Cancer NK Cell-Mediated Immunotherapy of T Regulatory Cells ...

The Journal of Immunology

CD4ⴙCD25ⴙ T Regulatory Cells Suppress NK Cell-Mediated Immunotherapy of Cancer1 Mark J. Smyth,2* Michele W. L. Teng,* Jeremy Swann,* Konstantinos Kyparissoudis,† Dale I. Godfrey,† and Yoshihiro Hayakawa* CD4ⴙCD25ⴙ regulatory T cells (Treg) that suppress T cell-mediated immune responses may also regulate other arms of an effective immune response. In particular, in this study we show that Treg directly inhibit NKG2D-mediated NK cell cytotoxicity in vitro and in vivo, effectively suppressing NK cell-mediated tumor rejection. In vitro, Treg were shown to inhibit NKG2Dmediated cytolysis largely by a TGF-␤-dependent mechanism and independently of IL-10. Adoptively transferred Treg suppressed NK cell antimetastatic function in RAG-1-deficient mice. Depletion of Treg before NK cell activation via NKG2D and the activating IL-12 cytokine, dramatically enhanced NK cell-mediated suppression of tumor growth and metastases. Our data illustrate at least one mechanism by which Treg can suppress NK cell antitumor activity and highlight the effectiveness of combining Treg inhibition with subsequent NK cell activation to promote strong innate antitumor immunity. The Journal of Immunology, 2006, 176: 1582–1587. D4⫹ CD25⫹ regulatory T cells (Treg)3 are a unique population of T cells that maintain immune tolerance and are critical in host suppression of organ-specific autoimmune diseases (1, 2). Treg also contribute to elicit a dominant tolerance during infections (3, 4) and allogeneic transplantation (5) and suppress immune responses to tumors at both the priming and effector phases (6 –9). Depletion of Treg in mice with anti-IL-2 receptor (PC61) Ab (7) or low dose cyclophosphamide depletion of Treg (9) improves T cell-based tumor clearance. In addition, cancer patients have increased numbers of peripheral and tumor-infiltrating Treg that functionally inhibit tumor-specific T cells and predict poor survival (10, 11). Tumor immunity is not only T cell dependent; recently, a number of studies in leukemia (12), lymphoma (13), and gastrointestinal stromal tumors (14) revealed that NK cell activation and cytotoxicity influence patient outcome. NK cells are regulated by cytokines in the environment, and when interacting with tumor directly, there is a delicate balance between inhibitory signals mediated by MHC class I molecules and activating signals triggered by specific ligands (15–17). In particular, the activating NKG2D receptor expressed by NK cells is an important mechanism of tumor recognition and suppression (18 –20). Tumors possess many mechanisms by which they might evade NK cell-mediated suppression (16), such as shedding of soluble ligands for activation

C

*Cancer Immunology Program, Trescowthick Laboratories, Peter MacCallum Cancer Centre, East Melbourne, Australia; and †Department of Microbiology and Immunology, University of Melbourne, Parkville, Australia Received for publication August 8, 2005. Accepted for publication November 12, 2005. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported by program grant funding from the National Health and Medical Research Council of Australia. M.J.S. and D.I.G. are currently supported by National Health and Medical Research Council of Australia Research Fellowships, J.S. by a University of Melbourne Australian Postgraduate Award, and Y.H. by a Cancer Research Institute Postdoctoral Fellowship. 2 Address correspondence and reprint requests to Dr. Mark Smyth, Cancer Immunology Program, Peter MacCallum Cancer Center, Locked Bag 1, A’Beckett Street, 8006 Victoria, Australia. E-mail address: [email protected] 3

Abbreviations used in this paper: Treg, regulatory T cell; m, mouse; WT, wild type.

Copyright © 2006 by The American Association of Immunologists, Inc.

receptors (21, 22) and secretion of inhibitory cytokines such as TGF-␤ (23–25); however, to date, the role that Treg might play in hampering NK cell activation has been largely ignored. In this study we present a series of experiments that convincingly demonstrate that Treg directly suppress NK cell cytotoxicity in vitro and in vivo, thus providing a novel rational approach to stimulating more powerful innate antitumor immunity.

Materials and Methods Mice Inbred C57BL/6 (B6) and BALB/c wild-type (WT) mice were purchased from the Walter and Eliza Hall Institute of Medical Research. The B6 RAG-1-gene-targeted mice were 10 generations backcrossed to B6 and were bred at the Peter MacCallum Cancer Center. Mice, 6 –12 wk of age, were used in all experiments that were performed according to animal experimental ethics committee guidelines.

Isolation of spleen NK cells, Treg, and cytotoxicity assay Spleen cells were depleted of B cells by labeling with anti-B220 (clone RA3-6B2), followed by incubation with sheep anti-rat IgG-conjugated immunomagnetic Dynabeads M-450 (Dynal Biotech). Bead-labeled cells were magnetically removed using a Dynal MPC-1 magnetic particle concentrator. Remaining cells were subsequently labeled with anti-FcR mAb (clone 2.4G2 grown in-house), CD4-FITC (clone RM4-5), anti-CD25PE (clone PC61), anti-NK1.1-PE-Cy7 (clone PK136), and anti-TCR␤-APC (clone H57-597; all purchased from BD Pharmingen), and specific cell populations were sorted using a FACSAria (BD Biosciences). Spleen NK cells were sorted as NK1.1⫹TCR␣␤⫺ cells, Treg were sorted as TCR␣␤highCD4⫹CD25high cells, and conventional CD4 T cells were sorted as TCR␣␤highCD4⫹CD25⫺ cells. The purities were standard (CD4⫹ T cells, ⬎99%; NK cells, ⬎98%; Treg, ⬎97%). The cytolytic activity of NK cells from PC61- or control Ig-treated mice was tested against tumor target cells at a number of E:T cell ratios against NK cell-sensitive targets in a standard 6-h 51Cr release assay as previously described (26). Where indicated, NK cells were cocultured at a 3:1, 1:1, or 1:3 ratio with CD4⫹CD25high or CD4⫹CD25⫺ T cells. Before coculture, in some groups CD4⫹CD25high or CD4⫹CD25⫺ T cells were preactivated with anti-CD3 and anti-CD28 bound to plastic plates for 3 h or with anti-CD3 and antiCD28 on plastic for 48 h with 500 U/ml IL-2 as previously described (27). In some experiments NK cells and activated CD4⫹CD25high T cells were cocultured or separated by Transwells, with targets and NK cells incubated together. In the same experiments the supernatants from 48-h activated T cells were cocultured with NK cells and tumor target cells. The coculture assay was performed in the presence of neutralizing purified anti-mouse 0022-1767/06/$02.00

The Journal of Immunology

1583 TGF-␤ (anti-mTGF-␤; 1D11.16), anti-mIL-10 (JES-2A5), or control Ig (30 ␮g/ml).

Flow cytometric analysis NK cells and Treg were assessed after PC61 treatment for activation and depletion, using the following reagents from BE Pharmingen: anti-␣␤TCRFITC or -allophycocyanin (H57-597), NK1.1-PE (PK136), anti-CD4 (GK1.5), and anti-CD25 (7D4). The stained cells were analyzed on a FACScan (BD Biosciences), and the data were processed using the CellQuest program (BD Biosciences).

Tumor cell lines The following standard experimental mouse tumor cell lines were used in vitro and in vivo: matched B16 melanoma (H-2b) and B16-Rae-1⑀ melanoma (provided by Dr. L. Lanier, University of California, San Francisco, CA), matched RMA-S lymphoma (H-2b) and RMA-S-Rae-1␤ lymphoma, and the 3LL Lewis lung carcinoma (H-2b). The maintenance of all tumor cell lines has been previously described (28).

Tumor models in vivo 3LL, B16, and B16-Rae1⑀ tumor cells were inoculated i.v. at the doses indicated, and on day 14, the surface metastases were counted with the aid of a dissecting microscope. Recombinant mouse IL-12 was provided by Genetics Institute, and the preparation of IL-12 was diluted in PBS immediately before use. RMA-S-Rae1␤ and RMA-S were inoculated s.c. at the doses indicated. Tumor growth was measured every second day with a caliper square as the product of two diameters.

NK and CD8 cell depletion To determine the effector subsets mediating tumor suppression, some groups of B6 and BALB/c mice were additionally depleted of NK cells or CD8⫹ cells using 100 ␮g of rabbit anti-asialoGM1 Ab i.p. (Wako Chemicals) or anti-CD8 (53-6.7) on days 0, 1, and 7 (after tumor inoculation), respectively, or both together as described previously (29). Specific NK cell and CD8⫹ T cell depletion was validated as previously described (29, 30).

Statistical analysis Significant differences in the number of metastases and tumor size were determined by unpaired Mann-Whitney U test, whereas significant differences in NK cell cytotoxicity were determined by Student’s t test. A value of p ⬍ 0.05 was considered significant.

Results Enhanced NK cell-mediated suppression of tumors expressing NKG2D ligands in the absence of Treg

FIGURE 1. Depletion of Treg enhances NKG2D-mediated tumor suppression by NK cells in vivo. A, Groups of five B6 WT mice received control Ig (250 ␮g i.p.) or anti-CD25 (250 ␮g i.p.) 1 day before tumor inoculation as indicated. All WT mice were then inoculated i.v. with increasing doses of 3LL tumor cells on day 0 as indicated. Some mice were additionally depleted of NK cells using anti-asGM1 treatment on days ⫺1, 0, and 7. Lungs were removed from mice on day 14, and metastatic nodules were quantified. Data are recorded as the mean ⫾ SEM, with the significance of Treg depletion defined by Mann-Whitney U test (ⴱ, p ⬍ 0.05). B, Groups of five B6 WT mice received control Ig (250 ␮g i.p.) or anti-CD25 (250 ␮g i.p.) 1 day before tumor inoculation as indicated. All WT mice were then inoculated i.v. with 106 and 105 doses of B16-Rae1⑀ or B16 tumor cells on day 0 as indicated. Lungs were removed from mice on day 14, and metastatic nodules were quantified. Data are recorded as the mean ⫾ SEM, with the significance of Treg depletion recorded as defined by Mann-Whitney U test (ⴱ, p ⬍ 0.05). C, Groups of five B6 WT mice received control Ig (250 ␮g i.p.) or anti-CD25 (250 ␮g i.p.) 1 day before tumor inoculation as indicated. All WT mice were then inoculated s.c. with 107 or 5 ⫻ 106 RMA-S-Rae1␤ tumor cells on day 0 as indicated. Mice were monitored every 2 days for tumor growth, and s.c. tumors were measured with a caliper along the perpendicular axes of the tumors. Mice were killed when tumors reached ⬎10 mm in diameter or the mice became moribund. Mice with no tumor growth were kept under observation for at least 100 days. Data presented are representative of two independent experiments for C.

To examine the importance of Treg in controlling NK cell-mediated immunity to tumors, we initially chose to examine the Rae-1⫹ 3LL experimental metastases model (31). Using a standard protocol for effectively depleting Treg using the anti-CD25 mAb (data not shown), we demonstrated that at lower tumor doses inoculated, groups of mice depleted of Treg displayed significantly lower numbers of 3LL lung metastases than mice receiving a control Ig (Fig. 1A). At the highest dose of 3LL inoculated, Treg depletion alone was without effect, suggesting that the natural host response was completely overwhelmed at this point. The enhanced rejection associated with Treg depletion was NK cell dependent, as illustrated by the loss of all host control of metastases with concomitant depletion of NK cells using either anti-asGM1 (Fig. 1A) or antiNK1.1 (data not shown). To specifically examine the effects of Treg depletion on NKG2D-mediated tumor suppression, B16-Rae1⑀ tumor cells were injected at two different doses that grew effectively in mice receiving control Ig (Fig. 1B). Depletion of Treg cells strikingly enabled host immunity to control and in some mice completely suppress the lung metastases of B16-Rae1⑀ tumor cells (Fig. 1B). Again, this suppression enabled by Treg depletion was NK cell dependent (data not shown). Suppression required NKG2D ligand on the tumor cells, because at these doses B16 tumor cells grew in B6 mice regardless of whether Treg had been depleted (Fig. 1C).

1584 To examine the effect of Treg depletion on tumor growth in a s.c. site and in a model system where tumor cells lack MHC class I expression and are rejected by NK cells using perforin, we used the RMA-SRae-1␤ tumor model (19). To examine the effects of Treg depletion, RMA-S-Rae1␤ tumor cells were injected at two different high doses that grew effectively in mice receiving control Ig (Fig. 1C). By contrast, depletion Treg cells strikingly enabled host immunity to control and in some mice reject the growth of large numbers of RMA-SRae1␤ tumor cells (Fig. 1C). At least some of the suppression was mediated by NKG2D ligand on the tumor cells at these doses, because RMA-S tumor cells grew just as effectively in B6 mice regardless of whether Treg cells had been depleted (data not shown). NKG2D-mediated NK cell cytotoxicity is suppressed by Treg cells We next examined whether Treg may directly inhibit NK cellmediated cytotoxicity. We first compared the abilities of naive and activated CD4⫹CD25high (Treg) and conventional CD4⫹CD25⫺ T cells to affect NK cell-mediated cytotoxicity in vitro. By sorting both NK and T cells from B6 mice and coculturing them at a ratio of 1:1, we were able to demonstrate the specific ability of in vitro 3-h activated Treg to suppress NK cell-mediated cytolysis of RMA-S-Rae1␤ target cells (Fig. 2A). Similar data were obtained at a ratio of 3:1, and consistent with the importance of cell density, at

FIGURE 2. Treg suppress NKG2Dmediated NK cell cytotoxicity in vitro. The cytolytic activity of purified NK cells from naive B6 WT mice was tested against tumor target cells at a number of E:T cell ratios against NK cell-sensitive targets, RMA-S-Rae-1␤ (A and E), RMA-S (B and F), B16Rae-1⑀ (C), and B16 (D) in a standard 6-h 51Cr release assay. In some samples, as indicated, NK cells were cocultured at a 1:1 ratio with CD4⫹CD25high or CD4⫹CD25⫺ T cells that had been concomitantly sorted by flow cytometry to 99% purity. In some experiments (A–D), the sorted CD4⫹CD25high or CD4⫹CD25⫺ T cells were preactivated with anti-CD3 and anti-CD28 bound to plastic plates for 3 h before coculture. In other experiments (E and F), the sorted CD4⫹CD25high or CD4⫹CD25⫺ T cells were preactivated with anti-CD3 and anti-CD28 bound to plastic plates and IL-2 (500 U/ml) for 48 h before coculture. Activated CD4⫹CD25high or CD4⫹CD25⫺ T cells were also examined for cytotoxicity in the absence of NK cells (A–F). Data are shown as the mean ⫾ SEM of triplicate samples, and the results presented are representative of two independent experiments for A–F. The significance of suppression by activated Treg was determined by Student’s t test (ⴱ, p ⬍ 0.05).

REGULATION OF NK CELL ACTIVATION a 1:3 ratio of Treg:NK cells, the suppression of Treg began to diminish (data not shown). Neither unstimulated Treg nor unstimulated or stimulated conventional CD4⫹ T cells had any effect on NK cell-mediated cytotoxicity in this assay (Fig. 2A). Surprisingly, 3-h activated Treg did not suppress NK cell-mediated cytotoxicity of RMA-S target cells to any detectable extent (Fig. 2B). In support of a specific effect of 3-h activated Treg on the NKG2Dmediated cytotoxicity of target cells, comparative data were obtained using B16-Rae1⑀ and B16 target cells (Fig. 2, C and D). In both RMA-S-Rae1␤ and B16-Rae-1⑀ experiments, the release from suppression of the NKG2D-mediated NK cell cytotoxicity was almost complete, given the residual non-NKG2D-mediated cytolysis (i.e., vs RMA-S or B16) displayed by NK cells against these targets. To examine whether additional activation of Treg might enhance and broaden Treg suppression of NK cell cytotoxicity, Treg were activated with anti-CD3/CD28 and IL-2 for 48 h before coculture (Fig. 2, E and F). Very similar data were obtained with Treg able to significantly and specifically suppress NK cellmediated cytolysis of RMA-S-Rae1␤ target cells. Treg suppression of NKG2D-mediated NK cell cytotoxicity is contact and TGF-␤ dependent Given the reported ability of TGF-␤ to inhibit lymphocyte cytotoxicity (23, 24), we next examined whether TGF-␤ produced by

The Journal of Immunology coculturing NK cells and 48-h activated Treg might be responsible for the suppression of NKG2D-mediated NK cell cytotoxicity of ligand-expressing tumor targets (Fig. 3A). Clearly, an anti-TGF-␤ mAb was able to restore NK cell-mediated cytotoxicity of RMARae-1␤ tumor cells, whereas IL-10, which is also produced by activated Treg (32), was without effect (Fig. 3A). Importantly, neutralization of TGF-␤ did not increase the NK cell cytotoxicity of these target cells in the absence of activated Treg or in the presence of activated conventional CD4⫹ T cells (Fig. 3B), suggesting that neither TGF-␤ production by NK cells or tumor cells was contributing to the suppression of NK cell-mediated cytotoxicity. Attempts to transfer suppression on NK cell cytotoxicity by using the supernatants of activated Treg were unsuccessful, suggesting that the functional TGF-␤ may be surface bound, rather than secreted by Treg (Fig. 3C). In concert with this supposition, when activated Treg were separated by Transwells from NK cells and tumor targets, suppression was not detected (Fig. 3C). Consistent with a need for continual direct exposure of Treg to NK cells to mediate suppression of cytotoxicity, spleen cells examined ex vivo from Treg-depleted and cIg-treated B6 mice displayed equivalent levels of cytotoxicity toward each of the target cells, RMA-S-Rae-1␤ and RMA-S (Fig. 3D), B16-Rae-1⑀, and B16 (data not shown). We did not detect any other alterations in spleen, liver, bone marrow, peripheral blood, or lymph node NK cell numbers, activation status (as determined by Mac-1 and CD69), or homeostatic proliferative capacity in naive B6 mice depleted of Treg (data not shown).

1585 NKG2D-mediated tumor metastasis suppression by NK cells is inhibited by Treg Because depletion with anti-CD25 mAb has been reported to lead to significant activation of CD4⫹ T cells (6), it was possible that such activated CD4⫹ T cells might be indirectly activating NK cells via their IL-2 production. To address whether Treg might directly inhibit NK cell-mediated tumor suppression of metastases, 48-h activated Treg or activated CD4⫹CD25⫺ T cells were adoptively transferred into B6 RAG-1-deficient mice 6 h after receiving i.v. B16-Rae-1⑀ tumor cells (Fig. 4). Strikingly, transfer of activated Treg, but not resting Treg or activated CD4⫹CD25⫺ T cells, enhanced B16-Rae-1⑀ tumor metastases, but was without effect upon B16 tumor metastasis in the lung (Fig. 4). Mice depleted of NK cells did not effectively control B16-Rae-1⑀ tumor metastases in RAG-1-deficient mice that had or had not received activated Treg (Fig. 4B). These data support the ability of activated Treg to directly suppress NK cell function both in vitro and in vivo. Treg depletion and IL-12 therapy synergize to promote NK cellmediated tumor suppression Recently, we have shown that cytokines such as IL-2 and IL-12 can mediate their antimetastatic activity via NK cells in large part via the NKG2D-NKG2D ligand pathway (31). Although IL-12 has been shown in some model systems to overcome the effects of Treg suppression of T cell responses, our data in this study raised the possibility that IL-12 therapy may be more effective against

FIGURE 3. Treg suppression of NKG2D-mediated NK cell cytotoxicity is contact and TGF-␤ dependent. A and B, Activated CD4⫹CD25high or CD4⫹CD25⫺ T cells were cocultured with NK cells at a 1:1 ratio in a 6-h 51Cr release assay using RMA-S-Rae-1␤ target cells (at the indicated E:T cell ratios: 20:1, 5:1, and 1:1) in the presence of neutralizing anti-mTGF-␤, anti-mIL-10, or control Ig (30 ␮g/ml). In some wells (B; E:T cell ratio of 20:1), NK cells were alternatively cocultured with activated CD4⫹CD25⫺ conventional T cells (Tcon). In following experiments (C), NK cells alone (NK) or cocultured with activated Treg (NK ⫹ Treg) or activated Treg supernatant (NK ⫹ Treg sup) were tested for cytotoxicity against RMA-S-Rae-1␤ tumor cells as described above. In the same experiment, NK cells and tumor cells were coincubated across Transwells from activated Treg (NK ⫹ Treg TW). D, Spleen cells harvested from mice treated with cIg or anti-CD25 mAb 24 h previously were examined in a 6-h 51Cr release assay against RMA-S or RMA-S-Rae-1␤ target cells at the E:T cell ratio shown. In all experiments (A–D) data are shown as the mean ⫾ SEM of triplicate samples. The significance of suppression by activated Treg was determined by Student’s t test (ⴱ, p ⬍ 0.05).

1586

FIGURE 4. NKG2D-mediated tumor metastasis suppression by NK cells is inhibited by Treg. Groups of five B6 RAG-1⫺/⫺ mice were inoculated i.v. with 105 B16-Rae-1e or 5 ⫻ 104 B16 tumor cells. A, Six hours later, mice were inoculated i.v. with 5 ⫻ 105 purified, activated (48 h anti-CD3/CD28 and IL-2) CD4⫹CD25⫹ T cells or similarly activated CD4⫹CD25⫺ T cells. Groups of tumor-inoculated mice were left untreated as controls (untreated). B, Six hours later, mice were inoculated i.v. with 5 ⫻ 105 purified resting or activated (48 h anti-CD3/CD28 and IL-2) CD4⫹CD25⫹ T cells or similarly resting or activated CD4⫹CD25⫺ T cells. Some groups of mice were additionally depleted of NK cells using antiNK1.1 mAb. Lungs were removed from mice on day 14, and metastatic nodules were quantified. Data are recorded as the mean ⫾ SEM, with the significance of Treg transfer or NK cell depletion determined by MannWhitney U test (ⴱ, p ⬍ 0.05). The results presented in A are representative of two independent experiments.

REGULATION OF NK CELL ACTIVATION

FIGURE 5. Depletion of Treg enhances IL-12 activation of NK cell antimetastatic activity in vivo. A, Groups of five B6 WT mice received control Ig (250 ␮g i.p.) or anti-CD25 (250 ␮g i.p.) 1 day before tumor inoculation as indicated. All WT mice were then inoculated i.v. with 5 ⫻ 105 3LL tumor cells on day 0 as indicated. Groups of mice were then treated with either PBS or 500 U of IL-12 i.p. on days 3–7 after tumor inoculation. B, Some groups of mice were additionally depleted of NK cells or CD8⫹ T cells on days ⫺1, 0, and 7. In both experiments, lungs were removed from mice on day 14, and metastatic nodules were quantified. Data are recorded as the mean ⫾ SEM, with the significance of Treg depletion determined by Mann-Whitney U test (ⴱ, p ⬍ 0.05). The results presented in A are representative of two independent experiments, whereas those in B are from a single experiment.

NK cells to the therapeutic activity of the NK cell-activating cytokine, IL-12.

Discussion NKG2D ligand-expressing tumors in a setting where Treg were depleted. Thus, we examined the effect of Treg depletion on IL-12 therapy at a dose of 3LL tumor cells at which Treg depletion alone was of no effect and the efficacy of IL-12 was limited. A combination of IL-12 and Treg depletion significantly reduced 3LL lung metastases compared with Treg depletion or IL-12 therapy alone (Fig. 5A). The combined effect of IL-12 and Treg depletion was completely dependent on NK cells and was independent of CD8⫹ cells (Fig. 5B). The combination of IL-12 and Treg depletion was shown to also be more effective than Treg depletion or IL-12 alone in BALB/c mice injected s.c. or when orthotopically implanted with the 4T1 mammary carcinoma that spontaneously metastasizes to lung and liver (data not shown). In this case the combination treatment suppressed s.c. tumors and distant metastases, but, again, the combined effect was NK cell dependent. In summary, these data from several different tumor models in two different inbred strains of mice indicated that Treg depletion can greatly sensitize

Although earlier studies have suggested a role for Treg in suppressing NK cell effector functions in vitro (6, 32, 33), we have defined at least one mechanism by which activated Treg can directly suppress NK cell function via the NKG2D pathways in vivo. We have extended the implications further by illustrating that relief of that Treg suppression can greatly enhance the functional activity of NK cells responding to the activating IL-12 cytokine, which is known to enhance the NKG2D pathway of NK cell activation (31). The suppressive effect of Treg-expressed TGF-␤ on NK cell cytotoxicity is not surprising given the previously reported ability of soluble TGF-␤ to reduce NK cell cytotoxicity and perforin gene transcription (24) and the role of perforin downstream of NKG2D receptor function (31). However, most consideration has previously been given to a likely scenario where tumor secretes TGF-␤ to suppress NK cell activation and effector functions, rather than a direct effect of Treg on NK cell cytotoxicity. Although we were unable to detect surface TGF-␤ on activated Treg

The Journal of Immunology (data not shown), functional Transwell and supernatant experiments suggested that activated Treg producing membrane-associated TGF-␤ suppressed the NKG2D-mediated cytotoxicity displayed by NK cells. Alternatively, because high E:T cell ratios were required to observe Treg suppression of NK cell cytotoxicity, a labile form of TGF-␤ may have been secreted by activated Treg, and this was not active enough for supernatants to suppress NK cell function. However, in vivo, even relatively few transferred activated Treg (compared with host NK cells) suppressed NK cellmediated control of tumor metastases. We do not want to suggest that the effect of Treg on NK cell function may be simplistic; indeed, to the contrary, Treg may suppress NK cell function by a number of different mechanisms. Interestingly, we were unable to detect any down-regulation of NKG2D on NK cells when exposed to activated Treg (data not shown). These data are somewhat in contrast with previous reports that showed that NKG2D expression on human NK cells was down-regulated by soluble TGF-␤ produced by tumor cells (25, 34). It remains to be determined whether Treg-expressed TGF-␤ might interfere with NKG2D signaling or NKG2D ligand expression on tumor cells. Treg may have direct, and possibly indirect, effects on NK cell homeostasis, activation, proliferation, migration, and function in the context of tumor challenge, and many additional experiments will be required to explore the consequences of depleting Treg cells on NK cell antitumor activity in vivo. Recent data also suggest that functional TGF-␤ can be provided by a non-Treg source (35), and thus, additional cell types that regulate NK cell antitumor activity must also be considered carefully when designing effective immunotherapies. NK cell function in cancer patients is often severely impaired, and it will be of interest to determine whether an inverse relationship between NK cell activation and cytotoxicity and Treg expansion is detected in these patients.

Acknowledgments We thank Mark Shannon for reagent acquisition, Shannon Griffiths for maintaining the mice, Nadeen Zerafa and Jonathan Coquet for producing and purifying mAb, and Dr. Lewis Lanier for kindly providing the B16 and B16-Rae1⑀ tumor cells.

Disclosures The authors have no financial conflict of interest.

References 1. Sakaguchi, S. 2004. Naturally arising CD4⫹ regulatory T cells for immunologic self-tolerance and negative control of immune responses. Annu. Rev. Immunol. 22: 531–562. 2. Shevach, E. M. 2002. CD4⫹ CD25⫹ suppressor T cells: more questions than answers. Nat. Rev. Immunol. 2: 389 – 400. 3. Kullberg, M. C., D. Jankovic, P. L. Gorelick, P. Caspar, J. J. Letterio, A. W. Cheever, and A. Sher. 2002. Bacteria-triggered CD4⫹ T regulatory cells suppress Helicobacter hepaticus-induced colitis. J. Exp. Med. 196: 505–515. 4. Belkaid, Y., C. A. Piccirillo, S. Mendez, E. M. Shevach, and D. L. Sacks. 2002. CD4⫹CD25⫹ regulatory T cells control Leishmania major persistence and immunity. Nature 420: 502–507. 5. Trenado, A., F. Charlotte, S. Fisson, M. Yagello, D. Klatzmann, B. L. Salomon, and J. L. Cohen. 2003. Recipient-type specific CD4⫹CD25⫹ regulatory T cells favor immune reconstitution and control graft-versus-host disease while maintaining graft-versus-leukemia. J. Clin. Invest. 112: 1688 –1696. 6. Shimizu, J., S. Yamazaki, and S. Sakaguchi. 1999. Induction of tumor immunity by removing CD25⫹CD4⫹ T cells: a common basis between tumor immunity and autoimmunity. J. Immunol. 163: 5211–5218. 7. Onizuka, S., I. Tawara, J. Shimizu, S. Sakaguchi, T. Fujita, and E. Nakayama. 1999. Tumor rejection by in vivo administration of anti-CD25 (interleukin-2 receptor ␣) monoclonal antibody. Cancer Res. 59: 3128 –3133. 8. Steitz, J., J. Bruck, J. Lenz, J. Knop, and T. Tuting. 2001. Depletion of CD25⫹ CD4⫹ T cells and treatment with tyrosinase-related protein 2-transduced dendritic cells enhance the interferon ␣-induced, CD8⫹ T-cell-dependent immune defense of B16 melanoma. Cancer Res. 61: 8643– 8646. 9. Turk, M. J., J. A. Guevara-Patino, G. A. Rizzuto, M. E. Engelhorn, S. Sakaguchi, and A. N. Houghton. 2004. Concomitant tumor immunity to a poorly immunogenic melanoma is prevented by regulatory T cells. J. Exp. Med. 200: 771–782.

1587 10. Woo, E. Y., C. S. Chu, T. J. Goletz, K. Schlienger, H. Yeh, G. Coukos, S. C. Rubin, L. R. Kaiser, and C. H. June. 2001. Regulatory CD4⫹CD25⫹ T cells in tumors from patients with early-stage non-small cell lung cancer and late-stage ovarian cancer. Cancer Res. 61: 4766 – 4772. 11. Curiel, T. J., G. Coukos, L. Zou, X. Alvarez, P. Cheng, P. Mottram, M. Evdemon-Hogan, J. R. Conejo-Garcia, L. Zhang, M. Burow, et al. 2004. Specific recruitment of regulatory T cells in ovarian carcinoma fosters immune privilege and predicts reduced survival. Nat. Med. 10: 942–949. 12. Ruggeri, L., M. Capanni, E. Urbani, K. Perruccio, W. D. Shlomchik, A. Tosti, S. Posati, D. Rogaia, F. Frassoni, F. Aversa, et al. 2002. Effectiveness of donor natural killer cell alloreactivity in mismatched hematopoietic transplants. Science 295: 2097–2100. 13. O’Hanlon, L. H. 2004. Natural born killers: NK cells drafted into the cancer fight. J. Natl. Cancer Inst. 96: 651– 653. 14. Borg, C., M. Terme, J. Taieb, C. Menard, C. Flament, C. Robert, K. Maruyama, H. Wakasugi, E. Angevin, K. Thielemans, et al. 2004. Novel mode of action of c-kit tyrosine kinase inhibitors leading to NK cell-dependent antitumor effects. J. Clin. Invest. 114: 379 –388. 15. Moretta, A., C. Bottino, M. C. Mingari, R. Biassoni, and L. Moretta. 2002. What is a natural killer cell? Nat. Immunol. 3: 6 – 8. 16. Smyth, M. J., Y. Hayakawa, K. Takeda, and H. Yagita. 2002. New aspects of natural-killer-cell surveillance and therapy of cancer. Nat. Rev. Cancer 2: 850 – 861. 17. Lanier, L. L. 2005. NK cell recognition. Annu. Rev. Immunol. 23: 225–274. 18. Diefenbach, A., E. R. Jensen, A. M. Jamieson, and D. H. Raulet. 2001. Rae1 and H60 ligands of the NKG2D receptor stimulate tumour immunity. Nature 413: 165–171. 19. Hayakawa, Y., J. M. Kelly, J. A. Westwood, P. K. Darcy, A. Diefenbach, D. Raulet, and M. J. Smyth. 2002. Cutting edge: tumor rejection mediated by NKG2D receptor-ligand interaction is dependent upon perforin. J. Immunol. 169: 5377–5381. 20. Cerwenka, A., J. L. Baron, and L. L. Lanier. 2001. Ectopic expression of retinoic acid early inducible-1 gene (RAE-1) permits natural killer cell-mediated rejection of a MHC class I-bearing tumor in vivo. Proc. Natl. Acad. Sci. USA 98: 11521–11526. 21. Wu, J. D., L. M. Higgins, A. Steinle, D. Cosman, K. Haugk, and S. R. Plymate. 2004. Prevalent expression of the immunostimulatory MHC class I chain-related molecule is counteracted by shedding in prostate cancer. J. Clin. Invest. 114: 560 –568. 22. Groh, V., J. Wu, C. Yee, and T. Spies. 2002. Tumour-derived soluble MIC ligands impair expression of NKG2D and T-cell activation. Nature 419: 734 –738. 23. Loeffler, C. M., M. J. Smyth, D. L. Longo, W. C. Kopp, L. K. Harvey, H. R. Tribble, J. E. Tase, W. J. Urba, A. S. Leonard, H. A. Young, et al. 1992. Immunoregulation in cancer-bearing hosts: down-regulation of gene expression and cytotoxic function in CD8⫹ T cells. J. Immunol. 149: 949 –956. 24. Smyth, M. J., S. L. Strobl, H. A. Young, J. R. Ortaldo, and A. C. Ochoa. 1991. Regulation of lymphokine-activated killer activity and pore-forming protein gene expression in human peripheral blood CD8⫹ T lymphocytes. Inhibition by transforming growth factor-␤. J. Immunol. 146: 3289 –3297. 25. Lee, J. C., K. M. Lee, D. W. Kim, and D. S. Heo. 2004. Elevated TGF-␤1 secretion and down-modulation of NKG2D underlies impaired NK cytotoxicity in cancer patients. J. Immunol. 172: 7335–7340. 26. Kelly, J. M., P. K. Darcy, J. L. Markby, D. I. Godfrey, K. Takeda, H. Yagita, and M. J. Smyth. 2002. Induction of tumor-specific T cell memory by NK cell-mediated tumor rejection. Nat. Immunol. 3: 83–90. 27. Thornton, A. M., and E. M. Shevach. 2000. Suppressor effector function of CD4⫹CD25⫹ immunoregulatory T cells is antigen nonspecific. J. Immunol. 164: 183–190. 28. Smyth, M. J., N. Y. Crowe, D. G. Pellicci, K. Kyparissoudis, J. M. Kelly, K. Takeda, H. Yagita, and D. I. Godfrey. 2002. Sequential production of interferon-␥ by NK1.1 ⫹ T cells and natural killer cells is essential for the antimetastatic effect of ␣-galactosylceramide. Blood 99: 1259 –1266. 29. Smyth, M. J., K. Y. Thia, E. Cretney, J. M. Kelly, M. B. Snook, C. A. Forbes, and A. A. Scalzo. 1999. Perforin is a major contributor to NK cell control of tumor metastasis. J. Immunol. 162: 6658 – 6662. 30. Smyth, M. J., N. Y. Crowe, and D. I. Godfrey. 2001. NK cells and NKT cells collaborate in host protection from methylcholanthrene-induced fibrosarcoma. Int. Immunol. 13: 459 – 463. 31. Smyth, M. J., J. Swann, J. M. Kelly, E. Cretney, W. M. Yokoyama, A. Diefenbach, T. J. Sayers, and Y. Hayakawa. 2004. NKG2D recognition and perforin effector function mediate effective cytokine immunotherapy of cancer. J. Exp. Med. 200: 1325–1335. 32. Trzonkowski, P., E. Szmit, J. Mysliwska, A. Dobyszuk, and A. Mysliwski. 2004. CD4⫹CD25⫹ T regulatory cells inhibit cytotoxic activity of T CD8⫹ and NK lymphocytes in the direct cell-to-cell interaction. Clin. Immunol. 112: 258 –267. 33. Wolf, A. M., D. Wolf, M. Steurer, G. Gastl, E. Gunsilius, and B. Grubeck-Loebenstein. 2003. Increase of regulatory T cells in the peripheral blood of cancer patients. Clin. Cancer Res. 9: 606 – 612. 34. Castriconi, R., C. Cantoni, M. Della Chiesa, M. Vitale, E. Marcenaro, R. Conte, R. Biassoni, C. Bottino, L. Moretta, and A. Moretta. 2003. Transforming growth factor ␤1 inhibits expression of NKp30 and NKG2D receptors: consequences for the NK-mediated killing of dendritic cells. Proc. Natl. Acad. Sci. USA 100: 4120 – 4125. 35. Fahlen, L., S. Read, L. Gorelik, S. D. Hurst, R. L. Coffman, R. A. Flavell, and F. Powrie. 2005. T cells that cannot respond to TGF-␤ escape control by CD4⫹CD25⫹ regulatory T cells. J. Exp. Med. 201: 737–746.