Biochem. J. (2008) 416, 129–136 (Printed in Great Britain)
129
doi:10.1042/BJ20081068
Mg2+ -dependent ATP occlusion at the first nucleotide-binding domain
(NBD1) of CFTR does not require the second (NBD2)
Luba ALEKSANDROV, Andrei ALEKSANDROV and John R. RIORDAN1
Department of Biochemistry/Biophysics and Cystic Fibrosis Center, University of North Carolina, Manning Drive CB 7248, Chapel Hill, NC 27599, U.S.A.
ATP binding to the first and second NBDs (nucleotide-binding
domains) of CFTR (cystic fibrosis transmembrane conductance
regulator) are bivalent-cation-independent and -dependent steps
respectively [Aleksandrov, Aleksandrov, Chang and Riordan
(2002) J. Biol. Chem. 277, 15419–15425]. Subsequent to the
initial binding, Mg2+ drives rapid hydrolysis at the second site,
while promoting non-exchangeable trapping of the nucleotide
at the first site. This occlusion at the first site of functional
wild-type CFTR is somewhat similar to that which occurs when
the catalytic glutamate residues in both of the hydrolytic sites
of P-glycoprotein are mutated, which has been proposed to be
the result of dimerization of the two NBDs and represents a
transient intermediate formed during ATP hydrolysis [Tombline
and Senior (2005) J. Bioenerg. Biomembr. 37, 497–500]. To
test the possible relevance of this interpretation to CFTR, we
have now characterized the process by which NBD1 occludes
[32 P]N3 ATP (8-azido-ATP) and [32 P]N3 ADP (8-azido-ADP). Only
N3 ATP, but not N3 ADP, can be bound initially at NBD1 in the
absence of Mg2+ . Despite the lack of a requirement for Mg2+
for ATP binding, retention of the NTP at 37 ◦C was dependent
on the cation. However, at reduced temperature (4 ◦C), N3 ATP
remains locked in the binding pocket with virtually no reduction
over a 1 h period, even in the absence of Mg2+ . Occlusion occurred
identically in a NBD2 construct, but not in purified recombinant
NBD1, indicating that the process is dependent on the influence
of regions of CFTR in addition to NBD1, but not NBD2.
INTRODUCTION
In wild-type CFTR (cystic fibrosis transmembrane conductance
regulator), there is a serine residue at the consensus position of
the catalytic glutamate residue in the Walker B motif of NBD1,
and ATP is stably bound at this site rather than turning over
rapidly as it does at the NBD2 site, where glutamate is present
[10,11]. In the present study, we have characterized this normal
trapping of the nucleotide with respect to Mg2+ - and temperaturedependence and determined whether it depended on NBD1 or
NBD2 association. The results indicate that the occlusion of ATP
at NBD1 is maintained by NBD1 in the absence of NBD2, but
is also dependent on the influence of some other portions of
CFTR.
ABC (ATP-binding cassette transporter) ATPases generally do
not exhibit high-affinity binding of ATP [1]; however, the
nucleotide may be trapped in the two composite binding sites
at the interface between these two opposed NBDs (nucleotidebinding domains) under some experimental conditions [2]. The
proposed catalytic intermediates are formed in the presence of
the so-called transition-state inhibitors orthovanadate, beryllium
fluoride and aluminum fluoride or by mutagenesis of catalytic
glutamate residues in the Walker B motifs of the NBDs [3]. Such
experimental manipulations have been used to great advantage to
study the mechanism of ATP hydrolysis [4] and, in some cases,
the relationship to the coupled transport process by several ABC
transporters [5]. For example, the use of these transition-state
intermediates has helped in dissecting the kinetics of ATP binding
and NBD dissociation by the homodimeric mitochondrial ABC
transporter Mdl1 (multidrug-resistance-like protein 1) [6]. In the
case of the P-gp (P-glycoprotein) multidrug transporter, mutation
of the catalytic Walker B glutamate residues in both NBDs results
in an ATP-occluded state in which the nucleotide is very tightly
bound [7,8]. Since the binding affinity of a drug transported by
P-gp is reduced in this state, it has been proposed that
conformational change involved in this tight ATP trapping is
coupled directly to drug transport [9]. It has also been suggested
that the occluded state is maintained by the dimerization of the
two NBDs [2,8]. In this view, the two NBDs brought together
by two ATPs in a so-called ‘nucleotide sandwich dimer’ are kept
together when hydrolysis cannot occur, and the bound nucleotides
are unable to escape.
Key words: ATP-binding cassette transporter protein (ABC
protein), ATP occlusion, cystic fibrosis, cystic fibrosis
transmembrane conductance regulator (CFTR), nucleotide
sandwich.
EXPERIMENTAL
Materials
Stable BHK-21 (baby-hamster kidney) cell lines expressing
full-length and NBD2 CFTR (terminating at residue 1172)
were generated and maintained as described previously
[12,13]. [α-32 P]N3 ATP (8-azido-ATP), [γ -32 P]N3 ATP and [α32
P]N3 ADP (8-azido-ADP) were obtained from Affinity
Labeling Technologies. Unlabelled nucleotide, Tos-Phe-CH2 Cl
(tosylphenylalanylchloromethane or ‘TPCK’)-treated trypsin,
protease inhibitors and other chemicals were obtained from
Sigma. Purified NBD1 protein was provided by P. J. Thomas
(University of Texas, Southwestern Medical Center, Dallas, TX,
U.S.A.).
Abbreviations used: ABC, ATP-binding cassette; BHK, baby-hamster kidney; CFTR, cystic fibrosis transmembrane conductance regulator; NBD,
nucleotide-binding domain; P-gp, P-glycoprotein; PKA, protein kinase A; N3 ADP, 8-azido-ADP; N3 ATP, 8-azido-ATP; Tos-Phe-CH2 Cl, tosylphenylalanylchloromethane (‘TPCK’).
1
To whom correspondence should be addressed (email [email protected]).
c The Authors Journal compilation c 2008 Biochemical Society
130
L. Aleksandrov, A. Aleksandrov and J. R. Riordan
Figure 1
Photoaffinity labelling of the two CFTR NBDs
(A) [γ - P]N3 ATP photolabelling at 4 or 37 ◦C at various MgCl2 concentrations. Membranes (20 μg) from CFTR-expressing BHK-21 cells were incubated for 5 min and then irradiated for 2 min in
a Stratalinker UV cross-linker (λ = 254 nm) on ice. Following limited trypsin digestion (15 min on ice) to generate fragments containing N- and C-terminal ‘halves’ and NBD1 and NBD2 of CFTR,
and solubization in RIPA buffer, immunoprecipitation was performed using two monoclonal antibodies (L12B4 recognizing an epitope in NBD1 and 596 recognizes one in NBD2). Washed immune
complexes were resolved by SDS/PAGE on 4–20 % acrylamide gels and visualized by autoradiography. (B) Photolabelling as in (A) with [γ -32 P]N3 ATP, [α-32 P]N3 ATP or [α-32 P]N3 ADP as indicated
in the absence (−) or presence (+) of 2 mM MgCl2 at either 4 or 37 ◦C. Results are representative of more than two experiments with different CFTR-containing membrane preparations. Molecular
masses (in kDa) are indicated in both panels.
32
Membrane isolation
CFTR-expressing BHK cells were harvested by scraping and
homogenized on ice in 10 mM Hepes (pH 7.2) and 1 mM
EDTA containing a protease inhibitor cocktail (benzamidine
at 120 μg/ml, E-64 [trans-epoxysuccinyl-L-leucylamido-(4guanido)butane] at 3.5 μg/ml, aprotinin at 2 μg/ml, leupeptin
at 1 μg/ml and Pefabloc at 50 μg/ml). Centrifugation at 600 g
for 15 min removed nuclei and undisrupted cells. The supernatant
was centrifuged at 100 000 g for 30 min to pellet membranes,
which were resuspended in 40 mM Tris/HCl (pH 7.4) and 0.1 mM
EGTA with or without 5 mM MgCl2 .
Photoaffinity labelling
Assays of azido-nucleotide binding and retention by the two
NBDs of CFTR were performed essentially as described
previously [10,12] with special attention paid to monitoring the
rate of turnover of the bound nucleotide at each site after the initial
binding. First, to monitor binding, membrane suspensions (20 μg
of protein) were incubated either at 4 or 37 ◦C for 5 min with
25 μM [32 P]azido-nucleotide as indicated in the Figure legends.
The suspension was then irradiated at 254 nm in a Stratalinker
UV cross-linker for 2 min either before or after pelleting to
wash away unbound nucleotide. The labelled membranes were
then either subjected to limited trypsin digestion, SDS/PAGE and
autoradiography in order to quantify binding to each of the sites
or re-incubated for various periods of time in nucleotide-free
buffer after washing away unbound nucleotide before irradiation
to monitor the stability of the bound nucleotide at each site.
c The Authors Journal compilation c 2008 Biochemical Society
Figure 2 Bivalent cation requirement for competition with N3 ATP binding
to NBD1 by ADP, but not by ATP
[γ -32 P]N3 ATP photolabelling at 4 ◦C as in Figure 1, in the presence of the ATP or ADP
concentrations indicated. The experiment was repeated several times with very similar results.
CFTR ATP occlusion is independent of NBD dimerization
Figure 3
131
Retention of N3 ATP at NBD1 at 37 ◦C is Mg2+ -dependent
Membranes were incubated with 25 μM [γ -32 P]N3 ATP for 5 min at 37 ◦C, pelleted and washed free of the radiolabelled nucleotide before re-incubation for the times indicated. The absence or
presence of 2 mM MgCl2 during each of the three steps is indicated above each panel. Immunoprecipitation was with the L12B4 antibody recognizing an NBD1 epitope. The relative amounts of 32 P
radioactivity in the NBD1-containing bands determined by electronic autoradiography are plotted in the graph. Representative results from several such experiments are shown. Molecular masses
(in kDa) are indicated in (A)–(D).
Limited trypsin digestion and immunoprecipitation
Labelled membranes following azido-nucleotide binding or
binding and release were subjected to limited trypsin digestion by
incubation with Tos-Phe-CH2 Cl-treated trypsin for 15 min on ice
with a protease/membrane protein mass ratio of 1:180. Digestion
was stopped with soya bean trypsin inhibitor and solubilization
was with radioimmune precipitation (RIPA) buffer (50 mM
Tris/HCl, pH 7.4, 150 mM NaCl, 1 % sodium deoxycholate,
1 % Triton X-100 and 0.1 % SDS). CFTR fragments containing
NBD1 and NBD2 [10,12] were immunoprecipitated with
immobilized mouse monoclonal antibodies L12B4 and 596 [13].
The immunoprecipitates were resolved by SDS/PAGE (4–20 %
acrylamide) and transferred on to nitrocellulose membranes
for autoradiography [X-ray films and Packard Instant Imager
(PerkinElmer) for quantification of 32 P radioactivity associated
with NBD1- and NBD2-containing bands].
RESULTS
Nucleotide binding to CFTR
The two composite NBDs of CFTR are quite distinct with
degenerate Walker B and signature sequences in the first and
strict consensus sequences at these locations in the second. As
a consequence, ATP is hydrolysed slowly or not at all at the
first site, but efficiently at the second site [10]. ATP binding
at both sites in intact membrane-bound CFTR can be monitored
by photolabelling with [γ -32 P]N3 ATP as shown in Figure 1(A). In
this experiment, CFTR-containing membranes were incubated
with the nucleotide in the presence of various concentrations
of Mg2+ for 5 min at 4 or 37 ◦C. UV irradiation then caused
attachment of the nucleotide to the NBD containing the Walker A
and B motifs of each composite site. Limited trypsin digestion
produced fragments containing each labelled NBD which
were immunoprecipitated with specific antibodies, separated by
SDS/PAGE and detected by autoradiography. The two larger
fragments containing NBD1 and NBD2 are produced by a
cleavage in the R domain near the middle of the CFTR sequence,
whereas the two smaller fragments are produced by additional
cleavages that excise the NBDs [12]. Since 32 P is present at
the γ -phosphate position of N3 ATP, the binding of the intact
NTP is detected. Most striking is the inverse bivalent-cationdependence of binding to the two sites. Binding to NBD1 clearly
does not require a bivalent cation, and increasing concentrations
actually reduce binding (Figure 1A). The addition of EDTA or
other chelating agents did not change the level of binding from
that with no added cation (results not shown). Binding to NBD2
appears to be entirely dependent on Mg2+ , increasing with elevated
concentrations at 4 ◦C and saturating by 0.5 mM (left-hand panel
of Figure 1A). At 37 ◦C, [γ 32 P]N3 ATP is very rapidly hydrolysed
at NBD2 and hence there is no signal in association with NBD2containing bands, whereas, even at this higher temperature, the
nucleotide is bound to NBD1 with or without Mg2+ (right-hand
panel of Figure 1A).
To determine whether the first site could bind ADP, as well
as ATP, in a bivalent-cation-independent manner, two types of
c The Authors Journal compilation c 2008 Biochemical Society
132
Figure 4
L. Aleksandrov, A. Aleksandrov and J. R. Riordan
Enhanced retention of N3 ATP at NBD1 at reduced temperature (4 ◦C)
Incubation (binding), washing and re-incubation (release) was as in Figure 3, but with 2 mM MgCl2 present in all cases. Binding temperatures indicate those of incubation. Release temperatures
indicate those of re-incubation. Processing and analysis of samples was as in Figure 3. The experiment was repeated with the release time extended to 1 h with essentially identical results. The values
shown in the graph are means +
− S.D. for two 30 min experiments.
experiment were performed. First, the binding of [γ -32 P]N3 ATP,
[α-32 P]N3 ATP and [α-32 P]N3 ADP were compared (Figure 1B).
As expected, non-hydrolytic [α-32 P]N3 ATP binding (middle
panel) like that of [γ -32 P]N3 ATP (left-hand panel) did not
require Mg2+ at either temperature. In contrast, [α-32 P]N3 ADP
binding (right-hand panel) appeared to be completely Mg2+ dependent. Secondly, the relative abilities of unlabelled ATP
and ADP to compete for [γ -32 P]N3 ATP binding were tested.
Although increasing concentrations of unlabelled ATP competed
effectively, both in the presence and absence of Mg2+ (Figure 2A),
ADP was only effective in the presence of the cation (Figure 2B).
Thus both types of experiment indicated that the binding of the
NDP to NBD1 requires Mg2+ , whereas binding of the NTP does
not. At NBD2, the bivalent cation is clearly required for the
binding of both the NDPs and NTPs (compare NBD2 bands in
the left- and right-hand panels of Figures 2A and 2B).
Nucleotide retention by NBD1
Having found that Mg2+ is unnecessary for the initial binding
of N3 ATP to NBD1, we examined whether it plays a role in
the long-term retention of the unhydrolysed nucleotide by this
site. To do this, the dissociation of [γ -32 P]N3 ATP was monitored
under conditions where Mg2+ was either present or absent during
the initial binding, during the washing away of free nucleotide
after binding and during re-incubation in nucleotide-free buffer
after washing. As observed previously [10], when Mg2+ is present
during all three steps, the amount of [γ -32 P]N3 ATP bound at the
first site decreases quite slowly (t1/2 ∼ 30 min; Figure 3B). In sharp
c The Authors Journal compilation c 2008 Biochemical Society
contrast, when the cation is absent at all stages, dissociation occurs
very rapidly (Figure 3A). This rapid loss of signal clearly is not
due to removal of the γ -phosphate from N3 ATP, because it occurs
in the absence rather than in the presence of the bivalent cation
which is required for hydrolysis. This is confirmed further by
the very similar results observed with [α-32 P]N3 ATP (results not
shown). In that case, the rapid loss of NBD1-bound signal in the
absence of Mg2+ cannot reflect the removal of the radioactive γ phosphate by hydrolysis. When Mg2+ was absent during both the
binding and the washing away of the free nucleotide, the presence
of the ion during re-incubation was sufficient to maintain this same
degree of retention (Figure 3C). Thus the independent sequential
binding of ATP and Mg2+ is able to secure the stable association of
the nucleotide. However, when the nucleotide and the cation were
both present during binding, the nucleotide was retained even
when free Mg2+ was washed away (Figure 3D). This may reflect
the retention of tightly bound Mg2+ . It is therefore apparent that,
although ATP and Mg2+ can bind independently to stabilize the
nucleotide in its binding pocket, the presence of the cation either
during or after the binding of nucleotide causes a change in the
pocket, resulting in trapping. All of the steps in these experiments
(Figure 3) were performed at 37 ◦C.
To examine the effect of temperature on the Mg2+ -dependent
ATP occlusion, we monitored the rate of dissociation at 4 ◦C as
well as at 37 ◦C. At the low releasing temperature, the amount of
bound [γ -32 P]N3 ATP did not diminish at all during a 30 min period
(Figures 4A and 4C), whereas dissociation at 37 ◦C (Figures 4B
and D) occurred at a rate similar to that in the presence of Mg2+ in
Figure 3(B). Even though the initial amount bound actually was
CFTR ATP occlusion is independent of NBD dimerization
Figure 5
133
N3 ATP is retained at NBD1 at low temperature even in the absence of bivalent cations
Incubation (binding), washing and re-incubation (release) were performed in the absence or presence of 2 mM MgCl2 at the temperatures indicated. The values shown in the graph are means +
− S.D.
for three experiments.
greater at 37 ◦C (Figures 4C and 4D) than at 4 ◦C (Figures 4A
and 4B), the extent of retention at either temperature was the
same for each binding temperature. Therefore a step following
binding, presumably a conformational change, is responsible for
the tight locking into the binding site. Even though it is quite
stable at physiological temperature, the occluded state appears
to be entirely frozen in place during continuing incubation at
4 ◦C (Figures 4A and 4C). The fact that this ‘frozen state’ is no
longer dependent on Mg2+ is shown clearly by experiments such
as that illustrated in Figure 5. [γ -32 P]N3 ATP remains stably bound
over a 30 min incubation at 4 ◦C whether or not Mg2+ is present
(Figures 5A and 5B), in contrast with incubation at 37 ◦C, where
there is much more rapid dissociation in the absence of Mg2+ than
in its presence (Figures 5C and 5D). Thus the ‘conformational
lock’ at low temperature is more complete and is able to overcome
the requirement for Mg2+ to maintain a partially occluded state.
Although it is not surprising that dissociation should be slower
at lower temperatures regardless of the underlying mechanism,
the very large difference observed is consistent with the notion
of tight retention by a restricted conformational state probably
involving regions of CFTR beyond NBD1 (see Figures 6 and 7
and results below).
ATP occlusion at NBD1 does not require NBD2
It has been shown recently that C-terminally truncated CFTR
constructs not containing NBD2 mature conformationally and
are transported to the cell surface where they generate low-level
Cl− channel activity [13]. To determine whether the presence of
NBD2 was required for ATP occlusion at NBD1, we compared
the rates of [γ -32 P]N3 ATP dissociation at 37 ◦C in the presence of
2 mM MgCl2 using full-length and NBD2 CFTR (Figure 6A).
These low rates are virtually identical in both cases. As with
the full-length protein, Mg2+ ions are not required at 4 ◦C for
N3 ATP binding to the truncated protein (Figure 6B). However,
it exhibits the same dependence on Mg2+ for the competition by
ADP of N3 ATP binding as does the full-length CFTR (Figure 6C).
Thus the properties of N3 ATP binding and occlusion at NBD1
appear to be identical in full-length and NBD2 CFTR. The
nucleotide occluded in NBD2 CFTR appears to be entirely nonexchangeable, as shown in Figures 6(D) and 6(E), where there is
no increase in the rate of dissociation in the presence of an excess
of unlabelled nucleotide (2 mM ATP).
To determine whether NBD1 alone is capable of occluding
ATP, the interaction of [γ -32 P]N3 ATP with purified recombinant
NBD1 also was measured (Figure 7). As can be seen, the
nucleotide binds in a Mg2+ -independent manner, but is not
retained when free nucleotide is washed away, in contrast with
either full-length or NBD2 CFTR, from which it is not removed
by washing [12]. Thus, although nucleotide retention at NBD1
occurs in full-length and NBD2 CFTR, additional regions of the
protein excluding NBD2 are required. This is consistent with the
finding that, although each CFTR domain may form a compact
structure individually [14] and NBD1 crystallizes as a conserved
c The Authors Journal compilation c 2008 Biochemical Society
134
Figure 6
L. Aleksandrov, A. Aleksandrov and J. R. Riordan
Retention of N3 ATP at NBD1 of NBD2 CFTR
Membranes from cells expressing CFTR truncated at residue 1172 between MSD2 and NBD2 [13] were incubated with [γ -32 P]N3 ATP and analysed in the same way as the full-length protein.
(A) Time course of retention by full-length and NBD2 CFTR at 37 ◦C. (B) Mg2+ -independence of [γ -32 P]N3 ATP binding to NBD1 of NBD2 CFTR at 4 ◦C. (C) Mg2+ -dependence of competition
by ADP of [γ -32 P]N3 ATP binding to NBD1 of NBD2 CFTR. (D) NBD1-occluded nucleotide is non-exchangeable, i.e. incubation with 2 mM ATP does not promote dissociation. (E) Quantification
of autoradiogram in (D) by electronic radiography. Each experiment was repeated at least twice. Results are means +
− S.D. for three experiments. Molecular masses (in kDa) are indicated in (A)–(D).
elaborately folded structure [15], interdomain contacts are crucial
to their achievement of a final mature functional state [13,16].
DISCUSSION
A distinguishing feature of ABC proteins is the formation
of two composite ATP-binding sites at the interface between
their two NBDs [17]. In some ABCs, two copies of a single
NBD subunit form a head-to-tail homodimer, whereas in others,
highly homologous, but distinct, NBDs form the two sites where
two ATPs are bound and hydrolysed. However, there are many
ABC proteins in which the NBDs are highly asymmetric, so that
only one of the two composite sites catalyses ATP hydrolysis
at a significant rate [18]. CFTR belongs to this latter group,
with degenerate sites at the positions of the Walker B glutamate
residue and histidine loop histidine residue in NBD1, as well as
the signature sequence in NBD2. Since residues at all three of
these positions contribute to ATP hydrolysis by ABC proteins
[19], it is not surprising that there is a very low rate of turnover at
NBD1. In the present study, we have examined the ability of this
site to bind and retain nucleotides.
c The Authors Journal compilation c 2008 Biochemical Society
Using the intact functional membrane-bound protein to monitor
nucleotide interactions at both sites simultaneously, NBD1 was
observed to bind N3 ATP in the absence of Mg2+ ions. In this
regard, it is of interest to mention the influence of re-addition
of Mg2+ ions to phosphorylated CFTR channels in guinea-pig
ventricular myocyte membranes from which ATP and Mg2+ had
been washed out [20]. Re-exposure to Mg2+ in the absence of
fresh bulk ATP was sufficient for closure of channels which
had remained opened for tens of seconds after both agents
were removed. This was interpreted as reflecting Mg2+ -supported
hydrolysis of ATP that had remained bound at NBD2 at least over
the short time scale of these experiments. Our current observations
with unphosphorylated membrane-bound human CFTR indicate
that ATP would not remain bound at NBD2 over our longer time
scale. This is because NBD2 requires bivalent cations to bind
ATP, and, in the presence of Mg2+ , all of the bound nucleotide
is hydrolysed rapidly. On the other hand, the initial bivalentcation-independent binding at NBD1 was rapidly reversible.
However, when Mg2+ was present, either during binding or
added afterwards, N3 ATP then dissociated much more slowly
(t1/2 ∼ 30 min at 37 ◦C). The reduction rather than an increase
CFTR ATP occlusion is independent of NBD dimerization
Figure 7
Isolated recombinant NBD1 binds, but does not occlude, N3 ATP
NBD1 protein (0.3 μg) was photolabelled with 25 μM [γ -32 P]N3 ATP in the presence or absence
of 2 mM MgCl2 as indicated. (A) Samples were subjected to SDS/PAGE and autoradiography
and Western blotting. (B) Protein was bound by immobilized antibody L12B4 before N3 ATP
binding and either washed free of the radiolabelled nucleotide or not, as indicated, before
UV irradiation and analysis. As only limited quantities of purified NBD1 are available, these
experiments were only been repeated once, but with identical results. Molecular masses (in kDa)
are indicated in both panels.
in the rate of diminution of the NBD1-associated 32 P signal by
the bivalent cation which is essential for hydrolysis attested to the
fact that dissociation of the whole NTP rather than hydrolysis was
being monitored. When the temperature was reduced to 4 ◦C, the
nucleotide was even more tenaciously trapped, with essentially
no dissociation detected over periods as long as 1 h. These
results suggest that NBD1 undergoes a nucleotide-ligand-induced
conformational change that traps ATP in a non-exchangeable
state. Direct evidence of the lack of exchange is provided by
Figures 6(D) and 6(E). This is reminiscent of what occurs in the
P-gp multidrug transporter when the catalytic glutamate residues
in the Walker B motifs of both NBDs are replaced [2]. This
occlusion, which has been characterized extensively, is proposed
to represent a stable form of a transient intermediate state in the
normal hydrolytic mechanism [8] in which the unhydrolysed ATP
is trapped between the two NBDs. However, with CFTR, this does
not appear to be the case, because the nucleotide is retained as
tightly by a truncated CFTR not containing NBD2 as by the fulllength protein containing both NBDs. On this basis, NBD1 alone
135
would seem to be responsible for the ligand-induced locking-in
of the nucleotide.
Thus the present findings add significantly to our knowledge
of the action of ATP at the NBDs of CFTR in several ways.
First, the long-lived occlusion at NBD1, which contrasts the rapid
hydrolytic turnover at NBD2, does not require the presence of
NBD2. This indicates clearly that NBD1 is entirely responsible for
the tight gripping of the nucleotide and that the NBD2 signature
sequence, envisaged in generic models of ABC proteins to interact
with the γ -phosphate of the NBD1-bound ATP, does not play any
role in the binding or occlusion reaction. Secondly, the ability of
NBD1 to clasp the occluded nucleotide requires the presence
of other CFTR domains because the isolated NBD that does
reversibly bind ATP is unable to trap it. Membrane-integrated
domains of CFTR contact NBD1 via their cytoplasmic loops [21]
and could influence its ability to grasp the nucleotide. Further
experiments will be required to investigate the possible role of
these contacts in ATP occlusion at NBD1. It is worth mentioning
that all of the results described were obtained in the absence of
PKA (protein kinase A) phosphorylation, which is obligatory for
activation of the CFTR channel. PKA phosphorylation does cause
a small decrease in the K m for hydrolysis of ATP at NBD2 [22] and
promotes association of the two NBDs [23]. In parallel with this
latter effect, we speculate that phosphorylation of the R domain
also may influence the occluded ATP state at NBD1 subsequently
to its formation.
The relevance of the stable interaction of ATP with NBD1
of CFTR to the physiological function of the protein also
remains to be fully elucidated. Findings of both the present
and previous studies [10,11], as well as the absence of key
catalytic residues (serine residues replace the so-called catalytic
glutamate residue in the Walker B motif and the His loop histidine)
argue strongly that significant hydrolysis does not occur at this
site. Replacements by in vitro mutagenesis of the consensus
lysine residue in the Walker A motif perturbs channel gating,
accelerating its closing [24]. The unhydrolysed ATP is released
with a t1/2 of approx. 30 min at 37 ◦C under our experimental
conditions. Nevertheless, we have not eliminated the possibility
that release from NBD1 might occur much more rapidly under
different conditions. Further extension of both biochemical and
electrophysiological analyses are necessary to clarify whether,
and to what extent, turnover of the nucleotide stably bound at
NBD1 may play a role in the mechanism of channel regulation.
We thank Dr P. J. Thomas, University of Texas, Southwestern Medical Center, Dallas, TX,
U.S.A., for providing the purified NBD1 protein. This work was supported by the National
Institute of Diabetes and Digestive and Kidney Disease of the National Institutes of Health
(DK051619).
REFERENCES
1 Al-Shawi, M. K. and Senior, A. E. (1993) Characterization of the adenosine triphosphatase
activity of Chinese hamster P-glycoprotein. J. Biol. Chem. 268, 4197–4206
2 Tombline, G. and Senior, A. E. (2005) The occluded nucleotide conformation of
p-glycoprotein. J. Bioenerg. Biomembr. 37, 497–500
3 Senior, A. E., al-Shawi, M. K. and Urbatsch, I. L. (1995) The catalytic cycle of
P-glycoprotein. FEBS Lett. 377, 285–289
4 Tombline, G., Muharemagic, A., White, L. B. and Senior, A. E. (2005) Involvement of the
“occluded nucleotide conformation” of P-glycoprotein in the catalytic pathway.
Biochemistry 44, 12879–12886
5 Sauna, Z. E., Smith, M. M., Muller, M., Kerr, K. M. and Ambudkar, S. V. (2001)
The mechanism of action of multidrug-resistance-linked P-glycoprotein.
J. Bioenerg. Biomembr. 33, 481–491
6 Van der Does, C., Presenti, C., Schulze, K., Dinkelaker, S. and Tampe, R. (2006) Kinetics
of the ATP hydrolysis cycle of the nucleotide-binding domain of Mdl1 studied by a novel
site-specific labeling technique. J. Biol. Chem. 281, 5694–5701
c The Authors Journal compilation c 2008 Biochemical Society
136
L. Aleksandrov, A. Aleksandrov and J. R. Riordan
7 Tombline, G., Bartholomew, L. A., Urbatsch, I. L. and Senior, A. E. (2004) Combined
mutation of catalytic glutamate residues in the two nucleotide binding domains of
P-glycoprotein generates a conformation that binds ATP and ADP tightly. J. Biol. Chem.
279, 31212–31220
8 Sauna, Z. E., Nandigama, K. and Ambudkar, S. V. (2006) Exploiting reaction intermediates
of the ATPase reaction to elucidate the mechanism of transport by P-glycoprotein
(ABCB1). J. Biol. Chem. 281, 26501–26511
9 Martin, C., Higgins, C. F. and Callaghan, R. (2001) The vinblastine binding site adopts
high- and low-affinity conformations during a transport cycle of P-glycoprotein.
Biochemistry 40, 15733–15742
10 Aleksandrov, L., Aleksandrov, A. A., Chang, X. B. and Riordan, J. R. (2002) The first
nucleotide binding domain of cystic fibrosis transmembrane conductance regulator is a
site of stable nucleotide interaction, whereas the second is a site of rapid turnover.
J. Biol. Chem. 277, 15419–15425
11 Basso, C., Vergani, P., Nairn, A. C. and Gadsby, D. C. (2003) Prolonged nonhydrolytic
interaction of nucleotide with CFTR’s NH2 -terminal nucleotide binding domain and its role
in channel gating. J. Gen. Physiol. 122, 333–348
12 Aleksandrov, L., Mengos, A., Chang, X., Aleksandrov, A. and Riordan, J. R. (2001)
Differential interactions of nucleotides at the two nucleotide binding domains of the cystic
fibrosis transmembrane conductance regulator. J. Biol. Chem. 276, 12918–12923
13 Cui, L., Aleksandrov, L., Chang, X. B., Hou, Y. X., He, L., Hegedus, T., Gentzsch, M.,
Aleksandrov, A., Balch, W. E. and Riordan, J. R. (2007) Domain interdependence in the
biosynthetic assembly of CFTR. J. Mol. Biol. 365, 981–994
14 Kleizen, B., van Vlijmen, T., de Jonge, H. R. and Braakman, I. (2005) Folding of CFTR is
predominantly cotranslational. Mol. Cell 20, 277–287
15 Lewis, H. A., Buchanan, S. G., Burley, S. K., Conners, K., Dickey, M., Dorwart,
M., Fowler, R., Gao, X., Guggino, W. B., Hendrickson, W. A. et al. (2004) Structure of
nucleotide-binding domain 1 of the cystic fibrosis transmembrane conductance regulator.
EMBO J. 23, 282–293
Received 27 May 2008/26 June 2008; accepted 8 July 2008
Published as BJ Immediate Publication 8 July 2008, doi:10.1042/BJ20081068
c The Authors Journal compilation c 2008 Biochemical Society
16 Younger, J. M., Chen, L., Ren, H. Y., Rosser, M. F., Turnbull, E. L., Fan, C. Y., Patterson, C.
and Cyr, D. M. (2006) Sequential quality-control checkpoints triage misfolded cystic
fibrosis transmembrane conductance regulator. Cell 126, 571–582
17 Karpowich, N., Martsinkevich, O., Millen, L., Yuan, Y. R., Dai, P. L., MacVey, K., Thomas,
P. J. and Hunt, J. F. (2001) Crystal structures of the MJ1267 ATP binding cassette reveal
an induced-fit effect at the ATPase active site of an ABC transporter. Structure 9, 571–586
18 Jones, P. M. and George, A. M. (2000) Symmetry and structure in P-glycoprotein and
ABC transporters: what goes around comes around. Eur. J. Biochem. 267,
5298–5305
19 Zaitseva, J., Jenewein, S., Jumpertz, T., Holland, I. B. and Schmitt, L. (2005) H662 is the
linchpin of ATP hydrolysis in the nucleotide-binding domain of the ABC transporter HlyB.
EMBO J. 24, 1901–1910
20 Dousmanis, A. G., Nairn, A. C. and Gadsby, D. C. (2002) Distinct Mg2+ -dependent steps
rate limit opening and closing of a single CFTR Cl− channel. J. Gen. Physiol. 119,
545–559
21 Serohijos, A. W., Hegedus, T., Aleksandrov, A. A., He, L., Cui, L., Dokholyan, N. V. and
Riordan, J. R. (2008) Phenylalanine-508 mediates a cytoplasmic-membrane domain
contact in the CFTR 3D structure crucial to assembly and channel function. Proc. Natl.
Acad. Sci. U.S.A. 105, 3256–3261
22 Li, C., Ramjeesingh, M. and Bear, C. E. (1996) Purified cystic fibrosis transmembrane
conductance regulator (CFTR) does not function as an ATP channel. J. Biol. Chem. 271,
11623–11626
23 Mense, M., Vergani, P., White, D. M., Altberg, G., Nairn, A. C. and Gadsby, D. C. (2006)
In vivo phosphorylation of CFTR promotes formation of a nucleotide-binding domain
heterodimer. EMBO J. 25, 4728–4739
24 Powe, Jr, A. C., Al-Nakkash, L., Li, M. and Hwang, T. C. (2002) Mutation of Walker-A
lysine 464 in cystic fibrosis transmembrane conductance regulator reveals functional
interaction between its nucleotide-binding domains. J. Physiol. 539,
333–346