Vaginal Lactobacillus fatty acid response mechanisms reveal a metabolite-targeted strategy for bacterial vaginosis treatment

Abstract

Bacterial vaginosis (BV), a common syndrome characterized by Lactobacillus-deficient vaginal microbiota, is associated with adverse health outcomes. BV often recurs after standard antibiotic therapy in part because antibiotics promote microbiota dominance by Lactobacillus iners instead of Lactobacillus crispatus, which has more beneficial health associations. Strategies to promote L. crispatus and inhibit L. iners are thus needed. We show that oleic acid (OA) and similar long-chain fatty acids simultaneously inhibit L. iners and enhance L. crispatus growth. These phenotypes require OA-inducible genes conserved in L. crispatus and related lactobacilli, including an oleate hydratase (ohyA) and putative fatty acid efflux pump (farE). FarE mediates OA resistance, while OhyA is robustly active in the vaginal microbiota and enhances bacterial fitness by biochemically sequestering OA in a derivative form only ohyA-harboring organisms can exploit. OA promotes L. crispatus dominance more effectively than antibiotics in an in vitro BV model, suggesting a metabolite-based treatment approach.

Keywords: Lactobacillus; bacterial vaginosis; female genital tract; metabolism; vaginal microbiome; women’s health.

Graphical abstract

Figure 1

cis-9-uLCFAs selectively inhibit L. iners and promote growth of L. crispatus and other FGT lactobacilli (A) Growth of representative L. crispatus, L. gasseri, L. iners, L. jensenii, and L. mulieris strains in modified Lactobacillus MRS broth (MRS + CQ broth³⁶) supplemented with varying concentrations of oleic acid (OA), linoleic acid (LOA), or palmitoleic acid (POA). Calculated as growth relative to the median of no-LCFA control cultures. (B) Relative growth of diverse non-iners FGT Lactobacillus (n = 30) and L. iners (n = 14) strains in MRS + CQ broth supplemented with varying OA concentrations. (C) Minimum bactericidal concentration (MBC) assays for representative L. crispatus, L. gasseri, L. iners, L. jensenii, and L. mulieris strains in MRS + CQ broth. Colony forming units (CFUs) were measured after 24 h of OA exposure and expressed relative to CFU from no-OA controls. (D) Transmission electron microscopy (TEM) images of L. crispatus (top) and L. iners (bottom) treated with 3.2 mM OA (right) or no OA (left) for 1 h. (E) Growth rescue of non-iners FGT Lactobacillus (n = 32) and L. iners (n = 5) strains in S-broth supplemented with varying OA concentrations. Calculated relative to growth in S-broth supplemented with 0.1% (v/v) Tween-80. (A, B, and E) Growth was measured by optical density at 600 nm (OD600) after 72 h. (A and C) Points represent 3 technical replicates per condition and are representative of ≥2 independent experiments. (B and E) Points represent median values for 3 technical replicates per condition and are representative of ≥2 independent experiments. Boxplots represent the 25^th and 75^th percentiles (lower and upper boundaries of boxes, respectively), the median (middle horizontal line), and measurements that fall within 1.5 times the interquartile range (IQR; distance between the 25^th and 75^th percentiles; whiskers). See also Figure S1.

Figure S1

cis-9-uLCFA structures, FGT Lactobacillus cis-9-uLCFA responses, and cis-9-uLCFA response gene presence in FGT Lactobacillus species, related to Figures 1 and 2 (A) Chemical structures of oleic acid, linoleic acid, and palmitoleic acid. (B) Growth of diverse non-iners FGT Lactobacillus (n = 13 strains) and L. iners (n = 14 strains) in NYCIII broth supplemented with varying concentrations of OA. Growth was measured by OD600 after 72 h of culture. Relative growth for each strain was calculated relative to the median OD600 measurement in its no-LCFA control. Points represent median relative growth for 3 technical replicates per condition. Boxplots represent the 25^th and 75^th percentiles (lower and upper boundaries of boxes, respectively), the median (middle horizontal line), and measurements that fall within 1.5 times the IQR (whiskers). (C) Cultures of L. iners and L. crispatus were grown to mid/late-log phase, and then parallel aliquots of each strain were exposed to varying concentrations of OA for the indicated amount of time, followed by centrifugation and separation of supernatants and pellets. ATP concentrations were determined for each supernatant and pellet. ATP release was calculated by dividing the supernatant ATP concentration by the pellet ATP concentration and expressed relative to the strains ATP release at T0. (D) Transcriptional responses of cultured L. crispatus (left), L. gasseri (middle), and L. jensenii (right) grown to exponential phase in MRS + CQ broth, then exposed to LOA (top, 3.2 mM) or POA (bottom, 3.2 mM) for 1 h. Data were analyzed as in Figure 2A. Consistently, DE genes included a predicted ohyA, putative farE, and its putative tetR. The untreated (no-uLCFA) controls for each species in Figures 2A and S1D are the same. (E) Venn diagram showing the shared sets of genes differentially expressed in response to each of the three cis-9-uLCFAs (OA, LOA, and POA) in each species and the overlap between these sets of DE gene functions that were shared among all three species. Three shared DE gene functions (ohyA, farE, and tetR; all upregulated) were observed in all species under all treatment conditions. (F–H) Dot plot showing the shared sets of DE genes induced by OA, LOA, and POA treatments in L. crispatus (F), L. gasseri (G), and L. jensenii (H). Each point depicts a DE gene with color representing log₂(FC) and size representing −log₁₀(adjusted p value). (I) Presence of gene functions predicted to encode oleate hydratase (ohyA) and putative fatty acid efflux pump (farE) activity in long-read sequenced, isolate genomes of strains from the indicated FGT Lactobacillus species, representing all Lactobacillus strains used experimentally in this study (see also Table S1). Presence of gene functions involved in exogenous fatty acid acquisition and utilization (fakAB, plsC, plsX, and plsY) is shown for comparison.

Figure 2

Non-iners FGT lactobacilli share a conserved set of OA response genes that L. iners lacks (A) Transcriptional responses of cultured L. crispatus (left), L. gasseri (middle), and L. jensenii (right) to OA (3.2 mM). Plots depict log₂(fold change [FC]) in OA relative to control and −log₁₀(adjusted p value) for each gene. Dotted lines indicate significant differential expression (FC ≥ −1 or FC ≥ 1 with adjusted p ≤ 0.05). (B) Venn diagram showing numbers of OA-regulated genes in each species and shared among species. Three gene functions (ohyA, COG4716; farE, COG2409; and tetR, COG1309) were consistently OA-regulated. (C) Presence of predicted ohyA and farE genes in isolate and metagenome-assembled genomes (MAGs) of FGT Lactobacillus species (n = 1,167 total).^, Predicted genes involved in exogenous fatty acid acquisition and utilization (fakAB, plsC, plsX, and plsY) are shown for comparison. See also Figures S1, S2, and S3.

Figure S2

OhyA and FarE ortholog diversity in FGT Lactobacillus species and homology to orthologs found in other human-adapted species, related to Figures 2 and 3 (A) EggNOG-predicted orthologs for OhyA and the number of copies of each ortholog per genome or MAG for each species. Numbers of genomes and MAGs for each species are shown. (B) Percent identity matrix of representative OhyA orthologs in FGT Lactobacillus species and other human-adapted bacteria, determined by protein sequence alignment (MUSCLE v5.1¹⁰⁸). (C) EggNOG-predicted orthologs for FarE and the number of copies of each ortholog per genome or MAG for each species. Numbers of genomes and MAGs for each species are shown. (D) FarE protein phylogenetic tree for representative farE orthologs from the indicated Lactobacillus species. Starred leaf tips indicate FGT Lactobacillus orthologs; ^∗ indicates confirmed OA-induced ortholog (Figure 2A). (E) Species phylogenetic tree of representative Lactobacillus genomes constructed based on core ribosomal protein sequences. Starred leaf tips indicate FGT Lactobacillus species. (F) Percent identity matrix of representative EggNOG-predicted FarE orthologs in FGT Lactobacillus species and other human-adapted bacteria, determined by protein sequence alignment (MUSCLE v5.1¹⁰⁸). (G) Schematic of farE and tetR genomic neighborhoods in representative L. crispatus, L. gasseri, L. jensenii, and S. aureus genomes (not to scale). (D and E) Trees were rooted to the FarE ortholog (D) or core genome (E) of the representative genome for Limosilactobacillus vaginalis (see STAR Methods).

Figure 3

FGT Lactobacillus OhyA enzymes are functional and physiologically active (A) OhyA protein phylogenetic tree for representative orthologs from the indicated species (see STAR Methods). Starred leaf tips indicate FGT Lactobacillus orthologs; ^∗ indicates confirmed OA-induced ortholog (Figure 2A). (B) Diagram of OhyA9 enzymatic activity with OA substrate and 10-hydroxystearic acid (10-HSA or h18:0) product. (C) Extracted ion chromatograms from supernatants of ohyA-gene deleted S. aureus complemented with empty vector (ΔSaohyA/empty vector), SaohyA-expressing plasmid (ΔSaohyA/pSaohyA), LCRIS_00558-expressing plasmid (ΔSaohyA/pLCRIS_00558), or LCRIS_00661-expressing plasmid (ΔSaohyA/pLCRIS_00661), cultured with OA for 1 h. Annotated peaks include OA (18:1) and 10-HSA (h18:0). (D) MS2 spectra with major fragmentation labels for the 10-HSA (h18:0) peak from ΔSaohyA/pLCRIS_00661 cultured with OA (Figure 3C, lower right). (E) Universally ¹³C-labeled 10-HSA (¹³C₁₈-10-HSA) concentrations in supernatants of L. crispatus, L. gasseri, and L. jensenii cultured for 72 h in NYCIII broth with or without universally ¹³C-labeled OA (¹³C₁₈-OA; 3.2 mM). (F) ¹³C₁₈-10-HSA concentrations in supernatants of L. crispatus and L. iners cultured for 72 h in NYCIII broth with or without ¹³C₁₈-OA (100 μM, a sublethal concentration for L. iners). (E and F) The same no-OA controls for media and L. crispatus are shown in (E) and (F). Points represent 3 technical replicates per condition. See also Figures S2 and S3.

Figure S3

ohyA and farE orthologs presence is widespread among Lactobacillaceae species, related to Figures 2 and 3 Species phylogenetic tree of representative species genomes from the Lactobacillaceae family, constructed based on core ribosomal genes contained in all species. The species genomes derive from a recent comprehensive review and taxonomic revision of the 48 genera within the Lactobacillaceae family.Staphylococcus aureus and Staphylococcus epidermidis genomes were included to root the phylogenetic reconstruction. Metadata rings mark the genomes of major FGT Lactobacillus species, genomes of other gram-positive host-adapted species, each organism’s lifestyle (if known), and presence or absence of farE and ohyA orthologs in each genome. With the exception of L. iners, all vertebrate-associated members of the Lactobacillus genus—including common mammalian intestinal Lactobacillus species (which also require the OA-containing media supplement Tween-80 for growth)—possessed both putative ohyA and farE genes. Trees were constructed from MUSCLE v5.1-aligned protein sequences using FastTree v2.1 (see STAR Methods).

Figure S4

Detection of enzymatic products from ohyA orthologs and characterization of ohyA9 and farE genetic knockouts in L. gasseri, related to Figures 4 and 5 (A) MS2 spectra with major fragmentation labels for 10-HSA standard (Ambeed, A125712-50MG). (B) Extracted ion chromatograms from supernatants of ΔSaohyA/empty vector, ΔSaohyA/pSaohyA, ΔSaohyA/pLCRIS_00558, and ΔSaohyA/pLCRIS_00661 cultured with LOA for 1 h. Annotated peaks include LOA (18:2) and the detected hydroxyFA (h18:1). (C) OhyA9 enzymatic activity reaction diagram with LOA substrate. (D) MS2 spectra with major fragmentation labels for the h18:1 peak from ΔSaohyA/pLCRIS_00661 cultured with LOA (lower right of Figure S4B), identified as 10-hydroxy-12-octadecenoic acid (h18:1). (E) OhyA12 enzymatic activity reaction diagram with LOA substrate. (F) MS2 spectra with major fragmentation labels for the h18:1 peak from ΔSaohyA/pLCRIS_00558 cultured with LOA (lower left of Figure S4B), identified as 13-hydroxy-9-octadecenoic acid. (G) ¹³C₁₈-10-HSA concentrations in cell pellets of L. crispatus, L. gasseri, and L. jensenii cultured for 72 h in NYCIII broth with and without ¹³C₁₈-OA (3.2 mM). The cell pellets are from the same cultures as the supernatants shown in Figure 3E. (H) ¹³C₁₈-10-HSA concentrations in cell pellets of L. crispatus and L. iners cultured for 72 h in NYCIII broth with and without ¹³C₁₈-OA (100 μM, which is a sublethal concentration for L. iners). The cell pellets are from the same cultures as the supernatants shown in Figure 3F. (I) Presence of gene functions predicted to encode oleate hydratase (ohyA) and putative fatty acid efflux pump (farE) activity in long-read sequenced, isolate genomes of the indicated FGT species. Presence of gene functions involved in exogenous fatty acid acquisition and utilization (fakAB, plsC, plsX, and plsY) is shown for comparison. (J) Untargeted lipidomics was performed on control (blank) media and spent media supernatants collected from strains of diverse FGT bacteria after 72 h of culture in NYCIII broth. Blank media supplemented with 100 μM OA is shown as a positive control. Changes in concentration of key LCFA metabolites for each bacterial species are shown as the log₁₀(fold change) of their median relative abundances compared with control media. The plot depicts median fold change values for 5 technical replicates per condition. Statistical significance was determined by unpaired t test using the Bonferroni method to correct for multiple hypothesis testing (^∗adjusted p < 0.05). (K) Representative hFA extracted ion chromatograms for human CVL samples with CT1 (L. crispatus-dominant, top) or CT2 (L. iners-dominant, bottom) bacterial communities, quantified via a targeted metabolomics approach employing picolylamine-based derivatization.^, (L) DNA gel of PCR products amplified from the primers flanking either LGAS_1630 (farE) or LGAS_1351 (ohyA9) from the L. gasseri ATCC 33323 wild-type (WT) strain and ΔohyA9 and ΔfarE genetic knockout (KO) strains. Expected amplicon lengths were 3.9 kb for LGAS_1630 in WT, 2.0 kb for LGAS_1351 in WT, while in-frame gene deletions of LGAS_1351 in ΔohyA9 and of LGAS_1630 in ΔfarE each had expected amplicon lengths of 1.2 kb. KO strains were additionally WGS verified. (M) Cultures of L. gasseri WT, L. gasseri ΔfarE, and L. gasseri ΔfarE/pfarE were grown to mid- to late-log phase and exposed to varying concentrations of OA, then ATP release assays were performed. (N) Detection of ¹³C-labeled OA (¹³C₁₈-OA) in pellets of L. gasseri ΔfarE and ΔfarE/pfarE genetic mutant strains treated with 100 and 400 μM ¹³C₁₈-OA in NYCIII broth for 15 min. Pellets were washed two times with ice-cold PBS before sample preparation for targeted ¹³C₁₈-OA detection. The fold change of [¹³C₁₈-OA] was calculated by dividing the relative abundance of ¹³C₁₈-OA detected in the 400 μM ¹³C₁₈-OA condition for each technical replicate by the median relative abundance of ¹³C₁₈-OA detected in 100 μM ¹³C₁₈-OA condition. Significance of the difference in fold change was determined by unpaired t test (^∗∗∗p < 0.001). Points represent 5 technical replicates per condition. (O) ¹³C₁₈-10-HSA relative abundance in cell pellets from L. gasseri WT, ΔohyA9, ΔohyA9/pohyA9, ΔfarE, and ΔfarE/pfarE cultured for 24 h in NYCIII broth with and without ¹³C₁₈-OA (100 μM). Pellets are from the same cultures as the supernatants in Figure 5D. Points represent 2 or 3 technical replicates per condition. (G and H) Points represent 3 technical replicates per condition.

Figure 4

Women with non-iners lactobacilli have uniquely elevated vaginal concentrations of OhyA products (A) FGT microbiota composition of 180 distinct vaginal swab samples from 106 women, determined by bacterial 16S rRNA gene sequencing (top, stacked barplot) and classified into cervicotypes (CTs) as previously described.^, Middle and bottom bar plots show relative concentrations of h18:0 (10-HSA) and h18:1, respectively, in paired cervicovaginal lavage (CVL) samples. The top colorbar shows Nugent score-based BV status. (B) h18:0 (top) and h18:1 (bottom) concentrations within each CT for the samples in (A). Significance determined by one-way ANOVA with post hoc Tukey’s test; selected pairwise differences are shown (^∗∗∗∗p < 0.0001; full statistical results in Table S3). Boxplots represent the 25^th and 75^th percentiles (lower and upper boundaries of boxes, respectively), the median (middle horizontal line), and measurements that fall within 1.5 times the IQR (whiskers). (C) Change in relative h18:0 (top) and h18:1 (bottom) concentrations within 74 paired serial samples in which microbiota transitioned to CT1 (n = 5), away from CT1 (n = 6), remained CT1 (n = 11), or remained non-CT1 (n = 52). Significance determined by paired t test on log-transformed values (^∗∗p < 0.01; ^∗∗∗p < 0.001; ^∗∗∗∗p < 0.0001; ns: p ≥ 0.05). See also Figure S4.

Figure 5

farE is required for resistance to OA inhibition, and ohyA9 is required for 10-HSA production (A) Relative growth of L. gasseri ATCC 33323 wild-type (WT) and mutant strains, including knockouts of ohyA9 (ΔohyA9) and farE (ΔfarE), and ΔfarE complemented with plasmid-overexpressed farE (ΔfarE/pfarE), cultured in MRS + CQ broth supplemented with varying OA concentrations. (B) Relative growth rescue of L. gasseri WT and mutant strains in lipid-depleted MRS + CQ broth supplemented with varying OA concentrations. (C) MBC assay results for L. gasseri WT and mutant strains in MRS + CQ broth. (D) ¹³C₁₈-10-HSA relative concentrations in blank media and supernatants from L. gasseri WT, ΔohyA9, ΔohyA9 complemented with plasmid-overexpressed ohyA9 (ΔohyA9/pohyA9), ΔfarE, and ΔfarE/pfarE cultured for 24 h in NYCIII broth with or without sub-inhibitory concentrations of ¹³C₁₈-OA (100 μM). (E) Schematic depicting a proposed model for FarE and OhyA9 activity on exogenous cis-9-uLCFAs in non-iners FGT Lactobacillus species. (A–D) Points represent 2–3 technical replicates per condition. (A–C) Results are representative of ≥2 independent experiments. See also Figure S4.

Figure 6

FGT lactobacilli are FA auxotrophs that exploit OA and its OhyA9-dependent derivative 10-HSA for phospholipid synthesis (A) Fatty acid synthesis II (FASII) and phospholipid synthesis pathways, annotated with predicted gene function presence in FGT Lactobacillus genomes. Predicted gene functions are annotated as missing in a species if absent in >50% of genomes. (B) Growth rescue of L. crispatus (n = 19), L. gasseri (n = 3), L. iners (n = 13), L. jensenii (n = 8), and L. mulieris (n = 5) strains in lipid-depleted MRS + CQ broth supplemented with acetate or OA (3.2 mM each), cultured for 72 h. (C) Phosphatidylglycerol (PG) profiles in cell pellets of L. crispatus (top) and L. iners (bottom) cultured for 72 h in NYCIII broth with no added OA (left) or supplemented with 100 μM (middle) or 3.2 mM (right, L. crispatus only) unlabeled OA (black) or ¹³C₁₈-OA (red). Plots depict representative MS1 spectra. Predominant unlabeled isotopologs of major PG species (black) and the differences in mass/charge (m/z) ratio of their corresponding ¹³C-labeled isotopologs (red) are annotated. (D) Growth rescue of non-iners FGT Lactobacillus (n = 17) and L. iners (n = 5) strains in lipid-depleted MRS + CQ broth supplemented with varying concentrations of OA (left) or 10-HSA (right), cultured for 72 h. (E) Growth rescue of L. gasseri ATCC 33323 WT and mutant strains in lipid-depleted MRS + CQ broth supplemented with varying concentrations of 10-HSA, cultured for 24 h. (F) Detection of PG lipids in pellets from L. gasseri WT (top), ΔohyA9 (middle), and ΔohyA9/pohyA9 (bottom), cultured for 24 h in lipid-depleted MRS + CQ broth containing 50 μM ¹³C₁₈-OA with (right) or without (left) 400 μM unlabeled 10-HSA (growth shown in Figure S5G). Plots depict representative MS1 spectra with major ¹³C-labeled (black) and partially labeled or unlabeled (red) PG species annotated. (G) L. gasseri genetic mutants strains, ΔohyA9 and ΔohyA9/pohyA9, were co-cultured in lipid-depleted MRS + CQ broth supplemented with varying concentrations of OA and 10-HSA, but no erythromycin selection. After 18 h, the ratio of CFU on MRS agar plates with and without erythromycin (respectively representing the ΔohyA9/pohyA9 CFU relative to the total CFU) and relative growth (Figure S5I) were determined. The dotted line represents the input CFU ratio. (B, D, and E) Relative growth rescue was calculated as growth relative to median OD600 measurement in non-lipid-depleted MRS + CQ broth. (B and D) Points represent median relative growth for 3 technical replicates per condition. Boxplots represent the 25^th and 75^th percentiles (lower and upper boundaries of boxes, respectively), the median (middle horizontal line), and measurements that fall within 1.5 times the IQR (whiskers). (E) Points represent 3 technical replicates per condition and are representative of ≥2 independent experiments. See also Figure S5.

Figure S5

Genomic analysis of FASII pathway in FGT Lactobacillus genomes, OA isotope tracing in cultured FGT lactobacilli, and 10-HSA growth effects, related to Figure 6 (A) Presence of gene functions predicted to encode FASII pathway genes in isolate genomes and MAGs of the indicated FGT Lactobacillus species (n = 1,167). (B) Relative growth of diverse L. crispatus (n = 19), L. gasseri (n = 3), L. iners (n = 13), L. jensenii (n = 8), and L. mulieris (n = 5) strains in MRS + CQ broth supplemented with 3.2 mM acetate or 3.2 mM OA. Growth was measured by OD600 after 72 h of culture. (C) Heatmap representing the median incorporation ratio of ¹³C₁₈-OA in detected diglycerides and central metabolites involved in the tricarboxylic acid (TCA) cycle in cell pellets from representative strains of FGT Lactobacillus species. Bacteria were cultured for 72 h in NYCIII broth with 3.2 mM (top) or 100 μM (bottom). ¹³C₁₈-OA incorporation ratio was calculated as the signal from the detected ¹³C-labeled metabolite relative to the signal of the detected unlabeled metabolite. (D) Heatmap representing the normalized signal of detected unlabeled and labeled phosphatidylglycerol in cell pellets from representative strains of FGT Lactobacillus species. Bacteria were cultured for 72 h in NYCIII broth with 3.2 mM unlabeled OA (left) or ¹³C₁₈-OA (right). Each row labeled A, B, and C represents a replicate culture of the indicated condition. (E) Relative growth of L. gasseri WT and mutant strains in MRS + CQ broth supplemented with varying concentrations of 10-HSA. Growth was measured by OD600 after 24 h of culture. (F) Relative growth of L. crispatus (n = 3) and L. iners (n = 4) strains in MRS + CQ broth supplemented with varying concentrations of OA (left) or 10-HSA (right). Growth was measured by OD600 after 72 h of culture. (G) L. gasseri genetic mutant strains (WT, ΔohyA9, and ΔohyA9/pohyA9) were grown for 24 h in lipid-depleted MRS + CQ broth supplemented with ¹³C₁₈-OA concentration (50 μM) alone and in combination with unlabeled 10-HSA (400 μM, corresponding to the isotopic tracing data in Figure 6F). Relative growth was calculated relative to the median OD600 measurement in non-lipid-depleted MRS + CQ broth. (H) L. gasseri ΔohyA9/pohyA9 was grown in lipid-depleted MRS + CQ broth supplemented with varying concentrations of OA and 10-HSA, but without erythromycin selection (the pohyA9 plasmid encodes erythromycin resistance). After 18 h of culture, the ratio of CFU on MRS agar plates with and without erythromycin (respectively representing the amount of the viable ΔohyA9/pohyA9 strain relative to total bacteria) was determined. (I) Growth (determined by OD600) of L. gasseri genetic mutants strains ΔohyA9 and ΔohyA9/pohyA9 grown in mono-culture or mixed and co-cultured in lipid-depleted MRS + CQ broth supplemented with varying concentrations of OA and 10-HSA, without erythromycin selection. Relative growth was calculated relative to the median OD600 measurement in the 100 μM HSA: 100 μM OA condition. The cultures correspond to the experiment shown in Figure 6G. (B, E, and F) Relative growth was calculated relative to the median OD600 measurement in the no supplementation control. (B and E) Boxplots represent the 25^th and 75^th percentiles (lower and upper boundaries of boxes, respectively), the median (middle horizontal line), and measurements that fall within 1.5 times the IQR (whiskers). (B and F) Points represent the median of 3 technical replicates per condition. (E) Points represent 3 technical replicates per condition.

Figure 7

OA and 10-HSA treatments (with or without MTZ) shift in vitro BV-like communities toward L. crispatus dominance (A) Relative growth of the indicated species in NYCIII broth with or without metronidazole (MTZ; 50 μg/mL) and/or OA (3.2 mM), cultured for 72 h. Points represent 3 technical replicates per condition. (B) Relative bacterial abundance in defined BV-like communities grown for 72 h in NYCIII broth with or without MTZ (50 μg/mL) and/or OA (3.2 mM). (C) Ratios of L. crispatus to the sum of all other taxa in the communities in (B). (D) Relative bacterial abundance in a defined BV-like community grown for 72 h in NYCIII broth with OA (3.2 mM) or 10-HSA (1.6 mM), with or without metronidazole (MTZ; 50 μg/mL). (E) Ratios of L. crispatus to the sum of all other taxa in the communities in (D). (B–E) Plots depict 6 technical replicates per condition. (B and D) Compositions of the cultured communities and input mixture (T0) were determined by 16S rRNA gene sequencing. (C and E) Dotted lines represent the ratios in the input mixtures (T0). Boxplots represent the 25^th and 75^th percentiles (lower and upper boundaries of boxes, respectively), the median (middle horizontal line), and measurements that fall within 1.5 times the IQR (whiskers). Between-group differences were determined by one-way ANOVA with post hoc Tukey’s test; selected significant pairwise differences are shown (ns: not significant; ^∗p ≤ 0.05; ^∗∗∗p < 0.001; ^∗∗∗∗p < 0.0001; full statistical results in Table S3). See also Figures S6 and S7.

Figure S6

Effects of OA and MTZ on growth of FGT bacterial species and defined, in vitro BV-like bacterial community experiment controls, related to Figure 7 (A) Relative growth of representative L. gasseri and L. jensenii strains in NYCIII broth with or without MTZ (50 μg/mL) and/or OA (3.2 mM). (B) Relative growth of the indicated bacterial species in S-broth with or without MTZ (50 μg/mL) and/or OA (3.2 mM). (C) Relative growth of representative F. vaginae, G. piotii, G. swindsinskii-leopoldii, G. vaginalis (2 strains), P. amnii, P. bivia, P. timonensis, and S. vaginalis strains in NYCIII broth with varying concentrations of OA. Points represent median relative growth for 3 technical replicates per condition. (D) Relative growth of the indicated bacterial species in NYCIII broth with OA (3.2 mM) or 10-HSA (1.6 mM), each alone or in combination with metronidazole (MTZ; 50 μg/mL), after 72 h of culture. (E) Cultures of P. bivia and P. timonensis were grown to mid- to late-log phase and then exposed to varying concentrations of OA, then ATP release assays were performed. (F) Total sequencing read counts from mock BV-like community experiments per sample type, including extraction controls (extract_ctrl), blank media controls (media_ctrl), community samples (C1–C6 corresponding to samples from Figures 7B, 7C, S7A, and S7B), input bacterial isolate mono-cultures (isolate_T0), 72-h cultured bacterial isolate mono-cultures (isolate_T2), and no template PCR controls (PCR_ctrl). (G) Taxonomic composition of bacterial isolate mono-cultures used to construct defined communities in Figures 7B, 7C, S7A, and S7B, determined by 16S rRNA gene sequencing to confirm mono-culture purity. Plot shows data for input (top) isolates and for the corresponding 72-h mono-culture controls (bottom). (H) OD600 for each total community (C1–C6) at 0 and 72 h of culture in NYCIII broth (top) and S-broth (bottom), corresponding to community cultures in Figures 7B, 7C, S7A, and S7B. (I) Total sequencing read counts from mock BV-like community experiments per sample type, including extraction controls (extract_ctrl), blank media controls (media_ctrl), community samples (C1 corresponding to samples from Figures 7D and 7E; C1P1–3 corresponding to samples from Figure S7C), input bacterial isolate mono-cultures (isolate_T0), and cultured bacterial isolate mono-cultures (isolate_T1, isolate_P2, and isolate_P3). (J) Taxonomic composition of bacterial isolate mono-cultures used to construct the defined communities in Figures 7D, 7E, and S7C, determined by 16S rRNA gene sequencing to confirm mono-culture purity. Plot shows data for input (top) isolates and the corresponding passaged cultures (passage 1–3), each cultured for 48 h (bottom). (K) OD600 for each total community (C1 corresponding to Figures 7D and 7E; C1P1–3 corresponding to Figure S7C) at 0 h for all communities, 72 h for C1, and 48 h for C1P1–3 of culture in NYCIII broth. (A–D) Growth was measured by OD600 after 72 h of culture. Relative growth was calculated relative to median OD600 measurement in the no-additive control. (A–E) Points represent 3 technical replicates per condition. (F and I) Boxplots represent the 25^th and 75^th percentiles (lower and upper boundaries of boxes, respectively), the median (middle horizontal line), and measurements that fall within 1.5 times the IQR (whiskers).

Figure S7

Community compositions for experiments in S-broth and OA-shifted L. crispatus dominance is stable over multiple passages in vitro, related to Figure 7 (A) Relative bacterial abundance in defined BV-like communities grown for 72 h in NYCIII broth or S-broth with or without MTZ (50 μg/mL) and/or OA (3.2 mM). Composition of the cultured communities and of the input mixtures (T0) was determined by bacterial 16S rRNA gene sequencing (see also Figures 7B and 7C). Plots depict 6 technical replicates per condition. Sequencing reads were not recovered from a single technical replicate of community 5 cultured in S-broth with MTZ and OA (due to a failed PCR reaction). This replicate is marked with an “x” in the plot, and its corresponding low read count is shown in Figure S6F. (B) Ratios of non-iners FGT Lactobacillus species taxa to the sum of all other taxa in the mock communities shown in (A). The gray dotted lines represent the ratios measured in the input inocula (T0). Boxplots represent the 25^th and 75^th percentiles (lower and upper boundaries of boxes, respectively), the median (middle horizonal line), and measurements that fall within 1.5 times the IQR (whiskers). Between-group differences were determined by one-way ANOVA with post hoc Tukey’s test; selected significant pairwise differences are shown (^∗∗∗p < 0.001, ^∗∗∗∗p < 0.0001; full statistical results in Table S3). (C) Relative bacterial abundance in a defined BV-like community grown for 48 h in NYCIII broth with or without OA (3.2 mM) alone or in combination with MTZ (50 μg/mL; “P1”). After 48 h of culture, 2.5% v/v of each technical replicate was passaged into NYCIII broth without additives (“P2”) and cultured for another 48 h. Each replicate culture was then passaged once more into NYCIII broth without additives (“P3”). Composition of the cultured communities and of the input mixture (T0) was determined by bacterial 16S rRNA gene sequencing. Plots depict 6 technical replicates per condition.