Partial restoration of gut-mucosal dysbiosis in late-treated HIV-infected subjects with CD4 T-cell recovery

Abstract

[...]we aimed to explore potential differences in specific dysbiosis in the gut-mucosal microbiome, comparing biopsies from two different locations (ileum and caecum), of HIV-infected subjects with different clinical phenotypes. SEE PDF] Sequencing data of 16S rRNA from the microbiota present in terminal ileum and caecum samples of study subjects were analysed and compared with bacterial taxonomic databases. [...]parameters of richness and alpha-diversity of the gut microbiome showed no statistically significant differences between healthy and HIV-groups (elite controller [EC], early-treated [ET], late-treated high recovery [LT-HR] and late-treated low recovery [LT-LR]) regardless the biopsied intestinal area (Figure 2A) or patient sex (Figure S1). [...]abundance analyses were performed without separating by gut location. [...]this is the first report to analyse the potential association of the clinical phenotype of HIV-subjects, regarding their CD4 status at cART onset and afterwards, with alterations in the microbiome composition.

Dear Editor,

Dysbiosis of the gut microbiome is commonly found in human immunodeficiency virus (HIV)-infected subjects,¹ related to alterations such as microbial translocation, inflammation and chronic immune activation.² These alterations especially affect late-diagnosed and/or late-treated individuals, who represent 40%–60% of HIV-subjects and have higher morbimortality.³ Furthermore, a quarter of the late-treated subjects do not properly recover CD4 T-cell levels after combined antiretroviral therapy (cART) and suffer an even worse clinical evolution (see Supporting Information-text for extended context).³ There is no consensus on a single sample type or gut location suitable for studying dysbiosis, but reports mostly rely on faecal samples.^4,5 However, interactions with the HIV-targeted gut-associated lymphoid tissue mainly depend on mucosal-specific microorganisms.⁶ Moreover, studies comparing alterations in different intestinal sites are scarce.⁷ Additionally, most studies have explored the effect of infection itself⁸ or from cART⁹ on the gut microbiota. However, potential associations between immune status and changes occurring in the gut microbiome of treated HIV-infected subjects have not been explored yet.

Our hypothesis was that clinical phenotypes with poorer immune status (in terms of CD4 counts before and after cART) would present a more dysbiotic microbiome in their gut mucosa. Thus, we aimed to explore potential differences in specific dysbiosis in the gut-mucosal microbiome, comparing biopsies from two different locations (ileum and caecum), of HIV-infected subjects with different clinical phenotypes. Associations with parameters of inflammation, immune activation and gut tissue damage were also explored.

Biopsies of terminal ileum and caecum mucosa, as well as blood samples, were taken from 35 virologically-suppressed HIV-infected subjects after at least a 2-year cART period during colonoscopies performed at Virgen del Rocío University Hospital between 2014 and 2017. These patients were classified into three groups depending on their CD4 levels at cART onset and their response upon receiving therapy, according to Figure 1. Ten healthy (non-HIV) people, as control subjects, and three elite controllers (ECs) were also enrolled (description of groups available at Table S3).

[IMAGE OMITTED. SEE PDF]

Sequencing data of 16S rRNA from the microbiota present in terminal ileum and caecum samples of study subjects were analysed and compared with bacterial taxonomic databases. Thus, parameters of richness and alpha-diversity of the gut microbiome showed no statistically significant differences between healthy and HIV-groups (elite controller [EC], early-treated [ET], late-treated high recovery [LT-HR] and late-treated low recovery [LT-LR]) regardless the biopsied intestinal area (Figure 2A) or patient sex (Figure S1). Even so, healthy and EC groups showed higher mean values of richness and diversity than ET and LT HIV-groups. A strong positive correlation was found (r > .8; p < .0001) when alpha-diversity was compared between ileum and caecum samples of all study groups (Figure 2B). In addition, estimation of beta-diversity through Non-metric multidimensional scaling (NMDS) analysis revealed a clustering of healthy and EC samples compared to the rest of HIV-groups (PERMANOVA, p < .001; HOMOVA, p < .041) (Figure 2C). These results were also significant and even clearer when distances were calculated by axis, especially in NMDS1 (Figure 2D). As for alpha diversity, beta diversity did not show relevant differences when comparing ileum and caecum samples (Figure S2). Thus, further abundance analyses were performed without separating by gut location.

[IMAGE OMITTED. SEE PDF]

When operational taxonomic units (OTUs) identified in gut mucosal samples were compared between groups at phylum level, EC showed the highest relative abundance of Firmicutes (p ≤ .07), but lower abundance of Actinobacteria compared to ET (p = .04), LT-HR (p = .006) and LT-LR (p = .01) groups (Figure 3A). At class level, Clostridia, Bacteroidia and Deltaproteobacteria were more significantly abundant in healthy and EC groups (p ≤ .03) than treated HIV-groups, who showed instead higher relative abundance of Bacilli, Erysipelotrichi, Actinobacteria or Coriobacteriia (p ≤ .03) (Figure S3). Using DESeq2 methodology, OTUs appearing as significantly different in relative abundance at family/genus level were compared among study groups (Figures 3B and S4A). For a visual simplification of all comparisons, profiles of taxonomic grouping of more abundant OTUs, in their respective genus and/or family taxa, in comparisons of each study group with the rest of groups was also performed (Figure S4B). Healthy and EC groups presented a similar pattern of significantly more abundant taxa dominated by Ruminococcaceae and Lachnospiraceae members. Strikingly, ET and LT-HR coincided in Propionibacterium, Carnobacterium, Pseudomonas, Butyricicoccus, Dorea and Rothia as exclusive abundant genera. By contrast, LT-LR showed a different pattern to the rest of groups, with greater abundance of Enterobacteriaceae pathobionts as Escherichia (Figures S4A,B). Interestingly, when ET or LT-HR were compared with LT-LR, both first concurred in most of their abundant OTUs belonging to Propionibacterium, Carnobacterium, Pseudomonas and Dorea, but also Blautia, Clostridium and Veillonella genera. In addition, comparison between ET and LT-HR showed the lowest number of differentially abundant OTUs (Figure 3B). All these results could be corroborated by an alternative method to calculate the effect size, LEfSe, obtaining the same characteristic abundant OTUs for each group (Figure 3C). Taxonomic relationships of phylogeny among all the OTUs obtained were displayed using a cladogram (Figure 3D). No significant additional information was obtained when abundance analyses were performed separating ileum and caecum samples (data not shown).

[IMAGE OMITTED. SEE PDF]

A random forest analysis was performed to find out OTUs whose relative abundance could classify treated-subjects based on their immune status, regardless the gut location analysed. Again, ET and LT-HR samples showed a large overlap when the three treated-groups were analysed separately, so were grouped for further analysis. We chose the 14 most relevant OTUs (mean decrease accuracy, MDA > 5) (Figure 4A) for a multivariable logistic regression model that yielded a nine OTUs’ signature as the best to predict samples belonging to ET/LT-HR or LT-LR groups (Figure 4B). Using an receiver operating characteristic (ROC) curve, predictions of this OTUs-based model gave an area under curve of .97.

[IMAGE OMITTED. SEE PDF]

Some peripheral (Figure S5, panel A) and gut mucosa-related (Figure S5, panel B) parameters were compared in available samples of subgroups of subjects belonging to the different treated-groups. Most of the inflammation markers, either soluble or tissue-scored (Tables S1 and S2), as well as microbial translocation, immune activation and mucosal barrier damage parameters were higher in late-treated groups, especially in non-recoverers (LT-LR).

A limitation of this study was the low number of samples, although being comparable with previous studies using gut-mucosal samples in groups of HIV-subjects. Such limitation made not possible to match groups by age and/or gender. Additionally, it would have been relevant to record data about transmission risk (i.e., men who have sex with men) and the additional study on metabolic pathways would have helped elucidate differential microbiome functions possibly associated to particular HIV-groups.

In conclusion, this is the first report to analyse the potential association of the clinical phenotype of HIV-subjects, regarding their CD4 status at cART onset and afterwards, with alterations in the microbiome composition. Also importantly, our study was performed in two different gut-mucosal locations when most related studies are based on faecal samples.^4,5 Late-treated-subjects recovering CD4 under cART (LT-HR) showed a partial restoration of the gut-mucosal dysbiosis produced by HIV-infection, being their microbiome composition similar to that of the more immunopreserved early-treated-subjects (ET). Such composition included Propionibacterium, Carnobacterium, Pseudomonas or Dorea, among others, as coinciding and exclusive abundant genera. In contrast, non-recoverers (LT-LR) appeared enriched in Escherichia, despite displaying all late-treated-subjects similar alpha-diversity values in their microbiota. In addition, a nine OTUs-based signature could be established for the non-recovery clinical phenotype, also affected by more inflammation, immune activation and gut tissue damage. Early treatment and optimal cART regimen election⁹ seem to impact both, the composition of the gut-mucosal microbiome and the clinical evolution of HIV-subjects. Our descriptive study does not allow concluding about causality, and further approaches are encouraged to determine if modifying the gut-microbiota composition could help the CD4 recovery. In that case, potential clinical interventions, as faecal microbiota transplant,¹⁰ are showing promising results at modifying gut microbiota structure to correct HIV-associated dysbiosis (extended discussion at SI-text).

ACKNOWLEDGEMENTS

We would like to thank all the subjects involved in this study who voluntarily underwent a colonoscopy. This work was supported by grants from the Fondo de Investigación Sanitaria (FIS; PI18/01216 and PI21/00357), which is co-funded by Fondos Europeos para el Desarrollo Regional (FEDER) “Una manera de hacer Europa” and the Junta de Andalucía, Consejería de Economía, Innovación, Ciencia y Empleo (Proyecto de Investigación de Excelencia; CTS2593). Israel Olivas-Martínez, Ángel Bulnes-Ramos and Vanesa Garrido-Rodríguez were supported by Instituto de Salud Carlos III (Rio Hortega program CM19/00051, Sara Borrell program CD19/00143 and PFIS program FI19/00298, respectively). María del Mar Pozo-Balado was supported by a postdoctoral contract from Consejería de Transformación Económica, Industria, Conocimiento y Universidades, Junta de Andalucía (DOC_01646). Yolanda María Pacheco was supported by the Consejería de Salud y Familias of Junta de Andalucía through the ‘‘Nicolás Monardes’’ program (C-0013-2017). The funders had no role in the study design, data collection and interpretation, or the decision to submit the work for publication.

CONFLICT OF INTEREST

The authors declare no conflict of interests.

References

Vujkovic-Cvijin I, Somsouk M. HIV and the gut microbiota: composition, consequences, and avenues for amelioration. Curr HIV/AIDS Rep. 2019; 16 : 204-213.

Vázquez-Castellanos JF, Serrano-Villar S, Latorre A, et al. Altered metabolism of gut microbiota contributes to chronic immune activation in HIV-infected individuals. Mucosal Immunol. 2015; 8 : 760-772.

The Late Presentation Working Groups in EuroSIDA and COHERE. Estimating the burden of HIV late presentation and its attributable morbidity and mortality across Europe 2010–2016. BMC Infect Dis. 2020; 20 : 728.

Lu W, Feng Y, Jing F, et al. Association between gut microbiota and CD4 recovery in HIV-1 infected patients. Front Microbiol. 2018; 9 : 1451.

Lee SC, Chua LL, Yap SH, et al. Enrichment of gut-derived Fusobacterium is associated with suboptimal immune recovery in HIV-infected individuals. Sci Rep. 2018; 8 : [eLocator: 14277].

Meyer-Myklestad MH, Medhus AW, Lorvik KB, et al. HIV-infected immunological non-responders have colon-restricted gut-mucosal immune dysfunction. J Infect Dis. 2020; 225 : 661-674.

Mutlu EA, Keshavarzian A, Losurdo J, et al. A compositional look at the human gastrointestinal microbiome and immune activation parameters in HIV infected subjects. PLoS Pathog. 2014; 10 : [eLocator: e1003829].

Lozupone CA, Li M, Campbell TB, et al. Alterations in the gut microbiota associated with HIV-1 infection. Cell Host Microbe. 2013; 14 : 329-339.

Pinto-Cardoso S, Klatt NR, Reyes-Terán G. Impact of antiretroviral drugs on the microbiome: unknown answers to important questions. Curr Opin HIV AIDS. 2018; 13 : 53-60.

Serrano-Villar S, Talavera-Rodríguez A, Gosalbes MJ, et al. Fecal microbiota transplantation in HIV: a pilot placebo-controlled study. Nat Commun. 2021; 12 : 1139.