Introduction

The vaginal microbiota is colonized by several microorganisms [1], where commensal Lactobacillus species act as defense mechanism [2]. Lactobacilli can adhere and biosynthesize antimicrobial compounds reducing colonization by pathogens [1] associated with bacterial vaginosis, aerobic vaginitis, and candidiasis [3]. Thus, probiotics could be an efficient alternative to antimicrobial treatment, which reduces commensal microbiota and increases resistance [4, 5]. Lactobacilli may restore healthy microbiota, as postulated by Mitrea and colleagues [5]. Although Candida spp. is commensal, this genus can become an opportunistic pathogen in high levels [6, 7] leading to vulvovaginal candidiasis and evolving in more serious urinary tract infections and venereal diseases [8, 9]. Studies reported the application of lactobacilli biofilms and biosurfactants against pathogens [10,11,12,13]. However, these approaches are designed to treat established infections and do not endure in vaginal microbiota [14]. Another approach could be the colonization of the mucosal epithelia by new and more probiotic lactobacilli [15], allowing a permanent integration in the microbiota. However, the inhibition of the initial adhesion of pathogens by lactobacilli is not fully understood [16, 17]. It is important to compare the variability of vaginal and non-vaginal lactobacilli to inhibit the adhesion of pathogens and to avoid their displacement. This study evaluated the probiotic ability of three vaginal L. gasseri and one non-vaginal L. plantarum to protect an abiotic surface in initial adhesion assays, assessing the lactobacilli displacement and the inhibition of three C. albicans through different scenarios of dysbiosis conditions, candidiasis, and healthy vaginal microbiota.

Main text

Methods

From previous studies [3, 18], three Lactobacillus gasseri (H59.2, IMAUFB014, and JCM1131), two Candida albicans (one isolate from a healthy vaginal microbiota, and another from candidiasis), and C. albicans ATCC10231 and L. plantarum ATCC14917 were used in this study. Lactobacillus strains were grown in Man, Rogosa and Sharpe agar for 48 h at 37 °C under microaerophilic conditions [19, 20]. C. albicans strains were grown in Sabouraud Dextrose Agar at 37 °C for 18 h [19,20,21]. Brain Heart Infusion (BHI) broth was used for initial adhesion assays [20].

Initial adhesion assays

Each microorganism was concentrated in 5 ml of sterile phosphate-buffered saline (PBS) solution. Both suspensions were collected by centrifugation (4000 g, 12 min, at room temperature), and washed twice with PBS. The pellet was resuspended according to the growth curves to 1.0E + 03 colony-forming unit (CFU)/ml and 1.00E + 09 CFU/ml (Additional files 1 and 2) by optical density at 600 nm (OD600). Four experimental settings (ES) were made from concentration combinations (Additional file 3), varying on low levels of lactobacilli (1.00E + 03 CFU/ml) against low and high levels of C. albicans (ES1 and ES2, respectively) and then on high levels of lactobacilli (1.00E + 09 CFU/ml) against low and high levels of C. albicans (ES3 and ES4, respectively). These ES mimicked dysbiosis conditions (ES1 and ES4), candidiasis (ES2), and healthy vaginal microbiota (ES3). Initial adhesion assays were realized using a preincubation of lactobacilli for 4 h at 37 °C with 120 rpm [22, 23] and then evaluating the initial adhesion of Candida albicans with the pre-adhered lactobacilli during 30 min at same conditions [23,24,25,26], as illustrated in the flowchart (Additional file 4) [27, 28]. Non-adherent microorganisms were removed by PBS washing. All experimental assays were repeated three times on different days.

Microscopy analysis and cell quantification

After adhesion assay, a PBS washing step was carried out on coverslips, which were fixed with ethanol (96%; v/v) and stained with crystal violet at 3% for 1 min [29]. From each coverslip, 15 random fields were photographed in Olympus BX50 microscope under 1000x [23, 30] using the AmScope MU633-FL camera. The number of Lactobacillus spp. and C. albicans were counted from each picture (Additional file 5), being expressed as the number of cells per glass surface ± standard deviation (Additional file 6 and Additional file 7).

Statistical analysis

The statistical analysis was realized through two-tailed ANOVA (ANalysis Of VAriance) with post-hoc Tukey HSD (Honestly Significant Difference) and Student t-test using JASP software version 0.13. ANOVA analysis evaluated differences in and between ES, post-hoc Tukey HSD test analyzed differences between species on the same ES, and Student t-test assessed differences between samples and controls. P values ≤ 0.050 were statistically significant.

Results

Initial adhesion assays were realized through different experimental settings, simulating dysbiosis conditions (ES1 and ES4), candidiasis (ES2), and healthy vaginal microbiota (ES3). The lactobacilli displacement was evaluated on low levels (ES1 and ES2) against low and high concentrations of C. albicans and then on high levels (ES3 and ES4), as shown in Additional file 6. On low levels of lactobacilli, the displacement was between 15 and 99% (see Fig. 1). At ES1, C. albicans ATCC10231 induced bigger displacement of L. gasseri IMAUFB014 (84%; P = 0.010, two‐way ANOVA) and H59.2 (83%; P < 0.001, two‐way ANOVA) showing significant differences among C. albicans (Tukey's post hoc, P < 0.05). Likewise, all C. albicans showed to be statistically different in their displacement ability among the L. gasseri. C. albicans from healthy vaginal microbiota was able to displace 99% of L. gasseri IMAUFB014, while C. albicans isolated from candidiasis demonstrated 99% of displacement against L. gasseri JCM1131. At ES2, no significant differences were found in the displacement among L. gasseri. At ES3, C. albicans isolated from candidiasis showed statistical differences, evidencing a greater ability to displace L. gasseri H59.2 (90%; P < 0.001, two‐way ANOVA). L. gasseri JCM1131 showed only 15% of displacement by C. albicans isolated from candidiasis, being statistically different when compared to C. albicans ATCC10231 (83%; P = 0.001, Tukey's post hoc) and C. albicans isolated from healthy vaginal microbiota (84%; P < 0.001, Tukey's post hoc). At ES4, L. gasseri JCM1131 showed 65% of displacement by C. albicans isolated from candidiasis, but it only evidenced a significant difference against C. albicans isolated from healthy vaginal microbiota (93%; P = 0.045, Tukey's post hoc).

Fig. 1
figure 1

Displacement of Lactobacillus gasseri by Candida albicans obtained through initial adhesion assays. Displacement of L. gasseri by C. albicans after initial adhesion treatments with the experimental setting of high and low inoculum in the glass surface. The percentage of adhesion of L. gasseri is the result of the variation in the adhesion of L. gasseri and C. albicans strains to coverslip in comparison to controls (CT, 100% of adhesion) when incubated alone at the same conditions. Statistical analysis: *P < 0.05 when using t-student statistical analysis (95% confidence interval) for comparison of lactobacilli control and sample tested in the adhesion assay; †P < 0.05 analyzed using two-tailed ANOVA statistical test (95% confidence interval) for comparison of displacement values from all lactobacilli strains induced by a certain C. albicans isolate tested in the adhesion assay; ‡P < 0.05 analyzed using two-tailed ANOVA statistical test (95% confidence interval) for comparison of displacement values from a certain strain of lactobacilli among all C. albicans isolates tested in the adhesion assay

Full size image

The adhesion inhibition of C. albicans by L. gasseri was also evaluated (see Fig. 2). At ES1, L. gasseri JCM1131 and L. gasseri IMAUFB014 showed statistical differences among C. albicans (L. gasseri JCM1131 P = 0.006, and L. gasseri IMAUFB014 P = 0.002; two‐way ANOVA). L. gasseri JCM1131 evidenced the lowest inhibition rate against C. albicans ATCC10231 (27%), illustrating significant values when compared against C. albicans isolated from candidiasis (60%; P = 0.016, Tukey's post hoc) and C. albicans isolated from healthy vaginal microbiota (67%; P = 0.006, Tukey's post hoc). While L. gasseri IMAUFB014 showed a more efficient inhibition rate against C. albicans isolated from candidiasis (76%; P = 0.002, Tukey's post hoc). At ES2, L. gasseri H59.2 was the only strain to show statistically inhibition values among C. albicans isolates (P = 0.008; two‐way ANOVA). Again, at ES3, only L. gasseri H59.2 demonstrated a statistical difference in its inhibition ability (P = 0.030; two‐way ANOVA) against C. albicans ATCC10231 (61%) and C. albicans isolated from candidiasis (89%; P = 0.034, Tukey's post hoc). At ES4, all C. albicans showed significant inhibition rates (C. albicans ATCC10231: P = 0.010; C. albicans isolated from candidiasis: P = 0.011; C. albicans isolated from healthy microbiota: P = 0.025, two‐way ANOVA analysis). L. gasseri IMAUFB014 showed the highest inhibition rate against C. albicans isolated from healthy microbiota (80%), being statistically different to C. albicans ATCC10231 (47%; P = 0.016, Tukey's post hoc).

Fig. 2
figure 2

The probiotic activity of Lactobacillus gasseri against Candida albicans showed in initial adhesion assays. Inhibition of C. albicans by L. gasseri after initial adhesion treatments with the experimental setting of high and low inoculum in the glass surface. The percentage of adhesion of C. albicans is the result of the variation in the adhesion of L. gasseri and C. albicans strains to coverslip in comparison to controls (CT, 100% of adhesion) when incubated alone at the same conditions. Statistical analysis: *P < 0.05 when using t-student statistical analysis (95% confidence interval) for comparison of candida control and sample tested in the adhesion assay; †P < 0.05 analyzed using two-tailed ANOVA statistical test (95% confidence interval) for comparison of inhibition values between experimental setting (ES) for each evaluated C. albicans isolated in the adhesion assay

Full size image

Probiotic ability of Lactobacillus plantarum ATCC14917 was realized against C. albicans ATCC10231 and compared with L. gasseri evidencing significant displacement and inhibition values (see Additional file 7 and Fig. 3A). The displacement values of L. plantarum were 23% and 54% against low (ES1) and high (ES2) levels of C. albicans, respectively. These values were significantly inferior to L. gasseri (ES1: 61–99% and ES2: 82–96%), more exactly: L. gasseri IMAUFB014 (ES1: P < 0.001and ES2: P = 0.002, Tukey's post hoc); L. gasseri JCM1131 (ES1: P = 0.003 and ES2: P = 0.025, Tukey's post hoc); and L. gasseri H59.2 (ES1: P < 0.001 and ES2: P = 0.012, Tukey's post hoc). At ES3, L. plantarum was only displaced by 31% evidencing again a better resistance when compared to L. gasseri (ES3: 83–95%), specifically: L. gasseri IMAUFB014 (P < 0.001, Tukey's post hoc); L. gasseri H59.2 (P < 0.001, Tukey's post hoc); and L. gasseri JCM1131 (P = 0.001, Tukey's post hoc). At ES4, L. plantarum was displaced 94% without statistical differences. As shown in Fig. 3B, the adhesion inhibition of C. albicans by L. plantarum demonstrated a superior activity in ES1 (81%) and ES2 (58%) when compared to L. gasseri (ES1:27–73% and ES2:1–49%; both P < 0.001, two‐way ANOVA). At ES3 and ES4, L. plantarum showed the lowest inhibition rate (50–56%) without statistical differences.

Fig. 3
figure 3

Preliminary analysis of the displacement of Lactobacillus plantarum by Candida albicans and its probiotic activity on C. albicans through initial adhesion assays. A Displacement of L. plantarum ATCC 14917 by C. albicans ATCC 10231 after initial adhesion treatments with the experimental setting of high and low inoculum in the glass surface. Statistical analysis: *P < 0.05 when using t-student statistical analysis (95% confidence interval) for comparison of lactobacilli control and sample tested in the adhesion assay; †P < 0.05 analyzed using two-tailed ANOVA statistical test (95% confidence interval) for comparison of displacement values between L. plantarum and L. gasseri strains in the adhesion assay at same experimental setting. B Inhibition of C. albicans by L. plantarum after initial adhesion treatments with the experimental setting of high and low inoculum in the glass surface. Statistical analysis: *P < 0.05 when using t-student statistical analysis (95% confidence interval) for comparison of candida control and sample tested in the adhesion assay; †P < 0.05 analyzed using two-tailed ANOVA statistical test (95% confidence interval) for comparison of inhibition values between experimental setting (ES) for each evaluated C. albicans ATCC 10231 isolated in the adhesion assay. No statistically significant values were found

Full size image

Discussion

The use of Lactobacillus is a low-risk alternative for antimicrobial resistance and its adverse effects [31]. It is well-known that a prerequisite for C. albicans' pathogenicity is the initial adhesion to host cells [32] before leading to genital or urinary tract infections [8, 9]. This study focused on the intrinsic probiotic activity of lactobacilli against C. albicans. Although studies reported lactobacilli biofilm/biosurfactant activities against pathogens [10,11,12,13], few authors evaluated the inhibition of the initial adhesion of pathogens [16, 17, 23, 33]. Some studies previously characterized the inhibition of the initial adhesion of certain pathogens, such as Gardnerella vaginalis, Prevotella bivia, Mobiluncus mulieris [23], Listeria monocytogenes [34], and Streptococcus mutans [17]. Recently, He et al. [35] evaluated the probiotic activities of Lactobacillus species on the inhibition of the initial adhesion of several pathogens. However, only L. gasseri demonstrated a more probiotic activity against C. albicans, when compared to L. crispatus. Our results agreed with He et al. [35], reporting differences in probiotic activity among Lactobacillus species/strains against C. albicans isolates. However, there is still scarce information about how these Candida albicans can be inhibited by Lactobacillus strains/species or even to displace different lactobacilli.

An alternative approach could be used by the colonization of new and more probiotic lactobacilli in this environment [15], being permanently assimilate in the vaginal microbiota. Once incorporated, certain lactobacilli species should be able to produce supernatant and eventually evolve in biofilm formation [15, 36], such as L. plantarum. So, the initial adhesion is a vital step for the human epithelial colonization and the inhibition of pathogens being worthy to study.

The present study evaluated three L. gasseri as inherent vaginal lactobacilli and a single L. plantarum (ATCC 14917) as a strong probiotic species atypical of the vaginal microbiota. Our results evidenced statistical differences between the displacement values of L. gasseri strains by the same C. albicans isolate and the variability of each Lactobacillus to inhibit different C. albicans. This variability agrees with a study realized by De Gregorio et al. [11] with Lactobacillus crispatus on the adhesion of Candida species, which evidenced the strain-specific probiotic activity of L. crispatus. So, it is plausible to assume that the remaining Lactobacillus species could also evidence discrepancies in their probiotic ability and displacement resistance against different C. albicans isolates, as proposed by Zangl et al. [15]. The application of different lactobacilli from other biological sources in the human epithelial colonization could increment the probiotic activity of the remaining commensal microbiota, as suggested in other studies [37,38,39]. These studies together with our results of low displacement values in L. plantarum against low or normal levels of C. albicans suggested the potential application of non-human lactobacilli to sustain a more resilient healthy microbiota. L. plantarum ATCC 14917 demonstrated high inhibition percentages of C. albicans ATCC 10231, being more efficient and statistically different when compared to L. gasseri. Our results surpassed the rates of inhibition on C. albicans reported by Dos Santos et al. [12]. Further studies should evaluate longitudinal colonization between non-vaginal lactobacilli and vaginal lactobacilli against Candida species in vitro assays.

Limitations

There are some major limitations in this study: (1) it is a preliminary study realized on an abiotic surface and unable to establish an efficient report on human epithelial colonization; (2) the study did not evaluate the longitudinal colonization between lactobacilli and C. albicans; and (3) the study only compared the probiotic activity of L. plantarum against a single Candida albicans.