Introduction

α-O-oligosaccharide compounds have important application value in pharmaceuticals, diagnostics, and food industries. In the clinical diagnosis of acute pancreatitis, the detection of blood amylase activity relies on a structurally precise oligosaccharide substrate, EPS-G7, enabling specific, stable, and highly sensitive detection of amylase. However, traditional chemical synthesis of EPS-G7 involves tedious steps with multiple rounds of protection and deprotection, yielding only ~5% overall, and is accompanied by problems such as contamination from metal catalysts and α/β isomeric mixtures, making purification difficult. Although enzymatic synthesis is milder and simpler, natural glycosyltransferases (such as CGTase) usually cause complex side reactions, making it hard to control glycan chain length and linkage specificity, resulting in high product heterogeneity. These challenges have led to the long-term monopoly of EPS-G7 by global giants such as Roche Diagnostics, with China relying entirely on imports, causing high prices and fragile supply chains.

Recently, the research team led by Researcher Yang Guangyu at Shanghai Jiao Tong University, in collaboration with Hanhai Xinzme and Tianwu Technology, published a study in Bioresource Technology titled “Protein language model-assisted directed evolution of cyclodextrinase enables precision α-O-oligosaccharide synthesis.” The team proposed a protein language model–based multi-objective optimization strategy, successfully achieving efficient directed evolution of cyclodextrinase (CDase). Using cyclodextrin as a donor, they developed a one-step enzymatic method for highly efficient synthesis of the EPS-G7 precursor—pNP-G7—achieving a yield of 161 g/L and a conversion rate of 85%, far surpassing traditional chemical synthesis. This provides a brand-new solution for the precision synthesis of α-O-oligosaccharides.

Author Profile

Name: Yang Guangyu
Affiliation: School of Life Sciences and Biotechnology, Shanghai Jiao Tong University / State Key Laboratory of Microbial Metabolism
Title: Researcher
Bio: Ph.D., Researcher at Shanghai Jiao Tong University, specializing in structure–function relationship analysis of enzymes, directed enzyme evolution, and in vitro synthetic biology.

Experimental Results

(1) A New Direction: From Cyclodextrin Glycosyltransferase to Cyclodextrinase

(2) AI-Guided Enzyme Mining: ESBS Motif Probe Identifies High-Potential Candidates

(3) AI Enzyme Design: Protein Language Model Enables Multi-Objective Optimization

(4) Mechanism Elucidation: Why AI-Designed Enzymes Are More “Precise”

(5) Industrial Application: Process Optimization and Scale-Up

Original link:

https://www.sciencedirect.com/science/article/pii/S0960852425011733

Data Display

1. Purity: 

≥95% (vs. ≥90% for leading foreign brands)

Impurity control: 

pNP ≤0.01%, pNPG7 ≤0.1%

Stability:

• Heat-stability for 15 days, deviation <2%
• Open-vial stability for 25 days, deviation <3.5%

Batch-to-batch consistency: 

Deviation ≤5% across clinical linear ranges

Comparison test with imported substrates on clinical samples shows excellent correlation, with an average deviation of <2.5%

Excellent stability

After 15 days of heat stability testing, the variation in clinical sample results was less than 2%

After 25 days of open-vial stability, the variation in clinical sample results was less than 3.5%

Assay compatibility: 

Seamless replacement for imported substrates, correlation bias <2.5%

Strict quality control, large production capacity, dual control, reducing batch-to-batch variation

Randomly select three batches of reagents and compare them using clinical samples covering the linear range; the average batch-to-batch variation does not exceed 5.0%.

Product Ordering