Natural killer cells : Tumour Cells tout trogocytosis

A new report in The EMBO Journal describes the transfer of HLA-G molecules by trogocytosis from tumour cells to activated natural killer (NK) cells, which might be a mechanism of protection for HLA-G-negative tumour cells that are in the vicinity of HLA-G-positive tumour cells.

Trogocytosis is the rapid transfer of intact cell-surface proteins between cells in contact with each other, and has been observed between NK cells and tumour cells. But why this process occurs between these two cell types has been unclear. In this study, Caumartin et al. examined HLA-G, a non-classical MHC class I molecule that is expressed by various tumour cells in vivo and for which engagement with inhibitory receptors on NK cells is known to block NK-cell-mediated cytolysis. They asked whether HLA-G could be transferred from tumour cells to NK cells and, if so, what the functional consequences might be.

Cell-culture experiments showed that activated but not resting NK cells display HLA-G1 on their cell surfaces following contact with HLA-G1-expressing tumour cells. But did the NK cells become HLA-G1-positive by trogocytosis? The authors first ruled out the possibilities of endogenous expression of HLA-G1 by the NK cells and binding of shed soluble HLA-G1 to receptors on NK cells, and then showed that cell–cell contact was necessary for NK-cell surface display of HLA-G1. They confirmed that trogocytosis was indeed the mechanism involved by demonstrating the fast kinetics of the transfer of HLA-G1 molecules from the tumour cells to the NK cells, and the limited lifetime of the transferred molecules, both of which, along with the dependence on cell–cell contact, are the main parameters of trogocytosis.

Caumartin et al. next investigated whether trogocytosis of HLA-G1 had functional consequences for NK cells and, in particular, whether it could constitute an immune-escape mechanism for tumour cells, given that HLA-G1 engages inhibitory receptors on NK cells resulting in immunosuppression. The authors showed that the acquisition of HLA-G1 by activated, cytotoxic NK effector cells blocked their proliferation and temporarily inhibited their cytolytic function. Further investigation revealed that the transfer of HLA-G1 molecules from tumour cells to NK cells in fact induced a temporary immunosuppressive state in the NK-cell population (that is, NK cells could cross-inhibit the cytotoxic functions of other NK cells), and that this was due to the interaction of trogocytosed HLA-G1 with the inhibitory receptor immunoglobulin-like transcript 2 (ILT2) on other NK cells.

So, once the activated NK cells acquire HLA-G1 from tumour cells by trogocytosis they stop proliferating, are no longer cytolytic and behave as suppressor cells, and so can potentially protect tumour cells from NK-cell-mediated cytolysis. The unique qualities of this process are that it effects a switch from effector to suppressor cell through the transfer of a molecule that the recipient cell does not endogenously express, and that it is temporary and does not alter the true nature of the cell. These data highlight the key role of the cellular microenvironment in the immune response