Design, synthesis, and characterization of a potent, nonpeptide, cell-permeable, bivalent smac mimetic that concurrently targets both the BIR2 and BIR3 domains in XIAP

Published on Web 11/14/2007

Design, Synthesis, and Characterization of a Potent,

Nonpeptide, Cell-Permeable, Bivalent Smac Mimetic That

Concurrently Targets Both the BIR2 and BIR3 Domains in

XIAP

Haiying Sun,⁺Zaneta Nikolovska-Coleska,⁺Jianfeng Lu,⁺Jennifer L. Meagher,^‡

Chao-Yie Yang,⁺Su Qiu,⁺York Tomita,^|Yumi Ueda,^|Sheng Jiang,^#

Krzysztof Krajewski,^#Peter P. Roller,^#Jeanne A. Stuckey,^†,‡and

Shaomeng Wang*^,+

Contribution from Departments of Internal Medicine, Pharmacology and Medicinal Chemistry

and ComprehensiVe Cancer Center, Life Sciences Institute, and Biological Chemistry, Biophysics

Research DiVision, UniVersity of Michigan, 1500 E. Medical Center DriVe, Ann Arbor, Michigan

48109, Lombardi Cancer Center, Georgetown UniVersity Medical Center, Washington DC

20007, and Laboratory of Medicinal Chemistry, National Cancer InstitutesFrederick, NIH,

Frederick, Maryland 21702

Received July 16, 2007; E-mail: shaomeng@med.umich.edu

Abstract: XIAP is a central apoptosis regulator that inhibits apoptosis by binding to and inhibiting the effectors

caspase-3/-7 and an initiator caspase-9 through its BIR2 and BIR3 domains, respectively. Smac protein in

its dimeric form effectively antagonizes XIAP by concurrently targeting both its BIR2 and BIR3 domains.

We report the design, synthesis, and characterization of a nonpeptide, cell-permeable, bivalent small-

molecule (SM-164) which mimics Smac protein for targeting XIAP. Our study shows that SM-164 binds to

XIAP containing both BIR domains with an IC₅₀value of 1.39 nM, being 300 and 7000 times more potent

than its monovalent counterparts and the natural Smac AVPI peptide, respectively. SM-164 concurrently

interacts with both BIR domains in XIAP and functions as an ultrapotent antagonist of XIAP in both cell-

free functional and cell-based assays. SM-164 targets cellular XIAP and effectively induces apoptosis at

concentrations as low as 1 nM in the HL-60 leukemia cell line. The potency of bivalent SM-164 in binding,

functional, and cellular assays is 2-3 orders of magnitude higher than its corresponding monovalent Smac

mimetics.

apoptosis.^4,5XIAP potently inhibits apoptosis by directly binding

to and effectively inhibiting three members of caspases, the two

Introduction

Apoptosis, or programmed cell death, is a critical cell process

in normal development and homeostasis of multicellular organ-

isms. Inappropriate regulation of apoptosis has been implicated

in many human diseases, including cancer.^1-3It is now

recognized that dysfunction of the apoptosis machinery is a

hallmark of cancer.²Accordingly, targeting critical apoptosis

regulators is an attractive approach for the development of new

classes of therapies for the treatment of cancer and other human

diseases.^1-3

effectors, caspase-3 and -7, and an initiator, caspase-9 (Figure

1A), whose activity is critical for execution of apoptosis. XIAP

contains three baculovirus IAP repeat (BIR) domains. The third

BIR domain (BIR3) of XIAP selectively targets caspase-9,

whereas the BIR2 domain, together with the linker preceding

BIR2, inhibits both caspase-3 and caspase-7 (Figure 1B).⁵

Consistent with its potent apoptosis-suppressing function, XIAP

is found to be highly expressed in many human tumor cell lines

and tumor samples from patients⁶and plays an important role

in conferring resistance on cancer cells to a variety of anticancer

drugs.⁷

The X-linked inhibitor of apoptosis protein (XIAP) is a

member of IAP proteins and a central apoptosis regulator,

although its role may not be limited to the regulation of

In cells, the antiapoptotic function of XIAP is antagonized

by Smac/DIABLO (second mitochondria-derived activator of

caspases or direct IAP binding protein with low pI) (Figure

1A).^8,9Smac/DIABLO has been identified as a protein released

⁺Departments of Internal Medicine, Pharmacology and Medicinal

Chemistry and Comprehensive Cancer Center, University of Michigan.

^‡Life Sciences Institute, University of Michigan.

^†Biological Chemistry, Biophysics Research Division, University of

Michigan.

^|Georgetown University Medical Center.

(5) Salvesen, G. S.; Duckett, C. S. Nat. ReV. Mol. Cell Biol. 2002, 3, 401-

410.

^#National Cancer Institute-Frederick.

(1) Ponder, B. A. Nature 2001, 411, 336-341.

(6) Tamm, I.; Kornblau, S. M.; Segall, H.; Krajewski, S.; Welsh, K.; Kitada,

S.; Scudiero, D. A.; Tudor, G.; Qui, Y. H.; Monks, A.; Andreeff, M.; Reed,

J. C. Clin. Cancer Res. 2000, 6, 1796-1803.

(2) Lowe, S. W.; Lin, A. W. Carcinogenesis 2000, 21, 485-495.

(3) Nicholson, D. W. Nature 2000, 407, 810-816.

(4) Deveraux Q. L.; Reed, J. C. Genes DeV. 13, 239-252.

(7) Holcik, M.; Gibson, H.; Korneluk, R. G. Apoptosis 2001, 6, 253-261.

9

J. AM. CHEM. SOC. 2007, 129, 15279-15294

15279

A R T I C L E S

Sun et al.

show that the linker preceding BIR2 in XIAP binds to caspase-

3/-7 and is the major energetic determinant, while the BIR2

domain itself plays a regulatory role. Modeling suggests that

Smac protein also binds to XIAP BIR2 through its AVPI motif

and prevents the binding of XIAP to caspase-3/-7.^16-19In this

manner, dimeric Smac protein effectively removes the inhibition

of XIAP to caspase-9 and to caspase-3/-7 by concurrently

targeting both the BIR2 and BIR3 domains in XIAP (Figure

1B).¹⁹

Because XIAP blocks apoptosis at the downstream effector

phase, a point where multiple signaling pathways converge, it

represents a particularly attractive molecular target for the design

of new classes of anticancer drugs aimed at overcoming the

apoptosis resistance of cancer cells.⁷To date, a number of

research laboratories, including ours, have pursued the design

of small-molecule antagonists of XIAP.^20-28One approach

focuses on the design of small molecules that target the XIAP

BIR3 domain, antagonizing the inhibition of XIAP to caspase-

9.^20-25These efforts have so far yielded small molecules that

bind to the XIAP BIR3 domain with high affinities. A number

of these compounds are effective in inhibition of cell growth

and induction of apoptosis in cancer cells.^20-25Another approach

is to design small-molecule inhibitors that target the BIR2

domain, blocking the interaction of XIAP with caspase-3/-7.^26,27

For example, polyurea compounds that target the caspase-3/

XIAP interaction have been shown to achieve broad anticancer

activity and are synergistic when used in combination with

chemotherapeutic agents.²⁷Collectively, these studies have

provided convincing evidence that targeting XIAP is an effective

strategy to overcome the apoptosis resistance of cancer cells.

Since both the BIR2 and BIR3 domains in XIAP effectively

suppress apoptosis by targeting caspase-3/-7 and caspase-9,

respectively, we reason that bivalent small molecules which

concurrently interact with both BIR domains could be particu-

larly efficient and potent XIAP antagonists (Figure 1B). Indeed,

a small-molecule mimic containing two Smac AVPI binding

motifs was shown to antagonize XIAP with a potency exceeding

that of Smac protein in a cell-free functional assay and to

enhance apoptosis induction of other therapeutic agents in cancer

cells.²⁸It was proposed²⁸that its high potency as an XIAP

antagonist is attributable to its bivalency, but its precise mode

of action in targeting XIAP remains unclear.

Figure 1. (A) A schematic, simplified apoptosis pathway. XIAP inhibits

apoptosis by directly binding to and inhibition of caspase-9 and caspase-3

and -7. Smac protein binds to XIAP and antagonizes XIAP to promote

activation of caspases and apoptosis. (B) Design of bivalent Smac mimetics

to target both the BIR2 and BIR3 domains of XIAP by mimicking the

binding of dimeric Smac protein.

from mitochondria into the cytosol in response to apoptotic

stimuli.^8,9It forms an elongated dimer¹⁰and targets both the

BIR2 and BIR3 domains in XIAP (Figure 1B).¹¹It removes

the XIAP inhibition of caspase-9 by binding to the BIR3 domain

in XIAP through its AVPI tetrapeptide binding motif and

directly competing with a similar ATPF tetrapeptide in the

processed caspase-9 (Figure 1B).^11-15Conclusive understanding

of the mechanism by which Smac removes the inhibition of

XIAP to caspase-3/-7 remains elusive.¹¹Crystal structures^16-18

In this study, we present structure-based design, synthesis,

and detailed characterization of a novel, nonpeptidic, bivalent

(18) Riedl, S. J.; Renatus, M.; Schwarzenbacher, R.; Zhou, Q.; Sun, C.; Fesik,

S. W.; Liddington, R. C.; Salvesen, G. S. Cell 2001, 104, 791-800.

(19) Huang, Y.; Rich, R. L.; Myszka, D. G.; Wu, H. J. Biol. Chem. 2003, 278,

49517-49522.

(8) Du, C.; Fang, M.; Li, Y.; Wang, X. Cell 2000, 102, 33-42.

(9) Verhagen, A. M.; Ekert, P. G.; Pakusch, M.; Silke, J.; Connolly, L. M.;

Reid, G. E.; Moritz, R. L.; Simpson, R. J.; Vaux, D. L. Cell 2000, 102,

43-53.

(20) Fulda, S.; Wick, W.; Weller, M.; Debatin, K. M. Nature Med. 2002, 8,

808-815.

(21) Sun, H.; Nikolovska-Coleska, Z.; Yang, C.-Y.; Xu, L.; Liu, M.; Tomita,

Y.; Pan, H.; Yoshioka, Y.; Krajewski, K.; Roller, P. P.; Wang, S. J. Am.

Chem. Soc. 2004, 126, 16686-16687.

(10) Chai, J.; Du, C.; Wu, J. W.; Kyin, S.; Wang, X.; Shi, Y. Nature 2000, 406,

855-62.

(11) Shiozaki, E. N.; Shi, Y. Trends Biochem. Sci. 2004, 29, 486-494.

(12) Wu, G.; Chai, J.; Suber, T. L.; Wu, J. W.; Du, C.; Wang, X.; Shi, Y. Nature

2000, 408, 1008-1012.

(22) Sun, H.; Nikolovska-Coleska, Z.; Yang, C.-Y.; Xu, L.; Tomita, Y.;

Krajewski, K.; Roller, P. P.; Wang, S. J. Med. Chem. 2004, 47, 4147-

4150.

(13) Liu, Z.; Sun, C.; Olejniczak, E. T.; Meadows, R. P.; Betz, S. F.; Oost, T.;

Herrmann, J.; Wu, J. C.; Fesik, S. W. Nature 2000, 408, 1004-1008.

(14) Srinivasula, S.; Hegde, R.; Saleh, A.; Datta, P.; Shiozaki, E.; Chai, J.; Lee,

R.-A.; Robbins, P. D.; Fernandes-Alnemri, T.; Shi, Y.; Alnemri, E. S.

Nature 2001, 410, 112-116.

(23) Sun, H.; Nikolovska-Coleska, Z.; Lu, J.; Qiu, S.; Yang, C.-Y.; Gao, W.;

Meagher, J.; Stuckey, J.; Wang, S. J. Med. Chem. 2006, 49, 7916-7920.

(24) Oost, T. K.; et al. J. Med. Chem. 2004, 47, 4417-4426.

(25) Zobel, K.; Wang, L.; Varfolomeev, E.; Franklin, M. C.; Elliott, L. O.;

Wallweber, H. J.; Okawa, D. C.; Flygare, J. A.; Vucic, D.; Fairbrother, W.

J.; Deshayes, K. ACS Chem. Biol. 2006, 1, 525-533.

(15) Shiozaki, E. N.; Chai, J.; Rigotti, D. J.; Riedl, S. J.; Li, P.; Srinivasula, S.

M.; Alnemri, E. S.; Fairman, R.; Shi, Y. Mol. Cell 2003, 11, 519-527.

(16) Chai, J.; Shiozaki, E.; Srinivasula, S. M.; Wu, Q.; Dataa, P.; Alnemri, E.

S.; Shi, Y. Cell 2001, 104, 769-780.

(26) Wu, T. Y.; Wagner, K. W.; Bursulaya, B.; Schultz, P. G.; Deveraux, Q. L.

Chem. Biol. 2003, 10, 759-67.

(27) Schimmer, A. D.; et al. Cancer Cell 2004, 5, 25-35.

(28) Li, L.; Thomas, R. M.; Suzuki, H.; De Brabander, J. K.; Wang, X, Harran,

P. G. Science 2004, 305, 1471-4.

(17) Huang, Y.; Park, Y. C.; Rich, R. L.; Segal, D.; Myszka, D. G.; Wu, H.

Cell 2001, 104, 781-790.

9

15280 J. AM. CHEM. SOC. VOL. 129, NO. 49, 2007

Bivalent Smac Mimetic as a Potent Antagonist of XIAP

A R T I C L E S

Figure 2. Design of a potent, conformationally constrained, monovalent Smac mimetic SM-122 and its biotinylated analogue.

small molecule (SM-164) using a series of complementary

biochemical, functional, and cellular assays. We demonstrate

that SM-164 concurrently targets the BIR2 and BIR3 domains

in the same XIAP molecule, achieves an extremely high binding

affinity to XIAP, and functions as an ultrapotent antagonist of

XIAP in cell-free functional and cellular assays. It binds potently

to cellular XIAP and effectively induces apoptosis in leukemia

cancer cells at concentrations as low as 1 nM. Its potency in

binding, functional and cellular assays is 2-3 orders of

magnitude higher than its corresponding monovalent Smac

mimetics.

Figure 3. Competitive binding of our designed monovalent Smac mimetics

1, 2, and 3 and the AVPI peptide to XIAP BIR3 (residues 240-356) as

determined using an FP-based assay.

Results

effect on the binding of compound 1 to XIAP BIR3 protein. In

comparison, the Smac AVPI peptide has an IC₅₀value of 1183

nM and its calculated K_ivalue is 425 nM, similar to the reported

K_dvalue for this peptide.³⁰Therefore, compound 1 binds to

XIAP BIR3 with an affinity 13 times higher than that of the

Smac AVPI peptide.

Design of a Cell-Permeable Monovalent Smac Mimetic.

Smac protein binds to both the BIR2 and BIR3 in XIAP Via its

AVPI binding motif.¹¹It was shown¹³that short Smac-based

peptides containing the AVPI sequence bind to recombinant

XIAP BIR3 and BIR2 proteins with K_dvalues of 0.4-0.7 µM

and 6-9 µM, respectively. The Smac-based AVPI peptide thus

provides a suitable template for the design of bivalent small

molecules that can concurrently interact with both the BIR2

and BIR3 domains. Unfortunately, Smac-based peptides are not

cell-permeable. Thus, for the design of cell-permeable, bivalent

Smac mimetics, it was essential to first derive cell-permeable,

monovalent Smac mimetics that can bind to both XIAP BIR2

and BIR3 with good affinities.

We have employed a structure-based strategy to design potent,

cell-permeable, small molecules (monovalent Smac mimetics)

that mimic the Smac AVPI binding motif, although our initial

purpose was to target the XIAP BIR3 domain.^21-23Using this

approach, we have designed compound 1 (SM-122) as a

conformationally constrained, Smac AVPI mimetic containing

a [8.5] bicyclic system (Figure 2) and evaluated its binding

affinities to XIAP BIR2 and BIR3 proteins. To facilitate the

investigation of the molecular mechanism of action for Smac

mimetics, we have also designed a biotinylated analogue of SM-

122 (2, named BL-SM-122, Figure 2).

We next evaluated the binding affinity of SM-122 (1) to XIAP

BIR2. For this purpose, we attempted to develop an FP assay

for XIAP BIR2, similar to that for XIAP BIR3.²⁹Unfortunately,

the fluorescently labeled Smac-based peptide tracer we used

previously to establish the competitive binding FP assay for

XIAP BIR3 domain²⁹has a low binding affinity to XIAP BIR2

and is not ideal as a tracer for the development of an FP assay

for XIAP BIR2. To address this limitation, we employed the

Biacore surface plasmon resonance (SPR) technique to directly

evaluate the binding affinity of biotinylated compound 2 to

XIAP BIR2 protein with compound 2 immobilized on the

streptavidin chip. Our analysis showed that compound 2 binds

to XIAP BIR2 with a K_dvalue of 1.85 µM (K_on) 1118 1/ms

and K_off) 0.0018 1/s) (Figure 4A). We next evaluated

compound 1 for its binding to XIAP BIR2 in a competitive

SPR assay with biotinylated compound 2 immobilized on the

streptavidin chip and XIAP BIR2 protein and compound 1 in

the mobile phase. Our results showed that compound 1 blocks

the binding of XIAP BIR2 to compound 2 in a dose-dependent

manner and has an IC₅₀value of 5.6 µM (Figure 4B). Therefore,

our SPR experiments showed that both compounds 1 and 2 bind

to XIAP BIR2 with a good affinity.

The binding affinities of compounds 1 and 2 and the Smac

AVPI peptide to the XIAP BIR3 domain protein were evaluated

using a previously established sensitive fluorescence-polarization

(FP) assay (Figure 3).²⁹Compound 1 binds to XIAP BIR3

protein with an IC₅₀value of 91 nM. Its biotinylated analogue

2 has an IC₅₀value of 112 nM, very close to that of compound

1, indicating that the biotin label does not have a significant

Our FP and SPR assays showed that compound 1 binds to

XIAP BIR2 and BIR3 proteins with good affinities. Importantly,

as will be demonstrated later in the paper, compound 1

effectively inhibits cell growth and induces apoptosis in cancer

(29) Nikolovska-Coleska, Z.; Wang, R.; Fang, X.; Pan, H.; Tomita, Y.; Li, P.;

Roller, P. P.; Krajewski, K.; Saito, N. G.; Stuckeys, J. A.; Wang, S. Anal.

Biochem. 2004, 332, 261-73.

(30) Kipp, R. A.; Case, M. A.; Wist, A. D.; Cresson, C. M.; Carrell, M.; Griner,

E.; Wiita, A.; Albiniak, P. A.; Chai, J.; Shi, Y.; Semmelhack, M. F.;

McLendon, G. L. Biochemistry 2002, 41, 7344-7349.

9

J. AM. CHEM. SOC. VOL. 129, NO. 49, 2007 15281

A R T I C L E S

Sun et al.

Figure 4. (A) Analysis of the binding of biotin-labeled compound 2 to

XIAP L-BIR2 protein (residues 120-240) using the Biacore surface plasmon

resonance (SPR) technique with different concentrations of the protein (from

the top: 50, 25, 12.5, 6.25, and 3.125 µM). (B) Competitive binding of

compound 1 to XIAP L-BIR2 as determined using the Biacore SPR

technique. Biotinylated SM-122-BL was immobilized on the streptavidin

chip.

Figure 5. Predicted binding model of Smac mimetic 1 in complex with

(A) XIAP BIR2 domain and (B) BIR3 domain.

cells, indicative of its excellent cell permeability. Hence,

compound 1 represents a promising monovalent Smac mimetic

template for the design of bivalent Smac mimetics to target

concurrently the BIR2 and BIR3 domains.

Design of Bivalent Smac Mimetics. We next identified

suitable sites in compound 1 for chemically tethering two

molecules together for the design of bivalent Smac mimetics.

For this purpose, we employed computational modeling to

predict the binding of 1 to BIR2 and BIR3 domains of XIAP

using high-resolution crystal structures of these individual

domains.^12,18The predicted binding models (Figure 5) showed

that one phenyl group of 1 in the pro-(S) configuration does

not insert into the binding pocket of either BIR2 or BIR3 and

consequently is exposed, making this phenyl ring suitable as

an anchoring site for tethering two molecules of 1 together to

produce bivalent Smac mimetics.

Modeling also suggested that the pro-(S) phenyl group in 1

may be replaced by other aromatic rings without compromising

the binding. Thus, compound 3 was designed and synthesized

to test if the pro-(S) phenyl group can be replaced by a [1,2,3]-

triazole so that the “click chemistry”, a highly efficient coupling

method,³¹can be readily applied to facilitate the synthesis of

bivalent Smac mimetics (Figure 6). Consistent with our model-

ing prediction, compound 3 was determined to bind to XIAP

BIR3 with an IC₅₀value of 57 nM, essentially the same as that

of compound 1 (91 nM) (Figure 3). In addition, compound 3

was also shown to bind to XIAP BIR2 with the same affinity

as that of compound 1 in our SPR assay (data not shown). Thus,

compounds 1 and 3 have similar binding affinities to either BIR2

or BIR3 domain.

Figure 6. Design of a new class of bivalent Smac mimetics based upon a

conformationally constrained monovalent Smac mimetic.

between BIR2 and BIR3 containing approximately 25 residues

is quite flexible. Accordingly, we designed a bivalent Smac

mimetic SM-164 (4, Figure 6), which contains two monovalent

Smac mimetics tethered together through a flexible linker which,

when extended, provides a distance of approximately 15 Å

between the two triazole rings. We reasoned that such a linker

would provide sufficient length and flexibility for the two

monovalent binding units in 3 to bind concurrently to both the

BIR2 and BIR3 domains in the same XIAP molecule. On the

basis of our predicted binding models for compound 1 (Figure

5), the methyl group in 1 inserts into a small hydrophobic cavity

in both BIR2 and BIR3, and the N-methyl amino group

participates in a hydrogen-bonding network with protein residues

in both models. To confirm our predicted binding models and

to determine the specificity of bivalent Smac mimetic 4,

compound 5 was designed as an inactive control (Figure 6). In

The experimentally determined three-dimensional structure

of XIAP containing both BIR2 and BIR3 domains has not been

reported, but our modeling analysis showed that the linker region

(31) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V. and Sharpless, K. B. Angew.

Chem., Int. Ed. 2002, 41, 2596-2599.

9

15282 J. AM. CHEM. SOC. VOL. 129, NO. 49, 2007

Bivalent Smac Mimetic as a Potent Antagonist of XIAP

A R T I C L E S

Scheme 1. Synthesis of Compound 1

Scheme 3. Synthesis of Compounds 4-6

^aReagents and conditions: (a) i. 10% Pd-C, H₂, MeOH; ii. 2 N LiOH,

1.4-dioxane, then 1 N HCl, 96% over two steps; (b) i. aminodiphenyl-

methane, EDC, HOBt, N,N-diisopropylethylamine, CH₂Cl₂, overnight; ii.

4 N HCl in 1,4-dioxane, MeOH; iii. Boc-N-methyl-^L-alanine, EDC, HOBt,

N,N-diisopropylethylamine, CH₂Cl₂; iv. 4 N HCl in 1,4-dioxane, MeOH,

72% over four steps.

Scheme 2. Synthesis of Compound 3

^aReagents and conditions: (a) i. 14, EDC, HOBt, N,N-diisopropylethy-

lamine, CH₂Cl₂, overnight; ii. 4 N HCl in 1,4-dioxane, MeOH; iii. Boc-

N-methyl-^L-alanine, EDC, HOBt, NN,-diisopropylethylamine, CH₂Cl₂, 73%

over three steps; (b) i. 4 N HCl in 1,4-dioxane; ii. N-Boc-^L-Trp, EDC, HOBt,

N,N-diisopropylethylamine, CH₂Cl₂; iii. 4 N HCl in 1,4-dioxane, MeOH,

iv. acetic anhydride, N,N-diisopropylethylamine, Ch₂Cl₂, 67% over four

steps; (c) i. 1,4-bis-(4-azido-butyl)-benzene, CuSO₄, (+)-sodium-L-ascorbate,

tert-butanol-H₂O 2:1; ii. 4 N HCl in 1,4-dioxane, MeOH, 68% for 4 and

66% for 6; (d) 1,4-bis-(4-azido-butyl)-benzene, CuSO₄, (+)-sodium-L-

ascorate, tert-butanol-H₂O 2:1, 76%.

^aReagents and conditions: (a) 11, EDC, HOBt, N,N-diisopropylethy-

lamine, CH₂Cl₂, overnight, 91%; (b) i. 4 N HCl in 1,4-dioxane, MeOH; ii.

Boc-N-methyl-^L-alanine, EDC, HOBt, N,N-diisopropylethylamine, Ch₂Cl₂,

86% over two steps; (c) i. azidomethylbenzene, CuSO₄, (+)-sodium-L-

ascorbate, tert-butanol-H₂O 2:1; ii. 4 N HCl in 1,4-dioxane, MeOH, 74%

over two steps.

L-alanine to afford amide 13. Cycloaddition of 13 with azi-

domethylbenzene catalyzed by CuSO₄-(+)-sodium L-ascorbate

yielded a triazole.³¹Removal of the Boc protecting group from

this triazole furnished compound 3 (SM-157).

The synthesis of the bivalent compounds 4 (SM-164), 5 (SM-

173), and 6 (SM-206) is shown in Scheme 3. The chiral amine

14 was prepared according to a reported method.³³Compound

15 was synthesized with the same method as that used for

compound 13. Removal of the Boc protecting group in 12

followed by condensation of the resulting ammonium salt with

L-NR-Boc-Trp yielded an amide. Removal of the two Boc

protecting groups in this amide followed by condensation of

the resulting salt with acetic anhydride provided compound 16.

Cycloaddition of 2 equiv of compound 13 or 15 with 1 equiv

of 1,4-bis-(4-azidobutyl)benzene respectively under the catalysis

of CuSO₄-(+)-sodium L-ascorbate gave two bis-triazoles.³¹

Removal of the two Boc protecting groups in these bis-triazoles

yielded compounds 4 and 6, respectively. Cycloaddition of 2

equiv of compound 16 with 1 equiv of 1,4-bis-(4-azido-butyl)-

benzene catalyzed by CuSO₄-(+)-sodium L-ascorbate gave

compound 5.³¹

The synthesis of biotinylated compound 2 (BL-SM-122) is

shown in Scheme 4. The chiral amine 18 was synthesized

according to our reported method.³⁴Condensation of acid 9 with

18 followed by removal of the Boc protecting group and

condensation of the resulting ammonium salt with L-N-Boc-N-

methyl-alanine yielded an amide. Cleavage of the Cbz protecting

group in this amide by hydrogenation followed by condensation

of the resulting amine with Cbz-6-aminohexanoic acid provided

amide 19. Removal of the Cbz protecting group in this amide

(19) followed by condensation with (+)-biotin N-hydroxy-

5, the methyl group in each monovalent binding unit was

replaced by a (1H-indolyl)methyl group, a much larger hydro-

phobic group, to disrupt the hydrophobic interactions at this

site and the methylamino group was replaced by an acetamido

group to block the hydrogen bond formation. To test if the

stereospecificity is critical for binding, compound 6, a stereoi-

somer of 4, was designed and synthesized.

Chemistry. The synthesis of compound 1 is shown in Scheme

1. The key intermediate 8 was synthesized from pyroglutamic

acid 7 in eight steps according to a reported method.³²

Hydrogenation of the CdC double bond in 8 catalyzed by 10%

Pd-C, followed by hydrolysis of the methyl ester gave the acid

9. Condensation of 9 with aminodiphenylmethane followed by

removal of the Boc protecting group yielded an ammonium salt,

condensation of which with Boc-N-methyl-L-alanine furnished

an amide. Removal of the Boc protecting group from this amide

afforded compound 1 (SM-122).

The synthesis of SM-157 (3) is shown in Scheme 2. The chiral

amine 11 was prepared according to a reported method from

the alcohol 10 in five steps.³³Condensation of acid 9 with amine

11 gave an amide 12. After removal of the Boc protecting group,

the resulting ammonium salt was condensed with Boc-N-methyl-

(32) Duggan, H. M. E.; Hitchcock, P. B.; Young, D. W. Org. Biomol. Chem.

2005, 3, 2287-2295.

(33) Messina, F.; Botta, M.; Corelli, F.; Schneider, M. P.; Fazio, F. J. Org.

Chem. 1999, 64, 3767-3769.

9

J. AM. CHEM. SOC. VOL. 129, NO. 49, 2007 15283

A R T I C L E S

Sun et al.

Scheme 4. Synthesis of Compound 2 (BL-SM-122)

^aReagents and conditions: (a) i. 18, EDC, HOBt, N,N-diisopropylethylamine, CH₂Cl₂, overnight; ii. 4 N HCl in 1,4-dioxane, MeOH; iii. Boc-N-methyl-

L-alanine, EDC, HOBt, N,N-diisopropylethylamine, CH₂Cl₂; iv. H₂, 10% Pd-C, MeOH; v. Cbz-6-aminohexanoic acid, EDC, HOBt, N,N-diisopropylethylamine,

CH₂Cl₂, 42% over five stesp; (b) i. H₂, 10% Pd-C, MeOH; ii. (+)-biotin N-hydroxyl-succinimide ester, N,N-diisopropylethylamine, Ch₂Cl₂; iii. 4 N HCl in

1,4-dioxane, MeOH, 74% over three steps.

Figure 8. Competitive binding of Smac mimetics to XIAP protein (residues

120-356) containing BIR2 and BIR3 domains determined in a competitive

FP-based assay.

Figure 7. (A) Chemical structure of Smac-1F, a bivalent Smac-based

peptide labeled with a fluorescence probe. (B) Saturation curve of XIAP

polarization (FP) assay to determine quantitatively the binding

(residues 120-356) containing both BIR2 and BIR3 domains to Smac-1F.

affinities of our designed Smac mimetics to XIAP containing

both BIR2 and BIR3 domains.

succinimide ester furnished a biotinylated amide. Removal of

the Boc protecting group in the amide gave our designed

biotinylated compound 2 (BL-SM-122).

Quantitative Determination of the Binding Affinities of

Designed Smac Mimetics to XIAP Protein. The binding

affinities of compounds 1, 3, 4, 5, and 6 and the Smac AVPI

peptide to XIAP L-BIR2-BIR3 protein were determined in a

competitive FP assay using Smac-1F as the tracer (Figure 8).

Compounds 1, 3, 4, and 6 and the Smac AVPI peptide bind to

XIAP with IC₅₀values of 438, 376, 1.39, 71.5, and 10,396 nM,

respectively, whereas the designed inactive control 5 shows no

appreciable binding at 100 µM. Hence, the bivalent Smac

mimetic 4 is 271-351 times more potent than the monovalent

compounds 1 and 3 and >7000 times more potent than the Smac

AVPI peptide, respectively. The stereoisomer 6 is 51 times less

potent than compound 4, confirming the importance of the

stereospecificity. There was no significant difference with

binding affinities for each of these Smac mimetics when either

XIAP L-BIR2-BIR3 or BIR2-BIR3 protein (residues 156-356)

was used in the binding assays (data not shown), indicating that

the linker preceding BIR2 domain does not interact with these

Smac mimetics. Our binding data demonstrate that the designed

bivalent Smac mimetic 4 achieves an extremely high binding

affinity to XIAP proteins containing both BIR2 and BIR3

domains and is much more potent than monovalent Smac

mimetics 1 and 2 and the natural Smac AVPI peptide.

Design of a Fluorescently Tagged, Bivalent Smac Peptide

and Development of a Competitive Binding Assay to XIAP

Containing BIR2 and BIR3 Domains. To evaluate accurately

the binding affinities of our designed bivalent Smac mimetics,

it was critical to develop a sensitive and reliable binding assay

for XIAP containing both BIR2 and BIR3 domains. For this

purpose, we designed and synthesized a fluorescently tagged

Smac-based peptide, Smac-1F (Figure 7), which contains two

basic binding units (AKPF) chemically tethered through two

lysines by a flexible linker. This design was based upon a

biochemical study,³⁰which showed that the second residue

(valine) in the Smac AVPI peptide can be replaced by a lysine

and the fourth residue by a phenylalanine to yield new peptides

with a better binding affinity than the original Smac AVPI

peptide to XIAP BIR3. Saturation experiments determined that

Smac-1F binds to L-BIR2-BIR3 (residues 120-356) with a K_d

value of 2.3 nM (Figure 7). We further determined that the linker

preceding BIR2 does not contribute to the binding to Smac-

1F.³⁵Our extensive investigations conclusively showed that

Smac-1F indeed interacts with both BIR2 and BIR3 domains

simutaneously in XIAP.³⁵Using Smac-1F and XIAP L-BIR2-

BIR3 protein, we have established a competitive fluorescence-

Probing the Binding Modes of Smac Mimetics to XIAP

Proteins by Analytical Gel Filtration. Although compound 4

was designed to target both the BIR2 and BIR3 domains in

XIAP simultaneously, it was unclear if this is indeed happening.

To probe the binding mode of 1, 3, and 4 to XIAP, we

(34) Sun, H.; Nikolovska, C. Z.; Yang, C. Y.; Wang, S. Tetrahedron Lett. 2005,

46, 7015-7018.

(35) Nikolovska-Coleska, Z.; Meagher, J. L.; Jiang, S.; Kawamoto, S. A.; Gao,

W.; Yi, H.; Qin, D.; Roller, P. R.; Stuckey, J. A.; Wang, S. Anal. Biochem.,

in press.

9

15284 J. AM. CHEM. SOC. VOL. 129, NO. 49, 2007

Bivalent Smac Mimetic as a Potent Antagonist of XIAP

A R T I C L E S

Figure 9. Analytical gel filtration of (A) XIAP BIR3-only protein or (B)

XIAP BIR2-BIR3 protein alone or in the presence of monovalent Smac

mimetics 1 and 3 and bivalent Smac mimetic 4. The apparent molecular

weights are shown in parentheses. The calculated molecular weight (MW)

for BIR3 and BIR2-BIR3 proteins in monomeric form are 15.7 and 28.0

kDa, respectively.

Figure 10. Analytical gel filtration of (A) wild-type XIAP BIR2-BIR3,

BIR2(E219R)BIR3 (BIR2M-BIR3), BIR2-BIR3(E314S,W323E) (BIR2-

BIR3M) proteins alone (B) or in the presence of bivalent Smac mimetic 4.

The apparent molecular weights (app MW) are shown in parentheses.

performed analytical gel filtration experiments using recombi-

nant XIAP BIR3 (residues 241-356) and BIR2-BIR3 (residues

156-356) proteins (Figure 9). The BIR2-BIR3 protein without

the linker preceding the BIR2 domain was used because this

protein has a better solubility than the L-BIR2-BIR3 protein

and compound 4 binds to these two proteins with the same

affinities in our FP binding assay.

to the bivalent Smac mimetic 4, we generated mutations in each

of the BIR domains. Based upon our modeling, E219 in the

BIR2 domain is involved in a charge-charge interaction with

compound 1 (Figure 5A), suggesting an important role of E219

in the interaction between 1 and the BIR2 domain. Our modeled

structure also showed that the E314 residue in the BIR3 domain

has a strong charge-charge interaction with 1, and W323 has

hydrophobic contacts with the eight-membered ring in 1,

suggesting that these two residues may play a key role in the

binding of 1 to XIAP BIR3. Accordingly, we have prepared

two mutated proteins, BIR2(E219R)-BIR3 and BIR2-BIR3-

(E314S,W323E), and used them to probe the involvement of

these two individual domains in the binding of compound 4 to

wild-type XIAP BIR2-BIR3. The gel filtration experiments using

wild-type and mutated XIAP proteins containing both BIR2 and

BIR3 domains showed that in the absence of 4, all these proteins

behave as a monomer (Figure 10A). In the presence of

compound 4, complete dimerization of the XIAP BIR2(E219R)-

BIR3 mutated protein is observed, which is in sharp contrast to

the absence of dimerization when the wild-type XIAP BIR2-

BIR3 protein was used (Figure 10B). These data clearly showed

that BIR2 is directly involved in the interaction of XIAP BIR2-

BIR3 with compound 4. The gel filtration experiments further

showed that compound 4 induces partial dimerization of the

XIAP BIR2-BIR3(E314S,W323E) mutated protein, presumably

Via binding to two native BIR2 domains (Figure 10B). The

incomplete dimerization of the XIAP BIR2-BIR3(E314S,-

W323E) protein may be due to the relatively weak binding of

its monovalent binding unit in 4 to the BIR2 domain but

nevertheless clearly demonstrates that BIR3 is also directly

involved in the binding of XIAP BIR2-BIR3 with compound

4. These gel filtration experiments using both native and mutated

proteins, together with the binding data, provide clear and

When the recombinant XIAP protein containing only the

BIR3 domain was used in the gel filtration experiment,

compound 4 induced dimerization of the protein (Figure 9A),

indicating that each of the two monovalent binding units in 4

binds to one XIAP BIR3-only protein and forms a 1:2 complex

between 4 and the XIAP BIR3-only protein. In sharp contrast,

when the recombinant XIAP protein containing both BIR2 and

BIR3 domains was used in the gel filtration experiment,

compound 4 did not induce dimerization of the protein (Figure

9B). Instead, compound 4 appeared to make the protein become

more compact, judging by the apparent molecular weight derived

from the gel filtration data. Furthermore, monovalent compounds

1 and 3 failed to cause dimerization of either BIR3 or BIR2-

BIR3 protein. These data suggest that when 4 interacts with

the BIR3-only protein, one molecule of compound 4 binds to

two molecules of XIAP BIR3 protein and induces dimerization

of the protein. But when presented with XIAP containing both

BIR2 and BIR3 domains, it binds simultaneously to both the

BIR2 and BIR3 domains in the same XIAP molecule.

Although our binding data and gel filtration experiments using

wild-type XIAP proteins strongly suggested that bivalent Smac

mimetic 4 interacts simultaneously with both the BIR2 and BIR3

domains in the same XIAP molecule, these data alone do not

conclusively show that either BIR2 or BIR3 is directly involved

in the binding. To confirm that both BIR2 and BIR3 domains

are directly involved in the binding of XIAP BIR2-BIR3 protein

9

J. AM. CHEM. SOC. VOL. 129, NO. 49, 2007 15285

A R T I C L E S

Sun et al.

Figure 11. Small regions of overplayed ¹⁵N HSQC NMR spectra of ¹⁵N-labeled XIAP BIR2-BIR3 with various concentrations of compounds 1 and 4 (blue:

0 µM, orange: 10 µM, red: 30 µM, and green: 150 µM (for panels e and f only). (a-c) Residue valine 298 (V298) in BIR3 domain of BIR2-BIR3 protein

with different concentrations of compounds 4 and 1, respectively; (d-f) a residue in BIR2 domain of BIR2-BIR3 proteins with different concentrations of

compounds 4 and 1, respectively.

convincing evidence that the bivalent Smac mimetic 4 interacts

simultaneously with both the BIR2 and BIR3 domains in XIAP,

achieving an extremely tight binding to the protein.

There are striking differences between the ¹⁵N HSQC spectra

of XIAP BIR2-BIR3 protein with compounds 1 and 4 (Sup-

porting Information). Overall, more of the protein residues are

affected by compound 4 than by compound 1, and the changes

are apparent at the lowest concentration (10 µM) of compound

4 tested, which is at 1:10 ratio of the protein concentration used

in the NMR experiments. In comparison, higher concentrations

of compound 1 are required for similar changes in chemical

shifts, indicating that compound 1 has a lower binding affinity

to XIAP than compound 4, consistent with FP binding data.

Analysis of the residues in BIR3 domain of BIR2-BIR3

protein showed that compounds 1 and 4 affected the same BIR3

residues (Supporting Information), indicating that they bind to

the same site in the BIR3 domain. However, compound 4 at

the same concentrations has a greater effect on these BIR3

residues than compound 1. For example, both compounds 4 and

1 induce significant chemical shifts on valine 298 (V298) (Figure

11a-c) and a number of other residues (Supporting Information)

located in the BIR3 domain at the lowest concentration tested,

but compound 4 at the same concentrations has a greater effect

than compound 1. These data showed that although both

compounds bind to BIR3 strongly, compound 4 binds to BIR3

with an even higher affinity than compound 1. Of note, V298

directly contacts with compound 1 in our predicted binding

model (Figure 5) and with the isoleucine residue in the Smac

AVPI peptide in both crystal and NMR solution structures.^12,13

The BIR2 residues affected with different concentrations of

compound 4 follow the same trends as seen with the BIR3

residues, and there are also significant changes in chemical shifts

for BIR2 residues when compound 4 was tested at 1:10 ratio

Analysis of the Binding of Smac Mimetics to XIAP BIR2-

BIR3 protein by HSQC NMR Spectroscopy. Heteronuclear

single quantum correlation (HSQC) NMR spectroscopy was

employed to further investigate if a single molecule of bivalent

compound 4 concurrently interacts with the BIR2 and BIR3

domains in XIAP and if monovalent compound 1 and bivalent

compound 4 have different binding modes to XIAP containing

both BIR domains. The BIR2-BIR3 protein (residues 156-356)

was also used in our NMR experiments for its better solubility

than the L-BIR2-BIR3 protein.

It was found that most of the resonances within the BIR3

domain in XIAP BIR2-BIR3 protein appear at identical positions

as those from the BIR3-only protein (residues 241-356), indi-

cating that the BIR3 domain structure is well preserved in XIAP

BIR2-BIR3 protein (data not shown). By comparing the spectra

of BIR2-only, BIR3-only, and BIR2-BIR3 proteins, we were

able to assign most of the peaks for the BIR3 domain and to

identify peaks associated with the BIR2 domain in the XIAP

BIR2-BIR3 protein.

When compound 4 was added to uniformly ¹⁵N-labeled BIR2-

BIR3 protein, ¹⁵N HSQC spectra showed that many residues in

the protein are affected by the compound (Supporting Informa-

tion). Line width analysis of the residues within the core

structural domains indicated that XIAP BIR2-BIR3 protein, with

or without compound 4, has approximately the same molecular

size, confirming that compound 4 does not cause dimerization

of the protein, consistent with our gel filtration results.

9

15286 J. AM. CHEM. SOC. VOL. 129, NO. 49, 2007

Bivalent Smac Mimetic as a Potent Antagonist of XIAP

A R T I C L E S

Figure 13. (A) Western blot analysis of expression levels of IAP proteins

in Jurkat cells transfected with vector control (Vec-JK), or full length of

XIAP (XIAP-JK). (B) Induction of cell death by TRAIL alone or in

combination with compounds 1 and 4. Cells were treated with TRAIL alone

or in combination for 24 h and cell viability was assessed by trypan blue

exclusion assay.

Figure 12. Functional antagonism of Smac mimetics to XIAP protein

(residues 120-356) in cell-free (A) caspase-9 and (B) caspase-3/-7 activity

assays.

of the protein concentration. In sharp contrast, even at the highest

concentration (150 µM) of compound 1 tested, which is 50%

in excess of the protein concentration, BIR2 residues are affected

much more weakly than the BIR3 residues (Figure 11 and

Supporting Information). These data showed that compound 1

binds to BIR2 with a much weaker affinity than to BIR3, but

compound 4 binds to both BIR domains with high affinities.

Taken together, our NMR data provide additional and strong

evidence that bivalent compound 4 interacts concurrently with

both BIR2 and BIR3 domains in the same XIAP molecule, and

monovalent compound 1 binds to the BIR3 domain with a higher

affinity than to the BIR2 domain. Furthermore, compound 4

binds to XIAP with a higher affinity than compound 1. These

NMR data are entirely consistent with binding data and the

results from gel filtration experiments using native and mutated

XIAP proteins.

Antagonism by Smac Mimetics of XIAP Activation of

Caspase-9 and -3/-7 in Cell Free Functional Assays. Since

XIAP functions as a potent inhibitor of caspase-9, -3 and -7,

we evaluated 1, 4, 5 and the AVPI peptide for their ability to

antagonize XIAP in cell-free caspase functional assays (Figure

12). In these assays, XIAP L-BIR2-BIR3 (residues 120-356)

was found to dose-dependently inhibit the activity of caspase-9

and caspase-3/-7 and achieve complete inhibition at 50 nM

(Figure 12). Both compounds 1 and 4 antagonize XIAP in a

dose-dependent manner and are capable of restoring the activity

of caspase-9, as well as that of caspase-3/-7. Consistent with

their binding affinities to XIAP, compound 4 is 100 times more

potent than 1, and significantly, at an equal molar concentration

of XIAP, it completely overcomes the inhibition of XIAP and

fully restores the activity of caspase-9 and -3/-7, indicating its

extremely high potency to antagonize XIAP. In comparison,

100 µM of the Smac AVPI peptide, (2000 times of the

concentration of XIAP protein) is needed to completely restore

the activity of caspase-9 and caspase-3/-7, and the inactive

control 5 has a minimal effect at 100 µM. These functional data

show that bivalent Smac mimetic 4 functions as an ultrapotent

antagonist of XIAP and is 100 and 2000 times more potent than

the corresponding monovalent Smac mimetic 1 and the Smac

AVPI peptide, respectively, entirely consistent with their binding

affinities to XIAP.

Cellular Antagonism to XIAP in Isogenic Cell Lines

Transfected with XIAP or Vector Control. Our cell-free

functional assays demonstrate that our designed bivalent Smac

mimetic 4 functions as an ultrapotent antagonist of XIAP and

is much more potent than its monovalent counterpart 1. We

next determined if compounds 1 and 4 are capable of antagoniz-

ing XIAP in cells and if 4 is also much more potent than 1. For

this purpose, we employed a pair of isogenic Jurkat cell lines

transfected with either XIAP (XIAP-JK) or vector control (Vec-

JK).³⁶Western blot analysis showed that XIAP-JK cells have

much higher levels of XIAP than Vec-JK cells, and transfection

of XIAP did not alter the levels of other IAP proteins, including

cIAP-1 and cIAP-2 (Figure 13A). XIAP-JK cells become highly

resistant to TRAIL (tumor necrosis factor-related apoptosis-

inducing ligand) as compared to Vec-JK cells (Figure 13B).

Both 1 and 4 can antagonize XIAP to cell death induction by

TRAIL, but 4 is 100 times more potent than 1 in overcoming

the resistance of XIAP to cell death induction by TRAIL. Of

note, both 1 and 4 have no significant effect on their own in

these two Jurkat cell lines. These data clearly showed that 4 is

a highly effective, cell-permeable antagonist of XIAP and is

100 times more potent than its monovalent counterpart 1.

Inhibition by Smac Mimetics of Cell Growth and Induc-

tion of Apoptosis in Leukemia Cells. Consistent with its role

as a potent inhibitor of apoptosis, XIAP is highly expressed in

many cancer cell lines and tumor samples from patients.⁶We

(36) Wilkinson, J. C.; Cepero, E.; Boise, L. H.; Duckett, C. S. Mol. Cell. Biol.

2004, 24, 7003-7014.

9

J. AM. CHEM. SOC. VOL. 129, NO. 49, 2007 15287

A R T I C L E S

Sun et al.

of 1400 and 2000 nM, respectively, the stereoisomer 6 has an

IC₅₀value of 90 nM, and the inactive control 5 has an IC₅₀

value >3000 nM. Therefore, the bivalent Smac mimetic 4 is

1000 times more potent than the monovalent compounds 1 and

3 in inhibition of cell growth, and their cell growth activities

correlate well with their binding affinities to XIAP.

Because cell growth inhibition is a combination of cell growth

arrest and cell death induction, we examined if compounds 1,

3, 4, and 5 can induce apoptosis in the HL-60 cell line by flow

cytometry (Figure 14B). Our data showed that compound 4

potently and effectively induces apoptosis in the HL-60 cell

line in a dose-dependent manner (Figure 14B). Compound 4 at

10 and 100 nM for 24 h induces 57% and 69% of HL-60 cells

to undergo apoptosis, respectively. In fact, compound 1 at

concentrations as low as 1 nM for 24 h induces 17% of HL-60

cells to undergo apoptosis, higher than the untreated control

(9%). Monovalent Smac mimetics 1 and 3 at 1000 nM induce

25% of HL-60 cells to undergo apoptosis but are at least 100

times less effective than compound 4. Importantly, the inactive

control compound 5 at 1000 nM has no significant effect,

indicating the specific effect by compound 4.

Since compound 4 effectively antagonizes XIAP and pro-

motes activation of caspase-9 and -3 in cell-free systems, we

evaluated if 4 also activates caspase-9 and -3 in the HL-60

leukemia cells. In addition, we investigated the ability of

compound 4 to induce cleavage of PARP (poly(ADP-ribose)-

polymerase), a substrate of caspase-3 and a biochemical marker

of apoptosis. The results are shown in Figure 14C. As shown,

within 24 h compound 4, at concentrations as low as 10 nM,

induces robust activation of caspase-9 and -3 and cleavage of

PARP and is 100 times more effective than the monovalent

Smac mimetics 1 and 3. The inactive control compound 5 has

no effect on caspase activation and PARP cleavage at 1000 nM.

Taken together, our data showed that these Smac mimetics

can effectively inhibit cell growth and induce apoptosis as single

agents, and the bivalent Smac mimetic 4 is 100-1000 times

more potent than the corresponding monovalent Smac mimetics

1 and 3. Compound 4 achieves an IC₅₀value of 1 nM in the

HL-60 leukemia cell line in the cell growth assay and induces

significant apoptosis at concentrations as low as 1 nM in a 24

h treatment. The activity for compound 4 is highly specific since

the inactive control compound 5 has no or minimal effect in all

these assays at concentrations as high as 1000 nM.

Figure 14. (A). Inhibition of cell growth by Smac mimetics in the HL-60

leukemia cancer cell line. HL-60 cells were treated with these Smac

mimetics for 4 days and cell growth was analyzed by the WST-8 assay.

(B) Induction of apoptosis by Smac mimetics 1, 3, 4, and 5 in the HL-60

leukemia cell line. HL-60 cells were treated with compounds 1, 3, 4, and

5 for 24 h. Cells were harvested, stained with annexin-V-FITC and

propidium iodide (PI) double staining and analyzed by flow cytometer. (C)

Western blot analysis of cleavage of PARP (CL-PARP), caspase-3 (CL-

C3), and -9 (CL-C9) in the HL-60 cell line induced by compounds 1, 3, 4,

and 5. HL-60 cells were treated with Smac mimetics for 24 h. Whole cell

lysates were analyzed with Western blotting using specific antibodies.

â-Actin was used as the loading control.

Intracellular Targets of Designed Smac Mimetics; Co-

immunoprecipitation Assay. Our data show that compound 4

potently induces apoptosis in the HL-60 leukemia cell line at

low nanomolar concentrations. To provide direct evidence that

Smac mimetics 1, 3, and 4 target XIAP in cells, we have

performed co-immunoprecipitation assays using biotinylated BL-

SM-122 in the HL-60 cell lysates (Figure 15).

hypothesize that in some, but not all, cancer cell lines, XIAP

may function as the final defensive mechanism to protect cancer

cells from apoptosis induction, and a potent Smac mimetic may

be capable of inducing apoptosis and inhibiting cell growth as

a single agent without using another apoptotic stimulus by

efficiently antagonizing the XIAP mediated protection.

Indeed, we found that compounds 1, 3, 4, and 6 effectively

inhibit cell growth in the HL-60 human leukemia cell line

(Figure 14A) and a number of other cancer cell lines with

diverse tumor types. Compound 4 achieves an IC₅₀value of 1

nM in inhibition of cell growth in the HL-60 cell line. In

comparison, monovalent compounds 1 and 3 have IC₅₀values

These studies showed that BL-SM-122 pulls down XIAP

protein in a dose-dependent manner in the HL-60 cell lysates.

Compound 4 effectively competes off the binding of BL-SM-

122 to XIAP dose-dependently. Compound 4 at 10 nM competes

off more than 50% of XIAP bound to BL-SM-122 at 10 µM,

and at 100 nM completely blocks the binding of XIAP to BL-

SM-122. Monovalent Smac mimetics 1 and 3 at 1 µM reduce

the amount of XIAP protein bound to BL-SM-122 only slightly,

whereas the inactive control compound 5 has no effect at 1 or

3 µM. These data clearly indicate that bivalent Smac mimetic

9

15288 J. AM. CHEM. SOC. VOL. 129, NO. 49, 2007

Bivalent Smac Mimetic as a Potent Antagonist of XIAP

A R T I C L E S

Figure 15. Probing the interaction of Smac mimetics to cellular XIAP in the HL-60 leukemia cell line by a competitive, co-immunoprecipitation pull-down

assay using biotinylated SM-122 (BL-SM-122). HL-60 whole cell lysates were incubated with BLSM-122 alone, or preincubated with compounds 1, 3, 4,

and 5, followed by coincubation with BL-SM-122. Complexes formed between BL-SM-122 and its targeted proteins were recovered by incubation with

Streptavidin-argarose beads. XIAP protein associated with beads was eluted by heating and detected by western blotting using a monoclonal XIAP antibody.

1

hexane (1:1). H NMR (CDCl₃) δ 5.62 (brs, 1H), 4.63 (m, 1H), 4.50

(m, 1H), 4.25 (m, 1H), 2.41-2.09 (m, 3H), 2.07-1.50 (m, 9H), 1.45

(brs, 9H); ¹³C NMR (75 MHz, CDCl₃) δ 174.77, 172.50, 155.24, 79.72,

59.96, 59.59, 50.97, 36.39, 32.21, 28.36, 26.56, 25.24, 22.65; ESI

MS: m/z 349.2 (M + Na)⁺; HR ESI MS for C₁₆H₂₆N₂O₅Na required:

349.1739, found: 349.1727.

4 binds to cellular XIAP with a much higher binding affinity

than BL-SM-122 and the unlabeled monovalent Smac mimetics

1 and 3.

Summary

We have designed compound 4 (SM-164) as a cell-permeable,

bivalent Smac mimetic to mimic the mode of binding of Smac

protein to XIAP. Our study showed that SM-164 binds to XIAP

protein containing both the BIR2 and BIR3 domains with an

IC₅₀value of 1.39 nM and functions as an extremely potent

antagonist of XIAP in our cell-free functional assays and in

cells. Our gel filtration experiments using wild-type and mutated

XIAP proteins, together with our NMR experiments, established

that SM-164 achieves an extremely high affinity to XIAP by

concurrently interacting with both the BIR2 and BIR3 domains.

It potently binds to cellular XIAP protein, promotes activation

of caspases and effectively induces apoptosis in leukemia cancer

cells at concentrations as low as 1 nM. Our study provides

convincing evidence that concurrent targeting of the BIR2 and

BIR3 domains in XIAP using a nonpeptide, bivalent small

molecule is a highly effective strategy to antagonize XIAP,

promoting apoptosis in cancer cells. SM-164 is a powerful tool

for further elucidation of the cellular functions of XIAP and a

very promising lead compound in the development of a new

class of anticancer therapy aimed at overcoming apoptosis

resistance of cancer cells.

(3S,6S,10aS)-N-Benzhydryl-6-((S)-2-(methylamino)propanamido)-

5-oxodecahydropyrrolo[1,2-a]azocine-3-carboxamide (1). To a solu-

tion of the acid 9 (160 mg, 0.5 mmol) in CH₂Cl₂(10 mL) were added

aminodiphenylmethane (140 mg, 0.77 mmol), EDC (150 mg, 0.77

mmol), HOBt (105 mg, 0.78 mmol), and N,N-diisopropylethylamine

(0.5 mL) at 0 °C with stirring. The mixture was stirred at room

temperature for 8 h and then condensed. The residue was purified by

chromatography to give an amide. To a solution of this amide in MeOH

(5 mL) was added HCl (4 N in 1,4-dioxane, 3 mL). The solution was

stirred at room temperature overnight and then evaporated to give an

ammonium salt. To a mixture of this salt in CH₂Cl₂(10 mL) were

added Boc-N-methyl-^L-alanine (150 mg, 0.74 mmol), EDC (154 mg,

0.8 mmol), HOBt (103 mg, 0.76 mmol), and N,N-diisopropylethylamine

(0.7 mL) at 0 °C with stirring. The mixture was stirred at room

temperature overnight and then condensed. The residue was purified

by chromatography to give an amide. To a solution of this amide in

MeOH (5 mL) was added HCl solution (4 N in 1,4-dioxane, 3 mL).

The solution was stirred at room temperature overnight and then

condensed to give crude compound 1 as a salt with HCl (184 mg, 72%

over four steps). The crude product was purified by C18 reversed-

phase semipreparative HPLC with a gradient from 70% of solvent A

and 30% of solvent B to 50% of solvent A and 50% of solvent B in 40

min. ¹H NMR (300 M Hz, D₂O) δ 7.30-7.18 (m, 10H), 5.95 (s, 1H),

4.74 (m, 1H), 4.34 (m, 1H), 4.24 (m, 1H), 3.83 (m, 1H), 2.57 (s, 3H),

2.21-1.50 (m, 11H), 1.40 (d, J ) 7.0 Hz, 3H), 1.38 (m, 1H); ¹³C NMR

(D₂O) δ 173.77, 172.62, 169.95, 141.56, 141.28, 129.54, 129.44, 128.40,

128.35, 128.01, 127.82, 62.50, 61.40, 58.19, 57.49, 51.42, 36.16, 33.30,

32.64, 31.61, 28.07, 25.36, 22.14, 15.95; ESI MS: m/z 477.3 (M +

H)⁺; HR ESI MS for C₂₈H₃₇N₄O₃required: 477.2866, found: 477.2849.

Experimental Section

1

Chemistry. General Methods. H NMR spectra were acquired at

1

300 MHz and ¹³C spectra at 75 MHz. H chemical shifts are reported

with CDCl₃(7.27 ppm) or HDO (4.70 ppm) as internal standards. ¹³

C

chemical shifts are reported relative to CDCl₃(77.00 ppm) or

1,4-dioxane (67.16 ppm) as internal standards. The final products were

purified by C18 reverse phase semipreparative HPLC column with

solvent A (0.1% of TFA in H₂O) and solvent B (0.1% of TFA in CH₃-

CN) as eluents.

(3S,6S,10aS)-6-(tert-Butoxycarbonylamino)-5-oxodecahydropyr-

rolo[1,2-a]azocine-3-carboxylic Acid (9). To a solution of compound

8 (1.35 g, 4 mmol) in MeOH (20 mL) was added 10% Pd-C (0.2 g).

The mixture was stirred under H₂for 8 h and filtered through celite,

and the filtrate was evaporated to give an ester. To a solution of this

ester in 1,4-dioxane (5 mL) was added 2 N LiOH (6 mL). The mixture

was stirred at room temperature for 2 h and then neutralized with 1 N

HCl until the pH was 4. After extraction of the mixture with ethyl

acetate (3 × 30 mL), the combined organic layers were dried over

Na₂SO₄and removal of the solvent gave the acid 9 (1.25 g. 96% over

two steps). This acid can be used in the following steps without

purification or can be purified by recrystallization from ethyl acetate/

tert-Butyl (3S,6S,10aS)-5-Oxo-3-((R)-1-phenylprop-2-ynylcar-

bamoyl)decahydropyrrolo[1,2-a]azocin-6-ylcarbamate (12). To a

solution of acid 9 (650 mg, 2 mmol) in CH₂Cl₂(20 mL) were added

the chiral amine 11 (380 mg, 2.9 mmol), EDC (580 mg, 3 mmol),

HOBt (405 mg, 3 mmol), and N,N-diisopropylethylamine (1.5 mL) at

0 °C with stirring. The mixture was stirred at room temperature for a

further 8 h and then condensed. The residue was purified by chroma-

tography to give amide 12 (798 mg, 91%). ¹H NMR (CDCl₃): δ 7.80

(brd, J ) 8.6 Hz, 1H), 7.50-7.40 (m, 2H), 7.33-7.20 (m, 3H), 5.91

(dd, J ) 8.6, 2.4 Hz, 1H), 5.45 (brd, J ) 7.9 Hz, 1H), 4.64 (dd, J )

8.3, 6.3 Hz, 1H), 4.52 (m, 1H), 4.10 (m, 1H), 2.60 (m, 1H), 2.45 (d, J

) 2.4 Hz, 1H), 2.20-1.60 (m, 5H), 1.40-1.22 (m, 14H), 1.05 (m,

1H); ¹³C NMR (CDCl₃): δ 172.25, 169.30, 154.95, 138.50, 128.55,

9

J. AM. CHEM. SOC. VOL. 129, NO. 49, 2007 15289

A R T I C L E S

Sun et al.

128.04, 127.04, 81.71, 79.53, 72.75, 59.50, 59.09, 50.95, 44.70, 36.48,

36.28, 31.92, 28.29, 24.89, 23.87, 22.94; ESI MS: m/z 439.3 (M +

Na)⁺; HR ESI MS for C₂₅H₃₃N₃O₄Na required: 462.2369, found:

462.2354.

NMR (D₂O): δ 7.40 (s, 2H), 7.20-7.05 (m, 10H), 6.60 (s, 4H), 6.05

(s, 2H), 4.70 (m, 2H), 4.28 (m, 2H), 4.15 (m, 2H), 4.06 (m, 4H), 3.80

(m, 2H), 2.55 (s, 6H), 2.28-1.04 (m, 38H); ¹³C NMR (D₂O): δ 172.75,

172.09, 169.49, 148.33, 139.53, 129.24, 128.54, 127.41, 123.44, 61.89,

60.89, 57.18, 51.02, 50.30, 35.95, 34.28, 33.09, 32.35, 31.34, 29.22,

27.86, 25.11, 21.92, 15.69; ESI MS: m/z 1121.6 (M + H)⁺; HR ESI

MS for C₆₂H₈₅N₁₄O₆required: 1121.6777, found: 1121.6777.

tert-Butylmethyl ((S)-1-Oxo-1-((3S,6S,10aS)-5-oxo-3-((R)-1-phe-

nylprop-2-ynylcarbamoyl)decahydropyrrolo[1,2-a]azocin-6-ylami-

no)propan-2-yl)carbamate (13). To a solution of compound 12 (440

mg, 1 mmol) in MeOH (5 mL) was added HCl solution (4 N in 1,4-

dioxane, 5 mL). The solution was stirred at room temperature overnight

and then condensed to give an ammonium salt. To a mixture of this

salt and CH₂Cl₂(10 mL) were added Boc-N-methyl-L-alanine (303 mg,

1.5 mmol), EDC (287 mg, 1.5 mmol), HOBt (204 mg, 1.5 mmol), and

N,N-diisopropylethylamine (1.2 mL) at 0 °C with stirring. The mixture

was stirred at room temperature overnight then condensed. The residue

was purified by chromatography to give amide 13 (450 mg, 86% over

(3S,6S,10aS)-6-((S)-2-Acetamido-3-(1H-indol-3-yl)propanamido)-

5-oxo-N-((R)-1-phenylprop-2-ynyl)decahydropyrrolo[1,2-a]azocine-

3-carboxamide (16). To a solution of compound 12 (434 mg, 1 mmol)

in MeOH (5 mL) was added HCl solution (4 N in 1,4-dioxane, 5 mL).

The solution was stirred at room temperature overnight and then

condensed to give an ammonium salt. To a mixture of this salt in CH₂-

Cl₂(10 mL) were added NR-Boc-L-Trp (360 mg, 1.2 mmol), EDC (290

mg, 1.5 mmol), HOBt (202 mg, 1.5 mmol), and N,N-diisopropylethy-

lamine (1.2 mL) at 0 °C with stirring. The mixture was stirred at room

temperature overnight and then condensed. The residue was purified

by chromatography to give an amide. To a solution of this amide in

MeOH (5 mL) was added HCl solution (4 N in 1,4-dioxane, 5 mL).

The solution was stirred at room temperature overnight and then

evaporated to give an ammonium salt. To a mixture of this salt were

added acetic anhydride (0.3 mL) and N,N-diisopropylethylamine (0.8

mL) at 0 °C with stirring. The mixture was stirred at room temperature

overnight then condensed, and the residue was purified by chroma-

tography to give the amide 16 (379 mg, 67% over four steps). ¹H NMR

(CDCl₃): δ 8.71 (brs, 1H), 8.13 (d, J ) 1.6 Hz, 1H), 8.01 (d, J ) 8.5

Hz, 1H), 7.45-7.15 (m, 7H), 7.10-6.92 (m, 2H), 6.77 (d, J ) 1.6 Hz,

1H), 6.39 (d, J ) 8.5 Hz, 1H), 5.91 (dd, J ) 8.5, 2.4 Hz, 1H), 5.20 (m,

1H), 5.05 (m, 1H), 4.34 (t, J ) 8.2 Hz, 1H), 4.12 (m, 1H), 3.20 (m,

1

two steps). H NMR (CDCl₃): δ 7.75 (brd, J ) 8.5 Hz, 1H), 7.50-

7.40 (m, 2H), 7.33-7.20 (m, 3H), 5.91 (dd, J ) 8.5, 2.3 Hz, 1H), 6.88

(brs, 1H), 4.80 (m, 1H), 4.60 (dd, J ) 8.4, 6.3 Hz, 1H), 4.58 (brm,

1H), 4.10 (m, 1H), 2.80 (s, 3H), 2.60 (m, 1H), 2.49 (d, J ) 2.3 Hz,

1H), 2.20-1.60 (m, 5H), 1.50-1.05 (m, 18H); ¹³C NMR (CDCl₃): δ

171.55, 170.51, 169.29, 154.95, 138.52, 128.59, 128.11, 127.09, 81.69,

80.57, 72.81, 59.59, 59.03, 49.88, 44.77, 36.50, 35.86, 31.92, 30.04,

28.36, 24.85, 23.98, 23.01, 13.72; ESI MS: m/z 547.3 (M + Na)⁺;

HR ESI MS for C₂₉H₄₀N₄O₅Na required: 547.2896, found: 547.2897.

(3S,6S,10aS)-N-((S)-(1-Benzyl-1H-1,2,3-triazol-4-yl)(phenyl)methyl)-

6-((S)-2-(methylamino)propanamido)-5-oxodecahydropyrrolo[1,2-

a]azocine-3-carboxamide (3). To a solution of compound 13 (105 mg,

0.2 mmol) and azidomethylbenzene (201 mg, 1.5 mmol) in tert-butyl

alcohol (10 mL) was added a mixture of CuSO₄(20 mg, 0.13 mmol)

and (+)-sodium ^L-ascorbate (60 mg) in H₂O (5 mL). The mixture was

stirred at room temperature overnight and then extracted with CH₂Cl₂

(3 × 15 mL). The combined organic layer was washed with brine,

dried over Na₂SO₄, and condensed. The residue was purified by

chromatography to give a triazole. To a solution of this triazole in

MeOH (5 mL) was added HCl solution (4 N in 1,4-dioxane, 2 mL).

The solution was stirred at room temperature overnight and then

condensed to give crude compound 3 as a salt with HCl (88 mg, 74%

over two steps). This crude product was purified by C18 reversed phase

semipreparative HPLC, and the purity was determined by analytical

HPLC to be over 98%. The gradient ran from 70% of solvent A and

30% of solvent B to 50% of solvent A and 50% of solvent B in 40

1H), 2.90 (m, 1H), 2.45 (d, J ) 2.4 Hz, 1H), 2.20-1.45 (m, 15H); ¹³

C

NMR (CDCl₃): δ 170.79, 170.72, 170.66, 170.11, 137.60, 135.81,

128.61, 127.83, 127.59, 127.06, 122.57, 121.69, 118.94, 118.30, 111.13,

110.55, 81.66, 72.60, 60.45, 60.18, 53.37, 49.38, 44.47, 36.59, 35.96,

31.99, 30.05, 26.44, 25.56, 22.98, 22.34, 20.63; ESI MS: m/z 590.3

(M + Na)⁺; HR ESI MS for C₃₃H₃₇N₅O₄Na required: 590.2743,

found: 590.2748.

(S,3S,3′S,6S,6′S,10aS,10a′S)-N,N′-((1S,1′S)-(1,1′-(4,4′-(1,4-Phe-

nylene)bis(butane-4,1-diyl))bis(1H-1,2,3-triazole-4,1-diyl))bis(phe-

nylmethylene))bis(6-((S)-2-acetamido-3-(1H-indol-3-yl)propanamido)-

5-oxodecahydropyrrolo[1,2-a]azocine-3-carboxamide) (5). To a solution

of compound 16 (190 mg, 0.34 mmol) and 1,4-bis-(4-azidobutyl)-

benzene (45 mg, 0.17 mmol) in tert-butyl alcohol (10 mL) was added

a mixture of CuSO₄(20 mg, 0.13 mmol) and (+)-sodium l-ascorbate

(60 mg) in H₂O (5 mL). The mixture was stirred at room temperature

overnight and then extracted with CH₂Cl₂(3 × 30 mL). The combined

organic layer was washed with brine, dried over Na₂SO₄, and evaporated

to give a residue which was purified by chromatography to give

1

min. H NMR (D₂O) δ 7.66 (s, 1H), 7.35-7.10 (m, 10H), 6.04 (m,

1H), 5.36 (s, 2H), 4.70 (m, 1H), 4.28 (m, 1H), 4.19 (m, 1H), 3.80 (m,

1H), 2.56 (s, 3H), 2.20-1.50 (m, 11H), 1.40 (d, J ) 7.1 Hz, 3H), 1.38

(m, 1H); ¹³C NMR (D₂O) δ 173.14, 172.21, 169.50, 148.55, 139.09,

134.99, 129.40, 129.27, 129.04, 128.50, 128.31, 127.38, 124.02, 61.97,

60.97, 57.20, 54.22, 51.08, 50.36, 35.91, 33.05, 32.28, 31.32, 27.65,

25.06, 21.92, 15.63; ESI MS: m/z 580.3 (M + Na)⁺; HR ESI MS for

C₃₁H₃₉N₇O₃Na required: 580.3012, found: 580.3017.

1

compound 5 (282 mg, 76%). H NMR (CDCl₃): δ 8.70 (s, 2H), 8.32

(d, J ) 8.0 Hz, 2H), 7.60 (d, J ) 7.4 Hz, 2H), 7.40-6.80 (m, 26 H),

6.47 (d, J ) 7.9 Hz, 2H), 6.25 (d, J ) 8.0 Hz, 2H), 4.81 (m, 2H), 4.70

(m, 2H), 4.50 (m, 2H), 4.27 (t, J ) 7.1 Hz, 4H), 3.95 (m, 2H), 3.30-

3.08 (m, 4H), 2.60 (t, J ) 2.61 Hz, 4H), 2.45-2.25 (m, 2H), 2.12-

1.20 (m, 36H); ¹³C NMR (CDCl₃): δ 170.78, 170.41, 170.14, 169.88,

148.00, 140.45, 138.98, 136.19, 128.53, 128.44, 127.68, 127.56, 127.28,

123.70, 121.67, 121.55, 119.13, 118.57, 111.29, 110.25, 59.98, 59.06,

54.01, 50.40, 50.20, 50.01, 47.76, 35.93, 34.61, 31.74, 29.53, 29.09,

28.08, 24.83, 23.30, 22.95, 20.82; ESI MS: m/z 1429.7 (M + Na)⁺;

HR ESI MS for C₈₀H₉₄N₁₆O₈Na required: 1429.7338, found: 1429.7341.

(S,3S,3′S,6S,6′S,10aS,10a′S)-N,N′-((1S,1′S)-(1,1′-(4,4′-(1,4-Phe-

nylene)bis(butane-4,1-diyl))bis(1H-1,2,3-triazole-4,1-diyl))bis(phe-

nylmethylene))bis(6-((S)-2-(methylamino)propanamido)-5-oxodecahy-

dropyrrolo[1,2-a]azocine-3-carboxamide) (4). To a solution of

compound 13 (280 mg, 0.53 mmol) and 1,4-bis-(4-azidobutyl)benzene

(71 mg, 0.26 mmol) in tert-butyl alcohol (10 mL) was added a mixture

of CuSO₄(20 mg, 0.13 mmol) and (+)-sodium L-ascorbate (60 mg) in

H₂O (5 mL). The mixture was stirred at room temperature overnight

and then extracted with CH₂Cl₂(3 × 30 mL). The combined organic

layer was washed with brine, dried over Na₂SO₄, and evaporated to

give a residue which was purified by chromatography to give a bis-

triazole. To a solution of this bis-triazole in MeOH (5 mL) was added

HCl solution (4 N in 1,4-dioxane, 5 mL). The solution was stirred at

room temperature overnight and then condensed to give crude

compound 4 as a salt with HCl (215 mg, 68% over two steps). This

crude product was purified by C18 reversed phase semipreparative

HPLC and the purity checked by analytical HPLC to be over 98%. ¹H

tert-Butylmethyl ((S)-1-Oxo-1-((3S,6S,10aS)-5-oxo-3-((S)-1-phe-

nylprop-2-ynylcarbamoyl)decahydropyrrolo[1,2-a]azocin-6-ylami-

no)propan-2-yl)carbamate (15). To a solution of the acid 9 (480 mg,

1.47 mmol) in CH₂Cl₂(15 mL) were added chiral amine 14 (290 mg,

2.21 mmol), EDC (430 mg, 2.25 mmol), HOBt (297 mg, 2.2 mmol)

and N,N-diisoopropylethylamine (1.5 mL) at 0 °C with stirring. The

mixture was stirred at room temperature for 8 h and then condensed.

The residue was purified by chromatography to give an amide. To a

9

15290 J. AM. CHEM. SOC. VOL. 129, NO. 49, 2007

Bivalent Smac Mimetic as a Potent Antagonist of XIAP

A R T I C L E S

solution of this amide in MeOH (5 mL) was added HCl solution (4 N

in 1,4-dioxane, 5 mL). The solution was stirred at room temperature

overnight and then condensed to give an ammonium salt. To a mixture

of this salt in CH₂Cl₂(10 mL) were added Boc-N-methyl-L-alanine

(404 mg, 2.0 mmol), EDC (390 mg, 2.0 mmol), HOBt (272 mg, 2.0

mmol), and N,N-diisopropylethylamine (2.0 mL) at 0 °C with stirring.

The mixture was stirred at room temperature overnight and then

concentrated. The residue was purified by chromatography to give the

to give a residue which was purified by chromatography to give

compound 19 (245 mg, 42% over five steps). ¹H NMR (CDCl₃) δ 7.95

(brd, J ) 8.5 Hz, 1H), 7.40-7.05 (m, 14H), 6.95 (brm, 1H), 6.19 (brd,

J ) 8.6 Hz, 1H), 5.49 (m, 1H), 5.10 (brs, 2H), 4.95-4.76 (m, 2H),

4.70 (m, 1H), 4.60 (brm, 1H), 4.14 (m, 1H), 3.33-3.12 (m, 4H), 2.86

(s, 3H), 2.68-2.49 (m, 3H), 2.20-1.15 (m, 33H); ¹³C NMR (CDCl₃)

δ172.71, 171.65, 170.58, 169.54, 156.41, 141.80, 140.50, 139.28,

136.59, 128.57, 128.47, 128.05, 127.45, 127.26, 127.14, 80.61, 66.54,

59.90, 59.16, 56.88, 50.00, 40.74, 39.11, 36.47, 35.91, 32.88, 31.94,

31.12, 30.04, 29.61, 28.37, 26.20, 25.14, 24.89, 24.12, 23.17, 13.77;

ESI MS: m/z 903.5 (M + Na)⁺; HR ESI MS for C₅₀H₆₈N₆O₈

required: 903.4996, found: 903.5009.

1

amide 15 (563 mg, 73% over three steps). H NMR (CDCl₃): δ 7.85

(brd, J ) 8.5 Hz, 1H), 7.50-7.42 (m, 2H), 7.35-7.20 (m, 3H), 6.98

(brd, 1H), 4.86 (m, 1H), 5.90 (dd, J ) 8.5, 2.4 Hz, 1H), 4.73 (brm,

1H), 4.62 (m, 1H), 4.21 (m, 1H), 2.81 (s, 3H), 2.55 (m, 1H), 2.42 (d,

J ) 2.4 Hz, 1H), 2.13 (m, 1H), 2.08-1.55 (m, 10H), 1.38 (brs, 9H),

1.32 (d, J ) 7.1 Hz, 3H); ¹³C NMR (CDCl₃): δ 172.23, 171.10, 170.27,

138.61, 129.13, 128.55, 127.35, 82.06, 73.18, 59.89, 59.54, 50.37, 44.93,

37.22, 36.60, 32.29, 30.53, 28.78, 25.66, 24.38, 23.92, 14.29; ESI

MS: m/z 547.3 (M + Na)⁺; HR ESI MS for C₂₉H₄₀N₄O₅Na required:

547.2896, found: 547.2906.

(3S,6S,10aS)-6-((S)-2-(Methylamino)propanamido)-5-oxo-N-((S)-

(4-(3-(6-(5-((3aS,4S,6aR)-2-oxohexahydro-1H-thieno[3,4-d]imidazol-

4-yl)pentanamido)hexanamido)propyl)phenyl)(phenyl)methyl)-

decahydropyrrolo[1,2-a]azocine-3-carboxamide (2, BL-SH-122). To

a solution of compound 19 (110 mg, 0.13 mmol) in MeOH (10 mL)

was added 10% Pd-C (50 mg). The solution was stirred under 1 atm

of H₂at room temperature overnight before filtering through celite and

concentration. The residue was dissolved in CH₂Cl₂(10 mL), and (+)-

biotin N-hydroxy-succinimide ester (48 mg, 0.15 mmol) and N,N-

diisoopropylethylamine (0.5 mL) were added to the solution. The

mixture was stirred overnight and then condensed. The residue was

purified by chromatography to yield a biotinylated amide. To a solution

of the amide in MeOH was added a solution of HCl in 1,4-dioxane.

The solution was stirred at room temperature overnight, and then the

solvent was evaporated to give a crude product which was purified by

C18 reversed phase semipreparative HPLC to give compound 19 (84

mg, 74% over three steps). The purity was checked by analytical HPLC

to be over 98%. The gradient ran from 70% of solvent A and 30% of

(S,3S,3′S,6S,6′S,10aS,10a′S)-N,N′-((1R,1′R)-(1,1′-(4,4′-(1,4-Phe-

nylene)bis(butane-4,1-diyl))bis(1H-1,2,3-triazole-4,1-diyl))bis(phe-

nylmethylene))bis(6-((S)-2-(methylamino)propanamido)-5-oxodecahy-

dropyrrolo[1,2-a]azocine-3-carboxamide) (6). To a solution of

compound 15 (240 mg, 0.46 mmol) and 1,4-bis-(4-azidobutyl)benzene

(60 mg, 0.22 mmol) in tert-butyl alcohol (10 mL) was added a mixture

of CuSO₄(20 mg, 0.13 mmol) and (+)-sodium L-ascorbate (60 mg) in

H₂O (5 mL). The mixture was stirred at room temperature overnight

and then extracted with CH₂Cl₂(3 × 30 mL). The combined organic

layer was washed with brine, dried over Na₂SO₄, and condensed. The

residue was purified by chromatography to give a bis-triazole. To a

solution of this bis-triazole in MeOH (5 mL) was added HCl solution

(4 N in 1,4-dioxane, 5 mL). The solution was stirred at room

temperature overnight and then condensed to give compound 6 as a

salt with HCl (180 mg, 66% over two steps). The product was purified

by C18 reversed phase semipreparative HPLC and the purity determined

by analytical HPLC to be over 98%. ¹H NMR (D₂O): δ 7.44 (s, 2H),

7.30-6.80 (m, 10H), 6.49 (s, 4H), 5.99 (s, 2H), 4.63 (m, 2H), 4.28

(m, 2H), 4.06 (m, 2H), 3.92 (m, 2H), 3.80 (m, 4H), 2.55 (s, 6H), 2.28-

0.95 (m, 42H); ¹³C NMR (D₂O): δ 175.04, 174.33, 171.97, 150.65,

143.08, 141.97, 131.63, 131.00, 130.59, 129.60, 126.25, 64.31, 63.33,

59.74, 53.47, 52.92, 38.47, 36.87, 35.85, 34.86, 33.87, 31.85, 30.42,

27.65, 24.52, 18.26. ESI MS: m/z 1143.7 (M + Na)⁺; HR ESI MS for

C₆₂H₈₅N₁₄O₆Na required: 1143.6596, found: 1143.6616.

1

solvent B to 50% of solvent A and 50% of solvent B in 40 min. H

NMR (D₂O): δ 7.22-7.04 (m, 5H), 7.02-6.90 (m, 4H), 5.87 (s, 1H),

4.65 (m, 1H), 4.32 (m, 1H), 4.22 (m, 1H), 4.12 (m, 1H), 4.01 (m, 1H),

3.80 (m, 1H), 3.05-2.82 (m, 5H), 2.62 (m, 1H), 2.52 (s, 3H), 2.46 (m,

1H), 2.45-2.30 (m, 2H), 2.18-1.05 (m, 33H); ¹³C NMR (D₂O): δ

176.33, 176.30, 172.70, 172.07, 169.46, 165.37, 141.42, 139.27, 129.10,

127.94, 127.63, 62.31, 62.03, 60.99, 60.46, 57.43, 57.19, 55.76, 51.00,

40.08, 39.40, 39.24, 35.98, 35.83, 32.59, 31.31, 30.56, 28.49, 28.37,

28.08, 25.99, 25.66, 25.47, 15.66; ESI MS: m/z 895.5 (M + Na)⁺;

HR ESI MS for C₄₇H₆₈N₈O₆S required: 895.4880, found: 895.4878.

Molecular Modeling. The crystal structure of XIAP BIR3 in a

complex with the Smac protein¹²(PDB code: 1G73) was used to predict

the binding models of XIAP BIR3 bound to designed compounds. The

XIAP BIR2 structure from the X-ray structure¹⁸of the complex of XIAP

BIR2 and caspase-3 (PDB code: 1I3O) was used to predict the binding

models of designed compounds to XIAP BIR2.

(3S,6S,10aS)-6-[2-(tert-Butoxycarbonylmethylamino)pro-

pionylamino]-5-oxodecahydropyrrolo[1,2-a]azocine-3-carbonic Acid

(S)-1-(4-(3-(4-benzoxycarbonylaminobutyrylamino)propyl)phenyl)-

1′-phenyl Amide (18). To a solution of acid 9 (220 mg, 0.67 mmol)

in CH₂Cl₂(10 mL) were added amine 17 (250 mg, 0.67 mmol), EDC

(190 mg, 1.0 mmol), HOBt (135 mg, 1.1 mmol), and N,N-diisopro-

pylethylamine (1 mL) at 0 °C with stirring. The mixture was stirred at

room temperature for 8 h and then evaporated. The residue was purified

by chromatography to give an amide. To a solution of this amide in

MeOH (5 mL) was added HCl (4 N in 1,4-dioxane, 3 mL). The solution

was stirred at room temperature overnight and then condensed to give

an ammonium salt. To a mixture of this salt in CH₂Cl₂(10 mL) were

added Boc-N-methyl-^L-alanine (160 mg, 0.79 mmol), EDC (191 mg,

1.0 mmol), HOBt (140 mg, 1.04 mmol), and N,N-diisopropylethylamine

(1 mL) at 0 °C with stirring. The mixture was stirred at room

temperature overnight and then condensed. The residue was purified

by chromatography to give an amide. To a solution of this amide in

MeOH (10 mL) was added 10% Pd-C (100 mg). The solution was

stirred under 1 atm of H₂at room temperature overnight before being

filtered through celite and condensed. The residue was dissolved in

CH₂Cl₂(10 mL), and Cbz-6-aminohexanoic acid (212 mg, 0.8 mmol),

EDC (193 mg, 1 mmol), HOBt (134 mg, 1 mmol), and N,N-

diisopropylethylamine (1 mL) were added at 0 °C with stirring. The

mixture was stirred at room temperature overnight and then concentrated

A previous study³⁷has demonstrated that the N-terminus of the small

subunit of caspase 3 (sequence: SGVDD) interacts with XIAP BIR2

in the same binding groove as is involved in the XIAP BIR2-Smac

peptide interaction. This groove can be seen clearly in the asymmetric

crystallographic unit between XIAP BIR2 and caspase 3. The Swiss

PDB viewer program was used to generate the asymmetric crystal-

lographic unit in 1I3O while retaining the structure of XIAP BIR2 and

the SGVDDDM peptide from caspase-3 for further molecular dynamics

simulations. Our initial experiments docking compound 1 with XIAP

BIR2 using the X-ray structure of XIAP BIR2 from 1I3O failed to

yield satisfactory results. Examination of the crystal structure of XIAP

BIR2 and comparison with that of XIAP BIR3 showed that the first

aspartic acid residue, D in the SGVDDDM peptide, interacts with the

side chain of Q197 of XIAP BIR2. Thus, the binding pocket in the

crystal structure of XIAP BIR2, which interacts with this aspartate

residue, may be forced to adopt a conformation that is incompatible

with its interaction with a hydrophobic residue in the fourth position

(37) Scott, F. L.; Denault, J.-B.; Riedl, S. J.; Shin, H.; Renatus, M.; Salvesen,

G. S. EMBO J. 2005, 24, 645-655.

9

J. AM. CHEM. SOC. VOL. 129, NO. 49, 2007 15291

A R T I C L E S

Sun et al.

of the Smac AVPI peptide. To resolve this problem, we performed an

extensive molecular dynamics simulation of the XIAP BIR2 structure.

In order to investigate if a hydrophobic residue in the fourth position

can induce a conformational change of the XIAP BIR2 and achieve

favorable interactions, we used Sybyl³⁸to mutate this residue, converting

SGVDDDM to SGVFDDM and added an N-methyl group to the

C-terminus of the peptide. A molecular dynamics simulation of XIAP

BIR2 in a complex with this mutated peptide was then performed to

refine the structure.

algorithm (GA) run, a maximum number of 200 000 operations were

performed on a population of 5 islands of 100 individuals. Operator

weights for crossover, mutation, and migration were set to 95, 95, and

10, respectively. The docking simulations were terminated after 20

runs for each ligand. GoldScore implemented in Gold 2.2 was used as

the fitness function to evaluate the docked conformations. The 20

conformations ranked highest by each fitness function were saved for

analysis of the predicted docking modes. For the docking poses reported,

these were the highest ranked conformations from the docking

simulations.

Prior to the MD simulation, counterions were added to neutralize

the peptide-protein complex before solvating the system with a 10 Å

cube of explicit waters. The four residues bonded with the Zn²⁺ion

were constrained by moderate harmonic forces to prevent unfolding

of the protein. The protocol for the MD simulation was as follows: A

1000-step minimization of the solvated system was performed and

followed by 6 ps of MD simulation to gradually heat the system from

0 K to 298 K. The system was then equilibrated and refined by a further

94 ps simulation at 298 K. During the simulation, the four residues

covalently bonded with the zinc ion and the backbone atoms of the

remaining residues in XIAP BIR2 are constrained by harmonic forces

with force constants equal to 15 kcal/mol. On the basis of this protocol,

only the side chain atoms of XIAP BIR2 are refined and adapted to

the mutated peptide (SGVFDDM) as compared to the SGVDDM

determined in the crystal structure. This refined structure of XIAP BIR2

was then used to predict the models of our designed compounds binding

to XIAP BIR2.

We used the Amber program suite³⁹(version 7) to perform all

molecular dynamics (MD) simulations. A recent version of the Amber

force field (ff96)⁴⁰was used for the natural amino acids in the complex,

and the TIP3P model⁴¹was used for water molecules. There is one

Zn²⁺ion covalently bound to C200, C203, H220, and C227 in the XIAP

BIR2 domain and to C300, C303, H320, and C327 in the XIAP BIR3

domain. This Zn²⁺ion, while important for structural integrity, has no

direct interaction with the ligands. We used parameters developed by

Ryde⁴²for the Zn²⁺ion and its coordination with the neighboring four

residues to model this chelating structure in our simulation. All the

MD simulations were carried out at NTP. The SHAKE algorithm⁴³

was used to fix the bonds involving hydrogen. The PME method⁴⁴was

used to account for long-range electrostatic interactions, and the

nonbonded cutoff distance was set at 10 Å. The time step was 2 fs,

and the neighboring pairs list was updated after every 20 steps. For

the refinement of the structure between SM-122 and XIAP BIR2 and

BIR3, the protocol is as follows: A 500-step minimization of the

solvated system was performed followed by 6 ps of MD simulation to

gradually heat the system from 0 K to 298 K . The system was then

equilibrated by another 34 ps simulation at 298 K . Finally, the 1 ns

production simulation was run and the snapshots of conformations

(typically 2000), evenly spaced in time, were collected for structural

analysis.

Protein Expression and Purification, Binding Assays to Different

Constructs of XIAP, Cell-Free Functional Assays for the Activity

of Caspase-9 and Caspase-3/-7, and Analytical Gel Filtration Assays.

1. Protein Expression and Purification. Different constructs of human

XIAP proteins, including linker-BIR2-BIR3 (residues 120-356), BIR3

(residues 241-356), and linker-BIR2 (residues 120-240) were cloned

into a pET28 vector (Novagen) containing an N-terminal 6xHis tag.

BIR2-BIR3 (residues 156-356), without the linker preceding BIR2,

was cloned into a modified HIS-TEV vector with an N-terminal 8xHis

tag. The mutated proteins, BIR2(E219R)-BIR3 and BIR2-BIR3(E314S,-

W323E), were created using QuikChange mutagensis (Stratagene) on

a BIR2-BIR3 (residues 156-356) template. Proteins were produced

in E. coli BL21(DE3) cells grown at 37 °C in 2xYT containing

kanamycin to an OD₆₀₀of 0.6. Protein expression was induced by 0.4

mM IPTG at 20 °C for 20 h (linker-BIR2-BIR3), 20 °C for 4 h (BIR2-

BIR3), 37 °C for 3 h (linker-BIR2), or 27 °C for 4 h (BIR3). Cells

were lysed by sonication in buffer containing Tris pH 7.5 (50 mM),

NaCl (200 mM), ZnAc (50 µM), 0.1% âME and leupectin/aprotin

protease inhibitors. Proteins were purified from the soluble fraction

using Ni-NTA resin (QIAGEN) followed by gel filtration on a Superdex

75 column in Tris pH 7.5 (20 mM), NaCl (200 mM), ZnAc (50 µM),

and dithiothreitol (DTT, 1 mM). After purification, DTT was added to

a final concentration of 10 mM.

2. Fluorescence-Polarization-Based Binding for XIAP BIR3

Protein. A sensitive in Vitro binding assay using the fluorescence

polarization (FP)-based method²⁹was used to determine the binding

affinity of Smac mimetics to XIAP BIR3 protein. In this assay,

5-carboxyfluorescein was coupled to the lysine side chain of a mutated

Smac peptide with the sequence (AbuRPFK-Fam). This fluorescently

tagged peptide, named SM5F, was used as the fluorescent tracer in the

FP-based binding assay of different compounds to the XIAP BIR3

protein. The recombinant XIAP BIR3 protein of human XIAP (residues

241-356) fused to His-tag was stable and soluble and was used for

the FP-based binding assay. The K_dvalue of SM5F peptide to XIAP

BIR3 protein was determined to be 17.9 nM.²⁹

Dose-dependent binding experiments were carried out with serial

dilutions of the tested compounds. An aliquot of the samples and

preincubated XIAP BIR3 protein (0.030 µM) and SM5F peptide (5

nM) in the assay buffer (100 mM potassium phosphate, pH 7.5; 100

µg/mL bovine gamma globulin; 0.02% sodium azide, purchased from

Invitrogen Life Technology), were added to Dynex 96-well, black,

round-bottom plates (Fisher Scientific). For each assay, the controls

included XIAP BIR3 protein and SM5F (equivalent to 0% inhibition),

and SM5F only (equivalent to 100% inhibition). The polarization values

were measured after 3 h of incubation using an Ultra Reader (Tecan

U.S. Inc., Research Triangle Park, NC). IC₅₀values, the inhibitor

concentration at which 50% of the bound tracer is displaced, were

determined from a plot using nonlinear least-squares analysis. Curve

fitting was performed using Graphpad Prism software (GraphPad

Software, Inc., San Diego, CA).

All the binding poses of designed ligands with XIAP BIR2 and BIR3

were developed with the GOLD program⁴⁵(version 2.2). The centers

of the binding sites for XIAP BIR2 and BIR3 were set at T208 and

T308, respectively, and the radius of the binding site was defined as

13 Å, large enough to cover the binding pockets. For each genetic

(38) Sybyl, a molecular modeling system, is supplied by Tripos, Inc., St. Louis,

MO 63144.

(39) Case, D. A.; et al. AMBER7, University of California, San Francisco, 2002.

(40) Kollman, P. A.; Dixon, R.; Cornell, W.; Fox, T.; Chipot, C.; Pohorille, A.

In Computer Simulation of Biomolecular Systems; Wilkinson, A., Weiner,

P., van Gunsteren, W. F., Eds.; Elsevier: New York, 1997; Vol. 3, p 83.

(41) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.; Klein,

M. L. J. Chem. Phys. 1983, 79, 926-935.

3. Surface Plasmon Resonance (SPR) Affinity Measurements and

SPR Competitive Binding Assays to XIAP BIR2 Domain. SPR

experiments were performed at room temperature on a Biacore 3000

biosensor with HBS-P [HEPES, (pH 7.4, 10 mM), NaCl (150 mM),

and 0.005% Tween 20] as the running buffer. Biotin-labeled Smac

mimetic, BL-SM-122, was immobilized onto streptavidin (SA) sensor

(42) Ryde, U. Proteins: Struct. Funct. Genet. 1995, 21, 41-56.

(43) Rychaert, J. P.; Ciccotti, G.; Berendsen, H. J. C. J. Comput. Phys. 1977,

23, 327-341.

(44) Darden, T. A.; York, D. M.; Pedersen, L. J. Chem. Phys. 1993, 98, 10089-

10092.

(45) Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R. J. Mol. Biol.

1997, 267, 727-748.

9

15292 J. AM. CHEM. SOC. VOL. 129, NO. 49, 2007

Bivalent Smac Mimetic as a Potent Antagonist of XIAP

A R T I C L E S

chips by using streptavidin-biotin coupling chemistry. BL-SM-122 was

immobilized on two chips with different densities, 272 and 105 response

units on the sensor chip surface.

(residues 156-356) protein, and BIR2-BIR3(E314S,W323E) protein.

Analytical gel filtration experiments were performed on a Superdex

75 column (GE Healthcare) attached to an AKTA Purifier-10 system

in Tris-HCl, (pH 7.5, 20 mM), NaCl (200 mM), zinc acetate (50 µM),

and DTT (1 mM). Recombinant protein was run on the column at a

concentration of 1 mg/mL alone or after incubation with a 1:1 molar

ratio of Smac mimetic. Molecular weight standards from Amersham-

Pharmacia were used to calibrate the column.

NMR HSQC Experiments. 1. Expression and Purification of ¹⁵N-

Labeled Proteins. The BIR3 domain (residues 241-356) of human

XIAP fused to His-tag (pET28b, Novagen), the BIR2 domain (residues

156-260), and the BIR2-BIR3 protein (residues 156-356) of human

XIAP in pHis-TEV vector were used to express respective proteins

from BL21(DE3) cells (Nogaven) in M9 medium containing ¹⁵NH₄Cl

to label the protein uniformly with ¹⁵N. Most of the proteins were found

in the soluble fraction, and they were purified using TALON (Clontech)

affinity chromatography, followed by Q-XL ion exchange and G75

size-exclusion chromatography (GE Healthcare) with AKTA purifier

(GE Healthcare).

2. NMR HSQC Experiments. ¹⁵N HSQC NMR spectra were

recorded on a Bruker AVANCE 500 MHz NMR spectrometer with

samples containing 100 µM of the ¹⁵N labeled proteins in 50 mM Tris

(pH 7.2), 50 µM ZnCl₂, 1 mM DTT at 25 °C with or without test

compound at a final concentration between 10 and 150 µM. The spectra

were then compared in order to identify residues affected by the

interaction with the test compound. Within the BIR3 domain, the known

chemical shift assignments of the backbone atoms^13,46were used to

identify the residues corresponding to the affected peaks. The backbone

atom resonance assignments of the BIR3 domain were also confirmed

by 3D NMR triple resonance experiments (HNCA, HNCACB, HN-

(CO)CBCA, HNCO, TOCSY-HSQC, C(CO)NH). All NMR data were

processed and analyzed using the nmrPipe/nmrDraw package (Dr. Frank

Delaglio, NIDDK, NIH).⁴⁷

To collect kinetic binding data of BIR2 XIAP protein for BL-SM-

122 directly, the BIR2 protein solution with different concentrations

in running buffer was injected over the ligand and reference flow cells

at a constant flow rate of 20 µL/min. During each injection, the ligand/

protein complex was allowed to associate/dissociate for 240 and 300

s, respectively. Residual bound protein was desorbed with NaOH (50

mM), followed by two washes with the running buffer. Binding kinetics

were derived from sensorgrams after subtraction of baseline responses

and the data were fit globally to a 1:1 interaction model (A + B )

AB) by using the BIA evaluation software.

For competitive binding assay, a fixed concentration of XIAP BIR2-

only protein (1 µM) was incubated with different concentrations of

the tested compound in HBS-P buffer at room temperature. The

mixtures were then injected over a sensor chip containing a channel

with BL-SM-122 immobilized on the chip and a control without BL-

SM-122. The baseline response (control) was subtracted to obtain the

specific binding response. Taking the response for the protein alone as

the maximal response (100%), the relative residual binding (%) in the

presence of different concentrations of the tested compound at a given

injection time point (225 s) was then calculated. The relative residual

responses were plotted against initial concentrations of the tested

compound and fitted using a nonlinear least-squares equation using

Graphpad Prism software (GraphPad Software, Inc.).

4. Fluorescence-Polarization-Based Binding for XIAP L-BIR2-

BIR3 Protein. A FP-based competitive binding assay was established

for quantitative determination of the binding affinities of our designed

Smac mimetics to XIAP containing both BIR2 and BIR3 domains. In

the competitive binding experiments, the tested compound was

incubated with XIAP protein (residues 120-356, 3 nM) and Smac-1F

(1.0 nM) in the assay buffer [potassium phosphate, pH 7.5 (100 mM);

bovine gamma globulin (100 µg/mL); 0.02% sodium azide] in Dynex

96-well, black, round-bottom plates (Fisher Scientific). In each experi-

ment, the controls included XIAP and Smac-1F peptide (equivalent to

0% inhibition), and Smac-1F only (100% of inhibition). Polarization

values were measured after 2 h incubation, using the Ultra Plate reader.

The IC₅₀value, the inhibitor concentration at which 50% of bound tracer

was displaced, was determined from the plot using nonlinear least-

squares analysis. For each assay, Smac AVPI peptide was used as the

control. Curve fitting was performed using Graphpad Prism software

(GraphPad Software, Inc.).

5. Cell-Free Caspase Functional Assays. MDA-MB-231 cell lysates

were prepared by solubilizing cells in ice cold buffer containing KCl

(50 mM), EGTA (5 mM), MgCl₂(2 mM) DTT (1 mM), 0.2% CHAPS

and HEPES (pH 7.5 50 mM), containing cocktail protease inhibitors,

incubating on ice for 10 min, and then freezing in liquid nitrogen.

Cytochrome c and dATP were added to the cell lysates, which were

then incubated at 30 °C in a water bath for 60 min to activate caspase-9

and -3/-7. Addition of recombinant XIAP L-BIR2-BIR3 protein (50

nM) in the cell lysates completely suppressed the activity of caspase-

9, and caspase-3/-7. Different concentrations of a tested Smac mimetic

(1 nM to 100 µM) were added to determine the restoration of the activity

of these caspases.

Assays for Cell Growth, Cell Viability, Apoptosis, Western

Blotting, and Biotin-Streptavidin Pull-Down. 1. Cell Lines. Human

HL-60 leukemia cell line was purchased from the American Type

Culture Collection (Manassas, VA), maintained in RPMI 1640 (Invit-

rogen), supplemented with 10% fetal bovine serum (Invitrogen) and

1% penicillin-streptomycin, at 37 °C in 5% CO₂. Human Jurkat T

leukemic cell lines stably transfected with pEBB-HA (VEC-JK) and

pEBB-HA-XIAP (XIAP-JK) were a kind gift of Dr. Colin Duckett

(Departments of Pathology and Internal Medicine, University of

Michigan). Jurkat cell lines were cultured in RPMI 1640 supplemented

with 10% fetal bovine serum (Invitrogen) and 1% penicillin-strepto-

mycin in the presence of puromycin from EMD Biosciences (2 µg/

mL).

2. Cell Growth Assay. The effect of Smac mimetics on HL-60 cell

growth was evaluated by a WST-8 [2-(2-methoxy-4-nitrophenyl)-3-

(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium, monosodium salt

assay (Dojindo Molecular Technologies, Inc). Cells (3000-4000 cells

in each well) were cultured in 96-well tissue culture plates in medium

(200 µL) containing various concentrations of Smac mimetics for the

indicated time. At the end of incubation, WST-8 dye (20 µL) was added

to each well and incubated for an additional 1-3 h, and then the

absorbance was measured in a microplate reader (Molecular Devices)

at 450 nm. Cell growth inhibition was evaluated as the ratio of the

absorbance of the sample to that of the control.

For determination of caspase activity, 25 µM of either caspase-9

substrate (Z-LEHD-AFC) or caspase-3/-7 substrate (Z-DEVD-AFC)

(BioVision Inc.) was added. Fluorescence detection of substrate

cleavage by caspase-9 or -3/-7 for their specific substrate was carried

out on an Ultra Reader using an excitation wavelength of 400 nm and

an emission wavelength of 505 nm. The reaction was monitored for

1-2 h.

3. Cell Viability Assay. Cell viability was quantitated by microscopic

examination in a trypan blue exclusion assay. Cells were treated in

triplicate for 24 h, harvested, and stained with an equal volume of 0.04%

trypan blue. Both blue cells and morphologically shrunk cells were

6. Analytical Gel Filtration Experiments. To probe the binding

mode of the designed Smac mimetics to XIAP proteins, we performed

analytical gel filtration experiments with BIR3 (residues 241-356)

protein, BIR2-BIR3 (residues 156-356) protein, BIR2(E219R)-BIR3

(46) Sun, C.; Cai, M.; Meadows, R. P.; Xu, N.; Gunasekera, A. H.; Herrmann,

J.; Wu, J. C.; Fesik, S. W. J. Biol. Chem. 2000, 275, 33777-33781.

(47) Delaglio, F.; Grzesiek, S.; Vuister, G. W.; Zhu, G.; Pfeifer, J.; Bax, A. J.

Biomol. NMR 1995, 6, 277-293.

9

J. AM. CHEM. SOC. VOL. 129, NO. 49, 2007 15293

A R T I C L E S

Sun et al.

scored as nonviable cells. At least 100 cells from each treatment,

performed in triplicate, were counted.

biotinylated SM-122, named BL-SM-122, alone for pull-down assay,

or preincubated with compound 1, 2, 3, and 4 for 5 min followed by

co-incubation with BL-SM-122 for competitive experiments. Complexes

formed between the Smac mimetics and the associated proteins were

recovered by incubation with streptavidin-agarose beads (100 µL) on

a shaker for 2 h at 4 °C and then centrifuged at 1000g for 1 min. The

complexes were then washed three times with lysis buffer at 4 °C and

eluted from the beads by boiling in SDS loading buffer (100 µL). The

eluted proteins were detected by Western blotting.

4. Apoptosis Assay. Apoptosis assays were performed with an

annexin-V/propidium iodide (PI) apoptosis detection kit (Roche)

according to manufacturer’s instructions. Briefly, cells were harvested,

washed with ice-cold PBS, and then stained with annexin-V-FITC

and PI for 15 min at room temperature in the dark. Stained cells were

analyzed in a FACS calibur flow cytometer. Annexin-V (+) cells were

measured as apoptotic cells; annexin V (-) and PI (+) cells were

measured as death cells.

Acknowledgment. We are grateful for the financial support

from the Breast Cancer Research Foundation, the Prostate

Cancer Foundation, the Department of Defense Prostate Cancer

Program (W81XWH-04-1-0213), Ascenta Therapeutics, and the

National Cancer Institute, NIH (R01CA109025). Jurkat cells

were kind gifts of Dr. Colin Duckett from the University of

Michigan. The authors thank Dr. Dennis Torchia at the National

Institute of Diabetes and Digestive and Kidney Diseases,

National Institutes of Health, for helpful discussions of the NMR

experiments, Dr. Jason Gestwicki at the Life Sciences Institute

and Department of Pathology at the University of Michigan for

assistance and discussions of the SPR experiments, and Dr. G.

W. A. Milne for critical reading of the manuscript.

5. Western Blotting. Cells were harvested and washed with cold

PBS. Cell pellets were lysed in double lysis buffer (DLB; 50 mmol/L

Tris, 150 mmol/L sodium chloride, (1 mmol)/L EDTA, 0.1% SDS and

1% NP-40) in the presence of L PMSF (1 mmol) and protease inhibitor

cocktail (Roche) for 10 min on ice and then centrifuged at 13 000 rpm

at 4 °C for 10 min. Protein concentrations were determined using a

Bio-Rad protein assay kit (Bio-Rad Laboratories).

Proteins were electrophoresed onto either 4% or 20% gradient SDS-

PAGE (Invitrogen) and then transferred to PVDF membranes. Follow-

ing blocking in 5% milk, membranes were incubated with a specific

primary antibody, washed, and incubated with horseradish peroxidase-

linked secondary antibody (Amersham). The signals were visualized

with the Chemiluminescent HRP antibody detection reagent (Denville

Scientific). When indicated, the blots were stripped and reprobed with

a different antibody. Primary antibodies against caspase-9, cleaved

caspase 3, and â-actin were purchased from Cell Signaling Technology;

primary antibody against cleaved PARP was purchased from Epitomics.

6. Biotin-Avidin Pull-Down Assay. The interaction between

compounds 1, 2, 3, and 4 and cellular XIAP was investigated using a

biotin-streptavidin pull-down assay. Cells were lysed in lysis buffer

[20 mmol/L Tris, NaCl (150 mM/L), and 1% NP40] for 20 min. Cell

lysates were precleared with streptavidin-agarose beads, incubated with

Supporting Information Available: Complete references of

24, 27, and 39, NMR HSQC spectra of XIAP BIR2-BIR3

protein with different concentrations of compounds 1 and 4,

and enlarged view for several BIR3 residues available. This

material is available free of charge via the Internet at

http://pubs.acs.org.

JA074725F

9

15294 J. AM. CHEM. SOC. VOL. 129, NO. 49, 2007

Article Doi

Design, synthesis, and characterization of a potent, nonpeptide, cell-permeable, bivalent smac mimetic that concurrently targets both the BIR2 and BIR3 domains in XIAP

DOI: 10.1021/ja074725f

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ja074725f

Full text of DOI:10.1021/ja074725f

Products guided by the article

R&D Labs maybe for 1001060-43-0

Relevant to this article

Hot Product