Background

While advancements in genome sequencing have uncovered numerous genetic variants, determining the actual biochemical changes these variants cause within cells remains a difficult challenge. Previous genomic studies have primarily focused on changes in the expression levels of single proteins or whether protein structure collapses due to amino acid substitutions. A representative approach is the fragmented assessment of whether a variant causes loss-of-function in disease-associated variants.

Cellular life processes are maintained not by the independent action of individual proteins, but by a complexly intertwined Protein-Protein Interaction (PPI) network. Most genetic variants tend to finely tune the binding affinity with specific partner proteins rather than completely destroying protein function. This has led to criticisms that it is difficult to identify the precise biological action of a variant without accurately measuring changes in binding strength at the complex interface.

Environmental factors also add complexity to these interactions. Biological systems are constantly exposed to external stimuli such as nutrient depletion, heat shock, and osmotic stress. The same amino acid mutation can exhibit entirely different binding patterns depending on the physiological state. Due to technical limitations, large-scale studies tracking the quantitative impact of genetic variants on complex binding affinity under various environmental perturbation conditions have been rare.

Key Findings

Researchers analyzed the multidimensional network between genetic variants, environmental perturbations, and protein binding strength using the model eukaryote Saccharomyces cerevisiae. The Protein-fragment Complementation Assay (PCA) technique was used to measure binding strength. This molecular tool is designed so that when two proteins physically bind, split receptor fragments recombine to emit a signal.

The researchers selected 61 protein pairs with already identified interactions and constructed a genotypic library encompassing both natural and artificial mutations. Subsequently, they calculated the quantitative changes in the binding strength of each protein pair under diverse environmental conditions, including increased incubation temperature, nutrient source switching, and treatment with chemical stress factors.

The analysis results clearly demonstrate that the effects of genetic variants are not fixed. In many of the 61 binding pairs, the change in binding affinity due to mutations fluctuated sharply depending on the surrounding culture conditions. For instance, a specific metabolic protein variant, which showed a negligible difference in binding strength of less than 5% compared to the wild type in standard media, exhibited a non-linear response with a decrease in binding strength of over 40% under high-temperature stress conditions.

Allosteric effects, where mutations located outside the protein binding interface disturb binding strength remotely when coupled with environmental stress, were also confirmed. This means that mutations classified as neutral by conventional structure prediction programs can act as factors that hinder complex formation in specific stress environments. Based on the collected binding dataset, the researchers succeeded in statistically classifying the proteins into a 'variable binding group' that responds sensitively to the environment and an 'invariant binding group' that maintains stable binding despite environmental changes.

Significance and Outlook

Moving beyond static genome sequence analysis to incorporate dynamic environmental variables into binding affinity expands the scope of systems biology research. This is because it demonstrates that when predicting the pathogenicity of variants in disease models, the physiological stress environment to which cells are exposed must be considered an essential input. It also shows potential as key training data for improving the accuracy of computer-based variant effect prediction algorithms.

It provides a new analytical framework for the reclassification of Variants of Uncertain Significance (VUS), a major challenge in human genome analysis. This is because it cannot be ruled out that patient-derived variants may form abnormal bindings only within specific tissue-specific metabolic or inflammatory microenvironments.

Expanding precision to the single-cell level and applying it to mammalian cell lines remain challenges to be addressed in the future. Further verification is required to see if the principles obtained from 61 protein pairs in budding yeast apply identically across the entire human interactome, which involves tens of thousands of intertwined proteins. The development of a next-generation quantitative screening platform capable of reflecting the diversity of post-translational modifications (PTM) in higher biological tissues is also identified as a subsequent challenge.