Reciprocal age related change in natural killer cell receptors for MHC class I

Abstract

Natural killer (NK) cells are essential for healthy aging. NK cell activation is controlled by MHC class I-specific CD94/NKG2 receptors and killer immunoglobulin-like receptors (KIR). To assess NK cytotoxic function in isolation from MHC receptor engagement, we measured the ability of purified NK cells to kill mouse P815 target cells in the presence of anti-CD16 mAb. CD16-mediated cytotoxicity did not change with age, indicating that NK activation and cytotoxic granule release remained functional. We then investigated MHC class I receptor expression on NK cells. There was an age related decrease in CD94 and NKG2A expression and a reciprocal age related increase in KIR expression. NKG2A expression also declined with age on CD56(+) T cells. CD94/NKG2A receptor function was proportional to expression, indicating that NK cell inhibitory signaling pathways were intact. NKG2A and KIR expression were complementary, suggesting that CD94/NKG2A function could substitute for inhibitory KIR function during polyclonal NK cell development in both young and elderly adults. The distinct roles of CD94/NKG2A and KIR receptors suggest that shifting MHC class I receptor expression patterns reflect age related changes in NK cell and CD56(+) T cell turnover and function in vivo.

Fig. 1

Flow cytometry gating strategy. Peripheral blood mononuclear cells from a young subject (top) and an elderly subject (bottom) were stained with Cy5 PE labeled anti-CD14 and anti-CD19, FITC labeled anti-CD3, APC labeled anti-CD56, and PE labeled anti-NKG2A mAb as described in Section 2. Left panel: After CD14⁺ monocytes and CD19⁺ B cells were excluded, CD3 and CD56 staining on a logarithmic scale revealed four major T and NK related populations. Right panel: NKG2A expression on CD3⁻CD56⁺ conventional NK cells. The bar indicates staining that is above background.

Fig. 2

The proportion of NK, T, and related cells in peripheral blood. Major T, NK, and related populations were quantified as illustrated in Fig. 1. Compared with young adult subjects (inner ring), elderly subjects (outer ring) had lower percentages of CD3⁺CD56⁻ conventional T cells (T, p ≤ 0.0001) and CD3⁺ total T cells (T and NKT, p ≤ 0.0006) and a higher percentage of CD3⁻CD14⁻CD19⁻CD56⁻ quadruple negative (QN, p ≤ 0.008) cells. The age differences were significant whether or not subject 12 was included in the analysis and the most conservative of the p-values are shown. Elderly subjects also had a higher percentage of CD3⁻CD56⁺ conventional NK cells (p = 0.052 with subject 12 included and p = 0.033 with subject 12 excluded).

Fig. 3

NK CD16-mediated cytotoxicity did not vary with age or gender. Enriched NK cells were incubated with 3G8 anti-CD16 mAb and ⁵¹Cr labeled mouse P815 target cells. Shown are the cytotoxicity values expressed as lytic units for each subject and the mean value without adjustment. One elderly male was not tested due to an insufficient purified NK cell yield.

Fig. 4

NK cell NKG2A expression decreased and KIR expression increased with age. NK cells were gated and the percentage of receptor positive cells was quantified as illustrated in Fig. 1. Shown are individual and mean age group values without gender correction. The young (Y) and elderly (E) subjects differed significantly in the percentage of conventional CD3⁻CD56⁺ NK cells that expressed cell surface NKG2A (A) and KIR (B). Differences were significant with or without correction for gender interaction.

Fig. 5

Correlation between MHC receptor expression and mAb-mediated NK cytotoxicity. For all figures the putative linear relationship was tested using gender-corrected values, but uncorrected values are shown for individual young females (YF), young males (YM), elderly females (EF), and elderly males (EM). Values for subject 12 are shown as diamonds. (A) The percentage NK cells expressing NKG2A correlated linearly with anti-NKG2A mAb inhibition of CD16-mediated cytotoxicity. Cytotoxicity of ⁵¹Cr labeled mouse P815 cells by purified NK cells was measured in the presence of 3G8 anti-CD16 mAb in combination with either control IgG mAb or Z199 anti-NKG2A mAb. Anti-NKG2A inhibition of CD16-mediated cytotoxicity was calculated in comparison to cytotoxicity of anti-CD16 and control IgG mAb. (B) The percentage NK cells expressing NKG2A correlated inversely with the percentage NK cells expressing KIR. (C) The percentage NK cells expressing KIR did not correlate with KIR-mediated cytotoxicity. Cytotoxicity of ⁵¹Cr labeled mouse P815 cells by purified NK cells was measured in the absence of mAb or in the presence of anti-KIR mAb cocktail, EB6, GL183, and 5.133. KIR stimulation is expressed as lytic units. (D) The percentage NK cells expressing NKG2A correlated linearly with KIR-mediated cytotoxicity, measured as in (C).

Fig. 6

CD94 and NKG2A expression showed age and gender interactions on CD56⁺ T cells. CD56⁺CD3⁺ cells were gated and anti-receptor mAb staining was quantified as illustrated in Fig. 1. Shown are individual and mean age group values without correction for age or gender interaction. The statistically significant differences were based on correction for age and gender interaction. (A) Young males (YM) had significantly higher percentage of CD56⁺ T cells that expressed CD94. (B) Young (Y) and elderly (E) adult subjects differed significantly in the percentage of CD56⁺ T cells that expressed cell surface NKG2A.