Microbial dynamics during processing and ripening of traditional cheeses such as registered designation of origin Salers cheese, an artisanal cheese produced in France, play an important role in the elaboration of sensory qualities. The aim of the present study was to obtain a picture of the dynamics of the microbial ecosystem of RDO Salers cheese by using culture-independent methods. This included DNA extraction, PCR, and single-strand conformation polymorphism (SSCP) analysis. Bacterial and high-GC% gram-positive bacterial primers were used to amplify V2 or V3 regions of the 16S rRNA gene. SSCP patterns revealed changes during the manufacturing of the cheese. Patterns of the ecosystems of cheeses that were provided by three farmers were also quite different. Cloning and sequencing of the 16S rRNA gene revealed sequences related to lactic acid bacteria (Lactococcus lactis, Streptococcus thermophilus, Enterococcus faecium, Leuconostoc mesenteroides, Leuconostoc pseudomesenteroides, Lactobacillus plantarum, and Lactobacillus pentosus), which were predominant during manufacturing and ripening. Bacteria belonging to the high-GC% gram-positive group (essentially corynebacteria) were found by using specific primers. The present molecular approach can effectively describe the ecosystem of artisanal dairy products.The typical sensorial qualities of traditional cheese ultimately depend on several factors, including traditional cheese-making practices, feeding of dairy cows, and the dynamics of microbial communities. The qualitative and quantitative composition of the microbial flora, its evolution, and its activity during ripening play an important role in the development of hygienic and sensorial qualities. In order to better understand the functions of the microbial community, a full description of the microbial ecosystem is required. Classically, this has been addressed by enumerating members of certain microbial groups by using various culture media, followed by identification of a number of dominant isolates by phenotypic tests (16,18,33) or molecular techniques such as ribotyping (21), random amplified polymorphic DNA analyses, and sequencing (4, 27). However, cultivation-dependent approaches do not necessarily provide reliable information on the composition of entire microbial communities. It is therefore difficult to assess the significance of cultured microorganisms in microbial ecosystems because of the disparity between culturable and in situ diversity. Indeed, microbial communities may contain viable but nonculturable bacteria and also bacteria that would be culturable if appropriate medium had been developed. Thus, in order to study interactions between microorganisms, it is important to study the ecosystem without dissociating it. It has been shown that a dual approach, e.g., inventory by using a cultivation-dependent method and di-

The microbial communities in milks from one herd were evaluated during 1-year of lactation, using molecular methods to evaluate their stability and the effect of breeding conditions on their composition. The diversity of microbial communities was measured using two approaches: molecular identification by 16S and 18S rDNA sequencing of isolates from counting media (two milks), and direct identification using 16S rDNA from clone libraries (six milks). The stability of these communities was evaluated by counting on selective media and by Single Strand Conformation Polymorphism (SSCP) analysis of variable region V3 of the 16S rRNA gene and variable region V4 of the 18S rRNA gene. One hundred and eighteen milk samples taken throughout the year were analyzed. Wide diversity among bacteria and yeasts in the milk was revealed. In addition to species commonly encountered in milk, such as Lactococcus lactis, Lactococcus garvieae, Enterococcus faecalis, Lactobacillus casei, Leuconostoc mesenteroides, Staphylococcus epidermidis, Staphylococcus simulans, Staphylococcus caprae, Staphylococcus equorum, Micrococcus sp., Kocuria sp., Pantoea agglomerans and Pseudomonas putida, sequences were affiliated to other species only described in cheeses, such as Corynebacterium variabile, Arthrobacter sp., Brachybacterium paraconglomeratum, Clostridium sp. and Rothia sp. Several halophilic species atypical in milk were found, belonging to Jeotgalicoccus psychrophilus, Salinicoccus sp., Dietza maris, Exiguobacterium, Ornithinicoccus sp. and Hahella chejuensis. The yeast community was composed of Debaryomyces hansenii, Kluyveromyces lactis, Trichosporon beigelii, Rhodotorula glutinis, Rhodotorula minuta, Candida pararugosa, Candida intermedia, Candida inconspicua, Cryptococcus curvatus and Cryptococcus magnus. The analyses of microbial counts and microbial SSCP profiles both distinguished four groups of milks corresponding to four periods defined by season and feeding regime. The microbial community was stable within each period. Milks from winter were characterized by Lactococcus and Pseudomonas, those from summer by P. agglomerans and Klebsiella and those from autumn by Chryseobacterium indologenes, Acinetobacter baumanii, Staphylococcus, Corynebacteria and yeasts. However, the composition of the community can vary according to factors other than feeding. This study opens new investigation fields in the field of raw milk microbial ecology.

Bacteria and yeasts are important sensory factors of raw-milk cheeses as they contribute to the sensory richness and diversity of these products. The diversity and succession of yeast populations in three traditional Registered Designation of Origin (R.D.O.) Salers cheeses have been determined by using phenotypic diagnoses and Single-Strand Conformation Polymorphism (SSCP) analysis. Isolates were identified by phenotypic tests and the sequencing of the D1-D2 domains of the 26S rRNA gene. Ninety-two percent of the isolates were identified as the same species in both tests. Yeast-specific primers were designed to amplify the V4 region of the 18S rRNA gene for SSCP analysis. The yeast species most frequently encountered in the three cheeses were Kluyveromyces lactis, Kluyveromyces marxianus, Saccharomyces cerevisiae, Candida zeylanoides and Debaryomyces hansenii. Detection of less common species, including Candida parapsilosis, Candida silvae, Candida intermedia, Candida rugosa, Saccharomyces unisporus, and Pichia guilliermondii was more efficient with the conventional method. SSCP analysis was accurate and could be used to rapidly assess the proportions and dynamics of the various species during cheese ripening. Each cheese was clearly distinguished by its own microbial community dynamics.

Raw milk cheeses show a wide diversity of sensorial characteristics, largely determined by the microflora of raw milk. Microbial dynamics in Registered Designation of Origin (R.D.O.) Salers cheese was assessed by DNA and RNA SSCP analysis on nine cheeses. These cheeses showed considerable diversity both in microbial dynamics and sensorial characteristics. Relationships between the sensorial characteristics and the microbial dynamics were studied. A global consideration of bacterial dynamics demonstrated that other bacteria than lactic acid bacteria can play a role in the elaboration of sensorial characteristics. Indeed, high CG% Gram-positive bacteria can be involved. DNA data, as well as RNA data, appeared relevant to attempts to explain sensorial variance. Correlations between sensorial and microbial data were rather complex. Several microbial variables for DNA and RNA analyses, noted at different times of analysis, were correlated to each sensory variable. A global view of cheese microbial community proved to be insufficient in explaining the diversity of the sensorial qualities of R.D.O. Salers cheese.

Extracting DNA from deep subsurface sediments is challenging given the complexity of sediments types, low biomasses, resting structures (spores, cysts) frequently encountered in deep sediments, and the potential presence of enzymatic inhibitors. Promising results for cell lysis efficiency were recently obtained by use of a cryogenic mill (Lipp et al., 2008). These findings encouraged us to devise a DNA extraction protocol using this tool. Thirteen procedures involving a combination of grinding in liquid nitrogen (for various durations and beating rates) with different chemical solutions (phenol, chloroform, SDS, sarkosyl, proteinase, GTC), or with use of DNA recovery kits (MagExtractor®) were compared. Effective DNA extraction was evaluated in terms of cell lysis efficiency, DNA extraction efficiency, DNA yield and determination of prokaryotic diversity. Results were compared to those obtained by standard protocols: the FastDNA®SPIN kit for soil and the Zhou protocol. For most sediment types grinding in a cryogenic mill at a low beating rate in combination with direct phenol-chloroform extraction resulted in much higher DNA yields than those obtained using classical procedures. In general (except for clay-rich sediments), this procedure provided high-quality crude extracts for direct downstream nested-PCR, from cell numbers as low as 1.1×10(6) cells/cm(3). This procedure is simple, rapid, low-cost, and could be used with minor modifications for large-scale DNA extractions for a variety of experimental goals.