Discovery of a Thermostable GH10 Xylanase with Broad Substrate Specificity from the Arctic Mid-Ocean Ridge Vent System

A two-domain GH10 xylanase-encoding gene (amor_gh10a) was discovered from a metagenomic data set, generated after in situ incubation of a lignocellulosic substrate in hot sediments on the sea floor of the Arctic Mid-Ocean Ridge (AMOR). AMOR_GH10A comprises a signal peptide, a carbohydrate-binding module belonging to a previously uncharacterized family, and a catalytic glycosyl hydrolase (GH10) domain. The enzyme shares the highest sequence identity (42%) with a hypothetical protein from a Verrucomicrobia bacterium, and its GH10 domain shares low identity (24 to 28%) with functionally characterized xylanases. Purified AMOR_GH10A showed thermophilic and halophilic properties and was active toward various xylans. Uniquely, the enzyme showed high activity toward amorphous cellulose, glucomannan, and xyloglucan and was more active toward cellopentaose than toward xylopentaose. Binding assays showed that the N-terminal domain of this broadspecificity GH10 binds strongly to amorphous cellulose, as well as to microcrystalline cellulose, birchwood glucuronoxylan, barley p-glucan, and konjac glucomannan, confirming its classification as a novel CBM (CBM85). IMPORTANCE Hot springs at the sea bottom harbor unique biodiversity and are a promising source of enzymes with interesting properties. We describe the functional characterization of a thermophilic and halophilic multidomain xylanase originating from the Arctic Mid-Ocean Ridge vent system, belonging to the well-studied family 10 of glycosyl hydrolases (GH10). This xylanase, AMOR_GH10A, has a surprisingly wide substrate range and is more active toward cellopentaose than toward xylopentaose. This substrate promiscuity is unique for the GH10 family and could prove useful in industrial applications. Emphasizing the versatility of AMOR_GH10A, its N-terminal domain binds to both xylans and glycans, while not showing significant sequence similarities to any known carbohydrate-binding module (CBM) in the CAZy database. Thus, this N-terminal domain lays the foundation for the new CBM85 family.