Rath, R.; Biswas, M.

In vivo protein-protein association is modulated by non-specific interactions with the cellular matrix, known as crowding. Replicating the heterogeneity and complexity of cellular conditions in experiments and simulations to quantify the crowding influences remains a formidable challenge. As an alternative, homogeneous mixtures of polymers or proteins as crowders are employed, although they provide limited insight to the actual effect. In this work, we present a coarse- grained model of GB1 dimerization in lysozyme crowders by adjusting the GB1-lysozyme effective potential parameters obtained through high-resolution Martini simulation. Employing these parameters it is shown that GB1 dimerization is destabilized in presence of lysozyme, as observed in experiments. The model is further extended to incorporate binary crowder mixtures for crowder species of various sizes and interaction potentials. It is found that depending on mixing conditions and crowder volume fraction, dimerization can either be stabilized or destabilized. In particular, in presence of crowders of different sizes, the smaller one plays the dominant role. Finally, we observe that the overall change of stability due to crowding emerges from a delicate balance between enthalpy and entropy that cannot be predicted by considering the property of single crowder species. These results provide invaluable insights for interpreting experimental observations under cell-like conditions.