Background:
HIV persists in the human body, including the gastrointestinal (GI) tract where the virus causes cellular damage, microbial translocation and inflammation. The GI tract also act as a source of rebounding virus upon cessation of ART. We investigate HIV dynamics within host between tissue reservoirs and evaluate biomarkers, viral reservoir characteristics and microbiome composition associated with viral dynamics in 8 participants of the Last Gift cohort.
Methods:
To infer HIV dynamics within each participant, Bayesian discrete trait analyses (DTA) were conducted using HIV DNA single genome full env sequencing data across 33 tissues (10 to 25 per participant) including 7 GI tissues. A panel of 44 cytokines were measured perimortem in blood (ELISA and Luminex assays). Microbiome was characterized in GI tissue (bacterial 16S-rDNA sequencing). Bayesian hierarchical models were used to interrogate the association between viral migration between tissues and (1) HIV reservoir characteristics (size measured by ddPCR, diversity and genetic distance), (2) microbiome composition and (3) soluble biomarkers. The models included predictors of the source and recipient tissues, whether the tissues are within the same biological system, and host- and tissue- level effects.
Results:
For inflammatory markers, our DTA showed that viral migration within, from or toward GI tissues (Fig.1A) was consistently associated with higher levels of markers of inflammation and homeostasis (e.g.IL22, RANTES) and lower blood levels of ICAM1, a regulator of cellular responses in inflammation (Fig.1B). For viral characteristics, our DTA showed strong associations between lower genetic distance between viral populations at anatomical sites and increased viral migration between those sites. Among GI tissues, viral diversity in source and recipient tissues was also positively associated with viral migration. Finally, there was a positive association between viral migration and the abundance of Fusobacteria (marker of inflammation) in the source tissue, and Actinobacteria in the recipient tissue. There was also more viral migration from tissues with lower Firmicutes/Bacteroides ratio, a marker of gut dysbiosis.
Conclusions:
This study further supports the impact of HIV reservoir composition on viral dynamics. We showed that systemic inflammation and microbial gut environment also impacted HIV dynamics. Our approach provides a robust way to evaluate host and tissue level markers associated with directional HIV dispersal within host.
A. Global host HIV dynamics of FL HIV env sequences from 2 participants. B.Posterior distribution of the estimates for three selected biomarkers (predictors) associated with viral movement (outcome) obtained from a Bayesian hierarchical model (BHM). ![]()