Oral Microbiome in HIV-Infected Women: Shifts in the Abundance of Pathogenic and Beneficial Bacteria Are Associated with Aging, HIV Load, CD4 Count, and Antiretroviral Therapy

Abstract

Human immunodeficiency virus (HIV)-associated nonacquired immunodeficiency syndrome (AIDS) conditions, such as cardiovascular disease, diabetes, osteoporosis, and dementia are more prevalent in older than in young adult HIV-infected subjects. Although the oral microbiome has been studied as a window into pathogenesis in aging populations, its relationship to HIV disease progression, opportunistic infections, and HIV-associated non-AIDS conditions is not well understood. We utilized 16S rDNA-based pyrosequencing to compare the salivary microbiome in three groups: (1) Chronically HIV-infected women >50 years of age (aging); (2) HIV-infected women <35 years of age (young adult); and (3) HIV-uninfected age-matched women. We also examined correlations between salivary dysbiosis, plasma HIV RNA, CD4⁺ T cell depletion, and opportunistic oral infections. In both aging and young adult women, HIV infection was associated with salivary dysbiosis characterized by increased abundance of Prevotella melaninogenica and Rothia mucilaginosa. Aging was associated with increased bacterial diversity in both uninfected and HIV-infected women. In HIV-infected women with oral coinfections, aging was also associated with reduced abundance of the common commensal Veillonella parvula. Patients taking antiretroviral therapy showed increased numbers of Neisseria and Haemophilus. High plasma HIV RNA levels correlated positively with the presence of Prevotella and Veillonella, and negatively with the abundance of potentially beneficial Streptococcus and Lactobacillus. Circulating CD4⁺ T cell numbers correlated positively with the abundance of Streptococcus and Lactobacillus. Our findings extend previous studies of the role of the microbiome in HIV pathogenesis, providing new evidence that HIV infection is associated with a shift toward an increased pathogenic footprint of the salivary microbiome. Taken together, the data suggest a complex relationship, worthy of additional study, between chronic dysbiosis in the oral cavity, aging, viral burden, CD4⁺ T cell depletion, and long-term antiretroviral therapy.

Keywords: HIV; aging; disease progression; opportunistic infection; oral microbiome; saliva.

FIG. 1.

Immunologic and virologic characteristics of HIV patient groups and HIV-uninfected women. (A) Comparison of absolute peripheral CD4⁺ T cell numbers in HIV patient groups and HIV-uninfected women quantitated by flow cytometry indicated a statistically significant depletion in each infected group. p-Values were calculated by Student's t-test. *p < .05, **p < .01. (B) Although young adult HIV-infected women displayed peripheral viral loads that were about one log higher than other infected groups, the difference was not statistically significant. The remaining HIV patient groups all displayed similar levels of plasma HIV RNA. UI, uninfected; HL, hairy leukoplakia; OC, oral candidiasis; HIV, human immunodeficiency virus; NS, not significant. Color images are available online.

FIG. 2.

Phylogenetic analysis of the impact of aging and HIV infection on the salivary microbiome. The richness and abundance of bacterial taxa in the salivary microbiome was evaluated and compared through alpha (Shannon index) diversity scores (A). HIV-infected young adult (orange line) and aging (top, green line) groups harbored significantly more diverse communities than their age-matched control counterparts (red and blue lines, respectively). (B) Mean phylogenetic distribution at the phylum level taxonomic resolution of bacteria comprising at least 0.5% of the total bacterial community in the saliva of young adult (<35) and aging (>50) untreated HIV-infected women and age- and gender-matched HIV-uninfected women. (C) Mean distribution of highest taxonomic levels detected within each phylum. Color images are available online.

FIG. 3.

Impact of ART and opportunistic oral infections on the salivary microbiome. Bar graphical representation of the mean phylogenetic distribution of bacteria comprising at least 0.5% of the total bacterial community in the saliva of untreated HIV-infected women, HIV-infected women on long-term ART, and HIV-infected women with diagnosed oral opportunistic infections. ART, antiretroviral therapy. Color images are available online.

FIG. 4.

Principal coordinates analysis (PCoA). PCoA was utilized to evaluate differences in bacterial community structure. Microbiome structure was not significantly different based on HIV incidence, age (in both HIV-infected and HIV-uninfected groups), or by coinfection (A–D). However, significant clustering (p = .024) was observed among HIV-infected subjects when age and coinfection status were considered collectively (E). Additionally among these two groups, HIV RNA loads significantly correlated (F) with young adult HIV-infected women (p = .013), whereas therapy status correlated significantly with HIV-positive aging subjects with coinfection (p = .012). Color images are available online.

FIG. 5.

Linear discriminant analysis effect size (LEfSe). LEfSe was performed to identify statistically different bacterial taxa in their relative abundance between young adult HIV-infected women versus aging HIV-infected women with coinfection. Kruskal–Wallis test was used to process dataset while LEfSe alpha values were set at 0.05 and logarithmic linear discriminant analysis score >2 for discriminative feature filtering. Color images are available online.

FIG. 6.

Comparison of salivary EBV and Candida albicans levels in aging and young adult HIV-infected individuals and HIV-uninfected women. The quantities (DNA copy numbers) of (A) EBV and (B) C. albicans in saliva were determined by quantitative polymerase chain reaction. A statistically significant increase in EBV was observed in the HIV-infected young adult group in comparison to age-matched HIV-uninfected women. No significant change in C. albicans was detected between experimental and control groups. EBV, Epstein–Barr virus. Color images are available online.

FIG. 7.

Correlation matrix of relationships between salivary dysbiosis, age, clinical status, and alcohol consumption. Correlative relationships were determined between age, plasma HIV RNA, peripheral CD4⁺ T cell depletion, salivary EBV and C. albicans copy numbers, alcohol consumption, and the predominant genus-level changes in the salivary microbiome of HIV-infected patients. Positive correlations are shown in red and negative correlations in blue with increasing color intensity corresponding with decreasing p-values. All of the bacterial genera (X-axis) showing statistically significant correlations to one or more of the study parameters (Y-axis) described above are depicted. Correlations were determined by Spearman's rank correlation coefficient test. Statistically significant (<0.05) p-values are shown. <.01. White squares are used to highlight trends toward correlation, where .05 < p-value < .2. Color images are available online.