Packing of apolar amino acids is not a strong stabilizing force in transmembrane helix dimerization.

The factors that stabilize the folding and oligomerization of membrane proteins are still not well understood. In particular, it remains unclear how the tight and complementary packing between apolar side chains observed in the core of membrane proteins contributes to their stability. Complementary packing is a necessary feature since packing defects are generally destabilizing for membrane proteins. The question is the extent by which packing of apolar side chains-and the resulting van der Waals interactions-is a sufficient driving force for stabilizing the interaction between transmembrane helices in the absence of hydrogen bonding and polar interactions. We addressed this question with an approach based on high-throughout protein design and the homodimerization of single-pass helices as the model system. We designed hundreds of transmembrane helix dimers mediated by apolar packing in the backbone configurations that are most commonly found in membrane proteins. We assessed the association propensity of the designs in the membrane of Escherichia coli and found that they were most often monomeric or, at best, weakly dimeric. Conversely, a set of controls designed in the backbone configuration of the GAS