Complestatin synthetic studies: The effect of the amino acid configuration on peptide backbone conformation in the common western BCD macrocycle

Article abstract of DOI:10.1016/j.bmcl.2004.01.056

The synthesis and structural analysis, involving X-ray crystallographic, nuclear magnetic resonance, and computational studies of four diastereomers of the common western BCD diarylether macrocycle of the complestatins, a family of HIV entry inhibitors, h

Full text of DOI:10.1016/j.bmcl.2004.01.056

Bioorganic & Medicinal Chemistry Letters 14 (2004) 1697–1702
Complestatin synthetic studies:
the effect of the amino acid configuration on peptide backbone
conformation in the common western BCD macrocycle
Amos B. Smith, III,* Jason J. Chruma, Qiang Han and Joseph Barbosa
Department of Chemistry, University of Pennsylvania, Philadelphia, PA 19104-6323, USA
Received 11 November 2003; accepted 20 January 2004
Abstract—The synthesis and structural analysis, involving X-ray crystallographic, nuclear magnetic resonance, and computational
studies of four diastereomers of the common western BCD diarylether macrocycle of the complestatins, a family of HIV entry
inhibitors, has been achieved exploiting a ruthenium-promoted intramolecular SNAr reaction. The stereogenicity of the individual
phenylglycines (residues C and D) results in remarkable effects on the backbone conformation.
# 2004 Elsevier Ltd. All rights reserved.
The initial event in cell entry of the human immunode-
ficiency virus (HIV) entails the binding of the 120 kDa
viral-envelope glycoprotein (gp120) to the human lym-
phocyte cell-surface cluster determinant 4 (CD4).1 Sub-
sequent conformational change in the viral glycoprotein
generates a secondary binding epitope specific for the
lymphocyte cell-surface chemokine receptors CCR5
and/or CXCR4. Once bound to both cell-surface pro-
teins, the virus is poised for membrane fusion. Recog-
nition of this cascade of events has led to a world-wide
campaign to identify compounds that abrogate CD4–
gp120 binding interactions, and thereby viral entry and
HIV infection.2 The complestatins, a small family of
bismacrocycle heptapeptides isolated from various soil
Streptomyces species, typified by chloropeptin I [(ꢀ)-1],3
represent a particularly intriguing class of CD4–gp120
binding inhibitors. Chloropeptin I [(ꢀ)-1], the most
plestatin family. Early in this venture, in collaboration
-
with Omura and Tanaka,5 we determined the absolute
stereochemistry and solution conformation of (ꢀ)-
chloropeptin I [(ꢀ)-1, Figure 1], exploiting a combin-
ation of computational and spectroscopic methods.
Subsequently, the synthetic community has shown con-
siderable interest in the complestatins,6 culminating
recently in the total synthesis of chloropeptin I.6h
In this paper, we report the synthesis of four diaster-
eomers of the complestatin western BCD macrocycle,
including the isostructural congener. The stereogenicity
of the two phenylglycines within the macrocyclic diaryl-
ether tripeptide were found to effect dramatically the
overall conformation of the peptide backbone, a result
expected to play a significant role in the future design of
complestatin based CD4–gp120 inhibitors.
soluble member of the complestatin family, was shown
-
by Omura et al. to inhibit the binding of CD4 to gp120
in the presence of bovine serum albumin (BSA) with an
IC50 value of 2.0 mM; in the absence of BSA the potency
is 32nM.
Given our interest in the development of novel gp120–
CD4 agonists and antagonists,4 we initiated a program
directed toward the synthesis of members of the com-
* Corresponding author. Tel.: +1-215-898-4860; fax: +1-215-898-
5129; e-mail: smithab@sas.upenn.edu
Figure 1. Chloropeptin I (ꢀ)-1.
0960-894X/$ - see front matter # 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.01.056

1698
A. B. Smith, III et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1697–1702
In support of this hypothesis, we constructed a compu-
tational docking model in which (ꢀ)-1 was bound to the
gp120 binding domain of CD4. The model indicates
that (ꢀ)-1 can form a b-sheet with the C0C00 region of
CD4, similar to the b15 strand of gp120 (Fig. 3a).
Comparable b-sheet interactions have been reported for
the related conformationally-constrained peptide anti-
bacterial agent vancomycin with the bacterial cell-wall
precursor l-Lys-d-Ala-d-Ala.8 Interestingly, in contrast
with chloropeptin I, 100 mM vancomycin does not inhi-
bit gp120/CD4 binding; the complestatins are equally
inactive against bacteria. In addition to possible b-sheet
interactions, multiple ionic hydrogen-bonding opportu-
nities between the deprotonated phenolic oxygens of
chloropeptin I with protonated side-chains of CD4
would increase the binding potential (Fig. 3b). Rotation
of the Phe-43 side-chain, which is most likely freely
rotating in solvent, into the diaryl ether hydrophobic
pocket of (ꢀ)-1, would result in additional binding
affinity.
Figure 2. C0C00 loop of CD4 (mostly pink) in relationship to the b-15
strand of gp120 (cyan). Critical residues Ser-42 and Phe-43, as well as
the backbone of Leu-44, Thr-45, and Lys-46 are presented in standard
atom colors, highlighting strand-to-strand interactions with b-15 of
gp120.
The crystal structure of a ternary complex composed of
laboratory-adapted gp120 core (HxBC2), the D1D2
portion of CD4, and 17b, a mouse antibody known to
bind to the chemokine receptor-binding epitope on
gp120 was reported in 1998 by Hendrickson et al.7 The
nine contiguous amino acid span of CD4, comprising
residues 40–48 (C0C00 strands), was suggested to be
responsible for 63% of all interfacial contacts, of which
23% were covered by the previously solvent-exposed
side-chain of Phe-43, which caps a hydrophobic pocket
in gp120. Half of the interacting CD4 residues, however,
make interfacial contributions only through their back-
bone atoms. The C0 strand of CD4 is the dominant
contributor of the main chain interactions participating
in the b-pleated sheet alignment with the b15 strand of
gp120 (Fig. 2). These observations strongly suggest that
interfering with the ability of gp120 to form a b-sheet
with CD4 would greatly reduce binding affinity.
Having determined the stereostructure of chloropeptin I
and developed a plausible hypothesis for the inhibition
of gp120 binding to CD4, we initiated a stereodivergent
synthesis of the complestatins and related diaster-
eomers. Of particular interest was how the configur-
ation of the phenylglycine stereocenters affects the
peptide backbone conformation in the rigid diphenyl-
ether and biaryl macrocyclic constructs. Our initial
efforts focused on the construction of the common wes-
tern BCD macrocycle. Given the commonality of this
structural unit throughout the complestatin family of
peptides and their potential biological importance, an
efficient route to the BCD macrocycle and congeners
thereof was viewed as central both to our unified
synthetic strategy for the complestatins and for the
development of novel agonists/antagonists of HIV-1
gp120-CD4 binding.
Chloropeptin I [(ꢀ)-1] exhibits a high degree of struc-
tural and electrostatic homology to the CD4-binding
domain of gp120 (e.g., a hydrophobic pocket and a rigid
peptide b-strand crowned by a hemisphere of acidic
sites). This structural homology suggests that (ꢀ)-1 and
gp120 may compete for a common binding site on CD4.
To construct the common complestatin western macro-
cycle, we selected the elegant arene-ruthenium chemistry
pioneered by Pearson9 and Rich.10 The requisite indivi-
dual amino acid residues (ꢀ)-3, (ꢀ)-5, and (ꢀ)-7 were
prepared as outlined in Scheme 1. Residue B [(ꢀ)-3] was
Figure 3. Computational Model of (ꢀ)-1 Docked to CD4. (a) Potential b-sheet interaction between (ꢀ)-1 (pink) and the C0C00 subregion of CD4
(cyan) [amide N=blue, a-C=black, carbonyl O=red], (b) Hydrogen bonding opportunities between (ꢀ)-1 (pink) and CD4 (cyan): the positively
charged amino acid side chains and Phe-43 of CD4, key to CD4-gp120 binding (yellow) are highlighted. [Figures created using MOLSCRIPT
(Kraulis, P. J. J. Appl. Crystallogr. 1991, 24, 946.)].

A. B. Smith, III et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1697–1702
1699
obtained in 86% yield11 upon N,O-bis-methylation of
Boc-protected (4-chlorophenyl)alanine (+)-2,12 fol-
lowed by ruthenium-arene complexation. For residue C
[(ꢀ)-5], bis-chlorination of N-Boc-d-(4-hydroxyphenyl)-
glycine [(ꢀ)-4]13 with Cl2 in 3:5H2O/CHCl3, followed in
14,15
turn by O,O-bismethylation with TMSCHN2
and
catalytic trans-esterification to the benzyl ester exploit-
ing the Otera protocol16 proceeded in good overall yield
and with high enantioselectivity (65% and ꢁ97% e.e.; 3
steps).17,18 The final amino acid, residue D [(ꢀ)-7], was
prepared in 7 steps and 31% overall yield (ꢁ90% e.e.)
from methyl (3,5-dihydroxy-4-methoxyphenyl)acetate
(6) following literature precedent.19
With the individual amino acids in hand, we turned to
the construction of the linear BCD tripeptide. The
initial route, involving a C-to-N protocol (residues B-to-
D), was quickly abandoned due both to our inability to
purify adequately the arene-ruthenium salt inter-
mediates, and to the near quantitative hydrolysis of the
newly formed N-methyl amide bond upon Boc-depro-
tection. We instead employed an N-to-C coupling
strategy (residues D-to-B) to form the linear tripeptide
(Scheme 2). Accordingly, N-Boc-phenylglycinate (ꢀ)-5
was deprotected with TMSOTf in dichloromethane, and
the resulting free amine was coupled to phenylglycine
(ꢀ)-7 employing HATU to afford dipeptide (ꢀ)-8 in
71% overall yield, with variable diastereomeric purity
(5–10:1 d.r.).20 Palladium-catalyzed hydrogenolysis of
benzyl ester (ꢀ)-8 next proceeded in near quantitative
yield to provide the CD dipeptide acid (ꢀ)-9. In CDCl3,
the N-Boc-protected acid (ꢀ)-9 exists as a concentration
Scheme 2. synthesis of CD-dipeptide (ꢀ)-9.
protected diphenylglycine (ꢀ)-9. Polar solvents would
be expected to disrupt the hydrogen-bonding network
that stabilizes the syn-rotamer; accordingly, in CD3CN
only one conformation was observed in the H NMR
spectra of (ꢀ)-9.
1
We next explored a variety of coupling protocols to
form the N-methyl amide linkage. Not unexpectedly,
near or complete epimerization of the C residue a-center
resulted (Scheme 3).22 From the perspective of yield, the
best results were obtained by removing the Boc-moiety
from the phenylalanine (ꢀ)-3 with TFA and then cou-
pling to the CD dipeptide (ꢀ)-9 using BOP-Cl in di-
chloromethane;23 acyclic tripeptide 10 was obtained in
83% yield as a mixture of diastereomers. Treatment of a
dilute solution (1–5 mM) of 10 in DMF with Cs2CO3,
followed by photolytic demetalation of the macrocycle
in acetonitrile led to a mixture of tripeptide macro-
cycle diastereomers and rotamers in 60% combined
yield (Scheme 4). Separation of the individual diaster-
eomers was accomplished by normal-phase HPLC.24
Two of the four diastereomers (vide infra) existed as
separable mixtures of N-methyl amide rotational iso-
mers; the remaining two showed no evidence for rota-
tional isomerism. In accord with literature precedents,
the rotameric ratios and rates of equilibration for the
two rotational isomers of both diastereomers were sol-
vent dependent.22,6a,h As noted by us and others,5,6a,h,25
the C-terminal residue of chloropeptin I (residue A)
locks the BCD-macrocycle of the natural product in the
cis N-methyl amide conformation. Presumably, this
increased conformational rigidity protects the residue C
chiral center of (ꢀ)-1 from the base induced-epimerization
suffered by (S,R,R)-BCD.
1
dependent mixture of rotamers, as determined by H
NMR. In this regard, Marcovici-Mizrahi et al. reported
that the sterically unfavored syn-carbamate rotamer of
N-Boc-protected mono-phenylglycines is stabilized in
apolar solvents by a hydrogen-bond dimer between the
free carboxylic acid and the carbamate carbonyl and
amine (see insert, Scheme 2).21 Presumably, this stabili-
zation also occurs, though to a lesser extent, in N-Boc
Interestingly, the ratio of diastereomers was sig-
nificantly different from that of the starting acyclic tri-
peptide 10, with the isostructural natural (S,R,R)-
diastereomer being the least favored product. Attempts
to epimerize the (S,S,S), (S,R,S), and (S,S,R)-BCD
Scheme 1. Synthesis of amino acids (ꢀ)-3, (ꢀ)-5, and (ꢀ)-7.

1700
A. B. Smith, III et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1697–1702
Scheme 3. Synthesis of acyclic BCD-tripeptide 10.
Scheme 4. Formation of BCD macrocycle diastereomers.
macrocycles (DBU, CD3CN, 50 ꢂC, 18 h; monitored by
1H NMR) proved unsuccesful.26 However, when the
(S,R,R)-diastereomer was submitted to similar condi-
tions, 75% of the material was converted into the
(S,S,R) congener within 3 h.
by NOE-constrained computational analyses, similar to
those employed to elucidate the stereostructure of (ꢀ)-
chloropeptin I (1).5 In short, distance constraints were
generated from experimental NOE values for both the
cis and trans rotamers of (S,R,S)-BCD and (S,R,R)-
BCD; these constraints were then employed in a Monte
Carlo conformation search (AMBER* force field,
chloroform solvation) and the resulting 75 lowest-
energy conformations were selected, superimposed and
shown to represent, except for (S,R,S)-trans-BCD (vide
infra), one major conformational group as shown in
Figure 4.28 Although the cis:trans ratio of the N-methyl
amide moiety for both the (S,R,S) and (S,R,R) diaster-
eomers was solvent dependent, the cis rotamer was
consistently the dominant conformational isomer. The
cis rotamer possessing the R configuration at residue C
was easily distinguished by a strong NOE crosspeak
between the a-CHs of residues B and C.29
Determination of the complete stereostructures of the
individual BCD macrocycle diastereomers was particu-
larly challenging. Gratifyingly, X-ray quality crystals of
the two diastereomers that did not show evidence of N-
methyl amide rotational isomerism were obtained. The
X-ray analysis confirmed these to be the (S,S,R) and
(S,S,S)-BCD tripeptide macrocycles. The unit cell of the
(S,S,R) diastereomer displayed two distinct conformers,
while three different conformers were observed for the
(S,S,S) diastereomer, each representing, for the most
part, different torsional orientations about the methyl ester
carbonyl-phenylalanine a-carbon bond or the dichloro-
hydroxy-phenylglycine sidechain-a-carbon linkage27
Structural analysis of the individual BCD stereoisomers
revealed that the stereogenicities of the amino acids
produce a diverse array of peptide backbone conforma-
tions. Shown in Figure 4 is a comparison of the con-
formations of the four BCD diastereomers, including
where appropriate the cis and trans rotamers, with the
previously determined stereostructure of chloropeptin I
Despite considerable efforts, X-ray quality crystals of
the two remaining diastereomers, which both displayed
N-methyl amide rotational isomerism, could not be
obtained. Their absolute configurations and resulting
conformations were deduced from extensive NMR ana-
lysis (TOCSY, HMBC, HMQC, and NOESY) followed

A. B. Smith, III et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1697–1702
1701
Figure 4. Conformational comparison of chloropeptin I [(ꢀ)-1] with the four synthesized BCD diastereomers. The structures for the (S,S,S) and
(S,S,R) diastereomers were determined by X-ray crystallography.27 The four other conformations were determined by NOE-constrained computa-
tional methods, as described earlier, and represent overlays of the 75 lowest-energy conformations, except for (S,R,S)-trans-BCD (30 lowest-energy
structures).28
[(ꢀ)-1].5 Unlike the western macrocycle of (ꢀ)-1 which
has the (S,R,R) configuration, the N-methyl amides of
(S,S,R) and (S,S,S)-BCD are clearly in the trans con-
formation about the N-methyl amide linkage. Other
structural distinctions between the (S,S,R)/(S,S,S) dia-
stereomers and (ꢀ)-1 include the opposite orientation of
the NH and carbonyl moieties in the macrocyclic
peptide backbone. These remarkable conformational
differences would be expected to elicit pronounced
changes in biological potency, based on our proposed
model for inhibition of CD4-gp120 binding by (ꢀ)-
chloropeptin I.
rotamer, the intracyclic region of the peptide backbone
of the cis-rotamer of (S,R,S)-BCD displays a direction-
ality similar to the (S,R,R) diastereomer as well as to
(ꢀ)-chloropeptin I (1).
In summary, a stereodivergent synthesis of four of the
possible diastereomers of the complestatin western
(BCD) macrocycle, including the isostructural natural
configuration, has been achieved, employing a ruthe-
nium-activated intramolecular SNAr cyclization tactic.
Pleasingly, the route provides direct access to useful
quantities of advanced intermediates to explore the role
of the stereogenicity of the C and D residues on peptide
backbone conformation. The observed complex rela-
tionships between phenylglycine configuration and pep-
tide backbone conformation/directionality in the rigid
macrocyclic constructs provides potentially important
information for the design of small molecule inhibitors
of CD4–gp120 binding. The results of these studies will
be reported in due course.
The cis amide rotamer of (S,R,R)-BCD, not surpris-
ingly, is nearly isostructural with the conformation of
the BCD macrocycle of chloropeptin I. Similarly, the
trans rotamer of (S,R,R)-BCD possesses the same pep-
tide backbone directionality between residues C and D
to (ꢀ)-1. The trans rotamer of (S,R,S)-BCD was found
to be conformationally the most different member of the
four diastereomers. Inspection of the 75 lowest- energy
conformations of trans-(S,R,S)-BCD revealed three
unique structural groups, of which the lowest energy
grouping (30 conformations) is shown in Figure 4. The
other two (not shown) were similar to the (S,R,S)-cis-
BCD conformation, save the conformation about the N-
methyl amide linkage. The directionality of the peptide
backbone between residues C and D in the lowest
energy conformation is opposite to that of chloropeptin
I. A more striking feature in the S,R,S-trans-congener is
placement of the residue C aromatic ring, with one edge
directly underneath the biaryl linkage. This unique
orientation seems reasonable since the residue C aryl
Acknowledgements
Financial support was provided by the National Insti-
tutes of Health through grant GM-57139 and GM-
56550. We also thank Drs. R. K. Kohli, G. Furst, and
P. J. Carroll, Directors of the University of Pennsylva-
nia Spectroscopic Facilities, for valuable assistance in
obtaining mass spectra, NMR data, and X-ray analyses,
respectively. J.C. also acknowledges the National Science
Foundation, the American Chemical Society, and Wyeth-
Ayerst for graduate fellowships. Finally, we thank Prof.
R. C. Breslow (Columbia University) for the use of his
computer in computational analysis studies.
1
hydrogens are split into two different peaks in the H
NMR spectra of (S,R,S)-BCD. Finally, unlike the trans

1702
A. B. Smith, III et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1697–1702
variable enantiopurity (0–95% e.e.), as determined by
chiral HPLC analysis.
References and notes
1. (a) McDougal, J. S.; Kennedy, M. S.; Sligh, J. M.; Cort,
S. P.; Mawle, A.; Nicholson, J. K. A. Science 1986, 231,
16. (a) Otera, J.; Yano, T.; Kawabata, A.; Nozaki, H. Tetra-
hedron Lett. 1986, 27, 2383. (b) Otera, J.; Dan-oh, N.;
Nozaki, H. J. Org. Chem. 1991, 56, 5307.
17. Despite previous reports (ref 6c,f), ester hydrolysis of
methyl N-Boc-d-(3,5-dichloro-4-methoxyphenyl)glycinate
under various conditions, including K2CO3 in MeOH,
resulted in immediate and complete racemization, as
determined by chiral HPLC analysis.
382. (b) Bour, S.; Geleziunas, R.; Wainberg, M. A.
Microbiol. Rev. 1995, 59, 63.
2. (a) Omura, S.; Tanaka, H.; Matsuzaki, K.; Ikeda, H.;
-
Masuma, R. J. Antibiotics 1993, 46, 1908. (b) Patil, A. D.;
Kumar, N. V.; Kokke, W. C.; Bean, M. F.; Freyer, A. J.;
De Brosse, C.; Mai, S.; Truneh, A.; Faulkner, D. J.;
Carte, B.; Breen, A. L.; Hertzberg, R. P.; Johnson, R. K.;
Westley, J. W.; Potts, B. C. M. J. Org. Chem. 1995, 60,
1182. (c) Nara, P. L.; Hwang, K. M.; Rausch, D. M.;
Lifson, J. D.; Eiden, L. E. Proc. Natl. Acad. Sci. U.S.A.
1989, 86, 7139. (d) Martin, L.; Strichner, F.; Misse, D.;
Sironi, F.; Pugniere, M.; Barthe, P.; Prado-Gotor, R.;
Freulon, X.; Roumestand, C.; Menez, A.; Lusso, P.; Veas,
F.; Vita, C. Nature-Biotech. 2003, 21, 71. (e) Ferrer, M.;
Harrison, S. C. J. Virol. 1999, 73, 5795.
18. Enantiomeric excess determined by chiral HPLC analysis:
Chiralpak AD1 (Chiral Technologies) 4.6ꢃ250 mm col-
umn, 3:97 i-PrOH/Hexanes, 1 mL/min, l=220 nm,
0
k1 =2.19, a=1.24, Rs=1.04.
19. (a) Pearson, A. J.; Chelliah, M. V.; Bignan, G. C. Synth-
esis 1997, 536. (b) Pearson, A. J.; Zhang, P. J. Org. Chem.
1996, 61, 9603.
20. Carpino, L. A. J. Am. Chem. Soc. 1993, 115, 4397.
21. Marcovici-Mizrahi, D.; Gottlieb, H. E.; Marks, V.;
Nudelman, A. J. Org. Chem. 1996, 61, 8402.
22. Attempts to couple the free amine of (ꢀ)-3 to N-Boc-(3,5-
dichloro-4-hydroxyphenyl)glycine with DEPBT (ref 6h)
led to competitive formation of phenolic esters.
3. Tanaka, H.; Matsuzaki, K.; Nakashima, H.; Ogino, T.;
Matsumoto, A.; Ikeda, H.; Woodruff, H. B.; Tanaka, H.;
-
Omura, S. J. Antibiot. 1997, 50, 58 and references cited
therein..
4. Smith, A. B., III; Savinov, S. N.; Manjapparra, U. V.;
Chaiken, I. M. Org. Lett. 2002, 4, 4041.
5. Gouda, H.; Matsuzaki, K.; Tanaka, H.; Hirono, S.;
23. Diago-Messeguer, J.; Palomo-Coll, A. L.; Fernandez-
Lizarbe, J. R.; Zugaza-Bilbao, A. Synthesis 1980, 547.
24. HPLC conditions: Waters Novapak Si 19ꢃ300 mm col-
umn, 45:50 EtOAc/Hexanes, 5 mL/min then 40:2:58
EtOAc/i-PrOH/Hexanes, 10 mL/min.
25. (a) Singh, S. B.; Jayasuriya, H.; Hazuda, D. L.; Felock,
P.; Homnick, C. F.; Sardana, M.; Patane, M. A. Tetra-
hedron Lett. 1998, 39, 8769. (b) Singh, S. B.; Jayasuriya,
H.; Salitruo, G. M.; Zink, D. L.; Shafiee, A.; Heimbuch,
B.; Silverman, K. C.; Lingham, R. B.; Genilloud, O.;
Teran, A.; Vilella, D.; Felock, P.; Hazuda, D. J. Nat.
Prod. 2001, 64, 874.
26. In all cases, starting material was recovered unaltered.
27. Crystallographic data (excluding structure factors) for the
structures in this paper have been deposited with the
Cambridge Crystallographic Data Centre as supplemen-
tary publication numbers CCDC 223088 and 223089.
Copies of data can be obtained, free of charge, on appli-
cation to CCDC, 12Union Road, Cambridge CB21EZ,
UK [fax: +44 (0)1223-336033 or email: deposit@
ccdc.cam.as.uk]
-
Omura, S.; McCauley, J. A.; Sprengler, P. A.; Furst, G. T.;
Smith, A. B. J. Am. Chem. Soc. 1996, 118, 13087.
6. (a) Kai, T.; Kajimoto, N.; Konda, Y.; Harigaya, Y.;
Takayanagi, H. Tetrahedron Lett. 1999, 40, 6289. (b)
Elder, A. M.; Rich, D. H. Org. Lett. 1999, 1, 1443. (c)
Roussi, G.; Zamora, E. G.; Carbonnelle, A. C.; Beugel-
mans, R. Heterocycles 1999, 51, 2041. (d) Carbonelle,
A. C.; Zamora, E. G.; Beugelmans, R.; Roussi, G. Tetra-
hedron Lett. 1998, 39, 4467. (e) Carbonelle, A. C.;
Zamora, E. G.; Beugelmans, R.; Roussi, G. Tetrahedron
Lett. 1998, 39, 4471. (f) Roussi, G.; Zamora, E. G.; Car-
bonelle, A. C.; Beugelmans, R. Tetrahedron Lett. 1997,
38, 4401. (g) Gurjar, M. K.; Tripathy, N. K. Tetrahedron
Lett. 1997, 38, 2163. (h) Deng, H.; Jung, J.-K.; Liu, T.;
Kuntz, K. W.; Snapper, M. L.; Hoveyda, A. H. J. Am.
Chem. Soc. 2003, 123, 9032.
7. Kwong, P. D.; Wyatt, R.; Robinson, J.; Sweet, R. W.;
Sodroski, J.; Hendrickson, W. A. Nature 1998, 393, 648.
8. (a) Sheldrick, G. M.; Jones, P. G.; Kennard, O.; Williams,
D. H.; Smith, G. A. Nature 1978, 271, 223. (b) Knox,
J. R.; Pratt, R. F. Antimicrob. Agents Chemother. 1990,
34, 1342. See also: Williams, D. H. Acc. Chem. Res. 1984,
17, 364.
9. (a) Pearson, A. J.; Park, J. G. J. Org. Chem. 1992, 57,
1744. (b) Pearson, A. J.; Zigmantas, S. Tetrahedron Lett.
2001, 42, 8765 and references cited therein..
10. Rich, D. H.; Janetka, J. W. J. Am. Chem. Soc. 1995, 117,
10585.
11. Reported yields are averages of at least three runs and
refer to chromatographically and spectroscopically pure
compounds, except as otherwise indicated.
12. RSP Amino Acid Analogues, Inc., Worcester, MA.
13. Miller, M. J.; Mattingly, P. G. Tetrahedron 1983, 39,
2563.
28. For all Monte Carlo conformational searches, the pro-
gram Maestro1 Molecular Modeling Interface version
4.1.012(Schrodinger, Inc.) was utilized and the following
parameters were employed: AMBER* force field with a
CHCl3 solvation model, PCRG method, maximum num-
ber of iterations=99,999, converge on gradient with a
convergence threshold of 0.05, 10,000 steps, 50.0 energy
window. NOE values were translated into distance con-
˚
straints as follows: strong (s)=1.5–2.5 A, medium
˚
˚
(m)=1.5–3.5 A, weak (w)=1.5–5.0 A, all with a force
constant of 1,000.
29. For a similar N-methyl diarylether tripeptide macrocycle
Kai proposed that the trans rotamer was dominant based
on an NOE signal between the N-Me group and the resi-
due C a-CH (ref 6a and X-ray: Kai, T.; Kajimoto, N.;
Yamada, Y.; Harigaya, Y.; Takayanagi, H. Anal. Sci.
2002, 18, 369) Our computational studies (this paper and
ref. 5) however, suggests that the strength of the NOE
between a-CHs of residues B and C is a more accurate
method for distinguishing between the cis and trans amide
rotamers of complestatin-like diarylether macrocycles.
14. (a) Hashimoto, N.; Aoyama, T.; Shioiri, T. Chem. Pharm.
Bull. 1981, 29, 1475. (b) Aoyama, T.; Terasawa, S.; Sudo,
K.; Shioiri, T. Chem. Pharm. Bull. 1984, 32, 3759.
15. O,O-Bismethylation with MeI and K2CO3 in DMF affor-
ded the desired methyl ester in 83% yield, but with highly