Wooden tools : complex reservoirs of microbial diversity for food fermentations

Food fermentation exists for millenaries (Salque et al., 2012). An incredible number of raw material either vegetal or animal can be fermented to increase shelf life and to create diversified flavors. Up to date more than 5000 fermented foods (including fermented beverages) are listed in the world and the daily consumption is estimated between 50 to 400g per day and per capita (Tamang J.P. and Kailasapathy K., 2010). By an empirical approach, only based on their “five senses” and their creativity, our ancestors have created all the kind of fermented foods we know today, which still have a crucial and heartening place in our diet: wine, bread, cheese, fermented milks and butter, vinegar, olives, beer and so on….Most of them have deep cultural and territory roots, and local declensions. However, most of them are for few decades the object of strong industrialization and standardization. Indeed, since Pasteur (1865) and his demonstration of the involvement of “microbes” in the fermentation process, starters were developed to better control it, to avoid defects and to make it more repeatable in order to be able to increase the scale of production and shelf life. Indeed, between 1880 and 1930, in Europe,a rapid urbanization and industrialization occurred changing food demand and ways of production. However, excellent fermented foods and beverages existed before the use of industrial starters. Unfortunately we tend to forget how elaborated and rich, almost artistic, are the artisanal practices of our ancestors, which were able to manage the microbial diversity, without even suspecting its existence! Most of these artisanal practices involved wooden tools: barrels, mess (kneading machine), vats, spoons, molds, cream separator, hoops, shelves,...Surprisingly the microbial ecology of wooden tools used in food fermentation has been rarely explored, as well as its involvement in the final quality. These last years several publications reported the presence of a rich biofilm on the surface of wooden vats used in dairy fermentations, in particular in French and Sicilian PDO cheeses (Licitra et al., 2007; Lortal et al; 2009; Didienne et al., 2012; Settanni et al., 2012; Scatassa et al., 2015) as well as the microbial ecology of shelves used in cheese ripening (Mariani et al., 2007). Thanks to recent molecular tools, a large microbial biodiversity was showed on the surface of the vats. Electron or confocal laser microscopy revealed extraordinary images of thick microbial biofilms covering the wood surface (figure 1), biofilm composed of many different species including lactic acid bacteria, (dominating), enterococci, high GC% bacteria like coryneforms, some gram negative; yeasts and moulds (Lortal et al., 2014). When raw milk is placed in the vat, a massive spontaneous inoculation by the vat biofilm occurs in few minutes as it was demonstrated by putting microfiltrated milk into tina wooden vats (Lortal et al., 2009); the inoculated lactic acid bacteria composing the dominant part of the biofilm, contribute then significantly to the acidification step and the whole biofilm ecosystem to the subsequent ripening. The contact with the vat represents quantitatively an enrichment of the milk by a ‘trained’ microflora, supporting the affirmation of many cheese makers that wooden vats directly impacts final sensorial qualities and typicity. The microbial ecosystem of a given wooden vat is related to the cheese technology and is different for Salers, Ragusano,caciocavallo Palermito or Vastedda della valle des Belice. This complex ecosystem is stable in time in terms of dominant species, and for each species, several strains co-exist. Interestingly, molecular typing revealed that strains are farm/vat specific. All the authors above underlined the absence or extremely low levels of undesirable microorganisms like coliforms, and the complete absence of pathogens like Salmonella or Listeria. Qualitatively the composition of wooden vat biofilm was deeply explored. From four tinas (wooden vat used in the Sicilian PDO Ragusano)(Licitra et al., 2007; Lortal et al., 2009), 200 clones of the dominant species S. thermophilus, which is also the most metabolically active as shown by RT-PCR-TTGE, were isolated and characterize by pulsed-field gel electrophoresis (PFGE) and Multilocus sequence typing (MLST) in order to assess the number of strains as several are cohabiting inside the same vat. By comparing these isolated strains to 160 other S. thermophilus coming from all over the world, it was found that Sicilian Tina strains forms a completely separate cluster (Valence et al., unpublished data) and were thus unique with 17 completely new sequence types. Whole genome sequencing of some of these strains will help in understanding their specificity. Some technological properties were explored like the ability to produce antimicrobial compounds and phage resistance. Interestingly they were shown to be at least 4 times more resistant to phages when compared to commercial starters. Since that first attempt, the microbial ecology of several others wooden vats used in pasta filata Sicilian cheeses, from cow and ewe raw milks, were deeply explored (Settanni et al., 2012; Scatassa et al., 2015a; Scatassa et al., 2015b). By 16S DNA sequencing, a total of 16 different lactic acid bacteria (LAB) species were identified at dominating levels in these wooden vats, including S. thermophilus, various lactobacilli and Enterococci. 2 to 5 LAB or enterococci species coexist in a predominant within one wooden vat species (depending on the cheese technology concerned), E.faecium being systematically present. Clustering of strains was performed by RAPD, and strains belonging to the same species clustered closely. Some technological properties were assess for isolated strains (acidification, diacetyl formation, autolysis, proteolytic activity, and production of antimicrobial compounds). Interestingly, many strains of LAB and Enterococci were shown to produce bacteriocin-like inhibitory substances against pathogens, which thus contribute for sure to the safety of wooden vats. In the last case explored up to date, the Gerle, used in the French PDO Salers (Didienne et al., 2012), the predominant lactic acid bacteria were shown to be lactobacilli and leuconostoc, with the presence of yeasts and molds. Again a large biodiversity in the biofilm composition was observed and was correlated with management procedures. All these molecular descriptions showed that this ancestral system is a reservoir of microbial diversity, obviously safe, and efficient in enriching raw milk. However, beyond these descriptions, many generic questions are still not solved. How the colonization of the wood takes place? How deep is the wood colonized and what is the long term “dialogue” in situ between the microbial biofilm and the wood? Is the wood itself, by its own compounds, contributing to the final safety by inhibiting pathogens (as suggested by Miller et al., 1996)? What is the functional implications of the microbial community structure (Smid et al., 2014) of this complex spontaneous wood biofilm and its exact contribution to final cheese quality and typicity? As some of these cheese making are not continuous and even can be seasonal, how the biofilm survives without nutriment within the wood? Aand finally, how to optimize the management, cleaning of the wood to reassure safety agencies? Indeed, despite the absence of pathogens, and the fact that wood has never been documented to be involved in any food borne disease outbreak, despite the invaluable technological qualities of this natural material (renewable, used from immemorial times, available everywhere, resistant, cheap, etc…), the Codex Alimentarius does not approve the use of wood in contact with food. The main argument is its irregular surface and porous structure, which make wood difficult to clean. Attempts by the FDA to forbid its contact with food and in particular milk and cheese is in the air. European harmonization is still not done and texts governing his use are mainly national; its use is for example under a provisory authorization in France and Italy at least, as the use of wood is mandatory in several PDO cheeses. Only recently appropriate technics to assess the surface contamination of wood were proposed in the literature (Ismael et al., 2014) . In conclusion, more science is urgently needed to reinforce safety and cleaning issues, to better understand the mechanisms underlying the establishment of this natural stable biofilm, the functional balance within the strains and their final contribution to the food typicity and nutritional value (Montel et al., 2014). Wooden tools is a source of unique strains for artisanal and PDO cheeses in many European countries. They are also at least absolutely crucial in many small scale fermentations in developing countries (Holzapfel et al., 2002; Motarjemi et al., 2012; Nout and Motarjemi, 1997). For all these reasons, to ban wooden tools for hypothetical safety reasons would be a very detrimental decision. Science has now revolutionary tools to explore and manage microbial diversity (Cocolin and Ercolini, 2015). This is one of the role of scientists to contribute to informed decision. The case of wooden tools is now urgent.