In this study, we have confirmed that anti–4-1BB mAb-induced cross-linking of the 4-1BB receptor in vitro provides a potent costimulatory effect on anti-CD3–induced polyclonal T cell proliferation. However, we also found that 4-1BB–mediated proliferation was highly biased toward the CD8+ subset of T cells. For instance, the addition of anti-4-1BB mAb 3H3 (10 μg/ml) reduced the threshold of anti-CD3 from 1,000 ng/ml to 10 ng/ml and increased the absolute value of cpm of [3H]thymidine incorporated (Fig. B) by CD8+ T cells 100-fold. In contrast, signaling through the 4-1BB receptor on CD4+ T cells resulted in a fourfold increase in proliferative capacity (Fig. A). In both cases, 4-1BB–mediated costimulation was maximal at 48 h. Stimulation of either population with varying doses of anti-CD3 alone demonstrated that these two populations are near-equivalent in their ability to respond to CD3-mediated proliferative signals, with CD4+ T cells being somewhat more responsive in these experiments. Furthermore, in contrast with 4-1BB, CD28-mediated costimulation was highly biased toward CD4+ T cell activation (Fig. ) and optimal activation occurred at 72 h. Thus, there is a reciprocal relationship of costimulation through these two receptor systems with regard to T cell subset proliferation.
The ability of anti–4-1BB mAbs to enhance T cell activation was found to be more dependent upon epitope recognition than antibody isotype or antibody avidity, because both antibodies of the IgM and IgG2A isotype were equally effective as costimulators, and within the IgG2A isotype, antibodies having similar avidity varied markedly with respect to their biological activity. For instance, all of the antibodies, regardless of avidity or isotype, that bound near or at the ligand binding site were most effective at inducing costimulation. This observation led us to question whether or not enhanced proliferation induced by these reagents was a consequence of their ability to block a negative proliferation signal transmitted by 4-1BBL binding to the 4-1BB receptor. This was not found to be the case (Fig. B). Rather, our data support the view that cross-linking the 4-1BB receptor on an activated T cell transmits a positive activation signal to the cell. This notion is further supported by the observation that 4-1BB cross-linking on activated CD8+ T cells led to rapid tyrosine phosphorylation of multiple substrates and that the pattern of phosphorylation was nearly identical to anti-CD3–induced protein tyrosine phosphorylation. This finding suggests that the 4-1BB receptor on CD8+ T cells may be structurally or functionally coupled to the TCR of these cells.
In light of the fact that CD8+
T cells preferentially respond to 4-1BB costimulation, we assessed the effects of several anti-4-1BB mAbs in vivo. For these studies, we chose to examine the effects of anti-4-1BB mAbs on CTL development as this activity represents a major function of the CD8+
T cell subset. To facilitate these studies, we chose acute GVHD, cardiac allograft, and MHC-mismatched skin transplantation as murine models for our studies. The ability of anti-4-1BB mAbs to enhance GVHD (Fig. A
) leading to marked depletion of host splenocytes (Fig. B
) appears to be the result of preferential costimulation and amplification of CD8+
T cells in vivo (Fig. C
). In contrast, the percentage of CD4+
T cells did not change (data not shown). In cardiac allograft (Fig. ) and skin transplant studies (data not shown) we found that transplants were rejected faster in mice treated with any of a variety of anti-4-1BB mAbs than what was observed in control mice injected with the 6E9 mAb. We also found certain functional differences in this system. For example, certain mAbs, such as 22B6 and 1D8, were poor inducers of proliferation, whereas mAbs such as 3E1 and 3H3 were strong inducers of proliferation. Nevertheless, all were equally effective at enhancing GVHD in mice and in rejecting MHC-mismatched skin transplants (data not shown). This suggests that the induction of CD8+
T cell proliferation was not critical to the observed functional effects in these systems. To explore further this issue, we examined lymphokine production by CD4+
T cells or unseparated T cells costimulated with anti-CD3 alone or via costimulation with either anti-4-1BB or anti-CD28 over a 96-h time course. The results of this experiment demonstrated that anti-4-1BB costimulation had a profound effect upon the upregulation of IFN-γ production by CD8+
T cells but had little or no effect upon IL-2, IL-4, or IL-10 production in either subset. In contrast, anti-CD28–mediated costimulation had only a slight effect upon IFN-γ production by either subset. In contrast, anti-CD28 costimulation markedly enhanced IL-2, IL-4, and IL-10 production in CD4+
T cells. When unseparated T
cells were activated as described above, the only notable difference was the level of IFN-γ produced after anti–4-1BB costimulation. Allowing that ~30–35% of these cells were CD8+
, on a cell by cell basis, CD8+
T cells in these cultures produced lower (~50%) levels of IFN-γ, which was probably caused by inhibition of Th1
cytokine production by IL-4 and IL-10.
From these studies, it would appear that 4-1BB/4-1BBL differs from other previously defined receptor–ligand systems known to be important for costimulation of T cells in that its primary effect appears to be the regulation of CD8+ T cell activation. In contrast, other systems such as CD28– CD80/CD86, CD27/CD27L, or gp39/CD40 primarily affect the function of CD4+ T cells. Finally, the increase in amplitude of the CTL response during GVHD, cardiac allograft, and skin rejection is not just limited to these forms of alloantigenic stimulation; we have also observed the same type of phenomena in mice injected with either nonmetastatic or highly metastatic, poorly immunogenic tumors (manuscript in preparation). Under these conditions, the administration of anti–4-1BB promoted rejection of large established tumors and markedly enhanced animal survival. Taken together, these observations point to a potential therapeutic role for this receptor in the management of cancer and chronic infectious diseases.