The microbiota of high-moisture mozzarella cheese produced with different acidification methods.

The microbiota of high-moisture Mozzarella cheese made from cow's milk and produced with different acidification methods was evaluated at the end of refrigerated storage by pyrosequencing of the 16S rRNA gene. The cheeses were clearly separated on the basis of the acidification methods. Cheeses produced with the addition of starters were dominated by Streptococcus thermophilus, but a variety of lactic acid bacteria and spoilage microorganisms appeared at low levels (0.01-1%). Cheeses produced by direct addition of citric acid were dominated by a diverse microbiota, including both lactic acid bacteria and psychrotrophic γ-proteobacteria. For five brands the acidification system was not declared on the label: the microbiota was dominated by thermophilic lactic acid bacteria (S. thermophilus, Lactobacillus delbrueckii, Lactobacillus helveticus) but a variety of other subdominant lactic acid bacteria, psychrotrophs and Enterobacteriaceae were present, with a diversity comparable or higher to cheeses produced by direct acid addition. This led to the conclusion that undefined starters were used for acidification. Both ordination methods and network analysis were used for the representation of beta-diversity: matrix cluster analysis, principal coordinate analysis and OTU networks uncovered different aspects of the microbial community structure. For three cheese brands both biological replicates (cheeses from different lots) and technical replicates (replicate cheeses from the same lot) were analyzed. Repeatability was acceptable for OTUs appearing at frequencies >1%, but was low otherwise. A linear mixed model showed that the starter system was responsible for most differences related to dairies, while difference due to psychrotrophic contaminants was more related to lot-to-lot variability.

Microbial community dynamics in thermophilic undefined milk starter cultures.

Model undefined thermophilic starter cultures were produced from raw milk of nine pasta-filata cheesemaking plants using a selective procedure based on pasteurization and incubation at high temperature with the objective of studying the microbial community dynamics and the variability in performances under repeated (7-13) reproduction cycles with backslopping. The traditional culture-dependent approach, based on random isolation and molecular characterization of isolates was coupled to the determination of pH and the evaluation of the ability to produce acid and fermentation metabolites. Moreover, a culture-independent approach based on amplicon-targeted next-generation sequencing was employed. The microbial diversity was evaluated by 16S rRNA gene sequencing (V1-V3 regions), while the microdiversity of Streptococcus thermophilus populations was explored by using novel approach based on sequencing of partial amplicons of the phosphoserine phosphatase gene (serB). In addition, the occurrence of bacteriophages was evaluated by qPCR and by multiplex PCR. Although it was relatively easy to select for a community dominated by thermophilic lactic acid bacteria (LAB) within a single reproduction cycle, final pH, LAB populations and acid production activity fluctuated over reproduction cycles. Both culture-dependent and -independent methods showed that the cultures were dominated by either S. thermophilus or Lactobacillus delbrueckii subsp. lactis or by both species. Nevertheless, subdominant mesophilic species, including lactococci and spoilage organisms, persisted at low levels. A limited number of serB sequence types (ST) were present in S. thermophilus populations. L. delbrueckii and Lactococcus lactis bacteriophages were below the detection limit of the method used and high titres of cos type S. thermophilus bacteriophages were detected in only two cases. In one case a high titre of bacteriophages was concurrent with a S. thermophilus biotype shift in the culture. This study largely confirms previous data on the composition of undefined thermophilic starters used for the production of traditional cheeses in Italy but it is the first one to systematically address the dynamics of the cultures under a repeated reproduction regime with backslopping.

Effect of adjuncts on microbiological and chemical properties of Scamorza cheese.

Scamorza is a semi-hard, pasta filata cheese resembling low-moisture Mozzarella cheese, with a short ripening time (<30d). Scamorza has a bland flavor and, to provide diversification from similar cheeses, it was manufactured using 2 types of milk in the current study: 100% Italian Friesian milk (F) or 90% F and 10% Jersey cow milk (mixed, M), and 2 types of starter: Streptococcus thermophilus or S. thermophilus with peptidolytic Lactococcus lactis, Lactobacillus helveticus, and Lactobacillus paracasei strains as adjuncts). The cheeses were ripened for 30d. The adjunct did not significantly affect acid production or growth of the primary starter; 2 of the species used in the adjunct (Lb. paracasei and Lb. helveticus) rapidly colonized the cheese and persisted until the end of ripening, whereas the counts of nonstarter lactic acid bacteria in the control cheese were low until the end of ripening. The use of adjuncts affected pH, microbial composition (as assessed by both culture-dependent and culture-independent methods), total free amino acid content, and volatile profile (measured using an electronic nose), whereas milk type had only a minor effect. Although differences in primary proteolysis were found, they were probably indirect and related to the effects on pH and moisture. We conclude that, even with a short ripening time (30d), use of a peptidolytic adjunct may significantly affect important features of Scamorza and may be used to create a product that is measurably different from competing products.

Genotypic diversity of stress response in Lactobacillus plantarum, Lactobacillus paraplantarum and Lactobacillus pentosus.

Lactobacillus plantarum, Lactobacillus pentosus and Lactobacillus paraplantarum are three closely related species which are widespread in food and non-food environments, and are important as starter bacteria or probiotics. In order to evaluate the phenotypic diversity of stress tolerance in the L. plantarum group and the ability to mount an adaptive heat shock response, the survival of exponential and stationary phase and of heat adapted exponential phase cells of six L. plantarum subsp. plantarum, one L. plantarum subsp. argentoratensis, one L. pentosus and two L. paraplantarum strains selected in a previous work upon exposure to oxidative, heat, detergent, starvation and acid stresses was compared to that of the L. plantarum WCFS1 strain. Furthermore, to evaluate the genotypic diversity in stress response genes, ten genes (encoding for chaperones DnaK, GroES and GroEL, regulators CtsR, HrcA and CcpA, ATPases/proteases ClpL, ClpP, ClpX and protease FtsH) were amplified using primers derived from the WCFS1 genome sequence and submitted to restriction with one or two endonucleases. The results were compared by univariate and multivariate statistical methods. In addition, the amplicons for hrcA and ctsR were sequenced and compared by multiple sequence alignment and polymorphism analysis. Although there was evidence of a generalized stress response in the stationary phase, with increase of oxidative, heat, and, to a lesser extent, starvation stress tolerance, and for adaptive heat stress response, with increased tolerance to heat, acid and detergent, different growth phases and adaptation patterns were found. Principal component analysis showed that while heat, acid and detergent stresses respond similarly to growth phase and adaptation, tolerance to oxidative and starvation stresses implies completely unrelated mechanisms. A dendrogram obtained using the data from multilocus restriction typing (MLRT) of stress response genes clearly separated two groups of L. plantarum strains from the other species but there was no correlation between genotypic grouping and grouping obtained on the basis of the stress response pattern, nor with the phylograms obtained from hrcA and ctsR sequences. Differences in sequence in L. plantarum strains were mostly due to single nucleotide polymorphisms with a high frequency of synonymous nucleotide changes and, while hrcA was characterized by an excess of low frequency polymorphism, very low diversity was found in ctsR sequences. Sequence alignment of hrcA allowed a correct discrimination of the strains at the species level, thus confirming the relevance of stress response genes for taxonomy.

Inactivation of ccpA and aeration affect growth, metabolite production and stress tolerance in Lactobacillus plantarum WCFS1.

The growth of Lactobacillus plantarum WCFS1 and of its ΔccpA ery mutant, WCFS1-2, was compared in batch fermentations in a complex medium at controlled pH (6.5) and temperature (30°C) with or without aeration, in order to evaluate the effect of ccpA inactivation and aeration on growth, metabolism and stress resistance. Inactivation of ccpA and, to a lesser extent, aeration, significantly affected growth, expression of proteins related to pyruvate metabolism and stress, and tolerance to heat, oxidative and cold/starvation stresses. The specific growth rate of the mutant was ca. 60% of that of the wild type strain. Inactivation of ccpA and aerobic growth significantly affected yield and production of lactic and acetic acid. Stationary phase cells were more stress tolerant than exponential phase cells with little or no effect of inactivation of ccpA or aeration. On the other hand, for exponential phase cells inactivation of ccpA impaired both heat stress and cold/starvation stress, but increased oxidative stress tolerance. For both strains, aerobically grown cells were more tolerant of stresses. Evidence for entry in a viable but non-culturable status upon prolonged exposure to cold and starvation was found. Preliminary results of a differential proteomic study further confirmed the role of ccpA in the regulation of carbohydrate catabolism and class I stress response genes and allow to gain further insight on the role of this pleiotropic regulator in metabolism and stress. This is the first study in which the impact of aerobic growth on stress tolerance of L. plantarum is evaluated. Although aerobic cultivation in batch fermentations does not improve growth it does improve stress tolerance, and may have significant technological relevance for the preservation of starter and probiotic cultures.