Cholinergic regulation of vascular endothelial function by human ChAT+ T cells

Abstract

Endothelial dysfunction and impaired vasodilation are linked with adverse cardiovascular events. T lymphocytes expressing choline acetyltransferase (ChAT), the enzyme catalyzing biosynthesis of the vasorelaxant acetylcholine (ACh), regulate vasodilation and are integral to the cholinergic antiinflammatory pathway in an inflammatory reflex in mice. Here, we found that human T cell ChAT mRNA expression was induced by T cell activation involving the PI3K signaling cascade. Mechanistically, we identified that ChAT mRNA expression was induced following the attenuation of RE-1 Silencing Transcription factor REST-mediated methylation of the ChAT promoter, and that ChAT mRNA expression levels were up-regulated by GATA3 in human T cells. In functional experiments, T cell-derived ACh increased endothelial nitric oxide-synthase activity, promoted vasorelaxation, and reduced vascular endothelial activation and promoted barrier integrity by a cholinergic mechanism. Further, we observed that survival in a cohort of patients with severe circulatory failure correlated with their relative frequency of ChAT ⁺CD4⁺ T cells in blood. These findings on ChAT⁺ human T cells provide a mechanism for cholinergic immune regulation of vascular endothelial function in human inflammation.

Keywords: acetylcholine; circulation; lymphocytes; vascular biology.

Fig. 1.

Activation-induced ChAT in primary human T cells. (A) ChAT mRNA expression following T cell activation in vitro. ChAT mRNA was quantified in activated primary human T cells harvested at 0 to 120 h (n = 3 to 8). Bars show mean ± SEM, normalized to PPIA and B2M, and expressed relative to the ChAT mRNA level at 72 h (Kruskal–Wallis test, post hoc Dunn’s multiple comparisons test). (B) Gates used for the isolation of activated primary human T cells (96 h). (C) ChAT mRNA levels in large and small T cells measured by qPCR in FACS isolated small and large activated (96 h) primary human (n = 3) T cells. Bars show mean ± SEM, normalized to PPIA, and graphed relative to average ChAT mRNA levels in large cells (two-tailed, unpaired Student’s t test). (D) Absolute ChAT quantification in primary activated (96 h) human (n = 54) T cells. Bars show log of the mean amount of ChAT mRNA/ng of total RNA. (E) Correlation between normalized ChAT mRNA levels (qPCR) and ChAT promoter methylation (MeDIP) in primary activated (96 h) human (n = 4) T cells. Circles represent mean ± SEM threshold values of ChAT normalized to PPIA versus mean ± SEM fold enrichment of methylated DNA [Pearson r (r_P) correlation]. (F) ChAT mRNA expression following T cell activation and 5-Aza-2-deoxycytidine (5-AZA) exposure in vitro. ChAT mRNA expression was quantified by qPCR in primary activated (96 h) human T cells exposed to vehicle (DMSO) (n = 24) or 5-AZA (n = 26) for 72 h. Bars show mean ± SEM, dots represent donors. Data are normalized to PPIA and expressed relative to DMSO-exposed cells (two-tailed, unpaired Student’s t test). (G) ChAT mRNA expression quantified by qPCR in primary activated (96 h) human T cells exposed to vehicle (DMSO) (n = 11), PMA (n = 6), or Forskolin (n = 12) for 72 h. Bars show mean ± SEM. Data are normalized to PPIA and expressed relative to DMSO-exposed cells (Kruskal–Wallis test, post hoc Dunn’s multiple comparisons test). (H) ChAT mRNA expression quantified by qPCR in primary activated (96 h) human T cells exposed to vehicle (DMSO) (n = 23), LY294002 (n = 15), U0126 (n = 9), SP600126 (n = 12), SB203580 (n = 14), rapamycin (n = 12), or H89 (n = 10) for 72 h. Bars show mean ± SEM. Data are normalized to PPIA and expressed relative to DMSO-exposed cells (Kruskal–Wallis test, post hoc Dunn’s multiple comparisons test). Outliers are indicated in red. nd–not detected, ns–not significant, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 2.

ChAT gene methylation in human lymphocytes was regulated by the transcription factor REST. (A) REST mRNA expression following T cell activation in vitro. REST mRNA was quantified by qPCR in activated primary human T cells harvested at 0 to 120 h (n = 2 to 7). Bars show mean ± SEM, normalized to PPIA and B2M, and expressed relative to the ChAT mRNA level at 0 h (One-way ANOVA, Dunnett’s multiple comparisons test). (B and C) mRNA levels of REST and ChAT after shRNA-mediated REST knockdown in transfected Jurkat T cells (n = 10 to 12). Bars show mean ± SEM, normalized to PPIA and B2M, and expressed relative to scrambled shRNA (two-tailed, unpaired Student’s t test). (Inset) Western blot of REST protein level following lentiviral transduction of Jurkat cells with scrambled or REST shRNA. Beta-actin was used as loading control. (D) Schematic representation of potential REST-binding motifs (gray box) within the ChAT promoter. (E) Western blot analysis of DAPA of potential REST-binding sites in the ChAT promoter. Nuclear proteins from Jurkat cells transfected with scrambled or shRNA against REST were examined for the five (A–E) potential ChAT promoter-binding sites of REST. Input was used as a loading control. (F) ChIP PCR analysis of REST binding to synapsin (Syn RE1) in Jurkat cells (n = 2), immune-precipitated with either anti-REST or nonspecific IgG antibodies. Promoter fragments containing RE1 element and GAPDH DNA quantified by qPCR. Bars show mean enrichment of fragments precipitated by anti-REST antibody relative to nonspecific IgG ± SEM. Data are normalized to input. (G) ChIP PCR analysis of REST binding to the ChAT promoter following primary human (n = 4) T cell activation (96 h). Cross-linked REST–DNA complex was immunoprecipitated with anti-REST or nonspecific IgG antibodies, and ChAT promoter and GAPDH genomic DNA fragments were quantified by qPCR. Bars show mean enrichment of fragments precipitated by anti-REST antibody relative to nonspecific IgG ± SEM. Data are normalized to input. (H) ChIP PCR analysis of REST binding to the ChAT promoter in Jurkat cells exposed to vehicle (DMSO) (n = 7) or LY294002 (n = 9). Cross-linked REST–DNA complex was immunoprecipitated with anti-REST or nonspecific IgG antibodies and ChAT promoter fragments were quantified by qPCR. GAPDH genomic DNA or 14th exon of ChAT were used as a negative control. Bars show mean enrichment of promoter fragments precipitated by anti-REST antibody relative to nonspecific IgG ± SEM. Data are normalized to input and expressed relative to the DMSO-exposed samples (one-way ANOVA, Uncorrected Fisher’s LSD). (I) ChAT mRNA expression after 5-Aza-2-deoxycytidine (5-AZA) treatment (n = 13 to 15) following lentiviral transduction of Jurkat cells with scrambled or REST shRNA exposed to vehicle (DMSO) or 5-AZA. mRNA was quantified by qPCR. Data are normalized to PPIA/B2M. Bars show percentage ± SEM and expressed as the 5-AZA:DMSO-treated sample ratio. (Two-tailed, unpaired Student’s t test) (J) Methylation ratio of CpG island 1 from the ChAT promoter following knockdown of REST using shRNA, quantified by bisulfite PCR in Jurkat cells exposed to vehicle (DMSO) (Scrambled: n = 8, shNRSF: n = 9) or 5-AZA (Scrambled: n = 9, shNRSF: n = 9). DNA content was amplified by PCR and quantified with ImageJ after DNA electrophoresis. Bars show the ratio of methylated to unmethylated DNA mean ± SEM. Data are expressed relative to the ratio of methylated to unmethylated DNA in scrambled shRNA-expressing cells exposed to DMSO (one-way ANOVA, Holm–Sidak’s multiple comparisons test). nd–not detected, ns–not significant, *P < 0.05, **P < 0.01, **** P < 0.0001.

Fig. 3.

ChAT mRNA expression was regulated by the Th2-associated master regulator GATA3. (A) ChAT mRNA levels in activated CD3⁺ T cells (n = 4), fractionated into CD4-enriched and CD4-depleted populations. Bars show mean ± SEM, normalized to PPIA/B2M and expressed as fold change compared with the same donor CD4-depleted cells (two-tailed, paired Student’s t test). (B) Frequency of ChAT⁺ cells within CD4⁺ cells following FACS isolation for Fluidigm Biomark single-cell (n = 1,631) gene expression profiling. Healthy donor CD4⁺ T cells activated for 4 (n = 3), 10 (n = 3), or 11 (n = 5) days were isolated using FACS and ChAT was measured by qPCR. The boxes indicate the interquartile range, the central bar indicates the median, and the whiskers indicate the 5th to 95th percentile range. Circles show ChAT frequency in individual donors. Correlation of the expression of the genes of interest (GOI) between ChAT⁺CD4⁺ and ChAT^-CD4⁺ T cells: (C) mRNA levels and (D) fraction of cells expressing the GOIs. mRNA levels were measured in activated single ChAT⁺ cells (n = 25) and ChAT^- cells (n = 153). Each dot represents one GOI. Pearson r (r_P), Spearman r (r_S), D’Agostino & Pearson normality test. (E) Correlation of ChAT and GOI-expression within activated single ChAT⁺CD4⁺ T cells (n = 25). Bars represent Pearson r (r_P) correlation coefficients, D’Agostino & Pearson normality test. (F) Frequency of GOI expression within activated single ChAT ⁺CD4⁺ cells. The bars show the percentage of ChAT⁺CD4⁺CD8⁻ cells (n = 19 to 25) that express GOI. (G) Relative ChAT mRNA levels during CD4⁺ T cell polarization. mRNA was isolated at 4, 11, and 14 d, and mRNA levels were measured in Th 1 (n = 2 to 4), Th 2 (n = 4 to 6), and Treg (n = 4) polarizing conditions. Bars show mean ± SEM, normalized to PPIA/B2M, and plotted relative to Th1 population at 4 d (Two-way ANOVA, FDR: two-stage linear step-up procedure of Benjamini, Krieger, and Yekutieli). (H) Correlation of ChAT to GATA3, IL-4, IFNG, and FOXP3 during CD4⁺ T cell polarization and activation. Circles represent threshold values normalized to PPIA/B2M. Spearman r (r_S)D’Agostino & Pearson normality test. (I) Flow cytometry analysis of human-activated T cells (n = 4) following siRNA-mediated GATA3 knockdown. Histogram shows the average signal from four donors following nontargeted or GATA3 siRNA knockdown at 7 d of knockdown. (J) mRNA levels of GATA3, IL-13, and ChAT after siRNA-mediated GATA3 knockdown. Primary human (n = 4) CD3⁺ T cells were activated and transfected with Accell nontargeted siRNA or siRNA targeting GATA3 mRNA for 96 h. Bars show mean ± SEM, normalized to PPIA/B2M, relative to nontargeted siRNA (two-tailed, paired Student’s t test). *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 4.

T cell-derived ACh attenuated vascular endothelial inflammation and promoted barrier integrity. (A) ACh released following activation using anti-CD3/CD28 antibodies of human T cells in vitro. Activated primary T cells (n = 6) harvested at 0 to 120 h were incubated in PBS+pyridostigmine bromide (PBr), and supernatants were analyzed for ACh concentration using mass spectrometry. Bars represent mean ng ACh level ± SEM per 10⁶ T cells (Kruskal–Wallis test, post hoc Dunn’s multiple comparisons test). (B and C) Quantification of immunointensities of (B) nuclear NF-κB p65 and (C) E-selectin in HUVECs following exposure to human CD3⁺ T cell-conditioned media. HUVECs were exposed to vehicle (media+PBr) (n = 4) or activated (48 h) T cell-conditioned media (n = 4), fixed and stained for NF-κB p65, E-selectin. Bars represent fluorescence integrated density as percentage of vehicle control (two-tailed, unpaired Student’s t test). (D–F and J) Quantification of immunointensities of (D) nuclear NF-κB p65, (E) E-selectin, (F and G) paracellular exposure of matrix-bound biotin (representative images in G), (J) 70 kDa dextran-TMR uptake assessing macromolecular pinocytosis in endothelial cells following exposure to human CD3⁺ T cell-conditioned media. HUVECs were exposed to vehicle (media+PBr) (n = 3 to 4) or activated (11 d) CD3⁺ T cell-conditioned media (n = 3), with and without atropine and mecamylamine (MEC). Bars represent fluorescence-integrated density as percentage of vehicle control (One-way ANOVA, Fisher’s LSD). (H and I) Representative images (H) and quantification (I) of immunointensities of the five different categories of VE-cadherin+ adherens junction morphologies in HUVECs. HUVECs were exposed to vehicle (media+PBr) (n = 3) or activated (11 d) T cell-conditioned media (n = 8), with and without atropine and mecamylamine (MEC). The cells were fixed and stained for VE-cadherin. (K) Quantification of 70 kDa dextran-TMR uptake in HUVECs. HUVECs were exposed to vehicle (media+PBr) (n = 3) or activated (11 d) T cell-conditioned media (n = 4), with and without atropine and mecamylamine (MEC), and activated using TNFα and IL-1β. (L) HUVECs were exposed to vehicle (media+PBr) (n = 3) or activated (11 d) T cell-conditioned media (n = 5), with and without atropine and mecamylamine (MEC), and activated using TNFα and IL-1β. The cells were fixed and stained for VE-cadherin. Bars represent integrated fluorescence intensity as percentage of TNFα/IL-1β ± SEM (one-way ANOVA, Fisher’s LSD). ns–not significant. *P ≤ 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 5.

The relative frequency of CD4⁺ ChAT⁺ T cells in blood correlated with survival. (A) Box plots indicating ChAT expression in single-cell transcriptomic analysis of peripheral blood CD4⁺ cells from patients in circulatory failure requiring veno-arterial extra corporeal life support (extracorporal membrane oxygenation, ECMO) (n = 33). Transcript levels are shown as log normalized counts. The boxes indicate the interquartile range, the central bar indicates the median, and the whiskers show the 5th to 95th percentile range. (B) Scatter dot plot indicating the relative frequency of ChAT⁺ cells within CD4⁺ T cells in healthy control individuals (n = 11) and patients in circulatory failure ECMO (n = 33). Each circle indicates an individual patient. nd–not detected. (C) Uniform manifold approximation and projection (UMAP) of transcriptomic analysis of single CD4⁺ cell from patients in circulatory failure ECMO (n = 33). Colors denote cell populations defined using Louvain clustering. (D) UMAP projection of ChAT expression. Expression (Exp.) is presented as log counts of ChAT. (E) Percentage of ChAT⁺ cells in each of the cell populations defined using Louvain clustering identified in C. (F) Violin plot of transcripts per million (TPM) of ChAT expression in each of the cell populations defined using Louvain clustering identified in C. Clusters (x axis) are ordered by percentage of ChAT⁺ cells. (G) Dot plot showing ChAT and the 16 transcripts most positively or negatively correlated with ChAT (r > 0.3, P < 0.05) or (r < −0.3, P < 0.05). Transcripts (x axis) are ordered by r. Clusters (y axis) are ordered by percentage of ChAT⁺ cells. The size of the dot corresponds to the percentage of positive cells in each cluster. The color Expression (Exp.) represents the average transcripts per million reads (TPMs) normalized by mean TPMs in all clusters and adjusted to 100%. (H) Violin plot of TPM of ChAT in single cells positive for at least one or negative for all eight significantly positively correlated genes identified in the correlation analysis in G. (I) The survival probability for patients (n = 32) with low ChAT⁺CD4⁺ T cell frequency (Low Strata, 25th percentile) and high ChAT⁺CD4⁺ T cell frequency (High Strata, 75th percentile) and all other significant variables (inotrope score and lactate) at their median.