Exploring Potential Signals of Selection for Disordered Residues in Naturally Occurring Prokaryotic and Eukaryotic Proteins

Intrinsically disordered proteins (IDPs) were recognized as an important class of proteins in all domains of life for their functional importance. However, how nature has shaped the disorder potential of prokaryotic and eukaryotic proteins is still not clearly known. Randomly generated sequences are free of any selective constraints thus these sequences are commonly used as null models. Considering different types of random protein models here we seek to understand how disorder potential of natural eukaryotic and prokaryotic proteins differs from random sequences. Comparing proteome-wide disorder content between real and random sequences of 12 model organisms we noticed that while in eukaryotes natural sequences tend to be more disordered than random sequences prokaryotes follow an opposite trend. By analyzing position-wise disorder profile, here we showed that there is a general trend of higher disorder near the N and C-terminal regions of eukaryotic proteins as compared to the random models; however, either no or a weak such trend was found in prokaryotic proteins. Moreover here we showed that this preference is not due to the biases either in the amino acid or nucleotide composition or other factors at the respective sites. Instead, these regions were found to be endowed with a higher fraction of protein-protein binding sites suggesting their functional importance. Here, we proposed various explanations for this pattern such as improving the efficiency of protein-protein interaction, ribosome movement, and post-translational modification, etc. However, further studies are needed to clearly understand the biophysical mechanisms causing the trend.