Functional and structural characterization of thermostable D-amino acid aminotransferases from Geobacillus spp

Abstract

D-amino acid aminotransferases (D-AATs) from Geobacillus toebii SK1 and Geobacillus sp. strain KLS1 were cloned and characterized from a genetic, catalytic, and structural aspect. Although the enzymes were highly thermostable, their catalytic capability was approximately one-third of that of highly active Bacilli enzymes, with respective turnover rates of 47 and 55 s(-1) at 50 degrees C. The Geobacillus enzymes were unique and shared limited sequence identities of below 45% with D-AATs from mesophilic and thermophilic Bacillus spp., except for a hypothetical protein with a 72% identity from the G. kaustophilus genome. Structural alignments showed that most key residues were conserved in the Geobacillus enzymes, although the conservative residues just before the catalytic lysine were distinctively changed: the 140-LRcD-143 sequence in Bacillus D-AATs was 144-EYcY-147 in the Geobacillus D-AATs. When the EYcY sequence from the SK1 enzyme was mutated into LRcD, a 68% increase in catalytic activity was observed, while the binding affinity toward alpha-ketoglutarate decreased by half. The mutant was very close to the wild-type in thermal stability, indicating that the mutations did not disturb the overall structure of the enzyme. Homology modeling also suggested that the two tyrosine residues in the EYcY sequence from the Geobacillus D-AATs had a pi/pi interaction that was replaceable with the salt bridge interaction between the arginine and aspartate residues in the LRcD sequence.

FIG. 1.

Comparison of d-amino acid aminotransferases from thermophilic Bacilli strains using Western blotting analyses. Lanes 1 to 7 are the immunostains for the cell extracts from different Bacillus strains: 1, YM1; 2, LK1; 3, LK2; 4, SK1; 5, KLS1; 6, KL1; and 7, SD1. The polyclonal antibody was prepared from an antiserum boosted by the YM1 d-AAT. Lanes 8 to 11 are the purified d-AATs: 8 and 10, Bacillus sp. strain YM1; 9 and 11, G. toebii SK1. Lanes 10 and 11 are the sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis of the purified YM1 and SK1 proteins.

FIG. 2.

Sequence alignment of Bacillus d-AATs. Abbreviations: G.SK1, G. toebii SK1; G.KLS1, Geobacillus sp. strain KLS1; G.kau, G. kaustophilus; B.YM1, Bacillus sp. strain YM1, B.sph, B. sphaericus; B.sub, B. subtilis; B.lic, B. licheniformis; B.cer, B. cereus; B.ant, B. anthracis; B.hal, B. halodurans; L.mon, Listeria monocytogenes; S.hae, Staphylococcus haemolyticus. The conserved residues are shaded dark, while similar residues are shaded light. The asterisk indicates the lysine residue that forms a sciff base with pyridoxal 5′-phosphate.

FIG. 3.

Effect of temperature and pH on SK1 d-AAT. The open symbols indicate the native enzyme, while the closed symbols with dotted lines indicate the mutant enzyme with an LRcD sequence instead of 144-EYcY-147. (A) The native and mutant enzymes were incubated at different temperatures for 20 min in a 0.1 M Tris-HCl buffer (pH 8.5), and the remaining activities determined at 50°C to evaluate the respective thermal stabilities. (B) The native and mutant enzymes were assayed at different reaction temperatures for 20 min. The concentrations of pyruvate from d-alanine were measured by using the salicylaldehyde method. (C) The native enzyme was incubated in different buffers for 1 h at 50°C, and the remaining activities were assayed to evaluate the pH stability. (D) The native enzyme was assayed for 20 min at different pHs with the following 0.1 M buffers: sodium acetate buffer (pH 4.0 to 6.0), potassium phosphate buffer (pH 6.0 to 8.0), Tris-HCl buffer (pH 7.5 to 9.0), and N-cyclohexyl-3-aminopropanesulfonic acid buffer (pH 9.0 to 12.0).

FIG. 4.

Homology model for G. toebii SK1 d-AAT. (A) 3-D Align analysis of the monomeric units of the wild and mutant enzymes; (B) magnified structures of the 143-wEYcYik-149 loop. The loop structure bears the cofactor binding lysine (underlined) and is located at the crevice between the N and C domains. The 143W is related to the intersubunit packing to form a catalytic dimer. The capital letters in EYcY indicate the residues substituted for LRcD in the mutant enzyme. The structures of the native and mutant enzymes are represented in green and blue, respectively. The gray molecule shows the structure of d-cycloserine pyridoxalphosphate, adopted from the PDB-entry 2DAA.