Influence of KIR gene copy number on natural killer cell education

Abstract

Natural killer (NK) cells are functionally tuned by education via killer cell immunoglobulin receptors (KIRs) interacting with HLA class I molecules. We examined the effect of KIR gene copy number variation on the education of human NK cells. The frequency of NK cells expressing a given KIR correlated with the copy number of that gene. However, coexpression of multiple copies from a single locus, or duplicated loci, was infrequent, which is in line with independent transcriptional regulation of each allele or copy. Intriguingly, coexpression of 2 KIR alleles, resulting in higher surface expression, did not lead to enhanced functional responses in vitro or to selective advantages during in vivo responses to cytomegalovirus infection, suggesting that receptor density does not influence NK education at the single cell level. However, individuals with multiple KIR gene copies had higher frequencies of responding cells, consistent with heightened overall responsiveness.

Figure 1

The KIR expression frequency correlates linearly with KIR gene copy number. Frequency of CD56^dim NK cells expressing the indicated inhibitory (A) and activating (B) KIR stratified based on KIR gene copy number. (C) Representative examples of expression of the indicated KIR in donors with distinct copy numbers. (D) Frequency of 2DL1⁺ (top) and 2DL3⁺ (bottom) NK cells within the indicated subsets and stratified based on KIR gene copy number. Red dots represent the donors with statistical outliers identified by the Chauvenet algorithm. **P < .01; ***P < .001.

Figure 2

Influence of KIR gene copy number on NK cell education. (A) Gating strategy to identify single and bi-allelic expression of 2DL3 in donors lacking 2DS1, 2DS2, and 2DL2 genes. First, gates were set on NKG2C⁻NKG2A⁻ NK cells to avoid differences in functionality due to NKG2C expression and/or education by NKG2A. Thereafter, 2DL3⁺2DL1⁻ cells were gated out for further stratification into 2DL3*005 single positive cells (180701⁻EB6⁺), KIR2DL3*xxx single positive cells (180701⁺EB6⁻) and 2DL3*005/2DL3*xxx double positive cells (180701⁺EB6⁺). KIR2DL3*xxx represents all 2DL3 alleles recognized by mAb 180701. All donors were confirmed to be 2DL3*005 at the genetic level (B) Representative histogram and (C) bar chart (n = 9), showing the mean fluorescence intensity (MFI) (^+/− standard deviation) of single and bi-allelic expression of 2DL3 (GL183+). (D) Expression levels (MFI) of 2DL1, 2DL3, and 3DL1 as a function of CNV. (E) Function (CD107a, IFN-γ and tumor necrosis factor-α) of single KIR-positive NK cells in donors stratified based on KIR gene CNV and the presence or absence of the corresponding KIR ligand. (F) One representative example and (G) recapitulative bar chart (mean of 3 donors) showing the allelic distribution in the expanded (NKG2C⁺) and the nonexpanded (NKG2C^-) subset after gating on GL183⁺2DL1⁻NKG2A⁻ CD56^dim NK cells. All donors had the 2DL3*005/2DL3*xxx genotype. (H) Average contribution of the indicated KIR expressing subset to the CD107a response in donors with 1 or 2 copies of the KIR gene and with the cognate HLA-class I ligand. (I) Overall effect of KIR gene CNV on degranulation in the CD56^dim NKG2A⁻NKG2C⁻ NK cell compartment after exclusion of 2DL2/S2⁺ donors. *P < .05; **P < .01; ***P < .001. ns, not significant.