On the variation of structural divergence among residues in enzyme evolution
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (Arcadia Science)
Abstract
Structural divergence varies among protein residues. Unlike the classic problem of substitution rate variation, this structural divergence variation has been largely ignored. Here we show that in enzymes structural divergence increases with both residue flexibility and distance from the active site. Although these factors are correlated, we demonstrate through modelling that the pattern arises from two independent types of constraints, non-functional and functional. Their relative importance varies across enzyme families: as functional constraints increase from 4% to 85%, non-functional constraints decrease from 96% to 15%, reshaping the divergence pattern. This analysis overturns two accepted views in protein evolution: First, evolutionary variation has been thought to mirror protein dynamics generally, but we show this similarity exists only when non-functional constraints dominate. Second, active site structural conservation has been attributed to functional constraints alone, but we show it stems largely from their location in rigid regions where non-functional constraints are high.
Article activity feed
-
First, evolutionary variation has been thought to mirror protein dynamics generally, but we show this similarity exists only when non-functional constraints dominate. Second, active site structural conservation has been attributed to functional constraints alone, but we show it stems largely from their location in rigid regions where non-functional constraints are high.
This is a really interesting paper looking at how different constraints contribute to evolutionary divergence within enzymes! There are some important findings that I'll definitely keep in mind when working with enzymes in the future.
-
AlphaFold
I wonder if you expanded outside of the PDB and used predicted structures if you would find similar trends. It could even be a useful way to evaluate AlphaFold.
-
While we could have detected and filtered out these cases by examining active site structural differences, doing so would have introduced circularity in our analysis of active site conservation. Instead, their detection as statistical outliers serves to validate our analytical approach by showing it can identify cases that violate its assumptions.
I like the inclusion of these outliers and the discussion about why they may be outliers.
-
Our results show that indeed, much of their enhanced conservation is explained by their lack of flexibility.
I think this is such a cool finding! I'm wondering if there are there any cases where the active site is in the more flexible region of the protein? Or if instead of using distance to the active site you used distance to the least flexible region (especially if it's separate from the active site)? That could maybe help sort this out even more.
-
Section ??)
Just pointing this out!
-
Thus, active sites are more conserved than average, but their conservation cannot be attributed solely to functional constraints. These results show that much of their conservation is due to their location in rigid regions, indicating the need to reconsider the assumption that active site conservation arises exclusively from functional constraints.
Very interesting!
-
As the relative contribution of distance increases, protein size decreases (from 322 to 107 residues) and mean distance to active site shortens (from 15.3Å to 9.8Å). These changes align with the shift from flexibility to distance contributions. The spread of local flexibility, sd(lRMSF), which we showed drives flexibility contributions, shows no consistent trend in these cases. This pattern illustrates our earlier finding that the spread of flexibility and mean distance from the active site explain only 38% of variation in relative contributions between constraints, with the remaining variation potentially arising from other factors such as the variation among families of selection pressure.
I may have missed this, but it would be interesting to see if these trends hold for all protein families studied.
-
