A tryptophan metabolite prevents depletion of circulating endothelial progenitor cells in systemic low-grade inflammation

Abstract

Background: Chronic systemic inflammation reduces the bioavailability of circulating endothelial progenitor cells (EPCs). Indoleamine 2,3-dioxygenase 1 (IDO1), a key enzyme of immune tolerance catalyzing the initial step of tryptophan degradation along the so-called l-kynurenine (l-kyn) pathway, that is induced by inflammatory stimuli and exerts anti-inflammatory effects. A specific relationship between IDO1 activity and circulating EPC numbers has not yet been investigated.

Methods: In this study, circulating EPCs were examined in mice treated with low doses of lipopolysaccharide (LPS) to mimic low-grade inflammation. Moreover, the association between IDO1 activity and circulating EPCs was studied in a cohort of 277 patients with variable systemic low-grade inflammation.

Results: Repeated low doses of LPS caused a decrease in circulating EPCs and l-kyn supplementation, mimicking IDO1 activation, significantly increased EPC numbers under homeostatic conditions preventing EPC decline in low-grade endotoxemia. Accordingly, in patients with variable systemic low-grade inflammation, there was a significant interaction between IDO1 activity and high-sensitivity C-reactive protein (hs-CRP) in predicting circulating EPCs, with high hs-CRP associated with significantly lower EPCs at low IDO1 activity but not at high IDO1 activity.

Interpretation: Overall, these findings demonstrate that systemic low-grade inflammation reduces circulating EPCs. However, high IDO1 activity and l-kyn supplementation limit circulating EPC loss in low-grade inflammation.

Keywords: EPC — endothelial progenitor cells; IDO1 enzyme; IL-6 (interleukin 6); hs-CRP (high sensitivity C-reactive protein); kynurenine (KYN); low-grade inflammation.

Figure 1

Low grade of inflammation reduces circulating EPCs. (A) WT mice were treated with low doses of LPS for 3 weeks. On day 0, day 15 and day 22 bleeding was performed for the determination of circulating EPCs, serum cytokines, and serum IDO1 activity; subsequently, mice were sacrificed for the determination of Ido1 mRNA and IDO1 protein expression in splenocytes. Created by BioRender.com. (B) WT mice were treated with low doses of LPS for 3 weeks as depicted in (A). On day 0, day 15 and day 22 bleeding was performed for the determination of IL-6 and TNF-α. Data are shown as mean ± S.D. ****P < 0.0001, one-way ANOVA with the Bonferroni post hoc analysis. 5 mice per group (n = 2). (C) Heat map representation of cytokine content in serum from mice treated as (A) at different time points. Shown is heatmap with colored boxes, representing cytokines amounts ranging from bright blue (lowest) to bright red (highest). Numbers represent cytokine concentrations (pg/ml). (D) WT mice were treated as in B and circulating EPCs were determined by flow cytometry. Data are shown as means ± S.D. *P < 0.05, paired samples t test. 5 mice per group (n=2). ns, not significant.

Figure 2

The tryptophan metabolism preserves EPCs numbers in vivo under homeostatic conditions. (A) Ido1 and Ido2 mRNA expression was assessed by RT-PCR on day 0, day 15 and day 22 in splenocytes derived from mice treated either with PBS or LPS. Data are represented as normalized transcript expression in the samples relative to normalized transcript expression in control mice. Data are shown as mean ± S.D. One-way ANOVA with the Bonferroni post hoc analysis. 5 mice per group (n=2). (B) Western analysis was carried out for IDO1 and IDO2 protein expression in total splenocytes derived from mice in (A) β-Tubulin was used ad loading control. One representative experiment (n = 3). (C) Serum l-kyn content serum was determined by HPLC in mice treated as in (A) Data are shown as mean ± S.D. One-way ANOVA with the Bonferroni post hoc analysis. 5 mice per group (n=2). (D) The number of circulating EPCs was evaluated in WT mice treated with IDO1 (Epacadostat) or TDO2 (680C91) inhibitors or vehicle as control under homeostatic conditions by flow cytometry. ***P < 0.001, one-way ANOVA followed by Bonferroni multiple comparison test. 10 mice per group (n=2). (E) Pharmacokinetic of oral administration of l-kyn. 2 mg of l-kyn were administrated by oral gavage and the l-kyn concentration was monitored in the plasma at indicates time points. Data are shown as mean ± S.D. ****P < 0.0001, two-way ANOVA with the Bonferroni post hoc analysis. 3 mice per group (n = 2). (F) The number of circulating EPCs was evaluated in WT and Ido1^–/– mice treated with PBS or l-kyn for 3 weeks by flow cytometry. *P < 0.5, ***P < 0.001, one-way ANOVA followed by Bonferroni multiple comparison test. 5 mice per group (n=2). ns, not significant.

Figure 3

l-kynurenine prevents TNF-α effects on the formation of capillary networks. (A) Representative images of tubulogenic assay on BOECs in presence or absence of TNF-α and/or l-kyn. Vehicle was used as control. Scale bar = 500 μm. (B) Quantification of number of nodes, branches, junctions and total length of the tubule networks. ****P < 0.0001; ***p < 0.001; **p 0.01; *p < 0.05. Data are expressed as mean ± S.D. and are representative of three independent experiments. ns, not significant.

Figure 4

l-kynurenine administration prevents EPC numbers reduction during low-grade inflammation. (A) WT mice were treated with low doses of LPS in combination with l-kyn for 3 weeks. On day 0, day 15 and day 22 bleeding was performed for the determination of circulating EPCs and pro-inflammatory cytokines. Created by BioRender.com. (B, C) The number of circulating EPCs was evaluated by flow cytometry on day 0, day 15 and day 22 in mice treated with either l-kyn or LPS or combination of both. Data are shown as means ± S.D. *P < 0.05, paired sample t test. 5 mice per group (n=2). (D) Serum derived from mice treated as in (A-C) was assessed for IL-6 production. Data are shown as means ± S.D. ****P < 0.0001, one-way ANOVA with the Bonferroni post hoc analysis. 5 mice per group (n=2). ns, not significant.

Figure 5

Low IDO1 activity is associated with a decrease of EPC numbers in patients characterized by low-grade inflammation. Correlation between lg-EPCs and lg-hs-CRP (A) and between lg-EPCs and lg-IDO1 activity (B) in the human study population. (C) Interaction between IDO1 activity and hs-CRP in the prediction of circulating EPCs. EPCs, endothelial progenitor cells; hs-CRP, high-sensitivity C-reactive protein; IDO1, indoleamine 2,3-dioxygenase 1; lg, logarithmic. Patients with variable low-grade systemic inflammation (n = 277) were stratified according to low versus high IDO1 activity (i.e., Kyn/Trp <40 versus Kyn/Trp ≥40) and high versus low hs-CRP (i.e., < 3 mg/L versus ≥3 mg/L). High hs-CRP is associated with significantly lower EPC numbers as compared with low hs-CRP at low IDO1 activity but not at high IDO1 activity. P values are from the independent sample t test.

Figure 6

Schematic representation of the role of tryptophan metabolism in the low-grade inflammation leading to the maintenance of EPC numbers. In a preclinical model of systemic low-grade inflammation, repetitive low-dose LPS administration reduces circulating EPC levels. Moreover, l-kyn supplementation limit circulating EPC loss (left). In human beings (right), high hs-CRP is associated with significantly lower EPC counts as compared to low hs-CRP in the presence of low Kyn/Trp ratio but not high Kyn/Trp ratio, while low Kyn/Trp ratio is associated with lower circulating EPCs as compared to high Kyn/Trp ratio in the presence of high hs-CRP but not low hs-CRP. Created by BioRender.com.