Mark Hilge
Structural Biology
Na,K-ATPase
Background
The Na,K-ATPase or sodium pump is a membrane protein that hydrolyses
ATP in order to drive the coupled extrusion and uptake of Na+ and
K+ ions across the plasma membrane. The thereby established electrochemical
gradient is essential for cellular function [1]. The Na,K-ATPase is
a family II P-type ATPase and is composed of a catalytic a subunit,
common to all P-type ATPases, and a regulatory b subunit that is unique
to Na,K- and H,K-ATPase.
Understanding of the molecular mechanism of P-type ATPases requires
high-resolution, 3D structures of the enzymes at various points in
the pump cycle. Recently, two structures of the Ca-ATPase of sarcoplasmic
reticulum (SERCA1a) have become available, describing the E1 and E2
states (see pump cycle [2,3]). These structures reveal that P-type-ATPases
consist of a transmembrane domain (TM) of 10 a-helices and three distinct
cytosolic domains (N-, P-and A-domain). While the N-domain binds ATP,
the P-domain contains an invariant aspartate (Asp369 in the case of
the sodium pump) that is phosphorylated during the pump cycle. The
A-domain is necessary for the structural integrity of the pump.
Because binding of ATP is an essential event in the pump cycle, we
wish to explore its structural consequences.
N-domain structure
We determined the structures of the native and ATP-bound N-domain
of rata1 Na,K-ATPase (NaKa1), encompassing residues Gln376 –
Pro588, using high-resolution, heteronuclear, multidimensional NMR
spectroscopy. The domain has the topology of a twisted, six-stranded,
antiparallel b-sheet that is flanked by two a-helices on either side.
Despite the low sequence similarity of only
21% the overall fold of the N-domain of NaKa1 is similar to that of
the N-domain of SERCA1a. Structural differences mainly occur where
SERCA1a has large inserts relative to NaKa1. The cluster of aromatic
residues formed by Phe558, Phe562, Phe564 and Phe571 in NaKa1 is part
of a scaffold covering the majority of the central b-sheet. The analogous
portion of the Ca-ATPase is rotated by approximately 60° relative
to NaKa1 and, with the exception of Tyr587, lacks the aromatic residues.
In SERCA2a, the cardiac isoform of the Ca-ATPase, residues Lys397
– Val402 are thought to interact with the unphosphorylated form
of the regulator phospholamban. On the basis of sequence alignments,
it was believed that this region is missing in the sodium pump. However,
in the structure of NaKa1 residues Ala393 – Thr407 form a largely
disordered, solvent-accessible loop that occupies a position similar
to the phospholamban interaction site in the Ca-ATPase. It is therefore
tempting to speculate that this region in the Na,K-ATPase may also
interact with an as yet unidentified regulator molecule.
The ATP-complex
In order to elucidate the structure of the ATP-protein complex we
measured a series of NMR spectra under conditions where ATP binding
was saturated by addition of an eleven-fold excess over protein. Twenty-five
ATP-protein distances (NOEs) allowed unambiguous positioning of ATP.
The nucleotide binds to the protein primarily through the base (H2
and H8 protons) and sugar ring (NOEs to Phe475, Lys480, Ala503 and
Leu546) leaving the phosphate groups solvent-exposed. ATP binds to
the N-domain of NaKa1 in the anti-conformation, in contrast to the
syn-conformation seen in the 4 Å complex of the Ca-ATPase with
the inhibitor 2',3'-O-(2,4,6-trinitrophenyl)-AMP [2]. It may be that
the resolution of the X-ray structure does not permit unequivocal
distinction between syn- and anti-conformation or that the bulky TNP-group
prevents the normal binding mode.
Compared to ATP, titration with MgATP and
ADP yields a substantially reduced affinity (Kd ~25 mM ) for
the N-domain.
Superposition of the native structure of NaKa1 and the structure of
SERCA1a shows that the arrangement of the crucial residues in the
binding pocket is well-preserved. In particular, the five residues
Lys501, Gly502, Ala503, Glu505, and Cys511, which are strictly conserved
in family II P-type ATPases, adopt very similar orientations. Contributions
to ATP-binding originate from four different locations in the N-domain
of NaKa1. Among these, only the highly conserved sequence K501GAPExxxDRC511
is well-defined in the native form of the sodium pump, whereas the
other three regions (Gly442 – Glu446, Phe475 – Lys480,
and Gly542 – Arg544) show a high degree of conformational variability.
In the ATP-complex displayed Phe475 – Lys480 adopt a rigid conformation.
The paucity of interactions found between the protein and ATP is in
good agreement with the low binding constant. The amino protons of
the base form the only unambiguously identified hydrogen bond to Oe1
of Gln482. On the opposite side of the pocket, Phe475 provides a hydrophobic
platform for a stacking interaction with the base of ATP. Mutation
of the corresponding residue Phe487 in SERCA1a reduced the affinity
for ATP 22-fold [5]. Lys480 points into the binding pocket, well-oriented
towards the a-phosphate group of the nucleotide. Mutational data [6,7]
suggest that Arg544 is responsible for the stabilization of the g-
and/or b-phosphate group of the substrate. Arg544 is well-positioned
to interact with either the g- and/or b-phosphate groups, however,
no direct restraints involving the phosphate groups can be measured.
Indicative of direct interactions between the protein and these phosphates
are the observed 31P chemical shift changes upon ATP titration with
the N-domain of NaKa1.
The pump cycle of Na,K-ATPase
In 1972 a scheme for the pump cycle of the Na,K-ATPase was proposed
[8]. With the availability of two X-ray structures of the related
Ca-ATPase [2,3], the 11 Å EM structure of the Na,K-ATPase [4],
the NMR and X-ray structures of the N-domain of Na,K-ATPase in the
free [0,9] and ATP-bound [0] state as well as numerous data from mutation
and Fe-cleavage experiments [10,11] allow for the construction of
a molecular movie of the reaction cycle.
Movie script
Assuming the reaction cycle starts with the sodium pump in the E2[2K]
state ([4,0]) ATP initially binds with low apparent affinity (Kd ~
200-400 mM). Binding of the nucleotide induces a conformational change
in the hinge region between the N- and P-domain [0] that is relayed
to the K+ binding sites within the membrane, leading to the release
of the K+ ions. Concomitantly or sequentially the A-domain disengages
from the N- and P-domain.
Binding of three Na+ ions induces the inclination of the N-domain
that brings the g-phosphate group into close proximity of the invariant
aspartate residue on the P-domain. The high-affinity state is formed
by mainly stabilising the g-phosphate group via the magnesium ion,
bound on the P-domain [11], and a strictly conserved threonine residue
in the phosphorylation loop [10].
After transfer of the phosphate group the affinity for the nucleotide
(ADP) is drastically reduced, inducing several conformational changes
in either a concerted or sequential fashion: The occluded sodium ions
are released to the outside of the cell one by one [12], the N-domain
snaps back and the A-domain engages again. At this moment the enzyme
is in the E2P state [3] that predominantly differs from the E2[2K]
state in the position of the N-domain. Hydrolysis of the phosphorylated
aspartate residue leads to the E2[2K] state where the cycle ends.
For recent reviews of the reaction cycle of Na,K- and Ca-ATPase see
[14,15] and [16], respectively.
Addendum 1: Comparison of the cytosolic
domains of the E1 and E2 forms of Ca-ATPase [2,3] reveals relatively
few structural changes in the domains themselves. However, large differences
for residues in the hinge regions between the N- and P- as well as
the A- and TM-domains, result in substantial domain movements.
Addendum 2: Although the early
description of E1 and E2 conformations was of great help it is likely
an oversimplification and it may be more accurate to describe the
reaction cycle as a series of unique intermediates separated by small,
reversible steps [16].
Reprint
[0] Hilge, M., Siegal, G., Vuister G.W., Güntert,
P., Gloor, S.M. & Abrahams, J.P. ATP-induced conformational
changes of the nucleotide-binding domain of Na,K-ATPase. Nat. Struct.
Biol. 10, 468-74 (2003).
Coordinates
Native structure of the N-domain of NaKa1: 1MO7
ATP-complex structure of the N-domain of NaKa1: 1MO8
Chemical shifts
Native structure of the N-domain of NaKa1: 5576
ATP-complex structure of the N-domain of NaKa1: 5577
References
[1] Horisberger, J.D., Lemas, V., Krähenbühl,
J.P. & Rossier, B.C. Structure-function relationship of Na,K-ATPase.
Annu. Rev. Physiol. 53, 565-584 (1991).
[2] Toyoshima, C., Nakasako, M., Nomura, H.
& Ogawa, H. Crystal structure of the calcium pump of sarcoplasmic
reticulum at 2.6 Å resolution. Nature 405, 647-655 (2000).
[3] Toyoshima, C. & Nomura, H. Structural
changes in the calcium pump accompanying the dissociation of calcium.
Nature 418, 605-611 (2002).
[4] Rice, W.J., Young, H.S., Martin, D.W.,
Sachs, J.R. & Stokes, D.L. Structure of Na+,K+-ATPase at 11-Å
resolution: comparison with Ca2+- ATPase in E1 and E2 states. Biophys.
J. 80, 2187-2197 (2001).
[5] McIntosh, D.B., Woolley, D.G., Vilsen,
B. & Andersen, J.P. Mutagenesis of segment 487Phe-Ser-Arg-Asp-Arg-Lys492
of sarcoplasmic reticulum Ca2+-ATPase produces pumps defective in
ATP binding. J. Biol. Chem. 271, 25778-25789 (1996).
[6] Jacobsen, M.D., Pedersen, P.A. & Jorgensen,
P.L. Importance of Na,K-ATPase residue a1-Arg544 in the segment Arg544-
Asp567 for high-affinity binding of ATP, ADP, or MgATP. Biochemistry
41, 1451-1456 (2002).
[7] Teramachi, S., Imagawa, T., Kaya, S. &
Taniguchi, K. Replacement of several single amino acid side chains
exposed to the inside of the ATP-binding pocket induces different
extents of affinity change in the high and low affinity ATP-binding
sites of rat Na/K- ATPase. J. Biol. Chem. 270, 37394-37400 (2002).
[8] Post, R.L., Hegyvary, C. & Kume, S.
Activation by adenosine triphosphate in the phosphorylation kinetics
of sodium and potassium ion transport adenosine triphosphatase. J.
Biol. Chem. 247, 6530-6540 (1972).
[9] Håkansson, K.O. The crystallographic
structure of Na,K-ATPase N-domain at 2.6 Å resolution. J. Mol.
Biol. 332, 1175-1182 (2003).
[10] Clausen, J.D., McIntosh, D.B., Woolley, D.G. &
Andersen, J.P. Importance of Thr-353 of the conserved phosphorylation
loop of the sarcoplasmic reticulum Ca2+-ATPase in MgATP binding and
catalytic activity. J. Biol. Chem. 276, 35741-50 (2001).
[11] Patchornik, G., Munson, K., Goldshleger, R., Shainskaya,
A., Sachs, G. & Karlish, S.J. ATP-Mg2+ binding site and cytoplasmic
domain interactions of Na+,K+-ATPase investigated with Fe2+-catalysed
oxidative Cleavage and Modeling. Biochemistry 41, 11740-11749 (2002).
[12] Holmgren, M., Wagg, J., Bezanilla, F., Rakowski,
R.F., De Weer, P. & Gadsby, D.C. Three distinct and sequential
steps in the release of sodium ions by the Na+/K+-ATPase. Nature 403,
898-901 (2000).
[13] Ogawa, H. & Toyoshima, C. Homology modeling of
the cation binding sites of Na+K+-ATPase. PNAS 99, 15977-15982 (2002).
[14] Jorgensen, J.R. & Pedersen, P.A. Role of phylogenetically
conserved amino acids in folding of Na,K- ATPase. Biochemistry 40,
7301-7308 (2001).
[15] Kaplan, J.H. Biochemistry of Na,K-ATPase. Annu. Rev.
Biochem. 71, 511-535 (2002).
[16] Stokes, D.L. & Green, N.M. Structure and Function
of the Calcium Pump. Annu. Rev. Biophys. Biomol. Struct. 32, 445-468
(2003).
|
|