Extending the osmophobic effect to protein side chains with a unified transfer model across osmolyte classes

Understanding the role of the protein backbone and side chains on cosolvent-induced stabilization is essential for a molecular picture of osmolyte action. The dominant view has been that protecting osmolytes stabilize proteins primarily through unfavorable interactions with the peptide backbone - the osmophobic effect - with side chains playing a minor or opposing role. By revisiting the decomposition of amino acid transfer free energies with proper account of the mutual shielding between backbone and side-chain groups, we derive a transfer model that is consistent with experimental denaturation m-values for urea and protecting osmolytes simultaneously - a feat neither the established nor the previously proposed universal-backbone models could achieve alone. A mechanism-dependent interpretation of backbone accessibility is proposed: geometric for excluded cosolvents, complete for binders where cosolvent-backbone interactions are specific. The model reveals that for all strong protecting osmolytes, including TMAO, sarcosine, sucrose, trehalose, and sorbitol, both backbone and side chains contribute favorably to protein stabilization, with side-chain contributions comparable to or exceeding those of the backbone. For urea, the model recovers the known balanced backbone and side-chain contributions to denaturation when the directional nature of urea-backbone hydrogen bonding is accounted for, which makes the backbone accessible to urea regardless of side-chain shielding. Weaker protectants such as proline, betaine, and glycerol are distinguished by competing backbone and side-chain effects that partially cancel. These results extend the osmophobic effect to protein side chains and establish a three-tier classification of osmolyte action: cooperative backbone and side-chain stabilization, cooperative destabilization, and competing contributions. The greater sensitivity of the model predictions to side-chain composition provides avenues for experimental validation of the underlying physical assumptions and for protein engineering.