Ryanodine receptor dysfunction and the resolution revolution: how Nobel Prize-winning techniques transform cardiovascular research

Dr. Niels Voigt is professor of Molecular Pharmacology at the Institute of Pharmacology at the Georg-August University Göttingen. He graduated and received his MD from the University of Dresden in 2007. He worked as a postdoctoral scientist at the Institute of Pharmacology in Dresden, at the Medical Faculty Mannheim, Heidelberg University and at the Institute of Pharmacology at the Duisburg-Essen University. In 2013, he obtained the specialization in Pharmacology and Toxicology. His field of interest is the regulation of ion channels and calcium signalling in the normal and diseased heart. He has received numerous research awards e.g. the Oskar-Lapp-Research Award (2015) from the German Cardiac Society and the Best of Basic Science Abstracts Award from the American Heart Association (2013). In 2016, he joined the nucleus of the Scientists of Tomorrow of the European Society of Cardiology (ESC). Dr Stephan E. Lehnart obtained a professorship for Translational Cardiology at the University of Göttingen, located in the Department of Cardiology & Pneumology (2008). He studied at the Universities of Freiburg, Basel, and Manchester and graduated as medical doctor (1997). He completed his dissertation in experimental medicine at the University of Freiburg (Dr. med., 1998). He worked as a clinical and research fellow at the Universities of Freiburg and Göttingen, and as a postdoctoral scientist at Johns Hopkins University and Columbia University. In 2008, he was appointed as adjunct professor and later as international member of the Center for Biomedical Imaging and Engineering (BioMET) at the University of Maryland. Since 2017 he coordinates the interdisciplinary Heart Research Center Göttingen as executive director. His scientific interests include allosteric mechanisms of ion channel regulation, intracellular calcium signalling, ryanodine receptors in excitable organ system and drug targeting, and accelerated ageing mechanisms in health and disease. Commentary on ‘Structural basis for gating and activation of RyR1’ by des Georges et al., Cell, 2016.1 Fundamental, unsolved mechanisms of transmembrane ion transport continue to strongly stimulate emerging research and therapeutic efforts. To control vitally important cardiac signalling functions with highest temporal precision, ion channels use dynamic gating structures that define distinct open and closed states. Hence, structural elucidation of the precise mechanisms that facilitate versus disrupt rapid ion conduction through the channel pore is crucial for disease modelling and translation. While the concept of ion channelopathies was traditionally applied to voltage-dependent channel gating in the plasma membrane, recent single-particle cryo-electron microscopy (cryo-EM) studies demonstrated that ryanodine receptors (RyRs) share a similar six-transmembrane segment (S1–S6) pore architecture and, thus, belong to the same ion channel superfamily.2 In addition, recent application of cryo-EM to study the gating mechanisms of native skeletal muscle RyR1 channels has led to astonishing structural details and a new allosteric concept with broad therapeutic implications.1 RyRs are intracellular Ca2+ release channels with essential functions in all excitable cells: cardiac and skeletal muscle contraction, pancreatic beta cell secretion, synaptic transmission in the brain, and vascular muscle relaxation.2 RyR channels are formed by four identical subunits, each regulated by a set of accessory proteins in a massive macromolecular complex, the largest known ion channel.3 In contrast, RyR2 patient mutations or heart failure have been shown to destabilize channel closure, leading to pathologically increased Ca2+ leak from the sarcoplasmic reticulum (SR) Ca2+ store.3,4 While hundreds of genetic variants were associated with different disease syndromes for distinct RyR isoforms, missense mutations of the cardiac RyR2 channel (Figure 1A) were primarily associated with stress-dependent arrhythmias and exceptionally high mortality rates.4 Recent cryo-EM approaches now resolve native RyR channels while bound to their regulatory ligands and proteins at near-atomic resolution, a major breakthrough for mechanistic and therapeutic translation.1 These observations became only possible through decades of rigorous development and application of cryo-EM techniques, pioneered by Joachim Frank, Richard Henderson, and Jacques Dubochet for high-resolution structure determination of functional biomolecules in solution, who received the Nobel Prize in Chemistry 2017. For complex molecular structures such as RyR channels, cryo-EM as compared to X-ray crystallography can be highly advantageous for several reasons. First, flash-freezing of RyR1 channels from skeletal muscle or RyR2 channels from the heart captures the molecular complexes in their close-to-native states, obviating crystallization, and potentially biased molecular conformations due to crystallographic constrains. Second, through a combination of (sample preserving) low-dose electron imaging and single-particle averaging methods complex native structures can be visualized as pioneered for the 40S ribosome subunit by Frank et al.5 Third, the structural heterogeneity inherent to native protein–ligand complexes can be directly analysed by image classification methods. The latter approach was particularly important to map the dynamic range of ligand-dependent RyR channel conformations by cryo-EM discussed below.1,6 In fact, while crystallization has been attempted, the structural variability and massive size (>2.3 MDa) of RyR channels has significantly limited this approach.2 Moreover, only recently has the fundamental barrier of structural atomic resolution (i.e. <5 Å) been overcome through general application of cryo-EM by using direct electron detection cameras.7 In 2015, within remarkable agreement, three independent teams published the first near-atomic structures of the RyR1 channel in a closed state.2 While these structures clearly resolved the RyR pore domain, the inner S5 and S6 helices of the central ion conduction pathway were surprisingly similar to the voltage-gated bacterial NavAb and Ca2+-permeable TRPV1 channels.2 Furthermore, unique features of RyR channels were identified: a cytosolic S2S3 domain insertion and a 30 Å long cytosolical S6 extension continuous with the C-terminal domain. These first high-resolution structures led to new models of RyR gating, where a large cytoplasmic ‘shell’ interacts with the pore through central, ligand-dependent domain interactions (Figure 1A), apparently orchestrated by a far-ranging network of allosteric mechanisms. Moreover, key findings about the mechanisms of allosteric coupling between the RyR cytosolic shell and the channel pore were provided by elegant ligand-binding studies (i.e. ATP, Ca2+, and caffeine) that directly correlate the variability of conformational changes with ligand occupation and channel activation.1,6 An important insight from these studies was that previously unknown modes of RyR gating exist, for example a primed channel state is induced by the synergistic effects of several activating ligands.2 Indeed, RyR channels are thought to bind to a large number of functional ligands including ions (Ca2+, Mg2+), small molecules (caffeine, drug compounds, and nucleotides), and proteins.2 As most RyR protein interactions occur through the extensive cytosolic shell, both inhibitory/stabilizing (i.e. calstabin2/FKBP12.6) and activating (i.e. mAKAP tethering of cAMP-dependent protein kinase A) conformational changes are functionally integrated through the same gating mechanisms.3,8,9 Hence, the emerging picture of RyR channels is that of a molecular machine defined by a large network of interconnected helices with vast (known and unknown) ligand interactions, engaging conformational gating changes through far-ranging allosteric mechanisms.2 How can these structural insights support new therapeutic ideas for the treatment of cardiovascular diseases? Chronic hyperadrenergic and oxidative stress, as observed in human atrial fibrillation and heart failure as well as genetic RyR2 mutations (Figure 1A) can destabilize channel closure and significantly increase diastolic Ca2+ leak from the SR Ca2+ store.10 Importantly, diastolic Ca2+ leak via RyR2 channels is now a commonly accepted mechanism of cardiac arrhythmias, occurring in the atria or ventricles, triggered through delayed after depolarizations (Figure 1B). Therefore, RyR2 represents an important and particularly promising (see below) target for antiarrhythmic therapy.11 Notably, dantrolene, a long established therapeutic agent to treat life-threatening episodes of malignant hyperthermia in RyR1 mutation carriers, was recently shown to reduce SR Ca2+ leak in living cardiomyocytes from patients.12 However, a decisive molecular proof of the presumed therapeutic and RyR channel-specific dantrolene action is overdue. Interestingly, new chemical structures with therapeutic RyR2 activities have emerged either based on clinically used or newly developed drug compounds (Figure 1B). The benzothiazepine derivative JTV-519 (K201) initially developed as a voltage-dependent Ca2+-channel blocker, was shown to stabilize RyR2 closed gating and increased binding of the calstabin2 subunit.13 Importantly, S107, a water soluble derivative of JTV-519 with significantly improved RyR2 selectivity, reduces cardiac SR Ca2+ leak in vivo in RyR2 knock-in mice that reproduce the phenotype of catecholaminergic arrhythmogenic patient mutations.14 Moreover, analogues of the β-adrenoceptor blocker carvedilol were shown to inhibit SR Ca2+ leak, the latter attributed to Ca2+ store-overload.15 Finally, clinically established antiarrhythmic drugs such as flecainide and ranolazine were shown to reduce intracellular Ca2+ leak and arrhythmogenic Ca2+ waves.16,17 However, the proposal of a direct flecainide effect on RyR2 channels was subsequently challenged,18 further emphasizing the need for improved structure–function analysis. Here, without doubt, the Nobel Prize to Jacques Dubouchet, Joachim Frank, and Richard Henderson recognizes not only revolutionary new imaging techniques, but opens avenues to unravel how, if at all, high-affinity drug compounds may stabilize RyR2 channels allosterically as compared to physiologically disruptive pore blocking effects. Therefore, direct structure–function analysis of ligand-bound RyR2 channels will strongly support the identification and refinement of lead compounds for treatment of cardiac arrhythmias.1,6 Notably, through the ground-breaking work of Andrew Marks, Wayne Hendrickson, and Joachim Frank the structure of the RyR1 channels and previously unknown modes of ligand interactions are emerging as a significant area of translational research.2 The fact that many RyR1 mutations are located in N-terminal hot spots suggested that the shell movements may become biased towards abnormally increased pore opening and uncontrolled Ca2+ leak, as observed for certain anaesthetic drugs and/or stress regulatory ligands in patients.2 Moreover, RyR2 mutations associated with catecholaminergic polymorphic ventricular tachycardia (CPVT) cluster in similar hot spots (Figure 1A), for example at N-terminal interfaces between neighbouring channel protomers.2,4,8 Therefore, the recent cryo-EM structures and new channel models represent critical progress towards our understanding about disease-causative mechanisms of channel dysregulation.1,2 Together with a fundamentally novel understanding of the allosteric ligand-binding and gating mechanisms, these insights provide lasting opportunities for targeted drug development, ultimately to treat a vast spectrum of acquired and genetic RyR disease syndromes, both in the heart and other organs. (A) Hot spot clustering of disease-causing RyR2 channel mutations indicated by coloured spheres. Each sphere represents an individual missense mutation identified in human carriers (http://www.uniprot.org/uniprot/ Q92736). A high-resolution side view through the pore region of the RyR1 channel (grey) is used as template to visualize different classes of hot spot locations: (i) the extreme N-terminal domain (blue spheres); (ii) the bridging solenoid (green spheres); (iii) the core solenoid (purple spheres); (iv) the transmembrane domain (orange spheres), and (v) the C-terminal domain (red spheres). Additionally, many RyR2 mutations affect domain–domain interfaces, for example between two neighbouring promoters in the channel tetramer. (B) Schematic illustration of arrhythmogenic RyR2 Ca2+ leak in diastole and therapeutic strategies. Pathologically increased RyR2 opening, i.e. due to CPVT missense mutations and increased stress, represents an established mechanism of pathogenic SR Ca2+ leak. In diastole, increased cytosolic Ca2+ is extruded via electrogenic Na+/Ca2+ exchange (NCX) at the cost of 3 Na+ ions entering the cell per Ca2+ ion. In cardiomyocytes, this generates a depolarizing transient inward current (ITI) leading to delayed afterdepolarizations (DADs) and Ca2+ triggered arrhythmias. A number of experimental or repurposed pharmacological substances (italic) were shown to either stabilize RyR2 channel closure (S107, RyCals) or to block ion conduction (carvedilol analogues, dantrolene, flecainide, and ranolazine), altogether resulting in decreased arrhythmogenic Ca2+ leak. However, the exact molecular action of specific compounds and their dynamic RyR2 interactions during gating are currently unknown. A, activation module; P, pore domain; see also main text and Des Georges et al., 2016.1 The authors would like to thank Ran Zalk for critical comments and provision of Figure 1A. Conflict of interest: none declared. This work was supported by grants from Deutsche Forschungsgemeinschaft to NV (VO 1568/3-1, IRTG1816, and SFB1002 project A13), to S.E.L (SFB1002 project A09 and SFB1190 project P03), by the Else-Kröner-Fresenius Foundation to N.V. (EKFS 2016_A20), and by the German Center for Cardiovascular Research to N.V. and S.E.L. (DZHK GOE MD3). S.E.L. is a principal investigator of DZHK (German Center for Cardiovascular Research).