Abstract
The efficient characterization of glycan structures remains a critical challenge in glycomics. While DNA and protein sequencing have advanced rapidly, glycan sequencing lags due to structural complexity and low charge density. This paper presents a novel "hydrolysis-assisted" nanopore sensing strategy using an engineered $\alpha$-hemolysin (M113R/T115A) pore. By combining the specificity of exoglycosidases with high-resolution nanopore discrimination, we achieved sequential identification of monosaccharide units and glycosidic linkages. Machine learning models classified hydrolysis fragments with over 90% accuracy. This approach offers a high-sensitivity, high-throughput alternative to traditional mass spectrometry and NMR for glycan sequencing.
Introduction: The Glycan Sequencing Bottleneck
Glycans—complex sugar chains—play vital roles in biological processes, yet their structural characterization is notoriously difficult. Unlike the linear sequences of DNA or proteins, glycans are often branched and feature diverse glycosidic linkages ($\alpha$ vs. $\beta$, 1-3 vs. 1-4). Furthermore, their neutral or low-charge nature makes them resistant to the electrophoretic forces typically used in nanopore sequencing.
Current gold standards like Mass Spectrometry (MS) and Nuclear Magnetic Resonance (NMR) are powerful but suffer from high operational complexity, long processing times, and limited sensitivity for certain structural isomers. There is an urgent need for direct glycan sequencing technologies with improved sensitivity and throughput.
Methods: A "Reverse-Sequencing" Approach
The researchers developed a strategy based on sequential enzymatic hydrolysis and nanopore fingerprinting.
Engineered Nanopore Sensing
The core of the system is a mutant $\alpha$-hemolysin pore (M113R/T115A). This engineering creates a sensing region specifically tuned to the volume and length of neutral glycan chains. As glycans translocate through the pore, they cause characteristic blockages in the ionic current ($I$).
Sequential Hydrolysis and Fingerprinting
Instead of attempting to read a long glycan chain in one pass, the team utilized exoglycosidases—enzymes that act as molecular scissors to snip off specific monosaccharides from the non-reducing end. As the enzymes shorten the chain (e.g., from a hexasaccharide "Hexa" to a pentasaccharide "Penta"), the nanopore detects a shift in the electrical current blockage ratio ($\Delta I_1/I_0$).
Crucially, the pore identifies the remaining chain's length and composition rather than the released monosaccharide. This "remaining chain" strategy avoids signal interference from small hydrolysis products.
Results and Discussion
The experimental setup demonstrated high sensitivity to glycan concentration and chain length. The figure below illustrates the current traces for a hexasaccharide (Hexa) at varying concentrations, showing a clear relationship between concentration and signal frequency.
Machine Learning Classification
To interpret the noisy translocation signals, the team employed machine learning models, including Neural Networks and Random Forests. These models were trained on over 1,500 translocation events for each glycan standard. The Neural Network model achieved a classification accuracy of 90.9% for various glycan fragments.
Performance Benchmarks
The nanopore method showed significant advantages over traditional techniques:
- Sensitivity: Required 100-fold lower concentrations of sample compared to Thin-Layer Chromatography (TLC).
- Speed: The identification process was at least three times faster than conventional HPLC and MS methods.
- Linkage Specificity: Because exoglycosidases are bond-specific, successful cleavage inherently identifies the glycosidic linkage type.
Conclusion and Future Directions
This proof-of-concept study successfully sequenced five consecutive building blocks in a model heptasaccharide. While the current work relied on a limited set of enzymes, expanding the "kit" of glycosidases will enable the sequencing of more complex mammalian glycans. The transition from single-channel setups to arrayed nanopores is the next logical step to achieve true high-throughput glycan sequencing.
Summary updated