The Role of a Loop in the Non-catalytic Domain B on the Hydrolysis/Transglycosylation Specificity of the 4-α-Glucanotransferase from Thermotoga maritima

Abstract

The mechanism by which glycoside hydrolases control the reaction specificity through hydrolysis or transglycosylation is a key element embedded in their chemical structures. The determinants of reaction specificity seem to be complex. We looked for structural differences in domain B between the 4-α-glucanotransferase from Thermotoga maritima (TmGTase) and the α-amylase from Thermotoga petrophila (TpAmylase) and found a longer loop in the former that extends towards the active site carrying a W residue at its tip. Based on these differences we constructed the variants W131G and the partial deletion of the loop at residues 120-124/128-131, which showed a 11.6 and 11.4-fold increased hydrolysis/transglycosylation (H/T) ratio relative to WT protein, respectively. These variants had a reduction in the maximum velocity of the transglycosylation reaction, while their affinity for maltose as the acceptor was not substantially affected. Molecular dynamics simulations allow us to rationalize the increase in H/T ratio in terms of the flexibility near the active site and the conformations of the catalytic acid residues and their associated pKas.

Keywords: Glucanotransferase; Glycosidases; Hydrolysis; Reaction-specificity; Transglycosylation.

Fig. 1

Topological map of secondary structure elements (left panel) and 3D-structure (right panel) of TmGTase (PDB ID 1LWJ). α-Helices are represented by rods and β-strands are indicated as arrows. Domain A: green and blue; Domain B: red and orange; Domain C: purple (Color figure online)

Fig. 2

General mechanism of glycosyl hydrolase-catalyzed hydrolysis (R′ = H) or transference (R′ = Glycosyl, alkyl)) reactions produced with retaining stereochemical configuration of anomeric carbon

Fig. 3

Sites targeted to change the reaction specificity in the TmGTase. Differences in the targeted loop from domain B of TmGTase (PDB ID 1LWJ, green), and TpAmylase (PDB ID 5M99, gray). The differential part of the protruding loop (shown in blue) with Trp131 in the tip is present only in the transferase enzyme. The competitive inhibitor acarbose –a delimiter of the active site– is presented as yellow sticks. The sequence alignment between TmGTase and TpAmylase for the loop region is presented in the lower panel (Color figure online)

Fig. 4

Structural characterization of TmGTase variants. a CD spectra for TmGTase WT (red), W131G (green), and truncated loop (blue) b SDS-PAGE of purified proteins lane 1 MW marker, lane 2 WT, lane 3 W131G and lane 4 truncated loop TmGTase variants c RMSD distribution of WT (red), W131G (green), and truncated (blue) TmGTase, d Representative cluster centers containing at least the 98% of the population for WT (red), W131G (green) and truncated (blue; more labile regions are shown in red). e Fluctuation of TmGTase in 100 ps intervals. RMSF for WT (upper panel), W131G (middle panel), and truncated loop variant (lower panel). f Structural representation with cartoon putty from Pymol based on RMSF values of WT (left), W131G (middle), and truncated loop variant (right). The acarbose (cyan stick) indicates the active site, placed by grafting from the original 1LWJ PDB structure. All structures are colored from N-terminus (red) to middle (green) to C-terminus (blue) (Color figure online)

Fig. 5

Conformational changes in TmGTase variants. a Initial structure from PDB 1LWJ b Key interactions in the WT protein result in a closed lid conformation c Key interactions in the WT protein result in an open lid conformation d truncated loop mutant loses all the WT interactions leading to lid opening and loop disorder e Interactions in W131G mutant in closed conformation f W131G mutant loses interactions resulting in an open lid conformation

Fig. 6

Transglycosylation activity pH profile of TmGTase variants. a WT, b W131 G, and c truncated TmGTase. Data are presented with errors corresponding to 1 SD and the data were fit to Eqs. 3, 3 and 1, respectively shown in the Materials and Methods section. Lower panels show the respective residuals from the experimental data and the model fit

Fig. 7

Changes in the computed pKa along the simulation for active site residues. Distribution of pKa values for D186 (a), E216 (b), H93 (c), and H94 (d) for WT (red), W131G (green), and truncated loop variant (blue) e representative structures of the distribution of D186 with pKa = 3.00 (left panel) and 10.00 (right panel) in WT f representative structures of the distribution of truncated loop variant D186 with pKa = 4.00 (left panel) and 6.50 (right panel), respectively. The distances between atoms are represented by broken lines (Color figure online)