NK Cells Negatively Regulate Antigen Presentation and Tumor ...

The Journal of Immunology

NK Cells Negatively Regulate Antigen Presentation and Tumor-Specific CTLs in a Syngeneic Lymphoma Model1 Melissa A. Barber, Tong Zhang, Bethany A. Gagne, and Charles L. Sentman2 NK cells are known to kill tumor cells and produce proinflammatory cytokines that lead to the generation of tumor-specific CTLs. Many studies have used MHC class I-deficient tumor cells and/or adjuvants that induce NK cell responses. In this study, the focus was on less-immunogenic lymphoma cells that express MHC class I as a model to study NK cell responses to tumors that do not directly stimulate NK cell activation. When RMA tumor cells that expressed a truncated version of OVA, or RMA cells alone, were injected into mice that were depleted of NK cells, the mice developed an increased number of tumor-specific CTLs, increased IFN-␥ responses, and a higher amount of Ag presentation in draining LNs compared with mice with intact NK cells. These data suggest that NK cells can inhibit the development of effective adaptive immunity in the absence of signals that trigger NK cell activation. The Journal of Immunology, 2007, 178: 6140 – 6147.

N

atural killer cells are cells of the innate immune system that are able to quickly respond to viral infections and tumors. Their main effector functions include cytokine secretion and cellular cytotoxicity. Although these functions can have a direct effect on a pathogen or tumor, NK cells also promote the activity of other cells of the immune system, including dendritic cells (DCs),3 macrophages, Th1 cells, and CTLs. Therefore, NK cells play an important role in the development and regulation of adaptive immunity. Previous studies have demonstrated a role of NK cells in the development of CTL activity against tumors (1– 4). NK cells are also known to promote the maturation and activation of DCs (5–10). Most studies investigating NK cell effects on immunity to cancer have used either MHC class I-negative or NKG2D ligand-positive tumor cells. These tumor cells cannot engage inhibitory receptors for MHC class I and engage activating NK cell receptors, respectively. Although human tumor cells can down-regulate all MHC class I molecules on the cell surface, it is common for there to be selective down-modulation of MHC class I HLA-A or HLA-B alleles but not the HLA-C or HLA-E molecules, which are recognized by NK cell inhibitory receptors (11). In some cases, NK cells have been reported to limit the immune response. It has been reported that NK cells impair immune responses in allograft rejection and cancer, and data suggest that NK cell lysis of APCs may be responsible for this process (12–14). It is well documented that activated NK cells can kill immature DCs (15–18). In one case, it was reported that NK cells lysed activated

Department of Microbiology and Immunology, Dartmouth Medical School, Lebanon, NH 03756 Received for publication December 22, 2006. Accepted for publication March 19, 2007. 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 Grants CA101748 and AI07363 from the National Institutes of Health. 2 Address correspondence and reprint requests to Dr. Charles L. Sentman, Dartmouth Medical School, Dartmouth-Hitchcock Medical Center, 6W Borwell Building, One Medical Center Drive, Lebanon, NH 03756. E-mail address: charles.sentman@ dartmouth.edu 3

Abbreviations used in this paper: DC, dendritic cell; LN, lymph node.

Copyright © 2007 by The American Association of Immunologists, Inc. 0022-1767/07/$2.00 www.jimmunol.org

syngeneic CD4⫹ and CD8⫹ T cells (19). It remains unclear under which conditions NK cells promote or inhibit immune responses. We were interested in examining the contribution of NK cells to the development of adaptive immune responses to a less-immunogenic tumor. The RMA lymphoma is a MHC class I-positive, NKG2D ligand-negative tumor that causes tumors in 100% of C57BL/6 (B6) mice at doses of 104 cells (20). RMA tumors grow in a similar manner in the presence or absence of NK cells, which suggests that NK cells may not play a role in responses to these tumors (21, 22). Primary tumors develop over a period of time, and although they may express molecules that can activate or can be recognized by the immune system, many spontaneous tumors elicit a limited immune response. This may be due to “immune editing” of tumors by the host’s immune system, so that tumors that develop to become clinical problems are those that have managed to evade or inhibit host immunity. The aim of this study was to determine whether NK cells can contribute to host immunity or actively prevent immunity to poorly immunogenic tumors. In this study, we describe a role for NK cells that limits Ag presentation and CTL development in response to a syngeneic MHC class I⫹ lymphoma.

Materials and Methods Mice and cell lines Female C57BL/6 mice were purchased from the National Cancer Institute and The Jackson Laboratory. B6.CD1⫺/⫺ mice were provided by Dr. M. Exley (Harvard University, Boston, MA) (23). B6.Rag1⫺/⫺ mice were bred and maintained at Dartmouth Medical School. OT1 mice were purchased from The Jackson Laboratory. Mice were used at 7–9 wk of age. All animal work was conducted at the Dartmouth Medical School Animal Facility in accordance with institutional guidelines. RMA is a C57BL/6 T cell lymphoma. RMA/t-OVA was made by retroviral transduction with a vector containing a truncated OVA gene where nucleotides encoding aa 1– 40 (the leader peptide) were removed (24, 25). P815 cells expressing H-2Kb were made by retroviral transduction. MHC class I-restricted B3Z T cell hybridoma cells recognize an epitope between residues 257–264 of the OVA Ag associated with H-2Kb, and upon activation, B3Z cells express the LacZ gene (26). Complete medium was as follows: RPMI 1640 with L-glutamine, supplemented with 10% FBS (HyClone), 20 U/ml penicillin, 20 ␮g/ml streptomycin, 1 mM sodium pyruvate, 10 mM HEPES, 0.1 mM nonessential amino acids, and 50 ␮M 2-ME.

Tumor inoculation RMA or RMA/t-OVA cells in 100 ␮l of PBS or HBSS were injected s.c. into the shaved right flank of C57BL/6 mice. Tumors were monitored for


6141

FIGURE 2. NK cell-depleted mice inoculated with RMA/t-OVA cells have increased amounts of IFN-␥ production in response to OVA peptides. Spleen (A) and draining LN (B) cells from RMA/t-OVA-bearing mice, which were treated with anti-NK1.1 mAbs or control Abs, were stimulated with OVA257–264 peptide or medium only. Conditioned media were collected on day 5 and IFN-␥ was measured by ELISA. Values are the average of three independent experiments and are represented as percentage of medium only control (⫽100) for each individual. Error bars represent SEM.

b

FIGURE 1. RMA/t-OVA cells express H-2K /OVA peptide complexes on their surfaces and do not express NKG2D ligands. RMA cells were retrovirally transduced with a vector containing a truncated OVA gene. Cells were stained with 25D1 Ab, which measures expression of H-2Kb/ OVA peptide complexes (A), as well as soluble NKG2D-Ig to measure NKG2D ligand expression (B). Dashed lines represent isotype controls and solid lines show specific staining.

growth starting at day 5 and measured (length and width) using a caliper every other day, with two independent measurements per mouse at each time point. Mean tumor areas were calculated for each animal and group. Single-cell suspensions of spleen and inguinal (draining) lymph node (LN) cells were prepared for further analysis.

NK cell depletion in vivo For NK cell depletion, 200 ␮g of anti-NK1.1 (PK136) or control mouse ␥ globulin (Jackson ImmunoResearch Laboratories) in 200 ␮l of PBS was injected into mice i.p. at days ⫺2 and ⫹3, and ⫹8 relative to tumor inoculation (day 0), or 30 ␮l of anti-asialo GM1 (WAKO) or control rabbit serum (Sigma-Aldrich) at day ⫺1, ⫹4, and ⫹9. Depletion was confirmed in splenocytes on day of sacrifice by flow cytometry for DX5⫹ and CD3⫺ cells, and no NK cells were detected.

Enrichment of DCs from LNs Draining LNs were mechanically teased and mashed in petri dishes. The single-cell suspensions were carefully collected, resuspended in PBS containing 5% FBS plus 10 mM EDTA (PBS-EDTA-FBS), and layered onto 14.5% (wt/vol) Nycodenz (Sigma-Aldrich) solution containing RPMI 1640, 10% FBS, and 10 mM EDTA. After centrifugation at 600 ⫻ g for 20 min, cells obtained from the interphase were washed twice in PBS-EDTAFBS. Usually, 40 – 60% of isolated cells were DCs (CD11c⫹), as determined by flow cytometry.

Ag presentation assay Nycodenz-enriched LN cells were cocultured with B3Z cells (105) at a ratio of 1:5 (LN cells:B3Z) in 96-well, round-bottom plates for 24 h. Cells were

then transferred to 96-well polypropylene PCR plates and subjected to an assay for LacZ activity using a LacZ staining kit (Invitrogen Life Technologies) according to the manufacturer’s instructions. The number of LacZ⫹ (blue) cells was counted using a hemacytometer. To determine whether there were any tumor cells in the DC preparation, some of the DC-enriched cells were cultured in vitro for 5– 8 days. The few samples with tumor cell growth were excluded.

Flow cytometry RMA/t-OVA cells were stained with purified 25D1 mAb, then PE-antimouse-IgG1 (A85-1; BD Pharmingen) (27). Soluble mouse NKG2D-human IgG fusion protein and FITC-goat anti-human IgG (secondary Ab; Caltag Laboratories) were used to stain RMA and RMA/t-OVA cells for detection of NKG2D ligands. To check for NK cell depletion, spleen cells were stained with FITC-anti-CD49b (DX5; BD Pharmingen) and PE-antiCD3␧ (145-2C11; BD Pharmingen), or isotype controls. To detect Agspecific CD8⫹ T cells in splenocytes and LN cells, FcRs were blocked, and cells were stained with 1 ␮g of DimerX: soluble mouse H-2Kb:Ig fusion protein (BD Pharmingen) that had been loaded with 0.6 ␮g of OVA257–264 SIINFEKL peptide or 1 ␮g of control mouse ␥ globulin (per 106 cells/ sample), and PE-anti-ms-IgG1 (A85-1; BD Pharmingen) and FITC-antiCD8␣ (Caltag Laboratories). Cells were analyzed using a FACSCalibur (BD Biosciences) and CellQuest software (BD Biosciences).

Secondary in vitro stimulation and ELISA To determine IFN-␥ production in RMA/t-OVA tumor experiments, 2.5 ⫻ 106 spleen or 0.5 ⫻ 106 LN cells were plated in triplicates in a 48- (750 ␮l) or 96-well plates (200 ␮l), respectively, with or without 10⫺6 M OVA peptide. Stimulation of IFN-␥ in RMA tumor experiments was done by culturing 3.5 ⫻ 106 spleen cells with 3 ⫻ 105 irradiated RMA cells or complete medium alone in 24-well plates. Supernatants were collected after 5 days. IFN-␥ secretion was determined using a murine-specific IFN-␥ Duoset ELISA kit (R&D Systems).

CTL assay Spleen cells (2 ⫻ 107) were cultured in 6-well plates with 10⫺6 M OVA peptide in 5 ml of complete medium (and 10 U/ml IL-2 was added after 2 days). Stimulated spleen cells were used as effectors, at 100:1 and 50:1 E:T ratios, with RMA/t-OVA, 10⫺6M OVA peptide-pulsed P815/H-2Kb cells, or P815/H-2Kb cells alone that had been labeled with 51Cr for 1 h. After

6142

NK CELLS LIMIT IMMUNITY TO MHC CLASS I⫹ SYNGENEIC LYMPHOMAS

FIGURE 3. Increased cytotoxicity of OVA-target cells by spleen cells from NK cell-depleted mice. Spleen cells from RMA/t-OVA-bearing mice depleted of NK cells (〫) or control Ab-treated (⽧) mice were stimulated with OVA257–264 SIINFEKL peptide for 5 days and used as effector cells in a chromium-release cytotoxicity assay. Target cells were RMA/t-OVA, OVA peptide-pulsed P815/H-2Kb, or P815/H-2Kb cells. Percentage of specific lysis is shown. Data are representative of two independent experiments. Error bars represent SEM.

5 h, cell-free supernatants were harvested and radioactivity was determined. The percentage of specific lysis was calculated.

Generation of bone marrow-derived DCs Single-cell suspensions of bone marrow obtained from C57BL/6 mice were depleted of RBC by ACK lysis buffer (0.16 M NH4Cl, 0.01 M KHCO3, and 0.1 mM EDTA). To generate DCs, the precursors were cultured in RPMI 1640 plus 10% FCS and 10 ng/ml recombinant mouse GM-CSF (PeproTech) in 100-mm petri dishes at a density of 2 ⫻ 106 cells/ml. One volume of fresh GM-CSF-containing medium was added every 3 days for a total of 6 –7 days. DCs were purified by CD11c⫹ magnetic beads (Miltenyi Biotec) according to the manufacturer’s instructions.

Purification of NK and T cells Purified NK cells were obtained from B6.Rag1⫺/⫺ spleen cells using magnetic bead selection (Miltenyi Biotec) and FITC-labeled anti-DX5 Abs according to the manufacturer’s instructions. CD8⫹ OT1 T cells were purified using a similar protocol with FITC-labeled anti-CD8␤ Abs. Cell purities were ⬎90% for NK cells (DX5⫹) and 95% for OT1 T cells (CD8␤⫹).

In vitro cross-presentation assay Purified CFSE-labeled OT1 T cells (5 ⫻ 104) were cultured with 2 ⫻ 104 bone marrow-derived DCs and 2 ⫻ 104 irradiated P815/t-OVA (100 Gys) in a total of 200 ␮l of complete medium. As a positive control, DCs were pulsed with OVA257–264 peptide (10⫺10 M) for 45 min at 37°C in complete medium. The DCs were washed three times before culture with OT1 T cells. In some wells, 2 ⫻ 103 or 104 NK cells were added to cell cultures. To determine the effects of NK cells on T cell proliferation in the absence of DCs, OT1 cells were stimulated with anti-CD3 (2 ␮g/ml) or OVA257–264 peptide (10⫺8 M) in the presence or absence of NK cells. Proliferation of T cells (loss of CFSE labeling in T cells) was determined by flow cytometry after 65– 68 h of culture.

FIGURE 4. NK cell depletion results in a higher percentage and number of Ag-specific T cells. Spleen or draining LN cells from NK cell-depleted (f) or control-treated (o) mice were analyzed for the presence of OVA257–264/H-2Kb-specific CD8⫹ T cells. Data are from three (spleen) or four (LN) independent experiments. Error bars represent SEM. Examples of flow cytometry histograms, gated on CD8⫹ cells, are from spleen.

Statistics Differences between groups were analyzed using the Student t test. A value of p ⬍ 0.05 was considered significant.

Results NK cell-depleted mice have more tumor-specific CTLs and increased tumor-specific lysis To determine the role of NK cells in the development of tumorspecific CTLs in a syngeneic tumor model, C57BL/6 mice were depleted of NK cells and inoculated s.c. with 104 RMA/t-OVA cells. RMA/t-OVA cells express H-2Kb/OVA peptide complexes


6143

FIGURE 6. Ag presentation is enhanced upon depletion of NK cells. Greater activation of OVA-specific B3Z cells was observed upon coculture with DCs from draining LNs of mice inoculated s.c. with RMA/t-OVA cells and given NK cell-depleting Abs (anti-NK1.1, f; anti-asialo GM1, 䡺) compared with those from mice given control Abs (o). DCs from CD1⫺/⫺ mice are also shown (s). Results are expressed as percentage of LacZ⫹ B3Z cells per 2 ⫻ 104 CD11c⫹ cells (B6 mice ⫽ 100) (A), or total APC activity per LN (B6 mice ⫽ 100) (B).

FIGURE 5. CD1⫺/⫺ and B6 mice respond similarly to RMA/t-OVA tumor cells. CD1⫺/⫺ and B6 mice were inoculated with 104 RMA/t-OVA cells s.c. Fourteen days later, spleen and draining LN cells were stimulated with OVA257–264 SIINFEKL peptide for 5 days, and specific cytotoxicity (A) and IFN-␥ production (B) were determined. IFN-␥ production is shown as percentage of medium control (⫽100). The percentages of OVA257–264/ H-2Kb-specific CD8⫹ T cells were determined in fresh spleen and tumordraining LNs (C).

on their surfaces (Fig. 1A) and were made by retrovirally transducing RMA cells with a vector containing a truncated OVA gene, so the OVA protein cannot be secreted and remains an intracellular tumor Ag (24, 25). Like the RMA parental cell line, RMA/t-OVA cells do not express NKG2D ligands (Fig. 1B). Fourteen days after RMA/t-OVA tumor inoculation, spleen and draining (inguinal) LN cells were restimulated with OVA257–264 SIINFEKL peptide for 5 days. As shown in Fig. 2, cells were stimulated by OVA peptide in both groups, as illustrated by the increase in IFN-␥ over medium alone. However, there was a greater amount of IFN-␥ produced by both spleen and LN cells from NK cell-depleted mice compared with those given control Abs. This result suggests that NK cells may have been limiting the generation of tumor-specific CD8⫹ T cells. Tumors appeared slightly larger by day 11 in mice without NK cells compared with intact mice, but this difference disappeared after day 15 and there was no overall difference in tumor growth (data not shown).

To explore the effect of NK cell depletion on tumor-specific CTL activity, lysis of tumor target cells by OVA peptide-stimulated spleen cells was measured. Killing of RMA/t-OVA and OVA peptide-pulsed P815-H-2Kb cells was higher using effector cells from NK cell-depleted animals (Fig. 3), indicative of increased functional activity of CTLs specific for both RMA and OVA tumor Ags. There was no difference in the lysis of P815-H-2Kb cells in the absence of specific OVA peptides. To confirm that these results concerning Ag-specific CTLs reflected Ag-specific CD8⫹ T cells in vivo, we analyzed spleen and LN for the presence of OVA257–264/H-2Kb-specific CD8⫹ T cells by flow cytometry. Compared with cells from control mice, there were significantly greater percentages and absolute numbers of OVA257–264/H-2Kb-specific CD8⫹ T cells in spleen and draining LN in animals that had been depleted of NK cells (Fig. 4). Thus, increases in Ag-specific CTLs were found in vivo upon NK cell depletion in this syngeneic tumor model. Tumor-specific CTLs and IFN-␥ production are similar in CD1⫺/⫺ mice and B6 mice To determine whether NKT cells might have a role in the inhibition of Ag-specific T cells in this tumor model, we inoculated CD1⫺/⫺ mice with RMA/t-OVA tumor cells. As shown in Fig. 5, CD1⫺/⫺ mice had similar cytotoxicity against RMA/t-OVA and OVA peptide-pulsed P815-H-2Kb cells as B6 mice. There were also similar amounts of IFN-␥ produced by spleen and LN cells from B6 and CD1⫺/⫺ mice. Moreover, the percentages of OVA257–264/H-2Kb-specific CD8⫹ T cells in the spleen and draining LN were slightly lower in CD1⫺/⫺ mice compared with B6

6144


FIGURE 7. NK cells inhibit DCmediated cross-presentation of Ag in vitro. A, CFSE-labeled OVA-specific OT1 T cells were cocultured with DCs and P815/t-OVA cells. After 2.5 days, proliferation of T cells was determined by flow cytometry as a reduction in CFSE labeling of T cells. Coculture of OT1 T cells with OVA peptide-pulsed DCs was used as a positive control. Numbers represent percentage of proliferating T cells. A representative result of three independent experiments is shown as mean ⫾ SD. B, Addition of NK cells to T cell-DC-P815/t-OVA cocultures led to significantly lower T cell proliferation (p ⬍ 0.05). Data are pooled from two independent experiments and shown as mean ⫾ SD. Proliferation in the absence of NK cells is set to 100%. C, NK cells do not inhibit T cell proliferation in a DC-free system. CFSE-labeled OT-1 CD8⫹ T cells were stimulated with anti-CD3␧ (f) or OVA peptide (⽧) in the presence or absence of NK cells. Proliferation of T cells was assessed by flow cytometry. Results are expressed as mean ⫾ SD. A representative result of three independent experiments is shown.

mice. These data indicate that the lack of CD1-restricted NKT cells did not enhance antitumor immunity or the generation of tumorspecific CD8⫹ T cells. Ag presentation is enhanced in the absence of NK cells To determine whether NK cells were affecting T cells or the APC function necessary to generate OVA-specific T cells, we developed a system that enabled us to measure Ag presentation of tumor Ags by DCs in tumor draining LNs. Animals were given anti-NK1.1, anti-asialo GM1, or control Abs i.p. and were inoculated with 106 RMA/t-OVA tumor cells s.c. DCs from draining LNs were isolated and cultured with B3Z cells. B3Z cells, a T cell hybridoma reporter cell line that is specific for an epitope of OVA257–264 presented by MHC class I H-2Kb, express LacZ (␤-galactosidase) upon activation (26). The number of LacZ⫹ cells correlated with the number of CD11c⫹ DCs from the draining LN, and there was no activation of B3Z cells by CD11c⫹ DCs taken from the contralateral LN (data not shown). CD11c⫹-enriched LN cells from both anti-NK1.1-treated or anti-asialo GM1-treated mice led to an increased number of LacZ⫹ B3Z cells compared with control mice (Fig. 6A). Moreover, the total “APC activity” in the draining LN was higher in NK cell-depleted mice as well (Fig. 6B). Because anti-asialo GM1 sera and anti-NK1.1 mAb treatments both increased APC function while CD1⫺/⫺ mice were no different from B6 mice, the increased activity observed was due to the absence of NK cells and not to a potential lack of NKT cells. These results were observed for Ag presentation function based on each DC and for the total APC activity in the draining LNs (Fig. 6, A and B). We did not observe differences with respect to MHC class II or CD86 expression on the DCs in the draining LNs (data not shown). The enhanced Ag presentation by DCs in the draining LNs of NK cell-

depleted mice suggests that NK cells may prevent the development of tumor-specific CTLs by limiting DC Ag presentation. NK cells inhibit syngeneic DC-mediated activation of Ag-specific T cells To determine whether NK cells were inhibiting T cells directly or altering DC-mediated cross-presentation of tumor Ags, we developed an in vitro system to study tumor Ag-derived cross-presentation. NK cell-resistant P815 (H-2d) tumor cells that expressed truncated OVA (P815/t-OVA) were combined with B6 bone marrow-derived DCs and OT1 T cells. OT1 T cells are CD8⫹ T cells that recognize OVA peptide associated with H-2Kb (28). As shown in Fig. 7A, OT1 T cells proliferated in the presence of P815/t-OVA and DCs. However, addition of syngeneic, freshly isolated NK cells resulted in a reduced proliferation of OT1 cells (Fig. 7B). NK cells were not inhibiting T cell proliferation to Ag itself because culture of NK cells and OT1 cells together in the presence of OVA peptide resulted in a vigorous T cell response (Fig. 7C). Taken together, these data indicate that NK cells can inhibit the crosspresentation of tumor Ags to T cells by syngeneic DCs. Mice inoculated with RMA tumor cells have increased IFN-␥ production upon NK cell depletion To determine whether these results would occur in the absence of a strong model Ag (OVA), C57BL/6 mice were inoculated s.c. with 104 RMA tumor cells and were depleted of NK cells by anti-NK1.1 or anti-asialo GM1 Abs. Upon sacrifice, spleen cells were stimulated with irradiated RMA cells or medium only for 5 days. In the absence of a strong tumor Ag, restimulation of the immune cells is required to generate sufficient responses. Spleen cells from both groups displayed higher IFN-␥ production when


FIGURE 8. Spleen cells from NK cell-depleted mice inoculated with RMA cells produce increased levels of IFN-␥. Mice received anti-NK1.1, anti-asialo GM1, or control Abs and 104 RMA cells s.c. A, IFN-␥ production of spleen cells stimulated with irradiated RMA tumor cells (f) or medium alone (䡺). B, Tumor growth in anti-NK1.1-treated mice (open symbols) and control murine IgG-treated mice (filled symbols). Error bars represent SEM. Four independent experiments were done. The average of all experiments (A) and a representative experiment (B) are shown.

cultured with RMA cells, compared with medium alone, indicating stimulation by the tumor cells. Similar to the RMA/t-OVA experiments, IFN-␥ was produced in significantly larger amounts by RMA-stimulated spleen cells from NK cell-depleted animals compared with control-treated animals (Fig. 8A). Therefore, NK cells can exert their inhibitory effects in response to both RMA and RMA/t-OVA. We observed no tumor size difference between mice given NK cell-depleting Abs and those given control Abs (Fig. 8B), indicating that NK cells did not alter tumor growth during the initial period of RMA tumor growth, which supports other findings (20 –22).

Discussion The aim of this study was to determine whether NK cells alter the generation of adaptive immunity against tumors that do not themselves induce significant NK cell activity. We have used the RMA tumor cell that expresses MHC class I and does not express NK cell-activating ligands for NKG2D. Upon NK cell depletion, enhanced immune responses against tumor Ags were observed in this murine lymphoma model. Ag-specific CD8⫹ T cells from NK celldeficient mice were more abundant and functionally active, compared with those from intact mice. There was also increased Ag presentation by DCs in mice depleted of NK cells. We used RMA tumor cells expressing a truncated version of OVA as a model tumor Ag that would not be secreted but could be recognized on tumor cells and presented by DCs due to crosspresentation. We determined that NK cells inhibited Ag-specific

6145 T cell activation by affecting DC-dependent APC function in draining LNs and in an in vitro cross-presentation assay. Similar findings were obtained in mice inoculated with RMA/tOVA or RMA tumor cells, which indicates that these findings were independent of the presence of OVA. There may be several means by which NK cells alter the function of DCs in vivo. It is known that DCs lose their ability to efficiently cross-present Ags as they mature, and there are data suggesting that maturation of human DCs can be induced by coculture with NK cells in vitro at low NK:DC ratios (1:5) (7). Mature DCs are believed to be inefficient in cross-presentation due to a loss of the ability to phagocytose exogenous Ags (29, 30). Therefore, a possible mechanism for NK inhibitory effects on DC-mediated cross-presentation may be that NK cells promote DC maturation, leading to lower levels of cross-presentation. However, we did not observe differences in the expression of MHC class II or CD86 on DCs taken from draining LNs between NK cell-depleted and NK cell-intact mice. It may be possible that cytokines from NK cells, such as IFN-␥, may make it less likely for DCs to move to regional LNs, as has been suggested previously (31). Activated NK cells have been shown to kill DCs in vitro (15, 16). Studies where cytotoxicity of DCs by NK cells was observed involved activated NK cells and may not occur with fresh NK cells. Killing of immature DCs has been observed in vitro at higher NK:DC ratios (5:1) but not at low ratios (1:5) (7). In our in vitro assay, we observed NK-mediated inhibitory effects even with 2 ⫻ 103 cells (NK:DC ratio ⫽ 1:10). It is less likely that fresh NK cells will kill DCs at a low NK:DC ratio used in our experiments. We have no data to support the fact that resting NK cells kill immature DCs in vitro during the cross-presentation assay used in this study. Many studies have implicated a stimulatory role for NK cells in antitumor immunity. NK cells have been shown to promote DC maturation and the generation of Th1 and CTL responses. In several studies, investigators used tumor cells that lacked expression of MHC class I or that expressed ligands for NK cell-activating receptors, such as NKG2D. Studies showed that antitumor immunity against MHC class I-negative tumors was dependent on DCactivated NK cells in vivo and that TNF-␣-stimulated mature DCs activated NK cells in vitro in a contact-dependent manner (32). Development of tumor-specific CTLs and Th1 responses to CD70⫹ or CD80⫹ MHC class I-deficient RMA-S tumor cells were dependent on NK cells, perforin, and IFN-␥ (1, 2). Protective CD8⫹ T cell memory responses induced by NK cell-activated DCs were observed in studies in which the NK cells were activated by tumor cells that lacked MHC class I (3). Mice treated with RMA tumor cells that expressed an NKG2D ligand, Rae1, were able to reject the tumor cells and developed the capacity to resist a challenge with RMA tumor cells alone (21). Other studies also showed that expression of NKG2D ligands resulted in NK cell-dependent resistance to tumor cell challenge (33). A s.c. NKG2D ligandpositive lymphoma model in AKR mice demonstrated a stimulatory role for NK cells in CTL generation (34). Depletion of NK cells in this AKR tumor model led to alternatively activated macrophage formation and subsequent suppression of CTL generation. Taken together, these studies show that when activated against tumor cells, NK cells can help to generate an effective antitumor immune response, including tumor-specific CTLs. Many tumor cells express MHC class I molecules and engage inhibitory receptors on NK cells. Some tumors do not express high amounts of ligands for NK cell-activating receptors. Spontaneous tumor cells are not particularly immunogenic, although they may express tumor-specific or tumor-associated Ags. We have modeled these “less-immunogenic” tumor cells using the RMA lymphoma. RMA is an aggressive tumor and grows progressively in B6 mice

6146


whether the mice have NK cells or not. In contrast to some other reported tumor models, we have found that removal of NK cells resulted in an increase in tumor-specific T cells and DC-mediated presentation of tumor Ags in the draining LNs. Our data suggest that when NK cells are not activated by a tumor (due to the presence of activating ligands and/or the absence of MHC class I), NK cells may limit immune responsiveness and, hence, promote tumor growth and survival. Thus, when NK cells are absent, Ag presentation by DCs and tumor-specific CTL generation are enhanced. Some reports indicate that NK cells may limit or prevent immunity under certain conditions. Recent data have shown that NK cells can have a negative role in allograft rejection in a perforindependent manner (12, 13). Perforin deficiency has been shown to increase transplanted DC survival in vivo, although TRAIL deficiency leads to a much larger increase in DC survival (14). The latter study also found that NK cell depletion increased T cell Ag-specific tumor immunity; however, this report involved immunization with tumor Ag-loaded immature DCs and may not reflect responses to tumor cells themselves. Activated NK cells have also been shown to kill syngeneic, APC-activated CD4⫹ and CD8⫹ T cells under certain in vitro conditions (19). There is evidence from multiple sclerosis patients treated with daclizumab that NK cells can kill activated autologous T cells, and this result supports a potential immunoregulatory role for NK cells (35). In one study, NK cells down-regulated CD4⫹ T cell responses during virus infection (36). However, our data using an in vitro cross-presentation system did not indicate that NK cells killed syngeneic T cells. Another possible reason is that NK cells compete for survival cytokines, such as IL-15, with CD8⫹ T cells. If this mechanism is involved, then elimination of NK cells would allow more such cytokines for CD8⫹ T cells and allow them to expand to greater numbers. In models of autoimmunity, NK cells may have a stimulatory or a regulatory role that limits disease depending on the particular disease model (37, 38). The conditions under which this negative regulation by NK cells occurs will be important to understand, with implications for tumor immunity, allograft rejection, autoimmunity, and tolerance. These findings using a less-immunogenic tumor may reflect the role that NK cells have in regulating some autoimmune responses and spontaneous tumors. In the absence of microorganism-derived signals or direct activating ligands for NK cell receptors, NK cells may prevent presentation of Ags by DCs under conditions of limited inflammation, such as injection of tumor cells. The NK cell activity may slow, but not eliminate, the development of CTLs. This process may reflect a function of NK cells that down-regulates the immune response as an infection becomes controlled and the need to modulate immune activity takes over. In the case of cancer, tumors may use similar mechanisms to limit tumor-specific CTL development and prevent antitumor immunity from developing. Tumors use many different mechanisms to avoid an immune response, and the data reported here suggest that NK cells may aid tumor survival by limiting the presentation of tumor Ags and subsequent activation of host CD8⫹ T cells.

Acknowledgment We thank Dr. Brent Berwin (Dartmouth Medical School, Lebanon, NH) for providing the 25D1 Abs and the Animal Resource Center for assistance with animal care.

Disclosures The authors have no financial conflict of interest.

References 1. 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. 2. Kelly, J. M., K. Takeda, P. K. Darcy, H. Yagita, and M. J. Smyth. 2002. A role for IFN-␥ in primary and secondary immunity generated by NK cell-sensitive tumor-expressing CD80 in vivo. J. Immunol. 168: 4472– 4479. 3. Mocikat, R., H. Braumuller, A. Gumy, O. Egeter, H. Ziegler, U. Reusch, A. Bubeck, J. Louis, R. Mailhammer, G. Riethmuller, et al. 2003. Natural killer cells activated by MHC class Ilow targets prime dendritic cells to induce protective CD8 T cell responses. Immunity 19: 561–569. 4. Strbo, N., S. Oizumi, V. Sotosek-Tokmadzic, and E. R. Podack. 2003. Perforin is required for innate and adaptive immunity induced by heat shock protein gp96. Immunity 18: 381–390. 5. Gerosa, F., B. Baldani-Guerra, C. Nisii, V. Marchesini, G. Carra, and G. Trinchieri. 2002. Reciprocal activating interaction between natural killer cells and dendritic cells. J. Exp. Med. 195: 327–333. 6. Gerosa, F., A. Gobbi, P. Zorzi, S. Burg, F. Briere, G. Carra, and G. Trinchieri. 2005. The reciprocal interaction of NK cells with plasmacytoid or myeloid dendritic cells profoundly affects innate resistance functions. J. Immunol. 174: 727–734. 7. Piccioli, D., S. Sbrana, E. Melandri, and N. M. Valiante. 2002. Contact-dependent stimulation and inhibition of dendritic cells by natural killer cells. J. Exp. Med. 195: 335–341. 8. Mailliard, R. B., Y. I. Son, R. Redlinger, P. T. Coates, A. Giermasz, P. A. Morel, W. J. Storkus, and P. Kalinski. 2003. Dendritic cells mediate NK cell help for Th1 and CTL responses: two-signal requirement for the induction of NK cell helper function. J. Immunol. 171: 2366 –2373. 9. Mailliard, R. B., S. M. Alber, H. Shen, S. C. Watkins, J. M. Kirkwood, R. B. Herberman, and P. Kalinski. 2005. IL-18-induced CD83⫹CCR7⫹ NK helper cells. J. Exp. Med. 202: 941–953. 10. Vitale, M., M. Della Chiesa, S. Carlomagno, D. Pende, M. Arico, L. Moretta, and A. Moretta. 2005. NK-dependent DC maturation is mediated by TNF␣ and IFN␥ released upon engagement of the NKp30 triggering receptor. Blood 106: 566 –571. 11. Chang, C. C., M. Campoli, and S. Ferrone. 2005. Classical and nonclassical HLA class I antigen and NK cell-activating ligand changes in malignant cells: current challenges and future directions. Adv. Cancer Res. 93: 189 –234. 12. Raziuddin, A., D. L. Longo, M. Bennett, R. Winkler-Pickett, J. R. Ortaldo, and W. J. Murphy. 2002. Increased bone marrow allograft rejection by depletion of NK cells expressing inhibitory Ly49 NK receptors for donor class I antigens. Blood 100: 3026 –3033. 13. Beilke, J. N., N. R. Kuhl, L. Van Kaer, and R. G. Gill. 2005. NK cells promote islet allograft tolerance via a perforin-dependent mechanism. Nat. Med. 11: 1059 –1065. 14. Hayakawa, Y., V. Screpanti, H. Yagita, A. Grandien, H. G. Ljunggren, M. J. Smyth, and B. J. Chambers. 2004. NK cell TRAIL eliminates immature dendritic cells in vivo and limits dendritic cell vaccination efficacy. J. Immunol. 172: 123–129. 15. Chambers, B. J., M. Salcedo, and H. G. Ljunggren. 1996. Triggering of natural killer cells by the costimulatory molecule CD80 (B7-1). Immunity 5: 311–317. 16. Geldhof, A. B., M. Moser, L. Lespagnard, K. Thielemans, and P. De Baetselier. 1998. Interleukin-12-activated natural killer cells recognize B7 costimulatory molecules on tumor cells and autologous dendritic cells. Blood 91: 196 –206. 17. Ferlazzo, G., M. L. Tsang, L. Moretta, G. Melioli, R. M. Steinman, and C. Munz. 2002. Human dendritic cells activate resting natural killer (NK) cells and are recognized via the NKp30 receptor by activated NK cells. J. Exp. Med. 195: 343–351. 18. Ferlazzo, G., B. Morandi, A. D’Agostino, R. Meazza, G. Melioli, A. Moretta, and L. Moretta. 2003. The interaction between NK cells and dendritic cells in bacterial infections results in rapid induction of NK cell activation and in the lysis of uninfected dendritic cells. Eur. J. Immunol. 33: 306 –313. 19. Rabinovich, B. A., J. Li, J. Shannon, R. Hurren, J. Chalupny, D. Cosman, and R. G. Miller. 2003. Activated, but not resting, T cells can be recognized and killed by syngeneic NK cells. J. Immunol. 170: 3572–3576. 20. Hoglund, P., H. G. Ljunggren, C. Ohlen, L. Ahrlund-Richter, G. Scangos, C. Bieberich, G. Jay, G. Klein, and K. Karre. 1988. Natural resistance against lymphoma grafts conveyed by H-2Dd transgene to C57BL mice. J. Exp. Med. 168: 1469 –1474. 21. 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. 22. Tahara-Hanaoka, S., K. Shibuya, H. Kai, A. Miyamoto, Y. Morikawa, N. Ohkochi, S. Honda, and A. Shibuya. 2006. Tumor rejection by the poliovirus receptor family ligands of the DNAM-1 (CD226) receptor. Blood 107: 1491–1496. 23. Sonoda, K. H., M. Exley, S. Snapper, S. P. Balk, and J. Stein-Streilein. 1999. CD1-reactive natural killer T cells are required for development of systemic tolerance through an immune-privileged site. J. Exp. Med. 190: 1215–1226. 24. Zhao, Y., D. Boczkowski, S. K. Nair, and E. Gilboa. 2003. Inhibition of invariant chain expression in dendritic cells presenting endogenous antigens stimulates CD4⫹ T-cell responses and tumor immunity. Blood 102: 4137– 4142. 25. De Veerman, M., C. Heirman, S. Van Meirvenne, S. Devos, J. Corthals, M. Moser, and K. Thielemans. 1999. Retrovirally transduced bone marrow-derived dendritic cells require CD4⫹ T cell help to elicit protective and therapeutic antitumor immunity. J. Immunol. 162: 144 –151.

The Journal of Immunology 26. Shastri, N., and F. Gonzalez. 1993. Endogenous generation and presentation of the ovalbumin peptide/Kb complex to T cells. J. Immunol. 150: 2724 –2736. 27. Porgador, A., J. W. Yewdell, Y. Deng, J. R. Bennink, and R. N. Germain. 1997. Localization, quantitation, and in situ detection of specific peptide-MHC class I complexes using a monoclonal antibody. Immunity 6: 715–726. 28. Hogquist, K. A., S. C. Jameson, W. R. Heath, J. L. Howard, M. J. Bevan, and F. R. Carbone. 1994. T cell receptor antagonist peptides induce positive selection. Cell 76: 17–27. 29. Wilson, N. S., G. M. Behrens, R. J. Lundie, C. M. Smith, J. Waithman, L. Young, S. P. Forehan, A. Mount, R. J. Steptoe, K. D. Shortman, et al. 2006. Systemic activation of dendritic cells by Toll-like receptor ligands or malaria infection impairs cross-presentation and antiviral immunity. Nat. Immunol. 7: 165–172. 30. Hickman-Miller, H. D., and J. W. Yewdell. 2006. Youth has its privileges: maturation inhibits DC cross-priming. Nat. Immunol. 7: 125–126. 31. Wu, X., W. Hou, S. Sun, E. Bi, Y. Wang, M. Shi, J. Zang, C. Dong, and B. Sun. 2006. Novel function of IFN-␥: negative regulation of dendritic cell migration and T cell priming. J. Immunol. 177: 934 –943. 32. Fernandez, N. C., A. Lozier, C. Flament, P. Ricciardi-Castagnoli, D. Bellet, M. Suter, M. Perricaudet, T. Tursz, E. Maraskovsky, and L. Zitvogel. 1999. Dendritic cells directly trigger NK cell functions: cross-talk relevant in innate anti-tumor immune responses in vivo. Nat. Med. 5: 405– 411.

6147 33. Cerwenka, A., J. L. Baron, and L. L. Lanier. 2001. Ectopic expression of retinoic acid early inducible-1 gene (RAE-1) permits natural killer cellmediated rejection of a MHC class I-bearing tumor in vivo. Proc. Natl. Acad. Sci. USA 98: 11521–11526. 34. Geldhof, A. B., J. A. Van Ginderachter, Y. Liu, W. Noel, G. Raes, and P. De Baetselier. 2002. Antagonistic effect of NK cells on alternatively activated monocytes: a contribution of NK cells to CTL generation. Blood 100: 4049 – 4058. 35. Bielekova, B., M. Catalfamo, S. Reichert-Scrivner, A. Packer, M. Cerna, T. A. Waldmann, H. McFarland, P. A. Henkart, and R. Martin. 2006. Regulatory CD56bright natural killer cells mediate immunomodulatory effects of IL-2R␣targeted therapy (daclizumab) in multiple sclerosis. Proc. Natl. Acad. Sci. USA 103: 5941–5946. 36. Su, H. C., K. B. Nguyen, T. P. Salazar-Mather, M. C. Ruzek, M. Y. Dalod, and C. A. Biron. 2001. NK cell functions restrain T cell responses during viral infections. Eur. J. Immunol. 31: 3048 –3055. 37. French, A. R., and W. M. Yokoyama. 2004. Natural killer cells and autoimmunity. Arthritis Res. Ther. 6: 8 –14. 38. Shi, F. D., and L. Van Kaer. 2006. Reciprocal regulation between natural killer cells and autoreactive T cells. Nat. Rev. Immunol. 6: 751–760.