Lactobacillus crispatus inhibits growth of Gardnerella vaginalis and Neisseria gonorrhoeae on a porcine vaginal mucosa model

Abstract

Background: The vaginal microbiota can impact the susceptibility of women to bacterial vaginosis (BV) and sexually transmitted infections (STIs). BV is characterized by depletion of Lactobacillus spp., an overgrowth of anaerobes (often dominated by Gardnerella vaginalis) and a pH > 4.5. BV is associated with an increased risk of acquiring STIs such as chlamydia and gonorrhea. While these associations have been identified, the molecular mechanism(s) driving the risk of infections are unknown. An ex vivo porcine vaginal mucosal model (PVM) was developed to explore the mechanistic role of Lactobacillus spp. in affecting colonization by G. vaginalis and Neisseria gonorrhoeae.

Results: The data presented here demonstrate that all organisms tested can colonize and grow on PVM to clinically relevant densities. Additionally, G. vaginalis and N. gonorrhoeae form biofilms on PVM. It was observed that lactic acid, acetic acid, and hydrochloric acid inhibit the growth of G. vaginalis on PVM in a pH-dependent manner. N. gonorrhoeae grows best in the presence of lactic acid at pH 5.5, but did not grow well at this pH in the presence of acetic acid. Finally, a clinical Lactobacillus crispatus isolate (24-9-7) produces lactic acid and inhibits growth of both G. vaginalis and N. gonorrhoeae on PVM.

Conclusions: These data reveal differences in the effects of pH, various acids and L. crispatus on the growth of G. vaginalis and N. gonorrhoeae on a live vaginal mucosal surface. The PVM is a useful model for studying the interactions of commensal vaginal microbes with pathogens and the mechanisms of biofilm formation on the vaginal mucosa.

Fig. 1

Growth of human clinical isolates on porcine vaginal mucosa (PVM). a PVM is obtained as large specimens (i) and 5 mm explants are trimmed (ii) and placed mucosal side up in transwells over liquid media (iii, iv). “Air” refers to either an aerobic or anaerobic environment depending on the experiment being performed. b L. crispatus, c G. vaginalis, and d N. gonorrhoeae were inoculated onto PVM explants at ~ 10⁴ CFU/explant (dotted lines). Explants were processed for CFU/explant at each time point to evaluate bacterial growth. Data was log₁₀ transformed and plotted on a log scale as the mean ± SD

Fig. 2

G. vaginalis and N. gonorrhoeae form biofilm on PVM. Strains were inoculated onto PVM explants at ~ 10⁴ CFU/explant and processed for microscopy at indicated times. The LIVE/DEAD stain allows for imaging of both the mucosal epithelium and bacteria. Green cells are alive while red cells are dead. a-d, i-l Uncolonized control (CNTL) tissue remains healthy throughout both experiments as evidenced by large green intact epithelial cells. e-h By 48 h post-colonization, G. vaginalis (GV) forms a patchy biofilm that persists and spreads over time (anaerobic growth). m-p In just 24 h N. gonorrhoeae (NG) forms a robust biofilm, which thickens and persists over time (aerobic growth). Epithelial cells that can be seen under and around the NG biofilm are alive as evidenced by their green staining, while those on GV-colonized explants are dead (large red cells under the green biofilm). Scale bars = 100 μm for all images

Fig. 3

Various acids inhibit growth of G. vaginalis and N. gonorrhoeae on PVM. RPMI was pH-adjusted with lactic acid, acetic acid or hydrochloric acid (HCl) and placed under transwells containing PVM explants. Bacteria were enumerated 48 h post-colonization. L. crispatus and G. vaginalis were incubated anaerobically, while N. gonorrhoeae was incubated aerobically. a The low pH produced by each of the three acids has no effect on growth of L. crispatus. b G. vaginalis growth is inhibited at pH 4.0 in the presence of all three acids. c N. gonorrhoeae growth peaked at pH 5.5 in the presence of lactic acid and HCl, while growth was inhibited at pH 5.5 in the presence of acetic acid. Asterisks indicate significant difference from inoculum (dotted lines) (p < 0.0001)

Fig. 4

Unbuffered RPMI does not support sufficient pH changes in the presence of L. crispatus on PVM. Unbuffered RPMI was placed under transwells and relevant explants were inoculated with L. crispatus. After 48 h, relevant explants were inoculated with G. vaginalis. Experiments were performed under anaerobic conditions and the pH of media below explants was monitored over time. a Media under uncolonized controls (CNTL) and those colonized with G. vaginalis (GV) alone reached pH 6.5 while those colonized with L. crispatus (LC) alone or co-colonized with both organisms (LC + GV) reached pH 6.0 (N = 4, n = 1). b The presence of L. crispatus had a statistically significant (asterisk, p < 0.016), yet minor effect on G. vaginalis growth. Dotted line represents GV inoculum

Fig. 5

Conditioned media supports L. crispatus-induced pH changes, inhibiting growth of G. vaginalis on PVM. Media from overnight aerobic broth culture of L. crispatus was sterile filtered and mixed with unbuffered RPMI to achieve pH 5.5. This conditioned media (CM) was placed under transwells and relevant explants were inoculated with L. crispatus. After 48 h, relevant explants were inoculated with G. vaginalis. Experiments were performed under anaerobic conditions and the pH of media below explants was monitored over time. a Media under uncolonized controls (CNTL) and those colonized with G. vaginalis (GV) alone remained at pH 5.0 – 5.5 while those colonized with L. crispatus (LC) alone or co-colonized with both organisms (LC + GV) were reduced to pH 4.0 (N = 6, n = 1). b The presence of L. crispatus inhibited G. vaginalis growth. Dotted line represents GV inoculum. Asterisk indicates significant difference in growth between the two groups shown (p < 0.0001). c, d D- and L-lactate levels in the CM below transwells from experiment (a) were analyzed at 96 h. CM below explants colonized with L. crispatus (LC and LC + GV) showed a significant increase in lactic acid when compared with uncolonized controls (CM) and G. vaginalis (GV) alone (p < 0.0001). e A comparison of D- and L- lactate levels in CM below LC alone infections shows that D-lactate is produced at ~2X the level of L-lactate (p < 0.002). (C-E) Data shown are combined (N = 5, n = 1)

Fig. 6

L. crispatus inhibits growth of N. gonorrhoeae. a, b Unbuffered RPMI or (c) L. crispatus CM was adjusted to pH 5.5 and placed under transwells. Relevant explants were inoculated with L. crispatus. After 48 h, relevant explants were inoculated with N. gonorrhoeae. Experiments were performed under aerobic conditions and the pH of media below explants was monitored over time. a RPMI under uncolonized controls (CNTL) and those colonized with N. gonorrhoeae (NG) alone remained at pH 5.5 while RPMI under explants colonized with L. crispatus (LC) alone or co-colonized with both organisms (LC + NG) was reduced to pH 4.5 (N = 4, n = 1). b The presence of L. crispatus inhibited 48 h N. gonorrhoeae growth. Asterisk indicates significant difference in growth between the two groups shown (p < 0.002). c CM under CNTL or NG remains at pH 5.5 while CM under explants colonized with LC alone or co-colonized with both organisms (LC + NG) was reduced to pH 4.0 (N = 3, n = 1). d While N. gonorrhoeae did exhibit a growth defect over NYC + RPMI media (the background media for CM) it was completely killed by aerobic or anaerobic CM at pH 5.5, even in the absence of a L. crispatus co-infection. Asterisks indicate significant difference from inoculum (p < 0.0001). For (b) and (d), dotted lines represent NG inocula