14 Jan 2003
14:58
AR
AR177-PH65-34.tex
AR177-PH65-34.sgm
LaTeX2e(2002/01/18)
P1: IBC
10.1146/annurev.physiol.65.092101.142558
Annu. Rev. Physiol. 2003. 65:817–49
doi: 10.1146/annurev.physiol.65.092101.142558
c 2003 by Annual Reviews. All rights reserved
Copyright °
First published online as a Review in Advance on January 9, 2003
STRUCTURE AND MECHANISM OF Na,K-ATPASE:
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
Functional Sites and Their Interactions
Peter L. Jorgensen,1 Kjell O. Håkansson,1 and
Steven J. D. Karlish2
1
Biomembrane Center, August Krogh Institute, Copenhagen University,
Universitetsparken 13, 2100 Copenhagen OE, Denmark; e-mail: [email protected];
[email protected]
2
Department of Biological Chemistry, Weizmann Institute of Sciences, Rehovot 76100,
Israel; e-mail: [email protected]
Key Words Na-K-pump homology models, ATP binding, Na+ and K+ ion
binding, energy transduction mechanism
■ Abstract The cell membrane Na,K-ATPase is a member of the P-type family of
active cation transport proteins. Recently the molecular structure of the related sarcoplasmic reticulum Ca-ATPase in an E1 conformation has been determined at 2.6 Å
resolution. Furthermore, theoretical models of the Ca-ATPase in E2 conformations are
available. As a result of these developments, these structural data have allowed construction of homology models that address the central questions of mechanism of active
cation transport by all P-type cation pumps. This review relates recent evidence on
functional sites of Na,K-ATPase for the substrate (ATP), the essential cofactor (Mg2+
ions), and the transported cations (Na+ and K+) to the molecular structure. The essential
elements of the Ca-ATPase structure, including 10 transmembrane helices and welldefined N, P, and A cytoplasmic domains, are common to all PII-type pumps such as
Na,K-ATPase and H,K-ATPases. However, for Na,K-ATPase and H,K-ATPase, which
consist of both α- and β-subunits, there may be some detailed differences in regions of
subunit interactions. Mutagenesis, proteolytic cleavage, and transition metal-catalyzed
oxidative cleavages are providing much evidence about residues involved in binding
of Na+, K+, ATP, and Mg2+ ions and changes accompanying E1-E2 or E1-P-E2-P conformational transitions. We discuss this evidence in relation to N, P, and A cytoplasmic
domain interactions, and long-range interactions between the active site and the Na+
and K+ sites in the transmembrane segments, for the different steps of the catalytic
cycle.
INTRODUCTION AND SCOPE
The Na,K-pump is essential for all mammalian cells. At rest it consumes 20–30%
of ATP production to actively transport Na+ out of and K+ into the cell. The Na
and K gradients are required for maintaining membrane potentials, cell volume,
0066-4278/03/0315-0817$14.00
817
14 Jan 2003
14:58
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
818
AR
AR177-PH65-34.tex
JORGENSEN
¥
AR177-PH65-34.sgm
HÅKANSSON
¥
LaTeX2e(2002/01/18)
P1: IBC
KARLISH
and secondary active transport of other solutes, e.g., the transcellular transport
processes in intestine, glands, and kidney.
Na,K-ATPase is the largest protein complex in the family of P-type cation
pumps. The minimum functional unit is a heterodimer of the α- and β-subunits
(1) and may also be coexpressed with small ion transport regulators of the FXYD
family (2, 3). Individual genes of four α-subunit isoforms and at least three
β-subunit isoforms of Na,K-ATPase have been identified in mammals (4, 5). The
isoforms combine to form a number of Na,K-ATPase isozymes that are expressed
in a tissue- and cell-specific manner. The central question in the comprehensive
physiology of active Na+ and K+ transport is the nature of the complex regulation
of the expression of these isozymes and the variety of systems for posttranslational
modification and short-term regulation.
Biochemical and spectroscopic data show that long-range E1-E2 conformational transitions in the α-subunit mediate interactions between cytoplasmic
domains and the cation sites in the intramembrane domain (1, 6–8). These transitions couple the scalar processes of ATP binding, phosphorylation, and dephosphorylation to the vectorial extrusion of three Na+ ions and uptake of two
K+ ions.
High-resolution molecular structural information is essential for understanding these complex reactions and the physiological regulation of the system. The
Na,K-pump was first visualized as distinct particles of 30–50 Å diameter in purified membrane preparations from mammalian kidney (9). Crystallization in the
membrane after incubation in Mg2+ and vanadate (10) led to a three-dimensional
reconstruction at 20–25 Å resolution, with information on the mass distribution
of the cytoplasmic protrusion of the α-subunit and the extracellular protrusion
of the β-subunit (11). Recently, electron microscopy image reconstructions of
Na,K-ATPase at 9–11 Å resolution were reported (12, 13). Crystallization of sarcoplasmic reticulum Ca-ATPase in vanadate (14) led to 8–9 Å resolution structures, which allowed tentative assignment of transmembrane helices (15). In 2000,
Toyoshima and coworkers presented the 2.6 Å high-resolution structure based on
X-ray diffraction of crystals grown in CaCl2 solution (16). In addition to this structure at atomic resolution (1EUL), two structures of the E2 conformation at 6 Å
resolution are available, based on electron microscopy of Ca-ATPase stabilized by
decavanadate and thapsigargin, 1FQU (16) and 1KJU (17).
The focus of this review is the structure-function relationship of binding sites
for Na+, K+, Mg2+, and ATP and the structural changes mediating interactions
of functional sites during the catalytic cycle. Ideally, resolution of these mechanisms requires a high-resolution structure of each intermediate of the reaction
cycle. In the absence of such information provisional concepts can be inferred by
comparing a high-resolution structure of the E1[2Ca] form (16) with the available models of the E2-forms [1FQU and 1KJU (17)]. The structural reconstructions of the E2-Mg-vanadate complexes of the α-subunit of Na,K-ATPase (12, 13)
show considerable similarity to the E2 forms of Ca-ATPase (16, 17). Homology modeling of the α-subunit of Na,K-ATPase on the backbone of the
14 Jan 2003
14:58
AR
AR177-PH65-34.tex
AR177-PH65-34.sgm
LaTeX2e(2002/01/18)
Na,K-PUMP STRUCTURE AND FUNCTION
P1: IBC
819
three-dimensional structures of Ca-ATPase therefore allows visualization of the
domain movements as large global changes of structure involving practically all
amino acid residues. In light of these models, data on proteolytic (1, 8) or metalcatalyzed (6, 7) peptide bond cleavage, mutagenesis, and chemical modification
can be interpreted to yield important information on the mechanism of energy
transduction.
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
THE KINETIC MECHANISM OF Na,K-ATPase
Extensive studies into the cation transport mechanism and partial reactions, including definition of the properties of the phosphorylated intermediates (18), the
occluded Rb+ or K+ after dephosphorylation (19), and the conformational changes
of the protein (1), provided the basis for the kinetic mechanism of
Na,K-ATPase (1, 20, 21). Known as the Albers-Post mechanism, it involves these
steps (cyt refers to cytoplasmic, exc refers to extracellular with respect to Na+
and K+):
a. E2[2K] ↔ E12K + ATP ↔ ATP · E1 + 2K+
cyt
b.
c.
d.
e.
f.
+
ATP · E1 + 3Na+
cyt ↔ E1ATP3Na
E1ATP3Na+ ↔ E1-P[3Na] + ADP
E1-P[3Na] ↔ E2-P[2Na] + Na+
exc
E2-P[2Na] + 2K+
↔
E
-P[2K]
+ 2Na+
2
exc
exc
E2-P[2K] ↔ E2[2K] + Pi.
The steps of this ping-pong transport mechanism include (a) ATP acting with low
affinity accelerates inward transport of two K+ ions coupled to the E2[2K] ↔ E1
conformational transition; (b) binding of three Na+ ions at cytoplasmic-oriented
+
sites; (c) Na+
cyt -dependent phosphorylation from ATP and occlusion of three Na
+
ions; (d) outward transport of three Na ions coupled to the E1-P ↔ E2-P conformational transition; (e) binding of two K+ ions at extracellularly oriented sites; and
+
( f ) K+
exc -activated dephosphorylation and occlusion of two K ions. As pointed
out by Jencks (22), an effective ion pump has strict cation specificity of the phosphorylation and dephosphorylation reactions, (steps c and f ), and tight coupling
of the E1/E2 conformational changes to cation movements, (steps a and d ), as well
as features that minimize energy losses via nonproductive pathways. The latter
include the cation occlusion in E1-P[3Na] and E2[2K] and change in specificity
from ADP reactivity of E1-P to water reactivity of E2-P. Other P-type pumps, such
as sarcoplasmic reticulum Ca-ATPase or gastric H,K-ATPase have different cation
specificities of the phosphorylation-dephosphorylation steps and stochiometries
of cation movements, but the basic mechanisms are the same. An explanation of
these mechanisms in terms of molecular structure is the question that needs to be
addressed.
21 Jan 2003
12:40
820
AR
AR177-PH65-34.tex
JORGENSEN
¥
AR177-PH65-34.sgm
HÅKANSSON
¥
LaTeX2e(2002/01/18)
P1: IBC
KARLISH
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
OVERALL STRUCTURE OF Ca- AND Na,K-ATPase
AND THE E1-E2 TRANSITIONS
The general characteristics of a P-type ATPase include a bundle of transmembrane
helices making up the cation transport path and three domains on the cytoplasmic
side of the membrane; the A (actuator), P (phosphorylation), and N (nucleotidebinding) domains. In Ca-ATPase and Na,K-ATPase, the A domain consists of
the N-terminal portion of the polypeptide chain and an insertion between M2
and M3, whereas the P and N domains are located between M4 and M5 (16)
(Figure 1).
Rabbit sarcoplasmic reticulum Ca-ATPase is the only member of the P-type
ATPase family for which an atomic resolution structure is available (16). Much of
the Na,K-ATPase shows sufficient sequence similarity to allow construction of a
homology model, although there are exceptions, notably in the N and P domains.
Key residues involved in ATP binding can be identified in both Ca-ATPase (23, 24)
and Na,K-ATPase (25), but there are several large insertions and deletions in the
N domain and in the first 70 residues of the P domain. The extracellular loops between M1 and M2, and between M7 and M8, contain an insertion in Na,K-ATPase,
and there is a deletion between M3 and M4. Nevertheless, at the extracellular face,
a negatively charged surface area, where the Na+ ions exit and the K+ ions enter,
is preserved in the two enzymes.
Very large movements of the cytoplasmic A, N, and P domains accompany
the E1-E2 transitions in Ca-ATPase, as visualized by comparing the 2.6 Å resolution atomic structure of Ca-ATPase in an E1[2Ca] conformation (1EUL) (16)
with the E2-models (1FQU, 1KJU) (16, 17) (Figure 1). The P domain is rotated
with reorientation of the catalytic site, including the important residues (Asp351,
Thr625, and Asp703), on a plane along the Rossmann folds. Movement of the N
domain comes mainly from rotation of the P domain (17). The A domain rotates
approximately 90◦ about a vertical axis, thus docking the conserved 181-TGES
sequence onto the P domain at the 625-TGD motif, the 701-TGDGVNDS motif,
and the phosphorylated Asp351 in the catalytic site of the E2 form (16, 17).
The N Domain
The N domain contains the ATP-binding site and extends from the phosphorylation
site Asp369, with the 370-KTGTL sequence, to the C-terminal 586-DPPR hinge
(Figure 2). This domain is inserted into the P domain and consists of a fairly
twisted antiparallel eight-stranded β-sheet, with a basic fold known only from the
Ca-ATPase structure (16). The strand topology is the same as found in the eightstranded β-barrels of the lipocalins, which are carriers of hydrophobic compounds.
(26, 27). Two long segments protrude from this β-sheet, one between the outermost
β-strands 3 and 4, and another between strands 6 and 7 (Figure 2). Each of these
protrusions contains a β-hairpin and two or three helices. The binding site for
the ATP adenine base is found in a hydrophobic pocket lined by strands 4, 5, 6,
14 Jan 2003
14:58
AR
AR177-PH65-34.tex
AR177-PH65-34.sgm
LaTeX2e(2002/01/18)
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
Na,K-PUMP STRUCTURE AND FUNCTION
P1: IBC
821
Figure 2 Topological representation of the N and P domains. The positions of some
residues around the ATP-binding site are shown. β-strands are shown as arrows but are
numbered only if they belong to any of the two central β-sheets; α-helices are shown
as cylinders.
and 7 on one side and the N-terminal part of helix Asp443-Cys457 on the other.
Na,K-ATPase has two large deletions relative to the Ca-ATPase structure, one
between strands 1 and 2 and the other on the C-terminal side of helix Asp443Cys457. In the tertiary structure, both deletions are far from the ATP-binding site.
The ATP-Binding Site
The ATP-binding site in the Ca-ATPase structure is detected by bound
TNP-AMP in the N domain (16), but precise interactions between ATP and protein
have not been characterized. The role of the N domain of Na,K-ATPase is also
illustrated by binding of nucleotides to the isolated expressed loop between M4
and M5 with the selectivity of native enzyme, although with much lower affinity
(28). Several residues in the ATP-binding pocket are conserved in Na,K-ATPase
14 Jan 2003
14:58
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
822
AR
AR177-PH65-34.tex
JORGENSEN
¥
AR177-PH65-34.sgm
HÅKANSSON
¥
LaTeX2e(2002/01/18)
P1: IBC
KARLISH
and Ca-ATPase. Lys501 (Lys515 in Ca-ATPase) in the conserved KGA motif is
covalently labeled with fluorescein isothiocyanate (FITC) (29). Lys480 (Lys492)
is the receptor for 8-azido ATP and pyridoxal 50 -diphospho-50 -adenosine (30, 31).
ATP protects against insertion, and labeling abolishes ATPase activity. However,
mutations of Lys515 in Ca-ATPase (32) and Lys480 in Na,K-ATPase (33) have
moderate effects on catalysis. Mutation of Lys480 appears to reduce the affinity
of the Na,K-ATPase for both ATP and phosphate, but it is not essential for ATP
binding or hydrolysis (33). Although insertion of bulky groups in the binding
cavity may prevent binding of ATP, these basic residues may not be essential for
coordination of nucleotide. Hydrogen bonding of Lys501 with Glu446 may be of
importance for the general structure of the ATP site. In Ca-ATPase, competition
against photolabeling of Lys492 (Lys480 in Na,K-ATPase) by [γ -32P]-TNP-8azido-ATP has been exploited to analyze ATP binding (34). The results implicate
Phe487 (Phe475 in Na,K-ATPase), Arg489, and Lys492 (Lys480) (34). Crosslinking of Lys492 and Lys515 also produces strong inhibition of ATPase activity
and phosphorylation in Ca-ATPase (24).
In Na,K-ATPase, Arg544 (Arg560 in Ca-ATPase) is positioned at the mouth of
the ATP-binding pocket near the interface with the P domain and initiates a Walker
B–like sequence (1, 25). Direct ATP binding or phosphorylation assays of mutants
shows that Arg544 contributes 7 to 8 kJ/mol to the change in Gibbs free energy
of ATP binding (11Gb). ATP binding is unaffected in the Arg544Lys mutation,
whereas the affinity for ADP is reduced, suggesting that Arg544 stabilizes the β
or γ phosphate moieties of ATP and aligns the γ phosphate for interaction with
the carboxylate group of Asp369 (25).
Several investigators have proposed that Na,K-ATPase has two ATP sites located either on the same α/β protomer, or on interacting protomers. A review of
this controversial topic is presented in Reference 35 with many original references.
The minimum asymmetric unit in membrane crystals and the minimal functional
unit in detergent solution are unequivocally α/β protomers (36–38), and in our
view, the most recent findings favor a single ATP site mechanism. First and
foremost, the Ca-ATPase structure (16) has only a single ATP-binding pocket
and, insofar as they have been compared, ATP-binding residues of Na,K-ATPase
and Ca,ATPase are similar. The two-ATP site hypothesis is based extensively on
binding of bulky ATP derivatives (39) and modification by reagents such as erythrosine isothiocyanate (Er-ITC) to FITC-modified Na,K-ATPase, which blocks
high-affinity ATP binding (40–42). Recently, pure duck salt gland Na,K-ATPase
fully modified with FITC (6.5 nmol/mg protein) was shown to bind TNP-ADP
and be modified by Er-ITC, although with much lower affinity than native enzyme
(43). Thus the fluorescein label attached to Lys515 hinders binding of ATP and
also TNP-ADP and Er-ITC, but the latter are not absolutely blocked and can still
enter the ATP pocket and bind with a low affinity. Duck salt gland microsomal
Na,K-ATPase is mostly monomeric (α/β). In pure Na,K-ATPase, higher oligomers
are detectable, but the molecular activity is the same as for the monomer (α/β) in
microsomes (44).
14 Jan 2003
14:58
AR
AR177-PH65-34.tex
AR177-PH65-34.sgm
LaTeX2e(2002/01/18)
Na,K-PUMP STRUCTURE AND FUNCTION
P1: IBC
823
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
The P Domain
The phosphorylation domain is homologous to the haloacid dehalogenase (HAD)
family of enzymes for which the structure is known to atomic resolution in the
case of L-2-haloacid halogenase (45), phosphonoacetaldehyde hydrolase (46), and
Methanococcus jannashii phosphoserine phosphatase (47). As anticipated from
the sequence similarities, these proteins share a common fold corresponding to
the phosphorylation domain in the Na,K-ATPase, although the substrate-binding
domain insertions are different from each other and from the N domain of the
ATPases. The fold can be described as a six-stranded parallel β-sheet made up
of two Rossmann folds (βαβαβ), which in Ca-ATPase and Na,K-ATPase are
flanked with one parallel and one antiparallel β-strand (Figure 2). The N domain
is inserted between the second and third strand. Three motifs are arranged next to
each other and outline the active site for binding of Mg2+ and phosphate. The first
of these motifs, 369-DKTGTL in Na,K-ATPase, contains the aspartic acid residue,
which is phosphorylated during catalysis (48–50). The second motif is 610-TGD
on strand 3, C terminal to the N domain. It associates with the phosphorylated
segment because Asp627 in Ca-ATPase is hydrogen bonded to Lys352 and the
Asp627Glu mutation blocks the Ca2+ transport at the E1-P-E2-P transition (32, 51).
The third motif, 708-TGDGVND, terminates the sixth β-strand of the P domain
where Asp710 is required for binding of Mg2+ and phosphorylation (52).
L-2-haloacid halogenase and phosphonoacetaldehyde hydrolase act on relatively small substrates, and Mg2+ is not required for catalysis (45, 46). Similar
to the phosphoserine phosphatase (47), they do not catalyze energy transfer from
phosphate bond ester cleavage or undergo sizeable domain-domain movements as
in P-type ATPases. Their use as a model system for conformational changes of the
ATPases is therefore limited.
More can be learned from the binding of Mg2+ and phosphate in these enzyme structures. In M. jannashii phosphoserine phosphatase (47), the Mg2+ ion is
coordinated to residues homologous to Na,K-ATPase residues Asp710, the main
chain carbonyl oxygen of Thr371, the phosphorylation residue Asp369, and to a
phosphate ion. The other phosphate oxygens interact with (a) Asn713, Lys691 and
the main chain NH group of Gly611, (b) Thr610 and the main chain NH groups
of Lys370, and (c) Thr371. Coordination of the phosphate group is most likely to
be similar to the γ phosphate of the ATP in P-type ATPases. The structure of the
DKT motif is well adapted to bind both a negatively charged phosphate ion and
a positively charged Mg2+ ion. Lys370 initiates a short, one-turn helix, and the
dipole moment is utilized for binding phosphate by the main chain NH groups of
residues Lys370 and Thr371. However, on one side of the helix, the peptide bonds
of residues Thr371-Gly372 and Thr375-Gln376 are flipped, allowing the carbonyl
oxygen of Thr371 to interact with the Mg2+ ion. Thus two antiparallel dipole
moments in this unusual helix are utilized to dock both the positively charged
Mg2+ and the negatively charged phosphate ion. Residues likely to interact with
the α and β phosphate groups toward the binding site on the N domain include
21 Jan 2003
12:41
824
AR
AR177-PH65-34.tex
JORGENSEN
¥
AR177-PH65-34.sgm
HÅKANSSON
¥
LaTeX2e(2002/01/18)
P1: IBC
KARLISH
Lys370, Thr375, and Asp586. The structure of phosphonoacetaldehyde hydrolase
complexed with tungstate is similar to the Mg-phosphate complex in the phosphoserine phosphatase, although Lys691 and Asn713 are replaced by arginine and
serine residues, respectively.
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
The A Domain
The A domain is the smallest cytoplasmic domain of the α-subunit of
Na,K-ATPase, consisting of the N-terminal segment (Gly1-Gln88) plus the loop
between M2 and M3 (Lys205-Glu312). In Ca-ATPase, the cytoplasmic loop forms
a distorted jelly roll structure, and there are two short helices in the 40 residues
of the N-terminal segment (16). In Na,K-ATPase, the N-terminal segment is 33
residues longer and the bond at Lys30 is a well-defined tryptic cleavage site (53).
The tryptic cleavage sequence of Lys30 and Arg438 implies that this part of the
A domain interacts with the N domain in the E2[2K] form (1). The conserved
212-TGES motif is located at an edge of the A domain (16). The critical importance of this loop has been demonstrated in iron cleavage experiments, as it is
cleaved in the E2-P form, but not in E1-P (54).
Exposure of proteolytic cleavage sites in the L2/3 loop in E1 forms and protection in E2 forms are parallel in Na,K-ATPase (1, 55) and Ca-ATPase (1, 56, 57),
suggesting that the conformational movements of the A domain are similar. The
importance of the A domain movements and the L2/3 loop are underlined by the
nature of its connections near the cytoplasmic end of M3 (Figure 3). In Ca-ATPase,
a triangular hydrogen-bonded connection is made between L2/3, helix-1 of S4 and
L6/7, involving Thr247 in 247-TPL of L2/3, Glu340 in 339-VETLG of the P1helix in S4, and Arg822 in 819-RXPK of L6/7 (16). These segments are conserved
in Na,K-ATPase 275-TPI, 357-VETLG, and Arg830 in 827-RXPK of L6/7 and
are modeled in Figure 3, as in Ca-ATPase. The connection to P1-helix of S4 may
function as a lever mechanism to alter the position of M4. L6/7 runs as a collar
around the P domain, and there are several conserved hydrogen bond connections
between L6/7 and S5, i.e., between Asp813 in L6/7 and Asn755 in S5 and between
the main chain carbonyl oxygens of Arg819 and Arg751 in S5. Additional hydrogen bonds and van der Waal interactions between L6/7 and M5 are described for
Ca-ATPase (16).
The Transmembrane Region
There are 10 transmembrane α-helices or 5 pairs of helices inserted from the
cytoplasmic side (37, 57). With the exception of M9-M10, the returning helix
is in close contact with the preceding one in Ca-ATPase, and the four M1-M2,
M3-M4, M5-M6, and M7-M8 helical pairs are sequentially juxtaposed in the
membrane. Three of the helical pairs contribute to cation binding via residues in
M4, M5, M6, and M8 (55, 58–60) (Figure 4a and Figure 5). The M2, M4, and M5
helices are extended to lengths of 41, 26, >27, and 60 Å, with several turns on
the cytoplasmic side where they connect with cytoplasmic domains. The M9-M10
14 Jan 2003
14:58
AR
AR177-PH65-34.tex
AR177-PH65-34.sgm
LaTeX2e(2002/01/18)
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
Na,K-PUMP STRUCTURE AND FUNCTION
P1: IBC
825
pair does not contribute directly to domain anchoring or cation binding, and its two
helices are not packed together but wrap around M8 (16). Determination of the
helical tilts in the bilayer is not straightforward because the membrane structure
is not determined or defined in crystallographic structures. The inclination of the
transmembrane helices in Ca-ATPase has been estimated by electron microscopy
(15) to range from less than 10◦ for M4 and M6 to ∼30◦ for M5 and M9. The
relative orientations of helices varies from almost parallel to an angle close to 50◦
between M5 and M10.
The transmembrane regions of P-type ATPases differ from those of known
ion channels by the absence of an obvious transport path in the form of an open
water-filled channel. Presumably, this reflects the difference between passive and
active transport mechanisms and the necessity of occluding the cations in the ion
pumps. Probing the structure with the Hole software (61) in Figure 4c shows that a
channel with dimensions suitable for passage of Na+ or K+ connects the proposed
entry site for Na+ with site I and II near the center of the membrane region. In the
extracellular half of the bilayer, the putative channel is constricted in a section that
appears to be too narrow for passage of a cation, as could be expected for an E1
conformation.
HELIX PACKING STABILITY AND α-β-SUBUNIT
INTERACTIONS
Because the Ca-ATPase structure or homology models are used extensively to
interpret data on Na,K-ATPase or H,K-ATPase, we need to know how α- and βsubunits interact and whether β-subunits affect organization of the α-subunit. A
similar question arises regarding the γ -subunit or other FXYD proteins. Transmembrane helix arrangement has been predicted to be similar in Ca-ATPase and
Na,K-ATPase (57, 62). However, as discussed below, there may be a significant
difference in the M7-M10 region where α- and β-subunits interact.
The α-Subunit
Several proximities and interactions between α- and β-subunits have been detected
with renal Na,K-ATPase, digested extensively with trypsin to shave off cytoplasmic N, P, and A domains of the α-subunit, which leave well-defined membraneembedded fragments (63–65). The β- and γ -subunits are largely intact. In these
19-kDa membranes, cation occlusion and ouabain binding are preserved whereas
the ATP-binding sites are removed. This shows directly that cation sites are located
within transmembrane segments and communicate with ATP sites via long-range
interactions. Cation occlusion is much less thermally stable in the 19-kDa membrane than in native Na,K-ATPase, although it is strongly protected by Rb+ or Na+
ions or ouabain. (66). Instability can be explained by loss of stabilizing interactions
of L6/7 after the cut at Arg830-Asn831 and removal of the P1 helix in S4. This
fits well with a proposal that L6/7 of Ca-ATPase and its interacting stalk segments
14 Jan 2003
14:58
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
826
AR
AR177-PH65-34.tex
JORGENSEN
¥
AR177-PH65-34.sgm
HÅKANSSON
¥
LaTeX2e(2002/01/18)
P1: IBC
KARLISH
stabilize both phosphorylation (P) and Ca-binding (M) domains, based on effects
of mutations to charged residues (67). The M5/M6 fragment dissociates from digested Na,K-ATPase or H,K-ATPase upon warming at 37◦ C, in the absence of
K(Rb+) ions (68, 69). This phenomenon indicates thermal lability of interactions
between M5/M6 and other segments, notably the M7-M10 segments (70), and
suggests that M5 and M6 are mobile segments in cation transport.
Cu-phenanthroline treatment of detergent-solubilized 19-kDa membranes
causes S-S bridge formation between Mβ (Cys44) and a cysteine in M8 (Cys911
or Cys930), an internal cross-link between M9 (Cys964) and M10 (Cys983) of
the α-subunit, and a cross-link of M1/M2 to M7/M10, which was not identified
(71, 72). A homology model shows that Cys983 and Cys964, which form the internal cross-link, lie at the ends of M9 and M10 but are not close to each other
(Figure 5). The observation (66) that in 19-kDa membranes Cys983 in M10 reacts with cysteine reagents only after dissociation of the M5/M6 fragment is
not self-explanatory because Cys983 already appears to be exposed (Figure 5).
These discrepancies imply that M9 and M10 may not be organized identically
in Ca-ATPase and Na,K-ATPase. In another study o-pthalaldehyde produced two
cross-links, one near the cytoplasmic entrance of M5 and M7, consistent with
the Ca-ATPase structure, and the other located between the extracellular entrance of M8 and the β-subunit, which fits other evidence on the α-β interaction.
(73).
Cysteine modifications and mutations are beginning to provide valuable information. Covalent modification of Cys964 and Cys911 in right-side-out renal
vesicles shows that these residues lie at the extracellular boundaries of M9 and
M8 and are conformationally mobile. Mobility of M9 is also demonstrated with
BIPM-labeled Cys964 (74). Na,K-ATPase was expressed in insect cells, Cys964
and Cys911 were mutated to non-cysteine residues, and residues in putative extracellular loops were mutated to cysteine: Pro118 in L1/2; Thr309Cys in L3/4;
Leu793Cys in L5/6; Leu876Cys in L7/8; and Met973Cys in L9/10. Each of the
mutants reacted with a membrane-impermeant cysteine reagent (MTSEA-biotin),
confirming their location in extracellular loops (74). These results provide a strong
indication for the model with 10 transmembrane segments (37) in the functional
enzyme. Val969Cys and Leu976Cys, located on either side of Met973, were not labeled, suggesting that they lie at the boundaries of M9 and M10. On the homology
model (Figure 5), Val969 and Leu976 are located in the center of the long L9/10,
where cysteines should be exposed to MTSEA-biotin. Thus again, the unexpected
observation raises the possibility of a different arrangement of M9 and M10 in
Na,K-ATPase and Ca-ATPase. A functional cysteine-less Na,K-ATPase has now
been expressed in insect cells (75), providing an important tool for defining helix
packing, which uses cysteine-scanning mutagenesis as applied to transporters such
as lac permease (76).
Covalent labeling using photoactivated hydrophobic reagents detects lipidexposed faces of transmembrane helices. Both α- and β-subunits of renal
Na,K-ATPase are labeled by INA (S-iodonaphtyl-L-azide) (77), TID
14 Jan 2003
14:58
AR
AR177-PH65-34.tex
AR177-PH65-34.sgm
LaTeX2e(2002/01/18)
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
Na,K-PUMP STRUCTURE AND FUNCTION
P1: IBC
827
[3-(trifluromethyl)-3-(m-iodophenyl)diazirine], and phosphatidylcholine-TID analogues (78, 79). Labeling of the α- but not β-subunit is amplified in the E2[K]
compared with the E1Na conformation. Recent work with Torpedo Na,K-ATPase
shows that TIDPC labels the hydrophobic segments M1, M3, M9, and M10 with
no difference in E1 or E2 conformations. M2 and M4-M8 segments were not labeled (80). This labeling pattern is expected for a helix arrangement such as that of
Ca-ATPase, except for the lack of labeling of M2 and M7. Mβ has been hypothesized to conceal M7. TID labels the same fragments as TIDPC and also an M5-M6
fragment; this labeling is enhanced in the E2 conformation. Thus the M5-M6 hairpin is conformationally mobile, whereas there is little conformation-dependent
change of the peripheral segments M1, M3, M9, and M10.
Specific oxidative cleavage catalyzed by bound transition metals provides information on proximity of cleaved polypeptide bonds (81). Fe2+ catalyzes specific
cleavages of either Na,K-ATPase or H,K-ATPase (6, 82, 83) at two sites. Site I
is in the P and A domains (discussed below). Site II is at the membrane-water
interface where Fe2+ mediates cleavages near the entrance of M3 and M1, as does
also a lipid-soluble complex of Cu2+ ions and bathophenanthroline (DPP) (54).
A homology model confirms proximity of M3 and M1 at one point where the
sequence 283-HFIH meets 81-EWVK to form a Fe2+ site II (85). Cleavages near
M3 and M1 are not affected by the E1/E2 conformations. This work indicates a
similar disposition of M3 and M1 in Ca,ATPase, Na,K-ATPase, and H,K-ATPase.
α-, β-, γ -Subunit Interactions
Several excellent reviews describe the role of the β-subunit in stabilizing the
α-subunit and facilitating its routing and insertion into the membrane, as well as
effects on functional properties, notably cation affinities (86–88). The strongest
subunit interactions exist between the extracellular loop L7/8 and the β-subunit
(86–88), but membrane and cytoplasmic sequences also play a part (86, 88). Recent
tryptophan-scanning mutagenesis shows that two faces of Mβ interact with the
α-subunit. Cys46 in the β-subunit interacts with the α-subunit near M8 to which it
can be cross-linked (71, 72). Tyr40 and Tyr44 in the β-subunit interact with other
α-subunit helices, which also affect K+ affinity (89).
Functional studies show that the γ - and α-subunits interact at more than one
position (90), but the interaction sites of α with γ or other FXYD proteins are
largely unknown. An interaction of the γ -subunit with a C-terminal portion of
the α-subunit has been proposed based on thermal denaturation (90). A direct
cross-link of γ to the M7-M10 fragment of 19-kDa membranes has been observed
recently (M. Fuzesi & S.J.D. Karlish, unpublished observation).
A Hypothesis Regarding Helix Packing
The helix packing of renal Na,K-ATPase proposed in the reconstruction based
on electron microscopy (13) assumes that the arrangement of the α-subunit and
Ca-ATPase are the same. Accordingly, Mβ is identified with a peripheral electron
21 Jan 2003
12:42
828
AR
AR177-PH65-34.tex
JORGENSEN
¥
AR177-PH65-34.sgm
HÅKANSSON
¥
LaTeX2e(2002/01/18)
P1: IBC
KARLISH
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
density (Figure 6A). This arrangement does not easily explain an interaction of
α with Mβ at more than one face (89). Also Mβ and M8 may be too far apart
for cross-linking (71). To resolve these inconsistencies, and those discussed above
in relation to L9/10, we propose an alternative arrangement in Figure 6B. Mβ
occupies the position of M10 of Ca-ATPase, whereas M10 is displaced toward
M9. Location of Mβ between M7 and M8 should allow Cys46 to approach M8,
and Tyr40 and Tyr44 to approach M7. M10 (Cys983) should be close enough to
M9 (Cys964) for S-S bridge formation. Finally, γ is close enough to M8 to permit
cross-linking.
THE Na+- AND K+-ION-BINDING SITES IN THE
TRANSMEMBRANE REGION
The passage of cations through the low dielectric of the lipid bilayer is facilitated
by oxygen-containing side chains near the center of the bilayer and by interruption
of helices M4 and M6 in Ca-ATPase (16) and probably helix M5 in Na,K-ATPase
(55). The helix dipoles of interrupted helices may impose a negative potential to
attract the cations, and in the open loops, residues at positions -3 and -4 relative to
prolines can expose main chain carbonyl oxygen groups for cation coordination
(91).
Ca2+ ions may enter from the cytoplasmic side, between M2, M4, and M6,
and be guided via a hydrophilic pathway to oxygen groups of the coordinating
residues oriented toward the cytoplasm, in particular Glu309 and Asp800 (16).
Electrostatic surface potential analysis, Figure 7a,b shows this region and the
corresponding region of Na,K-ATPase as red, negatively charged areas. On the
extracellular side, a negatively charged grove, including Ile788 in Ca-ATPase,
may serve as a cation outlet. The distance between these two putative entrance
areas is not excessive and indicates that they are close to the membrane-water
interface. The center of M4 and M6 are unwound loops with rows of exposed
oxygen atoms, from Pro312-Glu309 in M4 and from Gly801-Asp800 in M6. The
side chains of Glu309, Asn768, Glu771, Asn796, Thr799, Asp800, and Glu908
and several main chain carbonyl oxygens (Val304, Ala305, Ile307) contribute
six coordinating groups for each of two Ca2+ ions (16). In the α-subunit of
Na,K-ATPase, the M5 helix may also be interrupted because the sequence 777IPEITP is similar to the interrupted sequence 307-IPEGLP in M4 of Ca-ATPase
(60). At the exit, there is a ring of oxygens with water molecules for rehydration, but a narrow water-filled access channel or ion well is not apparent
(16).
In Na,K-ATPase, effects of several divalent or trivalent cations (Br-TITU),
which act as Na-like competitive inhibitors of Na+ or Rb+ occlusion, led to
the concept of a cytoplasmic cation entry port, consisting of negatively charged
residues in L6/7 that control access to the occlusion sites (92). Mutations of L6/7 of
Ca-ATPase (Asp813, Asp815, Asp818) strongly reduce Ca+ affinity for activation
of active Ca+ transport, suggesting a similar hypothesis (93). On the other hand,
14 Jan 2003
14:58
AR
AR177-PH65-34.tex
AR177-PH65-34.sgm
LaTeX2e(2002/01/18)
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
Na,K-PUMP STRUCTURE AND FUNCTION
Figure 6 (A) Helix arrangement as in Hebert et al. (2002). (B)
Alternative arrangement in the regions of α-β- and α-γ -subunit
interaction.
P1: IBC
829
14 Jan 2003
14:58
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
830
AR
AR177-PH65-34.tex
JORGENSEN
¥
AR177-PH65-34.sgm
HÅKANSSON
¥
LaTeX2e(2002/01/18)
P1: IBC
KARLISH
in the Ca-ATPase structure, Asp813, Asp815, and Asp818 appear too far apart to
create a Ca+-binding site, implying that the mutations could have indirect effects
(16). However, a peptide corresponding to L6/7 has been shown to form specific
1:1 complexes with Ca+ and La+, and all three residues (Asp813, Asp815, and
Asp818) are required (94). Thus L6/7 may have a dual role, mediating communication and stabilizing the P and M domains as suggested by (16), and also acting
as a port to control entry of Ca2+ ions (94).
Although the charges of Ca2+ and Na+ ions differ, their ionic radii and coordination properties are similar. In general, these hard cations coordinate to oxygen atoms of negatively charged or neutral groups. Two Na+-binding sites in
Na,K-ATPase can therefore be assumed to be homologous to the two Ca2+ sites
found in the Ca-ATPase structure. Sequence alignment and site-directed mutagenesis are also in agreement with the role of residues Glu327, Asn776, Glu779,
Asp804, and Asp808 in Na+ binding (55, 58–60). In site I, one of the Na+ ions is
coordinated to Asn776, Glu779, Thr807, and Gln923. In site II, the second Na+
ion coordinates to Glu327, Asp804, as well as to the carbonyl oxygen groups of
Val322, Ala323, and Val325, whereas the side chain of Asp808 coordinates to both
ions (16, 32) (Figure 4a).
Location of a Third Na+ Site?
The model discussed above identifies two sites for binding of Na+, but the E1
and E1-P forms are known to bind three Na+ ions, and a site for the third Na+
ion has not been identified. One possibility is that a third Na+ ion is bound near
the middle of the membrane bilayer at the level of site I and II in Ca-ATPase.
The coordinating groups might be located in M8 and M9, although mutagenesis
data have not revealed significant contributions to binding of Na+ from Glu953
or Glu954 in M9 (95). Asp925 in M8 may contribute up to 5 kJ/mol to the free
energy of Na+ binding (55, 96), but this residue is positioned in the cytoplasmic
half of the bilayer. In addition, the location of the third Na+ site in the middle of the
membrane, near sites I and II, may not be compatible with the electrogenic binding
of the third Na+ ion with a dielectric coefficient of 0.25 in a neutral binding site or
with the electrogenic release with a dielectric coefficient of 0.7 at the extracellular
surface of the first Na+ ion from the E2-P form. For references to the literature on
electrogenecity of Na+- and K+-transport, see earlier reviews (97–99).
Another possibility for locating a third Na+ site is a cytoplasmic neutral site for
binding of a partially dehydrated Na+ ion. In a model of Na,K-ATPase built on the
backbone of the high-resolution structure of Ca-ATPase, a cavity appears between
M4, M5, and M6 where Ser768 in Na,K-ATPase replaces Phe-760 in Ca-ATPase
(Figure 4b) (K.O. Håkansson & P.L. Jorgensen, submitted for publication). A partially dehydrated Na+ ion can be docked into this site, 9–10 Å on the cytoplasmic
side of the other two Na+ ions in sites I and II. The Na+ ion is coordinated by
the side chain and main chain oxygens of Ser768 and three water molecules and
makes contacts with Thr772, consistent with the selective effect on Na+ binding on
14 Jan 2003
14:58
AR
AR177-PH65-34.tex
AR177-PH65-34.sgm
LaTeX2e(2002/01/18)
Na,K-PUMP STRUCTURE AND FUNCTION
P1: IBC
831
mutation of Thr772 (60). Figure 4b illustrates how binding and partial dehydration
of a Na+ ion in a neutral site prior to binding in site I may explain both the electrogenic binding of the third Na+ (d = 0.25) (d is the dielectric distance explained
in References 97–99) and the electrogenic release of the first Na+ ion (d = 0.7)
(97–99). Charge movements detected by electrochromic dyes are compatible with
a Na+ site in the cytoplasmic membrane leaflet in Na,K-ATPase, whereas there is
no indication of a cytoplasmic Ca2+ site in Ca-ATPase (100).
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
K+-Binding Sites
Structural changes in the cation-binding site during E1-E2 transition cause a switch
in the selectivity from the Na+ ion (D = 1.9 Å) to the somewhat larger K+ ion
(D = 2.7Å) (55). Most of the residues contributing to Na+ binding also bind
K+ in the E2 form (55). The alkali metal ions do not move from one set of sites
to another; rather sites I and II are bifunctional and their coordination geometry
and orientation are rearranged during the E1-E2 transitions. The E2 models of CaATPase, 1FQU and 1KJU, show that the relative positions of M5 and M6 are
roughly the same in E1 and E2, but that helices M4 and M8 move relative to these
two helices. In the case of Na,K-ATPase, this could mean that Glu327 and the
coordinating carbonyl oxygen groups of residues 322, 323, and 325, as well as
Gln923, move away, creating a more spacious cavity for binding of the larger
K+ ion. The E2 structure of Na,K-ATPase, although not of sufficient resolution to
visualize individual helices, demonstrates that the E2 conformation is similar in
the two enzymes (12, 13).
Owing to the high K+ affinity, direct binding of two Rb+ or Tl+ ions per
α-subunit can be measured at equilibrium in recombinant Na,K-ATPase (59, 60)
after expression at high yield in Saccharomyces cerevisiae (101). Each of four
carboxylate groups, Glu327 in M4, Glu779 in M5, Asp804, and Asp808 in M6,
are essential for high-affinity binding Tl+ with Kd K 7–9 µM in the wild-type
enzyme (58, 59). Significant contributions to K+ or Tl+ binding are provided by
Tyr771 (60), Ser775 (60, 102), and Asn776 (55, 60, 103). Parallel assays of K+
binding show that Tl+ and Rb+ are adequate analogues for K+ (55). Mutations
of Asp804 and Asp808, except Asp808Glu, are devoid of Na,K-ATPase activity
(58, 104), whereas partial activities remain after mutations of Glu327 and Glu779
(105, 106). In mutations of Glu327 or Glu779, the E2 form is destabilized and
the conformational equilibrium is poised in favor of the E1 form, with high apparent affinity for ATP and a greatly increased Na-ATPase activity in the absence
of K+(103, 106). Large changes in the currents generated by the recombinant
Na,K-ATPase with Glu779Ala mutation (107) also implicate the 775-SNIPEITP
segment of M5 in Na+/K+ selectivity at the extracellular face.
Cation Specificity
The E1 and E2 conformations differ with respect to orientation and specificity of
cation sites for Na+ and K+. The adjustment of the cation sites from the E2 [2K]
14 Jan 2003
14:58
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
832
AR
AR177-PH65-34.tex
JORGENSEN
¥
AR177-PH65-34.sgm
HÅKANSSON
¥
LaTeX2e(2002/01/18)
P1: IBC
KARLISH
form with specificity for K+ over Na+ to the E1ATP form with specificity for Na+
over K+ is accompanied by changes of both the number and distances of coordinating groups from the bound cations. Twisting or tilting of the intramembrane
helices may adapt the distances between the oxygen atoms of the coordinating
groups of sites I and II (55, 60). In the E2 [2K] form, the dimensions are suitable
for coordinating two dehydrated K+ ions and orientation is toward the extracellular phase. In the E1ATP form, the geometry is suitable for coordination of three
dehydrated or partially dehydrated Na+ ions and the sites are oriented toward the
cytoplasm.
Relative cation specificities of the two conformations can be estimated from apparent affinities at the cytoplasmic and extracellular surfaces. The apparent affinities for Na+ and K+ at the cytoplasmic surface have been estimated to be 0.6 and
10 mM, respectively, corresponding to a 16-fold preference for Na+ over K+, at
the cytoplasmic surface. This requires that Na+ binding is stabilized by 6.4 kJ/mol
more than binding of K+ (60). Specific contributions in the range of 4–7 kJ/mol
from Thr774, Thr776, Ser775, and Tyr771 in the cytoplasmic part of M5 appear to be relevant for the selectivity for Na+ at the cytoplasmic surface (55, 60).
At the extracellular surface, the affinities for Na+ and K+ amount to 600 and
0.2 mM, respectively. The extracellular K/Na-affinity ratio of 3000-fold requires
that K+ binding is 18.4 kJ/mol more stable than binding of Na+. Glu327 contributes
4–8 kJ/mol more to binding of Tl+ or K+ than to binding of Na+, but the contribution of several other residues appears to be required (55, 60).
ENERGY TRANSDUCTION, CYTOPLASMIC DOMAIN
INTERACTIONS, AND THE CATALYTIC CYCLE
A characteristic feature of E1-E2 transitions is that the active site alters its size
owing to domain association and dissociation and its inclination through tilting
or rotation of the P domain. The relationship between the N and P domains does
not change much during the reaction, but association and dissociation of the A
domain is a prominent feature (16, 17). In the first step, ATP binding to the E2[2K]
conformation displaces the A from the N and P domains and stabilizes the E1ATP
conformation. In the E1 or E1-P conformations, the N domain remains docked onto
the P domain with the A domain displaced to one side. In the E2[2K] and E2-P
conformations, the A domain docks onto the P and N domains and interferes with
the tight N to P interactions. (8, 107a). The concepts, depicted in the schematic diagrams in Figure 8, are applicable to both Na,K-ATPase and H,K-ATPase. Similar
models for Ca-ATPase have been proposed recently (17, 108). Presumably, they
apply to all P-type ATPases.
N, P, and A Interactions in E1 and E2 Conformations
In light of the high-resolution structure, the exposure and protection of proteolytic
cleavage sites in Na,K-ATPase and Ca-ATPase can now be interpreted in terms of
14 Jan 2003
14:58
AR
AR177-PH65-34.tex
AR177-PH65-34.sgm
LaTeX2e(2002/01/18)
833
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
Na,K-PUMP STRUCTURE AND FUNCTION
P1: IBC
Figure 8 Schematic diagrams of N, P, and A interactions and ATP-Mg2+- or
Mg2+-binding sites in the different states of the catalytic cycle.
cytoplasmic domain interactions. Association and dissociation of the A domain
from P domain can be monitored by tryptic and chymotryptic cleavage (1) or
tryptic, V8, and proteinase-K cleavages of Ca-ATPase (56, 57, 108, 109). Dissociation of the A domain from the P domain in E1[3Na] or E1-P[3Na] forms of
Na,K-ATPase exposes the bond at Leu266 (Cl) in the L2/3 loop to chymotrypsin
(1, 110). Similarly, sites for tryptic, V8, and proteinase-K are exposed in the E1
[2Ca] form of Ca-ATPase (56, 57).
14 Jan 2003
14:58
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
834
AR
AR177-PH65-34.tex
JORGENSEN
¥
AR177-PH65-34.sgm
HÅKANSSON
¥
LaTeX2e(2002/01/18)
P1: IBC
KARLISH
Tryptic cleavage patterns reflect interaction between the A domain and the
N domain in the E2[2K] or E2-P[2Na] forms. Trypsin cleaves the α-subunit at
Arg438 (T1) in the N domain and subsequently at Lys30 in the A domain. Loss
of Na,K-ATPase activity is linear and associated with cleavage at Arg438. The
biphasic inactivation of K phosphatase has a typical delay indicating that cleavage
at Arg438 precedes cleavage at Lys30 (1, 111). Arg438 is exposed at the surface of
the N domain close to a loop pointing toward the A domain. High-affinity binding
of ATP interrupts this interaction and exposes the T2 site (K30) to rapid tryptic
cleavage (111). The cleavage patterns are compatible with an interaction between
the 438-RAVAGDA loop in the N domain and the segment around Lys30 in the
A domain in the E2 forms of Na,K-ATPase. The functional significance of these
connections is apparent from the changes in catalytic properties because cleavage at
Lys30 strongly reduces the rate of the E1-P to E2-P conversion, stabilizing the E1
conformations three- to fourfold (112).
Rotation of the A domain accompanying the E2 to E1 transition exposes Lys 30
(T2) in the N terminus to rapid cleavage, whereas Arg 262 (T3) in the exposed L2/3
loop is cleaved 30-fold more slowly, resulting in biphasic inactivation of Na,KATPase in the E1 conformations (1). Synergistic effects on function of a truncation
at M32 and a G233K mutation demonstrate the interplay of these segments in the
A domain (113).
Changes in N, P, and A domain interactions also explain effects of K+ or Na+
ions on covalent labeling of Lys501 by various reagents (114) or cross-linking
of Lys501 to Lys480 (115). ATP or low concentrations of K+ ions protect the
enzyme, whereas Na+ ions accelerate inactivation. Protection by ATP is expected
because Lys501 and Lys480 are contact residues. The cation effects are explained
most simply by altered accessibility of Lys501 and Lys480 to the reagents. K+
stabilizes E2[2K] with a closed organization of the N, P, and A domains. Na+ ions
stabilize E1Na with the open N, P, and A domain organization.
The most direct evidence for P and A domain interactions has come from Fe2+catalyzed cleavages of Na,K-ATPase or H,K-ATPase (116–118) α-subunits. In the
E2[2K] form, Fe2+ bound at site I catalyzes splits in the conserved sequences of
the A domain (at 214-ESE) and P domain (near 367-CSDK, near 608-MVTGD,
and at 712-VNDS). The observations led to the prediction that the A and P domains interact near these sequences. In E1 conformations, cleavages at site I are
absent, although those at Fe site II remain. This can be explained by separation
of the cytoplasmic domains so that Fe site I no longer exists. These inferences
are confirmed by the Ca-ATPase structure in the E1 form, or models in the E2
conformation which indeed demonstrate proximity of A to P and also A to N domain interactions (16, 17). Glu214 in the TGES sequence appears to be the linchpin of the A to P domain interaction because it is hydrogen bonded to Asp369,
Thr371, Lys370, Thr610, and Gly611. Suppression of different Fe2+-catalyzed
cleavages within site I by Pi or Pi, Mg2+, ouabain, and vanadate/Mg2+ suggest that
Pi interacts with residues in 608-MVTGD and 367-CSDK and Mg2+ with 212TGES and 712-VNDS in the E2-P conformation (116). These predictions fit well
with the structure of phosphoserine phosphatase (47) and phosphonoacetaldehyde
14 Jan 2003
14:58
AR
AR177-PH65-34.tex
AR177-PH65-34.sgm
LaTeX2e(2002/01/18)
Na,K-PUMP STRUCTURE AND FUNCTION
P1: IBC
835
hydrolase crystallized in the presence of Mg2+ and PO3−
4 ions (46) or the phos2−
phate analogues, BF−
3 and WO4 (45), except that involvement of 212-TGES cannot be tested for the HAD proteins, which lack an A domain and the homologous
residues.
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
Binding of ATP or MgATP to E1 (N ↔P Interactions)
and E2(K) (N ↔A Interactions)
In the initial step of the Na,K-ATPase reaction cycle, binding of ATP to the E2[2K]
conformation with low affinity (Km ∼ 0.2–0.4 mM) accelerates the E2[2K]-E1(2K)
transition, accompanied by release of K+ at the cytoplasmic surface. Mg2+ is
not required for this effect of ATP. ATP is bound with high affinity in the E1
conformation (Kd ∼ 30–100 nM) (1, 7) stabilizing the E1ATP form by comparison
with the E2[2K] form.
In Na,K-ATPase, binding of free ATP to E1 conformations in the absence of
Mg2+ elicits strong electrostatic repulsion between the γ phosphate and the negative charges at the surface of the P domain. Removal of the charge of the carboxylate
groups in the Asp369Ala mutation increases the equilibrium-binding affinity for
ATP 30- to 40-fold (119). Substitution of Ala for Asp710 causes a more moderate
two- to threefold increase of the ATP affinity (52). The apparent dissociation conapp
app
stant for wild-type Na,K-ATPase of K d = 39 nM is reduced to K d = 13 nM
app
in Asp710Ala, and in the Asp369Ala mutation, the K d is only 1.8 nM. The data
allow estimation of the free energy required to overcome the electrostatic interaction between the γ phosphate of ATP and the carboxylate groups to 2 kJ/mol for
Asp710 and 7.9 kJ/mol for Asp369.
The strong electrostatic interaction of the γ phosphate of the tightly bound ATP
with the negative charges of Asp369 and Asp710 at the surface of the P domain in
Na,K-ATPase shows that the N and P domains must be closed in the ATP-E1[3Na]
form, even in the absence of Mg2+ ions (52, 119). As expected, coordination of
Mg2+ in the negatively charged surface of the P domain reduces the electrostatic
repulsion of the phosphate groups and thus facilitates ATP binding. In presence
of Mg2+, the curve for ATP dependence of phosphorylation has a 0.5 K value of
12 nM ATP (25). This is at least threefold lower than the Kd = 38 ± 5 nM for
binding of free ATP at equilibrium in the absence of Mg2+. An important function
of the negative charges of Asp369 and Asp710 could be to counterbalance the very
high intrinsic binding affinity of ATP for Na,K-ATPase. As proposed by Jencks
(22), high-energy intermediates are likely to act as barriers to rapid turnover of
transport systems.
In order to explain ATP binding, a model of Ca-ATPase has been proposed in
which rotation of the N domain about flexible hinges at Asn359 and Arg604 brings
N and P domains into alignment (17). The model also explains cross-linking of
Lys492 (N domain) and Arg678 (P domain), which come into close proximity
(120). This implies that the N to P domain interaction may occur even without
ATP or ATP-Mg, although binding of the nucleotides should stabilize closure. The
fact that Na,K-ATPase binds free ATP with a high affinity in the E1 state whereas
14 Jan 2003
14:58
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
836
AR
AR177-PH65-34.tex
JORGENSEN
¥
AR177-PH65-34.sgm
HÅKANSSON
¥
LaTeX2e(2002/01/18)
P1: IBC
KARLISH
Ca-ATPase binds only MgATP (121) may be explained if the N to P interaction is
much stronger for Na,K-ATPase than for Ca-ATPase.
The complex of ATP or the non-phosphorylating analogues AMP-PNP with
Fe2+ act as affinity cleavage agents of Na,K-ATPase or H,K-ATPase (54, 83).
Fe2+ substitutes for Mg2+ ions in catalyzing ATPase activity and phosphorylation
(122). Thus the cleavages reveal points of contact of the Fe2+ or Mg2+ ions. In
E1 conformations, ATP-Fe or AMPPNP-Fe catalyzes two cleavages at and near
712-VNDS in the P domain and near 440-VAGDA in the N domain. The A domain
is not in contact with bound ATP-Fe2+ and is not cleaved. The data imply that
Mg2+ bound to ATP is in close proximity to both 708-TGDGVNDS in the P
domain, presumably Asp710 and Asp714, and 440-VAGDASE in the N domain.
This predicted N to P closure provides a testable constraint for model building.
Using an approach similar to that in Reference 17, N and P domains have been
brought into proximity by tilting the N domain about 80◦.(107a). Asp703 (the
homologue of Asp710 in Na,K-ATPase) and Glu439 (the homologue of Asp443
in 440-VAGDASE) can be brought within 4 Å of each other. With the ATP-Mg2+
docked, the model shows nucleotide interacting with Lys515, Lys492, Phe487,
Arg560, and phosphates are close to Asp351, Thr353, Thr625, Lys684, Asp703,
and Asp707, consistent with the known facts. The Mg2+ ion is close to both
Asp703 and Glu439. Thus cleavage and modeling fit well with the mutation data
in identifying of Asp710 in Mg+ binding and point to Asp443 as an additional
candidate.
The models of Ca-ATPase in the E2 conformation demonstrate N to A domain
interactions (16, 17). Specifically, Glu439 comes within 4 Å of Ser186 or 3 Å of
Arg174 (A), and residues Glu486 and Arg489 (N) interact with Ser170-Thr171
and Thr191-Glu192 (A). Recently, ATP-Fe2+ complex-mediated cleavages in the
E2[2Rb] conformation have been detected near 212-TGES (A domain), near 440VAGDA, and at a site between residues 460 and 490 (both in the N domain) (107a)
that demonstrates directly the N to A domain interaction.
Comparison of the models for E2 (1KJU)- and E1-bound ATP can, at least
in part, explain the difference in affinity of ATP in E1 and E2 conformations.
In E1, ATP is bound to residues in both N and P domains, whereas in the E2
conformation, the A domain interferes with the close N and P domain interaction.
The low ATP affinity could be explained by the lack of accessibility to Asp351,
Asp703, Asp707, Thr353, and Lys684 side chains within the P domain when these
residues are completely overlaid by the A domain in the E2 conformation. In the
E2 model, Glu183 appears to provide key contacts between the A to P domains
by H-bonding to Asp351, Lys352, Thr353, Thr625, and Gly626, several of which
(Lys352, Thr353, Thr625) are predicted to be in direct contact with ATP in the E1
conformation. In addition, the A domain hinders access of ATP to Lys684. Thus in
essence, stabilization of E1 by ATP can be understood as the result of competition
between the ATP and the A domain for contact residues in the P domain (107a). In
this way, binding of ATP to E2[2K] with a low affinity induces the large outward
movement of the A domain during the E2 → E1 transition and brings the P domain
back into close proximity with the N domain (17).
14 Jan 2003
14:58
AR
AR177-PH65-34.tex
AR177-PH65-34.sgm
LaTeX2e(2002/01/18)
Na,K-PUMP STRUCTURE AND FUNCTION
P1: IBC
837
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
Mg2+ Binding and Formation of the MgE1-P[3Na] Complex
In the 1KJU model of the E2 conformation of sarcoplasmic reticulum Ca-ATPase
(17), the transition from E2 to E1 involves a 53◦ rotation about an oblique axis
of the P domain, which carries the N domain with it in addition to the movement
of the A domain by 90◦ about a vertical axis. These motions can render the surface of
the Rossmann folds with Asp369 and Asp710 more reactive toward binding of ATP
and Mg2+ in the E1 conformation, but the transfer of the γ phosphate to Asp369
occurs only after binding of three Na+ ions in the sites on the transmembrane
segments.
Mg2+ is essential for all phosphoryl transfer reactions and for the E1-P to E2-P
conversion in Na,K-ATPase. To determine binding constants for Mg2+ binding it
is important to stabilize either the E1-P or the E2-P conformations. In pig kidney
Na,K-ATPase, the E1-P-E2-P equilibrium is poised heavily in favor of the E2-P,
and oligomycin is required to stabilize the E1-P form (1). The high affinity of the
MgE2-P complex for ouabain provides a convenient assay of Mg2+ binding in the
E2-P, but that conformation can be distinct from the MgE2-P[2Na] intermediate in
the catalytic cycle.
Analysis of mutations in the 708-TGDGVND loop shows that the Na,K-ATPase
activity is severely reduced in mutations of Thr708, Asp710, Asn713, and Asp714
(52). In plots of Na+-dependent phosphorylation versus free Mg2+ the apparent
Mg2+ affinity, [Mg2+]0.5, for wild-type is 24 ± 5 µM. A 27-fold reduction of Mg2+
affinity is observed for the Asp710Asn mutation, whereas phosphorylation is almost abolished for the Asp710Ala mutation. Removal of the carboxamide group
of Asn713 causes a moderate 4-fold reduction of Mg2+ affinity (52). Coordination
of Mg2+ to Asp710 can establish a link via the γ phosphate to Asp369. The role
of Asp710 in this scheme is to stabilize Mg2+ in a position that facilitates nucleophilic attack of Asp369 to form the acyl phosphate intermediate (123), and Mg2+
shields the negative charges of the phosphate moiety and the two carboxylates.
Another contribution to coordination of Mg2+ is the main chain carbonyl oxygen
of Thr-371 in the 369-DKTGGT segment of the phosphorylated residue (Thr353 in
Ca-ATPase); however, the side chain hydroxyl of Thr371 (Thr-353) may participate in interactions with nucleotide and phosphate (124).
Because ATP-Fe2+ serves as a substrate, the enzyme becomes phosphorylated
in the presence of ATP-Fe2+ and Na+ ions (83). In the presence of oligomycin,
the E1-P form is stabilized. In this state, bound Fe2+ catalyzes exactly the same
cleavages as in the E1 state (54). Thus the N to P interaction and Fe2+ or Mg2+
binding in 712-VNDS and near 440-VAGDA is assumed to be unchanged.
Transition from E1-P to E2-P (A to P Interaction)
During the transition to E2, the conserved 212-TGES loop engages with the surface
of the P domain. Mutagenesis studies have not yet identified residues ligating Mg2+
in the E2-P conformation, but the Ca-ATPase structure suggests alternative ligating
loops in the A domain.
14 Jan 2003
14:58
838
AR
AR177-PH65-34.tex
JORGENSEN
¥
AR177-PH65-34.sgm
HÅKANSSON
¥
LaTeX2e(2002/01/18)
P1: IBC
KARLISH
Upon phosphorylation of Na,K-ATPase (ATP-Fe2+, Na+) or H,K-ATPase (ATPFe , Tris) to E2-P, Fe2+ is tightly bound and a major cleavage occurs near
212-TGES in the A domain, whereas the cleavages near 712-VNDS are less prominent (54). This is indicative of the A to P domain interaction. In addition, the N to P
domain interaction is relaxed because the cleavage near 440-VAGDA (N domain)
is not seen in E2-P. The shift of Mg2+ coordination to the 212-TGES segment could
be important for forming a bridge to stabilize interactions between the P and A
domains, and it may also have a crucial functional significance. The active site
contains the Asp369 bound with phosphate, Mg2+ ions bound to the phosphate,
and the 212-TGES sequence. A change in geometry of ligands around the phosphate and the more hydrophobic environment produced by the domain closure may
be required for inducing reactivity of E2-P to water, compared with the reactivity
of E1-P to ADP. Mg2+ ions interact less strongly with Asp710, although Asp710,
Asp714, and Asn713 are still in close proximity.
Mutagenesis results imply that in E2-P the 708-TGDGVND loop is not required
for Mg2+ recognition but rather for stabilization of the transition state and for K+
stimulation of acyl-phosphate bond hydrolysis in the E2-P form. Asp710 may
contribute directly to coordination of pentacovalent vanadate in the transition state
(52). Removal of the Asp710 also interferes with transmission of the structural
change to the cation sites and ouabain-binding domain at the extracellular surface.
The Asp710Ala mutation causes a threefold reduction of the affinity for binding
of Tl+ (52). In the forward reaction, this corresponds to a role of Asp710 in
transmission of the K+ stimulation of phosphoenzyme hydrolysis from M5 to the
P domain. Overall, the effects of mutations in the 708-TGDGVND loop on binding
of Mg2+ in E1-P and E2-P forms complement the results of Fe2+-catalyzed cleavage.
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
2+
ENERGY TRANSDUCTION, LONG-RANGE
INTERACTIONS BETWEEN CYTOPLASMIC
AND MEMBRANE DOMAINS
ATP binding, phosphorylation, and cytoplasmic domain movements associated
with E1 ↔ E2 transitions are tightly coupled to the vectorial processes in the intramembrane domain via long-range structural changes. The E2[2K] → E1(2K)
transition is associated with an increase in binding energy of ATP, which drives K+
transport into the cell. The E1-P → E2-P transition in the P domain is coupled to
extrusion of Na+ from the cell. These long-range effects could be mediated directly
via the P domain and S4 and S5 or by an indirect route via the large movements
of the A domain and the loops (L2/3 and L6/7) connecting to the transmembrane
segments. The evidence for either path is not decisive, and it is likely that both
mechanisms could apply.
ATP-Driven K+ Uptake
The ATP-K+ antagonism is unique to Na,K-ATPase because this is the only
cation pump that possesses high-affinity binding sites for both ATP and K+. The
14 Jan 2003
14:58
AR
AR177-PH65-34.tex
AR177-PH65-34.sgm
LaTeX2e(2002/01/18)
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
Na,K-PUMP STRUCTURE AND FUNCTION
P1: IBC
839
structural adaptation following conversion of the E2[2K] → E1 transition driven
by high-affinity binding of ATP elicits an outward rotation of the A domain. Comparison of E1 and E2 structures (Figure 1) suggests that there is a particularly
large rearrangement of N domain structure in the region of the T1 cleavage site at
the 438-RAVAGDA loop, a known site of contact between the N and A domains
(8). The large outward movement of the A domain alters the position of the L2/3
loop and transmits the movement via the triangular hydrogen-bonded connection
to the P1-helix of S4 and the L6/7 loop. The significance of these connections
for energy transfer is apparent from the consequences of chymotryptic cleavage
at Leu266 (1). This split eliminates the normal ATP-K antagonism, preventing
ATP-driven uptake of K+, normal displacement of occluded Rb+ by 1 mM ATP,
and displacement of ATP binding by K+ ions, so that ATP is bound with high
affinity in both NaCl and KCl (110). These results show that there is only a very
limited contribution of direct energy transfer from the ATP-binding site in the
N domain to the release of K+ ions from the binding sites to the cytoplasmic
surface.
E1-P to E2-P Transitions and Na+ Movements
It is also a question whether the long-range conformational transitions coupling the
conversion of E1-P[3Na] → E2-P[2Na] to extrusion of Na+ are transmitted directly
or indirectly to the cation sites. With the architecture of the P domain in mind,
direct transmission via S4 and S5 to M4 and M5 is clearly an option, but there
is considerable evidence for indirect transmission via the A domain and the L2/3
loop and its connections to L6/7, M4, M5, and M6 (8). Chymotryptic cleavage
at Leu266, which causes interruption of the L2/3 loop (Figure 3), blocks Na+
translocation and stabilizes the protein in the E1-P[3Na] form (110), thus allowing
the first demonstration of occlusion of Na+ (125). Similar to proteolytic splits,
the mutations Gly263Ala and Arg264Ala (rat α1) stabilize the E1 conformation
(126). In Ca-ATPase, a number of residues in L2/3 are crucial for the E1-P to
E2-P transformation. They are located in the portion of L2/3 that connects the
A domain to transmembrane segment M3 (127) or in S4, the stalk that links the
P domain to M4 (128). Mutations in the P1-helix also prevent transition to the
E2-P form (129), thus underlining the importance of the triangular connection
from L2/3 to the P1-helix and to L6/7. Additional hydrogen bonds and van der
Waals interactions between L6/7 and M5 are described in the structure of CaATPase (16). Structural changes can also be transmitted from the N-terminal part
of the A domain through the helix loop connecting to M2. Both transmembrane
segments M2 and S2, the helix loop connecting to the A domain, are in close
contact with M4 and its extension S4.
A direct transmission route is favored by two sets of information. First, a role
of the S4 segment is implied by observations that mutations of several conserved
residues of Ca-ATPase S4 (T316-356) (124), or an analogous segment of S. cerevisiae H-ATPase (129), stabilize E1 conformations. Second, mutations of Asp369
or Asp710 to neutral residues stabilize E2 conformations, and double mutations
14 Jan 2003
14:58
840
AR
AR177-PH65-34.tex
JORGENSEN
¥
AR177-PH65-34.sgm
HÅKANSSON
¥
LaTeX2e(2002/01/18)
P1: IBC
KARLISH
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
(such as Asp369Ala, Asp710Asn) do not give greater effects (52, 119). Asp369
and Asp710 need to be close in E1 in order to bind the ATP-Mg2+ complex. One
can hypothesize that electrostatic repulsion between Asp369 and Asp710 stabilizes the E1 conformation, but when these are neutralized in E1-P-Mg2+ through Pi
and Mg2+ binding, interactions stabilizing E2-P predominate, notably those of the
212-TGES loop in A with N and P domains. The resulting change in Mg2+ ligation
and environment in E2-P makes it reactive with water, and the altered disposition of
Asp369 and Asp710 may trigger movement of S4 and S5 and the transmembrane
segments.
Na+-Dependent Phosphorylation and
K+-Activated Dephosphorylation
In contrast to the ion translocation reactions, long-range structural changes from
the cation sites to the P domain appear to be communicated exclusively by a direct
mechanism. Activation of phosphorylation from ATP by binding of the third Na+
ion involves nucleophilic attack of Asp369 on the γ phosphate of ATP, whereas
binding of two K+ ions to E2-P[2K] stimulates nucleophilic attack by water on the
Asp369–C–O–P bond and hydrolysis of phosphoenzyme. These cation-specific
signals are transmitted directly to the P domain without altering the association
with the A or N domains.
The signal for K+ activation of E2-P hydrolysis appears to be transmitted directly through M5 and S5 to the active site. This route of transmission has been
proposed after analysis of a series of mutations of residues in S5 of Ca-ATPase
(130). The mutations interfere with transmission of conformational changes from
the cation-binding sites in the membrane domain to the P domain controlling dephosphorylation. Arg751 (Arg758 in Na,K-ATPase) in S5 is hydrogen bonded to
L6/7 (Figure 3), and this residue is important for the structural integrity of the enzyme. Through this and other connections in the network of hydrogen bonds and
van der Waals interactions of L6/7, the structural change can also be transmitted
from K+ coordinating residues in M4 and M5 to the P domain without involving
the A domain. Alanine substitution of Asp369, Asp710, or Asn713 eliminates the
K+-stimulated p-nitrophenylphosphate hydrolysis. The mutations have dramatic
and specific effects, as removal of the carboxylate group of Asp710 eliminates
vanadate binding and removal of the carboxamide of Asn713 eliminates the phosphate dependence of ouabain binding (52). The 708-TGDGVND loop is therefore
required for stabilization of the transition state and for the dephosphorylation
reaction.
A Na+-specific conformational change in the N domain has been detected by
fluorescence (131). The analysis shows that the third Na+ ion binds to the neutral
site only after binding of two Na+ ions and triggers Na-dependent phosphorylation
presumably by inducing correct alignment of the γ phosphate of ATP with the
Asp369. Effects of the Na+ ions in E1 could also be transmitted by the direct route
via M5 and S5.
14 Jan 2003
14:58
AR
AR177-PH65-34.tex
AR177-PH65-34.sgm
LaTeX2e(2002/01/18)
Na,K-PUMP STRUCTURE AND FUNCTION
P1: IBC
841
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
CONCLUSIONS AND PERSPECTIVES
This review demonstrates the power of homology models of the α-subunit of Na,KATPase, based on the high-resolution structure of Ca-ATPase (16), for analysis of
structure-function relationships and molecular mechanisms of the conversion of
the free energy of ATP hydrolysis to active transport of Na+ and K+ ions. Information from site-directed mutagenesis and cleavage experiments has been combined
with molecular modeling, which allows us to discuss in detail the events in the
active sites, including binding of ATP and Mg2+ and the transfer of energy to
the cation sites for each reaction of the catalytic cycle. The progress in understanding the structure and function of the α-subunit underlines the gaps in our
knowledge of the structure and subunit interactions of the β- and γ -subunits,
CHIF (corticosteroid-induced factor), and other members of the family of small
ion transport regulators of the FXYD family, which are subunits of the Na,Kpump. Coexpression of α1β1-subunits with the γ -subunit or CHIF regulates the
activity of the Na,K-pump (132) in the medullary thick ascending limb of Henle
or inner medullary collecting duct. These nephron segments are important for
the regulation of water and salt excretion. High-resolution structural information
about these Na,K-pump protein complexes is essential for solving the molecular
mechanism of this regulation and is therefore a question of primary physiological
and pharmacological significance.
ACKNOWLEDGMENTS
Work in the authors’ laboratories is supported by grants from the Carlsberg Foundation (KOH), the Danish Natural Science Research Council (PLJ and KOH), and
the Israel Science Foundation (SJK).
The Annual Review of Physiology is online at http://physiol.annualreviews.org
LITERATURE CITED
1. Jorgensen PL, Andersen JP. 1988. Structural basis for E1-E2 conformational
transitions in Na,K-pump and Ca-pump
proteins. J. Membr. Biol. 103:95–120
2. Kuster B, Shainskaya A, Pu HX, Goldshleger R, Blostein R, et al. 2000. A
new variant of the gamma subunit of renal Na,K-ATPase. Identification by mass
spectrometry, antibody binding, and expression in cultured cells. J. Biol. Chem.
275:18441–46
3. Sweadner KJ, Rael E. 2000. The FXYD
gene family of small ion transport regu-
lators or channels: cDNA sequence, protein signature sequence, and expression.
Genomics 68:41–56
4. Lingrel JB, Orlowski J, Shull MM, Price
EM. 1990. Molecular genetics of Na,KATPase. Prog. Nucleic Acid Res. Mol.
Biol. 38:37–89
5. Axelsen KB, Palmgren MG. 1998. Evolution of substrate specificities in the
P-type ATPase superfamily. J. Mol. Evol.
46:84–101
6. Goldshleger R, Patchornik G, Shimon
MB, Tal DM, Post RL, Karlish SJ.
14 Jan 2003
14:58
842
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
7.
8.
9.
10.
11.
12.
13.
14.
15.
AR
AR177-PH65-34.tex
JORGENSEN
¥
AR177-PH65-34.sgm
HÅKANSSON
¥
2001. Structural organization and energy
transduction mechanism of Na+,K+ATPase studied with transition metalcatalyzed oxidative cleavage. J. Bioenerg. Biomembr. 33(5):387–99
Goldshleger R, Karlish SJ. 1999. The energy transduction mechanism of Na,KATPase studied with iron-catalyzed oxidative cleavage. J. Biol. Chem. 274:
16213–21
Jorgensen PL, Jorgensen JR, Pedersen PA. 2001. Role of conserved
TGDGVND-loop in Mg2+ binding,
phosphorylation, and energy transfer in
Na,K-ATPase. J. Bioenerg. Biomembr.
33:367–77
Deguchi N, Jorgensen PL, Maunsbach
AB. 1977. Ultrastructure of the sodium
pump. Comparison of thin sectioning,
negative staining, and freeze-fracture
of purified, membrane-bound Na+,K+ATPase. J. Cell Biol. 75:619–34
Skriver E, Maunsbach AB, Jorgensen
PL. 1981. Formation of two-dimensional
crystals in pure membrane-bound Na+,
K+-ATPase. FEBS Lett. 131:219–22
Maunsbach AB, Skriver E, Hebert H.
1991. Two-dimensional crystals and
three-dimensional structure of Na,KATPase analyzed by electron microscopy. Soc. Gen. Physiol. Ser. 46:
159–72
Rice WJ, Young HS, Martin DW,
Sachs JR, Stokes DL. 2001. Structure
of Na+,K+-ATPase at 11-Å resolution:
comparison with Ca2+-ATPase in E1 and
E2 states. Biophys. J. 80:2187–97
Hebert H, Purhonen P, Vorum H, Thomsen K, Maunsbach AB. 2001. Threedimensional structure of renal Na,KATPase from cryo-electron microscopy
of two-dimensional crystals. J. Mol. Biol.
314:479–94
Taylor K, Dux L, Martonosi A. 1984.
Structure of the vanadate-induced crystals of sarcoplasmic reticulum Ca2+ATPase. J. Mol. Biol. 174:193–204
Zhang P, Toyoshima C, Yonekura K,
LaTeX2e(2002/01/18)
P1: IBC
KARLISH
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
Green NM, Stokes DL. 1998. Structure
of the calcium pump from sarcoplasmic reticulum at 8-Å resolution. Nature
392:1835–39
Toyoshima C, Nakasako M, Nomura H,
Ogawa H. 2000. Crystal structure of the
calcium pump of sarcoplasmic reticulum
at 2.6 Å resolution. Nature 405:647–55
Xu C, Rice WJ, He W, Stokes DL.
2002. A structural model for the catalytic
cycle of Ca(2+)-ATPase. J. Mol. Biol.
316:201–11
Post RL, Kume S. 1973. Evidence for an
aspartyl phosphate residue at the active
site of sodium and potassium ion transport adenosine triphosphatase. J. Biol.
Chem. 248:6993–7000
Post RL, Hegyvary C, Kume S. 1972.
Activation by adenosine triphosphate in
the phosphorylation kinetics of sodium
and potassium ion transport adenosine triphosphatase. J. Biol. Chem. 247:
6530–40
Glynn IM, Karlish SJ. 1990. Occluded
cations in active transport. Annu. Rev.
Biochem. 59:171–205
Glynn IM. 1993. All hands to the sodium
pump. J. Physiol. 462:1–30
Jencks WP. 1989. Utilization of binding energy and coupling rules for active
transport and other coupled vectorial
processes. Methods Enzymol. 171:145–
64
McIntosh DB, Woolley DG, Vilsen B,
Andersen JP. 1996. Mutagenesis of segment 487Phe-Ser-Arg-Asp-Arg-Lys492
of sarcoplasmic reticulum Ca2+-ATPase
produces pumps defective in ATP binding. J. Biol. Chem. 271:25778–89
Hua S, Ma H, Lewis D, Inesi G,
Toyoshima C. 2002. Functional role of
“N” (nucleotide) and “P” (phosphorylation) domain interactions in the sarcoplasmic reticulum (SERCA) ATPase.
Biochemistry 41:2264–72
Jacobsen MD, Pedersen PA, Jorgensen
PL. 2002. Importance of Na,K-ATPase
residue alpha 1-Arg544 in the segment
14 Jan 2003
14:58
AR
AR177-PH65-34.tex
AR177-PH65-34.sgm
LaTeX2e(2002/01/18)
Na,K-PUMP STRUCTURE AND FUNCTION
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
26.
27.
28.
29.
30.
31.
32.
33.
Arg544-Asp567 for high-affinity binding of ATP, ADP, or MgATP. Biochemistry 41:1451–56
Francischetti IM, Andersen JF, Ribeiro
JM. 2002. Biochemical and functional
characterization of recombinant rhodnius prolixus platelet aggregation inhibitor 1 as a novel lipocalin with high
affinity for adenosine diphosphate and
other adenine nucleotides. Biochemistry
41:3810–18
Flower DR, North AC, Sansom CE.
2000. The lipocalin protein family: structural and sequence overview. Biochim.
Biophys. Acta 1482:9–24
Gatto C, Wang AX, Kaplan JH. 1998.
The M4M5 cytoplasmic loop of the
Na,K-ATPase, overexpressed in Escherichia coli, binds nucleoside triphosphates with the same selectivity as the
intact native protein. J. Biol. Chem.
273:10578–85
Karlish SJ. 1980. Characterization of
conformational changes in (Na,K) ATPase labeled with fluorescein at the active
site. J. Bioenerg. Biomembr. 12:111–36
Hinz HR, Kirley TL. 1990. Lysine 480
is an essential residue in the putative
ATP site of lamb kidney (Na,K)-ATPase.
Identification of the pyridoxal 50 diphospho-50 -adenosine and pyridoxal
phosphate reactive residue. J. Biol.
Chem. 265:10260–65
Tran CM, Scheiner-Bobis G, Schoner
W, Farley RA. 1994. Identification of an
amino acid in the ATP binding site of
Na+,K+-ATPase after photochemical labeling with 8-azido-ATP. Biochemistry
33:4140–47
Andersen JP. 1995. Dissection of the
functional domains of the sarcoplasmic reticulum Ca(2+)-ATPase by sitedirected mutagenesis. Biosci. Rep. 15:
243–61
Wang K, Farley RA. 1992. Lysine 480 is
not an essential residue for ATP binding
or hydrolysis by Na,K-ATPase. J. Biol.
Chem. 267:3577–80
P1: IBC
843
34. McIntosh DB, Woolley DG, Vilsen B,
Andersen JP. 1996. Mutagenesis of segment 487Phe-Ser-Arg-Asp-Arg-Lys492
of sarcoplasmic reticulum Ca2+-ATPase
produces pumps defective in ATP binding. J. Biol. Chem. 271:25778–89
35. Taniguchi K, Kaya S, Abe K, Mardh S.
2001. The oligomeric nature of Na/Ktransport ATPase. J. Biochem. 129:335–
42
36. Brotherus JR, Jacobsen L, Jorgensen PL.
1983. Soluble and enzymatically stable
Na,K-ATPase from mammalian kidney
consisting predominantly of protomer alpha beta-units. Preparation, assay and reconstitution of active Na, K-transport.
Biochim. Biophys. Acta 731:290–303
37. Jorgensen PL. 1992. Na,K-ATPase,
structure and transport mechanism. In
Molecular Aspects of Transport Proteins. New Comprehensive Biochemistry,
ed. JJHM DePont, pp. 1–26. Amsterdam:
Elsevier
38. Ward DG, Cavieres JD. 1993. Solubilized alpha beta Na,K-ATPase remains protomeric during turnover yet
shows apparent negative cooperativity
toward ATP. Proc. Natl. Acad. Sci. USA
90:5332–36
39. Thoenges D, Amler E, Eckert T, Schoner
W. 1999. Tight binding of bulky fluorescent derivatives of adenosine to the low
affinity E2ATP site leads to inhibition of
Na+,K+-ATPase. Analysis of structural
requirements of fluorescent ATP derivatives with a Koshland-Nemethy-Filmer
model of two interacting ATP sites. J.
Biol. Chem. 274:1971–78
40. Linnertz H, Urbanova P, Obsil T, Herman
P, Amler E, Schoner W. 1998. Molecular distance measurements reveal an (alpha beta)2 dimeric structure of Na+,K+ATPase. High affinity ATP binding site
and K+-activated phosphatase reside on
different alpha-subunits. J. Biol. Chem.
273:28813–21
41. Ward DG, Cavieres JD. 1996. Binding
of 20 (30 )-O-(2,4-6-trinitrophenyl) ADP
14 Jan 2003
14:58
844
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
42.
43.
44.
45.
46.
47.
AR
AR177-PH65-34.tex
JORGENSEN
¥
AR177-PH65-34.sgm
HÅKANSSON
¥
to soluble alpha beta protomers of Na,
K-ATPase modified with fluorescein
isothiocyanate. Evidence for two distinct nucleotide sites. J. Biol. Chem. 271:
2317–21
Ward DG, Cavieres JD. 1998. Affinity
labeling of two nucleotide sites on Na,
K-ATPase using 20 (30 )-O-(2,4,6-trinitrophenyl)8-azidoadenosine 50 -[alpha-32P]
diphosphate
(TNP-8N3-[alpha-32P]
ADP) as a photoactivatable probe.
Label incorporation before and after
blocking the high affinity ATP site
with fluorescein isothiocyanate. J. Biol.
Chem. 273:33759–65
Martin DW, Sachs JR. 2000. Ligands
presumed to label high affinity and low
affinity ATP binding sites do not interact
in an (alpha beta)2 diprotomer in duck
nasal gland Na+,K+-ATPase nor do the
sites coexist in native enzyme. J. Biol.
Chem. 275:24512–17
Martin DW, Marecek J, Scarlata S,
Sachs JR. 2000. Alpha beta protomers
of Na+,K+-ATPase from microsomes of
duck salt gland are mostly monomeric:
formation of higher oligomers does not
modify molecular activity. Proc. Natl.
Acad. Sci. USA 97:3195–200
Li YF, Hata Y, Fujii T, Hisano T,
Nishihara M, et al. 1998. Crystal structures of reaction intermediates of L-2haloacid dehalogenase and implications
for the reaction mechanism. J. Biol.
Chem. 273:15035–44
Morais MC, Zhang W, Baker AS, Zhang
G, Dunaway-Mariano D, Allen KN.
2000. The crystal structure of Bacillus
cereus phosphonoacetaldehyde hydrolase: insight into catalysis of phosphorus
bond cleavage and catalytic diversification within the HAD enzyme superfamily. Biochemistry 39:10385–96
Wang W, Kim R, Jancarik J, Yokota H,
Kim S. 2001. Crystal structure of phosphoserine phosphatase from Methanococcus jannaschii, a hyperthermophile,
at 1.8 Å resolution. Structure 9:65–72
LaTeX2e(2002/01/18)
P1: IBC
KARLISH
48. Walderhaug MO, Post RL, Saccomani G,
Leonard RT, Briskin DP. 1985. Structural
relatedness of three ion-transport adenosine triphosphatases around their active
sites of phosphorylation. J. Biol. Chem.
260:3852–59
49. Kuntzweiler TA, Wallick ET, Johnson
CL, Lingrel JB. 1995. Amino acid replacement of Asp369 in the sheep alpha
1 isoform eliminates ATP and phosphate
stimulation of [3H]ouabain binding to
the Na+,K+-ATPase without altering the
cation binding properties of the enzyme.
J. Biol. Chem. 270:16206–12
50. Pedersen PA, Rasmussen JH, Jorgensen
PL. 1996. Consequences of mutations
to the phosphorylation site of the alphasubunit of Na,K-ATPase for ATP binding
and E1-E2 conformational equilibrium.
Biochemistry 35:16085–93
51. Clarke DM, Loo TW, MacLennan DH.
1990. Functional consequences of alterations to amino acids located in the nucleotide binding domain of the Ca2+ATPase of sarcoplasmic reticulum. J.
Biol. Chem. 265:22223–27
52. Pedersen PA, Jorgensen JR, Jorgensen
PL. 2000. Importance of conserved
alpha-subunit segment 709GDGVND
for Mg2+ binding, phosphorylation, and
energy transduction in Na,K-ATPase. J.
Biol. Chem. 275:37588–95
53. Jorgensen PL, Collins JH. 1986. Tryptic
and chymotryptic cleavage sites in sequence of alpha-subunit of Na,K-ATPase
from outer medulla of mammalian kidney. Biochim. Biophys. Acta 860:570–
76
54. Patchornik G, Goldshleger R, Karlish SJ.
2000. The complex ATP-Fe(2+) serves
as a specific affinity cleavage reagent in
ATP-Mg(2+) sites of Na,K-ATPase: altered ligation of Fe(2+) Mg(2+) ions accompanies the E(1) → E(2) conformational change. Proc. Natl. Acad. Sci. USA
97(22):11954–59
55. Jorgensen PL, Pedersen PA. 2001.
Structure-function relationships of
14 Jan 2003
14:58
AR
AR177-PH65-34.tex
AR177-PH65-34.sgm
LaTeX2e(2002/01/18)
Na,K-PUMP STRUCTURE AND FUNCTION
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
56.
57.
58.
59.
60.
61.
62.
63.
Na(+), K(+), ATP, or Mg(2+) binding
and energy transduction in Na,KATPase. Biochim. Biophys. Acta 1505:
57–74
Andersen JP, Vilsen B, Collins JH, Jorgensen PL. 1986. Localization of E1-E2
conformational transitions of sarcoplasmic reticulum Ca-ATPase by tryptic
cleavage and hydrophobic labeling. J.
Membr. Biol. 93:85–92
Moller JV, Juul B, le Maire M. 1996.
Structural organization, ion transport,
and energy transduction of P-type ATPases. Biochim. Biophys. Acta 1286:1–
51
Kuntzweiler TA, Arguello JM, Lingrel JB. 1996. Asp804 and Asp808
in the transmembrane domain of the
Na,K-ATPase alpha subunit are cation
coordinating residues. J. Biol. Chem.
271:29682–87
Nielsen JM, Pedersen PA, Karlish SJ,
Jorgensen PL. 1998. Importance of intramembrane carboxylic acids for occlusion of K+ ions at equilibrium in
renal Na,K-ATPase. Biochemistry 37(7):
1961–68
Pedersen PA, Nielsen JM, Rasmussen
JH, Jorgensen PL. 1998. Contribution to
Tl+, K+, and Na+ binding of Asn776,
Ser775, Thr774, Thr772, and Tyr771 in
cytoplasmic part of fifth transmembrane
segment in alpha-subunit of renal Na,KATPase. Biochemistry 37:17818–27
Smart OS, Neduvelil JG, Wang X, Wallace BA, Sansom MS. 1996. HOLE: a
program for the analysis of the pore dimensions of ion channel structural models. J. Mol. Graph. 14:354–60
Sweadner KJ, Donnet C. 2001. Structural similarities of Na,K-ATPase and
SERCA, the Ca(2+)-ATPase of the sarcoplasmic reticulum. Biochem. J. 356:
685–704
Karlish SJD, Goldshleger R, Stein WD.
1990. Identification of a 19 KD Cterminal tryptic fragment of the alpha
chain of Na,K-ATPase, essential for oc-
64.
65.
66.
67.
68.
69.
70.
71.
P1: IBC
845
clusion and transport. Proc. Natl. Acad.
Sci. USA 87:4566–70
Capasso JM, Hoving S, Tal DM, Goldshleger R, Karlish SJD. 1992. Extensive digestion of Na,K-ATPase by specific and
non-specific proteases with preservation
of cation occlusion sites. J. Biol. Chem.
267:1150–58
Schwappach B, Stürmer W, Apell H-J,
Karlish SJD. 1994. Binding of sodium
ions and cardiotonic steroids to native
and selectively trypsinized Na,K pump,
detected by charge movements. J. Biol.
Chem. 269:21620–26
Shainskaya A, Nesaty V, Karlish SJ.
1998. Interactions between fragments
of trypsinized Na,K-ATPase detected by
thermal inactivation of Rb+ occlusion
and dissociation of the M5/M6 fragment.
J. Biol. Chem. 273:7311–19
Zhang ZS, Lewis D, Sumbilla C, Inesi
G, Toyoshima C. 2001. The role of the
M6-M7 loop (L6/7) in stabilization of
the phosphorylation and Ca(2+) binding domains of the sarcoplasmic reticulum Ca(2+)-ATPase (SERCA). J. Biol.
Chem. 276:15232–39
Lutsenko S, Anderko R, Kaplan JH.
1995. Membrane disposition of the M5M6 hairpin of Na+,K+-ATPase alpha
subunit is ligand dependent. Proc. Natl.
Acad. Sci. USA 92:7936–40
Gatto C, Lutsenko S, Shin JM, Sachs
G, Kaplan JH. 1999. Stabilization of the
H,K-ATPase M5M6 membrane hairpin
by K+ ions. Mechanistic significance
for p2-type ATPases. J. Biol. Chem.
274:13737–40
Lutsenko S, Daoud S, Kaplan JH. 1997.
Identification of two conformationally
sensitive cysteine residues at the extracellular surface of the Na,K-ATPase
alpha-subunit. J. Biol. Chem. 272:5249–
55
Or E, Goldshleger R, Karlish SJ. 1999.
Characterization of disulfide cross-links
between fragments of proteolyzed Na,
K-ATPase. Implications for spatial
14 Jan 2003
14:58
846
72.
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
73.
74.
75.
76.
77.
78.
79.
80.
AR
AR177-PH65-34.tex
JORGENSEN
¥
AR177-PH65-34.sgm
HÅKANSSON
¥
organization of trans-membrane helices.
J. Biol. Chem. 274:2802–9
Ivanov A, Zhao H, Modyanov NN. 2000.
Packing of the transmembrane helices
of Na,K-ATPase: direct contact between
beta-subunit and H8 segment of alphasubunit revealed by oxidative crosslinking. Biochemistry 39:9778–85
Or E, Goldshleger R, Shainskaya A,
Karlish SJ. 1998. Specific cross-links between fragments of proteolyzed Na,KATPase induced by O-phthalaldehyde.
Biochemistry 37(22):8197–207
Hu YK, Kaplan JH. 2000. Site-directed
chemical labeling of extracellular loops
in a membrane protein. The topology of
the Na,K-ATPase alpha-subunit. J. Biol.
Chem. 275:19185–91
Hu YK, Eisses JF, Kaplan JH. 2000. Expression of an active Na,K-ATPase with
an alpha-subunit lacking all twenty-three
native cysteine residues. J. Biol. Chem.
275:30734–39
Frillingos S, Sahin-Toth M, Wu J,
Kaback HR. 1998. Cys-scanning mutagenesis: a novel approach to structure
function relationships in polytopic membrane proteins. FASEB J. 12:1281–99
Jorgensen PL, Karlish SJ, Gitler C.
1982. Evidence for the organization of
the transmembrane segments of (Na,K)ATPase based on labelling lipid-embedded and surface domains of the alphasubunit. J. Biol. Chem. 257:7435–42
Jorgensen PL, Brunner J. 1983. Labeling
of intramembrane segments of the alphasubunit and beta-subunit of pure membrane-bound (Na+,K+)-ATPase with 3trifluoromethyl-3-(m-[125I]iodophenyl)
diazirine. Biochim. Biophys. Acta 735:
291–96
Modyanov NN, Lutsenko S, Chertova
R, Efremov R, Gulyaev D. 1992. Transmembrane organization of the Na,KATPase molecule. Acta Physiol. Scand.
146:49–58
Blanton MP, McCardy EA. 2000. Identifying the lipid-protein interface and
LaTeX2e(2002/01/18)
P1: IBC
KARLISH
81.
82.
83.
84.
85.
86.
87.
88.
89.
transmembrane structural transitions of
the Torpedo Na,K-ATPase using hydrophobic photoreactive probes. Biochemistry 39(44):13534–44
Goldshleger R, Karlish SJ. 1997. Fecatalyzed cleavage of the alpha subunit
of Na/K-ATPase: evidence for conformation-sensitive interactions between
cytoplasmic domains. Proc. Natl. Acad.
Sci. USA 94(18):9596–601
Goldshleger R, Karlish SJ. 1999. The energy transduction mechanism of Na,KATPase studied with iron-catalyzed oxidative cleavage. J. Biol. Chem. 274:
16213–21
Shin JM, Goldshleger R, Munson KB,
Sachs G, Karlish SJ. 2001. Selective
Fe2+-catalyzed oxidative cleavage of
gastric Na+,K+-ATPase: implications for
the energy transduction mechanism of
P-type cation pumps. J. Biol. Chem.
276:48440–504
Deleted in proof
Tal DM, Capasso JM, Munson K, Karlish SJ. 2001. Proximity of transmembrane segments M3 and M1 of the alpha subunit of Na+,K+-ATPase revealed
by specific oxidative cleavage mediated by a complex of Cu2+ ions and
4,7-diphenyl-1,10-phenanthroline. Biochemistry 40(42):12505–14
Fambrough DM, Lemas MV, Hamrick
M, Emerick M, Renaud KJ, et al. 1994.
Analysis of subunit assembly of the
Na-K-ATPase. Am. J. Physiol. Cell Physiol. 266:C579–C89
Colonna T, Kostich M, Hamrick M,
Hwang B, Rawn JD, Fambrough DM.
1997. Subunit interactions in the sodium
pump. Ann. NY Acad. Sci. 834:498–513
Geering K. 2001. The functional role
of beta subunits in oligomeric P-type
ATPases. J. Bioenerg. Biomembr. 33:
425–38
Hasler U, Crambert G, Horisberger JD,
Geering K. 2001. Structural and functional features of the transmembrane domain of the Na,K-ATPase beta subunit
14 Jan 2003
14:58
AR
AR177-PH65-34.tex
AR177-PH65-34.sgm
LaTeX2e(2002/01/18)
Na,K-PUMP STRUCTURE AND FUNCTION
90.
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
91.
92.
93.
94.
95.
96.
97.
98.
revealed by tryptophan scanning. J. Biol.
Chem. 276(19):16356–64
Therien AG, Pu HX, Karlish SJ, Blostein
R. 2001. Molecular and functional studies of the gamma subunit of the sodium
pump. J. Bioenerg. Biomembr. 33:407–
14
von Heijne G. 1991. Proline kinks in
transmembrane alpha-helices. J. Mol.
Biol. 218:499–503
Shainskaya A, Schneeberger A, Apell
HJ, Karlish SJ. 2000. Entrance port for
Na(+) and K(+) ions on Na(+),K(+)ATPase in the cytoplasmic loop between
trans-membrane segments M6 and M7 of
the alpha subunit. Proximity of the cytoplasmic segment of the beta subunit. J.
Biol. Chem. 275:2019–28
Menguy T, Corre F, Bouneau L, Deschamps S, Moller JV, et al. 1998.
The cytoplasmic loop located between
transmembrane segments 6 and 7 controls activation by Ca2+ of sarcoplasmic
reticulum Ca2+-ATPase. J. Biol. Chem.
273:20134–43
Menguy T, Corre F, Juul B, Bouneau L,
Lafitte D, et al. 2002. Involvement of the
cytoplasmic loop L6-7 in the entry mechanism for transport of Ca2+ through the
sarcoplasmic reticulum Ca2+-ATPase. J.
Biol. Chem. 277:13016–28
Van Huysse JW, Jewell EA, Lingrel
JB. 1993. Site-directed mutagenesis of
a predicted cation binding site of Na,KATPase. Biochemistry 32:819–26
Yamamoto S, Kuntzweiler TA, Wallick
ET, Sperelakis N, Yatani A. 1996. Amino
acid substitutions in the rat Na+,K+ATPase alpha2-subunit alter the cation
regulation of pump current expressed in
HeLa cells. J. Physiol. 495:733–42
Rakowski RF, Gadsby DC, De Weer P.
1997. Voltage dependence of the Na/K
pump. J. Membr. Biol. 155:105–12
Holmgren M, Wagg J, Bezanilla F,
Rakowski RF, De Weer P, Gadsby
DC. 2000. Three distinct and sequential
steps in the release of sodium ions by
99.
100.
101.
102.
103.
104.
105.
106.
P1: IBC
847
the Na+,K+-ATPase. Nature 403:898–
901
Apell HJ, Karlish SJD. 2001. Functional
properties of Na,K-ATPase, and their
structural implications, as detected with
biophysical techniques. J. Membr. Biol.
180:1–9
Pedersen M, Roudna M, Beutner S,
Birmes M, Reifers B, et al. 2002. Detection of charge movements in ion pumps
by a family of styryl dyes. J. Membr. Biol.
183:221–36
Pedersen PA, Rasmussen JH, Jorgensen
PL. 1996. Expression in high yield of pig
alpha 1 beta 1 Na,K-ATPase and inactive mutants D369N and D807N in Saccharomyces cerevisiae. J. Biol. Chem.
271:2514–22
Blostein R, Wilczynska A, Karlish SJ,
Arguello JM, Lingrel JB. 1997. Evidence that Ser775 in the alpha subunit
of the Na,K-ATPase is a residue in the
cation binding pocket. J. Biol. Chem.
272:24987–93
Arguello JM, Whitis J, Cheung MC, Lingrel JB. 1999. Functional role of oxygencontaining residues in the fifth transmembrane segment of the Na,K-ATPase
alpha subunit. Arch. Biochem. Biophys.
364:254–63
Pedersen PA, Rasmussen JH, Nielsen
JM, Jorgensen PL. 1997. Identification of Asp804 and Asp808 as Na+
and K+ coordinating residues in alphasubunit of renal Na,K-ATPase. FEBS
Lett. 400:206–10
Kuntzweiler TA, Wallick ET, Johnson
CL, Lingrel JB. 1995. Glutamic acid
327 in the sheep alpha 1 isoform of
Na+,K+-ATPase stabilizes a K+-induced
conformational change. J. Biol. Chem.
270:2993–3000
Vilsen B. 1995. Mutant Glu781 → Ala
of the rat kidney Na+,K+-ATPase displays low cation affinity and catalyzes
ATP hydrolysis at a high rate in the absence of potassium ions. Biochemistry
34:1455–63
14 Jan 2003
14:58
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
848
AR
AR177-PH65-34.tex
JORGENSEN
¥
AR177-PH65-34.sgm
HÅKANSSON
¥
107. Peluffo RD, Arguello JM, Lingrel JB,
Berlin JR. 2000. Electrogenic sodiumsodium exchange carried out by Na,KATPase containing the amino acid substitution Glu779Ala. J. Gen. Physiol.
116:61–73
107a. Patchornik G, Munson K, Goldshleger
R, Shainskaya A, Sachs G, Karlish
SJG. 2002. The ATP-Mg2+ binding site
and cytoplasmic domain interactions of
Na,K-ATPase investigated with Fe2+catalyzed oxidative cleavage and molecular modeling. Biochemistry. In press
108. Danko S, Yamasaki K, Daiho T, Suzuki
H, Toyoshima C. 2001. Organization
of cytoplasmic domains of sarcoplasmic
reticulum Ca(2+)-ATPase in E(1)P and
E(1)ATP states: a limited proteolysis
study. FEBS Lett. 505:129–35
109. Danko S, Daiho T, Yamasaki K, Kamidochi M, Suzuki H, Toyoshima C.
2001. ADP-insensitive phosphoenzyme
intermediate of sarcoplasmic reticulum
Ca(2+)-ATPase has a compact conformation resistant to proteinase K, V8 protease and trypsin. FEBS Lett. 489:277–
82
110. Jorgensen PL, Petersen J. 1985. Chymotryptic cleavage of alpha-subunit in
E1-forms of renal (Na+,K+)-ATPase:
effects on enzymatic properties, ligand
binding and cation exchange. Biochim.
Biophys. Acta 821:319–33
111. Jorgensen PL. 1977. Purification and
characterization of Na+,K+-ATPase. VI.
Differential tryptic modification of catalytic functions of the purified enzyme
in presence of NaCl and KCl . Biochim.
Biophys. Acta 466:97–108
112. Jorgensen PL, Karlish SJ. 1980. Defective conformational response in a
selectively trypsinized Na+,K+-ATPase
studied with tryptophan fluorescence.
Biochim. Biophys. Acta 597:305–17
113. Boxenbaum N, Daly SE, Javaid ZZ,
Lane LK, Blostein R. 1998. Changes in
steady-state conformational equilibrium
resulting from cytoplasmic mutations of
LaTeX2e(2002/01/18)
P1: IBC
KARLISH
114.
115.
116.
117.
118.
119.
120.
121.
122.
the Na,K-ATPase alpha-subunit. J. Biol.
Chem. 273:23086–92
Kaplan JH, Gatto C, Holden JP,
Thornewell SJ. 1998. Structural changes
associated with the coupling of ATP
hydrolysis and cation transport by the
Na pump. Acta Physiol. Scand. Suppl.
643:99–105
Gatto C, Lutsenko S, Kaplan JH. 1997.
Chemical modification with dihydro4,40 -diisothiocyanostilbene-2,20 -disulfonate reveals the distance between K480
and K501 in the ATP-binding domain
of the Na,K-ATPase. Arch. Biochem.
Biophys. 340:90–100
Goldshleger R, Karlish SJ. 1997. Fecatalyzed cleavage of the alpha subunit of Na/K-ATPase: evidence for
conformation-sensitive interactions between cytoplasmic domains. Proc. Natl.
Acad. Sci. USA 94:9596–601
Goldshleger R, Karlish SJ. 1999. The energy transduction mechanism of Na,KATPase studied with iron-catalyzed oxidative cleavage. J. Biol. Chem. 274:
16213–21
Deleted in proof
Pedersen PA, Rasmussen JH, Jorgensen
PL. 1996. Consequences of mutations
to the phosphorylation site of the alphasubunit of Na,K-ATPase for ATP binding
and E1-E2 conformational equilibrium.
Biochemistry 35:16085–93
McIntosh DB. 1992. Glutaraldehyde
cross-links Lys-492 and Arg-678 at the
active site of sarcoplasmic reticulum
Ca(2+)-ATPase. J. Biol. Chem. 267:
22328–35
Andersen JP, Moller JV, Jørgensen PL.
1982. The functional unit of sarcoplasmic reticulum Ca2+-ATPase. Active site
titration and fluorescence measurements.
J. Biol. Chem. 257:8300–7
Fukushima Y, Post RL. 1978. Binding
of divalent cation to phosphoenzyme of
sodium- and potassium-transport adenosine triphosphatase. J. Biol. Chem.
253:6853–62
14 Jan 2003
14:58
AR
AR177-PH65-34.tex
AR177-PH65-34.sgm
LaTeX2e(2002/01/18)
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
Na,K-PUMP STRUCTURE AND FUNCTION
123. Ridder IS, Dijkstra BW. 1999. Identification of the Mg2+-binding site in the
P-type ATPase and phosphatase members of the HAD (haloacid dehalogenase) superfamily by structural similarity to the response regulator protein
CheY. Biochem. J. 339:223–26
124. Clausen JD, McIntosh DB, Woolley DG,
Andersen JP. 2001. Importance of Thr353 of the conserved phosphorylation
loop of the sarcoplasmic reticulum Ca2+ATPase in MgATP binding and catalytic
activity. J. Biol. Chem. 276:35741–50
125. Glynn IM, Hara Y, Richards DE. 1984.
The occlusion of sodium ions within the
mammalian sodium-potassium pump: its
role in sodium transport. J. Physiol. 351:
531–47
126. Toustrup-Jensen M, Hauge M, Vilsen B.
2001. Mutational effects on conformational changes of the dephospho- and
phospho-forms of the Na+,K+-ATPase.
Biochemistry 40:5521–32
127. Andersen JP, Sorensen TL, Povlsen K,
Vilsen B. 2001. Importance of transmembrane segment M3 of the sarcoplasmic reticulum Ca2+-ATPase for control
of the gateway to the Ca2+sites. J. Biol.
Chem. 276:23312–21
P1: IBC
849
128. Zhang Z, Sumbilla C, Lewis D, Summers
S, Klein MG, Inesi G. 1995. Mutational
analysis of the peptide segment linking
phosphorylation and Ca(2+)-binding
domains in the sarcoplasmic reticulum
Ca(2+)-ATPase. J. Biol. Chem. 270:
16283–90
129. Ambesi A, Miranda M, Allen KE, Slayman CW. 2000. Stalk segment 4 of
the yeast plasma membrane H+-ATPase.
Mutational evidence for a role in the
E1-E2 conformational change. J. Biol.
Chem. 275:20545–50
130. Sorensen TL, Andersen JP. 2000. Importance of stalk segment S5 for intramolecular communication in the sarcoplasmic
reticulum Ca2+-ATPase. J. Biol. Chem.
275:28954–61
131. Schneeberger A, Apell HJ. 2001. Ion
selectivity of the cytoplasmic binding
sites of the Na,K-ATPase: II. Competition of various cations. J. Membr. Biol.
179:263–73
132. Beguin P, Crambert G, Guennoun S,
Garty H, Horisberger JD, Geering K.
2001. CHIF, a member of the FXYD
protein family, is a regulator of Na,KATPase distinct from the gammasubunit. EMBO J. 20:3993–4002
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
22 Jan 2003
21:37
AR
AR177-34-COLOR.tex
AR177-34-COLOR.SGM
LaTeX2e(2002/01/18)
P1: GDL
Figure 1 Models of E1 and E2 forms of the a-subunit of Na,K-ATPase based on
the high-resolution structure of Ca-ATPase (1EUL) (16) in the E1[2Ca] form and a
structure model (1KJU) based on the 6 Å electron microscopy structure of the E2 form
of Ca-ATPase stabilized by decavanadate and thapsigargin (17). Side chains in balland-stick are shown for Phe475, Lys501, and Arg 544 in the N-domain (green), for
D369 and D710 in the P-domain (blue), and for Glu327, Glu779, Asp804, and Asp808
in the cation-binding domain (gray). The position of the N-domain, which in the crystal
structure of Ca-ATPase might be displaced by crystal contacts (16), has been adjusted
to allow spanning of an ATP molecule over the N- and P-domains (8). The resulting
relative positions of the N and P domains conform well with those observed in the E2
form by electron microscopy (12, 13, 17).
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
22 Jan 2003
21:37
AR
AR177-34-COLOR.tex
AR177-34-COLOR.SGM
LaTeX2e(2002/01/18)
P1: GDL
Figure 3 Direct or indirect transmission of structural changes from P-domain to
cation sites? Homology model of the connections from the A-domain via L2/3 to M3,
P1-helix, and L6/7 to M4, M5, and M6. Side chains of the α-subunit of Na,K-ATPase
are modeled on the backbone of Ca-ATPase in the E1[Ca] form (16). CHYM indicates
the position of the peptide bond (Leu266), which is cleaved by chymotrypsin in the
E1[3Na] form of the α-subunit of Na,K-ATPase (1). T2-TRYP (Arg198) and V8 PROT
(Glu231) denote positions of bonds cleaved in the E1[2Ca] form of Ca-ATPase (56,
57).
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
22 Jan 2003
21:37
AR
AR177-34-COLOR.tex
AR177-34-COLOR.SGM
LaTeX2e(2002/01/18)
P1: GDL
Figure 4 (A) Ball-and-stick representation of Na+-binding sites I and II; homology model
of Na,K-pump side chains on the backbone of the high-resolution structure of SR Ca-ATPase
(1EUL) (16). (B ) Proposed mechanism of Na+ translocation. The entrance of the third Na+
from the cytoplasmic site and the release of the first ion on the extracellular side are electrogenic owing to a bucket brigade mechanism. Binding of the third Na+ will displace the
second Na+ to site I. The E1-P-E2-P transition displaces the third Na+ ion from the entry site
to site I and expels the second Na+ ion (K.O. Håkansson & P.L. Jorgensen, submitted). (C )
Probing for channels through the model of Na,K-ATPase structure with the HOLE software
package (K.O. Håkansson & P.L. Jorgensen, submitted).
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
22 Jan 2003
21:37
AR
AR177-34-COLOR.tex
AR177-34-COLOR.SGM
LaTeX2e(2002/01/18)
P1: GDL
Figure 5 A homology model of Na,K-ATPase with transmembrane helices seen from
the extracellular side. Note residues highlighted in extracellular loops, particularly
L9/10.
Figure 7 Grasp figure showing the electrostatic potential on the surface. (Left) CaATPase. (Right) A Na,K-ATPase homology model (K.O. Håkansson & P.L. Jorgensen,
submitted). Positive charge is shown in blue and negative charge in red. Arrow points
at the suggested site for Na+ entrance on the cytoplasmic side between M2, M4, and
M6.
P1: FDS
January 17, 2003
11:23
Annual Reviews
AR177-FM
Annual Review of Physiology,
Volume 65, 2003
CONTENTS
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
Frontispiece—Jean D. Wilson
xiv
PERSPECTIVES, Joseph F. Hoffman, Editor
A Double Life: Academic Physician and Androgen Physiologist,
Jean D. Wilson
1
CARDIOVASCULAR PHYSIOLOGY, Jeffrey Robbins, Section Editor
Lipid Receptors in Cardiovascular Development, Nick Osborne
and Didier Y.R. Stainier
Cardiac Hypertrophy: The Good, the Bad, and the Ugly, N. Frey
and E.N. Olson
Stress-Activated Cytokines and the Heart: From Adaptation to
Maladaptation, Douglas L. Mann
23
45
81
CELL PHYSIOLOGY, Paul De Weer, Section Editor
Cell Biology of Acid Secretion by the Parietal Cell, Xuebiao Yao
and John G. Forte
Permeation and Selectivity in Calcium Channels, William A. Sather
and Edwin W. McCleskey
Processive and Nonprocessive Models of Kinesin Movement,
Sharyn A. Endow and Douglas S. Barker
103
133
161
COMPARATIVE PHYSIOLOGY, George N. Somero, Section Editor
Origin and Consequences of Mitochondrial Variation in Vertebrate
Muscle, Christopher D. Moyes and David A. Hood
Functional Genomics and the Comparative Physiology of Hypoxia,
Frank L. Powell
Application of Microarray Technology in Environmental
and Comparative Physiology, Andrew Y. Gracey and
Andrew R. Cossins
177
203
231
ENDOCRINOLOGY, Bert W. O’Malley, Section Editor
Nuclear Receptors and the Control of Metabolism,
Gordon A. Francis, Elisabeth Fayard, Frédéric Picard, and
Johan Auwerx
261
vii
P1: FDS
January 17, 2003
viii
11:23
Annual Reviews
AR177-FM
CONTENTS
Insulin Receptor Knockout Mice, Tadahiro Kitamura, C. Ronald Kahn,
and Domenico Accili
The Physiology of Cellular Liporegulation, Roger H. Unger
313
333
GASTROINTESTINAL PHYSIOLOGY, John Williams, Section Editor
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
The Gastric Biology of Helicobacter pylori, George Sachs,
David L. Weeks, Klaus Melchers, and David R. Scott
Physiology of Gastric Enterochromaffin-Like Cells, Christian Prinz,
Robert Zanner, and Manfred Gratzl
Insights into the Regulation of Gastric Acid Secretion Through Analysis of
Genetically Engineered Mice, Linda C. Samuelson and Karen L. Hinkle
349
371
383
NEUROPHYSIOLOGY, Richard Aldrich, Section Editor
In Vivo NMR Studies of the Glutamate Neurotransmitter Flux
and Neuroenergetics: Implications for Brain Function,
Douglas L. Rothman, Kevin L. Behar, Fahmeed Hyder,
and Robert G. Shulman
401
Transducing Touch in Caenorhabditis elegans, Miriam B. Goodman
and Erich M. Schwarz
429
Hyperpolarization-Activated Cation Currents: From Molecules
to Physiological Function, Richard B. Robinson and
Steven A. Siegelbaum
453
RENAL AND ELECTROLYTE PHYSIOLOGY, Steven C. Hebert, Section Editor
Macula Densa Cell Signaling, P. Darwin Bell, Jean Yves Lapointe,
and János Peti-Peterdi
Paracrine Factors in Tubuloglomerular Feedback: Adenosine, ATP,
and Nitric Oxide, Jürgen Schnermann and David Z. Levine
Regulation of Na/Pi Transporter in the Proximal Tubule,
Heini Murer, Nati Hernando, Ian Forster, and Jürg Biber
Mammalian Urea Transporters, Jeff M. Sands
Terminal Differentiation of Intercalated Cells: The Role of Hensin,
Qais Al-Awqati
481
501
531
543
567
RESPIRATORY PHYSIOLOGY, Carole R. Mendelson, Section Editor
Current Status of Gene Therapy for Inherited Lung Diseases,
Ryan R. Driskell and John F. Engelhardt
The Role of Exogenous Surfactant in the Treatment of Acute Lung
Injury, James F. Lewis and Ruud Veldhuizen
Second Messenger Pathways in Pulmonary Host Defense,
Martha M. Monick and Gary W. Hunninghake
585
613
643
P1: FDS
January 17, 2003
11:23
Annual Reviews
AR177-FM
CONTENTS
Alveolar Type I Cells: Molecular Phenotype and Development,
Mary C. Williams
Annu. Rev. Physiol. 2003.65:817-849. Downloaded from arjournals.annualreviews.org
by Columbia University on 04/25/07. For personal use only.
SPECIAL TOPIC: LIPID RECEPTOR PROCESSES, Donald W. Hilgemann,
Special Topic Editor
Getting Ready for the Decade of the Lipids, Donald W. Hilgemann
Aminophospholipid Asymmetry: A Matter of Life and Death,
Krishnakumar Balasubramanian and Alan J. Schroit
Regulation of TRP Channels Via Lipid Second Messengers,
Roger C. Hardie
Phosphoinositide Regulation of the Actin Cytoskeleton,
Helen L. Yin and Paul A. Janmey
Dynamics of Phosphoinositides in Membrane Retrieval and
Insertion, Michael P. Czech
SPECIAL TOPIC: MEMBRANE PROTEIN STRUCTURE, H. Ronald Kaback,
Special Topic Editor
Structure and Mechanism of Na,K-ATPase: Functional Sites
and Their Interactions, Peter L. Jorgensen, Kjell O. Håkansson,
and Steven J. Karlish
G Protein-Coupled Receptor Rhodopsin: A Prospectus,
Slawomir Filipek, Ronald E. Stenkamp, David C. Teller, and
Krzysztof Palczewski
ix
669
697
701
735
761
791
817
851
INDEXES
Subject Index
Cumulative Index of Contributing Authors, Volumes 61–65
Cumulative Index of Chapter Titles, Volumes 61–65
ERRATA
An online log of corrections to Annual Review of Physiology chapters
may be found at http://physiol.annualreviews.org/errata.shtml
881
921
925