More Exquisitely Adapted Species Have Higher Structural Disorder in Vertebrate Protein Domains

Protein structural disorder helps avoid misfolding and aggregation, but in impeding protein folding might also impede function. The balance between these two selective pressures on protein biophysics might vary among species as a function of the effectiveness of weak selection.  We predicted the Intrinsic Structural Disorder (ISD) of Pfam domains across 118 fully-sequenced vertebrate species and estimated the effect of species identity on to control for differences in Pfam composition across species. We compared this to each species’ Codon Adaptation Index of Species (CAIS), a metric we developed to quantify for effectiveness of selection from synonymous codon usage, corrected for total genomic GC content and amino acid composition, to be comparable across species.
Simple correlations between ISD and CAIS indicate that well-adapted species tend to have high ISD (Spearman’s R = 0.67, p < 2e-16). To correct for phylogenetic confounding and resulting pseudoreplication, we transformed CAIS and ISD species effect data using Phylogenetic Independent Contrasts (PIC). Phylogenetically controlled linear models confirmed that better-adapted species have higher ISD (Spearman’s R = 0.43, p = 1.2e-7). This relationship is conserved across protein domains dated older than 2.1 Gya and protein domains dated younger than 1.5 Gya, though the relationship is stronger in recent domains than in ancient. This indicates selection for higher disorder, with the greater benefit to more recently emerged vertebrate protein domains.