Formation of the 7-oxa-1,4,10-triazatricyclo[8.2.25,12]tetradecane-2,14-dione ring system: Misrouted synthesis of a peptidomimetic

Article abstract of DOI:10.1021/jo961114p

An attempted synthesis of the tricyclic peptidomimetic 1, designed to imitate a β-turn tripeptide in tendamistat, afforded instead the 6,6,8-ring system of 2. The key step in the synthesis entailed acylation of the hindered α,α'-disubstituted morpholine 4.2, which was approached by acylative ring opening of the 3,6-oxazabicyclo[4.2.0]octane 4.3. However, transannular rather than exocyclic cleavage occurred, giving the 1,6-oxazacyclooctane isomer 4.5. Subsequent ring closures to form the bi- and tricyclic intermediates 7.3 and 8.5 were difficult because of the strain being built into the ring systems. After completion of the synthesis, the structures of the intermediates and final product were elucidated by NMR, with three-bond, heteronuclear multiple-bond correlation experiments providing unambiguous evidence for the ring connectivity, and by molecular modeling, which allowed assignment of the stereochemistry. Compound 2 is a modest inhibitor of the target enzyme α-amylase (K(i) = 170 μM in 5% DMSO/water), binding with similar affinity to the tripeptide Ac-Trp-Arg-Tyr-OMe. Although the side-chain attachment points in the ring system of 2 correspond closely to the relative Cα-positions in tendamistat (rmsd = 0.24 A), the alignment of the Cα-Cβ bonds is poor, illustrating the importance of side-chain orientation in a peptidomimetic.

Full text of DOI:10.1021/jo961114p

J . Org. Chem. 1996, 61, 7681-7696
7681
F or m a tion of th e
7-Oxa -1,4,10-tr ia za tr icyclo[8.2.25,12]tetr a d eca n e-2,14-d ion e Rin g
System : Misr ou ted Syn th esis of a P ep tid om im etic
Marisa C. Kozlowski and Paul A. Bartlett*
Department of Chemistry, University of California, Berkeley, California 94720-1460
Received J une 13, 1996X
An attempted synthesis of the tricyclic peptidomimetic 1, designed to imitate a â-turn tripeptide
in tendamistat, afforded instead the 6,6,8-ring system of 2. The key step in the synthesis entailed
acylation of the hindered R,R′-disubstituted morpholine 4.2, which was approached by acylative
ring opening of the 3,6-oxazabicyclo[4.2.0]octane 4.3. However, transannular rather than exocyclic
cleavage occurred, giving the 1,6-oxazacyclooctane isomer 4.5. Subsequent ring closures to form
the bi- and tricyclic intermediates 7.3 and 8.5 were difficult because of the strain being built into
the ring systems. After completion of the synthesis, the structures of the intermediates and final
product were elucidated by NMR, with three-bond, heteronuclear multiple-bond correlation
experiments providing unambiguous evidence for the ring connectivity, and by molecular modeling,
which allowed assignment of the stereochemistry. Compound 2 is a modest inhibitor of the target
enzyme R-amylase (Ki ) 170 µM in 5% DMSO/water), binding with similar affinity to the tripeptide
Ac-Trp-Arg-Tyr-OMe. Although the side-chain attachment points in the ring system of 2 correspond
closely to the relative CR-positions in tendamistat (rmsd ) 0.24 Å), the alignment of the CR-Câ
bonds is poor, illustrating the importance of side-chain orientation in a peptidomimetic.
Sch em e 1
The tricyclic peptidomimetic 1 was designed as a mimic
of the R-amylase inhibitor tendamistat.1 The 6,6,6-ring
system can be readily identified in CAVEAT2 searches
as a rigid template able to hold the side chains in the
same orientation observed around the â-turn at positions
18-20 in tendamistat.3,4 This Trp18Arg19Tyr20 loop is
central to the interaction between the two proteins,5 and
mimics incorporating this sequence are effective inhibi-
tors of R-amylase.6,7 A synthesis of this compound was
planned with the key ring closures as shown in Scheme
1. Execution of this route proved to be a significant
challenge and, indeed, went awry at a key step, leading
to a final product with the isomeric 6,6,8-ring system of
2. In this report, we detail the chemistry involved and
proof of structure of the compound produced.
Tricycle 1 incorporates the Trp-Arg-Tyr tripeptide unit
as well as a four-carbon backbone. To our knowledge,
no compound containing the tricyclic ring system repre-
sented by 1 has been reported, although several strate-
gies can be envisaged for its assembly. We considered
two primary routes: the “bottom-up” sequence depicted
in Scheme 1 and a “top-down” approach for stitching the
four-carbon backbone directly to a tripeptidyl unit. We
regarded the tryptophan indole as the most sensitive
functionality in the molecule and therefore wanted to
introduce it toward the end of the synthesis. This
consideration, as well as concern that the “top-down”
sequence could result in O- as opposed to N-alkylation
in the amide cyclization steps, led us to pursue the
“bottom-up” route. The side-chain configurations on the
X Abstract published in Advance ACS Abstracts, October 1, 1996.
(1) Vertesy, L.; Oeding, V.; Bender, R.; Zepf, R. H.; Nesemann, G.
Eur. J . Biochem. 1984, 141, 505.
(2) Lauri, G.; Bartlett, P. A. J . Comput.-Aided Mol. Des. 1994, 8,
51.
(3) Pflugrath, J . W.; Wiegand, G.; Huber, R. J . Mol. Biol. 1986, 189,
383.
(4) Kline, A. D.; Braun, W.; Wu¨thrich, K. J . J . Mol. Biol. 1988, 204,
675.
(5) Wiegand, G.; Epp, O.; Huber, R. J . Mol. Biol. 1995, 247, 99.
(6) Etzkorn, F. A.; Guo, T.; Lipton, M. A.; Goldberg, S. D.; Bartlett,
P. A. J . Am. Chem. Soc. 1994, 116, 10412.
(7) Matter, H.; Kessler, H. J . Am. Chem. Soc. 1995, 117, 3347.
S0022-3263(96)01114-0 CCC: $12.00 © 1996 American Chemical Society

7682 J . Org. Chem., Vol. 61, No. 22, 1996
Koslowski and Bartlett
Sch em e 2
Sch em e 3
amino epoxide 2.8 was successful; this material was
prepared by the sequence depicted in Scheme 2. The
amino epoxide 2.8 could be isolated and purified; it
cyclized slowly at room temperature but was converted
cleanly to the morpholine 2.9 in refluxing dioxane.
Further evidence of steric encumbrance around the
nitrogen of the trans-substituted morpholine was its
resistance to acylation even under forcing conditions. For
example, conventional peptide coupling reagents had no
effect on di-O-protected derivatives of 2.9, and trifluoro-
acetic anhydride was required even for formation of the
trifluoroacetamide. Despairing of carrying the morpho-
line on, we explored alternative strategies for assembly
of the tricyclic system that could be characterized as
“outside-in” sequences (Scheme 3). Ultimately, the acyl
bond to the morpholine was to be formed by lactamiza-
tion, after establishment of the backbone-nitrogen bond;
that linkage would also be formed intramolecularly by
walking it in from the 1° carbon via the aziridine.
F or m a tion a n d Rin g Op en in g of Azetid in e 4.3
(Sch em e 4). In one model study, attempted alkylation
of phenylalanine methyl ester with the 1° mesylate 4.2
failed to give the desired 2° amine 4.4 but led instead to
the 3-oxa-6-azabicyclo[4.2.0]octane 4.3. When prepared
intentionally, this material could be formed in 79% yield
with triethylamine in refluxing acetonitrile. Although
formation of azetidines by ring closure of linear precur-
sors is usually difficult,14 with alkyl substitution and
especially with a monocyclic amine, high yields can be
obtained, as exemplified by the synthesis of conidine from
R-(2-chloroethyl)piperidine.15 Although the bicyclic aze-
tidine 4.3 did not lie on a planned route to the desired
tricycle, its formation represented one of the few trans-
formations in which a new bond could be formed to the
morpholine nitrogen and it was a structure in which the
reactivity of the nitrogen itself was significantly altered.
There is considerable precedent for the ring opening
of strained, quaternary ammonium ions by nucleophilic
attack, and significant regioselectivity is observed for
unsymmetrical aziridinium and azetidinium ions.14-20
Most relevant is a study of the cleavage pattern of the
quaternized 1-azabicyclo[4.2.0]heptane and 1-azabicyclo-
tricyclic system would originate with the L-amino acid
starting materials, and asymmetric epoxidation of an
allylic alcohol intermediate8 was to be used to control the
stereochemistry of the backbone itself. Accordingly, our
first goal was to assemble the allylic ether from cis-2-
butene-1,4-diol and tyrosinol.
Syn th esis a n d Attem p ted Acyla tion of Mor p h o-
lin e 2.9 (Sch em e 2). The mono-tert-butyldimethylsilyl
ether of cis-2-butene-1,4-diol, 2.1, prepared according to
McDougal’s procedure,9 was converted to the mesylate
2.2 and coupled with O-benzyl-L-tyrosinol, 2.3,10,11 gener-
ated by reduction of the corresponding methyl ester.12
The use of sodium hydride or potassium tert- butoxide
as base led to mixtures of O- and N-alkylated material;
however, potassium hydride afforded the desired ether
2.4 as the exclusive isomer, albeit in 55% isolated yield
on a preparative scale. We undertook a number of
experiments designed to effect cyclization of the morpho-
line ring from N-acylamino epoxides. The substrates
were prepared from 2.4 by N-acylation, O-deprotection,
Sharpless epoxidation8,13 of the allylic alcohol, and,
finally, silylation or mesylation of the hydroxyl group.
Unfortunately, attempts to form the morpholine ring
from such substrates under a variety of basic conditions
failed. We presume that steric hindrance prevents
cyclization to the desired product, in which severe A1,3
-
strain cannot be avoided. However, cyclization of the
(8) Hanson, R. M.; Sharpless, K. B. J . Org. Chem. 1986, 51, 1922.
(9) McDougal, P. G.; Rico, J . G.; Oh, Y.-I.; Condon, B. D. J . Org.
Chem. 1986, 51, 3388.
(10) Itsuno, S.; Hirao, A.; Nakahama, N. J . Chem. Soc., Perkin
Trans. 1 1983, 1673.
(11) Rinaldi, P. L.; Wilk, M. J . Org. Chem. 1983, 48, 2141.
(12) Ramachadran, J .; Berger, A.; Katchalski, E. Biopolymers 1971,
10, 1829.
(14) Ballard, S. A.; Melstrom, D. S. In Heterocyclic Compounds; R.
C. Elderfield, R. C.; Ed.; Wiley: New York, 1950; Vol. 1; p 78.
(15) Moore, J . A. In The Chemistry of Heterocyclic Compounds;
Weissberger, A., Ed.; Wiley: New York, 1964; Vol. 19, Part 2, p 885.
(16) Tshiamala, K.; Vebrel, J .; Laude, B.; Mercier, R. Bull. Soc.
Chim. Fr. 1990, 127, 584.
(17) DiVona, M. L.; Illuminati, G.; Lilloci, C. J . Chem. Soc., Perkin
Trans. 2 1985, 1943.
(13) Gao, Y.; Hanson, R. M.; Klunder, J . M.; Ko, S. Y.; Masamune,
H.; Sharpless, K. B. J . Am. Chem. Soc. 1987, 109, 5765.
(18) Deyrup, J . A. In The Chemistry of Heterocyclic Compounds;
Hassner, A., Ed.; Wiley: New York, 1983; Vol. 42, Part 1, p 1.

Misrouted Synthesis of a Peptidomimetic
J . Org. Chem., Vol. 61, No. 22, 1996 7683
Sch em e 4
reagents. Although acetic anhydride gave a complex
mixture after formation of a highly polar intermediate,
treatment of the bicyclic azetidine with acetyl chloride
in methylene chloride at 0 °C led rapidly to a ring-cleaved
and chlorine-substituted acetamide in 79% yield. Fmoc-
alanine acid chloride23 was not sufficiently stable under
the reaction conditions, forming the Leuch’s anhydride,24
but the N-tosyl analog reacted with 4.3 to generate an
acylated compound in 44% yield. This product was
initially assigned the structure 4.6.
Hin d sigh t. The spectral differences to be expected for
the isomeric cleavage products from azetidine 4.3 are
subtle, since they have identical masses and identical 1H
NMR spin systems. Moreover, the presence of amide
conformers further complicated an already congested
NMR spectrum. From the precedent cited above, rein-
forced by our expectation that the mild cleavage condi-
tions would favor SN2-like attack at the primary carbon
of the azetidine ring, we assigned the structure of the
cleavage products as 4.6, in preference to the eight-
membered rings 4.5. As detailed below, this assignment
was incorrect.
We completed the synthesis of a tricyclic product,
incorporating the tyrosine, arginine, and tryptophan side
chains, with the belief that we were following the route
depicted in Scheme 1. With hindsight, many of the
difficulties we encountered in a number of the steps, in
particular the ring-closure reactions, are understandable,
but it was not until we had a fully elaborated tricyclic
ring system, with no conformational ambiguities, that
NOESY spectra suggested, and heteronuclear multiple-
bond correlation (HMBC) experiments confirmed, that we
had made the wrong assignment.
A Robu st P r otection Sch em e for Or n ith in e. The
requirement of an acid chloride for the acylative ring-
opening step placed severe limitations on the manner in
which the arginine residue could be introduced and on
the choice of protecting groups. Rather than carry the
guanidino function through the synthesis, we elected to
incorporate a protected ornithine at this stage and
introduce the guanidine at the end. The protecting group
on the R-nitrogen had to be stable to acid and strong
acylating conditions, tolerate the adjacent acid chloride
without formation of Leuch’s anhydride or the azlactone,
and yet be readily removed. These requirements were
satisfied by the recently described [2-(trimethylsilyl)-
ethyl]sulfonyl (SES) group,25 which can be fragmented
with fluoride ion. The protection on the δ-nitrogen
additionally had to survive strong base (for subsequent
ring closures) and fluoride ion (for cleavage of the
R-protecting group); moreover, it could not retain any NH
moiety or cyclization on the acid chloride would occur.
We chose the seldom used diphenyloxazolone (Ox) moiety,
which can be removed under oxidative or reductive
conditions.26
Sch em e 5
[3.2.0]octane systems in the presence of various nucleo-
philes: Ebno¨ther reported that the former affords com-
parable amounts of the isomeric azepine and pyrrolidine
products, while in reaction of the latter the azetidine ring
is cleaved exclusively at the primary carbon to give the
piperidine (Scheme 5).21 Less precedent exists for the
reactivity of azetidines toward acylating agents; the one
example we found was for an intramolecular system with
no regiochemical issue.22
The protected ornithine 6.4 was prepared in 48% yield
for four steps from the commercially available NR-Boc
derivative, as shown in Scheme 6.
Azetid in e Op en in g w ith th e P r otected Or n ith in e
Acid Ch lor id e. Initial experiments involving azetidine
acylation and cyclization of the second ring were carried
The premise of ring opening the strained azetidine 4.3
with concomitant acylation was tested with several
(19) Crist, D. R.; Leonard, N. J . Angew. Chem., Int. Ed. Engl. 1969,
81, 962.
(20) Fruton, J . S. In Heterocyclic Compounds; Elderfield, R. C., Ed.;
Wiley: New York, 1950; Vol. 1, p 61.
(23) Bodanzsky, M.; Bodanzsky, A. The Practice of Peptide Synthesis,
Vol. 21; Springer Verlag: Berlin, 1984.
(24) Bodanszky, M. Principles of Peptide Synthesis; Springer-Ver-
lag: New York, 1984.
(21) Ebno¨ther, A.; J ucker, E. Helv. Chim. Acta 1964, 47, 745.
(22) Bartholomew, D.; Stocks, M. J . Tetrahedron Lett. 1991, 32,
4799.
(25) Weinreb, S. M.; Demko, D. M.; Lessen, T. A. Tetrahedron Lett.
1986, 27, 2099.
(26) Sheehan, J . C.; Guzier, F. J ., J r. J . Org. Chem. 1973, 38, 3034.

7684 J . Org. Chem., Vol. 61, No. 22, 1996
Koslowski and Bartlett
Sch em e 6
refluxing acetonitrile or dioxane. Ring closure of 7.2t to
the bicyclic lactam 7.3t was effected in the more polar
solvent DMSO at 110 °C with DIEA. Interestingly,
sodium hydride in DMSO was ineffective in inducing the
desired cyclization. In contrast to the two monocyclic
precursors 7.1t and 7.2t, the product did not display a
mixture of amide rotamers in the NMR spectrum,
indicating that a rigid, bicyclic structure had formed.
When the SES-protected analog 6.4 was employed in
this sequence, 55-78% yields of the acylated product 4.5s
were obtained, and formation and cyclization of the
corresponding mesylate 7.2s proceeded in an overall yield
of 40-60%. However, in contrast to the tosylamide, the
SES derivative required more forcing conditions for
cyclization: with DIEA as base, an 8-h period at 130 °C
in DMSO was necessary, giving cyclic material 7.3s in
38% yield. The best conditions we found for cyclization
of 7.2s employed sodium hydride in DMSO, although
these strongly basic conditions also produced the elimi-
nated products 7.4s and 7.5.
Com p letion of th e Syn th esis. Removal of the SES
group from the bicyclic intermediate 7.3s proceeded with
1 M TBAF in THF at room temperature, surprisingly
mild conditions in comparison to those typically reported
for this transformation.25 We attribute this behavior,
along with the difficulty encountered in the preceding
cyclization step, to the steric crowding around the B-ring,
which is alleviated as the N-S bond is broken.
Sch em e 7
A significant amount of the eliminated byproduct 7.5
was also generated during removal of the SES group with
TBAF. This material could be recycled to the desired
product with sodium borohydride (NaBH4), reduction
occurring from the exo direction to restore the lost
stereocenter. Thus, treatment of sulfonamide 7.3s con-
secutively with TBAF and NaBH4 afforded the amine
7.6o in 80% yield. The imine 7.5 presumably results
from â-elimination of 2-(trimethylsilyl)ethyl sulfinate as
another consequence of the relatively weak N-S bond.
In spite of the apparent steric crowding of the B-ring
nitrogen, the amine 7.6o reacted readily with unhindered
acylating agents. The glycyl derivatives were formed
either from the acid chloride of N-tosylglycine or from
bromoacetyl bromide followed by ammonia. However,
neither of these derivatives could be induced to cyclize
under a variety of basic conditions.27 Consequently, our
approach to introduction of the C-ring involved displace-
ment of the halide first and ring closure to the lactam
last. The steric congestion of the SES amide prevented
substitution of the chloride, even under Finkelstein
conditions, but displacement with iodide proceeded in
quantitative yield with amine 7.6o. However, attempted
replacement of the iodide in turn with tryptophan methyl
ester (Trp-OMe) in refluxing acetonitrile only gave aziri-
dine 8.2; iodide-induced epimerization is presumably
out with the NR-tosyl analog of 6.4. When the acid
chloride of this derivative was isolated and added to the
azetidine 4.3, low yields (ca. 10%) of acylated material
were invariably obtained. However, if the acid chloride
(3 equiv) was generated in situ with excess PCl5 (8 equiv),
and then combined with diisopropylethyl amine (DIEA)
and the azetidine, ring-opened material 4.5t could be
isolated in 49% yield, with 46% recovery of starting
azetidine.
(27) While the resistance of i to cyclization can be understood in
the context of the strained bicyclic skeleton, it was puzzling when we
believed we were working with the fused ring system of ii. Moreover,
in this system there would not have been any stereochemical conse-
quences of the double displacement process for introduction of the
tryptophan moiety (R-Cl f R-I f R-Trp-OMe).
Removal of the silyl ether with tetrabutylammonium
fluoride (TBAF) afforded the chloro alcohol 7.1t as a
mixture of amide rotamers (Scheme 7). This material
was not cyclized under Mitsunobu conditions, nor was
the mesylate 7.2t cyclized in the presence of DIEA in

Misrouted Synthesis of a Peptidomimetic
J . Org. Chem., Vol. 61, No. 22, 1996 7685
Sch em e 8
Sch em e 9
had been cleaved. Although the Ox group is susceptible
to oxidative cleavage as well (e.g., MCPBA in TFA),26 this
approach was also precluded by the oxidative sensitivity
of the indole moiety.
The Ox group apparently had to be replaced prior to
introduction of the tryptophan unit. From examination
of the synthetic route, it was clear that exchange after
installation of the B-ring was the most advantageous
prospect: by this point, the most demanding reactions
had been executed and subsequent transformations
would tolerate a Boc or Cbz group. From intermediate
7.3s, we initially examined oxidative processes in order
to preserve the phenolic benzyl ether (Scheme 9). How-
ever, a variety of oxidative conditions provided the
desired product 9.1 in less than 50% yield; the remainder
of the material was recovered starting material and
benzamide 9.2 from Baeyer-Villiger oxidation of one of
the intermediates during breakdown of the Ox group.
The reductive approach proved to be more successful.
Hydrogenolysis of 7.3s in acidic methanol over palladium
black afforded phenolic amine 9.3 in quantitative yield.
Protection of the amine with N-((benzyloxycarbonyl)oxy)-
succinimide (CbzOSu) afforded the N-Cbz derivative in
87% yield after hydrolysis of the small amount of O-Cbz
byproduct, and re-benzylation proceeded smoothly with
benzyl bromide.36 What this roundabout approach lacked
in elegance it more than made up in efficiency, providing
the Cbz derivative 9.5 in 78% overall yield from the Ox
analog 7.3s.
With the more felicitous Cbz group in place, the same
steps were executed to introduce the C-ring as described
for the Ox-protected series. The Trp-OMe adduct 8.3z
and subsequent intermediates proved to be fairly sensi-
tive to acidic conditions, including halogenated solvents
and silica chromatography, as indicated by the develop-
ment of a red-orange color; this may be due to the
propensity of the indole nucleus to dimerize in acid.37
After conversion to the tricyclic product 8.5z, removal of
the protecting groups was accomplished by hydrogenoly-
sis, and the guanidino moiety of the arginine side chain
was introduced with aminoiminomethanesulfonic acid in
the presence of triethylamine (Scheme 8).38,39 Reverse
phase HPLC then afforded the final compound 2. The
overall sequence is summarized in Scheme 10.
responsible for the double inversion required for this
displacement.
The challenge at this point was to favor an inter-
molecular reaction (8.1 f 8.3) over an intramolecular
reaction (8.1 f 8.2) (Scheme 8). Increased nucleophile
concentration was one factor that we addressed, of course,
but the most important proved to be temperature. We
reasoned that the entropically favored intramolecular
process would be less favored at lower temperature.
Indeed, when iodide 8.1o was treated with an excess of
Trp-OMe in acetonitrile at room temperature, the alky-
lated product 8.3o was produced exclusively! Because the
tricyclic bislactam 8.5o is quite strained, it is not surpris-
ing that cyclization of the amino ester 8.3o does not occur
simply on heating. However, the amino acid 8.4o cyclizes
in 70% yield with bromotris(pyrrolidino)phosphonium
hexafluorophosphate (PyBrOP)28,29 or diphenylphosphoryl
azide (DPPA)30,31 in DMF.
From tricycle 8.5o, only deprotection steps remained,
but they were not to be accomplished directly. In spite
of a number of precedents for removal of the diphenyl-
oxazolinone group (Ox) under mild conditions,26,32-35 we
were unable to accomplish this step in the presence of
the indole moiety. Under hydrogenolytic conditions, the
phenolic benzyl ether was cleaved first, and under the
more forcing conditions required for removal of the Ox
group, the indole was also attacked. Dissolving metal
reduction of 8.5o yielded material in which a lactam bond
(28) Castro, B.; Dormoy, J . R. Tetrahedron Lett. 1973, 35, 3243.
(29) Coste, J .; Frerot, E.; J ouin, P. Tetrahedron Lett. 1991, 32, 1967.
(30) Brady, S. F.; Paleveda, W. J .; Arison, B. H.; Saperstein, R.;
Brady, E. J .; Raynor, K.; Reisine, T.; Veber, D. F.; Freidinger, R. M.
Tetrahedron 1993, 49, 3449.
(31) Laszlo, S. E.; Ley, S. V.; Porter, R. A. J . Chem. Soc., Chem.
Commun. 1986, 344.
Assign m en t of In ter m ed ia te a n d F in a l Str u c-
tu r es. Several observations led us to suspect that the
(32) Pansare, S. V.; Vederas, J . C. J . Org. Chem. 1987, 52, 4804.
(33) Salituro, G. M.; Townsend, C. A. J . Am. Chem. Soc. 1990, 112,
760.
(34) Basak, A.; Salowe, S. P.; Townsend, C. A. J . Am. Chem. Soc.
1990, 112, 1654.
(35) Sreenivasan, U.; Mishra, R. K.; J ohnson, R. L. J . Med. Chem.
1993, 36, 256.
(36) Schmidhammer, H.; Brossi, A. J . Org. Chem. 1983, 48, 1469.
(37) Houlihan, W. J ., Ed. Indoles Part I, Vol. 25; Wiley: New York,
1972.
(38) Miller, A. E.; Bischoff, J . J . Synthesis 1986, 777.
(39) Kim, K.; Lin, Y. T.; Mosher, H. S. Tetrahedron Lett. 1988, 29,
3183.

7686 J . Org. Chem., Vol. 61, No. 22, 1996
Koslowski and Bartlett
Sch em e 10
Sch em e 11
final product might not have the desired structure 1.
First, the difficulties we encountered in forming the B-
and C-rings were not in accord with our expectations for
the fused-ring structures. Second, the final material we
produced proved to be only a modest inhibitor of R-amy-
lase, comparable to the Trp-Arg-Tyr tripeptide. Most
significantly, cross peaks expected in the ROESY spec-
trum of 1, such as between H6 and H7, were missing,
while unanticipated and strong cross peaks were ob-
served between H3′ and H6 and between H7 and H21
(see Scheme 11 for numbering scheme). We therefore
reexamined our structural assignments for key interme-
diates in the sequence. The immediate products from
opening of the azetidine ring, amides 4.5, gave broad
peaks in the 1H NMR spectrum that did not coalesce
satisfactorily at higher temperature nor resolve at lower
temperature. These observations are consistent with
conformational interconversions that are slow on the
NMR time scale, of which both 4.5 and 4.6 are capable.
We therefore used the bicyclic intermediate 8.3z and the
final tricyclic product 2 for NMR studies.
The aromatic region of the spectra is not intrinsic to
the ring systems and hence will not be discussed in
relation to the structural elucidation. However, the data
for the aromatic region are consistent with the assign-
ments, confirming the presence and connectivity of one
indole and three phenyl groups for 8.3z and one indole
and one phenyl group for 2.
COSY and TOCSY spectra of the bicyclic molecule
identified four spin systems consistent with either struc-
ture iso-8.3 or 8.3 (Scheme 11): CH-CH2-CH2-CH2-
NH (10-15-16-17-18), CH-CH2 (21-23), CH2-CH-

Misrouted Synthesis of a Peptidomimetic
J . Org. Chem., Vol. 61, No. 22, 1996 7687
Ta ble 1. P r oton -Ca r bon Lon g-Ra n ge Cor r ela tion s for
Sch em e 12
8.3z
two bonds to carbon
three bonds to carbona
H
(H-X-C)
(H-X-Y-C)
2
3
5
6
7
3, 12
8, 11
5
3, 7
8, 21
5
2
6
7
8
7
2, 6, 11
7, 16
3, 14
11, 17
10
15, 19
6, 24
22, 25, 26
10
12
15
16
17
21
23
11, 15
2, 13
10, 16
15, 17
16
22, 23
21, 24
a
Correlations which establish the ring system are in boldface
type.
CH2 (3-2-12), and CH2-CH-CH-CH2 (5-6-7-8 or
8-7-6-5). All protons were assigned, although the
inherent ambiguity for the CH2-CH-CH-CH2 spin system
could not be resolved until a two- and three-bond proton-
carbon correlation experiment was carried out, as de-
scribed below. The carbons were assigned from the
multiplicity observed in DEPT experiments and from one-
bond correlations with their respective hydrogens identi-
fied by an HSQC (heteronuclear single quantum coher-
ence) experiment. A two- and three-bond proton-carbon
correlation experiment was carried out to establish the
connectivity of the spin systems rigorously. The HMBC
(heteronuclear multiple-bond correlation) experiment was
performed in such a way as to filter out one-bond proton-
carbon correlations, which would have complicated data
interpretation. The three-bond correlation data enabled
the two possible assignments for the CH2-CH-CH-CH2
spin system to be distinguished, with a number of key
correlations across the heteroatoms (Table 1) consistent
only with structure 8.3z.
Ta ble 2. Ca lcu la ted a n d Obser ved Vicin a l Cou p lin g
Con sta n ts for Bicycle 8.3z
calcd coupling constants (Hz)a
spin system
obsd
trans-gauche
trans-anti
cis
2-3
2-3′
5′-6
5-6
6-7
7-8′
7-8
3.5
10.6
<1.0
4.5
3.9
5.0
4.6
11.2
0.9
4.4
2.6
3.5
10.9
5.1
11.0
9.9
4.7
11.3
1.9
9.4
3.2
5.7
1.3
3.8
2.1
5.3
1.3
1.4
a
Calculated in Macromodel according to ref 55.
In addition to the three-bond correlation data, the
coupling constants of the ring protons and the NOEs
observed are more in accord with structure 8.3z than iso-
8.3z. These data in turn allowed us to determine the
stereochemistry of the bicyclic product and the course of
the cyclization reaction (7.6 f 8.1 f 8.3), which involved
iodide catalysis in displacement of the chloride. Two
inversions from 7.6 would provide the trans isomer 8.3,
while multiple displacements with iodide in formation
of 8.1 or in the next step, as postulated to explain
formation of the aziridine 8.2, could have resulted in the
cis isomer ep i-8.3 (Scheme 12).
The two ring systems were modeled with methyl
groups substituted for the side chains. The cis isomer
ep i-8.3m is lower in energy than the trans conformers,
8.3m , and it is less flexible conformationally. Low-energy
structures of the former are very similar, with the ether
oxygen of the eight-membered ring exhibiting the great-
est mobility. The trans isomer has two major low-energy
conformational sets which differ in the dihedral angle
between H6 and H7 and are designated as trans-gauche
and trans-anti (Figure 1). The trans-anti conformations
of 8.3m are generally higher in energy than the trans-
gauche, and both sets contain members with slightly
different conformations of the eight-membered ring near
the ether oxygen. The coupling constants and NOEs
observed for 8.3 are most consistent with the geometry
and interatomic distances predicted for the trans-gauche
model of 8.3m (Tables 2 and 3). The discrepancies in
the data arise from the flexibility of the eight-membered
Ta ble 3. Ca lcu la ted Dista n ces a n d NOEs Obser ved for
Bicycle 8.3z
calcd distance (Å)
spin system
NOE obsd
trans-gauche
trans-anti
cis
2.6
2.2
4.0
2.3
4.2
3-12
3′-5′
3-8
weak
2.6
2.2
2.6
2.4
3.9
2.4
3.1
obscureda
very weak
obscureda
very weak
medium
medium
medium
medium
4.0b
2.3
2-8
5-8
2.3
2.5
<3.0
2.2
6-7
3.1
2.4
6-21
7-21
8′-10
<3.0
<4.0
2.4
<3.0
<4.0
2.3
2.3
a
The expected cross peaks are obscured by overlapping reso-
b
nances. The eight-membered ring containing H1 is conforma-
tionally mobile, so this NOE could be due to another conformation.
ring about the ether oxygen and a minor population of
the trans-anti conformer.
Determination of the stereochemistry and conforma-
tion of the tricyclic product 2 confirmed the above
assignments in the bicyclic series. COSY and TOCSY
spectra identified the four aliphatic spin systems in the
tricyclic molecule and enabled assignment of all the
protons. The two possible arrangements of the CH2-
CH-CH-CH2 spin system were distinguished in this
case by the distinctive coupling patterns of H5 and H8,
as described below. 1D-decoupling experiments revealed
an H6-H7 coupling of 9.5 Hz, reflecting either an anti
relationship or an almost totally eclipsing syn arrange-

7688 J . Org. Chem., Vol. 61, No. 22, 1996
Koslowski and Bartlett
F igu r e 1. Minimum-energy conformations for the bicyclic models ep i-8.3m and 8.3m (energy values calculated in MacroModel
using MM2* force field).
F igu r e 2. Minimum-energy conformations for the tricyclic models ep i-2m and 2m (energy values calculated in MacroModel
using MM2* force field).
ment between these hydrogens. Because the size of 2
(MW ) 530) prevented observation of NOE cross peaks,
a ROESY spectrum was recorded to identify through-
space relationships. The absence from this spectrum of
a cross peak between H6-H7 argued for an anti arrange-
ment, but the peaks that actually were observed provided
more definitive evidence. The strongest ROESY cross
peaks (aside from geminal relationships) were between
H3′ and H6, between H7 and H21, and between H8′ and
H10, which could not be reconciled with any conformation
or stereoisomer of 1, but are completely consistent with
the structure of 2.
Both the cis and trans stereoisomers of 2 were modeled
with methyl groups substituted for the side chains
(Figure 2). In this series as well, the cis isomers were
found to be much lower in energy than the trans
compounds; both tricycles have very little flexibility, with
the only apparent mobility confined to the ether moiety.
As above, calculations of the coupling constants and
interatomic distances for the methyl side-chain analogs

Misrouted Synthesis of a Peptidomimetic
J . Org. Chem., Vol. 61, No. 22, 1996 7689
Ta ble 6. In h ibitor s of r-Am yla sea
Ta ble 4. Ca lcu la ted a n d Obser ved Vicin a l Cou p lin g
Con sta n ts for Tr icycle 2
inhibitor
Ki (µM)
calcd coupling constants (Hz)a
Trp-Arg-Tyr
Ac-Trp-Arg-Tyr-OMe
tricycle 2
870 ( 30
190 ( 10
170 ( 20
30b
spin system
obsd
trans
cis
2-3′
2-3
5′-6
5-6
6-7
7-8′
7-8
10.1
3.0
5.0
11.0
9.5
2.2
10.8
3.3
5.0
11.0
9.6
3.1
11.3
5.1
2.6
10.2
4.4
4.9
cyclo[DPFAWRY]
Determined in 5% DMSO/water at 25 °C, pH 7.6 Estimated
a
b
from two-point determination.
in this solvent for comparison (Table 6). These values
were found to be twice as large as in water, reflecting
the importance that desolvation of the hydrophobic
regions of these analogs plays in their affinity. Com-
pound 2 binds to R-amylase with the same affinity as the
capped tripeptide, Ac-Trp-Arg-Tyr-OMe, in spite of the
fact that its skeleton is quite different from that in the
original design 1.
2.2
2.8
1.5
a
Calculated in Macromodel according to ref 55.
Ta ble 5. Ca lcu la ted Dista n ces a n d ROEs Obser ved for
Tr icycle 2
calcd distances (Å)
spin system
ROE
trans
cis
Fortuitously, the ring systems of both 1 and 2 locate
the side-chain attachment points (corresponding to the
CR carbons) in a similar fashion to the â-turn region of
tendamistat itself (Figure 3). The rms deviations for
these three carbons in the modeled structures of 1 and 2
from the corresponding CR carbons in tendamistat-R-
amylase complex5 are 0.05 and 0.24 Å, respectively
(Figure 3). However, the simple distance relationships
between these points is less important than the relative
orientations of the side chains, which are aligned quite
differently in 2 than in 1. Moreover, the skeleton of 2 is
a poor match to the intervening atoms of the peptide
backbone, in contrast to 1. Whereas the lactams in 1
follow the zig-zag of the amides in the tendamistat
â-turn, those of 2 are perpendicular; as a result, side-
chain conformations that are staggered in tendamistat
would be eclipsed in 2.
3-12
3′-5′
3′-6
3-8
medium
obscureda
medium
weak
medium
absent
absent
strong
strong
2.5
2.5
2.2
3.9b
2.4
3.1
3.6
2.3
2.3
2.7
2.2
4.4
3.9
2.3
2.4
3.1
2.3
2.4
2-8
6-7
6-21
7-21
8′-10
a
This portion of the spectrum was obscured by contaminating
TOCSY cross peaks which have the opposite sign of ROESY cross
peaks and effectively cancel. b The eight-membered ring containing
H3 is mobile so this distance could be less in some conformations.
Sch em e 13
Con clu sion
The tricyclic structure of 2 embodies a novel complex
ring system. Little or no precedent for compounds of this
type exist according to structural searches of the Chemi-
cal Abstracts Service database. Although many examples
of the all-carbon skeleton exist, the 4-oxa-1,7,10-triaza-
tricyclo[8.2.2.06,11]tetradecane ring system itself appears
to be unknown. The closest heterocyclic structure we
found is 3, derived from an intramolecular azomethine
ylide cycloaddition.40
of 2m and ep i-2m indicate that the trans structure is
the most consistent with the experimental coupling
constants and observed ROEs (Tables 4 and 5).
Assignment of the tricyclic structure of the final
product as 2 explains some of the side reactions and
synthetic difficulties observed along the route. For
example, the difficulty of the ring closure reactions is now
understandable, given the strain that is being built into
the tricylic system. We had also found it disconcerting
that reduction of the imine 7.5 is so stereoselective; this
byproduct, formed by elimination during the deprotection
of 7.3, is reduced to a material identical to the normal
deprotection product 7.6. While such stereoselectivity in
reduction of the bicyclo[4.4.0] ring of iso-7.6 would be
unexpected, it is quite consistent with exo approach to
the bicyclo[5.3.1] structure (Scheme 13).
The route to the tricyclic compound 2 entails 31 steps
from the component pieces O-Bn-L-Tyr, but-2-ene-1,4-
diol, R-N-Boc-L-ornithine, and L-Trp-OMe with an overall
yield from O-Bn-L-Tyr of 0.54%. The synthesis is semi-
convergent with the longest linear sequence consisting
of 25 steps. The challenge of this molecule lies in the
large amount of functionality that is exposed during the
synthesis, including seven nitrogens and four oxygens,
most of which require protection given the current state
of synthetic technology. In particular, the troublesome
indole and guanidine functionalities were predictably
In h ibition of r-Am yla se. The ability of the tricyclic
compound 2 to inhibit R-amylase was determined in 5%
DMSO/water using a modified assay with p-nitrophenyl
maltotrioside as substrate.6 The inhibition constants for
previously reported peptide analogs were redetermined
(40) Garner, P.; Sunitha, K.; Ho, W.-B.; Youngs, W. J .; Kennedy, V.
O.; Djebli, A. J . Org. Chem. 1989, 54, 2041.

7690 J . Org. Chem., Vol. 61, No. 22, 1996
Koslowski and Bartlett
F igu r e 3. Superposition of the tricyclic ring systems of 2 (a) and 1 (b) on the Trp18Arg19Tyr20 loop of tendamistat.
wise to a 0 °C solution of alcohol 2.1 (10.0 g, 49.4 mmol) and
triethylamine (7.60 mL, 54.4 mmol) in CH2Cl2 (250 mL). The
ice bath was allowed to warm to rt, and after 3.5 h, the mixture
was diluted with ether (500 mL), washed with saturated NaCl
(3 × 250 mL), dried over MgSO4, and concentrated. The brown
oil was purified by chromatography with 30% ether/hexanes
to afford the mesylate 2.2 (13.0 g, 94%) as a clear oil: 1H NMR
(500 MHz, CDCl3) δ 5.82 (m, 1), 5.61 (m, 1), 4.85 (d, 2, J )
6.9), 4.27 (d, 2, J ) 5.2), 2.99 (s, 3), 0.88 (s, 9), 0.06 (s, 6); IR
(film) 1370, 1185 cm-1. Anal. Calcd for C11H24O4SSi: C, 47.11;
H, 8.63; S, 11.43. Found: C, 47.50; H, 8.75; S, 11.05.
4-[(2S)-2-Am in o-3-h yd r oxyp r op yl]-1-(p h en ylm eth oxy)-
ben zen e (2.3).10,11 O-Bn-L-Tyr (49.6 g, 0.18 mol) was dissolved
in methanol (1.2 L) and cooled to -10 °C with an ice/salt/water
bath. Thionyl chloride (32 mL, 0.44 mol) was added dropwise,
causing a highly exothermic reaction. After the addition, the
ice bath was removed and the mixture warmed to rt. After 3
d, the mixture was concentrated to a yellow solid, which was
precipitated from hot methanol with two volumes of ether to
afford O-Bn-L-TyrOMe‚HCl12 (57.8 g, 98%) as a white solid
which was used without further purification.
A suspension of O-Bn-L-TyrOMe•HCl (28.3 g, 88 mmol) in
EtOAc (290 mL) was cooled to 0 °C, and triethylamine (13.5
mL, 97 mmol) was added. After 1 h, the precipitated salt was
removed by filtration and the filtrate was concentrated. The
remaining oil was dissolved in THF (500 mL), and the solution
was added dropwise over 1 h to a suspension of LiAlH4 (10.0
g, 260 mmol) in THF (90 mL). After stirring for an additional
1 h, the mixture was quenched cautiously at 0 °C by successive
addition of water (10 mL), 20% aqueous NaOH (10 mL), and
water (30 mL). After the solution was stirred vigorously for 1
h at rt, the slurry was concentrated, filtered through Celite,
and rinsed with CH2Cl2 (3 × 350 mL). The filtrate was washed
with saturated NaCl (600 mL) and dried over Na2SO4. Re-
moval of the solvent afforded the amino alcohol 2.3 (22.2 g,
98%) as fine white crystals: mp 97-99 °C (lit.10 mp 95-97
°C); 1H NMR (500 MHz, CDCl3) δ 7.43 (d, 2, J ) 6.9), 7.39 (t,
2, J ) 7.0), 7.32 (t, 1, J ) 7.1), 7.10 (d, 2, J ) 8.6), 6.93 (d, 2,
J ) 8.6), 5.05 (s, 2), 3.63 (dd, 1, J ) 3.9, 10.6), 3.36 (dd, 1,
difficult to reconcile with the required reaction conditions.
The synthesis of 2 provides solutions to many of these
challenges, while highlighting the structural issue that
must still be surmounted in assembling its elusive isomer
1.
Exp er im en ta l Section 41
Syn th esis. (Z)-4-(ter t-Bu tyld im eth ylsiloxy)bu t-2-en -1-
ol (2.1). NaH (60% oil dispersion, 9.60 g, 0.24 mol) was
washed with hexanes (3 × 50 mL) and suspended in ether (400
mL). (Z)-2-Butene-1,4-diol (16.5 mL, 0.20 mol) was added
dropwise to this suspension and the mixture was stirred for 1
h at rt. The mixture was then cooled to 0 °C and tert-
butyldimethylsilyl chloride (30.1 g, 0.2 mol) was added in
portions. After 3 h, the reaction mixture was diluted with
ether (1.5 L), washed with saturated NaHCO3 (3 × 600 mL),
and dried over MgSO4. Removal of the solvent gave the
desired compound (40.7 g, 100%) as a beige oil. The crude
product was usually carried on directly, but analytical samples
could be obtained by chromatography with 30% EtOAc/
hexanes which gave the silyl ether 2.1 as a clear oil: 1H NMR
(500 MHz, CDCl3) δ 5.69 (m, 2), 4.25 (d, 2, J ) 5.3), 4.20 (d, 2,
J ) 5.3), 1.68 (s, 1), 0.91 (s, 9), 0.09 (s, 6); 13C NMR (100 MHz,
CDCl3) δ 131.2, 130.0, 59.5, 58.7, 25.8, 18.3, -5.3; IR (film)
3375 cm-1
. Anal. Calcd for C10H22O2Si: C, 59.35; H, 10.96.
Found: C, 59.54; H, 10.92.
(Z)-4-(ter t-Bu t yld im et h ylsiloxy)b u t -2-en yl Mesyla t e
(2.2). Mesyl chloride (4.20 mL, 54.4 mmol) was added drop-
(41) Gen er a l. Aminoiminomethanesulfonic acid38,39 and (trimeth-
ylsilyl)ethanesulfonyl chloride (SESCl)25 were prepared by following
literature procedures. Column chromatography was performed by the
method of Still, Kahn, and Mitra using 60-mesh silica gel.42 Mass
spectra and combustion analyses were obtained from the Mass
Spectrometry Laboratory and the Microanalytical Laboratory, Depart-
ment of Chemistry, University of California, Berkeley.
(42) Still, W. C.; Kahn, M.; Mitra, A. J . J . Org. Chem. 1978, 43,
2923.

Misrouted Synthesis of a Peptidomimetic
J . Org. Chem., Vol. 61, No. 22, 1996 7691
J ) 7.2, 10.6), 3.08 (m, 1), 2.73 (dd, 1, J ) 5.3, 13.6), 2.47 (dd,
1, J ) 8.5, 13.6), 1.65 (br s, 3); 13C NMR (100 MHz, CDCl3) δ
157.4, 137.0, 130.9, 130.1, 128.6, 127.9, 127.4, 114.9, 70.0, 66.2,
. Anal. Calcd for
C16H19NO2: C, 74.68; H, 7.44; N, 5.44. Found: C, 74.39; H,
7.32; N, 5.45.
tartrate (1.546 g, 6.60 mmol) in CH2Cl2 (40 mL) was cooled to
-30 °C. Titanium(IV) isopropoxide (1.475 g, 5.19 mmol) was
added, and the mixture was stirred at -30 °C for 45 min.
t-BuOOH (4.71 mL of 3.0 M solution in isooctane, 14.14 mmol)
was added, and the mixture was stirred for an additional 15
min at -30 °C. A solution of allylic alcohol 2.6 (2.592 g, 4.71
mmol) in CH2Cl2 (10 mL) was added, and the mixture was
stirred at -15 °C. After 37 h, water (20 mL) was added and
the resultant emulsion was stirred at rt for 1 h. The mixture
was then filtered through Celite, and the solids were thor-
oughly rinsed with CH2Cl2 (500 mL). The combined filtrates
were washed with water (250 mL), and the aqueous layer was
extracted with CH2Cl2 (3 × 250 mL). The combined organic
layers were dried with MgSO4 and concentrated, and the
residue was purified by chromatography with a gradient of
50-80% EtOAc/hexanes to afford the epoxy alcohol 2.7 (2.667
54.2, 39.9; IR (CH2Cl2) 3620, 3400 cm-1
(Z)-4-[(2S)-2-Am in o-3-(4-(p h en ylm eth oxy)p h en yl)p r o-
p oxy]-1-(ter t-bu tyld im eth ylsiloxy)bu t-2-en e (2.4). Alcohol
2.3 (6.00 g, 23.3 mmol) was added to a suspension of KH (35%
oil dispersion, 5.87 g, 51.3 mmol), washed with hexanes (3 ×
60 mL), in ether (180 mL). After the generation of H2 ceased,
the mixture was cooled to 0 °C and a solution of mesylate 3
(7.84 g, 28.0 mmol) in ether (50 mL) was added dropwise. The
ice bath was allowed to warm to rt and after 3 h, the reaction
mixture was diluted with saturated NH4Cl (300 mL) and
extracted with CH2Cl2 (3 × 500 mL). The combined organic
extracts were dried over MgSO4 and concentrated, and the
residue was purified by chromatography with 5% methanol/
CH2Cl2 to afford the amino ether 2.4 (6.65 g, 55%) as a clear
oil: 1H NMR (400 MHz, CDCl3) δ 7.36 (m, 5), 7.11 (d, 2, J )
8.6), 6.91 (d, 2, J ) 8.6), 5.69 (ddddd, 1, J ) 1.3, 1.3, 5.8, 5.8,
11.4), 5.59 (ddddd, 1, J ) 1.5, 1.5, 6.2, 6.2, 12.5), 5.04 (s, 2),
4.24 (dd, 2, J ) 1.0, 5.8), 4.05 (d, 2, J ) 6.0), 3.41 (dd, 1, J )
3.8, 8.8), 3.25 (dd, 1, J ) 7.2, 8.7), 3.18 (m, 1), 2.49 (dd, 1, J )
5.4, 13.6), 1.74 (s, 2), 0.90 (s, 9), 0.07 (s, 6); 13C NMR (100 MHz,
CDCl3) δ 157.3, 137.0, 132.5, 131.0, 130.1, 128.5, 127.9, 127.4,
126.8, 114.8, 74.9, 70.0, 66.9, 59.5, 52.4, 39.7, 25.9, 18.3, -5.2;
IR (film) 3300, 1615 cm-1. Anal. Calcd for C26H39NO3Si: C,
70.70; H, 8.90; N, 3.17. Found: C, 70.74; H, 8.81; N, 3.23.
(Z)-1-(ter t-Bu t yld im et h ylsiloxy)4-[(2S)-2-[(flu or en yl-
m eth oxyca r bon yl)a m in o]-3-(4-(p h en ylm eth oxy)p h en yl)-
p r op oxy]bu t-2-en e (2.5). Amino ether 2.4 (6.26 g, 14.2
mmol) and FmocOSu (7.17 g, 21.2 mmol) were dissolved in
THF (120 mL) and aqueous 10% Na2CO3 (22 mL) was added.
After 23 h, the THF was removed. The resultant slurry was
diluted with water (250 mL) and extracted with CH2Cl2 (5 ×
500 mL). The combined organic extracts were dried with
MgSO4 and concentrated, and the residue was purified by
chromatography with 15% EtOAc/hexanes to afford the pro-
tected product 2.5 (8.41 g, 89%) as a white powder: mp 68.5-
69.0 °C; 1H NMR (400 MHz, CDCl3) δ (rotamers) 7.79 (d, 2, J
) 7.0), 7.59 (d, 2, J ) 6.5), 7.36 (m, 9), 7.15 (m, 2), 6.93 (d, 2,
J ) 6.7), 5.74 (m, 1), 5.65 (m, 1), 5.11 (m, 1), 5.04 (s, 2), 4.46
(m, 1), 4.36 (m, 1), 4.25 (m, 3), 4.08 (m, 2), 4.03 (m, 1), 3.40
(m, 2), 2.88 (m, 2), 0.95 (s, 9), 0.12 (s, 6); 13C NMR (100 MHz,
CDCl3) δ (rotamers) 157.4, 155.8, 143.9, 141.2, 137.0, 132.7,
130.3, 130.2, 128.5, 127.9, 127.6, 127.4, 127.0, 126.6, 125.0,
119.9, 114.8, 70.0, 69.9, 66.9, 66.5, 59.5, 52.1, 47.2, 36.8, 25.9,
18.3, -5.2; IR (film) 3410, 3330, 2950, 2925, 2855, 1720, 1610,
1510, 1450, 1240, 1075, 840, 775, 740 cm-1. Anal. Calcd for
C41H49NO5Si: C, 74.17; H, 7.44; N, 2.11. Found: C, 73.82; H,
7.45; N, 2.04.
1
g, 100%) as a white solid: mp 95-97 °C; H NMR (500 MHz,
CDCl3) δ 7.76 (d, 2, J ) 7.5), 7.58 (t, 2, J ) 7.6), 7.37 (m, 9),
7.13 (d, 2, J ) 8.0), 6.91 (d, 2, J ) 8.0), 5.40 (d, 1, J ) 8.6),
4.99 (s, 2), 4.45 (t, 1, J ) 6.9), 4.36 (t, 1, J ) 6.9), 4.20 (t, 1, J
) 6.5), 4.03 (s, 1), 3.73 (d, 2, J ) 5.0), 3.67 (dd, 1, J ) 3.4,
11.3), 3.54 (dd, 1, J ) 5.7, 10.8), 3.44 (m, 2), 3.20 (m, 3), 2.82
(d, 1, J ) 6.9); 13C NMR (100 MHz, CDCl3) δ 157.5, 156.0,
143.8, 141.2, 136.9, 130.2, 129.8, 128.5, 127.9, 127.6, 127.4,
127.0, 125.0, 119.9, 114.8, 71.5, 69.9, 69.0, 66.4, 60.4, 55.6, 54.8,
52.0, 47.2, 36.7; IR (CH2Cl2) 3600, 3440, 1727, 1517, 920 cm-1
.
Anal. Calcd for C35H35NO6: C, 74.31; H, 6.24; N, 2.48.
Found: C, 74.26; H, 6.29; N, 2.48.
(2R,3S)-3-[[(2S)-2-Am in o-3-(4-(p h en ylm et h oxy)p h en -
yl)p r op oxy]m eth yl]oxir a n em eth a n ol (2.8). DBU (0.62
mL, 4.12 mmol) was added to a solution of epoxide 2.7 (2.12
g, 3.74 mmol) in CH2Cl2 (19 mL). After 1.5 h, the solvent was
removed, and the resultant oil was purified by chromatography
with a gradient of 10-20% methanol/CH2Cl2 to afford the
amino epoxide 2.8 (1.29 g, 100%) as a clear oil, which slowly
cyclized to the morpholine 2.9 at rt: 1H NMR (400 MHz,
CDCl3) δ 7.43 (d, 2, J ) 7.5), 7.38 (t, 2, J ) 7.6), 7.32 (t, 1, J
) 7.6), 7.09 (d, 2, J ) 8.6), 6.91 (d, 2, J ) 8.6), 5.03 (s, 2), 3.78
(dd, 1, J ) 5.7, 12.3), 3.71 (dd, 1, J ) 5.3, 12.3), 3.67 (dd, 2, J
) 3.6, 5.2), 3.48 (dd, 1, J ) 3.7, 9.4), 3.39 (dd, 1, J ) 7.1, 9.3),
3.22 (m, 2), 2.77 (br s, 4), 2.72 (dd, 1, J ) 5.6, 13.7), 2.53 (dd,
1, J ) 8.2, 13.6); 13C NMR (100 MHz, CDCl3) δ 157.5, 137.0,
130.1, 130.0, 128.5, 127.9, 127.4, 114.9, 75.0, 70.0, 68.9, 60.0,
55.9, 54.7, 52.4, 39.3; IR (film) 3340, 3280, 3150, 2900, 2850,
1795, 1605, 1565, 1505, 1230, 1100 cm-1
.
(2S ,6R )-6-[(1S )-1,2-Dih yd r o x ye t h yl]-2-[[4-(p h e n y l-
m eth oxy)p h en yl]m eth yl]m or p h olin e (2.9). A solution of
amino epoxide 2.8 (1.28 g, 3.73 mmol) in dry dioxane (20 mL)
was heated to reflux. After 9 h, the solvent was removed and
the resultant oil was purified by chromatography with 15%
methanol/ CH2Cl2 to afford the morpholine 2.9 (0.917 g, 72%)
as a white solid: mp 85-86 °C; 1H NMR (400 MHz, CDCl3) δ
7.38 (m, 5), 7.09 (d, 2, J ) 8.6), 6.91 (d, 2, J ) 8.6), 5.02 (s, 2),
3.70 (m, 4), 3.59 (dd, 1, J ) 4.4, 11.0), 3.50 (dd, 1, J ) 4.4,
11.4), 3.37 (dd, 1, J ) 6.4, 11.3), 3.10 (ddd, 1, J ) 3.0, 6.9,
13.5), 3.00 (m, 1), 2.71 (m, 2); 13C NMR (125 MHz, CDCl3) δ
157.3, 136.8, 130.1, 129.9, 128.4, 127.8, 127.3, 114.8, 70.4, 69.8,
68.9, 67.8, 63.9, 51.4, 51.4, 36.7; IR (CH2Cl2) 3680, 3600, 3340
cm-1; MS (EI) m/ z 344 (MH+), 282, 146. Anal. Calcd for
C20H25NO4: C, 69.95; H, 7.34; N, 4.08. Found: C, 69.72; H,
7.36; N, 3.98.
(2S,6R)-6-[(1S)-1-Hyd r oxy-2-(m eth a n esu lfon yloxy)eth -
yl]-2-[[4-(ph en ylm eth oxy)ph en yl]m eth yl]m or ph olin e (4.1).
Diol 2.9 (0.659 g, 1.92 mmol) was dissolved in CH2Cl2 (100
mL), and triethylamine (0.294 mL, 2.11 mmol) was added. The
mixture was cooled to -5 °C, and MsCl was added (0.156 mL,
2.01 mmol) over 1 h. After an additional 10 min, the mixture
was diluted with CH2Cl2 (200 mL), washed with saturated
NaCl (2 × 100 mL), dried with MgSO4, and concentrated, and
the residue was purified by chromatography with 5% methanol/
CH2Cl2 to afford the hydroxy mesylate 4.1 (0.649 g, 80%) as a
white powder: 1H NMR (500 MHz, CDCl3) δ 7.41 (d, 2, J )
7.1), 7.36 (t, 2, J ) 7.2), 7.31 (t, 1, J ) 7.2), 7.08 (d, 2, J ) 8.5),
6.90 (d, 2, J ) 8.5), 5.01 (s, 2), 4.29 (dd, 1, J ) 2.9, 11.0), 4.14
(dd, 1, J ) 5.2, 11.0), 4.00 (ddd, 1, J ) 3.0, 5.0, 7.9), 3.71 (dd,
1, J ) 3.2, 11.4), 3.68 (dd, 1, J ) 3.3, 11.9), 3.60 (dd, 1, J )
4.3, 11.9), 3.31 (dd, 1, J ) 7.3, 11.2), 3.08 (ddd, 1, J ) 3.0, 7.1,
(Z)-4-[(2S)-2-[(F lu or en ylm eth oxyca r bon yl)a m in o]-3-(4-
(p h en ylm eth oxy)p h en yl)p r op oxy]-2-bu ten -1-ol (2.6). HF
(40% in water, 11 mL) was added to a solution of silyl ether
2.5 (14.64 g, 22.0 mmol) in acetonitrile (110 mL) contained in
a polypropylene bottle. After 20 min, the reaction mixture was
poured into saturated NaHCO3 (400 mL) and extracted with
CHCl3 (5 × 500 mL). The combined organic extracts were
dried over MgSO4 and concentrated, and the residue was
purified by chromatography with a step gradient of 20-50%
EtOAc/hexanes to afford the alcohol 2.6 (11.06 g, 92%) as a
white solid: mp 97-98 °C; 1H NMR (500 MHz, CDCl3) δ 7.75
(d, 2, J ) 7.4), 7.56 (t, 2, J ) 7.6), 7.38 (m, 9), 7.11 (d, 2, J )
7.8), 6.90 (br d, 3, J ) 7.8), 5.79 (m, 1), 5.40 (m, 1), 5.15 (d, 1,
J ) 8.0), 5.00 (s, 2), 4.42 (t, 1, J ) 10.1), 4.33 (t, 1, J ) 10.1),
4.25-3.90 (m, 6), 3.38 (br s, 2), 2.81 (d, 2, J ) 6.7), 2.26 (s, 1);
13C NMR (125 MHz, CDCl3) δ 157.5, 156.0, 144.0, 141.4, 137.1,
132.4, 130.4, 130.1, 128.6, 128.0, 127.7, 127.6, 127.1, 125.4,
125.1, 120.0, 114.9, 70.4, 70.0, 66.6, 58.6, 52.5, 47.3, 43.5, 36.5;
IR (CH2Cl2) 3600, 3440, 1725, 1515 cm-1
. Anal. Calcd for
C35H35NO5: C, 73.36; H, 7.70; N, 4.28. Found: C, 73.18; H,
7.65; N, 4.22.
(2R,3S)-3-[[(2S)-2-((Flu or en ylm eth oxycar bon yl)am in o)-
3-(4-(p h e n y lm e t h o x y )p h e n y l)p r o p o x y ]m e t h y l]o x i -
r a n em eth a n ol (2.7). A suspension of powdered 4Å sieves
(0.510 g, dried in vacuo at 150 °C) and D-(-)-diisopropyl

7692 J . Org. Chem., Vol. 61, No. 22, 1996
Koslowski and Bartlett
10.1), 2.98 (s, 3), 2.92 (dd, 1, J ) 3.9, 7.9), 2.64 (d, 2, J ) 7.1);
13C NMR (125 MHz, CDCl3) δ 157.4, 136.9, 130.0, 129.9, 128.4,
127.8, 127.4, 114.9, 71.2, 71.1, 69.9, 67.3, 66.3, 51.1, 50.8, 37.3,
37.1; IR (film) 3500, 3380, 3300, 3020, 2900, 2850, 1605, 1505,
1450, 1345, 1235, 1170, 1100, 955 cm-1; HRMS (FAB) m/ z
calcd for C21H28NO6S 422.1637, found 422.1649, 422 (MH+),
326, 307, 282, 224, 154, 136.
CDCl3) δ 7.54 (m, 3), 7.43 (m, 2), 7.16-7.26 (m, 5), 3.68 (s, 3),
3.52 (dd, 2, J ) 7.0, 7.0), 3.34 (dd, 1, J ) 4.8, 7.7), 1.60 (m, 3),
1.43 (m, 3); 13C NMR (100 MHz, CDCl3) δ 175.9, 154.5, 134.4,
130.4, 130.1, 129.5, 128.4, 127.7, 127.6, 127.2, 124.2, 123.1,
53.8, 51.9, 41.6, 31.4, 25.0; IR (film) 3370, 3300, 3050, 2940,
1750, 1600, 1435, 1355, 1050 cm-1
.
Anal. Calcd for
C21H22N2O4: C, 68.84; H, 6.05; N, 7.65. Found: C, 68.59; H,
6.05; N, 7.40.
(2S,6R)-6-[(1S)-1-(ter t-Bu tyld im eth ylsiloxy)-2-((m eth -
a n esu lfon yloxy)et h yl]-2-[[4-(p h en ylm et h oxy)p h en yl]-
m eth yl]m or p h olin e (4.2). Note: use of 1 equiv of TBDM-
SOTf results in formation of the azetidine via intramolecular
reaction of the desired product. Compound 4.1 (0.645 g, 1.53
mmol) was dissolved in CH2Cl2 (15 mL), and collidine (0.81
mL, 6.12 mmol) and TBDMSOTf (1.05 mL, 4.59 mmol) were
added. After 75 min, the mixture was diluted with CH2Cl2
(200 mL), washed with saturated NaCl (2 × 80 mL), dried with
MgSO4, and concentrated, and the residue was purified by
chromatography with 2.5% methanol/CH2Cl2 to afford the silyl
ether 4.2 (0.811 g, 99%) as a white powder: 1H NMR (400
MHz, CDCl3) δ 7.36 (m, 5), 7.07 (d, 2, J ) 8.6), 6.89 (d, 2, J )
8.6), 5.02 (s, 2), 4.29 (dd, 1, J ) 3.8, 10.5), 4.09 (dd, 1, J ) 5.1,
10.5), 3.90 (m, 1), 3.79 (dd, 1, J ) 3.2, 11.3), 3.62 (dd, 1, J )
3.2, 11.2), 3.49 (d, 1, J ) 11.2), 3.48 (dd, 1, J ) 2.5, 11.1), 3.08
(ddd, 1, J ) 3.2, 6.8, 6.8), 3.02 (m, 1), 2.96 (s, 3), 2.79 (dd, 1, J
) 8.3, 13.6), 2.67 (dd, 1, J ) 6.5, 13.6), 1.95 (s, 1), 0.83 (s, 9),
0.09 (s, 3), -0.02 (s, 3); 13C NMR (100 MHz, CDCl3) δ 157.4,
137.1, 130.9, 130.1, 128.5, 127.8, 127.3, 115.0, 70.8, 70.4, 70.0,
70.0, 68.4, 52.1, 52.0, 37.3, 36.8, 25.6, 17.9, -4.4, -5.0; IR (film)
3420, 2930, 2855, 1610, 1515, 1190, 1175, 840 cm-1; HRMS
(FAB) m/ z calcd for MH+ C27H42NO6SiS 536.2502, found
536.2502, 536 (56), 478 (10), 440 (100), 281 (6), 242 (56), 197
(8), 115 (9).
Meth yl (2S)-5-(2,3-Dih yd r o-2-oxo-4,5-d ip h en yl-1,3-ox-
a zol-3-yl)-2-[[2-(t r im et h ylsilyl)et h yl]su lfon a m id o]p en t -
a n oa te (6.3). Compound 6.2 (1.06 g, 2.9 mmol) and trieth-
ylamine (4.0 mL, 29.0 mmol) were dissolved in DMF (7 mL),
and the mixture was cooled to 0 °C. 2-(Trimethylsilyl)-
ethanesulfonyl chloride (0.87 g, 4.35 mmol) dissolved in DMF
(8 mL) was added dropwise to the above mixture over 30 min.
After the reaction mixture was stirred an additional 2.5 h at
0 °C, the mixture was diluted with water (100 mL) and
extracted with ether (4 × 200 mL). The combined organic
extracts were dried over MgSO4 and concentrated, and the
residue was purified by chromatography with 40% EtOAc/
hexanes to afford compound 6.3 (1.30 g, 84%) as a white
solid: mp 49-50 °C; 1H NMR (400 MHz, CDCl3) δ 7.52 (m, 3),
7.38-7.41 (m, 2), 7.12-7.22 (m, 5), 5.28 (d, 1, J ) 9.2), 3.98
(m, 1), 3.70 (s, 3), 3.53 (dd, 2, J ) 6.3, 6.3), 2.87 (m, 2), 1.76
(m, 1), 1.58 (m, 3), 1.02 (m, 2), 0.01 (s, 9); 13C NMR (100 MHz,
CDCl3) δ 172.3, 154.5, 134.4, 130.3, 130.1, 129.5, 128.3, 127.5,
127.5, 126.9, 124.1, 123.0, 55.3, 52.5, 49.5, 41.1, 29.8, 24.7, 10.2,
-2.2; IR (film) 3250, 3050, 2940, 2245, 1740, 1600 cm-1. Anal.
Calcd for C26H34N2O6SiS: C, 58.84; H, 6.46; N, 5.28. Found:
C, 58.81; H, 6.38; N, 5.03.
(2S)-5-(2,3-Dih yd r o-2-oxo-4,5-d ip h en yl-1,3-oxa zol-3-yl)-
2-[[2-(tr im eth ylsilyl)eth yl]su lfon a m id o]p en ta n oic Acid
(6.4). Compound 6.3 (5.60 g, 10.6 mmol) was dissolved in THF
(100 mL) and water (50 mL), 2 N LiOH (53 mL, 106 mmol)
was added, and the mixture was stirred at rt. After 2.5 h,
the mixture was diluted with water (300 mL), acidified to pH
1 with concentrated HCl, and extracted with CH2Cl2 (4 × 400
mL). The combined organic extracts were dried over MgSO4
and concentrated, and the residue was purified by chroma-
tography with a gradient of 5-10% methanol/CH2Cl2 to afford
(2S,6R,7R)-7-(ter t-Bu tyld im eth ylsiloxy)-2-[[4-(p h en yl-
m et h oxy)p h en yl]m et h yl]-4,1-oxa za b icyclo[4.2.0]oct a n e
(4.3). Mesylate 4.2 (0.870 g, 1.62 mmol) and triethylamine
(1.13 mL, 8.12 mmol) were dissolved in acetonitrile (33 mL),
and the solution was heated to reflux. After 10 h, the reaction
mixture was concentrated and the oil was purified by chro-
matography with 40% EtOAc/hexanes to afford the bicyclic
azetidine 4.3 (0.568 g, 79%) as a white crystalline solid: mp
49-51 °C; 1H NMR (400 MHz, CDCl3) δ 7.38 (m, 5), 7.11 (d, 2,
J ) 8.5), 6.90 (d, 2, J ) 8.5), 5.04 (s, 2), 4.56 (dd, 1, J ) 6.4,
12.8), 4.22 (dd, 1, J ) 11.5, 11.5), 3.96 (dddd, 1, J ) 1.2, 6.0,
6.0, 11.4), 3.68 (dd, 1, J ) 6.0, 11.4), 3.66 (dd, 1, J ) 3.4, 8.2),
3.53 (m, 2), 3.47 (dd, 1, J ) 2.9, 11.6), 2.75 (dd, 1, J ) 8.9,
13.5), 2.70 (dd, 1, J ) 6.3, 13.6), 2.55 (m, 1), 0.88 (s, 9), 0.03
(s, 6); 13C NMR (125 MHz, CDCl3) δ 157.2, 137.2, 131.8, 130.3,
128.5, 127.8, 127.4, 114.7, 70.0, 64.0, 63.3, 61.8, 60.4, 59.2, 55.8,
37.7, 25.6, 17.9, -5.2, -5.3; IR (film) 2920, 2850, 1610, 1510,
1
acid 6.4 (5.27 g, 97%) as a white powder: mp 140-42 °C; H
NMR (400 MHz, 1:1 CDCl3/CD3OD) δ 7.46 (m, 3), 7.28 (m, 2),
7.13 (m, 5), 3.82 (s, 1), 3.51 (m, 1), 3.31 (m, 1), 2.87 (dd, 2, J )
8.8, 8.8), 1.77 (s, 1), 1.59 (s, 3), 0.94 (m, 2), -0.04 (s, 9); 13C
NMR (100 MHz, 1:1 CDCl3/ CD3OD) δ 177.9, 154.7, 134.5,
130.0, 129.9, 129.2, 128.0, 127.4, 127.0, 126.1, 123.8, 123.2,
56.2, 53.0, 41.0, 29.7, 24.5, 9.6, -2.8; IR (KBr) 3460, 3250,
3030, 2945, 1745, 1590, 1580, 1440, 1385, 1315, 1255, 1130,
855, 840, 750, 690 cm-1; HRMS (FAB) m/ z calcd for MH+
C25H32N2O6SiS 516.1750, found 516.1746, 555 (MK+), 539
(MNa+), 516 (M+), 493, 425, 366, 237, 172; UV-Vis (CH2Cl2)
λmax 220, 284 nm.
+
1335, 840 cm-1; MS (FAB) m/ z 441 (MH2 ), 440 (MH ), 439,
438, 281, 242, 197, 115. Anal. Calcd for C26H37NO3Si: C,
71.03; H, 8.48; N, 3.19. Found: C, 69.91; H, 8.30; N, 3.37.
Meth yl (2S)-5-(2,3-Dih yd r o-2-oxo-4,5-d ip h en yl-1,3-ox-
a zol-3-yl)-2-a m in op en ta n oa te (6.2). A mixture of Me4NOH
(20% in methanol w/v, 7.0 mL, 15.3 mmol) and NR-Boc-L-
ornithine (3.56 g, 15.3 mmol) was triturated with absolute
EtOH (3 × 30 mL). The residual solvent was removed to yield
the tetramethylammonium salt as a white powder. The salt
was dissolved in dry DMF (15 mL), and 4,5-diphenyl-1,3-
dioxolone (4.02 g, 16.8 mmol) was added. After being stirred
for 3 h, the reaction mixture was acidified to pH 1 with 2 N
HCl, diluted with EtOAc (200 mL), washed with 0.5 N HCl
(70 mL) and saturated NaCl (2 × 70 mL), and dried over
MgSO4. Removal of the solvent yielded the hydroxyoxazoli-
dinone 6.1 as a white foam.
+
(3S,4R,7S)-4-(ter t-Bu tyldim eth ylsiloxy)-3-ch lor o-6-[(2S)-
5-(2,3-d ih yd r o-2-oxo-4,5-d ip h en yl-1,3-oxa zol-3-yl)-2-[[2-
(t r im e t h ylsilyl)e t h yl]su lfon a m id o]p e n t a n o y l]-7-[[4-
(p h en ylm et h oxy)p h en yl]m et h yl]-1,6-oxa za cyclooct a n e
(4.5s). Azetidine 4.3 (0.292 g, 0.66 mmol) and acid 6.4 (1.028
g, 1.99 mmol) were suspended in THF (6.6 mL), and the
solution was cooled to 0 °C. PCl5 (1.105 g, 5.31 mmol) was
added, and the solution was stirred at 0 °C. After 1 h 45 min,
DIEA (0.11 mL, 0.64 mmol) was added and stirring was
continued at 0 °C. After an additional 15 min, the solution
was made basic with DIEA, diluted with EtOAc (200 mL), and
washed with 0.5 N HCl (2 × 70 mL) and saturated NaHCO3
(2 × 70 mL). The organic layer was dried with MgSO4 and
concentrated, and the residue was purified by chromatography
with a gradient of 20-30% EtOAc/hexanes to afford the ring-
opened compound 4.5s (0.429 g, 66%) as a white foam: 1H
NMR (400 MHz, CDCl 3) δ 7.54 (m, 3), 7.29-7.43 (m, 7), 7.17-
7.25 (m, 5), 7.00 (d, 2, J ) 8.6), 6.90 (d, 2, J ) 8.6), 5.11 (d, 1,
J ) 9.6), 5.05 (d, 1, J ) 11.7), 5.00 (d, 1, J ) 11.7), 4.29 (t, 1,
J ) 3.9), 4.20 (t, 1, J ) 10.9), 4.09 (m, 2), 4.01 (m, 2), 3.83 (m,
1), 3.48 (t, 3, J ) 6.8), 3.38 (d, 1, J ) 14.0), 3.26 (dd, 1, J )
10.2, 13.6), 2.91 (m, 3), 2.60 (dd, 1, J ) 4.7, 13.6), 1.69 (m, 2),
1.56 (m, 1), 1.26 (m, 1), 1.05 (m, 2), 0.92 (s, 9), 0.12 (s, 3), 0.04
(s, 3), 0.03 (s, 9); 13C NMR (100 MHz, CDCl3) δ 171.3, 157.4,
154.6, 136.9, 134.5, 130.3, 130.3, 130.2, 129.9, 129.7, 128.4,
The crude hydroxyoxazolidinone was dissolved in methanol
(340 mL) and cooled to -5 °C. Thionyl chloride (7.5 mL, 102.5
mmol) was added dropwise. The cooling bath was then
removed, and the reaction mixture was allowed to warm to
rt. After 23 h, the solvent was removed exhaustively at 50 °C
with a rotary evaporator. The resultant oil was dissolved in
saturated NaHCO3 (100 mL), and the mixture was extracted
with EtOAc (3 × 200 mL). The combined organic extracts were
dried over MgSO4 and concentrated, and the residue was
purified by chromatography with a gradient of 2.5-10%
methanol/CH2Cl2 to afford the Nδ-protected ornithine analog
6.2 (5.12 g, 93%) as a clear colorless oil: 1H NMR (400 MHz,

Misrouted Synthesis of a Peptidomimetic
J . Org. Chem., Vol. 61, No. 22, 1996 7693
128.4, 127.8, 127.6, 127.4, 127.3, 126.9, 124.2, 123.0, 114.8,
74.0, 70.5, 70.2, 69.9, 62.2, 54.1, 53.8, 49.6, 41.2, 33.1, 29.9,
26.4, 25.7, 25.2, 17.9, 10.1, -2.1, -4.4, -5.2; IR (film) 3240,
2951, 2858, 1754, 1648, 1510, 1250 cm-1; HRMS (FAB) m/ z
calcd for MH+ C51H69ClN3O8Si2S 974.4032, found 974.4053, 996
(MNa+), 974 (MH+), 956, 946, 916, 882, 808, 737, 684, 643,
573, 502, 476.
concentrated, and the residue was purified by chromatography
with 30% EtOAc/hexanes to afford the bicyclic product 7.3s
(0.062 g, 57%) as a white foam: 1H NMR (400 MHz, CDCl3) δ
7.51 (m, 5), 7.29-7.43 (m, 5), 7.24 (m, 2), 7.18 (m, 3), 7.02 (d,
2, J ) 8.6), 6.86 (d, 2, J ) 8.7), 5.02 (s, 2), 4.17 (s, 1), 4.08 (dd,
1, J ) 5.2, 13.4), 4.02 (m, 1), 3.89 (m, 2), 3.81 (m, 2), 3.70 (d,
1, J ) 13.4), 3.63 (dd, 1, J ) 3.6, 14.5), 3.59 (dd, 1, J ) 6.8,
14.2), 3.55 (dd, 1, J ) 6.7, 14.2), 3.38 (m, 2), 2.96 (ddd, 1, J )
1.4, 9.7, 9.7, 2), 2.81 (m, 1), 2.02 (m, 2), 1.78 (m, 1), 1.52 (m,
1), 0.97 (m, 2), 0.08 (s, 9); 13C NMR (100 MHz, CDCl3) δ 167.7,
157.4, 154.7, 136.9, 134.3, 130.6, 130.6, 130.0, 129.5, 129.4,
128.5, 128.4, 127.9, 127.9, 127.4, 127.4, 127.3, 124.2, 123.4,
114.9, 70.6, 70.1, 70.0, 70.0, 61.3, 59.5, 57.7, 49.8, 49.6, 41.8,
33.7, 27.4, 25.3, 9.9, -2.0; IR (CH2Cl2) 3061, 2954, 1750, 1677,
(3S,4R,7S)-3-Ch lor o-6-[(2S)-5-(2,3-d ih yd r o-2-oxo-4,5-
d ip h en yl-1,3-oxa zol-3-yl)-2-[[2-(tr im eth ylsilyl)eth yl]su l-
fon a m id o]p en ta n oyl]-4-h yd r oxy-7-[[4-(p h en ylm eth oxy)-
p h en yl]m eth yl]-1,6-oxa za cycloocta n e (7.1s). Compound
4.5s (0.220 g, 0.23 mmol) was dissolved in THF (5 mL), and
TBAF (1.0 M in THF, 0.23 mL, 0.23 mmol) was added. After
30 min, the solution was diluted with CH2Cl2 (100 mL). The
organic layer was washed with 0.5 N HCl (2 × 50 mL) and
saturated NaHCO3 (2 × 50 mL), dried with MgSO4, and
concentrated, and the residue was purified by chromatography
with 40% EtOAc/hexanes to afford alcohol 7.1s (0.159 g, 82%)
as a white foam: 1H NMR (400 MHz, CDCl3, rotamers 1:1) δ
7.52 (m, 3), 7.37 (m, 7), 7.21 (m, 5), 7.07 (d, 1, J ) 8.5), 6.99
(d, 1, J ) 8.6), 6.88 (d, 1, J ) 8.6), 6.82 (d, 1, J ) 8.6), 5.18 (d,
1, J ) 9.2), 4.96 (m, 2), 4.38 (m, 2), 4.28 (m, 1), 4.17 (m, 1),
4.07 (m, 2), 3.88 (m, 1), 3.73 (m, 2), 3.56 (m, 1), 3.40 (m, 2),
2.86 (m, 4), 1.71 (m, 2), 1.42 (m, 2), 1.00 (m, 2), 0.04 (s, 5),
0.01 (s, 4); 13C NMR (100 MHz, CDCl3, amide rotamers 1:1,
the rotameric carbons are listed parenthetically) δ 174.7
(173.0), 157.5 (157.2), 154.7 (154.3), 136.8 (136.6), 134.6
(134.1), 130.2, 130.2, 130.0, 129.9, 129.4 (129.4), 128.2, 128.2,
127.6 (127.6), 127.5 (127.5), 127.4 (127.3), 127.1 (127.1), 126.7
(126.6), 124.1 (123.9), 123.0 (122.9), 114.9 (114.5), 72.2 (72.0),
70.3 (69.9), 69.6 (69.5), 64.5 (62.0), 58.9, 53.3 (52.4), 49.8, 44.7
(44.2), 40.6, 34.1 (33.4), 29.3 (29.0), 26.8 (26.2), 24.9 (24.5), 10.1
(9.9), -2.2 (-2.2); IR (CH2Cl2) 3450, 3350, 3061, 2954, 1750,
1612, 1511, 1140 cm-1; HRMS (FAB) m/ z calcd for MH+
C45H55ClN3O8SiS 860.3167, found 860.3159, 860 (MH+), 832,
796, 768, 620, 570, 459.
1510, 1343, 1255 cm-1; HRMS (FAB) m/ z calcd for M+ C45H52
-
ClN3O7SiS 841.2983, found 841.2977, 864 (MNa+), 842 (MH+),
814, 767, 750, 676, 605, 586, 515.
(2S ,6S ,7S ,9S )-9-(3-Am in op r op yl)-6-ch lor o-2-(4-(h y-
d r oxyp h en yl)m eth yl)-10-oxo-8-[2-(tr im eth ylsilyl)eth a n e-
su lfon yl]-4,1,8-oxa d ia za bicyclo[5.3.1]u n d eca n e (9.3). A
solution of compound 7.3s (0.153 g, 0.181 mmol), Pd black
(0.096 g, 0.906 mmol), and HOAc (0.52 mL, 9.06 mmol) in
absolute EtOH (5 mL) and 5% water/methanol (52 mL) was
purged with H2 three times and stirred under an H2 atmo-
sphere. After 2 d, the reaction mixture was filtered through
Celite and concentrated. The resultant yellow oil was sus-
pended in water and lyophilized to remove the toluene and
benzil formed during the reaction. This procedure afforded
the amine phenol 9.3 (0.097 g, 100%) as a pale yellow oil which
was used without further purification: 1H NMR (400 MHz,
CD3OD) δ 7.01 (d, 2, J ) 8.4), 6.69 (d, 2, J ) 8.4), 4.16 (m, 1),
4.08 (m, 3), 4.01 (d, 1, J ) 10.9), 3.84 (dd, 1, J ) 3.7, 12.2),
3.78 (m, 3), 3.53 (m, 1), 3.30 (d, 1, J ) 10.0), 3.12 (m, 2), 2.91
(t, 2, J ) 7.3), 2.78 (dd, 1, J ) 6.0, 13.8), 2.13 (m, 2), 1.92 (m,
1), 1.78 (m, 1), 0.99 (m, 2), 0.06 (s, 9); 13C NMR (100 MHz,
CD3OD) δ 169.4, 157.1, 130.8, 130.6, 116.3, 71.9, 71.8, 71.1,
63.9, 61.2, 58.9, 49.8, 46.5, 40.9, 34.6, 28.3, 24.8, 10.3, -2.0;
IR (film) 3410, 3182, 2924, 2851, 1668 cm-1; HRMS (FAB) m/ z
calcd for MH+ C23H39ClN3O5SiS 532.2068, found 532.2057, 544
(MNa+), 532 (MH+), 498, 366, 349, 307.
(2S,6S,7S,9S)-6-Ch lor o-2-((4-h yd r oxyp h en yl)m et h yl)-
10-oxo-9-[3-[[(p h en ylm eth oxy)ca r bon yl]a m in o]p r op yl]-
8-[2-(tr im eth ylsilyl)eth an esu lfon yl]-4,1,8-oxadiazabicyclo-
[5.3.1]u n d eca n e (9.4). NaHCO3 (1 M, 8 mL) was added to a
solution of compound 9.3 (0.096 g, 0.181 mmol) and benzyl-
succinimidyl dicarbonate (0.452 g, 1.81 mmol) in THF (30 mL)
and water (8 mL). After 18 h, the THF was removed by rotary
evaporation and the reaction mixture was diluted with satu-
rated NaHCO3 (10 mL). The mixture was extracted with CH2-
Cl2 (3 × 40 mL), the combined organic layers were dried over
MgSO4 and concentrated, and the residue was purified by
chromatography with 40% EtOAc/hexanes to afford the Cbz
derivative 9.4 (0.105 g, 87%) as a white foam, as well as a small
amount of the di-Cbz compound which could be hydrolyzed
with LiOH in aqueous THF to give more of the desired
product: 1H NMR (400 MHz, CDCl3) δ 7.35 (m, 5), 6.97 (d, 2,
J ) 8.5), 6.75 (d, 2, J ) 8.5), 6.23 (s, 1), 5.24 (t, 1, J ) 4.6),
5.11 (s, 2), 4.20 (s, 1), 4.09 (dd, 1, J ) 5.4, 13.3), 4.01 (d, 1, J
) 10.5), 3.94 (t, 1, J ) 4.3), 3.89 (m, 2), 3.82 (dd, 1, J ) 1.5,
14.7), 3.76 (d, 1, J ) 13.4), 3.65 (dd, 1, J ) 2.7, 14.6), 3.38 (m,
2), 3.24 (dd, 2, J ) 6.1, 12.0), 2.96 (t, 2, J ) 7.7), 2.81 (dd, 1,
J ) 5.6, 13.1), 2.05 (m, 2), 1.76 (m, 1), 1.63 (m, 1), 1.01 (m, 2),
0.05 (s, 9); 13C NMR (100 MHz, CDCl3) δ 168.1, 156.7, 154.7,
136.6, 130.0, 129.7, 128.5, 128.0, 128.0, 115.5, 70.7, 70.2, 70.0,
66.6, 61.2, 59.8, 57.8, 50.1, 49.7, 40.7, 33.7, 27.6, 25.7, 9.9, -2.0;
IR (film) 3347, 2950, 1703, 1673, 1515, 1250 cm-1; HRMS
(FAB) m/ z calcd for MH+ C31H45ClN3O7SiS 666.2436, found
666.2448, 688 (MNa+), 666 (MH+), 622, 574, 532, 500, 466,
351.
(3S,4R,7S)-3-Ch lor o-6-[(2S)-5-(2,3-d ih yd r o-2-oxo-4,5-
d ip h en yl-1,3-oxa zol-3-yl)-2-[[2-(tr im eth ylsilyl)eth yl]su l-
fon a m id o]p e n t a n oyl]-4-(m e t h a n e su lfon yloxy)-7-[[4-
(p h en ylm et h oxy)p h en yl]m et h yl]-1,6-oxa za cyclooct a n e
(7.2s). A solution of alcohol 7.1s (0.169 g, 0.20 mmol) in CH2-
Cl2 (5 mL) was cooled to 0 °C; triethylamine (0.30 mL, 2.16
mmol) and mesyl chloride (0.15 mL, 1.96 mmol) were added
and the cooling bath was removed. After 1.5 h, the solution
was diluted with CH2Cl2 (100 mL), the organic layer was
washed with saturated NaCl (2 × 50 mL), dried, and concen-
trated, and the oily product was chromatographed (40%
EtOAc/hexanes) to afford 0.173 g (94%) of mesylate 7.2s as a
white foam: 1H NMR (400 MHz, CDCl3, 6:1 ratio of rotamers;
data reported only for major rotamer) δ 7.51 (m, 3), 7.35 (m,
7), 7.19 (m, 5), 6.99 (d, 2, J ) 8.5), 6.88 (d, 2, J ) 8.6), 5.39 (m,
1), 5.02 (d, 1, J ) 11.7), 4.98 (d, 1, J ) 11.8), 4.96 (m, 1), 4.37
(m, 1), 4.15 (m, 2), 4.08 (m, 2), 3.65 (m, 2), 3.38 (m, 3), 3.26
(dd, 1, J ) 11.0, 13.8), 3.14 (s, 3), 2.94 (t, 2, J ) 9.0), 2.86 (m,
1), 2.75 (dd, 1, J ) 5.7, 13.9), 1.68 (m, 2), 1.57 (m, 1), 1.47 (m,
1), 1.01 (m, 2), 0.01 (s, 9); 13C NMR (100 MHz, CDCl3, rotamers
1:1, the rotameric carbons are listed paranthetically) δ 172.7
(171.8), 157.5 (157.2), 154.3, 136.8 (136.6), 134.2 (134.1), 130.1,
130.0, 129.9 (129.8), 129.7, 129.4 (129.2), 128.2 (128.0), 127.6,
127.4, 127.2, 127.1, 126.7, 124.0 (123.9), 123.0 (122.9), 114.9
(114.6), 78.0, 69.6, 69.5, 62.1, 53.6 (53.3), 50.6 (50.2), 49.0, 41.0
(40.9), 39.2, 37.7, 34.0, 33.6, 31.3, 29.1, 25.2 (24.5), 9.9 (9.7),
-2.2 (-2.3); IR (film) 3257, 2948, 2350, 1754, 1657, 1511, 1361,
1317, 1171 cm-1; HRMS (FAB) m/ z calcd for MH+ C46H57
-
ClN3O10SiS2 938.2943, found 938.2940, 960 (MNa+), 938
(MH+), 910, 874, 860, 846, 772, 537, 471, 440.
(2S,6S,7S,9S)-6-Ch lor o-9-[3-(2,3-d ih yd r o-2-oxo-4,5-d i-
p h e n yl-1,3-oxa zol-3-yl)p r op yl]-10-oxo-2-[[4-(p h e n yl-
m eth oxy)ph en yl]m eth yl]-8-[2-(tr im eth ylsilyl)eth an esu lfo-
n yl]-4,1,8-oxa d ia za bicyclo[5.3.1]u n d eca n e (7.3s). Mesy-
late 7.2s (0.061 g, 0.065 mmol) was dissolved in DMSO (3.3
mL), and NaH (60% dispersion, 0.026 g, 0.65 mmol) was added.
Two such reactions were performed simultaneously and were
combined for the workup and purification. After 2 h, the
transparent light-orange solutions were diluted with EtOAc
(200 mL). The organic layer was washed with water (40 mL)
and saturated NaCl (2 × 80 mL), dried with MgSO4, and
(2S,6S,7S,9S)-6-Ch lor o-10-oxo-9-[3-[[(p h en ylm eth oxy)-
ca r bon yl]a m in o]p r op yl]-2-[[4-(p h en ylm eth oxy)p h en yl]-
m e t h y l]-8-[2-(t r im e t h y ls ily l)e t h a n e s u lfo n y l]-4,1,8-
oxa d ia za bicyclo[5.3.1]u n d eca n e (9.5). Benzyl bromide (72
µL, 0.603 mmol) was added to a suspension of compound 9.4
(0.040 g, 0.060 mmol) and K2CO3 (0.100 g, 0.724 mmol) in
CHCl3 (3 mL) and methanol (3 mL), and the mixture was
heated to reflux. After 37 h, the reaction mixture was diluted

7694 J . Org. Chem., Vol. 61, No. 22, 1996
Koslowski and Bartlett
with EtOAc (40 mL), washed with 1 N HCl (15 mL) and
saturated NaHCO3 (15 mL), dried over MgSO4, and concen-
trated, and the residue was purified by chromatography with
a gradient of 30-40% EtOAc/hexanes to afford the benzyl ether
9.5 (0.041 g, 90%) as a clear oil: 1H NMR (400 MHz, CDCl3)
δ 7.36 (m, 10), 7.05 (d, 2, J ) 8.6), 6.88 (d, 2, J ) 8.6), 5.12 (s,
1), 5.10 (s, 2), 5.03 (s, 2), 4.21 (s, 1), 4.10 (dd, 1, J ) 5.6, 13.4),
4.03 (d, 1, J ) 10.4), 3.91 (m, 3), 3.83 (dd, 1, J ) 1.4, 14.7),
3.78 (d, 1, J ) 13.5), 3.66 (dd, 1, J ) 2.6, 14.8), 3.40 (m, 2),
3.26 (dd, 2, J ) 5.7, 11.6), 2.97 (t, 2, J ) 7.8), 2.84 (m, 1), 2.08
(m, 2), 1.71 (m, 2), 1.02 (m, 2), 0.06 (s, 9); 13C NMR (100 MHz,
CDCl3) δ 168.1, 157.5, 156.5, 137.0, 136.8, 130.7, 129.6, 128.5,
128.4, 128.0, 128.0, 127.9, 127.4, 115.0, 70.7, 70.2, 70.1, 70.1,
66.5, 61.2, 59.8, 57.8, 50.2, 49.8, 40.7, 33.8, 27.7, 25.9, 10.0,
the resultant oil was purified by chromatography with EtOAc
to afford the adduct 8.3z (0.061 g, 93%) as a white foam: 1H
NMR (400 MHz, CDCl3) δ 8.16 (s, 1), 7.59 (d, 1, J ) 7.8), 7.44
(d, 2, J ) 7.2), 7.39 (t, 2, J ) 7.3), 7.34 (m, 7), 7.17 (dt, 1, J )
1.0, 7.6), 7.09 (dt, 1, J ) 1.0, 7.5), 7.03 (s, 1), 7.02 (d, 2, J )
8.5), 6.88 (d, 2, J ) 8.6), 5.15 (s, 1), 5.09 (s, 2), 5.04 (s, 2), 3.94
(dd, 1, J ) 4.0, 12.3), 3.84 (dd, 1, J ) 10.7, 12.2), 3.72 (dd, 1,
J ) 3.3, 12.3), 3.63 (s, 3), 3.59 (m, 2), 3.34 (dd, 1, J ) 9.0,
13.8), 3.20 (m, 3), 3.10 (m, 4), 3.00 (dd, 1, J ) 7.8, 14.4), 2.94
(t, 1, J ) 3.7), 2.79 (dd, 1, J ) 6.6, 13.9), 2.09 (t, 1, J ) 3.4),
1.82 (m, 1), 1.57 (quin, 2, J ) 7.2), 1.34 (m, 1); 13C NMR (100
MHz, CDCl3) δ 175.5, 172.1, 157.2, 156.5, 137.1, 136.7, 136.1,
131.6, 129.7, 128.6, 128.5, 128.0, 128.0, 127.9, 127.5, 127.4,
122.7, 122.1, 119.4, 118.8, 114.7, 111.9, 111.2, 72.0, 70.0, 70.0,
69.7, 66.5, 60.9, 60.2, 55.5, 52.7, 51.8, 49.2, 40.8, 34.2, 29.6,
27.3, 26.5; IR (film) 3415, 3333, 3033, 2928, 1716, 1651, 1511,
1239 cm-1; HRMS (FAB) m/ z calcd for MH+ C45H52N5O7
774.3867, found 774.3856, 774 (MH+), 556, 512.
(2S,6S,7S,9S)-6-[[(1S)-1-Car boxy-2-in dolyleth yl]am in o]-
10-oxo-9-[3-[[(p h en ylm eth oxy)ca r bon yl]a m in o]p r op yl]-
2-[[4-(p h en ylm et h oxy)p h en yl]m et h yl]-4,1,8-oxa d ia za -
bicyclo[5.3.1]u n d eca n e (8.4z). Two identical reactions were
performed simultaneously and were combined after the cy-
clization. Aqueous LiOH (2 N, 100 µL) and water (1 mL) were
added to a solution of ester 8.3z (16.0 mg, 20.7 µmol) in THF
(2 mL). After 3 h, 2 N HCl (120 µL) and water (10 mL) were
added. The resulting solution was lyophilized to afford the
free acid 8.4z as a white powder which was used without
further purification: 1H NMR (500 MHz, CD3OD) δ 7.62 (d,
1, J ) 7.7), 7.42 (d, 2, J ) 7.2), 7.36 (t, 2, J ) 7.6), 7.31 (m, 7),
7.25 (s, 1), 7.11 (t, 1, J ) 7.1), 7.04 (t, 1, J ) 7.3), 7.02 (d, 2, J
) 8.5), 6.90 (d, 2, J ) 8.5), 5.05 (s, 2), 4.92 (s, 2), 4.10 (m, 1),
3.95 (m, 1), 3.76 (m, 2), 3.52 (m, 2), 3.43 (m, 2), 3.21 (m, 2),
3.12 (t, 2, J ) 6.5), 2.92 (s, 1), 2.85 (d, 1, J ) 13.8), 2.69 (dd,
1, J ) 5.8, 13.6), 2.04 (m, 1), 1.88 (m, 1), 1.58 (m, 2), 1.51 (m,
1).
(3S,6S,8S,12S,14S)-2,5-Dioxo-6-(in d olylm et h yl)-3-[3-
[[(p h e n y lm e t h o x y )c a r b o n y l]a m i n o ]p r o p y l]-12-[[4-
(p h e n ylm e t h oxy)p h e n yl]m e t h yl]-10,1,4,7-oxa t r ia za -
tr icyclo[6.4.2.04,14]tetr a d eca n e (8.5z). DPPA (45 µL, 413
µmol) was added to a 0 °C solution of the free acid 8.4z (crude,
∼20.7 µmol) and NaHCO3 (0.070 g, 827 µmol) in DMF (6 mL,
3 µM), and the solution was stirred vigorously at 0 °C. After
4 h, more DPPA (45 µL, 413 µmol) was added. After an
additional 11 h, the two reaction mixtures were combined,
diluted with EtOAc (160 mL), washed with saturated NaHCO3
(2 × 60 mL), dried over MgSO4, and concentrated. The crude
mixture was partially purified by chromatography with EtOAc
to afford the tricyclic product 8.5z (11.8 mg, 38%) approxi-
mately 50-80% pure.
This material was usually used directly in the next reaction,
since this compound is particularly sensitive to acid and
degrades on silica or in the presence of CH2Cl2 or CHCl3. This
decomposition is especially noticeable since an orange color
develops due to dimerization of the indole groups. For
purposes of characterization, the material was subjected to
several purifications by preparative silica TLC, with EtOAc
as the eluent, to afford the 8.5z (5.9 mg, 19%) in pure form as
a clear oil: 1H NMR (400 MHz, CDCl3) δ 7.76 (s, 1), 7.59 (d, 1,
J ) 7.7), 7.38 (m, 2), 7.15 (m, 2), 7.08 (d, 2, J ) 8.5), 6.87 (d,
2, J ) 8.5), 5.62 (t, 1, J ) 5.1), 5.19 (d, 1, J ) 12.2), 5.11 (d, 1,
J ) 12.2), 5.04 (s, 2), 4.11 (d, 1, J ) 15.5), 3.91 (m, 2), 3.77 (m,
3), 3.55 (m, 3), 3.44 (quin, 2, J ) 7.2), 3.32 (m, 2), 3.25 (dd, 1,
J ) 5.6, 14.7), 3.14 (dd, 1, J ) 4.6, 14.9), 3.08 (dd, 1, J ) 4.6,
10.3), 2.93 (dd, 1, J ) 8.1, 14.7), 2.75 (m, 1), 2.20 (m, 1), 1.76
(m, 1), 1.62 (m, 1), 1.31 (m, 1); HRMS (FAB) m/ z calcd for
MH+ C44H48N5O6 742.3605, found 742.3601, 742 (MH+), 712,
698, 643, 612, 487, 429.
-2.0; IR (film) 3364, 3032, 2950, 1716, 1667, 1510, 1248 cm-1
;
HRMS (FAB) m/ z calcd for MH+ C38H51ClN3O7SiS 756.2906,
found 756.2890, 778 (MNa+), 756 (MH+), 712, 664, 622, 590.
(2S,6S,7S,9S)-6-Ch lor o-10-oxo-9-[3-[[(p h en ylm eth oxy)-
ca r bon yl]a m in o]p r op yl]-2-[[4-(p h en ylm eth oxy)p h en yl]-
m et h yl]-4,1,8-oxa d ia za b icyclo[5.3.1]u n d eca n e
(7.6z).
TBAF (82 µL of a 1.0 M solution in THF, 0.082 mmol) was
added to a solution of compound 9.5 (0.031 g, 0.041 mmol) in
THF (4 mL). After 2 h, methanol (4 mL) and NaBH4 (0.016 g,
0.408 mmol) were added. After an additional 1 h, the reaction
mixture was cooled to 0 °C and concentrated HCl was added
to give a pH 1 solution, which was poured slowly into saturated
NaHCO3 (20 mL). This mixture was extracted with CH2Cl2
(3 × 40 mL), the combined organic layers were dried over
MgSO4 and concentrated, and the residue was purified by
chromatography with 60% EtOAc/hexanes to afford the amine
7.6z (0.021 g, 86%) as a white foam: 1H NMR (500 MHz,
CDCl3) δ 7.42 (d, 2, J ) 7.1), 7.36 (m, 8), 7.06 (d, 2, J ) 8.6),
6.87 (d, 2, J ) 8.6), 5.11 (s, 1), 5.10 (s, 2), 5.03 (s, 2), 4.09 (dd,
1, J ) 3.8, 13.1), 3.94 (m, 2), 3.88 (dd, 1, J ) 3.0, 12.2), 3.68 (s,
2), 3.56 (s, 1), 3.47 (s, 1), 3.41 (m, 2), 3.34 (t, 1, J ) 5.4), 3.26
(m, 1), 3.18 (m, 1), 2.85 (m, 1), 1.89 (m, 1), 1.61 (m, 2), 1.41
(m, 1); 13C NMR (100 MHz, CDCl3) δ 171.6, 157.2, 156.4, 137.0,
136.6, 131.2, 129.6, 128.5, 128.4, 128.0, 127.9, 127.8, 127.4,
114.7, 72.5, 70.0, 69.9, 69.8, 66.5, 63.0, 55.5, 55.1, 49.0, 40.7,
34.1, 27.3, 26.3; IR (film) 3337, 3031, 2929, 1716, 1661, 1511,
1241 cm-1; HRMS (FAB) m/ z calcd for MH+ C33H39ClN3O5
592.2578, found 592.2575, 592 (MH+), 556, 487, 456, 307.
(2S,6R,7S,9S)-6-Iod o-10-oxo-9-[3-[[(p h en ylm et h oxy)-
ca r bon yl]a m in o]p r op yl]-2-[[4-(p h en ylm eth oxy)p h en yl]-
m eth yl]-4,1,8-oxa d ia za bicyclo[5.3.1]u n d eca n e (8.1z).
A
solution of compound 7.6z (0.055 g, 0.093 mmol) and NaI (1.39
g, 9.29 mmol) in acetone (20 mL) was heated to reflux. After
21 h, the reaction mixture was cooled, diluted with CH2Cl2
(80 mL), washed with 50% saturated NaCl (20 mL) and
saturated NaCl (20 mL), dried over MgSO4, and concentrated,
and the residue was purified by chromatography with a
gradient of 50-60% EtOAc/hexanes to afford the iodide 8.1z
(0.059 g, 94%) as a white foam: 1H NMR (500 MHz, CDCl3) δ
7.42 (d, 2, J ) 7.1), 7.36 (m, 8), 7.06 (d, 2, J ) 8.6), 6.87 (d, 2,
J ) 8.6), 5.14 (t, 1, J ) 5.8), 5.09 (s, 2), 5.02 (s, 2), 4.04 (d, 1,
J ) 9.6), 4.01 (d, 1, J ) 7.9), 3.85 (dd, 2, J ) 2.7, 12.0), 3.75 (t,
1, J ) 13.6), 3.69 (s, 1), 3.68 (s, 1), 3.40 (m, 2), 3.25 (t, 2, J )
6.2), 3.17 (m, 2), 2.84 (dd, 1, J ) 5.1, 12.4), 1.88 (m, 1), 1.61
(quin, 2, J ) 7.0), 1.40 (m, 1); 13C NMR (125 MHz, CDCl3) δ
172.0, 157.3, 156.5, 137.1, 136.7, 131.4, 129.7, 128.5, 128.4,
128.0, 128.0, 127.9, 127.4, 114.8, 71.6, 70.6, 70.3, 70.0, 66.5,
57.0, 56.0, 50.3, 40.8, 39.9, 34.2, 27.4, 26.4; IR (film) 3337, 3031,
2932, 1718, 1653, 1509, 1240 cm-1; HRMS (FAB) m/ z calcd
for MH+ C33H39IN3O5 684.1934, found 684.1923, 684 (MH+),
664, 556, 307.
(2S,6S,7S,9S)-6-[(1S)-2-In d olyl-1-[(m eth oxyca r bon yl)-
eth yl]am in o]-10-oxo-9-[3-[[(ph en ylm eth oxy)car bon yl]am i-
n o]p r op yl]-2-[[4-(p h en ylm eth oxy)p h en yl]m eth yl]-4,1,8-
oxa d ia za b icyclo[5.3.1]u n d eca n e (8.3z). L-TrpOMe‚HCl
(0.864 g, 3.39 mmol) was added to saturated NaHCO3 (4 mL)
and extracted with ether (3 × 6 mL). The organic extracts
were dried over MgSO4, and the solvent was removed to yield
L-TrpOMe as a clear oil which was used immediately without
further purification.
(3S,6S,8S,12S,14S)-3-(3-Am in op r op yl)-2,5-d ioxo-6-(in -
d olylm e t h yl)-12-[4-(p h e n ylm e t h oxy)-10,1,4,7-oxa t r i-
a za tr icyclo[6.4.2.04,14]tetr a d eca n e (8.6). Pd(OH)2/C (De-
gussa wet, 20% Pd, 7.0 mg, 27.0 µmol) was added to a solution
of crude compound 8.5z (50-80% pure, 10.0 mg, 13.5 µmol)
and HOAc (20 µL, 674 µmol) in absolute EtOH (1 mL) and 5%
water/methanol (5 mL). The mixture was purged three times
with H2 and stirred under an H2 atmosphere. After 1 h 40
min, the reaction mixture was filtered through Celite with
A solution of the L-TrpOMe in acetonitrile (8 mL) was added
to iodide 8.1z (0.058 g, 0.085 mmol) with a cannula. After
being stirred for 41 h at rt, the solution was evaporated and

Misrouted Synthesis of a Peptidomimetic
J . Org. Chem., Vol. 61, No. 22, 1996 7695
Ta ble 7. Alip h a tic P r oton Assign m en ts for 8.3z
methanol and water. The methanol was removed by rotary
evaporation at rt, and the water was lyophilized. The result-
ant yellow powder was purified by HPLC using a 10 µm C18
H
chem shift (δ)
mult
coupling constants (Hz)
2
3
3′
5
5′
6
7
8
8′
10
12
12′
15
15′
16
17
21
23
23′
3.15
3.66
3.81
3.88
3.65
2.13
3.02
3.22
3.27
3.17
2.74
3.38
1.34
1.80
1.58
3.12
3.61
3.10
3.00
m
dd
dd
dd
reverse phase preparative column (25 cm × 25 mm).
A
3.5, 12.1
10.6, 12.1
4.5, 12.2
<1, 12.4
(<1, 4, 4.5)
(1.4, 4, 5.0)
1.4, 14.2
5.0, 14.5
6, 7
6.2, 13.7
9.6, 13.6
7, 7, 14
7, 7, 14
7.2
gradient of 0-50% acetonitrile/water (0.1% TFA) over 30 min
was used with a flow rate of 8 mL/min. Detection at 280 nm
revealed that the product eluted at 22.4 min. Lyophilization
afforded the amine 8.6 (3.8 mg, 55%) as a white flocculent
powder, which was only partially soluble in water, methanol,
or acetonitrile but was completely soluble in 50% methanol/
water or 50% acetonitrile/water: 1H NMR (400 MHz, 50% CD3-
CN/D2O) δ 7.53 (d, 1, J ) 7.9), 7.35 (d, 1, J ) 8.1), 7.10 (s, 1),
7.09 (t, 1, J ) 7.3), 7.03 (d, 2, J ) 8.4), 7.00 (t, 1, J ) 7.2), 6.68
(d, 2, J ) 8.5), 3.98 (m, 2), 3.83 (m, 2), 3.71 (m, 2), 3.52 (m, 4),
3.36 (dd, 1, J ) 8.7, 14.0), 3.21 (dd, 1, J ) 5.5, 14.8), 3.03 (m,
1), 2.92 (m, 2), 2.84 (dd, 1, J ) 1.8, 7.0), 2.80 (d, 1, J ) 6.9),
2.53 (m, 1), 2.00 (m, 1), 1.58 (m, 2), 1.34 (m, 1); HRMS (FAB)
m/ z calcd for MH+ C29H36N5O4 518.2767, found 518.2775, 518
(MH+), 451, 429, 388, 307.
(3S,6S,8S,12S,14S)-2,5-Dioxo-3-(3-gu a n id in op r op yl)-6-
(in d olylm eth yl)-12-[[4-(p h en ylm eth oxy)p h en yl]m eth yl]-
10,1,4,7-oxatr iazatr icyclo[6.4.2.04,14]tetr adecan e (2). Com-
pound 8.6 (2.6 mg, 5.0 µmol) and aminoiminomethanesulfonic
acid38,39 (1.2 mg, 10.0 µmol) were dissolved in methanol (3 mL)
and triethylamine (3.5 µL, 25 µmol) was added. After being
stirred for 2.5 h, the solution was concentrated and the residue
was dissolved in methanol. The crude mixture was purified
by HPLC using a 10 µm C18 reverse phase preparative column
(25 cm × 25 mm) eluted with a gradient of 0-50% acetonitrile/
water (0.1% TFA) over 30 min at a flow rate of 8 mL/min.
Detection at 280 nm revealed that the product eluted at 23.8
min. Lyophilization afforded the final product 2 (2.4 mg, 86%)
as a white powder, which was insoluble in water or acetonitrile
but soluble in 20% DMSO/water, 50% methanol/ water, and
50% acetonitrile/water: 1H NMR (400 MHz, 50% CD3CN/D2O)
δ 7.53 (d, 1, J ) 8.0), 7.38 (d, 1, J ) 8.1), 7.16 (s, 1), 7.12 (t, 1,
J ) 7.5), 7.04 (t, 1, J ) 7.4), 7.03 (d, 2, J ) 8.4), 6.68 (d, 2, J
) 8.4), 4.00 (m, 4), 3.79 (m, 3), 3.67 (dd, 1, J ) 4.8, 12.6), 3.55
(m, 1), 3.39 (m, 2), 3.22 (m, 1), 3.11 (t, 2, J ) 6.9), 3.05 (dd, 1,
J ) 7.1, 14.3), 2.96 (dd, 1, J ) 8.4, 14.7), 2.81 (dd, 1, J ) 6.8,
14.2), 2.50 (m, 1), 1.99 (m, 1), 1.44 (m, 1), 1.36 (m, 1); HRMS
(FAB) m/ z calcd for MH+ C30H38N7O4 560.2985, found 560.3001,
560 (MH+), 465, 451, 441, 429, 373, 325.
Stu ctu r e Deter m in a tion . Gen er a l. Modeling was per-
formed using Macromodel with the MM2* forcefield and no
solvent.43 Conformations of various compounds were found
using a Monte Carlo search procedure, and conformational
space was considered well-searched if each conformation was
found five or more times.
NMR spectra were recorded with a Bruker AM 500 spec-
trometer and were processed on a Bruker X-32 data station
running UXNMR. Time proportional phase incrementation
(TPPI)44 was used to obtain phase-sensitive spectra. T1
relaxation times were measured using an inversion-recovery
experiment.
Bicyclic Com p ou n d 8.3z. Spectra of compound 8.3z were
recorded at 27 °C in d6-acetone. Chemical shifts were refer-
enced to acetone at δ 2.04 ppm for protons and δ 29.8 ppm for
carbons. A 1D 1H NMR spectrum was recorded and resolution
enhanced to allow for accurate measurement of individual
coupling constants (Table 7). A 1D 13C spectrum was recorded
as well as DEPT 90 and DEPT 135 spectra (Table 8).
A TOCSY spectrum was recorded using the MLEV-17
mixing sequence45 of 100 ms duration. All π/2 pulses were
39.2 µs. An acquisition time per scan of 524 ms, a size of 4 K,
and a spectral width of 3906 Hz in F2 and F1 were used; 320
experiments of four scans were performed with a relaxation
delay of 1.4 s. Zero filling was performed in F2 and F1 to 4K
complex × 1K real. Apodization in both dimensions used a
squared sinebell shifted by π/2.5 in F2 and π/2 in F1.
dd
br t
br t
dd
dd
dd
dd
dd
ddt
ddt
quintet
m
dd
ddd
ddd
6.1, 7.1
0.7, 6.0, 14.3
0.7, 7.1, 14.3
a
Spectrum recorded in d6-acetone at 27 °C.
Ta ble 8. Alip h a tic Ca r bon Assign m en ts for 8.3z
H
H chem shift (δ)
correlated carbon chem shift (δ)
2
3
5
6
7
3.15
70.6
72.5
70.6
61.5
53.7
50.1
56.4
34.6
28.4
27.4
41.6
62.0
29.3
3.66, 3.81
3.65, 3.88
2.13
3.02
3.22, 3.27
3.17
2.74, 3.38
1.34, 1.80
1.58
3.12
3.61
8
10
12
15
16
17
21
22
3.00, 3.10
a
Spectrum recorded in d6-acetone at 27 °C.
A DQF-COSY spectrum was obtained following the method
of Derome.46 All π/2 pulses were 13.3 µs. An acquisition time
per scan of 740 ms, a size of 8 K, and a spectral width of 5556
Hz in F2 and F1 were used; 512 experiments of eight scans
were performed with a relaxation delay of 1.6 s. Zero filling
was performed in F2 and F1 to 4K complex × 1K real.
Apodization consisted of exponential multiplication with 1.2
Hz of line broadening in F2 and a squared sinebell shifted by
π/2 in F1.
A NOESY spectrum was obtained using the pulse sequence
described by Bodenhausen47 with a mixing time of 600 ms.
All π/2 pulses were 13.3 µs. An acquisition time per scan of
524 ms, a size of 4 K, and a spectral width of 3906 Hz in F2
and F1 were used; 492 experiments of eight scans were
performed with a relaxation delay of 3 s. Zero filling was
performed in F2 and F1 to 4K complex × 1K real. Apodization
in both dimensions used a squared sinebell shifted by π/2.
An HSQC (heteronuclear single quantum coherence) spec-
trum was obtained with refocusing and 13C decoupling48 using
a GARP sequence.49 Protons bound to 12C were suppressed
using a BIRD sequence followed by an inversion-recovery
delay of 190 ms using the method of Bax et al.50 Proton π/2
pulses were 22.5 µs, 13C π/2 pulses were 19 µs, and a π/2 pulse
of 59 µs was used for 13C during the GARP sequence. An
acquisition time per scan of 260 ms, a size of 2K, and a spectral
width of 3937 Hz in F2 and 6410, centered on the aliphatic
region, in F1 were used; 512 experiments of eight scans were
performed with a relaxation delay of 100 ms. Delays were
(46) Derome, A. E.; Williamson, M. P. J . Magn. Reson. 1990, 88,
177.
(43) 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.
(44) Marion, D.; Wu¨thrich, K. Biochem. Biophys. Res. Commun.
1983, 113, 967.
(47) Bodenhausen, G.; Kogler, H.; Ernst, R. R. J . Magn. Reson. 1984,
58, 370.
(48) Norwood, T. J .; Boyd, J .; Heritage, J . E.; Soffe, N.; Campbell,
I. D. J . Magn. Reson. 1990, 87, 488.
(49) McCoy, M. A.; Mueller, L. J . Magn. Reson. 1993, 101, 122.
(50) Bax, A.; Subramanian, S. J . Magn. Reson. 1986, 67, 565.
(45) Bax, A.; Davis, D. G. J . Magn. Reson. 1985, 65, 355.

7696 J . Org. Chem., Vol. 61, No. 22, 1996
Koslowski and Bartlett
Ta ble 9. Alip h a tic P r oton Assign m en ts for 2
A ROESY spectrum was obtained using the pulse sequence
described by Rance with a Z-filtered pulsed spinlock of 300
ms duration and an effective field strength 2.2 kHz.52 All π/2
pulses were 22.5 µs; 588 experiments of 32 scans were
performed with a relaxation delay of 1.3 s.
H
chem shift (δ)
multiplicity
coupling constants (Hz)
2
3
3′
5
5′
6
7
8
10
12
12′
15
15′
16
17
21
23
23′
3.72
3.87
3.93
3.70
3.66
3.04
3.99
4.08
4.11
2.89
3.45
2.58
2.12
1.48
3.21
3.88
2.90
3.27
m
dd
dd
m
3.0, 13.0
10.1, 13.1
En zym ology. Porcine pancreatic R-amylase (type 1-A) was
obtained from Sigma; p-nitrophenyl maltotrioside (p-NPG3)
was obtained as a generous gift from Boehringer-Mannheim.
All solutions were prepared using doubly distilled deionized
water and were passed through a 0.45-µm filter. All assays
were performed at 30 °C. The forward reaction rate was
measured by monitoring the liberation of p-nitrophenol from
the p-NPG3 substrate at 405 nm (∆ꢀ ) 9500 M-1 cm-1).6
Enzyme dilutions were made with buffer containing 50 mM
HEPES (pH 7.0), 60 mM NaCl, 2 mM CaCl2, and 0.1 mg/mL
BSA. The assay mixture contained 40 nM enzyme in 25 mM
HEPES (pH 7.0), 30 mM NaCl, and 1 mM CaCl2 in 5% DMSO/
H2O in a total volume of 1 mL; the solution was equilibrated
for 3 min at 30 °C prior to addition of substrate.
m
ddd
ddd
m
5.0, 9.5, 11.0
2.2, 2.2, 9.4
m
dd
dd
m
m
m
6.3, 14.7
9.8, 14.6
dt
2.2, 7.1
5.6, 6.9
5.8, 15.0
6.6, 15.0
dd
dd
dd
Five substrate concentrations were used (0.40, 0.50, 0.70,
1.0, 3.0 mM) with two independent determinations done at
each concentration to determine Km. Ki measurements were
performed with substrate concentration equal to Km. Inhibitor
concentrations from DMSO stock solutions were measured
spectrophotometrically using the absorbance values for one
a
Spectrum recorded in 20% d6-DMSO/D2O at 28 °C.
optimized for a coupling constant of 131 Hz. Zero filling was
performed in F2 and F1 to 2K complex × 1K real. Apodization
consisted of exponential multiplication with 4 Hz of line
broadening in F2 and a squared sinebell shifted by π/2 in F1.
An HMBC (heteronuclear multiple-bond correlation) spec-
trum was obtained using the method of Bax.51 The experiment
was recorded in mixed mode using TPPI44 in F1. Proton π/2
pulses were 17.1 µs, and 13C π/2 pulses were 7.2 µs. Delays
were optimized for a long-range coupling constant of 8.3,
allowing one-bond C-H couplings to be filtered out. An
acquisition time per scan of 520 ms, a size of 4 K, and a
spectral width of 3937 Hz in F2 and 20,833 Hz in F1 were
used; 800 experiments of 16 scans were performed with a
relaxation delay of 1.4 s. Data were processed using a
magnitude calculation in the F2 dimension only. Zero filling
was performed in F2 and F1 to 2K complex × 1K real.
Apodization consisted of exponential multiplication with 2 Hz
of line broadening in F2 and a squared sinebell shifted by π/2
in F1.
Tr icyclic Com p ou n d 2. Spectra of compound 2 were
recorded at 28 °C in 20% d6-DMSO/D2O. Chemical shifts were
referenced to HDO at δ 4.74 ppm. All 2D spectra were
obtained with an acquisition time per scan of 490 ms, a size
of 4K, and a spectral width of 4167 Hz in F2 and F1. Zero
filling was performed in F2 and F1 to 4K complex × 1K real.
Apodization in both dimensions used a square sinebell shifted
by π/3 in F2 and π/2 in F1.
tyrosine and one tryptophan (λ ) 280 nm, ꢀ ) 6000 M-1 cm-1) 53
.
After substrate addition, 20 absorption points were recorded
in a 10-min period; the reaction was less than 2% complete at
this point and good first-order kinetics were observed. Initial
rates were calculated using the kinetics module of the Uvikon
860. No background correction was necessary, since a back-
ground rate was not observed when the assay was run either
without substrate or enzyme. Km was computed using the
Enzfitter program.54 Ki values were calculated by a Dixon
analysis, assuming competitive inhibition, performed in
Enzfitter using the equation Vo/Vi ) [I]/Ki(1 + [S]/Km) + 1.
Ack n ow led gm en t. This work was supported by a
grant (GM-30759) from the National Institutes of Health
and by fellowships to M.C.K. from the Department of
Education and from Syntex Corporation. We also
express our appreciation to Drs. Stephen Telfer, Karl-
Heinz Neff, and J effrey Charonnat for their contribu-
tions to the early stages of this project, to J . Hoyt Meyer
for preparation of key materials in the later stage of
the project, and to Dr. Graham Ball for assistance with
the NMR experiments.
A 1D 1H NMR of 1K scans was recorded and resolution
enhanced to allow for accurate measurement of individual
coupling constants (Table 9). A 1D decoupling experiment was
also performed to elucidate the magnitude of a critical coupling
constant. Irradiation at 3.06 ppm collapsed the doublet of
triplets at 4.00 ppm to a triplet, positively indicating that the
coupling constant between these two protons was 9.2 Hz.
A TOCSY spectrum was recorded using the MLEV-17
mixing sequence45 of 100 ms duration. All π/2 pulses were
34.8 µs; 512 experiments of 24 scans were performed with a
relaxation delay of 450 ms.
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectra for
compounds not submitted to elemental analysis (17 pages).
This material is contained in libraries on microfiche, im-
mediately follows this article in the microfilm version of the
journal, and can be ordered from the ACS; see any current
masthead page for ordering information.
J O961114P
(52) Rance, M. J . Magn. Reson. 1987, 74, 557.
A DQF-COSY spectrum was obtained following the method
of Derome.46 All π/2 pulses were 13.3 µs; 512 experiments of
48 scans were performed with a relaxation delay of 1.0 s.
(53) Wetlaufer, D. B. In Advances in Protein Chemistry, Vol. 17;
Anfinsen, C. B., J r., Anson, M. L., Bailey, K., Edsall, J . T., Eds.;
Academic Press: New York, 1962; p 303.
(54) Leatherbarrow, R. J . ENZFITTER; Elsevier Science: 1987.
(55) Haasnoot, C. A. G.; De Leeuw, F. A. A. M.; Altona, C.
Tetrahedron 1980, 36, 2783.
(51) Bax, A.; Marion, D. J . Magn. Reson. 1988, 78, 186.