Altered lipid and neurotransmitter metabolism as potential mechanisms of weight gain in women with HIV initiating INSTIs in a pilot study

Abstract

Objective: This pilot study used metabolomics to provide insight into metabolic alterations of INSTI-associated weight gain among women with HIV (WWH).

Methods: The study included 33 virally suppressed WWH who switched to or added an INSTI. Plasma samples collected 6-12 months pre- (Visit 1) and 1-6 months post INSTI add/switch (Visit 2) were analyzed with liquid chromatography-mass spectrometry-based high-resolution metabolomics. The baseline plasma metabolome and changes in metabolomic signatures (from Visit 1 to Visit 2) were compared in women who experienced ≥ 5% weight gain (n=18) over 1-2 years vs those who maintained/lost body weight (n=15).

Results: Median age was 53 (Q1 47, Q3 55) yrs, 94% were Black, baseline BMI was 34.2 (Q1 30.6, Q3 38.5) kg/m2. Median weight change was +9.20 kg (Q1-Q3 6.77-15.13 kg) and -0.68 kg (Q1-Q3 -5.14 to 0.00 kg) for the weight gain vs maintained/lost weight groups, respectively. A total of 820 metabolites spanning 9 enriched metabolic pathways, including amino acid (e.g., tryptophan) and micronutrient pathways (e.g., vitamin E) differed between weight groups before INSTI use (p<0.05). A total of 1147 metabolites spanning 10 enriched pathways, particularly lipid pathways, exhibited a significant group x time interaction effect (p<0.05), with an overall pattern of decreased free fatty acids over time among women who gained weight.

Conclusions: Several metabolic pathways at baseline and within 6 months were associated with weight gain in women initiating INSTIs. Acute changes in lipid metabolism following INSTI initiation provide potential insight into the pathophysiology of weight gain in this population.

Keywords: HIV; integrase strand transfer inhibitor; metabolism; metabolomics; obesity; overweight.

Figure 1.. Unsupervised hierarchical cluster analysis (HCA) discriminates between women who gained weight vs those who maintained/lost weight after adding INSTIs to their treatment regimen.

A total of 428 HILIC+ and 392 C18− metabolites (Y-axis) were different in abundance in plasma between the two groups at p<0.05. Two-way HCA analysis illustrates separation of participants based on weight change 1-2 years after starting INSTIs. Participants who gained (>5%) weight after starting INSTIs are represented in orange along upper X-axis, while those who maintained or lost weight are represented as the blue color. Panel A shows data from the HILIC positive chromatographic column. Panel B shows data from the C18 negative chromatographic column.

Figure 2.. Selected metabolites from enriched metabolic pathways differing between women on INSTI who later gained weight compared to those who maintained/lost weight reflect dopamine degradation.

Blue graphs represent women WWH who maintained or lost weight, while orange represents those who later gained weight. WWH who later gained weight following INSTI initiation exhibited elevated levels of metabolites that reflected dopamine degradation by several pathways, including 3,4-dihydroxyphenylacetaldehyde via monoamine oxidase (MAO), methoxytyramine* via catechol-O-methyltransferase (COMT), and salsolinol and N-methysalsolinol via direct oxidation of dopamine. Co-factors for enzymes involved in the conversion of tyrosine to dopamine were elevated, including 6,7-dihydrobiopterin and sepiapterin as metabolites along the BH4 salvage pathway (a cofactor for tyrosine hydroxylase) and pyridoxine* (vitamin B6) as a cofactor for aromatic amino acid decarboxylase (AADC). Starred (*) metabolites have been confirmed via ion dissociation mass spectrometry (MS/MS) and co-elution with authentic standards (Metabolomics Standards Initiative confidence level 1). Tyrosine* was detected and is a confirmed metabolite but did not differ between groups. Metabolites in grey boxes were not detected in this analysis but are part of the significantly enriched pathway.

Figure 3.. Selected metabolites from tryptophan metabolism pathway reflect upregulated kynurenine metabolite synthesis.

Blue graphs represent WWH who maintained or lost weight, while orange represents those who later gained weight. WWH who later gained weight following INSTI initiation had elevated levels of metabolites that reflected an upregulated kynurenine synthesis pathway, including 3-hydroxykynurenine* (via KMO) and 3-hydroxyanthranilic acid (via kynurenase). The serotonin precursor, 5-hydroxytryptophan* was lower among WWH who later gained weight compared to those that maintained or lost weight. Starred (*) metabolites have been confirmed via ion dissociation mass spectrometry (MS/MS) and co-elution with authentic standards (Metabolomics Standards Initiative confidence level 1). Tryptophan* and kynurenine* were detected and are confirmed metabolites but did not differ between groups. Metabolites in grey boxes were not detected in this analysis but are part of the pathway. Tryptophan may also be metabolized through bacterial indole synthesis pathways, although detected indole metabolites did not differ between groups.

Figure 4.. Representative plasma metabolites from lipid-related pathways that changed acutely as a function of time and group.

Metabolites shown are those with a significant group-by-time interaction term (*p<0.05, **p<0.01) in assessing weight group changes from baseline (pre-switch) to 1-6 months of INSTI initiation. All metabolites shown have been confirmed via ion dissociation mass spectrometry (MS/MS) and co-elution with authentic standards (Metabolomics Standards Initiative confidence level 1).