NMR Investigations of Allosteric Processes
in a Two-domain Thermus thermophilus Hsp70
Molecular Chaperone
Matthew Revington, Yongbo Zhang, Groverv N. B. Yip
Alexander V. Kurochkin and Erik R. P. Zuiderweg
*
Biophysics Research Division
and Departments of Biological
Chemistry and Chemistry
The University of Michigan
930 N, University Avenue
Ann Arbor, MI 48109-1055
USA
Hsp70 chaperones are two-domain proteins that assist in intra-cellular
protein (re) folding processes in all species. The protein folding activity of
the substrate binding domain of the Hsp70s is regulated by nucleotide
binding at the nucleotide-binding domain through an as yet undefined
heterotropic allosteric mechanism. The available structures of the isolated
domains of Hsp70s have given very limited indications of nucleotide-
induced conformational changes that could modulate the affinity for
substrate proteins. Here, we present a multi-dimensional NMR study of a
prokaryotic Hsp70 homolog, Thermus thermophilus DnaK, using a 54 kDa
construct containing both nucleotide binding domain and most of the
substrate binding domain. It is determined that the nucleotide binding
domain and substrate binding domain are closely associated in all ligand
states studied. Comparison of the assigned NMR spectra of the two-
domain construct with those of the previously studied isolated nucleotide
binding domain, allowed the identification of the nucleotide binding
domain-substrate binding domain interface. A global three-dimensional
structure was obtained for the two-domain construct on the basis of this
information and of NMR residual dipolar couplings measurements. This is
the first experimental elucidation of the relative positioning of the
nucleotide binding domain and substrate binding domain for any Hsp70
chaperone.
Comparisons of NMR data between various ligand states including
nucleotide-free, ATP, ADP.Pi and ADP.PiC peptide bound, identified
residues involved in the allosteric inter-domain communication. In
particular, peptide binding to the substrate binding domain was found to
cause conformational changes in the NBD extending to the nucleotide
binding pocket. Detailed analysis suggests that the inter-domain interface
becomes tighter in the (nucleotide binding domain ligation/substrate
binding domain ligation) order ATP/apo, ADP.Pi/apo ADP.Pi/peptide.
q
2005 Elsevier Ltd. All rights reserved.
Keywords: Hsp70; NMR; allostery; conformational change; residual dipolar
couplings
*Corresponding author
Introduction
The Hsp70 (70 kDa heat shock protein) class of
molecular chaperones is expressed ubiquitously in
both prokaryotic (called DnaK) and eukaryotic
organisms (called SSa, SSb, SSc in yeast and
Hsp70, Hsc70 and Bip in mammals). Under normal
growth conditions, the Hsp70s are directly engaged
in folding and re-folding of proteins, and protein
trafficking.
Under stress conditions, cells express
Hsp70 proteins at elevated levels in order to
0022-2836/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
Present addresses: M. Revington, Department of
Biochemistry, University of Western Ontario, London,
Ont., Canada N6A 5C1; Y. Zhang, Department of
Biochemistry, Molecular Biology, and Cell Biology,
Northwestern University, 2205 Tech Drive, Hogan 2-100,
Evanston, IL 60208, USA.
Abbreviations used: Hsp70, 70 kDa heatshock protein;
NBD, nucleotide binding domain; SBD, substrate binding
domain; RDC, residual dipolar coupling; GDO, general
degree of order; TROSY, transverse relaxation optimized
spectroscopy.
E-mail address of the corresponding author:
zuiderwe@umich.edu
doi:10.1016/j.jmb.2005.03.033
J. Mol. Biol. (2005) 349, 163–183
manage the high levels of damaged proteins; for
instance, in humans, 11 genes express closely
related Hsp70s forming up to 1.5% of intra-cellular
protein mass in stressed cells.
The chaperone
activity of Hsp70 proteins depends on the binding
of exposed hydrophobic sequences of unfolded
proteins, thereby preventing misfolding and the
formation of aggregates.
The binding of the
unfolded substrate occurs in a hydrophobic pocket
located in the 25 kDa C-terminal substrate binding
domain (SBD).
The affinity for substrate binding
increases by several orders of magnitude when ATP,
bound by the 42 kDa N-terminal nucleotide binding
domain (NBD), is hydrolyzed to ADP and inorganic
phosphate, characterizing the Hsp70s as hetero-
tropic allosteric systems.
The allostery is bi-direc-
tional with peptide binding increasing the rate of
ATP hydrolysis. In vivo this interdomain allosteric
regulation is enhanced by multiple co-chaperone
proteins that modulate substrate binding as well as
nucleotide hydrolysis and exchange.
High-resolution three-dimensional structures of
isolated forms of both major structural domains in
various ligand states are available,
but none
for any construct containing both domains. The
approximate relative spatial orientation of the two
domains and even the extent of interdomain contact
in any of the substrate states have not been clearly
defined. Small angle X-ray scattering studies
(SAXS) of Escherichia coli DnaK (DnaK
Eco
)
in the
ATP bound state suggested that the two domains
were associated in the apo state and dissociated
upon binding ATP. Conversely, SAXS analysis of
the highly homologous bovine Hsc70
indicated
that the domains were associated in the ATP state
and dissociated upon hydrolysis to ADP.Pi state.
A recent attempt was made
to model full-length
Hsc70 that was based largely on mutational data.
We investigate by solution NMR the Hsp70
protein of the thermophilic bacterium Thermus
thermophilus (DnaK
Tth
). No experimental three-
dimensional structures are known for this protein,
but DnaK
Tth
exhibits strong sequence and func-
tional homology to the well-studied DnaK
Eco
(see
) and bovine Hsc70.
The full-length
thermophilic protein is optimally functional at
75 8C and is stable to at least 90 8C.
In vivo, it
exists predominantly as part of a multimeric
complex with the co-factors DnaJ and DafA.
Recently, we presented an NMR study comparing
the ATP and ADP-Pi bound states of the isolated
42 kDa NBD of DnaK
TTH
.
It was found that the
ATP binding restricted the NBD to a single
conformational state while hydrolysis to ADP-Pi
resulted in a greater mobility of the subdomains in
the NBD that allowed them to interchange between
a conformation similar to the ATP bound state and a
second conformation unique to the ADP-Pi state.
Here, we extend these NMR studies to a 54 kDa
two-domain construct of DnaK
Tth
which encom-
passes the complete 42 kDa nucleotide binding
domain and a truncated 12 kDa substrate binding
domain (DnaK
Tth-NBD-SBD
). The current construct is
monomeric at concentrations less than 300 mM,
stable to at least 90 8C and shows allosteric
communication between the two domains upon
peptide and nucleotide binding. A model for the
global three-dimensional structure was obtained for
the two-domain construct at 55 8C on the basis of
NMR residual dipolar couplings (RDCs) and
chemical shift mapping. Our result is the first
experimental model reporting the relative position-
ing of the NBD and SBD for any Hsp70 chaperone.
Peptide binding to the SBD was found to cause
conformational changes in the NBD extending from
the interface to the nucleotide-binding pocket.
Detailed analysis suggests that the NBD–SBD
interface becomes tighter in the order (NBD
ligation/SBD ligation) ATP/apo, ADP.P
i
/apo
ADP.P
i
/peptide.
Results
Properties of the two-domain construct
For a detailed NMR study it was necessary to
produce a functional form of DnaK
Tth
that was
monodisperse and stable for a period of weeks or
months to allow multidimensional, triple resonance
spectra to be collected. Initial NMR spectra and
light scattering measurements indicated, however,
that at NMR concentrations (above w200 mM) the
full-length, wild-type DnaK
Tth
was largely in multi-
meric complexes, and was also subject to proteoly-
sis in the interdomain linker. The aggregation was
not surprising since wild-type DnaK
Tth
has been
found to be in an equilibrium between monomeric
and trigonal states.
Inter-domain proteolysis has
precluded detailed structural investigations of the
Hsp70 proteins in general; it is also a problem for
two-domain constructs of DnaK
Eco
and bovine
Hsc70 (Drs A. Joachimiak & D. McKay, personal
communications, and unpublished observations
from our laboratory). Our previously published
work
showed that the isolated NBD of DnaK
Tth
was monomeric and stable in solution. Several
reports have indicated that interactions that lead to
oligomerization in the Hsp70 family of chaperones
are mediated through residues near the C termi-
nus,
suggesting that deletion of the C-terminal
subdomain would decrease aggregation. More
importantly, several studies have shown that
deletion of the helical domain does not affect the
interdomain
allosteric
communication.
DnaK
Tth
was therefore truncated at residue 501
(out of 620 residues), resulting in a protein that is
homologous to the functional, truncated form of
DnaK
Eco
This truncated construct contained the
entire NBD, the ten-residue hydrophobic linker
(DVVLLDVTPL), the beta-strand sandwich section
of the SBD that contains the substrate binding cleft,
and a six-residue C-terminal histidine tag.
Light scattering studies showed that the trun-
cated construct had a reduced tendency to aggre-
gate, but still exhibited a significant fraction of its
164
Structural Indicators of Allostery in DnaK-Tth
Table 1.
DnaK-Eco DnaK-Tth sequence alignment and chemical shift perturbations
DnaK-Eco
Count
DnaK
-Tth
Count
Assigned
Dock
ATP
Pept
Rem
DnaK
-Eco
Count
DnaK
-Tth
Count
Assigned
Dock
ATP
Pept
Rem
1
M
1
M
61
N
60
N
X
X
2
G
2
A
X
XXX
62
P
61
P
3
K
3
K
X
XXX
XX
XX
63
Q
62
E
X
4
I
4
A
X
XXX
X
64
N
63
G
X
5
I
5
V
X
65
T
64
T
6
G
6
G
X
66
L
65
I
7
I
7
I
X
XX
X
X
67
F
66
F
8
D
8
D
X
XXX
XX
68
A
67
E
X
X
9
L
9
L
X
XXX
X
69
I
68
I
10
G
10
G
70
K
69
K
X
XX
X
ATP
11
T
11
T
ATP
71
R
70
R
X
XX
X
12
T
12
T
72
L
71
F
X
X
13
N
13
N
73
I
72
I
X
X
XX
X
14
S
14
S
74
G
73
G
X
XX
X
15
C
15
V
X
X
X
75
R
74
R
X
16
V
16
I
X
X
76
R
75
R
X
X
17
A
17
A
X
77
F
76
F
X
18
I
18
V
X
78
Q
–
19
M
19
L
X
XX
X
79
D
–
20
D
20
E
X
80
E
77
E
X
21
G
21
G
81
E
78
E
X
22
T
22
G
82
V
79
V
X
23
T
23
K
83
Q
80
Q
X
X
X
24
P
24
P
84
R
81
E
X
25
R
25
V
X
XX
85
D
82
E
X
26
V
26
V
X
86
V
83
A
X
27
L
27
L
X
87
S
84
K
X
28
E
28
E
X
X
X
X
88
I
85
R
X
29
N
29
N
X
XX
X
89
M
86
V
X
30
A
30
A
X
90
P
87
P
31
E
31
E
X
X
91
F
88
Y
32
G
32
G
X
92
K
89
K
X
33
D
33
E
X
93
I
90
V
X
34
R
34
R
X
X
94
I
91
V
X
X
35
T
35
V
95
A
92
P
36
T
36
T
96
A
93
G
X
37
P
37
P
97
D
94
P
38
S
38
S
X
98
N
95
D
39
I
39
V
X
X
99
G
96
G
X
40
I
40
V
X
X
100
D
97
G
X
41
A
41
A
X
101
A
98
V
X
42
Y
42
F
X
102
W
99
R
X
X
Fl
43
T
–
103
V
100
V
X
XX
X
44
Q
43
R
X
104
E
101
E
X
45
D
44
D
X
XXX
X
105
V
102
V
X
46
G
45
G
X
XXX
106
K
103
K
X
47
E
46
E
X
107
G
104
G
X
(continued on next page)
Table 1 (
continued)
DnaK-Eco
Count
DnaK
-Tth
Count
Assigned
Dock
ATP
Pept
Rem
DnaK
-Eco
Count
DnaK
-Tth
Count
Assigned
Dock
ATP
Pept
Rem
48
T
47
T
X
XX
108
Q
105
K
X
49
L
48
L
X
109
K
106
L
X
50
V
49
V
X
110
M
107
Y
X
51
G
50
G
X
XX
X
111
A
108
T
X
52
Q
51
R
X
112
P
109
P
X
53
P
52
M
X
XXX
113
P
110
E
X
X
XX
X
54
A
53
A
X
114
Q
111
E
X
55
K
54
K
X
115
I
112
I
X
X
56
R
55
R
X
116
S
113
S
X
57
Q
56
Q
X
117
A
114
A
X
58
A
57
A
X
118
E
115
M
X
59
V
58
V
X
X
119
V
116
I
X
60
T
59
L
X
120
L
117
L
X
XX
121
K
118
R
X
181
L
178
L
X
X
122
K
119
K
X
182
D
179
D
X
XXX
X
123
M
120
L
X
X
183
K
180
K
X
XX
124
K
121
V
X
XX
184
G
–
125
K
122
E
X
XXX
XX
185
T
181
K
X
126
T
123
D
X
X
186
G
182
G
X
X
127
A
124
A
X
X
187
N
183
N
X
128
E
125
S
X
188
R
184
E
X
XXX
X
129
D
126
K
X
X
189
T
185
T
X
XX
XX
130
Y
127
K
X
190
I
186
V
X
X
131
L
128
L
X
191
A
187
L
X
132
G
129
G
X
192
V
188
V
X
133
E
130
E
X
X
193
Y
189
F
X
XX
134
P
131
K
X
X
194
D
190
D
X
XX
X
XX
135
V
132
I
X
XX
195
L
191
L
X
XXX
XX
XX
136
T
133
T
X
XX
196
G
192
G
137
E
134
K
X
197
G
193
G
138
A
135
A
X
XX
XX
198
G
194
G
139
V
136
V
X
x
X
x
199
T
195
T
X
140
I
137
I
X
X
200
F
196
F
X
X
141
T
138
T
X
XX
X
201
D
197
D
X
X
X
142
V
139
V
X
XX
XX
X
202
I
198
V
X
XXX
X
143
P
140
P
203
S
199
T
X
X
144
A
141
A
X
XXX
204
I
200
I
X
X
145
Y
142
Y
X
X
dnaj
205
I
201
L
X
X
146
F
143
F
X
206
E
202
E
X
147
N
144
N
X
XXX
XX
dnaj
207
I
203
I
X
148
D
145
N
dnaj
208
D
–
149
A
146
A
X
X
209
E
–
150
Q
147
Q
X
X
X
210
V
–
151
R
148
R
X
X
211
D
–
152
Q
149
E
X
212
G
204
G
X
X
153
A
150
A
X
213
E
205
E
X
XX
XX
154
T
151
T
X
214
K
206
G
X
X
X
155
K
152
A
X
215
T
207
V
X
XXX
156
D
153
N
X
X
XX
X
216
F
208
F
X
X
XX
157
A
154
A
X
217
E
209
E
X
dnaj
158
G
155
G
X
218
V
210
V
X
X
dnaj
159
R
156
R
X
219
L
211
K
X
160
I
157
I
X
220
A
212
A
X
161
A
158
A
X
221
T
213
T
X
X
X
162
G
159
G
X
222
N
214
S
X
XXX
X
163
L
160
L
X
223
G
215
G
X
164
E
161
E
X
XXX
224
D
216
D
X
XX
X
X
165
V
162
V
X
225
T
217
T
X
X
166
K
163
L
X
XX
226
H
218
H
X
X
167
R
164
R
X
dnaj
227
L
219
L
X
X
X
X
168
I
165
I
X
228
G
220
G
X
X
169
I
166
I
X
X
X
X
229
G
221
G
X
170
N
167
N
X
XXX
X
dnaj
230
E
222
S
X
X
171
E
168
E
X
XX
XX
231
D
223
D
X
XX
XX
X
172
P
169
P
232
F
224
M
X
173
T
170
T
X
XX
XX
X
dnaj
233
D
225
D
X
174
A
171
A
X
X
234
S
226
H
X
175
A
172
A
X
235
R
227
A
X
176
A
173
A
X
XX
X
236
L
228
I
X
177
L
174
L
X
X
XX
237
I
229
V
X
178
A
175
A
X
238
N
230
N
X
X
179
Y
176
Y
X
X
XX
X
239
Y
231
W
X
180
G
177
G
X
X
240
L
232
L
X
241
V
233
A
X
299
K
293
K
X
X
242
E
234
E
X
300
V
294
L
X
X
243
E
235
E
X
301
T
295
T
X
244
F
236
F
X
302
R
296
R
X
245
K
237
K
X
303
A
297
A
X
246
K
238
K
X
304
K
298
K
X
247
D
239
E
X
XXX
305
L
299
F
X
248
Q
240
H
X
306
E
300
E
X
249
G
241
G
X
307
S
301
E
X
250
I
242
V
X
308
L
302
L
X
251
D
243
D
309
V
303
I
X
252
L
244
L
X
310
E
304
Q
X
253
R
245
K
X
311
D
305
P
254
N
246
A
X
312
L
306
L
X
255
D
247
D
X
313
V
307
L
X
256
P
248
R
314
N
308
K
X
X
257
L
249
Q
X
315
R
309
R
X
258
A
250
A
X
316
S
310
L
X
259
M
251
L
X
317
I
311
R
X
260
Q
252
Q
X
318
E
312
G
X
261
R
253
R
X
X
319
P
313
P
262
L
254
L
X
320
L
314
V
X
X
263
K
255
I
X
321
K
315
E
X
264
E
256
E
X
322
V
316
Q
X
265
A
257
A
X
323
A
317
A
X
266
A
258
A
X
324
L
318
L
X
(continued on next page)
Table 1 (
continued)
DnaK-Eco
Count
DnaK
-Tth
Count
Assigned
Dock
ATP
Pept
Rem
DnaK
-Eco
Count
DnaK
-Tth
Count
Assigned
Dock
ATP
Pept
Rem
267
E
259
E
X
325
Q
319
K
X
268
K
260
K
X
326
D
320
D
X
269
A
261
A
X
327
A
321
A
X
270
K
262
K
X
XX
328
G
322
G
X
271
I
263
I
X
329
L
323
L
X
272
E
264
E
X
330
S
324
T
X
X
273
L
265
L
X
331
V
325
P
274
S
266
S
X
332
S
326
A
275
S
267
S
X
X
333
D
327
Q
X
276
A
268
T
X
334
I
328
I
X
277
Q
269
L
335
D
329
D
X
278
Q
270
E
X
336
D
330
E
X
279
T
271
T
X
337
V
331
V
X
280
D
272
T
X
338
I
332
I
X
281
V
273
I
X
339
L
333
L
X
X
X
282
N
274
S
X
340
V
334
V
X
283
L
275
L
X
341
G
335
G
X
XXX
284
P
276
P
342
G
336
G
X
285
Y
277
F
343
Q
337
A
X
X
286
I
278
I
344
T
338
T
X
XX
X
X
287
T
279
A
X
345
R
339
R
X
288
A
280
L
X
346
M
340
V
X
289
D
281
D
X
347
P
341
P
290
A
282
P
348
M
342
A
X
291
T
283
A
X
XX
349
V
343
V
X
G
284
S
X
350
Q
344
Q
X
–
285
K
X
351
K
345
Q
X
X
292
–
286
T
X
352
K
346
V
293
P
287
P
353
V
347
V
294
K
288
L
X
354
A
348
R
X
295
H
289
H
X
355
E
349
E
X
296
M
290
L
X
356
F
350
L
X
297
N
291
E
357
F
351
L
X
298
I
292
K
X
358
G
352
G
X
359
K
353
K
X
419
P
413
P
360
E
354
E
X
420
T
414
T
X
361
P
355
P
421
K
415
R
X
X
362
R
356
N
422
H
416
K
X
XX
363
K
357
R
423
S
417
C
364
D
358
S
424
Q
418
E
365
V
359
V
425
V
419
I
366
N
360
N
426
F
420
F
367
P
361
P
427
S
421
T
368
D
362
D
X
428
T
422
T
DEL
369
E
363
E
X
429
A
423
A
A/E
370
A
364
V
X
430
E
424
E
371
V
365
V
X
X
431
D
425
H
372
A
366
A
432
N
426
N
373
I
367
M
433
Q
427
Q
374
G
368
G
X
X
434
S
428
T
375
A
369
A
X
X
435
A
429
A
X
XX
376
A
370
A
X
XXX
X
436
V
430
V
X
XX
377
V
371
I
X
437
T
431
E
X
XX
378
Q
372
Q
438
I
432
I
X
XX
379
G
373
A
X
x
x
439
H
433
H
X
XX
380
G
374
G
X
440
V
434
V
X
X
381
V
375
V
X
XXX
X
441
L
435
L
X
XX
382
L
376
L
X
X
442
Q
436
Q
X
383
T
377
M
X
XX
443
G
437
G
X
X
384
G
378
G
X
444
E
438
E
X
X
385
D
379
E
X
X
445
R
439
R
X
X
386
V
380
V
X
X
446
K
440
P
387
K
381
R
X
447
R
441
M
X
X
388
D
382
D
X
448
A
442
A
X
XX
389
V
383
V
449
A
443
Q
X
X
390
L
384
V
X
XX
X
450
D
444
D
X
391
L
385
L
X
451
N
445
N
X
392
L
386
L
X
452
K
446
K
X
393
D
387
D
X
X
453
S
447
S
X
394
V
388
V
X
454
L
448
L
X
395
T
389
T
X
X
455
G
449
G
X
X
396
P
390
P
456
Q
450
R
X
X
397
L
391
L
X
X
457
F
451
F
398
S
392
S
X
XX
458
N
452
R
X
XX
399
L
393
L
X
459
L
453
L
X
XX
400
G
394
G
X
X
460
D
454
E
X
X
X
401
I
395
V
X
X
X
461
G
455
G
X
X
402
E
396
E
X
XX
462
I
456
I
X
XX
403
T
397
T
X
X
463
N
457
P
404
M
398
K
ARCH
464
P
458
P
405
G
399
G
465
A
459
M
406
G
400
G
X
XX
466
P
460
P
407
V
401
V
X
XX
467
R
461
A
408
M
402
M
X
X
XX
468
G
462
G
X
409
T
403
T
X
X
469
M
463
V
X
410
T
404
V
X
470
P
464
P
X
411
L
405
L
X
XX
471
Q
465
Q
X
412
I
406
I
X
472
I
466
I
X
413
A
407
P
X
473
E
467
E
X
XX
414
K
408
R
X
X
MUT
474
V
468
V
X
XX
415
N
409
N
X
475
T
469
C
X
X
416
T
410
T
X
476
F
470
F
X
X
417
T
411
T
X
477
D
471
D
X
X
418
I
412
I
478
I
472
I
X
X
479
D
473
D
X
X
480
A
474
A
X
481
D
475
N
X
482
G
476
G
X
(continued on next page)
Table 1 (
continued)
DnaK-Eco
Count
DnaK
-Tth
Count
Assigned
Dock
ATP
Pept
Rem
DnaK
-Eco
Count
DnaK
-Tth
Count
Assigned
Dock
ATP
Pept
Rem
483
I
477
I
X
484
L
478
L
X
485
H
479
H
X
X
486
V
480
V
X
XX
487
S
481
T
X
X
488
A
482
A
X
X
489
K
483
K
X
XX
490
D
484
E
X
XX
491
K
485
R
X
XX
492
N
486
S
X
XX
493
S
487
T
X
XX
494
G
488
G
X
X
495
K
489
R
X
XX
496
E
490
E
X
497
Q
491
A
X
X
498
K
492
S
X
X
499
I
493
I
X
X
XX
500
T
494
T
X
501
I
495
I
X
XX
502
K
496
Q
X
XX
503
A
497
N
X
504
S
498
T
505
S
499
T
506
G
500
T
507
L
501
L/E
H
H
H
H
H
H
Fasta amino acid allignment of DnaK
Eco
and DnaK
Tth
with a Smith-Waterman score of 2133 and a 54.574% identity (57.096% ungapped) in 634 aa overlap (3–612 : 2–631). The column Assi identifies the
residues for which the NH NMR resonances were assigned. The column Dock identifies residues with NH NMR shifts depicted in
. The column ATP identifies residues with NH
NMR shifts depicted in
(b). The column Pep identifies residues with NH NMR shifts depicted in
(c). The column Rem identifies residues referred to in the text. The relative magnitude
of the shifts is indicated with the symbols X, XX and XXX (small, large, larger).
population in higher-order oligomers. Comparing
the amino acid sequences of DnaK
Eco
and DnaK
Tth
,
one observes a very important difference: the
hydrophobic arch over the substrate binding cleft,
M404–A429 in DnaK
Eco
is not conserved in
DnaK
Tth
, where the corresponding residues are
K398–A423 (see
). We reasoned that a
protective arch is likely never formed in DnaK
Tth
,
and hence the substrate binding cleft is easily
accessible at all times in our lid-less construct,
thus inducing aggregation by self-binding. Indeed,
the residual aggregation was abolished without
eliminating the peptide binding activity of the
construct (see below) by introducing a charged
residue, A423E, which together with K398 might be
forming a hydrophilic arch over the substrate
binding site. In addition, the T422 codon was
deleted during the introduction of the A423E
mutation. The resulting construct was monomeric
at 55 8C and at concentrations up to 400 mM. The
protein is stable in solution for several weeks and
gives rise to well-resolved NMR spectra (
The construct containing just the A423E mutation
showed similar behavior but a slightly increased
tendency to form aggregates. Here we demonstrate
that the DnaK
Tth
(1–501, DT422, A423E) mutant
(henceforth called DnaK
Tth-NBD-SBD
) retains peptide
binding activity with a K
D
of 100 mM and inter-
domain allosteric communication at 55 8C.
NMR probes
Using 3D HNCA, HNCOCA, HNCACB, HNCO
and HNCACO TROSY spectra, the backbone
assignments of DnaK
Tth-NBD-SBD
in the presence of
ADP-Pi but in the absence of peptide substrate
(referred to as DnaK
Tth-NBD-SBD
ADP.Pi/apo) at pH
7.4 and 55 8C were extended from the earlier work
on the isolated NBD.
A total of 415 of the 475 non-
proline residues in the native sequence (87%) could
be unambiguously assigned. The secondary struc-
ture was calculated based on the C
a
, C
b
and CO
chemical shifts.
All predicted secondary structure
elements aligned with the secondary structure
observed in the crystal structure of the bovine
Hsc70 NBD
and with the DnaK
Eco
SBD.
Although no high-resolution structure is available
for any part of DnaK
Tth
, the high level of sequence
and secondary structure homology with bovine
Hsc70 and DnaK
Eco
(see
) allowed us to
construct homology models of the individual
domains with a high degree of confidence.
As in
the isolated NBD,
many of the unassigned
residues in the NBD of the two-domain construct
Figure 1.
1
H–
15
N 800 MHz TROSY spectra of uniformly
2
H,
15
N,
13
C: labeled domains of the DnaK of T. thermophilus
recorded at 55 8C. The data of the 381-residue DnaK
Tth-NBD
, in the ADP.P
i
state (red) are superimposed on those of the
501-residue construct DnaK
Tth-NBD-SBD
ADP.P
i
/apo (black).
Structural Indicators of Allostery in DnaK-Tth
171
lie close to the nucleotide binding site and the NMR
signals are likely broadened by local dynamics,
which is common for enzyme active sites. In the
SBD, virtually all of the resonances of the loop
regions L3,4, L5,6 and L7,8, i.e. the grey colored
loops in
(b), could not be detected (for
nomenclature, see Zhu et al.
). This is not unex-
pected, since these residues on the “left-hand” side
of the substrate-binding cleft were found to be
disordered in the peptide-free SBD of DnaK
Eco
,
resulting in conformational exchange broadening.
Of great interest to report is that the inter-domain
hydrophobic linker, which has not been observed in
any previous structural study, has a defined
structure: the chemical shift secondary structure
prediction is helix for residues 367–371 and b-struc-
ture for the 380–388 region.
Relative domain orientation and motions
Since previous studies have not been able to
clearly define the degree or nature of the inter-
domain interactions, we initially examined the
assigned spectra at a qualitative level to assess
the differences in the signals from the two domains.
An analysis of the
1
H–
15
N TROSY spectra for the
(ADP.P
i
/apo) form, where the parenthesized text
indicates the ligation state of the NBD and SBD,
respectively, of the protein revealed that the line
shape and intensities for residues from both
domains are quite uniform (see
). Despite
quite different sizes of the two domains (42 kDa and
12 kDa), the similarities in the line shape indicate
that the domains are moving in solution as a single
unit and not as two individual structures connected
by a flexible linker.
These qualitative notions about relative mobility
are fully supported by RDC measurements. We
carried out RDC measurements
for the NH and
CO–C
a
vectors of triple-labeled DnaK
Tth-NBD-SBD
(ADP.P
/apo) in a dilute liquid crystal of Pf1
phage.
A total of 102 NH RDCs for the NBD and
43 NH RDCs for the SBD could be confidently
measured and fitted to the model structures of
the DnaK
Tth-SBD
(based on DnaK
Eco-SBD
) and
DnaK
Tth-NBD
(based on the NBD of bovine
Hsc70,
and DnaK
Eco-NBD
,
). 125 CO–C
a
RDCs
for the NBD and 38 CO–C
a
RDCs for the SBD were
also fitted to the model structures of the DnaK
Tth
domains. Analysis of the NH data sets yielded a
general degree of order (GDO;
) of 1.85!10
K
3
for
the NBD and 1.56!10
K
3
for the SBD. Analysis of
the CO–C
a
RDC data sets yielded GDOs of 1.06!
10
K
3
for the NBD and 0.77!10
K
3
for the SBD. The
similarity of the GDO values for the domains of
such significantly different sizes strongly confirms
the relatively rigid association of these domains (the
differences between the GDOs derived from the NH
and CO–C
a
RDCs are caused by a different overall
degree of alignment in the same, but aged sample in
which the phage concentration had decreased).
We have obtained excellent
1
H–
15
N TROSY
spectra of
15
N-labeled DnaK
Tth-NBD-SBD
in several
different ligation states (not shown). These were
(NBD/SBD): apo/apo; ADP.P
i
/apo; ATP/apo;
ADP.P
i
/pep and ATP/Pep, where Pep is a peptide
of the sequence KLIGVLSSLFRPK, known to bind
to the Hsp70 chaperones with an affinity in the
range of 10–100 mM, depending on species and
state.
The cross-peaks in the TROSY spectra of the
protein in these different states are equally uniform
in intensity as in the DnaK
Tth-NBD-SBD
ADP.P
i
/apo
for which the RDC measurements were carried out.
Accordingly, the two domains must be relatively
closely docked in all of the states studied. However,
detailed studies do suggest differences for the
rigidity of the interface depending on ligation
state (see below).
In addition to demonstrating the strong associ-
ation of the two domains, the RDC measurements
were also used to establish well-defined orien-
tations of the alignment tensor principal axes in the
molecular frames (i.e. PDB file coordinates) of the
domains in the ADP.Pi/apo state. Orientations for
all three axes could be obtained for the NBD, while
only the orientation of the S
zz
axis could be reliably
determined for the SBD. Given the strong evidence
that both domains are tightly docked, one may
co-align the S
zz
axes of both domains and obtain
their relative orientations in a common frame
(see
(a)). This exercise has a caveat: tensor axes
are rather double-headed arrows than vectors;
hence the dipolar information allows the z-axes of
the domains to be aligned in two opposite ways.
However, the fact that the C terminus of the NBD
must be connected to the N terminus of the SBD
rules out one of these possibilities. The relative S
zz
axes orientations of the two domains as derived
from the NH and CO–C
a
RDC data corresponded
within ten degrees (result not shown).
To define a potential interface between the NBD
and SBD, the previously reported NMR assign-
ments of the isolated NBD
were compared with
those of the NBD in the DnaK
Tth-NBD-SBD
construct.
shows an overlay of the
1
H–
15
N TROSY
spectra
of the isolated DnaK
Tth-NBD
in the ADP.P
i
state, with the TROSY of DnaK
Tth-NBD-SBD
, also in
the ADP.P
i
/apo state.
shows a histogram of
the differences in normalized
1
H–
15
N chemical
shifts
of the NBD in these spectra. The differences
represent a chemical shift mapping of the SBD on
the NBD, listed in
, and indicated on the
model structure of the NBD in
(a). It is
readily apparent that most of the larger shifts occur
in an area around the C terminus of the NBD. The
chemical shift changes in this region are somewhat
ambiguous, since differences could be either due to
direct SBD contacts or indirectly caused by struc-
tural rearrangements of the interdomain linker.
In
(a), we combined the “mapping”
information with the z-axis RDC information,
and have chosen an x/y orientation of the SBD
with respect to the NBD so that the (here missing)
a
-helical lid subdomain of the full Hsp70 may
transiently contact domain IB of the NBD (see
Flaherty et al.
for this nomenclature; in our Figures
172
Structural Indicators of Allostery in DnaK-Tth
it is the “left-top” of the NBD, see
(c)). Such
a contact would explain a change in the fluor-
escence of Trp102 in DnaK
Eco
, located in that
domain, upon nucleotide binding and peptide
binding.
In DnaK
Tth
, a corresponding residue
with the homologous position is R99, as indicated
in
(a).
Though the relative orientation of the two
domains is well-determined, the combined data
are still insufficient to complete their computer
docking, because the position of the (well-ordered)
ten-residue hydrophobic linker is unknown as of
yet. Furthermore, we lack chemical shift mapping
data for the SBD, which has not been assigned
separately from the NBD. Nevertheless,
affords, for the first time, the relative orientation
and position of both domains in the Hsp70
chaperones based on direct structural information.
Further refinement of this structure, and how it
changes in different ligation states, has to await
more experimental data.
Differences between the nucleotide states in the
absence of peptide substrate
To identify residues that are involved in the
allosteric communication between the domains
caused by nucleotide binding and hydrolysis, we
monitored the differences in chemical shifts in
1
H–
15
N TROSY spectra between the assigned
ADP.Pi/apo state and the apo/apo state or
ATP/apo state.
Figure 2.
Histogram of the back-
bone amide group chemical shift
differences, Dd (NH),
of the iso-
lated NBD DnaK
Tth-NBD
, and the
NBD in the two-domain construct
DnaK
Tth-NBD-SBD
. The root mean
square shift (RMSS) was calculated
from the data excluding the first ten
and last ten residues. The RMSS, 1.5
RMSS and two RMSS are indicated
with horizontal lines.
Structural Indicators of Allostery in DnaK-Tth
173
Figure 3.
Model of T. thermophilus DnaK(1–501). The model is based on NMR data described in the text and the crystal
structures of the NBD of bovine Hsc70 in the ADP.P
i
and the NBD of DnaK
Eco
in the apo state complexed with
and the crystal structure of the peptide-ligated SBD of DnaK
Eco
.
The Figures display residues M1-Q372 (C
terminus of the NBD) and D387-E501 (N terminus of the SBD). The structure for the 15-residue linking region (sequence
373
AGVLMGEVRDVVLL
386
) is not yet determined and is not shown. (a) Docking shifts. The N and C termini of the
domains are indicated with capitals for the NBD and lower case for the SBD. The linker between the domains is
schematically indicated with a broken line. The relative orientation of the NBD and SBD was partially determined from
NMR RDC measurements of DnaK
Tth-NBD-SBD
ADP.P
i
/apo in a dilute liquid crystal, by aligning the z-axes (indicated) of
the alignment tensors of the two domains as shown. The docking interface of the SBD on the NBD is suggested by the
color-coded surface residues that changed chemical shifts between isolated NBD of DnaK
Tth-NBD
, and the NBD in the
two-domain construct DnaK
Tth-NBD-SBD
. The relative orientations of the x and y axes of the domains was chosen to allow
an interaction of the (in this construct missing) helical lid domain with residue R99 in the NBD (indicated). The following
174
Structural Indicators of Allostery in DnaK-Tth
Addition of substoichiometric amounts of ADP.Pi
or ATP to an apo/apo sample of DnaK
Tth-NBD-SBD
resulted in a second set of resonances appearing
for many residues, indicating slow exchange for
nucleotide binding, as was expected from the
previous studies.
The TROSY spectrum of
ADP.P
i
/DnaK
Tth-NBD-SBD
ratio 5:1 (not shown) was
virtually identical to that with an equimolar amount
of ADP.P
i
(
), which is consistent with an
ADP equilibrium dissociation constant, K
D
, at least
an order of magnitude lower than the protein
concentration used in the experiment (about
200 mM), leading to an estimated value of K
D
!
10 mM. Almost identical results were obtained upon
addition of ATP.MgCl
2
to DnaK
Tth-NBD-SBD
(apo/
apo) (not shown). Thus, both nucleotide binding
affinities were relatively high, in full agreement
with Slepenkov & Witt.
A total of 28 peaks
showed significant chemical shift differences
between the saturated ADP.Pi/apo forms and
apo/apo state of DnaK
Tth-NBD-SBD
(see
This was substantially less than the 155 peaks that
shifted upon nucleotide binding to the isolated
NBD.
A total of 62 resonances strongly shifted
between the apo/apo and ATP/apo states. That
was also substantially less than what was observed
in case of the isolated NBD.
This strongly suggests
that interaction of the SBD with the NBD in the
apo/apo form induces changes in NBD towards a
conformation more similar to the nucleotide-bound
states. A greater number of differences in the
spectra between the apo and ATP than between
the apo and ADP bound states suggests that the
structures of the apo and ATP states are less related
than those of the apo and ADP states. This
correlates well with the results obtained by bio-
chemical studies.
In
(b) we color-coded the minimal
chemical shift changes per residue that must occur
to explain the differences between the TROSY
spectra of the ADP.P
i
/apo and ATP/apo states
(see also
). The majority of the shift changes
take place around the nucleotide-binding pocket,
but many, rather large, shifts also occur at the
“bottom” of the NBD domain. Several dramatic
shift changes also occur in the linker region between
the two domains (see
). Comparison of
(b) with
(a) shows that the areas
of shift changes in the NBD upon ADP5ATP
replacement are very similar to the areas with
different chemical shifts in the isolated NBD in
DnaK
Tth-NBD
and the NBD in the DnaK
Tth-NBD-SBD
construct (we call the latter “docking” or
“undocking” shifts). This strongly suggests that
the interface between the NBD and SBD is affected
by the nucleotide exchange. This result is, on the
one hand, to be expected for this allosteric molecule;
on the other hand, it is very exciting that for the first
time this is evidenced by actual data that reflect
structural (and/or dynamical) changes in that area.
The magnitude of the nucleotide-exchange shifts is
much smaller than the domain docking shifts,
confirming that the SBD does not become dislodged
from the NBD upon nucleotide exchange. Also,
detailed analysis shows that the nucleotide-
exchange shifts, while generally in the same area
as the docking shifts, involve different residues (see
). Other residues in the NBD undergoing
chemical shift changes were mainly located in
domain 1B.
In our NMR study of the isolated NBD,
almost
80 residues were observed to be undergoing slow
two-state-conformational exchange in the ADP.Pi
state while exhibiting a single state in the ATP
bound form. In the two-domain construct a total of
17 residues were observed to behave in a similar
manner and 13 of these were located in subdomain
IA with the rest being located close to the
nucleotide-binding site. It appears that part of
the effect of ATP hydrolysis is to allow a slow
conformational adaptation that mainly involves
subdomain IA.
As summarized in
(b), ADP5ATP
exchange causes significant shifts for five residues
in the SBD. This is the evidence of allosteric
communication between the two domains in the
construct. While these residues do not form any
identifiable interface or pathway within the SBD,
we note that two shifting residues (402 and 403) are
in a loop that makes up one half of the hydrophobic
arch over the substrate binding cleft. Lack of
assignments for the other half of the arch precludes
any conclusions for that area; nevertheless, the
available data lend credence to the proposal
that
the arch plays a major role in the regulation of
substrate binding in the cleft and that the allosteric
mechanism could be mediated through opening
and closing of the arch structure. The observation of
color coding was used: gray, the residues in DnaK-Tth that were unassigned; green, positions that shifted less than the
RMSS (root-mean-square-shift); yellow, RMSS!Dd (NH)!1.5 RMSS; orange, 1.5 RMSS!Dd (NH)!2 RMSS; and red, Dd
(NH)O2 RMSS as defined in
. (b) The minimal chemical shift changes necessary to account for the differences in
the TROSY spectra of DnaK
Tth-NBD-SBD
ADP.P
i
/apo and DnaK
Tth-NBD-SBD
ATP/apo. The location of the nucleotide
binding site is indicated with AdtP; the location of the substrate binding cleft and the nomenclature of the SBD loops is
identified as well. The color scale is: gray, unassigned; green, little or no shift; yellow, resonances that changed from
double to single, which narrowed, or had 10 Hz!Dd
min
(NH)!20 Hz; orange, resonances with a larger shift or which
completely disappeared. (c) The minimal chemical shift changes necessary to account for the differences in the TROSY
spectra of DnaK
Tth-NBD-SBD
ADP.P
i
/apo and DnaK
Tth-NBD-SBD
ADP.P
i
/peptide, where the peptide ligand is
KLIGVLSSLFRPK. The location of the peptide-binding site, the mutations A423E and deletion D422 in loop L3,4 and
the ADP binding site, with the catalytic residue K69 are indicated. The labels IA, IB, IIA, IIB identify the subdomains of
the NBD, as discussed in the text. The color scale is as for (b).
Structural Indicators of Allostery in DnaK-Tth
175
communication between the two active sites in this
construct also supports our earlier suggestions that
the alpha-helical lid subdomain (i.e. the residues
beyond residue 500) is not essential for the allosteric
mechanism.
However, so far the NMR data do not
allow for the delineation of the path of communi-
cation between the arch and the NBD.
Effect of peptide binding to the SBD on the SBD
Introduction of the mutations in the arch struc-
ture near the substrate binding site eliminated
aggregation of the sample, but could have also
interfered with substrate binding. To test the
binding of the mutated SBD, a peptide of the
sequence KLIGVLSSLFRPK, known to bind to
Hsp70s in general,
was titrated into samples of
DnaK
Tth-NBD-SBD
in the ATP/apo and ADP.Pi/apo
states. Both samples exhibited a very large number
of significant chemical shift changes for residues in
the SBD and, in particular, the residues close to the
substrate binding pocket, demonstrating that pep-
tide bound to that region. That the two-domain
construct is a functional allosteric molecule was also
further confirmed by the titration. As shown in
(a), peptide binding to DnaK
Tth-NBD-SBD
(ADP.P
i
/apo) causes resonance changes in the SBD
that are in slow exchange on the NMR time-scale,
suggesting
slow
peptide
release
kinetics.
We observe an approximately equal population of
the bound and free state of DnaK
Tth-NBD-SBD
when
the protein:peptide ratio is 1 : 1. With the given
concentration of DnaK
Tth-NBD-SBD
at 250 mM, we
compute a K
D
for the release of this peptide of
approximately 100 mM. This affinity is very close to
that measured for the binding of the same peptide
to BiP (the constitutive Hsp70 of the human
endoplasmic reticulum) or Hsc70 (the constitutive
Hsp70 of human cytosol) of 150 mM and 55 mM,
respectively.
The substrate binding pockets of the
mammalian Hsp70s and DnaK-Tth are virtually
identical. Thus the mutations made (DT422, A423E,
indicated in
(c)) have hardly or not at all
affected the affinity of the ADP state, which is likely
indicative of the flexibility of the loop (L3,4; see
(b) and (c)) in which these mutations are
located.
(b) shows intermediate-to-fast
exchange for substrate binding to DnaK
Tth-NBD-SBD
in the (ATP/apo) state, suggesting faster peptide
release kinetics. This is fully compatible with the
literature kinetics data,
and is a proof of NBD/
SBD kinetic allostery. Analysis of the chemical shift
titration data of the ATP state shows that the
peptide K
D
has (only) increased to 0.5 mM. Possibly,
this intrinsically weak-binding peptide is not a
good reporter of the thermodynamic allostery, or
the thermodynamic allostery from the NBD to the
SBD is impaired by the mutations.
(c) shows the mapping of the shift
differences induced by the addition of the peptide
to DnaK
Tth-NBD-SBD
in the ADP.P
i
state. Many
significant changes occur, especially for the reso-
nances in the SBD domain. Major changes in shifts
occur in the direct vicinity of the substrate binding
cleft. However, the shift changes in the SBD are not
limited to the peptide binding cleft and extend to
the bottom b-sheets and to the linker area with the
NBD: the entire loop Gln433-Gln436, residue R408
in loop L2, 3, residue V388, L391 and Ser392,
directly adjacent to the linker, are affected. These
findings partially agree with the effects seen in the
isolated SBD (489–507) in DnaK
Eco-SBD
, where
peptide binding also induced changes at the
N-terminal area,
but where the bottom b-sheet
was much less affected. This difference would
suggest that this bottom sheet of the SBD is part of
the interface with the NBD in DnaK
Tth-NBD-SBD
.
However, the RDC order tensor z-axis of the SBD
must point to the long axis of the two-domain
construct,
and hence towards the location of the
NBD. This proves that the NBD is located on the
right-hand side of the SBD in the representation of
. The following argument supports this
conclusion as well.
The DnaK
Eco
mutant K414I (in loop 2,3) cannot
support lambda replication in vivo and also shows
an absence of allosteric functions in vitro.
In
addition, proteolytic cleavage at Lys414 in DnaK
Eco
is much more pronounced in the ATP state than in
the ADP state, implicating this site in nucleotide-
induced conformational changes and allostery.
Significantly, DnaK
Eco
residue Lys414 corresponds
to residue R408 in DnaK
Tth
for which we do see a
shift in DnaK
Tth-NBD-SBD
upon peptide binding (see
). Together with the RDC data, this leads to
the conclusion that the contact area with the NBD is
at the right-upper side of the SBD. It is not unlikely
that the extended shifts in the b-sheet are due to a
need for larger structural adaptations on peptide
binding because of the residue deletion in the
substrate binding cleft. All together, our data
confirm that the substrate binding cleft commu-
nicates with the rather remote N-terminal area of
the SBD using a “pathway” of allosteric signal
transduction in the SBD towards the NBD within
the SBD itself, i.e. without need of the a-helical
domain.
Effects of peptide binding to the SBD on the NBD
By far the most exciting result is, obviously, that
peptide binding in the SBD also causes widespread
changes in NMR shifts in the NBD, as shown in
(c). This is very strong evidence of allostery
indeed. The peptide binding induced changes are
seen to propagate all the way to the nucleotide
binding site, which outlines a possible path of the
allosteric communication between the peptide and
nucleotide binding sites. Elegant mutagenesis/
functional/X-ray crystallographic studies on the
NBD of bovine Hsc70, showed that nucleotide
binding site residue K71, (K69 in DnaK
Tth
) is
essential to ATP hydrolysis
whereas residue T13
(T11 in DnaK
Tth
) is essential to couple ATP
hydrolysis to the overall allosterics.
Regretfully,
the loop area of T11 could not be assigned from our
176
Structural Indicators of Allostery in DnaK-Tth
Figure 4.
Detail of the effect of peptide binding on the TROSY spectrum DnaK
Tth-NBD-SBD
. (a) The protein was initially
in the ADP.P
i
/apo state and shows characteristics of slow chemical exchange on the NMR time-scale upon the binding of
the peptide KLIGVLSSLFRPK. (b) The protein was initially in the ATP/apo state and shows characteristics of
intermediate/fast chemical exchange on the NMR time-scale upon peptide binding. The color coding is: pink, no
peptide; green, 1 : 1 peptide:DnaK; blue, 4 : 1 peptide:DnaK for both panels.
Structural Indicators of Allostery in DnaK-Tth
177
NMR data, likely due to conformational mobility,
and we cannot make any statement about potential
changes therein. The loop that contacts the ATP-g-
phosphate side-chain, and that contains residue
K69 is assigned in the NMR spectra, and shows up,
not surprisingly, as a rather contiguous region of
sizeable shift changes upon ATP binding, but also,
and very significantly, as medium-sized chemical
shift differences upon peptide binding (indicated in
(c), and see
). This is strong evidence
that the allosteric communication runs from the
peptide binding site, over the interface, directly to
this loop.
Discussion
Global properties of the two-domain construct
The current construct of DnaK
Tth-NBD-SBD-1-501
is
an allosteric molecule, despite the fact that the
entire alpha-helical subdomain of the NBD is
missing. This supports similar findings for the
corresponding construct DnaK
Eco-1-507
and the
slightly larger construct DnaK
Eco-1-517
Full-length DnaK of T. thermophilus is known to
be thermally stable up to 90 8C. Earlier reports
demonstrated that DnaK
Tth
forms a trimeric com-
plex with 3x (DnaK:DnaJ:DafA) in solution. It was
speculated that the multimeric state was bestowing
the thermal stability. Our current data show that
such is not the case. With a rotational correlation
time of 15 ns, as determined from preliminary
15
N
relaxation experiments, and mono-disperse beha-
vior in light scattering experiments, we are confi-
dent that our construct has a hydrodynamic radius
corresponding to a monomeric species. Neverthe-
less, very few changes occur in the TROSY NMR
spectra between 25 8C and 90 8C (not shown),
demonstrating that the nucleotide binding domain
by itself
and the current construct DnaK
Tth-NBD-
SBD-1-501
, is stable up to at least that temperature.
The thermal stability is therefore intrinsically
embedded in the monomeric structure.
We combined the “mapping” information
together with the z-axis RDC information to obtain
a global structure for the two-domain construct. The
z-axis RDC data yield information on the relative
orientations of the two domains in the construct
and are reliable within a few degrees. However, the
used translational information obtained from the
mapping, as shown in
(a), is not unam-
biguous, especially in the light of the as of yet
unknown position of the (well-ordered) intervening
ten-residue hydrophobic linker. Furthermore, we
lack chemical shift mapping data on the SBD. We
thus assess the validity of our model for DnaK
Tth-
NBD-SBD
shown in
in the following
independent ways. First, as shown in
(a),
the long axis of the alignment tensor (S
zz
) of the
NBD points toward the proposed location of the
SBD and vice versa, both defining the long axis of
the model structure. This confirms
the proposed
SBD–NBD translational positions independently
from the chemical shift mapping data. Indeed,
when the proposed two-domain structure is used
for prediction of the alignment tensors of the NBD
and SBD as positioned in the model, using the
program PALES,
one obtains S
zz
tensor orien-
tations within 15 degrees of the observed tensors for
both domains. Second, we compute from our
current model for DnaK
Tth-NBD-SBD
in the ADP.Pi/
apo state a radius of gyration (R
g
) of 28 A
˚ , and a
longest axis (d
max
) of 108 A
˚ . These results may be
compared with small angle X-ray scattering studies
(SAXS) of Hsc70: Wilbanks et al.
observe a R
g
of
28 A
˚ , and a d
max
of 115 A
˚ for Hsc70(1–560) in the
MgATP state, and a R
g
of 32.5 A
˚ , and d
max
of 125 A
˚
for Hsc70(1–560) in the MgADP state. The agree-
ment between their and our data is reasonable,
given that their Hsc70 construct is larger. It provides
a second independent validation of our proposed
structure.
Our experimental model for DnaK
Tth-NBD-SBD
is
in disagreement with the recently proposed theo-
retical model for the complete Hsc70.
The latter
has the SBD in close contact with both the IA and IB
subdomains, completely “at the left” of the NBD in
the orientation of
shows, according to our model, the
locations of mutation sites that affected the binding
of the co-chaperone DnaJ to DnaK
Eco
together
with the residues in the NBD that change shifts
between the one and two-domain constructs. The
mutations Y142A (Y145 in DnaK
Eco
), N144A (N147),
Figure 5.
Space-filling model of DnaK
Tth-NBD-SBD
.
Indicated in yellow, orange and red are the docking shifts
as defined in
(a). Shown in dark blue are residues
Y143 (145 in DnaK
Eco
count; see
), N145 (147), N146
(148), E209 (217) and V210 (218) that, when mutated to A,
compromise the NBD–SBD allosteric communication and
suppress 100-fold the ATP-hydrolysis-rate enhancement
by DnaJ.
The residue R164 (167), colored light blue,
rescues upon mutation to H the impaired DnaJ binding of
DnaJ-D35N;
also in light blue are N167 (170) and T170
(173), that, when mutated to A, reduce the binding of
DnaJ to DnaK.
178
Structural Indicators of Allostery in DnaK-Tth
N145A (N148), E209A (E217) and V210A (V218),
suppress 100-fold the ATP-hydrolysis rate enhance-
ment by DnaJ,
suggesting that this area is
involved in DnaJ binding. The fact that these
residues are exposed in our model would be
compatible with this notion. Suh et al.
showed
that the DnaK
Eco
NBD mutation R164 H (R167 in
DnaK
Eco
) rescues the impaired DnaJ binding of
DnaJ-D35N, while N167A (N170) and T170A (T173)
reduce the binding of DnaJ. These residues are also
(barely) accessible for DnaJ binding in our model.
Thus, our model would be compatible with a non-
competitive DnaJ binding at those sites, but that is
not to say that the DnaJ binding could not happen to
other loci as well.
Interestingly, the (same) mutations Y142A,
N144A, N145A, E209A and V210A, also abrogate
the allosteric communication between the SBD and
NBD domains in the absence of DnaJ.
According
to our model, these residues are rather distant from
the likely NBD–SBD interface, and their dual roles
in interdomain communication and DnaJ binding
cannot be explained in simple steric terms. Likely
the deleterious effect of their mutation on the SBD–
NBD inter-domain communication is an indirect
effect perturbing the ATP-binding cleft opening
angle (see also below).
What is the nature of the allosteric changes?
The obvious question is, of course, what the
observed NMR chemical shift changes represent in
terms of structural and/or dynamical changes. We
can offer the following insights. First, it is evident,
from the comparison of
(a) and (c), that the
pattern of peptide-induced shift changes is virtually
identical to that of the differences caused by the
docking of the SBD to the NBD. Thus, the domains
clearly communicate through that interface. How-
ever, the peptide-induced shifts are much smaller
than the (un)docking shifts. When investigated in
detail, the great majority of the peptide-induced
shifts are in the opposite direction of the undocking
shifts. A particularly salient example of the direc-
tion of shift changes is shown in
(a). We
interpret this as to indicate that the inter-domain
interface is tighter in the ADP.P
i
/peptide con-
formation as compared to the ADP.Pi/apo
conformation.
The differences in NMR spectra between the
ADP.P
i
and ATP forms are best represented with the
NBD domain in the ADP state being in a roughly
1 : 1 equilibrium between two (closely related)
conformations, which encompasses already the
ATP-conformation (
(b)). Addition of ATP
drives the equilibrium to what must be called the
ATP conformation, while peptide binding to the
SBD drives the equilibrium in the NBD towards
the other conformation (
(b)). The data
suggest that the ADP state is characterized by a
dynamical equilibrium that already contains the
ATP form. These results are remarkably similar to
those observed for the isolated DnaK
Tth-NBD
and
to those proposed earlier on theoretical grounds.
In the latter hypothesis, one of the conformations
would open the substrate binding site to allow fast
binding and release of substrate while the other
would close it. ATP/ADP exchange and ATP
hydrolysis merely shifts the pre-existing equi-
librium between these states.
(a) strongly suggests that peptide bind-
ing to the ADP.P
i
state causes a rigidification of the
interdomain interface.
(b) indicates that
ATP exchange and peptide binding have the
opposite effects on the “split” NMR resonances
characteristic of the ADP.Pi state. This logically
suggests a loosening of the NBD–SBD interface in
the ATP-bound state as compared with the ADP.P
i
conformation. Our result is in agreement with
mutagenesis data
that proteolytic cleavage at the
(supposedly) interfacial residue Lys414 in DnaK
Eco
is more pronounced, and thus more exposed, in the
ATP state than in the ADP state. Our observations
are also supported by SAXS studies on full-length
DnaK
Eco
,
which show a modest increase of R
g
by
2 A
˚ in the ATP state as compared to the apo state,
while the longest axis increases by 10(G5) A
˚ (the
conformation of the apo state is commonly assumed
to be similar to the ADP state).
In contrast, the SAXS studies
for Hsc70 and for
Hsc70(1–560) show that R
g
decreases by 3–5 A
˚ in
the ATP state as compared to the ADP state, with a
d
max
decrease of 10 A
˚ (see also above). These data
have been interpreted as representing a loose NBD–
SBD interface in the ADP state, and a tight NBD–
SBD interface in the ATP state.
It is difficult to
reconcile the different results for conformational
changes upon ADP/ATP exchange in the highly
homologous Hsc70, DnaK
Eco
and DnaK
Tth
. Most
likely, the presence/absence of helical lid regions
in the different studies may account for the
differences, but the combined observations could
also point to an important difference in interface
allosterics between prokaryotic and eukaryotic
Hsp70s.
A detailed description of the allosteric confor-
mational and dynamical changes at the interface
and throughout the molecule, i.e. the molecular
mechanics of allostery, needs to await more studies.
However, our recent NMR-RDC study of sub-
domain reorientations in isolated NBD of
Hsc70 may shed some light on the mechanics in
DnaK
Tth-NBD-SBD
as well. There, we have com-
pared the solution conformation of the isolated
NBD of bovine Hsc70 (ADP.P
i
state) with the crystal
structure of the identical construct in the same
ligation state.
We showed that significant dif-
ferences exist in the nucleotide cleft opening angle
(up to 15 degrees) between solution and crystal.
Moreover, we found that the four individual
subdomains (see
(c)) that constitute the
NBD
all have slightly different relative positions
in solution than in the crystal. Significantly, all
subdomains interact with the nucleotide ligand.
We thus hypothesized that the structural communi-
cation with the internal nucleotide binding site from
Structural Indicators of Allostery in DnaK-Tth
179
Figure 6.
Details of TROSY spectra of DnaK
Tth-NBD-SBD
exemplifying changes upon ligand binding. (a) DnaK
Tth-NBD
ADP.P
i
(red), DnaK
Tth-NBD-SBD
ADP.Pi/apo (green) and DnaK
Tth-NBD-SBD
ADP.P
i
/peptide (blue). The NH cross-peaks of
V136, F208 and A373, shift from red to green to blue, suggesting a tightening of the NBD–SBD interface upon peptide
binding. R43 does not shift. (b) DnaK
Tth-NBD-SBD
ADP.Pi/apo, DnaK
Tth-NBD-SBD
ATP/apo and DnaK
Tth-NBD-SBD
ADP.P
i
/peptide. The NH of E205 gives rise to two cross-peaks in the ADP.P
i
/apo state (green) due to a slow
conformational equilibrium (indicated by two marks). In the ATP/apo state, the conformational equilibrium is driven to
one state (red), while peptide binding drives it to the other state (blue).
180
Structural Indicators of Allostery in DnaK-Tth
binding events at the “outside” of the protein can be
easily transmitted by induction of subtle changes in
the relative subdomain angles.
Our current insights in the likely location of the
NBD–SBD interface as being close to the subdomain
boundaries do support this notion. In other words,
the widespread, but in most cases fairly small
chemical shift changes seen in the NBD upon the
allosteric processes discussed are more likely due to
subtle global reorientations of the NBD subdomains
driven by interface changes with retention of local
structure, rather than to a domino-effect of local
structural changes.
Conclusion
A global structure of a two-domain construct of
the DnaK Hsp70 chaperone protein of T. thermo-
philus was obtained on the basis of chemical shift
information and residual dipolar couplings
measurements. This is the first experimental infor-
mation on the relative positioning of the nucleotide
and substrate binding domains for any Hsp70
chaperone. We find that the NBD and SBD,
including linker, form a relatively rigid structure.
The interface between the two domains is located in
the close vicinity of both the N and C terminus of
the NBD. The two-domain construct of DnaK
Tth
(1–501) is an allosteric molecule, despite the fact
that the entire alpha-helical subdomain is missing.
ATP/ADP exchange and peptide binding cause
subtle changes in the NBD–SBD interface area,
indicating that this interface transduces the allo-
steric signal. Detailed studies suggest that the
interface becomes tighter in the order ATP/apo,
ADP.Pi/apo ADP.Pi/peptide.
Materials and Methods
Sample preparation
The truncated 1–501 form of DnaK
Tth
was constructed
by PCR from a plasmid containing the full length DnaK
Tt
gene supplied by Dr A. Joachimiak at Argonne National
Laboratory. The PCR primer for the 1–501 region
introduced a 5
0
NcoI site and a 3
0
XhoI site that were
used to ligate into a pET28b vector digested with AflIII
and XhoI. Creation of the XhoI site required mutation of
Leu501 to Glu. The resulting plasmid, pMR23, had the
1–500 sequence of wild-type DnaK
Tth
with the L501E
mutation and a six-residue histidine tail. Introduction of
the DT422, A423E mutations was accomplished using
mutagenic primers and the Quikchange mutagenesis Kit
(Stratagene). All constructs were sequenced to confirm
that the desired mutations were introduced. The location
of the mutation is shown in
(c).
Escherichia coli strain BL21 (DE3) cells containing the
pMR23 plasmid were grown at 37 8C in M9 media with
15
NH
4
Cl or with 99%
2
H
2
O, 98%
15
NH
4
Cl, and 99%
[
13
C,
1
H]glucose. All isotopes were purchased from
Cambridge Isotope Laboratory (Andover, MA). The His-
tagged protein was purified as described for the DnaK
Tth
NBD.
Guanidine-HCl and heating was required to
denature DnaK
Tth-NBD-SBD
to allow re-protonation of all
amide groups in the perdeuterated protein and complete
removal of the bound nucleotide. TROSY spectra of the
protein confirmed that it was properly re-folded from
these conditions.
NMR experiments
NMR spectra were collected at 55 8C using an 800 MHz
Varian Inova spectrometer, using a triple-resonance
Z-gradient probe. Samples were approximately 400 mM
DnaK
Tth-NBD-SBD
in 50 mM Hepes (pH 7.4), 10 mM KCl,
5 mM MgCl
2
, 5 mM ADP, 5 mM Na
2
PO
4
. At these
conditions, the protein was monomeric as derived from
the rotational correlation time of approximately 15 ns
determined from 1D
15
N NMR R
1
and R
2
relaxation data.
Backbone resonance assignments (
1
HN,
15
N,
13
C
a
,
13
CO
and
13
C
b
) of the 501-residue DnaK
Tth-NBD-SBD
in the
presence of ADP.P
i
and in the absence of peptide
substrate at pH 7.4 and 55 8C were extended from earlier
work on the 381-residue DnaK
Tth
A total of 415 of
the 475 non-proline residues in the native sequence (87%)
could be unambiguously assigned, using 3D HNCA,
HNCOCA, HNCACB, HNCO and HNCACO, all opti-
mized with TROSY detection and
2
H decoupling.
The
average data collection time was 48 hours per 3D data set.
All spectra were processed with NMRPipe
and ana-
lyzed using SPARKY.
Assignments were formally made, i.e. using triple-
resonance experiments, for the ADP.P
i
/apo state of
DnaK
Tth-NBD-SBD
only. Titrations of ADP, ATP and peptide
were all in slow exchange or intermediate exchange on
the chemical shift time-scale (i.e. slower than 100 s
K
1
).
The TROSY spectra of the other ligation states were,
therefore, not formally assigned by either titration or
triple resonance. However, the differences were suf-
ficiently subtle that the following, limited, mapping of
chemical shift changes could be carried out. We deter-
mined for every assigned
1
H–
15
N TROSY resonance of the
ADP.P
i
/apo state a resonance that was the closest in the
TROSY spectrum of the formally unassigned state. In
several cases, the resonance of the ADP/apo state
disappeared and no clear candidate new resonance was
found in its vicinity in the apo/apo form. For these cases
we assumed that the resonance of the apo/apo state has
shifted away at least as far as the closest peak, even if that
latter peak would normally be associated with another
residue. In our experience, this procedure somewhat
underestimates the actual shift differences.
DnaK
Tth-NBD-SBD
ADP.P
i
/apo behaves well in a solu-
tion of 20 mg/ml Pf1 phage at 50 8C, allowing the
measurement of
15
N–
1
H RDCs using JD-shifted 2D
TROSY
and
3D
JD-shifted
HNCO-TROSY
experiments.
The
15
N–
1
H RDCs of DnaK
Tth-NBD-SBD
varied from
K
28 Hz to C44 Hz. Analysis of the dipolar couplings was
carried out with the programs PALES
and REDCAT.
102 NH RDCs from the secondary structure elements in
the NBD and 43 dipolar couplings of both the b-sheets
and a well-defined loop in the SBD could be fitted to the
structural models of the domains with an RMSD of 3.2 Hz
and 2.5 Hz, respectively. The
13
CO–
13
C
a
residual dipolar
couplings were measured from a double constant-time 3D
HNCO-TROSY experiment without C
a
decoupling
using a Bruker 600 MHz spectrometer equipped with a
triple-resonance Z-gradient cryo-probe. The
13
CO–
13
C
a
RDCs of DnaK
Tth-NBD-SBD
varied from K5.8 Hz to
C
5.7 Hz. 125
13
CO–
13
C
a
RDCs from the secondary
structure elements in the NBD and 38
13
CO–
13
C
a
RDCs
Structural Indicators of Allostery in DnaK-Tth
181
of both the b-sheets and a well-defined loop in the SBD
could be fitted to the structural models of the domains
with an RMSD of 1.2 Hz and 0.9 Hz, respectively.
The GDOs for the domains
were calculated according
to the equation GDOZ
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
3
ðS
xx
C
S
yy
C
S
zz
Þ
q
. A PDB
format file with coordinates for the global model of
DnaK
Tth-NBD-SBD
based on these data will be provided
upon request.
Titrations of ATP and ADP-P
i
were carried out on
200 mM samples of DnaK
Tth-NBD-SBD
apo/apo and TROSY
spectra were collected at protein to nucleotide ratios of
3 : 0, 3 : 1, 3 : 2, 3 : 3, 3 : 6 and 3 : 15. Titrations of peptide
KLIGVLSSLFRPK were carried out on 250 mM samples of
DnaK
Tth-NBD-SBD
ADP.P
i
/apo and DnaK
Tth-NBD-SBD
ATP/apo and TROSY spectra were collected for samples
with protein to peptide ratios of 3 : 0, 3 : 1, 3 : 2, 3 : 3, 3 : 6
and 3 : 24.
SDS-polyacrylamide gel electrophoresis analysis was
carried out on the NMR samples after each series of NMR
experiments. In all cases, the protein was O95% in the
54 kDa (two-domain) form, proving that no or negligible
hydrolysis of the construct occurred.
Acknowledgements
We thank Dr A. Joachimiak at Argonne National
Laboratory for the gift of the plasmid containing the
full DnaK of T. thermophilus. We thank Dr Carol
Fierke for the use of the PCR equipment, Mr Kai
Keliikuli and Ms Tina Holder for their contributions
to the pioneering stages of this work. We thank
Dr H. M. Al-Hashimi for valuable discussions
concerning the RDC analysis. We thank Dr R.
Sousa (UTHSC, San Antonio) for providing us
with the coordinates and manuscript for the X-ray
structure of Hsc(1-554) prior to publication. We
acknowledge the W. F. Keck foundation, NSF and
the NIH for funds for the purchase of the 800 MHz
spectrometer. The authors gratefully acknowledge
the Michigan Economic Development Corporation
and the Michigan Technology Tri-Corridor for the
support of the purchase of the 600 MHz spec-
trometer. This work was supported by NIH grant
GM63027.
References
1. Hartl, F. U. (1996). Molecular chaperones in cellular
protein folding. Nature, 381, 571–580.
2. Bukau, B. & Horwich, A. L. (1998). The Hsp70 and
Hsp60 chaperone machines. Cell, 92, 351–366.
3. Craig, E. A., Gambill, B. D. & Nelson, R. J. (1993).
Heat-shock proteins—molecular chaperones of pro-
tein biogenesis. Microbiol. Rev. 57, 402–414.
4. Buchberger, A., Theyssen, H., Schro¨der, H., McCarty,
J. S., Virgallita, G., Milkereit, P. et al. (1995). Nucleo-
tide-induced conformational changes in the ATPase
and substrate binding domains of the DnaK chaper-
one provide evidence for interdomain communi-
cation. J. Biol. Chem. 270, 16903–16910.
5. Zylicz, M., King, F. W. & Wawrzynow, A. (2001).
Hsp70 interactions with the p53 tumor suppressor
protein. EMBO J. 20, 4634–4638.
6. Gragerov, A., Zeng, L., Zhao, X., Burkholder, W. &
Gottesman, M. E. (1994). Specificity of DnaK-peptide
binding. J. Mol. Biol. 235, 848–854.
7. Zhu, X., Zhao, X., Burkholder, W. F., Gragerov, A.,
Ogata, C. M., Gottesman, M. E. & Hendrickson, W. A.
(1996). Structural analysis of substrate binding by the
molecular chaperone DnaK. Science, 272, 1606–1614.
8. McCarty, J. S., Buchberger, A., Reinstein, J. & Bukau,
B. (1995). The role of ATP in the functional cycle of the
DnaK chaperone system. J. Mol. Biol. 249, 126–137.
9. Pierpaoli, E. V., Sandmeier, E., Baici, A., Scho¨nfeld,
H.-J., Gisler, S. & Christen, P. (1997). The powerstroke
of the DnaK/DnaJ/GrpeE chaperone system. J. Mol.
Biol. 269, 757–768.
10. Flaherty, K. M., DeLuca-Flaherty, C. & McKay, D. B.
(1990). Three-dimensional structure the ATPase frag-
ment of a 70 kDa heat shock cognate protein. Nature,
346
, 623–628.
11. Harrison, C. J., Hayer-Hartl, M., Di Liberto, M., Hartl,
F. & Kuriyan, J. (1997). Crystal structure of the
nucleotide exchange factor GrpE bound to the ATPase
domain of the molecular chaperone DnaK. Science,
276
, 431–435.
12. Morshauser, R. C., Hu, W., Wang, H., Pang, Y., Flynn,
G. C. & Zuiderweg, E. R. P. (1999). High resolution
NMR solution structure of the substrate-binding
domain of the mammalian chaperone protein Hsc70.
J. Mol. Biol. 289, 1387–1403.
13. Pellecchia, M., Montgomery, D. L., Stevens, S. Y.,
Vander Kooi, C. W., Feng, H. P., Gierasch, L. M. &
Zuiderweg, E. R. P. (2000). Structural insights into
substrate binding by the molecular chaperone DnaK.
Nature Struct. Biol. 7, 298–303.
14. Stevens, S. Y., Cai, S., Pellecchia, M. & Zuiderweg,
E. R. P. (2003). The solution structure of the bacterial
HSP70 chaperone protein domain DnaK(393–507) in
complex with the peptide NRLLLTG. Protein Sci. 12,
2588–2596.
15. Shi, L., Kataoka, M. & Fink, A. L. (1996). Conformational
characterization of DnaK and its complexes by small-
angle X-ray scattering. Biochemistry, 35, 3297–3330.
16. Wilbanks, S. M., Chen, L., Tsuruta, H., Hodgson, K. O.
& McKay, D. B. (1995). Solution small-angle X-ray
scattering study of the molecular chaperone Hsc70
and its subfragments. Biochemistry, 34, 12095–12106.
17. Gruschus, J. M., Greene, L. E., Eisenberg, E. & Ferretti,
J. A. (2004). Experimentally biased model structure
of the Hsc70/auxilin complex: substrate transfer
and interdomain structural change. Protein Sci. 13,
2029–2044.
18. Klostermeier, D., Seidel, R. & Reinstein, J. (1999). The
functional cycle and regulation of the Thermus
thermophilus DnaK chaperone system. J. Mol. Biol.
287
, 511–525.
19. Groemping, Y., Klostermeier, D., Herrmann, C., Veit,
T., Seidel, R. & Reinstein, J. (2001). Regulation of
ATPase and chaperone cycle of DnaK from Thermus
thermophilus by the nucleotide exchange factor GrpE.
J. Mol. Biol. 305, 1173–1183.
20. Motohashi, K., Yohda, M., Endo, I. & Yoshida, M.
(1996). A novel factor required for the assembly of the
DnaK and DnaJ chaperones of Thermus thermophilus.
J. Biol. Chem. 271, 17343–17348.
21. Watanabe, Y. H. & Yoshida, M. (2004). Trigonal DnaK–
DnaJ complex versus free DnaK and DnaJ: heat stress
182
Structural Indicators of Allostery in DnaK-Tth
converts the former to the latter, and only the latter
can do disaggregation in cooperation with ClpB.
J. Biol. Chem. 279, 15723–15727.
22. Revington, M. & Zuiderweg, E. R. P. (2004). TROSY-
driven NMR backbone assignments of the 381-residue
nucleotide-binding domain of the Thermus Thermo-
philus DnaK molecular chaperone. J. Biomol. NMR, 30,
113–114.
23. Revington, M., Holder, T. M. & Zuiderweg, E. R. P.
(2004). NMR study of nucleotide-induced changes in
the nucleotide binding domain of Thermus thermo-
philus Hsp70 chaperone DnaK: implications for the
allosteric mechanism. J. Biol. Chem. 279, 33958–33967.
24. Chou, C. C., Forouhar, F., Yeh, Y. H., Shr, H. L., Wang,
C. & Hsiao, C. D. (2003). Crystal structure of the
C-terminal 10-kDa subdomain of Hsc70. J. Biol. Chem.
278
, 30311–30316.
25. Misselwitz, B., Staeck, O. & Rapoport, T. A. (1998). J
proteins catalytically activate Hsp70 molecules to
trap a wide range of peptide sequences. Mol. Cell. 2,
593–603.
26. Buczynski, G., Slepenkov, S. V., Sehorn, M. G. & Witt,
S. N. (2001). Characterization of a lidless form of
the molecular chaperone DnaK. J. Biol. Chem. 276,
27231–27236.
27. Wishart, D. S., Sykes, B. D. & Richards, F. M. (1992). The
chemical shift index: a fast and simple method for the
assignment of protein secondary structure through
NMR spectroscopy. Biochemistry, 31, 1647–1651.
28. Fischer, M. W., Losonczi, J. A., Weaver, J. L. &
Prestegard, J. H. (1999). Domain orientation and
dynamics in multidomain proteins from residual
dipolar couplings. Biochemistry, 38, 9013–9022.
29. Hansen, M. R., Mueller, L. & Pardi, A. (1998). Tunable
alignment of macromolecules by filamentous phage
yields dipolar coupling interactions. Nature Struct.
Biol. 5, 1065–1074.
30. Prestegard, J. H., Al-Hashimi, H. M. & Tolman, J. R.
(2000). NMR structures of biomolecules using field
oriented media and residual dipolar couplings. Quart.
Rev. Biophys. 33, 371–424.
31. Flynn, G. C., Chappell, T. G. & Rothman, J. E. (1989).
Peptide binding and release by proteins implicated as
catalysts of protein assembly. Science, 245, 385–390.
32. Pervushin, K., Riek, R., Wider, G. & Wu¨thrich, K.
(1997). Attenuated T2 relaxation by mutual cancella-
tion of dipole–dipole coupling and chemical shift
anisotropy indicates an avenue to NMR structures of
very large biological macromolecules in solution.
Proc. Natl Acad. Sci. USA, 94, 12366–12367.
33. Garrett, D. S., Seok, Y. J., Peterkofsky, A., Clore, G. M.
& Gronenborn, A. M. (1997). Identification by NMR of
the binding surface for the histidine-containing
phosphocarrier protein HPr on the N-terminal
domain of enzyme I of the Escherichia coli phospho-
transferase system. Biochemistry, 36, 4393–4398.
34. Slepenkov, S. V. & Witt, S. N. (1998). Kinetics of the
reactions of the Escherichia coli molecular chaperone
DnaK with ATP: evidence that a three-step reaction
precedes ATP hydrolysis. Biochemistry, 37, 1015–1024.
35. Montgomery, D. L., Morimoto, R. I. & Gierasch, L. M.
(1999). Mutations in the substrate binding domain of
the E. coli 70 kDa molecular chaperone, DnaK, which
alter substrate affinity or interdomain coupling. J. Mol.
Biol. 286, 915–932.
36. O’Brien, M. C., Flaherty, K. M. & McKay, D. B. (1996).
Lysine 71 of the chaperone protein Hsc70 is essential
for ATP hydrolysis. J. Biol. Chem. 271, 15874–15878.
37. Sousa, M. C. & McKay, D. B. (1998). The hydroxyl of
threonine 13 of the bovine 70-kDa heat shock cognate
protein is essential for transducing the ATP-induced
conformational change. Biochemistry, 37, 15392–15399.
38. Zweckstetter, M. & Bax, A. (2000). Prediction of
sterically induced alignment in a dilute liquid crystal-
line phase: aid to protein structure determination by
NMR. J. Am. Chem. Soc. 122, 3791–3792.
39. Gassler, C. S., Buchberger, A., Laufen, T., Mayer, M. P.,
Schroder, H., Valencia, A. & Bukau, B. (1998).
Mutations in the DnaK chaperone affecting inter-
action with the DnaJ cochaperone. Proc. Natl Acad. Sci.
USA, 26, 15229–15234.
40. Suh, W. C., Burkholder, W. F., Lu, C. Z., Zhao, X.,
Gottesman, M. E. & Gross, C. A. (1998). Interaction of
the Hsp70 molecular chaperone DnaK, with its
cochaperone DnaJ. Proc. Natl Acad. Sci. USA, 95,
15223–15228.
41. Mayer, M. P., Rudiger, S. & Bukau, B. (2000).
Molecular basis for interactions of the DnaK chaper-
one with substrates. J. Biol. Chem. 381, 877–885.
42. Zhang, Y. & Zuiderweg, E. R. P. (2004). The 70-kDa
heat shock protein chaperone nucleotide-binding
domain in solution unveiled as a molecular machine
that can reorient its functional subdomains. Proc. Natl
Acad. Sci. USA, 101, 10272–10277.
43. Salzmann, M., Wider, G., Pervushin, K. & Wu¨thrich,
K. (1999). Improved sensitivity and coherence selec-
tion for [
15
N,
1
H]-TROSY elements in triple resonance
experiments. J. Biomol. NMR, 15, 181–184.
44. Delaglio, F., Grzesiek, S., Vuister, G. W., Zhu, G.,
Pfeifer, J. & Bax, A. (1995). NMRPipe: a multidimen-
sional spectral processing system based on UNIX
pipes. J. Biomol. NMR, 6, 277–293.
45. Goddard, T. D. & Kneller, D. G. (2004). SPARKY 3,
University of California, San Francisco.
46. Yang, D., Venters, R. A., Mueller, G. A., Choy, W. Y. &
Kay, L. E. (1999). Trosy-based HNCO pulse sequences
for
the
measurement
of
1
HN–
15
N,
15
N–
13
CO,
1
HN–
13
CO,
13
CO–
13
Ca dipolar couplings in
15
N,
13
C,
2
H-labeled proteins. J. Biomol. NMR, 14, 333–343.
47. Valafar, H. & Prestegard, J. H. (2004). REDCAT: a
residual dipolar coupling analysis tool. J. Magn. Reson.
167
, 228–241.
Edited by M. F. Summers
(Received 12 January 2005; received in revised form 8 March 2005; accepted 10 March 2005)
Note added in proof: While the manuscript was in review, the coordinates for a crystal structure of bovine
hsc70(1-554) have been deposited in the Protein Data Bank. While the relative positioning of the SBD at the
“bottom” of the NBD is the same as in our structure, the centers of the interface on the NBD in their and our
structure are shifted by as much as 20 A
˚ .
Structural Indicators of Allostery in DnaK-Tth
183
Document Outline
- NMR Investigations of Allosteric Processes in a Two-domain Thermus thermophilus Hsp70 Molecular Chaperone
Wyszukiwarka
Podobne podstrony:
1 s2 0 S0022283610008843 main
1 s2 0 S0020025512001946 main
1 s2 0 S0378382002000085 main
1 s2 0 S0304397599001000 main
1 s2 0 S0006291X05021595 main
1 s2 0 S0040603111000104 main 2
1 s2 0 S0944501312001358 main
1 s2 0 S0166218X96000583 main
1 s2 0 S0005273614000303 main
1 s2 0 S0304397502001342 main
1 s2 0 S0377221798003622 main (1)
1 s2 0 S0022169496031423 main
1 s2 0 S1046592814002101 main
1 s2 0 0166218X93E0153P main
1 s2 0 S0022000006001474 main
1 s2 0 S000925099800520X main
1 s2 0 S0006291X07005785 main
1 s2 0 S0960852409006385 main
więcej podobnych podstron