People living with HIV display increased anti-apolipoprotein A1 auto-antibodies, inflammation, and kynurenine metabolites: a case-control study

Abstract

Objective: This study aimed to study the relationship between auto-antibodies against apolipoprotein A1 (anti-apoA1 IgG), human immunodeficiency virus (HIV) infection, anti-retroviral therapy (ART), and the tryptophan pathways in HIV-related cardiovascular disease.

Design: This case-control study conducted in South Africa consisted of control volunteers (n = 50), people living with HIV (PLWH) on ART (n = 50), and untreated PLWH (n = 44). Cardiovascular risk scores were determined, vascular measures were performed, and an extensive biochemical characterisation (routine, metabolomic, and inflammatory systemic profiles) was performed.

Methods: Anti-apoA1 IgG levels were assessed by an in-house ELISA. Inflammatory biomarkers were measured with the Meso Scale Discovery® platform, and kynurenine pathway metabolites were assessed using targeted metabolomic profiling conducted by liquid chromatography-multiple reaction monitoring/mass spectrometry (LC-MRM/MS).

Results: Cardiovascular risk scores and vascular measures exhibited similarities across the three groups, while important differences were observed in systemic inflammatory and tryptophan pathways. Anti-apoA1 IgG seropositivity rates were 15%, 40%, and 70% in control volunteers, PLWH ART-treated, and PLWH ART-naïve, respectively. Circulating anti-apoA1 IgG levels were significantly negatively associated with CD4+ cell counts and positively associated with viremia and pro-inflammatory biomarkers (IFNγ, TNFα, MIPα, ICAM-1, VCAM-1). While circulating anti-apoA1 IgG levels were associated with increased levels of kynurenine in both control volunteers and PLWH, the kynurenine/tryptophan ratio was significantly increased in PLWH ART-treated.

Conclusion: HIV infection increases the humoral response against apoA1, which is associated with established HIV severity criteria and kynurenine pathway activation.

Keywords: HIV; anti-apolipoprotein A1 auto-antibodies; anti-retroviral therapy; autoimmunity; cardiovascular disease; kynurenine pathway metabolites.

Figure 1

Study design. Clinical parameters and samples from 144 patients were analysed. These subjects were divided into three groups: control volunteers without HIV (n = 50), PLWH on ART (PLWH ART+) (n = 50), and PLWH ART-naïve (PLWH ART−) (n = 44). The analysis included clinical parameters, inflammatory profile, autoimmune response, and kynurenine pathway metabolites.

Figure 2

Cardiovascular risk estimation by Spearman's rank correlation. (A) Spearman's rank correlation between the FRS and clinical parameters. The heatmap is divided into subject groups of PLWH (n = 94), PLWH ART-treated (PLWH-ART+, n = 50), PLWH ART-naïve (PLWH-ART−, n = 44), and control volunteers (n = 49). (B) Spearman's rank correlation between cIMT and clinical parameters. The heatmap is divided into subject groups of PLWH (n = 74), PLWH ART-treated (PLWH-ART+, n = 46), PLWH ART-naïve (PLWH-ART−, n = 28), and control volunteers (n = 48). (C) Spearman's rank correlation between anti-apoA1 IgG and clinical parameters. The heatmap is divided into subject groups of PLWH (n = 94), PLWH ART-treated (PLWH-ART+, n = 50), PLWH ART-naïve (PLWH-ART−, n = 44), and control volunteers (n = 49). FRS, cIMT, and anti-apoA1 IgG were correlated to the following categories: HIV profile, CV risk estimation, inflammatory cytokines, and kynurenine pathway metabolites. In each category, parameters were top-down ranked according to the r value. Statistic difference was evaluated using non-parametric Spearman ranking tests: * p-values <0.05, ** < 0.01, *** < 0.001, and **** < 0.0001 were considered significant.

Figure 3

Spearman's rank correlation between the HIV profile and kynurenine pathway metabolites. The heatmap is divided into subjects groups of PLWH (n = 94), PLWH ART-treated (PLWH-ART+, n = 50), PLWH ART-naïve (PLWH-ART−, n = 44) and by the following parameters: CD4+ cell count (cells/µL) and viral load (RNA copies/mL). In each category, parameters were top-down ranked according to the r value. Statistic difference was evaluated using non-parametric Spearman ranking tests: * p-values <0.05, ** < 0.01, *** < 0.001, and **** < 0.0001 were considered significant.

Figure 4

Anti-apoA1 IgG treatment induces ICAM-1 and VCAM-1 expression but not tryptophan pathway metabolites. HAECs were incubated with anti-apoA1 IgG (40 µg/mL) or control IgG (40 µg/mL) for 24 h. Cells were lysed, mRNA was extracted, and the expression of VCAM-1 and ICAM-1 was analysed (A,B). The incubation medium was collected and the levels of VCAM-1 and ICAM-1 were quantified using the MSD platform (C,D). The incubation medium was collected, and the levels of kynurenine, tryptophan, kynurenine/tryptophan ratio (Kyn/Tryp), and kynurenic acid were quantified using LC/MRM-MS (E–H). Results are expressed as medians with interquartile ranges. Statistical difference was evaluated using non-parametric Wilcoxon tests; p < 0.05 was considered significant.