Abstract
The molecular architecture of bacteriorhodopsin (BR) is commonly regarded as a structural template for the three-dimensional structure of membrane receptors that are functionally coupled to guanine nucleotide-binding regulatory proteins (GPCR). More recently, specific molecular models of such GPCR were constructed on the basis of the functional and structural relation of rhodopsin to BR as well as the sequence homology between rhodopsin and the GPCR. Such models of GPCR leave unresolved the difficulty caused by the apparent lack of any significant degree of sequence homology between the seven transmembrane helices (TMH) of BR and the portions in the sequence of the various GPCR that are considered to constitute their transmembrane domains. Evolutionary arguments offered in favor of the structural relation between BR and the opsins, and hence the GPCR, prompted our investigation of the possibility that the sequence homology, including any similarity in the distribution of kink-inducing proline residues among the helices, might have been obscured by the assumption that the TMH maintained their sequential order from BR in the evolution of the mammalian proteins. With a definition of the TMH in the neurotransmitter GPCR guided by hydropathicity predictions, and additional criteria used to define the span of each helix, optimal alignment of each pair of sequences was determined with no gaps allowed in the matching. The resulting alignment proposed here reveals considerable homology between the TMH in BR and those in GPCR, if the sequential order of the helices is ignored. These findings suggest the possibility that exon shuffling could have occurred in the proposed evolution of the GPCR gene from BR and point to a modification of the BR template to account for the correct packing of the helices in the tertiary structures of GPCR. These findings could guide the construction of three-dimensional models of the neurotransmitter GPCR on the basis of specific interhelical interactions observed in BR.