Enzymatic hydrolysis of corn bran arabinoxylan: - theory versus practice

This thesis concerns enzymatic hydrolysis of corn bran arabinoxylan. The work has focused on understanding the composition and structure of corn bran with specific interest in arabinoxylan with the main purpose of targeting enzymatic hydrolysis for increased yields. Corn bran has been used as a model substrate because it represents a readily available agroindustrial side product with upgrading potentials. Corn bran originates from the wet-milling process in corn starch processing, is the outmost layers of the corn kernel and is particularly rich in pentose monosaccharides comprising the major components of arabinoxylan. Corn bran is one of the most recalcitrant cereal byproducts with arabinoxylans of particular heterogeneous nature. It is also rich in feruloyl derived substitutions, which are responsible for extensive cross-linking between arabinoxylan molecules and thereby participate in a complex and ridig cell wall structure. This thesis contains a thorough examination of the monosaccharide and structural composition of corn bran, which is used to assess and apply the relevant mono component enzyme preparations. In this way, the aim is to obtain the most effective minimal enzymatic requirements for hydrolyzing corn bran. The off set of the work has been a basic set of four hemicellulases consisting of an endo-β-1,4-xylanase (GH10 from H. insolens), a β-xylosidase (GH3 from T. reesei) and two α-L-arabinofuranosidases (GH43 and GH51 from H. insolens and M. giganteus respectively). This set of enzymes have proven efficient in degrading arabinoxylan structures from wheat arabinoxylan and it is also verified in this study that it probably is among the best available hemicellulases for increasing the hydrolysis of corn bran arabinoxylan at present. This set of enzymes creates a solid starting point for hydrolysis of the arabinoxylan structure but is not alone capable of catalyzing complete hydrolysis. Auxiliary enzyme activities that catalyse the hydrolysis of various substitutions are also necessary and several of such enzymes are investigated. This results in the identification of a suitable feruloyl esterase from A. niger (FAE-III) for catalyzing the release of free ferulic acid and diferulic acids to a certain extent. Furthermore, a novel acetyl xylan esterase from Flavolaschia sp. is also found to be important for obtaining higher release of xylose from the arabinoxylan structure. Structural analysis of a soluble fraction of corn bran also confirms the presence of highly acetylated pento-oligosaccharides. All these enzymes together with a commercial cellulase preparation (Cellic™ CTec) are capable of catalyzing the release of up to 36% xylose from a soluble fraction of hydrothermally pretreated corn bran. Yet enzymatic hydrolysis of corn bran is far from complete and in order to improve the yields, this thesis has thoroughly investigated the need and impact of different pretreatment conditions. Corn bran is a special substrate when it comes to pretreatment conditions because the biomass is mainly composed by heat, acid and alkali labile linkages in arabinoxylan. It therefore becomes a balancing task to find optimum conditions that compromise the advantages and disadvantages. Acidic pretreatments (pH 1.5-2) are found to be particularly effective in promoting the enzymatic hydrolysis, especially with respect to xylose and glucose release, but vast amounts of the valuable monosaccharides are lost during this pretreatment and this is especially evident for arabinose. From a scientific point of view acid catalysed pretreatment renders the substrate in a state of disruption where assessment of correct enzyme administration becomes difficult and enzymatic hydrolysis becomes a secondary route to disintegration. Alkaline pretreatments are less efficient in promoting the enzymatic hydrolysis, but still serve an academic purpose because those conditions chemically remove diferulate cross-linkings between arabinoxylans, which have been believed to be a major obstical for enzymatic hydrolysis. The chemical removal of these cross-links allows for the interpretation of hindering effects of cross-linking and it is concluded that they do not pose a significant barrier for enzymatic hydrolysis. By this conclusion a major hypothesis of this thesis is rejected. Because chemically catalysed pretreatments has obvious disadvantages, milder mechanical pretreatments has also be investigated and results show that decreasing the particle size of the insoluble substrate renders it more accessible to enzymatic hydrolysis. The hydrolysis improves with a factor of 3-8 for xylose, arabinose and glucose when comparing the yields in the largest particle size fraction to the yields in the smallest size fraction for native destarched corn bran. This is related to an increased substrate surface area, but it is also observed that different particle size fractions from corn bran are not uniformly composed. The content of monosaccharides varies and results in differences in content and composition of cellulose and arabinoxylan. These differences in biomass composition may very well also be part of the explanation why increased enzymatic hydrolysis is obtained. To further investigate the influence of particle size and other physical parameters on enzymatic hydrolysis, theoretic estimations of how changing particle size influences the enzymatic hydrolysis is made. These estimations point to the observation that other factors than particle size alone governs the enzymatic hydrolysis. It is observed that enzymatic hydrolysis is promoted in certain particle size fractions and inhibted in others. This is likely to be related to the biomass composition. Corn bran is a recalcitrant substrate and complete hydrolysis is not achieved in this thesis. Instead explanations as to what causes the recalcitrance are sought and it most likely lies within a combination of factors. Firstly, corn bran has an exceptional rigid and tight exterior that leaves it virtually impenetrable to enzymes. Disruption of this outside structure is important if the hydrolysis is at all to commence. In that sense it is important to obtain a higher understanding of the cell wall matrix, the packing of polysaccharides and how they interact with other polymeric structures in the cell wall, eg proteins and lignin. Especially proteins associated with the cell wall may play a significant role in maintaining cell wall strength and preventing enzymatic hydrolysis. Secondly, the heterogeneous nature of arabinoxylan from corn bran makes it difficult even for the correct enzymes to catalyse complete hydrolysis as observed for hydrolysis in a soluble corn bran fraction. Once the arabinoxylan structure is free of the cell wall matrix the hydrolysis seem to be restricted due to steric hindrance or lack of additional enzymes to catalyse the hydrolysis of certain unusual bonds. In particular, it is of outmost importance to target arabinosyl substitutions of arabinoxylan and other possible configurations of arabinose, as this in particular may hold part of the reason for corn bran recalcitrance. Generally, increased arabinose release will most likely also lead to increase in the overall release of xylose. Obstructions by heterogeneous arabinoxylan may be overcome by completing the knowledge about corn bran arabinoxylan, which can then lead to the identification of missing, central enzyme activities, and thereby also make the work on corn bran generic. The thesis is based upon the scientific publications produced during the last four years and they represent the development and achievements of this work. To ease the reading the thesis will highlight some of the findings and interpretations from the publications, but also from unpublished work and thereby establish the mindset and progress behind the project.