This is a very daring structural proposal, which will stimulate and spark many discussions. The monomeric rhodopsin structure imposes significant problems for the homology modelling of dimeric GPCRs. Here it is employed for the structure prediction of a far more complex system involving additional TM domains and many protein protein interactions. The supporting movie displays a very appealing pore, but the substrate must enter from the side. Furthermore, this pore must be regulated or blocked under normal conditions, otherwise it would be an unregulated aqueaporin with a giant opening. Further structures based on Asp KOs may be "caught in the act" with the substrate.

View all comments by Boris Schmidt