In vitro and in vivo activities of lactic acid bacteria from Italian mountain cheese" and their exploitation in dairy production

The use of health-promoting lactic acid bacteria (LAB) strains as starter or adjunct cultures for dairy productions could facilitate the in situ bio-synthesis of bioactive molecules during the fermentation process, increasing the interest towards dairy products as multifunctional foods. Currently, there is much research about genotypic and technological characterization of raw milk cheeses microbiota, which is rich in biodiversity and could be exploited for improving the sensory attributes and add healthy benefits to the cheese. Traditional Mountain (TM) cheese is made from raw cow’s milk and spontaneously fermented in small farms called “Malga” located in the alpine areas of Trentino region. For the first time, the microbial population of TM-cheese has been characterized in order to select cocci and non-starter LAB suitable for developing new starter or adjunct cultures, respectively. Samples (n = 120) of milk, curd and cheese at different ripening times (24 hours, 1 month and 7 months) were enumerated in selective culture media. Mesophilic and thermophilic cocci dominated during the first 24 hours following production, and mesophilic lactobacilli were dominant at the end of ripening. Six hundred and forty colonies were isolated from curd and cheese 24 hours following production, and 95 more colonies were isolated from cheese after 7 months of ripening. All isolates were genotypically characterized by Randomly Amplified Polymorphic DNA-Polymerase Chain Reaction (RAPD-PCR) with two primers, species-specific PCR and partial sequencing of 16S rRNA gene. Cocci clustered in 231 biotypes belonging to 16 different species, and non-starter LAB (NSLAB) clustered in 70 biotypes belonging to 13 different species. Lactococcus lactis, Streptococcus thermophilus and Enterococcus faecalis were dominant in curd and 24h-cheese; Pediococcus pentosaceus and Lactobacillus paracasei were the main species at the end of ripening. The phenotypic, technological and health-promoting activities of all strains were investigated. In particular, lactococci, streptococci and enterococci were tested for their acidification and proteolytic activity, ability to growth at not optimal temperatures, acetoin production, development of olfactory flavour notes, autolysis rate and ability to inhibit the growth of coliforms. Forty percent of enterococci showed the ability to inhibit raw milk resident coliforms in vitro, but they were excluded as possible starters, owing to the presence of associated risk factors. Among lactococci and streptococci, 4 Lc. lactis subsp. lactis and 2 Sc. thermophilus were fast acidifiers, produced pleasant flavours, and were subjected to the freezedrying stability test. Lc. lactis subsp. lactis 68 and Sc. thermophilus 93 showed the best properties and might be appropriate for cheese production. NSLAB strains were tested for their growth properties, carbohydrate metabolism, acidifying ability, proteolytic and lipolytic activities, acetoin production, amino-peptidase activity (AP) and biogenic amines production. Concerning the health-promoting properties, the bile salts hydrolysis (BSH) activity was tested qualitatively, the conjugated linoleic acid (CLA) production was measured spectrophotometrically, and the γ-aminobutyric acid (GABA) production was quantified by UHPLC (Ultra High Performance Liquid Cromatography). Lb. paracasei isolates resulted to be well adapted to the Malga environment and showed the highest AP activity and acetoin production. Some strains harbored very interesting health-promoting properties and produced bioactive substances. In particular, Lb. rhamnosus BT68, Lb. paracasei BT18, BT25, BT31, Pc. pentosaceus BT3, BT13, BT51 produced between 70 and 130 mg/mL of total CLA in vitro. Lb. brevis BT66 converted L-glutamate to a high concentration of GABA (129 ± 8.6 mg/L) and showed BSH activity. These first results revealed that TM-cheese is a reservoir of a high microbial diversity, and the resident LAB could be exploited not only for the applicability in dairy production but also for their health-promoting properties. Lc. lactis subsp. lactis 68 and Sc. thermophilus 93, which showed to be the best performing strains, were tested as starter and adjunct cultures, for the production of 9 experimental TM-cheese wheels in a Malga-farm, respectively. Three control (CTRL) cheeses were produced according to the tradition and any starter or adjunct culture was not added; three starter (STR) and three commercial starter (CMS) cheeses were produced inoculating the vat milk with both selected strains and a commercial Sc. thermophilus strain, respectively. After 24 hours, 1 month and 7 months of ripening the microbial content of all experimental cheeses was investigated. Mesophilic cocci and lactobacilli dominated in cheese samples after 24 hours and 1 month of ripening, while cocci dominated in full-ripened cheese. The total genomic DNA was extracted, and a fragment of the V1-V3 region was amplified and pyrosequenced. Lactococci and streptococci were the most abundant species in CTRL and STR cheese, and Lc. lactis subsp. lactis 68 affected the proliferation of the (raw milk) indigenous Lc. lactis subsp. cremoris during the early fermentation. Moreover, the commercial Sc. thermophilus showed to be dominant towards Lc. lactis subsp. lactis and cremoris naturally present in raw milk and to be responsible in decreasing the abundance of Lactobacillus subp. and Enterococcus sp. The survival of TM-cheese microbiota in vitro was investigated under simulated human gastro-intestinal (GI) conditions. The 9 full ripened experimental TM-cheeses were subjected to a model system that simulates digestive processes in the mouth, stomach and small intestine, comprising sequential incubation in human gastric and duodenal juices. Bacterial counts were performed before and after the simulation: total bacterial count and thermophilic cocci significantly decreased after the simulated digestion. Thirty-six lactobacilli were isolated from cheese after digestion: among them 1 Lb. paracasei, 1 Lb. parabuchneri and 1 Lb. fermentum were tested for their survival after GI transit. Lc. lactis subsp. lactis 68 and Lb. parabuchneri D34 strains were used to ferment whole milk and digested. The load of Lb. parabuchneri D34 decreased by about one log more when grown as pure culture than fermented milk after simulated digestion, suggesting that Lb. parabuchneri D34 had in itself the ability to survive digestion, but the fat content and the cheese structure might protect LAB during the GI transit. Furthermore, our interest towards the GABA producing strains lead us to test the ability of Lb. brevis BT66 to produce GABA in situ during cheese production, through the decarboxylation of glutamate. Twenty experimental micro-cheeses were produced using a commercial starter strain (107 CFU/mL) and Lb. brevis BT66 as adjunct culture. Four different concentrations (102 , 103, 104, 105 CFU/mL) of Lb. brevis BT66 were tested in quadruplicate. In order to follow the microbial evolution, samples of milk, curd and cheese after 20 days of ripening were enumerated in selective media. The control and experimental samples showed a similar trend, suggesting that both milk-resident and starter strains grew during ripening. However, the load of mesophilic lactobacilli in all experimental curd samples was higher than the control ones. The concentration of GABA and glutamic acid in cheese samples after 20 days of ripening was quantified by UHPLC-HQOMS. The amino acidic profiles showed that while the concentration of Lb. brevis BT66 in milk increased, the amount of both glutamic acid (from 324 ± 37 to 202 ± 32 mg/kg) and GABA (from 154 ± 31 to 91 ± 20 mg/kg) significantly decreased in cheese. These results suggested that the experimental strain converted glutamic acid to GABA, but that GABA may have subsequently been converted to succinate by GABA transaminases. The non-protein amino acid GABA has been reported to impact on brain function through the gut:brain axis system, to harbor an anti-obesity and antidiabetogenic effect, to regulate the immune system, the inflammation process and the energy metabolism in mammals including induction of hypotension, diuretic and tranquilizer effects, stimulation of immune cells. Owing to its ability to produce high concentrations of GABA and its BSH activity in vitro, Lb. brevis BT66 was selected to be tested in vivo in mice suffering obesity-associated type-2-diabetes. Another Lb. brevis (strain DPC6108), isolated from the human GI tract and harboring the same properties, was simultaneously investigated. The corresponding rifampicin resistant mutants (rif) were generated; their genotypic profile was obtained by RAPD-PCR and PFGE (Pulsed-Field Gel Electrophoresis) and was identical to the native strain. The conversion rates of monosodium glutamate to GABA were investigated by nextgeneration amino acid analysis: Lb. brevis BT66rif produced 840.5 ± 266 µg/mL of GABA with about 73% of bioconversion and Lb. brevis DPC6108rif produced 1,218.0 ± 393.2 µg/mL with about 87% of bioconversion. The BSH activity was positive to both qualitative and quantitative assays and the results were similar in both native and mutant strains. The rifampicin resistant strains were freeze-dried and tested for their stability at room temperature, +4 and -20 °C. Both spectrophotometer and plate count methods revealed that freeze-dried strains survived at room temperature during 24 hours after suspending in sterile water. The stability of freezedried strains at +4 and -20 °C was investigated enumerating the viable cells in selective medium during 10 weeks and any significant load reduction was not detected in the first 4 weeks following freeze-drying. Both pharmabiotic-producing Lb. brevis BT66rif and DPC6108rif were resistant to freeze-drying, survived transit through mouse GI tract (as proven by a pilot study), and their therapeutic efficiency is being assessed in vivo to treat metabolic obesity and type-2-diabetes