Design, synthesis, and conformational analysis of a proposed type I β- turn mimic

Article abstract of DOI:10.1021/ja974023y

In an effort to design a dipeptide structural mimic of protein and peptide β-turns, we have prepared and evaluated the conformation of derivatives of the novel, highly constrained ten-membered lactam, (3S,10S)- (6E)-2-azacyclodec-6-enone (1). A synthetic

Full text of DOI:10.1021/ja974023y

4334
J. Am. Chem. Soc. 1998, 120, 4334-4344
Design, Synthesis, and Conformational Analysis of a Proposed Type I
â-Turn Mimic
Brian E. Fink, Phil R. Kym,† and John A. Katzenellenbogen*
Contribution from the Department of Chemistry, UniVersity of Illinois, 600 S. Mathews AVenue,
Urbana, Illinois 61801
ReceiVed NoVember 25, 1997
Abstract: In an effort to design a dipeptide structural mimic of protein and peptide â-turns, we have prepared
and evaluated the conformation of derivatives of the novel, highly constrained ten-membered lactam, (3S,-
10S)-(6E)-2-azacyclodec-6-enone (1). A synthetic route utilizing ring-closing olefin metathesis (RCM) has
been used to prepare this novel ten-membered ring in high yield. X-ray crystallography and 1H NMR analysis
have established that ring closure proceeds to give the trans-olefin and that 1 exists in two conformations, that
of a chair-chair and chair-boat. Monte Carlo-molecular mechanics conformational searching has indicated
that this ring system would be a good mimic of a type I â-turn. The synthesis of model tri- and tetrapeptide
analogues based on 1 is reported. NMR studies indicate that the tetrapeptide derivatives constrain the i+1
and i+2 torsion angles to within 30° of those predicted for an ideal type I â-turn (φ1 ) -82°, ψ1 ) -20°, φ2
) -107°, ψ2 ) -18°) and that this conformation was shown to be stable in both hydrogen-bonding solvents
as well as non-hydrogen bonding solvents at various temperatures.
Introduction
great strides have been made, the direct determination of the
receptor-bound conformation adopted by a bioactive peptide is
still a formidable endeavor.
Peptide ligands and protein receptors play critical roles in
the regulation of nearly every biological process. Despite this
central role, identifying the basic elements necessary for recog-
nition between a peptide ligand and its receptor at the molecular
level remains a formidable task. While advances in site-directed
mutagenesis and peptide synthesis have provided dramatic in-
sights into the functional group requirements for binding, they
offer little information regarding the spatial orientation of both
ligand and receptor upon binding. Knowledge of the intimate
details of the three-dimensional interaction between peptide
ligand and protein receptor could be invaluable in understanding
bioactivity and in the design of analogues during the drug
discovery process.
In most cases, determining the spatial orientation of important
functionality has involved the conformational analysis of the
native peptide ligand using such techniques as NMR spectros-
copy, X-ray crystallography, circular dichroism, and molecular
modeling.1,2 However, peptides are characteristically highly
flexible molecules that exist in multiple conformations. The
central issue is then whether the conformation found in solution
has any real physical meaning or whether it represents an
average of the numerous conformations available to the peptide.
Perhaps the more important question is whether any of these
solution conformations correspond to that which is adopted by
the peptide ligand when it is bound to the receptor.3 While
One method for the indirect determination of the bioactive
conformation of a peptide ligand is through the implementation
of conformational restrictions which limit the torsional space
available to the native peptide.3,4 A highly flexible molecule
undergoes a large loss in entropy upon binding its receptor.
However, an analogue of the native peptide which properly
restricts important recognition elements to an appropriate three-
dimensional geometry should experience less of an entropic
change and may thereby bind more effectively to a receptor
than its more flexible natural counterpart.4 In this way, con-
formationally restricted peptides offer a means to probe peptide
ligand conformation and provide indirect information about the
three-dimensional interplay between receptor and ligand. In the
pursuit of this strategy, many efforts have been made to prepare
compounds that mimic certain secondary structural features of
peptides which are thought to play important roles in recognition
and biological activity.3-5
One common structural feature observed in small peptides
and proteins is the â-turn (Figure 1).1 By definition, the â-turn
consists of a tetraresidue sequence in which the peptide chain
reverses direction by approximately 180°.6 Most â-turns contain
an intramolecular hydrogen bond between the carbonyl oxygen
of the first residue (i) and the amide NH proton of the fourth
residue (i+3), which forms a pseudo-ten-membered ring (Figure
1). â-Turns that do not show an intramolecular hydrogen bond
are termed as “open turns”.1,7 Of the most commonly found
â-turns (see below), approximately 60% show an intramolecular
hydrogen bond.4
* Address correspondence to John A. Katzenellenbogen, Department of
Chemistry, University of Illinois, 461 Roger Adams Laboratory, Box 37-
5, 600 S. Mathews Avenue, Urbana, IL 61801. Telephone: (217) 333-
6310. Fax: (217) 333-7325. E-mail: jkatzene@uiuc.edu.
† Current address: Abbott Laboratories, Department 4NB, Bldg. AP10,
100 Abbott Park Road, Abbott Park, IL 60064.
(1) Smith, J. A.; Pease, L. G. CRC Crit. ReV. Biochem. 1980, 8, 314-
398.
(2) Rose, G. D.; Gierash, L. M.; Smith, J. A. AdV. Prot. Chem. 1985,
37, 1-109.
(3) Hruby, V. Life Sci. 1982, 31, 189-199.
(4) Ball, J. B.; Alewood, P. F. J. Mol. Recog. 1990, 3, 55-64.
(5) Kemp, D. S. Trends Biotechnol. 1990, 8, 249-255.
(6) Lewis, P. N.; Momany, F. A.; Scheraga, H. A. Biochim. Biophys.
Acta 1973, 303, 211-229.
(7) Crawford, J. L.; Lipscomb, W. N.; Schellman, C. G. Proc. Natl. Acad.
Sci. U.S.A. 1973, 70, 538-542.
S0002-7863(97)04023-7 CCC: $15.00 © 1998 American Chemical Society
Published on Web 04/28/1998

Synthesis of a Proposed Type I â-Turn Mimic
J. Am. Chem. Soc., Vol. 120, No. 18, 1998 4335
Figure 1. The structure of a â-turn.
Table 1. Torsion Angles for Classical â-Turns
turn type
torsion angle
I
I′
II
II′
III
III′
φ1
ψ1
φ2
-60
-30
-90
0
60
30
90
0
-60
120
80
0
15
60
120
80
0
-60
-30
-60
-30
18
60
30
60
30
3
ψ2
% occurrence
42
3
5
The most widely accepted system for the classification of
â-turns is based upon the φ and ψ peptide backbone torsion
angles of residues i+1 and i+2 (Figure 1). The ideal torsion
angles predicted by Venkatachalam are given in Table 1,
although these are rarely realized in the â-turns actually observed
in proteins.8 A stringent criteria for assigning an observed â-turn
to a given class has been set, allowing three of the four torsion
angles to deviate less than 30° from the ideal and one angle to
deviate less than 45°.6 The type I â-turn is by far the most
common, followed by type III and type II. The mirror image
of these turns are also possible, denoted in Table 1 with a prime,
although they occur far less frequently.9-11
The â-turn is found in all proteins, comprising approximately
25% of the amino acid residues in most proteins.1,2 The location
of â-turns at the surface of proteins and the predominance of
amino acids with reactive side chains suggest they may act as
key recognition elements for initiation of biological events.2 The
prevalence of this structural motif in small peptide ligands as
well as in proteins has led to numerous efforts toward the
development of analogues designed to stabilize a peptide chain
in a â-turn conformation.4 Examples of cyclic and bicyclic
modifications which stabilize â-turns include dipeptide lac-
tams,12-14 spirolactam-bicyclic and tricyclic proline based
systems,15-18 the bicyclic-turned dipeptide (BTD),19,20 and a
number of medium-ring heterocyclic compounds.21-26 In some
Figure 2. Proposed type I â-turn mimic (1) and Substance P analogue.
cases, incorporation of these structural mimics in the native
peptide has led to analogues with increased biological activity
or stability.14,19,27-31 To date, the majority of â-turn mimics
reported are for type II or II′ turns. To our knowledge, only
two examples exist of mimics for type I â-turns,22 while these
turns account for approximately 40% of the turns found in
protein crystal structures. In this paper, we present our approach
toward the design of a general â-turn mimic which is able to
mimic a number of â-turn types, including type I â-turns.
Additionally, the synthesis and conformational analysis of some
derivatives of the mimic will be discussed.
Results and Discussion
Design of â-Turn Mimic. Our initial interest in designing
a â-turn mimic focused on the neuropeptide Substance P.
Substance P is a biologically active undecapeptide that has been
shown to play important roles in pain transmission, bronchial
constriction, and vasodilatation (Figure 2).32,33 A â-turn has
been observed in the final four C-terminal residues Phe8-Gly9-
(8) Venkatachalam, C. M. Biopolymers 1968, 6, 1425-1436.
(9) Hutchinson, E. G.; Thornton, J. M. Protein Sci. 1994, 3, 2207-2216.
(10) Chou, P. Y.; Fasman, G. D. J. Mol. Biol. 1977, 115, 135-175.
(11) Wilmot, C. M.; Thornton, J. M. J. Mol. Biol. 1988, 203, 221-232.
(12) Freidinger, R. M.; Perlow, D. S.; Veber, D. F. J. Org. Chem. 1982,
47, 104-109.
(22) Kahn, M.; Wilke, S.; Chen, B.; Fujita, K.; Lee, Y.-H.; Johnson, M.
E. J. Mol. Recog. 1988, 1, 75-79.
(23) Kahn, M.; Nakanishi, H.; Chrusciel, R. A.; Fitzpatrick, D.; Johnson,
M. E. J. Med. Chem. 1991, 34, 3395-3399.
(24) Olson, G. L.; Voss, M. E.; Hill, D. E.; Kahn, M.; Madison, V. S.;
Cook, C. M. J. Am. Chem. Soc. 1990, 112, 323-333.
(25) Kemp, D. S.; McNamara, P. Tetrahedron Lett. 1982, 23, 3761-
3764.
(13) Freidinger, R. M.; Brady, S. F.; Paleveda, W. J.; Perlow, D. S.;
Colton, C. D.; Whitter, W. L.; Saperstein, R.; Brady, E. J.; Cascieri, M.
A.; Veber, D. F. In Clinical Pharmacology in Psychiatry; Dahl, Gram, Paul,
and Potter, Eds.; Springer-Verlag: Berlin, 1987; pp 12-19.
(14) Freidinger, R. M.; Veber, D. F.; Perlow, D. S.; Brooks, J. R.;
Saperstein, R. Science 1980, 210, 656-658.
(15) Genin, M. J.; Johnson, R. L. J. Am. Chem. Soc. 1992, 114, 8778-
8783.
(26) Kemp, D. S.; Stites, W. E. Tetrahedron Lett. 1988, 29, 5057-5060.
(27) Ward, P.; Ewan, G. B.; Jordan, C. C.; Ireland, S. J.; Hagan, R. M.;
Brown, J. R. J. Med. Chem. 1990, 33, 1848-1851.
(28) Sato, K.; Hotta, M.; Dong, M.-H.; Hu, H.-Y.; Taulene, J. P.;
Goodman, M.; Nagai, U.; Ling, N. Int. J. Peptide Protein Res. 1991, 38,
340-345.
(16) Genin, M. J.; Ojala, W. H.; Gleason, W. B.; Johnson, R. L. J. Org.
Chem. 1993, 58, 2334-2337.
(17) Hinds, M. G.; Richards, N. G. J.; Robinson, J. A. J. Chem. Soc.,
Chem. Commun. 1988, 1447-1449.
(18) Hinds, M. G.; Welsh, J. H.; Brennand, D. M.; Fisher, J.; Glennie,
M. J.; Richards, N. G. J.; Turner, D. L.; Robinson, J. A. J. Med. Chem.
1991, 34, 1777-1789.
(29) Baca, M.; Alewood, P. F.; Keny, S. B. H. Protein Struct. 1993, 2,
1085-1091.
(19) Nagai, U.; Sato, K.; Nakamura, R.; Kato, R. Tetrahedron 1993, 49,
3577-3592.
(30) Kahn, M.; Nakanishi, H.; Chrusciel, R. A.; Fitzpatrock, D.; Johnson,
M. E. J. Med. Chem. 1991, 34, 3395-3399.
(20) Nagai, U.; Sato, K. Tetrahedron Lett. 1985, 26, 647-650.
(21) Kahn, M.; Wilke, S.; Chen, B.; Fujika, K. J. Am. Chem. Soc. 1988,
110, 1638-1639.
(31) Su, T.; Nakanishi, H.; Xue, L.; Chen, B.; Tuladhar, S.; Johnson,
M. E.; Kahn, M. Bioorg. Med. Chem. Lett. 1993, 3, 835-840.
(32) Payan, D. G. Annu. ReV. Med. 1989, 40, 341-352.

4336 J. Am. Chem. Soc., Vol. 120, No. 18, 1998
Fink et al.
Leu10-Met11 by 1H NMR experiments, yet its presence has not
been unambiguously established, since replacement of Gly9-
Leu10 with a type II or II′ â-turn mimic did not result in a
biologically active compound.13,27 We envisioned that incor-
poration of a type I â-turn mimic into a substance P analogue
might help to define its bioactive conformation. If Substance
P adopted a type I â-turn, a proper analogue incorporating a
type I â-turn mimic should show increased biological activity.
We hypothesized that the Gly9-Leu10 dipeptide of Substance P
played structural roles and helped to orient the Phe8 and Met11
side chains;34 hence, we elected to constrain our mimic
externally in a ten-membered ring (1), through covalent attach-
ment through the i+1 and i+2 residues (Figure 2).35 While
this hydrocarbon tether was designed to mimic the side chains
of NK-1 selective petides, we were more concerned with the
ability of our mimic to adopt the appropriate backbone torsion
angles and its potential ability to influence peptide chain
direction and conformation.
Molecular Modeling of the â-Turn Mimic Core. Molec-
ular modeling techniques were used to investigate the low ener-
gy conformations of two diastereomers of the proposed â-turn
mimic. Monte Carlo conformational searching36 and energy
minimization of the 2-azacyclodec-6-enone ring systems 1a and
1b, where R ) CO2Me and R′ ) NHMe, were performed using
the MM2*37 force-field in Macromodel v. 5.5 (see Experimental
Section).38 The two lowest energy structures calculated for the
(3S,10S) diastereomer 1a are shown in Table 2. As predicted,
the expected chair-chair conformation is the lowest in energy,
whereas the second lowest energy conformation is a chair-
boat. Analysis of the φ and ψ torsion angles for the chair-
chair conformation and comparison to the φ and ψ torsion angles
for an ideal type I â-turn indicate that in this conformation, 1a
would be a good type I â-turn mimic (Table 2).
Table 2. Low Energy Conformers of (3S,10S)-â-Turn Mimic 1a
rel energy
conf
type I
chair-chair
chair-boat
(kcal/mol) φ1 (deg) ψ1 (deg) φ2 (deg) ψ2 (deg)
-60
-59
56
-30
-24
-80
-90
-101
-106
0
11
-12
0.0
2.04
Table 3. Low Energy Conformers of (3S,10R)-â-Turn Mimic 1b
rel energy
conf
(kcal/mol) φ1 (deg) ψ1 (deg) φ2 (deg) ψ2 (deg)
type II
type III
C1
C2
C3
-60
-60
-80
-55
53
120
-30
-14
-26
-82
128
80
-60
-65
-67
-73
106
0
-30
-30
-22
-28
-20
0.0
0.98
1.47
2.15
C4
-72
Scheme 1
A similar analysis of the (3S,10R) diastereomer 1b shows
that the lowest energy conformation (C1, Table 3) is a chair-
boat and does not mimic a known type of â-turn. The expected
chair-chair was found to be the fourth lowest in energy (C4).
However, the second lowest energy (C2), at 0.98 kcal/mol, does
accurately mimic a type III turn, and the third lowest in energy
(C3) mimics a type II turn, within the criteria set by Lewis.6
As the three lowest energy conformations are separated by less
than 2 kcal/mol, the (3S,10R) diastereomer seems to be more
flexible than the (3S,10S) diastereomer and should be able to
access either a type II conformation or type III â-turn turn
conformation depending on the environment at a receptor.
From these initial modeling experiments, it appeared that the
(6E)-2-azacyclodec-6-enone ring system would be able to
constrain a tetrapeptide sequence to the torsion angles found in
(33) Logan, M. E.; Goswami, R.; Tomczuk, B. E.; Venepalli, B. R. Annu.
Reports Med. Chem. 1991, 26, 43.
the three most common classes of â-turn. Thus, this would be
useful not only as a potential Substance P mimic but also as a
standard scaffold to probe the bioactive conformation of other
biologically active peptides where â-turn structures are thought
to be important.
Synthetic Approaches. A number of routes were explored
toward the synthesis of the 2-azacyclodec-6-enone core 1
(Scheme 1), including the macrolactamization of an ω-amino
acid derivative 2 followed by Curtius rearrangement to introduce
the C3 amino group and ring closing olefin metathesis of a
dipeptide diolefin 3.
(34) Srinivasan, R. Ph.D. Thesis, University of Illinois, Urbana, 1994.
(35) Although designed independently, our proposed turn mimic bears
considerable resemblance to the type II′ â-turn mimic proposed by Kemp.
However, conformational analysis and attempts at incorporation of Kemp’s
mimic into biologically active peptides were unsuccessful due to lack of
crystallinity and problems with poor solubility and tendency towards gel
formation. Kemp concluded that these problems were due to the two
macrocyclic amide functionalities, which were oriented perpendicular to
the plane of the turn and probably facilitated â-sheet formation through
intermolecular associations. Our proposed mimic should alleviate these
problems through replacement of one amide functionality with a surrogate
trans-olefin. Kemp, D. S.; Stites, W. E. Tetrahedron Lett. 1988, 29, 5057-
5060.
Macrolactamization Route. The most straightforward method
for the construction of the core 1 seemed to be macrolactam-
ization. However, the number of reported syntheses of medium-
sized ring (8-12-membered) lactams is few,39-41 compared to
the extensive literature on the preparation of small ring (3-7-
(36) Goodman, J. M.; Still, W. C. J. Comput. Chem. 1991, 12, 1110-
1117.
(37) Allinger, N. L. J. Am. Chem. Soc. 1977, 99, 8127-8134.
(38) Mohamadi, F.; Richards, N. G. J.; Guida, W. C.; Liskamp, R.;
Lipton, M.; Caufield, C.; Chang, G.; Hendrickson, T.; Still, W. C. J. Comput.
Chem. 1990, 11, 440-467.

Synthesis of a Proposed Type I â-Turn Mimic
J. Am. Chem. Soc., Vol. 120, No. 18, 1998 4337
Scheme 2
Scheme 4
Scheme 3
membered)42 and large ring (13-30-membered) lactams.43-46
The difficulty in preparing medium-sized ring lactams by
cyclization of ω-amino acids can be attributed to ring strain
that develops in the transition state during formation of the
medium-sized ring.47
After surveying a number of methods for the cyclization of
ω-amino acids with various activating strategies, we found that
under basic conditions, refluxing aminotriester 4 in DMF at high
dilution would yield a single diastereomer of 5 in modest yield
(Scheme 2). Only the diastereomer in which both the C3 and
C10 esters were oriented equatorially was observed, probably
reflecting the higher strain energy required to force either
substituent into an axial position.
The formation of a single diastereomer in the cyclization
reaction encouraged us to explore an asymmetric synthesis of
5. The stereogenic center R to the amine was created by an
asymmetric alkylation of Scho¨llkopf’s chiral glycine equivalent48
with a homoallylic diiodide. The preparation of the requisite
unsaturated diiodide is described in Scheme 3. Commercially
available trans-â-hydromuconic acid was refluxed in ethanol
with thionyl chloride to give the diester 6 in 97% yield.
Reduction of the ester with LiAlH4 afforded the diol 7, which
was subsequently mesylated with methanesulfonyl chloride and
Et3N to provide dimesylate 8. Conversion of 8 to diiodide 9
was accomplished using a modified Finklestein reaction involv-
ing NaI.
13, which was subjected to the previously described macrolac-
tamization conditions to afford lactam 14 in moderate yield.
At this point, selective hydrolysis of the C3 ester was attempted.
However, treatment with KOH selectively hydrolyzed the ester
at C10 to give acid 15 and gave as well a small amount of C3
epimerized 14, both effects presumably arising through eno-
lization at C3. Although a number of hydrolysis conditions were
used, we were unable to selectively hydrolyze the C3 ester.
Attempts at differentially protecting the C10 ester and then
hydrolyzing the ester at C3 were also unsuccessful owing, in
large part, to the insolubility of 15. With these difficulties
effectively preventing attainment of our target compound as well
as the inability of this route to access the (3S,10S) diastereomer
we sought a more general route to the target core structure 1.
Ring-Closing Olefin Metathesis Route. As an alternative
to the classic methods for the formation of lactams, cyclization
through the olefin was explored. In recent years, one of the
most successfully used methods for ring closure through an
olefin has been ring-closing olefin metathesis (RCM).49-52 In
addition to its successful application to a number of ring sizes,
Grubbs’ ruthenium alkylidene catalysts also have a very wide
range of functional group tolerance. Numerous examples are
known for cyclization to form lactams, lactones, and car-
bocycles.53-55 Yet, despite its success in the cyclization of small
and large ring systems, RCM has not been used extensively to
Alkylation of Scho¨llkopf’s auxiliary 10 with diiodide 9
proceeded with good yield and 84% de to afford iodide 11
(Scheme 4). After separation of the diastereomers by column
chromatography, dimethyl malonate was alkylated with the
desired (major) diastereomer 11 to afford diester 12. Hydrolysis
of the auxiliary in dilute aqueous HCl afforded the aminotriester
(39) Evans, P. A.; Holmes, A. B.; Russell, K. Tetrahedron Lett. 1992,
33, 6857-6858.
(40) Lease, T. G.; Shea, K. J. J. Am. Chem. Soc. 1993, 115, 2248-
2260.
(41) Koch, T.; Hesse, M. Synthesis 1992, 931-932.
(42) Ogliasrusso, M. A.; Wolfe, J. F. Synthesis of Lactones and Lactams;
John Wiley & Sons: New York, 1993.
(43) Paterson, I.; Mansuir, M. M. Tetrahedron 1985, 41, 3569-3624.
(44) Corey, E. J.; Weigel, L. O.; Floyd, D.; Bock, M. G. J. Am. Chem.
Soc. 1978, 100, 2916-2918.
(45) Corey, E. J.; Weigel, L. O.; Chamberlin, A. R.; Cho, H.; Hua, D.
H. J. Am. Chem. Soc. 1980, 102, 6613-6615.
(46) Nakatsuka, M.; Ragan, J. A.; Sammakia, T.; Smith, D. B.; Uehling,
D. E.; Schreiber, S. L. J. Am. Chem. Soc. 1990, 112, 5583-5601.
(47) Illuminati, G.; Mandolini, L. Acc. Chem. Res. 1981, 14, 95-102.
(48) Scho¨llkopf, U.; Neubauer, H.-J. Synthesis 1982, 861-863.
(49) Nguyen, S. T.; Johnson, L. K.; Grubbs, R. H.; Ziller, J. W. J. Am.
Chem. Soc. 1992, 114, 3974.
(50) Nguyen, S. T.; Grubbs, R. H.; Ziller, J. W. J. Am. Chem. Soc. 1993,
115, 9858-9859.
(51) Schwab, P.; Grubbs, R. H.; Ziller, J. W. J. Am. Chem. Soc. 1996,
118, 100-110.
(52) Schuster, M.; Blechert, S. Angew. Chem., Int. Ed. Engl. 1997, 36,
2036-2056.
(53) Grubbs, R. H.; Miller, S. J.; Fu, G. C. Acc. Chem. Res. 1995, 28,
446-452.
(54) Schmalz, H. G. Angew. Chem., Int. Ed. Engl. 1995, 34, 1833-
1836.

4338 J. Am. Chem. Soc., Vol. 120, No. 18, 1998
Fink et al.
Scheme 5
form medium-sized rings.56 However, we reasoned that RCM
might offer advantages in the formation of medium-sized rings,
because the size of ruthenium might relieve ring strain during
the cyclization.57
pattern around the olefin, the high functional group tolerance
and mildness of the reaction suggest the prospect of including
additional reactive functionality at the i+1 and i+2 positions,
contingent on the starting amino acids. Additionally, it has been
demonstrated that RCM is compatible with solid-phase organic
synthesis.58 Since the starting monomers are simple amino
acids, the wealth of methodology available for solid-phase
peptide synthesis could be used to create derivatives with any
peptide sequence desired. RCM, as the final step, would allow
for easy incorporation of our proposed mimic into peptide probes
or drug candidates.
With the core ring system in hand, tetrapeptide derivatives
were prepared that would be amenable to conformational
analysis and biological evaluation (Scheme 6). Removal of the
Boc group with TFA and coupling to Boc-L-phenylalanine using
PyBop afforded 22. Saponification of the methyl ester in dilute
NaOH was followed by PyBop coupling with L-methionamide
to afford the Substance P analogue 23. In addition to biological
data, we had hoped to analyze this derivative through X-ray
crystallography; however, at this time we have not successfully
grown crystals of these peptide derivatives of the macrolactam
that are suitable for analysis.
Compound 23 was only soluble in polar solvents (MeOH,
DMSO). To conduct solution conformational analysis, deriva-
tives that were soluble in weaker hydrogen bonding solvents
were prepared (Scheme 7). Removal of the Boc-protecting
group of 21 with trifluoroacetic acid, followed by PyBop
coupling to Boc-L-phenylalanine or Boc-L-alanyl-L-phenylala-
nine provided the tripeptide 22 and tetrapeptide 24, respectively.
Saponification of the methyl ester followed by peptide coupling
to L-phenylalanine methyl ester afforded the tetrapeptide 25a
and pentapeptide 25b, respectively, in good yield.
Tetrapeptide 25a and pentapeptide 25b proved to be soluble
in a wide variety of solvents, allowing conformational analysis
in both polar and nonpolar environments. The 1H NMR spectra
of 25a showed no spectral overlap, allowing complete assign-
ment of all peptide chain resonances and determination of
coupling constants. The pentapeptide 25b was prepared in order
Our synthetic approach to the core structure 1 using RCM is
shown in Scheme 5. Alkylation of Scho¨llkopf’s auxiliary 1048
with 1-bromo-3-butene afforded the alkene 16 in good yield
and good diastereoselectivity. The diastereomers were readily
separated by silica gel column chromatography. Hydrolysis of
the auxiliary gave a mixture of the desired methyl 2-amino-5-
hexenoate, along with valine methyl ester from the auxiliary.
This mixture proved to be inseparable either as the free amines
or the hydrochloride salts. However, upon treatment with
NaHCO3 and Boc2O, in a one pot procedure, the fully protected
amino acid 17 was readily separable from the Boc-valine methyl
ester in good yield, using argentation chromatography. Amino
acid 17 was deprotected with HCl in dioxane to afford amine
18 or saponified with dilute NaOH to afford the free acid 19.
These were then coupled, without purification, using the PyBop
reagent to afford the dipeptide 20 in 70% yield from 17.
Initial attempts to carry out RCM on dipeptide 20 in benzene
at ambient temperature or at reflux afforded only a mixture of
dimers with low conversion. The use of methylene chloride at
room temperature likewise afforded only a dimeric mixture.
However, at reflux temperature and high dilution (0.8 mM),
conversion of 20 to the ten-membered lactam 21 was achieved
in good yield, the amount of dimer being limited to only 10-
15%. The synthesis of lactam 21 through ring closing olefin
metathesis represents the first example of a ten-membered
lactam cyclized using RCM. Not only does RCM allow for all
of the necessary functionality of our proposed mimic but it also
allows access to the (3S,10S) diastereomer that was unattainable
through the macrolactamization route. Since this diastereomer
should be the most difficult to form (one substituent oriented
axially), the other diastereomer (3S,10R) should be readily
prepared in enantiomerically pure form, beginning from the
opposite enantiomer of Schollkopf’s auxiliary. Although the
success of RCM may be highly dependent on the substitution
(55) Snapper, M. L.; Tallarico, J. A.; Randall, M. L. J. Am. Chem. Soc.
1997, 119, 1478-1479.
(56) Furstner, A.; Muller, T. Synlett 1997, 1010-1012.
(57) Furstner, A.; Langemann, K. Synthesis 1997, 792-803.
(58) Miller, S. J.; Blackwell, H. E.; Grubbs, R. H. J. Am. Chem. Soc.
1996, 118, 9606-9614.

Synthesis of a Proposed Type I â-Turn Mimic
J. Am. Chem. Soc., Vol. 120, No. 18, 1998 4339
Scheme 6
Scheme 7
Figure 3. ORTEP diagram of (3S,10S)-21.
Table 4. Calculated, X-ray, and Ideal Type I â-Turn Torsion
Angles
rel energy
conf
type I
(kcal/mol) φ1 (deg) ψ1 (deg) φ2 (deg) ψ2 (deg)
-60
-30
-90
0
Calculated Lowest Energy Structures for 1a
chair-chair
chair-boat
0.0
-59
56
-24
-80
-101
-106
11
-12
2.04
to determine what effect an additional amino acid substituent
at the C-terminus would have on molecular conformation.
Conformational Analysis of the Macrolactam â-Turn
Mimic and Derivatives. There are a number of approaches
for the determination of peptide conformation. Typical methods
such as X-ray crystallography, NMR spectroscopy, and molec-
ular modeling all provide insight; however, peptide conformation
is strongly influenced by its environment. As such, the
conformation found in solution, in the solid state, or as a
theoretical minimum in a vacuum or in a simulated solvent is
not necessarily the true conformation, nor the one adopted when
the peptide is receptor bound. Integration of the results from
all three methods of analysis, along with evaluation of biological
activity, can provide the best idea of peptide conformation. Since
we were unable to prepare crystals of the peptide derivatives
of the macrolactam core 23 or 25 suitable for X-ray analysis,
compounds 21, 22, and 25a were analyzed through as many of
the above-mentioned techniques as possible, to try to gain insight
as to the final conformation.
The core macrolactam 21 gave crystals suitable for X-ray
analysis by the slow evaporation of a methanolic solution. Two
different conformations were observed in the unit cell, that of
a chair-chair and that of a chair-boat, related through a ring
flip at C3 (Figure 3).
This is nicely in accord with our initial modeling, as described
earlier, in which we found the chair-chair and chair-boat to
be the two lowest energy structures. Analysis of the torsion
X-ray Structures of 21
chair-chair
chair-boat
-107
120
-1
-64
-130
-131
23
-43
angles found in the X-ray structure and comparison with the
angles calculated from the lowest energy structures are shown
in Table 4.
Although the torsion angles were not as close as expected,
the central two angles ψ1 and φ2 are within 30° of the calculated
angles. When assigning a â-turn to a given class, Lewis has
used a criteria of less than 30° deviation from three of the
standard angles and deviation of one angle by less than 45°.9
When compared in this manner, the ten-membered lactam does
constrict the central two angles close to those of a type I â-turn,
although not ideally.
In solution, the 1H NMR spectrum of 21 showed some signals
that were unusually broad (Figure 4). Upon cooling to -30
°C, two different conformational isomers were observed,
although poor resolution prevented rigorous assignment of
resonances to any specific conformation. However, at 25 °C
the proton at C10 (H10) shows a large coupling constant (12.1
Hz) to the axial proton at C9 (H9ax), indicating that the methyl
ester was oriented equatorially, and a large coupling constant
to the ring amide proton H1 (10.0 Hz), indicating that the H10
and H1 protons are approximately 180° to one another. These
data would seem to indicate a chairlike conformation for this
portion of the molecule, which is in complete accord with the

4340 J. Am. Chem. Soc., Vol. 120, No. 18, 1998
Fink et al.
Table 6. Proposed Intramolecular Hydrogen-Bonding Pattern for
Tripeptide 22
∆δ/∆T
∆δ/∆C
(CDCl3)b
ppm/M
∆δ/∆solvent
(CD3CN-CDCl3)
ppm/solvent
Figure 4. 1H NMR spectrum of 21 at 25 °C.
(CD3CN)a
103 ppm/K
proton
Table 5. Analysis of Vicinal Coupling Constants for Tripeptide
22
H1
H1′
H3′
1.5
2.1
4.5
0.15
0.85
1.23
0.03
0.22
0.60
a ∆T ) 25-60 °C, 0.028 M, 500 MHz. b ∆C ) 0.25-0.001 M, 25
°C, 500 MHz.
Table 7. Analysis of Intramolecular Hydrogen-Bonding Pattern
for Tetrapeptide 25a
protons
J obsd, Hz
protons
J obsd, Hz
H3-H4ax
H3-H4eq
H3-H1′
2.90
4.90
8.24
H10-H9ax
H10-H9eq
H10-H1
11.90
2.20
10.10
results of the X-ray analysis. The resonances for H3 and H1′
are broadened by relatively slow conformational averaging
between a chair and boat conformation for this portion of the
molecule, again in accord with the X-ray analysis.
∆δ/∆T
∆δ/∆C
(CDCl3)b
ppm/M
∆δ/∆solvent
(CD3CN-CDCl3)
ppm/solvent
(CD3CN)a
103 ppm/K
proton
H1
1.6
3.2
3.8
2.5
0.07
0.25
0.40
0.17
0.00
0.17
0.72
0.28
Conformational analysis, through X-ray crystallography and
1H NMR, has shown that 21 may adopt either a chair-chair
or a chair-boat conformation, both in solution and in the solid
state. In addition, these results demonstrate that the methods
used for conformational analysis are self-consistent and that
we obserVe the same dynamics by either method. While such
a conformational equilibrium is undesirable in a peptidomimetic,
21 is a dipeptide without the potential intramolecular hydrogen
bond donors of a tetrapeptide â-turn mimic. The inclusion of
potential hydrogen bond donating groups at the i and i+3
positions may help to stabilize a single conformation.
H1′
H3′
H4′
a ∆T ) 25-65 °C, 0.031 M, 500 MHz. b ∆C ) (0.25-0.001 M),
25 °C, 500 MHz.
possibility for the formation of the same seven-membered ring,
it is likely that the carbonyl of the Boc group forms a
ten-membered ring as shown in Table 6.
It is well-known that type I â-turns show a preference for
amino acids in the i position that have strongly hydrogen bond
accepting side chains (Asn, Asp, Cys, Ser).9 These residues
stabilize the turn by forming an intramolecular hydrogen bond
with the main-chain nitrogen of the third residue, creating a
ten-membered ring. In these cases, the side chain and main
chain of residue i, along with the i+1 and i+2 residues, form
another turn type structure.9,59 A similar turn may be formed
between the carbonyl oxygen of the Boc-protecting group and
the amide proton H1 of the ten-membered lactam.
1
The H NMR spectrum of compound 22 showed a single
conformation at low temperature, indicating that some intramo-
lecular hydrogen bonding was probably taking place, stabilizing
one conformation. All resonances were now clearly resolved,
and coupling constants were obtained that matched well with
the expected chair-chair conformation. Proton H10 still showed
a large coupling to H9ax (11.9 Hz) and a large coupling to the
amide proton H1 (10.1 Hz); however, now H3 is resolved and
shows two small couplings (2.9 and 4.9 Hz) to vicinal neighbors,
indicating the amine substituent at C3 is in an axial position
(Table 5).
These results were encouraging, since they indicate that 22
seems to mimic the behaVior of naturally occurring type I
â-turns. This intramolecular hydrogen bond probably slows the
1
ring flip that led to broadening of the H NMR resonances for
To establish whether any intramolecular hydrogen bonds were
H3 and H1′ in compound 21 and stabilizes a single conformation
for 22. Through analysis of coupling constants, we are now
able to assign a chair-chair conformation to the ten-membered
lactam portion of 22.
1
forming, a series of H NMR experiments were conducted to
determine the effect of solvent, concentration, and temperature
on amide proton chemical shift. In all three cases (Table 6),
the amide proton of the ring (H1) shows a smaller chemical
shift dependence on temperature, concentration, and solvent.
The only possible intramolecular hydrogen bond acceptors
available to this hydrogen are the carbonyl oxygen of the Boc
group (forming a ten-membered ring) or the carbonyl of the i
residue (forming a seven-membered ring). Since no intramo-
lecular hydrogen bond was observed in 21, where there is the
A similar analysis of potential hydrogen bonding patterns for
tetrapeptide 25a was conducted, and the results are shown in
Table 7. Again, the chemical shift for H1 shows the smallest
dependence on concentration, temperature, and solvent. This
(59) Rees, D. C.; Lewis, M.; Lipscomb, W. N. J. Mol. Biol. 1983, 168,
367-387.

Synthesis of a Proposed Type I â-Turn Mimic
J. Am. Chem. Soc., Vol. 120, No. 18, 1998 4341
Table 8. Observed Coupling Constants and Calculated Dihedral
Angles
calcd
dihedral
angles
calcd
dihedral
angles
J obsd,
Hz
J obsd,
Hz
protons
protons
H3-H4ax
H3-H4eq
H3-H1′
2.93
5.12
6.59
13.80
3.42
49.10 H10-H9ax
-65.90 H10-H9eq
-138.19 H9ax-H8ax
H9ax-H8eq
11.53
1.83
14.89
2.93
177.06
-70.90
H4ax-H5ax
H4ax-H5eq
H10-H1
9.10
-166.57
indicates that the hydrogen bond formed between the Boc
carbonyl oxygen and the ring amide proton is not disrupted by
the addition of the i+3 residue. Although the existence of an
intramolecular hydrogen bond between the carbonyl of the i
residue and the NH of the i+3 residue is a good indication of
the formation of a â-turn, it is not necessary; many examples
of “open” turns exist.
Figure 5. Ten lowest energy structures calculated using Monte Carlo-
molecular mechanics and torsional restrictions derived form vicinal
coupling constants. The phenylalanine side chains and Boc-group have
been removed for clarity.
Table 9. Interresidue NOEs for Compound 25a
Analysis of vicinal coupling constants for 25a along the
peptide backbone indicated that the ten-membered lactam
remained in the expected chair-chair conformation (Table 8).
The protons at H3 showed small couplings to protons at H4,
indicating it was oriented in an equatorial manner, while the
proton H10 showed one large coupling (11.53 Hz) to the axial
proton H9ax and one small coupling (1.83 Hz) to the equatorial
proton H9eq. The remaining coupling constants around the ring
support a chair-chair conformation with typical gauche interac-
tions. Using the modified Karplus equations of Pardi,60 which
relates peptide dihedral angles to coupling constants, we
calculated dihedral angles from the vicinal coupling constants
of the ring and from the two JNR coupling constants (Table 8).
These dihedral angles were used as torsional constraints for
molecular modeling on the tetrapeptide mimic (see Experimental
Section).
proton percent
calcd
proton percent
calcd
The 10 lowest energy structures (within 1 kcal/mol of the
global minimum) are overlaid in Figure 5. It is immediately
apparent that a hydrogen bond is formed between the Boc
carbonyl oxygen and ring amide proton in the 10 lowest energy
structures. This is supported by the above 1H NMR data. The
calculated distance between this carbonyl oxygen and the
nitrogen of the ten-membered lactam is 2.5 Å. Equally apparent
is that the 10 lowest energy structures differ very little in the
torsional bonds constrained by the ten-membered ring. Minor
differences in energy arise from slight twists of the ψ2 angle
and rotations around the benzyl side chains of the i and i+3
phenylalanine residues (not shown).
Although vicinal coupling constants allow the determination
of dihedral angles and the geometry about a given residue, they
offer little information about the overall conformation of our
mimic, i.e., between residues. To investigate the relative
orientation between residues and possibly lend support to our
model, NOE experiments were conducted. The results of 1-D
NOE experiments are shown in Table 9.
pair
H1′-H5ax
H4′-H1
NOE
distance (Å)
pair
H4′-H10
H1′-H4eq
NOE
distance (Å)
4.75
1.02
2.1
2.2
2.66
1.12
3.0
2.9
the axial proton H5ax. This result indicates that H1′ is oriented
beneath the ten-membered ring, closer to H5ax than to H4eq. This
is strong support for our model in which the distance between
H1′ and H5ax is 2.1 Å and the distance between H1′ and H4eq is
2.8 Å. Assuming a trans-amide at the i residue, these NOEs
define the φ1 torsion angle. At the i+3 position, an NOE
between H4′ and H1, and between H4′ and H10, but not between
H4′ and H9eq, indicates the NH of the i+3 residue is oriented
toward the interior of the pseudo-ten-membered ring and above
the plane of the turn. Again, this is seen in the calculated model
in which both H1, H10, and H4′ are within 3 Å of one another,
but H9eq is 4 Å from H4′.
Thus, the experimental 1H NMR data, both in terms of
coupling constants and interresidue NOEs, support the calcu-
lated model. The torsion angles calculated from this model are
shown in Table 10, along with comparisons to an ideal type I
turn and our initial design modeling. Although there are
differences between the torsion angles initially calculated and
A number of informative interresidue NOEs are observed.
Of particular interest is the very strong NOE between H1′ and
(60) Pardi, A.; Billeter, M.; Wuthrich, K. J. Mol. Biol. 1984, 180, 741-
751.

4342 J. Am. Chem. Soc., Vol. 120, No. 18, 1998
Fink et al.
Table 10. Comparison of Ideal Type I â-Turn, 1a, and 25a
All reactions using water- or air-sensitive reagents were conducted
under an Ar atmosphere with dry solvents. Solvents were distilled
under N2 as follows: CH2Cl2 from CaH2, THF from sodium benzophe-
none ketyl, DMF from MgSO4, and hexanes from CaSO4. The
following reagents were purchased from commercial sources and puri-
fied before use as follows: triethylamine and diisopropylamine distilled
from CaH2, 1-bromo-3-butene fractionally distilled, and dimethyl
malonate fractionally distilled. All other reagents were purchased from
commercial suppliers and used without further purification.
conf
φ1 (deg)
ψ1 (deg)
φ2 (deg)
ψ2 (deg)
ideal type I
initial model 1a
25a
-60
-59
-82
-30
-24
-20
-90
-101
-107
0
11
-18
Table 11. Comparison of Vicinal Coupling Constants for 23 and
25a,b
Molecular Modeling. Conformational searches and energy mini-
mizations were performed using Macromodel version 5.5.38 The
Macromodel implementation of either the AMBER62 all-atom force field
or MM2 were used (respectively denoted AMBER* and MM2*). All
calculations were performed using the implicit water or CHCl3 GB/
SA solvation model of Still et al.63 Only small differences were noted
between force-fields used or between solvation models. The data
reported in this paper is using MM2* with a water solvation.
Conformational searches were performed using the Monte Carlo method
of Goodman and Still.36 All amide bonds and olefins were required to
be trans, those deviating more than 90° being rejected as energetically
improbable. For each search, 10 000 starting structures were generated
and minimized to an energy convergence of 0.001 (kcal/mol)/Å using
the truncated Newton-Raphson method implemented in Macromodel.
Duplicate structures and those greater than 20 kcal/mol above the global
minimum were discarded. In modeling the tetrapeptide 25a, the
dihedral angles constraints found in Table 8 were used with the FXTA
command in Batchmin version 5.5 with a force constant of 9999.0 kcal/
mol.
3JNR, 3JRâ1, 3JRâ2 (Hz)
compd (solvent)
H3
H10
23 (DMF)
25a (CD2Cl2)
25b (CD2Cl2)
7.57, 4.64, 3.17
6.60, 5.12, 2.93
6.90, 4.70, 3.10
9.69, 11.23, 1.76
9.10, 11.53, 1.83
9.50, 11.70, 2.10
those determined from modeling using a number of experimental
constraints, the differences are small, less than 30°.
Analysis of the coupling constants for the ring R-protons and
hydrogen bonding patterns for the tetrapeptide 23 and pen-
tapeptide 25b shows no significant differences from those of
25a in different solvents varying in polarity (Table 11). That
all derivatives maintain a similar conformation under a variety
of conditions lends support for the effectiveness of the 2-aza-
cyclodec-6-enone ring system 1 to stabilize a turn conformation.
Conclusions
(2R)-2-(3-Butenyl)-5-isopropyl-3,6-dimethoxy-2,5-dihydropyra-
zine (16). To a solution of Scho¨llkopf’s auxiliary 10 (2.0 g, 10.9 mmol)
in THF (50 mL) at -78 °C was added a 1.15 M solution of nBuLi
(12.3 mL, 14.1 mmol) in hexanes. The amber solution was allowed
to stir at -78 °C for 1 h. A solution of 1-bromo-3-butene (4.4 mL,
43.6 mmol) in THF (20 mL) was slowly added via syringe to the above
anion. The reaction mixture was allowed to stir for 12 h at -78 °C
and then warmed to room temperature for 2 h. Water (10 mL) was
added, and the reaction mixture was concentrated to a viscous oil. The
residue was extracted 6 × 10 mL with ether. The pooled organic
extracts were dried over MgSO4, filtered, and concentrated to a yellow
oil. Flash chromatography (7% ether/hexanes) afforded 16 as a clear
oil (2.07 g, 80%). Rf: 0.33 (7% ether/hexanes); [R]25D +7.89 (c ) 1,
CHCl3); IR (film) υ 2958.4, 2945.7, 2930.7, 1702.8, 1697.7, 1692.5,
In this study we have presented a rational design of a â-turn
mimic (1) capable of restricting the φ and ψ torsion angles of
a tetrapeptide to those found in type I â-turns. Ring-closing
olefin metathesis was shown to be an effectiVe method for
closing a medium-ring lactam under mild conditions with good
yields, representing the first example of cyclization to form a
ten-membered lactam using RCM. We anticipate that RCM will
greatly facilitate the synthesis of previously difficult-to-cyclize
medium-ring macrocycles. A combination of NMR experi-
ments, X-ray crystallography and molecular modeling was used
to analyze the conformation of a series of derivatives of 1. The
results of this analysis showed that the tetrapeptide deriVatiVe
25a restricted the central φ and ψ torsion angles to within 30°
of the ideal angles for a type I â-turn. This is, to our knowledge,
only the third example of a type I â-turn mimic reported in
the literature and the first to allow for Versatile incorporation
into biological systems. We believe that this system may serve
as a useful conformational constraint when incorporated into
other biologically significant peptides where a â-turn is thought
to be important.61
1
1687.6, 1681.7; H NMR (CDCl3, 500 MHz) δ 5.80 (ddt, 1H, J )
17.0, 10.3, 6.6 Hz), 5.02 (ddt, 1H, J ) 17.2, 2.2, 1.5 Hz), 4.93 (ddt,
1H, J ) 10.3, 2.5, 1.3 Hz), 4.0 (dt, 1H, J ) 6.4, 4.0 Hz), 3.91 (t, 1H,
J ) 3.5 Hz), 3.67 (s, 3H), 3.66 (s, 3H), 2.24 (sept d, 1H, J ) 6.7, 3.3
Hz), 2.04 (m, 2H), 1.90 (dddd, 1H, J ) 14.1, 9.5, 6.4, 4.2 Hz), 1.76
(dddd, 1H, J ) 13.2, 10.1, 6.6, 5.3 Hz), 1.02 (d, 3H, J ) 7.0 Hz), 0.67
(d, 3H, J ) 6.8 Hz); 13C NMR (CDCl3, 100 MHz) 163.8, 163.6, 138.5,
114.5, 60.8, 54.9, 52.3, 33.4, 31.7, 28.9, 19.0, 16.6; MS (EI, 70 eV)
m/z (relative intensity, %) 238.1(11), 223.1 (41), 195.1 (100). Anal.
Calcd for C13H22N2O2: C, 65.52; H, 9.30; N, 11.7. Found: C, 65.25;
H, 9.32; N, 11.83.
Experimental Section
Minor diastereomer Rf: 0.28 (7% ether/hexanes); [R]25 -83.13
D
1
(c ) 1, CHCl3); H NMR (CDCl3, 500 MHz) δ 5.84 (ddt, 1H, J )
General Procedures. Reaction progress was monitored by analyti-
cal thin-layer chromatography (TLC), using 0.25 mm silica gel glass-
backed plates with a UV sensitive indicator. Flash chromatography
was performed using 32-63 µm silica gel packing unless otherwise
noted. Visualization was accomplished by potassium permanganate
spray reagent, iodine vapors, or UV illumination (254 nm). 1H NMR
and 13C NMR spectra were recorded at 400, 500, or 750 MHz. HMBC
and HMQC spectra were recorded at 500 MHz. Chemical shifts (δ)
are reported as parts per million from an internal tetramethylsilane
(TMS) standard in the solvent indicated at 25 °C unless otherwise noted.
Coupling constants were determined from 1H NMR, {1H}13C decoupled
NMR, or through phase-sensitive DQF NMR spectroscopy. High-
pressure liquid chromatography (HPLC) was performed using ultraviolet
detection at 254 nm. For analytical purposes a 4 mm × 32 cm C-18
column was used with the solvent system indicated. Elemental analysis
was performed by the Microanalytical Service Laboratory at the
University of Illinois.
17.0, 10.4, 6.6 Hz), 5.03 (dq, 1H, J ) 17.0, 1.7 Hz), 4.95 (ddt, 1H, J
) 10.3, 2.2, 1.1 Hz), 3.97 (dt, 1H, J ) 8.6, 4.8 Hz), 3.91 (dd, 1H, J )
4.8, 3.8 Hz), 3.66 (s, 3H), 3.65 (s, 3H), 2.20 (m, 3H), 1.96 (dddd, 1H,
J ) 14.1, 10.1, 6.4, 4.4 Hz), 1.55 (m, 1H), 1.04 (d, 3H, J ) 6.8 Hz),
0.71 (d, 3H, J ) 6.8 Hz); 13C NMR (CDCl3, 100 MHz) 163.71, 163.02,
60.83, 55.08, 52.18, 34.77, 31.21, 30.21, 19.51, 17.40.
Methyl (2R)-2-(tert-Butoxycarbonylamino)-5-hexenoate (17). To
bis-lactim ether 16 (2.0 g, 8.40 mmol) at room temperature was added
(61) Compound 23 was tested for biological activity in a competitive
inhibition assay versus radiolabeled Substance P. Tetrapeptides 23 showed
no activity, which may mean that Substance P adopts a different conforma-
tion at the receptor than that adopted by our proposed mimic.
(62) Weiner, S. J.; Kollman, P. A.; Case, D. A.; Singh, U. C.; Ghio, C.;
Alagona, G.; Profeta, S.; Weiner, P. J. Am. Chem. Soc. 1984, 106, 765-
784.
(63) Still, W. C.; Tempczyk, A.; Hawely, R. C.; Hendrickson, T. J. Am.
Chem. Soc. 1990, 112, 6127-6129.

Synthesis of a Proposed Type I â-Turn Mimic
J. Am. Chem. Soc., Vol. 120, No. 18, 1998 4343
0.25 N HCl (84 mL, 21 mmol). The suspension was allowed to stir at
room temperature for 3 days. Solid NaHCO3 (3.53 g, 42 mmol) was
added in two equal portions at 0 °C. Dioxane (80 mL) was added
followed by Boc2O (2.2 g, 10.08 mmol) at 0 °C. Reaction mixture
was allowed to stir for 18 h. Dioxane was removed under reduced
pressure, and residue was extracted 10 × 20 mL with EtOAc. Organic
extracts were pooled, dried over Na2SO4, filtered, and concentrated to
a clear/colorless oil. Flash chromatography on 15% AgNO3 impreg-
nated silica gel (15% ether/hexanes) afforded 17 as a colorless oil (1.4
mg, 0.17 mmol) in CH2Cl2 (5 mL) at 0 °C was added Et3SiH (0.03
mL, 0.17 mmol). TFA (1 mL) was added, and the clear yellow solution
was warmed to room temperature and stirred for 2 h. The reaction
mixture was concentrated under reduced pressure and dried under
vacuum for 24 h. The crude amine salt was taken up in CH2Cl2 (5
mL) at room temperature and Boc-Phe-ONa (73.2 mg, 0.26 mmol) was
added, followed by DIEA (0.09 mL, 0.51 mmol). To this homogeneous
solution was added PyBop (132 mg, 0.26 mmol). After 1 h at room
temperature, a fine white precipitate developed. Stirring was continued
under Ar overnight, after which the reaction mixture was concentrated
to a white solid. The residue was taken up in EtOAc (20 mL) and
washed consecutively with 5% KHSO4 (10 mL), 5% NaHCO3 (10 mL),
and brine (10 mL) and dried over Na2SO4. The organics were then
filtered and concentrated to yield a solid which, upon flash chroma-
tography (45:50:5 EtOAc/hexanes/MeOH) afforded 22 (61.5 mg 76%).
Rf: 0.28 (40% EtOAc/hexanes); mp: 168-170 °C (EtOAc/hexanes);
[R]25D +27.7° (c ) 1.0, CHCl3); RP-HPLC Whatman ODS2 C-18 70:
30 ACN:H2O(0.1% TFA) tR ) 6.86 min; IR (CHCl3 solution) υ 3428.6,
g, 70%). Rf: 0.30 (20% ether/hexanes); [R]25D -22.4 (c ) 1, MeOH);
1
lit.61 [R]25 -17 (c ) 1.2, MeOH); H NMR (CDCl3, 500 MHz) δ
D
5.85 (ddt, 1H, J ) 17.1, 10.3, 6.6 Hz), 5.10 (dq, 1H, J ) 17.1, 3.2, 1.5
Hz), 5.08 (bs, 1H), 5.04 (dq, 1H, J ) 10.3, 2.8, 1.2 Hz), 2.14 (q, 2H,
J ) 7.3 Hz), 1.92 (dtd, 1H, J ) 13.7, 8.1, 5.1 Hz), 1.73 (dtd, 1H, J )
13.4, 8.3, 6.6, Hz), 1.42 (s, 9H); 13C NMR (CDCl3, 100 MHz) 173.33,
155.33, 136.96, 115.68, 79.86, 52.95, 52.23, 31.95,293.45, 28.28; MS
(EI, 70 EV) m/z (relative intensity) 238.1 (M+, 20), 223 (25), 207 (14),
195 (100), 181 (15), 166 (25); Anal. Calcd for C12H21NO4: C, 58.59;
H, 8.73; N, 5.69. Found: C, 58.65; H, 8.92; N, 5.63.
1
3009.8, 1738.2, 1686.1, 1680.6, 1497.9; H NMR (CDCl3, 500 MHz)
δ 7.26-7.36 (m, 5H), 6.66 (d, 1H, J ) 7.3 Hz), 6.35 (d, 1H, J ) 9.7
Hz), 5.34 (ddd, 1H, J ) 15.0, 11.2, 3.1 Hz), 5.16 (ddd, 1H, J ) 14.1,
10.6, 2.9 Hz), 4.99 (d, 1H, J ) 5.9 Hz), 4.74 (ddd, 1H, J ) 11.9,
10.07, 2.20 Hz), 4.52 (ddd, 1H, J ) 8.24, 4.94, 2.93 Hz), 4.40 (X of
ABX, 1H, JAX ) 6.6, JBX ) 7.5, JNHX ) 5.9 Hz), 3.70 (s, 3H), 3.16
(AB of ABX, 2H, JAB ) 14.3, JAX ) 6.6, JBX ) 7.5 Hz), 2.42 (tt, 1H,
J ) 14.1, 3.5 Hz), 2.25 (m, 1H), 2.11 (m, 2H), 1.99 (m, 1H), 1.85
(tdd, 1H, J ) 13.9, 11.0, 3.1 Hz), 1.55 (m, 2H), 1.44 (s, 9H); 13C NMR
(CDCl3, 125 MHz) 172.18, 171.66, 171.24, 156.53,136.52,131.61,
129.13, 129.05, 128.12, 127.35, 81.10, 56.40, 52.90, 52.28, 51.65, 37.27,
31.32, 30.33, 29.81, 28.36 28.34; MS (FAB) m/z (relative intensity,
%) 474.1 (MH+, 100), 418.1 (95), 374.1 (50), 308 (50); HRMS calcd
for C25H36N3O6: 474.2605, found 474.2604. Anal. Calcd for
C25H35N3O6: C, 63.40; H, 7.45; N, 8.87. Found: C, 63.22; H, 7.55;
N, 8.52.
Methyl 2-((2-(tert-Butoxycarbonylamino)-5-hexenoyl)amino)-5-
hexenoate (20). To a solution of 17 (470 mg, 1.90 mmol) in THF (10
mL) at 0 °C was added a 0.25 N solution of NaOH (23 mL, 5.80 mmol).
The reaction mixture was warmed to room temperature and stirred for
6 h. After neutralizing with 0.50 N HCl, the reaction mixture was
extracted 5 × 10 mL with EtOAc. The organic extracts were dried
over Na2SO4, filtered, and concentrated to a yellow oil. In a separate
flask was placed 17 (470 mg, 1.90 mmol) and dioxane (5 mL). The
flask was cooled to 0 °C and Et3SiH (0.3 mL, 1.9 mmol) was added.
To this clear solution was added 4 N HCl in dioxane (3 mL). The re-
action mixture was warmed to room temperature for 2 h and concen-
trated to a white solid. After drying both the free acid and amine salt
under vacuum for 24 h, they were combined and dissolved in CH2Cl2
(10 mL). DIEA (0.99 mL, 5.7 mmol) was added, followed by PyBop.
The reaction was allowed to stir for 9 h and then concentrated to an
orange oil. The residue was taken up in EtOAc and washed consecu-
tively with 5% KHSO4 (10 mL), 5% NaHCO3 (10 mL), brine (10 mL),
and dried over Na2SO4. After filtering and concentrating to an orange
oil, flash chromatography (25% EtOAc/hexanes) afforded 20 as a white
solid (476 mg, 71%). Rf: 0.32 (25% EtOAc/hexanes); mp: 131-132
°C (EtOAc/hexanes); [R]25D -5.84 (c ) 1.15, CHCl3); 1H NMR (CDCl3,
500 MHz) δ 6.60 (d, 1H, 8.1 Hz), 5.17 (d, 1H, J ) 7.4 Hz), 5.01 (dt,
2H, J ) 17.4, 3.2, 1.7 Hz), 4.96 (ddq, 2H, J ) 10.1, 5.3, 3.1, 1.1 Hz),
4.59 (dt, 1H, J ) 7.9, 4.9 Hz), 4.09 (q, 1H, J ) 7.2), 3.72 (s, 3H), 2.08
(m, 4H), 1.93 (ddt, 1H, J ) 13.4, 8.2, 5.1 Hz), 1.85 (m, 1H), 1.76
(dtd, 1H, J ) 13.9, 7.9, 6.2 Hz), 1.69 (dq, 1H, J ) 15.4, 7.7 Hz), 1.42
(s, 9H); 13C NMR (CDCl3, 100 MHz) 172.53, 171.86, 155.63, 137.26,
136.76, 115.84, 115.71, 79.99, 53.84, 52.31, 51.63, 31.42, 29.63, 29.35,
28.28, 28.26; MS (FAB) m/z (relative intensity, %) 355.2 (MH+, 80),
299.2 (100), 255.2 (90), 243.2 (20). Anal. Calcd for C18H30N2O5: C,
61.00; H, 8.53; N, 7.90. Found: C, 60.78; H, 8.74; N, 7.82.
(3S,10S)-(6E)-10-((S)-2-((N-tert-Butoxycarbonyl)phenylalanyl)-
amino-3-(S)-methionamide)carboxy-2-azacyclodec-6-en-1-one (23).
Compound 22 (24 mg, 0.05 mmol) was dissolved in MeOH (5 mL)
and cooled to 0 °C. A 0.25 N solution of NaOH (0.4 mL, 0.10 mmol)
was slowly added. The reaction mixture was slowly warmed to room
temperature and stirred for 8 h. After the disappearance of starting
material was noted by TLC the reaction mixture was concentrated to
a turbid suspension. The suspension was extracted 3 × 10 mL with
EtOAc, the organic layers were dried over Na2SO4, filtered, and
concentrated to a clear film. After drying overnight under vacuum,
the film was dissolved in CH2Cl2 (5 mL) and L-methionamide
hydrochloride (13.8 mg, 0.075 mmol) was added. DMF (0.2 mL) was
added to solubilize the methionine. DIEA (0.03 mL, 0.15 mmol), and
PyBop (31.2 mg, 0.06 mmol) were added. The reaction was stirred
overnight and then concentrated to a white solid. The solid was taken
up in EtOAc (10 mL) and washed consecutively with 5% KHSO4 (5
mL), 5% NaHCO3 (5 mL), and brine (5 mL) and dried over Na2SO4.
After filtering and concentrating, flash chromatography (5% MeOH/
CH2Cl2) afforded a white solid that was recrystallized from EtOAc/
petroleum ether to give 22 mg (75%) of 23. mp 194-195.5 °C; Rf:
0.35 (7% MeOH/CHCl3); tR ) 14.81 min (70:30, MeOH/H2O, 0.1%
TFA); 1H NMR (DMF-d6, 500 MHz) δ 8.25 (d, 1H, J ) 6.0 Hz), 7.68
(d, 1H, J ) 7.9 Hz), 7.43 (br s, 2H), 7.21-7.34 (m, 6H), 7.11 (br s,
1H), 7.08 (d, 1H, J ) 7.5 Hz), 5.61 (ddd, 1H, J ) 13.5, 10.6, 1.8 Hz),
5.25 (ddd, 1H, J ) 14.1, 10.4, 1.8 Hz), 4.59 (ddd, 1H, J ) 11.2, 9.7,
1.7 Hz), 4.51 (ddd, 1H, J ) 9.7, 7.5, 4.6 Hz), 4.42 (ddd, 1H, J ) 7.6,
4.6, 3.2 Hz), 4.34 (dt, 1H, J ) 7.6, 4.1 Hz), 3.21 (dd, 1H J ) 14.3, 4.4
Hz), 2.99 (dd, 1H, J ) 14.1, 10.0 Hz), 2.51 (m, 1H), 2.38 (m, 1H),
2.26 (m, 2H), 2.02-2.08 (m, 7H), 1.91 (dtd, 1H, J ) 13.9, 9.2, 5.3
Hz), 1.65 (m, 2H), 1.36 (s, 9H); 13C NMR (DMF-d6, 125 MHz) 173.8,
173.6, 172.8, 172.3, 157.0, 139.1, 133.2, 129.9, 128.9, 128.5, 126.9,
79.3, 56.7, 54.8, 53.4, 52.9, 37.5, 33.0, 32.4, 30.5, 30.4, 28.7, 28.5,
28.4, 15.0; MS (FAB) m/z (relative intensity, %) 628 (MK+, 40), 612
(MNa+, 40), 590 (MH+, 70), 490 (50), 386 (100), 342 (60); HRMS
calcd for C29H45N5O6S 590.3009, found 590.3012. Anal. Calcd for
C29H44N5O6S: C, 59.06; H, 7.35; N, 11.87; S, 5.44. Found: C, 58.66;
H, 7.59; N, 11.61; S, 5.32.
(3S,10S)-(6E)-10-((N-tert-Butoxycarbonyl)amino)-3-methoxycar-
bonyl-2-azacyclodec-6-en-1-one (21). To 500 mL of degassed
CH2Cl2 at reflux was added (PCy3)2Cl2Ru benzylidene (35 mg, 0.042
mmol). To this light purple solution at reflux was added a solution of
20 in degassed CH2Cl2 (10 mL) via cannula. The purple color
immediately faded to orange, and the reaction was allowed to reflux
for 18 h. The reaction mixture was then exposed to air until black,
and the solvent was removed under vacuum to afford a black solid.
Flash chromatography (20% acetone/hexanes) afforded 21 as an off-
white solid (62.2 mg, 68%). mp 132-133 °C (EtOAc/hexanes); Rf:
0.28 (40% EtOAc/hexanes); IR (CH3Cl solution) υ 1739.4, 1734.1,
1
1729.8, 1723.1, 1718.2, 1700.6; H NMR (CDCl3, 500 MHz) δ 6.54
(d, 1H, J ) 10.0 Hz), 5.39 (m, 2H), 4.92 (bs, 1H), 4.79 (ddd, 1H, J )
12.1, 10.1, 2.4 Hz), 4.21 (bs, 1H), 3.70 (s, 3H), 2.44 (m, 1H), 2.32 (m,
1H), 2.18 (m, 2H), 2.06 (ddq, 1H, J ) 14.5, 5.2, 2.6 Hz), 2.0 (m, 1H),
1.71 (m, 1H), 1.59 (dtd, 1H, J ) 14.7, 12.1, 3.1 Hz), 1.48 (s, 9H); MS
(FAB) m/z (relative intensity, %) 327.2 (MH+, 70), 307.1 (40), 271.1-
(100). Anal. Calcd for C16H26N2O5: C, 58.88; H, 8.03; N, 8.58.
Found: C, 58.87; H, 7.96; N, 8.58.
(3S,10S)-(6E)-10-((2S)-2-((N-tert-Butoxycarbonyl)phenylalanyl)-
amino-3-methoxycarbonyl-2-azacyclodec-6-en-1-one (22). To 21 (56

4344 J. Am. Chem. Soc., Vol. 120, No. 18, 1998
Fink et al.
(3S,10S)-(6E)-10-((S)-((N-tert-Butoxycarbonyl)alanyl)-(S)-
phenylalanyl)amino-3-methoxycarbonyl-2-azacyclodec-6-en-1-one
(24). Compound 24 was prepared from 21 and Boc-Ala-Phe-OH in a
Table 12. Crystallographic Data for 21
empirical
formula
formula mass 326.39
C16H26N2O5 vol, Å3
1795.4(4)
Z
4
1
absorption coeffs
0.090 mm-1
198(2) K
manner analogous to 22 (51 mg, 55%); H NMR (CD2Cl2, 500 MHz)
δ 7.20-7.40 (m, 5H), 6.97 (d, 1H, J ) 9.3 Hz), 6.87 (d, 1H, J ) 8.1
Hz), 6.73 (d, 1H, J ) 4.9 Hz), 5.48 (ddd, 1H, J ) 14.6, 11.1, 2.2 Hz),
5.31 (dddd, 1H, J ) 14.8, 10.6, 3.8, 1.7 Hz), 5.14 (d, 1H, J ) 2.8 Hz),
4.67 (ddd, 1H, J ) 11.7, 9.7, 2.0 Hz), 4.94 (X of ABX 1H, JNX ) 4.9,
JAX ) 7.1, JBX ) 4.1 Hz), 4.45 (ddd, 1H, J ) 8.1, 5.1, 3.1 Hz), 4.02
crystal system monoclinic T (K)
space group
a, Å
b, Å
P2(1)
9.3911(11)
10.0489(12) R indices (all data)
reflcns collected
9469 [R(int) )
0.1248]
R1 ) 0.2603,
wR2 ) 0.3714
c, Å
9.509(2)
90
102.782(3)
90
R, deg
â, deg
γ, deg
goodness-of-fit on F2 1.151
F
(dq, 2H, J ) 7.3, 3.7 Hz), 3.69 (s, 3H), 3.14 (AB of ABX 2H, JAB
)
calc, Mg/m3
µ, mm-1
1.207
0.090
14.4, JAX ) 7.1, JBX ) 4.9 Hz), 2.49 (tdd, 1H, J ) 14.3, 10.6, 3.7 Hz),
2.38 (tt, 1H, J ) 14.1, 3.5 Hz), 2.31 (m, 1H), 2.38 (tdd, 1H, J ) 13.5,
11.0, 2.7 Hz), 2.0 (m, 1H), 1.96 (ddq, 1H, J ) 14.1, 4.8, 2.4 Hz), 1.72
(tdd, 1H, J ) 14.5, 12.3, 2.8 Hz), 1.52 (ddt, 1H, J ) 13.9, 4.8, 3.1
Hz), 1.36 (s, 9H), 1.33 (d, 3H, J ) 7.1 Hz); 13C NMR (CD2Cl2, 125
MHz) 174.4, 172.4, 172.0, 170.3, 156.1, 136.0, 132.2, 129.1, 128.9,
128.0, 127.5, 81.0, 55.0, 53.9, 53.4, 52.2, 52.0, 36.4, 32.1, 30.1, 29.2,
27.8, 27.1, 17.2; MS (Type: FAB) m/z (relative intensity, %) 545.2
(MH+, 30), 489.2 (25), 445.2 (20), 386.2 (20), 338.3 (25), 284.3 (15),
227.1 (32), 199.2 (20), 120.0 (100); HRMS calcd for C28H41N4O7
545.2974, found 545.2975.
by 3d-profile analysis using SAINT and corrected for Lorentz-polari-
zation effects and for absorption. Scattering factors and anomalous
dispersion terms were taken from standard tables.64 The structure was
solved by direct methods;65 the correct C, N, O atom positions were
deduced from a vector/electron density map. Subsequent cycles of
isotropic least-squares refinements followed by an unweighted differ-
ence Fourier synthesis revealed positions for the remaining non-H
atoms. Methyl H atom positions, 4-CH∼3∼, were optimized by rota-
tion about 2-C bonds with idealized C-H and H- -H distances. Re-
maining H atoms were included as fixed idealized contributors. H atom
U’s were assigned as 1.2 times Ueq of adjacent C atoms. Non-H atoms
were refined with anisotropic thermal coefficients. Successful con-
vergence of the full-matrix least-squares refinement on F∧2∧ was indi-
cated by the maximum shift/error for the last cycle. The highest peaks
in the final difference Fourier map were in the vicinity of the C, N,
and O atoms; the final map had no other significant features. A final
analysis of variance between observed and calculated structure factors
showed no dependence on amplitude or resolution. Full details of the
crystallographic results are included in the Supporting Information.
(3S,10S)-(6E)-10-((S)-((N-tert-Butoxycarbonyl)phenylalanyl)amino-
3-((S)-carboxymethylphenylalanyl)carboxy-2-azacyclodec-6-en-1-
one (25a). Compound 25a was prepared from 22 and L-phenylalanine
in a manner analogous to 23 (44 mg, 79%). Crystallized from EtOAc/
1
hexanes, mp 114-115 °C; Rf: 0.34 (70% EtOAc/hexanes); H NMR
(CD2Cl2, 500 MHz) δ 7.19-7.39 (m, 5H), 6.78 (d, 1H, J ) 7.9 Hz),
6.70 (d, 1H, J ) 5.3 Hz), 6.58 (d, 1H, J ) 7.9 Hz), 5.30 (2H, m), 4.98
(d, 1H, J ) 2.6 Hz), 4.70 (dt, 1H, J ) 7.5, 5.5 Hz), 4.41 (ddd, 1H, J
) 11.5, 9.2, 1.8 Hz), 4.32 (ddd, 1H, J ) 6.6, 5.1, 2.9 Hz), 4.20 (ddd,
1H, J ) 8.4, 4.9, 3.7 Hz), 3.69 (s, 3H), 2.36 (tdd, 1H, J ) 14.4, 7.3,
4.0 Hz), 2.28 (m, 1H), 2.15-2.20 (m, 2H), 2.05 (m, 1H), 1.94 (m,
1H), 1.62 (ddt, 1H, J ) 14.4, 7.3, 3.4 Hz), 1.50 (ddd, 1H, J ) 14.9,
12.7, 2.9 Hz), 1.48 (s, 9H); 13C NMR (CD2Cl2, 125 MHz) 173.05,
172.04, 171.90, 171.58, 157.33, 136.94, 136.39, 133.23, 129.70, 129.37
(2C), 128.79, 127.82, 127.79, 127.08, 81.71, 57.46, 54.94, 53.87, 53.80,
52.47, 38.34, 37.60, 32.22, 29.26, 29.21, 28.98, 28.50; MS (Type: FAB)
m/z (relative intensity, %) 659 (MK+, 20), 643 (MNa+, 20), 621 (MH+,
20), 605 (20), 521 (20), 386 (40), 358 (35), 342 (20), 180 (60), 120
(100); HRMS calcd for C34H44N4O7 621.3290, found 621.3288.
(3S,10S)-(6E)-10-((S)-((N-tert-Butoxycarbonyl)alanyl)-(S)-
phenylalanyl)amino-3-((S)-carboxymethylphenylalanyl)carboxy-2-
azacyclodec-6-en-1-one (25b). Compound 25b was prepared from 24
and L-phenylalanine in a manner analogous to 23 (48 mg, 78%). 1H
NMR (CD2Cl2, 750 MHz) δ 7.15-7.40 (m, 10H), 6.97 (d, 1H, J )
7.0 Hz), 6.96 (d, 1H, J ) 9.5 Hz), 6.87 (d, 1H, J ) 7.9 Hz), 6.71 (d,
1H, J ) 5.5 Hz), 5.47 (ddd, 1H, J ) 14.5, 11.1, 2.1 Hz), 5.27 (dddd,
1H, J ) 15.0, 11.0, 3.6, 1.6 Hz), 5.12 (d, 1H, J ) 2.7 Hz), 4.67 (X of
ABX 1H, JNX ) 7.9, JAX ) 5.5, JBX ) 7.6 Hz), 4.36 (ddt, 1H, J )
11.7, 9.5, 2.2 Hz), 4.35 (ddd, 1H, J ) 7.1, 3.3, 4.5 Hz), 4.27 (X of
ABX 1H, JNX ) 4.6, JAX ) 7.1, JBX ) 4.9 Hz), 3.98 (dq, 2H, J ) 7.2,
2.7 Hz), 3.66 (s, 3H), 3.14 (AB of ABX 2H, JAB ) 14.4, JAX ) 7.1,
Acknowledgment. Support for this work was provided by
the Eli Lilly Corporation and the National Institutes of Health
(PHS 5R01 DK15556 and PHS 5R01 CA33091). We thank Prof.
Peter A. Petillo and Dr. Howard Robinson for many helpful
discussions and Dr. Bruce Gitter and Tiffanie Lamar for
conducting the Substance P binding assays. Dr. Feng Lin and
Dr. Vera Mainz of the University of Illinois Molecular Spec-
troscopy Lab are gratefully acknowledged for assistance in
conducting NMR spectroscopic experiments and Dr. Scott
Wilson and Teresa Prussak-Wieckowski for obtaining X-ray
crystallographic data. NMR spectra were obtained in the Varian
Oxford Instrument Center for Excellence in NMR Laboratory.
Funding for this instrumentation was provided in part from the
W. M. Keck Foundation, the National Institutes of Health (PHS
1 S10 RR104444-01), and the National Science Foundation
(NSF CHE 96-10502). Mass spectra were obtained on instru-
ments supported by grants from the National Institute of General
Medical Sciences (GM 27029), the National Institutes of Heath
(RR 01575), and the National Science Foundation (PCM
8121494).
JBX ) 4.9 Hz), 3.03 (AB of ABX 2H, JAB ) 13.9, JAX ) 5.5, JBX
)
7.6 Hz), 2.60 (tdd, 1H, J ) 14.3, 11.6, 3.4 Hz), 2.31 (tt, 1H, J ) 14.4,
4.4 Hz), 2.28 (m, 2H), 2.08 (tdd, 1H, J ) 14.5, 4.9, 2.6 Hz), 2.0 (m,
2H), 1.34 (s, 9H), 1.33 (d, 3H, J ) 7.3 Hz); 13C NMR (CD2Cl2, 125
MHz) 174.6, 173.0, 171.7, 171.6, 171.0, 156.4, 136.7, 135.7, 133.1,
129.4, 129.1, 129.0, 128.3, 127.9, 127.6, 126.6, 81.2, 55.4, 54.9, 53.5,
53.2, 52.6, 52.1, 38.1, 36.5, 32.5, 29.1, 28.7, 27.9, 27.2, 17.1; MS
(Type: FAB) m/z (relative intensity, %) 692.3 (MH+, 20), 614 (25),
592.3 (30), 515.3 (25), 457.3 (30), 258.2 (100), 187 (70); HRMS calcd
for C37H50N5O8 692.3662, found 692.3659.
Supporting Information Available: 1H and 13C NMR
spectra for compounds 11, 13, 14, 22, 24, 25a (Figures S1-
S12); Tables S1-S6 listing full crystallographic information,
atomic coordinates, anisotropic displacement parameters, and
intramolecular bond distances and angles; complete ORTEP
plots for compound 21 (Figures S13-S15), and complete
experimental procedures and characterization for compounds
6-9, 11-15, 24, 25a, and 25b (S16-S22) (35 pages). See
any current masthead page for ordering information and Web
access instructions.
X-ray Crystallography for 21. A suitable crystal was mounted
using oil (Paratone-N, Exxon) to a thin glass fiber. The sample was
bound by faces (1 0 -1), (-1 0 1), (1 0 1), (-1 1 0), (0 1 1), and (0 -
1 -1). Distances from the crystal center to these facial boundaries
were 0.010, 0.010, 0.080, 0.080, 0.190, and 0.190 mm, respectively.
Crystal and refinement details are given in Table 12. Systematic
conditions suggested the ambiguous/unambiguous space group P2(1).
Standard intensities monitored during frame collection showed no
decay; decay correction was not applied. Intensity data were reduced
JA974023Y
(64) Wilson, A. J. C. International Tables for X-ray Crystallography;
Wilson, A. J. C., Ed.; Kluwer Academic Publishers: Dordrecht, 1992; Vol.
C, pp 500-502, 219-222.
(65) Sheldrick, G. M. Acta Crystallogr. 1990, A46, 467-473.