Gut Bacterial Communities in HIV-Infected Individuals with Metabolic Syndrome: Effects of the Therapy with Integrase Strand Transfer Inhibitor-Based and Protease Inhibitor-Based Regimens

Abstract

Antiretroviral therapies (ART) are strongly associated with weight gain and metabolic syndrome (MetS) development in HIV-infected patients. Few studies have evaluated the association between gut microbiota and integrase strand transfer inhibitor (INSTI)-based and protease inhibitor (PI)-based regimens in HIV-infected patients with MetS. To assess this, fecal samples were obtained from HIV-infected patients treated with different regimens (16 PI + MetS or 30 INSTI + MetS) and 18 healthy controls (HCs). The microbial composition was characterized using 16S rRNA amplicon sequencing. The INSTI-based and PI-based regimens were associated with a significant decrease in α-diversity compared to HCs. The INSTI + MetS group showed the lowest α-diversity between both regimens. A significant increase in the abundance of short-chain fatty acid (SCFA)-producing genera (Roseburia, Dorea, Ruminococcus torques, and Coprococcus) was observed in the PI + MetS group, while Prevotella, Fusobacterium, and Succinivibrio were significantly increased in the INSTI + MetS group. Moreover, the Proteobacteria/Firmicutes ratio was overrepresented, and functional pathways related to the biosynthesis of LPS components were increased in the INSTI + MetS group. The gut microbiota of patients receiving INSTIs showed a more pronounced dysbiosis orchestrated by decreased bacterial richness and diversity, with an almost complete absence of SCFA-producing bacteria and alterations in gut microbiota functional pathways. These findings have not been previously observed.

Keywords: HIV infection; antiretroviral therapy; gut dysbiosis; gut microbiota; inflammation; metabolic syndrome.

Figure 1

Alpha diversity metrics comparing the HC, PI + MetS, and INSTI + MetS groups. The boxes extend from the 25th to the 75th percentile (interquartile range, IQR), and the lines inside the boxes represent median values. The vertical lines represent the lowest and the highest values. Individual sample values are shown as dots. * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 2

Beta diversity. PCoA plots for weighted (A) and unweighted UniFrac (B) distances in the HC, PI + MetS, and INSTI + MetS groups.

Figure 3

Relative abundances of the plots of the HC, PI + MetS, and INSTI + MetS groups. (A) Donut chart representing the relative abundance of different phyla in the three groups. Phyla are accompanied by relative abundance percentage. (B) Stacked bar plot of bacterial families in the three groups. Top three families include relative abundance percentage. (C) Stacked bar plot of bacterial genera. Top five genus include relative abundance percentage.

Figure 4

Scatter plot of Firmicutes/Bacteroidetes, Proteobacteria/Firmicutes, and Prevotella/Bacteroides ratios in in HC, PI + MetS, and INSTI + MetS groups. Data were first transformed by means of the centered log-ratio (clr). Then, relative abundances were obtained and calculated according to the material and methods section. Results are showed as mean ± SEM. Analyzed using the Kruskal–Wallis test with Benjamini–Hochberg (BH) multiple testing correction, * p <0.05, ** p < 0.01, *** p < 0.001.

Figure 5

Negative correlations between alpha diversity metrics and relative abundance of Prevotella among the HC, PI + MetS, and INSTI + MetS groups. Spearman’s ρ (rho) and p-values (two-tailed) are showed below each diagram. Relative abundances were first transformed using clr as previously described.

Figure 6

(A) LEfSe bar plot showing LDA scores of differentially abundant taxa among the HC (red), INSTI + MetS (green), and PI + MetS (blue) groups (LDA > 3.90, p < 0.05). (B) Cladogram showing differentially abundant taxa at phylum, class, family, and genus levels between the HC, INSTI + MetS, and PI + MetS groups. Red circles indicate the remarkable taxa in the HC group, while green and blue circles indicate the INSTI + MetS and PI + MetS groups, respectively. (LDA > 3.90, p < 0.05).

Figure 7

Volcano plot depiction of ANCOM analysis. The horizontal axis is the centered log ratio transformation (clr) representative of the difference in abundance of a significant taxonomical unit. A green color indicates an overlap of ASVs between groups. The W statistic indicates the value of the statistical test corresponding to the number of times the null hypothesis was rejected for each taxon. (Upper) Comparison among the HC vs INSTI + MetS groups, (middle) HC and PI + MetS groups, and (bottom) INSTI + MetS vs PI + MetS groups.

Figure 8

Volcano plot depiction of ANCOM methodology applied to PICRUSt2 results. Each dot represents a statistically significant pathway enriched in each group. Green dots indicate non-significant pathways. Upper: Healthy controls vs the INSTI + MetS groups; middle: Healthy controls vs. the PI + MetS groups; bottom: the INSTI + MetS vs PI + MetS groups.

Figure 9

Correlations between blood biochemical parameters and genus of interest in the HC, PI + MetS and INSTI + MetS groups. Spearman’s ρ (rho) and p-values (two-tailed) are showed below each diagram. TG: Triglycerides, VLDL: very-low-density lipoprotein. CRP: C-reactive protein. Values on the X-axis are expressed as clr-transformed total abundances.