Structure of integral membrane proteins

National Institute Of Neurological Disorders And Stroke

Investigators

Linked publications & trials

Abstract

Our knowledge about GPCR structures has advanced considerably since the first GPCR structure (bovine rhodopsin) was published in 2000 by Palczewski and colleagues. For crystallization trials, receptors must be purified in the presence of detergents; the choice of detergent becomes critical to maintain the GPCR in a functional, correctly folded form. Rhodopsin shows remarkable stability in detergent solution as long as it is kept in the dark to maintain its inactive state; this detergent tolerance allowed extensive crystallization screens and led to diffraction-quality crystals. The beta-1 adrenergic receptor is much less stable in detergent solution; an extensive alanine/leucine scanning mutagenesis approach was used to identify a mutant receptor suitable for crystallization. The beta-2 adrenergic and adenosine A2a receptors were engineered with T4 lysozyme replacing most of the flexible intracellular loop 3. The receptors were crystallized with an inverse agonist or antagonist bound to promote the inactive state. The transmembrane cores look similar but not identical in these receptor structures. Differences are seen in the N- and C-terminal receptor regions and in the loops connecting the helices. The interpretation of these differences remains uncertain since they may originate from the respective receptor sequences, disorder in flexible regions, or possibly from protein engineering. Recently, the structures of active GPCRs, and a complex of receptor with G-protein have been solved. NTS1 is very stable in an optimized detergent mixture for purification, but has reduced stability in detergents preferred for crystallization. Mutational approaches must be applied to obtain a mutant receptor which stays in a biologically relevant, single conformation, long enough for crystallization to occur. A structure of NTS1 alone and in complex with an antagonist will show whether different inactive conformations exist and how these relate to the currently known receptor structures with emphasis on Gq-coupled receptors. Our recent structure of NTS1 in complex with neurotensin in its active form shows the conformation of such a key signaling state and how agonist binding effects the changes on the intracellular receptor surface needed for G-protein engagement, and allows the comparison with other receptor structures thought to represent active state conformations.