вход по аккаунту


Modified Triple Resonance NMR Experiments with Optimized Sensitivity for Rapidly Exchanging Protons.

код для вставкиСкачать
Chrm. I n / . Ed. Engl. 1981,2U. 692; 0. I. Shchegolikhina, A. A. Zhdanov, V. A.
Igonin, Yu. E. Ovchinnikov, V. E. Shklover, Yu. T. Struchkov, Metalloorgan.
Khrm. 1991. 4. 74: V. A. Igonin. 0. I. Shchegolikhina, S. V. Lindeman, M. M.
Levitsky. Yu. T. Struchkov, A. A. Zhddnov, J. Organornet. Chem. 1992,
423. 351.
[I31 P. Woodward, L. E Dahl, E. W. Abel, B. C. Crosse, J. Am. Chem. Soc. 1965,87,
5251; R. 0 . Gould. M. M. Harding, J. Chem. Soc. A 1970, 875; N. R.
Kunchur, Actu Crjstullogr. Sert. B 1968, 24, 1623; E. W. Abel. B. C. Crosse,
1 Chm7. So<,.A 1966. 1377; I. G. Dance, M. L. Scudder, R. Secomb. Inorg.
Chem.1985, 24. 3201
[14] J.-F. You, a.C . Papaefthymiou, R. H. Holm. J. Am. Chem. Soc. 1992. 114.
2697. J.-F. You, B. S. Snyder, G. C. PdpdefthymiOu, R. H. Holm, ibid. 1990,
112, 1067; W. Saak. G. Henkel, S. Pohl, Angen. Chem. 1984, 96. 153; Angew.
Cl7c.m. Int. Ed. Engl. 1984, 23, 150.
[I51 K. L. Taft. C . D. Delfs. G. C. Papdefthymiou. S. Foner, D. Gdtteschi, S. J.
Lippard. J. .4m. Chrm. Soc. 1994. 116, 821.
[I61 A J. Blake, R. 0. Gould, P. E. Y Milne, R. E. P. Winpenny, J. Chem. Soc.
Chem Cornmiin. 1991, 1453; A. J. Blake. R. 0. Gould, C. M. Grant, P. E. Y
Milne, D. Reed. R. E. P. Winpenny, Angew. Chem. 1994, 106. 208; Angew.
Chem. Int. EX. EngL 1994, 33, 195.
[I71 M. S. Lah. M. L. Kirk, W. Hatfield. V. L. Pecoraro, 1 Chem. Sor. Chem.
Cummun. 1989. 1606; M. S. Lah, V. L. Pecoraro. J. Am. Chem. Soc. 1989,111,
7258; B. R. Gibney. A. J. Stemmler, S. Pilotek, J. W. Kampf, V. L. Pecoraro,
Inorg. Cheni 1993, 32, 6008.
[181 H. T. Evans. Jr.. Artu Crjslallogr. Seer. B 1974, 30, 2095.
[I91 V. S. Nair, A. Kitaygorodskiy, K. S. Hagen. Abstracts, 206th ACS National
M w l i n g (Chicago). 1993.
Modified Triple Resonance NMR Experiments
with Optimized Sensitivity for Rapidly
Exchanging Protons**
Wolfgang Jahnke and Horst Kessler*
Multidimensional N M R spectroscopy has developed into the
premier method for determining high-resolution structures of
molecules in solution. A major concern in the development of
pulse sequences is improving the sensitivity of the N M R experiments.['] The weakest signals in N M R spectra often arise from
amide protons that exchange rapidly with the solvent water, and
these weak signals often constitute the limiting factor in signal
assignment and structure determination of proteins. The intensity of these resonances can be enhanced when the water signal
is not saturated during the pulse s e q ~ e n c e . ' ~ We
- ~ ] describe here
modified triple resonance experiments in which the water signal
is not saturated and composite pulse proton decoupling is used.
The modified pulse sequences yield a sensitivity gain of up to
50 % for rapidly exchanging amide protons, where enhanced
sensitivity is of outstanding importance.
Biological macromolecules like proteins and nucleic acids are
usually investigated in aqueous solution and under semiphysiological conditions, in other words at roughly neutral pH. Under
these conditions chemical exchange takes place between protons
of the molecule under investigation (e.g. a protein) and the solvent water. Besides protons of ionizable groups in the side chains,
backbone amide protons exchange with the solvent protons at
rates ranging from reciprocal milliseconds to reciprocal years.
Hydrogen exchange has been the subject of several investigat i o n ~ , [ ~and
* ~ it] can be expected that the dynamic feature of
[*] Prof. Dr. H. Kessler, DipLChem. W. Jahnke
lnstitut fur Organische Chemie und Biochemie
der Technischen Universitit Munchen
Lichtenbergstrasse 4. 85747 Garching (Germany)
Telefax: Int code (89)3209-3210
[**I This work was supported by the Deutsche Forschungsgemeinschaft and the
Fonds der Chemischen Industrie (H. K.), and the Studienstiftung des
deutschen Volkes (fellowship for W. J.). We thank Dr. Gerd Gemmecker for
stimulating discussions.
Angen Chem. In/. Ed. Engl. 1995, 34, N o . 4
hydrogen exchange can enhance our understanding of structure
and dynamics of proteins and nucleic acids. Hydrogen exchange, however, is not a pure blessing. Very rapidly exchanging
protons such as hydroxyl groups of protein side chains or amide
protons at the protein N-terminus are not found in the spectrum
under normal conditions, and crucial amide proton signals are
attenuated if presaturation is used.
It has recently been pointed out'', 31 that the signals of rapidly
exchanging protons are attenuated not only by presaturation,
but also by other methods used to suppress the water signal such
as spinlock pulses['] and gradients,[*] when the water signal is
saturated prior to acquisition. This is due to the long TI relaxation time of water protons (4-5 s), which is much longer than
the TI time of protein protons. The delay between successive
scans is optimized for the relaxation of the protein protons and
is too short for the magnetization of the water protons to relax
back near thermal equilibrium. Instead, the water magnetization remains in a semisaturated state during the entire experiment, typically reaching less than 3 0 % of its equilibrium value
just before the next scan.
Several methods have been described for overcoming this
problem by not saturating the water signal prior to acquisition,
while still ensuring that water does not interfere with the detection of the desired protein signals. These methods include selective pulses on the water resonance,121improved phase cycling of
proton pulse^,[^.^] and the constructive use of radiation damping;[41all of these methods result in alignment of water magnetization on the + z axis before or shortly after acquisition.
While selective pulses may be undesirable due to imperfect excitation profiles,['] neither of the other two methods describes the
possibility for proton decoupling by a composite pulse decoupling sequence. This type of decoupling is desirable when the
desired coherence resides on nitrogen or carbon because of more
favorable relaxation behavior." It is not immediately obvious
how composite pulse decoupling can be incorporated into pulse
sequences where water magnetization must end up in a defined
state (e.g. at the + z axis) before acquisition. First, it seems that
the decoupling time must be an integer multiple of the time required for one decoupling cycle, but this is hard to achieve if the
decoupling time includes an incrementable delay such as the t ,
time. Second, B, field gradients applied during composite pulse
decoupling interfere with the decoupling and should be avoided.
We propose here a modification of pulse sequences that allows
the use of composite pulse decoupling for any desired time period,
while improved sensitivity of rapidly exchanging amide protons is
still achieved since water magnetization is realigned along the + z
axis prior to acquisition. The modification is illustrated with the
constant time (CT) HNCA and HN(C0)CA experiment["3 1 2 ]
in Figure 1, but can be applied to many N M R experiments.
The CT-HNCA sequence (Fig. 1 a) starts with excitation of the
amide protons, whose magnetization is subsequently transferred
to 5N by an INEPT transfer. During the delay needed to develop
the I5N-l3Cacoupling, 15Ncoherence refocuses with respect to
protons, and is completely refocused after l j 2 'J(N,H), approximately 5.5 ms. At this time composite pulse decoupling should
start. The decoupling phase is irrelevant for protein protons, but
is crucial to keep water magnetization in a defined state. The
water magnetization has been aligned at the - J axis by the first
90" proton pulse with phase x; it stays there since its resonance
corresponds to the carrier frequency and is thus not affected by
the following two proton pulses with phase / - y . Proton decoupling is now possible if a sequence is selected that applies
pulses only along one axis, and if this axis is chosen to be the y
axis. The DIPSI-2 sequence," 31 for example, can be applied
successfully along the y axis in this case. After 13Cafrequency
Verlagsgesellschrrfrfi mbH. 0-69451 Weinhrim, 1995
U57U-0833195/0404-046Y $ lO.OU+ .35,'0
HAM"" at 310K and p H 7.5.['4J This homodimer of 31 kDa
contains a number of rapidly exchanging amide protons,['] for
which a considerable increase in sensitivity is expected when the
water magnetization is flipped back to thermal equilibrium before
acquisition. Figure 2 shows rows through the first 2D planes of
E 135
b) CT-HN(C0)CA
6 ('H)
Fig. 1. Modified pulse sequences of a) CT-HNCA and b) CT-HN(C0)CA experiments. The strength of the B , decoupling field was 8.3 kHz. Thin and thick bars
refer to 90' and 180" pulses. respectively. Pulses are applied along the .Y axis. unless
otherwise indicated. Variable pulse phases are as follows: a)
= s, --I; $ 2 = v.
-J'; 4, = .Y: $+ = 4(1). 4 ( ~ ) .4( - x). 4( - J'); $s = 16(.~), 16( - s ) : $7 = 2(.\).
2( - s): receiver = ~(.Y.-.Y.-.Y.x,Y,xJ-.Y),
2( - .~.r,u.-u.s.-.~.-s..~). b)
4, = s, -.r: 4, =,v, -,v;
= I: $+ = 4 ( ~ ~4c~3).
) , 4(- s),4(-,v): + 5 = 2 ( s ) ,
2(- x): $(, = J'. - v ; receiver = 2 ( . ~ ) ,4(- .Y), 2 ( s ) . Pulsed B,, field gradients are
sine-shaped and are applied for 3 ms. followed by a delay o f 0 3 ms. Their relative
strength is G , = - 39.6. G , = 39.6 and G , = 4. with 100 being approximately
50 Gausscm-' at the center of the gradient. Gradient G, must be carefully optimized for inaximam refocusing of water. Delays are as follows: A , = 2.25 ms, 1. =
5.5ms. d, =11 ms. A', = A 2 - T ~ ~ J ~ ~ D
C , O=lOms,
) ,
A; = A , - 7,x0('3CO).
A , = 6ms.
labeling in t , with decoupling of 15Nand I3CO, a Bloch-Siegert
shift compensating for the 180" pulse on I3CO, and C T ' * N
evolution in t , , the coherence is transferred back to the amide
protons, and is refocused and detected. Gradients G I and G, are
included to select for the desired coherence pathway.
Water magnetization is spinlocked along the - y axis by the
DIPSI-2 decoupling sequence and is defocused by gradient G I .
It is not possible to refocus water magnetization by splitting G I
into two gradients of the same strength but opposite sign on
each side of a "N 180" pulse, since one of the gradients would
have to be applied while composite pulse decoupling is active.
Instead, the gradient for refocusing water magnetization is applied after the following "N 90" pulse, at a time when the protein
magnetization is in an ZzSzstate but water magnetization is in the
transverse plane. After the refocusing gradient G,, the water magnetization is aligned along the -y axis. It is then turned to the --i
axis by the 90" pulse with phase & = x, and turned to the + z
axis by the following 180" proton pulse. A one-dimensional experiment shows that the longitudinal water magnetization reaches as
much as 80% of its equilibrium value after this sequence. That
means that exchanging amide protons obtain almost the full
longitudinal magnetization from water protons and their signals
are not a priori attenuated by the exchange process. Any water
magnetization that is not realigned along the + z axis (due to
imperfect pulses or radiation damping) is defocused by gradient
G,, which supplies excellent suppression of the water signal.
The modified pulse sequence has been tested on a 2 m sample
of completely "N- and 3C-labeled mannose permease domain
:c*~ VCH
V e r l u ~ s ~ e s e l l s ~ l mhH,
i u f ~ 0-69451 Wrinheirn, 199.5
Fig. 2. ' H rows from the first planes ofCT-HNCA spectra taken at a "N frequency of 128.7. It can be seen that the proposed pulse sequence of Fig. 1 a gives the best
sensitivity for the rapidly exchanging amide proton. Residues A128 and E l 3 5
resonate at a slightly different I5N frequencies. All experiments were recorded and
processed under identical conditions on a Bruker AMX6OO spectrometer with triple
resonance equipment and self-shielded pulsed field gradients. 256 transients were
accumulated for each of the 80 i , increments. The recycle time between successive
Scans including acquisition was 1.1 s.
spectra obtained with different versions of HNCA experiments :
the C T experiment using 180" pulses for proton decoupIing["]
(curve a), the C T experiment with composite pulse proton decoupling["] and saturation of the water signal (48= y in Fig. 1 a)
by pulsed field gradients (curve b), and CT-HNCA (as described
in Fig. 1a) using composite pulse proton decoupling and flipping
water magnetization back to thermal equilibrium (curve c). In
addition the corresponding row through a MEXICO spectrum
is shown,"] in which only rapidly exchanging protons give rise
to cross peaks (curve d). It is obvious that proton decoupling by
a composite pulse decoupling sequence (curve b) yields considerably higher sensitivity than decoupling by 180" pulses (curve a),
as expected for a protein of this size.["' An additional substantial increase in sensitivity is gained for rapidly exchanging amide
protons if the water signal is not saturated prior to acquisition.
This is seen in curve c for residue E 135, which has a rapidly
exchanging amide proton (curve d). The signal-to-noise ratio
for this peak is about 50 % higher when the water magnetization
is flipped back to equilibrium before acquisition. Residue E 135
is located near the C-terminus of the protein in a flexible part of
the molecule that is not involved in regular secondary structure.
Thus, E 135 gives rise to a strong cross peak in spite of its rapidly
exchanging amide proton. It is important to note, however, that
since most other rapidly exchanging amide protons yield much
weaker signals, the sensitivity enhancement obtained specifically for these signals is of outstanding importance.
In conjunction with the described possibility of using composite pulse proton decoupling, the technique of flipping water
magnetization back to thermal equilibrium before acquisition
can be applied without compromise to a wide variety of N M R
experiments. Since the technique enhances the signal intensity
predominantly for the weak resonances, for which enhanced
sensitivity is most important, it can be expected that assignment
of N M R spectra is greatly facilitated by this technique.
Received: August 12. 1994 [Z7235IE]
German version: A n g w . Cheni. 1995. 107, 536
Keywords : N M R spectroscopy . structure elucidation . proteins
0570-0833/Y5/0404-0470 $ 10.00+ 2510
A n R i w Cheni. Int. €d. €nn/. 1995, 34. No. 4
( I ] A. G Palmcr. J. Cavanagh. P. E. Wright, M. Rance. J. Mugfl. Reson. 1991. Y3.
151 170. J. Schleucher. M. Sattler. C. Griesinger. Angpw. Chrm. 1993, 105,
151X-1521: .4n,qw. Cheni. I n f . Ed. EngI. 1993.32. 1489-1492; L.E. Kay, P.
Keil'er. T. Saarinen. J. Am. Chrm. S i c . 1992. 114, 10663-10665.
[2] S. Grzcsiek. A . Bax. J Am. Chrm. So(,.1993, f15,12593 -12594: N . A. Farrow,
R . Muhandiram. A. U. Singer, S. M. Pascal. C. M. Kay. G. Gish. S . E. Shoelson. T. Pawson. J. D. Forman-Kay. L. E. Kay. Biochemistr?. 1994, 33, 5984-
[3] J. Stonchouse. C i . L. Shaw, J. Keeler. E. D. Laue. J Mugn. Rcson. &r. A 1994.
107. 178 - 1 X4
[4] W Jahnke. H. Kessler. J. Biomoi. N M R , 1994, 4. 735---740.
[5] K. Wuthrich. V M R u/ Proteins und Nideic Acids. Wiley-Interscience, New
York. 1986: E. Liepinsh, G. Otting. K. Wuthrich. J. Biomol. N M R 1992, 2,
447 465. R . Baldwin. Curr. Opm. Slruci. Biol. 1993. 3. 84-91; C. K. Woodward. ;hid 1994. 4. 112- 116; s.W. Englander. L. Mayne. Annu. Rrs. Binphy.7.
B I O I S!ru[r
~ .
1992. 2f. 243-265; C. Chydn. C. Wormald, C. Dobson. P.
Evans. J Baum. Biuchemistrj 1993.32. 5681 -5691; C. A. Rohl. R. L. Baldwin.
h d 1994. -13, 7760-7767: N. J. Skelton. J. Kiirdel. M. Akke. W. J. Chazin, J.
Mol. B i d 1992. 227. 1100- 1117: M H. Werner. D. E. Wemmer. J ihid. 1992,
225. X73 XX9: J. Craplicki. C. Arrowsmith. 0. Jardetzky, J Biomol. N M R
1991. I, 349-361; S. Spera. M. Ikura. A. Bax, ihid. 1991, 1 , 155-165: S.
Grzesiek, .A. Bax. ihid. 1993. 3. 627-638.
(61 G. Gemmecker. W. Jahnke, H. Kessler, J. Am. Chrm. Sot. 1993. f l S . 1162011621
[7] B A. Meacerle. G. Wider, G. O t t i h . C. Weber. K. Wiithrich, J. Mugn. Resoti.
1989. 85. hit8 613.
[XI A. Bax. P. ( i . De Jong. A. F. Mehlkopf, J. Smidt. Chem. Phyy. Letr. 1980, 69.
567 570: R . E. Hurd, J Mugn. ReJon. 1990. 87, 422-428
[9] R . Freeman. C'hiw? Rev. 1991, Y1. 1397-1412; H. Kessler. S. Mrongd, G.
Gemmecker. .Mugn. R i w n . Chem. 1991, 29. 527 --557.
[lo] A. Bnx. M. Ikura. L. E. Kay. D. A. Torchia, R. Tschudin. J. Mugn. Reson.
1990. H6. 304- 318.
[ l l ] S. Grxesiek. A. Bax, J. Mugn. Rrson. 1992, 96. 432-440.
[I21 B. T. Farmer. R . A. Venters. L. D. Spicer. M. G. Wittekmd. L. Muller. J
Bionrrd. h'MR 1992, 2. 195- 202.
[I31 A. J. Shaka. C. J. Lee. A. Pines. J. Mugn. ReJon. 1988, 77, 274-293.
(141 S. Sctp. J. Balbach. S. Behrens, H. Kessler, K. Flukiger, R. de Mayer. B. E m ,
B ~ o c h r n n ~ 1994.
! r ~ ~ 33. 7174-7183.
The Arndt -Eistert Reaction in Peptide
Chemistry: A Facile Access to Homopeptides**
Joachim Podlech and Dieter Seebach*
1) NEt3 / CIC0,Et
[*] Prol: Dr. D. Seebach. Dr. J. Podlech
Laboratorium fur Orgdnische Chemie
der Eidgeniissischen Technischen Hochschule
ETH-Zentrum. Universititstrasse 16, CH-8092 Zurich (Switzerland)
Telefax: Int. code + (1)632-1144
We thank the Deutsche Forschungsgemeinschaft (fellowship for J. P.)
Aiig~il'.( % m i . In!. Ed. Engl. 1995. 34. N o . 4
10 % Ag benzoate/ NEt3
Scheme I . Ketene A and diazoketones 1-3 generated as intermediates in the
Arndt-Eistert reaction with amino acid and peptide derivatives. and coupling with
homologdtion of two amino acids or two peptides to homopeptides (4-7).The
formula of one of the synthesized homohexapeptides 7 is given. For R1. R2, R3.and
X see Table 1.
Table 1. Diaroketones 1-3 and homopeptide derivatives 4 7 synthesized according to Scheme 1. Abbreviations: Bz1 = PhCH,, 2 = BzlOCO, Boc = tBuOCO,
iBu = CH,CHMe,. Optical rotations were measured in CHCI, with c c 1.
Cmpd. R'
P-Amino acids, though less abundant than a-amino acids, are
components of natural peptides."' In recent years their importance for the preparation of modified peptidesl2] and p-lactam
antibiotic^'^] has increased enormously. Quite a number of
methods for the synthesis of enantiopure p-amino acids exists
already.[41Still it would be a conceptually new, simple approach
for the construction of peptides containing p-amino acids, if the
generation of the p-amino acid moiety could be combined with
a peptide coupling step. To this end, we have now applied to
peptides the chain extension of carboxylic acids developed by
Arndt and E i ~ t e r t .61~ ~ .
When we started this work, it was not certain whether the
peptidic ketenes A formed in this reaction could be trapped
intermolecularly, and if so, whether they would couple selectively with the free amino function of a second peptide.
At first we simply coupled two amino acids following the
sequence of steps outlined in Scheme 1 . Activation of Z-protected alanine with ethyl chlorocarbonate to give the mixed anhydride and reaction with diazomethane in T H F or Et,O afforded
the corresponding diazoketone 1 (R' = Z, R2 = Me) in 80%
yield. Like all compounds of this type obtained in the course of
3) isolation of diazoketone
- 9.45
this investigation, the diazo compound 1 is a solid and could be
purified by chromatography on silica gel or by recrystallization.
A solution of 1 and 2-4 equivalents of valine benzyl ester['] in
T H F was treated with Et,N and a catalytic amount of silver
benzoate at low temperature, yielding the homodipeptide 4. In
the same manner the protected di- and tripeptides Z-Ala-AIaOH and Boc-Leu-Sar-Leu-OH were coupled via diazoketones 2
and 3, respectively, with the tripeptides H-Ala-Sar-MeLeu-OBzl
and H-Leu-Sar-Leu-OMeI*] (see Table 1). The formation of
both the diazoketones and the homopeptides proceeds with
good to very good yields. The products were characterized by
mass spectrometry, elemental analysis, and by their optical rotation. Owing to the presence of rotamers, the NMR spectra are
not informative, except for those of the protected dipeptides. All
components of the peptide derivatives obtained are enantiopure, as proved by degradation to the free a- and p-amino acids
and analysis by gas chromatography on chiral columns.[91
To test whether the ketene of type A is actually the immediate
precursor to the homopeptide derivatives, we decomposed
(12% AgOCOPh, 6 h) the diazoketone 2 in THF in the absence
of a nucleophile and then added MeOH. The methyl ester 9 was
formed in virtually the same yield as when the reaction was
(Q VCH Verlugsgesellschaji mbH, D-69451 Wivnherm, 1995
0570-0833/9510404-0471 $ fO.OO+ .25/0
47 I
Без категории
Размер файла
404 Кб
experimentov, protons, nmr, optimize, sensitivity, exchanging, modified, triple, rapidly, resonance
Пожаловаться на содержимое документа