Human memory CD8 T cell effector potential is epigenetically preserved during in vivo homeostasis

Abstract

Antigen-independent homeostasis of memory CD8 T cells is vital for sustaining long-lived T cell-mediated immunity. In this study, we report that maintenance of human memory CD8 T cell effector potential during in vitro and in vivo homeostatic proliferation is coupled to preservation of acquired DNA methylation programs. Whole-genome bisulfite sequencing of primary human naive, short-lived effector memory (T_EM), and longer-lived central memory (T_CM) and stem cell memory (T_SCM) CD8 T cells identified effector molecules with demethylated promoters and poised for expression. Effector-loci demethylation was heritably preserved during IL-7- and IL-15-mediated in vitro cell proliferation. Conversely, cytokine-driven proliferation of T_CM and T_SCM memory cells resulted in phenotypic conversion into T_EM cells and was coupled to increased methylation of the CCR7 and Tcf7 loci. Furthermore, haploidentical donor memory CD8 T cells undergoing in vivo proliferation in lymphodepleted recipients also maintained their effector-associated demethylated status but acquired T_EM-associated programs. These data demonstrate that effector-associated epigenetic programs are preserved during cytokine-driven subset interconversion of human memory CD8 T cells.

Figure 1.

Genome-wide changes in DNA-methylation programming are coupled to human memory CD8 T cell subset–specific differentiation. (A) Flow cytometry–based strategy for isolating a sufficient quantity of naive and memory CD8 T cell subsets from apheresis blood unit of healthy donors for phenotypic, functional, and whole-genome epigenetic characterization. The cell subsets were identified based on the expression of three cell surface markers as follows: naive: CCR7⁺, CD45RO⁻, and CD95⁻; T_EM: CCR7⁻ and CD45RO⁺; T_CM: CCR7⁺ and CD45RO⁺; T_SCM: CCR7⁺, CD45RO⁻, and CD95⁺. (B) Circos plot showing DNA methylation levels for naive and memory CD8 T cell subsets across the whole genome from two donors. CpG methylation levels were averaged over 10-Mb genomic intervals and are represented as histogram tracks. Heat map shows changes in the levels of DNA methylation with respect to the sample across the entire genome. Dark red indicates high levels of methylation, and light red indicates low levels of methylation. Dotted black circle separates two methylation patterns: dark red (high methylation, as in naive and T_SCM) and light red (low methylation, as in T_CM and T_EM). (C) PCA of methylation status of total CpGs with more than five times the coverage. (D) Summary graph of the number of methylated and demethylated regions in the T_EM, T_CM, and T_SCM genomes relative to that in the naive CD8⁺ T cell genome. The number of demethylated regions was calculated based on difference ≥30% methylation between two populations. The number of methylated regions was calculated based on ≤30% methylation difference between the two populations. (E) Pie charts showing the percentage of demethylated and methylated regions across the genomes of T_EM, T_CM, and T_SCM cells relative to that of naive CD8 T cells. (F) Normalized plot of CpG methylation at sites surrounding and within DMRs of effector molecules (IFNγ, PRF1, GZMB, and GZMK), tissue-homing molecules (CD62L and CCR7), and transcription factors (T-BET and EOMES) obtained from WGBS analysis. Red and blue lines depict methylation and demethylation of CpG sites, respectively.

Figure 2.

Stable maintenance of poised effector programs during in vitro cytokine-driven proliferation of human memory CD8 T cell subsets. (A, left) Recall response of undivided and divided, CFSE-labeled, naive, T_EM, T_CM, and T_SCM CD8 T cells expressing IFNγ during exposure to IL-7/IL-15 in culture for 7 d, followed by anti-CD3/CD28 stimulation (1:1 ratio) for 4 h. Gates indicate the percentage of undivided and divided cells. (right) Bar graph showing cumulative data from four independent experiments presented as the percentage means ± SEM. CD8 T cells expressing IFNγ during homeostatic proliferation after TCR stimulation (n = 4). (B) Freshly isolated CD8 T cell subsets were labeled with CFSE and subsequently maintained in culture in the presence of IL-7/IL-15 for 7 d. Undivided and divided cell subpopulations were then sorted, and genomic DNA was extracted for bisulfite sequencing analysis. (C, top) Representative bisulfite sequencing analysis of effector molecules in undivided and divided cells for each CD8 T cell subset. (bottom) Bar graph showing the percentage of CpG methylation (means ± SEM) at each site of the effector loci in undivided and divided, naive, T_EM, T_CM, and T_SCM cells (n = 4 healthy donors). Mann-Whitney U test was used. P < 0.05 was considered significant. NS, not significant. Statistical comparison was based on the mean value of all CpG sites.

Figure 3.

Plasticity of tissue-homing programs during homeostatic proliferation of human memory CD8 T cell subsets. (A, top) Representative DNA methylation profile analysis of CCR7 and SELL DMRs from ex vivo isolated CD8 T cell subsets. Each horizontal line represents a clone, and each vertical line represents a CpG site. (bottom) Bar graphs showing the percentage of CpG methylation (means ± SEM) for each site (n = 4–5 independently sorted and analyzed healthy donor samples). Mann-Whitney U test was used. *, P < 0.03; and **, P < 0.01 were considered significant. NS, not significant. Statistical comparison was based on the mean value of all CpG sites. (B) CCR7 expression in undivided and divided, CFSE-labeled, naive, T_EM, T_CM, and T_SCM CD8⁺ T cells after exposure to IL-7/IL-15 in culture for 7 d. Gates indicate the percentage of undivided and divided cells (n = 4). (C) Paired analysis for CCR7 expression in undivided and divided populations of naive, T_EM, T_CM, and T_SCM CD8⁺ T cells after culture in IL-7/IL-15 for 7 d (n = 4 independently sorted and analyzed healthy donor samples). Paired Student’s t test was used. **, P < 0.01 was considered significant. Error bars indicate SEM. (D, top) Representative bisulfite sequencing analysis of tissue-homing loci in undivided and divided cells for the indicated CD8 T cell subsets. (bottom) Bar graphs showing the percentage of CpG methylation (means ± SEM) for each site of the tissue-homing loci in undivided and divided, naive, T_EM, T_CM, and T_SCM cells (n = 5 independently sorted and analyzed healthy donor samples). Mann-Whitney U test was used. *, P < 0.03 was considered significant. NS, not significant. Statistical comparison was based on the mean value of all CpG sites.

Figure 4.

Tcf1 expression is down-regulated during memory CD8 T cell in vitro homeostatic proliferation. (A) Histogram showing Tcf1 expression levels in sorted, human, naive, T_EM, T_CM, and T_SCM. (B, left) Representative bisulfite sequencing analysis of TCF7 DMR in ex vivo isolated CD8 T cell subsets from one representative donor. Each horizontal line represents a clone, and each vertical line represents a CpG site. (right) Bar graph showing the mean percentage of CpG methylation (means ± SEM) for each site (n = 3 healthy donors) *, P < 0.05; and ****, P < 0.0001 were considered significant. Unpaired Student’s t test was used. Statistical comparison was based on the mean value of all CpG sites. (C) Mean fluorescence intensity (MFI) of Tcf1 in CCR7^lo and CCR7^hi naive and memory CD8 T cell subsets after 7 d of IL-7/IL-15 culture (n = 4 healthy donors). Mann-Whitney U test was used. *, P < 0.03 was considered significant. NS, not significant. Error bars indicate SEM. (D) Representative bisulfite sequencing analysis of TCF7 in undivided and divided cells for the indicated CD8 T cell subsets.

Figure 5.

In vivo stability of effector-associated programs in memory CD8 T cells from immunocompromised patients after donor lymphocyte infusion. (A) Schematic representation for the donor lymphocyte infusion protocol showing infusion of CD45RO⁺ memory T cells from healthy, haploidentical donors into bone marrow transplant (BMT) recipient. Representative FACS plots show the frequency of naive, T_EM, and T_CM CD8 T cells in PBMCs from a healthy, haploidentical donor and a patient on day 32 after transplant. (B, top) Bar graph showing the percentage of T_EM cells among CD45RO⁺ CD8 T cells (means ± SEM) in donor and bone marrow transplant patient (n = 5). (bottom) Ki-67 staining for T_EM CD8 T cells from a healthy, haploidentical donor and a patient on day 53 after transplant. (C) Mean fluorescence intensity (MFI) of PD-1 in naive and T_EM CD8 T cells from a healthy, haploidentical donor versus T_EM from a bone marrow transplant patient (n = 4). Error bars indicate SEM. *, P < 0.02 was considered significant. (D) Representative FACS plots showing pre- and postsort purity for T_EM and T_CM in haploidentical donor and transplant patient. (E, top) Representative bisulfite sequencing analysis of effector-associated loci in donor naive and T_EM CD8 T cells and BMT patient T_EM CD8 T cells. (bottom) Bar graph showing the percentage of CpG methylation (means ± SEM) for each site of tissue homing–associated loci in naive and T_EM CD8 T cells from both donor and patient. (F, top) Representative bisulfite sequencing analysis of tissue homing–associated loci in donor naive and T_EM CD8 T cells and patient T_EM CD8 T cells. (bottom) Bar graph showing the percentage of CpG methylation (means ± SEM) for each site of tissue homing–associated loci in naive and T_EM CD8 T cells from both donor and patient (n = 4 healthy donors and recipients). Mann-Whitney U test was used.