Protein knockdown using methyl bestatin-ligand hybrid molecules: Design and synthesis of inducers of ubiquitination-mediated degradation of cellular retinoic acid-binding proteins

Article abstract of DOI:10.1021/ja100691p

Induction of selective degradation of target proteins by small molecules (protein knockdown) would be useful for biological research and treatment of various diseases. To achieve protein knockdown, we utilized the ubiquitin ligase activity of cellular inhibitor of apoptosis protein 1 (cIAP1), which is activated by methyl bestatin (MeBS, 2). We speculated that formation of an artificial (nonphysiological) complex of cIAP1 and a target protein would be induced by a hybrid molecule consisting of MeBS (2) linked to a ligand of the target protein, and this would lead to cIAP1-mediated ubiquitination and subsequent proteasomal degradation of the target protein. To verify this hypothesis, we focused on cellular retinoic acid-binding proteins (CRABP-I and -II) and designed hybrid molecules (compounds 4) consisting of MeBS (2) coupled via spacers of various lengths to all-trans retinoic acid (ATRA, 3), a ligand of CRABPs. Compounds 4 induced selective loss of CRABP-I and -II proteins in cells. We confirmed that 4b induced formation of a complex of cIAP1 and CRABP-II in vitro and induced proteasomal degradation of CRABP-II in cells. When neuroblastoma IMR-32 cells were treated with 4b, the level of CRABP-II was reduced and cell migration was inhibited, suggesting potential value of CRABP-II-targeting therapy for controlling tumor metastasis. Our results indicate that 4b possesses sufficient activity, permeability, and stability in cells to be employed in cellular assays. Hybrid molecules such as 4 should be useful not only as chemical tools for studying the biological/physiological functions of CRABPs but also as candidate therapeutic agents targeting CRABPs.

Full text of DOI:10.1021/ja100691p

Published on Web 04/06/2010
Protein Knockdown Using Methyl Bestatin-Ligand Hybrid
Molecules: Design and Synthesis of Inducers of
Ubiquitination-Mediated Degradation of Cellular Retinoic
Acid-Binding Proteins
Yukihiro Itoh,† Minoru Ishikawa,† Mikihiko Naito,‡ and Yuichi Hashimoto*,†
Institute of Molecular and Cellular Biosciences, The UniVersity of Tokyo, 1-1-1 Yayoi,
Bunkyo-ku, Tokyo 113-0032, Japan, and National Institute of Health Sciences, 1-18-1 Kamiyoga,
Setagaya-ku, Tokyo 158-8501, Japan
Received January 26, 2010; E-mail: hashimot@iam.u-tokyo.ac.jp
Abstract: Induction of selective degradation of target proteins by small molecules (protein knockdown)
would be useful for biological research and treatment of various diseases. To achieve protein knockdown,
we utilized the ubiquitin ligase activity of cellular inhibitor of apoptosis protein 1 (cIAP1), which is activated
by methyl bestatin (MeBS, 2). We speculated that formation of an artificial (nonphysiological) complex of
cIAP1 and a target protein would be induced by a hybrid molecule consisting of MeBS (2) linked to a
ligand of the target protein, and this would lead to cIAP1-mediated ubiquitination and subsequent
proteasomal degradation of the target protein. To verify this hypothesis, we focused on cellular retinoic
acid-binding proteins (CRABP-I and -II) and designed hybrid molecules (compounds 4) consisting of MeBS
(2) coupled via spacers of various lengths to all-trans retinoic acid (ATRA, 3), a ligand of CRABPs.
Compounds 4 induced selective loss of CRABP-I and -II proteins in cells. We confirmed that 4b induced
formation of a complex of cIAP1 and CRABP-II in vitro and induced proteasomal degradation of CRABP-II
in cells. When neuroblastoma IMR-32 cells were treated with 4b, the level of CRABP-II was reduced and
cell migration was inhibited, suggesting potential value of CRABP-II-targeting therapy for controlling tumor
metastasis. Our results indicate that 4b possesses sufficient activity, permeability, and stability in cells to
be employed in cellular assays. Hybrid molecules such as 4 should be useful not only as chemical tools
for studying the biological/physiological functions of CRABPs but also as candidate therapeutic agents
targeting CRABPs.
Introduction
a new therapeutic strategy in cases where expression of target
proteins is closely related to specific diseases.
Physiological degradation of proteins via the ubiquitin-
proteasome system is crucial for regulating cellular functions,
including the cell cycle, immunoresponses, and signal trans-
duction.1 In general, protein ubiquitination or polyubiquitination
is mediated by sequential reactions of ubiquitin-activating
enzyme (E1), ubiquitin-conjugating enzyme (E2), and ubiquitin
ligase (E3). Polyubiquitinated proteins are recognized and
degraded by proteasome.2 Many ubiquitin ligases (E3) have been
reported, and it is thought that different E3 ligases have different
specificities; i.e., they distinguish various proteins which are to
be ubiquitinated.3 A method for simply and selectively inducing
posttranslational degradation of target proteins by regulating this
system, which we call protein knockdown, would be useful for
biological studies and medical research. It might also provide
Genetic techniques such as gene knockout and gene knock-
down have been widely used for ablating target proteins and
have been useful to uncover the biological functions of
numerous proteins in cells or animals. However, complicated
and time-consuming genetic manipulation is required for gene
knockout. Gene knockdown using RNA interference is an easy
method but cannot remove existing proteins and so is especially
ineffective in the case of proteins with a long half-life. Therefore,
other techniques which can rapidly remove or down-regulate
proteins post-translationally would be desirable. One such
technique is application of proteolysis-targeting chimeric mol-
ecules (protacs)4 based on peptide structure.5 Protacs have been
reported to degrade target proteins. However, protacs possess
peptide structure and must be polyargininated to endow them
with sufficient membrane permeability for use in cellular
systems.6 Moreover, stability issues associated with their high
molecular weight and vulnerable peptide bonds limit their broad
† The University of Tokyo.
‡ National Institute of Health Sciences.
(1) Konstantinova, I. M.; Tsimokha, A. S.; Mittenberg, A. G. Int. ReV.
Cell Mol. Biol. 2008, 267, 59–124.
(4) (a) Sakamoto, K. M.; Kim, K. B.; Kumagai, A.; Mercurio, F.; Crews,
C. M.; Deshaies, R. J. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 8554–
8559. (b) Sakamoto, K. M.; Kim, K. B.; Verma, R.; Ransick, A.; Stein,
B.; Crews, C. M.; Deshaies, R. J. Mol. Cell. Proteomics 2003, 2, 1350–
1358. (c) Schneekloth, A. R.; Pucheault, M.; Tae, H. S.; Crews, C. M.
Bioorg. Med. Chem. Lett. 2008, 18, 5904–5908.
(2) Ciechanover, A.; Orian, A.; Schwartz, A. L. BioEssays 2000, 22, 442–
451.
(3) (a) Kee, Y.; Huibregtse, J. M. Biochem. Biophys. Res. Commun. 2007,
354, 329–333. (b) Deshaies, R. J.; Joazeiro, C. A. Annu. ReV. Biochem.
2009, 78, 399–434.
9
5820 J. AM. CHEM. SOC. 2010, 132, 5820–5826
10.1021/ja100691p  2010 American Chemical Society

Protein Knockdown Using Methyl Bestatin-Ligand Molecules
A R T I C L E S
Scheme 1. (a) cIAP1 Induces Degradation of Binding Proteins. (b) Auto-Ubiquitination and Degradation of cIAP1. (c) Protein Knockdown
Strategy
applicability, and other general methods to remove target
proteins post-translationally in cells and animals would be
desirable. Therefore, we attempted to develop a new approach,
which we call protein knockdown, using small molecules to
induce selective degradation of target proteins post-translation-
ally.
To develop such a protein knockdown approach, we focused
on cellular inhibitor of apoptosis protein 1 (cIAP1), which is
one of the inhibitor of apoptosis proteins (IAPs) and is
overexpressed in certain tumor cells.7 It inhibits apoptosis
induced by a variety of stimuli.8 cIAP1 contains (i) three
baculoviral IAP repeat (BIR) domains which interact with its
binding proteins, including caspases, and (ii) one really interest-
ing new gene (RING) finger domain involved in ubiquitin ligase
activity.7-9 cIAP1 promotes ubiquitination and proteasomal
degradation of its binding proteins (Scheme 1a).9 Furthermore,
Naito’s group reported that a class of bestatin ester analogues
represented by methyl bestatin (MeBS, 2) bind to the BIR3
domain of cIAP1 and promote autoubiquitination and degrada-
tion of cIAP1 (Scheme 1b).10 Based on these observations, we
hypothesized that a hybrid molecule consisting of MeBS (2)
coupled to a ligand for a target protein might be able to induce
cIAP1-mediated ubiquitination and proteasomal degradation of
the target protein.
For a proof-of-concept study, we chose cellular retinoic acid
binding proteins (CRABP-I and -II) as target proteins. These
proteins reside in cytoplasm and specifically bind to all-trans
retinoic acid (ATRA, 3), an endogenous ligand of retinoic acid
receptors (RARs).11 CRABP-I is thought to be related to
metabolism of retinoic acid (RA) and resistance to RA in cancer
cells,12 while CRABP-II is suggested to be associated with
nuclear transportation of RA.13 It has also been reported that
(5) A non-peptide protac, utilizing murine double minute 2 (MDM2) as
an E3 ligase and a non-peptide ligand (nutlin-3) as a MDM2 ligand,
degraded androgen receptor (AR) (ref 4c). However, MDM2 naturally
induces ubiquitination of AR when Nutlin-3 binds to MDM2. Gaughan,
L.; Logan, I. R.; Neal, D. E.; Robson, C. N. Nucleic Acids Res. 2005,
33, 13–26. Logan, I. R.; McNeill, H. V.; Cook, S.; Lu, X.; Lunec, J.;
Robson, C. N. Prostate 2007, 67, 900–906. Therefore, additional data
to compare AR levels with nutlin-3 treatment and protac treatment
would be needed to determine whether the decrease of AR is associated
with the nutlin-3 derivative or the protac.
(9) Vaux, D. L.; Silke, J. Nat. ReV. Mol. Cell. Biol. 2005, 6, 287–297.
(10) (a) Sekine, K.; Takubo, K.; Kikuchi, R.; Nishimoto, M.; Kitagawa,
M.; Abe, F.; Nishikawa, K.; Tsuruo, T.; Naito, M. J. Biol. Chem. 2008,
283, 8961–8968. (b) Sato, S.; Aoyama, H.; Miyachi, H.; Naito, M.;
Hashimoto, Y. Bioorg. Med. Chem. Lett. 2008, 18, 3354–3358.
(11) (a) Donovan, M.; Olofsson, B.; Gustafson, A. L.; Dencker, L.;
Eriksson, U. J. Steroid Biochem. Mol. Biol. 1995, 53, 459–465. (b)
Fogh, K.; Voorhees, J. J.; Astro¨m, A. Arch. Biochem. Biophys. 1993,
300, 751–755.
(6) (a) Schneekloth, J. S.; Fonseca, F. M.; Koldobskiy, M.; Mandal, A.;
Deshaies, R.; Sakamoto, K. M.; Crews, C. M. J. Am. Chem. Soc. 2004,
126, 3748–3754. (b) Rodriguez-Gonzalez, A.; Cyrus, K.; Salcius, M.;
Kim, K.; Crews, C. M.; Deshaies, R. J.; Sakamoto, K. M. Oncogene
2008, 27, 7201–7211. (c) Lee, H.; Puppala, D.; Choi, E. Y.; Swanson,
H.; Kim, K. B. ChemBioChem 2007, 8, 2058–2062. (d) Puppala, D.;
Lee, H.; Kim, K. B.; Swanson, H. I. Mol. Pharmacol. 2008, 73, 1064–
1071. (e) Tang, Y. Q.; Han, B. M.; Yao, X. Q.; Hong, Y.; Wang, Y.;
Zhao, F. J.; Yu, S. Q.; Sun, X. W.; Xia, S. J. Asian J. Androl. 2009,
11, 119–126.
(12) (a) Fiorella, P. D.; Napoli, J. L. J. Biol. Chem. 1991, 266, 16572–
16579. (b) Kizaki, M.; Ueno, H.; Matsushita, H.; Takayama, N.; Muto,
A.; Awaya, N.; Ikeda, Y. Leuk. Lymphoma 1997, 25, 427–434.
(13) (a) Schug, T. T.; Berry, D. C.; Toshkov, I. A.; Cheng, L.; Nikitin,
A. Y.; Noy, N. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 7546–7551.
(b) Schug, T. T.; Berry, D. C.; Shaw, N. S.; Travis, S. N.; Noy, N.
Cell 2007, 129, 723–733. (c) Budhu, A. S.; Noy, N. Mol. Cell. Biol.
2002, 22, 2632–2641.
(7) (a) Deveraux, Q. L.; Reed, J. C. Genes DeV. 1999, 13, 239–252. (b)
Salvesen, G. S.; Duckett, C. S. Nat. ReV. Mol. Cell Biol. 2002, 3,
401–410.
(8) Roy, N.; Deveraux, Q. L.; Takahashi, R.; Salvesen, G. S.; Reed, J. C.
EMBO J. 1997, 16, 6914–6925.
9
J. AM. CHEM. SOC. VOL. 132, NO. 16, 2010 5821

A R T I C L E S
Itoh et al.
CRABP-I is related to Alzheimer’s disease14 and CRABP-II is
associated with neuroblastoma,15 Wilms tumor,15 and head and
neck squamous cell carcinoma (HNSCC).16 Therefore, CRABPs
could be target proteins for treatment of these diseases. However,
compounds which directly control the function(s) of CRABPs
have never been reported, and the biological/physiological roles
of these proteins remain unclear. Thus, small molecules that
suppress the function of CRABPs would be of great interest,
not only as tools for probing the biological/physiological roles
of CRABPs but also as potential therapeutic agents for Alzhe-
imer’s disease and cancer.
Herein, we describe a protein knockdown approach targeting
CRABPs focusing on (i) design and synthesis of hybrid
molecules consisting of MeBS (2) and ATRA (3) connected
via a spacer, (ii) CRABPs-degrading activity of the hybrid
molecules, (iii) the mechanism through which CRABP-degrada-
tion is induced by the hybrid molecules, and (iv) inhibitory
activity of the hybrid molecules on tumor invasion in cell-based
assay.
Results and Discussion
Figure 1. Structures of bestatin (1), MeBS (2), ATRA (3), and comp-
ounds 4.
Molecular Design. We speculated that the E3 ligase activity
of cIAP1 might be harnessed for targeted protein degradation.
In other words, if cIAP1 and a target protein could be induced
to form an artificial (nonphysiological) complex, the target
protein should be ubiquitinated by cIAP1, as cIAP1 ubiquitinates
proteins that are bound to it. Thus, we hypothesized that hybrid
small molecules, i.e., conjugates of ligands for target proteins
with membrane-permeable MeBS (2), might induce selective
degradation of the target proteins (Scheme 1c). Such hybrid
small molecules were expected to strictly mimic the cognate
complex for ubiquitination and to have sufficient membrane
permeability for practical application.
To test this idea, we designed the hybrid molecules 4,
consisting of MeBS (2) coupled via a spacer moiety to ATRA
(3), the specific ligand of CRABPs. The spacer was linked to
the ester position of MeBS (2) and to the C4 position of ATRA
(3). These positions were selected on the basis of our previous
studies showing that (i) introduction of a bulky substituent at
the ester moiety of MeBS (2) does not affect the binding affinity
for cIAP1,10b (ii) introduction of a substituent at the C4 position
of ATRA (3) does not affect the binding affinity for CRABP,17
and (iii) the latter modification causes loss of ATRA’s binding
affinity for RARs (see our previous report and X-ray structure
in the Supporting Information, Figure S1).18 Observations (ii)
and (iii) suggest that 4 would selectively bind to CRABPs, but
not to RARs. Since the length of the spacer is likely to influence
the efficiency of ubiquitination of target proteins, we designed
and synthesized three compounds 4a-c (Figure 1) with different
spacer lengths.
Synthesis. The designed compounds were synthesized as
illustrated in Scheme 2. Compound 5 was prepared from ATRA
(3) by esterification with 2-cyanoethyl alcohol, oxidation with
excess MnO2, and oximation with O-(carboxylmethyl)hydroxyl-
amine.18d,19 Compound 6 was prepared by protection of the
amino group of bestatin (1). Condensation of compound 6 with
alcohols 7a-c20 in the presence of 1-ethyl-3-(3-dimethylami-
nopropyl)carbodiimide (EDCI) and 1-hydroxybenzotriazole
hydrate (HOBt ·H2O) afforded esters 8a-c. Amines 9a-c were
obtained from compounds 8a-c by removal of the Boc group
under acidic conditions. Amidation of compound 5 and amines
9a-c gave amides 10a-c in good yield. Deprotection of the
9-fluorenylmethyloxycarbonyl (Fmoc) groups and 2-cyanoethyl
groups of compounds 10a-c with tetrabutylammonium fluoride
(TBAF)21 and 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU)22 gave
compounds 4a-c.
Decrease of CRABPs by the Synthesized Compounds. MOLT-4
cells express cIAP1 and CRABP-I (but not CRABP-II), and
HT1080 cells express cIAP1 and CRABP-II (but not CRABP-
I), as examined by Western blotting analysis (Supporting
Information, Figure S2). Therefore, we examined the effects of
4 on the levels of CRABP-I in MOLT-4 cells and CRABP-II
in HT1080 cells by Western blot analysis. As shown in Figure
2a, compounds 4 induced a dose-dependent decrease of
CRABP-I protein in MOLT-4 cells. Compound 4b seemed to
be the most potent, and 4a seemed to be the least potent among
(14) Uhrig, M.; Brechlin, P.; Jahn, O.; Knyazev, Y.; Weninger, A.; Busia,
L.; Honarnejad, K.; Otto, M.; Hartmann, T. BMC Med. 2008, 6, 38–
49.
(15) (a) Gupta, A.; Williams, B. R.; Hanash, S. M.; Rawwas, J. Cancer
Res. 2006, 66, 8100–8108. (b) Gupta, A.; Kessler, P.; Rawwas, J.;
Williams, B. R. Exp. Cell Res. 2008, 314, 3663–3668.
(16) Vo, H. P.; Crowe, D. L. Anticancer Res. 1998, 18, 217–224.
(17) Chapman, J. M.; Curley, R. W., Jr. Protein Expr. Purif. 1990, 1, 63–
69.
(19) Patel, J. B.; Huynh, C. K.; Handratta, V. D.; Gediya, L. K.; Brodie,
A. M.; Goloubeva, O. G.; Clement, O. O.; Nanne, I. P.; Soprano, D. R.;
Njar, V. C. J. Med. Chem. 2004, 47, 6716–6729.
(20) (a) Dekker, F. J.; de Mol, N. J.; van Ameijde, J.; Fischer, M. J. E.;
Ruijtenbeek, R.; Redegeld, F. A. M.; Liskamp, R. M. J. ChemBioChem
2002, 3, 238–242. (b) Lebeau, L.; Oudet, P.; Mioskowski, C. HelV.
Chim. Acta 1991, 74, 1697–1706.
(18) (a) Crettaz, M.; Baron, A.; Siegenthaler, G.; Hunziker, W. Biochem.
J. 1990, 272, 391–397. (b) Kleywegt, G. J.; Bergfors, T.; Senn, H.;
Le, M. P.; Gesell, B.; Shudo, K.; Jones, T. A. Structure 1994, 2, 1241–
1258PDB ID 1CBS. (c) Renaud, J. P.; Rochel, N.; Ruff, M.; Vivat,
V.; Chambon, P.; Gronemeyer, H.; Moras, D. Nature 1995, 378, 681–
689PDB ID 2LBD. (d) Shimazawa, R.; Sanda, R.; Mizoguchi, H.;
Hashimoto, Y.; Iwasaki, S.; Tanaka, H.; Kagechika, H.; Shudo, K.
Biochem. Biophys. Res. Commun. 1991, 179, 259–265.
(21) Kita, Y.; Maeda, H.; Takahashi, F.; Fujui, S.; Ogawa, T.; Hatayama,
K. Chem. Pharm. Bull. 1994, 42, 147–50.
(22) Bernardes, G. J.; Grayson, E. J.; Thompson, S.; Chalker, J. M.; Errey,
J. C.; El Oualid, F.; Claridge, T. D.; Davis, B. G. Angew. Chem., Int.
Ed. 2008, 47, 2244–2247.
9
5822 J. AM. CHEM. SOC. VOL. 132, NO. 16, 2010

Protein Knockdown Using Methyl Bestatin-Ligand Molecules
Scheme 2. Synthesis of Compounds 4a
A R T I C L E S
a Reagents and conditions: (a) 2-cyanoethyl alcohol, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDCI), N,N-dimethylaminopyridine (DMAP), CH2Cl2,
room temperature, 18 h, 90%; (b) MnO2, CH2Cl2, room temperature, 22 h, 36%; (c) O-(carboxylmethyl)hydroxylamine, pyridine, room temperature, 17 h,
100%; (d) 9-fluorenylmethyloxycarbonyl chloride (FmocCl), K2CO3, THF, H2O, room temperature, 24 h, 97%; (e) EDCI,1-hydroxybenzotriazole hydrate
(HOBt·H2O), i-Pr2NEt, CH2Cl2, room temperature, 24 h, 14-51% (from 6); (f) HCl, 1,4-dioxane, room temperature, 1 h, quant; (g) EDCI, HOBt ·H2O,
Et3N, CH2Cl2, room temperature, 14 h; (h) tetrabutylammonium fluoride (TBAF), MeOH, THF, room temperature, 1 h; (i) 1,8-diazabicyclo[5.4.0]undec-
7-ene (DBU), n-C12H25SH, CH2Cl2, room temperature, 3 h, 25-46% (three steps from 5).
of 4a and 4c suggests that the longer linker is better for down-
regulation of CRABP-I, whereas the shorter linker is better for
down-regulation of CRABP-II. The differences between CRABP-I
and CRABP-II responses to treatment with 4a and 4c presum-
ably reflect the structural differences between CRABP-I and
-II.
Mechanism of Decrease in CRABPs. We next investigated
the mechanism by which CRABPs are downregulated. First,
we prepared HT1080 cells expressing FLAG-tagged cIAP1 and
examined the influence of 4 on cIAP1 level by Western blot
analysis, since esterified analogues of bestatin (1) induce
autoubiquitination of cIAP1. All the compounds 4a-c showed
a dose-dependent cIAP1-decreasing effect, but their efficacy was
lower than that of MeBS (2) (Figure 3a, Supporting Information,
Figure S3). The reason for the weaker activity of 4 compared
with MeBS (2) is not clear, but it might be due to the bulky
ester group or to competing ubiquitination activity toward
CRABP-II and cIAP1 itself. A decrease in cIAP1 level was also
seen in MOLT-4 cells (Supporting Information, Figure S4).
Figure 2. Down-regulation of CRABP-I and CRABP-II by treatment with
compounds 4a-c: (a) Western blot detection of CRABP-I levels in MOLT-4
cells after 16 h treatment; (b) Western blot detection of CRABP-II levels
in HT1080 cells after 6 h treatment.
the three compounds, though the differences were not large.
Compounds 4 also induced a dose-dependent decrease of
CRABP-II in HT1080 cells. In this case, 4a and 4b strongly
reduced the level of CRABP-II (4b seemed to be more potent
than 4a), while 4c was relatively ineffective (Figure 2b). These
results suggested that the linker with n ) 2 is preferable among
those investigated. A possible interpretation of this result would
be that the structure of compound 4b is most suitable to bring
the ubiquitination sites in CRABPs and the RING domain of
cIAP1 into an appropriate spatial relationship for effective
ubiquitination to occur. The difference in CRABP-I/II selectivity
Next, we pretreated the cells with an excess amount of MeBS
(2) to investigate whether the reduction of CRABP-II by 4 was
mediated by cIAP1 (Figure 3b). The pretreatment with an excess
amount of MeBS (2) resulted in complete disappearance of
cIAP1 but did not influence the CRABP-II level (lane 2). Both
CRABP-II and cIAP1 levels were decreased by treatment with
4b [without pretreatment with MeBS (2)] as mentioned above
(lane 3). On the other hand, compound 4b had no effect on
CRABP-II levels in cells pretreated with 1000 µM MeBS (2)
(lane 4). These results indicate that 4b reduces the CRABP-II
9
J. AM. CHEM. SOC. VOL. 132, NO. 16, 2010 5823

A R T I C L E S
Itoh et al.
Figure 3. Western blot detection of CRABP-II and cIAP1 levels in HT1080 cells expressing FLAG-tagged cIAP1: (a) CRABP-II and cIAP1 levels after
6 h treatment with 4; (b) influence of pretreatment of MeBS (2) on CRABP-II degradation induction. The cells were treated with 1 µM 4b for 3 h. MeBS
(2) (1000 µM) was added to the culture 1 h prior to the addition of 4b.
Figure 4. Western blot detection of GST-BIR3, GST, and CRABP-II levels
of samples prepared by pull-down assay in vitro: lane 1, CRABP-II levels;
lane 2, mixture of GST-BIR3 and CRABP-II; lane 3, mixture of GST,
Figure 6. Influence of combination of MeBS (2) and ATRA (3). Western
CRABP-II, and 4b; lane 4, mixture of GST-BIR3, CRABP-II, and 4b.
blot detection of CRABP-II and cIAP1 levels in HT1080 cells expressing
FLAG-tagged cIAP1. The cells were treated with 1 µM MeBS (2), 1 µM
ATRA (3), or 1 µM 4b for 6 h.
or MOLT-4 cells were treated with 10 µM MeBS (2) and ATRA
(3) (Supporting Information, Figure S5). Thus, conjunction of
MeBS (2) and ATRA (3) within a single molecule is essential
for CRABP degradation-inducing activity.
It was reported that ATRA (3) influences CRABPs levels by
transactivation of RAR23 or that ATRA (3) shows effects on
Figure 5. Influence of pretreatment with proteasome inhibitors on CRABP-
II degradation induction. Western blot detection of CRABP-II and cIAP1
HL-60 cell differentiation.24 However, 4b did not show agonistic
levels in HT1080 cells expressing FLAG-tagged cIAP1. The cells were
activity toward RARs in reporter gene assay (Supporting
Information, Figure S6). Compound 4b did not show effects
on HL-60 cell differentiation or did not enhance cell differentia-
tion induced by ATRA (3) (data not shown). Addtionally, 4b
did not induce degradation of RARR (Supporting Information,
Figure S7) and did not bind to RARs (Supporting Information,
Figure S8). Thus, 4b showed selective degradation-inducing
activity for CRABPs, but not RARR or ꢀ-actin, as judged from
Western blot experiments with equal loadings of total protein.
These results indicated that 4b has high specificity for degrada-
tion of CRABPs. All the results are consistent with our
hypothesis that the hybrid molecules 4 form an artificial ternary
complex with cIAP1 and CRABP, in which cIAP1 ubiquinates
CRABP, leading to its degradation by proteasome.
These small molecules 4 have sufficient membrane permeability
and stability to be used in cell systems and are effective in low
concentration. Therefore, they might be suitable for a variety of
studies in cells and/or animals. In addition, this protein knockdown
strategy might be generally adaptable to a wide range of proteins
by replacing ATRA (3) with a specific ligand for the target protein.
Therefore, protein knockdown could be a simple and easy technique
to complement genetic ablation at the DNA level (gene knockout)
or the mRNA level (gene knockdown).
treated with 1 µM 4b for 6 h. MG132 (10 µM) and lactacystin (10 µg/mL)
were added to the culture 30 min prior to the addition of 4b.
level in cIAP1-expressing cells but not in cIAP1-depleted cells,
suggesting that cIAP1 mediates the ubiquitination of CRABP-
II protein for proteasomal degradation in these cells.
We next examined the formation of a ternary complex
consisting of 4b, cIAP1, and CRABP-II by means of GST pull-
down assay using the GST-tagged BIR3 domain of cIAP1, to
which MeBS (2) binds. As shown in Figure 4, GST-BIR3
coprecipitated CRABP-II in the presence of 4b (lane 4) but not
in the absence of 4b (lane 2). GST did not pull down CRABP-
II even in the presence of 4b (lane 3). This result indicates that
CRABP-II is held in proximity to cIAP1 by 4b, as we had
hoped.
Next, we investigated the influence of proteasome inhibitors
on the CRABP-II level in 4b-treated cells. (Figure 5). The
reduction of CRABP-II by 4b was blocked by proteasome
inhibitors MG132 and lactacystin. Thus, the decrease of
CRABP-II induced by 4b can be attributed to proteasomal
degradation.
In addition, we confirmed that the decrease in CRABPs was
not caused by a partial structure of 4 or by a mere mixture of
MeBS (2) and ATRA (3) (Figure 6). Compound 4b induced a
decrease of CRABP-II, whereas the mixture of MeBS (2) and
ATRA (3) did not. The treatments with 1 µM MeBS (2), and
the combination of 1 µM MeBS (2) and 1 µM ATRA (3)
decreased the cIAP1 level but did not affect the CRABPs level
in HT1080. ATRA (3) at 1 µM did not cause any decrease of
cIAP1 or CRABP. Similar results were obtained when HT1080
(23) (a) Astro¨m, A.; Tavakkol, A.; Pettersson, U.; Cromie, M.; Elder, J. T.;
Voorhees, J. J. J. Biol. Chem. 1991, 266, 17662–17666. (b) Geisen,
C.; Denk, C.; Gremm, B.; Baust, C.; Karger, A.; Bollag, W.; Schwarz,
E. Cancer Res. 1997, 57, 1460–1467. (c) Dutta, A.; Sen, T.; Banerji,
A.; Das, S.; Chatterjee, A. J. Oncol. 2009, 2009, 627840.
(24) Kim, S. H.; Kim, S. W.; Choi, S. J.; Kim, Y. C.; Kim, T. S. Bioorg.
Med. Chem. 2006, 19, 6752–6757.
9
5824 J. AM. CHEM. SOC. VOL. 132, NO. 16, 2010

Protein Knockdown Using Methyl Bestatin-Ligand Molecules
A R T I C L E S
Figure 7. Suppression of cell migration by compound 4b and degradation of CRABP-II in IMR-32 cells. (a) Relative IMR-32 cell migration. Results are
presented as means ( SEM (performed in triplicate). (b) Western blot detection of CRABP-II and cIAP1 levels in IMR-32 cells after 24 h treatment with
each reagent.
tumor,28 HNSCC,16 ovarian cancer,29 gastric cancer,30 uterine
Inhibitory Activity on Tumor Invasion in Cell-Based Assay.
Finally, we tested the effect of 4b on neuroblastoma cells to
explore the potential clinical usefulness of CRABP-II degrada-
tion inducers. It has been reported that CRABP-II is closely
associated with development of neuroblastoma, Wilms tumor
and HNSCC, and a reduction of CRABP-II level suppresses
the migration of tumor cells.15,16 Therefore, we evaluated the
migration-suppressing activity of 4b toward human neuroblas-
toma IMR-32 cells (Figure 7). Treatment of the IMR-32 cells
with 30 µM MeBS (2) slightly reduced the cell migration by
approximately 30%, in agreement with previous reports that
MeBS (2) shows antitumor activity.10a,25 Interestingly, treatment
of the cells with ATRA (3) alone or the combination of MeBS
(2) and ATRA (3) enhanced the cell migration by 30-40%,
possibly as a result of up-regulation of CRABP-II expression
by ATRA (3). As expected, treatment of the cells with 10 and
30 µM 4b remarkably reduced the cell migration by ap-
proximately 75% and 95%, respectively. Moreover, the extent
of migration inhibition was well correlated with the CRABP-II
level in the cells. MeBS (2) did not affect CRABP-II levels in
IMR-32 cells, while ATRA (3) or the combination of MeBS
(2) and ATRA (3) up-regulated CRABP-II levels. Since
compound 4b induced degradation of CRABP-II, it might be a
useful tool for studying the function(s) of CRABP-II.
leiomyoma,31 and melphalan- or phorbol-ester-resistant cell
lines,32 suggesting that it may play a role in cancer development.
In Wilms tumor, CRABP-II overexpression has been reported
to correlate with poor clinical outcome.15 Here, we found that
the small-molecular CRABP-II degradation inducer 4b inhibited
migration of neuroblastoma cells. Thus, CRABP-II degradation
inducers may be effective for therapy of neuroblastoma and
other CRABP-II-overexpressing cancers. Further studies on
various types of cancer cells, including cell proliferation assay,
are in progress.
The strategy described in this paper might also be adaptable
to a range of cancer-related proteins by replacing ATRA (3)
with specific ligands for the target proteins. In addition,
suppression of cIAP1 function is thought to be favorable for
cancer treatment,33 and disruption of the cIAP1 gene in mice
results in no obvious abnormality.34 Therefore, suppression or
degradation of cIAP1, which is overexpressed in several human
cancers,7 should not vitiate the anticancer effect. On the other
hand, protacs have utilized peptides recognized by two kinds
of E3 ligase complex, von Hippel-Lindau tumor suppressor
(VHL), which is deficient or mutated in several cancer,35 and
(28) (a) Li, W.; Kessler, P.; Williams, B. R. Oncogene 2005, 24, 457–
468. (b) Li, C. M.; Guo, M.; Borczuk, A.; Powell, C. A.; Wei, M.;
Thaker, H. M.; Friedman, R.; Klein, U.; Tycko, B. Am. J. Pathol.
2002, 160, 2181–2190.
Targeted cancer therapy blocks the growth of cancer cells
by interfering with specific molecules required for carcinogen-
esis and tumor growth, rather than by simply interfering with
rapid cell division, as is done with traditional chemotherapy.26
Targeted cancer therapies may be more effective than traditional
chemotherapy and less harmful to normal cells. Many kinds of
target proteins related to cancer have been identified so far.
Further, small-molecular inhibitors for target enzymes or
modulators for target receptors have been discovered, and some
of them are used for targeted cancer therapy. But many cancer-
related proteins, including CRABP-II, can not currently be
functionally regulated with small molecules. CRABP-II is
expressed in several cancers, including breast cancer,27 Wilms
(29) Hibbs, K.; Skubitz, K. M.; Pambuccian, S. E.; Casey, R. C.; Burleson,
K. M.; Oegema, T. R. J.; Thiele, J. J.; Grindle, S. M.; Bliss, R. L.;
Skubitz, A. P. Am. J. Pathol. 2004, 165, 397–414.
(30) Nishigaki, M.; Aoyagi, K.; Danjoh, I.; Fukaya, M.; Yanagihara, K.;
Sakamoto, H.; Yoshida, T.; Sasaki, H. Cancer Res. 2005, 65, 2115–
2124.
(31) Tsibris, J. C.; Segars, J.; Coppola, D.; Mane, S.; Wilbanks, G. D.;
O’Brien, W. F.; Spellacy, W. N. Fertil. Steril. 2002, 78, 114–121.
(32) Samuel, S.; Bernstein, L. R. Mol. Carcinog. 2004, 39, 34–60.
(33) (a) Imoto, I.; Yang, Z. Q.; Pimkhaokham, A.; Tsuda, H.; Shimada,
Y.; Imamura, M.; Ohki, M.; Inazawa, J. Cancer Res. 2001, 61, 6629–
6634. (b) Imoto, I.; Tsuda, H.; Hirasawa, A.; Miura, M.; Sakamoto,
M.; Hirohashi, S.; Inazawa, J. Cancer Res. 2002, 62, 4860–4866. (c)
Dai, Z.; Zhu, W. G.; Morrison, C. D.; Brena, R. M.; Smiraglia, D. J.;
Raval, A.; Wu, Y. Z.; Rush, L. J.; Ross, P.; Molina, J. R.; Otterson,
G. A.; Plass, C. Hum. Mol. Genet. 2003, 12, 791–80. (d) Zender, L.;
Spector, M. S.; Xue, W.; Flemming, P.; Cordon-Cardo, C.; Silke, J.;
Fan, S. T.; Luk, J. M.; Wigler, M.; Hannon, G. J.; Mu, D.; Lucito, R.;
Powers, S.; Lowe, S. W. Cell 2006, 125, 1253–1267.
(25) Sekine, K.; Fujii, H.; Abe, F.; Nishikawa, K. Int. J. Cancer. 2001,
94, 485–491.
(26) Schultz, R. M., Ed; AdVances in Targeted Cancer Therapy; Birkhae-
user: Basel, 2005.
(34) Conze, D. B.; Albert, L.; Ferrick, D. A.; Goeddel, D. V.; Yeh, W. C.;
Mak, T.; Ashwell, J. D. Mol. Cell. Biol. 2005, 25, 3348–3356.
(35) (a) Hon, W.; Wilson, M. I.; Harlos, K.; Claridge, T. D. W.; Schofield,
C. J.; Pugh, C. W.; Mazwell, P. H.; Ratcliffe, P. J.; Stuart, D. I.; Jones,
E. Y. Nature 2002, 417, 975–978. (b) Kapitsinou, P. P.; Haase, V. H.
Cell Death Differ. 2008, 15, 650–659. (c) Gnarra, J. R.; et al. Nat.
Genet. 1994, 7, 85–90.
(27) (a) Korkola, J. E.; DeVries, S.; Fridlyand, J.; Hwang, E. S.; Estep,
A. L.; Chen, Y. Y.; Chew, K. L.; Dairkee, S. H.; Jensen, R. M.;
Waldman, F. M. Cancer Res. 2003, 63, 7167–7175. (b) Bertucci, F.;
Houlgatte, R.; Benziane, A.; Granjeaud, S.; Ade´la¨ıde, J.; Tagett, R.;
Loriod, B.; Jacquemier, J.; Viens, P.; Jordan, B.; Birnbaum, D.;
Nguyen, C. Hum. Mol. Genet. 2000, 9, 2981–2991.
9
J. AM. CHEM. SOC. VOL. 132, NO. 16, 2010 5825

A R T I C L E S
Itoh et al.
ꢀ-transducin repeat-containing protein (ꢀ-TRCP) Skp1-Cullin-F
box (SCF) complex (SCFꢀ-TRCP), which plays important roles
in many tissues.36 Those peptide-based protacs, especially those
This strategy for protein knockdown should be widely
applicable by replacing ATRA (3) with specific ligands for other
target proteins and should provide a methodology to complement
gene knockout and gene knockdown techniques. It is expected
to be useful for probing biological functions of proteins and
might also be applicable for targeted cancer therapy using MeBS
(2) conjugated with ligands for cancer-related proteins.
utilizing SCFꢀ-TRCP 4a,b seem best suited to cell-free systems and
,
might be difficult to use in cancer treatment due to stability
issues. Evaluation of our protein degradation inducers, MeBS
(2)-ligand hybrid molecules, for targeted cancer therapy seems
warranted.
Acknowledgment. The work described in this paper was
partially supported by Grants-in-Aid for Scientific Research from
The Ministry of Education, Culture, Sports, Science and Technol-
ogy, Japan, and the Japan Society for the Promotion of Science.
We are grateful to Nippon Kayaku Co., especially Dr. Keiko Sekine,
for providing bestatin (1) and MeBS (2), to Dr. Yukihide Tomari,
Dr. Tomoko Kawamata, Mr. Shintaro Iwasaki, and Mr. Shinichi
Sato for their help with Western blot detection, to Prof. Makoto
Makishima with providing RAR plasmids, and to Dr. Tomomi
Noguchi-Yachide for HL-60 differentiation assay.
Conclusion
We designed and synthesized CRABPs degradation inducers
4, in which a ligand for CRABPs is conjugated with MeBS (2)
via a spacer, aiming to make use of the ubiquitin E3 ligase
activity of cIAP1. Compounds 4 induced down-regulation of
CRABPs in cells. Our results indicate that these small molecules
induce the formation of an artificial (nonphysiological) ternary
complex of cIAP1 and CRABPs, leading to degradation of
CRABPs via the ubiquitin-proteasome pathway. Compound 4b
has sufficient activity, permeability, and stability for use in
cellular systems, and we confirmed that it inhibited migration
of neuroblastoma IMR-32 cells.
Supporting Information Available: Experimental procedure,
1
copies of H NMR spectra, Figures S1-S8, and complete ref
35c. This material is available free of charge via the Internet at
http://pubs.acs.org.
(36) Deshaies, R. J. Annu. ReV. Cell DeV. Biol. 1999, 15, 435–467.
JA100691P
9
5826 J. AM. CHEM. SOC. VOL. 132, NO. 16, 2010