Blockade of B7-H1 on Macrophages Suppresses CD4⁺ T Cell Proliferation by Augmenting IFN-γ-Induced Nitric Oxide Production¹

T cell activation is regulated by a balance between positive and negative signals. This balance is mainly kept by the interaction of costimulatory or coinhibitory receptors on T cells with their ligands on APCs. Programmed death-1 (PD-1)³ is a type I transmembrane protein that was originally identified in a T cell line undergoing programmed cell death (1), but additional studies have shown that its expression is associated with lymphocyte activation rather than cell death (2, 3). Structurally, PD-1 belongs to the CD28/CTLA-4 subfamily of the Ig superfamily and contains a single Ig V-like domain in its extracellular region (4, 5). The PD-1 cytoplasmic region contains two tyrosine residues, with the N-terminal tyrosine being located in an ITIM and the C-terminal tyrosine being located in an immunoreceptor tyrosine-based switch motif (6). In vivo studies using PD-1-deficient mice have shown that PD-1 plays an important role in the regulation of peripheral tolerance and prevention of autoimmunity. PD-1-deficient C57BL/6 mice developed a lupus-like disease and arthritis (7). In addition, PD-1-deficient BALB/c mice developed autoantibody-mediated dilated cardiomyopathy (8). Moreover, PD-1-deficient mice crossed with H-2L^d-specific TCR transgenic mice on an H-2^b/d background exhibited graft-vs-host disease-like pathologies (7). All these results have suggested that PD-1 acts as an important negative regulator of autoimmune responses.

Two new members of the B7 family, B7-H1 (PD-L1) and B7-DC (PD-L2), have been identified to be the ligands for PD-1 (9, 10, 11, 12). In vitro studies have shown that the engagement of PD-1 by B7-H1 or B7-DC inhibited TCR-mediated T cell proliferation and cytokine (IFN-γ, IL-10, IL-4, and IL-2) production (10, 11). These results indicated that the engagement of PD-1 by B7-H1 or B7-DC led to down-regulation of T cell responses. However, not all studies have supported the inhibitory role for B7-H1 and B7-DC. A previous report has shown that naive T cells stimulated with immobilized anti-CD3 and B7-DC-Ig exhibited enhanced proliferation and IFN-γ production (12). Moreover, when T cells were stimulated with low doses of anti-CD3 and immobilized B7-H1-Ig, proliferation and production of IFN-γ, GM-CSF, and IL-10 were enhanced (9, 13). These results indicated that B7-H1 and B7-DC could costimulate T cell proliferation and cytokine production. The reason for these contradictory results remains unknown. One possible explanation may be the presence of a second receptor for B7-H1 and B7-DC, which delivers a costimulatory signal.

B7-H1 was expressed on immunocompetent cells such as T cells, B cells, dendritic cells (DCs), and macrophages (14). In contrast, B7-DC expression was highly restricted to DCs and activated macrophages (14). PD-1 was expressed on activated T and B cells and a subset of thymocytes (2, 14). These results have suggested that B7-H1 and B7-DC probably contribute to T cell-APC interactions. To address this possibility, we examined the contributions of B7-H1 and B7-DC to CD4⁺ T cell activation by B cells, DCs, and macrophages using blocking mAbs against B7-H1, B7-DC, and PD-1. We found that the blockade of B7-H1 and PD-1 markedly inhibited the proliferation, but enhanced the IL-2 and IFN-γ production, of naive CD4⁺ T cells when costimulated by macrophages. This inhibition of CD4⁺ T cell proliferation was due to accumulation of NO produced by macrophages in response to CD4⁺ T cell-derived IFN-γ. These results suggested that B7-H1 on macrophages might play a critical role in the control of NO production by macrophages through the regulation of IFN-γ production by CD4⁺ T cells. The pathophysiological relevance of this finding is discussed.

Male BALB/c and C57BL/6 mice were purchased from Charles River Breeding Laboratories Japan. IFN-γ-deficient C57BL/6 mice were provided by Dr. Y. Iwakura (University of Tokyo, Tokyo, Japan). IFN-γR-deficient C57BL/6 mice were provided by Dr. T. Taniguchi (University of Tokyo). All mice were maintained under specific pathogen-free conditions and used at 6–7 wk of age according to institutional guidelines. A baby hamster kidney cell line, BHK, was purchased from American Type Culture Collection. The cells were cultured in RPMI 1640 medium containing 10% FCS, 10 mM HEPES, 2 mM l-glutamine, 0.1 mg/ml penicillin and streptomycin, and 50 μM 2-ME.

Purified anti-mouse CD3 mAb (145-2C11), anti-mouse IFN-γ mAb (R4-6A2), rat IgG2a isotype control, and PE-labeled streptavidin were purchased from BD Pharmingen. Goat anti-mouse IgM F(ab′)₂ Ab was purchased from Jackson ImmunoResearch Laboratories. Anti-mouse B7-H1 mAb (MIH6), anti-mouse B7-DC mAb (TY25), and anti-mouse CD40 mAb (HM40-3) were generated in our laboratory as previously described (14, 15). Fabs of anti-B7-H1 mAb were prepared using immobilized papain and a protein A column (Pierce). The purity of Fab was verified by SDS-PAGE analysis. A new anti-mouse PD-1 mAb (RMP1-14) was also generated in our laboratory. A Sprague Dawley rat was immunized with mouse PD-1-transfected BHK cells (2). Three days after the final immunization, spleen cells were fused with P3U1 myeloma cells as previously described (14). After hypoxanthine-aminopterin-thymidine selection, one hybridoma producing RMP1-14 mAb (rat IgG2a, κ) was identified by its strong reactivity with PD-1-transfected BHK cells, but not with parental BHK cells, by flow cytometry (Fig. 1,A). RMP1-14 blocked the binding of both B7-H1-Ig and B7-DC-Ig fusion proteins (14) to PD-1-transfected BHK cells (Fig. 1B). The inducible NO synthase (iNOS) inhibitors, N^G-monomethyl-l-arginine acetate (l-NMMA) and N⁵-(1-iminoethyl)-l-omithine dihydrochloride (l-NIO), were purchased from Wako Pure Chemicals. LPS and the indoleamine 2,3-dioxygenase (IDO) inhibitor, 1-methyl-d,l-tryptophan (1-MT), were purchased from Sigma-Aldrich. Recombinant mouse IFN-γ was purchased from BD Biosciences.

CD62L⁺ CD4⁺ naive T cells were purified from splenocytes by passage through nylon wool columns (Wako Biochemicals) and by using autoMACS columns with a CD4⁺ T cell isolation kit and anti-CD62L-coupled microbeads (Miltenyi Biotec) according to the manufacturer’s instructions (>95% CD4⁺CD62L⁺). Small resting B cells were also purified from splenocytes as previously described (16). Briefly, spleen cells were treated with a mixture of hybridoma supernatants (anti-Thy-1.2, anti-CD4, and anti-CD8) and Low-Tox rabbit complement (Cedarlane Laboratories). After Percoll (Amersham Biosciences) gradient centrifugation, small B cells were collected from the 60/70% interface. Purified B cells (>95% B220⁺; 3 × 10⁶/ml) were cultured with anti-IgM Ab (5 μg/ml) and anti-CD40 mAb (5 μg/ml) for 24 h, then used as APCs. For isolating splenic DCs, spleens were digested with 400 U/ml collagenase (Wako Biochemicals) in the presence of 5 mM EDTA and separated into low and high density fractions on Optiprep gradient (Axis-Shield). Low-density cells were incubated with anti-CD11c-coupled microbeads (Miltenyi Biotec), and the bound cells were isolated using autoMACS columns (>95% CD11c⁺). Peritoneal macrophages were obtained from mice that had received 2 ml of 4% thioglycolate (Sigma-Aldrich) i.p. 4 days previously. Peritoneal exudate cells were harvested by peritoneal lavage with ice-cold PBS and were depleted of nonadherent cells after 1-h culture on a plastic dish (>95% CD11b⁺).

Purified naive CD4⁺ T cells (1 × 10⁵/well) were cocultured with irradiated B cells (1 × 10⁵/well), DCs (1 × 10⁴/well), or macrophages (1 × 10⁵/well) in the presence or the absence of anti-CD3 mAb (2 μg/ml) in 96-well, flat-bottom plates for 24–72 h. Each blocking mAb (5 μg/ml) or control rat IgG (5 μg/ml) and iNOS inhibitors (100 μM) or IDO inhibitor (1 mM) were added at the start of the cultures. For estimating proliferative responses, the cultures were pulsed with 0.5 μCi/well [³H]thymidine (DuPont-NEN) for the last 6 h and then harvested using a Micro 96 Harvester (Skatron). Incorporated radioactivity was measured in a Micro β counter (Micro β Plus). To determine the production of cytokines, cell-free supernatants were collected at 24–72 h and assayed for IL-2, IL-4, IL-5, and IL-10 by ELISA using OptEIA kits (BD Pharmingen) and for IFN-γ using the mouse IFN-γ ELISA Ready-SET-Go kit (eBioscience) according to the manufacturer’s instructions.

The nitrite concentration in the culture supernatants was measured by a colorimetric assay using the Griess reagent. Cell-free supernatants (100 μl) were added to a microtiter plate containing 100 μl of 1% sulfanilamide and 0.1% naphthylethylenediamine in 2.5% H₃PO₄. After 10 min, the OD was measured at 550 nm and compared with a standard curve of NaNO₃.

Significant differences between two experimental groups were analyzed by unpaired Student’s t test. A value of p < 0.05 was considered significant.

To investigate the roles of B7-H1 and B7-DC in T-APC interactions, we first examined the contributions of B7-H1 and B7-DC to T cell activation by B cells, DCs, and macrophages. Purified naive CD4⁺ T cells from the spleens of BALB/c mice were cocultured with preactivated B cells, splenic DCs, or thioglycolate-elicited peritoneal macrophages in the presence of anti-CD3 mAb and anti-B7-H1 mAb, anti-B7-DC mAb, anti-PD-1 mAb, or control rat IgG for 24–72 h, and then the proliferative response was assessed by [³H]thymidine uptake (Fig. 2). In the absence of APCs, no significant proliferation (<1000 cpm) of naive CD4⁺ T cells was observed upon anti-CD3 stimulation (data not shown). The addition of activated B cells, DCs, or macrophages markedly costimulated the proliferation at 48 h. DCs were the most potent, in that the costimulatory effect of 1 × 10⁴ DCs was almost comparable to that of 1 × 10⁵ B cells or macrophages. The anti-B7-H1, anti-B7-DC, and anti-PD-1 blocking mAbs did not apparently affect the CD4⁺ T cell proliferation when costimulated by B cells or DCs. In contrast, when costimulated by macrophages, the proliferative response was markedly inhibited by either anti-B7-H1 or anti-PD-1 mAb alone, but not by anti-B7-DC mAb. Similar inhibitory effects of anti-B7-H1 and anti-PD-1 mAbs were also observed when CD4⁺ T cells were cocultured with resident peritoneal macrophages or splenic macrophages from naive mice (data not shown). These results suggested a critical contribution of PD-1/B7-H1 interaction to the costimulation of CD4⁺ T cell proliferation by macrophages.

To exclude a possible contribution of FcRs on macrophages to the inhibitory effect of anti-B7-H1 mAb, naive CD4⁺ T cells were cocultured with macrophages in the presence of anti-CD3 mAb and intact or Fab of anti-B7-H1 mAb. As shown in Fig. 3,A, CD4⁺ T cell proliferation was comparably inhibited by both intact and Fab of anti-B7-H1 mAb. We also examined cytokine production in these cultures. Unexpectedly, the production of IL-2 and IFN-γ was enhanced by either intact or Fab of anti-B7-H1 mAb (Fig. 3, B and C). IL-4, IL-5, and IL-10 were not detectable in this experiment (data not shown). These results indicated that the inhibitory effect of anti-B7-H1 mAb on CD4⁺ T cell proliferation was not due to a lack of IL-2 production.

The inhibitory effect of anti-B7-H1 mAb on CD4⁺ T cell proliferation might be explained by a down-regulation of IL-2R expression on CD4⁺ T cells. However, the expression levels of IL-2R α-, β-, and γ-chains on anti-CD3-stimulated CD4⁺ T cells in the presence of macrophages were not significantly affected by anti-B7-H1 mAb compared with control IgG (data not shown). Alternatively, the enhanced IFN-γ production might be responsible for the inhibition of CD4⁺ T cell proliferation. To address this possibility, we examined the effect of neutralizing anti-IFN-γ mAb on the inhibition of CD4⁺ T cell proliferation by anti-B7-H1 and anti-PD-1 mAbs. Purified naive CD4⁺ T cells were cocultured with macrophages in the presence of anti-CD3 mAb and anti-B7-H1 or anti-PD-1 mAb with or without anti-IFN-γ mAb for 48 h, and then the proliferative response was assessed. As shown in Fig. 4,A, anti-IFN-γ mAb alone significantly enhanced the proliferation of CD4⁺ T cells, indicating a suppressive effect of endogenously produced IFN-γ. More importantly, the inhibitory effects of both anti-B7-H1 mAb and anti-PD-1 mAbs were totally abrogated by anti-IFN-γ mAb. These results indicated that the inhibitory effects of anti-B7-H1 and anti-PD-1 mAbs on CD4⁺ T cell proliferation were dependent on the increased IFN-γ production. To further substantiate the inhibitory effect of the increased IFN-γ, we added a physiological concentration of rIFN-γ (2 ng/ml). As shown in Fig. 4 B, the addition of rIFN-γ inhibited the proliferation of CD4⁺ T cells to a comparable level with anti-B7-H1 mAb.

To determine the source of IFN-γ, CD4⁺ T cells and macrophages were prepared from IFN-γ-deficient or wild-type mice and cocultured with anti-CD3 mAb in the presence of anti-B7-H1 mAb or control IgG. The proliferative response was assessed on 48 h. As shown in Fig. 5, anti-B7-H1 mAb markedly inhibited the proliferation when both CD4⁺ T cells and macrophages were prepared from wild-type mice. In contrast, this inhibitory effect was mostly abrogated when IFN-γ-deficient CD4⁺ T cells were cocultured with wild-type macrophages. In contrast, anti-B7-H1 mAb still partially inhibited CD4⁺ T cell proliferation when wild-type CD4⁺ T cells were cocultured with IFN-γ-deficient macrophages. These results indicated that CD4⁺ T cells, rather than macrophages, were the major source of IFN-γ responsible for the inhibition of CD4⁺ T cell proliferation by anti-B7-H1 mAb.

We also examined the involvement of IFN-γR in the anti-B7-H1 mAb-mediated suppression. As shown in Fig. 6, when macrophages were prepared from IFN-γR^−/− mice, the proliferation of CD4⁺ T cells was not significantly inhibited by anti-B7-H1 mAb, indicating a critical contribution of IFN-γR to macrophages.

It is known that macrophages produce NO after induction of iNOS by IFN-γ, and that NO inhibits T cell proliferation (17, 18, 19). Therefore, we next examined NO production in the coculture of CD4⁺ T cells and macrophages in the presence of anti-CD3 mAb. As shown in Fig. 7,B, the addition of anti-B7-H1 mAb markedly induced NO production, which was totally abrogated by anti-IFN-γ mAb. NO production was apparently associated with the inhibition of CD4⁺ T cell proliferation (Fig. 7 A). Moreover, NO production was not observed when macrophages were derived from IFN-γR^−/− mice (data not shown). These results indicated that the B7-H1 blockade induced IFN-γ-dependent NO production by macrophages.

We next examined the effects of iNOS inhibitors, l-NMMA and l-NIO, on the inhibition of CD4⁺ T cell proliferation by anti-B7-H1 mAb. As shown in Fig. 8 A, l-NMMA alone significantly enhanced the proliferation, indicating a suppression by endogenously produced NO. More importantly, the inhibition by anti-B7-H1 mAb was totally abrogated by l-NMMA. l-NIO exhibited a similar effect (data not shown). The expression level of B7-H1 on macrophages was not affected by l-NMMA, as estimated by flow cytometry (data not shown). These results indicated that the inhibitory effect of anti-B7-H1 mAb on CD4⁺ T cell proliferation was mediated by iNOS-induced NO production by macrophages.

It has recently been reported that IDO is a tryptophan-catabolizing enzyme expressed by IFN-γ-stimulated macrophages that inhibits T cell proliferation by tryptophan depletion (20, 21, 22, 23). Thus, we also examined the effect of a specific inhibitor of IDO, 1-MT, on the inhibition of CD4⁺ T cell proliferation by anti-B7-H1 mAb. As shown in Fig. 8 A, 1-MT partially reversed the inhibition of CD4⁺ T cell proliferation by anti-B7-H1 mAb, suggesting a minor contribution of IDO. However, the combination of 1-MT with l-NMMA did not exert an additive effect, indicating the predominant contribution of iNOS, rather than IDO.

We also examined the effects of l-NMMA and/or 1-MT on cytokine production. As shown in Fig. 8,B, the increased IL-2 level caused by anti-B7-H1 mAb was not significantly affected by l-NMMA and/or 1-MT. In contrast, the increased IFN-γ level caused by anti-B7-H1 mAb was further increased by l-NMMA, suggesting the expansion of IFN-γ-producing T cells (Fig. 8 C).

In this study we found that the blockade of the PD-1/B7-H1 interaction markedly inhibited the proliferation of anti-CD3-stimulated naive CD4⁺ T cells when costimulated by macrophages in vitro. This inhibition was not due to blockade of a positive costimulation by PD-1/B7-H1 for priming T cell activation, because the production of IL-2 and IFN-γ was, instead, enhanced by the blockade of B7-H1. Additional experiments showed that the enhanced IFN-γ production by CD4⁺ T cells induced iNOS-mediated NO production by macrophages, which, in turn, inhibited the proliferation of CD4⁺ T cells. These results suggest that the PD-1/B7-H1 interaction plays a critical role in regulating IFN-γ production, NO production, and T cell expansion through the cognate interaction between CD4⁺ T cells and macrophages.

The blockade of B7-H1 enhanced IFN-γ production by CD4⁺ T cells, suggesting a critical role for B7-H1 on macrophages to suppress IFN-γ production by CD4⁺ T cells. We have previously demonstrated that B7-H1 is constitutively expressed on macrophages at a high level and is further up-regulated by T cell-derived cytokines, including IFN-γ (14). Therefore, the B7-H1-mediated suppression may be a negative feedback mechanism to prevent overproduction of IFN-γ. The inhibition of IFN-γ production by B7-H1 might be directly mediated by negative signaling via PD-1, because it has been shown that the immunoreceptor tyrosine-based switch motif of PD-1 recruits tyrosine phosphatases Src homology protein tyrosine phosphatase-1 and -2, which counteract the TCR-mediated T cell activation, in primary human T cells (24). Alternatively, the suppression of IFN-γ production by B7-H1 might be indirectly mediated by IL-10, because it has been reported that B7-H1 enhances IL-10 production by CD4⁺ T cells (9). However, we could not detect IL-10 in the coculture of CD4⁺ T cells and macrophages in the present study. Therefore, the negative signaling via PD-1 seems to be more likely to be responsible for the suppression of IFN-γ production by B7-H1. The critical role of B7-H1 in regulating IFN-γ production by naive CD4⁺ T cells is consistent with a recent study using B7-H1-deficient mice (25).

The inhibition of T cell proliferation by anti-B7-H1 mAb was abolished by anti-IFN-γ mAb and iNOS inhibitors, indicating that it was mediated by IFN-γ-induced and iNOS-mediated NO production by macrophages. It has previously been reported that NO is a potent inhibitor of T cell proliferation, but does not affect the secretion of IL-2 or the expression of IL-2R (26). The NO-mediated inhibition of T cell proliferation was associated with markedly reduced tyrosine phosphorylation of Jak3 and STAT5, suggesting a blockade of IL-2R signaling for proliferation (27). This is consistent with our present observation that the B7-H1 blockade resulted in the accumulation of IL-2 in culture supernatants, suggesting a blockade of the IL-2/IL-2R-mediated proliferation pathway. Actually, the tyrosine phosphorylation of STAT5 in anti-CD3-stimulated naive CD4⁺ T cells in the presence of macrophages was markedly decreased by the addition of anti-B7-H1 mAb, as estimated by Western blotting (data not shown). Moreover, it has been reported that activated T cells were arrested in the G₀/G₁ phase of the cell cycle in the presence of NO-producing macrophages (28). We confirmed that anti-CD3-stimulated CD4⁺ T cells in the presence of macrophages were mostly arrested in the G₀/G₁ phase when anti-B7-H1 mAb was added, as estimated by nuclear staining with propidium iodide and flow cytometry (data not shown). We also noticed that sub-G₀/G₁ population, representing apoptotic nuclei, was not increased by the addition of anti-B7-H1 mAb (data not shown). These results indicated that the blockade of cell cycle progression, but not the induction of apoptosis, was primarily responsible for the inhibition of T cell proliferation by anti-B7-H1 mAb.

The CD4⁺ T cell-derived, IFN-γ-mediated NO production by macrophages plays a pivotal role in the protective immunity against microbial pathogens (29). We have previously demonstrated that the B7-H1 expression on macrophages is also up-regulated by microbial products such as LPS (14). If the PD-1/B7-H1 interaction down-regulates NO production by macrophages through cognate interaction with Ag-specific CD4⁺ T cells, the blockade of PD-1/B7-H1 interaction may be beneficial for augmenting the protective immunity against microbial pathogens. In contrast, IFN-γ, macrophages, and NO have been also implicated in the pathogenesis of various inflammatory diseases, such as rheumatoid arthritis (30, 31, 32), multiple sclerosis (33), psoriasis (34), and type 1 diabetes (35, 36, 37, 38, 39, 40, 41). It has been reported that PD-1-dificient mice develop lupus-like arthritis (7) and that B7-H1-dificient mice develop exacerbated experimental autoimmune encephalomyelitis (25). We have also demonstrated that the blockade of PD-1 and B7-H1 exacerbated contact hypersensitivity (42) and autoimmune diabetes (43) in mice. If the PD-1/B7-H1-mediated down-regulation of IFN-γ and NO is involved in the suppression of inflammatory responses, blockade of the PD-1/B7-H1 interaction may exacerbate the infection-associated pathologies, such as pneumonitis and hepatitis. Therefore, the potentially beneficial or detrimental effect of the PD-1/B7-H1 blockade remains to be determined by additional studies using mouse models of microbial infection.

The original descriptions of cytokine-inducible, NO-mediated inhibition of lymphocyte proliferation were published over a decade ago (44, 45). These observations have been verified in various systems (27, 46, 47, 48, 49). It has been reported that in murine Trypanosoma brucei infection, the activation of macrophages to produce NO leads to impaired lymphocyte responses and immunosuppression (50). Moreover, NO modulates the severity of autoimmune diseases by exerting a protective effect through inhibition of autoreactive T cells. In experimental autoimmune encephalomyelitis, mice with a targeted disruption of the iNOS gene show increased severity due to increased proliferation of pathogenic T cells (51, 52). In a rat model of autoimmune nephritis, pharmacologic inhibition of iNOS with 1-N⁶-(1-iminoethyl)-lysine intensified the renal injury, suggesting that inhibition of iNOS may alter the expansion of nephritogenic T cells (53). The production of large amounts of NO by iNOS is a double-edged sword, eliciting pro- or anti-inflammatory effects in different situations. We have demonstrated in this study that the PD-1/B7-H1 interaction plays an important role in regulating NO production by macrophages. Therefore, more beneficial effects of the PD-1/B7-H1 blockade to augment immune responses against pathogens and tumors may be achieved by combination with the iNOS blockade.

We thank Drs. Y. Iwakura and T. Taniguchi for the mice, and X. Xue and Dr. H. Nakano for their helpful suggestions.

The authors have no financial conflict of interest.