Terphenyl-based Bak BH3 α-helical proteomimetics as low-molecular-weight antagonists of Bcl-xL

Article abstract of DOI:10.1021/ja050122x

We describe a general method for the mimicry of one face of an α-helix based on a terphenyl scaffold that spatially projects functionality in a manner similar to that of two turns of an α-helix. The synthetic scaffold reduces the flexibility and molecular

Full text of DOI:10.1021/ja050122x

Published on Web 07/02/2005
Terphenyl-Based Bak BH3 r-Helical Proteomimetics as
Low-Molecular-Weight Antagonists of Bcl-xL
Hang Yin,† Gui-in Lee,† Kristine A. Sedey,‡ Olaf Kutzki,† Hyung Soon Park,†
Brendan P. Orner,† Justin T. Ernst,† Hong-Gang Wang,‡ Said M. Sebti,‡ and
Andrew D. Hamilton*,†
Contribution from the Department of Chemistry, Yale UniVersity, P. O. Box 208107, New
HaVen, Connecticut 06520-8107, and Drug DiscoVery Program, H. Lee Moffitt Cancer Center
and Research Institute, Departments of Oncology and Biochemistry and Molecular Biology,
UniVersity of South Florida, Tampa, Florida 33612
Received January 8, 2005; E-mail: Andrew.Hamilton@yale.edu
Abstract: We describe a general method for the mimicry of one face of an R-helix based on a terphenyl
scaffold that spatially projects functionality in a manner similar to that of two turns of an R-helix. The synthetic
scaffold reduces the flexibility and molecular weight of the mimicked protein secondary structure. We have
applied this design to the development of antagonists of the R-helix binding protein Bcl-xL. Using a sequential
synthetic strategy, we have prepared a library of terphenyl derivatives to mimic the helical region of the
Bak BH3 domain that binds Bcl-xL. Fluorescence polarization assays were carried out to evaluate the ability
of terphenyl derivatives to displace the Bcl-xL-bound Bak peptide. Terphenyl 14 exhibited good in vitro
affinity with a Ki value of 0.114 µM. These terphenyl derivatives were more selective at disrupting the
Bcl-xL/Bak over the HDM2/p53 interaction, which involves binding of the N-terminal R-helix of p53 to HDM2.
Structural studies using NMR spectroscopy and computer-aided docking simulations suggested that the
helix binding area on the surface of Bcl-xL is the target for the synthetic ligands. Treatment of human
embryonic kidney 293 (HEK293) cells with terphenyl derivatives resulted in the disruption of the binding of
Bcl-xL to Bax in intact cells.
many protein-protein interfaces.2,4 However, only a few
Introduction
examples of such molecules have been prepared. In pioneering
The development of low-molecular-weight agents that disrupt
protein-protein interactions is still a challenging goal in
medicinal chemistry because the protein interfacial surfaces
involved are relatively large and featureless compared to those
of enzyme active sites.1 Conventional methods for identifying
inhibitors of protein-protein interactions have generally in-
volved the preparation and screening of chemical libraries to
discover lead compounds although often with little success.2
The rational design of such inhibitors offers a compelling
alternative as it can often be based on a structural knowledge
of the interface. In particular, synthetic scaffolds that mimic
the key elements of a protein surface can potentially lead to
small molecules with the full activity of a protein domain, a
fraction of the molecular weight, and no peptide bonds.
work, functionalized indanes have been shown to match the i
and i + 1 residues of an R-helix and give potent inhibitors of
the tachykinin receptors.5 Kahne et al. reported an R-helix
mimic, based on an oligosaccharide scaffold, which binds the
minor groove of DNA with selectivity over RNA.6 We have
recently introduced a proteomimetic strategy based on molecules
that mimic the structure and recognition function of discon-
tiguous stretches of an R-helix.7 We have extended this notion
by using a terphenyl scaffold with appended functionality to
mimic substituents projecting from the surface of an R-helix.8-10
In this design, the terphenyl is expected to adopt a staggered
conformation and closely reproduce the position and angular
(4) Cochran, A. G. Curr. Opin. Chem. Biol. 2001, 5, 654.
(5) Horwell, D.; Pritchard, M.; Raphy, J.; Ratcliffe, G. Immunopharmacology
1996, 33, 68. Horwell, D. C.; Howson, W.; Ratcliffe, G. S.; Willems, H.
M. G. Bioorg. Med. Chem. 1996, 4, 33.
(6) Xuereb, H.; Maletic, M.; Gildersleeve, J.; Pelczer, I.; Kahne, D. J. Am.
Chem. Soc. 2000, 122, 1883.
(7) Orner, B. P.; Salvatella, X.; Quesada, J. S.; de Mendoza, J.; Giralt, E.;
Hamilton, A. D. Angew. Chem., Int. Ed. 2002, 41, 117. Peczuh, M. W.;
Hamilton, A. D.; SanchezQuesada, J.; deMendoza, J.; Haack, T.; Giralt,
E. J. Am. Chem. Soc. 1997, 119, 9327.
(8) Orner, B. P.; Ernst, J. T.; Hamilton, A. D. J. Am. Chem. Soc. 2001, 123,
5382.
R-Helices are the most common protein secondary structure,
accounting for over 40% of polypeptide amino acids in natural
proteins.3 The pharmaceutical potential for synthetic mimics of
R-helices is immense because R-helices play pivotal roles in
† Yale University.
‡ University of South Florida.
(1) Stites, W. E. Chem. ReV. 1997, 97, 1233. Peczuh, M. W.; Hamilton, A. D.
Chem. ReV. 2000, 100, 2479. Toogood, P. L. J. Med. Chem. 2002, 45,
1543. Yin, H.; Hamilton, A. D. Angew. Chem., Int. Ed. 2005, 44, 4130.
(2) Cochran, A. G. Chem. Biol. 2000, 7, R85.
(3) Ruan, F. Q.; Chen, Y. Q.; Hopkins, P. B. J. Am. Chem. Soc. 1990, 112,
9403.
(9) Ernst, J. T.; Kutzki, O.; Debnath, A. K.; Jiang, S.; Lu, H.; Hamilton, A. D.
Angew. Chem., Int. Ed. 2001, 41, 278. Yin, H.; Lee, G. I.; Park, H. S.;
Payne, G. A.; Rodriguez, J. M.; Sebti, S. M.; Hamilton, A. D. Angew.
Chem., Int. Ed. 2005, 44, 2704.
(10) Kutzki, O.; Park, H. S.; Ernst, J. T.; Orner, B. P.; Yin, H.; Hamilton, A. D.
J. Am. Chem. Soc. 2002, 124, 11838.
9
10.1021/ja050122x CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 10191-10196
10191

A R T I C L E S
Yin et al.
Figure 1. (A) Proteomimetics. (B) Overlay of a polyalanine R-helix and the energetically minimized conformation of 3,2′,2′′-trimethylterphenylene (1;
RMSD ) 0.851).
orientation of functionality on the surface of an R-helix. We
have reported functionalized terphenyls as mimetics of the
R-helical region of the Bak BH3 domain involved in apoptosis.10
In a further extension of this approach, we have developed other
low-molecular-weight helical mimetics, such as terephthalamide
and oligoamide-foldamer derivatives, as inhibitors of the Bcl-
xL/Bak interaction.11
Apoptosis (programmed cell death) is essential in many
physiological processes.12 The B-cell lymphoma-2 (Bcl-2)
protein family, composed of pro-apoptotic (Bak, Bax, Bad, Bid)
and antiapoptotic subfamilies (Bcl-2, Bcl-xL), are critical
regulators of apoptosis.13,14 The sensitivity of cells to apoptotic
stimuli depends on the balance of pro- and antiapoptotic Bcl-2
proteins. The heterodimerization of anti- and proapoptotic Bcl-2
proteins leads to the formation of pores in the mitochondrial
membrane, resulting in the release of cytochrome c, which in
turn activates the caspase cascade.15 Previous studies have shown
that overexpressed Bcl-2 or Bcl-xL can block the apoptotic
pathway and impede the functions of many anticancer agents.16
Mimicking the function of the proapoptotic Bcl-2 proteins could
lead to a new class of anticancer agents and thus is of immense
interest in medicinal chemistry. Recently, the groups of Schep-
artz and Verdine have successfully developed constrained
peptides that mimic the R-helical region of the death-promoting
Bak BH3 domain;17 the rational design of low-molecular-weight
inhibitors that recognize the Bak-binding region on Bcl-xL with
high specificity remains a challenging goal.
As an extension of our previous study,10 we herein report a
group of Bcl-xL antagonists, based on the terphenyl scaffold,
designed to mimic the Bak BH3 peptide. Using a fluorescence
polarization assay, we have observed high in vitro inhibitory
potencies in disrupting the Bcl-xL/Bak complexation. NMR
spectroscopic analyses and computer-aided docking simulations
suggested that the helix binding area on the surface of Bcl-xL
is the target for the synthetic ligands. Cellular assays demon-
strated that the terphenyl derivatives are effective in disrupting
the Bcl-xL/Bax interaction in intact cells.
Our design is based on an extended 3,2′,2′′-trisubstituted
terphenylene scaffold in which the para-coupling of a series of
ortho-substituted aryl subunits results in molecules that mimic
the functionality at i, i + 4, and i + 7 positions of an R-helix
(Figure 1A). Although there is some flexibility due to aryl-
aryl bond rotation, the scaffold balances this with a rigid core
structure allowing some optimization of binding interactions due
to induction of fit. Energy minimization of 3,2′,2′′-trimethylter-
phenylene (1) showed that the terphenyl scaffold takes up a
staggered conformation, with the projected ortho-substituents
in a good agreement with the targeted amino acid side chains
at the i, i + 4, and i + 7 positions of an R-helix (Figure 1B).18
(11) Yin, H.; Hamilton, A. D. Bioorg. Med. Chem. Lett. 2004, 14, 1375. Yin,
H.; Lee, G. I.; Sedey, K. A.; Rodriguez, J. M.; Wang, H. G.; Sebti, S. M.;
Hamilton, A. D. J. Am. Chem. Soc. 2005, 127, 5463. Ernst, J. T.; Becerril,
J.; Park, H. S.; Yin, H.; Hamilton, A. D. Angew. Chem., Int. Ed. 2003, 42,
535.
(12) Hengartner, M. O. Nature 2000, 407, 770.
(13) Adams, J. M.; Cory, S. Science 1998, 281, 1322. Reed, J. C. Nature 1997,
387, 773. Adams, J. M.; Cory, S. Trends Biochem. Sci. 2001, 26, 61. Adams,
J. M.; Cory, S. Curr. Opin. Cell Biol. 2002, 14, 715.
(17) Rutledge, S. E.; Chin, J. W.; Schepartz, A. Curr. Opin. Chem. Biol. 2002,
6, 479. Chin, J. W.; Schepartz, A. Angew. Chem., Int. Ed. 2001, 40, 3806.
Walensky, L. D.; Kung, A. L.; Escher, I.; Malia, T. J.; Barbuto, S.; Wright,
R. D.; Wagner, G.; Verdine, G. L.; Korsmeyer, S. J. Science 2004, 305,
1466.
(18) Energy minimization was carried out using MM2 force field within the
Macromodel 7.0 program.
(14) Graeber, T. G.; Osmanian, C.; Jacks, T.; Housman, D. E.; Koch, C. J.;
Lowe, S. W.; Giaccia, A. J. Nature 1996, 379, 88. Fearon, E. R.; Vogelstein,
B. Cell 1990, 61, 759.
(15) Chao, D. T.; Korsmeyer, S. J. Annu. ReV. Immunol. 1998, 16, 395.
(16) Strasser, A.; Huang, D. C. S.; Vaux, D. L. Biochim. Biophys. Acta 1997,
1333, F151.
9
10192 J. AM. CHEM. SOC. VOL. 127, NO. 29, 2005

Terphenyl-Based Bak-BH3
R
-Helical Proteomimetics
A R T I C L E S
propionitrile 5. The phenol group was deprotected using BBr3
and activated to triflate 7, which in turn underwent a Suzuki
cross-coupling reaction with 2-(4-methoxy-3-naphthalen-1-
ylmethyl-phenyl)-4,4,5,5-tetramethyl[1,3,2]dioxaborolane.
Deprotection and triflation of the resulting biphenyl 8 was
followed by a second Suzuki coupling, giving the terphenyl 11.
The final steps involved deprotection and alkylation of the
hydroxyl group with chloroacetonitrile, affording (cyanomethoxy)-
terphenyl 13. Hydrolysis of the terminal cyano groups produced
14 with two carboxyl groups on both ends of the terphenyl
backbone. The resulting acid groups were converted into their
corresponding ammonium salts to increase the solubility of the
terphenyl derivatives in aqueous solution.
Crystallographic analysis has confirmed that in the solid state
the terphenyl derivatives take up a staggered conformation
similar to that in Figure 1B.8 Figure 3 shows that 3,2′,2′′-
trimethyl-4-nitro-4′′-hydroxyterphenyl (2) projects all three
ortho-side chains on the same face. The distances between these
side chains are 5.10, 6.28, and 8.83 Å, which compare to those
of the side chains at the i, i + 4, and i + 7 positions in an
R-helical peptide (5.6, 6.6, and 10.1 Å, respectively). This solid-
state structure shows that terphenyl derivatives of this type can
adopt a conformation that reproduces the spatial arrangement
of the side chains in R-helices. However, the aryl-aryl bond
rotational barrier is low, suggesting that minor conformational
changes can occur to optimize contact to the helix-binding
domain of a protein target.22
In Vitro Evaluation. The inhibitory effects of these terphenyl
derivatives on the Bcl-xL/Bak complex were evaluated using a
previously reported fluorescence polarization assay.23 The Bak
BH3 peptide was labeled with fluorescein as the probe to
monitor the competitive binding of the terphenyl compounds
to Bcl-xL. Regression analysis was conducted to determine the
IC50 values, which in turn can be related to the known affinity
of the 16-mer Bak BH3 peptide to acquire the inhibitory constant
(Ki) of the inhibitors.10,24 To test the validity of this assay, we
used a nonlabeled Bak BH3 peptide as the competitive inhibitor
to bind Bcl-xL, giving Ki ) 0.122 µM, which closely corre-
sponds to the Kd value (0.120 µM) obtained from a saturation
titration experiment.
Structure-activity studies have shown that the terphenyl
derivatives recognize the Bak-binding site through specific
binding. Terphenyl derivative 14, with an isobutyl, 1-naphth-
ylmethyl, isobutyl side chain sequence, was identified as a potent
inhibitor of Bcl-xL with Ki ) 0.114 µM (Table 1). Analogues
15 and 16, which contain smaller side chains, showed lower
activities (Ki ) 2.09 and 1.82 µM, respectively), indicating that
hydrophobic surface area is critical for binding to the helix-
binding cleft on Bcl-xL. Compound 19, an isomer with a
different arrangement of the same side chains as 14, is
significantly less potent, confirming the importance of effective
shape complementarity. Nonspecific binding by the terphenyl
backbone does not appear to be significant since 22 lacking
the interacting side chains shows only weak binding to Bcl-xL.
We have further studied the role of the carboxyl groups by
converting them to ammonium groups (24), which gave lowered
Figure 2. Interface of the Bcl-xL/Bak BH3 domain complex. Hydrophobic
residues (V74, L78, I81) and D83 of the Bak BH3 domain play critical
roles in the binding.
The distances between the methyl groups are 5.2 (3, 2′), 6.1
(2′, 2′′), and 9.0 (3, 2′′) Å, closely corresponding to the side
chains at i, i + 4, and i + 7 positions in an R-helix (5.6, 6.6,
and 10.1 Å, respectively). Aryl-aryl torsion angles of 56.0°
(A-B) and 55.9° (B-C) give a conformation with close corre-
spondence of the positions of the three ortho-substituents and
the i, i + 4, and i + 7 side chains in an R-helix. Computational
modeling shows the superimposition of 1 on the i, i + 4, and
i + 7 side chains of a polyalanine R-helix with a root-mean-
square deviation (RMSD) value of 0.851 Å, suggesting good
stereochemical similarity between the pair (Figure 1B).
The NMR-derived structure of the Bcl-xL/Bak BH3 domain
complex indicated that the Bak peptide is an amphipathic R-helix
that interacts with Bcl-xL by projecting its side chains of V74,
L78, and I81, on one face of the helical backbone, into a hydro-
phobic cleft of Bcl-xL (Figure 2).19 In addition, Bak D83 forms
an ion pair with a lysine residue of Bcl-xL. A 26-mer peptide,
derived from the related protein Bad, binds even better to Bcl-
xL,20,21 exploiting larger hydrophobic residues (Y, F) to induce
a slight structural change in the binding region of Bcl-xL. Fur-
thermore, it has been shown that the R-helical propensity of
these peptides is decisive for strong binding to Bcl-xL.20 On
the basis of these structural requirements, we designed a series
of terphenyl molecules containing alkyl or aryl substituents on
the three ortho-positions to mimic the key hydrophobic substit-
uents on the helical exterior of Bak and carboxylic acid sub-
stituents on one or both ends to mimic the additional ion pair.
Results and Discussion
Synthesis and Crystallographic Analysis. A flexible and
modular synthesis was developed using O-alkylated phenols as
building blocks. The terphenyl backbone was synthesized by
sequential Suzuki aryl-aryl cross-coupling of the corresponding
p-methoxyphenylboronate and phenyltriflate derivatives. Car-
boxyl groups were attached on one or both ends of the terphenyl
backbone in order to mimic the D83 residues in the Bak BH3
peptide. Scheme 1 shows a modular synthesis of terphenyl
derivative 14. A 2-cyanoethyl substituent was introduced at the
4-position of 4-iodo-2-isobutyl-1-anisole (3) using a Heck reac-
tion, followed by the reduction of the resulting olefin to generate
(22) Carreira, L. A.; Towns, T. G. J. Mol. Struct. 1977, 41, 1.
(23) Wang, J. L.; Liu, D. X.; Zhang, Z. J.; Shan, S. M.; Han, X. B.; Srinivasula,
S. M.; Croce, C. M.; Alnemri, E. S.; Huang, Z. W. Proc. Natl. Acad. Sci.
U.S.A. 2000, 97, 7124.
(19) Muchmore, S. W.; et al. Nature 1996, 381, 335.
(20) Petros, A. M.; et al. Protein Sci. 2000, 9, 2528.
(21) Kelekar, A.; Chang, B. S.; Harlan, J. E.; Fesik, S. W.; Thompson, C. B.
Mol. Cell. Biol. 1997, 17, 7040.
(24) Cheng, Y.; Prusoff, W. H. Biochem. Pharmacol. 1973, 22, 3099.
9
J. AM. CHEM. SOC. VOL. 127, NO. 29, 2005 10193

A R T I C L E S
Yin et al.
Scheme 1. Modular Synthesis of Terphenyl Derivative 14a
a (a) Acrylonitrile, Pd(OAc)2, Bu4NCl, KOAc, DMF, 63%. (b) Mg, MeOH, 100%. (c) BBr3, CH2Cl2, 98%. (d) Tf2O, DIEA, CH2Cl2, 90%. (e) 2-(4-
Methoxy-3-naphthalen-1-ylmethyl-phenyl)-4,4,5,5-tetramethyl[1,3,2]dioxaborolane, Pd(PPh3)4, Na2CO3, DMF, 78%. (f) BBr3, CH2Cl2, 85%. (g) Tf2O, DIEA,
CH2Cl2, 82%. (h) 2-(3-Isobutyl-4-methoxyphenyl)-4,4,5,5-tetramethyl[1,3,2]dioxaborolane, Pd(PPh3)4, Cs2CO3, DMF, 47%. (i) BBr3, CH2Cl2, 76%. (j)
ClCH2CN, K2CO3, DMF, 96%. (k) NaOH, Bu4NOH(aq.), 1,4-dioxane, 100%.
the Bak D83 and interacts with a Bcl-xL lysine residue,
enhancing the binding affinity. Compound 21 with a third
carboxylic acid on the middle phenyl ring did not show strong
potency in disrupting the Bcl-xL/Bak binding, ruling out the
possibility that the terphenyl compounds recognize the protein
by nonspecific interaction with the positive charges on the Bcl-
xL exterior surface.
A key issue is whether the terphenyl scaffold could demon-
strate selectivity among different helix-binding proteins. Schep-
artz et al. have shown that aPP-templated miniature proteins
can selectively bind to Bcl-xL or Bcl-2, two paralog proteins.25
To this end, we have determined the ability of the terphenyl
derivatives to disrupt the binding of the N-terminal R-helix of
the tumor suppressor p53 to HDM2, as well as the binding of
Bak BH3 to Bcl-2, a Bcl-xL homolog. Terphenyl 14 selectively
binds to Bcl-xL and Bcl-2 over HDM2 (225- and 212-fold,
Figure 3. Stereoview of the X-ray crystal structure of 3,2′,2′′-trimethyl-
respectively), while its isomer 17, with a 2-naphthylmethyl side
4-nitro-4′′-hydroxyterphenylene (2).
chain on the middle ring, is more potent in disrupting the p53/
activity (Ki ) 13.6 µM), or to methyl esters (23), which showed
poor solubility in aqueous solution. Terphenyl derivatives with
a carboxyl group at only one end of the backbone, such as 25-
30, showed weaker binding affinity compared to the terphenyl
derivatives with two flanking acid functionalities at each end
of the scaffold. The second carboxyl group possibly mimics
HDM2 interaction under the same assay conditions (Figure 4).
This selectivity may provide unique applications in the recogni-
tion of a specific protein in the presence of other structurally
similar targets in a complex cellular environment.
(25) Gemperli, A. C.; Rutledge, S. E.; Maranda, A.; Schepartz, A. J. Am. Chem.
Soc. 2005, 127, 1596.
9
10194 J. AM. CHEM. SOC. VOL. 127, NO. 29, 2005

Terphenyl-Based Bak-BH3
R
-Helical Proteomimetics
A R T I C L E S
Table 1. Results of the Fluorescence Polarization Assay using Fluorescein-Labeled Bak (BH3 domain 72-86) Binding to Bcl-xL
and F146 (shown in magenta in Figure 5) showed significant
chemical shift changes upon the addition of the synthetic
inhibitor 14. Some other residues, including G94, L112, S122,
G134, K157, E158, M159 (shown in yellow in Figure 5) showed
moderate chemical shift changes under the same conditions.
These affected residues all lie near the shallow cleft on the
protein surface into which the Bak BH3 helix binds. The targeted
residues V74, L78, and I81 of Bak BH3 are within 4 Å distance
of residues F97, R102, L108, L130, I140, A142, and F146 of
Bcl-xL, most of which showed significant chemical shift changes
(F97 overlapped with N5), confirming that 14 and Bak BH3
target the same area on the exterior surface of Bcl-xL. Overlay
of 14 and the Bak BH3 peptide (Figure 5B) suggested that the
terphenyl indeed adopts a staggered conformation, mimicking
the cylindrical shape of the helix with the substituents making
a series of hydrophobic contacts with the protein surface.
R-Helical domains achieve high selectivity for their protein
targets through the precise projection of recognition groups from
their surfaces. Comparison of different terphenyl derivatives (14,
16-18) has revealed a good correlation between the terphenyl/
Bcl-xL contacts and their binding affinities. Table 2 lists the
amino acid residues affected within the Bak BH3 binding cleft
on the Bcl-xL surface upon the ligand binding. Terphenyl 14
affected more residues in the Bak BH3 binding pocket,
suggesting that its side chains recognize the target pockets with
good spatial complementarity. In contrast, 17 induced fewer
residues showing chemical shift changes, indicating that the
weaker affinity of 17 is due to a less effective fit.
Figure 4. Comparison of terphenyl derivatives that bind to different helix-
binding proteins.
Binding Mode Studies. A computational docking simulation
of 14 using the Autodock 3.0 program suggested that the binding
cleft for the Bak BH3 domain on the surface of Bcl-xL is the
target area for the synthetic inhibitors (Figure 5A).26 The top-
ranked binding modes all lie within the Bak BH3 binding pocket.
15N-HSQC experiments lent further support to the proposed
binding site of the terphenyl derivatives in the Bak BH3 binding
region. The residues of A89, L99, L108, T109, S110, Q111,
I114, Q125, L130, F131, W137, G138, R139, I140, A142, S145,
Inhibition of the Bcl-xL/Bak Association in Intact Human
Cells. To determine if terphenyl derivatives inhibit the BH3-
mediated interaction with Bcl-xL in intact cells, HEK293 cells
transfected with both HA-Bcl-xL and flag-Bax (a close family
member of Bak) were treated with 100 µM of the indicated
(26) Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart, W. E.;
Belew, R. K.; Olson, A. J. J. Comput. Chem. 1998, 19, 1639. Morris, G.
M.; Goodsell, D. S.; Huey, R.; Olson, A. J. J. Comput.-Aid. Mol. Des.
1996, 10, 293. Goodsell, D. S.; Olson, A. J. Proteins: Struct., Funct., Genet.
1990, 8, 195.
9
J. AM. CHEM. SOC. VOL. 127, NO. 29, 2005 10195

A R T I C L E S
Yin et al.
Figure 5. Results of the 15N-HSQC and the molecular-docking experiments of 14 binding to Bcl-xL. The residues that showed significant and moderate
chemical shift changes upon the addition of 14 are shown in magenta and yellow, respectively. (A) The three top-ranked binding modes of 14 are shown
in blue (rank 1), orange (rank 2), and cyan (rank 3). (B) Overlay of the top-ranked docking result of 14 and the Bak BH3 peptide (blue). The key hydrophobic
side chains of V74, L78, and I81 are shown in stick representations.
Table 2. Comparison of the Amino Acid Residues Affected by Different Terphenyl Derivatives within the Bak BH3 Binding Pocket in
15N-HSQC Experiments
ligand
significantly shifted residuesa
moderately shifted residuesa
Bak peptide
14
S106, L112, A142, A148, A199. E202
A89, L99, L108, T109, S110, Q111, I114, Q125, L130, F131, W137,
R100, Y101, S122, R132, D133, V192, L194, A100
G94, L112, S122, G134, K157, E158, M159
G138, R139, I140, A142, S145, F146
16
17
18
R100, F105, L108, L112, I114, Y120, L194
R100, L112
L112, I114, Y120, S122, F143, L194, Y195, A200
R100, Q121
D95, Q121, Q125
Q121, V192
max
max
b 0.5∆ωNH
max
a ∆ωNH > 0.5∆ωNH
.
> ∆ωNH > 0.3∆ωNH
.
69.8 ( 13.4% inhibition of Bcl-xL/Bax association, which was
similar to the Bax BH3 peptide, perhaps due to improved
membrane permeability in 26. Consistent with this, computa-
tional analysis of Caco-2 permeability suggested that 26 has
better membrane penetration (327 nm/s) than 14 (14.5 nm/s).27
In conclusion, a strategy of helix mimicry based on a
substituted terphenyl scaffold was successfully applied to the
design of a Bcl-xL antagonist. Terphenyl derivative 14 exhibited
binding affinity in the nanomolar range as seen in the fluores-
cence polarization assay. The binding modes of the terphenyl
derivatives were confirmed by means of NMR spectroscopy and
computational docking simulations. Co-immunoprecipitation
experiments showed that terphenyl derivatives are effective in
disrupting the Bcl-xL/Bax interaction in intact cells. We have
also shown that the terphenyl serves as a general scaffold to
mimic R-helical segments with good selectivity in recognition
of different helix-binding proteins.
Figure 6. Inhibition of the Bcl-xL/Bax association in whole cells.
terphenyl derivatives or control Bax BH3 peptide. After 24 h
incubation, the cells were harvested, lysed, and subjected to
immunoprecipitation with anti-HA antibody. The resulting
mixture was loaded onto a 12.5% SDS-PAGE gel, and proteins
transferred to nitrocellulose for Western blot analysis. The
presence of flag-Bax protein in the HA-Bcl-xL immune com-
plexes was detected with antiflag antibody. The inhibitory
potencies of the terphenyl compounds were determined by
measuring the relative intensity of flag-Bax protein co-immu-
noprecipitated with HA-Bcl-xL. As shown in Figure 6, the
terphenyl derivative 14 blocked the Bcl-xL/Bax complexation
by 50.9 ( 34.5%. Terphenyl derivative 26, with isopropyl,
methyl, methyl side chains, showed the strongest effect with
Acknowledgment. We thank the NIH (Grant GM69850) for
financial support of this work and Deutsche Forschungsgemein-
schaft (DFG) for a fellowship to O.K.
Supporting Information Available: Experimental section and
complete refs 19 and 20. This material is available free of charge
via the Internet at http://pubs.acs.org.
JA050122X
(27) The permeablility prediction analysis was carried out using: QikProp V2.1;
Schrodinger Inc.: New York, 2003.
9
10196 J. AM. CHEM. SOC. VOL. 127, NO. 29, 2005