard, M. Terme, C. Flament, J. Taieb, N. Chaput, 18. Takeda, K., H. Tsutsui, T. Yoshimoto, O. Adachi, N. Yoshida, T. Kishimoto, P. E. Puig, S. Novault, B. Escudier, E. Vivier, et al. 2005. CD4+CD25+ regu- H. Okamura, K. Nakanishi, and S. Akira. 1998. Defective NK cell activity and latory T cells inhibit natural killer cell functions in a transforming growth factor- Th1 response in IL-18-deficient mice. Immunity 8: 383–390. beta-dependent manner. J. Exp. Med. 202: 1075–1085. 19. Vosshenrich, C. A., T. Ranson, S. I. Samson, E. Corcuff, F. Colucci, 45. Smyth, M. J., M. W. Teng, J. Swann, K. Kyparissoudis, D. I. Godfrey, and E. E. Rosmaraki, and J. P. Di Santo. 2005. Roles for common cytokine receptor Y. Hayakawa. 2006. CD4+CD25+ T regulatory cells suppress NK cell-mediated gamma-chain-dependent cytokines in the generation, differentiation, and matu- immunotherapy of cancer. J. Immunol. 176: 1582–1587. ration of NK cell precursors and peripheral NK cells in vivo. J. Immunol. 174: 46.Burgess,S.J.,A.I.Marusina,I.Pathmanathan,F.Borrego,andJ.E.Coligan. 1213–1221. 2006. IL-21 down-regulates NKG2D/DAP10 expression on human NK and 20. Strengell, M., S. Matikainen, J. Sire´n, A. Lehtonen, D. Foster, I. Julkunen, and CD8+ T cells. J. Immunol. 176: 1490–1497. T. Sareneva. 2003. IL-21 in synergy with IL-15 or IL-18 enhances IFN-gamma 47. Glimcher, L. H., M. J. Townsend, B. M. Sullivan, and G. M. Lord. 2004. Recent production in human NK and T cells. J. Immunol. 170: 5464–5469. developments in the transcriptional regulation of cytolytic effector cells. Nat. 21. Ortaldo, J. R., R. Winkler-Pickett, J. Wigginton, M. Horner, E. W. Bere, Rev. Immunol. 4: 900–911. A. T. Mason, N. Bhat, J. Cherry, M. Sanford, D. L. Hodge, and H. A. Young. 48. Zhu, J., B. Min, J. Hu-Li, C. J. Watson, A. Grinberg, Q. Wang, N. Killeen, 2006. Regulation of ITAM-positive receptors: role of IL-12 and IL-18. Blood J. F. Urban, Jr., L. Guo, and W. E. Paul. 2004. Conditional deletion of Gata3 107: 1468–1475. shows its essential function in T(H)1-T(H)2 responses. Nat. Immunol. 5: 1157– 22. Brady, J., Y. Hayakawa, M. J. Smyth, and S. L. Nutt. 2004. IL-21 induces the 1165. functional maturation of murine NK cells. J. Immunol. 172: 2048–2058. 49. Ivanov, I. I., B. S. McKenzie, L. Zhou, C. E. Tadokoro, A. Lepelley, J. J. Lafaille, 23. Agaugue´, S., E. Marcenaro, B. Ferranti, L. Moretta, and A. Moretta. 2008. D. J. Cua, and D. R. Littman. 2006. The orphan nuclear receptor RORgammat Human natural killer cells exposed to IL-2, IL-12, IL-18, or IL-4 differently directs the differentiation program of proinflammatory IL-17+ T helper cells. modulate priming of naive T cells by monocyte-derived dendritic cells. Blood Cell 126: 1121–1133. 112: 1776–1783. 50. Samson, S. I., O. Richard, M. Tavian, T. Ranson, C. A. Vosshenrich, F. Colucci, 24. Loza, M. J., and B. Perussia. 2001. Final steps of natural killer cell maturation: J. Buer, F. Grosveld, I. Godin, and J. P. Di Santo. 2003. GATA-3 promotes a model for type 1-type 2 differentiation? Nat. Immunol. 2: 917–924. maturation, IFN-gamma production, and liver-specific homing of NK cells. 25. Mailliard, R. B., S. M. Alber, H. Shen, S. C. Watkins, J. M. Kirkwood, Immunity 19: 701–711. R. B. Herberman, and P. Kalinski. 2005. IL-18-induced CD83+CCR7+ NK 51. Marcenaro, E., M. Della Chiesa, F. Bellora, S. Parolini, R. Millo, L. Moretta, and helper cells. J. Exp. Med. 202: 941–953. A. Moretta. 2005. IL-12 or IL-4 prime human NK cells to mediate functionally 26. Castriconi, R., C. Cantoni, M. Della Chiesa, M. Vitale, E. Marcenaro, R. Conte, divergent interactions with dendritic cells or tumors. J. Immunol. 174: 3992– R. Biassoni, C. Bottino, L. Moretta, and A. Moretta. 2003. Transforming growth 3998. factor beta 1 inhibits expression of NKp30 and NKG2D receptors: consequences 52. Morris, S. C., T. Orekhova, M. J. Meadows, S. M. Heidorn, J. Yang, and for the NK-mediated killing of dendritic cells. Proc. Natl. Acad. Sci. USA 100: F. D. Finkelman. 2006. IL-4 induces in vivo production of IFN-gamma by NK 4120–4125. and NKT cells. J. Immunol. 176: 5299–5305. 6688 NK CELL DIFFERENTIATION IN VITRO

53. Hyodo, Y., K. Matsui, N. Hayashi, H. Tsutsui, S. Kashiwamura, H. Yamauchi, 58. Raulet, D. H. 1999. Development and tolerance of natural killer cells. Curr. K. Hiroishi, K. Takeda, Y. Tagawa, Y. Iwakura, et al. 1999. IL-18 up-regulates Opin. Immunol. 11: 129–134. perforin-mediated NK activity without increasing perforin messenger RNA ex- 59. Epling-Burnette, P. K., F. Bai, J. S. Painter, D. E. Rollison, H. R. Salih, pression by binding to constitutively expressed IL-18 receptor. J. Immunol. 162: M. Krusch, J. Zou, E. Ku, B. Zhong, D. Boulware, et al. 2007. Reduced natural 1662–1668. killer (NK) function associated with high-risk myelodysplastic syndrome (MDS) 54. Maroof, A., L. Beattie, S. Zubairi, M. Svensson, S. Stager, and P. M. Kaye. 2008. and reduced expression of activating NK receptors. Blood 109: 4816–4824. Posttranscriptional regulation of II10 gene expression allows natural killer cells 60. Zhou, L., M. M. Chong, and D. R. Littman. 2009. Plasticity of CD4+ T cell to express immunoregulatory function. Immunity 29: 295–305. lineage differentiation. Immunity 30: 646–655. 55. Tomura, M., X. Y. Zhou, S. Maruo, H. J. Ahn, T. Hamaoka, H. Okamura, 61. Koch, M. A., G. Tucker-Heard, N. R. Perdue, J. R. Killebrew, K. B. Urdahl, and K. Nakanishi, T. Tanimoto, M. Kurimoto, and H. Fujiwara. 1998. A critical role D. J. Campbell. 2009. The transcription factor T-bet controls regulatory T cell for IL-18 in the proliferation and activation of NK1.1+ CD3- cells. J. Immunol. 160: 4738–4746. homeostasis and function during type 1 inflammation. Nat. Immunol. 10: 595– 56. Arase, H., T. Saito, J. H. Phillips, and L. L. Lanier. 2001. Cutting edge: the 602. mouse NK cell-associated antigen recognized by DX5 is 62. Zheng, Y., A. Chaudhry, A. Kas, P. deRoos, J. M. Kim, T. T. Chu, L. Corcoran, CD49b (alpha 2 integrin, very late antigen-2). J. Immunol. 167: 1141–1144. P. Treuting, U. Klein, and A. Y. Rudensky. 2009. Regulatory T-cell suppressor 57. Crowe, N. Y., A. P. Uldrich, K. Kyparissoudis, K. J. Hammond, Y. Hayakawa, program co-opts transcription factor IRF4 to control T(H)2 responses. Nature S. Sidobre, R. Keating, M. Kronenberg, M. J. Smyth, and D. I. Godfrey. 2003. 458: 351–356. Glycolipid antigen drives rapid expansion and sustained cytokine production by 63. Lian, R. H., and V. Kumar. 2002. Murine natural killer cell progenitors and their NK T cells. J. Immunol. 171: 4020–4027. requirements for development. Semin. Immunol. 14: 453–460. Downloaded from http://www.jimmunol.org/ by guest on September 24, 2021