Background: Recent studies demonstrated that natural killer (NK) cells play a regulatory role in immune responses. As for the interaction with T cells, NK cells coordinate T cell responses through not only influencing dendritic cell function, but also directly acting on T cells. Besides secreting several kinds of cytokines, NK cells also kill activated autologous T cells. However, whether NK cell-mediated cytotoxicity has any role in the primary T cell response remains elusive.

Methods: Peripheral blood samples were obtained from healthy volunteers with written informed consent. Naïve CD25− CD4+ T cells, NK cells, and monocytes were isolated with magnetic beads. Monocyte-derived dendritic cells (MoDCs) were generated and either immature MoDCs (ImMoDCs) or LPS-stimulated MoDCs (LPS-MoDCs) were used. Naïve CD4+ T cells cultured with allogeneic MoDCs (allo-MoDCs) in the presence or absence of autologous NK cells pre-activated by IL-18 and IL-15. Naïve CD4+ T cells were labeled with PKH67 dye before culture and proliferation of CD4+T cells was detected by the dilution of PKH67 signals by flow cytometry.

Results: We found that activated NK cells substantially inhibited CD4+ T cell proliferation in response to allo-MoDCs. The addition of cytokines such as IL-2 could not restore CD4+ T cell proliferation. However, NK cells were not able to inhibit CD4+ T cell growth when they were separated by a transmembrane. Even when naïve CD4+ T cells were cultured with LPS-MoDCs, which were resistant to NK-lysis, CD4+ T cell proliferation was also inhibited. Furthermore, even when activated NK cells were added to the culture after naïve CD4+ T cells were stimulated with allo-MoDCs for 24 hours, CD4+ T cell proliferation were also inhibited. These data suggested that NK cells directly acted on CD4+ T cells in this assay. Then we performed 4-hour cyototoxicity assays, in which activated (CD25+) CD4+ T cells isolated from the culture of naïve CD4+ T cells and allo-MoDCs at various time points were used as target cells and activated NK cells were used as effecter cells. Resting naïve CD4+ T cells were resistant to NK-lysis. However, at day 1, activated CD4+ T cells became susceptible. ImMoDC-stimulated CD4+ T cells were remarkably susceptible to NK-lysis, whereas LPS-DC-stimulated CD4+ T cells were relatively resistant. NK cells pretreated with concanamycin A could not lyse activated CD4+ T cells, indicating that killing was mediated by release of lytic granules. In contrast, at day 3, CD4+ T cells activated with either ImMoDCs or LPS-MoDCs were completely resistant to NK-lysis. We then analyzed the kinetics of MICA/B and HLA-E expression on naïve CD4+ T cells stimulated with allo-MoDCs. MICA/B expression, which was slightly induced at day 1 and peaked at day 3, was comparable between CD4+ T cells stimulated with ImMoDCs and those with LPS-MoDCs. Addition of blocking antibody against NKG2D to cytotoxic assays did not affect the susceptibility of activated CD4+ T cells to NK-lysis. In contrast, in agreement with the susceptibility to NK-lysis, HLA-E expression at day 1 was considerably higher on LPS-MoDC-stimulated CD4+ T cells than ImMoDC-stimulated CD4+ T cells. At day 3, HLA-E expression was further up-regulated, and similar expression levels were observed in both ImMoDC-stimulated CD4+ T cells and LPS-MoDC-stimulated CD4+ T cells. In addition, antibody-dependent blockade of the HLA-E-NKG2A interaction abrogated the relative resistance to NK-lysis of LPS-MoDC- stimulated CD4+ T cells at day 1.

Conclusions and Discussion: Naïve CD4+ T cells became transiently susceptible to NK-lysis shortly after stimulated with allo-MoDCs, and, in our assay, NK-lysis of activated CD4+ T cells is the major mechanisms by which NK cells inhibited the proliferation of CD4+ T cells. High level of HLA-E expression on activated CD4+ T cells prevents NK-lysis. These data suggest that the expression level of HLA-E on activated CD4+ T cells determines their fate when activated NK cells are nearby present. HLA-E and NKG2A may become a new target of immunoregulation.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author

Sign in via your Institution