Relatively large numbers of nonimmune spleen cells do not protect against the local growth of two lymphomas. However, this heterogeneous population of splenic lymphocytes contains a subset of cells that efficiently protects against in vivo tumor growth. This cell population (cell-surface phenotype Thyl.2(-)Ig(-)Ly5.1(+)) represents less than 5 percent of the spleen cell population and is responsible for in vitro NK-mediated lysis.
Although these studies clearly and directly demonstrate that Ly5(+) NK cells selected from a heterogeneous lymphoid population from nonimmune mice can protect syngeneic mice against local in vivo growth of two different types of tumor cells (in contrast to other lymphocyte sets within the spleen), they do not directly bear upon the role of NK cells in immunosurveillance. They do indicate that highly enriched Ig(-)Thyl(-)Ly5(+) cells, which account for virtually all in vitro NK activity, can retard tumor growth in vivo.
It is difficult to ascribe all anti-tumor surveillance activity to NK cells, because they probably do not recirculate freely throughout the various organ systems of the body. Perhaps NK ceils may play a role in prevention of neoplastic growth within discrete anatomic compartments where there is rapid differentiation of stem cells to mature progeny (e.g., bone marrow, spleen, and portions of the gastrointestinal tract)and may normally act to regulate the growth and differentiation of non-neoplastic stem cells. Long-term observation of chimeric mice repopulated with bone marrow from congenic or mutant donors expressing very low or very high NK activity may help to answer these questions.