Structural mimicry of two cytochrome b562 interhelical loops using macrocycles constrained by oxazoles and thiazoles

Article abstract of DOI:10.1021/ja0455300

A major chemical challenge is the structural mimicry of discontinuous protein surfaces brought into close proximity through polypeptide folding. We report the design, synthesis, and solution structure of a highly functionalized saddle-shaped macrocyclic s

Full text of DOI:10.1021/ja0455300

Published on Web 04/13/2005
Structural Mimicry of Two Cytochrome b562 Interhelical Loops
Using Macrocycles Constrained by Oxazoles and Thiazoles
Yogendra Singh, Martin J. Stoermer, Andrew J. Lucke, Tom Guthrie, and
David P. Fairlie*
Contribution from the Centre for Drug Design and DeVelopment, Institute for Molecular
Bioscience, UniVersity of Queensland, Brisbane Qld 4072, Australia
Received July 26, 2004; E-mail: d.fairlie@imb.uq.edu.au
Abstract: A major chemical challenge is the structural mimicry of discontinuous protein surfaces brought
into close proximity through polypeptide folding. We report the design, synthesis, and solution structure of
a highly functionalized saddle-shaped macrocyclic scaffold, constrained by oxazoles and thiazoles,
supporting two short peptide loops projecting orthogonally from the same face of the scaffold. This structural
mimetic of two interhelical loops of cytochrome b562 illustrates a promising approach to structurally mimicking
discontinuous loops of proteins.
helical tripeptide loops of the well-studied11 cytochrome b562
Introduction
(Figure 1), its loops being fairly rigidly positioned at the ends
Current scientific, industrial, and medical uses of proteins
and peptides are significantly limited by conformational flex-
ibility, chemical and biological instability, low bioavailability,
and high manufacturing costs.1 Protein function is often medi-
ated through small regions of folded polypeptide surfaces,2 but
short isolated peptides (e15 amino acids) corresponding to
bioactive surfaces of proteins typically lack well-defined
structure in water.3 If such peptides could be restrained to shapes
that structurally mimic protein surfaces, they could conceivably
lead to cheaper, conformationally restrained, chemically stable,
new nanomaterials with potential uses as supramolecular
building blocks, artificial proteins, biological probes, drug leads,
catalysts, and sensors. Toward this goal, small structural
mimetics have been devised for peptide strands,4 sheets,4,5
helices,6 turns,7 and loops,8 but there are very few examples of
mimetics for discontinuous surfaces9 of proteins, such as
multiple loops (Figure 1)10 formed by residues well separated
in sequence but brought close together in three-dimensional
space by polypeptide folding. To test a promising approach to
multi-loop mimicry, we chose to structurally mimic two inter-
of the helices and well defined in the crystal structure.12
(6) Helices: (a) Andrews, M. J. I.; Tabor, A. B. Tetrahedron 1999, 55, 11711-
11743. (b) Taylor, J. W. Biopolymers 2002, 66, 49-75. (c) Kelso, M. J.;
Hoang, H.; Appleton, T. G.; Fairlie, D. P. J. Am. Chem. Soc. 2000, 122,
10488-10489. (d) Kelso, M. J.; Hoang, H. N.; Oliver, W. N.; Sokolenko,
N.; March, D. R.; Appleton, T. G.; Fairlie, D. P. Angew. Chem, Int. Ed.
2003, 42, 421-424. (e) Kelso, M. J.; Beyer, R. L.; Hoang, H. N.;
Lakdawala, A. S.; Snyder, J. P.; Oliver, W. P.; Robertson, T. A.; Appleton,
T. G.; Fairlie, D. P. J. Am. Chem. Soc. 2004, 126, 4828-4842. (f) Shepherd,
N. E.; Abbenante, G.; Fairlie, D. P. Angew. Chem., Int. Ed. 2004, 43, 2687-
2690. (g) Fasan, R.; Dias, R. L. A.; Moehle, K.; Zerbe, O.; Vrijbloed, J.
W.; Obrecht, D.; Robinson, J. A. Angew. Chem., Int. Ed. 2004, 43, 2109-
2112. (h) Shepherd, N. E.; Hoang, H. N.; Abbenante, G.; Fairlie, D. P. J.
Am. Chem. Soc. 2005, 127, 2974-2983.
(7) Turns: (a) Hruby, V. J.; al-Obeidi, F.; Kazmierski, W. Biochem. J. 1990,
268, 249-262. (b) Toniolo, C. Int. J. Pept. Protein Res. 1990, 35, 287-
300. (c) Hruby, V. J.; Sharma, S. D. Curr. Opin. Biotechnol. 1991, 2, 599-
605. (d) Holzemann, G. Kontakte (Darmstadt) 1991, 3-12 and 55. (e)
Giannis, A. Angew. Chem., Int. Ed. Engl. 1993, 32, 1244-1267. (f) Fairlie,
D. P., Abbenante, G.; March, D. R. Curr. Med. Chem. 1995, 2, 654-686.
(g) Jones, R. M.; Boatman, P. D.; Semple, G.; Shin, Y. J.; Tamura, S. Y.
Curr. Opin. Pharmacol. 2003, 3, 530-543. (h) Kee, S.; Jois, S. D. S. Curr.
Pharm. Des. 2003, 9, 1209-1224.
(8) Loops: (a) Bisang, C.; Weber, C.; Robinson, J. A. HelV. Chim. Acta 1996,
79, 1825-1841. (b) Pfeifer, M. E.; Robinson, J. A. Chem. Commun. 1998,
1977-1978. (c) Bisang, C.; Jiang, L.; Freund, E.; Emery, F.; Bauch, C.;
Matile, H.; Pluschke, G.; Robinson, J. A. J. Am. Chem. Soc. 1998, 120,
7439-7449. (d) Favre, M.; Moehle, K.; Jiang, L.; Pfeiffer, B.; Robinson,
J. A. J. Am. Chem. Soc. 1999, 121, 2679-2685. (e) Wittelsberger, A.;
Keller, M.; Scarpellino, L.; Patiny, L.; Acha-Orbea, H.; Mutter, M Angew.
Chem., Int. Ed. 2000, 39, 1111-1115. (f) Peri, F.; Grell, D.; Dumy, P.;
Yokokawa, Y.; Welzenbach, K.; Weitz-Schmidt, G.; Mutter, M. J. Pept.
Sci. 1999, 5, 313-322. (g) Mathieu, M.; Lehmann, C.; Razaname, A.;
Tuchscherer, G. Lett. Pept. Sci. 1997, 4, 95-100.
(9) Discontinuous surfaces: (a) Mutter, M.; Vuillemeumier, S. Angew. Chem.,
Int. Ed. Engl. 1989, 28, 535-676. (b) Hauert, J.; Fernandez-Carneado, J.;
Michielin, O.; Mathieu, S.; Grell, D.; Schapira, M.; Spertini, O.; Mutter,
M.; Tuchscherer, G.; Kovacsovics, T. ChemBioChem 2004, 5, 856-864.
(10) Multiloops: (a) Mutter, M.; Dumy, P.; Garrouste, P.; Lehmann, C.; Mathieu,
M.; Peggion, C.; Peluso, S.; Razaname, A.; Tuchscherer, G. Angew. Chem.,
Int. Ed. Engl. 1996, 35, 1482-1485. (b) Hamuro, Y.; Calama, M. C.; Park,
H. S.; Hamilton, A. D. Angew. Chem., Int. Ed. Engl. 1997, 36, 2680-
2683. (c) Peluso, S.; Dumy, P.; Eggleston, Im.; Garrouste, P.; Mutter, M.
Tetrahedron 1997, 53, 7231-7236. (d) Peluso, S.; Dumy, P.; Nkubana,
C.; Yokokawa, Y.; Mutter, M. J. Org. Chem. 1999, 64, 7114-7120. (e)
Reineke, U.; Schneider-Mergener, J. Angew. Chem., Int. Ed. 1998, 37, 769-
771.
(1) (a) Milner-White, E. J. Trends Pharmacol. Sci. 1989, 10, 70-74. (b) West,
M. L.; Fairlie, D. P. Trends Pharmacol. Sci. 1995, 16, 67-75. (c) Lee, H.
J. Arch. Pharm. Res. 2002, 25, 572-584. (d) Witt, K. A.; Gillespie, T. J.;
Huber, J. D.; Egleton, R. D.; Davis, T. P. Peptides 2001, 22, 2329-2343.
(e) Topp, E. M. Med. Chem. Res. 1997, 7, 493-508. (f) Chan, O. H.;
Stewart, B. H. Drug DiscoVery Today 1996, 11, 461-473.
(2) (a) Giannis, A. Angew. Chem., Int. Ed. Engl. 1993, 32, 1244-1267. (b)
Gante, J. Angew. Chem., Int. Ed. Engl. 1994, 33, 1699-1720. (c) Schneider,
J. P.; Kelly, J. W. Chem. ReV. 1995, 95, 2169-2187. (d) Fairlie, D. P.,
West, M. L.; Wong, A. K. Curr. Med. Chem. 1998, 5, 29-62.
(3) (a) Zimm, B.; Bragg, J. J. Chem. Phys. 1959, 31, 526-535. (b) Scholtz, J.
M.; Baldwin, R. L. Annu. ReV. Biophys. Biomol. Struct. 1992, 21, 95-
118. (c) Dyson, H. J.; Wright, P. E. Ann. ReV. Biophys. Biophys. Chem.
1991, 20, 519-538.
(4) (a) Glenn, M. P.; Fairlie, D. P. Mini-ReV. Med. Chem. 2002, 2, 433-445.
(b) Loughlin, W. A.; Tyndall, J. D. A.; Glenn, M. P.; Fairlie, D. P. Chem.
ReV. 2004, 104, 6085-6117.
(5) Sheets: (a) Nowick, J. S. Acc. Chem. Res. 1999, 32, 287-296. (b) Moriuchi,
T.; Hirao, T. Chem. Soc. ReV. 2004, 33, 294-301.
(11) Jones, D. D.; Barker, P. D. ChemBioChem 2004, 5, 964-971.
(12) Hamada, K.; Bethge, P. H.; Mathews, F. S. J. Mol. Biol. 1995, 247, 947-
962.
9
10.1021/ja0455300 CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 6563-6572
6563

A R T I C L E S
Singh et al.
We recently reported the use of 4 and 5 to synthesize the
highly functionalized scaffold 6.17 Controlled condensation of
lysine with glutamate side chains in 6 enabled stereoselective
Figure 1. Discontinuous loop surfaces of proteins. (left) Theoretical scaffold
planes that intersect with protein loops define dimensions of both peptide
loops and putative scaffold onto which they will be grafted, providing
scaffold coordinates (X1-4) for attachment of loops; (right) crystal structure12
of cytochrome b562 (pdb: 256b) showing helices (green), heme (purple),
and the specific interhelix tripeptide loop sequences (red and blue) to be
mimicked.
ring closure to form the novel tris-macrocyclic molecule 7,17
a
prototype for two-loop protein surface mimetics. We now
demonstrate that macrocycle 6 has the appropriate dimensions
and side-chain directionality to be a suitable scaffold for building
larger and more complex tris-macrocyclic compounds as pro-
spective protein loop mimetics. However, condensation of the
unequal length side chains of the Lys and Glu residues in 6
had the effect in 7 of staggering the loops with respect to one
another.17 We therefore decided to modify 6, replacing Lys with
Dap (2,3-diaminopropanoic acid) and Glu with Asp, to create
side chains of uniform and reduced length. Herein, we specif-
ically describe the design, synthesis, and three-dimensional solu-
tion structure, determined in water by 2D 1H NMR spectroscopy,
of the densely functionalized tris-macrocycle 8. We also
compare this structure with the corresponding interhelical loops
in the crystal structure of cytochrome b562. We show that
compound 8 projects two tripeptide loops perpendicularly from
the same face of the constrained macrocyclic analogue of
scaffold 6, due to like chirality of all four amino acids inserted
between oxazoles/thiazoles, and we reveal that 8 is a success-
fully designed structural mimetic for the two interhelical loops
The first objective was to develop a scaffold, highly func-
tionalized at appropriate positions (e.g., X1, X2, X3, X4 in Figure
1) to support peptide sequences corresponding to two tripeptide
loops of cytochrome b562. The scaffold needed to have the
capacity to direct the attached loops orthogonally from the same
face of the scaffold into adjacent three-dimensional space, and
to be able to hold the loops in reasonable proximity to one
another. Inspired by Nature’s use of heterocyclic five-membered
ring constraints such as oxazole/oxazoline/thiazole/thiazoline/
thiazolidine (e.g.,1-3, X ) O, S) to regulate peptide shapes in
fungi, bacteria, marine organisms, and plants,13 we have been
exploring structural effects of such constraints in cyclic
peptides.14-18 We14-17 and others19 have synthesized a number
of such constrained macrocycles, those containing oxazole/
thiazole dipeptide surrogates (e.g., 4, 5) tending to be rigid and
psuedo-planar, the amino acid side-chains projecting from either
face depending on the R- or S-configuration at the R-carbon.
Molecular modeling studies17,18 suggested that such cyclic
peptides, with predetermined tethering points, might be ideal
scaffolds on which to mount peptide helices or loops to
structurally mimic discontinuous protein surfaces.
of cytochrome b562
.
(13) (a) Wipf, P. Chem. ReV. 1995, 95, 2115-2134. (b) Ishida, T.; Inoue, M.;
Hamada, Y.; Kato, S.; Shiori, T. J. Chem. Soc., Chem. Commun. 1987,
370-371. (c) Todorova, A. K.; Juttner, F.; Linden, A.; Pluss, T.; von
Philipsborn, W. J. Org. Chem. 1995, 60, 7891-7895. (d) Toske, S. G.;
Fenical, W. Tetrahedron Lett. 1995, 36, 8355-8. (e) Ireland, C. M.;
Scheuer, P. J. J. Am. Chem. Soc. 1980, 102, 5688-5691. (f) Sowinski, J.;
Toogood, P. L. Tetrahedron Lett. 1995, 36, 67-70. (g) Fate, G. D.; Benner,
C. P.; Grode, S. H.; Gilbertson, T. J. J. Am. Chem. Soc. 1996, 118, 11363-
11368. (h) McGeary, R. P.; Fairlie, D. P. Curr. Opin. Drug Discuss. DeV.
1998, 1, 208-217. (i) Lewis, J. R. Nat. Prod. Rep. 2002, 19, 223-258.
(14) Abbenante, G.; Fairlie, D. P.; Gahan, L. R.; Hanson, G. R.; Pierens, G.;
van den Brenk, A. L. J. Am. Chem. Soc. 1996, 118, 10384-10388.
(15) Sokolenko, N.; Abbenante, G.; Scanlon, M. J.; Jones, A.; Gahan, L. R.;
Hanson, G. R.; Fairlie, D. P. J. Am. Chem. Soc. 1999, 121, 2603-2604.
(16) Singh, Y.; Sokolenko, N.; Kelso, M. J.; Gahan, L. R.; Abbenante, G.; Fairlie,
D. P. J. Am. Chem. Soc. 2001, 123, 333-334.
(17) Singh, Y.; Stoermer, M. J.; Lucke, A. J.; Glenn, M. P.; Fairlie, D. P. Org.
Lett. 2002, 4, 3367-3370.
(18) Lucke, A. J.; Tyndall, J. D. A.; Singh, Y.; Fairlie, D. P. J. Mol. Graphics
Modell. 2003, 21, 341-355.
(19) (a) Wipf, P.; Venkatraman, S. J. Org. Chem. 1995, 60, 7224-7229. (b)
Wipf, P.; Fritch, P. C.; Geib, S. J.; Sefler, A. M. J. Am. Chem. Soc. 1998,
120, 4105-4112. (c) Mink, D.; Mecozzi, S.; Rebek, J., Jr. Tetrahedron
Lett. 1998, 39, 5709-5712. (d) Haberhauer, G.; Somogyi, L.; Rebek, J.,
Jr. Tetrahedron Lett. 2000, 41, 5013-5016. (e) Somogyi, L.; Haberhauer,
G.; Rebek, J., Jr. Tetrahedron Lett. 2001, 57, 1699-1708. (f) Boss, C.;
Rasmussen, P. H.; Wartini, A. R.; Waldvogel, S. R. Tetrahedron Lett. 2000,
41, 6327-6331. (g) Wipf, P.; Miller, C. P.; Grant, C. M. Tetrahedron 2000,
56, 9143-9150. (h) Bertram, A.; Blake, A. J.; de Turiso, F. G. L.; Hannam,
J. S.; Jolliffe, K. A.; Pattenden, G.; Skae, M. Tetrahedron 2003, 59, 6979-
6990.
9
6564 J. AM. CHEM. SOC. VOL. 127, NO. 18, 2005

Protein Surface Mimetic of Cytochrome b562 Loops
A R T I C L E S
Figure 2. Design of a protein surface mimetic. End (left) and side (right) views of superimposition of modeled 2-loop mimetic 8, featuring a side-chain
modified scaffold 6 (green) condensed with tripeptides Asp-Asn-Ala (blue) and Glu-Gly-Lys (red) to form loops, and the corresponding interhelix loop
sequences Asp21-Asn22-Ala23 and Glu81-Gly82-Lys83 (purple ribbons) from the crystal structure of cytochrome b562
.
Results
Design of Mimetic for Cytochrome b562 Loops. Computer
OH (9a) (Boc is tert-butoxycarbonyl, Ox is oxazole), Boc-L-
Dap(Alloc)(Ox)-OH (9b) (Alloc is allyloxycarbonyl), H-L-
Asp(OtBu)(Thz)-OH (10a) (Thz is thiazole), and H-L-
Asp(OAllyl)(Thz)-OtBu (10b), in turn respectively prepared
from oxazole 9c and thiazoles 10c and 10d.
modeling was first used to predict whether scaffold 6 might be
of approximate size and shape to support peptides that could
structurally mimic loop regions of a protein. The oxidized form
of cytochrome b562, a monomeric heme-binding four R-helix
bundle protein from the periplasm of Escherichia coli, was
chosen for mimicry on the basis that it represents a typical
4-helix bundle that fairly rigidly defines the locations of its
interhelical loops.12 Specifically its two interhelical loops
Asp21-Asn22-Ala23 (blue) and Glu81-Gly82-Lys83 (red)
(Figure 1b) that connect R-helices R1-R2 and R3-R4, respec-
tively, were grafted in silico from its reported high-resolution
crystal structure (b256)12 onto the energy minimized, and side-
chain modified, scaffold 6 using InsightII. We were particularly
conscious of the fact that the tripeptide loops in cytochrome
b562 were directly tethered to the ends of the four helices, and
so we wished to use as short a linker as possible between the
loops and the scaffold in our target molecule.
Tricyclic construct 8 was energy minimized before superim-
position with template forcing of all 24 tripeptide backbone
heavy atoms (N, CR, C, O) onto the corresponding tripeptide
atoms of cytochrome b562. Template forcing (minimization with
pairwise atom force restraints) was used because it can be useful
for identifying conformations that are of biological relevance.
Figure 2 shows the resulting conformation of 8 after template
forcing using default parameters, the scaffold (green) being of
approximately the correct size and geometry to project the loops
orthogonally from the pseudo-planar scaffold, demonstrated by
the good match (RMSD 1.21 Å) between the atoms of the
tripeptide components of the tris-macrocyclic construct (green)
and the cytochrome (purple ribbons).
Oxazole 9c was conveniently synthesized from the dipeptide
Boc-L-Dap(Z)Ser-OMe (Z is benzyloxycarbonyl) by cyclode-
hydration using diethylaminosulfur trifluoride (DAST)20 fol-
lowed by oxidation with a mixture of bromotrichloromethane
and a strong base 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU).17
It was necessary to replace the Z protecting group of 9c with
Alloc and Fmoc groups, respectively, for compatibility with
subsequent steps. This was accomplished using a deprotection
and reprotection strategy. Catalytic hydrogenolysis of the side
chain Z protecting group followed by reaction of the free amino
group with allyl chloroformate in the presence of DIPEA
(diisopropylethylamine) as base, and hydrolysis of the methyl
ester with LiOH gave Boc-L-Dap(Alloc)(Ox)-OH (9b). Boc-L-
Dap(Fmoc)(Ox)-OH (9a) was easily obtained from 9c after
sequential hydrolysis of the methyl ester, removal of the Z
protecting group by catalytic hydrogenolysis, and subsequent
treatment of the free amino acid (9d) with Fmoc-OSu.
Thiazoles 10c and 10d were synthesized from commercially
available amino acids Fmoc-L-Asp(OtBu)-OH and Fmoc-L-Asp-
(OAllyl)-OH, following conversion to the corresponding thioa-
mides using Lawesson’s reagent and a modified Hantzch
synthesis16 using ethyl bromopyruvate and tert-butyl bromopy-
ruvate, respectively (Supporting Information). Selective removal
of the Fmoc protecting group of 10d with piperidine in
dichloromethane furnished thiazole 10b [H-Glu(OAllyl)(Thz)-
OtBu]. Simultaneous removal of Fmoc and ethyl ester groups
from thiazole 10c using LiOH gave 10a [H-Asp(OtBu)(Thz)-
OH] in 96% yield.
Synthesis of Tris-macrocycle 8. The synthesis of 8 involved
first creating a scaffold 13 (a side-chain modified analogue of
6) with four differentially protected Asp/Dap side chains
(Scheme 1), followed by grafting the tripeptide chains onto the
scaffolds and cyclizing them. This approach involved sequen-
tially forming the two loops, coupling one tripeptide to the
scaffold and cyclizing it, and then coupling the scaffold-loop
conjugate to the second tripeptide which was subsequently
cyclized to produce 8 (Scheme 2).
Synthesis of Macrocyclic Scaffold 13. Scaffold 13, with
side-chain functional groups differentially protected, was syn-
thesized using the dipeptide surrogates Boc-L-Dap(Fmoc)(Ox)-
(20) Phillips, A. J.; Uto, Y.; Wipf, P.; Reno, M. J.; Williams, D. R. Org. Lett.
2000, 2, 1165-1168.
9
J. AM. CHEM. SOC. VOL. 127, NO. 18, 2005 6565

A R T I C L E S
Singh et al.
Scheme 1. Synthesis of Differentially Protected Scaffold 13a
a Reagents: (a) BOP, DIPEA, DMF, room temperature; (b) TFA/DCM, room temperature; (c) formic acid, room temperature; (d) 30% piperidine/DCM,
room temperature.
Pairs of dipeptide surrogates 9a and 10a, and 9b and 10b,
were elaborated to tetrapeptide analogues 11a [Boc-Dap(Fmoc)-
(Ox)Asp(OtBu)(Thz)-OH] and 11b [Boc-Dap(Alloc)(Ox)Asp-
(OAllyl)(Thz)-OtBu], respectively (Scheme 1). Tetrapeptide
analogue 11b was then treated with TFA (trifluoroacetic acid)
and coupled to 11a using BOP ([benzotriazol-1-yl-oxytris-
(dimethylamino) phosphonium] hexafluorophosphate) reagent
in the presence of DIPEA to give linear octapeptide analogue
12a in high yield (91%). Selective removal of the Boc protecting
group of 12a using neat formic acid gave 12b with free N- and
C-terminal functional groups required for final cyclization.
Macrolactamization of 12b under high dilution (1 × 10-3 M)
with BOP using DIPEA and DMF (N,N-dimethylformamide)
resulted in a high isolated yield of cyclic octapeptide 13 (81%),
after removal of the Dap side-chain Fmoc protecting group
(Scheme 1). Cyclization of 12b is favored over cyclooligomer-
ization by the presence of four turn-inducing heterocyclic
oxazole/thiazole constraints, which preorganize it for cycliza-
tion.16
acid side chains of the scaffold. Tripeptides Boc-Glu(OcHx)-
GlyLys(Z)-OH and Boc-Asp(OcHx)Asn(NHTrt)-Ala-OH were
obtained by synthesis on TCP resin using Fmoc protocols and
subsequent cleavage with 1% TFA in dichloromethane as
solvent. The coupling of the tripeptide Boc-Glu(OcHx)GlyLys-
(Z)-OH to the scaffold 13 was best achieved by using DPPA
(diphenylphosphoryl azide)/DIPEA at low temperature (∼5 °C),
with no epimerization detected at the lysine R-carbon atom of
the tripeptide.
Simultaneous removal of Boc and tert-butyl groups with TFA
in the presence of scavengers (water and triisopropylsilane) led
to isolation of compound 14 in 88% yield. Intramolecular loop
formation under dilute conditions (9 × 10-4 M) and at low
temperature (∼5 °C) in DMF using DPPA gave compound 15
in 86% yield, after simultaneous cleavage of Alloc and allyl
ester groups with Pd(PPh3)4 in the presence of 1,3-dimethyl-
barbituric acid and acetic acid in dichloromethane as solvent.
The use of DPPA instead of DOP avoids formation of hydroxy-
benzotriazole which tends to be trapped in the cycle.15
Coupling of the second tripeptide Boc-Asp(OcHx)Asn-
(NHTrt)Ala-OH to 15 was challenging as it required activation
of a peptidyl alanine residue which, unlike N-terminus protected
amino acid, is prone to epimerization at the R-carbon atom. In
addition, both the amine and the carboxylic acid functionalities
in 15 are not protected. The coupling was achieved with minimal
Elaboration of 13 to 8. The next step in the synthesis of
tris-macrocycle 8 involved the grafting of the two tripeptide
loops onto scaffold 13. We developed a strategy where the
C-terminus of the tripeptides was coupled to the scaffold Dap
side chains, prior to intramolecular loop formation by condensa-
tion of the N-terminus of the tripeptides with the carboxylic
9
6566 J. AM. CHEM. SOC. VOL. 127, NO. 18, 2005

Protein Surface Mimetic of Cytochrome b562 Loops
A R T I C L E S
Scheme 2. Synthesis of Tris-macrocyclic Protein Mimetic 8a
a Reagents: (a) Boc-Glu(OcHx)GlyLys(Z)-OH, DPPA, DIPEA, DMF, ∼5 °C; (b) TFA, TIPS, H2O; (c) DPPA, DIPEA, DMF (1 × 10-3 M), ∼5 °C; (d)
Pd(PPh3)4, 1,3-dimethylbarbituric acid, AcOH, DCM, N2, dark, room temperature; (e) Boc-Asp(OcHx)Asn(NHTrt)Ala-OH, DPPA, DIPEA, DMF, ∼5 °C;
(f) HF, p-cresol.
racemization (∼5%) when the tripeptide was activated with
DPPA21 in the presence of just the required amount of DIPEA
as base for about 15 min at ∼5 °C prior to addition of 15;
however, the reaction was very slow. This is not surprising in
view of the fact that the reactive functional groups on the Dap-
(Ox) and Asp(Thz) side chains are only 2 bond lengths from
the scaffold, sterically interfering with the intermolecular
reaction. Compound 16 was isolated in 65% yield after stirring
the mixture of 15, Boc-Asp(OcHx)Asn(NHTrt)Ala-OH, and
DPPA in the presence of DIPEA as base and DMF as solvent
for 4 days followed by treatment with TFA. Conversion to 8
involved stirring 16 with DPPA and DIPEA in DMF for 8 days
to produce the fully protected tris-macrocycle, before removing
the cyclohexyl and Z protecting groups with HF in the presence
of p-cresol as scavenger and purification to 8 by rp-HPLC (35%
isolated yield).
Solution Structure of Scaffold 6. The macrocycle in scaffold
6 possessed a small measure of structure in water. The diagnostic
large 3JRH-NH coupling constants for the Glu(Thz) (8.9 Hz) and
Lys(Ox) (9.1 Hz), when used as dihedral restraints in an
XPLOR-based NMR solution structure determination, resulted
in a moderately defined square macrocycle (RMSD 0.8 Å) with
randomly distributed side chains. The 20 lowest energy struc-
tures (Figure 3) contained no NOE distance (>0.1 Å) or dihedral
(21) Eichler, J.; Lucka, A. W.; Pinilla, C.; Houghten, R. A. Mol. DiVersity 1996,
1, 233-240.
9
J. AM. CHEM. SOC. VOL. 127, NO. 18, 2005 6567

A R T I C L E S
Singh et al.
Figure 3. Top and side views of the 20 lowest energy, NMR-derived, solution structures of scaffold 6 (â carbons colored purple, remaining side-chain
atoms not displayed for clarity).
Figure 4. Newman projections of CR-Câ bonds showing the three possible
ø1 rotamers. ø1 was assigned -60° for Asp(Thz) and +60° for Dap(Ox).
angle (>3°) violations. There were a number of weak NOEs,
notably between scaffold components Lys(Ox) Hâs and Glu-
(Thz) NH, and between Glu(Thz) Hâs and Lys(Ox) NH, that
could not be unambiguously assigned due to the symmetrical
nature of the molecule. These were omitted from the structure
calculation, so as not to bias the macrocycle fold. Variable-
temperature NMR experiments (Supporting Information, Figure
S1) indicated that the four macrocyclic amide NH protons were
not hydrogen bonded based on the moderate temperature
dependence of their chemical shifts (∆δ/T 4.5-7.3 ppb/K).
Solution Structure of Tris-macrocycle 8. Distinguishing
features in the 1H NMR spectra of 8 were large 3JRH-NH coupling
constants for one Asp(Thz) (9.1 Hz) and two Dap(Ox) (8.2 Hz,
3
8.2 Hz) residues, and large (Asn, JRH-NH 8.1 Hz) and small
3
(Ala, JRH-NH 3.0 Hz) coupling constants within the loop
regions. By careful examination of 3JRH-âH coupling constants
for the 2 Asp(Thz) AMX spin systems, in conjunction with NOE
intensities from NOESY spectra, it was possible to obtain
Figure 5. Comparison of 10 lowest energy NMR-derived solution structures
superimposed for tricyclic loop assembly 8 in water: (A), end view; (B),
top view; (C), side view of backbones (side chains not displayed) showing
stereospecific â-proton assignments and thus 2 ø1 restraints
(-60° ( 30°) for these two residues (Figure 4). Additionally,
after D2O exchange the Dap(Ox) residues contain AMX spin
systems, which provided another 2 ø1 restraints (+60° ( 30°)
for these two residues.
macrocyclic scaffold including linkers analogous to 6 (green); and Asp-
Asn-Ala (blue) and Lys-Gly-Glu (red) loops analogous to cytochrome
b562. (D) top view shows the macrocycle with loops omitted.
NOESY spectra at 600 MHz yielded a total of 59 NOE
distance restraints (3 strong short range, 24 medium, 32 long
range), which, together with the 9 dihedral angle restraints
two Dap-oxazole and two Asp-thiazole units are directed
perpendicularly to the macrocycle, which is pseudo-planar
(Figure 5C,D). The NMR structure calculation for 8 was highly
sensitive to dihedral angle, particularly ø1 restraints. Additional
structure calculations performed without ø1 restraints (Support-
ing Information) gave structures with staggered, instead of
eclipsed, loop conformations that were of only slightly (2-4
kcal) lower energy.
3
derived from the ø1 restraints and JRH-NH coupling constants,
were used in an XPLOR-based NMR solution structure deter-
mination. The 10 lowest energy structures of 8 (Figure 5) show
a moderately convergent (RMSD 0.43 Å) saddle-shaped mac-
rocyclic scaffold (Figure 5, green) supporting two tripeptide
loops (red, blue) that project orthogonally from the same face
of the scaffold (Figure 5A). There were no constraining NOEs
between the loops, which consequently displayed some vari-
ability in their positions (Figure 5B). The loops project from
the same face of the macrocycle and are in an eclipsed
conformation (Figure 5C). The R-â side-chain vectors of the
Comparison between Macrocycles and Cytochrome b562
.
The macrocycle that is common to 6 and 8 has a slightly
different structure in each case. In 6, it displays a conforma-
tionally averaged structure (Figure 3) that is almost planar. In
8, the loops appear to alter the conformation of the macrocyclic
9
6568 J. AM. CHEM. SOC. VOL. 127, NO. 18, 2005

Protein Surface Mimetic of Cytochrome b562 Loops
A R T I C L E S
Figure 6. Comparison of loop regions of 8 with those from cytochrome b562. End (left) and side view of an ensemble of 10 NMR solution structures of 8
(only one scaffold (green) shown for clarity) with the backbone loop atoms of Asp21-Asn22-Ala23 (blue) and Glu81-Gly82-Lys83 (red) superimposed
onto corresponding loop atoms (purple ribbons) from the crystal structure of cytochrome b562
.
scaffold analogous to 6, making it more saddle shaped (Figure
5). The loops in this solution structure of 8 are not in close
contact, as evidenced by the lack of interloop NOEs, and this
supporting 4-helix bundles,22 that the geometric requirements
of a scaffold do not need very high precision, so long as the
scaffold plus linkers permit the peptide surfaces access to the
desired three-dimensional space. Although it remains to be
determined for specific examples of protein surface mimicry
just how much flexibility should be introduced into the scaffold
or linkers, we think that some degree of flexibility will generally
be necessary to accommodate the induced fit needed for
cooperative interactions with proteins that mediate function.
is similar to the situation observed in the crystal structure of
12
cytochrome b562
.
In comparison with our reported structure
of 7, involving very short loops17 that bias the macrocycle
toward a rhomboid shape, the larger loops in 8 allow the
macrocyclic scaffold to become slightly less rhombohedral and
a little squarer. This highlights the fact that the constrained
macrocyclic scaffold has some degree of flexibility, despite the
presence of four heterocyclic five-membered rings conjugated
to four planar amide bonds in the 24-membered cycle, and is
susceptible to conformational change induced by the attached
loops in 7 and 8.
Conclusions
Using dipeptide surrogates H-Dap(Ox)-OH and H-Asp(Thz)-
OH and multiple differential protection strategies, we success-
fully constructed a densely functionalized macrocyclic octapep-
tide analogue that has proven to be a useful scaffold for creating
a protein surface mimetic. Two tripeptide sequences, corre-
sponding to the R1-R2 (Lys21-Gly22-Glu23) and R3-R4
InsightII was used to compare the NMR solution structure
of 8 with the cytochrome b562 crystal structure by superimposing
the two tripeptide loop regions. The superimpose function
performs a best fit alignment of two structures, calculating the
root mean square deviation (RMSD) between corresponding
pairs of atoms of a source and target structure. Therefore, the
backbone heavy atoms of the tripeptide loops of 8 (source) were
superimposed on to the corresponding atoms of cytochrome b562
(target). Figure 6 shows a superimposition of all 18 heavy atoms
(backbone tripeptide loop atoms, N, CR, C) from 10 low energy
solution structures of 8 onto the corresponding interhelical
(Asp81-Asn82-Ala83) interhelical loops of cytochrome b562
,
were sequentially grafted and cyclized onto the differentially
protected scaffold by condensation with its Dap and Asp side
chains, producing a tris-macrocyclic compound 8. An NMR-
based structure determination revealed that the tris-macrocycle
structure in water featured two loops projecting orthogonally
from the same face of a saddle-shaped macrocyclic scaffold.
Comparison with the crystal structure of cytochrome b562
indicated a very good match between the location of each
tripeptide loop of 8 and the positions of the corresponding
interhelical tripeptide loops defined in the solid-state structure
of the cytochrome. The ensemble of solution structures for 8
also indicated some loop flexibility, which may be very
important to accommodate the induced fit in protein-protein
interactions. However, just how much flexibility and how long
the linkers between scaffold and loops need to be in specific
protein surface mimetics, to impart optimal interactions with
macromolecular receptors, remains to be determined through
comparative studies on other protein surface mimics. We
conclude that this approach, and even this type of scaffold,
appears to be very promising for the structural mimicry of
discontinuous loop surfaces of proteins.
tripeptide loops from the crystal structure of cytochrome b562
.
Individually, each loop (9 backbone atoms) of 8 superimposed
quite well onto the respective loop atoms in the cytochrome
crystal structure (Asp21-Asn22-Ala23 (blue), mean RMSD
0.5 Å; Glu81-Gly82-Lys83 (red), mean RMSD 0.7 Å). In the
static crystal structure of the native cytochrome protein, the loop
regions are somewhat rigid, being highly constrained by being
joined directly to the ends of R-helices, the distance between
them varying according to the dynamics of the R-helical bundle.
In the case of 8, the flexible aliphatic components of Dap and
Asp linkers provided more mobility to the loops. The combined
mean RMSD for both loops of 8 on the cytochrome loops was
2.4 Å.
Experimental Section
It is clear that this template approach has successfully
anchored the loops of 8 within comparable three-dimensional
space to that occupied by the interhelical loops of cytochrome
b562. This is consistent with our previous assertion for scaffolds
Materials and Methods. Materials obtained commercially were
reagent grade unless otherwise stated. Preparative scale reverse phase
(22) Wong, A.; Jacobsen, M. P.; Winzor, D. J.; Fairlie, D. P. J. Am. Chem.
Soc. 1998, 120, 3836-3841.
9
J. AM. CHEM. SOC. VOL. 127, NO. 18, 2005 6569

A R T I C L E S
Singh et al.
HPLC separations were performed on a Vydac 218TP101550 C18
column; analytical reverse phase HPLC was performed on a Vydac
218TP54 C18 column [Rt(1)] and Phenomenex Luna 5µ C18 column
[Rt(2)], using gradient mixtures of water/0.1%TFA (solvent system A)
and water 10%/acetonitrile 90% /TFA 0.1% (solvent system B). Mass
spectra were obtained on a hybrid quadrupole TOF mass spectrometer
(PE SCIEX API QSTAR Pulsar) equipped with an Ionspray (pneumati-
cally assisted electrospray) operating at ambient temperature (ISMS).
Molecular Modeling. Simulations were performed with InsightII
software23 running on Silicon Graphics R10000 Octane workstations.
InsightII modules used were Builder, Search and Compare, Discover
and Discover 3. Scaffold 6 was truncated to convert the Lys and Glu
side chains into Dap and Asp, and the tripeptide interhelical loops were
grafted onto the scaffold using the Builder module. The N-terminus of
tripeptide Asp-Asn-Ala was connected to an aspartic acid side chain
of modified 6 forming an amide bond, and the C-terminus of the
tripeptide connected to the Dap amine side chain of 6 forming a loop
that included oxazole. The N-terminus of Glu-Gly-Lys was connected
to the other aspartic acid side chain of 6 to form an amide bond, and
the C-terminus of this tripeptide was connected to the other Dap amine
side chain of 6 to form a second loop that also included oxazole. The
resulting structure 8 was then optimized within Builder using default
parameters. Force field CFF91 was then selected, and the potentials,
partial charges, and formal charges of the 8 were fixed. Discover 3
was then used to perform energy minimization using default parameters
and gradient methods. Energy minimization was stopped once the final
convergence value reached 0.001. Compound 8 was then subjected to
minimization with template forcing constraints (template forcing).
Backbone atoms of the flexible tripeptide loop regions of 8 were
minimized with template forcing to constrain atoms to the same
positions as corresponding backbone atoms of the tripeptide interhelical
loops of cytochrome b562. This was performed as described in the
InsightII module using a series of decreasing force constraints over
the loop tripeptide backbone atoms, repeated until 8 could be minimized
without template forcing, to produce a structure with an RMSD of 1.21
in combination with NOE peak intensities at short mixing times (40,
100, 120 ms). Other diastereotopic protons were not stereospecifically
assigned, and corresponding distance restraints were adjusted with
standard pseudo-atom corrections.26 Solution structures were calculated
using simulated annealing and energy minimization protocols within
XPLOR 3.851.27 Initial structures were generated using random φ,ψ
dihedral angles and energy-minimized. Preliminary structures were
generated by torsion angle simulated annealing involving a high-
temperature (50 000 K) phase comprising 1000 steps of 0.015 ps of
torsion angle dynamics, a cooling phase with 1000 steps of 0.015 ps
of torsion angle dynamics during which the temperature was lowered
to 0 K, and an energy minimization phase comprising 2000 steps of
Powell minimization. Typically, 50 structures were calculated, and the
20 structures of lowest energy were superimposed and compared using
Insight II.23
Structural Superimpositions. InsightII superimposition calculates
the root mean square deviation (RMSD) between corresponding pairs
of atoms of a source and target structure. For each individual NMR
structure of 8, the tripeptide backbone (N, CR, CO) atoms of the two
loops (24 source atoms) were superimposed on to the corresponding
backbone atoms (24 target atoms) of the cytochrome b562 crystal
structure. The average superimposition RMSD of 10 structures when
compared to the crystal structure was 2.4 Å over the 24 backbone atoms
of both loops. However, superimposition of the 12 backbone atoms of
each loop gave lower average RMSD values of 0.5 Å (Asp-Asn-
Ala, R1-R2) and 0.7 Å (Glu-Gly-Lys, R3-R4).
Chemical Synthesis. Cyclo[-Dap(Alloc)(Ox)Asp(OAllyl)(Thz)-
Dap(H)(Ox)Asp(OtBu)(Thz)-] (13). Boc-Dap(Fmoc)(Ox)Asp(OtBu)-
(Thz)Dap(Alloc)(Ox)Asp(OAllyl)(Thz)-OH (12a, 245 g, 0.20 mmol)
was stirred with HCO2H (15 mL) at room temperature for 15 min. The
reaction mixture was concentrated under vacuum at ambient temper-
ature, and the residue was purified by rp-HPLC to give TFA‚H-Dap-
(Fmoc)(Ox)Asp(OtBu)(Thz)Dap(Alloc)(Ox)Asp(OAllyl)(Thz)-OH (12b)
as a white solid (202 mg, 82%). ISMS: M + H ) 1123.3. HRMS: M
+ H ) experimental 1123.3264, calculated 1123.3284. HPLC: Rt(1)
) 24.7 min and Rt(2) ) 25.9 min, 3.33%/min linear gradient starting
from 0% B. TFA‚H-Dap(Fmoc)(Ox)Asp(OtBu)(Thz)Dap(Alloc)Asp-
(OAllyl)(Thz)-OH (12b, 200 mg, 0.16 mmol) and BOP (110 mg, 0.25
mmol) were dissolved in DMF (125 mL) and stirred with DIPEA (0.1
mL, 0.57 mmol) at room temperature for 48 h. The solvent was
evaporated in vacuo. The residue was stirred with 30% piperidine/DCM
(25 mL) at room temperature for 1 h and purified by rpHPLC to yield
cyclo[-Dap(Alloc)(Ox)Asp(OAllyl)(Thz)Dap(H)(Ox)Asp(OtBu)(Thz)-
] (13) as a white solid (130 mg, 81%). ISMS: M + H ) 883.2.
HRMS: M + H ) experimental 883.2476, calculated 883.2498.
HPLC: Rt(1) ) 21.1 min and Rt(2) ) 22.5 min, 3.33%/min linear
Å over the 24 backbone (CO, CR, N) atoms of 8 and cytochrome b562
.
1
NMR Spectroscopy. H, NMR spectra were recorded on a Bruker
Avance 600 spectrometer on samples containing 1-4 mM peptide in
water at pH 4.0. Proton and carbon assignments were made using
TOCSY (80 ms mixing time), DQF-COSY, NOESY (40, 100, 350 ms
mixing time), HSQC and HMBC spectra according to the sequential
assignment method.24 Water suppression in 2D experiments was
performed using a 3-9-19 Watergate pulse sequence. Variable-temper-
ature 1D 1H and TOCSY spectra were typically collected at 5 K
increments from 278 to 313 K. For identification of slowly exchanging
1
amides, a series of 1D H and TOCSY spectra were run immediately
1
after dissolving the peptide (3 mg) in D2O (600 µL). All spectra were
gradient starting from 0% B. H NMR (DMSO-d6): δ 8.97, m, 2H,
analyzed in Xwinnmr.25
Asp-RNH & Dap-RNH; 8.76, s, 1H, Ox-H; 8.70, d, 1H, 3JAspRNH-AspRH
) 8.4 Hz, Asp-RNH; 8.63, s, 1H, Ox-H; 8.53, d, 1H, 3JDapRNH-DapRH
)
Structure Calculations. Backbone dihedral angle restraints were
derived from 3JRH-NH coupling constants measured from high-resolution
6.7 Hz, Dap-RNH; 8.27, s, 1H, Thz-H; 8.26, s, 1H, Thz-H; 8.13, br s,
3H, Dap-γNH3+; 7.57, t, 1H, 3JDapγNH-DapγH ) 6.0 Hz, Dap-γNH; 5.85,
m, 2H, 2 × H2CdCH-CH2-O; 5.76, m, 2H, 2 × Asp-RH; 5.70, m,
1H, Dap-RH; 5.26, d, 1H, J ) 17.3 Hz, H2CdCH-CH2-O; 5.23, d,
1H, J ) 17.4 Hz, H2CdCH-CH2-O; 5.17, d, 1H, J ) 10.5 Hz, H2Cd
CH-CH2-O; 5.15, d, 1H, J ) 10.5 Hz, H2CdCH-CH2-O; 5.11, m,
1H, Dap-RH; 4.51, d, 2H, J ) 5.0 Hz, ester H2CdCH-CH2-O; 4.44,
br s, 2H, Alloc H2CdCH-CH2-O; 3.40-3.62, m, 4H, Dap-âH; 3.34,
dd, 2H, J ) 12.0, 7.3 Hz, Asp-âH; 3.08, dd, 1H, J ) 16.4, 8.6 Hz,
Asp-âH; 2.97, dd, 1H, J ) 16.4, 6.3 Hz, Asp-âH; 1.30, s, 9H, t-Bu.
13C NMR (DMSO-d6): δ 169.72 (Asp γ-CO2Allyl); 169.68 (Asp
γ-CO2tBu); 168.80 (Thz C-2); 168.45 (Thz C-2); 163.22 (Ox C-2);
160.67 (Ox C-2); 160.55 (Thz 4-CONH); 160.36 (Thz 4-CONH);
159.12 (Ox 4-CONH); 159.03 (Ox 4-CONH); 156.24 (Alloc-CO);
1
1D H NMR spectra. φ angles were restrained to -120° ( 20° for
3JRH-NH > 8.0 Hz and to -60° ( 20° for 3JRH-NH < 6 Hz. No explicit
hydrogen bonds were used as distance restraints in structure calculations.
NOE distance restraints were derived from spectra at 300 K (compound
6) and 303 K (compound 8), but lower or higher temperature NOESY
spectra were also used to resolve ambiguities arising from NH overlap.
Upper distance restraints of 2.7, 3.5, 5.0, and 6.0 Å were used for strong,
medium, weak, and very weak nOe’s respectively. Stereospecific
assignments of â-methylene protons and ø-1 dihedral angles for AMX
3
spin systems (DTZ) were determined by examining JHR,Hâ coupling
constants in high-resolution 1D 1H NMR before and after D2O exchange
(23) InsightII Modeling EnVironment, Release 2000; Accelrys Inc., San Diego,
CA, 2001.
(24) Wuthrich, K. NMR of Proteins and Nucleic Acids; Wiley-Interscience:
New York, 1986.
(26) Wuthrich, K.; Billeter, M.; Braun, W. J. J. Mol. Biol. 1983, 169, 949-961
(27) Bru¨nger, A. T. X-PLOR Manual Version 3.1; Yale University, New Haven,
CT, 1992.
(25) Xwinnmr v2.6 and 3.5, Copyright 2004; Bruker BioSpin GmbH, Rhein-
stetten, Federal Republic of Germany.
9
6570 J. AM. CHEM. SOC. VOL. 127, NO. 18, 2005

Protein Surface Mimetic of Cytochrome b562 Loops
A R T I C L E S
148.47 (Thz C-4); 148.25 (Thz C-4); 142.66 (Ox C-5); 142.61 (Ox
C-5); 136.20 (Ox C-4); 135.60 (Ox C-4); 133.43 (Alloc H2CdCH-
CH2-O); 132.34 (ester H2CdCH-CH2-O); 126.00 (Thz C-5); 125.01
(Thz C-5); 117.79 (ester H2CdCH-CH2-O); 117.05 (Alloc H2Cd
CH-CH2-O); 80.63 (-CO2C(CH3)3); 64.63 (H2CdCH-CH2-O);
64.60 (H2CdCH-CH2-O); 48.32 (Dap-RCH); 46.40 (Asp-RCH);
45.14 (Asp-RCH); 44.57 (Dap-RCH); 42.28 (Dap-âCH2); 39.50 (Dap-
âCH2); 39.08 (Asp-âCH2); 36.63 (Asp-âCH2); 27.53 (-CO2C(CH3)3.
CH-O; 4.42, m, 1H, Glu-RH; 3.90-3.97, m, 2H, Gly-RH & Lys-RH;
3.74, m, 1H, Dap-âH; 3.41-3.55, m, 4H, 3 × Dap-âH & Gly-RH;
3.00-3.17, m, 4H, Asp-âH; 2.95, m, 2H, Lys-ꢀH; 2.25, m, 2H, Glu-
γH; 1.92, m, 1H, Glu-âH; 1.73, m, 3H, Glu-âH & 2 × cyclohexyl-H;
1.65, m, 2H, cyclohexyl-H; 1.57, m, 1H, Lys-âH; 1.47, m, 2H, Lys-
âH & Lys-δH; 1.28-1.38, m, 7H, Lys-δH & 6 × cyclohexyl-H; 1.14-
1.24, m, 2H, Lys-γH.
Compound 16. A mixture of Boc-Asp(OcHx)Asn(NHTrt)Ala-OH
(86 mg, 0.12 mmol), DPPA (25 µL, 0.12 mmol), and DIPEA (21 µL,
0.12 mmol) in DMF (5 mL) was stirred at ∼5 °C for 15 min and then
added to a solution of 15 (60 mg, 0.045 mmol) and DIPEA (15.8 µL,
0.091 mmol) in DMF (10 mL) at ∼5 °C under an atmosphere of
nitrogen. The stirring was continued at this temperature for 4 days.
The reaction mixture was concentrated, and the residue was freeze-
dried. ISMS: M + H ) 1939.7. The freeze-dried material was stirred
with 75% TFA/DCM (20 mL) for 2 h at room temperature. The reaction
mixture was concentrated under reduced pressure, and the residue was
purified by rpHPLC to give 16 as a white solid (50.2 mg, 65%).
ISMS: M + H ) 1597.6. HRMS: M + H ) experimental 1597.5821,
calculated 1597.5835. HPLC: Rt(1) ) 22.8 min and Rt(2) ) 23.7 min,
Compound 14. A solution of cyclo[-Dap(Alloc)(Ox)Asp(OAllyl)-
(Thz)-Dap(H)(Ox)Asp(OtBu)(Thz)-] (13, 125 mg, 0.13 mmol) and
Boc-Glu(OcHx)GlyLys(Z)-OH (136 mg, 0.21 mmol) in DMF (25 mL)
was cooled to ∼5 °C under an atmosphere of nitrogen and stirred with
a mixture of DPPA (70 µL, 0.33 mmol) and DIPEA (70 µL, 0.4 mmol)
for 12 h at the above temperature. The reaction mixture was then
warmed to 20 °C and stirred for further 36 h. The solvent was
evaporated in vacuo, and the residue was dissolved in water/acetonitrile
(1:1, 50 mL), and freeze-dried. ISMS: M + H ) 1513.6. The freeze-
dried material was stirred with 75% TFA/DCM (20 mL) for 2 h at
room temperature. The reaction mixture was concentrated under reduced
pressure, and the residue was purified by rpHPLC to give 14 as a white
solid (155 mg, 84%). ISMS: M + H ) 1357.5. HRMS: M + H )
experimental 1357.4621, calculated 1357.4612. HPLC: Rt(1) ) 23.6
min and Rt(2) ) 24.8 min, 3.33%/min linear gradient starting from
1
3.33%/min linear gradient starting from 0% B. H NMR (DMSO-d6):
3
δ 8.93, d, 1H, JDapRNH-DapRH ) 7.6 Hz, Dap-RNH; 8.80, d, 1H,
3JDapRNH-DapRH ) 8.5 Hz, Dap-RNH; 8.63, s, 1H, Ox-H; 8.60, m, 2H,
1
3
3
0% B. H NMR (DMSO-d6): δ 8.84, d, 1H, JAspRNH-AspRH ) 8.7 Hz,
Ox-H & Glu-RNH; 8.56, d, 1H, JAspRNH-AspRH ) 7.6 Hz, Asp-RNH;
3
3
Asp-RNH; 8.79, d, 2H, JRNH-RH ) 8.6 Hz, Asp-RNH & Dap-RNH;
8.48, d, 1H, JAspRNH-AspRH ) 7.5 Hz, Asp-RNH; 8.29, s, 1H, Thz-H;
8.76, d, 1H, 3JDapRNH-DapRH ) 8.6 Hz, Dap-RNH; 8.61-8.63, m, 3H, 2
× Ox-H & Gly-RNH; 8.34, m, 1H, Dap-γNH; 8.25, s, 2H, 2 × Thz-
8.25, s, 1H, Thz-H; 8.11, m, 2H, Dap-γNH & Asn-RNH; 8.02, m, 1H,
Gly-RNH; 7.95, m, 4H, Ala-RNH & Asp-NH3+; 7.90, d, 1H,
H; 8.13, br s, 3H, Glu-RNH3+; 8.07, d, 1H, JLysNH-LysRH ) 8.2 Hz,
3JLysRNH-LysRH ) 7.6 Hz, Lys-RNH; 7.81, t, 1H, JDapγNH-DapγH ) 5.6
3
3
Lys-RNH; 7.56, t, 1H, 3JDapγNH-DapγH ) 5.6 Hz, Dap-γNH; 7.29-7.38,
m, 5H, Ar-H; 7.17, t, 1H, 3JLysꢀNH-LysꢀH ) 5.7 Hz, Lys-ꢀNH; 5.82, m,
2H, H2CdCH-CH2-O; 5.77, m, 1H, Asp-RH; 5.71, m, 1H, Asp-RH;
5.40, m, 1H, Dap-RH; 5.37, m, 1H, Dap-RH; 5.24, dd, 1H, J ) 15.6,
1.7 Hz, H2CdCH-CH2-O; 5.21, dd, 1H, J ) 12.1, 1.6 Hz, H2Cd
CH-CH2-O; 5.15, dd, 1H, J ) 10.5, 1.4 Hz, H2CdCH-CH2-O;
5.11, dd, 1H, J ) 10.5, 1.4 Hz, H2CdCH-CH2-O; 5.00, s, 2H, PhCH2;
4.67, m, 1H, cyclohexyl CH-O; 4.52, d, 2H, J ) 4.6 Hz, ester H2Cd
CH-CH2-O; 4.44, br s, 2H, Alloc H2CdCH-CH2-O; 4.16, m, 1H,
Lys-RH; 3.91, m, 1H, Dap-âH; 3.84-3.89, m, 2H, Gly-RH & Glu-
RH; 3.75-3.80, m, 2H, Gly-RH & Dap-âH; 3.53, m, 1H, Dap-âH;
3.45, m, 1H, Dap-âH; 3.00-3.26, m, 4H, Asp-âH; 2.89, m, 2H, Lys-
ꢀH; 2.43, m, 2H, Glu-γH; 1.97, m, 2H, Glu-âH; 1.75, m, 2H,
cyclohexyl-H; 1.64, m, 2H, cyclohexyl-H; 1.45-1.52, m, 3H, Lys-âH
& 2 × cyclohexyl-H; 1.27-1.41, m, 5H, Lys-âH, 2 × cyclohexyl-H
& 2 × Lys-δH; 1.10-1.24, m, 4H, 2 × Lys-γH & 2 × cyclohexyl-H.
Hz, Dap-γNH; 7.51, br s, 1H, Asn-CONH; 7.25-7.37, m, 5H, Ar-H;
7.19, t, 1H, JLysꢀNH-LysꢀH ) 5.7 Hz, Lys-ꢀNH; 6.97, br s, 1H, Asn-
3
CONH; 5.90, m, 1H, Asp-RH; 5.48-5.54, m, 3H, 2 × Dap-RH & Asp-
RH; 5.00, s, 2H, PhCH2; 4.63, m, 2H, cyclohexyl CH-O; 4.32, m,
1H, Asn-RH; 3.88-4.15, m, 6H, Glu-RH, 2 × Asp-âH, Ala-RH, Dap-
âH & Lys-RH; 3.69, m, 2H, Gly-RH & Dap-âH; 3.56, m, 2H, Gly-RH
& Dap-âH; 3.34-3.48, m, 4H, 2 × Dap-âH & 2 × Asp-âH; 3.10-
3.27, m, 2H, Asp-âH; 2.96, m, 2H, Lys-ꢀH; 2.79-2.90, m, 2H, Asp-
âH; 2.55-2.67, m, 2H, Asn-âH; 2.38, m, 1H, Glu-γH; 2.27, m, 1H,
Glu-γH; 1.98, m, 1H, Glu-âH; 1.87, m, 1H, Glu-âH; 1.74, m, 2H,
cyclohexyl-H; 1.64, m, 2H, cyclohexyl-H; 1.55, m, 1H, Lys-âH; 1.47,
m, 2H, Lys-âH & cyclohexyl-H; 1.16-1.40, m, 9H, 2 × Lys-δH, 5 ×
Cyclohexyl-H & 2 × Lys-γH; 1.08, d, 3H, J ) 7.2 Hz, Ala-âH.
Compound 8. A solution of 16 (42 mg, 0.025 mmol) in DMF (40
mL) was cooled to ∼5 °C and stirred with a mixture of DPPA (15 µL,
0.07 mmol) and DIPEA (15 µL, 0.086 mmol) under an atmosphere of
nitrogen for 48 h. The reaction mixture was then allowed to warm to
∼25 °C and stirred for further 8 days. The solvent was evaporated
under reduced pressure at ambient temperature, and the residue was
purified by rp-HPLC to give the protected tris-macrocycle as a white
solid in quantitative yield. ISMS: M + H ) 1579.6. HRMS: M + H
) experimental 1579.5746, calculated 1579.5729. HPLC: Rt(1) ) 24.5
min and Rt(2) ) 26.4 min, 3.33%/min linear gradient starting from
0% B. The protected tris-macrocycle (12 mg, 8.2 × 10-6 mol) was
treated with HF in the presence of p-cresol and purified by rp-HPLC
to give 8 as a white solid (4 mg, 35%). ISMS: M + H ) 1281.4.
HRMS: M + H ) experimental 1282.3776, calculated 1281.3796.
HPLC: Rt(1) ) 14.8 min and Rt(2) ) 15.9 min, 3.33%/min linear
Compound 15. A solution of 14 (150 mg, 0.10 mmol) in DMF (110
mL) was cooled to ∼5 °C under an atmosphere of nitrogen and stirred
with a mixture of DPPA (70 µL, 0.33 mmol) and DIPEA (62 µL, 0.36
mmol) for 42 h. The solvent was evaporated under vacuum, and the
residue was freeze-dried. ISMS: M + H ) 1339.5. The freeze-dried
material was dissolved in dichloromethane (20 mL) and stirred with a
mixture of Pd(PPh3)4 (cat.), 1,3-dimethylbarbituric acid (88 mg, 0.55
mmol, 98%), and acetic acid (8 µL, 0.14 mmol) under an atmosphere
of nitrogen and in the absence of light for 2 h. The solvent was removed
under reduced pressure, and the residue was purified by rp-HPLC to
give 15 as a white solid (116 mg, 86%). ISMS: M + H ) 1215.4.
HRMS: M + H ) experimental 1215.3962, calculated 1215.3982.
HPLC: Rt(1) ) 21.4 min and Rt(2) ) 22.6 min, 3.33%/min linear
1
gradient starting from 0% B. H NMR (H2O/D2O, 303 K): δ 9.37, d,
1
3
gradient starting from 0% B. H NMR (DMSO-d6): δ 8.94, d, 1H,
1H, JAsp(Thz)NH-Asp(Thz)RH ) 9.1 Hz, Asp(Thz)3-RNH; 8.92, d, 1H,
3JDapRNH-DapRH ) 8.6 Hz, Dap-RNH; 8.88, d, 1H, 3JAspRNH-AspRH ) 8.2
Hz, Asp-RNH; 8.82, d, 1H, 3JDapRNH-DapRH ) 9.0 Hz, Dap-RNH; 8.66,
s, 1H, Ox-H; 8.44, d, 1H, 3JAspRNH-AspRH ) 8.9 Hz, Asp-RNH; 8.37, s,
1H, Ox-H; 8.29, s, 1H, Thz-H; 8.20, s, 1H, Thz-H; 8.11, m, 2H, Glu-
RNH & Dap-γNH; 8.03, m, 4H, Lys-RNH & Dap-γNH3+; 7.80, m,
3JAsp(Thz)NH-Asp(Thz)RH ) 8.1 Hz, Asp(Thz)1-R NH; 8.91, d, 1H, 3JAspNH-AspRH
) 6.9 Hz, Asp-RNH; 8.90, d, 1H, 3JDap(Ox)NH-Dap(Ox)RH ) 8.2 Hz, Dap-
3
(Ox)4-RNH; 8.70, d, 1H, JDap(Ox)NH-Dap(Ox)RH ) 8.2 Hz, Dap(Ox)2-R
NH; 8.58, d, 1H, 3JGluNH-GluRH ) 7.4 Hz, Glu-RNH; 8.389, s, 1H, Ox-
3
H; 8.387, s, 1H, Ox-H; 8.35, d, 1H, JAlaNH-AlaRH ) 4.8 Hz, AlaNH;
1H, Gly-RNH; 7.25-7.40, m, 5H, Ar-H; 7.22, t, 1H, 3JLysꢀNH-LysꢀH
)
8.33, m, 1H, Dap(Ox)4-γNH; 8.31, m, 1H, Dap(Ox)2-γNH; 8.24, t,
3
5.8 Hz, Lys-ꢀNH; 5.71, m, 1H, Asp-RH; 5.65, m, 2H, Asp-RH & Dap-
RH; 5.59, m, 1H, Dap-RH; 5.00, s, 2H, PhCH2; 4.64, m, 1H, cyclohexyl
1H, JGlyNH-GlyRH ) 6.3 Hz, Gly-RNH; 8.21, s, 1H, Thz-H; 8.207, s,
3
1H, Thz-H; 8.11, d, 1H, JDapNH-DapRH ) 7.1 Hz, LysRNH; 7.81, d,
9
J. AM. CHEM. SOC. VOL. 127, NO. 18, 2005 6571

A R T I C L E S
Singh et al.
1H, 3JAsnNH-AsnRH ) 8.1 Hz, Asn-RNH; 7.69, s, 1H, AsnδNH; 7.51, br
s, 3H, Lys-zNH3; 6.91, s, 1H, AsnδNH; 6.06, m, 1H, Asp(Thz)3-RH;,
5.69, m, 1H, Asp(Thz)1-RH; 5.46, m, 1H, Dap(Ox)4-RH; 5.39, m, 1H,
Dap(Ox)2-RH; 4.65, m, 1H, Asn-RH; 4.50, m, 1H, Glu-RH; 4.49, m,
1H, Asp-RH; 4.18, m, 1H, Lys-RH; 4.16, m, 1H, Dap(Ox)2-âH; 4.03,
m, 1H, Ala-RH; 3.98, m, 1H, Asp(Thz)1-âH; 3.97, m, 1H, Dap(Ox)4-
âH; 3.96, m, 1H, Gly-RH; 3.88, m, 1H, Gly-RH; 3.85, m, 1H, Dap-
Acknowledgment. We thank the ARC and NHMRC for
financial support and the ARC Special Research Centre for
Functional and Applied Genomics for technical advice.
Supporting Information Available: Further details of chemi-
cal synthesis and compound characterization; 1D and 2D NMR
spectra for 6, 8, 13, and 14; structure calculations, variable-
temperature NMR data, D2O exchange data for 6 and 8 (Figures
S1-S14); tables of NOE distance and dihedral angle restraints
for 6 and 8, RMSD comparisons of structures for 8 with
cytochrome b562 (Tables S1-S3). This material is available free
of charge via the Internet at http://pubs.acs.org.
2
3
(Ox)4-âH; 3.71, dd, 1H, JDapâH-DapâH 16.4 Hz, JDapâH-DapRH ) 11.5
Hz, Asp(Thz)3-âH; 3.64, m, 1H, Dap(Ox)2-âH; 3.18, dd, 1H,
3
2JDapâH-DapâH 14.7 Hz, JDapâH-DapRH 3.6 Hz, Asp(Thz)1-âH; 3.11, dd,
2
3
1H, JDapâH-DapâH 16.4 Hz, JDapâH-DapRH 3.0 Hz, Asp(Thz)3-âH; 3.05,
m, 1H, Asp-âH; 3.00, m, 2H, Lys-ꢀH; 2.81, m, 1H, Asp-âH; 2.51, m,
1H, Asn-âH; 2.50, m, 2H, Glu-γH; 2.14, m, 1H, Glu-âH; 2.10, m,
1H, Asn-âH; 2.02, m, 1H, Glu-âH; 1.80, m, 1H, Lys-âH; 1.66, m,
2H, Lys-δH; 1.66, m, 1H, Lys-âH; 1.44, m, 1H, Lys-γH; 1.37, m, 1H,
Lys-γH; 1.27, d, 7.4 Hz, Ala-âH.
JA0455300
9
6572 J. AM. CHEM. SOC. VOL. 127, NO. 18, 2005