Neurotransmitter signalling via NMDA receptors leads to decreased T helper type 1-like and enhanced T helper type 2-like immune balance in humans

Abstract

Given the pivotal roles that CD4⁺ T cell imbalance plays in human immune disorders, much interest centres on better understanding influences that regulate human helper T-cell subset dominance in vivo. Here, using primary CD4⁺ T cells and short-term T helper type 1 (Th1) and Th2-like lines, we investigated roles and mechanisms by which neurotransmitter receptors may influence human type 1 versus type 2 immunity. We hypothesized that N-methyl-d-aspartate receptors (NMDA-R), which play key roles in memory and learning, can also regulate human CD4⁺ T cell function through induction of excitotoxicity. Fresh primary CD4⁺ T cells from healthy donors express functional NMDA-R that are strongly up-regulated upon T cell receptor (TCR) mediated activation. Synthetic and physiological NMDA-R agonists elicited Ca²⁺ flux and led to marked inhibition of type 1 but not type 2 or interleukin-10 cytokine responses. Among CD4⁺ lines, NMDA and quinolinic acid preferentially reduced cytokine production, Ca²⁺ flux, proliferation and survival of Th1-like cells through increased induction of cell death whereas Th2-like cells were largely spared. Collectively, the findings demonstrate that (i) NMDA-R is rapidly up-regulated upon CD4⁺ T cell activation in humans and (ii) Th1 versus Th2 cell functions such as proliferation, cytokine production and cell survival are differentially affected by NMDA-R agonists. Differential cytokine production and proliferative capacity of Th1 versus Th2 cells is attributable in part to increased physiological cell death among fully committed Th1 versus Th2 cells, leading to increased Th2-like dominance. Hence, excitotoxicity, beyond its roles in neuronal plasticity, may contribute to ongoing modulation of human T cell responses.

Keywords: NMDA; CD4+ T cells; NMDA-R; Th1, Th2; glutamate; glutamic acid; human; immune regulation.

Figure 1

Anti‐CD3/CD28 activation induces elevated N‐methyl‐d‐aspartate receptor 1 (NMDA‐R1) expression by peripheral CD4⁺ T cells. (a) Scheme of kynurenine pathway in tryptophan catabolism. (b) Freshly isolated CD4⁺ T cells, activated with anti‐CD3/CD28 antibody for the indicated times, were examined for NMDA‐R1 expression by quantitative PCR (n = 9 for detailed kinetics) with medians (Mann–Whitney P‐values) shown. (c) Flow cytometric analysis of NMDA‐R1 protein expression on fresh CD3⁺/CD4⁺ primary T‐cell populations that were resting or stimulated with anti‐CD3/CD28 for 24 hr. Data are from one individual, representative of eight individuals examined. (d) Confocal laser‐scanning immunofluorescence microscopy of NMDA‐R1 expression at 24 hr on anti‐CD3/CD28 activated primary CD4⁺ T cells. NMDA‐R1 (green); DAPI (blue, right). Data are from one individual, representative of three examined.

Figure 2

Specificity and functional activity of N‐methyl‐d‐aspartate receptor (NMDA‐R) expression on CD4⁺ T cells. Dose‐dependent intracellular Ca²⁺ influx by enriched primary CD4⁺ T cells following (a) NMDA stimulation or (b) glutamate stimulation (200 μm) alone and in the presence of NMDA‐R antagonist MK‐801 (10 μm) or (c, d) stimulation with AMPA (100 μm) or ACPD, NMDA‐R‐independent glutamate receptor subtype‐specific agonists. Data are from one individual, representative of three individuals examined.

Figure 3

Type 1 but not type 2 cytokine responses by fresh human T cells are inhibited in the presence of N‐methyl‐d‐aspartate (NMDA) or kynurenine metabolite, quinolinic acid (QA). (a) Freshly obtained peripheral blood mononuclear cells were stimulated with anti‐CD3/CD28 for 24 hr then NMDA receptor (NMDA‐R) agonists (100 μm NMDA or 40 μm QA) were added. Cytokine levels were assessed at 48 hr. Each point represents a unique individual. Median values shown with paired non‐parametric analyses. (b) T helper type 2 (Th2) cytokine responses from anti‐CD3/CD28 cultures or, in separate wells, phytohaemagglutinin (10 μg/ml) in the presence/absence of NMDA‐R agonists were quantified. All supernatants in 12 experiments with unique donors were harvested at 48 hr. Median values and Wilcoxon analyses are shown. All other comparisons were not statistically significant. (c) CD4⁺ cells were anti‐CD3/CD28 activated for 24 hr then exposed to NMDA for 3 days, with MTT added for the final 5 hr. Cell viability was calculated as a ratio by setting the absorbance of cells in the absence of NMDA as 100% (medians and Friedman's test shown for n = 4 individuals).

Figure 4

Differential Ca²⁺ flux responses occur in activated T helper type 1 (Th1) versus Th2 cells upon N‐methyl‐d‐aspartate receptor (NMDA‐R) stimulation. Intracellular Ca²⁺ responses in anti‐CD3/CD28‐activated Th1 and Th2 short‐term lines pre‐loaded with Ca²⁺‐sensitive dyes then stimulated with (a) 500 μm and 1 mm NMDA or (b) 40 μm quinolinic acid (QA). Date are representative of 8 and 11 experiments for Th1 and Th2 respectively in (a) for NMDA and of five Th1 and Th2 lines in (b) for QA.

Figure 5

Differential responses of T helper type 1 (Th1) versus Th2 cell lines to N‐methyl‐d‐aspartate receptor (NMDA‐R) stimulation. (a) Th1 and (b) Th2 cells were anti‐CD3/CD28 stimulated for 24 hr then NMDA (100 μm) or quinolinic acid (QA) (40 μm) was added for 24 hr after which cytokine production was determined. Median values (Wilcoxon matched pairs tests) from replicate experiments, using Th1 or Th2 cell lines derived from five independent donors for each Th phenotype are shown.

Figure 6

(a) Reduced proliferation of T‐cell receptor (TCR) ‐activated T helper type 1 (Th1) versus Th2 cell lines in the presence of N‐methyl‐d‐aspartate (NMDA). CFSE‐labelled Th1 and Th2 cells stimulated with anti‐CD3/CD28 for 24 hr then incubated with NMDA at the concentrations shown. Proliferation was determined at day 4 by flow cytometry. Relative proliferation indices were calculated in comparison to the proliferation index for groups lacking NMDA exposure as 100%. n = 6 experiments in total, ***P < 0·001. (b) CFSE FACS plots for representative Th1 and Th2 lines stimulated with anti‐CD3/CD28 in the presence/absence of 200 μm NMDA treatment. (c) Live cell number transition of Th1 and Th2 cell lines. Th1‐ or Th2‐driven cells obtained from five unique donors for each Th phenotype were anti‐CD3/CD28 stimulated as in (a), then incubated with 200 μm NMDA. Live cell numbers were determined by trypan blue counting at the indicated times. Paired t‐test is shown.

Figure 7

T helper type 1 (Th1) cells differ from Th2 cells in their sensitivity to cell death upon N‐methyl‐d‐aspartate receptor (NMDA‐R) ligation. Kinetic analysis of cell death among Th1 versus Th2 cells stimulated with anti‐CD3/CD28 for 24 hr then cultured as indicated with (a, b) NMDA for the period indicated plus/minus the inhibitor MK‐801 (10 mm) or (c) quinolinic acid (QA) (40 μm). The total percentage of physiological cell death was determined by adding Annexin V⁺ TO‐PRO‐3^‐ cells (early stage of apoptosis) and Annexin V⁺ TO‐PRO‐3⁺ cells (late stage of apoptosis or necrosis) at each time‐point. n = 4 independent experiments, each with unique Th1 and Th2.