Background
We assessed whether the gut bacteriome affects the size and activity of HIV reservoir in the human gut. Since each gut segment has a different immune and bacteriome environment, we also aimed at identifying a core bacteriome that is associated with HIV reservoir consistently and similarly across all the segments
Methods
Tissue samples of 5 segments (small bowel: duodenum, jejunum, ileum; large bowel: right colon, rectum) were collected during rapid autopsies (≤6 hour from death) from people with HIV (PWH). For each segment, we determined the bacteriome composition by 16SrRNA sequencing and measured cell-associated HIV DNA and RNA by ddPCR. The α and β diversity were analyzed (Phyloseq) after dimension reduction of taxa. Mixed-effect models assessing the associations between the relative abundance (RA) of taxa and HIV DNA and RNA were adjusted for relevant factors (e.g., age, CD4+ count, antibiotics) and performed at the gut level (models A adjusted for segments; Fig.1) and then within each gut segment (models B)
Results
Twenty-four participants were included (87.5% male, 63±13 years, median CD4+ T cells 194/µL, 87.5% virally suppressed). Main cause of death was cancer (66.7%; 4 of the gut). We observed a large heterogeneity in the bacteriome composition at each taxonomic level, even between anatomically adjacent segments. The large bowel had the highest α diversity, but significant differences were found also between segments within large and small bowel; β diversity showed divergence in bacterial communities between small and large bowel. HIV DNA levels (but not HIV RNA) were higher in small than large bowel segments (101 vs 40 cp/106 cells, p=0.003). In models A, 15 genera were associated with HIV burden (Fig.1). In segment specific models B, none of the taxa was consistently associated with HIV burden across all segments: e.g., Clostridium was associated with higher HIV RNA in duodenum, ileum, and rectum only (p≤0.03 all). Some taxa showed opposite associations across segments: e.g., Lachnospira was associated with lower HIV DNA in colon but higher in ileum (p≤0.01 both); Ruminococcus was associated with higher HIV RNA in duodenum but lower in rectum (p≤0.01 both)
Conclusions
The local bacteriome has an impact on the size and activity of HIV reservoir in the gut. However, the segment-specificity of the bacteriome and its relationship with local HIV burden suggest the lack of a core bacteriome that could be univocally targeted by therapeutic microbiome modulation across the whole gut
Table or Figure
