Molecular development of a thermostable β-glucosidase for modification of natural products

Sammanfattning: Popular Abstract in English The gene encoding a β-glucosidase originating from the extreme thermophile Thermotoga neopolitana has been cloned and expressed in Escherichia coli. The aim was to produce a thermostable enzyme that could be used to remove sugar residues from glucosylated natural products classified as flavonoids by applying a method combining extraction in hot pressurized water with enzymatic hydrolysis. The β-glucosidase (termed TnBgl1A in this thesis) is a member of family 1 of glycoside hydrolase (GH1). The enzyme has an apparent unfolding temperature of 101.9 °C and a molecular weight of 52.6 kDa. The activity of TnBgl1A was first analysed using the model substrate para-nitrophenyl-β-D-glucopyranoside (pNPGlc), demonstrating that a single glucosyl residue (typical for β-glucosidases) is released by this enzyme. Hydrolysis of glucosylated forms of the natural product quercetin (the major flavonoid in yellow onion) was found to be dependent on the position of the glucosylation (see below). Expression in E. coli resulted in a relatively large fraction of insoluble target protein. To improve the folding of TnBgl1A during production, different strategies were then applied: I) the gene was constructed synthetically and codons were optimized to match codon usage in E.coli, II) the gene was cloned in frame with a signal peptide to translocate the protein to the periplasmic space, and III) the gene was co-expressed with genes encoding molecular chaperones. Among these strategies the co-expression of the gene with chaperones worked best, and the improved folding resulted in an increased fraction of soluble, active enzyme. The TnBgl1A was tested for the hydrolysis of quercetin-glucosides, which are antioxidants classified as flavonoids present in onion, and composed of a polyphenolic backbone glucosylated at two different positions (Q3, Q4´, and the diglucoside Q3,4´). The aglycone form of quercetin is a more potent antioxidant than the glucosylated forms. The activity of TnBgl1A for Q3 was lower compared to its activity on quercetin-4´-glucoside (Q4´). To improve hydrolysis of Q3, mutations were introduced in the enzyme, based on a structure model of TnBgl1A. The mutant N221S/P342L showed increased efficiency towards Q3 as well as for Q4´ compared to wild-type. This showed that the position and nature of this residue in the active site is important for substrate specificity and that by careful selection the specificity can be changed for different substrates. Therefore, the mutation studies were extended and the active site region was targeted for further mutagenesis. Among the changes introduced, the mutagenesis of the neighbouring residue, N220S, was also found to influence activity, and this variant had a higher specific activity for quercetin-glucosides. TnBgl1A and one of the best performance mutants (N221S/P342L) were immobilized on acrylic support to allow recycling of the enzyme in experiments coupling hot water extraction of quercetin-glucosides with enzymatic hydrolysis. The activity of the immobilised enzyme was analysed in batch experiments using pNPGlc as model substrate and showed not only that the enzyme remained active after immobilisation, but that the thermal stability of the enzyme improved slightly. The effect of additives on immobilisation was studied and, as glucose is an activator for the enzyme, the addition of glucose during immobilisation resulted in a slight increase in the specific activity. In the case of bovine serum albumin (BSA), the original activity was recovered after three months of storage by incubation with BSA for 24 hours. In conclusion, the work in the present thesis shows that TnBgl1A is an effective biocatalyst for the conversion of quercetin-glucosides to quercetin. It was also found that, in the production step, more soluble enzymes can be obtained if they are co-expressed with chaperones. The specificity of the enzyme can be changed by changing a single amino acid in the active site, and the improved hydrolysis of Q3 found for two single mutants (N220S and N221S) is caused by indirect changes in interactions, which may lead to a better fit of the quercetin backbone in the active site. As the modification of amino acids in the active site requires a deep understanding of the structure, it is hoped that the results reported here can contribute to the creation of, new mutants with better activity guided by predictions based on the current results. In this research, we have shown that both free and immobilised enzyme can be coupled to the hot water extraction process and that the immobilised enzyme can be used in an on-line process for hydrolysis of flavonoid glucosides. Further improvements can also be made in such combined processes, both concerning the conditions for the extraction and hydrolysis.