Expression and site-directed mutagenesis at labile asparagine residues of cyclodextrin glycosyltransferase from Paenibacillus sp. RB01

Abstract

Cyclodextrin glycosyltransferase (CGTase) from a thermotolerant Paenibacillus sp. RB01 exhibits 3 isoforms (I, II, and III) with equal mass but different charge. The aim of this work is to determine the cause of formation of multiple forms of the enzyme. Starting with cloning of CGTase gene (cgt) into pET19b vector and expressed in E. coli BL21 (DE3), the cloned enzyme showed multiple forms as in the wild type. The pI of major isoform I and II from IEF-PAGE were 4.87 and 4.75. Preliminary investigation led us propose that deamidation of Asn to Asp may be the cause of isoform formation. Deamidation rate of each Asn (expressed as Coefficient of Deamidation, CD value) was predicted by the computer algorithm (www.deamidation.org), by using the 3D-structure of CGTase from alkalophilic Bacillus sp. 1011 (PDB code 1PAM) which has 98% similarity in amino acid sequence with CGTase from Paenibacillus sp. RB01. Then, site-directed mutagenesis of labile Asn residues (with CD values of 1 to 158) into Asp and expression was performed, creating 13 single mutants, and one each of double and triple mutants. From native-PAGE, the isoform II of cloned showed the same migration with the major band of N336D, N415D, N427D, and N567D. While CGTase from double mutant (N336D/N415D) and triple mutant (N336D/N415D/N326D) moved similarly with the cloned isoform III. Separation of isoforms was carried out by preparative gel electrophoresis or FPLC with Mono Q or Mono P, or IEF-PAGE. Two-dimensional gel electrophoresis was then used to show which mutated enzyme most resembled to isoform II. By mixing mutated CGTase with isoform II, only the N336D, N427D, and N567D showed no effect on migration of isoform II. Then, the effect of pH and incubation time on the increase in number of isoforms was investigated. Deamidation buffer at pH 9.0 induced faster isoform formation than at pH 6.0, 37ºC. And when the content of isoAsp, an intermediate of deamidation reaction, was determined, the value at pH 9.0 was clearly higher than at pH 6.0 after 15 days of incubation. Then, each isoform was digested with trypsin and separation pattern of peptide analyzed by reverse phase HPLC with C18 column showed no difference between those of isoform I, II and mutated enzymes. MALDI-TOF was then used to investigate size difference in peptides. For mutated enzymes, N415D, N427D, and N567D showed the peptide fragments containing mutated Asn (each has an increase of 1 Da when Asn was mutated to Asp) of the size 1434.74, 1273.70, 2695.33 Da, respectively. N336D/N415D double mutant also showed a 1 Da increase of the fragment size 933.96 and 1434.74. And when compared the cloned isoform I and II, the difference in m/z of peptide containing Asn 427 was observed. Furthermore, from the analysis of isotopic distribution in other interesting peptides of isoform I and II, the difference was observed at two more positions, Asn 336 and Asn 567. Then, the product ratio α: β: γ CD was determined by HPAEC-PAD. Isoform I, II, and mutated enzymes N415D, N427D, and N567D gave similar ratio around 0.25:1.00:0.60. All enzyme forms still produced β-CD as major product. From kinetics of coupling reaction with β-CD, catalytic efficiency (kcat/Km) and turnover number (kcat) of isoform I was not much higher than those of isoform II. The mutated N336D and N567D showed similar values to those of isoform I, while kcat/Km of N427D was significantly lower than both isoforms. The overall results support that Asn deamidation is the cause of formation of multiple forms of CGTase, and Asn 336, Asn 427, and Asn 567 are among the susceptible residues responsible for the formation of CGTase isoforms.