Fig. 6. Effects of NKG2A and KIR3DL1 education apply in transplant donor/recipient pairs.

a Distribution of HLA-E expression in donor stimulator cells. HLA-E expression on donor cells was measured in triplicate by flow cytometry. Median fluorescence intensity was normalized by isotype control. b Stimulator groups were defined by HLA-E expression where 75th percentile was high (n=51) and remaining were low (n=16). Hierarchical clustering shows similar expression profile across HLA-E low donors with exception of two indicated in red in heatmap legend. These two donors expressed higher CD112 and HLA-F; additional analysis excluding these two donors show stronger associations (Supplementary Fig. 5). c Change in percent of XCL1+ and Granzyme B+ (delta between stimulated and PBMC only condition) in recipient NKG2A+NKG2C− NK cells was greater when stimulated by donors with lower HLA-E (n=16) than donors with higher HLA-E (n-51). Two donors with higher CD112/HLA-F (red dots) induced more Granzyme B, XCL1 and CD107a production in NKG2A+NKG2C− NK cells compared to other donors with high HLA-E expression. d Bw4 loss was defined as lower Bw4 copy number in transplant donor compared to recipient. No loss of Bw4 was defined as equal or greater copy number of Bw4 on HLA-A/B alleles in donor compared to recipient or recipient was KIR3DL1–/–. When gating on KIR3DL1+ NK cells, percent of CD107a+ and XCL1+ and degranulation of Ksp37 was higher when there was loss (n=17) than when there was no loss (n=48). P value was calculated using 2-sided unpaired Student’s t-test; nominal p values shown.