KLRG1+ natural killer cells protect against pulmonary metastatic disease by immunosurveillance.

Natural killer cells (NK cells) play a key role in cancer immunosurveillance. However, their activity is highly dependent upon their maturation stage, which in turn relates to organ distribution. Here, we discuss the role of intrinsic master transcription factors and extrinsic IL-15 signaling on NK cell-mediated immune protection against murine pulmonary metastasis.

One of the most intriguing questions in cancer immunology is how cancer cells escape immunosurveillance. Considering the course of most malignancies, cancer cells escape immune regulation at least twice. The first occurrence is when healthy cells transform into malignant cells without being destroyed immediately, and the second takes place when some of these cancer cells leave their origin to successfully colonize distant organs.

Natural killer (NK) cells are considered potent sentinels for cancer, and their relevance has been confirmed in several studies. In experimental models, NK cells protect mice against tumors arising from adoptively transferred cancer cells or tumors induced by chemical carcinogens. In humans, NK cell infiltration into tumor tissue is generally associated with a better clinical outcome, whereas suppressed NK cell activity can be a negative prognostic factor for cancer development or disease progression. However, the NK cell compartment does not consist of a homogeneous cell population, but rather of phenotypically and functionally distinct subsets. We, and others, have recently found that expression of the maturation markers CD27 (a member of the TNF-receptor superfamily) and CD11b, as well as KLRG1 (the killer cell lectin-like receptor subfamily G, member 1), allows for the discrimination of murine NK cell subpopulations with distinct levels of antitumor activity.,

The expression of CD27 and KLRG1 is regulated by cell-intrinsic pathways that are integral to the differentiation status of NK cells. This differentiation process is controlled through expression of the T-box transcription factors T-bet (Tbx21) and Eomes (eomesodermin). Importantly, we observed that T-bet deficient mice lack CD27loKLRG1+ NK cells leading to a loss of protection against pulmonary colonization after tail vein injection of CT26 colorectal cancer cells (Fig. 1). Considering that T-bet also plays an important role in regulating the fate and functions of adaptive immune cells, we next excluded a potential effector role for T and B cells in our model by evaluating T-bet−/− mice on a Rag1−/− background. We found that Rag1−/− mice that possess fully functional CD27loKLRG1+ NK cells but lack T and B cells were equally protected against CT26 pulmonary metastasis as wild-type mice, whereas Rag1−/−T-bet−/− mice showed extensive colonization of tumor cells in the lung. The importance of CD27loKLRG1+ NK cells in the prevention of metastatic disease was further corroborated by adoptive transfer experiments in which a reconstitution of T-bet deficient mice with wild-type CD27loKLRG1+ NK cells could restore antitumor protection (Fig. 1). Interestingly, immunosurveillance in T-bet−/− mice could also be recovered by reconstitution with wild-type CD27hiKLRG1- NK cells that became CD27loKLRG1+ upon adoptive transfer in vivo. Our findings revealing the essential role of T-bet in cancer immunosurveillance is in line with published data by other groups who found that T-bet deficiency is also associated with an augmented tumor-burden in a B16.F10 melanoma model and an increased rate of metastases in the TRAMP prostate cancer model.,

Figure 1. Distribution of natural killer (NK) cell subsets with distinct anticancer properties in a metastatic disease model. Natural killer (NK) cells with tumor-inhibitory activities are CD27loKLRG1+ and dependent upon the T-box transcription factors T-bet and Eomes. CT26 colorectal carcinoma cells were injected intravenously into T-bet−/− mice (center), Balb/c wild-type mice (left), and T-bet−/− mice treated with IL-15 (right) and monitored for metastatic burden and heterogeneous NK cell subsets. Eomes, eomesodermin; KLRG1, killer cell lectin-like receptor subfamily G, member 1; T-bet, T box 21.

To compensate for the immaturity of T-bet deficient NK cells and, in premise, thereby improving their anticancer activity, we treated T-bet deficient animals with recombinant IL-15 that was precomplexed with a fusion protein consisting of the ectodomain of the mouse IL-15-Receptor-α-chain and the Fc domain of human IgG1 (rIL-15/IL-15Rα/Fc). It is known that NK cells are highly responsive to common-γ chain cytokines. For example, NK cells stimulated with trans-presented IL-15 in vivo overexpress the stimulatory receptor killer cell lectin-like receptor K1 (KLRK1, better known as NKG2D), as well as effector molecules (e.g., granzyme B, perforin) that mediate tumor immunosurveillance. Indeed, IL-15-treated T-bet−/− mice were protected from lung colonization by CT26 carcinoma cells with an efficiency similar to that of their T-bet competent littermates (Fig. 1). This effect was accompanied by a rapid in vivo expansion of NK cells as well as upregulation of Eomes and KLRG1. These observations make IL-15 an interesting treatment option for immunotherapies in cancer patients. In fact, the first clinical trials are currently being conducted to evaluate the potency of IL-15 in the clinical setting (e.g., http://www.clinicaltrials.gov: NCT01021059, NCT01369888, NCT01189383, and NCT01337544).

To date, it can be argued that insufficient emphasis has been placed on experiments aimed at finding the causes of very early spread of cancer cells, some of which may progress to metastases that account for most cancer deaths. In this respect NK cell subpopulations may be critical to prevent the development of early metastases due to their anti-tumor effector functions and their differential distribution in specific organs. For instance, CD27high NK cells in mice predominate in bone marrow and lymph nodes, and rapidly respond to IL-15 stimulation as a result of their high expression levels of IL-15 receptor-β. Conversely, CD27low NK cells predominate in extra-lymphoid organs such as lungs and are more tightly regulated (relative to CD27high cells) in terms of effector functions because of their relatively low IL-15 receptor-β levels and high NK inhibitory receptor expression. We are currently exploring whether this NK cell heterogeneity results in distinct tissue-specific NK cell-mediated antitumor responses, thereby impacting the pattern and early onset of metastatic disease.

Further research is clearly required to elucidate the homing, effector functions, and anticancer properties of distinct NK cell subsets, especially with regards to their regulation by the T-box transcription factors T-bet and Eomes as well as exogenous IL-15. We believe that future studies should emphasize the use of spontaneous tumor models so that the role of NK cell subsets in the genesis of metastatic disease can be better understood, perhaps leading to novel therapeutic strategies.