Reversal of tumor acidosis by systemic buffering reactivates NK cells to express IFN-γ and induces NK cell-dependent lymphoma control without other immunotherapies.

Like other immune cells, natural killer (NK) cells show impaired effector functions in the microenvironment of tumors, but little is known on the underlying mechanisms. Since lactate acidosis, a hallmark of malignant tissue, was shown to contribute to suppression of effective antitumor immune responses, we investigated the impact of tissue pH and lactate concentration on NK-cell functions in an aggressive model of endogenously arising B-cell lymphoma. The progressive loss of IFN-γ production by NK cells observed during development of this disease could be ascribed to decreased pH values and lactate accumulation in the microenvironment of growing tumors. Interestingly, IFN-γ expression by lymphoma-derived NK cells could be restored by transfer of these cells into a normal micromilieu. Likewise, systemic alkalization by oral delivery of bicarbonate to lymphoma-developing mice was capable of enhancing IFN-γ expression in NK cells and increasing the NK-cell numbers in the lymphoid organs where tumors were growing. By contrast, NK-cell cytotoxicity was dampened in vivo by tumor-dependent mechanisms that seemed to be different from lactate acidosis and could not be restored in a normal milieu. Most importantly, alkalization and the concomitant IFN-γ upregulation in NK cells were sufficient to significantly delay tumor growth without any other immunotherapy. This effect was strictly dependent on NK cells.

Recruitment of natural killer cells in advanced stages of endogenously arising B-cell lymphoma: implications for therapeutic cell transfer.

During inflammation and in transplantable tumor models, natural killer (NK) cells are recruited to pathologic tissues and activated to produce proinflammatory cytokines favoring adaptive immune responses of the T-helper type 1 (Th1) type. Interferon (IFN)-γ is needed to induce chemokines that attract NK cells in transplanted tumors. Nothing, however, is known on NK-cell migration in spontaneous tumors. As effective recruitment is a prerequisite for therapeutic NK-cell transfer, we investigated the cytokine milieu and the mechanisms that are instrumental for NK-cell accumulation in an endogenous tumor model. We make use of λ-myc transgenic mice that harbor the c-myc oncogene and develop spontaneous B-cell lymphoma. In contrast to lymphomas induced by tumor cell injection, virtually no IFN-γ produced by NK or by other cells was present in the tumor environment, particularly in advanced stages. Dendritic cells showed an impaired expression of interleukin-12, which is suggestive of deficient Th1 priming. The IFN-γ-dependent chemokines CXCL9 and CXCL10 were pivotal for NK-cell migration in the endogenous lymphoma model. Although IFN-γ was absent in late tumor stages, there was still expression of CXCL9 and CXCL10 with an ongoing influx of NK cells. The results demonstrate that transplantable tumor models do not reflect the situation as found in endogenously arising neoplasia, because in the latter, effective Th1 and cytotoxic T-lymphocyte responses are presumably not induced because of impaired IFN-γ production. The data also suggest that CXCL9 and CXCL10 production and NK-cell migration become independent of IFN-γ during tumor progression, and therefore support approaches of adoptive NK-cell transfer that hold promise for treatment of cancer.

Requirements for control of B-cell lymphoma by NK cells.

The role of NK cells in the control of endogenously arising tumors is still unclear. We monitored activation and effector functions of NK cells in a c-myc-transgenic mouse model of spontaneously arising lymphoma. At early stages, tumors demonstrated reduced MHC class I expression and increased expression of natural killer group 2D ligands (NKG2D-L). NK cells in these tumors showed an activated phenotype that correlated with the loss of tumor MHC class I. With increasing tumor load however, NK-cell effector functions became progressively paralyzed or exhausted. In later stages of disease, tumors re-expressed MHC class I and lost NKG2D-L, suggesting a role of these two signals for NK cell-mediated tumor control. Testing a panel of lymphoma cell lines expressing various MHC class I and NKG2D-L levels suggested that NK cell-dependent tumor control required a priming and a triggering signal that were provided by MHC class I down-regulation and by NKG2D-L, respectively. Deleting either of the "two signals" resulted in tumor escape. At early disease stages, immune stimulation through TLR-ligands in vivo efficiently delayed lymphoma growth in a strictly NK cell-dependent manner. Thus, NK-receptor coengagement is crucial for NK-cell functions in vivo and especially for NK cell-mediated tumor surveillance.