Identification of key residues for the functional differentiation of glycosyl hydrolase family 1

IRAZOQUI, JOSE MATÍAS; EBERHARDT MA. FLORENCIA; AMADIO, ARIEL

Workshop; Tercer encuentro de la Red de tecnología enzimática; 2021

Red de Tecnología Enzimática

Carbohydrates are widely distributed in nature, where they mediate a multitude of biological functions. The number of possible combinations for a small oligosaccharide results in an astronomical structural and functional diversity. Therefore, the enzymes hydrolysing these sugars, glycoside hydrolases, must also present a wide diversity of structures and functionalities. CAZy is a widely used database that proposes a family classification for carbohydrate related enzymes, based on sequence identity, with at least one biochemically characterized member in each family. This system of classification keeps in a single family those enzymes that show similar sequence and also has conserved sequence motifs. This kind of classification enables the study of the evolutionary relationships, the structural features, and mechanisms of catalysis. In consequence this database provides useful bioinformatic resources to classify new enzymes. However, the occurrence of enzymes that act on different substrates in the same family is a significant problem for the automated functional annotation of CAZyme-related genes. The glycosyl hydrolase 1 family (GH1), is a broad group of enzymes with activities that are of interest for the industry, like β-glucosidase, β-galactosidase, and β-D-fucosidase. Even though this kind of enzymes have been used for decades now, little is known about what defines their substrate specificity.Here, we present a first effort for an in depth analysis of the GH1 family. First, we downloaded all GH1 sequences from the CAZy database classified as "characterized", and all the information available for each entry (i.e. taxonomic classification, EC number). Then, using a profile obtained from the dbCAN database, we extracted the GH1 domain from each sequence, aligned them using MAFFT and the resulting alignment was used to construct a phylogenetic tree, using RAxML. The main driver for clustering in the phylogeny was taxonomical, but within the clade containing most bacterial enzymes, sequences were grouped together according to their EC number, suggesting a functional clustering. These clusters were extracted, realigned together, and then we searched for key residues that could explain the differentiation between groups, in the context of a crystal structure, downloaded from the PDB. We use VMD to visualize the structure and WebLogo to create logos for each group.We identify several positions in the catalytic pocket of the enzyme that vary between groups of sequences, namely 3.2.1.21, 3.2.1.85 and 3.2.1.86, suggesting that these amino acids would be key in the selection of substrates. We consider this work a first attempt to identify what drives the activity differentiation within the GH1 family and would help to avoid unwanted substrate specificity when using an automatic annotation of genes to detect novel enzymes. The addition of more, well characterized sequences and wet lab testing of new candidates for each group would help improve this model even further.