This is a preparation of choice in studies of protein structure and membrane organization [S], conformational transitions coupled to ion translocation [6] and identification of sites for
Trang 3New Comprehensive Biochemistry
Trang 4Molecular Aspects of Transport
Proteins
Editor
Department of Biochemistry, University of Nijmegen,
6500 H B Nijmegen, The Netherlands
1992 ELSEVIER Amsterdam London New York Tokyo
Trang 5Elsevier Science Publishers B.V
No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the Publisher, Elsevier Science Publishers B.V., Copyright &
N o responsibility is assumed by the publisher for any injury and/or damage to persons or
tion of any methods, products, instructions or ideas contained in the material herein Because
of the rapid advances in the medical sciences, the publisher recommends that independent
Special regulations f o r readers in the USA - This publication has been registered with the Copyright Clearance Center Inc (CCC) Salem, Massachusetts Information can be obtained from the CCC about conditions under which photocopies of parts of this publication may be made in the USA All other copyright questions, including photocopying outside of the USA, should be referred to the Publisher
Printed o n acid-free paper
Printed in The Netherlands
Trang 6Preface
The first two volumes in the series New Comprehensive Biochemistry appeared in
1981 Volume 1 dealt with membrane structure and Volume 2 with membrane trans- port The editors of the last volume (the present editor being one of them) tried to provide a n overview of the state of the art of the research in that field Most of the chapters dealt with kinetic approaches aiming to understand the mechanism of the various types of transport of ions and metabolites across biological membranes Although these methods have not lost their significance, the development of mole- cular biological techniques and their application in this field has given to the area of membrane transport such a new dimension that the appearance of a volume in the series New Comprehensive Biochemistry devoted to molecular aspects of membrane proteins is warranted
During the last decade hundreds of primary structures of membrane proteins have been published and each month several new sequences of transport proteins appear in the data banks From these sequences global models for the structure of membrane proteins can be made using several type of algorithms These models are very useful for a partial understanding of the structure of these proteins and may help us with understanding part of the mechanism of action They d o not, however, provide us with complete answers of how these pumps, carriers and channels actually function The combination of biochemical (site-specific reagents), molecular biological (site- directed mutagenesis) and genetic approaches of which this volume gives numerous examples in combination with such biophysical techniques as X-ray analysis and
N M R will eventually lead to a complete elucidation of the mechanism of action of these transport proteins
It is clearly impossible to give a comprehensive overview of this rapidly expanding field I have chosen a few experts in their field to discuss one (class of) transport protein(s) in detail In the first five chapters pumps involved in primary active transport are discussed These proteins use direct chemical energy, mostly ATP, to drive transport The next three chapters describe carriers which either transport metabolites passively or by secondary active transport In the last three chapters channels are described which allow selective passive transport of particular ions The progress in the latter field would be unthinkable without the development of the patch clamp technique The combination of this technique with molecular biological approaches has yielded very detailed information of the structure-function relation- ship of these channels
Despite the limitation in the choice of membrane proteins, I hope that this volume will be useful for teachers, students and investigators in this field Although only a limited number of transport proteins is discussed in this volume in detail, the
Trang 8List of contributors
Stephen A Baldwin,
Departments of Biochemistry and Chemistry, and Protein and Molecular Biology, Royal Free Hospital School of Medicine (University of London), London NW3 2PF, U.K
Rebecca M Brawley,
Department of Pharmacology, Northwestern University Medical School, Chicago, TL
60611, U S A
Chan Fong Chang,
Department of Pharmacology, Northwestern University Medical School, Chicago, I L
6061 I , U.S.A
Jan Joep H.H.M De Pont,
Department of Biochemistry, University of Nijmegen, 6.500 H B Nijmegen, The Neth- erlands
Peter Leth Jerrgensen,
Biomembrane Research Centre, August Krogh Institute, University of Copenhagen,
Trang 9Tom J.F Van Uem,
Department of Biochemistry, University of Nijmegen, 6500 H B Nijmegen, The Neth- erlands
Trang 10Contents
P r ~ f u c e
List of contributors
Chapter 1 Nu K.A TPase structure and transport mechanism Peter Leth Jmgensen
1 Introduction
1.1 The Na, K-pump
1.2 Recent review articles on Na K-ATPase structure and function
Structure of Na, K-ATPase
2.1 Purified membrane-bound and soluble Na.K-ATPase
2.1.1 Enzymatic properties
2.1.2 Electron microscopy and crystal analysis
2.1.3 Three-dimensional models
2.2 Cytoskeletal associations
2.3 Proteolytic dissection of membrane-bound Na, K-ATPase
2.4 Membrane organization of the c1 subunit
2.5 Structure of the fl subunit of Na, K-ATPase
Nucleotide binding and phosphorylation
3.1 The nucleotide binding domain in the alp1 unit
3.1.1 Comparison with the nucleotide binding sites in adenylate kinase
3.1.2 Selective chemical labelling with ATP analogues
Conformations of the nucleotide binding area
The phosphorylation site, high- and low-energy phosphoforms, ElP-E2P 2 3 3.2 3.3 4.1 4.2 Isolation of the cation occlusion and transport path after tryptic 4.3 Transport stoichiometry and net charge of N a + and K + complexes with Na, K-ATPase
Structural transitions in the protein related to energy transformation and Na, K-transport
5.1 5.2 5.3 5.4 5.5 5.6 4 Cation binding and occlusion
Capacity for binding and occlusion of Na' or K + ( R b + )
digestion
5 Conformation dependent proteolytic cleavage of Na, K-ATPase
Tryptophan fluorescence and secondary structure changes
Effect of C3 cleavage on EIP-E2P transition and cation exchange
Cleaved derivatives; cleavage of bond 2 and the regulatory function of the N-terminus
Mutagenesis in yeast H-ATPase and Ca-ATPase from sarcoplasmic reticulum
Coupling to ion translocation
References
V
vii
1
i
1
2
2
2
3
3
5
6
7
7
10
11
12
12
12
13
13
15
16
17
17
18
18
19
20
20
21
22
23
Trang 11X
Chapter 2 Structure and function of gastric H K.ATPase
Tom J.F Van Uem and Jan Joep H.H.M De Pont
1 Introduction
2 Tissue and cell distribution
3 Structure
3.1 The catalytic a subunit
3.2 The fl subunit
3.3 Molecular organization
3.4 Conformations of H, K-ATPase
4 Kinetics of H, K-ATPase
4.1 Overall reaction
4.2 Phosphorylation from ATP
4.3 Characteristics of ATP hydrolysis
4.4 Hydrolysis of p-nitrophenylphosphate
4.5 Phosphorylation from inorganic phosphate
5 Transport by H, K-ATPase
5.1 The H+-ATP stoichiometry
5.2 Ion selectivity
5.3 Electrogenicity of ion transport
6 Lipid dependency of H, K-ATPase
7 Solubilization and reconstitution
8 Inhibitors of H, K-ATPase
9 Conclusions References
Chapter 3 The Ca2+ transport ATPases of sarco(endo)plasmic reticulum and plasma membranes Anthony Martonosi
1 Introduction
2 The classification of Ca2'-ATPase isoenzymes
2.1 The Ca2+ transport ATPases of sarco(endo)plasmic reticulum (SERCA) 2.1.1 SERCAl
2.1.2 SERCA2
2.1.3 SERCA3
2.1.4 SERCA-type Ca2+ -ATPases from non-mammalian cells (SERCAMED)
The plasma membrane Ca2+ transport ATPases (PMCA)
2.2.1 rPMCA1 and rPMCA2
2.2.2 rPMCA3
2.2.3 rPMCA4
2.2.4 hPMCA
The deduced amino acid sequences of the fast-twitch and slow-twitch isoforms of the sarcoplasmic reticulum Ca2'-ATPase
The predicted topology of the Ca2+-ATPases
4.1 The Ca*'-ATPase of the sarcoplasmic reticulum
4.1.1 The cytoplasmic headpiece
2.2
3
4
27
27
28
28
28
31
34
34
36
36
38
39
40
41
42
42
42
43
44
45
46
49
49
57
51
58
58
58
58
59
59
59
62
62
62
62
62
64
64
65
Trang 124.1.1.1 The phosphorylation and nucleotide binding domains
4.1.1.2 The transduction or B domain
4.1.1.3 The hinge domain
4.1.1.4 The stalk region
4.1.2 The transmembrane domain
4.2 The predicted domains of the plasma membrane Ca2’.- ATPase
Reconstruction of Ca”-ATPase structure by electron microscopy
5.1 The vanadate-induced E2-type crystals
5.1.1 Image reconstruction in three dimensions from negatively stained and frozen hydrated crystals
5.2 Crystallization of Ca’+-ATPase by Ca” and lanthanides in the E l state 5.3 Crystallization of Ca2 ’ -ATPase in detergent-solubilized sarcoplasmic reticulum
X-ray and neutron diffraction analysis of the Ca2+-ATPase of sarcoplasmic reticulum
Site specific mutagenesis of sarcoplasmic reticulum Ca2’-ATPase
7.1 The search for the Ca2+ binding site
7.1.1 Mutation of amino acids in the stalk sector
7.1.2 The probable location of Ca2+ binding sites in the transmembrane domain
Mutations in the putative catalytic site
7.2 1 Mutations around Asp35 1
7.2.2 The mutations around Lys5 I S
7.2.3 The role of sequences 601-604 in ATP binding and Ca’ ‘ transport
7.2.4 Mutations in the R616-K629 region of the Ca’ -ATPase (Thr625, Gly626, Asp627)
7.2.5 Mutations in the 701-707 region
7.2.6 Mutations of Lys712
7.2.7 The structure of the ATP binding site
7.3 The /3 strand sector Conformational change mutants
7.4 The transmembrane segments of the Ca2 ’ -ATPase
In situ proteolysis of Ca’&-ATPase
8.1 Hydrolysis of Ca2+-ATPase by trypsin
8.1.1 The T I cleavage
8.1.2 The T2 cleavage
8.1.3 The cleavage of Ca’ ‘-ATPase by trypsin at the T3 and T4 sites The effect of other proteolytic enzymes on the Ca’.‘-ATPase
8.2 I Chymotrypsin
8.2.2 Thermolysin
8.2.3 Staphylococcal V8 protease
8.3 Vanadate-catalyzed photocleavage of the Ca’ ’~ -ATPase
Monoclonal and polyclonal anti-ATPase antibodies
9.1 Antibodies reacting with the N- and C-terminal regions of the C a 2 + -ATPase
9.2 Distribution of epitopes in the cytoplasmic domain of Ca”-ATPase
9.3 Antibodies reacting with the putative luminal domain of the Ca’- - ATPase
10 Covalent modification of side-chain groups in the Ca’+-ATPase
10.1 Sulfhydryl groups
5
6
7
7.2
8
8.2
9
65
67
67
67
68
68
70
70
71
73
73
77
78
78
78
79
79
79
80
80
81
81
81
81
82
83
84
84
84
85
86
87
87
87
87
87
88
89
89
90
91
91
Trang 13xii
10.1 1 Identification of cysteine residues that react with N-ethylmaleimide
10.1.2 The reaction of iodoacetamide and its N-substituted derivatives
with the Ca2+-ATPase
10.1.2.1 Iodoacetamide (IAA) and 5-(2-acetamidoethyl)aminonaphthalene- 1 -sulfonate (IAEDANS)
10.1.2.2 The reaction of 6-(iodoacetamido)fluorescein (IAF) with the Ca2+ -ATPase
10.1.3 Modification of Ca2+-ATPase with 7-chloro-4-nitrobenzo-2-oxa- 1, 3-diazole (NBD-C1)
10.1.4 The disulfide of 3’(2’)-O-biotinyl-thioinosine triphosphate (biotin- (MalNEt)
6 yl-S -1TP2)
10.2 Modification of lysine residues
10.3 Modification of arginine residues
10.4 Modification of histidine
10.5 Modification of carboxyl groups
10.5.1 The reaction of Ca2+-ATPase with dicyclohexylcarbodiimide
10.5.2 The reaction of Ca2+ -ATPase with N-cyclohexyl-N’-(4-dimethyl- amino-a-naphthyl) carbodiimide (NCD-4)
10.5.3 Reaction of Ca2+-ATPase with the carbodiimide derivative of ATP
11 Spatial relationships between functional sites in the sarcoplasmic reticulum Ca2+-ATPase
1 1.1 Intramolecular distances determined by fluorescence energy transfer
11.1.1 The location of the high-affinity Ca2+ binding site
11.1.2 The ATP binding site
1 1.1.3 The use of IAEDANS as reference point for distance measure- ments
11.2 Thermal fluctuations in the structure of the Ca*+-ATPase
References
Chapter 4 The Neurospora crassa plasma membrane H + -ATPase Gene A Scarborough
1 Introduction
2 Structural features of the H+-ATPase molecule
2.1 H+-ATPase conformational changes
2.2 The purified ATPase preparation
2.3 Subunit composition of the H+-ATPase
2.4 The minimum functional unit
2.5 Primary structure of the H+-ATPase
2.6 Secondary structure of the H + -ATPase
2.7 Protein chemistry of the H+-ATPase
2.8 Chemical state of the H + -ATPase cysteines
2.9 Transmembrane topography of the H+-ATPase
2.10 A first-generation model for the tertiary structure of the H+-ATPase
3 The molecular mechanism of transport
References
91
92
92
92
92
93
93
94
95
96
96
97
97
98
98
100
101
102
103
105
117
117
118
118
119
119
120
121
121
122
122
123
124
129
131
Trang 14Chapter 5 The Enzymes 11 of the phosphoenolpyruvate-dependent carbohy-
drate transport systems
J.S Lolkema and G.T Robillard
1 Introduction
1 1 PTS carbohydrate specificity
1.2 PTS components
1.3 PTS nomenclature
2 Enzyme I1 structure
2.1 Sequence homology
2.2 Domain structure
2.3 Domain function
2.3.1 The A domain
2.3.2 The B domain
2.3.3 The A and B domains of E coli HIMa"
2.3.4 The C domain
2.4 Domain interactions
2.4.1 Association state of E-I1
2.4.1.1, E-II'"
2.4.1.2 E-IIGLC
2.4.2 Kinetics of domain interaction
3 Binding studies
3.1 General considerations
3.2 3.3 Orientation of the binding site
3.4 Kinetics of binding
The coupling between transport and phosphorylation
4.1 General considerations
4.2 Phosphorylation of free cytoplasmic carbohydrates
4.3 Facilitated diffusion catalyzed by E-I1
4.3.1 Diffusion in uptake studies
4.3.2 Diffusion in efflux studies
4.3.3 Regulation of efflux
4.4 Coupling in vectorial phosphorylation
Steady-state kinetics of carbohydrate phosphorylation
5.1 General considerations
5.2 The R sphaeroides IIF'" model
5.3 The E coli IIMtL model
References
Equilibrium binding to E-I1
4 5 Chapter 6 Mechanisms of active and passive transport in a family of homologous sugar transporters found in both prokaryotes and eukaryotes Stephen A Baldwin
1 Introduction
2 The kinetics of sugar transport in mammalian cells
2.1 Substrate specificity
2.2 Specific inhibitors of transport
2.3 Kinetics of transport in the erythrocyte
135
135
135
135
136
138
138
138
140
140
142
143
143
143
144
144
145
146
147
147
149
149
151
153
153
154
155
155
156
157
158
160
160
161
163
164
169
169
170
170
172
174
Trang 15xiv
2.3.1 General properties and methods of investigation
2.3.2 Transport asymmetry and the effect of cytoplasmic ATP
2.4 Kinetic models for the transport process
2.5 Measurements of individual rate constants for steps in the transport cycle Characterization of the isolated human erythrocyte transporter
3.1 Purification and kinetic properties of the transporter protein
3.2 Molecular properties of the isolated protein
3.2.1 Polypeptide composition and glycosylation state
3.2.2 Secondary structure
3.2.3 Oligomeric state
The structure of the human erythrocyte glucose transport protein
4.1 Amino acid sequence
4.2 Arrangement in the membrane
4.2.1 Topology
4.2.2 Three-dimensional arrangement
4.3 Location of the substrate-binding site(s)
4.3.1 Insights from proteolytic digestion
4.3.2 Photoaffinity labelling with cytochalasin B
4.3.3 Photoaffinity labelling with bis-mannose derivatives
4.3.4 Photoaffinity labelling with forskolin and its derivatives
4.3.5 Photoaffinity labelling with miscellaneous inhibitors
Conformational changes and the mechanism of transport
5.1 Influence of substrates and inhibitors on reactivity towards group-specific reagents
5.2 Biophysical studies
5.3 Differential susceptibility of conformers t o proteolysis
Homologous transporters and their distribution in mammalian tissues
6.1 GLUT-1
6.2 GLUT-2
6.3 GLUT-3
6.4 GLUT-4
6.5 GLUT-5
Homologous transporters in other organisms
7.1 Fungal transporters
1.2 Protozoan transporters
7.3 The transporters of photosynthetic organisms
7.4.1 7.4.1.1 The D-xylose/Hf transporter
7.4.1.2 The L-arabinose/H + transporter
7.4.1.3 The D-galactose/H + transporter
7.4.2 The citrate and tetracycline transporters of E coli
7.4.3 The lactose transporter of E coli
8 Clues to the mechanism of transport from comparison of the homologous transporters
9 Summary
References
3 4 5 6 7 7.4 Bacterial transporters
The galactose, arabinose and xylose transporters of E coli
174
176
177
179
182
182
184
184
184
185
185
185
186
186
189
189
189
189
190
191
191
192
192
194
195
196
197
198
199
199
200
200
200
201
201
202
202
202
202
202
203
207
208
210
21 1
Trang 16Chapter 7 Amino acid transporters in yeast: structure .function and regulation
M.Grenson
1 Introduction
2 3 Physiological background: assimilation of exogenous nitrogen compounds used as a source of nitrogen or as building blocks
General characteristics of amino acid transporters in Saccharomyces cerevisiae 3.1 Accumulation of amino acids
3.2 Multiplicity and specificity of amino acid transporters in Saccharomyces cerevisiae
3.3 Functional specialization of amino acid transporters
3.4 Irreversibility of amino acid accumulation
3.5 Role of the vacuole in amino acid retention
3.6 Efflux of amino acids
Identifying transport systems
4.1 Isolating mutants affected in uptake systems
4.2 An example of transporter identification in a complex case: the three GABA transport systems of Saccharomyces cerevisiae
Structure and evolution of amino acid transporters
5.1 Molecular cloning and nucleotide sequencing of amino acid permease genes
5.2 A family of amino acid transporters with amino acid sequence homologies Regulation of amino acid transport
6.1 Regulation of permease activity
6.2 Regulation of permease synthesis
6.2.1 Case of the NCR-insensitive amino acid permeases
6.2.2 Case of the NCR-sensitive amino acid permeases
6.2.2.1 Constitutive expression of permease genes
6.2.2.2 Inducible permeases
Nitrogen-catabolite repression (NCR) and nitrogen-catabolite inactivation (NCI): two superimposed regulatory mechanisms affecting uptake systems for nitrogenous compounds
6.3.1 4 5 6 6.3 Regulation of amino acid permease activity as a function of 6.3.1.1 6.3.1.2 6.3.1.3 6.3.2 6.3.2.1 6.3.2.2 6.3.2.3 6.3.2.4 6.3.2.5 nitrogen availability
Nitrogen-catabolite inactivation (NCI): negative control of GAP1 Positive control of GAP1 activity
How is GAP1 activity regulated?
Nitrogen-catabolite repression (NCR)
N C R affects permease gene transcription o r transcript accumula- tion
Glutamine as a n effector of N C R
The UREZIGDHCR gene product as a negative regulatory protein which participates in the repression of permease synthesis
The GLN3 gene product as a possible target for the URE2I GDHCR gene product
Double regulation of the ammonia-sensitive permeases
activity
I 8 The A P F l gene product a common factor of unknown function which increases the activity of amino acid permeases
Summary and prospects
219
219
220
222
222
222
223
223
224
225
225
226
226
221
221
23 1
232
232
234
234
234
234
235
231
238
238
239
239
240
240
240
240
240
241
241
241
Trang 17XVI
References 243
Chapter 8 Structure and function of plasma membrane Na+ lHt exchangers Peter Igarashi
1 Introduction
1.1 The N a t / H + exchanger
1.2 Functional heterogeneity
Biochemical properties of N a + / H + exchangers
2.1 ‘Group-specific’ modification
2.1.1 Imidazolium
2.1.2 Carboxyl
2.1.4 Amino
2.1.5 Carbohydrate
Identification and characterization of candidate transport protein(s)
2.2.1 Covalent labeling
2.2.2 Affinity chromatography
2.2.3 Other
cDNA cloning and primary structure
3.1.1 Human N a + / H + exchanger cDNA
3.1.2 Other species
3.2 Tissue and membrane localization
3.3 Isofoms
3.4 Genomic cloning
Summary and future directions
2 2.1.3 Sulfhydryl
2.2 3 Molecular cloning of N a + / H + exchangers
3.1 4 References
Chapter 9 Cl- -channels Rainer Greger
1 Introduction
2 Different types of CI-.channels
2.1 C1- channels in the nervous system
2.2 C1- channels of muscle and electric organ
2.3 C1- channels in apolar non-excitable cells
2.4 The problem of detecting small C1- channels
The structure and molecular basis of CI- channels
3.1 The GABAA-receptor and glycine-receptor channels
3.2 The Torpedo marmorata C1- channel
3.3 Muscle Cl-.channels
3.4 Epithelial C1- channels
Pharmacological modulation of CI-.channels
4.1 Pharmacological modulation of GABAA-receptor and glycine-receptor channels
4.2 Inhibition of epithelial Cl-.channels
2.5 Epithelial C1- channels
3
4
247
247
247
248
249
249
250
251
252
253
254
255
255
257
258
260
260
261
263
265
267
268
269
270
273
273
274
275
276
276
277
278
280
281
281
282
282
283
283
284
Trang 185 Regulation of epithelial CI channels
5.1 The CI- channel defect in cystic fibrosis
5.2 Mechanisms of CI- channel activation in epithelia
References
Chapter 10 Voltage-gated K + channels 0 Pongs
1 Introduction
2 2.1 K' channels of the Shokrr family
2.2 K' channels of the MBK/RCK/HBK family
2.3 K + channels of Shaker relatives
2.4 Pharmacology of K + channels
Structure of K' channel genes
3.1 Genes in Drosophila
3.2 Vertebrate genes
The basis of K + channel diversity
4.1 Properties of homo- and heteromultimers
4.2 Functional domains in K + channels
5 General structural implications
References
Structure and biophysical properties of cloned voltage-gated K+channels
3 4 Chapter I 1 Structure and regulation of voltage-dependent L-type calcium channels M Marlene Hosey Rebecca M Brawley Chan Fong Chang Luis M Gutierrez and Cecilia Mundina- Weilenmann
1 Introduction
1.1 Subtypes of Ca2+ channels
1.2 Functions of L-type C a l i channels
L-type Ca2+ channels
2.1 Pharmacology
2.2 Biochemical and molecular characterization
2.2.1 Isolation and purification of the multisubunit dihydropyridine- sensitive Ca2+ channels from skeletal muscle
2.2.2 Identification and purification of L-type channel proteins from other cells
2.2.3 DNA cloning and expression of channel proteins
2.2.3.1 Isoforms of the a , subunit
2.2.3.2 Cloning of DNAs for other putative channel subunits
2.2.4 Roles of subunits of L-type Ca2+ channels
2.2.5 Reconstitution of C a 2 + channels
Regulation of C a 2 + channels by protein phosphorylation and G-proteins 2.3.1 Phosphorylation by CAMP-dependent protein kinase
2
2.3
287
288
290
291
297
297
298
298
301
302
302
306
306
307
307
307
308
309
311
315
315
315
317
318
318
319
320
321
322
322
323
324
325
326
327
329
330
Trang 19337
Trang 20CHAPTER 1
Na,K-ATPase, structure and transport
mechanism
PETER LETH JORGENSEN
Biomembrane Research Centre, August Krogh Institute, University of Copenhagen,
2100 Copenhugen OE, Denmark
1 Introduction
1.1 The Na,K-pump
The Na,K-pump is ubiquitous and located at the surface membrane of most animal cells Primary active Na,K-pumping is a key process for the active uptake of nutri- ents, salts and water and for the regulation of fluid and electrolyte homeostasis in mammals The pump maintains electrochemical gradients for Na ’- (ApNa) for uti- lization in carrier mediated secondary active transport processes in kidney, intestine, lung and other epithelia In coupling the hydrolysis of ATP to active transport of 3Na+ out and 2 K + into the cell, the pump is electrogenic and maintains ion gradi- ents required for regulation of cell volume and the pump works as a battery for the electrical activity of excitable cells
The Na,K-pump was first purified from the outer medulla of mammalian kidney [ 1,2] and the protein was characterized in membrane-bound form and in detergent solution [3,4] This is a preparation of choice in studies of protein structure and membrane organization [S], conformational transitions coupled to ion translocation [6] and identification of sites for binding nucleotides and cations [7] By incubation in vanadate medium, the proteins of the purified membrane-bound pump protein can be organized in crystalline arrays with the afl unit as the minimum asymmetric unit [5,8]
Low resolution models of the overall structure of the Na,K-pump molecule can be constructed on the basis of diffraction analysis of p l and p12 crystals [9-111 This knowledge of pump structure and function has been important for understanding the physiological function and regulation of the Na,K-pump in kidney [12,13]
Application of recombinant DNA techniques led to the primary structure of the x
subunit [ 14,151 and p subunit [ 15,161 of the Na,K-pump in mammalian kidney and a number of tissues and species (for [ 171 and [ 181) The unitary concept that the Na,K-
Trang 212
pump was essentially the same protein in all tissues and cells has been abandoned as three structurally distinct a subunit isoforms were identified in several species [17,19] Three j subunit isoforms have now also been disclosed and extensive studies of tissue specific and developmental expression of the genes encoding the isoforms and the hormonal regulation of their expression are now being reported In functional and regulatory terms, the significance of the expression of the different combinations of
the a l , a2, and a3 or j 1 , j 2 , and j 3 subunits in brain, skeletal muscle, heart and other tissues remains obscure, mainly because so little is known about the Na,K-transport and enzymatic properties of isozymes other than the renal ( a l p l ) Na,K-pump The a1 j l isozyme of Na,K-ATPase remains the ‘household’ pump that is expressed
in kidney, other epithelia and most other cells This chapter is focused on the structural organization of this renal Na,K-pump and the molecular mechanisms behind the transformation of chemical energy to movement of N a t and K + across the membrane Particular emphasis is placed on the organization of the proteins in the membrane, their interaction with cytoskeletal components, and identificition of protein segments that are engaged in binding of nucleotides or cations and in conformational changes in the protein that bring about a reorientation of the cation binding sites
1.2 Recent review articles on Na,K-ATPase structure and function
The expansion of our knowledge of the structure and function of Na,K-ATPase is reflected in a rapid succession of reviews on Na,K-ATPase genes and regulation of expression [17], subunit assembly and functional maturation [20], the isozymes of Na,K-ATPase [18], and the stability of c1 subunit isoforms during evolution [21], physiological aspects and regulation of Na,K-ATPase [22], reconstitution and ca- tion exchange [23], chemical modification [24], and occlusion of cations [25] Other valuable sources are the review articles [26] and recent developments [27] reported at the International Na,K-pump Conference in September 1990
2 Structure of Na,K-ATPase
2.1 Purified membrane-bound and soluble Na,K-A TPase
The procedure for purification of Na,K-ATPase in membrane-bound form from the outer renal medulla of mammalian kidney offers the opportunity of exploring the structure of the Na,K-pump proteins in their native membrane environment The protein remains embedded in the membrane bilayer throughout the purification procedure thus maintaining the asymmetric orientation of the protein in the baso- lateral membrane of the kidney cell in the purified preparation This preparation has been particularly useful in studies of ultrastructure, protein conformation and for
Trang 22identification of sites for binding of ATP and cations [5-71 A further advantage is that the preparation from outer medulla contains only the a lp1 isozyme of Na,K- ATPase, while most other preparations consist of two or three a subunit and 8- subunit isoforms
The membrane-bound preparation from kidney is easily solubilized in non-ionic detergent and analytical ultracentrifugation shows that the preparation consists predominantly (80-85%) of soluble aB units with M , 143000 [28] The soluble ap
unit maintains full Na,K-ATPase activity, and can undergo the cation or nucleotide induced conformational transitions that are observed in the membrane-bound pre- paration A cavity for occlusion of 2Kf or 3Na+ ions can be demonstrated within the structure of the soluble aP unit [29], as an indication that the cation pathway is organized in a pore through the afi unit rather than in the interphase between subunits
unit [29], when fractions with maximum specific activities of Na,K-ATPase [40-
50 pmol P,/min mg protein) are selected for assay
2.1.2 Electron microscopy and crystal analysis
The purified membrane bound Na,K-ATPase consists of disc-shaped membrane fragments, 1 000-3 000 A in diameter, with no tendency for vesicle formation The densely packed protein particles with diameters of 30-50 A represent ap units that
can be visualized by negative staining with phosphotungstic acid or uranyl acetate They are arranged in irregular clusters or strands and appear to be free to move in the plane of the membrane without formation of well defined oligomeric structures
[30] From negatively stained images similar to those shown in Fig 1, the average
density of protein in the membrane is estimated to be 12000 ap units/pm2 This corresponds to a concentration of a subunit in the lipid bilayer of about 7mM or 0.5-1 g protein/ml of lipid phase These are conditions for supersaturation and for- mation of crystalline arrays of the protein units in the membrane fragments is rapidly induced in the presence of vanadate that stabilize the protein in a state similar to the E2P conformation [ 3 11 In Na,K-ATPase, the predominant crystal
form, shown in Fig 2a, has the two-sided plane group symmetry, p l , and contains one protomeric aP unit per unit cell Crystals with two-sided plane group symmetry, p21, with two UP units occupying one unit cell, are transient and less frequent
Trang 234
Trang 24a b
Fig 2 Crystalline arrays of Na,K-ATPase in the membrane with (a) a protomeric ab unit a s minimum asymmetric unit in a p l crystal or (b) with a n oligomeric (a& unit in the unit cell of a p21 crystal The p l crystal was formed after incubation of purified membrane-bound Na,K-ATPase in 0.25 mM sodium monovanadate, 1 mM MgC12 a t 4°C For formation of the p21 crystal the purified membrane-bound Na,K-ATPase was incubated in 12.5mM phosphate, 1 mM MgC12 and lOmM Tris-HCI, pH 7.5 at 4°C The membranes were negatively stained with uranyl acetate and micrographs were obtained a t 235000 x
magnification Images suitable for further analysis were densitometered a t 20-pm intervals Projection maps were calculated using the Fourier transform amplitudes and phases collected a t the reciprocal lattice points The protein-rich regions are drawn with unbroken contour lines, while negative stain regions have dashed lines In the reconstructed images 1 mm corresponds to 2.8 A The unit cell dimensions are in a:
2.1.3 Three-dimensional models
Low resolution models (20-30 A) based on diffraction analysis of membrane crystals
of Na,K-ATPase [34,35,39] and Ca-ATPase [40,41] show that the cytoplasmic pro-
trusions of the proteins are remarkably similar A notable difference is a 1&20A
Fig 1 Negative staining by phosphotungstic acid of Na,K-ATPase purified in membrane-bound form The membrane surfaces are covered by particles arranged in clusters between smooth areas From [2]
procedure a s described by Deguchi et al [30]
Trang 25protrusion on the extracellular surface of the model for Na,K-ATPase while the Ca- ATPase model has a smooth extracytoplasmic surface
The limitation of the resolution of these reconstructions is the internal order of the Na,K-ATPase crystals At the given resolution of 20-25 A some interactions and basic structure characteristics can be resolved, but there is no assignment of structural detail Higher resolution has been obtained in studies of Ca-ATPase from sarcoplas- mic reticulum in lamellar arrays consisting of sheets of protein arrays separated by lipid layers that are prepared from soluble Ca-ATPase in non-ionic detergent [42,43] Thin three-dimensional crystals were grown by adding purified Ca-ATPase to appropriate mixtures of detergent, lipid and calcium [44] They are rapidly frozen and maintained in frozen-hydrated state during electron microscopy Electron diffraction extends to 4 A and images contain phase data to 6 A resolution Based
on these projections and the previously determined low-resolution structure of Ca- ATPase a packing diagram for the three-dimensional crystals is presented A model with a specific arrangement for ten transmembrane a-helices is proposed [45]
2.2 Cytoskeletal associations
In the polarized tubule cells of mammalian kidney, the specific associations of Na,K-ATPase with cytoskeletal components and the cellkell contacts appear to be important for the induction of polarity of the ap unit between luminal and basolat- era1 membranes The epithelial cell adhesion molecule (CAM) uvomorulin (cadher- in) functions as an inducer of cell polarity for the constitutively expressed alp1 units through cytoplasmic linkage to the membrane cytoskeleton Loss of polarity with incorrect localization of Na,K-ATPase to apical membranes has been associated with a number of diseases including polycystic kidney disease [46,47] One link to the cytoskeleton is a high affinity binding site for ankyrin (& = lop8) that has been demonstrated in the purified renal Na,K-ATPase [48] A fraction of the alp1 units seems to associated in Na,K-ATPase-ankyrin-fodrin complexes with similarity to the capnophorin-ankyrin-spectrin complexes in the cytoskeleton of the human ery- throcyte Induction of cellkell contact alters the properties and distribution of these proteins Before contact between the epithelial cells, the ankyrin-fodrin tetramers form complexes with the membrane proteins, either Na,K-ATPase or uvomorulin
On cellkell contact, uvomorulin seems to mediate redistribution so that the Na,K- ATPase-ankyrin-fodrin complexes accumulate at the sites of cellkcell contact [49,50] It is proposed that the cellkell contacts via uvomorulin induce the specific distribution at the cell surface of Na,K-ATPase during development of the polarized epithelial cells Once polarity has been established, the proteins are replaced by targeting to the appropriate membrane from the Golgi complex
Neuron-glial adhesion in nerve cell cultures is mediated by the 82 subunit AMOG (adhesion molecule on glia) in the a2p2 isozyme of Na,K-ATPase [51] Antibodies to the 82 subunit dissociate cellkcell associations and also increase the rate of active
Trang 26transport suggesting that the p2 subunit may be part of a system for regulation of pump activity in the brain
2.3 Proteolytic dissection of membrane-bound Na,K-A TPase
Proteolytic cleavage has proven to be an efficient tool for exploring the structure and function of the Na,K-ATPase Exposure and protection of bonds on the surface
of the cytoplasmic protrusion provides unequivocal evidence for structural changes
in the a subunit accompanying E1-EZ transition in Na,K-ATPase [52] Localization
of the proteolytic splits provided a shortcut to identification of residues involved in E1-E2 transition [33,53,54] and to detection of structure-function correlations [33] Further proteolysis identifies segments at the surface of the protein and as the cytoplasmic protrusion is shaved off all ATP-dependent reactions are abolished Proteases seem to be unable to shave away protein closer than 10-20 residues from the membrane embedded segments Extensive proteolysis with trypsin or thermolysin
in KC1 medium removes 5&60% of the total a/l unit protein without reducing the content in the membrane of peptides that have been labelled with ['251]-iodonaphthy- lazide (INA) [55] or ['251]-trifluoromethyl-iodo-phenyldiazirine (TID) prior to pro- teolysis [3,56] Table I gives a list of peptide fragments of the a subunit remaining in the membrane after extensive trypsinolysis [7] for comparison with the limit peptides after digestion of the denatured protein [57] The p subunit is remarkably resistant to
proteolysis probably because it is protected by the carbohydrate moieties It is not clear to what extent the /3 subunit protects parts of the CI subunit from digestion Remarkably, the remnants of the aj? unit in the heavily digested membranes retain their ability for binding and occluding Na' or R b + ions which indicates that the coordinating groups of the cation sites are contributed by residues in the intramem- brane segments of the aj? unit [7]
2.4 Membrane organization of the a subunit
The presence of four transmembrane segments in the N-terminal half of the a sub- unit of Na,K-ATPase was predicted from the results of controlled proteolysis of three cytoplasmic sites (Fig 3A) combined with selective chemical labelling with photosensitive ouabain, phosphorylation [5,59] and insertion of small hydrophobic probes, INA [55] o r TID [56,60] In contrast, neither chemical labelling experiments nor hydroplot analysis lead to a decision as to whether two, four, or six transmem- brane segments are formed by the C-terminal part of the a subunit (residues 779- 1016) The model in Fig 3 has the maximum number of transmembrane segments
and their localization in the sequence is given in Table I Comparison of the hydro-
phobic labels [59] shows that T I D provides a reliable random labelling of amino acid side chains that are in contact with the lipid bilayer [56] while INA or adaman-
tanyl diazirine are prone to nucleophilic attachment A D labels exclusively the C-
Trang 27TABLE I
Transmembrane segments in a subunit and B subunit of pig kidney Na,K-ATPase in the model of Fig 3
Segment in model Hydro- Negatively
1.5
1.6 1.2
Trang 28subunit is based on sequencing of surface peptides and identification of S-S bridges [64,65] T I , Tz and C3
show location of proteolytic splits CHO are glycosylated asparagines in the subunit (B) Peptide fragments remaining in the membrane after extensive tryptic digestion of membrane-bound Na,K-ATPase from outer medulla of pig kidney as described by Karlish et al [7,58]
terminal half of the CI subunit and does not label the p subunit (see [59]) INA in low concentration labels only the N-terminal half of the CI subunit probably in a reaction with Cys'02 and Cys13' in M1 and M2, respectively The INA-labelled intramem- brane segment appeared as a 12-kDa fragment in SDS gels [59], and has now been identified in the A ~ p ~ ' - L y s ' ~ ~ peptide which can also be labelled with TID [57]
The intramembrane segments consist of 21-25 amino acid residues with over- representation of the hydrophobic residues Phe, Ile, Leu, Val, Trp, Tyr, but also of Pro and Cys This may suggest that S-S bridge formation is part of stabilizing
intramembrane structures Prolines or glycines break the continuity of membrane
Trang 2910
helices and the excess of proline is interesting in view of the demonstration that membrane-buried proline residues are found in transport proteins, while they are excluded from the membranous domains in non-transport proteins [61] A few basic
or acidic side chains are found in the transmembrane segments and the segments carry charged residues close to their cytoplasmic ends These charges may react with headgroups of lipids to stabilize the structure in the membrane [62] During biosynth- esis the charged residues may have served as stop signals preventing transfer across the membrane
One argument for a model with an equal number of eight or ten transmembrane segments in the a subunit is that the C-terminal seems to be exposed to the cytoplasm
as it is in the Ca-ATPase of the sarcoplasmic reticulum [63] Recent immunological studies suggest the presence of a cytoplasmic epitope near the C-terminus of the a subunit In the model of the a1 subunit of Na,K-ATPase in Fig 3, the transmembrane segments M1, M2, M3, M4, M5 are predicted by a hydropathic index of >2.0 and three segments M6, M8, and M9 have peak indices of > 1.5 using a window of 19
residues or the hydroplot, Table I Transmembrane segment M8 is included because it has a peak index of 1.6 in the hydroplot, but this segment does not seem to be labelled
by the hydrophobic labels In contrast the peptide Va1545-Arg589 is consistently labelled by hydrophobic labels [57] This peptide has a maximum hydropathic index
of 0.9 and possesses only short hydrophobic stretches It may form a hydrophobic pocket in a b-sheet structure of the ATP binding area as proposed on the basis of homologies with adenylate kinase and other ATP binding proteins [6,78]
2.5 Structure of the p subunit of Nu,K-ATPuse
In the family of cation pumps, only the Na,K-ATPase and H,K-ATPase possess a p
subunit glycoprotein (Table II), while the Ca-ATPase and H-ATPase only consist of
an a subunit with close to 1000 amino acid residues It is tempting to propose that the b subunit should be involved in binding and transport of potassium, but the functional domains related to catalysis in Na,K-ATPase seem to be contributed exclusively by the a subunit The functional role of the b subunit is related to biosynthesis, intracellular transport and cell-cell contacts The p subunit is required for assembly of the a,O unit in the endoplasmic reticulum [20] Association with a fi
subunit is required for maturation of the a subunit and for intracellular transport of the ap unit to the plasma membrane In the p1-subunit isoform, three disulphide bonds are formed by residues C y ~ ' * ~ - C y s ' ~ * C y ~ ' ~ ~ - C y s ~ ~ ~ , and Cys212-Cys275 [64,65] and their reduction is accompanied by loss of Na,K-ATPase activity [64,66] The 82-subunit isoform seems to have a function as a recognition element for cell adhesion in the brain This association may also mediate fine regulation of transport because antibodies to the /I2 subunit (AMOG = association molecule of glia) cause dissociation of cellkell adhesion and stimulate Na,K-transport [5 11 The bulk of the hydrophilic residues of the fi subunit are exposed on the extra-
Trang 30Homology of p-subunit isoforms of Na,K-ATPase and H,K-ATPase corresponding to beta signatures no
1 and no 2 of the Prosite database [70] Reference to sequences: PI rat [71], PI mouse [72], p2 rat [73], a2 mouse [51], 83 Xenopus [74], p subunit of H,K-ATPase [75]
Beta signature no 1
trifluoromethyl-iodo-phenyldiazirine [60] As an alternative to this model three
transmembrane segments are proposed on the basis of papain digestion [68] and
immunological studies [69] It is remarkable that both the Phe'4-Arg'42 and G I Y ' ~ ~ - Ser302 segments remain associated with the membrane even after trypsinolysis and reduction of the C y ~ ' ~ ~ - C y s ' ~ * disulphide bridge [7], but there is no evidence for additional hydrophobic segments in the Gly'43-Ser302 segment
The fi subunit is not as well conserved as the a subunit, with 91 YO overall homology between the j subunit of sheep, pig, and human and 61% between the j subunit of
human and Torpedo Homologies between P-subunit isoforms and the j subunit of
H,K-ATPase are moderate, 25-30%, but some segments are well conserved As shown in Table 11, beta signatures 1 and 2 are preserved in the j-subunit isoforms and j subunit of H,K-ATPase [70] as an indication that the positions of tryptophans and cysteines are well conserved elements
3 Nucleotide binding and phosphorylution
A characteristic structural feature of the renal Na,K-pump protein is a cytoplasmic protrusion with approximate dimensions 45 x 6 5 A in the plane of the membrane and a length of 5&60A in the plane perpendicular to the membrane The bulk of the protrusion is formed by the large central domain (residues 340-780 in cx subunit)
Trang 3112
that forms sites for ATP binding and phosphorylation The second cytoplasmic
domain (142-284 in a subunit) between M2 and M3 contains peptide bonds suscep-
tible to proteolytic cleavage in El forms and this domain is proposed to be involved
in energy transduction [6,33] The N-terminus is attached to M1 and involved in control of EI-E2 transition in Na,K-ATPase and of the rate of Na,K-pumping [6]
3.1 The nucleotide binding domain in the a1Pl unit
3.1.1 Comparison with the nucleotide binding sites in adenylate kinase
An indication of the functional groups required to form a nucleotide site may be obtained from examining sites in dehydrogenases, phosphofructo-kinase, and ade- nylate kinase [7&78] The comparison with adenylate kinase is particularly interest- ing in view of the information available on the structure of the nucleotide binding area in that protein [78] and the homologies demonstrated earlier [6] An important feature in the model of the ATP site in adenylate kinase is a hydrophobic pocket for accommodation of the adenine and ribose moieties which is formed by Ile, Val, His and Leu residues The triphosphate moiety is flanked by a hydrophobic strand of parallel P-pleated sheet terminated by Asp [78] This segment in adenylate kinase and in the P subunit of F1-ATPase shows some homology with respect to charges
and hydrophobic residues to segments in a subunit of Na,K-ATPase (543-561) The
observation that this segment is labelled by T I D or A D [57] shows that it may also form a hydrophobic pocket in the nucleotide binding domain of Na,K-ATPase Such a pocket may explain the high affinity of the reaction of the nucleotide site of Na,K-ATPase for reaction with compounds like trinitrophenyl-ATP [79] and tetra- bromo-fluorescein (eosin) [80]
Transferred nuclear Overhauser effect (TRNOE) measurements complement the sequence data in the sense that part of the ATP moiety is organized in a manner similar to that in adenylate kinase [78], while there is a significant difference in the torsion angle of the bond between adenine and ribose In Na,K-ATPase [81] the bound ATP adopts an anti conformation for the adenine ring with respect to the ribose (6 = 0 f 90") with a glycosidic torsion angle (6) of 35" The conformation of the ribose ring is N-type (C2,-exo, E3,-endo) with a torsion angle 6 = 100" The orientation of O5 relative to the ribose is determined by the torsion angle, 6 = 178", a typical value for protein bound Mg-ATP [78]
3.1.2 Selective chemical labelling with A T P analogues
Labelling Na,K-ATPase with ATP analogues provides evidence for contribution from charged residues that are widely separated in the sequence of CY subunit of Na,K-ATPase The first indication came from ATP sensitive covalent insertion of fluorescein-isothiocyanate (FITC) into Lyssol in the a subunit [90] The strong fluor- escence signal provides a convenient probe for monitoring conformational transi- tions in the proteins Site-directed mutagenesis of LysSo1 reduces the activity of
Trang 32Na,K-ATPase [82] Asp7l0 in the highly conserved 704-722 segment is covalently labelled by CIR-ATP (a-[4-(N-2-chloroethyl-N-methyl-amino)]benzoyl-amide-ATP)
[83] Another conserved sequence, 470487, is labelled by 8-azido-ATP and Lys480 in
this segment is targeted by pyridoxal-ATP [84] The lysine selective reagent (N-iso- thiocyano-phenyl-imidazole) also reacts in a conformation specific manner to block ATP binding, but the target residue has not been identified [85] In these studies it has not been excluded that the chemical modification alters the conformational adaptability of the protein rather than blocking side chains that are contributing coordinating groups for the nucleotides
3.2 Conformations of the nucleotide binding area
The segments contributing to nucleotide binding and phosphorylation domains un- dergo structural changes accompanying E ,-E2 transition as the nucleotide binding region adapts for tight binding in the El form with K D 0.1 pM for ATP, while binding t o the E2 form is weak requiring millimolar concentrations of ATP for saturation
The CI subunit provides the necessary segments for formation of a nucleotide binding area, but the possibility of more than one nucleotide site per C I ~ unit is often raised Two separate ATP sites with high and low affinity have been proposed on the basis of kinetic studies or inactivation experiments [86] The different affinities for ATP can be explained [79] by the alternating El and E2 conformations with high and low ATP affinities, respectively Since the maximum capacity of ATP binding in equilibrium experiments never exceeds one ATP bound per C I ~ unit, one may assume that a presumptive additional site has a low affinity and fewer coordinating groups than a site for high affinity binding
3.3 The phosphorylation site, high- und low-energy phosphoforms, EIP-E2P
A sequence of ten amino acids (ICS-D-KTGTLT) around the phosphorylation site
of Na,K-ATPase (Asp369) is highly conserved among the Na,K-, H,K-, Ca-, and H- pumps [6] There is also homology with the a subunit of F1-ATP synthetase of mitochondria and chloroplasts (see [6]) except that Asp is replaced by Thr Accord-
ingly a covalent phosphorylated intermediate is not formed in F1-ATPase Muta- genesis of the phosphorylated aspartate residue in Na,K-ATPase [82], Ca-ATPase
[87], or H-ATPase [88] completely blocks activity
Transition from the ‘high-energy’ phosphoform ElP[3Na] to the K-sensitive E2P[2Na] of Na,K-ATPase are accompanied by conformational transitions in pro- tein structure and changes of the capacity and orientation of cation sites In the El form of Na,K-ATPase, the exposure of Chy3 (Leu266) and Try3 (Arg262) to cleavage reflects that the cation sites of the phosphoprotein are in a conformation oriented towards the cytoplasm with a capacity for occlusion of three N a + ions The E2 form
Trang 3314
with exposed Try, (Arg438) and protected Chy3 and Try3 occludes either 2Naf or 2Rb'(K+) in the phosphoform or 2 R b t ( K + ) in the unliganded enzyme and it exchanges cations at the extracellular surface [6,33,89]
In the scheme in Fig 4 [6,89], the EIP-E2P transition releases a single Na+ ion at the extracellular surface and E2P[2Na] represents an occluded state in transition to E2P-2Na with N a t leaving the sites making them accessible for binding of K f from the extracellular phase In a scheme involving two cycles the term E*P was used for this intermediate [90] and it was observed that ouabain reacts with E*P 1911 Studies of protein conformation and cation occlusion show that the correct notation for E*P is E2P[2Na], since E2P can occlude either 2Na' or 2 K + , without altering protein conformation as detected by proteolysis In Na + medium, the E2P[2Na] intermedi- ate is sensitive to ADP, because binding of one N a + allows it to return to the E1P[3Na] form for reaction with ADP and formation of ATP After addition of
K t , exchange of Na+ for K + at the extracellular surface would lead to dephos- phorylation The apparent ambiguity of the E2P[2Na] form with respect to reactivity
to ADP and K f is therefore explained by the cation site occupancy, while the protein conformation of the E2P[2Na] intermediate is the same as that of other E2 forms With these properties of the E2P[2Na] complex, the ADP sensitive fraction of the phosphoenzyme comprises both E l , and E2 forms, namely the Na-occluded, E1P[3Na] and E2P[2Na] The conventional definition that the ADP-sensitive fraction
of the phosphoenzyme corresponds to the amount of the E I P form is therefore no longer valid The redefinition of the ADP sensitive phosphoenzyme solves an objec- tion that has repeatedly been raised towards the role of the phosphoenzyme in the Na,K-pump reaction, that the sum of the ADP- and K-sensitive phosphoenzymes exceed the total amount of phosphoenzyme (see [92]) Using the definitions above and
Trang 34in Fig 4, the sum of the amounts of ADP-sensitive and K-sensitive is equal to ElP[2Na] plus 2 x E2P[2Na] plus E2P[O] and thus exceeds the total phosphoenzyme
by the amount of the E2P[2Na] intermediate
Based on a series of studies of the effect of organic solvent on the reaction of Ca- ATPase with P, and ATP synthesis, De Meis et al proposed that a different solvent structure in the phosphate microenvironment in El and E2 forms the basis for existence of high- and low-energy forms of the aspartyl phosphate [93] Acyl
phosphates have relatively low free energy of hydrolysis when the activity of water
is reduced, due to the change of solvation energy The covalently bound phosphate may also reside in a hydrophobic environment in E2P of Na,K-ATPase since increased partition of Pi into the site is observed in presence of organic solvent [6]
in the same manner as in Ca-ATPase
4 Cation binding and occlusion
To understand the Na,K-pump mechanism it is obviously important to identify the cation pathway and the sites for binding and occlusion of N a + and K + relative to the intramembrane portion of the protein The groups coordinating the cations should be identified and it should be known if the pump has independent sites for
N a + and K + or if the cations bind alternately to the same set of sites With a stoichiometry of 3 N a + / 2 K C per ATP split this would mean that two sites bind
N a + and K + alternately, while one site only binds Na'
Known structures of other proteins or ionophores can be useful models for the cation sites in the Na,K-pump In models for ionophores, cation binding sites consist
of electrophilic carbonyl groups that are located in a fixed cavity to accommodate the size of the selected cations inside the ionophore Transition from an open to a closed
or occluded configuration of the cation binding sites in the ionophore involves only limited changes in conformation of carbonyl residues and a substitution of solvent molecules from the inner coordination sphere of the cation [94] This arrangement allows for excellent cation specificities, e.g., the affinity of valinomycin for K + is about six orders of magnitude higher than for N a + The complex of a crown ether (cyclohexyl- 18-crown-6) with Na has five nearly coplanar oxygens surrounding the
N a + with the ion coordination completed by the remaining oxygen of the hexaether and a water molecule, and with Na-0 distances of 2.5-2.6A [95] The corresponding complex with K + can be formed with minimal distortion of the cyclic ether and with bond lengths of 2.8 A
In the Ca-ATPase from sarcoplasmic reticulum, oligonucleotide-directed, site- specific mutagenesis has been applied to identify amino acids involved in C a 2 + binding Mutation of 30 glutamate and aspartate residues, singly or in groups, in a stalk sector near the transmembrane domain has little effect on Ca'+-transport In contrast mutations to G1u309, G I u ~ ~ ' , Thr799, Aspsoo or Gh9" resulted in loss
Trang 3516
of Ca2+-transport [96] As an indication that these mutations abolished C a 2 + - binding, phosphorylation from inorganic phosphate was observed even in the pre- sence of Ca2+ These data suggest that the carboxylate or carboxamide side chains or hydroxyl groups form coordinating groups in high affinity binding sites for Ca2+ near the centre of the transmembrane domain The homology around these residues between Ca-ATPase and a subunit of Na,K-ATPase and other pump proteins suggest that their cation sites may have a similar location and overall structure
4.1 Capacity for binding and occlusion of Na+ or K+ ( R b + )
It is generally accepted that Na' ions can be occluded in E I P forms Occlusion of 3Na+ ions per EP has been demonstrated in chymotrypsin cleaved enzyme and in the Cr-ADP-E1P[3Na] complex [29] Three N a + ions can also be occluded per EP in
a complex stabilized by oligomycin in the absence of Mg2+ or phosphate [97] while
a maximum of two N a f ions are occluded per a subunit in the ouabain complex
In preparations of high purity, the cation binding data above therefore correspond
to 2Rb+ and 3Naf occluded per a j unit both in the soluble and in the membrane bound state Fig 5 shows that the capacity for occlusion is either 2Rbf or 2Naf per
a j unit The apparent affinity of these E2 forms for the cations varies over a wide range from KIiz( Rb) N 9 pM to Kli2(Na) - 1.7 mM, but without changes in proteo- lytic cleavage patterns The apparent affinity of *6Rb for formation of the Mg- ouabain-E2P[Rb] complex is > 10-fold higher than for Na [89] The difference in apparent affinity for the complexes in Fig 5 is sufficient for allowing exchange of Na for Rb(K) at the extracellular surface, but the transition from E2P[2Na] to E2[2Rb] has no influence on the protein conformation as judged by the patterns of proteolytic digestion or the fluorescence levels
0 4
N
0.001 0.01 0.1 1 10 , ou
RbCl or NaCl (mM)
Fig 5 Occlusion of ouabain complexes with ( 0 ) "Na or (a) "Rb and unliganded Na,K-ATPase with
(A) "Rb Procedure as described before [89] using incubation at increasing concentrations of N a t or
86Rb for 15min a t 20°C with 1 mM ouabain, 1 m M MgClz and 1 mM PI-Tris
Trang 364.2 Isolation of the cation occlusion und transport path after tryptic digestion
Using the renal membrane bound Na,K-ATPase, extensive tryptic digestion pro- vides a method for isolating the cation sites and the pathway for cations across the membrane as a structure quite separate from the ATP binding and phosphorylation sites The intramembrane part of Na,K-ATPase is protected from tryptic digestion
by occlusion of either K + or Na’ while ATP binding and phosphorylation are rapidly abolished as the entire cytoplasmic protrusion is digested away As the frag- ments remaining in the membrane are the same whether R b + or N a + ions are protecting the sites against inactivation, the two ions may alternately bind to the same coordinating groups in the intrainembrane part of the enzyme The sites seem
to remain accessible from both membrane surfaces after the extensive cleavage [7,581
The intramembrane fragments capable of occluding cations in the digested mem- branes (Table I, Fig 3B) comprise a 19-kDa fragment plus 8-kDa, 9-kDa and 1 1-kDa fragments of the a subunit and a 14-kDa fragment of the /3 subunit The presence of residues coordinating cations is demonstrated by [I4C]DCCD binding to the frag- ments Thus K+-protected labelling was demonstrated in the 19-kDa fragment [7,58] suggesting that G ~and Glu9” are candidates for K u ~ ~ ~ + -binding carboxyls
4.3 Transport stoichiometry and net charge of N a + and K + complexes with Na,K-
A TPase
The membrane potential can be used as a tool for characterization of the properties and net charge of the major conformations of Na,K-ATPase and their complexes with the cations The potential difference across the membrane represents large electric fields of up to 500 kV/cm If the conformational states of the Na,K-pump have different electrical properties with respect to net charge or movable dipoles, the equilibria between conformations should interact with the field The field may also affect binding of N a + or K’ by altering the concentration of cation near the pump protein o r by altering the position of charged groups in the cation sites
In phospholipid vesicles reconstituted with renal Na,K-pumps, the membrane potential can be controlled using gradients of K + or Li+ and appropriate iono- phores At high or low ATP concentrations, the conformational change EIP3Na- E2P3Na is accelerated by voltage, whereas the E2[2K]-EI [2K] step is independent of potential [97] At limiting low Na ’ concentrations the imposed potential stimulates Na-transport showing that the electric field has an effect on the intrinsic binding affinity for N a + This may mean that N a + has to cross part of the electric field in order to reach its sites, thus becoming concentrated by the field Alternatively, the voltage may cause charges to move or dipoles to reorient in the cation binding area of the protein Studies of the conformational transitions using fluorescein attached covalently to Lyssol [98] also shows that the E1P-E2P isomerization is voltage
Trang 3718
dependent while the E ~ K - E I K step is electrically silent
The purified membrane bound Na,K-ATPase from kidney adsorbs to planar lipid bilayers and transient currents can be elicited upon release of caged ATP [99,100] The transient pump currents depend on Mg2+ and N a f , but not K + , and they are abolished by chymotryptic cleavage This agrees with the inference that N a f trans- location and charge movements precede the K + -transport step Similar conclusions are reached from studies of presteady state fluxes of Na+-ions in plasma membrane
or reconstituted vesicles [ 1011 With the assumption that 3Naf-ions are transported into and 2K+ ions out of the vesicles in a ping-pong reaction, these results mean that the net charge is + 1 with 3Na' ions bound and zero with 2Kf ions bound [98]
5 Structural transitions in the protein related to energy transformation and Na,K-transport
Transduction of the energy from ATP to movement of the cations may involve long- range structural transitions in the protein since ATP binding and phosphorylation takes place in the large cytoplasmic protrusion of the a subunit, while cation sites may be located in intramembrane domains It is therefore important to establish relationships between the structural changes in the CI subunit and ion binding or occlusion to see if the different exposure of bonds to proteolysis reflect the orienta- tion and specificity of the cation sites
The conformational transitions in Na,K-ATPase are large and easy to detect In addition to bonds exposed to proteolysis [6], the transition involves tryptophans, sulfhydryl groups, protonizable groups and residues binding FITC [ 1021 and iodoa- cetamide fluorescein [ 1031, and the conformational changes involve residues in ATP binding and phosphorylation domains and they are transmitted to the extracytoplas- mic surface with changes in binding affinity for cations and ouabain, see previous reviews for details [6,92]
5.1 Conformation dependent proteolytic cleavage of Na,K-ATPase
Definition of El and E2 conformations of the a subunit of Na,K-ATPase involves
identification of cleavage points in the protein as well as association of cleavage with different rates of inactivation of Na,K-ATPase and K-phosphatase activities [104,105] In the E l form of Na,K-ATPase the cleavage patterns of the two serine proteases are clearly distinct Chymotrypsin cleaves at Leu266 (C,), Fig 3A, and both Na,K-ATPase and K-phosphatase are inactivated in a monoexponential pat- tern [33,106] Trypsin cleaves the E l form rapidly at Lys30 (T2) and more slowly at Arg262 (T3) to produce the characteristic biphasic pattern of inactivation Localiza- tion of these splits was determined by sequencing N-termini of fragments after iso- lation on high resolution gel filtration columns [ 1071
Trang 38The E2 form is not cleaved by chymotrypsin, but trypsin cleaves at Arg438 (TJ and subsequently at Lys3' (T2) and tryptic inactivation of E2K or E2P forms is linear and associated with cleavage at Arg438 (TI) [ 104,1081 Inactivation of K-phosphatase is delayed because cleavage of T I and T2 in sequence is required for inactivation of K- phosphatase activity [105]
Thus, transition from El to E2 consists of an integrated structural change involving protection of bond C3 or T3 in the second cytoplasmic domain and exposure of TI in the central domain, while the position of T2 in the N-terminus is altered relative to the central domain (TI) so that cleavage of T2 becomes secondary to cleavage of T I within the same a subunit in the E2 form
5.2 Tryptophan ,fluorescence and secondary structure changes
Movement of peptide segments from cytoplasmic to hydrophobic environments accompanying EI-E2 transitions in Na,K-ATPase may be related to changes in intensity of intrinsic fluorescence in the proteins Quenching analysis of Na,K-ATP- ase suggests that the tryptophan fluorescence changes are directly related to change
of overall protein conformation [ 1081 Tryptophan fluorescence is increased 2-3%
by transition from El to E2 in Na,K-ATPase [cf 1091 This quantitative difference in
fluorescence of E l and E2 forms may be understood in terms of different locations of the tryptophans in the protein structures In Na,K-ATPase, only two of a total of twelve tryptophans in the CI subunit are located in predicted transmembrane seg- ments
Circular dichroism (CD) spectroscopy may detect changes in the ratio between secondary structure elements, but even large shifts in position of a-helices and B-
sheets relative to each other do not affect C D spectra The results of experiments addressing the question whether EI-E2 transitions are accompanied by changes in the ratio among secondary structure elements raised some controversy C D spectroscopy shows that Na,K-ATPase contains a roughly equal mixture of a-helical, fi-sheet and random coil structures [110] and changes in C D spectra accompanying exchange of
N a i for K + are interpreted to involve a-helix-&sheet transition of about 7% A
conversion in the opposite direction of 100 residues from fi- to a-helical conformation
is suggested by Raman spectroscopy [ l l l ] In apparent contrast to this, another group found that addition of K + to a Tris-HC1 medium did not cause changes in C D spectra [112] Infrared spectroscopy in the amide I region shows that N a + and K + bound forms of renal Na,K-ATPase have almost the same structure [113], but it is not known if the enzyme is active in D 2 0 It is not clear to what extent these contradictory results concerning secondary structure transitions are related to the special structure
of the membrane bound Na.K-ATPase
Trang 3920
5.3 Cleaved derivatives; cleavage of bond 2 and the regulatory function of the N -
terminus
Selective cleavage of bonds in the a subunit of Na,K-ATPase is important for ex-
amining structure-function relationships for the protein The N-terminus of the a
subunit is strongly hydrophilic with clusters of alternating positive and negative residues between residue 15 and 60 It is a flexible structure with a strong propensi-
ty for a-helix formation and several predicted turns, notably at residues 14-16, 35-
36, 49-50 and 70-72 The removal of residues 1-30 by selective tryptic cleavage at Lys3' (T2) is possible because the rate of cleavage of this bond is up to 60-fold higher than the rate of tryptic cleavage at Arg262 (T3) in the second slow phase of inactiva- tion T2 cleavage reduces Na,K-ATPase activity by 5&60% [lo41 with a parallel loss
of Na,K-transport [log] The loss of activity is explained by a poise of equilibria between cation bound (EINa-E2K) and phosphoforms (EIP-E2P) in the direction
of El forms [114] These properties of the selectively cleaved derivative suggested that charged residues in the N-terminus engage in salt bridge formation as part of EI-E2 transition and that this is important for control of the rate of active Na,K- transport by pumps containing the a1 isoform
These data suggest that the substantial differences between the amino acid sequences of the N-terminal region in the a 1, a2, and a3 isoforms [ 191 reflect different
regulatory functions of these segments The first 30 residues of the a1 isoform have a high frequency of charges with eight lysins and two arginins In the a2 isoform there are two negative and three positive charges fewer in this segment than in the crl isoform suggesting that the strength of the salt bridges formed with other segments may be weaker The first 11 residues of the a3 isoform are not homologous with the other isoforms, but all isoforms of the a subunit have a stretch of lysins around
residue 30 Also the N-terminus of H,K-ATPase [115] possesses a lysine-rich sequence
with strong homology with the a l subunit around T2 Biphasic cleavage patterns
resembling those in Na,K-ATPase have also been demonstrated in H,K-ATPase [I 161, but so far without identification of cleavage points
5.4 Effect of C j cleavage on EIP-E2P transition and cation exchange
The alternating exposure of C3 (Leu266) or T3 (Arg262) in the El form and T1 (Arg438) in the E2 form reflects that motion within the segment ( M , = 18 170) be- tween these bonds including the phosphorylated residue (Asp369) is an important element in E1-E2 transition This is illustrated by the widely different consequences
of selective cleavage of C3 and T I for El-Ez transition and cation exchange C3 cleavage is a selective and particularly efficient tool for examining structure- function relationships of the second cytoplasmic domain Binding affinities for ADP and ATP are reduced 4-5-fold, while TNP-ATP binds with the same affinity as in native Na,K-ATPase Nucleotide binding is not affected by K f or R b ' although
Trang 40TABLE 111
Properties of the C3-cleaved derivative of Na.K-ATPase
Enzymatic activity, ligand binding Cleavage C3 Control
Na,K-ATPase
ATP-ADP exchange
ADP binding capacity
ADP binding affinity (KD, pM)
Phosphorylation capacity
EIP/E2P ratio lOOj0
Vanadate binding capacity
Rb-binding capacity
Rb-binding affinity (KD, pM)
0%
4-5OOYo 100%
0.075 100%
100 Yo
100 Y O
9-12
Compiled from [6] and [33]
cation sites are undamaged Conversely the cleaved enzyme also binds "Rb with high affinity and it occludes the cations, but cation binding and occlusion are unaffected
by nucleotides, Table 111
Transport studies in reconstituted vesicles show that C3 cleavage blocks the relatively fast Na-Na or K-K exchange (2040s -I) and Na-K exchange (500s-'), but the slow passive ouabain-sensitive Rb-Rb exchange (1 s- I ) and occlusion of K +
or N a + are only partially affected C3 cleavage allows formation of EIP[3Na], but prevents charge transfer coupled to Na + translocation in purified Na,K-ATPase after adsorption to a planar lipid bilayer [I 171 In combination with the observation
of the occlusion of 2Na+ in the E2P[2Na] these data show that E2P-E2P transition represents a charge translocating step and that the single N a t ion released at the extracellular surface represents this transfer of charge C 3 cleavage thus interferes with structural changes that alter the capacity of the cation sites for occlusion of Na +
and presumably their orientation
5.5 Mutagenesis in yeast H-A TPase und Cu-A TPase ,from sarcoplasmic reticulum
The notion that the segment containing C3 and T3 is important for conformational adaptability of the protein is supported by mutations in yeast Mutations of the genes of the H-ATPase of Saccharomyces cerevisiae resulted in a thermo-sensitive
mutant ( G l ~ ~ ~ ~ - + S e r ) [118] In Schizosaccharomyces pombe the G l ~ ~ ~ ' + A s p substi- tution is responsible for a mutant phenotype with reduced ATPase and proton pumping activity and vanadate resistance [ I 191 The mutation may thus produce a higher concentration of E l forms and less of the vanadate binding E2 form during steady state ATP hydrolysis In interpreting the work o n mutations in H-ATPase, the view was forwarded that mutations in the second cytoplasmic loop disrupt an endogenous phosphatase activity [ 1 181 Inactivation of phosphatase activity after