Local barrier dysfunction identified by confocal laser endomicroscopy predicts bacterial translocation in HIV infection.

Abstract

Several studies link bacterial translocation in HIV infection to inflammation, immune system activation [1], lymphoid tissue fibrosis [2,3] and a worse clinical outcome [4–6]. There is little evidence in the literature on the relationship between microscopic changes such as epithelial barrier dysfunction and the amount of intramucosal bacteria and the bacterial translocation, activation and inflammatory biomarkers observed in HIV-1-infected patients [7]. Confocal laser endomicroscopy has been used to detect local barrier dysfunction, epithelial gaps and recognition of intramucosal bacteria in vivo[8–10]. The objective of the study was to describe morphological and dynamic findings of the intestinal mucosa by direct visualization with confocal endomicroscopy in patients with chronic HIV infection and correlate these findings with markers of inflammation, bacterial translocation, activated lymphocytes and myeloid cell (CD11c+) in mucosa. The study was approved by the institutional ethical review board. Demographic and clinical characteristics, pathological changes of rectal mucosa biopsies, confocal endomicroscopy findings (amount of intramucosal bacteria, amount of fluorescence in the crypt lumen and in lamina propria), microbial translocation markers (lipopolysaccharide binding protein, soluble CD14 and endotoxin core IgM antibody) and inflammation (TNF-α, IL-6, ultrasensitive C-reactive protein, D-dimer) in plasma, T-cell and myeloid subsets in rectal biopsy and peripheral blood were analyzed in 10 HIV-1 individuals (Fig. 1a). According to previous studies [8], the amount of intramucosal bacteria was described as none, scanty (in one of the quadrants) or abundant (in more than one quadrant). The amount of fluorescencein in crypt lumen and in lamina propria [9,11] was defined as predominantly black, focally black, darker than cells, same intensity as cells, brighter than cells, focally white or predominantly white. A correlation between microbial translocation and other factors was also performed.Fig. 1: Study design (a); histological cut of normal colonic mucosa (b); confocal endomicroscopy from an HIV-uninfected patient with normal colonic mucosa: scarce intramucosal bacteria, slight fluorescence leakage in crypt lumen and lamina propria and goblet structures (dark ‘dots’ within the epithelial cells) (c); confocal endomicroscopy from an HIV-infected patient with abnormal mucosal barrier: abundant intramucosal bacteria and intense fluorescence leakage in crypt lumen and in lamina propria (d); graphical representation of correlations between lipopolysaccharide binding protein (LBP) and CD4+ cells in peripheral blood (e); graphical representation of correlations between LBP and CD4+CD38+HLA-DR+ cells in peripheral blood (f); graphical representation of correlations between LBP and CD8+CD38+HLA-DR+ cells in peripheral blood (g).We recruited nine men and one woman with median age of 37 years, nine homosexual and one heterosexual. The median CD4+ nadir and current CD4+ was 572 and 767 cells/μl, respectively. All of them had an undetectable viral load (three of them received antiretroviral therapy and seven were considered elite controllers). Regarding changes in rectal mucosa, the lymphoid tissue architecture was preserved in all patients, only one of the 10 patients (10%) showed fibrosis. Moreover, in most of the biopsies analyzed, only mild chronic inflammation was observed (8/10 individuals) (data not shown). Regarding confocal endomicroscopy, the fluorescence in lamina propria was increased and the amount of intramucosal bacteria was high in most individuals, suggesting an abnormality of the mucosal barriers (Fig. 1b–d). Changes of gut mucosa assessed by confocal endomicroscopy were associated with translocation markers and myeloid subsets in mucosa. The amount of fluorescence in lamina propria, which determines local barrier dysfunction [9], positively correlated with CD14s (rho = 0.73, P = 0.015) and also with the proportion of CD11clowHLA-DRlowCD14− cells (rho = 0.801, P = 0.005), which likely correspond to mucosal macrophages [12]. Myeloid cells and especially macrophages (CD11clowHLA-DRlowCD163+) have been described to be accumulated in gut mucosa in HIV [13] and simian immunodeficiency virus (SIV) infection [14,15] driving persistent inflammation, allowing HIV replication and acting as long-lasting viral reservoirs in tissues [15,16]. The amount of intramucosal bacteria was positively associated with endotoxin core IgM antibody (rho = 0.64, P = 0.04) but negatively correlated with the proportion of CD11chighCD83+CCR5+ cells (rho = −0.844, P = 0.004), suggesting that the presence of intramucosa bacteria is associated with a lower amount of a myeloid DC-like subset that expresses CD83 and CCR5. Intramucosal bacteria have been described to be more common in inflammatory bowel disease than in patients with normal mucosa [8]. In addition, Dillon et al.[17] demonstrated that HIV-infected patients had a lower amount of myeloid DC (CD11chighCD83+) in colonic mucosa comparing with uninfected individuals. This subpopulation could be implicated in suppressing CD4+ T cells activation in mucosa from HIV-infected patients. Our data suggest that intramucosal bacteria are related to a higher mucosa immune activation and, as it has been described in inflammatory bowel diseases, it could be a valuable tool in the elucidation of the pathogenesis of HIV-1 gut disfunction. Furthermore, as expected we found that translocation markers were correlated with markers of inflammation and CD4+ and CD8+ T-cell subsets in peripheral blood (Fig. 1e–g). To our knowledge, no previous reports have described that epithelial barrier dysfunction measured as the amount of fluorescence extravasation in crypts and in the own lamina in HIV-1 infection. Significantly, they were associated with increased bacterial translocation and changes in the mucosal T-cell and myeloid subsets. Thus, confocal endomicroscopy provides advantages over conventional endoscopy as it allows direct visualization of barrier defects and bacteria from the intestinal lumen. Our data suggest that confocal endomicroscopy could be a good tool to further study gut epithelial damage and microbial translocation in HIV-infected patients. Acknowledgements The study was partially supported by grants: European Comission (grants: H2020-SC1-2016-RTD Proposal: 731626), the Spanish Ministry of Economy (MINECO) (grants: SAF2015-66193-R, RTI2018-096309-B-I00), the Fondo de Investigación Sanitaria (FIS) PI15/00480, AC16/00051 and PI18/00699, the Fondo Europeo para el Desarrollo Regional (FEDER), the SPANISH AIDS Research Network RD16/0025/0002 – ISCIII – FEDER (RIS) and the CERCA Programme/Generalitat de Catalunya SGR 615 and SGR 653. CIBERehd is funded by the Instituto de Salud Carlos III. F.G. has received the support of José María Segovia de Arana contracts. HIVACAT: HIV development program in Catalonia. IDIBAPS: Institut d’Investigacions Biomèdiques August Pi I Sunyer. Conflicts of interest The authors do not have a commercial or other association that might pose a conflict of interest.