Sortase A-mediated crosslinked short-chain dehydrogenases/reductases as novel biocatalysts with improved thermostability and catalytic efficiency

Abstract

(S)-carbonyl reductase II (SCRII) from Candida parapsilosis is a short-chain alcohol dehydrogenase/reductase. It catalyses the conversion of 2-hydroxyacetophenone to (S)-1-phenyl-1,2-ethanediol with low efficiency. Sortase was reported as a molecular "stapler" for site-specific protein conjugation to strengthen or add protein functionality. Here, we describe Staphylococcus aureus sortase A-mediated crosslinking of SCRII to produce stable catalysts for efficient biotransformation. Via a native N-terminal glycine and an added GGGGSLPETGG peptide at C-terminus of SCRII, SCRII subunits were conjugated by sortase A to form crosslinked SCRII, mainly dimers and trimers. The crosslinked SCRII showed over 6-fold and 4-fold increases, respectively, in activity and k _cat/K _m values toward 2-hydroxyacetophenone compared with wild-type SCRII. Moreover, crosslinked SCRII was much more thermostable with its denaturation temperature (T_m) increased to 60 °C. Biotransformation result showed that crosslinked SCRII gave a product optical purity of 100% and a yield of >99.9% within 3 h, a 16-fold decrease in transformation duration with respect to Escherichia coli/pET-SCRII. Sortase A-catalysed ligation also obviously improved T_ms and product yields of eight other short-chain alcohol dehydrogenases/reductases. This work demonstrates a generic technology to improve enzyme function and thermostability through sortase A-mediated crosslinking of oxidoreductases.

Figure 1

Schematic illustration showing the SrtA-mediated crosslinking of SCRII. GGGGSLPETGG tag was fused into the C-terminal end of SCRII. SrtA recognized and cleaved the tag, followed by the attack from the glycine at the N-terminal end to form crosslinked protein complexes (mainly dimers and trimers). The covalently bond dimers and trimers were used as units to form further oligomerisation, of which the intermolecular conformation was assumed in this work. The dotted line meaned possible covalent bond to form cyclic crosslinked SCRII. The crosslinked SCRII was subsequently applied to the asymmetric reduction of 2-HAP to (S)-PED.

Figure 2

Determination of SrtA-mediated ligation products. (a) Left: Optimization of the temperature for the SrtA-mediated ligation. SCRII-mtf and SrtA were incubated at different temperatures for 8 h. Right: Ligation result after 36 h for WT SCRII and SCRII-mtf. Lanes 1 and 2, WT SCRII and SCRII-mtf were incubated with SrtA for 36 h at 25 °C, respectively. Lane 3, Purified crosslinked SCRII. (b) MALDI-TOF-MS analysis of crosslinked SCRII. (c) The elution profiles of crosslinked SCRII, SCRII-mtf and WT SCRII by size exclusion chromatography. The red dotted lines were used for alignment of peaks.

Figure 3

CD spectra (a) and thermal shift assays (b) of WT SCRII, SCRII-mtf and crosslinked SCRII. The CD spectra were recorded by measuring the ellipticity as a function of wavelength at 0.1-nm increment between 190 and 250 nm at 20 °C. Thermal shift assay was conducted by heating the three samples from 20 to 90 °C and then cooling to 20 °C by a Jasco programmable Peltier element with a rate of 1 °C/min. The wavelength of 209 nm that characterized the α-helix within protein was used to monitor the unfolding rate of the protein structures. The buffer (5 mM phosphate buffer pH 6.0, 0.1 mg/mL enzymes.) showed pH shift with no more than 0.5 while it was heated from 20 °C to 100 °C.

Figure 4

Effects of temperature on enzyme activity for WT SCRII, SCRII-mtf and crosslinked SCRII. Enzyme assay was performed in 100 mM potassium phosphate buffer (pH 6.0), 0.5 mM NADPH, 5 mM substrate, and appropriate enzymes (mg/mL) at 20–70 °C. The relative activity was described as percentage of maximum activity under experimental conditions. All experiments were repeated three times.

Figure 5

Substrate specificities of WT SCRII, SCRII-mtf and SrtA-mediated Crosslinked SCRII. ^aStandard assay conditions: 100 mM phosphate buffer (pH 6.0), 0.5 mM NADPH, 5 mM substrate, 1 mg/mL enzymes at 35 °C. The data was the representative of three independent experiments and standard errors were not more than 5%. ^bThe biotransformation was conducted with 5 g/L substrate and sufficient NADPH for 6 h. S, substrate; Yd, yield; Cf, configuration, ee, enantiomeric excess.

Figure 6

Michaelis-Menten and Lineweaver-Burk plots of WT SCRII, SCRII-mtf and Crosslinked SCRII toward 2-HAP. Each value was calculated depending on three independent measurements and all standard errors of fits were not more than 5%. The data obtained were fitted to the function v = V_max[s]/(K _m + [s]).

Figure 7

SDS-PAGE analysis of SrtA-mediated ligation of 8 oxidoreductases after incubating for 36 h. Mono-oxidoreductases included oxidoreductase-mtf, oxidoreductase-GGGGSLPET and cyclized oxidoreductase-GGGGSLPET. Ligation products, mainly dimers and trimers, were labelled as crosslinked oxidoreductases.

Figure 8

Thermal shift curves of 8 oxidoreductases and their crosslinked oxidoreductases. The buffer (5 mM phosphate buffer pH7.0, 0.1 mg/mL enzymes.) showed pH shift with no more than 0.5 while it was heated from 20 °C to 100 °C.