GH70 dextransucases : Insights on the molecular determinants of dextran molar mass control

Glucansucrases (GS) from glycoside hydrolase family 70 (GH70) are -transglucosylases produced by lactic acid bacteria. From sucrose, an economical and abundant agro resource, they catalyze the polymerization of glucosyl residues. Depending on the enzyme specificity, α-glucans vary in terms of size, types of glucosidic bonds and degree of branching and have found multiple industrial applications mainly related to their molar mass (MM). However synthesizing polymers of controlled size with average MM ranging from 1 kg/mol to several millions g/mol and low polydispersity using one single enzyme remains challenging. Indeed, the molecular mechanisms underpinning the control of polymer size have been scarcely explored. To tackle this question, two GSs producing dextran (glucan composed of a majority of α-(1,6) linkages) were selected, and their mode of action explored via biochemical and structural analyses coupled to mutagenesis. The first enzyme selected, called DSR-M synthesizes only low molar mass (LMM) dextran (28 kg/mol) exclusively composed of -(1→6) linkages without any trace of HMM dextran (105 to 108 g/mol). In contrast, DSR-OK (second model), produces the highest MM dextran (>109 g/mol) described to date. Several 3D crystallographic structures of a truncated form of DSR-M (DSR-M2), either free or in complex with its substrate or product (isomaltotetraose) in the domain V or in the active site were solved. Such complexes were never obtained before. Noteworthy, one structure encompassed the most complete domain V reported to date. Analyses of these structures coupled to dextran synthesis monitoring, showed that the LMM dextran specificity of DSR-M2 is explained by a distributive elongation mode due to the weak affinity of its two sugar binding pockets in the domain V which interact with the growing dextran chains and allow the synthesis of dextran longer than 16 kg/mol. 15N1H NMR analyses (HSQC), for the first time performed with such a big protein, further revealed the crucial role of aromatic residues in the catalytic domain for the production of dextran from 2 to 16 kg/mol. In comparison, synthesis of HMM dextran by DSR-OK was shown to be mainly due to the sugar binding pockets of its domain V, ensuring much stronger interactions with growing dextran chains. The role of these pockets was evidenced for both enzymes, their functionality proposed to be linked to the presence of one aromatic stacking residue. Their positioning along domain V relatively to the active site is also important to promote efficient binding. All these findings highlight the cooperation between domain V and the catalytic domain for dextran elongation, offer new perspectives to acquire a deeper knowledge on this interplay and open promising strategies for GH70 enzyme engineering aiming at modulating glucan size.