Functional characterization and immunomodulatory properties of Lactobacillus helveticus strains isolated from Italian hard cheeses

Lactobacillus helveticus carries many properties such as the ability to survive gastrointestinal transit, modulate the host immune response, accumulate biopeptides in milk, and adhere to the epithelial cells that could contribute to improving host health. In this study, the applicability as functional cultures of four L. helveticus strains isolated from Italian hard cheeses was investigated. A preliminary strain characterization showed that the ability to produce folate was generally low while antioxidant, proteolytic, peptidase, and β-galactosidase activities resulted high, although very variable, between strains. When stimulated moDCs were incubated in the presence of live cells, a dose-dependent release of both the pro-inflammatory cytokine IL-12p70 and the anti-inflammatory cytokine IL-10, was shown for all the four strains. In the presence of cell-free culture supernatants (postbiotics), a dose-dependent, decrease of IL-12p70 and an increase of IL-10 was generally observed. The immunomodulatory effect took place also in Caciotta-like cheese made with strains SIM12 and SIS16 as bifunctional (i.e., immunomodulant and acidifying) starter cultures, thus confirming tests in culture media. Given that the growth of bacteria in the cheese was not necessary (they were killed by pasteurization), the results indicated that some constituents of non-viable bacteria had immunomodulatory properties. This study adds additional evidence for the positive role of L. helveticus on human health and suggests cheese as a suitable food for delivering candidate strains and modulating their anti-inflammatory properties.


Introduction
Lactic acid bacteria (LAB) are so far the main used microorganisms in functional foods, which can be defined as natural or processed foods that contain biologically active compounds which

Folate production
Lactobacillus helveticus strains were screened for the ability to produce folate. After overnight growth at 32˚C in MRS, re-activated cultures were washed three times with a NaCl solution (0.85% w/v). Resuspended cells were used to inoculate (2%; v/v) a folate-free medium (Folic Acid Casei Medium, Difco, Sparks, Maryland, USA). Cultures able to grow in the folate-free mediaum were sub-cultured three times in the same medium before the determination of folate concentration in the supernatants (ng/ml). Tests were carried out in triplicate by a microplate microbiological assay according to Laino et al. [17], using Lacticaseibacillus rhamnosus ATCC 7469 as auxotrophic strain in folate determination, as it can growth only in presence of folate. Briefly, cells from an overnight culture of the indicator strain were harvested by centrifugation, washed with 0.85% of NaCl, inoculated (2% v/v) in the folate-free medium, and loaded into 96-wells microplates. A calibration curve was performed, adding folic acid (Merck, Darmstadt, Germany) standard solution to each inoculated well. Filter-sterilized (0.22 mm) culture supernatants were diluted with the extraction buffer and added to each inoculated well. Both standards and extracted samples were loaded in triplicate. Microplates were incubated for 48 h at 32˚C and the OD650 was scored. Total folates were quantified according to the cubic polynomial function given by the standard curve. The values obtained were expressed as ng/ml.

Antioxidant activity in milk
After overnight growth at 42˚C in MRS broth, the lactobacilli were sub-cultured twice in sterile, reconstituted skimmed milk (10% w/v) to prevent any interference of the MRS medium in the determination of the antioxidant activity (AO). After 48-72 hours of incubation, the coagulated cultures were centrifuged and the supernatants used for the determination of AO, which was carried out by measuring the radical scavenging activity, the ferric reduction capacity (FRAP), and the accumulation of free thiol groups according to Perna et al. [18]. The radical scavenging activity was carried out by the ABTS (2,2'-azino-bis-3-ethylbenzothiazoline-6-sulfonic acid) method and expressed as percentage inhibition (I%) of the ABTS�+ radical, whereas FRAP and free thiol groups values were expressed as μM Fe(II) and μM-SH, respectively. The cultures in reconstituted skimmed milk were carried out in triplicate and the values were averaged.

Enzymatic activities
The o-nitrophenyl-ß-D-galactopyranoside (Sigma-Aldrich, Steinheim, Germany) substrate was used to determine the ß-galactosidase activity as described by Vinderola and Reinheimer [19]. The proteolytic activity in sterile skimmed milk was evaluated by the o-phthaldialdehyde (OPA) method [20]. Briefly, after incubation of the strains in sterile skimmed milk at 37˚C for 48 h, proteins in 1 ml milk culture samples were precipitated. After centrifugation and filtration of the resulting supernatants, the o-phthaldialdehyde reagent (Sigma-Aldrich, Steinheim, Germany) was added and the proteolytic activity was measured by OD reading at 340 nm (Jasco, Great Dunmow, UK). The results, which were the average of triplicate samples, were expressed as μg/ml of α-amino groups released in the medium. Glycine (Fluka, Steinheim, Germany) was used as standard to build the calibration curve (concentration range 0-75 μg/ ml). Aminopeptidase activity was evaluated as described by Gatti et al. [21]. Briefly, cells were resuspended at an OD 550 among 1.2-1.5 and incubated with 0.656 mM/l solutions of Phe-Pro-βNa, Arg-βNa, Lys-βNa, Leu-βNa and Pro-βNa substrates (Bachem Feinchemikalien AG, Bubendorf, Switzerland) for 30 min at 40˚C. The activities were evaluated after OD 580 reading and results expressed as μM/l β-naphthylamine/ml culture/h. The determinations were performed in duplicate in independent experiments.

Cheese production and sampling
Strains SIM12, SIS16, Lh43 and 1734 were evaluated for their acidification performance, which was tested during growth in pasteurized microfiltered whole milk (PM milk

Experiments with human PBMC and dendritic cells
Human peripheral blood mononuclear cells (PBMC) were isolated from buffy coats of healthy donors after informed consent, according to the institutional guidelines and approved by the institutional Ethical Committee of the Humanitas Clinical and Research Center (approval date 28 th January 2016). Monocytes-derived DC (moDCs) were derived from human peripheral blood monocytes selected with anti-CD14 antibodies coupled to magnetic beads (Miltenyi Biotec, Bologna, Italy). Positive selected CD14 + cells were incubated for 6-7 days in complete medium containing 1000 U/ml granulocyte-macrophage colony stimulating factor and 400 U/ ml interleukin-4 (Miltenyi Biotec) to obtain immature moDCs. MoDCs were stimulated with strain FB62 (MOI bacteria:moDCs = 10:1) or 10 ng/ml lipopolysaccharides (LPS) (Sigma-Aldrich, USA) and incubated in the presence of bacterial cell cultures (or their postbiotics, prepared as described above) of strains SIM12, SIS16, Lh43, and 1734 cultivated in IMDM medium. Stimulated moDCs without bacterial cells or postbiotics were used as controls. Bacterial cells were also incubated with non-stimulated moDCs. In other experiments, total PBMCs were incubated in the presence of CS of both CNTR (see above) and experimental cheeses, i.e., cheeses fermented by strain SIM12 and SIS16. The CS of experimental cheeses were used directly (TQ-CS) or after pasteurization at 63˚C for 30 min (P-CS). Viable cells in TQ-CS and P-CS samples were enumerated in MRS agar. After 1 h incubation and extensive wash, gentamicin (100 mg/ml) was added. After 24 h incubation, the amount of accumulated IL-12p70 heterodimer (moDC) or IL-12p40 chain (PBMC) and IL-10 cytokines was determined by ELISA according to the manufacturer's instruction (R&D Systems).

Statistical analyses
Statistical analyses for all figures (Figs 1-3) were performed using GraphPad Prism version 8 for macOS (La Jolla California USA). The comparison of multiple groups was carried out by two-way ANOVA, followed by Bonferroni's or Tukey's post-test for correction of multiple comparisons as indicated in figures legend. The data probability value of � P < 0.05 was considered significant.

Strain characterization
Results on strain characterization are shown in Table 1

Cytokine production in culture media
When moDCs were stimulated by LPS (or Salmonella FB62) and incubated in the presence of L. helveticus cells, an increased release of both IL-12p70, a pro-inflammatory cytokine which acts as a bridge between innate immunity and adaptive Th1 type immune response (Fig 1, panels b, c), and the anti-inflammatory cytokine IL-10 was retrieved (Fig 1, panels e, f). The production of IL-10 was approx. 10-times higher than that of IL-12p70 in LPS-stimulated moDCs. A dose-dependent effect was observed for all the bacterial strains: the smaller was the ratio bacteria:moDC, the lower was the cytokine production. In the absence of stimulus, strain SIM12 and, especially SIS16, induced a detectable production of IL12p70 with bacteria:cells ratio of 10:1, while a very low production of IL-10 was detected (Fig 1, panels a, d). Fig 2 shows the production of IL-12p70 and IL-10 cytokines (expressed as fold change normalized to LPS and FB62) when moDCs stimulated by LPS or Salmonella FB62 were treated with different percentages of culture supernatants (postbiotics) of strains SIM12, Lh43, SIS16, and 1734. Postbiotics showed a dose-dependent, and anti-inflammatory effect on both LPS- (Fig 2, panels a,  b) and Salmonella driven (Fig 2, panels d, e) inflammatory response, by inhibiting the production of IL-12p70, while strongly increasing the release of IL-10. Postbiotics from all the strains showed different potency in the modulation of IL-12p70 and IL-10 release with SIM12 and Lh43 having the highest anti-inflammatory index (Fig 2, panels c, f).

Cytokine production in cheese
Although all four strains showed in vitro immunomodulatory activity, their ability to acidify in milk was preliminary tested to see if they could be used in cheesemaking with the dual function of starter and functional cultures. Strains SIM12 and SIS16 showed a fast-acidifying activity, reaching pH 4.0 in about 10 h., while strains Lh43 and 1734, which were very slow acidifiers, were discharged (S1 Fig). The immunomodulating capacity of the cheeses obtained with SIM12 and SIS16 (S2 Fig) was then tested. To this end, cheese samples were solubilized with Na-citrate to obtain cheese suspensions (CS), which were tested both untreated (TQ-CS, with a viable count of approx. 10 8 CFU/ml, S1 Table) or after pasteurization (P-CS, 10 3 CFU/ml, S1 Table). Results (expressed as % response normalized to LPS) showed that cheese samples, , or homogenized and pasteurized cheese (P-CS) inoculated with Lactobacillus helveticus strains SIM12 and SIS16 and expressed as % of added supernatants (% SN-12,5%, 6,25%, 3,12%, 1,56%, 0,78% -) were showed. Unfermented milk samples acidified at pH 5.3 with lactic acid were used as controls (CNTR). The anti-inflammatory index (IL-10/IL-12p40, panels c, f) is indicated. Statistical significance of single bacteria strain was evaluated by using two-way ANOVA followed by multiple comparisons (Tukey) against the control. Values represent the means ± SD, 95% CI, ( � P<0.05, �� P<0.01).
https://doi.org/10.1371/journal.pone.0245903.g003 however treated and coming from cheeses fermented by both SIM12 and SIS16, decreased the production of the pro-inflammatory IL-12p40 and increased that of the anti-inflammatory IL-10 cytokines by total PBMC (Fig 3, panels a, b, d, e). The effect, which took place in both TQ-CS and P-CS samples, was dose-dependent, i.e., increased with increasing the % of added samples, as confirmed by the statistical significance of the anti-inflammatory index (Fig 3, panels c, f). The highest production of IL-10 was achieved with a 12,5% concentration of P-CS from cheeses made with SIS16, achieving a percentage of response to LPS of around 300 (Fig  3, panel e). It is interesting to underline the low production of cytokines in CNTR, i.e., milk coagulated by acidification at pH 5.3 with lactic acid, compared to milk fermented using L. helveticus strains.

Discussion
In recent years, there has been a marked increase of studies showing that cheese may have a great potential to act as a carrier for the administration of functional bacteria [23][24][25][26]. However, several factors should be considered to demonstrate the actual functionality of strains when included in cheese. Among them are stress resistance to which micro-organisms are subjected during the various phases of cheese processing, their viability in the product, and the maintenance of the health-promoting characteristics during storage and until food is consumed consumption [27]. Not least, few data are available to distinguish the contribution to health effects related to microbial metabolism from those intrinsically related to the food itself [2,28]. This study had three purposes: (i) to search for functional features in L. helveticus strains and test their ability to develop and maintain health benefits (in particular, the immunomodulatory capacity) in cheese; to this regard, we focused on some of the traits (e.g. folate accumulation, proteolytic, peptidase, and AO activities) most often involved in immunomodulatory effects; (ii) to distinguish between the direct immunomodulant effect of the cheese matrix and that related to the microbial growth; (iii) linked to the latter, to assess whether the immunomodulation effect was attributable to the cells (live or dead) trapped into cheese or to the products of their metabolism. Our attention was focused on L. helveticus, a species that, thanks to the ability to survive various environmental stresses such as high temperatures and osmotic pressure or low pH and oxygen concentration, is of increasing industrial importance as it is easily adaptable to fermentation conditions, thus facilitating its integration into new health promoting foods [6]. We selected cheese isolates to increase the probability of using strains already adapted to the dairy matrix, therefore more suitable to be successfully applied in cheesemaking trials.
Only strain SIM12 showed the ability to produce a moderate level (19.25 ng/ml) of folate. The ability to accumulate similar amounts (27 ng/ml) of folate by strains grown in chemically defined folate-free medium has been reported in L. plantarum [29]. The folate pathway is responsible for multiple functions, including energy (ATP) production, methylation reactions for DNA and protein synthesis, and the production of immunomodulatory molecules, inosine and adenosine, which play an important role in immune signaling and cytotoxicity [30]. Also, many peptides deriving from milk proteins and accumulated by highly proteolytic L. helveticus strains were shown to stimulate the host's immune system or display a marked antioxidant activity [5,31]. The four studied strains showed a variable, although generally high, proteolytic and aminopeptidase activity in milk. In particular, our attention was focused on aminopeptidase activities typical of LAB isolated from cheese (leu-, arg-, lys-, and X prolyl dipeptidyl-aminopeptidase), therefore more likely to be involved in the release of (bioactive) peptides and free amino acids from milk proteins [32]. The AO activity, which was evaluated in milk culture supernatants using the ABTS, FTG, and FRAP assays, was also very variable. Notably, the literature currently reports a considerable variability of data due to the wide variety of poorly comparable analytical methods, which implies a difficulty in setting a threshold value for classifying strains with high or low AO activity [33]. However, the AO values for L. helveticus were low when compared with literature data on milks fermented by probiotic strains of L. plantarum [18,34,35]. Interestingly, strains SIS16 and Lh43, which expressed the highest FRAP and ABTS values, respectively, resulted also the most proteolytic ones. Finally, very high β-galactosidase activity was expressed from the tested L. helveticus strains. The evaluation of βgalactosidase (β-gal) activity had a dual purpose. Firstly, β-gal activity is an important functional feature. Lactose intolerance is found in people lacking β-gal which determines, in the upper regions of the small intestine, an inability to hydrolyze and assimilate this sugar, which is instead used from gas-producing microorganisms in the gut [36]. Secondly, a high β-gal activity allows LAB to produce lactic acid rapidly and reliably [37], thus facilitating their inclusion in starter cultures.
The immunomodulatory activity of the four strains and their cellular soluble postbiotics on moDCs, which are pivotal in the initiation of the adaptive immune response [8], was studied. To this regard, moDCs need to be activated by lipopolysaccharides (LPS) or pathogenic bacteria and, once activated, they produce cytokines [4]. Traditionally, IL10/IL12 ratio produced by moDCs is one of the main indexes for the classification of strains as potentially anti-inflammatory [4]. Therefore, we quantified the production of the anti-inflammatory IL-10 and the proinflammatory IL-12p70 by LPS-or Salmonella-stimulated moDCs, incubated in the presence of the four L. helveticus strains (or their postbiotics) [35]. In culture media, bacterial cells exerted a prevalent anti-inflammatory effect when inflammation was caused by LPS whereas, in the absence of stimulus, strains SIM12 and SIS16 strains induced an inflammatory response local that resembled that induced by Salmonella. Thus, SIM12 and SIS16 could be relevant under steady-state conditions to boost the immune response (immuno-enhancing effect) and increase resistance to pathogen infection or for vaccination purposes [8]. An immunomodulatory effect (i.e., an increase of the Il-10/IL12p70 ratio) on whatever induced inflammation was more marked using postbiotics of the four strains.
The immunomodulating activity was then tested in Caciotta-like cheese, which was chosen as a model cheese. Caciotta belongs to a large category of Italian traditional soft cheeses, is relatively easy to produce and has a very limited maturation period (www.assolatte.it). Cheeses were produced using individually the strains SIM12 and SIS16 which, differently from Lh43 and 1734, had the dual function of acidifying and immunomodulating cultures. This choice avoided to include an additional acidifying culture, which could have interfered with the immunomodulatory activity of L. helveticus. Data showed that immunomodulators could be present also in cheese, thus confirming tests in culture media. The substantial similarity of the anti-inflammatory index between TQ-CS (containing a high number of alive cells) and P-CS (where 99.999% of the viable cells were inactivated by pasteurization) suggested that the immunomodulation was essentially due to either a paraprobiotic effect, i.e. beyond cell viability, or soluble metabolites (postbiotics) accumulated by the L. helveticus strains through either cheese fermentation (e.g. folic acid, antioxidants, lactic acid) or substrate modification (e.g. proteolysis and peptide-lysis leading to the possible accumulation of bioactive, small peptides) or both [4,5,38]. Importantly, no cytokine production was induced by unfermented cheese, confirming that the immunomodulation was exclusively related to the presence and/or metabolic activity of strains SIM12 and SIS16. Several studies reported the ability of strains belonging to L. helveticus species to exert immunostimulatory effects, both when used alone or in combination with other bacterial species, but few are the reports on such effects in dairy products and, even less, in cheese [6]. In very early attempts, Laffineur et al. [39] demonstrated that the cell-free supernatant of L. helveticus-fermented ß-casein-enriched medium modulated in vitro the proliferation of lymphocytes. Ng and Griffiths [40] observed that cultured macrophages conditioned with L. helveticus-fermented milk, or its cell-free supernatant, were stimulated to secrete IL-6. Stuknyte et al. [41] observed a reduction of the immunomodulatory properties of casein hydrolysates when the bacterial cells of L. helveticus MIMLh5, used for the digestion of casein, were removed by ultrafiltration. More recently, a strain of L. helveticus isolated from fresh artisan Pasta Filata cheese showed in vivo anti-inflammatory effects as it stimulated the production of interleukin 10 (IL-10) and lowered that of the tumor necrosis factor alpha (TNF-α) by peripheral blood mononuclear cells (PBMC) [28]. This latter study and our results seem to stress that traditional dairy products should be more thoroughly and systematically considered as a natural source of health-promoting microorganisms [28]. In conclusion, this study adds further evidence on the potential benefits of L. helveticus for human health. Once validated in in vivo models, strains SIM12 and SIS16 could be good candidates for inclusion as functional cultures in cheese production.  Table. Enumeration of viable cells in cheese suspensions.