Bacterial cultures and growth conditions
For this study, a total of 95 lactobacilli were investigated; 48 strains were isolated from dairy products, 31 from fruits and vegetables, 15 from sourdough, and 1 from other sources (Table 5). These strains, together with the pathogenic strain S. agalactiae 88II3 and C. albicans UNIBZ54, belong to the Micro4Food collection from the University of Bolzano-Bozen. All the cultures were maintained as frozen stocks at -20°C in their specific broth medium with 20% glycerol for subsequent analysis. Before their use, the lactobacilli were propagated twice in MRS broth (Sigma Aldrich) at 37°C (body temperature) for 24 h. C. albicans UNIBZ54 was refreshed in Sabouraud broth (Scharlau, Spain) with 5% Tween®80 (Sigma Aldrich) and incubated overnight at 37°C. S. agalactiae 88II3 was refreshed in Brain-Heart Infusion (BHI) broth and incubated at 37°C until it reached the stationary phase (ca. 24 h).
Table 5
Lactobacilli strains (n = 95) used in this study.
|
Number of strains
|
Species
|
Source and sub-source of isolation
|
|
50
|
Lacticaseibacillus paracasei
|
| |
31a, 25h, 100i, 104g, 22e, 25e, 25g, 28g, 41j, 45j,
50a, 52i, 76d, 83e, 84f, 93j, 99a, AAI9, AII8, BBII10, bI5, dII1, eI3, GII3, HHI10, iiII4, iiII9, JJI8, kI12,
LII1, MMII7, wI10, WWI9, zzI10, zzI3, zzI4, zzI6
|
Dairy (Milk)
|
| |
AFI10, AFI7, AFII5, ALII8
|
Fruits and Vegetables (Apple-by-products)
|
| |
S4d8
|
Fruits and Vegetables (Sauerkraut)
|
| |
F1, F10, F12, F13, F2, F23, F25, F5
|
Dairy (Cheese)
|
|
3
|
Lactiplantibacillus pentosus
|
|
| |
D2.15, E1.4, E3.10
|
Sourdough
|
|
40
|
Lactiplantibacillus plantarum
|
| |
P1, CB5
|
Dairy (Cheese)
|
| |
11j
|
Dairy (Milk)
|
| |
C5
|
Fruits and Vegetables (Carrot)
|
| |
CIL6
|
Fruits and Vegetables (Cherry)
|
| |
Fin6, Fin10
|
Fruits and Vegetables (Fennel)
|
| |
IT1, IT5
|
Fruits and Vegetables (Grape)
|
| |
K1, K13, K2, K9, KI-5
|
Fruits and Vegetables (Kiwi)
|
| |
OE1
|
Fruits and Vegetables (Olives)
|
| |
P3
|
Fruits and Vegetables (Papaya)
|
| |
1LS16
|
Fruits and Vegetables (Pineapple)
|
| |
PR14, PR3, PR6
|
Fruits and Vegetables (Prune)
|
| |
S6w5, AFI5
|
Fruits and Vegetables (Sauerkraut)
|
| |
POM1, POM20, POM27, POM35, POM42, POM43, POM40
|
Fruits and Vegetables (Tomato)
|
| |
DM
|
Other
|
| |
E3.13, E3.19, D9.30, D9.40, D9.46, D3.15, C5.10, D9.18, E3.8, D2.6
|
Sourdough
|
|
2
|
Lacticaseibacillus rhamnosus
|
| |
B6.19, B4.2
|
Sourdough
|
Cell-free supernatants collection
Lactobacillus strains were grown in MRS broth for 24 h at 37°C, and then the CFSs were recovered by centrifugation (7,500 rpm, 10 min). Further, CFSs were fractionated in two aliquots. One aliquot was neutralized at pH 7.0 (n-CFSs) with NaOH and sterilized by using a 0.22 µm filter. The other aliquot was preserved at the original acidic pH of ca. 4.0 (o-CFSs) and was further divided into two aliquots: one was sterilized by using a 0.22 µm filter, and one was kept not sterile. All the aliquots of CFSs were stored at -20°C until needed. Sterile CFSs were used for pathogen inhibition screening, while non-sterile CFSs were used for functional assays (H2O2 and lactic acid quantification).
Lactic acid isomers and hydrogen peroxide quantification
The quantification of lactic acid and H2O2 in the CFSs was done using two commercial kits, respectively the Megazyme D-Lactate and L-Lactate Assay Kit (Megazyme International Ireland Ltd., Wicklow) and the Peroxide assay kit (Sigma Aldrich), following the instructions provided by the manufacturers.
Auto-aggregation capacity and hydrophobicity characterization
The candidate strains were grown in MRS at 37°C for 18–22 hours. The pellets were harvested by centrifugation (10,000 g, 10 mins, 5°C), washed with PBS (phosphate buffer solution) (pH 7.0), and re-suspended in the same buffer. Cell suspensions were adjusted to an optical density (OD) of 620 nm of ca. 0.25 and used for both analyses (A0 and H0). Cell auto-aggregation was performed according to Gil-Rodríguez, Carrascosa, and Requena (2015) as modified by Di Cagno et al. (2020). Each cell suspension was left to settle at room temperature, then the OD620 was measured after 2, 4, and 24 h (At). The percentage of auto-aggregation (A) was calculated according to the following formulae:
Estimation of hydrophobicity was performed according to Burns et al. (2008) as modified by Di Cagno et al. (2020). Xylene was used to determine the hydrophobicity of the cell surface. A total volume of 0.4 mL of xylene (Sigma Aldrich) was added to the cell suspension (2 mL), and vortexed for 120s. After phase stabilization and separation (1 h, 37°C), the OD of the aqueous phase was measured at 620 nm (Ht). The percentage of hydrophobicity (H) was calculated according to the following formulae:
High-throughput pathogen inhibition screening
Growth inhibitions of C. albicans UNIBZ54 and S. agalactiae 88II3 by CFSs, previously collected and stored, were screened. Briefly, overnight cultures of the pathogens were washed twice with saline solution (0.9% NaCl) and resuspended in saline solution to a final OD600 of 0.1 for C. albicans UNIBZ54, or OD620 of 0.25 for S. agalactiae 88II3. The 96-well plates were set using one volume of CFSs and three volumes of each pathogen. The analysis was done in triplicate using CFSs previously neutralized (pH 7.0) and the acidic CFSs (pH ca. 4.0). Control wells were prepared using MRS medium at pH 4.0, and MRS at pH 7.0, and run in triplicates. Control MRS was acidified using a racemic solution of lactic acid to pH 4.0 and adjusted back to pH 7.0 using NaOH 1M to exclude the osmolyte effect. Growth kinetics of the pathogens were recorded for 46 h at 37°C measuring the absorbance (at a wavelength of 600 nm for C. albicans UNIBZ54, and 620 nm for S. agalactiae 88II3) every 15 min with the Infinite® M Nano + Spectrophotometer (TECAN, Austria).
Candidate selection approach and scoring procedure
The best-performing lactobacilli were selected through a scoring procedure based on the results of the previously mentioned tests and considering the measure of central tendency along with quartiles. Strains were considered “positive” if the value of each strain was higher than the third quartile (e.g., a value ≥ 25% of the highest values in the dataset for one assay).
$$Score= \frac{\sum positive}{total number of essays} \times 100$$
Strains were given a score according to their relative performance on each functional assay: score 1 if Q2 < X > Q3 and score 2 if X > Q3.
Double hits (DH = 1 point) of inhibition by the n-CFSs against S. agalactiae 88II3 (SA) or C. albicans UNIBZ54 (CA) corresponded to a significant effect (P < 0.05) towards a given pathogen detectable for two or more growth parameters. Strains whose n-CFSs at pH 7.0 showed a double hit (DH) inhibition and o-CFSs that showed a single inhibition effect (H) on A at pH 4.0 were scored according to the following calculation:
$$Inhibition Score =\left(4\times {DH}_{CA}+ {DH}_{SA}\right)+({4 \times H}_{CA}+ {H}_{SA})$$
Double hit (DH) or hit (H) on A for C. albicans UNIBZ54 were multiplied by 4 since the frequency of a double hit of inhibition on C. albicans UNIBZ54 was ca. 4 times lower than the one on S. agalactiae 88II3.
Biofilm preparation
Overnight planktonic cultures of the lactobacilli were centrifuged (7,500 rpm, 10 min, 4°C) and the pellets from each strain were collected and washed twice in saline solution. The suspensions were diluted to an OD620 of ca. 0.25 with saline solution. One 5-µL drop of diluted culture was used to inoculate individual sterile membrane filters (pore size, 0.22 µm, Whatman) resting on MRS agar Petri dishes. The membranes were sterilized by UV exposure (15 min per side) before inoculation. The plates were inverted after the inoculum and incubated at 37°C, with the membrane-supported biofilms transferred to a fresh MRS agar Petri dish every 8 to 10 h (Anderl, Franklin, and Stewart 2000). Five membranes were prepared for each strain.
Biofilm imaging by confocal laser scanning microscopy
One membrane-supported biofilm for each strain culture was visually inspected and photographed by CLSM (Leica SP8LIA, Leica Microsystems). The membrane-supported biofilms were carefully mounted on glass slides. Bacterial cells and the polysaccharide fraction of ECM were stained with 15 µM SYTO® 9 (Invitrogen) and 200 µg mL− 1 Texas Red®-labeled Concanavalin A (ConA, Invitrogen, stock solution, 5 mg mL− 1 in 0.1 M sodium bicarbonate) solution in PBS (pH 7.5). Samples were incubated under dark conditions for 1 h at room temperature. Then, biofilm images were collected with a CLSM with excitation at 488 nm and emission > 552 nm lasers. Fluorescence emission was observed between 500–565 nm (for SYTO® 9) and 565–645 nm (for ConA). Images were captured with a 40x lens using immersion oil and analyzed with the software LAS X (Leica).
Biofilm growth and ECM characterization
At the end of the incubation, two membrane-supported biofilms for each strain were inserted in separate falcon tubes containing saline solution (9 mL) to quantify the biomass (mg). Cells were then detached and suspended through the vortex (1 min at maximum speed) and serially diluted. Dilutions were plated on MRS agar Petri dishes and the cell density (Log colony forming unit [CFU] mL− 1) was enumerated after 48 h of incubation at 37°C. The ECM components of the biofilm colonies were collected and characterized following the method described by Chiba et al. (2015). Briefly, the biofilm colonies were scraped from the membrane filters and suspended in NaCl solution (1.5 M). The suspensions were centrifuged (5,000 g, 10 min, 25°C) and the supernatants were collected as ECM fractions for the quantification of proteins, total saccharides, and eDNA. The concentration of protein was measured following the Bradford assay (Bradford, 1976), using bovine serum albumin (BSA) as a standard. The protein concentration was measured at 590 nm with the UV-1800 Spectrophotometer (SHIMADZU). The total saccharide concentration in the ECMs was measured by the phenol sulfuric acid method (Chiba et al. 2015), using glucose as a standard. Briefly, the isolated ECM fractions (20 µL) were mixed with 5% phenol (20 µL) in a 96-well plate, before adding sulfuric acid (100 µL). The plate was then incubated for 10 min at room temperature, and then the absorbance was read at 492 nm with an Infinite® M Nano + Spectrophotometer (TECAN, Austria). The concentration of the eDNA in the ECM fractions was measured with NanoDrop 2000 (Thermo Fisher Scientific, Waltham, MA, USA).
Gel formulation
From the results of the second screening, L. plantarum C5 and POM1 were selected to be inserted in gel formulations that were developed in collaboration with the R&D Innovation center of Giuliani S.p.A. (Milan, Italy). The cell densities of overnight cultures of the selected lactobacilli were calculated by measuring the OD620 using the 6715 UV/VIS Spectrophotometer (Jenway, UK). A final pellet of ca. 11 Log CFU mL-1 was collected, then resuspended in 10 mL of vegetal glycerol (Acef s.p.a., Italy), to a final cell density of ca. 10 Log CFU mL-1. The glycerol suspension containing lactobacilli was then inserted in an aqueous-based gel formulation. The final formulation contained deionized water (77%), vegetal glycerol suspension (20%) (Solagum tara, Seppic, France), containing lactobacilli (2%), and EUXYL K712 (1%), a liquid cosmetic preservative, containing sodium benzoate and potassium sorbate. The final cell density of the lactobacilli in the probiotic gel was about 9 Log CFU mL-1. The viability of the lactobacilli in the gel was assessed by preparing serial dilutions and by plating these on MRS agar medium before any treatment. A total of four gels were prepared, one for each strain previously selected, L. plantarum C5 and POM1, one with the combination of the two strains, and one with no cell suspensions.
Human Vaginal Epithelium infection and treatment
The SkinEthic™ HVE model was obtained from EpiSkin (Lyon, France). The HVE model is based on the vulvar epidermoid carcinoma cell line A431 cells which form a 3-D tissue like the human in vivo vaginal mucosa when cultivated in vitro on a polycarbonate filter in a chemically defined medium (de Brugerolle 2007). After the arrival of the HVEs, the inserts were placed in a 24-well plate, containing maintenance medium (SkinEthic, Episkin) (1 mL), and re-equilibrated for 24 h at 37°C, in a humidified, 5% carbon dioxide (CO2) atmosphere. The HVEs were inoculated with C. albicans UNIBZ54 previously adjusted to an OD600 ca. 6 Log CFU mL-1 using a maintenance medium (SkinEthic, Episkin). The HVEs were previously incubated with C. albicans UNIBZ54 (30 µL) in a humidified, 5% CO2 atmosphere, at 37°C for 24 h, to allow the pathogen adhesion and infection. At the end of the incubation period, the HVEs were treated with the gel formula previously prepared (within 24 h from the moment of preparation): two gel formulations containing the single strains of L. plantarum C5 and POM1, one gel containing the combination of these two strains, and a gel not containing any strain. Also, some HVEs were left untreated, while HVEs infected with C. albicans UNIBZ54 were used as a positive control. After the analysis, the inserts were all incubated for the same time and in the same conditions previously described.
The MTT assay for cell viability determination
The cell viability of HVEs after infection with C. albicans UNIBZ54 and treatment with the gels were determined by the MTT assay (Grela, Kozłowska, and Grabowiecka 2018). Briefly, the yellow water-soluble salt is reduced by mitochondrial dehydrogenases to purple water-insoluble formazan, according to the viability of cells. A final formazan extraction step is required, using an organic solvent (e.g., isopropanol). A stock solution of MTT (5 mg mL-1 in PBS) was diluted (1:10) in the cell culture medium to prepare the MTT solution. The MTT assay was assessed using untreated HVEs (negative control), HVEs treated with blank gel (blank), HVEs infected with C. albicans UNIBZ54 (positive control), and HVEs infected with C. albicans UNIBZ54 and treated separately with L. plantarum C5 and POM1, and with the gel containing a combination of these two strains. After 24 h of incubation, the inserts were rinsed with PBS and placed in a new 12-well plate containing MTT solution (300 µL). After 2 h of incubation in a humidified, 5% CO2 atmosphere, at 37°C, isopropanol (800 µL) was added to each well. The plate was incubated again for 1.5 h, in the same conditions above mentioned. Two aliquots (200 µL) were then taken from each well and placed in 96-well plates. The OD570 was measured with a BioTek Micro-volume Plate Reader (BioTek Instruments Inc., Bad Friedrichshall, Germany) and elaborated with the ELX808 software (BioTek Instruments Inc., Bad Friedrichshall, Germany) (reference filter: 630 nm). Results were expressed as a percentage of viability compared to the negative control (mean ± standard error of triplicate cultures), using the following formula:
viability (%) = [OD (570 nm – 630 nm) test product / OD (570 nm – 630 nm) negative control] x 100
Gene expression profiling
The total RNA was extracted from the HVEs, previously infected with C. albicans UNIBZ54 and treated with the gels, using the RNeasy mini kit (Qiagen, Valencia, CA, USA), according to De Vuyst (2014). Briefly, the circumference of the polycarbonate filter was dissected from the bottom of the insert using a sharp surgical blade and then transferred into a 12-well culture plate containing RLT buffer (600 µL). After 1 or 2 min, the stratum corneum, detached from the epidermis, was removed using a pair of tweezers and discarded. For disrupting keratinocytes, the epidermis was gently scratched with a micropipette tip in the lysis buffer, provided by the kit. The lysate was homogenized by pipetting and then transferred into a spin column placed in a 2.0 mL collection tube. This procedure allows the recovery of enough RNA from the HVEs for real-time (RT) polymerase chain reaction (PCR) analysis of gene expression. Following the instructions provided by the manufacturer, 2 µg of RNA templates were used to synthesize complementary DNA (cDNA) in a 20 µL reaction volume, using the PrimeScriptTM RT Reagent Kit (TakaraBioInc., Japan). The cDNA was amplified and detected by the Stratagene Mx3000P RT-PCR System (Agilent Technologies Italia S.p.A., Milan, Italy). PCR conditions were the following: 37°C for 15 min, 85°C for 5 sec, and 25°C for 2 min. Afterward, the TaqMan® Gene Expression Assays were carried out for RT-PCR using the following genes: the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) Hs99999905_m1, IL-8 Hs00174103_m1, the cadherin 1 (CDH1) Hs01023894_m1. The GAPDH was used as a housekeeping gene. PCR amplifications were carried out in 20 µL of total volume. The mixture of reaction contained 10 µL of 2X Premix Ex Taq (Takara, Japan), 1 µL of 20× TaqMan Gene Expression assay, 0.4 µL of RoX Reference Dye II (Takara, Japan), 4.6 µL of water, and 4 µL of cDNA.
Data and statistical analysis
All the analyses were performed considering three biological replicates analyzed in triplicate. The growth parameters of each pathogen were determined using the grofit R package (Kahm et al. 2010) for each replicate of CFSs and controls. Growth parameters were determined using the free splines approach and bootstrapping with 100 resamplings. The inhibition effect was assessed through a non-parametric one-way Kruskal-Wallis followed by the Dunn Control post hoc test provided by the PMCMR R package (Pohlert, 2016). The resulting p-values were adjusted for multiple hypothesis testing using Benjamini and Hochberg false discovery rate correction (FDR). Data from the determination of biofilm formation were submitted for analysis of variance by the General Linear Model (GLM) of R statistical package (R, version 1.6.2 rcompanion.org/handbook/). Multi-comparison of treatment means was achieved by a Tukey-adjusted comparison procedure with a p-value < 0.05 (Mangiafico 2016). For the gene expression profiling, the average value of the target gene was normalized using the GAPDH gene, and the relative quantification of the levels of gene expression was determined by comparing the Δ cycle threshold (ΔCt) value (Vigetti et al. 2008). The statistical analysis was performed using GraphPad Prism 6 (GraphPad Software Inc). Data are expressed as the mean, mean ± standard error of the mean (SEM), or mean fold change ± SEM. p-values < 0.05 were considered as statistically significant.
