Characterising vaginal Lactobacillus strains from young South African women with persistently optimal vaginal microbiota – Developing the framework for an African vaginal probiotic product development platform for reproductive health

Thesis / Dissertation

2024

Permanent link to this Item
Authors
Journal Title
Link to Journal
Journal ISSN
Volume Title
Publisher
Publisher

University of Cape Town

License
Series
Abstract
Background: Bacterial vaginosis (BV) is a genital condition with a high prevalence in sub-Saharan Africa, which can cause adverse sexual and reproductive health outcomes. Treatment with antibiotics such as metronidazole results in only a brief resolution of the vaginal dysbiosis present in BV and the recurrence rates are high. Probiotics have been explored as a possible adjunctive treatment option, but most commercially available formulations do not contain vaginally-derived strains and there is limited evidence for their efficacy in local populations. Therefore, there is an urgent need for the development of vaginal probiotics tailored towards sub-Saharan African women containing strains with useful probiotic characteristics that have the potential to persistently colonise the female genital tract. Aim: This project aimed to isolate and characterise Lactobacillus strains from young South African women with a stable optimal vaginal microbiota for consideration as candidate strains in a vaginal probiotic product. Methods: Samples from a previously completed study cohort where hormonal contraceptive preferences were examined in adolescent and young women in Cape Town (UChoose) were used for this study. Samples were collected from healthy, HIV uninfected young women (aged 15-19) at three timepoints, 16 weeks apart. To identify women with a stable and optimal vaginal microbiota over time, vaginal microbial communities were classified using VALENCIA. Lactobacillus spp. were selectively isolated from participants with longitudinally stable Lactobacillus crispatus- dominant communities and assayed for their ability to inhibit Prevotella bivia, a BV-associated organism using a soft agar overlay method. Candidates with strong inhibition potential against P. bivia were characterised further fortheir inhibition potential against regionally relevant BV-associated organismsisolated from women within the same study cohort, as well astheir growth at physiologically relevant pH (pH 4, 4.5 and 5), their ability to produce lactic acid and finally their resistance to commonly used BV and urinary tract infection antibiotics. Additionally, whole genome sequencing (WGS) data were available for three strains and these were assessed to determine additional safety characteristics of the probiotic candidates. The strains were screened for antimicrobial resistance genes using several online databases (ResFinder, ARG-ANNOT, CARD, NCBI AMRFinderPlus and VFDB), the presence of integrated prophages using VirSorter and phage defence mechanisms using the DefenseFinder tool. Lastly, putative bacteriocins were identified using BAGEL4. Results: The most common community state type (CST) identified across all participants and visits was the highly diverse CST IV, which is dominated by a variety of anaerobes and shows a strong association with BV. In total 3 participants had an optimal, L. crispatus-dominant vaginal microbiota (CST I) across all three study visits and did not use any antibiotics during the study. Samples from these participants yielded a combined 337 isolates. Isolates displayed a wide variety in inhibition potential against P. bivia ATCC 29303T . The isolates selected for further characterisation were all identified as L. crispatus using 16S rRNA sequencing. More than half of the selected isolates inhibited local Prevotella strains more than they inhibited the P. bivia ATCC 29303T type strain. All L. crispatusisolates produced some of each lactic acid isomer, with the top three L-lactate producers all isolated from the same participant (UC101). No correlation was found between the amount of lactate produced and antimicrobial activity against Prevotella strains suggesting that the inhibition was not lactate dependent. All isolates were resistant to metronidazole, but susceptible to other BV and urinary tract infection treatment antibiotics indicating that co administration with metronidazole is a possibility for these candidate probiotic strains. Of the three genomes analysed, two of the isolates harboured plasmids. No previously characterised antimicrobial resistance determinants were identified. Each of the three strains harboured five putative bacteriocins, which may play a role in activity against competing bacterial strains. Additionally, 12 putative prophages were identified, with none found on the plasmid. Overall, seven different potential phage defence mechanisms were identified on the bacterial chromosomes. Conclusion: The proportion of women with longitudinally stable, optimal vaginal communities in our setting was relatively low. However, targeting samples provided by these women for the isolation of potential probiotic bacteria yielded a large number of isolates with inhibitory activity against a BV-associated pathogen. Phenotypic variation among isolates illustrates the importance of screening multiple strains of the same species per sample. The WGS screen did not identify any virulence or antimicrobial resistance genes suggesting that the probiotic candidates are safe to use. The putative bacteriocins found in the isolates could be targeted for future research. This pilot study provides proof-of-principle for a pipeline that will screen L. crispatus strains with probiotic potential from African women to treat BV and improve reproductive health in Africa.
Description
Keywords

Reference:

Collections