Amino Acid\nResidue-Specific Ramachandran Distributions\nDerived from a Simple Mean Field Potential

Protein dynamics in the unfolded state, in the context\nof early\nstage protein folding or intrinsically disordered proteins (IDPs),\nis not well understood. The discovery of IDPs, and their sequence-dependent\ndynamics, has led to many computational and experimental investigations\nregarding the conformational preferences of short oligopeptides and\nindividual amino acid residues in the unfolded state. As proteins\nconsist of sequences of amino acid residues, characterizing the intrinsic\nconformational preferences of the individual residues in the unfolded\nstate is crucial for understanding the emergent conformations of peptides\nand proteins. While advances have been made in understanding conformational\npreferences, the atomistic mechanisms driving these preferences remain\nunresolved. In this work, we show that the distributions of atomic\noverlaps between backbone and side chain atoms in Ramachandran space\nare unique for amino acid residue mimetic structures alanine, valine,\nleucine, and isoleucine in Ramachandran space indicating unique intrapeptide\nenergy landscapes for each residue. We then construct a mean field\npotential consisting of only an empirical peptide backbone–water\nand average intrapeptide Lennard-Jones contributions to explore their\ninfluence on the conformational preferences. With this fairly simple\nmodel, we were able to produce Ramachandran distributions that qualitatively\nagree with previously reported experimental and computational predictions\nabout the conformational preferences of these amino acid residues\nin the unfolded state in water. Our results indicate these conformational\npreferences are the result of the balance between pPII-stabilizing\nbackbone–water interactions and repulsive side chain–backbone\ninteractions where the latter will depend uniquely on the atomic makeup\nand geometry of the side chain.