ProtASR: An Evolutionary Framework for Ancestral Protein Reconstruction with Selection on Folding Stability

Area of Science:

• Evolutionary biology
• Biochemistry
• Computational biology

Background:

• Traditional ancestral sequence reconstruction (ASR) methods often use empirical models that assume uniform evolutionary rates across protein sites.
• This assumption is frequently violated due to varying selective constraints, leading to inaccurate reconstructions.
• Protein evolution is influenced by factors like folding stability, which is not adequately addressed by current ASR tools.

Purpose of the Study:

• To develop a novel evolutionary framework, ProtASR, for inferring ancestral protein sequences that accounts for selection on protein stability.
• To improve the accuracy of ancestral protein reconstruction by incorporating site-specific substitution models that consider protein folding.
• To address the known bias in maximum-likelihood ASR methods that tend to overestimate ancestral protein stability.

Main Methods:

• ProtASR employs a structurally constrained mean-field (MF) substitution model to generate site-specific substitution matrices, considering both unfolding and misfolding stability.
• MF models have previously demonstrated superior performance over empirical and other structurally constrained models in likelihood and amino acid distribution inference.
• A maximum-likelihood (ML) ASR procedure is adapted to infer ancestral proteins using these MF models, overcoming biases of standard ML approaches.

Main Results:

• ProtASR, using MF models, produced reconstructed proteins with less biased stabilities compared to those reconstructed using empirical models (JTT and CAT).
• The stabilities of proteins reconstructed by ProtASR were significantly closer to the true stabilities of simulated ancestral proteins.
• Analysis of extant protein families revealed that folding stability evolves dynamically over time, with some families showing more constant stability than others.

Conclusions:

• ProtASR provides a more accurate method for ancestral sequence reconstruction by incorporating protein folding stability.
• The framework offers practical implications for biomedicine and biotechnology by enabling more reliable insights into past biological events.
• The study highlights the dynamic nature of protein folding stability evolution across different protein families.