FcγRIIB controls antibody-mediated target cell depletion by ITIM-independent mechanisms.

Many therapeutic antibodies deplete target cells and elicit immunotherapy by engaging activating Fc gamma receptors (FcγRs) on host effector cells. These antibodies are negatively regulated by the inhibitory FcγRIIB (CD32B). Dogma suggests inhibition is mediated through the FcγRIIB immunoreceptor tyrosine-based inhibition motif (ITIM), negatively regulating immunoreceptor tyrosine-based activation motif (ITAM)-mediated signaling from activating FcγR. To assess this, we generated experimental models expressing human (h)FcγRIIB on targets or effectors, lacking or retaining ITIM signaling capacity. We demonstrate that signaling through the hFcγRIIB ITIM is dispensable for impairing monoclonal antibody (mAb)-mediated depletion of normal and malignant murine target cells through three therapeutically relevant surface receptors (CD20, CD25, and OX40) affecting immunotherapy. We demonstrate that hFcγRIIB competition with activating FcγRs for antibody Fc, rather than ITIM signaling, is sufficient to impair activating FcγR engagement, inhibiting effector function and immunotherapy.

Domain binding and isotype dictate the activity of anti-human OX40 antibodies.

BACKGROUND: Previous data suggests that anti-OX40 mAb can elicit anti-tumor effects in mice through deletion of Tregs. However, OX40 also has powerful costimulatory effects on T cells which could evoke therapeutic responses. Human trials with anti-OX40 antibodies have shown that these entities are well tolerated but to date have delivered disappointing clinical responses, indicating that the rules for the optimal use of anti-human OX40 (hOX40) antibodies is not yet fully understood. Changes to timing and dosages may lead to improved outcomes; however, here we focus on addressing the role of agonism versus depleting activity in determining therapeutic outcomes. We investigated a novel panel of anti-hOX40 mAb to understand how these reagents and mechanisms may be optimized for therapeutic benefit. METHODS: This study examines the binding activity and in vitro activity of a panel of anti-hOX40 antibodies. They were further evaluated in several in vivo models to address how isotype and epitope determine mechanism of action and efficacy of anti-hOX40 mAb. RESULTS: Binding analysis revealed the antibodies to be high affinity, with epitopes spanning all four cysteine-rich domains of the OX40 extracellular domain. In vivo analysis showed that their activities relate directly to two key properties: (1) isotype-with mIgG1 mAb evoking receptor agonism and CD8+ T-cell expansion and mIgG2a mAb evoking deletion of Treg and (2) epitope-with membrane-proximal mAb delivering more powerful agonism. Intriguingly, both isotypes acted therapeutically in tumor models by engaging these different mechanisms. CONCLUSION: These findings highlight the significant impact of isotype and epitope on the modulation of anti-hOX40 mAb therapy, and indicate that CD8+ T-cell expansion or Treg depletion might be preferred according to the composition of different tumors. As many of the current clinical trials using OX40 antibodies are now using combination therapies, this understanding of how to manipulate therapeutic activity will be vital in directing new combinations that are more likely to improve efficacy and clinical outcomes.

Augmentation of CD134 (OX40)-dependent NK anti-tumour activity is dependent on antibody cross-linking.

CD134 (OX40) is a member of the tumour necrosis factor receptor superfamily (TNFRSF). It acts as a costimulatory receptor on T cells, but its role on NK cells is poorly understood. CD137, another TNFRSF member has been shown to enhance the anti-tumour activity of NK cells in various malignancies. Here, we examine the expression and function of CD134 on human and mouse NK cells in B-cell lymphoma. CD134 was transiently upregulated upon activation of NK cells in both species. In contrast to CD137, induction of CD134 on human NK cells was dependent on close proximity to, or cell-to-cell contact with, monocytes or T cells. Stimulation with an agonistic anti-CD134 mAb but not CD134 ligand, increased IFNγ production and cytotoxicity of human NK cells, but this was dependent on simultaneous antibody:Fcγ receptor binding. In complementary murine studies, intravenous inoculation with BCL1 lymphoma into immunocompetent syngeneic mice resulted in transient upregulation of CD134 on NK cells. Combination treatment with anti-CD20 and anti-CD134 mAb produced a synergistic effect with durable remissions. This therapeutic benefit was abrogated by NK cell depletion and in Fcγ chain -/- mice. Hence, anti-CD134 agonists may enhance NK-mediated anti-tumour activity in an Fcγ receptor dependent fashion.

Detection of Experimental and Clinical Immune Complexes by Measuring SHIP-1 Recruitment to the Inhibitory FcγRIIB.

Fc γ receptors (FcγR) are involved in multiple aspects of immune cell regulation, are central to the success of mAb therapeutics, and underpin the pathology of several autoimmune diseases. However, reliable assays capable of accurately measuring FcγR interactions with their physiological ligands, IgG immune complexes (IC), are limited. A method to study and detect IC interactions with FcγRs was therefore developed. This method, designed to model the signaling pathway of the inhibitory FcγRIIB (CD32B), used NanoLuc Binary Interaction Technology to measure recruitment of the Src homology 2 domain-containing inositol phosphatase 1 to the ITIM of this receptor. Such recruitment required prior cross-linking of an ITAM-containing activatory receptor, and evoked luciferase activity in discrete clusters at the cell surface, recapitulating the known biology of CD32B signaling. The assay detected varying forms of experimental IC, including heat-aggregated IgG, rituximab-anti-idiotype complexes, and anti-trinitrophenol-trinitrophenol complexes in a sensitive manner (≤1 µg/ml), and discriminated between complexes of varying size and isotype. Proof-of-concept for the detection of circulating ICs in autoimmune disease was provided, as responses to sera from patients with systemic lupus erythematosus and rheumatoid arthritis were detected in small pilot studies. Finally, the method was translated to a stable cell line system. In conclusion, a rapid and robust method for the detection of IC was developed, which has numerous potential applications including the monitoring of IC in autoimmune diseases and the study of underlying FcγR biology.

Development and Characterization of Monoclonal Antibodies Specific for Mouse and Human Fcγ Receptors.

FcγRs are key regulators of the immune response, capable of binding to the Fc portion of IgG Abs and manipulating the behavior of numerous cell types. Through a variety of receptors, isoforms, and cellular expression patterns, they are able to fine-tune and direct appropriate responses. Furthermore, they are key determinants of mAb immunotherapy, with mAb isotype and FcγR interaction governing therapeutic efficacy. Critical to understanding the biology of this complex family of receptors are reagents that are robust and highly specific for each receptor. In this study, we describe the development and characterization of mAb panels specific for both mouse and human FcγR for use in flow cytometry, immunofluorescence, and immunocytochemistry. We highlight key differences in expression between the two species and also patterns of expression that will likely impact on immunotherapeutic efficacy and translation of therapeutic agents from mouse to clinic.

Immunotherapy targeting inhibitory Fcγ receptor IIB (CD32b) in the mouse is limited by monoclonal antibody consumption and receptor internalization.

Genetic deficiency of the inhibitory Fc receptor, FcγRIIB (CD32b), has been shown to augment the activity of activatory FcγR and promote mAb immunotherapy. To investigate whether mAbs capable of blocking FcγRIIB have similar capacity, we recently generated a panel of specific anti-mouse FcγRIIB mAbs that do not cross-react with other FcRs, allowing us to study the potential of FcγRIIB as a therapeutic target. Previous work revealed a number of these mAbs capable of eliciting programmed cell death of targets, and in the present study we demonstrated their ability to promote target cell phagocytosis. However, in a variety of murine tumor models, anti-FcγRIIB mAbs demonstrated limited therapeutic activity despite optimized treatment regimens. Unexpectedly, we observed that the anti-FcγRIIB mAbs are rapidly and extensively consumed in vivo, both by the tumor and host cells, including B cells, leading to a precipitous loss from the circulation. Closer analysis revealed that the anti-FcγRIIB mAbs become extensively internalized from the cell surface within 24 h in vivo, likely explaining their suboptimal efficacy. Subsequent studies revealed that anti-FcγRIIB mAb immunotherapy was effective when used against FcγRIIB(+) tumors in FcγRIIB(-/-) recipients, indicating that consumption of the mAb by nontumor cells is the primary limitation of these reagents. Importantly, similar rates of internalization were not seen on human target cells, at least in vitro. These studies further highlight the need to determine the propensity of mAb therapeutics to internalize target receptors and also identify potential key differences between human and mouse cells in this respect.

Development and characterisation of monoclonal antibodies specific for the murine inhibitory FcγRIIB (CD32B).

Fc receptors (FcRs) play a key role in regulating and coordinating responses from both innate and adaptive arms of the immune system. The inhibitory Fc gamma receptor II (FcγRIIB; CD32) is central to this regulation with FcγRIIB(-/-) mice demonstrating augmented responses to mAb immunotherapy, elevated incidence and severity of auto-immunity, and increased response to mAb-mediated cancer therapy. To date, these observations have remained unexploited therapeutically, partly through a lack of specific mAb reagents capable of exclusively binding mouse FcγRIIB. Thus almost all of the FcγRIIB-binding mAb currently available, such as 2.4G2, also bind FcγRIII (CD16), and polyclonal reagents have limited availability and are of unproven specificity and avidity, making in vivo manipulation of FcγRIIB impossible. Following an extensive immunisation protocol using FcγRIIB(-/-) mice, we recently produced three unique mAb that are suitable for this purpose. Here we characterise these novel reagents and demonstrate that they fall into two distinct categories; those which cause phosphorylation and subsequent activation of FcγRIIB (agonistic) and those that block receptor phosphorylation (antagonistic). These two types of mAb exhibit different characteristics in a range of biochemical, cellular, and functional assays relevant to FcγRIIB activity and mAb therapy.

Enhanced FcαRI-mediated neutrophil migration towards tumour colonies in the presence of endothelial cells.

Neutrophils potently kill tumour cells in the presence of anti-tumour antibodies in vitro. However, for in vivo targeting, the neutrophils need to extravasate from the circulation by passing through endothelial barriers. To study neutrophil migration in the presence of endothelial cells in vitro, we established a three-dimensional collagen culture in which SK-BR-3 tumour colonies were grown in the presence or absence of an endothelial barrier. We demonstrated that - in contrast to targeting FcγR on neutrophils with mAbs - targeting the immunoglobulin A Fc receptor (FcαRI) instead triggered neutrophil migration and degranulation leading to tumour destruction, which coincided with release of the pro-inflammatory cytokines interleukin (IL)-1ß and tumour necrosis factor (TNF)-α. Interestingly, neutrophil migration was enhanced in the presence of endothelial cells, which coincided with production of significant levels of the neutrophil chemokine IL-8. This supports the idea that stimulation of neutrophil FcαRI, but not FcγR, initiates cross-talk between neutrophils and endothelial cells, leading to enhanced neutrophil migration towards tumour colonies and subsequent tumour killing.

Interaction with FcγRIIB is critical for the agonistic activity of anti-CD40 monoclonal antibody.

A high activatory/inhibitory FcγR binding ratio is critical for the activity of mAb such as rituximab and alemtuzumab that attack cancer cells directly and eliminate them by recruiting immune effectors. Optimal FcγR binding profiles of other anti-cancer mAb, such as immunostimulatory mAb that stimulate or block immune receptors, are less clear. In this study, we analyzed the importance of isotype and FcγR interactions in controlling the agonistic activity of the anti-mouse CD40 mAb 3/23. Mouse IgG1 (m1) and IgG2a (m2a) variants of the parental 3/23 (rat IgG2a) were engineered and used to promote humoral and cellular responses against OVA. The mouse IgG1 3/23 was highly agonistic and outperformed the parental Ab when promoting Ab (10-100-fold) and T cell (OTI and OTII) responses (2- to >10-fold). In contrast, m2a was almost completely inactive. Studies in FcγR knockout mice demonstrated a critical role for the inhibitory FcγRIIB in 3/23 activity, whereas activatory FcγR (FcγRI, -III, and -IV) was dispensable. In vitro experiments established that the stimulatory effect of FcγRIIB was mediated through Ab cross-linking delivered in trans between neighboring cells and did not require intracellular signaling. Intriguingly, activatory FcγR provided effective cross-linking of 3/23 m2a in vitro, suggesting the critical role of FcγRIIB in vivo reflects its cellular distribution and bioavailability as much as its affinity for a particular Ab isotype. In conclusion, we demonstrate an essential cross-linking role for the inhibitory FcγRIIB in anti-CD40 immunostimulatory activity and suggest that isotype will be an important issue when optimizing reagents for clinical use.

Systemic inflammation modulates Fc receptor expression on microglia during chronic neurodegeneration.

Chronic neurodegeneration is a major worldwide health problem, and it has been suggested that systemic inflammation can accelerate the onset and progression of clinical symptoms. A possible explanation is that systemic inflammation "switches" the phenotype of microglia from a relatively benign to a highly aggressive and tissue-damaging phenotype. The current study investigated the molecular mechanism underlying this microglia phenotype "switching." We show in mice with chronic neurodegeneration (ME7 prion model) that there is increased expression of receptors that have a key role in macrophage activation and associated signaling pathways, including TREM-2, Siglec-F, CD200R, and FcγRs. Systemic inflammation induced by LPS further increased protein levels of the activating FcγRIII and FcγRIV, but not of other microglial receptors, including the inhibitory FcγRII. In addition to these changes in receptor expression, IgG levels in the brain parenchyma were increased during chronic neurodegeneration, and these IgG levels further increased after systemic inflammation. γ-Chain-deficient mice show modified proinflammatory cytokine expression in the brain after systemic inflammation. We conclude that systemic inflammation during chronic neurodegeneration increases the expression levels of activating FcγR on microglia and thereby lowers the signaling threshold for Ab-mediated cell activation. At the same time, IgG influx into the brain could provide a cross-linking ligand resulting in excessive microglia activation that is detrimental to neurons already under threat by misfolded protein.

Control of established melanoma by CD27 stimulation is associated with enhanced effector function and persistence, and reduced PD-1 expression of tumor infiltrating CD8(+) T cells.

The immune response to the tumor can be enhanced by targeting costimulatory molecules on T cells. As the CD70-CD27 costimulatory axis plays an important role in the activation, survival, and differentiation of lymphocytes, we have examined the efficacy of agonistic anti-CD27 antibodies as monotherapies for established melanoma in a murine model. We show that this approach leads to a substantial reduction in the outgrowth of both experimental lung metastases and subcutaneous tumors. Anti-CD27 treatment supports the maintenance of tumor-specific CD8(+) T cells within the tumor, reduces the frequency of FoxP3-expressing CD4(+) T cells within tumors, and potentiates the ability of NK1.1(+) and CD8(+) tumor infiltrating cells to secrete IFNγ upon coculture with tumor cells. The enhanced effector function correlated with lower levels of PD-1 expression on CD8(+) T cells from anti-CD27-treated mice. Despite the modulating effect of anti-CD27 on multiple cell types, only CD8(+) T cells were absolutely required for tumor control. The CD4(+) T cells were dispensable, whereas NK1.1(+) cells were needed during early stages of tumor growth but not for the effectiveness of anti-CD27. Thus, CD27-mediated costimulation provides a potent boost to multiple aspects of the endogenous responses to tumor, and may be exploited to enhance tumor immunity.

Ligation of CD11c during vaccination promotes germinal centre induction and robust humoral responses without adjuvant.

In this study, we investigated the mouse dendritic cell (DC) receptor, complement receptor 4 (CR4; CD11c/CD18), as an immunotarget for triggering humoral immunity. Comparison of antibody titres generated against a panel of 13 anti-antigen-presenting cell receptor monoclonal antibodies, with or without conjugated ovalbumin (OVA), revealed uniquely rapid and robust responses following CR4 targeting, with antibody titres approaching 1 : 100 000 7 days after a single dose of antigen. Furthermore, using just 100 ng OVA conjugated to anti-CD11c Fab', we generated anti-OVA titres greater than those produced by a 100-fold higher dose of OVA in complete Freund's adjuvant at day 28. These anti-OVA antibody titres were sustained and could be boosted further with targeted OVA on day 21. Investigations to explain this vaccine potency showed that, in addition to targeting splenic DC, anti-CDl1c antibodies delivered a powerful adjuvant effect and could boost humoral immunity against OVA even when the OVA was targeted to other molecules on DC, such as major histocompatibility complex class II, CD11a and CD11b. However, interestingly, this adjuvant effect was lost if OVA was targeted to other cells such as B cells via CD21 or CD19. The adjuvant effect was mediated through a marked enhancement of both germinal centre and extrafollicular plasma cell formation in responding spleens. These results demonstrate that anti-CD11c monoclonal antibody can both target antigen and act as a powerful adjuvant for rapid and sustained antibody responses. They also point to an interesting role for CR4 on DC in triggering B cells during humoral immunity.

CD11c provides an effective immunotarget for the generation of both CD4 and CD8 T cell responses.

The magnitude and quality of T cell responses generated when Ag is targeted to receptors on DC is influenced by both the specific receptor targeted and its distribution among DC subsets. Here we examine the targeting of the model Ag OVA to potential DC targets, including CD11c, CD205, MHC class II, CD40, TLR2 and FcgammaRII/III, using a panel of (Fab' x OVA) conjugates. In vitro studies identified CD11c, CD205 and MHC class II as superior and comparably effective immunotargets for the delivery of OVA to APC for presentation to T cells. In vivo studies, however, showed a marked advantage of targeting Ag to CD11c for both CD4 (OT-II) and CD8 (OT-I) responses, with robust stimulation after a single, low dose (equivalent to 0.5 microg OVA); in contrast, (anti-CD205 x OVA) and (anti-MHC class II x OVA) resulted in markedly less proliferation of both OT-I and OT-II cells. Biodistribution and immunohistochemical studies suggest that the exceptional ability of CD11c to capture Ag in lymphoid tissues may, at least partially, explain its ability to promote T cell responses. These results suggest that targeting antigen via CD11c offers a previously unappreciated strategy for vaccine development which, unlike most targets, delivers robust responses of both CD4 and CD8 T cells.

Eradication of lymphoma by CD8 T cells following anti-CD40 monoclonal antibody therapy is critically dependent on CD27 costimulation.

Growing evidence points to the potential of agonistic anti-CD40 mAbs as adjuvants for vaccination against cancer. These appear to act by maturing dendritic cells (DCs) and allowing them to prime CD8 cytotoxic T lymphocytes (CTLs). Although it is well established that optimal T-cell priming requires costimulation via B7:CD28, recent studies emphasize the contribution of TNF receptors to this process. To understand how anti-CD40 mAbs trigger effective antitumor immunity, we investigated the role of TNFR superfamily members CD27 and 4-1BB in the generation of this immunity and showed that, although partially dependent on 4-1BB:4-1BBL engagement, it is completely reliant on CD27:CD70 interactions. Importantly, blocking CD70, and to some extent 4-1BBL, during anti-CD40 treatment prevented accumulation of tumor-reactive T cells and subsequent tumor protection. However, it did not influence changes in DC number, phenotype, nor the activity of CTLs once immunity was established. We conclude that CD27:CD70 and 4-1BB:4-1BBL interactions are needed for DC-driven accumulation of antitumor CTLs following anti-CD40 mAb treatment. Finally, in support of the critical role for CD70:CD27, we show for the first time that agonistic anti-CD27 mAbs given without a DC maturation signal completely protect tumor-bearing mice and provide a highly potent reagent for boosting antitumor T-cell immunity.

A new anti-idiotype antibody capable of binding rituximab on the surface of lymphoma cells.

The chimeric anti-CD20 monoclonal antibody (mAb), rituximab, is an established part of the management of many non-Hodgkin lymphomas. The in vivo action of rituximab remains elusive, and this partially reflects a lack of highly specific reagents to detect rituximab binding at the cell surface. Here we report a new high-affinity mAb (MB2A4) with fine specificity for the idiotype of rituximab. It is able to detect rituximab in vitro, in the presence of high levels of human immunoglobulin G (IgG), in the serum of patients receiving rituximab therapy, and, surprisingly, when rituximab is bound to CD20 on the cell surface. We propose that the anti-idiotype (Id) binds to rituximab molecules bound univalently at the cell surface, facilitated by the relatively high off-rate of rituximab. This reagent provides new insights into the binding of rituximab at the cell surface and demonstrates a mode of binding that could be exploited for the surface detection of other mAbs with clinical and biologic applications.

Multivalent conjugates of poly-gamma-D-glutamic acid from Bacillus licheniformis with antibody F(ab') and glycopeptide ligands.

Poly-gamma-D-glutamic acid from Bacillus licheniformis is a water-soluble, nontoxic, nonimmunogenic exopolymer. Using synthetic linkers, the alpha-carboxylate side chains of PGA were conjugated to an exposed thiol side chain of an antibody F(ab') fragment, Mc109F4. Analysis of the PGA-Mc109F4 conjugate by gel filtration HPLC revealed a mixture of multivalent conjugates. The PGA-Mc109F4 conjugate retained biological activity, but showed a lower binding affinity to target BCL3B3 cells than free Mc109F4 F(ab')(2) by flow cytometry, and a lower efficacy for BCL3B3 growth inhibition than free Mc109F4 F(ab')(2). PGA was also conjugated with the free amino group of glycopeptide antibiotic vancomycin. The PGA-vancomycin conjugate showed slightly lower antibacterial activity than free vancomycin versus susceptible Bacillus subtilis, but slightly higher activity versus intrinsically resistant Leuconostoc mesenteroides.

CD20-induced lymphoma cell death is independent of both caspases and its redistribution into triton X-100 insoluble membrane rafts.

Rituximab is routinely used for the treatment of neoplasia, although the mechanism of action remains uncertain. In the current study, CD20-induced apoptosis was investigated with a panel of anti-CD20 monoclonal antibodies (mAb) in a wide range of cell lines. A hierarchy of mAb activity was apparent, with the B1 mAb generally the most potent. Apoptosis through CD20 was dependent on the nature of mAb binding and correlated with the extent of homotypic cell adhesion induced. However, using anti-CD20 mAb, which vary in the extent to which they redistribute wild-type and mutant CD20 molecules to membrane rafts, we showed that CD20-induced apoptosis was independent of translocation to TX-100 insoluble rafts. Using crmA-transfected cells and caspase inhibitors, we showed that phosphatidylserine translocation and mitochondrial permeability transition evoked during CD20-induced apoptosis appeared caspase independent. Furthermore, in cytoplasts which lack mitochondria and in Bcl(2)-transfected cells, phosphatidylserine was still translocated to the cell surface after CD20 stimulation. Together, these data imply that CD20 can evoke apoptosis without the involvement of mitochondria and caspases and irrespective of redistribution into TX-100 insoluble membrane rafts.

T cell immunity to lymphoma following treatment with anti-CD40 monoclonal antibody.

In this study we demonstrate that treatment with anti-CD40 mAb eradicates a range of mouse lymphomas (BCL(1), A31, A20, and EL4), but only when used against i.v. tumor doses in excess of 10(7) cells. Only partial protection was seen against smaller tumor loads. We saw no evidence that anti-CD40 mAb changed the phenotype of the lymphomas or inhibited their growth in the initial period following treatment, but it did result in a rapid expansion of cytotoxic CD8(+) cells that was able to clear the neoplastic disease and provide long-term protection against tumor rechallenge. The CTL responses were blocked by mAb against a range of coreceptors and cytokines, including CD8, B7-1, B7-2, LFA-1, and IFN-gamma, but not CD4 or CTLA-4, indicating the presence of a conventional cellular Th1 response. Furthermore, we found evidence of cross-recognition between lymphomas (BCL(1) and A20) as measured by cytotoxicity and IFN-gamma responses in vitro and using tumor rechallenge experiments, suggesting common target Ags. Finally, although anti-CD40 was shown to stimulate NK cell killing, we could find no role for these cells in controlling tumor growth. These data underline the ability of anti-CD40 mAb to potentiate CTL responses and the potency of cellular immunity in eradicating large quantities of syngeneic tumor.

Expression and costimulatory effects of the TNF receptor superfamily members CD134 (OX40) and CD137 (4-1BB), and their role in the generation of anti-tumor immune responses.

This study addresses the relative importance of CD134 (OX40) and CD137 (4-1BB) in the costimulation of CD4+ and CD8+ T cells under comparable conditions of antigenic stimulation. We demonstrate that CD134 is capable of directly costimulating CD8+ T cells. However, costimulation of CD8+ T cells by CD134 is less potent than that triggered by CD137. The higher costimulatory activity of CD137, when compared with CD134, correlates well with its faster expression kinetics and higher levels on CD8+ T cells. Furthermore, induction of CD137 expression on CD8+ T cells is highly sensitive to low levels of TCR stimulation, which is in contrast with CD134. Conversely, CD134 is more effective than CD137 in costimulating CD4+ T cells. This, however, could not be attributed to differential expression. We also demonstrate that the transient nature of CD134 and CD137 expression on activated CD4+ T cells is the resultof proteolytic shedding. Consistent with the greater ability of CD137 to costimulate CD8+ T cells, stimulation of CD137 in vivo is considerably more effective than CD134 in augmenting anti-tumor immune responses. Therefore, agents that stimulate signaling via CD137 are likely to be more useful in clinical conditions where highly effective CD8+ CTL responses are required.