Abstract :
[en] Thesis summary
In the last decade, numerous single-species biofilm reactors of various configurations have been implemented at lab and pilot scale for the production of chemicals and biological products. Compared to their counterparts in submerged cultures, these processes benefit from the specific physiology of biofilms, i.e. high robustness of the microbial system, long-term activity, continuous implementation and low ratio size / productivity. However, the risks of biofouling and the lack of analytical tools for the control and the monitoring of biofilms are obstacles for scale-up strategies. Up to now, single-species biofilm reactors have been mainly confined to the production of metabolites ranging from low (bulk chemicals) to medium (fine chemicals) added values. In this way, there is a need to design efficient single-species biofilm reactors exhibiting good scalability potentials and intended for the production of high added value compounds.
In this work, an experimental single-species biofilm reactor has been designed for the production of target molecules derived from metabolic pathways involved in biofilm physiology. On the basis of these criteria, three biological models having good abilities of biofilm formation and secretion performances were selected :
- the gram positive bacterium Bacillus subtilis for the production of surfactin, a surface active metabolite involved in biofilm formation.
- the filamentous fungus Trichoderma reesei for the production of hydrophobin (HFBII), a surface active protein (7kDa) involved in adhesion process of spores and mycelium on solid surface.
- the filamentous fungus Aspergillus oryaze (engineered strain) for the production of a recombinant protein (Gla::GFP) under the control of the glaB promoter specifically activated in solid-state fermentation.
The proposed experimental biofilm reactor has the configuration of a trickle-bed bioreactor. The agitation axis of a stirred tank reactor has been removed and replaced by a stainless steel structured packing filling the top of the vessel. The liquid medium, located in the bottom of the vessel is continuously recirculated on the packing element thanks to a peristaltic pump. An ascending air flow is performed above the liquid phase just under the packing element.
This thesis reports the screening of the three biological models in the experimental biofilm reactor. The results include the characterization of process performances in terms of biofilm formation and secretion of the target molecule under different operating conditions.
An original methodology based on high energy X-ray tomography has been developed to non-invasively visualize and quantify the biofilm colonization inside the packing element. This technique has highlighted that biofilm colonization and liquid phase distribution across the packing are strongly interrelated phenomena. The biofilm of B. subtilis occurring by cell aggregation preferentially developed on solid areas wetted by the liquid. Accordingly, optimal operating conditions improving liquid phase distribution have been defined for biofilm colonization. The fungal biofilm of A. oryzae and T. reesei occuring by cell filamentation equally colonize submerged and aerial surfaces of the packing element. Consequently, another configuration of biofilm reactor comprising a packing element totally immersed in the liquid medium has been investigated. The production yields of surfactin and hydrophobin in the experimental biofilm reactor are respectively 1.25 and 2.64 times greater than those of a submerged culture in a stirred tank reactor. This suggests that surface-active molecules involved in biofilm formation have a real interest for the design of single-species biofilm reactors. Although the Gla::GFP fusion protein is greater produced in the stirred tank culture, its integrity was preserved in the biofilm reactor despite the presence of proteases. This suggests that the quality and the stability of heterologous proteins produced in a fungal biofilm reactor are improved compared with a submerged culture. Finally, the implementation of the biofilm reactor has led to technological progresses including low energy consumption, no foam formation, continuous processing and simplification of downstream process operations.
Further experiments should deepen the understanding of structured phenotypic heterogeneity impact on secretion performances in the biofilm reactor. These experiments should consider development of operating conditions allowing for the growth of a thin biofilm homogeneously distributed on the whole surface provided by the packing element in order to optimize nutrients and metabolites mass transfers. The scale-up and the continuous implementation of the process should be also investigated.
