All the 14 Lactobacillus strains included in this study were previously isolated from vaginal swabs of healthy premenopausal Caucasian women (Parolin et al., 2015). The isolation technique, the species identification and the functional characterization of these strains are described elsewhere (Parolin et al., 2015; Foschi et al., 2017; Siroli et al., 2017).

In particular, 7 strains of L. crispatus (BC1, BC3-BC8), 5 strains of L. gasseri (BC9, BC10, BC12-BC14), and 2 strains of L. vaginalis (BC16, BC17) were employed.

Lactobacilli were cultured in de Man, Rogosa, and Sharpe (MRS) broth (Becton Dickinson and Company, Sparks, MD, USA) supplemented with 0.05% L-cysteine, incubated at 37°C. Anaerobic conditions were achieved by using anaerobic jars supplemented with GazPack EZ (Becton Dickinson and Company).

The turbidity of 18-h lactobacilli cultures was measured by a McFarland (McF) densitometer (DEN 1-B, BioScientifica, Italy), considering that a turbidity of 0.5 McF corresponds approximately to a cell concentration of 1.5 × 10⁸ colony forming unit (CFU)/mL. Lactobacilli cultures were adjusted to 1.6 McF with sterile saline (cell concentration: 5 × 10⁸ CFU/mL) and centrifuged at 5,000 × g for 10 min at 4°C. Supernatants were filtered through a 0.22 μm membrane filter to obtain stock cell free supernatants (CFS) and used for “inhibition” experiments. The pH values of CFS were measured by a pH Meter, after appropriate calibration (Table 1). Cell pellets were washed and re-suspended in phosphate-buffered saline (PBS) to obtain stock suspensions of 5 × 10⁸ CFU/mL for “inhibition” experiments.

Neisseria gonorrhoeae MS11 (ATCC BAA-1833) (Schoolnik et al., 1984) was grown at 37°C in a humidified 5% CO₂ environment on Thayer Martin-VCNT (VCNT = vancomycin, colistin methane sulfonate, nystatine, and trimethoprim) agar plates (MEUS SRL, Piove di Sacco, Padova, Italy). Piliation and Opa phenotype were distinguished by colony morphology under a stereomicroscope (Swanson et al., 1971; Swanson, 1978; Bergström et al., 1986) and only piliated Opa-negative variants were used.

To exclude a significant reduction of GC viability due to a spontaneous autolysis, the survival rate of gonococci in control conditions (e.g., PBS, MRS medium) was checked prior to each experiment, proving to be comprised between 75 and 85%.

“Inhibition” experiments were carried out incubating 1 × 10⁸ lactobacilli cells or the corresponding supernatants with the same amounts of gonococcal cells (ratio gonococci: lactobacilli = 1), in order to evaluate a direct effect of lactobacilli fractions on GC viability. In details, 1 × 10⁸ gonococcal cells (200 μL of a PBS stock solution with a concentration of 5 × 10⁸ CFU/mL) were pelleted in an eppendorf tube and the supernatant was discarded. Afterward, 200 μL of the stock suspensions of lactobacilli cells or CFS were added and the mixes were incubated for 7 or 60 min at 37°C in 5% CO₂ atmosphere. At the end of the incubation time, the tubes were centrifuged at 20,000 × g for 10 min. Pellets were re-suspended in PBS, serially-diluted (1:50, 1:100, 1:200, 1:500), and 10 μL of the diluted suspensions were seeded on Thayer-Martin agar plates. After 48 h of incubation (37°C, 5% CO₂), gonococcal colonies were visually counted and expressed in number of gonococci/mL. Each experiment was conducted in duplicate and a control tube (1 × 10⁸ gonococcal cells incubated with 200 μL of PBS or MRS, without lactobacilli fractions) was always included.

Results were expressed as the percentage of the average number of gonococci/mL. The gonococcal growth in the presence of lactobacilli fractions was compared with the corresponding control (gonococcal growth in the absence of lactobacilli fractions), taken as 100%.

Finally, in order to verify the viability of lactobacilli cells after the interaction with gonococci, in case of cell pellets experiments, 10 μL of the same dilutions were seeded in parallel on MRS agar plates and incubated at 37°C in anaerobic conditions. After 48 h, lactobacilli colonies were numbered and compared with the growth of lactobacilli incubated without gonococci.

Solutions of lactic acid and hydrochloric acid at a concentration of 10 mM in 150 mM NaCl and buffered to different pH values (3.4, 3.7, 4.0, 4.4, 4.7, 5.0, 5.3) with NaOH 1M were used for “inhibition” experiments against GC, as described above. One hundred and fifty millimolars of NaCl solution (pH 5.8) was used as control. Similarly, gonococcal inhibition experiments were performed with MRS broth buffered to different pH values (3.4, 3.7, 4.0, 4.4, 4.7, 5.0, 5.3) with lactic acid and hydrochloric acid solutions (molar range of acids: 5–50 mM), using MRS broth (pH: 6.0) as control.

The supernatants of highly active Lactobacillus strains (BC1, BC4, BC5, BC6, BC7, BC8, BC13) were buffered to pH 6.0 with NaOH 1M and “inhibition” experiments were conducted in the same conditions described above.

L. crispatus BC1 and BC3 suspensions (5 × 10⁸ lactobacilli CFU/mL) were divided into two 200 μL-aliquots. Aliquots were treated with 500 μL of ice-cold methanol for 10 min and then centrifuged at 14,000 × g for 3 min to pellet the bacteria. The supernatant was removed and replaced with 200 μl of sterile saline (aliquot 1, methanol treated) or 160 μl of saline and 40 μl proteinase K (20 mg/ml) (aliquot 2, proteinase K treated). After an incubation for 2 h at 37°C, aliquots were boiled for 3 min and then centrifuged at 14,000 × g for 3 min to pellet the bacteria. The supernatants were removed and lactobacilli were directly re-suspended in 200 μl of a suspension of 5 × 10⁸ gonococci/mL, then “inhibition” experiments were carried out as previously described. Microscopic visualization of Gram-stained samples was performed to evaluate the effect of methanol and proteinase K on lactobacilli.

Released surface-associated components were isolated from BC1 and BC10 lactobacilli strains, following previous published protocol with slight modifications (Spurbeck and Arvidson, 2010; Sambanthamoorthy et al., 2014). Considering the cell pellet activity against GC, BC1 and BC10 lactobacilli were chosen as examples of a highly active and a non-active strain, respectively.

Briefly, a 1-liter culture of lactobacilli was grown for ~48 h in MRS broth to a turbidity of 8–10 McF, and the bacteria were harvested by centrifugation (10,000 × g for 10 min at 4°C). The bacterial cell pellet was washed twice with sterile water and then re-suspended in 100 mL of PBS. This suspension was gently stirred for 2 h at room temperature to release cell-bound bio-molecules and then centrifuged for 20 min at 3,000 × g. The supernatant was filtered through a 0.22 μm filter, to remove the remaining bacteria. This cell-free preparation was concentrated to ~10 ml by Amicon ultraconcentration using a 10,000-molecular-weight-cutoff filter membrane. The pH values of RSC were measured by a pH Meter, after appropriate calibration.

The biosurfactant activity of the RSC was evaluated by performing an oil spreading assay, as described elsewhere (Sambanthamoorthy et al., 2014). Briefly, 50 mL of demineralized water were poured into a 150 mm Petri dish and 20 μL of motor oil were added to the surface of the water. Ten microliters of RSC from either L. crispatus BC1 and L. gasseri BC10 were then added to the surface of the oil. Water was used as a negative control. The diameters of clear zones of triplicate assays from the same sample were determined.

RSC were finally employed in “inhibition” experiments, as previously described. Serial dilutions of BC1 and BC10 RSC were tested (1:1, 1:2, 1:4), in order to verify a dose-response effect. Gonococcal cells (1 × 10⁸) incubated in PBS were used as control.

Lactobacilli were screened for their capacity to interact and aggregate with piliated gonococci: 1 × 10⁸ gonococcal cells and 1 × 10⁸ lactobacilli CFU were mixed together in PBS in a final volume of 200 μl and incubated for 7 or 60 min at 37°C. Afterwards, an aliquot of the suspension was fixed on a microscope slide and Gram stained, in order to evaluate the bacterial spatial arrangement.

To strengthen the hypothesis of a specific interaction between lactobacilli and gonococci, a “highly aggregating” Lactobacillus strain (L. crispatus BC1) was used as a model for additional experiments. In particular, after 60 min contact time between BC1 and GC and prior to the Gram-staining, the bacterial pellet was subjected to 2 times PBS washing and low speed-centrifugation (1,200 × g for 10 min), to remove lactobacilli not specifically bound to gonococcal cells.

All statistical analysis were performed by using GraphPad Prism software (GraphPad Prism version 5.02 for Windows, GraphPad Software, San Diego California USA, www.graphpad.com). Statistical analysis for “inhibition” assays were performed by using one-way analysis of variance (ANOVA) test, followed by Dunnett's Multiple Comparison test. Differences in the anti-GC activity between lactic acid and hydrochloric acid were searched by means of a paired t-test. Results were expressed as mean ± Standard Error of the Mean (SEM). Statistical significance was determined at ^*P < 0.05, ^**P < 0.01, and ^***P < 0.0001.

In order to explore the anti-gonococcal activity of bio-molecules secreted by vaginal Lactobacillus strains, “inhibition” experiments were carried out with lactobacilli culture supernatants. The results obtained with CFS are shown in Figure 1. L. crispatus BC1, BC4, BC5, BC6, BC7, BC8 and L. gasseri BC13 CFS were able to completely abolish GC viability both at 7 and 60 min contact times. L. crispatus BC3, L. gasseri BC9, BC12, BC14, and L. vaginalis BC17 caused a 100%-reduction of GC viability only after 60 min, whereas L. gasseri BC10 and L. vaginalis BC16 were unable to fully prevent GC growth at both contact times, although a significant reduction was observed at 60 min. Globally, L. crispatus species showed the best antimicrobial profile, considering that all the lactobacilli supernatants of that species reduced GC viability, in a highly significant manner, both at 7 and 60 min.

As shown in Table 1, the effect of lactobacilli supernatants on GC viability was strictly related to the pH values and time of contact. Indeed, in case of pH values < 4.0 lactobacilli supernatants completely abolished GC growth at both contact times, whereas for pH values ranging from 4.0 to 4.4, GC was totally killed only at 60 min. Finally, when supernatant pH values were ≥ 4.5, no complete reduction of GC viability was observed at any time of contact.

On the basis of the results of “inhibition” experiments with lactobacilli CFS, the same assays were carried out with lactic acid and hydrochloric acid solutions, buffered at different pH values, in order to evaluate the effect of organic/inorganic acids on GC viability. The concentration of acids (10 mM) was chosen because corresponding to the mean titer of lactate found in Lactobacillus CFS, as reported elsewhere (Nardini et al., 2016).

As shown in Figure 2, for pH values ≤ 4.0, a complete abolishment of GC viability was noticed both for hydrochloric acid and for lactic acid solutions. Differently, in case of pH values in the range 4.4–5.3, lactic acid showed a higher anti-GC activity compared to hydrochloric acid, both at 7 and 60 min contact time. Globally, considering all the pH values and both the contact times, lactic acid was significantly more active than hydrochloric acid in reducing GC viability (P = 0.003).

Comparable results were obtained when “inhibition” experiments were carried out with MRS broth buffered to different pH values with lactic or hydrochloric acid (Figure S1).

To verify the central role of the acidic environment in counteracting GC viability and to exclude the involvement of other bioactive substances (e.g., H₂O₂) in anti-gonococcus activity, we performed “inhibition” experiments with the CSF of Lactobacillus active strains, buffered to pH 6.0.

Compared to controls, the lactobacilli supernatants buffered at pH values of 6.0 led to a not-significant reduction of GC viability, both at 7 and 60 min contact times (Figure S2).

The capacity of lactobacilli surface components to reduce GC viability was investigated by means of “inhibition” experiments with 1 × 10⁸ lactobacilli. The results of “inhibition” experiments with lactobacilli are shown in Figure 3. Globally, cell pellets were far less effective compared to the corresponding supernatants, in particular at short contact times (7 min). In fact, no cell pellet was able to completely abolish GC viability at any time point. L. crispatus BC1, BC3 and BC7 were the only strains able to significantly reduce GC growth at 7 min, whereas all the lactobacilli, except for L. gasseri BC10 and BC13 and L. vaginalis BC16, were effective at 60 min, although at various levels. Overall, L. crispatus species showed the best activity against GC, in agreement with the results obtained with lactobacilli CFS. Lactobacilli cell viability was completely retained after the interaction with GC cells, at both contact times (data not shown).

“Inhibition” experiments were also carried out with lactobacilli cells treated with methanol and proteinase K prior to the interaction with gonococci. These assays were performed to evaluate if lactobacilli surface components were involved in GC inhibition and, finally, to determine the chemical nature of such components. On the basis of the anti-gonococcus activity of cell pellets, L. crispatus BC1 and BC3 were chosen as examples of highly active strains. After methanol and proteinase K treatment of BC1 and BC3 cells, the lactobacilli completely lost their anti-gonococcal activity, indicating that lactobacilli lipidic/proteic surface components are involved in the inhibition and that viable lactobacilli are needed for the GC killing (Figure S3).

Microscopic visualization of Gram-stained samples confirmed that methanol and proteinase K treatment left the lactobacilli intact (data not shown).

In order to evaluate the role of released surface-associated components in the inhibition of GC showed by cellular fraction, RSC were isolated from BC1 and BC10 lactobacilli strains. Considering their activity against GC, these two strains were chosen as examples of a highly active and a non-active strain, respectively.

L. crispatus BC1 RSC showed a pH value of 5.5 and possessed a high biosurfactant activity, evaluated by the oil spreading assay (diameter of the clear zone on the oil-water surface: 14 ± 0.5 mm), whereas BC10 RSC were characterized by a pH value of 6.4 and by a lower biosurfactant activity (diameter of the clear zone on the oil-water surface: 8 ± 0.1 mm).

The effect of lactobacilli RSC on GC viability is shown in details in Figure 4. BC1-derived RSC were able to abolish completely GC viability both at short (7 min) and longer contact times (60 min), when tested as a whole. When diluted 1:2, the RSC retained a partial activity against GC (15%-reduction of GC viability at 7 min and 50%-reduction after 60 min), whereas in case of 1:4 dilution, the GC-killing activity was lost.

On the other hand, BC10 RSC were less effective compared to BC1 in reducing GC viability: after 7 min, a 50%-reduction of GC viability was found, whereas at 60 min contact time, the reduction reached the 70%. When diluted (both 1:2 and 1:4) BC10 RSC completely lost their activity against GC, at both contact times.

The capacity of lactobacilli to aggregate with piliated gonococci was screened, in order to evaluate if the killing effect displayed by lactobacilli cells could be associated to the ability to physically interact with gonococci.

At short contact times, L. crispatus BC1 and BC3 showed the highest levels of co-aggregation with GC, indeed chains of lactobacilli were incorporated into GC cells. On the contrary, for the remaining Lactobacillus strains, the ability to aggregate was partially or totally excluded. Figure 5 shows some examples of lactobacilli-GC interaction after 7 min contact time.

After an interaction of 60 min, all the strains belonging to L. crispatus species, as well as some strains of the other species (L. gasseri BC9 and BC14, L. vaginalis BC17) were able to aggregate with piliated gonococci, often arranged in microcolonies. Figure S4 shows Gram-stained examples of the interaction between lactobacilli (BC1 and BC10) and GC, after 60 min contact time. It is worth to be underlined that the ability of BC1 to aggregate with gonococcal cells was retained even after repeated cycles of washes and centrifugation, thus corroborating the hypothesis of a specific bacterial interaction (Figure S5).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.