EventsThe 6th International Electronic Conference on Foods
Published
This submission belongs to the session A. Food Technology and Engineering of the event The 6th International Electronic Conference on Foods
Published date
27 Oct, 2025
Academic Editor
author-avatarElsa Gonçalves
Citation
Hoa Dang Khanh Chu, COMPUTATIONAL SCREENING OF ENZYME VARIANTS OF 2-OXOISOVALERATE DEHYDROGENASE SUBUNIT ALPHA, in Proceedings of The 6th International Electronic Conference on Foods, 28 October–30 October 2025, MDPI: Basel, Switzerland
Share
Email
Facebook
Twitter
LinkedIn

COMPUTATIONAL SCREENING OF ENZYME VARIANTS OF 2-OXOISOVALERATE DEHYDROGENASE SUBUNIT ALPHA

1. Department of Bioproducts and Biosystems, School of Chemical Engineering, Aalto University, Espoo, 02150, Finland, Finland
Abstract

Microbial engineering is a powerful tool for advancing sustainable food biotechnology, particularly through expanding the metabolic capacity of established hosts such as Saccharomyces cerevisiae. One key limitation of yeast metabolism is the absence of branched-chain α-keto acid dehydrogenase activity, which restricts its ability to process branched-chain substrates and produce valuable compounds. To address this, we conducted a computational search for enzyme variants of 2-oxoisovalerate dehydrogenase subunit alpha (EC 1.2.4.4) that could be functionally expressed in yeast.

Using PSI-BLAST, we identified 1,286 candidates for Yarrowia lipolytica A0A1H6QCU5 and 1,358 for Aspergillus niger A0A370BTM3. After redundancy removal with CD-HIT, the numbers were reduced to 684 and 691, respectively. Protein structures were predicted with AlphaFold, yielding 404 and 450 models for the two seeds. Structural alignment with TM-align showed strong similarity, with most candidates scoring between 0.85–0.95 for A0A1H6QCU5 and 0.86–0.99 for A0A370BTM3, well above the functional threshold of 0.7. Active site mapping with crystal structure 1QS0 (from Pseudomonas putida) confirmed conservation of key residues across several variants.

Phylogenetic analysis revealed that while most candidates originated from filamentous fungi and vertebrates, a smaller set came from Saccharomycotina species, which are better suited for yeast engineering. Promising examples include Blastobotrys adeninivorans (TM-score 0.92, 90.5% coverage), Geotrichum candidum (0.95, 89.0%), and Yarrowia lipolytica (0.87, 88.8%). In total, fewer than 10 high-confidence yeast variants were shortlisted for future testing.

This work demonstrates how computational pipelines combining sequence mining, structure prediction, and phylogenetics can guide enzyme discovery for food biotechnology. The next stage will involve validating these candidates experimentally in S. cerevisiae, with potential applications in flavor synthesis, carbon cycling, and sustainable fermentation.

Keywords
Enzyme Engineering
Metabolic Pathway Optimization
Yeast Biotechnology
Functional Food Ingredients
Sustainable Food Production
Computational Protein Design
Bioinformatics in Food Sciences
Synthetic Biology Applications
Branched-Chain Amino Acid Me
Traditional vs. innovative rice cakes: the contribution of Alcobaça apple and Spirulina
Changes in the Textural and Pasting Properties and the Microstructure of Two Early Indica Rice Varieties During Storage Using General Linear Model Univariate (GLMU) Analysis