Interaction between lactobacillus plantarum and food related microorganisms by proteomics and bioinformatics

Lactobacillus plantarum is a versatile and widespread microorganism found in materials and environments ranging from vegetable, dairy products and meat fermentations to the human gastrointestinal (GI) tract. Some strains are marketed as probiotics that are claimed to provide a health benefit for the consumer. Furthermore, certain strains of Lb. plantarum are known for their ability to produce several natural antimicrobial substances. The production of these metabolites could represent stress conditions that strongly affects the development of undesirable microbial species. There are many scientific reports that highlight antimicrobial effects of Lb.plantarum strains on undesirable bacteria. Several strains of Lb. plantarum showed a broad spectrum of antibacterial activity (including Bacillus cereus, S. aureus, Listeriamonocytogenes, Salmonella enterica, E. coli, and Enterobacteraerogenes) and carries several plantaricin genes of the pln locus. Moreover various other bacteriocins produced by Lb. plantarum species isolated from fermented food are well known. Nevertheless, the effectiveness of bacteriocin-producing strains in foods can be limited by several factors including narrow activity spectrum, limited, diffusion in solid matrices, inactivation through proteolytic enzymes or binding to food ingredients such as lipids, low production level and the emergence of bacteriocin-resistant bacteria. Although, the use of class IIa bacteriocins or bacteriocins-producing strains represent a promising alternative for the control of spoilage or pathogenic microorganisms in foods, their efficacy could be compromised by onset of bacteriocins resistant strains and cross-resistance between bacteriocins. However, the use of bacteriocins in combination with other hurdles (e.g. salt, acid, other natural substances etc.) may result extremely effective for inhibit L. monocytogenes and reduce its resistance. A number of experiments have been showing the anti-listeria effect due to production of acid organics, including lactic acid by Lb. plantarum and 3-Phenyllactic acid (2-hydroxy-3-phenylpropanoic acid, PLA). The genomic architecture and the induced metabolic consequencesare central to the success of Lb. plantarum in industrial applications. Moreover, the most of Lb. plantarum selected for their antimicrobial activity has been isolated from fresh or fermented food. Therefore, the characterization and selection of food-borne Lb. plantarum strains remains a topic of great interest for applied research. On the basis of this last finding, the first part of this PhD study (Chapter II) was devoted to isolate and identify food borne Lb. plantarum as well as to evaluate their antimicrobial range. Thirty-two samples from three type of traditional fermented food were subjected to microbiological analyses in order to isolate and select Lb. plantarum strains to be used as antagonistic strains (producers) against undesirable food-stuff microorganisms. To identify LAB isolates, several approaches were used, consisting of the DGGE analysis and 16S rRNA gene sequencing. While the antimicrobial activity, exerted by cells or cell-free supernatants of Lb. plantarum strains, were evaluated by spot on the lawn test and by agar well diffusion assay test. The results evidenced that Lb. plantarum represents the prevailing lactobacilli species in sourdough andred wine, while this species were detected only in few sample of fermented sausages. In detail 60 Lb. plantarum strains were isolated from red wines, 36 strains from sourdoughs, and 10 from fermented sausages. Out 106 Lb. plantarum strains, seven strains evidenced the ability to inhibit Gram negative and Gram positive bacteria as well as moulds strains. These inhibitory effect was not attributable to pH decrease, sincein the presence of neutralized CFS of producer strains were also detected with a strong antimicrobial activity. Noticeablewas the data that evidenced a strong antimicrobial activity produced by Lb. plantarum RTB strain against L. innocua ATCC 33090. Since L. innocua has been deemed a suitable biological indicator for L. monocytogenes and it revealed a similar sensitivity to different stress condition. In the last years, great attention was focused on the inhibitory action against Listeria monocytogenes exerted by Lb. plantarum strains. The interest towards this topic is due to diseases caused by L. monocytogenes and which are known as “listeriosis” (causative agent of abortions, gastrointestinal diseases or septicaemia, thatoften lead to the death of infected individuals). This pathogen bacterium, growing at low pH, at refrigeration temperature, and at very high salt concentrations, is isolable from several food products, albeit in low numbers. Several studies reported the characterization of antimicrobial substances produced by certain Lb. plantarum strains. However it is well known that the knowledge of the undesirable strains response to these antimicrobial substances (stress conditions) represents a crucial step for the definition of an effective bio-control tool. Several mechanisms, can be developed by Listeria spp. in order to resist the injuries caused by stress conditions (temperature, acidity, NaCl). In detail, the stress seems to induce variations in the synthesis of certain cell components, especially proteins. Nevertheless the literature is very poor in studies focusing on the mechanisms of response, in terms of susceptibility or resistance, expressed by L. monocytogenes against antimicrobial substances produced by Lb. plantarum. Therefore, the second part of the present PhD activity focused the attention on the stress response of Listeria to the presence of Lb. plantarum (Chapter III e Chapter IV). In detail, a commercial L. innocua strain was used as a pathogen surrogate throughout this study. For this purpose a multiple technique approach was adopted in the study, consisting of microbiological (dynamic model to predict the growth, cell counts) and proteomics (SDS-PAGE and 2D-E) approaches Results showed that both cell and cell free supernatant of Lb. plantarum strain RTB represent a strong stress factor for L. innocua ATCC 33090, expressed through its growth inhibition. In detail, the inhibition was not attributable to organic acids produced by Lb. plantarum, since L. innocua ATCC 33090 expressed a series of new protein including Universal Stress Protein (USP) in the presence of lactic acid alone, that allowed to react to the acidic environment. On the other hand, the presence of Lb. plantarum RTB produced on L. innocua not only the expression of USP, but also the degradation or non-expression of other proteins. This phenomenon could be due to several antimicrobial substances and mechanisms carried out by the producer strain, and they could be responsible for the inhibition exerted by Lb. plantarum RTB against L. innocua ATCC 33090. Particular attention was focused on the neo-expressed USP, a group of proteins induced by different stress conditions and which are found in numerous prokaryotic as well as eukaryotic organisms. The majority of UPS genes are monocistronically expressed, and different transcription factors, promote transcription of USPs. The significance of USPs in the resistance or susceptibility model of L. monocytogenes is presently unknown. Moreover few information are available in literature about the biochemical function and 3D-structures of USPs in bacteria and there are no 3D-structures for USP ofListeria. available. Bioinformatics approach can help to get more information about the structure of USPs and the function of these proteins. Therefore in the 3rd phase of this PhD study (Chapter V), the three-dimensional (3D) structure of a USP (EHN60729.1) belonging to L. innocua was predicted on the basis of the available template (PDB code:3S3T ; structure deposited by Osipiuk et al., 2011) homologues from Protein Data Bank. The Comparative Homology Modeling procedure uses the structure of proteins experimentally determined (template) to predict the 3D structure of a protein that has a similar amino acid sequence (target). The Comparative Homology Modeling approach can be used when the template and target possess at least 30% identity. In the present study the hypothetical USP of L. innocua shares 31% amino acid with the template 3S3T which corresponds to about 85% of the C-α with 3.5 Å from the correct position. The accuracy of the model is confirmed by the values of the torsion angles phi and psi showed in the Ramachandran plot as well as the QMEAN Z-score. The RMSD (0.3 Å) of the final refined model confirms the evolutionary relationship between the model and the template. Of interest are the results regarding the analyses of the interfaces carried out with both PISA WebServer and with the multiple structural alignment (MUSTANG). The surface of the interfaces (ΔG <0, see Figure 5.6) is amongst the average values of expected forhomologous proteins, but even more interesting is the presence of highly conserved residues in the region involved in the formation of the dimer and of residues that represent the ATP-binding motif, which is fundamental in the formation of the tetramer. This work suggests that L. innocua possesses a UspFG-Type and that this protein can assemble in a tetrameric structure. These results, although to be confirmed experimentally, provide important information about a poorly studied protein and may stimulate experimental investigations. Overall, the results obtained in this study improved the knowledge both in stress response of L. innocuaand in develop of bio-control (anti-listeria adjunct starter or protective cultures) useful in food bio-preservation.