Defense Date


Document Type


Degree Name

Doctor of Philosophy


Microbiology & Immunology

First Advisor

Kimberly K. Jefferson


Bacterial vaginosis (BV) describes a dysbiotic state of the vaginal mucosa, during which the dominance by beneficial lactobacilli species is lost, the vaginal pH increases, and the flora comprising the vaginal microbiome becomes more diverse. Bacterial taxa associated with this dysbiotic state interact to form a dense polymicrobial biofilm on the surface of the vaginal epithelium. This shift in the vaginal microflora is clinically important, even when symptoms are not noted, as women with BV are at increased risk for acquisition of other sexually transmitted infections, including HIV, and are more than two times as likely to experience preterm birth. Moreover, BV has been associated with numerous other adverse outcomes, including pelvic inflammatory disease, recurring urinary tract infections, chorioamnionitis, postpartum endometritis, infertility, and adverse neonatal outcomes, even among full term infants. The prevalence of BV, in the United States alone, is approximately 30% for women of reproductive age, and is significantly increased in Hispanic and African American women. Standard metronidazole treatment carries a recurrence rate of 52% within 12 months of treatment and recent reports indicate that BV recurrence can, at least in part, be attributed to the inability to reestablish the beneficial lactobacilli population after antibiotic treatment. Thus, it is clear that more targeted approaches to BV treatment are needed.

While the last few decades of research have given us a better understanding of the organisms that comprise the vaginal microbiome, there are still a lot of gaps in our understanding of what they are doing and how they are interacting with the host and each other, both in the context of health and disease. This is due to inadequacies of animal models and difficulty in culturing and genetically manipulating vaginally relevant microorganisms. The overarching aim of this dissertation research was to characterize the factors that contribute to the establishment and maintenance of vaginal dysbiosis, focusing on the role(s) of one organism, Gardnerella vaginalis. Various tools were developed to further characterize putative virulence determinants produced by Gardnerella to better understand how the bacterium interacts with the human host, with emphasis on the cholesterol-dependent toxin, vaginolysin (VLY), and an opacity phenotype presumptively driven by differential pilus production. Through the development of a human vaginal epithelial model of Gardnerella infection, in which Gardnerella spp. grow to in vivo densities, we uncovered a previously undescribed influence of tissue polarity on VLY activity. This influence was mediated by differential expression of the proposed VLY co-receptor, CD59. Further investigation of VLY in different Gardnerella strains revealed conserved amino acid differences within regions of the toxin known to interact with host cells, allowing for the characterization of distinct VLY types. Two predominant VLY types exhibited differential interaction with vaginal epithelial cells, putatively related to CD59 availability. Bioinformatic analyses showed that certain VLY types are restricted to select Gardnerella species and the dominant VLY type in vaginal samples has implications for overall Gardnerella abundance and symptom frequency during BV. We then characterized a colony opacity phenotype potentially susceptible to phase variation. Both opaque and translucent variants were detected in all tested Gardnerella isolates, representing six different species. Opaque variants exhibited increased bacterial aggregation in suspension as well as increased surface biofilm formation, to suggest that opacity may promote resistance to host defenses. Finally, we report the development of the first method for performing site-specific mutagenesis in Gardnerella spp. After optimization, this oligonucleotide-based protocol yielded transformants at an efficiency similar to that observed in other bacterial species. Using these optimized methods, we were able to transform 3 strains of Gardnerella representing 3 different species. While this represents a significant advance in the field, continued optimization is necessary to further improve transformation efficiency in order to increase the practicality of using this tool to generate non-selectable transformants. In sum, continuing to build upon the knowledge gleaned from this work will allow us to develop more targeted therapies for BV and ultimately reduce its global health burden.


© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission