HLA-G and CD86 Expression on Malignant Plasma Cells Convey a Poor Prognosis and Immunoregulate T Cells In Patients with Myeloma

Abstract 983

HLA-G is a “non-classical” HLA class I molecule which is considered to be responsible for the immune tolerance of the fetus, allografts or tumors by inhibiting the antigen-specific cytotoxic T lymphocyte response and decreasing NK cell function. CD86 is a costimulatory molecule which binds to a T cell counter receptor (eg CD28) during antigen stimulation and T cell receptor engagement. T cells do not express CD86mRNA but can acquire CD86 and HLA-G by trogocytosis. We have studied the incidence and prognostic significance of HLA-G and CD86 on malignant plasma cells, the incidence of HLA-G and CD86 on T cells and the immunosuppressive function of HLA-G on malignant plasma cells and T cells. HLA-G and CD86 expression were determined on T cells (CD3+) and plasma cells (CD38++) of patients with myeloma (MM). Flow-sorted CD3+ HLA-G- cells were labeled with carboxyfluorescein succinimidyl ester (CFSE) and stimulated with anti-CD3/CD2/CD28 beads in the presence and absence of either HLA-G+ T cells or HLA-G+ plasma cells. Suppression of the proliferation of CFSE-tracked HLA-G- T cells by HLA-G+ T cells or plasma cells was determined by flow cytometry. CD86+ plasma cells are present in 54% of patients and are associated with a poor prognosis (χ²=4.6;p<0.03) (Blood 96:1274). We now report that the incidence of HLA-G expression on plasma cells was heterogeneous (0.2 to 96%; mean = 21.3%) and that the clinical relevance of HLA-G+ plasma cells was demonstrated by a significant reduction in overall survival (13 months vs median not achieved) for the 11/46 patients with HLA-G+ plasma cells (>5% CD38++ cells and expressing >12% HLA-G; χ2=12.4; p<0.0004) who had no difference in β₂M, age or M protein isotype. Biotinylated cell membrane protein transfer from plasma cells to T cells (trogocytosis) was demonstrated using flow cytometry and confocal microscopy. Trogocytosis was unidirectional, common in MM but not from CLL or WM cells and there was with a significantly greater acquisition of biotinylated proteins by CD3+ cells (mean=13.6%) than B cells (mean=2.4%; t=2.80; p<0.05) or NK cells (mean=3.1%; t=2.57; p<0.05). Low HLA-G expression was found on both CD4 and CD8 normal T cells (n=15; mean=0.19%). Twenty six % of MM patients had HLA-G-expressing CD3+ T cells above the normal range (χ²=4.9;p<0.03). There was a significant acquired expression of CD86 by T cells (t=6.06; p<0.004) after T cell activation (CD3/2/28 beads) and exposure to the CD86+ cell line RPMI8226. Trogocytosis of HLA-G has been reported by others (Blood 109:2040). Increased CD86 surface (but not mRNA) expression was found on the CD3+ cells of 32% of patients (n=98; F=38.6; p<0.0001). Flow-sorted HLA-G+ T cells significantly reduced the proliferation of CFSE-labeled T cells in 4 day cultures. Inhibition by HLA-G T cells was greater than the inhibition due to HLA-G+ plasma cells (t=3.03; p=0.039) and was neutralised by anti HLA-G. HLA-G+ T cells were both CD4 and CD8 but were CD25- and thus were not Treg cells. In conclusion, HLA-G and CD86 expression on plasma cells and T cells are clinically relevant in patients with myeloma. Tumor escape in patients with myeloma may be due to the inhibition of normal T cell proliferation by distinct T cell cohorts which develop novel immunoregulatory functions following the acquisition and ectopic expression of antigens which include HLA-G and CD86.