A review of enzyme design in catalytic stability by artificial intelligence.

The design of enzyme catalytic stability is of great significance in medicine and industry. However, traditional methods are time-consuming and costly. Hence, a growing number of complementary computational tools have been developed, e.g. ESMFold, AlphaFold2, Rosetta, RosettaFold, FireProt, ProteinMPNN. They are proposed for algorithm-driven and data-driven enzyme design through artificial intelligence (AI) algorithms including natural language processing, machine learning, deep learning, variational autoencoder/generative adversarial network, message passing neural network (MPNN). In addition, the challenges of design of enzyme catalytic stability include insufficient structured data, large sequence search space, inaccurate quantitative prediction, low efficiency in experimental validation and a cumbersome design process. The first principle of the enzyme catalytic stability design is to treat amino acids as the basic element. By designing the sequence of an enzyme, the flexibility and stability of the structure are adjusted, thus controlling the catalytic stability of the enzyme in a specific industrial environment or in an organism. Common indicators of design goals include the change in denaturation energy (ΔΔG), melting temperature (ΔTm), optimal temperature (Topt), optimal pH (pHopt), etc. In this review, we summarized and evaluated the enzyme design in catalytic stability by AI in terms of mechanism, strategy, data, labeling, coding, prediction, testing, unit, integration and prospect.

Enhancing lactose recognition of a key enzyme in 2'-fucosyllactose synthesis: α-1,2-fucosyltransferase.

BACKGROUND: 2'-Fucosyllactose, a representative oligosaccharide in human milk, is an emerging and promising food and pharmaceutical ingredient due to its powerful health benefits, such as participating in immune regulation, regulation of intestinal flora, etc. To enable economically viable production of 2'-fucosyllactose, different biosynthesis strategies using precursors and pathway enzymes have been developed. The α-1,2-fucosyltransferases are an essential part involved in these strategies, but their strict substrate selectivity and unsatisfactory substrate tolerance are one of the key roadblocks limiting biosynthesis. RESULTS: To tackle this issue, a semi-rational manipulation combining computer-aided designing and screening with biochemical experiments were adopted. The mutant had a 100-fold increase in catalytic efficiency compared to the wild-type. The highest 2'-fucosyllactose yield was up to 0.65 mol mol-1 lactose with a productivity of 2.56 g mL-1  h-1 performed by enzymatic catalysis in vitro. Further analysis revealed that the interactions between the mutant and substrates were reduced. The crucial contributions of wild-type and mutant to substrate recognition ability were closely related to their distinct phylotypes in terms of amino acid preference. CONCLUSION: It is envisioned that the engineered α-1,2-fucosyltransferase could be harnessed to relieve constraints imposed on the bioproduction of 2'-fucosyllactose and lay a theoretical foundation for elucidating the substrate recognition mechanisms of fucosyltransferases. © 2022 Society of Chemical Industry.

Enhancing heterologous expression of a key enzyme for the biosynthesis of 2'-fucosyllactose.

BACKGROUND: 2'-Fucosyllactose (2'-FL) is the most abundant human milk oligosaccharide (HMO) in human milk and has important physiological functions. The market demand of 2'-FL is continuing to grow, but high production cost has limited its availability. To solve the dilemma, biosynthesis of 2'-FL has been proposed and is considered the most promising pathway for massive production. α-1,2-Fucosyltransferase is one of the key elements involved in its biosynthesis, but the limited intracellular accumulation and unstable properties of α-1,2-fucosyltransferases when expressed in host strains have become a major hurdle for the effective biosynthesis of 2'-FL. RESULTS: A combinatorial engineering strategy of synergic modification of ribosome binding site, fusion peptide and enzyme gene was leveraged to enhance the soluble expression of α-1,2-fucosyltransferases and promote enzyme activity. The preferable combination was to employ an optimized ribosome binding site region to drive 3 × FLAG as a fusion partner along with the α-1,2-fucosyltransferase for expression in Escherichia coli (DE3) PlySs, and protein yield and enzyme activity were remarkably improved by 11.51-fold and 13.72-fold, respectively. CONCLUSION: After finely tuning the synergy among different elements, the abundant protein yield and high enzyme activity confirmed that the drawbacks of heterologous expression in α-1,2-fucosyltransferase had been properly addressed. A suitable external environment further drives the efficient synthesis of α-1,2-fucosyltransferases. To our knowledge, this is the first report of a systematic and effective modification of α-1,2-fucosyltransferase expression, which could potentially serve as a guideline for industrial application. © 2022 Society of Chemical Industry.

Overexpression of an endogenous raw starch digesting mesophilic α-amylase gene in Bacillus amyloliquefaciens Z3 by in vitro methylation protocol.

BACKGROUND: Mesophilic α-amylases function effectively at low temperatures with high rates of catalysis and require less energy for starch hydrolysis. Bacillus amyloliquefaciens is an essential producer of mesophilic α-amylases. However, because of the existence of the restriction-modification system, introducing exogenous DNAs into wild-type B. amyloliquefaciens is especially tricky. RESULTS: α-Amylase producer B. amyloliquefaciens strain Z3 was screened and used as host for endogenous α-amylase gene expression. In vitro methylation was performed in recombinant plasmid pWB980-amyZ3. With the in vitro methylation, the transformation efficiency was increased to 0.96 × 102 colony-forming units µg-1 plasmid DNA. A positive transformant BAZ3-16 with the highest α-amylase secreting capacity was chosen for further experiments. The α-amylase activity of strain BAZ3-16 reached 288.70 ± 16.15 U mL-1 in the flask and 386.03 ± 16.25 U mL-1 in the 5-L stirred-tank fermenter, respectively. The Bacillus amyloliquefaciens Z3 expression system shows excellent genetic stability and high-level extracellular production of the target protein. Moreover, the synergistic interaction of AmyZ3 with amyloglucosidase was determined during the hydrolysis of raw starch. The hydrolysis degree reached 92.34 ± 3.41% for 100 g L-1 raw corn starch and 81.30 ± 2.92% for 100 g L-1 raw cassava starch after 24 h, respectively. CONCLUSION: Methylation of the plasmid DNA removes a substantial barrier for transformation of B. amyloliquefaciens strain Z3. Furthermore, the exceptional ability to hydrolyze starch makes α-amylase AmyZ3 and strain BAZ3-16 valuable in the starch industry. © 2020 Society of Chemical Industry.