Among the glycoside hydrolases (GHs) classified within the Carbohydrate-Active enZymes (CAZy) database, the α-amylase family GH13 [1] containing ~30 different enzyme specificities and more than 24,000 sequences, represents one of the largest GH families [2]. The family GH13 has officially been divided into 40 subfamilies [3]. The two closely related so-called oligo-1,6-glucosidase and neopullulanase subfamilies described earlier [4] and defined based on a characteristic sequence motif in their 5^th conserved sequence region, cover several CAZy GH13 subfamilies [5].

The present bioinformatics study has been focused on the neopullulanase subfamily (the subfamily GH13_20) in an effort to shed some light on the evolution of its modular arrangement, i.e. to reveal the evolutionary relationships between the catalytic domain representing the enzyme specificity and the non-catalytic SBDs. The neopullulanase subfamily GH13_20 is formed by the enzyme specificities of neopullulanase, cyclomaltodextrinase and maltogenic amylase that in most cases possess at their N-terminus a starch-binding domain (SBD) classified as the carbohydrate-binding module (CBM) family CBM34 [1].

The studied set that was extracted from the CAZy database subfamily GH13_20 and family CBM34, consisted of 69 family GH13 members representing biochemically characterized amylolytic enzymes. It is worth mentioning that the CBM34 was found to be present also in members of GH13 subfamilies other than GH13_20 (or the subfamily has been unassigned yet). Interestingly, several proteins from the collected set possessed the SBD from the family CBM20 (instead of CBM34) positioned N-terminally, although the CBM20 is usually seen in the GH13 at the C-terminal end of the enzymes with the α-amylase and cyclodextrin glucanotransferase specificities [6]. Further, some representatives contain a CBM48 preceding the typical CBM34-GH13_20 modular arrangement. In addition, there are also GH13_20 members lacking any CBM.