Zeng Bo, Zhao ShuYan, Zhou Rui, Zhou YanHong, Jin WenHui, Yi ZhiWei, Zhang GuangYa
Department of Bioengineering and Biotechnology, Huaqiao University, Xiamen, Fujian Province 361021, China.
Technology Innovation Center for Exploitation of Marine Biological Resources, Third Institute of Oceanography, Ministry of Natural Resources, Xiamen, Fujian Province 361005, China.
Comput Struct Biotechnol J. 2022 Jun 27;20:3313-3321. doi: 10.1016/j.csbj.2022.06.050. eCollection 2022.
Engineering of hydrolases to shift their hydrolysate types has not been attempted so far, though computer-assisted enzyme design has been successful. A novel integrative strategy for engineering and screening the β-1,3-xylanase with desired hydrolysate types was proposed, with the purpose to solve problems that the separation and preparation of β-1,3--oligosaccharides was in high cost yet in low yield as monosaccharides existed in the hydrolysates. By classifying the hydrolysate types and coding them into numerical values, two robust mathematical models with five selected attributes from molecular docking were established based on LogitBoost and partial least squares regression with overall accuracy of 83.3% and 100%, respectively. Then, they were adopted for efficient screening the potential mutagenesis library of β-1,3-xylanases that only product oligosaccharides. The virtually designed AncXyl10 was selected and experimentally verified to produce only β-1,3-xylobiose (60.38%) and β-1,3-xylotriose (39.62%), which facilitated the preparation of oligosaccharides with high purity. The underlying mechanism of AncXyl10 may associated with the gap processing and ancestral amino acid substitution in the process of ancestral sequence reconstruction. Since many carbohydrate-active enzymes have highly conserved active sites, the strategy and their biomolecular basis will shield a new light for engineering carbohydrates hydrolase to produce specific oligosaccharides.
尽管计算机辅助酶设计已取得成功，但迄今为止尚未尝试通过工程手段改变水解酶的水解产物类型。本文提出了一种新颖的综合策略，用于工程改造和筛选具有所需水解产物类型的β-1,3-木聚糖酶，旨在解决β-1,3-寡糖的分离制备成本高、产率低的问题，因为水解产物中存在单糖。通过对水解产物类型进行分类并将其编码为数值，基于LogitBoost和偏最小二乘回归建立了两个稳健的数学模型，从分子对接中选择了五个属性，总体准确率分别为83.3%和100%。然后，将它们用于高效筛选仅产生寡糖的β-1,3-木聚糖酶的潜在诱变文库。选择了虚拟设计的AncXyl10并通过实验验证，其仅产生β-1,3-木二糖（60.38%）和β-1,3-木三糖（39.62%），这有助于制备高纯度的寡糖。AncXyl10的潜在机制可能与祖先序列重建过程中的缺口处理和祖先氨基酸取代有关。由于许多碳水化合物活性酶具有高度保守的活性位点，该策略及其生物分子基础将为工程改造碳水化合物水解酶以产生特定寡糖提供新的思路。
