Reviews of 411 Physiology yrtsimehcoiB dna ygolocamrahP Editors M.P. Blaustein, Baltimore • O. Creutzfeldt, G6ttingen H. Grunicke, Innsbruck • E. Habermann, GieBen H. Neurath, Seattle • S. Numa, Kyoto D. Pette, Konstanz ' B. Sakmann, Heidelberg M. Schweiger, Innsbruck • U. Trendelenburg, Wtirzburg K.J. Ullrich, Frankfurt/M • E.M. Wright, Los Angeles With 21 Figures and 9 Tables galreV-regnirpS Berlin Heidelberg New York London Paris Tokyo Hong Kong ISBN 3-540-51693-X galreV-regnirpS Berlin Heidelberg New York ISBN 0-387-51693-X galreV-regnirpS New York Berlin Heidelberg Library of Congress-Catalog-Card Number 74-3674 This work is subject to copyright. All rights are reserved, whether the whole or part of the ma- terial is concerned, specifically the rights of translation, reprinting, re-use of illustrations, recitation, broadcasting, reproduction on microfilms or in other ways, and storage in data banks. Duplication of this publication or parts thereof is only permitted under the provisions of the German Copyright Law of September 9, 1965, in its version of June 24, 1985, and a copyright fee must always be paid. Violations fall under the prosecution act of the German Copyright Law. © Springer-Verlag Berlin Heidelberg 1990 Printed in the United States of America The use of registered names, trademarks, et~ in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Product Liability: The publisher can give no guarantee for information about drug dosage and application thereof contained in this book. In every individual case the respective user must check its accuracy by consulting other pharmaceutical literature. Typesetting: K+V Fotosatz GmbH, Beerfelden 2127/3130-543210 - Printed on acid-free paper Contents Molecular Properties of Calcium Channels By H. GLOSSMANN and J. ,GINSSEIRTS Innsbruck, Austria With 21 Figures and 6 Tables ................... Properties and Regulation of Calcium Channels in Muscle Ceils By D. ,REZLEP .S ,REZLEP Homburg/Saar, FRG, and E T. ,DLANODCM Halifax, Nova Scotia, Canada ...................................... 107 Pharmacological Modulation of Voltage-Dependent Calcium Channels in Intact Cells By H. ,GIZROP Bern, Switzerland With 3 Tables ................................. 209 Subject Index .................................... 263 Indexed in Current Contents Rev. Physiol. Biochem. Pharmacol., Vol. 114 © Springer-Verlag 1990 Molecular Properties of Calcium Channels* HARTMUT GLOSSMANN and JORG STRIESSNIG 1 Contents 1 Introduction .............................................................. 2 2 Drugs and Toxins as Molecular Probes for Calcium Channels ................... 3 2.1 The Voltage-Dependent Calcium Channel in Comparison with Other Ion Channels ............................................................ 3 2.2 Drugs - Specific Probes for L-Type Channels .......................... 5 2.3 Toxins ............................................................. 71 2.3.1 Toxins as Probes - A General Comment .............................. 71 2.3.2 Toxins with Claimed but Unproven Action on Calcium Channels .......... 81 2.3.3 Toxins Which Are Putative Candidates for Structural Research ............ 21 2.3.40mega-Conotoxins - N-Type Channel Probes .......................... 23 2.4 Endogenous Ligands ................................................ 34 2.4.1 General Remarks .................................................... 34 2.4.2 Antralin ........................................................... 34 2.4.3 Endothelin ......................................................... 53 3 Probing the Calcium Channel with Target Size Analysis ........................ 36 4 Identification of Calcium Channel-Associated Drug Receptors in Membranes of Excitable Tissues by Photoaffinity Labelling .................................. 38 5 Calcium Channel Structure (L-Type Channels) ................................. 41 5.1 Purification of Calcium Channels ..................................... 14 5.1.1 General Remarks .................................................... 14 5.1.2 Hydrodynamic Properties of Solubilized Calcium Channels ............... 45 5.1.3 Subunit Composition of the Calcium Channel in Skeletal Muscle ......... 46 5.1.4 Subunit Properties of the Isolated Skeletal Muscle Calcium Channel ....... 48 5.1.5 Regulatory Domains ................................................. 48 5.1.6 Hydrophobic Labelling ............................................... 52 5.1.7 Glycosylation ....................................................... 52 5.1.8 The AlpharSubunit is Sensitive to Proteolysis .......................... 35 5.2 Interaction of Purified Skeletal Muscle Calcium Channels with Calcium Channel Drugs ..................................................... 55 5.3 A Structural Model for the Skeletal Muscle Calcium Channel Complex .... 56 5.4 Evidence for L-Type Channel Subunits in Other Tissues .................. 57 5.5 AntibodiesA gainst Calcium Channel Subunits Modulate Channel Function 60 * This work is dedicated to Professor Emeritus Heribert Konzett 1 Institut fiir Biochemische Pharmakologie der Leopold-Franzens-Universittit Innsbruck, Peter- Mayr-Strage ,1 A-6020 Innsbruck, Austria 2 H. Glossmann and J. Striessnig 6 Structural Features of Omega-Conotoxin GVIA-Sensitive Calcium Channels ...... 6t 7 The Calcium Channel Structure and Excitation-Contraction Coupling (EEC) ..... 64 7.1 Introductory Remarks ............................................... 64 7.2 ECC in Skeletal Muscle - Effects of Cations and Drugs ................ 64 7.3 The Triad Junction and the Calcium Release Channels ................... 66 7.4 The Foot Structure Is an Oligomer of Calcium Release Channels .......... 69 7.5 ECC in Skeletal Muscle - Functional Association of Two Otigomeric Calcium Channels Across Membranes ................................. 71 7.6 Pathology of the Triad Junction and Restoration of ECC in the mdg/mdg Mouse by Alphal-Subunit cDNA ...................................... 74 8 Molecular Cloning, Models and Expression ................................... 76 9 Puture Prospects .......................................................... 86 References .................................................................. 89 Note Added in Proof ........................................................ 104 Abbreviations ,xamB maximal density of binding sites; CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-l- propanesulphonate; DHP, dihydropyridine; ECC, excitation-contraction coupling; G-proteins, GTP-binding proteins; kb, kilobase; ld)a, kilodalton; PAGE, polyacrylamide gel electrophore- sis; Irl?X, pertussis toxin; Kd, dissociation constant; k_l, dissociation rate constant; k+l, asso- ciation rate constant; SDS, sodium dodecyl sulphate; T-tubule, transverse tubule; SR, sarco- plasmic reticulum Drugs: BAY K 8644, methyl-l,4-dihydro-2,6-dimethyl-3-nitro-4-(2-trifluoromethylphenyl)-pyri- dine-5-carboxylate; DPI 201-106, 4-3'-(4"-benzhydril-l"-piperazinyl)-2'-hydroxy-propoxy-lH- indole-2-carbonitrile - BDF 8784 carries a methyl group instead of the CN group; IN-methyl- aHII~49888 ((-)-5-[(3 -azidophenethyl)[N-methyl.3H]methylamino]-2-(3,4,5-trimethoxy-phe - nyl)-2-isopropylvalero nitrile; PN200-110, isopropyl-4-(2,1,3-benzoxadiazol-4-yl)-l,4-dihydro- 2,6-dimethyl-5-methoxy-carbonyl-pyridine-3-carboxylate; 202-791, isopropyl-4-(2,1,3-benzoxa- diazol-4-yl)- 1,4-dihydro-2,6-dimethyl-5-nitro-3-pyridine carboxylate 1 Introduction Our review deals with the molecular properties of voltage-dependent calcium channels. Compared with the voltage-dependent sodium or ligand-activated ion channels, structural information on channels selective for calcium is limit- ed. It is virtually nonexistent on T-type channels, scarce on N-type channels and, with respect to L-type channels, restricted to the skeletal muscle. L-type channels have distinct but allosterically coupled receptor sites for drugs (e.g. the i,4 dihydropyridines such as nifedipine, the phenylalkylamines such as verapamil and the benzothiazepines such as (+)-cis-diltiazem). These drugs, especially those of the 1,4 dihydropyridine class, were essential for the purifi- cation (and cloning) of the channel proteins. In contrast to the recently char- acterised archetype of cation-selective channels, the (A current) ÷ K channel raluceloM seitreporP of muiclaC slennahC 3 (where first the gene coding for the Shaker mutation was mapped in Droso- phila, cDNA clones were isolated, and the protein later was expressed func- tionally in a heterologous system) isolation of the L-type channel followed a conventional path - in analogy to the sodium channel. It was perhaps this analogy which misled researchers initially to a large glycoprotein in purified preparations from skeletal muscle (drug receptors are here abundant com- pared with other tissues) as the pore-forming calcium channel "alpha'sub- unit. This glycoprotein released a 25- to 35-kDa set of glycopeptides upon re- duction of disulphide bonds. The rat brain sodium channel (heavily glycosy- lated) alpha-subunit is linked to (glycosylated) fl2-subunits via disulphide bonds. Quite unexpectedly, the purified calcium channel turned out to be a complex of (four to) five subunits, where the pore-forming drug receptor-car- rying alpha-subunit was neither disulphide linked nor (heavily) glycosylatedl Almost simultaneously with the skeletal muscle "Ca +2 antagonist drug re- ceptor" the ryanodine-sensitive calcium release channel from sarcoplasmic reticulum (SR) was isolated, characterised and reconstituted. This channel forms the foot structure, bridging the gap between the transverse tubule mem- brane and the SR membrane. Ironically, it was once believed that the feet had a solely structural role. Quite similarly, the Ca2+-antagonist receptors in skel- etal muscle were long viewed with suspicion as functionally silent drug-bind- ing sites. Both structures are now regarded as essential constituents of a novel transmembrane communication pathway. For that reason, and despite the mysteries which still surround the process of excitation-contraction coupling, the skeletal muscle is a main theme in this review. First we present the tools which are proven or suggested to be molecular probes for calcium channels; we then mention target size analysis and photoaffinity labelling, discuss the L-type channel properties (mainly but not exclusively) from skeletal muscle in great detail and provide a critical overview of N-type channels. In the final chapter we discuss the (deduced) primary structures, models and expression. 2 Drugs and Toxins as Molecular Probes for Calcium Channels 2.1 The Voltage-Dependent Calcium Channel in Comparison with Other Ion Channels The nicotinic acetylcholine receptor was the first ligand-activated ion channel to be characterised and even localised with radioactive toxins (see Waser 1986). Small protein toxins such as the alpha-neurotoxins from various Naja species and from Bungarus multinctus were subsequently essential in the puri- fication of the channel and led to the elucidation of the complete amino acid sequence of its four subunits (Conti-Tronconi and Raftery 1982). Toxins are 4 .H nnamssolG dna .J ginsseirtS also helpful in the differentiation, biochemical characterisation and purifica- tion of potassium channels by conventional schemes. Alternative approaches with methods providedb y molecular biology and exemplified by the analysis of Shaker mutants (potassium channels from Drosophila) are equally success- ful (Papazian et al. 1987; Tempel et al. 1987; Schwarz et al. 1988; Timpe et al. 1988). The key role of the voltage-dependent sodium channel in information transfer has also made it a prime target for potent neurotoxins. These toxins have been classified in either six (Lazdunski et al. 1986a; Barchi 1988) or four categories (Catterall 1986) on the basis of binding sites and/or physiological effects. Polypeptide toxins which can be iodinated with iodine 125, such as the Tityus gamma toxin (Lazdunski et al. 1986a), or small, naturally occur- ring nonprotein compounds such as tetrodotoxin and saxitoxin (which can be labelled with tritium) are extremely useful probes. Theyc an be employed for the characterisation of receptor sites in membranes from electrically excitable ceils and (after chemical modification to yield affinity or photoaffinity pro- bes) to identitfhye receptor-carrying polypeptides by irreversible labelling, to follow solubilization and purification and, finally, to probe for alteration of the conductance behaviour in reconstituted or even in mRNA expression sys- tems. Identification of the toxin binding domains within the primary struc- ture and on crystallized channel proteins are pursued. Still another aspect of the toxins is the discrimination of subtypes within the sodiumc hannel family which can complement the rapidly increasing member of channel structures deduced by molecular biology techniques. In contrast to voltage-dependent sodium or potassium channels and the acetylcholine-activated channel, naturally occurring toxins have not played any role in the characterisation of L-type calcium channels. Instead drugs originally synthesized as therapeutics are the keys for structural research, as they still are for differentiation of different subtypes within the calcium chan- nel family. In contrast, some omega-conotoxins, e.g. GVIA and MVIA, are useful probes for the neuronal (N-type) calcium channel. The apparent (but perhaps not complete)n eglect of the (L-type) calcium channel as a target in the everlasting struggle between organisms is not well understood. It may re- late to the fact that the calcium signal has different inputs, e.g. from the extra- cellular space, from intracellular stores, or by changing the sensitivity of in- tracellular calcium-binding proteins. A blockade of the channel could be more or less compensated for by other mechanisms. Transient initial Ca ÷2 signals are often from internal stores. Only prolonged Ca ÷2 signals may re- quire influx (Putney 1987), and there is a tissue- and species-specific variation even foro ne neurotransmitter to utilize these different sources. Noradrenaline contracts the rat spleen by activation of alphal-adrenoceptors (alpha~B type) even when the L-type Ca ÷2 channels are blocked by nifedipine. In the vas deferens (alphalA type) from the same species there is almost complete inhi- raluceloM seitreporP of muiclaC slennahC 5 bition of contraction by the same concentration of this L-type Ca +2 channel- specific blocker (Han et al. 1987). Other aspects are that neuronal L-type channels (in the majority of systems investigated) have no important role in neurotransmitter release (see review by Miller 1987), and that the overwhelm- ing majority of L-type channels in all vertebrates so far investigated reside deeply hidden in skeletal muscle transverse tubules and have a very specialised function, where calcium influx is not required to elicit contraction. However, contractions of invertebrate skeletal muscle are highly dependent on extracel- lular calcium, and the membrane action potential is generated by voltage-de- pendent calcium channels (Fatt and Ginsborg 1958) first recorded in crab leg fibres by Fatt and Katz (1953). Blockade of these channels may be an attrac- tive method of paralysing the prey. Ca ÷2 influx induced by mechanisms simi- lar to those seen with many of the sodium-channel toxins (e.g. by persistent activation, enhancing activation or slowing inactivation) may be another principle of poisoning. Feedback mechanisms, Ca ÷2 pump activity, the Na+/Ca +z exchanger and intracellular storage may, at least in part, protect against the disastrous metabolic consequences of intracellular Ca ÷2 excess. L-type channels are opened by depolarisation. Receptors (e.g. the alphalA- adrenoceptor) - most likely directly via GTP-binding proteins (Yatani et al. 1987; Brown and Birnbaumer 1988) - and second messenger systems (Reuter 1983; Hofmann et al. 1987) can modulate channel activity. Therefore, toxins with alleged activator actions at the L-type channels may act indirectly via different mechanisms, e.g. second messengers, depolarisation, selective pore formation. L-type channel selective agents (e.g. 1,4 DHPs, (+)-cis-diltiazem or verapamil) often block venom-induced smooth muscle contraction, posi- tive inotropic effects and hormone, mediator or neurotransmitter release, to name some of the bioassay methods. Even if it can be shown that extracellular calcium is required, this is by no means proof that the toxic principle acts di- rectly on the calcium channel, as the cation could be crucial for binding only. 2.2 Drugs - Specific Probes for L-Type Channels Tools to characterise, isolate and purify L-type calcium channels have been found among low-molecular-weight synthetic organic compounds, termed "Ca ÷2 antagonists" by Fleckenstein (see e.g. Fleckenstein 1983; Godfraind et al. 1986; Janis et al. 1987; Triggle and Janis 1987). They can be classified according to criteria derived from physiology, phar- macology or therapeutics. For the present purposes, a chemical classification is appropriate. We divide the compounds (for typical structures consult Fig. )1 into five classes, named according to their basic structure(s). This division also reflects the current view (CatteraU et al. 1988; Glossmann and Striessnig 1988a,b; Janis et aL 1987) that each class may recognise a distinct binding do- H. Glossmann and J. Striessnig ~,.NO~ c,~c,,o~-.~.-~-oc~ 0 H L, o ~.-oc-~, - ,'~r-CO(C..).oc.. ~HC =HC~,,,.N.~3HC • CIH~-C~HN3 H H NITRENDIPINE NIMODIPINE [ ~CF~ .~125 I O O c._c. o_c 3 -- C 3HCO "~HC ~.N ~- 3HC" O CH~N~"CH: ) H H IODIPINE NIFEDIPINE • H F'~'--N-. ° o j~. o, . c.,o o~.~-t ";HC / "~N ~HC" ~ - - =:HC"~N~"3HC H H AZlDOPINE PN 200-110 3~c 3 CH3--C--CH I 3F C -0 0 I ,, 0 HN 11 n I H I • O,N-ff~COOCH; 3HC - --2HC --0 C--~ ~C-N-2)2HC(--O-C 2HC(--HC-- )~'S- 3HC CH~#,,"~CH3 ,-:'~-HC N H-~., 3 0 H a SADOPINE BAY K 8644 Fig. 1 a- e. L-type channel drugs. a 1,4 DHP structures, including [12sI]- or [3sS]-labelled ligands (sadopine), the arylazide photoaffinity ligand azidopine and an agonistic 1,4 DHP (Bay K 8644). Note that in the text PN200-1 l0 is sometimes referred to as isradipine. b Drugs which bind to the phenylalkylamine-selective domain. (-)-Desmethoxyverapamil and LU 49888 (a reversible and photoaffinity ligand) are employed for structural research as (opti- caily pure) tritium-labelled compounds. e Drugs claimed to bind to the benzothiazepine-setective domain are shown (trans-diclofurime, MDL 12330A, Fostedil) together with the diltiazem structure. Of the four diltiazem diastereo- mers only (+)-cis-diltiazem binds with high affinity to L-type channels, and it is the tritium-la- belled standard radioligand for receptor domain "3" (see Fig. 2) Molecular Properties of Calcium Channels 7 ~ aHC,~HC - CH,,- NCH,~- CHCH,~OCH,,CH j CH3 HC~' ~HC '~':~..~ ~ -. ~ CH 3 O=HC =:HCO )-+( VERAPAMIL BEPRIDIL H3C--O . . . . N= "CH C =H H=C--O /x C3H ~HC )-( DESMETHOXYVERAPAMIL LU 49888 ococ, ~[~ v -N~,. O ~N /~ O I /CH3 \~( ~ ~-c,.-p-oc.,. HC~,HC'HC 3 FOSTEDIL DILTIAZEM ~ OCH 3 CI N OCH2CH2N (C2H5) 2 trans - DICLOFURIME MDL 12,330 A c main on the alphal-subunit of the L-type channel (see below), although all drug-receptor domains interact with each other in a heterotropic allosteric manner in in vitro ligand binding experiments. A schematic view of the ob- served interactions is given in Fig. 2. Only within the 1,4 DHP class do com- pounds exist (e.g. S-(-)-Bay K 8644, (S)-(+)-202-791, (-)-Bay F6653) which activate L-type Ca +2 channels (Bechem et al. 1988). These "calcium channel agonists" have not been useful for direct structural studies (probably because