| dc.description.abstract | Background: Pregnant women are a key target population in the prevention of new HIV infections, and their protection can be extrapolated to their unborn children during pregnancy and at birth. The role of vaginal microbiota in shaping host susceptibility to HIV infection is not fully understood, particularly in pregnant African women. We hypothesized that vaginal microbiota affects epithelial barrier integrity and contributes to host susceptibility to HIV infection.
Objectives: To investigate vaginal microbial diversity and its immuno-modulatory effects on host susceptibility to HIV infection among pregnant women in Uganda.
Methods: A formative review of vaginal microbiome diversity among African women was conducted. This was followed by a cross-sectional study of 360 pregnant mothers, aged 18 years and over, who were randomly selected among women attending antenatal care at Kawempe National Referral Hospital. We screened for herpes simplex virus-2 (HSV-2) by enzyme-linked immunosorbent assay (ELISA), and gonorrhea, Chlamydia and Mycoplasma genitalium were all detected by a multiplex real-time qualitative polymerase chain reaction (RT-PCR). The vaginal microbiome was characterized using 16S rRNA sequencing on the Illumina platform. Species richness and abundances of vaginal commensal bacteria were analyzed from a random sample of 179 pregnant women (100 HIV-negative and 79 HIV-infected). The concentrations of cytokines IL-1β, IL-1RA, IL-6, IL-8, IL-10, and TNF-α within the cervicovaginal lavage were determined using multiplex ELISA.
To determine the effect of vaginal microbiota on the integrity of the vaginal epithelium, we conducted a sub-study of pregnant women with sequenced microbiome. We measured concentrations of tight junction proteins zonula occludens-1 (ZO-1) and claudin-1 in cervicovaginal lavages by ELISA. Thereafter, we set up ex-vivo experiments to measure the effect of vaginal microbiota on epithelial tight junctions and immune responses using both immortalized vaginal cells VK2 (E6/E7, ATCC® CRL-2616™) and a vaginal epithelial tissue model. Bacterial cell-free supernatants, derived from cultures of the dominant bacteria in the vaginal tract of the study population, were used to mimic the in-vivo vaginal milieu. From vaginal tissue lysates treated with bacterial supernatants, changes in mRNA of tight junction genes of ZO-1, claudin-1, claudin-4, JAM-A, occludin relative to the house-keeping gene GADPH were measured by QuantiGene™ Plex Gene Expression Assay. Expression of tight junction proteins in vaginal epithelial cells and tissues in response to bacterial cell-free supernatants was quantified by western blot. The cytokine response of vaginal epithelial cells, VK2 to bacterial cell-free supernatants was also measured by ELISA method. Immuno-staining of claudin-1 and claudin-4 expression enabled visualization of the effect of bacterial cell-free supernatants on the tight junctions of treated vaginal cells.
Results: From the literature review, a few studies had characterized the vaginal microbiome among African women (mostly non-pregnant) and in general, the vaginal microbiome was more diverse in Africans relative to Western women. The vaginal microbiome of pregnant women in this study was categorized into four distinct cervicotypes (“CTs”). CT1 which was predominantly non-iners Lactobacillus (6%, 11/179), CT2 which was dominated by L. iners (13%, 24/179), CT3 that was Gardnerella dominant (49%, 87/179) and CT4, a mixed CT co-dominated by L. iners, Gardnerella and Atopobium. Few of the pregnant women, 35/179 (19%) had a Lactobacillus-dominant vaginal microbiota profile. G. vaginalis was present in 90% of the samples and L. iners was the most prevalent Lactobacillus species; predominant in 24/35 (68%) of the women with Lactobacillus-dominant microbiota.
Women with a non-Lactobacillus dominant microbiome (CT3 &CT4) had significantly higher levels of inflammatory cytokines IL-1beta, TNF-alpha, and chemokines IL-6 and IL-8 in their cervicovaginal lavages. Mean concentration of claudin-1 in cervicovaginal lavages of HIV-negative women with CT1 was 0.06332 (SD+ 0.03218), CT2 was 1.336 (SD+ 1.999), CT3 was 0.5912 (SD+ 1.002) and 1.225 (SD+2.54) among women with CT4. HIV-infected pregnant women had lower claudin-1 and higher ZO-1 concentration compared to the HIV-negative pregnant women.
In the ex-vivo experiments, bacterial cell-free supernatants from L. iners and L. crispatus elicited low levels of pro-inflammatory cytokines IL-1alpha, IL-8, IL-6 and IL-1 beta in comparison to media and G. vaginalis. Soluble factors produced by G. vaginalis lowered claudin-1 and claudin-4 expression but L. crispatus or L. iners reversed this effect of G. vaginalis; as observed by immuno-staining of treated vaginal cells.
Conclusions: There were limited studies on the vaginal microbiome of pregnant women in sub-Saharan Africa and African women had a more diverse vaginal microbiome relative to women in the America and Europe as observed in the literature review. In this study, there was a high prevalence (80%) of non-Lactobacillus-dominant vaginal microbiota among pregnant women receiving antenatal care at Mulago hospital. The non-Lactobacillus-dominant vaginal microbiota was associated with elevated levels of pro-inflammatory cytokines and alterations of the tight junctions of vaginal epithelium. These findings were reproduced in-vitro when bacterial cell-free supernatants were applied onto vaginal epithelial cells or tissue models. Given that both inflammation and weakened vaginal epithelium favor HIV acquisition, pregnant women with a non-Lactobacillus dominant microbiota may be at greater risk of HIV acquisition than those with a Lactobacillus dominant microbiota. This data present promising opportunities to design and study HIV prevention interventions that may modify the non-Lactobacillus dominant vaginal microbiota to a more protective Lactobacillus-dominant microbiota in order to reduce the high rate of HIV-infections observed in young women of sub-Saharan Africa.
| en_US |
