Perforin Is Important For Both CD4+ and CD8+ T Cell-Mediated Graft-Versus-Tumor Effect But Plays Differential Roles In CD4+ and CD8+ T Cell Expansion After Allogeneic Transplantation
Abstract Graft versus tumor (GVT) effect is the desired and integral outcome for successful allogeneic bone marrow transplantation (allo-BMT) for cancer patients. This effect is dependent on T cell mediated recognition and elimination of residual host tumor cells present after allo-BMT. T cell killing is mediated primarily via three pathways: perforin/granzymes, Fas/FasL, and cytotoxic cytokines. Recent work from our lab has revealed a detrimental role for granzyme B (GzmB) in GVT effect due to its role in activation induced cell death (AICD) of CD8+ T cells. As a result, GzmB-/- CD8+ T cells exhibited higher expansion after allo-BMT and subsequently provided better tumor control. Our current study sought to determine the role of perforin (Prf1) in GVT effect mediated by both CD4+ and CD8+ T cells. Using the MHC-mismatched C57BL/6 (H-2b) to BALB/c (H-2d) allo-BMT model, we first confirmed previous findings that when transplanting CD8+ T cells along with T cell depleted (TCD) BM cells, donor CD8+ T cells require Prf1 to mediate GVT effect against allogeneic A20 lymphoma (Fig 1A, Prf1-/- (n=4) vs WT (n=4), *P<0.05). In addition, our data suggest that Prf1 is also required for CD4+ T cells to effectively mediate GVT effect against A20, as transplant with Prf1-/- CD4+CD25- T cells does not control tumor growth as well as WT controls (Fig 1B). Our previous work showed that GzmB deficiency allows for less AICD and subsequently more CD8+ T cell expansion. New data now show a similar effect for Prf1 in CD8+ T cell accumulation, as Prf1-/- CD8+ T cells outcompete WT CD8+ T cells (CD45.1+) when these two genotypes are mixed in equal numbers and transplanted into tumor bearing BALB/c mice (n=5/time point, *P=0.02 day 9)(Fig 1C). This competitive advantage was due to less AICD in the Prf1-/- CD8+ T cells. However, Prf1 appears to be required for efficient GVT activity, because the higher number of Prf1-/- CD8+ T cells are still less capable than WT counterparts in controlling tumor growth. We next tested the effect of Prf1 in AICD in CD4+CD25- T cells, and again co-transplanted WT CD45.1+ and Prf1-/- CD4+CD25- T cells into tumor bearing mice for a competition assay. Unexpectedly, WT CD4+CD25- T cells accumulate to significantly higher numbers when in direct competition with Prf1-/- CD4+CD25- T cells (n=4/time point, **,P<0.01)(Fig 1D). When we measured apoptotic cells with Annexin V staining, we found that WT CD4+CD25- T cells still had significantly more AICD (Prf1-/- 38.3 ± 4.2% vs. WT 48.1 ± 5.1%, P<0.01 on day 7 post-BMT; Prf1-/- 12.7 ± 1.0% vs. WT 18.1 ± 3.4%, P<0.03 on day 9 post-BMT). This result suggests that while Prf1 has an important role in AICD, it may also play a role in another feature of CD4+ T cell biology. We then explored the hypothesis that may Prf1 promote CD4+ T cell proliferation by evaluating Hoescht staining on day 9 post-BMT. Preliminary results suggest that Prf1 may enhance T cell proliferation, as Prf1-/- CD4+ T cells have less actively dividing cells at this time point. Therefore, Prf1 appears to have a surprising role after allo-BMT in sustaining T cell expansion for CD4+ T cells, but not for CD8+ T cells. Another factor influencing GVT effect may be T cell phenotype. Our previous work with CD8+ T cells suggests that more effector memory (CD62LLOWCD44HIGH) T cells accumulate in the absence of GzmB, and that GzmB-/- CD8+ T cells exhibited higher GVT activity than WT controls. We now found that while Prf1-/- CD4+ T cells also skewed towards the effector memory phenotype (CD62LLOWCD44HIGH), loss of Prf1 still reduced the ability of CD4+ T cells to control tumor growth in this model of allo-BMT. In summary, our results suggest that Prf1 plays an important role in GVT responses mediated not only by CD8+ T cells but also by CD4+ T cells, which were shown in previous literature to mainly utilize Fas ligand and cytokine systems to mediate GVT activity. In addition, Prf1 can cause AICD to both CD4+ and CD8+ T cells after allo-BMT. While Prf1-induced AICD reduces CD8+ T cell expansion, Prf1 appears to play a previously unrecognized role enhancing CD4+ T cell proliferation via an unidentified mechanism. Disclosures: No relevant conflicts of interest to declare.