Inhibitory effect of a dimerization-arm-mimetic peptide on EGF receptor activation

Article abstract of DOI:10.1016/j.bmcl.2009.04.080

A cyclic decapeptide was chemically synthesized that mimics the loop structure of a β-hairpin arm of the EGF receptor, which is highly involved in receptor dimerization upon activation by ligand binding. This peptide was revealed to reduce dimer formation

Full text of DOI:10.1016/j.bmcl.2009.04.080

Bioorganic & Medicinal Chemistry Letters 19 (2009) 3279–3282
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Inhibitory effect of a dimerization-arm-mimetic peptide on EGF
receptor activation
Takaaki Mizuguchi a, Hiromasa Uchimura b, Taeko Kakizawa a, Tooru Kimura a, Shigeyuki Yokoyama c,d
,
Yoshiaki Kiso a, Kazuki Saito b,c,
*
a Department of Medicinal Chemistry, Center for Frontier Research in Medicinal Science and 21st Century COE Program, Kyoto Pharmaceutical University, Yamashina-ku,
Kyoto 607-8412, Japan
b Laboratory of Proteomic Sciences, 21st Century COE Program, Kyoto Pharmaceutical University, Yamashina-ku, Kyoto 607-8412, Japan
c RIKEN Systems and Structural Biology Center, Tsurumi, Yokohama 230-0045, Japan
d Department of Biophysics and Biochemistry, Graduate School of Science, University of Tokyo, Bunkyo-ku, Tokyo 113-0033, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
A cyclic decapeptide was chemically synthesized that mimics the loop structure of a b-hairpin arm of the
EGF receptor, which is highly involved in receptor dimerization upon activation by ligand binding. This
peptide was revealed to reduce dimer formation of the receptor in a detergent-solubilized extract of epi-
Received 8 January 2009
Revised 15 April 2009
Accepted 20 April 2009
Available online 23 April 2009
dermoid carcinoma A431 cells and to inhibit receptor autophosphorylation at less than 10
intact cells.
lM in the
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
A431 cells
Autophosphorylation
Dimerization-arm-mimetic peptide
Dimerization inhibitor
EGF receptor
In response to ligand binding, epidermal growth factor (EGF)
receptor on the cellular membrane is activated, causing prolifera-
tion and differentiation of various cells. The overexpression and
unregulated activation of the receptor are known to initiate can-
cers in cells, as often observed in lung carcinoma.1–4 Therefore,
the receptor has recently become a good medicinal target for
designing anti-cancer drugs. For example, gefitinib (IressaÒ) has
been developed as a molecular-targeted agent to directly inhibit
the intracellular tyrosine kinase of the receptor.5,6 In addition, anti-
body drugs against the receptor have been developed, such as
cetuximab (ErbituxÒ), which is a monoclonal antibody recognizing
the extracellular ligand-binding region of the receptor.7–9 How-
ever, the use of such agents is hindered by some serious problems,
such as side effects10 and the exorbitant medication cost,11,12
although they have shown certain efficacy in cancer therapeutics.
A new type of inhibitor against the EGF receptor is now eagerly
anticipated, to further improve the quality of life of cancer patients.
The EGF receptor is a single-spanning membrane protein on the
cell surface. Ligand binding to the receptor extracellular region
dramatically changes its conformation, and consequently drives
the receptor molecules into dimers. The receptor dimerization
successively causes intracellular autophosphorylation, in which
the kinase transphosphorylates tyrosine residues in the C-terminal
tail of the dimerization partner.13–15 Since dimerization may be
one of the most crucial steps in receptor activation, in this study,
we designed and synthesized a peptide that would prevent the
dimerization and inhibit the autophosphorylation of the EGF
receptor.
The structure of the EGF receptor extracellular region, dimer-
ized by binding the EGF ligand, was previously elucidated by
X-ray crystallography (PDB ID: 1IVO) (Fig. 1).16 The extracellular
region of the receptor consists of four domains (I–IV), and in the
dimerized structure, the interface between the receptor monomers
is mostly occupied by domain II. Especially, a b-hairpin arm (ami-
no-acid residues 242–259) protruding from this domain seems to
be most responsible for the dimerization, since a considerable
number of hydrogen-bonds are observed between the residues at
a turn head of the arm and those of the partner receptor molecule
(Fig. 1, inset). Therefore, the b-hairpin structure is called a ‘dimer-
ization arm’.16
To develop dimerization inhibitors against the EGF receptor, a
decapeptide mimicking the loop structure of the arm head,
(1), was designed. Peptide 1 has the amino-acid
CYNPTTYQMC
sequence of the turn head and two additional cysteine residues
at both the N- and C-termini, for cyclization with a disulfide bond
between them, and it was chemically synthesized by conventional
Fmoc-based solid-phase methodology. Namely, the first C-terminal
* Corresponding author. Tel.: +81 75 595 4714; fax: +81 75 595 4798.
E-mail address: saito@mb.kyoto-phu.ac.jp (K. Saito).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.04.080

3280
T. Mizuguchi et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3279–3282
Figure 1. Crystal structure of dimerized EGF receptor ectodomains (left) and the intermolecular hydrogen-bond network at the turn head of the dimerization arm (right
inset) (PDB ID: 1IVO). The dimer interface is mostly occupied by domain II, and several hydrogen-bonds are observed between the residues at the turn head of the protruding
dimerization arm and those of the partner receptor molecule. Four representative hydrogen-bonds are drawn in the inset: between Tyr246A and Cys283B, Thr249A and
Asn86B, Tyr251A and Arg285B, and Gln252A and Ala286B (superscript suffixes, A and B, on residue numbers indicate each of the receptor molecules in the dimer, with
monomer structures drawn in blue and green, respectively).
residue, Fmoc-
ride resin using diisopropylethylamine in 1,2-dichloroethane.
Fmoc-protected -amino acids (2.5 equiv) were successively
L
-Cys(Trt)-OH, was attached to 2-chlorotrityl chlo-
(peak b) was detected as well as the starting material without a
disulfide bond (peak a). The amount of the resultant peptide 1 grad-
ually increased, and the reaction was mostly completed within 3 h.
The peptide sequence was quite readily cyclized to yield the product
1. After purification by preparative HPLC, formation of the disulfide
bond was confirmed with MALDI-TOF-MS (monoisotopic mass:
found, 1243.50; calcd for [M+Na]+, 1243.42). As a reference com-
pound, a linear octapeptide lacking the cysteine residues, YNPT-
TYQM (2), was similarly prepared by the Fmoc-based solid-phase
synthesis. After cleavage from the resin and precipitation with cold
ether, the peptide was directly subjected to preparative HPLC purifi-
cation, without the oxidative cyclization step. The purified product
of 2 showed a quite acceptable mass value (monoisotopic mass:
found, 1017.36; calcd for [M+H]+, 1017.43).
First of all, the inhibitory activities of the synthetic peptides
against the EGF receptor dimerization were assessed. Formation
of the dimer was monitored by cross-linking of the receptor mole-
cules in the ligand-induced dimer state. In this study, as fully
described in Supplementary data, a hydrophilic, membrane-imper-
meant cross-linker, bis(sulfosuccinimidyl) suberate (BS3),17 was
employed, since Staros and his co-workers had successfully
trapped the EGF receptor dimer in extracts solubilized from human
epidermal carcinoma A431 cells by the use of this cross-linker.18
The carcinoma cells abundantly express the EGF receptor on their
cellular surface, and the receptor, in a Triton X-100-solubilized
membrane fraction of the cells, was easily detected as a protein
band at ꢀ170 kDa on SDS–PAGE gels, by immunoblotting with an
anti-EGFR antibody (Fig. 3A). In response to stimulation by the
ligand EGF, an additional band obviously appeared at ꢀ350 kDa,
corresponding to the molecular mass of the BS3-trapped receptor
dimer. The dimer band became slightly thinner, as larger amounts
of peptide 1 were co-incubated during the ligand stimulation,
indicating that the dimerization-arm-mimetic peptide reduced
the EGF-induced dimerization of the receptor. The degrees of
reduction in the dimer formation were quantified by densitometry
L
coupled with the aid of 1-hydroxybenzotriazole (2.5 equiv) and
diisopropylcarbodiimide (2.5 equiv) in DMF, and completion of
the coupling reactions was monitored by Kaiser’s ninhydrin test.
A treatment with 20% (v/v) piperidine was utilized for the removal
of the Fmoc group. Finally, the desired peptide was cleaved and re-
leased from the completed resin using a solution of trifluoroacetic
acid (TFA)–ethanedithiol–triisopropylsilane–H2O (94.5:2.5:1:2.5,
v/v) at room temperature for 90 min. After removal of the resin
with a glass filter, an excess volume of cold diethyl ether was
added to precipitate the cleaved peptide, which was then redis-
solved in H2O and lyophilized. A solution (0.1 mg/mL) of the lyoph-
ilized peptide in 100 mM NH4HCO3 (pH ꢀ8) was stirred in air at
room temperature, and oxidative cyclization of the peptide to yield
1 was monitored by analytical reversed-phase HPLC (Fig. 2). Even
immediately after the dissolution at 0 h, the cyclized peptide 1
Figure 2. Oxidative disulfide bond formation in peptide 1. The parent linear peptide
CYNPTTYQMC was dissolved in 100 mM NH4HCO3, and the solution (0.1 mg/mL)
was stirred in air at room temperature. The cyclization reaction was monitored by
analytical HPLC [Column, YMC-Pack ODS-A A-312 (6.0 Â 150 mm); solvent, 0.05%
TFA/CH3CN–H2O (gradient from 15:85 to 35:65 over a period of 40 min) at a flow
rate of 1.0 mL/min]. Even immediately after the dissolution of the starting material
at 0 h, both the parent peptide (peak a; m/z 1245.41) and the resultant cyclic one
(peak b; m/z 1243.50) were detected. The peak intensity of the product peptide 1 (b)
gradually increased at 1 h, and, finally, the peak of the parent peptide (a) had mostly
disappeared by 3 h. Calculated monoisotopic m/z values for [M+Na]+ ion of the
parent linear and product cyclic peptides are 1245.43 and 1243.42, respectively.
of the bands, and 1
dimer formation to less than 80% of that without the peptide
(Fig. 3B). Unfortunately, amounts larger than 1 M of peptide 1
generated only poorly-reproducible data with large deviations,
perhaps due to perturbation of the solubilized state of the receptor
itself, and thus complete quenching of the dimer formation could
not be accomplished by peptide 1. On the other hand, the linear
peptide 2 hardly inhibited the receptor dimerization in the concen-
tration range studied here (Fig. 3B). Since the inhibitory activity of
lM peptide 1 was revealed to decrease the
l

T. Mizuguchi et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3279–3282
3281
Figure 4. Inhibitory effect of peptide 1 on EGF receptor autophosphorylation in
intact A431 cells. The cells were serum-starved for 24 h in serum-free medium
containing 1 mg/mL BSA. After the starvation, they were stimulated with 20 ng/mL
EGF for 5 min in the presence of 0, 0.1, 1, or 10 lM peptide 1. Cell lysates were
Figure 3. Inhibitory effects of the cyclic (1) and linear (2) peptides on dimerization
of the EGF receptor in a solubilized membrane fraction prepared from human
epidermal carcinoma A431 cells. Solubilized extracts containing the EGF receptor
subjected to immunoprecipitation (IP) with an anti-EGFR antibody. Precipitates
were resolved by SDS–PAGE, transferred to a PVDF membrane, and visualized by
immunoblotting (IB) with an anti-phosphotyrosine antibody. Membranes were
reprobed with an anti-EGFR antibody to control for protein loading. (A) Typical
immunoblots for autophosphorylation assays of the EGF receptor. (B) Inhibitory
activities of peptide 1 against EGF receptor autophosphorylation. Data represent the
mean standard deviation values of at least three independent experiments, and
were normalized such that the positive control (EGF-stimulated but in the absence
of the peptide) represents 100% receptor autophosphorylation.
were incubated with or without 7
lg/mL EGF for 10 min in the presence of 0, 0.01,
0.1, or 1
l
M peptide 1 or 2. Then, the extracts were treated with 1 mM BS3 for
10 min to covalently cross-link the EGF receptor molecules in the dimer state. After
termination of the cross-linking reaction with 50 mM glycine, the complexes were
separated by SDS–PAGE, transferred to
a PVDF membrane, and visualized by
immunoblotting with an anti-EGFR antibody. (A) Typical immunoblots for inhibi-
tion assays of the EGF receptor dimerization by peptide 1. (B) Inhibitory activities of
peptides 1 (open bars) and 2 (gray bars) against the receptor dimerization. Relative
ratios of apparent band densities between receptor dimer and monomer are shown
as the mean standard deviation values of at least three independent experiments,
and were normalized such that the positive control (EGF-stimulated but in the
absence of the peptides) represents a ratio = 1.0.
Our simple strategy for the inhibition of EGF receptor activation
via that of the dimerization by the arm-mimetic cyclic peptide 1
seems quite successful, although the peptide-binding site on the
receptor should be confirmed in the future. The cyclic peptide
can bind to the extracellular region of the receptor, at least, since
it might not penetrate the hydrophobic cellular membrane to the
cytoplasm of A431 cells, but it obviously inhibited the receptor
autophosphorylation in the intact cells. On the other hand,
25–40-residue fragments of a negative regulatory protein of the
EGF receptor, MIG6, were reported to directly inhibit the receptor
autophosphorylation by preventing the interaction between a
phosphorylating kinase domain and a phosphorylated substrate
domain; that is, the asymmetric dimer formation of the intracellu-
lar kinase domain.20 However, it is still unknown whether the frag-
ments can inhibit the dimerization of not only the kinase domain
but also a ‘whole’ receptor molecule.
peptide 2 was considerably lower than that of peptide 1, the loop
structure of the arm-mimetic peptide, maintained by the disulfide
bond, may play an important role in retaining the affinity for the
EGF receptor molecule to competitively inhibit the receptor
dimerization.
Next, the cyclic peptide 1, which showed weak but certain
inhibitory activity against the receptor dimerization, was subjected
to another assay of receptor autophosphorylation, using intact
A431 cells. The cell-based assay was performed as described previ-
ously19 with minor modifications (for details of the assay proce-
dure, see Supplementary data). After the cells were stimulated
with EGF in the presence of an appropriate amount of the peptide,
the receptor was immunoprecipitated with an anti-EGFR antibody,
and autophosphorylation of the receptor was observed by immu-
noblotting with an anti-phosphotyrosine antibody (Fig. 4A). The
cyclic peptide 1 showed obvious inhibition of the receptor auto-
phosphorylation. Quantitative densitometric analyses of the
immunostained bands revealed that the residual autophosphoryla-
tion level of the EGF receptor was suppressed to 61%, 51%, and 30%,
Another cyclic peptide,
, was designed and
FCDGFYACYMDV
found to inhibit the activity of ErbB-2 (HER2), a homologue of
the EGF receptor, by Murali and his co-workers.21,22 However, in
contrast to this study, the sequence of their peptide was derived
from the structure of the CDR-H3 loop of the anti-p185HER2/neu
monoclonal antibody, 4D5 (herceptin; TrastuzumabÒ), and thus
the peptide is referred to as an anti-HER2/neu peptide (AHNP).
AHNP may bind domain IV, not domain II, because its parent
antibody binds domain IV of ErbB-2. When inhibiting the ErbB-2
activity, it does not compete with the dimerization arm of the
receptor, while our cyclic peptide might do so with that of the
by the addition of 0.1, 1, and 10 lM peptide 1, respectively, to the
culture medium (Fig. 4B). The dimerization-inhibiting peptide
absolutely reduced the EGF-stimulated receptor activation, in
terms of the intracellular autophosphorylation in living cells.

3282
T. Mizuguchi et al. / Bioorg. Med. Chem. Lett. 19 (2009) 3279–3282
EGF receptor. Therefore, peptide 1 is the first example of a peptide
that was designed based on the loop structure of the dimerization
arm and exhibits inhibitory activity against the EGF receptor.
In conclusion, the cyclic peptide 1 that mimics the loop of the
dimerization arm in the crystal structure was revealed to inhibit
References and notes
1. Hendler, F. J.; Ozanne, B. W. J. Clin. Invest. 1984, 74, 647.
2. Veale, D.; Ashcroft, T.; Marsh, C.; Gibson, G. J.; Harris, A. L. Br. J. Cancer 1987, 55,
513.
3. Sozzi, G.; Miozzo, M.; Tagliabue, E.; Calderone, C.; Lombardi, L.; Pilotti, S.;
Pastorino, U.; Pierotti, M. A.; Della Porta, G. Cancer Res. 1991, 51, 400.
4. Rusch, V.; Baselga, J.; Cordon-Cardo, C.; Orazem, J.; Zaman, M.; Hoda, S.;
McIntosh, J.; Kurie, J.; Dmitrovsky, E. Cancer Res. 1993, 53, 2379.
5. Ciardiello, F.; Caputo, R.; Bianco, R.; Damiano, V.; Pomatico, G.; De Placido, S.;
Bianco, A. R.; Tortora, G. Clin. Cancer Res. 2000, 6, 2053.
the EGF receptor activity at less than 10 lM. Although the pep-
tide should be optimized for higher inhibitory activity, it repre-
sents a promising lead for developing a new type of inhibitor
against cancers arising from uncontrolled activation of the EGF
receptor.
6. Anderson, N. G.; Ahmad, T.; Chan, K.; Dobson, R.; Bundred, N. J. Int. J. Cancer
2001, 94, 774.
7. Kawamoto, T.; Sato, J. D.; Le, A.; Polikoff, J.; Sato, G. H.; Mendelsohn, J. Proc. Natl.
Acad. Sci. U.S.A. 1983, 80, 1337.
8. Sato, J. D.; Kawamoto, T.; Le, A. D.; Mendelsohn, J.; Polikoff, J.; Sato, G. H. Mol.
Biol. Med. 1983, 1, 511.
Acknowledgments
9. Goldstein, N. I.; Prewett, M.; Zuklys, K.; Rockwell, P.; Mendelsohn, J. Clin. Cancer
Res. 1995, 1, 1311.
10. Inoue, A.; Saijo, Y.; Maemondo, M.; Gomi, K.; Tokue, Y.; Kimura, Y.; Ebina, M.;
Kikuchi, T.; Moriya, T.; Nukiwa, T. Lancet 2003, 361, 137.
11. Norum, J. J. Chemother. 2006, 18, 532.
The authors would like to thank Professor Koji Owada and Dr.
Akihiro Hiraki for their help with ultracentrifugation. They are also
grateful to Professor Masayuki Yoshikawa and Dr. Hisashi Matsuda
for their kind support in quantification of the immunostained
bands with the image analyzer. This study was supported in part
by the 21st Century COE Program and the ‘Academic Frontier’ Pro-
ject for Private Universities: matching fund subsidy, from the Min-
istry of Education, Culture, Sports, Science and Technology (MEXT)
of Japan.
12. Starling, N.; Tilden, D.; White, J.; Cunningham, D. Br. J. Cancer 2007, 96, 206.
13. Schlessinger, J. J. Cell Biol. 1986, 103, 2067.
14. Yarden, Y.; Schlessinger, J. Biochemistry 1987, 26, 1434.
15. Yarden, Y.; Schlessinger, J. Biochemistry 1987, 26, 1443.
16. Ogiso, H.; Ishitani, R.; Nureki, O.; Fukai, S.; Yamanaka, M.; Kim, J.-H.; Saito, K.;
Sakamoto, A.; Inoue, M.; Shirouzu, M.; Yokoyama, S. Cell 2002, 110, 775.
17. Staros, J. V. Biochemistry 1982, 21, 3950.
18. Fanger, B. O.; Stephens, J. E.; Staros, J. V. FASEB J. 1989, 3, 71.
19. Kim, J.-H.; Saito, K.; Yokoyama, S. Eur. J. Biochem. 2002, 269, 2323.
20. Zhang, X.; Pickin, K. A.; Bose, R.; Jura, N.; Cole, P. A.; Kuriyan, J. Nature 2007,
450, 741.
Supplementary data
21. Park, B.-W.; Zhang, H.-T.; Wu, C.; Berezov, A.; Zhang, X.; Dua, R.; Wang, Q.; Kao,
G.; O’Rourke, D. M.; Greene, M. I.; Murali, R. Nat. Biotechnol. 2000, 18, 194.
22. Berezov, A.; Zhang, H.-T.; Greene, M. I.; Murali, R. J. Med. Chem. 2001, 44, 2565.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2009.04.080.