Final Report Summary - BAR STRUCTURE (Microcrystallography of stabilized adrenergic receptors)
The aim of this project was to advance the knowledge about b1-adrenergic receptor using X-ray crystallography. The nature of the crystals of the receptor being thin small needles requires corresponding approach to data collection while performing diffraction experiment. The method for such a data collection was established previously while working on the structures of rhodopsin in Dr Schertler group. It was possible to find satisfactorily the structure of the rhodopsin using this technique and therefore it was planned to apply this microsrystallography technique for the b1-adrenergic receptor. As was envisaged in the proposal it was not easy to get the phases for the structure, which could reveal it after refinement.. This was due to the absence of good homology to model used for molecular replacement and not sufficient resolution of the data. It has happened that both these shortcomings were removed at the same time and the very first structure of b1-adrenergic receptor could be obtained. Thus the major goal of the project was achieved even before the actual term of the fellowship has started. Nonetheless this appeared to be only the starting point for an exciting work, which could lead to a further progress in the receptor field. Even after the actual fellowship has ended there are still a number of questions requiring microcrystallography, which has to be answered to fulfil the suggested in the proposal line of the research. The most prominent one is regarding the improvement of the data analysis method with current data collection strategy. Altogether the accumulation of the new knowledge on the receptor and advancing the technical site of the approach for obtaining the structures are the good results of the fellowship term.
As for the new knowledge obtained for the receptor in course of the fellowship, then it was in the obtaining of the multitude of structures with antagonist and agonist ligands. The microcrystallography approach was routinely used for obtaining these structures. It showed to produce the satisfactorily results with a structures in its original formulation. Altogether 6 different structures of antagonist bound receptor and 5 structures of the agonist bound receptor were produced. This was for the following antagonists: carazolol, cyanopindolol, iodocyanopindolol, nadolol and timolol; and following agonists: isoproterenol, salbutamol, dobutamine and carmoterol. Below the main implications and conclusions are given separately for antagonist and agonist structures.
The work on antagonist structures unexpectedly brought the interesting fact of the receptor nature in to light. One important part of the structure, namely cytoplasmic loop 3 (CL3), was unresolved in the very first structure of the receptor. And we were looking into the structure of the missing bit from other data, which was collected for different antagonists. The nature of the receptor crystals is so that the CL3 is involved in crystal contacts and to be convinced that the conformation of the loop found in the structure is genuine it is necessary to assure that in different crystal environments this conformation is the same. The crystals of the antagonist bound receptor have grown in several space groups and could be classified in several groups with significantly different lattice parameters. In the selected antagonist datasets the 5 different crystal environments for the CL3 were observed and among 8 different monomers of the receptor 4 was found to have the one conformation for the loop and 4 was found to have the different conformation. Each of the two conformations was completely reproduced among 4 structures in corresponding group in spite of the different crystal environment. This convinced us that the conformations are genuine and not induced by the crystal but rather selected on the basis of producing good contacts for crystal growth. Therefore two different conformations of the CL3 can be attributed to the inactive state. The conformations are not correlated to the nature of the ligand, which also refers them to the phenomena of basal activity. This was an exciting finding especially since one of the conformations had the famous ionic lock region closed and the other opened. The one with closed ionic lock could be related to the most inactive state of the receptor whereas the other with ionic lock open could be related to the conformation one step further in the activation process. The receptor is known to have some activity in the absence of the ligand (basal activity). And two conformations of the CL3 found in antagonist structures possibly shed the light on the basal activity phenomena. Therefore the systematic analysis of the antagonist datasets revealed the important information on the receptor nature. However the structure of the receptor binding pocket with two further antagonists nadolol and timolol in combination with iodocyanopindolol, cyanopindolol and carazolol used for CL3 structure determination is on its own a valuable source of information about the receptor-ligand interaction. The results of this work on antagonist structures are prepared for the publication. Paper entitled "Two distinct conformations of helix 6 observed in antagonist-bound structures of the b1-adrenergic receptor" is submitted to PNAS journal and one further paper on antagonist ligand binding is in preparation.
The work on agonist structures has revealed the important initial rearrangements in the binding pocket leading to the signalling state of the receptor. Full and partial agonists (isoproterenol, salbutamol, dobutamine and carmoterol) were crystallized bound to the receptor. Microcrystallography technique was used to collect and process the data. Binding of all four agonists induces a 1 Å contraction of the catecholamine binding pocket. Full agonists form hydrogen bonds with two conserved serine residues in transmembrane helix 5 (Ser5.42 and Ser5.46) but partial agonists only interact with Ser5.42. The structures provide an understanding of the pharmacological differences between different ligand classes, which illuminates how GPCRs function and provides a solid foundation for the structure-based design of novel ligands with predictable efficacies. Unfortunately the significant structural changes like those found in opsin structure were not observed, which could be due to the fact that the receptor was thermostabilized with several point mutations towards the inactive state and also due to the absence of stabilizing effect of the G-protein or any its fragment. Nonetheless the analysis of the ligand binding for these agonists shows the basis for efficacy modulation and specificity of binding to b1 receptor rather than b2 or b3. The structures also underlined the importance of several helical interfaces in activation process, like TM4 and TM5 interface for example. Results of this work were published in Nature, January 2011.
As for the new knowledge obtained for the receptor in course of the fellowship, then it was in the obtaining of the multitude of structures with antagonist and agonist ligands. The microcrystallography approach was routinely used for obtaining these structures. It showed to produce the satisfactorily results with a structures in its original formulation. Altogether 6 different structures of antagonist bound receptor and 5 structures of the agonist bound receptor were produced. This was for the following antagonists: carazolol, cyanopindolol, iodocyanopindolol, nadolol and timolol; and following agonists: isoproterenol, salbutamol, dobutamine and carmoterol. Below the main implications and conclusions are given separately for antagonist and agonist structures.
The work on antagonist structures unexpectedly brought the interesting fact of the receptor nature in to light. One important part of the structure, namely cytoplasmic loop 3 (CL3), was unresolved in the very first structure of the receptor. And we were looking into the structure of the missing bit from other data, which was collected for different antagonists. The nature of the receptor crystals is so that the CL3 is involved in crystal contacts and to be convinced that the conformation of the loop found in the structure is genuine it is necessary to assure that in different crystal environments this conformation is the same. The crystals of the antagonist bound receptor have grown in several space groups and could be classified in several groups with significantly different lattice parameters. In the selected antagonist datasets the 5 different crystal environments for the CL3 were observed and among 8 different monomers of the receptor 4 was found to have the one conformation for the loop and 4 was found to have the different conformation. Each of the two conformations was completely reproduced among 4 structures in corresponding group in spite of the different crystal environment. This convinced us that the conformations are genuine and not induced by the crystal but rather selected on the basis of producing good contacts for crystal growth. Therefore two different conformations of the CL3 can be attributed to the inactive state. The conformations are not correlated to the nature of the ligand, which also refers them to the phenomena of basal activity. This was an exciting finding especially since one of the conformations had the famous ionic lock region closed and the other opened. The one with closed ionic lock could be related to the most inactive state of the receptor whereas the other with ionic lock open could be related to the conformation one step further in the activation process. The receptor is known to have some activity in the absence of the ligand (basal activity). And two conformations of the CL3 found in antagonist structures possibly shed the light on the basal activity phenomena. Therefore the systematic analysis of the antagonist datasets revealed the important information on the receptor nature. However the structure of the receptor binding pocket with two further antagonists nadolol and timolol in combination with iodocyanopindolol, cyanopindolol and carazolol used for CL3 structure determination is on its own a valuable source of information about the receptor-ligand interaction. The results of this work on antagonist structures are prepared for the publication. Paper entitled "Two distinct conformations of helix 6 observed in antagonist-bound structures of the b1-adrenergic receptor" is submitted to PNAS journal and one further paper on antagonist ligand binding is in preparation.
The work on agonist structures has revealed the important initial rearrangements in the binding pocket leading to the signalling state of the receptor. Full and partial agonists (isoproterenol, salbutamol, dobutamine and carmoterol) were crystallized bound to the receptor. Microcrystallography technique was used to collect and process the data. Binding of all four agonists induces a 1 Å contraction of the catecholamine binding pocket. Full agonists form hydrogen bonds with two conserved serine residues in transmembrane helix 5 (Ser5.42 and Ser5.46) but partial agonists only interact with Ser5.42. The structures provide an understanding of the pharmacological differences between different ligand classes, which illuminates how GPCRs function and provides a solid foundation for the structure-based design of novel ligands with predictable efficacies. Unfortunately the significant structural changes like those found in opsin structure were not observed, which could be due to the fact that the receptor was thermostabilized with several point mutations towards the inactive state and also due to the absence of stabilizing effect of the G-protein or any its fragment. Nonetheless the analysis of the ligand binding for these agonists shows the basis for efficacy modulation and specificity of binding to b1 receptor rather than b2 or b3. The structures also underlined the importance of several helical interfaces in activation process, like TM4 and TM5 interface for example. Results of this work were published in Nature, January 2011.