Application of Computational Methods for Fermentative Hydrogen Production

Energy and environment are inseparable, since the production and use of energy always affects the environment. Current energy production relies on nonrenewable energy sources such as oil, coal and natural gas. However, the continuous production of energy from limited resources is not sustainable. This creates an urgent need to develop new methods for the production of energy from renewable sources. One possible solution is fermentative hydrogen (H₂) production. H₂ is seen as a future energy carrier. Fermentative H₂ production has many environmental advantages such as ability to use wastes as the source of energy and possibility to apply ambient temperature and pressure. Drawbacks are rather low yields and slow H₂ production rates. In order to overcome these issues vast amount of research has been conducted. Under anaerobic conditions, various anaerobic and facultatively anaerobic bacteria utilize organic compounds by fermentation and excrete H₂ as a byproduct. In the nature, bacteria exist as mixed cultures. With appropriate pretreatments and culture conditions, H2 producing bacteria can be enriched. Microscopy can be used for visual examination of bacterial communities, which can reveal their diversity and dominant bacterial species. Additionally wide range of fluorescent staining methods can be employed in the microscopic analysis of bacterial groups. The manual analysis of the microscopy images is user dependent and laborious. Moreover, the visual quantification of fluorescence intensities and morphological features is impossible. Therefore, automated image analysis methods were developed, e.g., for monitoring culture compositions in the H2 producing bioreactors. The highest H₂ production rates have been achieved with undefined mixed cultures, where the role of each bacterium to H₂ production is not exactly known. In this work, the properties of Escherichia coli and Clostridium butyricum that often coexists in mixed bacterial cultures are described. Additionally the effect of coculture of E. coli and C. butyricum was investigated and found to enhance the utilization of the given substrate. Moreover, the effects of growth conditions and possibilities of genetic modification to H₂ production by E. coli and C. butyricum are presented. The biological approach to the design of experiments often relies on intuition. However, with computational methods higher understanding over fermentative H₂ production can be achieved. Computational methods in this work mostly focus on the modeling of bacterial metabolism and some emphasis is also given to the systematic design of experiments. Metabolic models are interaction based presentations of reactions occurring within metabolic pathways, in which the knowledge of molecules and enzymes taking part to reactions is combined. The largest metabolic models are based on the complete genome of bacteria. Metabolic models can be used to help in designing mutations and cultivation conditions to enhance bioprocesses. Various approaches, such as flux balance analysis, can be used to simulate and analyze metabolic models. Here, the existing genome-scale metabolic model is utilized with flux balance analysis for analysis and enhancement of fermentative H₂ production. Increasing amount of knowledge and the need to make the processes as efficient as possible has made the utilization of computational tools inevitable. Therefore, cooperation between experts with biological and computational skills is encouraged. Commonly, the aid of a computational expert is requested when data mining from an overwhelming amount of existing measurements is needed. Actually, the cooperation should start from experimental design to gain most information over the system by applying statistical design-of-experiment methods. This thesis gives an overview of computational methods applied to fermentative H₂ production and describes the use of genome-scale metabolic models to experimental design, analysis and modeling.