John W. Lammons

Background: Bacterial vaginosis (BV) is a vaginal dysbiosis associated with adverse reproductive and infectious outcomes. Current diagnostics identify BV after symptom onset. We evaluated whether artificial neural network (ANN) models of the vaginal microbiome could predict incident BV (iBV) up to 14 days before clinical diagnosis. Methods: ANNs were trained using 16S rRNA gene sequencing data from 1,201 longitudinal vaginal specimens collected from 58 women across two prospective cohorts. Models classified individual specimens as pre-iBV or healthy using the relative abundance of vaginal bacterial taxa. Model performance was assessed using a held-out participant-level test set, participant-level cross-validation, and external validation. SHAP analysis was used to identify taxa associated with model predictions. Results: On the held-out participant-level test set, the ANN achieved 93% accuracy (AUC = 0.97, sensitivity = 95%, specificity = 92%). Models using only five taxa ( Lactobacillus crispatus , Gardnerella spp., L. iners , L. mulieris , and Megamonas spp.) maintained >91% accuracy, sensitivity, and specificity. SHAP analysis identified Lactobacillus spp. and Gardnerella spp. as the taxa most strongly associated with model predictions. Participant-level cross-validation produced lower and more variable performance (AUC = 0.826 ± 0.076; accuracy = 71.5% ± 8.2%), while external validation achieved approximately 80% balanced accuracy. Interpretation: Vaginal microbiome composition contains predictive information that can identify women at risk of iBV before clinical onset. Lower performance during participant-level cross-validation and external validation indicates that additional validation in larger, more diverse cohorts is required before clinical implementation.

Chlamydia trachomatis (Ct) is a causal agent of upper reproductive tract pathology. There is a broad spectrum of cervical Ct load in infected women, and upper tract infection is associated with higher cervical Ct load. Recent studies indicate that bacterial vaginosis (BV) can modulate host-Ct outcomes. To identify features associated with BV status and Ct load, we performed an integrated multi-omics analysis of the cervicovaginal microbiome, tryptophan metabolome, and cytokines. Samples were analyzed using 16S rRNA gene sequencing, targeted UPLC-MS/MS quantification of tryptophan metabolites, and multiplex cytokine profiling. Ordination analyses showed that BV status was separated by the microbiome, metabolome, and cytokines, whereas Ct load was separated only by cytokines. K-means clustering of tryptophan metabolites defined three metabolome state types (MSTs). MST I, associated primarily with Lactobacillus crispatus -dominated community state type (CST) I, exhibited high tryptophan availability, indole-3-lactic acid, and complete kynurenine-pathway activity. Both MST II and MST III were associated with BV-associated CST IV and showed marked tryptophan depletion. MST II was broadly depleted of most tryptophan metabolites, while MST III was enriched in downstream microbially derived indole pathway metabolites and kynurenic acid. Hierarchical all-against-all association testing revealed coordinated relationships linking clusters of bacterial taxa, metabolites, and cytokines. Importantly, multi-omics network analyses identified integrated microbial-metabolic-immune modules that predicted high versus low Ct load, highlighting CXCL9, CXCL10, IL-17, BV-associated taxa, and indole pathway metabolites as key discriminative features. Results demonstrate that cervical Ct load reflects coordinated microbial-metabolic-immune ecological states rather than microbiome composition alone and refine current models of Ct-BV interactions.

Key bacterial vaginosis (BV)-associated bacteria implicated in biofilm formation include Gardnerella species, Prevotella bivia , and Fannyhessea vaginae . We investigated their spatial organization in the BV biofilm over time from longitudinal vaginal specimens obtained from women with incident BV (iBV) using peptide nucleic acid-fluorescence in situ hybridization (PNA-FISH). Heterosexual women with optimal vaginal microbiota self-collected vaginal specimens twice daily for 60 days or until iBV development (Nugent score 7–10 on ≥4 consecutive specimens). Women who developed iBV were matched to healthy controls by age, race, and contraceptive method. Gardnerella spp., P. bivia , and F. vaginae were quantified using PNA-FISH 2 days pre-iBV, the day of iBV, and 2 days post-iBV across five optical layers ( z , z + 2, z + 4, z + 6, and z + 8 μm). Total counts of all three bacterial species were significantly higher on the day of iBV compared to 2 days pre-iBV ( P = 0.011) and remained elevated 2 days post-iBV. Across most layers and time points, pooled mean Gardnerella spp. counts were significantly higher than F. vaginae counts ( P ≤ 0.022-0.0003). On the day of iBV and 2 days post-iBV, pooled mean counts of Gardnerella spp. and F. vaginae progressively increased across most biofilm layers ( P ≤ 0.043-0.0012). Controls had significantly lower counts of Gardnerella spp. and F. vaginae. P. bivia had low counts in all specimens. During the critical time period surrounding iBV, Gardnerella spp. are abundant throughout the developing biofilm and facilitate F. vaginae incorporation at later time points and higher biofilm layers. Additional research, including other Prevotella spp., is needed. IMPORTANCE: Bacterial vaginosis (BV) is the most common vaginal infection in reproductive-age women worldwide with a global prevalence of 30%. Recurrence rates can be up to 60% within 1 year of treatment. While BV is characterized as a polymicrobial biofilm infection, the exact etiology remains unknown. The BV biofilm may persist after antibiotic treatment, possibly due to incomplete eradication by current antimicrobial therapies, contributing to recurrent infection. Data are limited in evaluating the spatial formation of the BV biofilm around the time of incident BV. Providing a better understanding of this critical time period in incident BV pathogenesis is necessary to inform the development of prevention methods aimed at inhibiting biofilm formation and improving long-term treatment outcomes.

Frailty is a major health concern associated with aging and has been linked to gut microbiome composition, especially in elderly individuals needing care. People with HIV (PWH) present high risk of early-onset frailty. This study examines the relationship between frailty and the gut microbiome, with an emphasis on sex-based differences in PWH. Data were drawn from 268 participants in the New Orleans Alcohol Use in HIV (NOAH) Study, including 16S microbiome sequencing from stool samples, cytokine levels, viral load, and T cell counts. Frailty was assessed using phenotypic frailty index (PFI) and a 58-item deficit index (DI-58). Fourteen taxa were significantly linked to both PFI and DI-58. Among females, butyrate-producing genera (e.g., Eubacterium ventriosum and Butyricimonas) were inversely associated with frailty, while genera such as Paraprevotella and Romboutsia showed positive associations. In males, Prevotella and Erysipelotrichaceae UCG-003 were inversely associated, whereas Christensenellaceae R-7, Clostridium sensu stricto, and others were positively associated with frailty. Fifteen genera were also linked to circulating cytokine levels (IL-1β, IL-17A, IL-2, IFN-γ). We identified three community state types (CSTs) with unique bacterial structures. Community state analysis identified CST 2, characterized by lower diversity and more pathobionts, to be highly predictive of increased frailty. These findings underscore the relevance of sex-specific differences in the gut microbiota relations with frailty, supporting tailored interventions for frailty in PWH. Graphical Abstract

Background: Bacterial vaginosis (BV) is a vaginal dysbiosis characterized by polymicrobial communities of BV-associated bacteria (BVAB) adhered to the vaginal epithelium. Despite decades of research, its etiology remains unknown. We aimed to investigate BV biofilm formation over time among women who developed incident BV (iBV) using peptide nucleic acid-fluorescence in situ hybridization (PNA-FISH), focusing on 3 key BVAB (Gardnerella species, Prevotella bivia, and Fannyhessea vaginae). Methods: Heterosexual, nonpregnant women ages 18–45 with optimal vaginal microbiota were enrolled to self-collect twice-daily vaginal specimens for 60 days. iBV was defined as a Nugent score of 7–10 on ≥4 consecutive specimens. For women who developed iBV (cases), Gardnerella spp., P. bivia, and F. vaginae were visualized and quantified by PNA-FISH for up to 14 days prior to iBV, the day of iBV, and 3 days post-iBV. Cases were matched to women maintaining optimal vaginal microbiota (controls) based on age, race, and contraceptive method. Control specimens were matched to case specimens by day of menses. Results: Among 135 women enrolled, 18 developed iBV and were matched to 18 controls. Pooled median Gardnerella spp. counts significantly increased starting 5 days before iBV, while pooled median F. vaginae counts significantly increased on the day of iBV diagnosis. In contrast, pooled median P. bivia counts were not significantly different between groups. Conclusions: These data suggest that Gardnerella spp. are early colonizers of the BV biofilm while F. vaginae is a secondary colonizer. P. bivia was not found to be significantly different between iBV case and control specimens.