Mapping and Characterization of HCMV-Specific Unconventional HLA-E-Restricted CD8 T Cell Populations and Associated NK and T Cell Responses Using HLA/Peptide Tetramers and Spectral Flow Cytometry

Abstract

HCMV drives complex and multiple cellular immune responses, which causes a persistent immune imprint in hosts. This study aimed to achieve both a quantitative determination of the frequency for various anti-HCMV immune cell subsets, including CD8 T, γδT, NK cells, and a qualitative analysis of their phenotype. To map the various anti-HCMV cellular responses, we used a combination of three HLA_peptide tetramer complexes (HLA-E_VMAPRTLIL, HLA-E_VMAPRSLLL, and HLA-A2_NLVPMVATV) and antibodies for 18 surface markers (CD3, CD4, CD8, CD16, CD19, CD45RA, CD56, CD57, CD158, NKG2A, NKG2C, CCR7, TCRγδ, TCRγδ2, CX3CR1, KLRG1, 2B4, and PD-1) in a 20-color spectral flow cytometry analysis. This immunostaining protocol was applied to PBMCs isolated from HCMV^- and HCMV⁺ individuals. Our workflow allows the efficient determination of events featuring HCMV infection such as CD4/CD8 ratio, CD8 inflation and differentiation, HCMV peptide-specific HLA-E_UL40 and HLA-A2_pp65CD8 T cells, and expansion of γδT and NK subsets including δ2^-γT and memory-like NKG2C⁺CD57⁺ NK cells. Each subset can be further characterized by the expression of 2B4, PD-1, KLRG1, CD45RA, CCR7, CD158, and NKG2A to achieve a fine-tuned mapping of HCMV immune responses. This assay should be useful for the analysis and monitoring of T-and NK cell responses to HCMV infection or vaccines.

Keywords: CD8 T cells; HCMV; HLA-E; NK; pHLA tetramers; spectral flow cytometry; γδT.

Figure 1

(A) A representative nested gating strategy illustrating lymphocyte population subgated by the expression of CD3 and γδTCR. Cells were gated first on FSC-A vs. FSC-H plots and then on SSC-H vs. SSC-A plots to eliminate doublets. Lymphocytes were gated on an SSC-A vs. FSC-A dot plot. Viable lymphocytes were selected using fixable viability stain (FVS) 440UV staining. Lymphocytes were subgated using CD3 and γδTCR staining. (B,C) Analysis of CD4 and CD8 lymphocyte subsets. A representative density plot showing CD4 and CD8 staining among the CD3⁺ γδTCR- lymphocytes. (C) A graphical and statistical analysis of CD4 and CD8 lymphocyte subsets from independent HCMV⁻ (n = 4) and HCMV⁺ (n = 4) individuals. (D,E) Detection and quantification of peptide-specific conventional and non-conventional CD8 T cell populations using HLA class I /peptide tetramers (pHLA). Representative detection of anti-HCMV peptide-specific, conventional (HLA-A2_pp65, peptide NLVPMVATV) and two unconventional (HLA-E_UL40, peptides VMAPRSLLL and VMAPRTLIL), CD8 T cells from three HCMV⁺ hosts is shown. The percentages of tetramer positive cells (tet+) among total CD8 T cells are indicated. FSC-H: Forward scatter height. FSC-A: Forward scatter area. SSC: Side scatter. TCR: T cell receptor. p value: * for p < 0.05.

Figure 2

Analysis of CD8 T cell differentiation. (A) Representative density plots showing the sequential gating of CD8 T cells negatively (tet-) and positively (tet+) stained by pMHC class I tetramer (HLA-E_{UL40 VMAPTSLLL,}upper panel or HLA-A2_pp65, lower panel) after a selection according to CD4 and CD8 costaining among the CD3⁺ TCRγδ⁻ lymphocytes. The expression pattern for CD45RA/CCR7 of each population (tet⁻ and tet₊) is shown. (B) A schematic representation of CD8 differentiation stages including CD8 naive T cells (TN), central memory T cells (TCM), effector memory T cells (TEM), and terminally differentiated T cells (TEMRA) according to CD45RA and CCR7 expression is indicated (upper panel). A graphical and statistical analysis of differentiation for total CD8 T cell pool from independent HCMV⁻ (n = 4) and HCMV⁺ (n = 4) individuals. Data are expressed as box plot with median and interquartile values. (C) Immunophenotyping of CD8 T cells for receptors for T cell activation and inhibition (2B4, PD-1, CD158, NKG2A, NKG2C, KLRG1), migration (CX3CR1) and cytotoxic capacity (CD56, CD57). The coexpression of receptors are shown for the 4 differentiation states (TN, TCM, TEM, TEMRA) analyzed for total CD8 T cells and representative of a single HCMV+ individual. p value: * for p < 0.05.

Figure 3

Analysis of γδT and Vδ2⁻γδT cells. (A) Representative density plots from a single HCMV⁻(upper panel) and HCMV⁺ (lower panel) individuals showing the sequential selection of CD3⁺ γδT cells using CD3 and γδTCR costaining and of Vδ2⁻γδT and Vδ2⁺γδT subsets using Vδ2γδT and CD8 costaining used to quantify the γδT cell subsets. Immunophenotyping of γδT cell subsets showing costaining for CD45RA/CCR7, CX3CR1/KLRG1 and CD56/CD16. The coexpression of immune receptors for Vδ2⁻γδT vs. Vδ2⁺γδT cell populations are shown. For comparison, cell frequency (%) is indicated for some costainings. (B,C) Box plots with median and interquartile values were used to represent the percentages of (B) CD3⁺ γδ⁺T among lymphocytes and (C) Vδ2⁻γδT vs. Vδ2⁺γδT cells obtained from independent HCMV⁻ (n = 4) and HCMV⁺ (n = 4) individuals. p value: * for p < 0.05.

Figure 4

Analysis of HCMV-induced NK cell subsets. (A,B) Representative density plots showing the sequential selection (upper panel, left to right) CD3⁻ γδ^-T cells, CD19⁻ cells using CD19/CD56 costaining, and CD56/CD16 costaining, enabling the determination of 4 major NK cell subsets. Co-expression for CD57/NKG2C is shown for all NK (upper panel) and for each NK subset (lower panel). The data shown are from one representative HCMV⁻ HV (A) and HCMV⁺ kidney transplant recipient (B). For comparison, cell frequency (%) is indicated for some costainings. Quantitative analyses are represented as box plots comparing the percentages of total NK cells (C), the percentages of NK subsets (D), and the percentages of NKG2C⁺ NK cells (E) for HCMV⁻ (n = 4) and HCMV⁺ (n = 4) individuals. p value: * for p < 0.05.