Comparison of amplified ribosomal DNA restriction analysis, peptidoglycan hydrolase and biochemical profiles for rapid dairy propionibacteria species identification

Species of dairy propionibacteria are used as cheese-ripening cultures as well as probiotics. However, no rapid identification methods are currently available. With this in mind, the present study compared three methods, (i) carbohydrate fermentation, (ii) ARDRA (amplified ribosomal DNA restriction analysis) and (iii) peptidoglycan hydrolase (PGH) activity profiles to improve the identification of Propionibacterium thoenii, Propionibacterium jensenii, Propionibacterium acidipropionici and Propionibacterium microaerophilum. The species Propionibacterium freudenreichii and Propionibacterium cyclohexanicum have previously been shown to be easily distinguishable from the other species. Principal component analysis of the carbohydrate fermentation profiles of 113 P. thoenii, P. jensenii, P. acidipropionici and P. microaerophilum strains correctly classified 85% of the strains based on the fermentation of seven carbohydrates. Regarding PGH profiles, optimized conditions of PGH-renaturing SDS–PAGE were applied to 34 of the strains. The PGH profiles of P. acidipropionici and P. microaerophilum were indistinguishable from one another, but were easily distinguished from P. jensenii and P. thoenii. However, four strains exhibited atypical profiles. Hence, in general, the PGH profiles were shown to be conserved within a species, with some exceptions. Four endonucleases were tested for ARDRA and the four species differentiated by combining the profiles obtained with MspI and HaeIII. P. freudenreichii and P. cyclohexanicum profiles were also performed but showed wide differences. Consequently, ARDRA was shown to be the most appropriate method for rapidly distinguishing strains of propionibacteria. Carbohydrate fermentation and peptidoglycan hydrolase activity profiles are useful as complementary identification tools, since about 15% of the 34 strains tested showed atypical profiles.