Welland, J. W. J. | Search bioRxiv

bioRxiv2026

Ensemble Machine Learning Approaches Predict Survival in Lower-Grade Glioma Based on Glycosphingolipid Gene Expression and Metabolic Modelling

Glycosphingolipids (GSLs) are essential components of biological membranes with important roles in cell signalling. Disrupted GSL metabolism is associated with malignancy across a range of cancers, with different GSLs implicated in distinct tumours. GSLs have potential mechanistic roles in cancer; however, their functions in Lower Grade Gliomas (LGGs) remain poorly understood. We present ensemble machine learning approaches using transcriptomic data from LGG, combined with GSL-specific metabolic simulations, to predict survival outcomes. The ensemble approach demonstrates effective risk stratification for LGG patients based on GSL gene expression. Pathway analysis of model-derived risk groups highlighted potential association of GSLs with cell motility, division and Wnt signalling in LGG pathology. Given the strong performance of machine learning approaches to predict outcomes and that GSLs are shed into the tumour microenvironment, GSL-based diagnostics and prognostics may prove clinically beneficial. A Python package enabling GSL-specific metabolic modelling and risk prediction from RNA-seq data is provided.

bioinformatics↗

bioRxiv2025

Conformational Dynamics and Membrane Insertion Mechanism of B4GALNT1 in Ganglioside Synthesis

Glycosphingolipids (GSLs) are crucial membrane components involved in essential cellular pathways. Complex GSLs, known as gangliosides, are synthesised in the ER/Golgi by a series of glycosyltransferase enzymes. Imbalances in GSL metabolism can lead to severe diseases, often affecting the nervous system. B4GALNT1 is a key enzyme in the ganglioside synthetic pathway, synthesising complex gangliosides including GM2 and GD2 from GM3 and GD3. These products are precursors to the major brain gangliosides. Loss of B4GALNT1 function causes hereditary spastic paraplegia 26 (HSP26), while its overexpression is linked to cancers including childhood neuroblastoma. Here, we present crystal structures of the homodimeric B4GALNT1 enzyme demonstrating conformational changes upon binding of donor substrate ligands and product. These structures support a catalytic mechanism that involves dynamic remodelling of the substrate binding site during catalysis. We also demonstrate that processing of lipid substrates by B4GALNT1 is severely compromised when surface loops flanking the active site are mutated from hydrophobic residues to polar. Molecular dynamics simulations support that these loops can insert into the lipid bilayer explaining how B4GALNT1 accesses and processes lipid substrates. By combining structure prediction and molecular simulations we propose that this mechanism of dynamic membrane insertion is exploited by other, structurally distinct GSL synthesising enzymes.

biochemistry↗