A template for stabilizatgion of a peptide α-helix: Synthesis and evaluation of conformational effects by circular dichroism and NMR

Article abstract of DOI:10.1021/ja964231a

The bicyclic diacid 1 was designed as a semi-rigid template for the hydrogen-bonding pattern of a peptide α-helix. The protected precursor 7 was synthesized in eight steps from tert-butyl 3,5-dimethoxybenzoate and linked to L-alanine and L-lactic acid to

Full text of DOI:10.1021/ja964231a

J. Am. Chem. Soc. 1997, 119, 6461-6472
6461
A Template for Stabilization of a Peptide R-Helix: Synthesis
and Evaluation of Conformational Effects by Circular
Dichroism and NMR
Richard E. Austin,† Rachael A. Maplestone,† Andrea M. Sefler,† Kim Liu,†
Witold N. Hruzewicz,† Corey W. Liu,‡ Ho S. Cho,‡ David E. Wemmer,‡ and
Paul A. Bartlett*,†
Contribution from the Department of Chemistry, UniVersity of California,
Berkeley, California 94720-1460, and Structural Biology DiVision,
Lawrence Berkeley National Laboratory, 1 Cyclotron Road, Berkeley, California 94720-5230
ReceiVed December 9, 1996. ReVised Manuscript ReceiVed April 21, 1997X
Abstract: The bicyclic diacid 1 was designed as a semi-rigid template for the hydrogen-bonding pattern of a peptide
R-helix. The protected precursor 7 was synthesized in eight steps from tert-butyl 3,5-dimethoxybenzoate and linked
to L-alanine and L-lactic acid to provide derivatives appropriate for coupling to a peptide. Both the amide 8-N and
the ester 12-O were obtained in each of the four diastereomeric forms. The structure of R,R-8-N was determined
by X-ray crystallography, which facilitated assignment of the diastereomers and confirmed the intended conformational
effects of the quaternary methyl groups. The bicyclic amide and ester derivatives were appended to the peptide
EALAKA-NH2, and their influence on the conformation was evaluated in aqueous solution using circular dichroism
and NMR. The amide analogs have only a slight effect on the appended peptide, whereas the ester-linked template
in S,S-9-O induces 32-50% helical character at 23 °C and 49-77% at 0 °C, depending on the method of determination;
significant helical character persists even at 70 °C. The ability of the template 1 to induce the helical conformation
is related to its structural and electronic complementarity to the N-terminus of the peptide; templating ability disappears
when the carboxylate in S,S-9-O is protonated, and it is not observed in the dimethylamide S,S-9-N-a. The structural
and dynamical properties of conjugate S,S-9-O were studied by NMR and compared with those of the acetylated
heptapeptide 13. The dispersion and temperature dependence of amide hydrogen chemical shifts and the pattern
observed for intra- and inter-residue nuclear Overhauser enhancements are all consistent with a significant population
of helical conformers within the conformational ensemble of conjugate S,S-9-O, in contrast to the unstructured
peptide 13. The generalized order parameter S2 was derived for each residue from the 15N T1 and T2 relaxation rate
1
constants and H-15N heteronuclear NOEs determined for the 15N-labeled derivative of S,S-9-O; these parameters
demonstrate a high degree of conformational rigidity along the peptide chain at 4 °C, with relative motion increasing
for the C-terminal residues. These data are consistent with the chiroptical studies and demonstrate that the template
is exceptionally effective in inducing helical behavior in an appended peptide.
The R-helix, a classic element of protein secondary structure,
represents an unstable conformation of a small peptide in the
general case;1-3 without additional constraints, the modest
advantage of internal hydrogen bonding and side chain-side
chain interactions in the helix is insufficient to overcome the
penalty of conformational restriction.4-6 In peptides that do
show significant helical character, this instability is manifested
at the end of the helix where considerable “fraying” of the
terminal residues occurs.7,8 Just as a shoelace is stabilized
against unraveling by the aglet, so can an R-helix be stabilized
by side chain-side chain cross-links9-13 or by other structural
elements that reduce the conformational freedom of the terminal
hydrogen-bonding groups.14 Synthetic templating moieties have
been reported by Kemp15-17 and by Mu¨ller18 and their co-
workers, and “capping boxes” comprised of normal amino acid
sequences have been elucidated as well.19,20
In this paper, we describe the synthesis and evaluation of
the hexahydroindol-4-one 3,6-diacid 1 and some of its deriva-
tives as helix templates, to be appended to the N-terminus of a
peptide chain (Figure 1b). The carboxylate at C-6, vinylogous
(10) Ghadiri, M. R.; Fernholz, A. K. J. Am. Chem. Soc. 1990, 112, 9633-
9635.
(11) Jackson, D. Y.; King, D. S.; Chmielewski, J.; Singh, S.; Schultz, P.
G. J. Am. Chem. Soc. 1991, 113, 9391-9392.
(12) Chorev, M.; Roubini, E.; McKee, R. L.; Gibbons, M. E.; Caulfield,
M. P.; Rosenblatt, M. Biochemistry 1991, 30, 5968-5974.
(13) Bracken, C.; Gulya´s, J.; Taylor, J. W.; Baum, J. J. Am. Chem. Soc.
1994, 116, 6431-6432.
X Abstract published in AdVance ACS Abstracts, June 15, 1997.
(1) Naider, F.; Goodman, M. In Bioorganic Chemistry; Van Tamelen,
E. E., Ed., Academic Press: New York, 1977; Vol III, p 177.
(2) Goodman, M.; Schmitt, E. E.; Yphantis, D. A. J. Am. Chem. Soc.
1962, 84, 1962.
(14) Schneider, J. P.; Kelly, J. W. Chem. ReV. 1995, 95, 2169-2187.
(15) Kemp, D. S.; Curran, T. P.; Davis, W. M.; Boyd, J. G.; Muendel,
C. J. Org. Chem. 1991, 56, 6672-6682.
(3) Zimm, B. H.; Bragg, J. K. J. Chem. Phys. 1959, 31, 526-535.
(4) Yang, A.-S.; Honig, B. J. Mol. Biol. 1995, 252, 351-365.
(5) Scholtz, J. M.; Baldwin, R. L. Annu. ReV. Biophys. Biomol. Struct.
1992, 21, 95-118.
(16) Kemp, D. S.; Curran, T. P.; Boyd, J. G.; Allen, T. J. J. Org. Chem.
1991, 56, 6683-6697.
(17) Kemp, D. S.; Allen, T. J.; Oslick, S. L. J. Am. Chem. Soc. 1995,
117, 6641-6657.
(6) Creamer, T. P.; Rose, G. D. Proc. Nat. Acad. Sci. U.S.A. 1992, 89,
5937-5941.
(18) Mu¨ller, K.; Obrecht, D.; Knierzinger, A.; Stankovic, C.; Spiegler,
C.; Bannwarth, W.; Trzeciak, A.; Englert, G.; Labhardt, A. M.; Schoen-
holzer, P. Perspect. Med. Chem. 1993, 513-31.
(7) Waltho, J. P.; Feher, V. A.; Merutka, G.; Dyson, H. J.; Wright, P. E.
Biochemistry 1993, 32, 6337-6347.
(8) Young, W. S.; Brooks, C. L., III J. Mol. Biol. 1996, 259, 560-572.
(9) Ruan, F.; Chen, Y.; Hopkins, P. B. J. Am. Chem. Soc. 1990, 112,
9403-9404.
(19) Zhou, H. X.; Lyu, P.; Wemmer, D. E.; Kallenbach, N. R. Proteins
1994, 18, 1-7.
(20) Harper, E.; Rose, G. D. Biochemistry 1993, 32, 7605-7609.
S0002-7863(96)04231-X CCC: $14.00 © 1997 American Chemical Society

6462 J. Am. Chem. Soc., Vol. 119, No. 28, 1997
Austin et al.
Scheme 2
Figure 1. The helix “aglet”: (a) structure of parent compound 1 (S,S-
isomer); (b) model of R-helical conformation of peptide conjugate.
Scheme 1
This active methylene compound was added to methyl â-ni-
troacrylate,26 with the nitro group as the Michael acceptor, to
give adduct 3. Catalytic reduction of the nitro group was
followed by spontaneous cyclization to the hexahydroindolone
4. After protection of the nitrogen (f 5), the quaternary methyl
was introduced by enolate alkylation to give 6.27 Alkaline
hydrolysis of 6 removed the Boc group from the vinylogous
amide as well as the methyl ester, affording the monoacid 7
which was coupled to L-alanine methyl ester with carbodiimide
activation.
No stereochemical control was exerted during the racemic
synthesis of the bicyclic template, but the cis and trans isomers
of diester 6 were separated by MPLC prior to coupling with
L-alanine. The diastereomers of 8-N (Scheme 2) were separated
in turn so that all four stereoisomers were available in pure form.
An X-ray structure of one of these isomers, which proved to be
R,R-8-N (Figure 2), allowed us to assign this and the S,S-
diastereomer. This structure also confirmed that the quaternary
amide, and carbonyl group at C-3 were designed to mimic three
successive peptide carbonyls in the R-helical conformation, and
the negative charge was expected to complement the positive
helix dipole.21-25 The methyl substituents were placed at C-6
and C-3 to favor the pseudoaxial disposition of the carboxylate
moiety and to disfavor the undesired torsional conformer at the
C-3 carbonyl position.
Synthesis
The synthesis of a protected form of 1 is depicted in Scheme
1. Birch-type reductive alkylation of tert-butyl 3,5-dimethoxy-
benzoate was followed by hydrolysis of the enol ethers to give
the functionalized cyclohexane-1,3-dione 2 in 34% overall yield.
(26) McMurry, J. E.; Musser, J. H. Org. Synth. 1977, 56, 65-68.
(27) Direct incorporation of this methyl group at the Michael addition
step, by using methyl â-nitromethacrylate,28 was thwarted by isomerization
of this material under the basic reaction conditions to methyl R-(nitro-
methyl)acrylate, with ensuing addition-elimination to give i.
(21) Wada, A. AdV. Biophys. 1976, 9, 1-63.
(22) Hol, W. G. J.; Van Duijnen, P. T.; Berendsen, H. J. C. Nature
(London) 1978, 273, 443-446.
(23) Åqvist, J.; Kuecke, H.; Quiocho, F. A.; Warshel, A. Proc. Natl.
Acad. Sci. U.S.A. 1991, 88, 2026-2030.
(24) Nicholson, H.; Anderson, D. E.; Dao-pin, S.; Matthews, B. W.
Biochemistry 1991, 30, 9816-9828.
(25) Lockhart, D. J.; Kim, P. S. Science 1993, 260, 198-202.
(28) Bonetti, G. A.; DeSavigny, C. B.; Michalski, C.; Rosenthal, R. J.
Org. Chem. 1968, 33, 237-243.

Template for Stabilization of a Peptide R-Helix
J. Am. Chem. Soc., Vol. 119, No. 28, 1997 6463
Figure 2. Solid state conformation of R,R-6-N.
Figure 3. Potential 310-type hydrogen-bonding pattern for amide-linked
conjugate S,S-9-N.
methyls induce the desired axial orientation of the carboxyl
group at C-6. No structural information was obtained that
allowed us to assign the R,S- and S,R-diastereomers directly.29
30,31
The sequence EALAKA-NH2
was selected as the test
sequence, since it incorporates residues that are predominantly
helix-stabilizing (Ala and Glu-Lys salt bridge); the helix-neutral
leucine was included to provide some differentiation for NMR
analysis, and the charged residues to aid in solubilization as
well. This peptide was synthesized on Rink-amide resin
according to standard protocols, with tert-butyl ester and Boc
protection on the glutamate and lysine residues, respectively,
and then coupled with the diastereomeric alanine adducts 8-N.
Trifluoroacetic acid detached and deprotected the peptide
conjugates 9-N, which were purified by reverse phase HPLC.
Figure 4. Circular dichroism spectra of templated peptides 9-O at
0 °C (1 mM phosphate buffer, pH 5) corrected for the contributions
of the corresponding diastereomers of 10-O: (a) S,S-9-O (63 µM)
and R,R-9-O (64 µM); (b) R,S-9-O (83 µM), S,R-9-O (94 µM),
and acetylated heptapeptide (137 µM) (the asterisk denotes that
the stereochemical assignments of R,S- and S,R-diastereomers are
tentative).
Scheme 3
Characterization by Circular Dichroism
The helical character of the template peptides 9-N in aqueous
solution was evaluated from their CD spectra, correcting for
contributions from the templates themselves by subtracting the
spectra of the alanine-coupled derivatives 10-N. All four
diastereomers show ca. 10-12% helical character at 23 °C; there
is increased helicity at 0 °C, but still no significant difference
between the stereoisomers. The template, in short, does not
work in this context. Modeling studies with S,S-8-N suggested
that an intratemplate hydrogen-bonded conformation (Figure 3)
is readily accessible energetically; this conformation corresponds
to the hydrogen-bonding pattern of a 310 helix, rather than the
desired R-pattern, and could be responsible for the ineffective-
ness of the template in stabilizing the helix. The ester-linked
derivatives 9-O were therefore synthesized to circumvent this
possibility.
Because of the lower nucleophilicity of the lactate hydroxyl
group and the more vigorous conditions required for the
coupling reaction, the vinylogous amide nitrogen of 7 had to
be protected as the Boc derivative, 11, prior to esterification
(Scheme 3). The carboxyl group was activated with trichlo-
robenzoyl chloride and the coupling with methyl L-lactate was
catalyzed by 4-(dimethylamino)pyridine (DMAP) to give 12-
O. These diastereomers were separated by MPLC and assigned
by correlation with the amides 8-N by alkaline hydrolysis to 7
and coupling to L-alanine methyl ester. The lactate methyl esters
of 12-O were carefully saponified, the templates were coupled
to the hexapeptide, and the oxygen-linked conjugates 9-O were
isolated and purified as described above for the amides 9-N.
(29) The tentative basis for assignment of R,S- and S,R-diastereomers
was the slightly greater ellipticity observed when the former was conjugated
to a longer peptide as {template}-AEAAAKEAAAKY-amide (data not
shown).
(30) Marqusee, S.; Baldwin, R. L. Proc. Natl. Acad. Sci. U.S.A. 1987,
84, 8898-8902.
(31) Shoemaker, K. R.; Kim, P. S.; York, E. J.; Stewart, J. M.; Baldwin,
R. L. Nature 1987, 326, 563-567.

6464 J. Am. Chem. Soc., Vol. 119, No. 28, 1997
Austin et al.
Table 1. Molar Ellipticities at 222 nm and Percent Helicity for
Template Peptides 9-N and 9-O
[
]
helicity
(%)d
[
]
helicity
(%)d
222
222
peptide
(23 °C)c
(0 °C)c
S,S-9-Na
R,R-9-Na
S,S-9-Ob
-3290
-2913
-8109
-1450
13
12
32
6
-6200
-3100
-11143
-2
25
12
49
0
R,R-9-Ob
Ac-AEAKALA-NH2 (13)
-1966
8
a Measured in 0.1 M KCl, 50 mM potassium phosphate buffer, pH
5.2. b Measured in 1.05 mM potassium phosphate buffer, pH 5. c In
(deg cm2)/dmol, corrected for contribution from corresponding template-
Ala or template-Lac derivative, 10-N or 10-O. d Calculated from ratio
[
]
222/[
]
max, where [
]
) -39 500 × [1 - (2.57/n)] ()25 000
max
for n ) 7).37,38,76
Replacement of the amide NH with the ester oxygen has a
dramatic effect on the helical character of the conjugates (Figure
4). In aqueous buffer at pH 5.2, the S,S-conjugate of 9-O is
32% helical at 23 °C, increasing to 50% at 0 °C, according to
the molar ellipticity at 222 nm (Table 1). Moreover, there are
significant differences between the diastereomers, since the CD
spectrum of the conjugate with the enantiomeric template, R,R-
9-O, is that of a random coil. The S,R- and R,S-diastereomers
also show no significant helical character, nor does the N-acetyl
analog. The ellipticity measured for S,S-9-O at 222 nm is
directly proportional to concentration over the range 8.6-272
µM, which suggests that the observed helicity does not result
from aggregation. Moreover, such a large distinction between
the diastereomers and the ester- and amide-linked analogs would
not be expected if the helicity observed in S,S-9-O were due to
aggregation.
The form of the CD curve for S,S-9-O below 250 nm is that
expected for an R-helix and quite distinct from the 310-pattern
recently described.32 The helical conjugate S,S-9-O shows
positive ellipticity from 270 to 290 nm. This band disappears
when the helical character is lost (e.g., on titration; see Figure
6), and it is not observed for the diastereomeric conjugate R,R-
9-O, even when a helix is induced in the latter by trifluoro-
ethanol (see Figure 5b). The vinylogous amide of the bicyclic
cap (λmax ) 305 nm) is presumably responsible for this band,
which reflects the different asymmetric environment of the
chromophore in the templated helix than in the control compund
10-O.33
Figure 5. Comparison of CD spectra of (a) S,S-9-O and (b) R,R-9-O
in aqueous buffer and 50:50 buffer/2,2,2-trifluoroethanol (0 °C).
character in 50% TFE/buffer, as reflected in their molar
ellipticities at 222 nm, indicating that the peptides behave
normally and adopt the R-helical conformation under these
conditions. TFE titration indicates that the helicity observed
in 50% TFE/buffer is the maximum that these conjugates can
attain; the molar ellipticity determined under these conditions
thus provides an alternative benchmark for estimating helical
character under different conditions. In fact, the ellipticity
exhibited by S,S-9-O in aqueous buffer at 0 °C is 77% of
the limiting value observed in 50% TFE, which is consider-
ably higher than the value calculated from the traditional
formula.37,38
The divergent behavior of S,S-9-O and the R,R-stereoisomer
supports a clear role for the template in inducing helical behavior
in aqueous solution. However, the protonation state of the
template carboxylate is also a critical feature: comparison of
the CD spectra of S,S-9-O at different pH values shows that
the templating effect is lost when the carboxylate is protonated
(Figure 6). Spectra taken as a function of pH show two
isodichroic points, consistent with a simple two-state equilibrium
between helix and random coil. Protonation of the glutamate
side chain at lower pH and loss of any salt-bridging interaction
with lysine may also contribute to the decrease in helical
character as well.
The degree of helical character observed for this conjugate
is unprecedented for such a short peptide in aqueous solution:
the seven residues (if lactate is counted) constitute less than
two turns of an R-helix. Indeed, it is not clear if the equation
for helicity is appropriate in the present case (Table 1), since it
is parameterized from considerably longer peptides. An alterna-
tive estimation of the percent helicity can be obtained by
comparison of the conjugates under conditions in which
maximal helicity can be expected, e.g., in the presence of
trifluoroethanol (TFE) (Figure 5).34-36 Both the S,S-9-O and
R,R-9-O conjugates show enhanced, and comparable, helical
(32) Toniolo, C.; Polese, A.; Formaggio, F.; Crisma, M.; Kamphuis, J.
J. Am. Chem. Soc. 1996, 118, 2744-2745.
(33) The ellipticity measured at 222 nm for the helical conjugates arises
from the peptide only, since the template ester does not absorb significantly
in the region 200-270 nm (see Figure S4 in the Supporting Information).
The negative ellipticity at 275 nm observed in the peptide conjugate of one
of the diastereomeric templates, tentatively assigned as R,S-9-O (Figure
4b), arises from subtraction of a positive peak in the spectrum of R,S-10-
O.
As reported before for other templated helical conjugates,11,17
there is no dramatic temperature dependence of the molar
ellipticity (Figure 7). In particular, there is no inflection
representative of a cooperative melting process, and significant
helicity persists up to 70 °C.
(34) Storrs, R. W.; Truckses, D.; Wemmer, D. E. Biopolymers 1992,
32, 1695-1702.
(35) Jasanoff, A.; Fersht, A. R. Biochemistry 1994, 33, 2129-2135.
(36) Cammers-Goodwin, A.; Allen, T. J.; Oslick, S. L.; McClure, K. F.;
Lee, J. H.; Kemp, D. S. J. Am. Chem. Soc. 1996, 118, 3082-3090.
(37) Chen, Y. H.; Yang, J. T.; Chau, K. H. Biochemistry 1974, 13, 3350-
3359.
(38) Engel, M.; Williams, R. W.; Erickson, B. W. Biochemistry 1991,
30, 3161-3169.

Template for Stabilization of a Peptide R-Helix
J. Am. Chem. Soc., Vol. 119, No. 28, 1997 6465
Figure 8. Proposed templating conformation of dimethylamide
conjugate S,S-9-N-a.
response to TFE (data not shown). This result, as well as the
pH dependence for conjugate S,S-9-O (Figure 6), underscores
the importance of a negatively charged carboxylate at the C-6
position. This charge presumably helps to complement the
macrodipole of the helix, as generally observed for negatively
charged amino acids at the N-terminus.39
Figure 6. Effect of pH on CD spectrum of S,S-9-O in aqueous solution
(0 °C). Shown are spectra obtained at pH 5, 4, 3, 2, and 1; spectra at
pH 8, 7, 8, and 9 were superimposable with that at pH 5.
The absolute strength of the hydrogen bonds formed by the
various templates is not, to a first approximation, the critical
factor in determining the position of equilibrium between the
helical and random coil forms of the peptide conjugates. As
the NH-to-carbonyl bond strength decreases from S,S-9-N to
S,S-9-O in the templated conformation, so does the strength of
the hydrogen bond(s) between that oxygen and water in the
solvated form (eq 1). For this reason, we interpret the
ineffectiveness of the amide S,S-9-N as arising from the
hydrogen-bonding pattern of Figure 3 and that of amide S,S-
9-N-a from the lack of charge complementarity.
(1)
Figure 7. Effect of temperature on CD spectrum of S,S-9-O in aqueous
solution at 0, 30, 50, and 70 °C.
The success of the ester-linked template and failure of the
amide-linked design lend credence to the rationale behind this
alteration, namely, prevention of the hydrogen-bonding pattern
illustrated in Figure 3. We envisaged an alternative strategy
for preventing this interaction sterically, in the form of the
neutral dimethylamide 9-N-a. This template would be able to
Characterization by NMR
Although CD gives quantitative information on conforma-
tional equilibria, it is time-averaged and global in nature. NMR,
in contrast, provides structural information at the single-residue
level as well as insight into the dynamical behavior of the
oligopeptide chain. The nuclear Overhauser effect (NOE)
provides the most important parameter for elucidating the
conformation of a peptide since it has a strong, 1/r6 dependence
on the internuclear distance. The observation of a direct NOE
between a pair of protons is evidence for a significant population
of conformers in which the protons are close, typically within
3.5 Å for flexible peptides in solution.40 The distance between
the CRH of residue i and the NH of the following residue (i+1),
represented by dRN(i, i+1), is a short 2.2 Å in the extended
conformation of a peptide, and this contact can be detected even
if the population of this conformer is relatively small. In
contrast, the threshold population for observation of the medium-
range dRN(i, i+3) NOEs characteristic of helical conformations
is relatively high, since the distance is 3.4 Å.41 Such medium
range NOEs are typically weak and high signal-to-noise NOESY
or ROESY spectra are required for their detection.42
enter into the desired hydrogen-bonding interaction, with the
dimethylamino group bisecting the substituents at C-6; however,
the undesired, transannular interaction would force one of the
amide methyls to eclipse the C-6 methyl group (Figure 8).
Although it is the most straightforward route to this analog,
direct formation of the dimethylamide at C-6, after selective
cleavage of the tert-butyl ester in 6, proved difficult to effect
in good yield. The S,S- and R,R-diastereomers of the L-alanine
amide 8-N-a were therefore synthesized from N,N-dimethyl-
3,5-dimethoxybenzamide by a route analogous to that shown
in Scheme 2.
Certain inter-residue NOE patterns are characteristic of
specific secondary structures, since these structural elements
have regular geometries and defined hydrogen-hydrogen
(39) Doig, A. J.; Baldwin, R. L. Protein Sci. 1995, 4, 1325-1336.
(40) Wu¨thrich, K. NMR of Proteins and Nucleic Acids; John Wiley &
Sons: New York, 1986.
(41) Waltho, J. P.; Feher, V. A.; Lerner, R. A.; Wright, P. E. FEBS Lett.
1989, 250, 400-404.
The dimethylamide derivative was not effective as a helix
template. The CD spectra indicate that both the S,S- and R,R-
diastereomers of 9-N-a are predominantly random coil in
aqueous solution, although they are otherwise normal in their

6466 J. Am. Chem. Soc., Vol. 119, No. 28, 1997
Austin et al.
distances. For example, the geometry of the R-helix is
characterized by short distances between residues i and i+3 and
i and i+4; in the more tightly wound 310-helix, short distances
occur between residues i and i+2 and i and i+3; and in tight
turns, close contacts also occur between residues i and i+2.
â-Sheet structures, in contrast, consist of extended polypeptide
segments in which medium-range distances are absent.44
Short dRN(i, i+1) distances and thus strong dRN(i, i+1) NOE
connectivities indicate backbone dihedral angles in the â
(extended chain) region of (φ, ψ) space; these dRN(i, i+1) NOEs
normally dominate NOESY spectra of unfolded (“random coil”)
peptides in aqueous solution. In contrast, helical residues are
connected throughout by short dNN(i, i+1) distances; hence, the
observation of dNN(i, i+1) NOE connectivities is evidence that
the conformational ensemble of the peptide includes R-helical
or turnlike structures. The simultaneous observation of extended
regions of dRN(i, i+1) and dNN(i, i+1) connectivities for an
oligopeptide indicates conformational averaging between the R-
and â-regions of (φ, ψ) space. Adjacent (i to i+1) dRN and
dNN NOE connectivities provide information only on individual
backbone dihedral angles and do not identify extended regions
of helical or â-strand structure.43 However, a sequence of
consecutive NOEs is more robust evidence for a particular
secondary structure; for example, sequences of 3-5 successive
distance constraints dNN < 3.6 Å enable helical segments to be
identified with high reliability,44 especially when some medium-
range NOEs can also be identified in this segment.
Assignment of Resonances. The one dimensional (1D) 1H
NMR spectra of the N-terminally capped peptide S,S-9-O and
the acetylated comparison peptide (Ac-AEALAKA-NH2, 13)
were recorded at 4 °C in 50 mM phosphate, 100 mM KCl buffer
in 10% D2O/H2O at pH 4.2, to permit the observation of
exchangeable protons. A jump-and-return selective excitation
sequence45 was used to avoid saturation transfer from water to
the exchangeable protons. The 1H NMR spectrum of S,S-9-O
is characterized by narrow line widths, good signal-to-noise
ratio, and a large chemical shift dispersion within the envelope
of the amide protons (Figure 9a), which is indicative of a
significant population of stable, compact conformations. On
the other hand, the chemical shift dispersion within the envelope
of the CRH protons is poor, as typically seen in small helical
proteins rather than in regions of â-structure or in random coil
peptides.46 The acetylated analog 13 shows a much narrower
chemical shift range for the amide protons, as expected for a
more flexible, unstructured peptide.
RH intraresidue connectivities were identified in spectra acquired
in 90% H2O/10% D2O by direct connectivity or by TOCSY
relay to resonances farther out the side chain. In this manner,
the parent spin systems of all the amide protons were identified
(Figure 9b).47
The sequential assignment of resonances for S,S-9-O was
obtained from NOESY and TOCSY spectra recorded under the
standard conditions (pH 4.2, 4 °C in 90% H2O/10% D2O). In
spite of the low chemical shift dispersion in the CRH proton
envelope of the reference 1D spectrum, intraresidue NH-RH
connectivities in the TOCSY spectrum gave rise to a series of
nonoverlapping cross peaks (see Figure 9b). Although the
signal-to-noise ratio of the NOESY spectrum used for the
sequential assignment was lower than that of the TOCSY
spectrum, a complete set of sequential dNN(i, i+1) connectivities
could nonetheless be obtained (Figure 11). However, the jump-
and-return read sequence in the former pulse program gave rise
to less-efficient suppression of the residual water signal, and as
a result, the NH-RH connectivities could not be observed in
the f2 dimension. For conjugate S,S-9-O, dNN(i, i+1) connec-
tivities were used for the assignment because poor digital
resolution in the f1 dimension (256 t1 increments), as well as
the low chemical shift dispersion for the R-hydrogens, made it
difficult to identify dRN(i, i+1) connectivities. In contrast, the
inter-residue NOEs observed for the acetylated peptide 13 were
dRN(i, i+1) rather than dNN(i,i+1) connectivities (data not
shown), as expected for an unstructured peptide. All chemical
shifts assigned for the peptides and the bicyclic cap of S,S-9-O
and 13 are listed in Table 2.
The observation of a series of dNN(i, i+1) connectivities
spanning the entire six-residue sequence in the NOESY spectra
of S,S-9-O (Figure 11) suggests that helical conformers
dominate the population for this conjugate; the distribution of
dâN(i, i+1) NOEs is also consistent with this interpretation.44
However, the relative intensities of dRN(i, i+1) and dNN(i, i+1)
NOEs indicate that extended chain conformers are also part of
the ensemble. As unambiguous evidence for the helical
conformer, we sought medium-range NOEs, e.g., dRN(i, i+3),
dRâ(i, i+3), dRâ(i, i+4) etc. The dRN NOE region of the NOESY
spectrum of S,S-9-O is depicted in Figure 12. The severe
resonance overlap of CHâ protons in the region δ ) 1.44-1.46
ppm and CHR protons in the region δ ) 4.15-4.22 ppm made
it difficult to distinguish medium-range and side chain-side
chain NOEs from intraresidue connectivities. As a result, only
one of four possible dRN(i, i+3) NOEs was observed (Ala-7/
Leu-4 NH-RH).48
We assigned the 1D 1H NMR spectra in the sequence specific
manner devised by Wu¨thrich,40 tracing a continuous pathway
along the peptide backbone using alternating COSY connec-
tivities within a residue (NH-RH) and sequential NOESY
connectivities, dRN(i, i+1) and dNN(i, i+1), between adjacent
residues, as illustrated in Figure 10.
Amide Proton Temperature Coefficients. The magnitude
of the temperature dependence of an amide proton chemical
shift is generally accepted as a measure of the degree to which
it is engaged in an intramolecular hydrogen bond;49,50 smaller
(47) In routine 1D 1H NMR spectra that were recorded, several low-
intensity signals were observed within the amide envelope (see Figure 9a);
TOCSY spectra confirmed the peptidic nature of these resonances, as spin
systems corresponding to alanine and glutamate were identified. Since
neither the broadening of these resonances over the range of 277-323 K
nor the chemical exchange of the cross peaks was observed between the
two sets of amide proton resonances in the NOESY spectrum, we concluded
that a small amount of uncapped peptide as a contaminant, rather than a
minor conformer of S,S-9-O, was responsible. In any event, the resonances
from the impurity did not interfere with those of the conjugate S,S-9-O or
complicate their interpretation.
(48) The molecular mass (907 Da) of the capped peptide is such that
the molecular correlation time may fall near the zero NOE region (ωoτc )
1). At 4 °C, tumbling is slowed to the extent that the peptide falls into the
negative NOE regime, as evidenced by cross peaks of the same sign as the
diagonal in the 2D NOESY spectra recorded at this temperature. 2D ROESY
spectra were acquired under spin-locked conditions to overcome this
limitation; however, additional medium-range NOEs were not apparent in
spectra recorded with mixing times (Tm) of 250 and 400 ms.
The side chain spin systems for leucine, lysine, and glutamate
were assigned from their scalar coupling networks and the
relayed connectivities observed in TOCSY spectra. The NH-
(42) While NOEs are sensitive indicators of folded structures within a
conformational ensemble, quantitative estimation of a population size from
relative NOE build up rates is complicated by conformational averaging
and the 1/r6 dependence of the NOE.43 Consequently, we have used the
2D spectra qualitatively, simply to identify conformations whose populations
in aqueous solution exceed the threshold value for observation of diagnostic
NOEs.
(43) Dyson, H. J.; Wright, P. E. Ann. ReV. Biophys. Chem. 1991, 20,
519-538.
(44) Wu¨thrich, K.; Billeter, M.; Braun, W. J. Mol. Biol. 1984, 180, 715-
740.
(45) Plateau, P.; Gueron, M. J. Am. Chem. Soc. 1982, 104, 7310-7311.
(46) Wishart, D. S.; Sykes, B. D.; Richards, F. M. J. Mol. Biol. 1991,
222, 311-334.

Template for Stabilization of a Peptide R-Helix
J. Am. Chem. Soc., Vol. 119, No. 28, 1997 6467
Figure 9. (a) Amide NH region of 1D spectrum of conjugate S,S-9-O (3.8 mM, 90% H2O/10% D2O, pH 4.2, 4 °C) (the asterisk denotes cross
peaks from uncapped peptide); (b) TOCSY spectrum showing NH to side chain cross peaks.
peptide conjugate S,S-9-O and the comparison analog 13 were
determined from spectra acquired in 90% H2O/10% D2O over
the temperature range of 4-62 °C (Figure 13 and Table 3);
solvent suppression was achieved with a jump-and-return read
sequence to ensure that the NH resonances were not attenuated
by saturation transfer from water.
The values calculated for the temperature coefficients rep-
resent averages over the conformational ensemble of the peptide
and conjugate; as a result, intermediate values are expected if
both helical and extended conformations are represented sig-
nificantly. The low temperature coefficients observed for all
of the residues of S,S-9-O except Ala-3 are consistent with a
Figure 10. Intra- and inter-residue connectivities employed in the
sequential assignment of resonances.
dominant conformation in which the amide protons are part of
values ( δ < 3-4 ppb/K) are observed if the NH is involved
in an intrachain hydrogen bond as part of a defined secondary
structure than if it is exposed to solvent water ( δ ≈ 8 ppb/K).
The temperature coefficients for the amide hydrogens in the
hydrogen-bonding networks of a defined secondary structure.
The higher temperature dependence of the Ala-3 NH chemical
shift is anomalous in this series, with a value that corresponds
to a solvated amide. It is unclear whether the high value arises
from dislocation of this amide from the anticipated helical
structure or from the vinylogous amide carbonyl as an atypical
hydrogen-bonding partner.
(49) Kessler, H. Angew. Chem., Int. Ed. Engl. 1982, 21, 512-523.
(50) Dyson, H. J.; Cross, K. J.; Houghten, R. A.; Wilson, I. A.; Lerner,
R. A. Nature 1985, 318, 480.

6468 J. Am. Chem. Soc., Vol. 119, No. 28, 1997
Austin et al.
a set of six relaxation measurements.54,55 However, a simplified,
model-free formalism introduced by Lipari et al.56,57 allows
analysis of backbone dynamics from only three relaxation
measurements (T1, T2, and NOE), which are represented by eqs
2-4.58
1
T1
) d2[J(ωH - ωN) + 3J(ωN) + 6J(ωH + ωN)] + c2J(ωN)
(2)
1
T2
) 0.5d2[4J(0) + J(ωH - ωN) + 3J(ωN) + 6J(ωH +
c2[3J(ωN) + 4J(0)]
ωN)] +
(3)
6
γH
1γN
NOE ) 1 + T
d2[6J(ωH + ωN) - J(ωH - ωN)] (4)
where
γ2Hγ2N(h/2π)2
2
15
d2 )
and c2 ) ω2N(σ| - σ )2
r3HN
Figure 11. Jump-and-return NOESY spectrum of S,S-9-O illustrating
d
NN(i, i+1) sequential NOEs (Tm ) 800 ms, pH 4.2, 4 °C).
p is Planck’s constant, γH and γN are the 1H and 15N
3
3JHNR Coupling Constants. The magnitude of the JHNR
1
gyromagnetic ratios, ωH and ωN are the H and 15N Larmor
coupling constant observed in a polypeptide chain is dependent
on the φ-angle and therefore on the local conformation of the
polypeptide backbone (Figure 14). While JHNR values for
â-sheet conformations fall in the range from 8 to 10 Hz (-150°
< φ < -90°), helices are characterized by 3JHNR values between
4 and 5 Hz (-70° < φ < -30°).51 A series of three or more
coupling constants below 6 Hz has been taken as diagnostic of
helical structure.43
frequencies, rHN is the 1H-15N internuclear distance (1.02 Å53),
and σ| and σ are the parallel and perpendicular components of
the assumed axially symmetric 15N shift tensor (σ| - σ ) )
-160 ppm.59
In the model-free formalism, these three relaxation parameters
provide a measure of dynamic behavior that is specific to each
residue. The approach is based on a simplified spectral density
function (eq 5), with the assumption that the molecule undergoes
isotropic tumbling.
3
3
The JHNR values for conjugate S,S-9-O could be measured
directly from the 1D spectrum over the temperature range of
4-50 °C because of the excellent chemical shift resolution
within the amide envelope. The coupling constants determined
at 4 and 25 °C are presented with the corresponding φ torsion
angles in Table 4.
For all residues, the coupling constants measured at 4 °C
average 4.1 Hz, which is close to the value of 3.9 Hz predicted
for the ideal helix geometry. Larger values are observed at 25
°C, consistent with an increase in the proportion of extended
and random coil conformers relative to helix as thermal
unfolding occurs. Nevertheless, the coupling constants remain
less than 7 Hz, even up to 42 °C, suggesting that a significant
population of helical conformers persists. The absence of a
defined unfolding or melting temperature was observed in the
CD spectrum of this conjugate and has been described previ-
ously for other templated peptides.11,17
Dynamical Behavior of Peptide Conjugate S,S-9-O. The
ability of the cap to prevent N-terminal fraying of the R-helix
in S,S-9-O can be assessed by analyzing the dynamical behavior
of the peptide backbone, which is revealed by the relaxation
properties of the amide bond vector (1H-15N). Of particular
use in interpreting protein backbone dynamics are the NOE and
T1 and T2 relaxation rates,52,53 which are dependent on the
spectral density function at five characteristic frequencies (J(0),
J(ωN), J(ωH), J(ωH - ωN), and J(ωH + ωN)). Sampling the
spectral density function fully at these five frequencies requires
S 2τm
(1 - S2)τ
2
5
1
τ
1
1
J(ω) )
+
, where
)
+
2
2
[
]
τm τe
(5)
1 + (ωτm)
1 + (ωτ)
The effective correlation time, τe, represents the reorientation
of the N-H bond vector with respect to the overall tumbling
time of the molecule itself, τm. S2, the generalized order
parameter, is a measure of the extent to which internal motion
is coupled to that of the molecule as a whole; S2 can range
from 0, for independent isotropic motion, to 1, for completely
restricted motion. In proteins, these values generally extend
from ca. 0.5 for disordered loop regions to ca. 0.85 for well-
defined secondary structures.53 These parameters are obtained
from a fit of the spectral density function (eq 5) with the
relationships describing the experimentally determined relaxation
rates (eqs 2-4).60,61
The data acquired at 4 °C on a sample of S,S-9-O uniformly
15N-labeled from Glu-2 to Ala-7 indicate limited backbone
motion for residues 3-5, with increasing motion for Lys-6 and
(54) Peng, J. W.; Wagner, G. J. Magn. Reson. 1992, 98, 308-332.
(55) Peng, J. W.; Wagner, G. Biochemistry 1992, 31, 8571-8586.
(56) Lipari, G.; Szabo, A. J. Am. Chem. Soc. 1982, 104, 4559-4570.
(57) Lipari, G.; Szabo, A. J. Am. Chem. Soc. 1982, 104, 4546-4559.
(58) Abragham, A. The Principles of Nuclear Magnetism; Claredon
Press: Oxford, England, 1961.
(59) Hiyama, Y.; Niu, C.; Silverton, J. V.; Bavaso, A.; Torchia, D. A. J.
Am. Chem. Soc. 1988, 110, 2378-2383.
(51) Pardi, A.; Billeter, M.; Wu¨thrich, K. J. Mol. Biol. 1984, 180, 741.
(52) Kay, L. E.; Torchia, D. A.; Bax, A. Biochemistry 1989, 28, 8972-
8979.
(53) Clore, G. M.; Driscoll, P. C.; Wingfield, P. T.; Gronenborn, A. M.
Biochemistry 1990, 29, 7387-7401.
(60) Palmer, A. G. I.; Rance, M.; Wright, P. E. J. Am. Chem. Soc. 1991,
113, 4371-4380.
(61) Orekhov, V. Y.; Nolde, D. E.; Golovanov, A. P.; Korzhnev, D. M.;
Arseniev, A. S. In Shemyakin and OVchinnikoV; Institute of Bio-Organic
Chemistry, Russian Academy of Sciences: Moscow, 1995.

Template for Stabilization of a Peptide R-Helix
J. Am. Chem. Soc., Vol. 119, No. 28, 1997 6469
Table 2. Assignment of Template Peptide Conjugate S,S-9-O (3.5 mM) and, in Parentheses, Acetylated Peptide 13 (5.8 mM)a
residue
NH
RHb
â,â′H
γ,γ′H
δ,δ′H
ꢀ,ꢀ′H
Lac-1/(Ala-1)
Glu-2
Ala-3
Leu-4
Ala-5
Lys-6
Ala-7
n/a (8.48)
4.810 (4.167)
4.160 (4.195)
4.150 (4.222)
4.207 (4.252)
4.195 (4.219)
4.185 (4.233)
4.215 (4.227)
1.46 (1.36)
2.16 (1.96, 2.03)
1.48 (1.37)
1.71 (1.65)
1.45 (1.39)
1.84 (1.77, 1.85)
1.44 (1.41)
7.76 (8.69)
8.54 (8.31)
7.68 (8.12)
7.88 (8.22)
8.02 (8.25)
7.91 (8.26)
2.46, 2.50 (2.30)
1.60 (1.58)
0.87 (0.86, 0.91)
1.46, 1.54 (1.45)
1.68 (1.67)
2.98 (2.99)
C-1,1′
C-5,5′
C-7,7′
C-2-Me
1.33
C-6-Me
1.44
2.43, 2.64
3.59, 4.15
2.61, 3.00
a In 90% H2O/10% D2O, pH 4.2, at 4 °C; chemical shifts referenced to internal dioxane (δ ) 3.74 ppm). b Recorded to three decimal places
because of low dispersion in RH chemical shifts.
Figure 12. Jump-and-return NOESY spectrum of S,S-9-O illustrating
dRN connectivities (Tm ) 800 ms, pH 4.2, 4 °C).
Figure 14. Relationship between secondary structure and φ torsion
angle.40,51
3
Table 4. JRN Coupling Constants (Hertz) and Corresponding
Average φ Torsional Angle, Determined from the Amide Proton
Resonances for Conjugate S,S-9-O
Glu-2
Ala-3
Leu-4
Ala-5
Lys-6
Ala-7
4 °C
φb
25 °C
φb
4.1
-59
5.1
a
4.1
-59
5.8
3.6
-55
5.0
5.2
-68
6.2
3.7
-56
5.3
3.2
-51
-67
-73
-67
-76
-69
a Coupling constant not resolved. b The corresponding φ-angles are
3
determined from the equation: JRN ) 6.4 cos2 θ - 1.4 cos θ + 1.9 (θ
) φ - 60).50
The low S2 value for Glu-2 may reflect its proximity to the
template-lactate ester linkage; the ester is not as strongly
hydrogen bonded as the peptide amides and may undergo a
crankshaft-like rotation that increases the mobility of the
neighboring NH bond vector.
Conclusions
Figure 13. Temperature dependence of the chemical shifts of the amide
protons of S,S-9-O (3.8 mM, pH 4.2).
The bicyclic template S,S-9-O, when linked to a peptide
through an ester linkage, is unusually effective in inducing the
R-helical conformation, with some 50-75% helicity observed
in the conjugate with lactyl-Glu-Ala-Leu-Ala-Lys-Ala-amide,
as determined by circular dichroism spectroscopy. The tem-
plating ability is directly related to the stereochemistry of the
template, and it is characterized by a relative insensitivity to
temperature and a strong dependence on the ionization state of
the terminal carboxyl group. It is apparent that the template
must complement the N-terminus of the R-helix both sterically
as well as electronically to be effective. By the global, time-
averaged view afforded by CD spectroscopy, the template
induces helical character in a short peptide to an unprecedented
extent.
Table 3. Temperature Dependence of Amide Proton Chemical
Shifts ( δ, ppb/K) for Conjugate S,S-9-O and Acetylated Peptide
13a
Ala-1 Glu-2 Ala-3 Leu-4 Ala-5 Lys-6 Ala-7
S,S-9-O
13
>-1.2b -8.5 -2.4 -2.7 -3.9 >-1.4b
-6.5 -6.2 -6.2 -6.3 -7.0 -6.2
-8.0
a Coefficients were determined over the temperature range of 4-52
°C. A negative sign indicates a shift to lower frequency with increasing
temperature. b Nonlinear relationship observed for δ vs 1/T for this
residue (see Figure 13).
Ala-7 and, interestingly, Glu-2 (Figure 15). The S2 values
suggest the presence of a well-defined secondary structure in
the center of the peptide, with the motion at Ala-7 approaching
that of a dynamic loop. This behavior is consistent with S2
profiles typically observed for R-helices, which show lower
order parameters for the ends of the helix where fraying occurs.62
The NMR investigations provide insight into the effects of
the helix template on the local conformation and dynamical
(62) Redfield, C.; Boyd, J.; Smith, L. J.; Smith, R. A. G.; Dobson, C.
M. Biochemistry 1992, 31, 10431-10437.

6470 J. Am. Chem. Soc., Vol. 119, No. 28, 1997
Austin et al.
Cl2/methanol) to give 3.88 g (86%) of tert-butyl 1-methyl-3,5-
dioxocyclohexanecarboxylate (2) as a light brown solid: IR (CHCl3)
2970, 1720, 1590 cm-1; 1H NMR δ 3.35 (dt, 1, J ) 18.0, 1.7), 3.28 (d,
1, J ) 18.0), 2.83 (dd, 2, J ) 1.7, 15.6), 2.51 (dd, 2, J ) 1.0, 15.6),
1.37 (s, 9), 1.31 (s, 3); 13C NMR δ 201.7, 172.9, 83.6, 56.1, 50.0, 41.8,
27.6, 24.2; MS (EI) m/z 226 (3.38), 211 (6.21), 202 (8.33), 186 (12.19),
170 (33.26), 153 (34.24), 125 (58.52), 57 (100), 56 (99.38).
A solution of methyl â-nitroacrylate26,64 (2.03 g, 15.5 mmol) as a
solution in methanol (62 mL) was added Via cannula to a solution of
the diketone 2 (3.50 g, 15.5 mmol) and triethylamine (1.56 g, 2.15
mL, 15.5 mmol) in methanol (85 mL) at 0 °C. After 2 h, the cooling
bath was removed and the solution was stirred at room temperature
for 24 h. The mixture was concentrated under reduced pressure, and
the residue was dissolved in CH2Cl2 (100 mL) and washed successively
with 0.5 N HCl (25 mL) and brine (25 mL). The organic phase was
dried (MgSO4) and concentrated in Vacuo to give the Michael adduct
3 (5.46 g, 15.2 mmol, 99% yield): IR (film) 2980, 1730, 1550, 1380
cm-1; 1H NMR δ 4.99 (dd, 1, J ) 8.3, 13.8), 4.71 (dd, 1, J ) 6.1, 8.3),
4.29 (dd, 1, J ) 6.0, 13.8), 3.66 (s, 3), 2.95 (d, 1, J ) 17.0), 2.94 (d,
1, J ) 17.0), 2.44 (d, 1, J ) 17.0), 2.42 (d, 1, J ) 17.0), 1.40 (s, 9),
1.31 (s, 3); 13C NMR δ 173.9, 171.7, 109.6, 81.9, 73.4, 53.4, 52.4,
43.6, 42.1, 42.0, 37.1, 27.6, 27.5, 24.5. Anal. Calcd for C16H23NO8:
C, 53.78; H, 6.49; N, 3.92. Found: C, 53.50; H, 6.60; N, 3.59.
Methyl 6-tert-Butoxycarbonyl-6-methyl-2,3,4,5,6,7-hexahydroin-
dol-4-one-3-carboxylate (4). A slurry of the Michael adduct 3 (2.67
g, 7.48 mmol) and Raney Nickel (2.24 g, washed with water until the
filtrate was neutral to pH paper, followed by washing with methanol)
in methanol (25 mL) was placed under a 50 psi atmosphere of H2 and
shaken for 65 h. The mixture was filtered through Celite, the filtrate
was evaporated, and the residue was chromatographed (95:5, CH2Cl2/
methanol) to give the vinylogous amide 4 (1.94 g, 6.28 mmol, 84%
yield) as a mixture of diastereomers: IR (film) 3200, 2985, 1730, 1575,
1
1160 cm-1; H NMR δ 3.71-3.83 (m, 3), 3.59 (s, 3), 2.80 (d, 1, J )
16.7), 2.55 (d, 1, J ) 16.1), 2.27 (d, 1, J ) 16.7), 2.19 (d, 1, J )
16.2), 1.29 (s, 9), 1.20 (s, 3); 13C NMR δ 187.6, 174.5, 174.2, 169.3,
106.4, 80.8, 51.9, 50.8, 45.6, 45.3, 42.9, 33.0, 27.6, 24.0; MS (EI) m/z
309 (24.84), 250 (46.99), 236 (33.26), 208 (56.36), 194 (78.11), 148
(74.71), 120 (60.55), 106 (42.79), 80 (85.98), 57 (97.90), 41 (100);
HRMS (EI) calcd for C16H23NO5 309.1577; found 309.1587.
Methyl 1,6-Di-tert-butoxycarbonyl-6-methyl-2,3,4,5,6,7-hexahy-
droindol-4-one-3-carboxylate (5). To the bicyclic ester 4 (1.27 g,
4.13 mmol) in 200 mL of CH2Cl2 was added 2-((tert-butoxycarbony-
loxy)imino)-2-phenylacetonitrile (“Boc-ON”, 1.22 g, 4.96 mmol) and
KH (199 mg, 4.96 mmol). The mixture was stirred under N2 at room
temperature for 22 h and then washed with 1:1 satd. K2CO3 until no
more UV activity was observed in the aqueous layer. The organic layer
was washed with brine, dried (MgSO4), and concentrated, and the
residue was chromatographed (silica, 2:1 hexanes/EtOAc) to give 5 as
two diastereomers.
Figure 15. Relaxation rates and characterization of dynamical behavior
of 15N-labeled S,S-9-O (pH 4.2, 4 °C).
behavior of conjugate S,S-9-O. The data are consistent with a
conformational ensemble for S,S-9-O in which helical conform-
ers are dominant, and they indicate that the template works by
reducing the fraying at the N-terminus of the peptide, as
intended. The NMR results are thus consistent with the
chiroptical studies, which together suggest that the template will
be useful in probing structural effects on helical behavior and
the role of helical peptide conformations in ligand-receptor and
protein-protein interactions.
Experimental Section
tert-Butyl 4-(1-Methoxycarbonyl-2-nitroethyl)-1-methyl-3,5-di-
oxocyclohexanecarboxylate (3). tert-Butyl 3,5-dimethoxybenzoate63
(11.51 g, 44 mmol) and tert-butyl alcohol (4.59 mL, 48 mmol) were
dissolved in 50 mL of THF. Ammonia (200 mL) was distilled from
sodium into the flask at -78 °C, and potassium (4.5 g, 15 mmol) was
added in small pieces until the solution remained blue. Methyl iodide
(14.9 mL, 240 mmol) was filtered through alumina, diluted with 50
mL of THF, and added to the solution, which turned from blue to brown
and then to a light yellow. The solution was stirred at -78 °C for 2
h, the ammonia was allowed to evaporate under a stream of N2, and
the reaction mixture was concentrated under reduced pressure. The
residue was dissolved in water and extracted several times with ether,
and the organic layer was dried (MgSO4) and concentrated to 10.3 g
(84%) of tert-butyl 3,5-dimethoxy-1-methylcyclohexa-2,5-dienecar-
boxylate, which was carried on directly to the hydrolysis step.
A solution of the alkylation product (5.09 g, 20 mmol) and mercuric
nitrate (2.26 g, 6.6 mmol) in 140 mL of CH3CN and 30 mL of water
was heated to reflux under N2 for 12 h. The reaction mixture was
concentrated, dissolved in aqueous NaCl solution, and extracted with
three portions of EtOAc. The organic layer was dried (MgSO4) and
concentrated, and the crude product was chromatographed (95:5 CH2-
1
Higher Rf diastereomer: 944 mg (56%); H NMR δ 4.00 (m, 2),
3.80 (m, 1), 3.71 (s, 3), 2.76 (d, 1, J ) 4.2), 2.71 (s, 1), 2.33 (d, 1, J
) 16.2), 1.49 (s, 9), 1.37 (s, 9), 1.32 (s, 3); 13C NMR δ 191.8, 174.5,
173.4, 150.7, 116.6, 83.0, 81.3, 60.3, 52.5, 46.1, 45.6, 41.6, 34.9, 28.1,
27.8, 24.7; MS (EI) m/z 409 (2), 336 (2), 309 (5), 280 (7), 250 (15),
208 (31), 194 (45), 57 (100). Anal. Calcd for C21H31NO7: C, 61.60;
H, 7.63; N, 3.42. Found: C, 61.45; H, 7.71; N, 3.35.
Lower Rf diastereomer: 601 mg, 36%; 1H NMR δ 4.05 (s, 1), 4.03
(d, 1, J ) 1.6), 3.87 (m, 1), 3.68 (s, 3), 2.82 (dd, 1, J ) 1.2, 16.4),
2.62 (dd, 1, J ) 2.2, 18.6), 2.23 (d, 1, J ) 16.4), 1.50 (s, 9), 1.39 (s,
9), 1.31 (s, 3).
Methyl (3RS,6RS)- and (3RS,6SR)-1,6-Di-tert-butoxycarbonyl-
3,6-dimethyl-2,3,4,5,6,7-hexahydroindol-4-one-3-carboxylate (cis-6
and trans-6). To a solution of the Boc-protected diester (mixture of
isomers, 3.49 g, 8.55 mmol), methyl iodide (6.07 g, 2.66 mL, 42.8
mmol), and hexamethylphosphoric triamide (HMPA) (3.83 g, 3.72 mL,
21.4 mmol) in THF (67 mL) at -78 °C was added lithium hexameth-
yldisilazide (LiHMDS) (12.0 mL, 12.0 mmol, 1.0 M in THF) Via
addition funnel over a period of 15 min. After 30 min, acetic acid
(721 mg, 687 µL, 12.0 mmol) was added, and the mixture was warmed
to room temperature and concentrated under reduced pressure. The
(63) Crowther, G. P.; Kaiser, E. M.; Woodruff, R. A.; Hauser, C. R.
Org. Synth. 1971, 51, 96-100.
(64) Shechter, H.; Conrad, F. J. Am. Chem. Soc. 1953, 75, 5610-5613.

Template for Stabilization of a Peptide R-Helix
J. Am. Chem. Soc., Vol. 119, No. 28, 1997 6471
1
residue was dissolved in EtOAc (100 mL), washed with brine (25 mL),
dried (MgSO4), and concentrated, and the residue was chromatographed
(2:1, hexanes/EtOAc) to give the monomethylated product (3.03 g, 7.18
mmol, 84% yield) as an oil. The diastereomers were separated by
MPLC on a Lobar Lichroprep Si 60 column, size C, using a 3:1 hexane/
EtOAc mixture as solvent, monitoring at 254 nm; the maximum amount
injected under these conditions was 1.4 g. For cis-6 (higher Rf
diastereomer): 1H NMR δ 4.05 (d, 1, J ) 11.6), 3.63 (s, 3), 3.59 (d,
1, J ) 11.5), 2.74 (d, 1, J ) 16.2), 2.66 (d, 1, J ) 18.5), 2.23 (d, 1, J
) 16.2), 1.49 (s, 9), 1.44 (s, 3), 1.39 (s, 9), 1.28 (s, 3); 13C NMR δ
191.5, 174.4, 174.2, 150.9, 82.8, 81.2, 60.7, 52.4, 48.1, 46.5, 45.3, 34.7,
28.1, 27.7, 24.9, 22.3. For trans-6 (lower Rf diastereomer): 1H NMR
δ 4.03 (d, 1, J ) 11.7), 3.68 (s, 3), 3.60 (d, 1, J ) 11.7), 2.72 (d, 1, J
) 16.1), 2.63 (d, 1, J ) 18.4), 2.26 (d, 1, J ) 16.2), 1.49 (s, 9), 1.41
(s, 3), 1.37 (s, 9), 1.30 (s, 3); 13C NMR δ 191.5, 174.8, 174.4, 150.9,
82.9, 81.3, 60.8, 52.6, 47.9, 46.4, 45.7, 36.6, 35.0, 28.1, 27.9, 27.8,
25.0, 22.5; MS (FAB) m/z 424 (MH)+. Anal. Calcd for C22H33NO7:
C, 62.39; H, 7.85; N, 3.31. Found: C, 62.17; H, 7.88; N, 3.34.
(3RS,6RS)-6-tert-Butoxycarbonyl-3,6-dimethyl-2,3,4,5,6,7-hexahy-
droindol-4-one-3-carboxylic Acid (cis-7). A solution of the ester cis-6
(725 mg, 1.72 mmol) and 2 N NaOH (1.89 mL, 3.78 mmol) in 50%
aqueous methanol (28 mL) was heated at 80 °C for 24 h. After cooling
the reaction to room temperature, the solution was evaporated, and the
residue was dissolved in water (15 mL). The aqueous solution was
carefully acidified to pH ) 2 by the addition of 1 N HCl and extracted
with CH2Cl2 (3 × 15 mL), and the combined organic fractions were
dried (MgSO4) and evaporated to give the carboxylic acid cis-7 (454
mg, 1.47 mmol, 85% yield). 1H NMR indicated >95% purity as well
as complete removal of the Boc group. For cis-7: 1H NMR (400 MHz)
δ 6.78 (s, 1), 4.31 (d, 1, J ) 11.8), 3.45 (d, 1, J ) 11.8), 2.92 (d, 1, J
) 16.8), 2.76 (d, 1, J ) 16.5), 2.39 (d, 1, J ) 16.5), 2.33 (d, 1, J )
16.8), 1.53 (s, 3), 1.36 (s, 9), 1.29 (s, 3); 13C NMR δ 190.2, 177.0,
174.0, 170.5, 108.6, 82.0, 57.3, 51.5, 45.9, 44.6, 33.8, 27.7, 24.9, 24.4;
MS (FAB) m/z 310 (MH)+; HRMS (FAB) calcd for C16H24NO5
310.1655, Found 310.1656.
Compound trans-7 was prepared from trans-6 in a similar fashion:
1H NMR (400 MHz) δ 7.20 (s, 1), 4.28 (d, 1, J ) 12.0), 3.46 (d, 1, J
) 12.0), 2.95 (d, 1, J ) 16.9), 2.84 (d, 1, J ) 16.9), 2.32 (d, 1, J )
16.9), 2.26 (d, 1, J ) 16.6), 1.45 (s, 3), 1.39 (s, 9), 1.31 (s, 3); 13C
NMR δ 189.1, 177.5, 174.0, 171.6, 108.8, 81.7, 57.5, 51.4, 46.0, 44.4,
33.8, 27.8, 25.3, 24.9.
(3R,6R)- and (3S,6S)-O-(1,6-Di-tert-butoxycarbonyl-3,6-dimethyl-
3-oxo-2,3,4,5,6,7-hexahydroindol-4-one)-L-lactate Methyl Ester (R,R-
12-O and S,S-12-O). A slurry of KH (115 mg, 2.87 mmol, washed
with hexane) and Boc-ON (292 mg, 1.19 mmol) in THF (11 mL) was
stirred at 0 °C, and a solution of cis-7 (334 mg, 1.08 mmol) in THF
(11 mL) was added Via cannula. On completion of the addition, the
cooling bath was removed and the reaction mixture was stirred at room
temperature for 24 h. A solution of 0.5 M HCl (25 mL) was added,
the resulting mixture was extracted with CH2Cl2 (3 × 50 mL), and the
combined extracts were dried (MgSO4) and evaporated to give a mixture
of the BOC-protected compound cis-11 and oxime side product (621
mg).
For S,S-12-O: H NMR (500 MHz, CDCl3) δ 5.09 (q, 1, J ) 7.1),
4.20 (d, 1, J ) 11.7), 3.71 (s, 3), 3.65 (d, 1, J ) 11.7), 3.54 (d, 1, J )
18.4), 2.82-2.74 (m, 2), 2.26 (d, 1, J ) 16.2), 1.54 (s, 3), 1.53-1.50
(m, 12), 1.40 (s, 9), 1.30 (s, 3).
For R,R-12-O: 1H NMR (500 MHz, CDCl3) δ 5.12 (q, 1, J ) 7.1),
4.20 (d, 1, J ) 11.6), 3.71 (s, 3), 3.69-3.66 (m, 1), 3.54 (d, 1, J )
18.5), 2.80 (d, 1, J ) 18.7), 2.75 (d, 1, J ) 16.4), 2.26 (d, 1, J )
16.4), 1.52 (s, 9), 1.45 (s, 3), 1.42-1.40 (m, 12), 1.31 (s, 3).
HRMS (FAB) characterization of the R,S-diastereomer, prepared
from trans-6 by the same procedure: calcd for C25H38NO9 496.2547,
found 496.2533.
General Procedure for the Preparation of Templated Peptides
9-O by Solid Phase Peptide Synthesis. To a solution of the methyl
ester 12-O (0.0323-0.114 mmol) in 25% aqueous methanol (1-2 mL)
was added 1 N LiOH (1.2 equiv). The resulting mixture was stirred at
room temperature for 1-2 h and then concentrated under reduced
pressure. The residue was dissolved in water (15 mL) and carefully
acidified to pH ) 2 with 1 N HCl. The resulting solution was extracted
with CH2Cl2 (3 × 15 mL), and the combined organic fractions were
dried (MgSO4) and concentrated in Vacuo to give the monocarboxylic
acid (61-88% yield).
The Fmoc-EALAKA-Rink resin65 (gift of Reyna Simon and Simon
Ng of Chiron Corporation; 1.25 equiv) was allowed to swell in DMF
(1.5 mL) for 1 h. DMF was removed by filtration, and the resin was
washed with additional DMF (1.5 mL). The resin was treated with
20% piperidine in DMF (2 × 1.5 mL, 3 and 10 min) to remove the
Fmoc protecting group, filtered, and washed with additional DMF (10
× 1.5 mL). A solution of the carboxylic acid prepared above, ethyl
2-(dimethylamino)ethyl carbodiimide (EDC, 1 equiv), and 1-hydroxy-
benzotriazole (HOBT, 1 equiv) in DMF (1.5 mL) was added to the
resin and the slurry was stirred for 1 h. The resin was filtered, washed
with DMF (2 × 1.5 mL) followed by CH2Cl2 (2 × 1.5 mL), and stored
under vacuum overnight.
The resin was treated with cleavage reagent R (90% trifluoroacetic
acid (TFA), 5% thioanisole, 3% ethanedithiol, 2% anisole; 5 mL) under
N2 for 2.5 h.66 The mixture was filtered through glass wool into cold
ether, and the solution was stored at 0 °C overnight. The fluffy white
precipitate that formed was collected on a fine frit, washed with cold
ether, and stored under vacuum overnight. The peptide conjugate was
purified by reverse phase HPLC using a two-solvent system (solvent
A 90% H20, 10% CH3CN, 0.1% TFA; solvent B 10% H20, 90% CH3-
CN, 0.1% TFA) with programmed elution (0 min, 100% A; 30 min,
50% A, 50% B; 40 min, 100% B; 50 min, 100% B; 60 min, 100% A)
on a Vydac C18 prep column, monitored at 304 nm.
(3S,6S)-(3,6-Dimethyl-6-(hydroxycarbonyl)-3-oxo-2,3,4,5,6,7-
hexahydroindol-4-one)-L-lactyl-Glu-Ala-Leu-Ala-Lys-Ala-NH2 (S,S-
9-O): LRMS (FAB) 908.4 (MH+); HRMS (FAB) calcd for C41H66N9O14
907.4729, found 907.4725.
(3R,6R)-(3,6-Dimethyl-6-(hydroxycarbonyl)-3-oxo-2,3,4,5,6,7-
hexahydroindol-4-one)-L-lactate-Glu-Ala-Leu-Ala-Lys-Ala-NH2 (R,R-
9-O): LRMS (FAB) 908.5 (MH+).
Circular Dichroism (CD) Measurements. All stock solutions were
prepared with doubly distilled water. Spectra of pure buffer were
obtained for each temperature and subtracted from the spectra for each
peptide. The concentrations of stock solutions of the 9-N, 10-N, 9-O,
and 10-O diastereomers were determined by measuring the UV
absorbance in water at 307 nm (ꢀ307 ) 14 400). The ꢀ value was
determined by careful dilution of a precisely weighed sample of S,R-
9-N in CH3CN and measurement of the absorption at 304 nm (red shift
due to decreased solvent polarity). The concentration of the control
peptide Ac-AEALAKA-NH2 was determined from amino acid analysis.
The observed ellipticities were corrected by subtraction of the measured
ellipticities of reference samples of the corresponding cap itself in buffer
(10-N or 10-O), either measured at the same concentration as the peptide
conjugate or scaled to the same concentration prior to subtraction.
(A) Diastereomers of 9-N and 10-N. The standard buffer solution
was 0.1 M KCl and 50 mM potassium phosphate, with the pH adjusted
to 5.2 by the addition of 1.0 M KOH. Spectra were obtained on a
The crude Boc-protected product was dissolved in THF (7 mL) with
triethylamine (240 mg, 0.331 mL, 2.38 mmol) and 2,4,6-trichloroben-
zoyl chloride (580 mg, 0.372 mL, 2.38 mmol). After 2 h, the cloudy
mixture was filtered through a plug of glass wool, the amine
hydrochloride salt was washed with THF, and the filtrate was
concentrated in Vacuo. The residue was dissolved in toluene (10 mL)
and methyl L-lactate (337 mg, 0.309 mL, 3.24 mmol) and DMAP (791
mg, 6.48 mmol) were added. The flask was equipped with a reflux
condenser and immersed in a preheated (110 °C) oil bath. After 3.5
h, the solution was cooled, diluted with ether (40 mL), washed
sequentially with 0.5 N HCl (20 mL), water (20 mL), saturated NaHCO3
(20 mL), and water (20 mL), dried (MgSO4), and evaporated. The
residue was first purified by flash chromatography (2:1, hexane/EtOAc)
to give the template lactate ester 12-O as a mixture of diastereomers
(400 mg, 806 mmol, 70% yield for four steps). The diastereomers
were separated by MPLC on a Lobar Lichroprep Si 60 column, size
C, employing 2.5:1 hexane/EtOAc as mobile phase, monitoring at 254
nm. The less-polar compound was identified as R,R-12-O by chemical
correlation to R,R-8-N (as described in the Supporting Information).
(65) Rink, H.; Sieber, P. In Peptides 1988; Jung, G., Bayer, E., Eds.;
Walter de Gruyter & Company: Berlin, 1989; pp 139-141.
(66) Hudson, D. In Peptides 1988; Jung, G., Bayer, E., Eds.; Walter de
Gruyter & Company: Berlin, 1989; pp 211-213.

6472 J. Am. Chem. Soc., Vol. 119, No. 28, 1997
Austin et al.
Jasco J-600 spectropolarimeter at 15 µM concentration in a 1.0 cm
Hellma cylindrical cell. A Neslab Instruments circulating ethylene
glycol/water cooling unit connected to the jacketed cell was used for
temperature regulation. With this arrangement, high absorbance
prevented accurate measurements below 200 nm. The temperature was
allowed to stabilize for at least 15 min before a spectrum was recorded.
The spectra were obtained from 200 to 260 nm at a scan speed of 20
nm/min and 4 scans per run. The sensitivity was (50 mdeg, the time
constant was 2.0 s, and the band width was 1.0 nm with a 0.2 nm step
resolution.
(B) Diastereomers of 9-O and 10-O. The standard buffer was 1.05
mM KH2PO4 buffer at pH 5.0; spectra were recorded in 0.2 cm, strain-
free cells on an Aviv 62 DS spectropolarimeter with thermojacketed
cell holder. Spectra were obtained over the range of 185-300 nm; 5
scans were typically recorded with a band width of 1.5 nm, 0.5 nm
step size, and 1 s time constant. Temperature melts were recorded
over the range of 0.5-70 °C at 222 nm with a 30 s averaging time,
2-5 °C step size, and 3-5 min equilibration time.
For NMR spectra recorded over a range of temperatures, the
temperature of the probe was calibrated using solutions of analytical
grade methanol or ethylene glycol by the method of Van Geet.72
Chemical shifts were referenced to internal dioxane (δ ) 3.74 ppm).
NMR measurements of 15N-labeled material were recorded at 600.14
MHz (1H) and 60.819 MHz (15N) on a Bruker AMX-600 at 4 and 37
°C. Spectral widths of 1094.74 and 6024.10 Hz were used in F1 and
F2, respectively, with a total of 2K complex data points collected in
F2. Phase sensitivity was achieved using TPPI.67 Water-suppression
was achieved using a combination of trim pulses73 and low-power
presaturation. Measurements of the 15N T1 and T2 values were carried
out using the pulse programs described by Barbato74 with relaxation
delays of 26, 54, 96, 250, 503, 797, and 1499 ms for T1 (each time
point was recorded twice) and 8, 39, 79, 181, 605, 909, and 1500 ms
for T2 (each time point was recorded one or two times). A recycle
delay of 2.5 s was used and 29 complex t1 increments of 32 scans each
were acquired. The relaxation times T1 and T2 were obtained from
exponential fits of the peak height data. NOE experiments were carried
out according to Skelton et al.75 in which two data sets were collected,
one with the NOE and one without. A recycle delay of 5 s was used,
and 29 complex t1 increments of 64 scans were acquired. The NOE
effect was calculated as the ratio of peak heights in the two spectra
collected with and without NOE.
Synthesis of 15N-Labeled Conjugate. 15NR-Labeled amino acids
were obtained from Cambridge Isotopes. 15NR-Glutamic acid and 15NR-
lysine were converted to the NR-Fmoc-derivatives with tert-butyl and
Boc side chain protection, respectively, as described in the Supporting
Information; 15NR-leucine was converted to the Fmoc derivative by the
standard procedure, and 15NR-Fmoc-alanine was available commercially.
The side chain-protected, 15NR-containing hexapeptide EALAKA was
prepared on Rink resin and coupled to the cap-L-lactic acid as described
above. After purification by reverse phase chromatography (with the
desired product eluting at ∼24% CH3CN/H2O), the 15NR-labeled
conjugate S,S-9-O was obtained as a white powder and characterized
15N NMR Relaxation Calculations. The 15N relaxation data for
conjugate S,S-9-O was analyzed using the formalism of Lipari and
Szabo for isotropic tumbling.56,57 Model-free parameters were obtained
by substituting the spectral density expression (eq 5) into those
describing the dependence of the relaxation rate constants T1 and T2,
and the heteronuclear NOE on the spectral density values (eqs 2-4);
the experimentally measured values were used to optimize the fit with
the software programs MODELFREE3.1 and DNMR.60,61
1
by mass spectrometry (FAB m/z 914 (MH+)) and H NMR spectro-
scopy.
Sample Preparation. Solutions for the observation of exchangeable
protons were prepared by dissolving the conjugate in 500 µL of buffer
containing 50 mM potassium phosphate and 0.1 M KCl, lyophilizing,
and dissolving the residue in 450 µL of H2O and 50 µL of D2O at a
concentration of 3.5 mM in conjugate. The sample pH recorded within
the NMR tube was 4.2; unless otherwise stated, all spectra were
recorded at this pH.
Acknowledgment. This work was supported by a grant from
the National Institutes of Health (no. GM-30759 to P.A.B.);
A.M.S. and W.N.H. received support from NIH Training Grants
nos. GM08295 and GM-08352, respectively. P.A.B. has been
on appointment as a Miller Research Professor in the Miller
Institute for Basic Research in Science. D.E.W., C.W.L., and
H.S.C are supported through the Director, Office of Energy
Research, Office of Biological & Environmental Research,
General Life Science Division of the U.S. Department of Energy
under contract no. DE-AC03-76F00098. An instrumentation
grant to D.E.W. was provided by the NSF (BBS 87-20134).
We thank Prof. Susan Marqusee (Molecular and Cell Biology,
UC Berkeley) for assistance with the CD measurements, Dr.
Reyna Simon and Simon Ng (Chiron Corporation) for solid
phase synthesis of the hexapeptide precursor, Georges Lauri
for molecular dynamics simulations of the template peptides,
and Kathy Lee for bringing the word “aglet” to our attention.
NMR Experimental Details. 1H NMR spectra were recorded on a
Bruker AM 500 spectrometer equipped with an Aspect 3000 computer,
and data were processed using the Bruker UXNMR software package.
1D spectra in 90% H2O/10% D2O were recorded using a jump-and-
return selective excitation sequence, (π/2)-D2-(π/2)-acquire.45 The
carrier frequency was set on the water signal with a delay time, D2, of
140 µs to ensure that there would be effective excitation over the
envelope of amide protons. 2D experiments were recorded at a probe
temperature of 4 °C in the phase-sensitive mode using the time-
proportional phase incrementation method of Marion and Wu¨thrich67
for quadrature detection in f1. The same spectral widths were used as
in the 1D experiments. Between 8 and 32 transients of 2048 data points
each were typically recorded in the f2 dimension and 256-512 t1
increments were recorded and zero-filled to 1024 data points. In cases
where high digital resolution was required in f1, up to 1024 t1 increments
were collected and zero-filled to 2048 data points. Prior to Fourier
transformation, spectra were zero-filled once in the f1 dimension and
subjected to a phase-shifted sine bell weighting function in both
dimensions.
Supporting Information Available: Experimental proce-
dures and characterization of the compounds not described
above; representative CD spectra of the diastereomers of 9-N
and 9-N-a; temperature and concentration dependence of CD
spectrum of S,S-9-O; preparation of R-Fmoc-R-15N-glutamic
acid γ-tert-butyl ester and R-Fmoc-ꢀ-Boc-R-15N-lysine; and full
TOCSY spectrum of S,S-9-O (17 pages). See any current
masthead page for ordering and Internet access instructions.
Phase-sensitive NOESY spectra were recorded using a jump-and-
return45 read pulse with the overall sequence (π/2)-t1-(π/2)-τm-(π/
2)-D2-(π/2)-acquire (D2 ) 140 µs). Mixing times (Tm) of 200,
400, and 800 ms were used with a 20 ms z-filter to remove zero
quantum artifacts and appropriate phase cycling to select z-magnetiza-
tion. The initial value of t1 was set to 3 µs.
JA964231A
TOCSY68 spectra were recorded with a 50 ms MLEV-17 spin-locking
pulse.69 Phase coherent solvent suppression was achieved using a
DANTE70,71 sequence (pulse width 25 µs and inter-pulse delay of 250
µs) applied for 1.7 s before the start of each transient.
(71) Zuiderweg, E. R. P.; Hallenga, K.; Olejniczak, E. T. J. Magn. Reson.
1986, 70, 336-343.
(72) Van Geet, A. L. Anal. Chem. 1968, 40, 2227-2229.
(73) Messerle, B. A.; Wider, G.; Otting, G.; Weber, C.; Wu¨thrich, K. J.
Magn. Res. 1989, 85, 608-613.
(67) Marion, D.; Wu¨thrich, K. Biochem. Biophys. Res. Commun. 1983,
113, 967-974.
(68) Braunschweiler, L.; Ernst, R. R. J. Magn. Reson. 1983, 53, 521-
528.
(69) Bax, A.; Davis, D. G. J. Magn. Reson. 1985, 65, 355-360.
(70) Morris, G. A.; Freeman, R. J. Magn. Reson. 1978, 29, 433-462.
(74) Barbato, G.; Ikura, M.; Kay, L. E.; Pastor, R. W.; Bax, A.
Biochemistry 1992, 31, 5269-5278.
(75) Skelton, N. J.; Palmer, A. G. I.; Akke, M.; Ko¨rdel, J.; Rance, M.;
Chazin, W. J. J. Magn. Res., Ser. B 1993, 102, 253-264.
(76) The more conservative n ) 7 was chosen, although the contribution
from the first Ala residue may be compensated by the controls 7-N or 7-O.