Over-expression of CD200 in Acute Myeloid Leukemia Mediates the Expansion of Regulatory T-lymphocytes and Directly Inhibits Natural Killer Cell Tumor Immunity

CD200 is a type-1 transmembrane glycoprotein which suppresses inflammatory and autoimmune responses by signalling through its cognate transmembrane receptor homologue (CD200R). Normally, CD200 expression is restricted to immune privileged sites where it enhances immune tolerance through mechanisms that include modulating the expansion of FOXP3+ regulatory T-lymphocytes (T-regs) and suppressing macrophage cytolytic activity. Furthermore, leukocyte associated CD200 has been reported to suppress Natural Killer (NK) cell activity in vivo. Pathologically, we have previously shown that CD200 over-expression on leukemic blasts in around 50% of acute myeloid leukemia (AML) patients is significantly associated with a poor overall survival (Tonks et al, Leukemia, 2007). Given the existing evidence that T-reg frequency and NK cell function influence blast clearance and long-term survival in AML, we investigated the possibility that CD200 expression in AML may be directly suppressing anti-tumor immunity in this disease. Here we present evidence that CD200+ AML can suppress host anti-tumor responses by augmenting the frequency of AML patient T-regs and by direct inhibition of NK cell anti-tumor activity. We also show that targeting the interaction between CD200 and its receptor might provide a new strategy for the treatment of AML. Bone marrow aspirates from 91 diagnostic AML patients were analysed by multiparameter flow cytometry for blast CD200 protein expression. We found that the level of blast CD200 expression directly correlated with an increased frequency of T-regs (CD4+CD25++FoxP3+; R=0.78, p=0.0008). Measuring 3H-thymidine incorporation, we show that T-regs isolated from AML patients by MACS® separation inhibited T-cell proliferation (induced by CD3 and CD28 stimulation) at ratios <0.1%, thus confirming that patients T-regs were functional. In contrast to T-regs, NK cell frequency (CD45+CD19−CD3−CD56+) did not correlate with the level of AML blast CD200 expression (R=0.15, p=0.851), however, NK cell subpopulation bivariate analysis using CD56 and CD16 demonstrated that the CD56dimCD16+ (the principle active NK population) was significantly reduced by over 50% in CD200+ AML patients (36±5% compared to 15±5%, p=0.009). Furthermore, CD200 expression on target cells appeared to have a direct effect on the cytotoxic activity of NK cells; co-culture of NK cells with CD200+ targets resulted in decreased CD107a expression (a marker for cytolytic granules) in NK cells (23±4% vs 12±5%, p=0.038) and decreased apoptosis of the target cells (19±1% vs 10±1%, p=0.041). Since CD200R was detected on NK cells in AML patients, it was likely that CD200 was having a direct effect on suppression of NK cytotoxicity. This was supported by the significant recovery of NK cytolytic activity against CD200+ blasts in the presence of a CD200 blocking antibody (5±1% vs 11±2% CD107a+ NK cells, p=0.046) whereas there was no change seen with CD200− blasts (19±4% vs 19±3%). In conclusion, these findings suggest that CD200 expression on leukemic blasts plays an influential role in suppressing anti-tumor immunity in AML patients through modulating the expansion of functionally suppressive T-regs and directly suppressing NK cell cytolytic activity. In this study blocking CD200 interaction with its receptor was able to recover a significant proportion of patient NK activity, making CD200 a potential therapeutic target for CD200+ AML.