Psychological stress-induced, IDO1-dependent tryptophan catabolism: implications on immunosuppression in mice and humans

Abstract

It is increasingly recognized that psychological stress influences inflammatory responses and mood. Here, we investigated whether psychological stress (combined acoustic and restraint stress) activates the tryptophan (Trp) catabolizing enzyme indoleamine 2,3-dioxygenase 1(IDO1) and thereby alters the immune homeostasis and behavior in mice. We measured IDO1 mRNA expression and plasma levels of Trp catabolites after a single 2-h stress session and in repeatedly stressed (4.5-days stress, 2-h twice a day) naïve BALB/c mice. A role of cytokines in acute stress-induced IDO1 activation was studied after IFNgamma and TNFalpha blockade and in IDO1(-/-) mice. RU486 and 1-Methyl-L-tryptophan (1-MT) were used to study role of glucocorticoids and IDO1 on Trp depletion in altering the immune and behavioral response in repeatedly stressed animals. Clinical relevance was addressed by analyzing IDO1 activity in patients expecting abdominal surgery. Acute stress increased the IDO1 mRNA expression in brain, lung, spleen and Peyer's patches (max. 14.1+/-4.9-fold in brain 6-h after stress) and resulted in a transient depletion of Trp (-25.2+/-6.6%) and serotonin (-27.3+/-4.6%) from the plasma measured 6-h after stress while kynurenine levels increased 6-h later (11.2+/-9.3%). IDO1 mRNA up-regulation was blocked by anti-TNFalpha and anti-IFNgamma treatment. Continuous IDO1 blockade by 1-MT but not RU486 treatment normalized the anti-bacterial defense and attenuated increased IL-10 inducibility in splenocytes after repeated stress as it reduced the loss of body weight and behavioral alterations. Moreover, kynurenic acid which remained increased in 1-MT treated repeatedly stressed mice was identified to reduce the TNFalpha inducibility of splenocytes in vitro and in vivo. Thus, psychological stress stimulates cytokine-driven IDO1 activation and Trp depletion which seems to have a central role for developing stress-induced immunosuppression and behavioral alteration. Since patients showed Trp catabolism already prior to surgery, IDO is also a possible target enzyme for humans modulating immune homeostasis and mood.

Figure 1. Major pathways of tryptophan (Trp) catabolism.

Dietary Trp is essential for protein biosynthesis, but the majority is catabolized along the kynurenine pathway (bold arrows) providing NAD⁺ for cellular redox- and energy homeostasis. Alternative pathways are the conversion of Trp to serotonin (serotonin) and then to melatonin, or to tryptamine and then to the kyurenamines (not shown). Indoleamines analysed in this study are shaded in grey. Abbreviations: 3-HAO, 3-hydrox-anthranilic acid oxidase; 5-HIAA, 5-hydroxy-indole acetic acid; IDO1 and 2, indoleamine 2,3-dioxgenase 1 and 2; KATs, kynurenine aminotransferases; NAD+,nicotinamide adenine dinucleotide; MAO monoamine oxidase; QPRT, quinolinic acid phosphoribosyltransferase; TDO, tryptophan 2,3-dioxygenase.

Figure 2. Transient activation of Trp catabolism after acute psychological stress.

A–E. Kinetics of IDO1 mRNA expression induced in brain (A), lung (B), spleen (C) and Peyer's patches (D) within 24-h after termination of 2-h-stress exposure (n = 6 mice/time, n = 6 controls; average values of basal expression levels in non-stressed mice were assigned as value of 1), and IDO1 mRNA expression in the brain of TNFα antiserum- and IFNγ antiserum treated vs. vehicle-treated animals 6-h after acute stress (E, n = 6 mice/group, average values of basal expression levels in vehicle treated, non-stressed mice assigned as value of 1). F–I. Kinetics of plasma concentrations of Trp (F), Kyn (G), the Kyn/Trp ratio (H) and of the levels of serotonin (I), Quin (J) and Kyna (K) following 2-h-stress exposure (n = 6 mice/time) compared with basal levels of healthy controls (n = 15 mice/group). All panels depict data of one representative experiment of two: *p<.05; **p<.01; ***p<.001 compared with non-stressed controls; ^#p<.05; ^##p<.01; ^###p<.001 compared with mice immediately after acute stress (0-h) and ^⊥<.05 ^⊥⊥<0.01, ^⊥⊥⊥<0.001 compared with mice 24-h after stress exposure by Kruskal-Wallis testing with post-hoc Dunn Multiple comparison testing (KW- and p-values are indicated in the graph).

Figure 3. Effects of pharmacological IDO1 blockade on Trp catabolism, IL-10 inducibility and anti-bacterial defense in repeatedly stressed mice.

A–E. Plasma concentrations of tryptophan (A) and the Trp/Kyn ratio (B), Stress Severity Score evaluating the behavioral stress response (C), change of body weight during repeated stress (D) and IL-10 levels in the supernatant of splenocytes after ex vivo stimulation with LPS (from S. abortus equi, 1-µg/ml) (E) of repeatedly stressed and non-stressed mice, with or without 1-MT treatment which were measured immediately after the termination after the ninth stress exposure. F–H. Bacterial burden in spleen (F), liver (G) and blood (H) 24-h after ip infection with E. coli ATCC 25922 in mice with or without 1-MT treatment after repeated stress and of non-stressed mice (injection was performed immediately after the ninth stress exposure); n = 8 mice/group, summary of two independent experiments; *p<.05, **p<.01, ***p<.001 compared with non-stressed, without 1-MT-treatment, ^#p<.05; ^##p<.01; ^###p<.001 compared with repeatedly stressed mice by Kruskal-Wallis testing with post-hoc Dunn's Multiple comparison testing. Two-way ANOVA and post-hoc Bonferroni's Multiple comparison test were used for testing the overall influences of stress and treatment (ANOVA F- and p-values for the influence of stress and 1-MT are indicated in the graph).

Figure 4. 1-MT treatment: effects on Kyna levels, TNFα inducibility, lymphocytes and thymus cortex atrophy in repeatedly stressed mice.

Plasma kynurenic acid concentrations (A), TNFα levels in the supernatant of splenocytes (B) after ex vivo stimulation with LPS (from S. abortus equi, 1-µg/ml), absolute lymphocyte counts in the blood (C) and thickness of the thymic cortex (D) of repeatedly stressed and non-stressed mice with or without 1-MT treatment; n = 8 mice/group, summary of two independent experiments; *p<.05, **p<.01, ***p<.001 compared with non-stressed, without 1-MT-treatment, ^#p<.05; ^##p<.01; ^###p<.001 compared with repeatedly stressed mice by Kruskal-Wallis testing with post-hoc Dunn's Multiple comparison testing. Two-way ANOVA and post-hoc Bonferroni's Multiple comparison test were used for testing the overall influences of stress and treatment (ANOVA F- and p-values for the influence of stress and 1-MT are indicated in the graph).

Figure 5. In vivo and in vitro Kyna effects on TNFα inducibility, in naïve mice.

Dose-dependent effects of Kyna on TNFα secretion into the supernatant (A) 16-h after stimulation with LPS (E. coli O55:B5, 500-ng/ml) or on spontaneous release in PBS-treated splenocyte cultures (B), and of LPS-stimulated splenocytes obtained from mice that were pre-treated with 250-ng Kyna/g BW or saline 6-h prior to ex vivo stimulation of splenocytes (C) n = 6 mice/group, summary of two independent experiments. *p<.05, **p<.01, ***p<.001 compared with NaCl control treatment, ^⊥⊥<0.01 compared LPS control treatment by Kruskal-Wallis testing with post-hoc Dunn's Multiple comparison testing (KW- and p-values are indicated in the graph).

Figure 6. Tryptophan catabolism along the kynurenine pathway during the peri-operative stress response of patients undergoing elective abdominal surgery without complications.

Preoperative, intraoperative and postoperative Kyn/Trp ratios in the plasma of 18 patients compared with Kyn/Trp ratios of 20 healthy volunteers ***p<.001 compared with healthy controls, ^#p<.05, ^##p<.01, ^###p<.001, ^####p<.0001 compared with values from pre-operative day -1.