close

Вход

Забыли?

вход по аккаунту

?

High-Resolution Solid-State NMR Spectroscopy of the Prion Protein HET-s in Its Amyloid Conformation.

код для вставкиСкачать
Angewandte
Chemie
Protein Structures
High-Resolution Solid-State NMR Spectroscopy
of the Prion Protein HET-s in Its Amyloid
Conformation**
Ansgar B. Siemer, Christiane Ritter, Matthias Ernst,
Roland Riek, and Beat H. Meier*
High-resolution solid-state NMR spectroscopy is being established as a technique to enable the determination of biological
macromolecular structure with atomic-level resolution. The
main focus is on microcrystalline and noncrystalline solids
with particular emphasis on the latter for which there are
presently no suitable structure-determination methods available. For nanocrystalline compounds, the potential of solidstate NMR spectroscopy has been clearly demonstrated by
the complete resonance assignment of several proteins and
peptides with sizes of up to 100 residues[1–5] and by the
structure determination of an SH3 domain with 62 amino
acids.[6] Herein we demonstrate the applicability of highresolution solid-state NMR-spectroscopic techniques to the
structure determination of amyloid fibrils of a prion protein.
The aforementioned pilot applications of solid-state NMR
spectroscopy to structure determination were made possible
by considerable technical progress over the past few years and
through new developments in pulse-sequence methods.[7] In
particular, the line width in the 13C and 15N NMR spectra of
uniformly labeled microcrystalline compounds has been
decreased to 0.5–0.8 ppm.[5, 8] These line widths are partially
limited by homonuclear J couplings, and in samples that lack
directly bound 13C atoms, line widths as small as 0.2 ppm have
been observed.[8] Herein we demonstrate that approximately
the same resolution (up to 0.25 ppm for resonances with
resolved J couplings or 0.5 ppm for those without resolved J
couplings) can also be obtained in spectra of amyloid fibrils, in
our case the amyloidogenic fragment 218–289 of the HET-s
prion protein from the fungus Podospora anserina.
The HET-s protein is involved in a genetically controlled
programmed-cell-death phenomenon termed heterokaryon
incompatibility.[9, 10] It is encoded by the het-s gene, one of at
least nine het loci that are involved in heterokaryon incompatibility, and exists in two different forms: a nonprion form
and a prion form. The prion form aggregates in vivo to
generate amyloid fibrils.[11] Recombinant HET-s can also
[*] A. B. Siemer, Dr. M. Ernst, Prof. Dr. B. H. Meier
Laboratory of Physical Chemistry, ETH Zurich
ETH-Hnggerberg, 8093 Zurich (Switzerland)
Fax: (+ 41) 1632-1621
E-mail: beme@ethz.ch
Dr. C. Ritter, Prof. Dr. R. Riek
Structural Biology Laboratory, The Salk Institute
La Jolla, 92037 CA (USA)
[**] We are grateful to Andreas Hunkeler and Urban Meier for their
technical support, to Dr. Ren Verel and Rochus Keller for scientific
advice, and to Dr. Thorsten Luehrs for the EM pictures. This work
was supported by the Swiss National Science Foundation (SNF)
and the ETH Zurich.
Angew. Chem. Int. Ed. 2005, 44, 2441 –2444
aggregate in vitro to form amyloid fibrils,[12] in which the Cterminal fragment that consists of residues 218–289 (HETs(218–289)) forms the protease-resistant part of the fibrils.
This fragment has the sequence KI DAIVGRNSAK DIRTEERARV QLGNVVTAAA LHGGIRISDQ TTNSVETVVG KGESRVLIGN EYGGKGFWDN and alone forms
amyloid fibrils in vitro.[13] Strains that express HET-s in the
nonprion form ([Het-s*]) can undergo a transition to the
prion-expressing state ([Het-s]) through contact with [Het-s]
strains, by introduction of HET-s fibrils composed of fulllength recombinant HET-s,[14] or fibrils formed by the fragment HET-s(218–289).[13] Therefore this fragment is thought
to be not only the amyloid-forming part, but also the prionpropagating portion of the HET-s protein.
To our knowledge, there are as yet no published solidstate NMR spectra of prions with sufficient resolution to
permit the assignment of resonances in a uniformly labeled
sample. In contrast, similarly well-resolved spectra have been
recorded for other amyloid fibrils such as the A-b-peptide[15, 16] and transthyretin(105–115),[17] in which line widths of
1–2 ppm were found as well as even narrower lines for certain
methyl groups.[16] Solid-state NMR spectra of fibrils from a
fragment of the mammalian prion protein (PrP) have been
obtained[18, 19] and show a lower spectral resolution.
Figure 1 shows the homonuclear 13C,13C-correlation spectra of fibrils derived from HET-s(218–289) with a C-terminal
His tag collected by using a DREAM polarization transfer
(DREAM = dipolar recoupling enhanced by amplitude modulation)[20, 21] at a MAS frequency of 40 kHz (MAS = magicangle spinning). The C’ Ca and C’ Ca Cb correlations in
Figure 1 a were recorded in two separate experiments in
which the tangential amplitude ramp in the DREAM pulse
sequence was taken from low to high amplitude in one case
and from high to low amplitude in the other. The labels given
in Figure 1 correspond to the tentative assignment of the
resonance frequencies as listed in Table 1. The assignment
was obtained from 2D and 3D heteronuclear 15N,13C,13Ccorrelation spectra with adiabatic polarization-transfer techniques.[1, 22] This correlation identifies two separate but contiguous segments of the peptide, namely the 23 residues of
positions 226–248 and the 20 residues of positions 262–282
(see also Table 1), in which amino acid 274 of the second
segment is missing. These resonances explain almost all the
signals in the spectra, except for some unassigned cross-peak
intensities in the aromatic region and very few unassigned
cross-peaks in other regions, leaving 29 residues whose signals
cannot be detected. The resonances from the six residues in
the His tag are also unassigned. The missing resonances are
not a consequence of inefficient cross-polarization. Protondriven spin-diffusion (PDSD) correlation spectra without
cross-polarization (direct excitation of carbon coherence by a
908 pulse) showed additional diagonal peak intensity in the
aromatic region, but no new cross-peaks that could lead to the
assignment of the backbone atoms of the missing 29 residues.
Interestingly, NOE presaturation provided significant signal
enhancement for our samples, but no additional resonances
were observed.
The line width of the observed resonances is astonishingly
narrow. Figure 2 is an extracted portion of Figure 1 which
DOI: 10.1002/anie.200462952
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2441
Communications
Figure 1. Homonuclear 13C DREAM correlation spectra of fibrillized HET-s(218–289) recorded at 40 kHz MAS. The lower level of the contours was
chosen just above the noise level (5 % of the maximal intensity); adjacent contour levels differ by a factor of 1.4. Positive contours are shown in
blue, negative contours in red. Spectra underwent Fourier transformation with a shifted sin2 window function with zero filling to 2048 data points.
a) Carbonyl region of the DREAM spectra. Positive contours were taken from a spectrum recorded with a tangential DREAM sweep from high- to
low-amplitude modulation; negative contours were taken from a similar spectrum but with a sweep from low- to high-amplitude modulation.
b) Aliphatic region of the DREAM spectrum recorded with a tangential DREAM sweep from low- to high-amplitude modulation.
shows the well-isolated cross-peaks of Gly 271 and Ala 228
and a 1D cut through their centers. In the d2 dimension, the Jsplit doublet is clearly visible, and several other carbonyl
resonances are also resolved sufficiently to show the J
splitting. The line shape of Gly 271, for example, can be
fitted by two Gaussian lines with a full width at half height of
30 Hz or 0.2 ppm (left multiplet line) and 37 Hz or 0.24 ppm
(right multiplet line) with a splitting of 54 Hz. This line is as
narrow as corresponding resonances from microcrystalline
samples which indicates that the fibrils in HET-s are highly
ordered on a molecular scale.
This observation of high molecular order for part of the
molecule stands in striking contrast to the absence of signals
from more than one third of the residues from the spectrum
which, almost certainly, must be explained by pronounced
molecular disorder. This disorder can either be of static or
dynamic origin and the spectra measured so far do not permit
an unequivocal distinction. The disorder affects only parts of
the molecule and must be distinguished from the helical pitch
variation observed, for example, by cryoelectron micro-
2442
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
copy.[23] Static disorder can lead to line widths of up to
10 ppm in proteins, which is the range of resonance
frequencies for the backbone carbon resonances when
the two backbone torsion angles vary over the entire
range.[24] For such an extreme disorder, the line width
would increase by a factor of 20 in a 1D spectrum and
up to 202 in 2D spectra; such lines may escape detection.
Dynamic disorder can also lead to very broad lines if the
molecular motion has spectral components in the kHz range,
and interferes with the averaging by MAS or with the
decoupling. We can exclude T11 effects as the cause of the
missing signals of the 29 residues, owing to the absence of
additional cross-peaks in the PDSD spectra as discussed
above.
An attempt was made to freeze out the dynamics of the
missing residues. The resulting 1D variable-temperature
13
C CP-MAS spectra (CP = cross-polarization) of HETs(218–289) fibrils are shown in Figure 3. Significant line
broadening is observed below 20 8C, but the number of
observed resonances seems to be independent of temper-
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 2441 –2444
Angewandte
Chemie
Table 1: Chemical shifts (13C and 15N) of the peptide backbone of the
assigned regions of HET-s(218–289).[a]
Residue
N
C’
Ca
Residue
N
C’
Ca
Asn 226
Ser 227
Ala 228
Lys 229
Asp 230
Ile 231
Arg 232
Thr 233
Glu 234
Glu 235
Arg 236
Ala 237
Arg 238
Vla 239
Gln 240
Leu 241
Gly 242
Asn 243
Val 244
Val 245
Thr 246
Ala 247
Ala 248
125.5
119.2
122.7
123.6
117.7
122.7
129.8
113.5
120.0
117.3
122.8
126.0
117.9
122.7
126.7
131.4
113.7
109.9
122.8
129.3
116.7
121.1
119.3
–
169.9
174.9
–
–
172.6
–
172.5
171.8
172.2
174.3
173.9
172.7
171.1
171.1
–
168.7
174.3
172.5
–
173.3
177.4
177.6
49.9
54.7
47.4
57.7
51.1
58.9
52.1
57.6
51.7
56.6
52.6
51.1
52.9
58.0
50.7
50.6
41.8
49.2
60.0
59.2
60.4
54.1
53.0
Asn 262
Ser 263
Val 264
Glu 265
Thr 266
Val 267
Val 268
Gly 269
Lys 270
Gly 271
Glu 272
Ser 273
128.2
117.6
125.6
127.3
112.8
124.1
128.6
113.4
121.5
115.6
128.6
116.3
–
169.8
172.5
174.0
171.0
–
173.2
170.0
174.7
170.1
174.3
171.5
50.3
54.3
55.0
57.4
58.9
58.5
59.2
42.5
52.3
46.4
52.1
57.2
Val 275
Leu 276
Ile 277
Gly 278
Asn 279
Glu 280
Tyr 281
Gly 282
122.9
130.6
130.3
111.3
114.5
120.4
128.9
109.9
170.7
–
171.3
170.0
171.0
–
174.4
172.3
58.2
50.6
56.6
41.9
49.8
52.6
54.0
44.0
[a] 13C NMR chemical shifts were referenced to tetramethylsilane,
15
N NMR chemical shifts to liquid ammonia.
Figure 3. Variable-temperature 13C-CP spectra of HET-s(218–289)
recorded at 25 kHz MAS: a) 19 8C, 256 scans; b) 10 8C, 256 scans;
c) 31 8C, 256 scans; d) 19 8C (after cooling down to 31 8C), 430
scans.
Experimental Section
Figure 2. Detail of Figure 1 showing the Ca C’ cross-peaks of Ala 228
and Gly 271 and 1D cuts through these signals. To get maximal resolution, no window function was applied before the Fourier transformation. The J-split doublet of Gly 271 is best fit with a pair of Gaussian
lines of line width = 30 Hz (left multiplet line), 37 Hz (right multiplet
line), and a J coupling of 54 Hz. Such differences in line width of a Jsplit doublet have also been observed in simple model compounds
but is, to our knowledge, not yet understood.
ature. The changes in spectra with temperature were found to
be reversible.
We have shown that the HET-s(218–289) prion protein
fibrils are highly ordered for approximately two-thirds of the
structure and have found strong evidence for pronounced
disorder in the remaining region. The ordered and disordered
parts of the protein can be assigned to different protein
segments (Table 1). The observation of narrow lines in the
NMR spectra and the assignment of the resonances demonstrate the potential of solid-state NMR spectroscopy to
elucidate the structures of these molecules in the amyloid
form. The collection of structural-constraint information from
solid-state NMR spectroscopy is currently in progress.
Angew. Chem. Int. Ed. 2005, 44, 2441 –2444
All spectra were recorded with a Bruker AV600 spectrometer (1Hresonance frequency = 600.13 MHz) at a static field of 14.09 T. The
1D CP-MAS spectra were recorded with a Varian probe (2.5 mm) at a
MAS frequency of 25 kHz. The initial cross-polarization was done for
1.5 ms with RF-field amplitudes of 70 and 95 kHz on the 13C and 1H
channels, respectively. Decoupling was performed at an RF-field
amplitude of 130 kHz by using the XiX decoupling Scheme (tp =
74 ms, XiX = X inverse X).[25] 13C,13C-correlation spectra were
recorded with a 1.8-mm probe[26] at a MAS frequency of 40 kHz
and a probe temperature of 15 8C. The initial cross-polarization was
done for 1 ms with RF-field amplitudes of 90 and 130 kHz on the 13C
and 1H channels, respectively. During the DREAM[20, 21] recoupling
period of 7 ms the RF-field amplitude was tangentially swept between
15 and 26 kHz under the following parameters:[21] DRF/2p = 4.5 kHz,
dest = 1.5 kHz and w̄1/2p = 20 kHz. Continuous-wave decoupling was
applied during the DREAM recoupling, and XiX decoupling (tp =
70 ms) was applied during t1 and t2, both with a RF-field amplitude of
150 kHz. The spectral width was 40 kHz and 62 scans were added
together for each of the 1024 t1 increments which led to a total
acquisition time of 55 h. Data were processed with XWINMR
(Bruker Biospin) and analyzed with CARA.[27]
Received: December 15, 2004
Published online: March 16, 2005
.
Keywords: fibrils · NMR spectroscopy · prions ·
protein structures · structure elucidation
[1] A. Detken, E. H. Hardy, M. Ernst, M. Kainosho, T. Kawakami,
S. Aimoto, B. H. Meier, J. Biomol. NMR 2001, 20, 203.
[2] A. McDermott, T. Polenova, A. Bckmann, K. W. Zilm, E. K.
Paulsen, R. W. Martin, G. T. Montelione, J. Biomol. NMR 2000,
16, 209.
www.angewandte.org
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2443
Communications
[3] A. Bckmann, A. Lange, A. Galinier, S. Luca, N. Giraud, M. Juy,
H. Heise, R. Montserret, F. Penin, M. Baldus, J. Biomol. NMR
2003, 27, 323.
[4] J. Pauli, M. Baldus, B. van Rossum, H. de Groot, H. Oschkinat,
ChemBioChem 2001, 2, 272.
[5] T. I. Igumenova, A. E. McDermott, K. W. Zilm, R. W. Martin,
E. K. Paulson, A. J. Wand, J. Am. Chem. Soc. 2004, 126, 6720.
[6] F. Castellani, B. van Rossum, A. Diehl, M. Schubert, K.
Rehbein, H. Oschkinat, Nature 2002, 420, 98.
[7] S. Dusold, A. Sebald, Annu. Rep. NMR Spectrosc. 2000, 41, 185.
[8] F. Castellani, B. J. van Rossum, A. Diehl, K. Rehbein, H.
Oschkinat, Biochemistry 2003, 42, 11 476.
[9] N. L. Glass, I. Kaneko, Eukaryotic Cell 2003, 2, 1.
[10] S. J. Saupe, Microbiol. Mol. Biol. Rev. 2000, 64, 489.
[11] V. Coustou-Linares, M.-L. Maddelein, J. Begueret, S. J. Saupe,
Mol. Microbiol. 2001, 42, 1325.
[12] S. Dos Reis, B. Coulary-Salin, V. Forge, I. Lascu, J. Begueret, S. J.
Saupe, J. Biol. Chem. 2002, 277, 5703.
[13] A. Balguerie, S. Dos Reis, C. Ritter, S. Chaignepain, B. CoularySalin, V. Forge, K. Bathany, I. Lascu, J.-M. Schmitter, R. Riek,
S. J. Saupe, EMBO J. 2003, 22, 2071.
[14] M. L. Maddelein, S. Dos Reis, S. Duvezin-Caubet, B. CoularySalin, S. J. Saupe, Proc. Natl. Acad. Sci. USA 2002, 99, 7402.
[15] R. Tycko, Curr. Opin. Struct. Biol. 2004, 14, 96.
[16] A. T. Petkova, G. Buntkowsky, F. Dyda, R. D. Leapman, W. M.
Yau, R. Tycko, J. Mol. Biol. 2004, 335, 247.
[17] C. P. Jaroniec, C. E. MacPhee, V. S. Bajaj, M. T. McMahon, C. M.
Dobson, R. G. Griffin, Proc. Natl. Acad. Sci. USA 2004, 101, 711.
[18] J. Heller, A. C. Kolbert, R. Larsen, M. Ernst, T. Bekker, M.
Baldwin, S. B. Prusiner, A. Pines, D. E. Wemmer, Protein Sci.
1996, 5, 1655.
[19] D. D. Laws, H. M. L. Bitter, K. Liu, H. L. Ball, K. Kaneko, H.
Wille, F. E. Cohen, S. B. Prusiner, A. Pines, D. E. Wemmer, Proc.
Natl. Acad. Sci. USA 2001, 98, 11 686.
[20] R. Verel, M. Baldus, M. Ernst, B. H. Meier, Chem. Phys. Lett.
1998, 287, 421.
[21] R. Verel, M. Ernst, B. H. Meier, J. Magn. Reson. 2001, 150, 81.
[22] M. Ernst, B. H. Meier in Encyclopedia of Nuclear Magnetic
Resonance, Vol. 9 (Eds.: D. M. Grant, R. K. Harris), Wiley,
Chichester, 2002, pp. 23.
[23] J. L. Jimenez, J. I. Guijarro, E. Orlova, J. Zurdo, C. M. Dobson,
M. Sunde, H. R. Saibil, EMBO J. 1999, 18, 815.
[24] A. C. Dedios, J. G. Pearson, E. Oldfield, Science 1993, 260, 1491.
[25] A. Detken, E. H. Hardy, M. Ernst, B. H. Meier, Chem. Phys.
Lett. 2002, 356, 298.
[26] A. Samoson, T. Tuherm, J. Past, Chem. Phys. Lett. 2002, 365, 292.
[27] R. L. J. Keller, The Computer Aided Resonance Assignment
Tutorial, Cantina, Goldau, 2004.
2444
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 2441 –2444
Документ
Категория
Без категории
Просмотров
0
Размер файла
231 Кб
Теги
conformational, spectroscopy, resolution, solis, prior, het, nmr, high, protein, amyloid, state
1/--страниц
Пожаловаться на содержимое документа