Chebon-Bore, L. | Search bioRxiv

bioRxiv2026

GBAviewer: a structural database of GBA1 variants in Parkinson's disease

Background: GBA1 variants are common risk factors for Parkinson's disease (PD), yet the structural consequences of most variants remain uncharacterized. No centralized resource currently integrates the growing number of GCase structures, compounds, and PD-associated variants. Objectives: To develop an interactive tool mapping GBA1-PD missense variants onto GCase structures, alongside interactors and therapeutic compounds. Methods: We built GBAviewer, an R/Shiny web server integrating GBA1-PD Browser variants with crystal structures, cryo-EM complexes, AlphaFold3 models, and AlphaMissense scores. We also modelled a putative GCase-Saposin C-glucosylceramide ternary complex using AlphaFold3. Results: GBAviewer maps variants in 3D across the GCase active site, LIMP-2 and SapC surfaces, chaperone pockets, and dimerization interface. Our ternary model corroborates SapC binding to the GCase active site entrance, where several variants of unknown significance cluster. Conclusions: GBAviewer is a freely accessible platform for structural analysis of GBA1 variants, supporting mechanistic studies and therapeutic development. Available at: https://g-can.shinyapps.io/GBAviewer/

neuroscience↗

bioRxiv2021

Force field parameters for Fe2+4S2-4 clusters of dihydropyrimidine dehydrogenase, the 5-fluorouracil cancer drug deactivation protein: a step towards in silico pharmacogenomics studies

The dimeric dihydropyrimidine dehydrogenase (DPD) metalloenzyme, an adjunct anti-cancer drug target contains highly specialized 4 x Fe2+4S2-4 clusters per chain. These clusters facilitate the catalysis of the rate-limiting step in the pyrimidine degradation pathway through a harmonized electron transfer cascade that triggers a redox catabolic reaction. In the process, majority of administered 5-fluorouracil (5-FU) cancer drug is inactivated while a small proportion is activated to nucleic acid antimetabolites. The occurrence of missense mutations in DPD protein within the general population, including those of African descent, has adverse toxicity effects due to altered 5-FU metabolism. Thus, deciphering mutation effects on protein structure and function is vital, especially for precision medicine purposes. We previously proposed combined molecular dynamics (MD) and dynamic residue network (DRN) analysis to decipher the molecular mechanisms of missense mutations in other proteins. However, the presence of Fe2+4S2-4 clusters in DPD poses a challenge for such in silico studies. The existing AMBER force field parameters cannot accurately describe the Fe2+ center coordination exhibited by this enzyme. Therefore, this study aimed to derive AMBER force field parameters for DPD enzyme Fe2+ centers, using the original Seminario method and collation features Visual Force Field Derivation Toolkit as a supportive approach. All-atom MD simulations were performed to validate the results. Both approaches generated similar force field parameters which accurately described the human DPD protein Fe2+4S2-4 clusters architecture. This information is crucial and opens new avenues for in silico cancer pharmacogenomics and drug discovery related research on 5-FU drug efficacy and toxicity issues.

bioinformatics↗