Direct Glycan Analysis of Biological Samples and Intact Glycoproteins by Integrating Machine Learning-Driven...

Area of Science:

• Analytical Chemistry
• Biochemistry
• Spectroscopy

Background:

• Glycan pattern monitoring is crucial for understanding cellular processes.
• Existing glycan analysis tools are resource-intensive and not suitable for routine laboratory use.

Purpose of the Study:

• To develop a novel, accessible glycan detection platform for routine laboratory use.
• To overcome limitations of current resource-intensive glycan analysis methods.

Main Methods:

• Integration of surface-enhanced Raman spectroscopy (SERS) with boronic acid (BA) receptors and machine learning.
• Utilizing BA receptors for stereoselective binding to cis-diol-containing glycans, generating unique vibrational spectra.
• Combining Raman spectra from multiple BA receptors to enhance glycan structural information for classification and quantification.

Main Results:

• Direct analysis of biological samples like whole milk and intact glycoproteins (fetuin, asialofetuin) without glycan release or purification.
• High classification accuracy for milk oligosaccharides, even with background interference.
• Accurate quantification of sialylation levels in fetuin/asialofetuin mixtures.
• Differentiation of α2,3 and α2,6 sialic acid linkages by selecting appropriate BA receptors.

Conclusions:

• The developed SERS-based sensor offers a low-cost, rapid, and highly accessible tool for routine glycan screening.
• This platform enables direct analysis of complex biological samples, simplifying glycan analysis.
• The sensor demonstrates significant potential for advancing glycomics research in standard laboratory settings.