Lactobacillus rhamnosus AD3 as a Promising Alternative for Probiotic Products

Lactobacillus strains dominate the vaginal habitat and they are associated with a lower risk of genital infections. In addition, they contribute to the conservation of the vaginal microbiota by producing postbiotic agents. Previous studies have shown that their predominance involves antimicrobial activity against urogenital pathologies. In this context, probiotics may improve treatment outcomes. The aim of this study was to evaluate the probiotic properties of lactobacilli strains of vaginal origin using a multidisciplinary approach. For this purpose, safety criteria, ability to resist at low pH and bile salts, antimicrobial activity, ability to produce biofilm, capacity to produce hydrogen peroxide and more importantly, auto-aggregation, co-aggregation (with Candida spp.) and adhesion to human cells were evaluated. The strains belonged to the species of L. crispatus, L. gasseri, L. rhamnosus and L. delbruckii. Among these, a strain of L. rhamnosus named AD3 showed the best probiotic properties. As probiotics are already in use in many clinical practice and there are no major safety concerns, L. rhamnosus AD3 showed promise in becoming a prevention and complementary treatment option for urogenital diseases. Indeed, these results suggest that strain L. rhamnosus AD3 is non-pathogenic and likely to be safe for human consumption. This study revealed the great amensalistic properties of a new L. rhamnosus strain which can aim to be used as probiotic in pharmaceutical applications.


Introduction
Urogenital infections are common concerns in young women [1,2]. Among all sexually transmitted diseases, bacterial vaginosis and vulvovaginal candidiasis are the most frequent [3,4]. These infections make-up the most common reasons for a woman to need to visit to gynaecologist, with the outcome of discomfort and also enormous costs for health care treatment [5]. Moreover, these kinds of diseases most often include treatment with antibiotics and/or antimycotics, which have led to a raised concern regarding drug resistance among pathogenic species [6,7].
The human vaginal ecosystem is dominated by Lactobacillus species and an altered vaginal microbiota can cause a greater chance of symptomatic conditions. Furthermore, there is a clear relationship between vaginal microbiota and urogenital infections [8], in fact most of the microorganisms in the vagina also derive from the gastrointestinal tract, and this does not depend only on sexual activity but also on personal hygiene and nutrition [9]. The protective role of lactobacilli is increasingly accepted and several clinical studies are underway to evaluate the use of lactobacilli as probiotics. Further studies have shown that the composition of the vaginal microbiota can undergo changes throughout life. The vaginal environment is dynamic and it changes in relation to events such as sexual activity, the menstrual cycle or menopause [10]. During the fertile age, there is a vaginal epithelium rich in glycogen and mucins that act as nutrients and that allow colonization and domination 2.2. LAB Identification 2.2.1. DNA Extraction RTP Bacterial DNA Mini Kit (Stratec Biomedical AG, Birkenfeld, Germany) was used to extract bacterial DNA following manufacturer instruction. Then, DNA samples obtained were visualized by electrophoresis (70 V) in agarose gel (1.5%).

Molecular Identification
16S rDNA PCR/RFLP and, multiplex-PCRs to amplify different loci of the 16S-23S gene region and ITSregion flanking the 23S rDNA gene were used to make molecular identification as previously described [21].

Safety Criteria of Potential Probiotics 2.3.1. Antibiotic Resistance
Ampicillin, erythromycin, tetracycline, gentamicin, clindamycin, chloramphenicol, streptomycin and kanamycin, were used to assess antibiotic resistance according to European Food Safety Authority [22]. Moreover, ciprofloxacin was used as internal negative control.

Gelatinase Activity and Haemolytic Activity
Gelatinase activity and haemolytic activity were evaluated as previously described [23]. Briefly, overnight cultures of strains were streaked onto Nutrient Gelatin agar plates (Oxoid, Thermo Fisher Scientific Inc., Rodano (MI), Italy, CM0635) and Columbia +5% sheep blood agar plates (bioMérieux Corporate, Bagno a Ripoli (FI), Italy). Plates were incubated at 37 • C under micro-aerobic conditions for 48 h. Bacillus subtilis ATTC 23857 and Streptococcus pyogenes ATCC 19615 were used as internal positive control for gelatinase activity and haemolytic activity, respectively.

Ability to Resist at Low pH and Bile Salts
The evaluation of the resistance to low pH and bile salts was carried out by bacterial counts and relative survival rate after stressed growth [24]. Briefly, Lactobacillus strains and E. coli ATCC 25922 (as positive internal control) were grown in LSM broth and MH broth, respectively, at 37 • C overnight. Then, two different conditions were evaluated: growth in bile salts agar, and growth in bile salts agar after one hour in acidic condition (pH 3.0). Subsequently, serial ten-fold dilutions in NaCl (0.85% w/v) were made and plated on agar for enumeration. Each experiment was made six times in duplicate and the growth in LSM broth (and MH broth for E. coli) without stress was used as control.

Antimicrobial Activity
Cells-free supernatants produced by Lactobacillus strains (grown at 37 • C, 18 h in micro-aerobic condition) were neutralized with NaOH 1 mol·L −1 and inoculated into wells in pre-inoculated MH agar plates and Sabouraud dextrose agar with chloramphenicol, with bacteria and fungi, respectively [25][26][27]. For quality control ciprofloxacin and itraconazole were used as positive internal controls for antibacterial and antifungal activity, respectively. MH plates were incubated at 37 • C overnight in aerobic condition to evaluate antibacterial activity, and SAB plates were incubated at 33 • C for 24 h in aerobic condition to evaluate antifungal activity. Inhibition zones were measured by a gauge [23,28]. Results were expressed in mm ± SD.

Ability to Produce Biofilm
Lactobacillus strains were tested for the ability to form biofilm. An inoculum (1.5 × 10 5 CFU mL −1 ) of each strain was incubated in MRS broth at 37 • C under microaerobic condition. Then, the absorbance at 600 nm (BioRad Microplate Reader Model 680, ©2021 Bio-Rad Laboratories, Inc., Segrate (MI), Italy) was recorded and the broth was discharged. The plates were washed, with PBS in order to remove suspended cells, and in-cubated at 60 • C for 1 h. Subsequently, 200 µL of crystal violet were added in each well and after 10/15 min crystal violet residues were dissolved with a solution of methanol/acetic acid 2:1.5. A second read was recorded at 570 nm. The following formula OD 570 /OD 600 × 0.4 was used to calculate biofilm and OD 0.061 was considered as cut-off value.

Capacity to Produce Hydrogen Peroxide
Horseradish peroxidase (Sigma, Merck KGaA, Darmstadt, Germany, P6782) and 3,3 ,5,5 -tetramethylbenzidine (Sigma, Merck KGaA, Darmstadt, Germany, T2885) were added to MRS agar to evaluate and quantify hydrogen peroxide production by Lactobacillus strains. Briefly, a bacterial inoculum (1.5 × 10 8 CFU mL −1 ) was streaked on this MRS modified agar and, the plates were incubated under micro-aerobic condition for 48 h at 3 • C, and further at room temperature for other 24 h in the dark. Negative producers gave colourless colonies, while positive producers gave colour colonies from brown to dark blue.
One day before the experiment, cells were trypsinized, counted and inoculated in 129AX-1 tubes with a slide (Sterilin TM , Thermo Fisher Scientific Inc., Rodano (MI), Italy); approximately 1.5 × 10 5 cells/tube. After 48 h the cells were washed in PBS, and a bacterial suspension (1.0 × 10 8 CFU mL −1 ) was added and the mixture incubated for 30 min at room temperature. Then, three washes with PBS were done and cells were fixed with ethanol 1 mL/tube for 15 min at 3 • C, stained with 20% Giemsa (v/v in phosphate-buffered water) for 20 min (in the dark) and examined microscopically under oil immersion. For each slide 200 cells were observed and the results are expressed as the average of bacteria per cells. The adhesion test was repeated at least six times and data were summarized using mean (±SD/cell) [29].

Auto-Aggregation Study
Auto-aggregation study was performed with slightly modified protocol previously described by Gil et al. [30]. Briefly, the strains were grown overnight at 37 • C in MRS agar, then suspensions of 0.6 at 600 nm (BioRad Microplate Reader Model 680, © 2021 Bio-Rad Laboratories, Inc., Segrate (MI), Italy) in PBS were done. Variation of OD was monitored every hour for 4 h. Then, the percentage of auto-aggregation was calculated as follows: Gram staining was used to visualize the aggregates under oil immersion microscopy (1000×).

Co-Aggregation Study
Co-aggregation study was evaluated as follows: 129AX-1 tubes with slides (Sterilin TM , Thermo Fisher Scientific Inc., Rodano (MI), Italy) were added with 500 µL of an overnight culture of Lactobacillus strains grown at 37 • C in MRS broth (Oxoid, Thermo Fisher Scientific Inc., Rodano (MI), Italy, CM0359) under micro-aerobic conditions and 500 µL of an overnight culture of Candida spp. grown at 33 • C in DMEM (Dulbecco's MEM-Biochrom GmbH, Berlin, Germany). Tubes were incubated at 37 • C for 4 h. Then, Gram staining was used to visualize the co-aggregates under oil immersion microscopy (1000×) [30]. Scoring was done according to Reid et al. [31]. Moreover, the co-aggregation were also evaluated on Hep-2 and Caco-2 cells. Tubes with monolayer cells were incubated at 37 • C for 4 h after inoculation with Candida and Lactobacillus strains, then Giemsa staining was used to visualize co-aggregates and adhesion on cells.

Statistical Analysis
All experiments were performed at least three times and data were summarized using mean (±SD). Where applicable, data were analysed by one-way ANOVA with correction for multiple comparisons by Bonferroni. All results with p-value < 0.05 were considered significant. All results and graphs were generated using GraphPad ® Prism ver. 6 software.

LAB Identification
The identification of the species of Lactobacillus isolates was conducted with the amplification of 16S rDNA. Four species were identified among which L. crispatus was the most representative (6 strains), followed by L. gasseri (4 strains), L. rhamnosus (1 strain) and L. delbrueckii (1 strain).

Antibiotic Resistance
Antibiotic resistance was evaluated to verify if the twelve Lactobacillus strains could be accepted to use as potential probiotic. Strains were inhibited by the eight antibiotics tested according to the microbiological cut-off values defined by EFSA (Table 1). Therefore, all strains were susceptible to all antibiotics. Concurrently, the strains resulted resistant for the antibiotic CIP used as quality control.

Gelatinase Activity and Haemolytic Activity
Gelatinase activity was not found for the Lactobacillus strains tested. Furthermore, lysis on blood agar plates were not shown for any Lactobacillus strain under examination.

Ability to Resist at Low and Bile Salts
As already stated, according to FAO [32], bacterial strains must resist to gastrointestinal conditions and remain viable to be considered probiotics. Figures 1 and 2 show the reduction to exposition to low pH and different concentrations of bile salts, respectively.
All isolates showed good tolerance to bile salts at difference concentration pre-exposition at low pH. In particular, L. rhamnosus AD3 proved to be the most resistant, with a survival of 69.2% with bile salts at 0.5% and even 130.8% with bile salts at 0.12%. Oddly, the only strain that showed significant reduction in bile salt treatment was L. rhamnosus GG (−93.7% with BS 0.5%; −90% with BS 0.12%). After exposure to acid pH for one hour, the consequent exposure to bile salts showed a slight decrease for all strains including L. rhamnosus GG. Additionally in this case, L. rhamnosus AD3 proved to be the most resistant strain (−0%). This means that after GI tract passage, all strains were able to survive but not at the same initial concentration of inoculum.   All isolates showed good tolerance to bile salts at difference concentration pre-exposition at low pH. In particular, L. rhamnosus AD3 proved to be the most resistant, with a survival of 69.2% with bile salts at 0.5% and even 130.8% with bile salts at 0.12%. Oddly, the only strain that showed significant reduction in bile salt treatment was L. rhamnosus GG (−93.7% with BS 0.5%; −90% with BS 0.12%). After exposure to acid pH for one hour, the consequent exposure to bile salts showed a slight decrease for all strains including L. rhamnosus GG. Additionally in this case, L. rhamnosus AD3 proved to be the most resistant strain (−0%). This means that after GI tract passage, all strains were able to survive but not at the same initial concentration of inoculum.

Antimicrobial Activity
The antagonistic activity of putative probiotic strains was assessed against the following target microorganisms: Escherichia coli ATCC 25922, Enterococcus faecalis ATCC 29212, Pseudomonas aeruginosa ATCC 27853, Klebsiella pneumoniae ATCC 700603, a clinical isolate of Streptococcus agalactiae, three clinical isolates of Candida (C. albicans 10, C. glabrata 14, C. tropicalis 21). The strains showed good activity both against Gram-positive and Gramnegative pathogens, and also excellent activity against Candida strains ( Table 2). L. gasseri SG3 showed the highest antagonistic activity against both bacteria and fungi, followed by L. rhamnosus AD3 which however had no effect against E. coli and C. tropicalis. Finally, no inhibition zones were shown by L. rhamnosus GG. Table 2. Antimicrobial activity of cell-free supernatants produced by Lactobacillus strains against bacterial and fungal urogenital pathogens. Results were expressed in mm (mean ± SD).

Strains
Escherichia coli

Capacity to Produce Hydrogen Peroxide
Production of hydrogen peroxide was tested in 12 vaginal lactobacilli strains and L. rhamnosus GG by the semi-qualitative TMB peroxidase assay (Figure 4). Only three strains out of 13 produced great amount of hydrogen peroxide: two L. crispatus (BG2, EB3) and one L. gasseri (BA3). L. rhamnosus GG showed a moderate production. The ability of a bacterial strain to adhere to human epithelial cells is one of the most important properties for such a strain to be considered probiotic. Indeed, this feature will allow it to colonize the mucous membrane and perform important modulating functions on the immune system. In this study, Lactobacillus strains were tested for adhesion to

Capacity to Produce Hydrogen Peroxide
Production of hydrogen peroxide was tested in 12 vaginal lactobacilli strains and L. rhamnosus GG by the semi-qualitative TMB peroxidase assay (Figure 4). Only three strains out of 13 produced great amount of hydrogen peroxide: two L. crispatus (BG2, EB3) and one L. gasseri (BA3). L. rhamnosus GG showed a moderate production.

Capacity to Produce Hydrogen Peroxide
Production of hydrogen peroxide was tested in 12 vaginal lactobacilli strains and L. rhamnosus GG by the semi-qualitative TMB peroxidase assay (Figure 4). Only three strains out of 13 produced great amount of hydrogen peroxide: two L. crispatus (BG2, EB3) and one L. gasseri (BA3). L. rhamnosus GG showed a moderate production.

Ability to Adhere to Human Hep-2 and Caco-2 Cell Lines
The ability of a bacterial strain to adhere to human epithelial cells is one of the most important properties for such a strain to be considered probiotic. Indeed, this feature will allow it to colonize the mucous membrane and perform important modulating functions on the immune system. In this study, Lactobacillus strains were tested for adhesion to

Ability to Adhere to Human Hep-2 and Caco-2 Cell Lines
The ability of a bacterial strain to adhere to human epithelial cells is one of the most important properties for such a strain to be considered probiotic. Indeed, this feature will allow it to colonize the mucous membrane and perform important modulating functions on the immune system. In this study, Lactobacillus strains were tested for adhesion to Hep-2 and Caco-2 cells (Table 3). Among all isolates tested, L. rhamnosus AD3 and L. crispatus EB1 showed heavy in vitro adhesion towards both epithelial cell lines compared to L. rhamnosus GG (≈47/cell).  (1); aggregates with larger numbers of bacteria (2); macroscopically visible clumps with larger groups of bacteria (3) and macroscopically visible clumps (4).

Auto-Aggregation Study
The ability of auto-aggregation of vaginal lactobacilli was evaluated at five time-points (0-4 h). This ability was confirmed at all time-points but the strains showed the highest degree of auto-aggregation after 4 h of incubation. Some strains have a greater ability to self-aggregate than others: in this study L. rhamnosus AD3 and L. crispatus CC4 showed the highest auto-aggregating aptitude, even greater than L. rhamnosus GG. Figure 5 showed auto-aggregation observed for L. rhamnosus AD3, L. crispatus CC4 and L. rhamnosus GG in two time-points (T0 and T4). Hep-2 and Caco-2 cells (Table 3). Among all isolates tested, L. rhamnosus AD3 and L. crispatus EB1 showed heavy in vitro adhesion towards both epithelial cell lines compared to L. rhamnosus GG (≈47/cell).  (1); aggregates with larger numbers of bacteria (2); macroscopically visible clumps with larger groups of bacteria (3) and macroscopically visible clumps (4).

Auto-Aggregation Study
The ability of auto-aggregation of vaginal lactobacilli was evaluated at five time-points (0-4 h). This ability was confirmed at all time-points but the strains showed the highest degree of auto-aggregation after 4 h of incubation. Some strains have a greater ability to self-aggregate than others: in this study L. rhamnosus AD3 and L. crispatus CC4 showed the highest auto-aggregating aptitude, even greater than L. rhamnosus GG. Figure 5 showed auto-aggregation observed for L. rhamnosus AD3, L. crispatus CC4 and L. rhamnosus GG in two time-points (T0 and T4).

Co-Aggregation Study
The auto-aggregation ability of Lactobacillus strains was reflected in the ability to create co-aggregates with Candida cells. Vaginal strains were evaluated with Candida albicans (Figure 6A), Candida krusei ( Figure 6B) and Candida tropicalis ( Figure 6C). The co-aggregation scores were obtained after 4 h of incubation at 37 • C (Table 3). Figure 7 showed macroscopically visible clumps when L. rhamnosus AD3 and GG have been tested with all three Candida strains. It is possible that co-aggregates are able to create microenvironments around the pathogen capable of blocking its diffusion towards epithelial cells. Indeed, the study of the co-aggregation between Lactobacillus strains and Candida strains on Hep-2 and Caco-2 monolayers has shown, in some cases, how the bacteria-fungus co-aggregates inhibited the adhesion of Candida on epithelial cells (Figure 8 AD3).
Biomolecules 2021, 11, x FOR PEER REVIEW 10 of 14 Figure 5. Auto-aggregation of vaginal lactobacilli at two time-points: T0 and T4. L. rhamnosus AD3, L. crispatus CC4 and L. rhamnosus GG showed greater ability to self-aggregate than others strains. Strains were observed under the microscope (1000×) after Gram staining.

Co-Aggregation Study
The auto-aggregation ability of Lactobacillus strains was reflected in the ability to create co-aggregates with Candida cells. Vaginal strains were evaluated with Candida albicans ( Figure 6A), Candida krusei ( Figure 6B) and Candida tropicalis ( Figure 6C). The co-aggregation scores were obtained after 4 h of incubation at 37 °C (Table 3). Figure 7 showed macroscopically visible clumps when L. rhamnosus AD3 and GG have been tested with all three Candida strains. It is possible that co-aggregates are able to create microenvironments around the pathogen capable of blocking its diffusion towards epithelial cells. Indeed, the study of the co-aggregation between Lactobacillus strains and Candida strains on Hep-2 and Caco-2 monolayers has shown, in some cases, how the bacteria-fungus co-aggregates inhibited the adhesion of Candida on epithelial cells (Figure 8  AD3).

Co-Aggregation Study
The auto-aggregation ability of Lactobacillus strains was reflected in the ability to create co-aggregates with Candida cells. Vaginal strains were evaluated with Candida albicans ( Figure 6A), Candida krusei ( Figure 6B) and Candida tropicalis ( Figure 6C). The co-aggregation scores were obtained after 4 h of incubation at 37 °C (Table 3). Figure 7 showed macroscopically visible clumps when L. rhamnosus AD3 and GG have been tested with all three Candida strains. It is possible that co-aggregates are able to create microenvironments around the pathogen capable of blocking its diffusion towards epithelial cells. Indeed, the study of the co-aggregation between Lactobacillus strains and Candida strains on Hep-2 and Caco-2 monolayers has shown, in some cases, how the bacteria-fungus co-aggregates inhibited the adhesion of Candida on epithelial cells (Figure 8  AD3).

Discussion
Maintaining an adequate concentration of Lactobacilli is a necessary requirement to ensure the balance and health of the urogenital tract, especially in women [33].
Apparently, as already mentioned, alterations in the composition of the microbiota in women determine a greater susceptibility to infections. Therefore, re-establishing the Lactobacilli prevalence in the vaginal microbiota is necessary for the maintenance of the woman's state of health [34].
The aim of this work was to find strains of Lactobacilli with probiotic characteristics adequate to respond to the physiological demands of women.
Firstly, the strains were tested for their susceptibility to antibiotics as well as for haemolytic and gelatinase activity. These tests were necessary to evaluate the safety of the strains, in fact no strain having antibiotic resistance genes or haemolytic/gelatinase activity can be marketed [22,32]. All the strains tested met these safety qualities.
The strains studied belonged to the following species: L. crispatus, L. gasseri, L. rhamnosus and L. delbrueckii. These species are among the most encountered in the vaginas of healthy women [35,36]. Usually, L. crispatus and L. gasseri are strong producers of lactic acid and inhibitory molecules against pathogenic species that cause urethritis, vaginitis or candidiasis [37,38]. In our study, a strain of L. gasseri, named SG3, showed the highest antagonistic activity against both bacteria and fungi, followed by L. rhamnosus AD3. Oddly enough, the positive control strain L. rhamnosus GG did not show any inhibitory activity. In addition, L. gasseri and L. rhamnosus are the species most resistant to gastrointestinal conditions, as well as strong producers of biofilm [23,24].
In this study, the strain that showed the greatest resistance at low pH and bile salts (pre-and post-exposition) was L. rhamnosus AD3, followed by L. gasseri SG3. In contrast, the positive control L. rhamnosus GG showed low resistance to these factors.
As far as biofilm production is concerned, all strains have been shown to be moderate or strong (AD3 and GG) producers except for CC1, BG2 and RB1. The latter did not even show resistance to gastrointestinal conditions proving to be the weakest strain among those tested and therefore the strain with less probiotic properties.
Moreover, the most of the strains under examination showed moderate accumulation of H2O2, with the exception of SG3 (very low) and BG2, EB3 and BA3 which showed to be great producers. However, the role of hydrogen peroxide as a protective agent is controversial. Indeed, despite the statement that Lactobacillus spp. can exert its probiotic properties also through the H2O2 production is accepted, there are several evidences as

Discussion
Maintaining an adequate concentration of Lactobacilli is a necessary requirement to ensure the balance and health of the urogenital tract, especially in women [33].
Apparently, as already mentioned, alterations in the composition of the microbiota in women determine a greater susceptibility to infections. Therefore, re-establishing the Lactobacilli prevalence in the vaginal microbiota is necessary for the maintenance of the woman's state of health [34].
The aim of this work was to find strains of Lactobacilli with probiotic characteristics adequate to respond to the physiological demands of women.
Firstly, the strains were tested for their susceptibility to antibiotics as well as for haemolytic and gelatinase activity. These tests were necessary to evaluate the safety of the strains, in fact no strain having antibiotic resistance genes or haemolytic/gelatinase activity can be marketed [22,32]. All the strains tested met these safety qualities.
The strains studied belonged to the following species: L. crispatus, L. gasseri, L. rhamnosus and L. delbrueckii. These species are among the most encountered in the vaginas of healthy women [35,36]. Usually, L. crispatus and L. gasseri are strong producers of lactic acid and inhibitory molecules against pathogenic species that cause urethritis, vaginitis or candidiasis [37,38]. In our study, a strain of L. gasseri, named SG3, showed the highest antagonistic activity against both bacteria and fungi, followed by L. rhamnosus AD3. Oddly enough, the positive control strain L. rhamnosus GG did not show any inhibitory activity. In addition, L. gasseri and L. rhamnosus are the species most resistant to gastrointestinal conditions, as well as strong producers of biofilm [23,24].
In this study, the strain that showed the greatest resistance at low pH and bile salts (pre-and post-exposition) was L. rhamnosus AD3, followed by L. gasseri SG3. In contrast, the positive control L. rhamnosus GG showed low resistance to these factors.
As far as biofilm production is concerned, all strains have been shown to be moderate or strong (AD3 and GG) producers except for CC1, BG2 and RB1. The latter did not even show resistance to gastrointestinal conditions proving to be the weakest strain among those tested and therefore the strain with less probiotic properties.
Moreover, the most of the strains under examination showed moderate accumulation of H 2 O 2 , with the exception of SG3 (very low) and BG2, EB3 and BA3 which showed to be great producers. However, the role of hydrogen peroxide as a protective agent is controversial. Indeed, despite the statement that Lactobacillus spp. can exert its probiotic properties also through the H 2 O 2 production is accepted, there are several evidences as support the implausibility of "in vivo" action as an antimicrobial factor in the vaginal environment [39]. Hence, this property cannot be taken as a predisposing factor for a strain's probiotic activity.
The ability of auto-aggregation of Lactobacillus strains can help to increase the colonization of vaginal environment by adhesion and sub-sequentially by biofilm formation and postbiotic production to inhibit the overgrowth pathogenic microorganisms [30,40]. Moreover, the ability to create co-aggregates with pathogenic microorganisms can modulate the destiny of pathologies among which stand out recurrent vulvovaginal candidiasis [41,42]. L. rhamnosus AD3 and L. crispatus CC4 showed the highest capacity to adhesion and autoaggregation ( Figure 5). Moreover, L. rhamnosus AD3 has shown an intended ability to form co-aggregates with all three Candida species, especially C. tropicalis (Figure 7). Although L. rhamnosus GG's co-aggregation capacity is visibly greater (Figure 7), L. rhamnosus AD3 would seem capable of binding Candida cells more consistently. This property is evident from Candida absence on monolayer cells after co-culture exposure (Figure 8 AD3). On the contrary, bound Candida cells can be seen on cells despite the formation of co-aggregates with L. rhamonsus GG (Figure 8 GG).
In summary, our results indicated that among all vaginal lactobacilli strains tested, L. rhamnosus AD3 revealed the required properties as potential probiotic strain. Indeed, it has responded to all safety features, and it has shown the highest levels of resistance to the gastrointestinal tract, production of active postbiotics and biofilm against fungi and bacteria, ability to adhere to epithelial mucosa, capacity to form strong aggregates and co-aggregates with Candida. Instead, L. delbrueckii RB1 proved to be the weakest strain. Strangely, as already said, the strain used as a positive control, L. rhamonosus GG, already on the market, showed low resistance to the gastrointestinal tract and no production of postbiotics active against pathogenic strains. Nonetheless, it showed high adhesion, autoaggregation and co-aggregation ability, although candida cells were still present on the epithelial mucosa.
These results suggest that L. rhamnosus AD3 may be a promising strain to aid vaginal microbiota recovery as complementary treatment option for urogenital diseases, and to enhance the normal vaginal microbiota in healthy young women.
Future studies are required to assess technological method to make pharmaceutical preparations able to respond to market demands.