Macrocyclization in the Design of Non-Phosphorus-Containing Grb2 SH2 Domain-Binding Ligands

Article abstract of DOI:10.1021/jm030510e

Macrocyclization from the phosphotyrosyl (pTyr) mimetic's β-position has previously been shown to enhance Grb2 SH2 domain-binding affinity of phosphonate-based analogues. The current study examined the effects of such macrocyclization using a dicarboxymethyl-based pTyr mimetic. In extracellular assays affinity was enhanced approximately 5-fold relative to an open-chain congener. Enhancement was also observed in whole-cell assays examining blockade of Grb2 binding to the erbB-2 protein-tyrosine kinase.

Full text of DOI:10.1021/jm030510e

2166
J . Med. Chem. 2004, 47, 2166-2169
Brief Articles
Ma cr ocycliza tion in th e Design of Non -P h osp h or u s-Con ta in in g Gr b2 SH2
Dom a in -Bin d in g Liga n d s
Zhen-Dan Shi,† Chang-Qing Wei,† Kyeong Lee,† Hongpeng Liu,‡ Manchao Zhang,‡ Toshiyuki Araki,§
Lindsey R. Roberts,# Karen M. Worthy,# Robert J . Fisher,# Benjamin G. Neel,§ J ames A. Kelley,†
Dajun Yang,‡ and Terrence R. Burke, J r.*,†
Laboratory of Medicinal Chemistry, CCR, National Cancer Institute, National Institutes of Health, Frederick, Maryland 21702,
University of Michigan Medical School, Ann Arbor, Michigan 48109, Beth Israel Deaconess Medical Center,
Boston, Massachusetts 02215, and Protein Chemistry Laboratory, SAICsFrederick, Frederick, Maryland 21702
Received October 8, 2003
Macrocyclization from the phosphotyrosyl (pTyr) mimetic’s â-position has previously been shown
to enhance Grb2 SH2 domain-binding affinity of phosphonate-based analogues. The current
study examined the effects of such macrocyclization using a dicarboxymethyl-based pTyr
mimetic. In extracellular assays affinity was enhanced approximately 5-fold relative to an open-
chain congener. Enhancement was also observed in whole-cell assays examining blockade of
Grb2 binding to the erbB-2 protein-tyrosine kinase.
In our efforts to develop Grb2 SH2 domain-signaling
antagonists, we have examined non-phosphorus-con-
taining, carboxy-based pTyr mimetics displayed in a
high-affinity Grb2 SH2 domain-binding platform (1a ,
Figure 1) first disclosed by Novartis Corporation.1 Of
several carboxy-based analogues,2-5 the greatest po-
tency was obtained using a 4-malonylphenylalanine res-
idue bearing an NR-oxalyl substitutent (1b).4 In separate
work using phosphonate-containing pTyr mimetics, re-
cently we have employed ring-closing olefin metathesis
(RCM) to prepare macrocyclic tetrapeptide mimetics of
type 2 as globally constrained congeners that exhibited
enhanced Grb2 SH2 domain-binding potency and ef-
ficacy in whole cells.6 However, it had remained to be
shown what the combined effects of macrocyclization
would be in conjunction with carboxy-based pTyr mi-
metics, since macrocycle-induced alterations of binding
within the pTyr pocket could potentially have unex-
pected results. To address these uncertainties, reported
herein is a non-phosphorus-containing macrocycle (3
and its sodium salt 3s) along with biological evaluation
in both in vitro and in vivo systems.
F igu r e 1. Structures of open-chain and macrocyclic Grb2 SH2
domain-binding analogues.
modeling studies were carried out using the previously
reported X-ray structure of a phosphohexapeptide bound
to the Grb2 SH2 domain.7 Low-energy binding modes
were compared for the open-chain 1b and its corre-
sponding macrocycle 3 (Figure 2). Significant differences
were noted in the hydrogen-bonding interactions of
certain oxygens comprising the phosphate-mimicking
malonyl groups. It was found that the malonyl group of
macrocyle 3 did not fully participate in the network of
hydrogen bonds within the pTyr pocket that were
displayed by 1b.
Mea su r em en t of in Vitr o Gr b2 SH2 Dom a in -
Bin d in g Affin ity. By use of extracellular ELISA-based
Grb2 SH2 domain binding assays,4,6 open-chain ana-
logues 1b and 1c were compared with their correspond-
ing macrocyclic congeners 3s and 2, respectively (Table
1). The IC50 values obtained for 1b, 1c, and 2 (22, 10,
and 1.4 nM, respectively) were of the same order as the
previously reported data (70,4 9,8 and 2,6 respectively).
The newly prepared macrocycle 3s exhibited an IC50 of
4.3 ( 2.4 nM, which is approximately 5-fold more potent
Syn th esis
Synthesis of the final product 3 was by a procedure
similar to that recently reported for the preparation of
the analogous phosphonate-containing macrocycle 2.6
Preparation of the required tri-(OBut)-containing 10 was
achieved as shown in Scheme 1.
Resu lts a n d Discu ssion
To approximate potential binding interactions of
macrocycle 3 with the Grb2 SH2 domain, molecular
* To whom correspondence should be addressed. Phone: (301) 846-
5906. Fax: (301) 846-6033. E-mail: tburke@helix.nih.gov.
† National Cancer Institute.
‡ University of Michigan Medical School.
§ Beth Israel Deaconess Medical Center.
#
SAICsFrederick.
10.1021/jm030510e CCC: $27.50 © 2004 American Chemical Society
Published on Web 03/11/2004

Brief Articles
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 8 2167
Sch em e 1a
Ta ble 1. Grb2 SH2 Domain-Binding Affinities Determined by
SPR and ELISA Techniques
SPR-deriveda
kon
koff
Kd ( SD
(nM)
ELISA-deriveda
IC50 ( SD (nM)
compd (M-1 s-1
)
(s-1
)
1b
1c
2
rapid equilibrium
22.1 ( 7.1 (n ) 5)
8.9 × 105 1.1 × 10-1 128.1 ( 92.9 (n ) 8)b 10.0 ( 4.8 (n ) 3)
6.7 × 106 5.6 × 10-3 0.9 ( 0.4
(n ) 4)b 1.4 ( 0.7 (n ) 4)
(n ) 10) 4.3 ( 2.4 (n ) 8)
3s
5.5 × 106 4.4 × 10-2 12.1 ( 9.8
a
Values were determined as outlined in the Experimental
b
Section. Previously reported in ref 9.
(Table 1). The Kd value obtained for macrocycle 2
compared favorably with affinity as determined above
by ELISA techniques (IC50 ) 1.4 nM). Similarly, the Kd
value for 1c was of the same order as the IC50 value
previously obtained using SPR (IC50 ≈ 80 nM).2 Sur-
prisingly, however, the SPR value for open-chain 1c was
significantly higher (∼10-fold) than the ELISA-derived
IC50 value (10.0 ( 4.8 nM, Table 1). As with macrocycle
2, the SPR-derived Kd value for macrocycle 3s (12.1 nM)
was only slightly higher than the corresponding ELISA-
derived IC50 value (4 nM). Undesirably fast off rates
were observed despite high affinity, consistent with
previous reports that fast off rates are characteristic of
SH2 domain-binding pTyr-containing peptides.10
a
Reagents and conditions: (1) NaH, CuBr, dioxane, HMPA,
reflux, 3 h (75% yield); (ii) Pd(OAc)2, (o-tolyl)3P, Et3N, reflux, 16
h (87% yield); (iii) PhSCu, Et2O, THF, -40 °C, 3 h (58% yield)
[73% de]; (iv) NaHMDS, THF, -78 °C, 2 h (53% yield); (v) LiOH,
H2O2, 0 °C, 2.5 h, then room temp, 2.5 h (73% yield).
In h ibition of Gr b2 Bin d in g to p 185er bB-2 in Cells
in Cu ltu r e. Affinity constants obtained by in vitro
methods may not accurately reflect behavior in whole
cells, where passage across lipid bilayers must occur
prior to competing with phosphorylated protein ligand
for binding to Grb2 SH2 domains. Therefore, studies
were conducted by treating cells with varying concen-
trations of 1b, 1c, 2, and 3s in solution and then
measuring the levels of intracellular association of Grb2
with cytoplasmic erbB-2 PTK (in MDA-MB-453 breast
cancer cells). As shown in Figure 3, the greater potency
of 2 relative to 1c is consistent with those from previous
reports.6 A comparison of data for 3s versus 1b indicates
that the same can now be concluded for non-phosphorus-
containing macrocycles, namely, that macrocyclization
enhances apparent Grb2 SH2 domain inhibitory potency
in whole cell systems.
F igu r e 2. Molecular modeling overlay of 1b and 3 complexed
to the Grb2 SH2 domain with protein backbone atoms super-
imposed. Ligands are shown in larger diameter stick rendering
with 1b colored red and 3 colored by atom, with green being
carbon. For clarity, protein and amino acid side chains/
backbone amides for the 1b complex only are shown. Hydrogen
bonds are represented as dashed lines (yellow for protein-1b
and pink for protein-3).
Con clu sion s
The intent of the current study was to examine the
effects of macrocyclization on a non-phosphorus-con-
taining, carboxy-based ligand where the effects of con-
formational constraints on binding orientations within
the critical pTyr-binding pocket were unknown. Result-
ant affinity enhancements in both extracellular and
whole-cell assays further validated the potential utility
of macrocyclization in the design of Grb2 SH2 domain-
binding antagonists.
than its open-chain homologue 1b. To our knowledge,
3s is the most potent non-phosphorus-containing Grb2
SH2 domain-binding antagonist reported to date. Sur-
face plasmon resonance (SPR) derived Kd measurement,
determined by analyzing the direct binding of synthetic
compounds to Grb2 SH2 domain protein, provided Kd
values for 1c and 2 of 1289 and 0.9 nM,9 respectively
F igu r e 3. Inhibition of Grb2 binding in whole cells. Anti-phosphotyrosine Western blots of MDA-MB-453 breast cancer cell
lysates following treatment with analogues 1b, 1c, 2, and 3s at the indicated concentrations, cell lysis, and immunoprecipitation
with Grb2 polyclonal antibody. Probing of the same blot with Grb2 MAb was performed as control.

2168 J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 8
Brief Articles
4.21 (dd, 1H, J ) 3.3 and 8.8 Hz), 3.90 (m, 1H), 3.52 (dd, 1H,
J ) 8.3 and 16.5 Hz), 3.31 (dd, 1H, J ) 6.6 and 16.4 Hz), 1.47
(s, 18H). FABMS m/z 536 [MH+]. Anal. (C31H37NO7) C, H, N.
2-[4-(-3-((4S)-2-Oxo-4-ph en yl(1,3-oxazolidin -3-yl))(1S,2S)-
2-{[(ter t-bu tyl)oxyca r bon ylm eth yl}-3-oxo-1-vin ylp r op yl)-
p h en yl]p r op a n e-1,3-d ioic Acid Bis(1,1-d im et h ylet h yl)
Ester (9). To a 1 M stirred solution of N-sodiohexamethyl-
disilazane (NaHMDS) in THF (24.00 mL, 24.00 mmol) was
added dropwise a solution of 8 (5.58 g, 10.4 mmol) in THF (60
mL) at -78 °C. The solution was stirred at -78 °C (1 h), then
iodoacetate tert-butyl ester (2.55 g, 10.5 mmol) in THF (30 mL)
was added, and the mixture was stirred for an additional 4 h
at -78 °C. The mixture was subjected to an extractive workup
and purified by silica gel flash chromatography to provide 9
as a colorless oil (3.60 g, 53% yield). 1H NMR (CDCl3) δ 7.43-
7.22 (m, 9H), 6.08 (m, 1H), 5.21-5.17 (m, 2H), 4.83-4.76 (m,
2H), 4.41 (s, 1H), 3.97 (t, 1H, J ) 8.4 Hz), 3.86 (dd, 1H, J )
2.5 and 8.6 Hz), 3.32 (t, 1H, J ) 10.0 Hz), 2.77 (dd, 1H, J )
11.3 and 17.2 Hz), 2.64 (dd, 1H, J ) 3.7 and 17.2 Hz), 1.47 (s,
18H), 1.31 (s, 9H). FABMS m/z 650 [MH+]. HRFAB-MS calcd
for C29H32O9N (MH+ - 2C4H8): 538.2077. Found: 538.2022.
(2S ,3S )-3-(4-{Bis[(t er t -b u t yl)oxyca r b on yl]m e t h yl}-
p h en yl)-2-{[(ter t-bu tyl)oxyca r bon yl]m eth yl}p en t-4-en o-
ic Acid (10). A solution of 30% H2O2 (2.80 mL, 27.5 mmol)
was added via syringe over 10 min to a solution of 9 (3.58 g,
5.51 mmol) in THF/H2O (v:v, 3:1) (84 mL) at 0 °C. This was
followed by addition of LiOH (463 mg, 11.03 mmol) in H2O
(11 mL). After being stirred at 0 °C for 2.5 h, the solution was
warmed to room temperature and stirring was continued for
an additional 2.5 h. Sodium sulfite (3.47 g, 27.5 mmol) in H2O
(23 mL) was added to the mixture followed by addition of 1 N
HCl (27 mL). Evaporation of solvent in vacuo provided a
residue, which was subjected to an extractive workup and
purified by silica gel flash chromatography to give 10 as a
yellow solid (2.02 g, 73% yield). 1H NMR (CDCl3) δ 7.32 (d,
2H, J ) 8.4 Hz), 7.21 (d, 2H, J ) 8.2 Hz), 5.95 (m, 1H), 5.17-
5.13 (m, 2H), 4.40 (s, 1H), 3.68 (t, 1H, J ) 8.6 Hz), 3.23 (m,
1H), 2.63 (dd, 1H, J ) 10.7 and 16.8 Hz), 2.44 (dd, 1H, J ) 3.9
and 17.0 Hz), 1.46 (s, 18H), 1.28 (s, 9H). FABMS (negative
ion) m/z 503 [M - H]-. Anal. (C28H40O8‚1.25H2O) C, H.
Exp er im en ta l Section
Molecu la r Mod elin g. Molecular modeling studies were
carried out using MacroModel 8.0 (Schro¨dinger, L.L.C.) and
Sybyl 6.9 (Tripos, Inc.) on
a Silicon Graphics Octane 2
workstation. Construction of the protein-ligand complexes
was based on the X-ray structure of the Grb2 SH2 domain
complexed with a hexapeptide inhibitor (PDB entry 1TZE).7
The complex structure was minimized on MacroModel using
MMFF94s force field with a continuum solvation model. The
SD method was used with convergence criteria set to Gradient.
Ligand and protein residues within 5 Å were allowed to move
freely during energy minimization, while residues at a distance
between 5 and 10 Å were constrained by a parabolic force
constant of 50 kJ /Å. Residues beyond 10 Å were frozen. Results
are shown in Figure 2.
Extr a cellu la r Bin d in g Assa ys. Plasmon resonance analy-
ses of antagonist binding to Grb2 SH2 domain protein were
performed on BIACORE 2000 and BIACORE 3000 instru-
ments (Biacore Inc., Piscataway NJ ). Data were fit to a simple
1:1 interaction model with a correction for mass transport
using the global data analysis program CLAMP.11 Results are
shown in Table 1. Experimental details are contained in the
Supporting Information. ELISA-based binding assays were as
conducted as previously reported.4 Results are shown in Table
1.
In h ib it ion of Gr b 2 SH2 Dom a in Bin d in g in Wh ole
Cells. Analyses of Grb2 binding to p185erbB-2 in erbB-2
overexpressing MDA-MB-453 breast cancer cells were con-
ducted as previously reported.4 Results are shown in Figure
3.
2-(4-Br om op h en yl)p r op a n e-1,3-d ioic Acid Bis(1,1-d i-
m eth yleth yl) Ester (5). To a suspension of sodium hydride,
60% in oil (1.88 g, 46.9 mmol) in anhydrous dioxane (80 mL),
was added hexamethylphosphoramide (HMPA) (6 mL) and di-
tert-butyl malonate (10.14 g, 46.9 mmol). The mixture was
stirred at room temperature (1 h), charged with copper(I)
bromide (8.08 g, 56.3 mmol) and 1-bromo-4-iodobenzene (4)
(15.00 g, 53.02 mmol), and refluxed under argon (3 h). After
cooling to room temperature, the mixture was subjected to an
extractive workup and purified by by silica gel flash chroma-
tography to provide 5 as a white solid (13.22 g, 75% yield). 1H
NMR (CDCl3) δ 7.47 (dd, 2 H, J ) 6.66 and 2.0 Hz), 7.26 (dd,
2H, J ) 6.5 and 1.9 Hz), 4.40 (s, 1H), 1.46 (s, 18 H). FABMS
m/z 371 [MH+]. Anal. (C17H23O4Br) C, H.
Ack n ow led gm en t. T.A. is a Fellow of the Leukemia
and Lymphoma Society (Grant No. 5094-03). Work was
supported in part by NIH Grant R01 CA49152 (to
B.G.N.).
2-{4-[(1E)-3-((4S)-2-Oxo-4-p h en yl(1,3-oxa zolid in -3-yl))-
3-oxop r op -1-en yl]p h en yl}p r op a n e-1,3-d ioic Acid Bis(1,1-
d im eth yleth yl) Ester (7). To a solution of 5 (7.77 g, 20.95
mmol) and (4S)-3-(1-oxo-2-propenyl)-4-phenyl-2-oxazolidinone
(6)6 (4.55 g, 20.95 mmol) in Et3N (87 mL) was added Pd(OAc)2
(233 mg, 0.993 mmol) and (o-toly)3P (1.30 g, 4.27 mmol). The
mixture was refluxed under argon (16 h), then cooled to room
temperature, diluted with CH2Cl2, and filtered through Celite.
The mixture was subjected to an extractive workup and
purified by silica gel flash chromatography to provide 7 as a
white solid (9.19 g, 87% yield). 1H NMR (CDCl3) δ 7.91 (d, 1H,
J ) 15.6 Hz), 7.76 (d, 1H, J ) 15.6 Hz), 7.56 (d, 2H, J ) 8.4
Hz), 7.41-7.33 (m, 7H), 5.55 (dd, 1H, J ) 4.0 and 8.8 Hz),
4.73 (t, 1H, J ) 8.8 Hz), 4.43 (s, 1H), 4.31 (J ) 4.0 and 8.8
Hz), 1.45 (s, 18 H). FABMS m/z 508 [MH+]. Anal. (C29H35NO7‚
0.25H2O) C, H, N.
2-{4-[(1R)-3-((4S)-2-Oxo-4-p h en yl(1,3-oxa zolid in -3-yl))-
3-oxo-1-vin ylp r op yl]p h en yl}p r op a n e-1,3-d ioic Acid Bis-
(1,1-d im eth yleth yl) Ester (8). To a slurry of PhSCu (3.10 g,
17.94 mmol) in dry ether (250 mL) under argon at -40 °C was
added dropwise a solution of vinylmagnisium bromide, 1.0 M
in THF (53.8 mL, 53.8 mmol), over 40 min. The mixture was
stirred at -40 °C (20 min), a precooled solution of 7 (9.10 g,
17.94 mmol) in THF (150 mL) was added over 1 h, and the
resulting mixture was stirred at -40 °C (3 h). The mixture
was subjected to an extractive workup and purified by silica
gel flash chromatography to provide 8 as a yellow solid (5.60
g, 58% yield). 1H NMR (CDCl3) δ 7.36-7.20 (m, 9H), 5.96 (ddd,
1H, J ) 7.0, 10.5, and 17.0 Hz), 5.29 (dd, 1H, J ) 3.3 and 8.6
Hz), 5.00-4.94 (m, 2H), 4.57 (t, 1H, J ) 8.7 Hz), 4.40 (s, 1H),
Su p p or tin g In for m a tion Ava ila ble: Detailed experi-
mental protocols for SPR binding analysis as well as the
detailed modifications of procedures in ref 6 used to prepare
macrocycle 3 from pTyr mimetic 10. This material is available
free of charge via the Internet at http://pubs.acs.org.
Refer en ces
(1) Furet, P.; Gay, B.; Caravatti, G.; Garcia-Echeverria, C.; Rahuel,
J .; Schoepfer, J .; Fretz, H. Structure-based design and synthesis
of high affinity tripeptide ligands of the Grb2-SH2 domain. J .
Med. Chem. 1998, 41, 3442-3449.
(2) Yao, Z. J .; King, C. R.; Cao, T.; Kelley, J .; Milne, G. W. A.; Voigt,
J . H.; Burke, T. R. Potent inhibition of Grb2 SH2 domain binding
by non-phosphate-containing ligands. J . Med. Chem. 1999, 42,
25-35.
(3) Burke, T. R., J r.; Luo, J .; Yao, Z.-J .; Gao, Y.; Milne, G. W. A.;
Guo, R.; Voigt, J . H.; King, C. R.; Yang, D. Monocarboxylic
phosphotyrosyl mimetics in the design of Grb2 SH2 domain
inhibitors. Bioorg. Med. Chem. Lett. 1999, 9, 347-352.
(4) Gao, Y.; Luo, J .; Yao, Z.-J .; Guo, R.; Zou, H.; Kelley, J .; Voigt, J .
H.; Yang, D.; Burke, T. R., J r. Inhibition of Grb2 SH2 domain
binding by non-phosphate containing ligands. 2. 4-(2-Malonyl)-
phenylalanine as a potent phosphotyrosyl mimetic. J . Med.
Chem. 2000, 43, 911-920.
(5) Burke, T. R., J r.; Yao, Z. J .; Gao, Y.; Wu, J . X.; Zhu, X.; Luo, J .
H.; Guo, R.; Yang, D. N-Terminal carboxyl and tetrazole-
containing amides as adjuvants to Grb2 SH2 domain ligand
binding. Bioorg. Med. Chem. 2001, 9, 1439-1445.
(6) Wei, C.-Q.; Gao, Y.; Lee, K.; Guo, R.; Li, B.; Zhang, M.; Yang,
D.; Burke, T. R., J r. Macrocyclization in the design of Grb2 SH2
domain-binding ligands exhibiting high potency in whole cell
systems. J . Med. Chem. 2003, 46, 244-254.

Brief Articles
J ournal of Medicinal Chemistry, 2004, Vol. 47, No. 8 2169
(7) Rahuel, J .; Gay, B.; Erdmann, D.; Strauss, A.; GarciaEcheverria,
C.; Furet, P.; Caravatti, G.; Fretz, H.; Schoepfer, J .; Grutter, M.
G. Structural basis for specificity of GRB2-SH2 revealed by a
novel ligand binding mode. Nat. Struct. Biol. 1996, 3, 586-589.
(8) Wei, C.-Q.; Li, B.; Guo, R.; Yang, D.; Burke, T. R., J r. Develop-
ment of a phosphatase-stable phosphotyrosyl mimetic suitably
protected for the synthesis of high affinity Grb2 SH2 domain-
binding ligands. Bioorg. Med. Chem. Lett. 2002, 12, 2781-2784.
(9) Shi, Z.-D.; Lee, K.; Liu, H.; Zhang, M.; Roberts, L. R.; Worthy,
K. M.; Fivash, M. J .; Fisher, R. J .; Yang, D.; Burke, T. R., J r. A
novel macrocyclic tetrapeptide mimetic that exhibits low-pico-
molar Grb2 SH2 domain-binding affinity. Biochem. Biophys. Res.
Commun. 2003, 310, 378-383.
(10) Felder, S.; Zhou, M.; Hu, P.; Urena, J .; Ullrich, A.; Chaudhuri,
M.; White, M.; Shoelson, S. E.; Schlessinger, J . SH2 domains
exhibit high-affinity binding to tyrosine-phosphorylated peptides
yet also exhibit rapid dissociation and exchange. Mol. Cell. Biol.
1993, 13, 1449-1455.
(11) Myszka, D. G.; Morton, T. A. CLAMP: a biosensor kinetic data
analysis program. Trends Biochem. Sci. 1998, 23, 149-150.
J M030510E