Use of Lactobacillus crispatus to produce a probiotic cheese as potential gender food for preventing gynaecological infections

This research is aimed to evaluate the suitability of Squacquerone cheese to support the viability of Lactobacillus crispatus BC4, a vaginal strain endowed with a strong antimicrobial activity against urogenital pathogens and foodborne microorganisms, in order to recommend a gender food for woman wellbeing. The viability of L. crispatus BC4, used as adjunct culture, was evaluated during the refrigerated storage of Squacquerone cheese, as well as when the cheese was subjected to simulated stomach-duodenum passage tested by the patented Simulator of the Human Intestinal Microbial Ecosystem (SHIME). Moreover, the effects of L. crispatus BC4 addition were evaluated on product hydrolytic patterns, in terms of proteolysis, lipolysis and volatile molecule profiles. The data showed that L. crispatus BC4 maintained high viability, also in presence of physiological stress conditions, until the end of the refrigerated storage. Moreover, the inclusion of L. crispatus BC4 gave rise to cheese product with higher score of overall acceptability when compared to control cheese. In addition, the survival of L. crispatus BC4, carried in test cheese, in gastro intestinal conditions was confirmed by SHIME. The results showed that the vaginal Lactobacillus strain was more affected by the low pH of the stomach, simulated by the SHIME reactor, rather than to bile salts and pancreatic juices. Although only in vivo trials will be able to confirm the functionality of the cheese in the vaginal environment, these data represent a first step towards the employment of the Squacquerone cheese as probiotic food able to promote the woman’s health by preventing gynaecological infections.


Introduction
The demand of the consumers in food production has considerably changed throughout the years. In this time, foods can be perceived not only to satisfy hunger but also to directly contribute to consumer's health providing necessary nutrients, and preventing nutrition-related diseases and improving the general human well-being [1]. In this framework, one of the fastest pathogenic microorganisms of food interest, their fermentation kinetics in milk and their ability to survive at 4˚C suggested their potential use as adjunct cultures for the production of female gender foods. In this view, the production of a functional gender food based on L. crispatus represents a new strategy to vehicle a health-promoting strain to the vaginal environment. This idea is supported by the already demonstrated ability of probiotic strains to pass from intestine to vagina [37][38][39][40][41]. In this scenario, the main aim of this research was the evaluation of the suitability of a soft cheese like Squaquerone to serve as a carrier of L. crispatus BC4, a selected vaginal strain gifted of interesting antimicrobial and technological properties [31,36]. The effects of the inclusion of this Lactobacillus strain on the cheese safety, shelf-life, proteolytic and lipolytic patterns, volatile molecule and sensory profiles were evaluated, and compared to the traditional product. In addition, the viability of this strain during refrigerated storage, simulated gastro-duodenum passage and in the patented simulator of the Human Intestinal Microbial Ecosystem (SHIME) was evaluated since the ability to survive through the gastrointestinal tract is notoriously fundamental for the strain functionalities in the host. About this, SHIME system has been used as for the accurate simulation in vitro on the ability of probiotic bacteria to merge into the intestinal environment [42] Materials and methods

L. crispatus and cheese starter culture conditions
L. crispatus BC4, a vaginal isolate able to contrast Candida spp. and other pathogens of the urogenital tract [31,33,36], was used in this study for the production of a test cheese. Overnight cultures (37˚C for 16 h) of L. crispatus were obtained in de Man-Rogosa-Sharpe (MRS) broth (Oxoid, Basingstoke, UK) in anaerobic condition by using Oxoid gas generating kit. Cells were harvested by centrifugation at 8,000 × g for 20 min at 4˚C. The resultant pellet was washed twice with saline solution (0.9% NaCl in distilled water) and resuspended in commercial milk for the inoculums in industrial environment. A commercial freeze-dried culture of Streptococcus thermophilus St 0.20 (Sacco S.R.L., Como, Italy) was used as starter for Squacquerone cheese production. The culture was inoculated in milk in the optimal conditions and in the proportions indicated by the producer.

Squacquerone cheesemaking
The production of the Squacquerone cheeses (test cheese supplemented with L. crispatus BC4 and control cheese) was carried out in a pilot-scale plant of a local dairy farm (Mambelli, Bertinoro, Italy) For each cheese type, 100 liters of milk were heated and maintained at 42˚C and inoculated with the commercial starter culture of S. thermophilus (final concentration: 6 log CFU/mL). To obtain test Squacquerone cheese, L. crispatus BC4 was added (final concentration: 6.8 log CFU/mL). After 40 minutes, the two batches were added of NaCl (0.7%) and 37 mL of rennet (12000 U, 80% chymosin and 20% pepsin, Bellucci Modena, Italy). Twenty min after coagulation, the curd was cut and transferred into traditional moulds. The cheeses were let to rest until the reaching of pH 5.15 and stored at 4˚C. After 1 day, the cheeses were packed in modified atmosphere and stored at 4˚C for 18 days.

Microbiological analyses, pH and water activity
The test cheese supplemented with L. crispatus BC4 and the control cheese (without L. crispatus BC4) were subjected to microbiological, pH and water activity analyses. For the microbiological analyses, 20 grams of cheese was placed into 180 mL sodium citrate sterile solution (20 g/L) and homogenised in a stomacher (Lab-blender 80, Pbi International, Milan, Italy) for 3 min. Decimal dilutions were performed in Ringer's solution (NaCl 0.9%), and 0.1 mL of appropriate dilutions was spread onto different agar media. S. thermophilus was counted on M17 agar (Oxoid, Basingstoke, Hampshire, UK) (42˚C, 48 h), Lactobacillus spp. were detected on MRS agar (37˚C, 48 h, anaerobic conditions) and the presence of L. crispatus was assured also by checking the bacterial colony morphology. To confirm its presence, genomic DNA was extracted from each L. crispatus presumptive colony using the InstaGene Matrix kit (Bio-Rad Laboratories, Milano,Italy) and by sequencing the 16S rRNA region according to the protocol described by De Angelis et al. [43]. Enterobacteriaceae and yeasts were also counted on VRBGA (Oxoid, Basingstoke, Hampshire, UK) (37˚C, 24 h) and YPD (Oxoid, Basingstoke, Hampshire, UK) (25˚C, 48 h) plates, respectively. The presence of pathogenic species such as L. monocytogenes, S. enteritidis and E. coli was checked in both the cheeses during their refrigerated storage, according to the ISO methods 11290, 6579, and 16649, respectively. pH and water activity were measured by diluting 10 g of cheese with 10 mL of distilled water. pH was registered by using a pH-meter Hanna Instruments 8519 (Incofar, Modena, Italy). Water activity (Aw) was measured by using an Aqualab Series 4TE (Decagon Device, Inc. Pullman,WA, USA).

Proteolysis, lipolysis, volatile profiles
The test cheese supplemented with L. crispatus BC4 and the control cheese (without L. crispatus BC4) were subjected to proteolysis, lipolysis and volatile profile.
The proteolysis of the two cheese types was monitored by SDS-PAGE. The soluble proteins at pH 4.6 were extracted from the supernatants of the two cheese types following the method proposed by Kuchroo and Fox

Textural profile analyses
The test cheese supplemented with L. crispatus BC4 and the control cheese (without L. crispatus BC4) were subjected to textural analysis over the refrigerated storage. Texture analyses were performed after 1, 4, 6, 8 and 13 days of refrigerated storage using a Texture Analyser TA DHI (Stable MicroSystem, Godalming, UK) and according to the method proposed by Patrignani et al. [2].

Organoleptic evaluation
The test cheese supplemented with L. crispatus BC4 and the control cheese (without L. crispatus BC4) were subjected to sensory evaluation over the refrigerated storage. The sensory evaluation was performed throughout a panel test after 4 and 13 days of storage at 4˚C. Twenty-five trained panellists were enrolled within the employees of Mambelli (Bertinoro, Italy) local dairy farm (5 panellists) and of the Department of Agricultural and Food Sciences, University of Bologna, Campus Food Science, Cesena, Italy (20 panellists). The Campus of Food Science, where the test was organized, is provided by sensory-laboratories certified according to regional requirements. The panellists were verbally informed on the cheese product manufacture and the strains used in the research by the committee organizing the test. The committee was formed by Dr. Patrignani, Dr. Siroli and Prof. Lanciotti. The panellists, after oral consent to perform the test, tasted 20 g of each sample served at 15˚C under controlled conditions according to Standard 8589 (ISO, 1988), as suggested by Gallardo-Escamilla et al. [48]. The assessors evaluated cheese colour, flavour, creaminess, off-flavours, bitter and overall acceptance attributing a score ranging from 0 (low or poor) to 5 (high or very excellent).

Evaluation of the fate of Lactobacilli population (mainly represented by L. crispatus BC4) by Simulated stomach duodenum passage and the Simulator of the human intestinal microbial ecosystem (SHIME)
The fate of L. crispatus BC4 was performed by the simulation of the passage through stomach and duodenum, developed according to the method proposed by Vinderola et al. [49]. Briefly, 25 g of tested cheese was mixed with the same volume of a 'saliva-gastric' solution. Saliva-gastric solution contained CaCl 2 (0.22 g/L), NaCl (16.2 g/L), KCl (2.2 g/L), NaHCO 3 (1.2 g/L) and 0.3% (w/v) porcine pepsin. Porcine pepsins from two different and undisclosed manufacturers were used on a g/mL basis because no information about the enzymatic units was available on the label of the products. A 1 mL sample was removed for cell counts immediately after admixture and pH was quickly brought to 3.00, 2.70 or 2.50, with 5 M and 0.1 M HCl. Samples were brought to 37˚C in a water bath. Aliquots (1 mL) were taken at the time fixed by the method and serial dilutions were plated for cell viability assessment.
The destiny of the vaginal strain L. crispatus BC4 during passage in the human upper gastrointestinal tract when delivered within a cheese matrix was evaluated also in an adapted SHIME system [50]. This system, composed of a reactor used in a sequential setup to simulate over time first the stomach and then the small intestinal environment, has been used as for the accurate simulation in vitro on the ability of probiotic bacteria to merge into the intestinal environment [42].
Appropriate retention time and pH were chosen in order to resemble in vivo conditions in the different parts of the gastrointestinal tract in fed state. Control and test cheeses were incubated as reported below to simulate gastric and upper intestinal phases, both working in fed state. 2.5 g of each cheese were used in the simulation, all experiments were conducted in triplicate to account for biological variability. Along the simulation, the following samples were collected: time 0, end of stomach, small intestine after 1, 2 and 3h. Viability of the L. crispatus strain was evaluated by means of conventional culturing techniques using selective medium (MRS). To confirm its presence, the procedure previously described was used [43]. The SHIME consists of a succession of reactors simulating the different parts of the human gastrointestinal tract. In this research, the SHIME reactor simulates over time, firstly the stomach and then the small intestine. The applied conditions used in the simulator are summarized below: Gastric phase (fed state): the cheese matrix was incubated at 37˚C for 2 h at the SHIME patent conditions [50], while mixing via stirring, with sigmoidal decrease of the pH profile. Pepsin was supplied with the activity being standardized by measuring absorbance increase at 280nm of TCA-soluble products upon digestion of hemoglobin (reference protein). Phosphatidylcholine was added and followed by the addition of SHIME complex nutritional medium. The salt (NaCl and KCl) levels were implemented according to Mackie et al. [51].
Small intestinal phase (fed state): The small intestine phase was characterized by an incubation of the cheese matrix, after gastric phase, at 37˚C for 3 h, while mixing via stirring. In this step, pH increased to 7.4. Pancreatic enzymes, in a specific and defined ratio, were used according to the patent. 10 mM bovine bile extract was supplemented.

Statistical analysis
The data are expressed as the mean of three repetitions and two independent experiments. Microbiological, chemico-physical and texture data were analyzed using Statistica software (version 8.0; StatSoft, Tulsa, Oklahoma, USA). Means were compared using ANOVA followed by LSD test at p < 0.05 level to mark differences between samples with the same storage times. The volatile molecule profiles were analyzed using ANOVA followed by a principal component analysis (PCA) performed by Statistica software.

Microbiological analyses
Test Squacquerone cheeses were produced in the pilot-scale plant of a local Italian cheesemaker using L. crispatus BC4 strain as adjunct culture with the starter one of S. thermophilus. The test cheeses were compared to the control ones, produced in the same conditions with the starter culture alone. The cell loads of the starter (S. thermophilus) and of Lactobacillus population were monitored during the refrigerated storage (13 days) and reported in Fig 1, respectively. The results highlighted the constant presence of S. thermophilus over the whole storage period, independently on the presence of the adjunct culture. Notably, the inclusion of L. crispatus BC4 in the cheese positively affected the starter culture viability over the storage. In fact, in the test cheese, the initial count (1d) of S. thermophilus was 7.4 log CFU/g and decreased to 5.9 log CFU/g at the end of the storage (13d), while in the control one, S. thermophilus had an initial cell load of 6.1 log CFU/mL and it reduced to 5.0 log CFU/g ( Fig 1A). Fig 1B reports the cell loads of Lactobacillus population over the storage period. The levels of lactobacilli in the control cheese increased from 2.1 log CFU/g (1d) to 5.0 log CFU/g (13d). By contrast, lactobacilli cell load was quite constant in the test cheese and always over 7.0 log CFU/g. According to the morphological and molecular analysis, the most abundant Lactobacillus population was represented by L. crispatus BC4, that was inoculated at a level of 6.8 log CFU/g. On the contrary, L. crispatus was not found among the Lactobacillus population detected in the control cheese. No contamination by yeasts, Enterobacteriaceae and pathogenic species was found.

Physico-chemical analyses
The pH and water activity values of Squacquerone cheeses, in relation to storage time and L. crispatus BC4 presence, are reported in Table 1. During the storage, the pH decreased both in test and control cheese, even if, at the end of the storage, the pH of the cheese containing L. crispatus BC4 was lower than the control. Regarding water activity, the two cheese types presented significant different values at the beginning of the shelf-life while after 13 d of storage no significant difference was recorded between the two cheese types.

Cheese ripening profiles
The proteolytic profiles of the two types of Squacquerone cheese, containing or not L. crispatus BC4, are represented in the heat map reported in Fig 2. The dendrogram of the cheese proteins soluble at pH 4.6 showed similar weight molecular bands for the two cheese types up to 6 days of storage. As the storage time increased, significant differences between the test and control cheeses were detected. In fact, after 13 days of refrigerated storage, the cheese containing L. crispatus BC4 showed a proteolytic profile very similar to that of the control cheese after 8 days of storage, with the exception of the band corresponding to 66 KDa. This protein band, corresponding to the molecular weight of bovine serum albumin, characterized the test cheese while it was absent in the control cheese, suggesting that the activity of the adjunct culture affected the matrix protein network and the whey separation and, consequently, the retention of whey proteins. This hypothesis is also corroborated by the rheological data showing that cheese containing L. crispatus BC4 was characterized by a significantly higher creaminess, with Lactobacillus crispatus for probiotic cheese production respect to the control cheese, also immediately after cheesemaking. Also the faster decrease of the pH during storage in test cheese probably contributed to modify the water binding capacity of the proteins and their interaction with the other macromolecules of the system. In fact, the pH has a key role in the interaction between proteins and other macromolecules [52]. By contrast, the low molecular weight bands differed between the two cheese typologies only for their abundance (Row-Z-score). More specifically, peptides having molecular weights ranging between 21.07 and 14.18 KDa were more abundant in the control cheese.
The cheese FFA contents recorded between 4 and 13 days of storage at 4˚C, are shown in Table 2. After 4 days of storage, the control cheese presented FFA profiles characterized by C12:0, C14:0, C16:0, and C18:1, while the test cheese was characterized by C10:0, C14:0, C16:0, C18:0 and C18:1.Over the storage, in both cheese typologies the release of FFA increased, even though with different trends in relation to the presence of L. crispatus. The C18:0 was detected in the control cheese after 6 days of storage. Branched anteiso 17 carbon atoms, as well as long-chain fatty acids such as C18:2 and C20:0 appeared in test cheese type between 6 and 13 days of storage. Both in the control and in the test cheese, also 15 carbon atom FFAs and C18:2 appeared after 6 days of storage. However, the presence of L. crispatus BC4 fastened the release of FFAs, whose concentrations peaked at 8 days of storage, and then decreased probably due to their further transformation in aroma compounds. By contrast, in control cheese the accumulation of the reported FFAs continued over time and peaked at the end of storage.
The volatile molecule profiles of the two types of Squacquerone cheese were monitored over the refrigerated storage using the GC/MS/SPME technique. The detected molecules belonged to different classes of compounds such as alcohols, esters, acids and ketones. To evaluate the consequences of the inclusion of L. crispatus BC4 over time, the detected volatile compounds were analysed using a principal component analysis (PCA). In Fig 3, the projection of samples and detected molecules is reported. The PCA was able to explain over 60% of the total variance among the samples over time. All the test cheese samples, independently on the time of storage, clustered together, with the exception of the test cheese after 13 days of storage that was very similar to the control one at the same time of storage. At 1, 4 and 6 days of storage, the test cheese was principally characterized by ketones and alcohols, while the control cheese was characterized by ketones, short fatty acids and esters. At the end of the shelf-life, 2-butanone, propanone and, at minor amount, ethanol were found in both types of cheese.

Panel test
The sensory analysis was performed on the two types of cheese after 4 and 13 days of storage and showed that test cheese was preferred compared to the control (Fig 4). In fact, after 4 days, test cheese received a significant higher score compared to the control cheese for the overall acceptance. This positive evaluation was principally affected by sensory features such as creaminess, flavour, and lacking of off-flavours, balancing the low scores received for the colour attribute. After 13 days of storage, the sensory differences perceived by the panellists were significantly reduced, although the test cheese still remained more appreciated than the control for all the parameters considered.

Textural profile
The textural analyses of hardness and consistency over refrigerated storage is reported in Fig 5, respectively. Regarding hardness, significant differences between the two types of Squacquerone cheese were assessed for any storage time considered (1-13 days). Consistency was found significantly different between the two cheese products up to 8 days of refrigerated storage.

Evaluation of the fate of test cheese Lactobacilli population (mainly represented by L. crispatus BC4) in simulated stomach-duodenum passage and in the Simulator of the Human Intestinal Microbial Ecosystem (SHIME)
A simulation of the passage of the test cheese through stomach and duodenum was carried out, and the surviving levels of Lactobacilli (mainly represented by L. crispatus BC4) were evaluated. The presence of L. crispatus was verified by genomic DNA extraction from colonies and sequencing the 16S rRNA region according to the method previously indicated. At 4 and 13 days of refrigerated storage, test cheese were subjected to different chemical stresses. The simulated physiological stresses applied on the test cheese after 4 days of storage showed that L. crispatus BC4 (added in the test cheese at level of 7 log CFU/g) was more affected by the application of gastric stress conditions than the duodedum stress. In fact, when gastric condition was applied to the test cheese, containing Lb. crispatus BC4, after 4 days of storage, its Lactobacilli cell load values decreased of 1 log cycle (Fig 6). When the test cheese after 13 days of storage was subjected to the stomach duodenum passage, the viability of Lactobacilli population were more affected by the application of intestinal juice stress.
Results related to the survival of the Lactobacillus population in the upper gastro intestinal tract under simulated fed conditions in SHIME are shown in Fig 7, which reports the plate count cell loads of lactobacilli detected before and after the exposure to stomach condition, and 1, 2 and 3 hours after the exposure to the small intestine conditions. Plate count data on MRS medium showed that the transit in the stomach led to a decrease of approx. 3 Log units in lactobacilli viability in the test cheese. During transition in the small intestine, Lactobacillus population levels remained constant in the test cheese. The presence of L. crispatus, at the highest cell loads, was verified by genomic DNA extraction and sequencing the 16S rRNA region according to the method previously indicated. These data showed that the vaginal strain was initially significant affected by the low pH in the stomach while it was resistant to bile salts and pancreatic juices.  (1, 4, 6, 13 days). CC1: control cheese after 1 day of storage; CC4: control cheese after 4 days of storage; CC6: control cheese after 6 days of storage; CC13: control cheese after 13 days of storage; PC1: Test cheese after 1 day of storage; PC4: Test cheese after 4 days of storage; PC6: Test cheese after 6 days of storage; PC13: Test cheese after 13 days of storage. https://doi.org/10.1371/journal.pone.0208906.g003

Discussion
The present research is part of a wider project aimed to use food as strategy to increase woman well-being. Generally, since vaginal dysbiosis and other gynaecological infections are generally characterised by a reduction in vaginal lactobacilli content, the administration of probiotics, especially lactobacilli, has been proposed as a valid treatment for vaginal disorders. In fact, a wide literature reports that the administration of probiotics is able to restore the normal vaginal microbiota [53][54][55][56]. Beside the vaginal application, a wide literature shows that probiotic bacteria can be administered orally, also through foods, due to their ability to pass from the intestine to the vagina, with a consequent beneficial impact on the vaginal habitat [37-41]. As an alternative to pharmaceutical or nutraceutical preparations, the present research aimed to develop a probiotic soft cheese as dietary strategy to increase woman wellbeing and particularly to prevent or solve dysbiosis disorders.
In this framework the results of the present study show that Squacquerone cheese can represent a suitable carrier of L. crispatus, able to preserve the strain viability during the refrigerated storage and the simulated gastro intestinal digestion, respecting the hedonistic features of foods. In fact, L. crispatus BC4 maintained levels of about 7 log CFU/g up to 13 days of refrigerated storage. These cell loads allow a daily intake of 100 g of cheese to introduce the probiotic cell amount recommended by FIL-IDF (International Dairy Federation) that, depending on the probiotic strain, range between 8 and 12 log CFU/day. Consequently, the strain herein considered, fulfills two of the main criteria to select probiotic microorganisms to be included in food systems: i) the compatibility with eventual starter culture during cheese making and ii) the maintenance of cell loads of at least 6 log CFU/g throughout the shelf life [61,62]. In addition, the Squacquerone cheese matrix resulted able to assure the protection of the L. crispatus against the simulated physiological stresses (low pH, high bile salt concentration, presence of digestive enzymes) characterizing the digestion process. In fact, at the end of the simulated stomach-duodenum passage, the strain showed a viability higher of 6 log CFU/g, independently on the storage time considered. Soft cheeses like Squacquerone or Crescenza are considered optimal carrier and dietary strategy to transfer probiotic strains as they are characterised by a protective protein and fat matrix, neutral or sub-alkaline pH, and high buffering capacity, properties able to protect the probiotic bacteria during their gastro-intestinal transit [47,62,63]. In addition, these cheeses can represent a good source of calcium and vitamins, particularly important for the woman wellbeing. The test cheese containing L. crispatus BC4 was also subjected to the Simulator of the Human Intestinal Microbial Ecosystem (SHIME). These reactors simulated, over time, firstly the stomach and then the small intestine under fed conditions. Differently from the simulated stomach duodenum passage, the incubation conditions and timing are optimized in the SHIME system in order to resemble in vivo situation in the different regions of the gastrointestinal tract. The data obtained confirmed the viability of L. crispatus BC4, to survive to the in vivo stress that the strain can encounter during the digestion process, evidencing its higher sensitiveness toward gastric conditions with respect to bile salts and pancreatic juice.
Notably, the inclusion of the L. crispatus BC4 improved the organoleptic properties of the cheese in terms of creaminess, flavour and overall acceptance. This is an important feature since the probiotic products are often characterized by lack of good sensory properties and several strategies were proposed to overcome such probiotic food drawbacks [64,65]. In fact, L. crispatus BC4 strain affected, although at levels compatible with the short ripening time, both the proteolytic and lipolytic patterns of cheese and, consequently, the volatile molecule profiles. These modifications resulted in a product appreciated by the panellists involved in the sensory analysis.

Conclusions
Although only in vivo clinical trials can definitely confirm the functional properties of a food, these data represent the first important tassel in the use of the Squacquerone cheese, containing L. crispatus BC4, as dietary strategy for woman well-being. The novelty of this work consists in the use of a Lactobacillus species, exclusively associated with the vaginal habitat and recognized as healthy marker, for the production of a probiotic food that can promote the woman's health by preventing gynaecological infections. At present the use of L. crispatus has been reported only in pharmaceutical formulations; for this reason, our work is a pioneer in the proposal of using this health-promoting vaginal species in the preparation of functional foods of gender.
However, next step of this research will include clinical studies to verify the potential of this functional cheese in humans to prevent vaginal diseases.