Abstract
Natural killer (NK) cells were first identified by their ability to kill tumour or virally infected cells without prior sensitization. In spite of this, the actual role of NK cells in tumour immunotherapy remains controversial. This study therefore set out to investigate the potential of Streptococcus salivarius K12, a gram-positive bacterium that has a history of commercial application as a probiotic in New Zealand, for use as a NK cell adjuvant, applying the therapy using B16.OVA melanoma as a model. To confirm that S. salivarius K12 was able to induce efficient activation of NK cells, I first screened a number of gram-positive and gram-negative bacteria for their ability to induce IFNγ release from NK cells. Using ELISA and fluorescence activated cell sorting (FACS) I found that gram-positive bacteria stimulated a rapid release (<24 hours) of IFNγ from dendritic cell: NK cell co-cultures. Cytotoxicity assays showed that despite optimal activation of NK cells by S. salivarius K12, their cytotoxic activity was not enhanced above that of the naive NK cells. Dissecting NK cells into two subsets based on their CD27 expression using FACS, it was discovered that S. salivarius K12-NK activation was predominantly exerted on the mature CD27high NK cell subset and was dependent on membrane-contact with DC and IL-12/IL-18 expression. NK cell activation was found to be independent of Ly49A expression, a marker that can be used in C57BL/6 mice to discriminate between 'unlicensed' NK cells, and those that had been 'licensed' through interaction with self MHC during development. Therefore having a setting where the addition of S. salivarius K12 activates NK cells, I investigated whether these NK cells were recruited to the draining lymph nodes where they could potentially influence the adaptive immune response. A range of adjuvant-activated and S. salivarius K12-activated DC were injected subcutaneously into the flanks of mice and tested their ability to recruit NK cells to the draining lymph node. The adjuvants differed markedly in their ability to recruit NK cells with S. salivarius K12 being the most effective. To determine if activated NK cells would be of benefit in tumour immunotherapy, I investigated the ability of bacterially activated DCs to elicit anti-tumour responses in a B16.OVA melanoma model. Utilizing a therapeutic tumour model where treatment was started three days following tumour inoculation, I found a significant delay of tumour growth in mice that were immunized with ovalbumin-pulsed DC that had been treated for 4 hours with S. salivarius K12 as opposed to other adjuvants tested. I also determined that in vivo depletion of NK cells completely abolished the benefit of DC immunotherapy. A therapeutic tumour experiment where DC were primed in the presence or absence of tumour antigen showed that while NK cells were critical for the antigen-dependent anti-tumour response they did not appear to exert an effector function. To investigate the role of NK cells in priming the anti-tumour response I next utilized a prophylactic setting, where mice were challenged with tumours sixty days after DC immunization. By depleting CD4+/CD8+ T cells and NK cells before time of priming or challenge, I tentatively showed that all three subsets of cells play a role in the anti-tumour response, although NK cells may play a greater role at time of challenge.