Improved amino acid sequence fingerprints to identify proteins with dehalogenase activity through screening of microbial genomes

Dehalogenases are environmentally important enzymes that detoxify organohalogens by cleaving their carbon-halogen bonds. Sequencing of microbial genomes reveals that many harbour enzyme families containing potential dehalogenases. However, characterization of dehalogenases on a greater scale is challenging, because little information is available to pinpoint proteins with this catalytic activity amongst thousands of family members having significant overall sequence similarity while displaying other hydrolytic activities. Reliable sequence markers are needed to fully profit from genome sequencing efforts for accelerating the discovery of novel dehalogenases with improved activities. Using enzymatic screens, 7 dehalogenases, 20 esterases and 32 phosphatases were identified amongst 174 candidates selected from five relevant microbial genomes. Subsequent protein sequence comparisons have identified dehalogenase fingerprints which improved the selection accuracy 16-fold over sequence similarity searches alone. Their broader applicability was demonstrated when an independently performed search discovered 11 dehalogenases from only 27 candidates using the improved procedure identified in this thesis.