In our recent publication in The Journal of Immunology, we proposed that NK cells shape CD8+ T cell responses by migrating to s.c. lymph nodes and by killing Ag-primed CD8+ T cells. We are happy to reply to the questions raised by Dr. Manjili and Dr. Toor regarding potential mechanisms behind our observations. Following data by Medema and colleagues (1), Majili and Toor suggest that T cells produce protease inhibitors to protect themselves from perforin/granzyme self-killing. We found this explanation appealing but confusing at the same time. Indeed, Medema's paper (1) shows that tumor cells, not T cells, produce these molecules as a means of a tumor escape mechanism. Whether Ag-primed T cells produce such inhibitors to avoid NK cell killing remains to be established.
We showed that activated CD8 T cells expressed NKG2D ligands, rendering Ag-primed CD8+ T cells susceptible to NK cell-mediated lysis. In our opinion, this finding would explain the increased frequency of CD8+ T cells differentiating in the absence of NK cells. Nevertheless, it could also be possible, as suggested by Manjili and Toor, that CD8+ T cells surviving NK cell-mediated killing could proliferate less than CD8 T cells produced in the absence of NK cells. The reason for this possibility is unclear at this stage, and further experimental work would be required to address this hypothesis.
In bone marrow transplantation settings, it has been shown that NK cell alloreactivity influenced engraftment and relapse rates after HLA-mismatched myeloablative hematopoietic cell transplants (2). In addition, the effectiveness of donor NK cell alloreactivity to enhance antileukemic effects in mismatched haploidentical transplants is being explored (3). On the basis of work by Billiau (4) and Sandmaier (5), Manjili and Toor suggest positive effects of NK cells on T cells after bone marrow transplants. In a retrospective study, Panse and colleagues (6) correlated the percentage of NK cells present in the infused HLA-matched transplant with T cell chimerism in the host, a parameter that is associated with a positive transplantation outcome. In contrast, De Somer and colleagues (4) have developed a mouse model to investigate lymphohematopoietic host-versus-graft reactivity and antileukemic responses in MHC-matched chimeras, a setting that may be reminiscent of the majority of clinical transplants. This study suggests that lymphohematopoietic alloreactivity provoked by adoptively transferred T cells provides a cytokine environment leading to NK cell activation and subsequent NK cell-mediated killing of recipient tumor cells. Overall, none of these papers have addressed whether donor and recipient NK cells do indeed kill infused or resident CD8+ T cells. Therefore, comparing bone marrow transplants with local events occurring in s.c. lymph nodes may not lead to definitive conclusions.
We showed increased numbers of CD8 T cells primed in the absence of NK cells, accounting for more efficient secondary T cell-mediated responses. Competition for available cytokines may certainly explain some aspects of T cell differentiation and survival (6). Nevertheless, it has to be considered that the number of CD8 T cells differentiating in the absence of NK cells is a log scale higher than in controls (NK-sufficient mice). Therefore, CD8–CD8 T cell competition, rather than NK cell–CD8 T cell competition, could be proposed as a determinant in the outcome of T cell memory differentiation.
- Copyright © 2011 by The American Association of Immunologists, Inc.