Selectivity and mechanism of action of a growth factor receptor-bound protein 2 Src homology 2 domain binding antagonist

Article abstract of DOI:10.1021/jm800523u

We have shown previously that a potent synthetic antagonist of growth factor receptor-bound protein 2 (Grb2) Src homology 2 (SH2) domain binding (1) blocks growth factor stimulated motility, invasion, and angiogenesis in cultured cell models, as well as tumor metastasis in animals. To characterize the selectivity of 1 for the SH2 domain of Grb2 over other proteins containing similar structural binding motifs, we synthesized a biotinylated derivative (3) that retained high affinity Grb2 SH2 domain binding and potent biological activity. To investigate the selectivity of 1 and 3 for Grb2, the biotinylated antagonist 3 was used to immobilize target proteins from cell extracts for subsequent identification by mass spectrometry. Non-specific binding was identified in parallel using a biotinylated analogue that lacked a single critical binding determinant. The mechanism of action of the antagonist was further characterized by immunoprecipitation, immunoblotting, and light microscopy. This approach to defining protein binding antagonist selectivity and molecular basis of action should be widely applicable in drug development.

Full text of DOI:10.1021/jm800523u

J. Med. Chem. 2008, 51, 7459–7468
7459
Selectivity and Mechanism of Action of a Growth Factor Receptor-Bound Protein 2 Src
Homology 2 Domain Binding Antagonist
Alessio Giubellino,† Zhen-Dan Shi,‡ Lisa M. Miller Jenkins,§ Karen M. Worthy,| Lakshman K. Bindu,| Gagani Athauda,†
Benedetta Peruzzi,† Robert J. Fisher,| Ettore Appella,§ Terrence R. Burke, Jr.,‡ and Donald P. Bottaro*,†
Urologic Oncology Branch and Laboratory of Cell Biology, National Cancer Institute, Bethesda, Maryland 20892, Laboratory of Medicinal
Chemistry, National Cancer Institute at Frederick, Frederick, Maryland 21702 and Protein Chemistry Laboratory-SAIC,
Frederick, Maryland 21702
ReceiVed May 7, 2008
We have shown previously that a potent synthetic antagonist of growth factor receptor-bound protein 2
(Grb2) Src homology 2 (SH2) domain binding (1) blocks growth factor stimulated motility, invasion, and
angiogenesis in cultured cell models, as well as tumor metastasis in animals. To characterize the selectivity
of 1 for the SH2 domain of Grb2 over other proteins containing similar structural binding motifs, we
synthesized a biotinylated derivative (3) that retained high affinity Grb2 SH2 domain binding and potent
biological activity. To investigate the selectivity of 1 and 3 for Grb2, the biotinylated antagonist 3 was used
to immobilize target proteins from cell extracts for subsequent identification by mass spectrometry. Non-
specific binding was identified in parallel using a biotinylated analogue that lacked a single critical binding
determinant. The mechanism of action of the antagonist was further characterized by immunoprecipitation,
immunoblotting, and light microscopy. This approach to defining protein binding antagonist selectivity and
molecular basis of action should be widely applicable in drug development.
Introduction
design of improved therapeutic and diagnostic agents. The
modification of investigational compounds with chemical tags
such as biotin is among the most direct strategies yet devised
to address these questions, and this approach was utilized in
our present investigation.
Antagonists of protein-protein interactions have significant
therapeutic potential in a broad range of indications, including
cancer. However, defining target selectivity and defining the
molecular basis of therapeutic effects are important challenges
to their development and use. For this class of drug in particular,
systematic exploration of these subjects at the earliest stages of
development can profoundly improve prediction, measurement
and avoidance of drug toxicity, identification of appropriate
patient populations, and drug efficacy assessment. In the long
term, systematically defining selectivity and mechanism of
action may also accelerate target validation and facilitate the
Ligand binding, receptor overexpression, and/or oncogenic
mutations can induce receptor tyrosine kinase (RTKa) activation
and autophosphorylation of specific tyrosine residues within
RTK intracellular domains. A prominent consequence of ty-
rosine phosphorylation is the formation of docking sites for
proteins containing Src homology 2 (SH2) domains.1 One of
the best characterized proteins of this class is growth factor
receptor bound protein 2 (Grb2), an adapter that acts as a critical
downstream intermediary in several oncogenic signaling path-
ways. Originally isolated through screening for epidermal
growth factor receptor (EGFR) interacting proteins,2 Grb2
associates with several signaling and regulatory proteins.
Through its SH2 domain, which is a conserved sequence of
approximately 100 amino acids, Grb2 can interact directly with
RTKs (e.g., hepatocyte growth factor (HGF) receptor, platelet-
derived growth factor receptor) and non-receptor tyrosine
kinases (e.g., focal adhesion kinase (FAK) and Bcr/Abl) by
preferential binding to phosphopeptide motifs of the form
pYXNX, where pY represent phosphotyrosine, N is asparagine,
and X is any residue.3 The amino- and carboxyl-terminal Src
homology 3 (SH3) domains of Grb2, which have a conserved
sequence of around 50 amino acids, bind proline-rich regions
within additional interacting proteins. Through these two SH3
domains, Grb2 links activated RTKs with several key intracel-
lular regulatory networks, including the Ras/Erk pathway
controlling cell cycle progression, and the p21-activating kinase
(PAK1) and Arp2/3/WASp pathways regulating the actin
filament system, cell shape change, and motility.4,5 Grb2 is also
a key intermediate of integrin signaling through its interaction
with activated FAK at pY925, which resides within a canonical
recognition motif (pYXNX) for the Grb2 SH2 domain.6 Grb2-
FAK binding triggers a signaling sequence involved in angio-
* To whom correspondence should be addressed. Address: Urologic
Oncology Branch, CCR, NCI, Buildingg 10 CRC 1, West Room 3961, 10
Center Drivem MSC 1107, Bethesda, MD 20892-1107. Telephone: (301)
402-6499. Fax: (301) 402-0922. E-mail: dbottaro@helix.nih.gov.
†
Urologic Oncology Branch, National Cancer Institute.
Laboratory of Medicinal Chemistry, National Cancer Institute at
‡
Frederick.
§
Laboratory of Cell Biology, National Cancer Institute.
Protein Chemistry Laboratory, SAIC-Frederick, Inc.
|
a Abbreviations: Grb2, growth factor receptor-bound protein 2; SH2, Src
homology 2; RTK, receptor tyrosine kinase; EGFR, epidermal growth factor
receptor; HGF, hepatocyte growth factor; PDGFR, platelet-derived growth
factor receptor; FAK, focal adhesion kinase; Bcr, breakpoint cluster region;
Abl, Abelson murine leukemia viral oncogene; SH3, Src homology 3; Ras,
rat sarcoma virus oncogene; Erk, extracellular signal-regulated kinase;
PAK1, p21-activating kinase; Arp2/3, actin related protein 2/3; WASp,
Wiskott-Aldrich syndrome protein; EMT, epithelial-mesenchymal transi-
tion; SPR, surface plasmon resonance; SA, streptavidin; Nck, noncatalytic
region of tyrosine kinase adaptor protein; Gab1, growth factor receptor
protein 2 associated protein 1; SOS1, son of sevenless homologue 1; ESI-
MS/MS, electrospray ionization tandem mass spectrometry; Rac, rat sarcoma
virus oncogene related C3 botulinum toxin substrate; CDC42, cell division
cycle 42; cAMP, cyclic adenosine monophosphate; CREB, cyclic adenosine
monophosphate response element binding; MDCK, Madin-Darby canine
kidney cell; DAPI, 4′,6-diamidino-2-phenylindole; DPPP, 1,3-bis(diphe-
nylphosphine)propane; EDC, 1-ethyl-3-(3-dimethylaminopropyl)carbodi-
imide; HOBt, 1-hydroxybenzotriazole; DMAP, 4-dimethylaminopyridine;
DMF, dimethylformamide; THF, tetrahydrofuran; DIPCDI, diisopropyl-
carbodiimide; TFA-TES, trifluoroacetic acid-N-tris(hydroxymethyl)methyl-
2-aminoethanesulfonic acid.
10.1021/jm800523u CCC: $40.75
2008 American Chemical Society
Published on Web 11/07/2008

7460 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23
Giubellino et al.
primary amine 9 was obtained in 54% yield. By use of standard
coupling procedures, asparagine and 1-aminocyclohexenecar-
boxylic acid were added to yield the dipeptide 10.14
Coupling of the upper peptide segment 10 and the bottom
residues 11a,b (11a was prepared as previously described;17
11b was available from Novobiochem Corp.) was performed
in the presence of 1-ethyl-3-(3-dimethylaminopropyl)carbodi-
imide hydrochloride (EDC) and hydroxybenzotriazole (HOBt)
to give 12a and 12b in 79% yield and 74% yield, respectively
(Scheme 2). Piperidine-mediated cleavage of N-Fmoc followed
by protection of the resulting free amino group with tert-
butyloxalyl chloride produced 13a and 13b in 43% yield and
80% yield, respectively. Of note, the 5-phenol remained
unaffected by the coupling reaction conditions and was subse-
quently coupled to the biotinylated linker in the next step. Global
deprotection of tert-butyl ester of 13a gave final product 2.
Coupling of 13a and 13b to the biotinylated linker was
performed in the presence of EDC and 4-dimethylaminopyridine
(DMAP) to provide the desired products 14a and 14b in 78%
and 42% yield, respectively. These were converted to the
corresponding free acids 3 and 4 after cleavage of the tert-butyl
esters.
To characterize the selectivity of the Grb2 SH2 domain
antagonist compound 1 for the SH2 domain of Grb2 over other
proteins containing similar motifs, analogues 2 and 3 were
prepared, the latter bearing biotin functionality at a carboxyl-
terminal position. The compound was biotinylated by attaching
the biotin to the naphthyl ring as described above. Biotin is
commonly attached to biologically active molecules by means
of a 6-aminocaproic acid ester spacer without unduly affecting
their biological activities; the spacer minimizes potential steric
interactions between avidin and the Grb2 SH2 domain-ligand
complex. A biotinylated variant of 1 that had very poor Grb2
SH2 domain binding affinity (4) was prepared in parallel and
used as a negative control throughout the course of these studies.
Analogue 4 is similar to 3, except that 4 does not contain a
phosphoryl mimetic at the 4-position of the tyrosyl mimetic aryl
ring, which is a critical requirement for high affinity SH2 domain
binding.
Figure 1. Chemical structures of Grb2-SH2 domain binding antagonists
and analogues.
genesis and epithelial-mesenchymal transition (EMT), both of
which are important contributors to tumor progression.7,8
The critical roles served by Grb2 in cell motility and
angiogenesis make it a logical therapeutic target for pathological
processes leading to the spread of solid tumors through local
invasion and metastasis.9 Potent, synthetic, low molecular weight
antagonists of Grb2 SH2 domain binding have been developed
that block RTK-Grb2 interaction and growth factor-stimulated
motility and matrix invasion by several tumor cell lines.10 We
have previously reported that the Grb2 SH2 domain binding
antagonist 1 (Figure 1) can inhibit cell motility,11 angiogenesis,12
and tumor metastasis in ViVo.13 To better understand the SH2
domain selectivity and mechanism of action of this class of
compounds, we designed and synthesized biotinylated deriva-
tives of two biologically active structures and established that
high affinity Grb2 binding was maintained. The synthesis and
binding evaluation of a macrocyclic Grb2 SH2 domain binding
antagonist that bears the biotin functionality at a carboxyl-
terminal position were described in a prior report.14 In the
current work, we present the synthesis of a hydroxylated variant
of 1 (2) as well as an open-chain biotinylated version of 1 (3)
(Figure 1). We use 2 and 3 as tools to explore ligand selectivity
through immobilization by streptavidin-coated beads and sub-
sequent mass spectrometry analysis. We further evaluated the
biological mechanism of action of the compound in intact cells
and report herein the inhibition of lamellipodia formation by 1
as a molecular basis for its antimotility activity.
Binding Affinities of Grb2 SH2 Domain Antagonists. The
Grb2 SH2 domain binding affinities of 2 and its biotin-
containing counterpart 3 were determined by surface plasmon
resonance (SPR). Binding constants for SH2 domain-small
molecule interactions typically have been determined indirectly
by measuring the ability of a small molecule to compete with
sensor-bound reference pTyr-containing peptide for the binding
of SH2 domain-containing protein in the mobile phase. Sig-
nificant improvements in the sensitivity and operation of SPR
instrumentation now permit accurate direct binding measure-
ments of small compounds such as 2 and 3 to SH2 domain
proteins immobilized on the sensor chip. Data were analyzed
to a one-site binding model using Scrubber software (Experi-
mental Section). As shown in Table 1, compound 2 bound to
amine-coupled Grb2 SH2 domain with a steady state affinity
of 132 nM while compound 3 displayed 3-fold lower binding
affinity (405 nM), indicating only modest perturbation of
antagonist-SH2domaininteractionbytheaddedbiotinfunctionality.
Results and Discussion
Syntheses. Syntheses of 2 and the biotinylated conjugates 3
and 4 were similar to the previously reported procedures14 and
differed mainly in the preparation of segment 9. Starting from
1,5-dihydroxylnaphthalene 5, the monohydroxyl group was
protected as the triflate using N-phenyltrifluoromethanesulfo-
nimide and 2,4,6-collidine (Scheme 1).15 Subsequent Heck
reaction proved troublesome because the C-O bond is more
stable compared to the C-halide bond and requires higher
cleavage energy. Thus, no reaction was observed under the
commonly used Heck reaction conditions of palladium(II)
acetate and tri-O-tolylphosphine. However, with the more active
phosphine ligand 1,3-bis(diphenylphosphine)propane (DPPP),16
two coupling products were obtained as the expected adduct 7
and its reductive product 8 in combined 81% yield. After
reduction of both adducts with lithium aluminum hydride, the
As part of our target selectivity assessment of these SH2
antagonists, we compared their binding to the SH2 domains of
Grb2 and Shc by SPR. The Shc SH2 domain recognizes the
same general sequence motif as all known SH2 domains
(pYXXX), but unlike Grb2 does not require asparagine (N) at
the pY + 2 position (pYXNX). While prior studies indicate
that compounds such as 3 are unlikely to bind Src-like SH2

Grb2 SH2 Domain Binding Antagonist
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23 7461
Scheme 1. Synthesis of the Upper Segment 10a
a Reagents, conditions and yields: (a) Tf2NPh, 2,4,6-collidine, DMAP, DMF-CH2Cl2, reflux, 12 h, 47%; (b) Pd(OAc)2, DPPP, Et3N, reflux, 14 h, 81%;
(c) LiAlH4, THF, room temp, overnight, 54%; (d) (i) Boc-Asn-OH, DIPCDI, HOBt, DMF, room temp, 12 h, 87%; (ii) TFA, CH2Cl2, room temp, 1 h, 79%;
(iii) Fmoc-1-aminocyclohexenecarboxylic acid, EDC, HOBt, DMF, room temp, 12 h, 79%; (iv) piperidine, CH3CN, room temp, 2 h, 86%.
Scheme 2. Synthesis of Inhibitor 2 and Biotinylated Ligands 3 and 4a
a Reagents, conditions, and yields: (a) EDC, HOBt, DMF, room temp, 12 h (79% for 12a, 74% for 12b); (b) (i) piperidine, CH3CN, room temp, 2 h; (ii)
i
tert-butyloxalyl chloride, Pr2NEt, DMF, room temp, 12 h (two steps, 43% for 13a; 80% for 13b); (c) TFA-TES-H2O, room temp, 1 h, 42%; (d) EDC,
DMAP, DMF, room temp, 12 h (78% for 14a, 42% for 14b); (e) TFA-HS(CH2)2SH-H2O, room temp, 1 h (41% for 3, 56% for 4).
domains with high affinity, the potential for Shc SH2 domain
binding is not easily predicted because its three-dimensional
structure differs significantly from those of the Src-like kinases
and Grb2.18 For these SPR studies, the Grb2 and Shc SH2
domains were recombinantly engineered to contain a biotin tag
at the encoded 5′ end.19 Immobilization of these constructs on

7462 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23
Giubellino et al.
Table 1. Grb2 and Shc SH2 Domain Steady State Binding Affinities
Determined by Surface Plasmon Resonance Spectrometry (Biacore)
compd
SH2 domain
couplinga
affinity (nM)
ratio
2
3
1
1
2
2
Grb2
Grb2
biotin-Grb2
biotin-Shc
biotin-Grb2
biotin-Shc
amine
amine
SA
SA
SA
132
405
370
24000
93
9400
1
3
1
65
1
101
SA
a Coupling to the chip surface via amine linkage or streptavidin (SA)
capture of biotinylated recombinant proteins.
Figure 3. Protein capture by 3 and 4. (A) SDS-PAGE and anti-Grb2
immunoblot analysis of lysates prepared from the human leiomyosa-
rcoma cell line SK-LMS-1. Following target protein capture using 3
(left) or 4 (right), SA-coated beads were washed with buffer containing
NaCl at the concentrations (M) indicated above each lane. (B)
SDS-PAGE and immunoblot analysis of SK LMS-1 lysates (left lane)
for known Grb2-associated proteins (top to bottom) Nck1, Nck2, Gab1,
and Sos1 captured using 4 (middle lane) or 3 (right lane).
to treatment with the negative control analogue 4 at any
concentration tested (data not shown).
In order to verify that 3 could efficiently capture Grb2 from
detergent extracts of cultured cells, SA-coated beads were treated
with biotin-containing 3 or 4 under conditions designed to
saturate available SA sites. The compound-conjugated SA beads
were then incubated with non-ionic detergent extracts prepared
from cultured tumor cells and washed with buffers of increasing
ionic strength (150-500 mM NaCl) to remove loosely bound
proteins (monitored by silver staining, data not shown). The
washes were analyzed for retention of Grb2 by SDS-PAGE
and immunoblotting (Figure 3A). Grb2 was clearly retained on
compound 3-conjugated SA beads, but in contrast, ligand 4 did
not capture Grb2 under any conditions tested. We also confirmed
the ability of 3 to capture proteins that are known to bind to
the SH3 domains of Grb2; SDS-PAGE and immunoblot
analysis of proteins captured by SA-bound 3 revealed the
presence of Nck1, Nck2, Gab1, and Sos1 (Figure 3B).
Proteins captured from cell extracts using compound 3-con-
jugated SA beads under conditions where Grb2 binding was
detected by immunoblotting were subjected to exhaustive trypsin
digestion; the resulting peptides were then separated chromato-
graphically by HPLC and analyzed by tandem mass spectrom-
etry (ESI-MS/MS) on a linear ion trap mass spectrometer.
Representative ESI-MS/MS spectra of two [M + 2H]2+ tryptic
peptides from Grb2 (NH2-GACHGQTGMFPR-COOH and
NH2-ATADDELSFK-COOH; NCBI sequence accession NM_
203506.2) that were captured by compound 3-conjugated SA
beads, but not by compound 4-conjugated beads, and which
represent 10.1% sequence coverage, are shown in Figure 4. The
b ion series is shown in red and the y ion series in blue.
Statistical analysis using the Sequest algorithm yielded peptide
correlation (Xcorr) scores of 3.707 (p ) 1.9 × 10-9) (Figure
4A) and 2.98 (p ) 1.4 × 10-5) (Figure 4B) for the two peptides.
The difference in correlation between the top two peptide
matches (∆Cn) were 0.44 and 0.28 for the spectra shown in
parts A and B of Figure 4, respectively; together with the peptide
correlation scores, these values confirm robust protein identi-
fication. Importantly, Grb2 was the only SH2-domain containing
protein detected by ESI-MS/MS among the proteins captured
from cell lysates by 3 or the negative control 4. As shown in
Table 2, five other proteins were captured differentially by 3
with Grb2; further research will be needed to determine whether
Figure 2. Inhibition of cell migration by 1 and 3. (A) Inhibition of
HGF-stimulated (50 ng/mL, white bar) or unstimulated (black bar)
PC3M cell migration by 1 and 3. Values represent mean number of
migrated cells per 10× microscopic field, expressed as percent
maximum, (standard deviation. (B) Dose-response analysis of inhibi-
tion of PC3M cell migration by 3.
sensor chips coated with streptavidin (SA) provides uniform
orientation on the chip surface and thereby less variability in
ligand binding. The binding affinity of 2 for SA-immobilized
biotin-Grb2 was in good agreement with the value obtained
when Grb2 was immobilized by amine coupling (93 vs 132 nM,
respectively, Table 1). Of note, compound 2 displayed 100-
fold lower binding affinity for biotin-Shc than for biotin-Grb2
(93 vs 9400 nM, Table 1), indicating a high degree of Grb2
selectivity. Consistent with its overall similarity to compound
2, compound 1 also displayed dramatically lower binding affinity
for the SH2 domain of Shc relative to that of Grb2 (370 vs
24 000 nM, Table 1).
Biotinylated 3 Retains Anti-motility Activity and Is Highly
Selective for Grb2. We have shown previously that the SH2
domain binding antagonist 1 blocks cell motility in a dose
dependent fashion.11,13 To confirm that biotin-containing 3
retains the biological activity of the parent compound 1 in intact
cells, we performed cell migration studies using modified
Boyden chambers in the presence or absence of the protein
motility factor HGF (50 ng/mL). As shown in Figure 2A, 3 has
anti-migratory activity comparable to that of the parent com-
pound 1, resulting in significant and potent inhibition of
migration by the prostate tumor-derived cell line PC3M. Detailed
dose-response analysis of inhibition of migration by 3 yielded
an estimated IC50 value of 14 nM (Figure 2B; R2 ) 0.9, P <
0.01). No inhibition of cell migration was observed in response

Grb2 SH2 Domain Binding Antagonist
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23 7463
Figure 4. Representative MS/MS spectra of two [M + 2H]2+ tryptic peptides from Grb2, which represent 10.1% sequence coverage, captured by
3-conjugated beads but not 4-conjugated beads. The b ion series is shown in red and the y ion series in blue. Sequest statistics are as follows: (A)
Xcorr score ) 3.707, ∆Cn ) 0.44, and p ) 1.9 × 10-9; (B) Xcorr score ) 2.98, ∆Cn ) 0.28, and p ) 1.4 × 10-5, where Xcorr is the peptide
correlation score and ∆Cn is the difference in correlation between the top two peptide matches.
their capture was mediated by Grb2 or by direct interaction with
3. Overall, these results confirm and extend our prior findings
that 1 and closely related structures selectively block the Grb2
SH2 domain and not structurally distinct SH2 domains such as
those present in Shc or phosphatidylinositol-3 kinase.11,13
Molecular and Cellular Mechanism of Action of 1. Having
established the selectivity of 1 for the SH2 domain of Grb2,
we further investigated its molecular mechanism of action by
determining its effect on the activation state and subcellular
localization of oncogenically relevant downstream signaling
proteins. We focused on proteins most likely to be involved in
the activities blocked by 1, namely, cell adhesion and motility.
The p21-activated kinase 1 (PAK1) is an effector of Rac/Cdc42
GTPases that regulates cytoskeletal dynamics and is essential
for cell migration. PAK1 binds directly to Grb2 SH3 domains,
and its linkage to Grb2 is critical for PAK1 activation,
autophosphorylation, and lamellipodia formation downstream
of growth factor-stimulated RTKs.20 We theorized that disrupt-
ing the SH2-mediated link between Grb2 and initiators of
motility signaling, such as activated RTKs and integrin receptors,
might also block PAK1 activation and autophosphorylation
without affecting the SH3 domain-mediated PAK1-Grb2 as-
sociation. We investigated the effect of 1 on growth factor-
stimulated PAK1 activation by treating intact PC3M cells with
1 (1 µM) for 16 h (vs leaving them untreated), then stimulating
the cells briefly with HGF and monitoring PAK1 activation by
immunoblotting with a specific anti-phospho-PAK1 antibody
(Figure 5A). Treatment with 1 had little effect on the steady
state level of PAK1 autophosphorylation, which is not associated
with cell shape change or motility (Figure 5A, left 2 lanes). In
the absence of 1, HGF stimulation results in an initial decrease
and then increase in PAK1 autophosphorylation after 30-60
min, which is correlated with cell shape change and lamellipodia
extension prior to increased motility (Figure 5A, middle lanes).
This latter process of HGF-stimulated PAK1 activation, which
is critical to motility, was completely inhibited by treatment
with 1 (Figure 5A, right lanes).
Growth factor-stimulated activation of PAK1 has been
associated with nuclear translocation, histone protein phospho-
rylation, and modulation of expression of several gene tran-
scripts.21 Increased nuclear PAK1 activity has been reported in
human renal cell carcinoma cells22 and in estrogen receptor-

7464 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23
Giubellino et al.
Table 2. Peptides Identified by Mass Spectrometry Analysis of Samples Captured from Cell Lysates by Streptavidin-Immobilized 3 but Not by
Streptavidin-Immobilized 4a
peptide sequences observed
NCBI accession
4506693
protein name
probability
9.53 × 10-12
Xcorr score
48.31
K.LLDFGSLSNLQVTQPTVGMNFK.T
R.VCEEIAIIPSKK.L
ribosomal protein S17
R.VIIEKYYTR.L
K.IAGYVTHLMKR.I
R.LGNDFHTNK.R
K.IYVDDGLISLQVK.Q
R.NTGIICTIGPASR.S
R.GDLGIEIPAEK.V
R.VNFAMNVGK.A
33286418, 33286420
pyruvate kinase 3b
1.10 × 10-6
40.24
K.ATADDELSFK.R
K.GACHGQTGMFPR.N
45359859
Grb2
2.17 × 10-6
6.51 × 10-6
2.12 × 10-7
5.16 × 10-5
20.83
20.23
20.2
K.AMGIMNSFVNDIFER.I
R.KESYSIYVYK.V
4504257, 4504271
4506701
histone H2B familyc
ribosomal protein S23
ribosomal protein L17
K.VANVSLLALYK.G
K.AHLGTALK.A
K.SAEFLLHMLK.N
K.GLDVDSLVIEHIQVNK.A
4506617
10.16
a Only those proteins for which two or more unique peptides were observed are shown. Probability is the p-value for the identification of the protein
context of the peptide. Xcorr score is the calculated correlation value for the protein based on the combined Xcorr values of the peptides, obtained using the
Sequest algorithm. b Unable to distinguish isoforms. c Unable to distinguish H2B family members.
Figure 5. Compound 1 inhibits PAK1 activation and nuclear trans-
location. (A) PC3M cells were serum-deprived and treated with 1 (+,
1 µM) or left untreated (-), as indicated, for 16 h, then left without
further treatment (left lanes) or stimulated briefly with HGF (50 ng/
mL, right lanes) for the times indicated. Cell lysates were analyzed by
SDS-PAGE, and immunoblot analysis was done with anti-phospho-
PAK1 antibody (upper panel) or anti-PAK1 antibody to control for
protein loading (lower panel). (B) PC3M cells were treated with
compound 1 for 16 h, and nuclear extracts were prepared using high-
salt buffer. Samples were analyzed by SDS-PAGE, and immunoblot
analysis was done with anti-PAK1 antibody (upper panel) or anti-CREB
antibody (lower panel).
positive breast cancer,23 where it was correlated with tamoxifen
resistance. We found that treatment of intact cells with 1
decreased PAK1 nuclear translocation (Figure 5B, upper panel)
relative to nuclear cAMP response element-binding (CREB)
protein (Figure 5B, lower panel), which is not modulated by
growth factor stimulation.
Reorganization of the actin cytoskeleton is an early and
Figure 6. Compound 1 inhibits lamellipodia formation: Alexa 488-
critical step in lamellipodia formation prior to cell migration.
We examined whether antagonism of Grb2 SH2 domain binding
affected growth-factor stimulated actin reorganization and
lamellipodia formation by staining cells with fluorescently
labeled phalloidin, a toxin that interacts specifically with
polymerized F-actin bundles. Lamellipodia formation is con-
veniently analyzed in monolayer cell cultures by directing their
formation at an artificially created wound edge. To this end,
confluent monolayers were scratched with a pipet tip, stimulated
with HGF in the presence or absence of 1 for 6 h, then stained
with fluorescently tagged phalloidin. Because lamellipodia
formation is cell-type dependent, we examined this process in
the prostate tumor-derived cell line PC3M and in the highly
organized, normal epithelial cell line MDCK (Figure 6). HGF-
stimulated PC3M cells displayed abundant pointed edges of
advancing plasma membrane in the direction of cell migration
phalloidin (green) and DAPI staining (blue) of PC3M cells (left panels)
and MDCK cells (center and right panels) treated with HGF alone (top
panels, control) or HGF + 1 (1 µM, lower panels). For PC3M cells,
arrowheads indicate regions of cellular extension into the artificial
wound. For MDCK cells, advancing lamellipodia observed in center
panels are outlined in white in the right panels for clarity.
(Figure 6, top left panel), while treatment with 1 markedly
reduced phalloidin staining at the leading edge, indicative of a
reduced actin bundle reorganization and membrane extension
(Figure 6, bottom left panel). In contrast to PC3M cells, the
actin cytoskeletal structure of MDCK cells is more uniformly
organized (Figure 5, right panels). HGF stimulation produced
the formation of well-structured advancing lamellipodia (Figure
6 top center and right panels). Treatment with 1 significantly
reduced the area and distance that lamellipodia extended into

Grb2 SH2 Domain Binding Antagonist
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23 7465
running buffer to concentrations ranging from 1.25 to 1500 nM
and serially injected at 25 °C at a flow rate of 30 µL/min for 2
min. Each concentration was followed by two blank buffer
injections, and every injection was performed in duplicate within
each experiment. To subtract background noise from each data set,
all samples were also run over Neutravidin reference surfaces to
allow double referencing. Data were fit to a simple 1:1 interaction
model using the global data analysis program Scrubber (http://
www.biologic.com.au/).
Protein Capture and Immunoblotting. SK-LMS-1 cells were
serum-deprived for 16 h, stimulated with recombinant human HGF
(25 ng/mL; R&D Systems, Minneapolis, MN) for 20 min, and lysed
in cold nonionic detergent containing buffer. Compound 3 (3 µM)
or 4 (3 µM) was added, and the lysate was incubated on ice for
1 h. SA-coated beads (streptavidin on UltraLink resin, Pierce
Biotechnology, Rockford, IL) were added, and lysates were mixed
at 4 °C for 1 h. SA-coated beads were then washed 5 times in lysis
buffer containing increasing concentrations of NaCl (150-500
mM), then stored at -80 °C in PBS. Samples were analyzed by
SDS-PAGE and immunoblotting or, in parallel, subjected to
exhaustive tryptic digestion for mass spectrometry analysis. Im-
munoblotting was performed as described using an anti-Grb2
antibody (Santa Cruz Biotech).13
Mass Spectrometry. Proteins captured on SA-coated beads were
washed three times with 100 mM ammonium bicarbonate to
exchange the buffer, reduced with 10 mM DTT for 1 h at 51 °C,
and alkylated with 55 mM iodoacetamide for 45 min at 25 °C.
The proteins were then digested with trypsin for 6 h at 25 °C with
agitation. The beads were pelleted, and the supernatant was
removed. Peptides in the supernatant were acidified by adding 0.1%
formic acid. Separation of the peptides was performed at 500 nL/
min and was coupled to online analysis by tandem mass spectrom-
etry (nLC-ESI-MS/MS) on an LTQ ion trap mass spectrometer
(ThermoElectron, San Jose, CA) equipped with a nanospray ion
source. From the eluted peptides, 10 µL were loaded onto a 0.1
mm × 150 mm Magic C18AQ column (Michrom Bioresources,
Inc., Auburn, CA) inline after a nanotrap column using the Paradigm
MS4 MDLC (Michrom Bioresources, Inc., Auburn, CA). Elution
of the peptides into the mass spectrometer was performed with a
linear gradient from 95% mobile phase A (2% acetonitrile, 0.1%
formic acid, 97.9% water) to 65% mobile phase B (10% water,
0.1% acetic acid, 89.9% acetonitrile) over 45 min and then to 95%
mobile phase B in 5 min. The peptides were detected in positive
ion mode using a data-dependent method in which the top seven
ions detected in an initial survey scan were selected for MS/MS
analysis. The MS/MS spectra were searched against the human
NCBI database using TurboSEQUEST in BioWorks, version 3.2
(ThermoElectron, San Jose, CA).
the wound, consistent with inhibition of directional cell migra-
tion (Figure 6, bottom center and right panels).11,13
Conclusions
We describe herein the synthesis and biochemical character-
ization of a biotinylated derivative of the Grb2-SH2 domain
inhibitor 1 and used this as a tool to assess its SH2 domain
binding selectivity. It is estimated that more than 100 different
SH2 domains are present in the human genome,24 each of which
recognizes phosphotyrosine (pTyr) in the context of three to
five additional residues, usually carboxyl-terminal to pTyr.25
Compound 1 is a peptidomimetic that was originally designed
to mimic the sequence Asn-pTyr-Val-Asn-Ile-Ile-Glu, where
pTyr has been substituted with a hydrolytically stable 4-(2′-
malonyl)phenylalanine residue.17,26 Several studies have re-
vealed the importance of Asn at the position Y + 2 within this
sequence for selective recognition by the SH2 domain of Grb2. In
addition to design criteria, it is essential to identify bona fide
intracellular targets for synthetic antagonists of protein-protein
interactions. Prior co-immunoprecipitation experiments provided
indirect evidence that Grb2 is the primary target of 1 and that 1
was able to disrupt Grb2-EGFR and Grb2-c-Met interactions.11,27
As a more efficient and direct alternative to screening 1 for
binding against all known SH2 domain containing proteins, we
devised a capture strategy coupled with mass spectral analysis
using a biotinylated analogue of the Grb2 SH2 domain binding
antagonist 1. This approach may be broadly useful to confirm
the selectivity of small molecule protein binding within the
growing field of targeted drug development. Biotin/streptavidin
based capture combined with proteomic analysis should be a
powerful tool for identifying adapter proteins along with their
associated proteins.
The findings that Grb2 SH2 domain binding antagonists cause
significant remodeling of the actin cytoskeleton and inhibit
PAK1 further confirm the importance of Grb2 as a regulator of
actin-based cell motility. Coincident with these effects, we have
previously shown that treatment of cells with 1 disrupts HGF-
stimulated physical association between Grb2 and focal adhesion
kinase (FAK), thereby inhibiting focal adhesion formation.13
Together, these results suggest that selective disruption of Grb2
SH2 domain-mediated signaling by a single agent may be a
viable strategy to target several molecular events required for
cell migration, which is critical to pathologic processes such as
tumor invasion, tumor angiogenesis, and metastasis.
Motility Assay. Cell migration was measured using modified
Boyden chambers as described.11 PC3M cells were seeded at
200 000 cells per chamber after treatment with the respective
compounds for 24 h. Cell migration was stimulated using 50 ng/
mL recombinant human HGF (R&D Systems, Minneapolis, MN).
Mean values from four fields observed by light microscopy (1 ×
1.4 mm) were calculated for each of triplicate wells per condition.
Immunofluorescence Videomicroscopy. PC3M and MDCK
cells plated in chamber slides were serum-deprived in the presence
or absence of 1 (1 µM) for 24 h. Cells were then removed from a
small portion of the monolayer using a pipet tip to simulate
wounding, and cells were incubated in the presence or absence of
HGF (50 ng/mL) to stimulate motility prior to fixation and
permeabilization. Slides were blocked with 5% bovine serum
albumin in PBS for 1 h and then incubated overnight at 4 °C with
Alexa488-conjugated phalloidin (Molecular Probes). Slides were
washed with PBS, nuclei were stained with 4′,6-diamidino-2-
phenylindole (DAPI, Molecular Probes), and digital epifluorescence
images were acquired using an inverted Olympus videomicroscope
and IPLab software (Scanalytics).
Experimental Section
Surface Plasmon Resonance Analysis of Drug-Protein Binding.
Grb2 SH2 domain-binding experiments were performed on a
Biacore T100 instrument (Biacore Inc., Piscataway, NJ) using
methods reported previously.28 All biotinylated Grb2 SH2 domain
proteins were expressed and purified by the Protein Expression
Laboratory and Protein Chemistry Laboratory, SAIC-Frederick. For
the T100 analysis using biotin-capture surfaces, streptavidin (Neu-
travidin, Pierce catalog number 31000) dissolved in 10 mM sodium
acetate, pH 4.5, was immobilized onto carboxymethyl 5′ dextran
surface (CM5 sensor chip, Biacore, Inc.) by amine coupling using
the immobilization wizard supplied by the Biacore T100 software
with a 5000 RU target on each of four flow cells. Lyophilized
biotinylated Grb2 SH2 domain protein was reconstituted in 50%
DMSO in H2O to make a stock solution of 1 mg/mL and stored at
-80 °C. A 20 µg aliquot of this solution was used for immobiliza-
tion by diluting in 1× PBS (phosphate-buffered saline, pH 7.4),
which was also used as the running buffer for Grb2 SH2 domain
capture. A manual method for building a capture surface was used
with a flow rate of 20 µL/min and a capture target of 2500 RU for
flow cells 2 and 4. Synthetic inhibitors were serially diluted in
Synthesis. Trifluoromethanesulfonic Acid 5-Hydroxynaph-
thalen-1-yl Ester (6). A solution of 5-dihydroxynaphthalene (10.18
g, 63.6 mmol), N-phenyltrifluoromethanesulfonimide (22.7 g, 63.6

7466 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23
Giubellino et al.
mmol), 2,4,6-collidine (8.4 mL, 63.6 mmol), and DMAP (930 mg,
7.62 mmol) in CH2Cl2/DMF (240 mL, 19:5 v/v) was refluxed for
12 h. The solvent was evaporated in vacuo, and the residue was
dissolved in EtOAc, washed with 5% HCl solution, H2O, brine,
and dried over Na2SO4. After evaporation of solvent, the remaining
residue was purified by silica gel flash chromatography to give 6
as brown solid (8.70 g, 47% yield). 1H NMR (DMSO-d6) δ 10.70
(s, 1H), 8.27 (d, 1H, J ) 8.4 Hz), 7.63 (d, 1H, J ) 7.2 Hz),
7.59-7.54 (m, 2H), 7.35 (d, 1H, J ) 8.4 Hz), 7.06 (d, 1H, J ) 7.4
Hz). MS (FAB) m/z 291 [M - H]-.
were treated as the preparative procedure of N-Boc Asn-9 to provide
N-Fmoc-protected 10 as a yellow solid (1.49 g, 79% yield). 1H
NMR (CD3OD) δ 8.30 (d, 1 H, J ) 8.1 Hz), 8.05 (dd, 1H, J ) 2.0
and 7.7 Hz), 7.82 (m, 1H). 7.74 (dd, 1 H, J ) 3.3 and 7.5 Hz),
7.50-7.42 (m, 3H), 7.36-7.31 (m, 3 H), 7.26-7.18 (m, 5H), 6.78
(dd, 1H, J ) 0.7 and 7.5 Hz), 4.62 (m, 1H), 4.34 (dd, 1H, J ) 6.5
and 10.6 Hz), 4.22 (dd, 1H, J ) 6.2 and 10.7 Hz), 4.05 (t, 1H, J
) 6.1 Hz), 3.27-3.21 (m, 2H), 3.00 (t, 2H, J ) 7.8 Hz), 2.86 (dd,
1H, J ) 6.7 and 15.5 Hz), 2.71 (dd, 1H, J ) 4.7 and 15.7 Hz),
1.96-1.83 (m, 5H), 1.75-1.27 (m, 7H). MS (FAB) m/z 663 (MH)+.
HRMS calcd for C39H42N4O6Na (M + Na)+, 685.2997; found,
685.3012.
3-(5-Hydroxynaphthalen-1-yl)acrylonitrile (7). To a solution
of 6 (7.10 g, 24.2 mmol), acrylonitrile (3.3 mL, 50.1 mmol), and
Et3N (7.5 mL, 50.1 mmol) were added Pd(OAc)2 (146 mg, 0.66
mmol) and DPPP (296 mg, 0.66 mmol). The mixture was refluxed
under argon (24 h), then cooled to room temperature, diluted with
EtOAc, and filtered through Celite. The filtrate was washed with
saturated NH4Cl solution and brine and dried over Na2SO4. After
evaporation of solvent, the remaining residue was purified by silica
gel flash chromatography to provide 7 (2.33 g, 49% yield) as a
(2S)-2-[(1-Aminocyclohexanecarbonyl)amino]-N1-[3-(5-hy-
droxynaphthalen-1-yl)propyl]succinamide (10). To a solution of
N-Fmoc-protected 10 (1.49 g, 2.25 mmol) in CH3CN (30 mL) was
added piperidine (3 mL), and the resulting solution was stirred at
room temperature for 2 h. After evaporation of solvent, the
remaining residue was purified by silica gel flash chromatography
to provide 10 as a yellow solid (847 mg, 86% yield). 1H NMR
(CD3OD) δ 8.07 (dd, 1H, J ) 3.1 and 6.3 Hz), 7.52 (d, 1H, J )
8.6 Hz), 7.34-7.27 (m, 3 H), 6.80 (dd, 1H, J ) 0.7 and 7.5 Hz),
4.65 (dd, 1H, J ) 5.9 and 6.6 Hz), 3.30-3.27 (m, 2H), 3.07-3.03
(m, 2H), 2.73 (dd, 1H, J ) 6.7 and 15.4 Hz), 2.66 (dd, 1H, J ) 5.9
and 15.4 Hz), 1.96-1.81 (m, 4H), 1.63-1.53 (m, 5H), 1.46-1.27
(m, 3H). MS (FAB) m/z 441 (MH)+. HRMS calcd for C24H33N4O4
(M + H)+, 441.2502; found, 441.2505.
1
yellow solid and 8 (1.52 g, 32% yield) as a yellow oil. H NMR
(CDCl3) δ 8.28 (d, 1H, J ) 8.5 Hz), 8.13 (d, 1H, J ) 16.4 Hz),
7.53 (d, 1H, J ) 8.6 Hz), 7.43-7.33 (m, 3H), 6.82 (d, 1H, J ) 7.5
Hz), 5.89 (d, 1H, J ) 16.4 Hz). MS (FAB) m/z 195 (M)+; 196
(MH)+.
5-(3-Aminopropyl)naphthalen-1-ol (9). Compound 7 (2.62 g,
13.4 mmol) was dissolved in THF (300 mL). To the stirred solution
was added LiAlH4 (1.09 g, 26.8 mmol) in several portions. The
mixture was stirred at room temperature overnight. EtOAc (40 mL)
was added followed by 10% NaOHaq (2 mL). The mixture was
stirred for additional 1 h. After filtration, the filtrate was removed
and the residue was purified by silica gel flash chromatography to
2-{4-[2-(1-(1S)-{2-Carbamoyl-1-[3-(5-hydroxynaphthalen-1-
yl)propylcarbamoyl]ethylcarbamoyl}cyclohexylcarbamoyl)-2-
(2S)-(9H-fluoren-9-ylmethoxycarbonylamino)ethyl]phenyl}malonic
Acid Di-tert-butyl Ester (12a). Compound 10 (603 mg, 1.37
mmol), compound 11a (826 mg, 1.37 mmol), HOBt (185 mg, 1.37
mmol), and DIPCDI (0.215 mL, 1.37 mmol) were treated as the
preparative procedure of N-Boc Asn-9 to provide 12a as a yellow
solid (1.058 g, 79% effective yield based on 247 mg of recovered
1
provide 9 as a yellow oil (1.47 g, 54% yield). H NMR (CD3OD)
δ 8.09 (dd, 1H, J ) 3.2 and 6.4 Hz), 7.51 (dd, 1H, J ) 0.8 and 8.7
Hz), 7.31-7.25 (m, 3H), 6.80 (dd, 1H, J ) 0.9 and 7.4 Hz), 3.05
(t, 2H, J ) 7.8 Hz), 2.72 (t, 2H, J ) 7.2 Hz). 1.92-1.85 (m, 2H);
MS (FAB) m/z 202 (MH)+.
1
10). H NMR (CD3OD) δ 8.18 (d, 1H, J ) 7.8 Hz), 8.05 (d, 1H,
J ) 9.1 Hz), 7.82-7.75 (m, 3H), 7.56-7.51 (m, 3H), 7.36-7.18
(m, 9H), 7.77 (d, 1H, J ) 7.5 Hz), 4.52 (m, 1H), 4.45 (s, 1H), 4.42
(m, 1H), 4.32-4.21 (m, 2H), 4.08 (t, 1H, J ) 6.7 Hz), 3.27-3.21
(m, 2H), 3.07-3.02 (m, 3H), 2.93-2.84 (m, 2H), 2.79 (dd, 1H, J
) 5.5 and 14.8 Hz), 1.99-1.92 (m, 4H), 1.77-1.70 (m, 2H), 1.44
(s, 18H), 1.42-1.24 (m, 4H). MS (FAB) m/z 1024.5 (MH)+. HRMS
calcd for C59H69N5O11Na (M + Na)+, 1046.4886; found, 1046.4886.
[2-(4-tert-Butoxyphenyl)-1-(1S)-{2-carbamoyl-1-[3-(5-hydrox-
ynaphthalen-1-yl)propylcarbamoyl]ethylcarbamoyl}cyclohexyl-
carbamoyl)ethyl]carbamic Acid 9H-Fluoren-9-ylmethyl Ester
(12b). Compound 10 (185 mg, 0.42 mmol), compound 11b (230
mg, 0.50 mmol), HOBt (68 mg, 0.50 mmol), and DIPCDI (0.079
mL, 0.50 mmol) were treated as the preparative procedure of N-Boc
(S)-tert-Butyl 4-Amino-1-(3-(5-hydroxynaphthalen-1-yl)pro-
pylamino)-1,4-dioxobutan-2-ylcarbamate (N-Boc Asn-9). To a
solution of 9 (874 mg, 4.34 mmol) in DMF (21 mL) was added an
active ester solution prepared by the reaction of N-Boc-Asn-OH
(1.00 g, 4.34 mmol), HOBt (586 mg, 4.34 mmol), and N,N′-
diisopropylcarbodiimide (DIPCDI) (0.682 mL, 4.34 mmol) in DMF
(14 mL) at room temperature for 10 min. The resulting solution
was stirred at room temperature (12 h). Solvent was evaporated,
and the remaining residue was dissolved in EtOAc, washed with
H2O, brine, and dried over Na2SO4. After evaporation of solvent,
the remaining residue was purified by silica gel flash chromatog-
raphy to provide N-Boc Asn-9 (1.56 g, 87% yield) as a yellow oil.
1H NMR (CD3OD) δ 8.08 (m, 1H), 7.52 (d, 1H, J ) 8.5 Hz),
7.32-7.27 (m, 3H), 6.80 (d, 1H, J ) 7.5 Hz), 4.42 (m, 1H),
3.32-3.26 (m, 2H), 3.04 (t, 2H, J ) 7.8 Hz), 2.68-2.61 (m, 2H),
1.91 (t, 2H, J ) 7.4 Hz), 1.43 (s, 9H). MS (FAB) m/z 416 (MH)+.
HRMS calcd for C22H30N3O5 (M + H)+, 416.2185; found,
416.2183.
1
Asn-9 to provide 12b as a yellow solid (274 mg, 74% yield). H
NMR (CDCl3) δ 8.02-7.95 (m, 2H), 7.69-7.60 (m, 2H), 7.42-7.38
(m, 2H), 7.32-7.28 (m, 2H), 7.21-7.17 (m, 6H), 7.00 (t, 1H, J )
7.5 Hz), 6.90-6.86 (m, 2H), 6.79-6.76 (m, 2H), 6.67 (d, 1H, J )
7.3 Hz), 4.60 (m, 1H), 4.30-4.25 (m, 3H), 4.03 (m, 1H), 3.32-3.21
(m, 2H), 2.98-2.89 (m, 2H), 2.74 (m, 1H), 2.48 (m, 1H), 2.02-1.06
(m, 14H), 1.22 (s, 9H). MS (FAB) m/z 882 (MH)+.
(2S)-2-Amino-N1-[3-(5-hydroxynaphthalen-1-yl)propyl]succi-
namide (N-Asn-9). N-Boc Asn-9 (1.56 g, 3.76 mmol) was dissolved
in TFA-CH2Cl2 (8 mL, 1:1 v/v), and the resulting solution was
stirred at room temperature for 1 h. After evaporation of solvent,
the remaining residue was purified by silica gel flash chromatog-
raphy to provide N-Asn-9 as a yellow oil (936 mg, 79% yield). 1H
NMR (CD3OD) δ 8.08 (dd, 1H, J ) 3.2 and 6.6 Hz), 7.53 (d, 1H,
J ) 8.5 Hz), 7.32-7.27 (m, 3H), 6.80 (dd, 1H, J ) 0.6 and 7.5
Hz), 3.70 (dd, 1H, J ) 5.0 and 8.1 Hz), 3.15-3.05 (m, 3H), 2.64
(dd, 1H, J ) 5.0 and 15.4 Hz), 2.48 (dd, 1H, J ) 8.1 and 15.4
Hz), 1.98-1.91 (m, 2H). MS (FAB) m/z 316 (MH)+. HRMS
calculated for C17H22N3O3 (M + H)+, 316.1675; found, 316.1661.
(S)-(9H-Fluoren-9-yl)methyl 1-(4-Amino-1-(3-(5-hydroxynaph-
thalen-1-yl)propylamino)-1,4-dioxobutan-2-ylcarbamoyl)cyclo-
hexylcarbamate (N-Fmoc 10). N-Asn-9 (901 mg, 2.86 mmol),
N-Fmoc-1-aminocyclohexenecarboxylic acid (1.043 g, 2.86 mmol),
HOBt (390 mg, 2.86 mmol), and DIPCDI (0.451 mL, 2.86 mmol)
2-{4-[2-(2S)-(tert-Butoxyoxalylamino)-2-(1-(1S)-{2-carbamoyl-
1-[3-(5-hydroxy-naphthalen-1-yl)propylcarbamoyl]ethylcarba-
moyl}cyclohexylcarbamoyl)ethyl]phenyl}malonic Acid Di-tert-
butyl Ester (13a). Compound 12a (609 mg, 0.595 mmol) and
piperidine (0.47 mL) were treated as preparative procedure of 10
to provide deprotected Fmoc 12a (477 mg). To a solution of this
crude material (477 mg, 0.595 mmol) in DMF (24 mL) was added
iPr2NEt (0.124 mL, 0.714 mmol) and tert-butyloxalyl (0.091 mL,
0.714 mmol) at 0 °C. The solution was stirred at room temperature
overnight. After evaporation of solvent, the remaining residue was
purified by silica gel flash chromatography to provide 13a as a
1
yellow oil (240 mg, 43% yield). H NMR (CDCl3) δ 8.05-8.01
(m, 2H), 7.65-7.61 (m, 2H), 7.46 (d, 1H, J ) 8.6 Hz), 7.27-7.25
(m, 4H), 7.08 (t, 1H, J ) 8.0 Hz), 7.04-7.01 (m, 2H), 6.74 (s,
1H), 6.62 (d, 1H, J ) 7.3 Hz), 6.46 (s, 1H), 5.58 (s, 1H), 4.67 (m,
1H), 4.56 (m, 1H), 4.41 (s, 1H), 3.39 (m, 1H), 3.28 (m, 1H),

Grb2 SH2 Domain Binding Antagonist
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23 7467
3.05-2.95 (m, 4H), 2.85 (dd, 1H, J ) 8.2 and 14.3 Hz), 2.46 (dd,
1H, J ) 4.9 and 15.0 Hz), 2.00-1.62 (m, 8H), 1.47 (s, 9H), 1.46
(s, 9H), 1.45 (s, 9H), 1.44-1.24 (m, 4H). MS (FAB) m/z 930
(MH)+.
(4R,5S,6S)-2-{4-[2-(1-{2-Carbamoyl-1-[3-(3S)-(5-{6-[5-(2-oxohexahy-
drothieno[3,4-d]imidazol-6-yl)pentanoylamino]hexanoyloxy}naphthalen-
1-yl)propylcarbamoyl}cyclohexylcarbamoyl)-2-(2S)-(oxalylamino)-
ethyl]phenyl}malonic Acid (3). A solution of 14a (127 mg, 0.1 mmol)
in a mixture of TFA-ethanedithol-H2O (1.0 mL, 3.8:0.1:0.1 v/v)
was stirred at room temperature for 1 h. The mixture was reduced
under vacuum to a volume of 0.25 mL. Diethyl ether (10 mL) was
added, giving a white solid. The solid was collected and purified
by HPLC as follows: a linear gradient over 25 min from 0.1% TFA
in 5% CH3CN/95% H2O to 0.1% TFA in 95% CH3CN/5% H2O;
analytical retention time ) 24.6 min; preparative retention time )
16.6 min. Liophilization provided 3 as a white solid (44.7 mg, 41%
N-[2-(4-tert-Butoxyphenyl)-(1-(1S)-{2-carbamoyl-1-[3-(5-hy-
droxynaphthalen-1-yl)propylcarbamoyl]ethylcarbamoyl}cyclo-
hexylcarbamoyl)ethyl]oxalamic Acid tert-Butyl Ester (13b).
Compound 12b (258 mg, 0.293 mmol) was treated as preparative
procedure of 13a to provide 13b as a yellow oil (174 mg, 80%
1
yield). H NMR (CD3OD) δ 8.03 (d, 1H, J ) 8.2 Hz), 7.52 (d,
1H, J ) 8.5 Hz), 7.32-7.21 (m, 3H), 7.05-7.02 (m, 2H),
6.85-6.81 (m, 2H), 6.75 (dd, 1H, J ) 0.7 and 7.5 Hz), 4.63 (dd,
1H, J ) 6.9 and 8.5 Hz), 4.52 (dd, 1H, J ) 5.0 and 7.2 Hz),
3.37-3.21 (m, 2H), 3.11-3.03 (m, 3H), 2.91 (dd, 1H, J ) 8.6 and
14.0 Hz), 2.86 (dd, 1H, J ) 7.2 and 15.4 Hz), 2.71 (dd, 1H, J )
5.0 and 15.5 Hz), 1.98-1.18 (m, 12H), 1.42 (s, 9H), 1.26 (s, 9H).
MS (FAB) m/z 788 (MH)+.
1
yield) in 99% purity as determined by HPLC. H NMR (DMSO-
d6) δ 8.75 (d, 1H, J ) 7.8 Hz), 8.24 (s, 1H), 7.93 (t, 2H, J ) 8.2
Hz), 7.71 (t, 1H, J ) 5.6 Hz), 7.63 (t, 1H, J ) 4.8 Hz), 7.45-7.41
(m, 2H), 7.37-7.35 (m, 2H), 7.30 (s, 1H), 7.20-7.15 (m, 5H),
6.84 (s, 1H), 6.34 (s, 1H), 4.66 (m, 1H), 4.52 (m, 1H), 4.31 (m,
1H), 4.21 (dd, 1H, J ) 5.0 and 7.7 Hz), 4.04 (dd, 1H, J ) 4.4 and
7.7 Hz), 3.13-3.07 (m, 3H), 3.04-2.93 (m, 6H), 2.75-2.69 (m,
3H), 2.62 (dd, 1H, J ) 6.2 and 15.5 Hz), 2.51-2.45 (m, 2H), 1.99
(t, 2H, J ) 7.4 Hz), 1.92-1.08 (m, 24H). MS (FAB) m/z 1100
(M)+, 1099 (M - H)-. HRMS calcd for C54H65N8O15Na3S (M -
3Na + 4H)+, 1101.4598; found, 1101.4502.
2-{4-[2-(1-(1S)-{2-Carbamoyl-1-[3-(5-hydroxy-naphthalen-1-
yl)propylcarbamoyl]ethylcarbamoyl}cyclohexylcarbamoyl)-2-(2S)-
(oxalylamino)ethyl]phenyl}malonic Acid (2). A solution of 13a (110
mg, 0.118 mmol) in a mixture of TFA-TES-H2O (2.0 mL, 3.7:
0.1:0.2 v/v) was stirred at room temperature for 1 h. After
evaporation of solvent, the remaining residue was purified by HPLC
as follows: a linear gradient over 25 min of from 0.1% TFA in 5%
CH3CN/95% H2O to 0.1% TFA in 95% CH3CN/5% H2O; analytical
retention time ) 24.4 min; preparative retention time ) 14.1 min.
Lyophilization provided 2 as a yellow solid (38 mg, 42% yield) in
97.6% purity as determined by HPLC. 1H NMR (DMSO-d6) δ 9.92
(s, 1H), 8.73 (d, 1H, J ) 7.8 Hz), 8.22 (s, 1H), 7.93-7.89 (m,
2H), 7.43-7.40 (m, 2H), 7.29-7.15 (m, 9H), 6.83 (s, 1H), 6.77
(d, 1H, J ) 7.3 Hz), 4.66 (m, 1H), 4.52 (s, 1H), 4.30 (m, 1H),
3.12-3.08 (m, 3H), 2.95-2.89 (m, 3H), 2.61 (dd, 1H, J ) 6.5 and
15.5 Hz), 2.43 (dd, 1H, J ) 7.0 and 15.5 Hz), 1.91-1.08 (m, 12H).
MS (FAB) m/z 761 (M)+, 760 (M - H)-. HRMS calcd for
C38H43N5O12Na (M - 2Na + 3H)+, 784.2800; found, 784.2810.
(4R,5S,6S)-2-{4-[2-(1-{2-Carbamoyl-1-[3-(3S)-(5-{6-[5-(2-oxo-
hexahydrothieno[3,4-d]imidazol-6-yl)pentanoylamino]hexanoyl-
oxy}naphthalen-1-yl)propylcarbamoyl}cyclohexylcarbamoyl)-2-
(2S)-(oxalylamino)ethyl]phenyl}malonic Acid Tri-tert-butyl Ester
(14a). To a solution of 13a (119 mg, 0.13 mmol) and 6-biotina-
midocaproic acid (94 mg, 0.26 mmol) in DMF (7 mL) was added
EDC (50 mg, 0.26 mmol) and DMAP (2.4 mg, 0.02 mmol) at 0
°C, and the resulting solution was stirred at room temperature for
12 h. The solvent was evaporated, and the residue was dissolved
in EtOAc, washed with H2O, brine, and dried over Na2SO4. After
evaporation of the solvent, the remaining residue was purified by
silica gel flash chromatography to provide 14a as a white solid
(127 mg, 78% yield). 1H NMR (CD3OD) δ 7.71 (dd, 1H, J ) 1.5
and 7.6 Hz), 7.50-7.39 (m, 4H), 7.29-7.19 (m, 5H), 4.72 (dd,
1H, J ) 6.7 and 8.5 Hz), 4.57 (dd, 1H, J ) 5.0 and 7.1 Hz), 4.47
(s, 1H), 4.46 (m, 1H), 4.32 (dd, 1H, J ) 4.8 and 7.7 Hz), 4.27 (dd,
1H, J ) 4.4 and 7.8 Hz), 4.13 (dd, 1H, J ) 4.4 and 7.9 Hz),
3.35-3.32 (m, 2H), 3.23-3.05 (m, 5H), 2.93-2.77 (m, 5H),
2.70-2.62 (m, 3H), 2.21-2.17 (m, 3H), 2.03-1.94 (m, 3H),
1.86-1.48 (m, 24H), 1.45 (s, 9H), 1.44 (s, 18H). MS (FAB) m/z
1269 (MH)+.
(4R,5S,6S)-6-[5-(2-Oxohexahydrothieno[3,4-d]imidazol-6-yl)-
pentanoylamino]hexanoic Acid 5-{3-[3-Carbamoyl-2-({1-[3-(4-
hydroxyphenyl)-2-(oxylylamino)propionylamino]cyclohexylcarbon-
yl}amino)propionylamino]propyl}naphthalen-1-yl Ester (4).
Compound 14b (9.0 mg) was treated as preparative procedure of
14a to provide 4 as a white solid (4.5 mg, 56% yield). Analytical
retention time ) 26.0 min; preparative retention time ) 18.1 min
1
(HPLC conditions as indicated for 14a). H NMR (DMSO-d6) δ
8.58 (d, 1H, J ) 7.8 Hz), 8.19 (s, 1H), 7.95 (d, 1H, J ) 8.8 Hz),
7.88 (d, 1H, J ) 7.6 Hz), 7.71 (m, 1H), 7.64 (m, 1H), 7.45 (d, 2H,
J ) 7.3 Hz), 7.40-7.36 (m, 2H), 7.30 (s, 1H), 7.19 (d, 1H, J )
7.4 Hz), 6.95 (d, 2H, J ) 7.4 Hz), 6.83 (s, 1H), 6.55 (d, 2H, J )
7.0 Hz), 6.38-6.32 (m, 2H), 4.56 (m, 1H), 4.31 (m, 1H), 4.21 (m,
1H), 4.04 (dd, 1H, J ) 4.6 and 7.5 Hz), 3.16-2.96 (m, 9H),
2.81-2.69 (m, 3H), 2.60 (dd, 1H, J ) 6.7 and 15.5 Hz), 2.61-2.56
(m, 2H), 2.01-1.10 (m, 26H). MS (FAB) m/z 1013 (M - H)-.
Acknowledgment. This research was supported by the
Intramural Research Program of the NIH, National Cancer
Institute, Center for Cancer Research. This project has been
funded in part with federal funds from the National Cancer
Institute, National Institutes of Health, under Contract N01-CO-
12400. The content of this publication does not necessarily
reflect the views or policies of the Department of Health and
Human Services, nor does mention of trade names, commercial
products, or organizations imply endorsement by the U.S.
Government.
References
(1) Blume-Jensen, P.; Hunter, T. Oncogenic kinase signalling. Nature
2001, 411, 355–365.
(2) Lowenstein, E. J.; Daly, R. J.; Batzer, A. G.; Li, W.; Margolis, B.;
Lammers, R.; Ullrich, A.; Skolnik, E. Y.; Bar-Sagi, D.; Schlessinger,
J. The SH2 and SH3 domain-containing protein GRB2 links receptor
tyrosine kinases to ras signaling. Cell 1992, 70, 431–442.
(3) Songyang, Z.; Shoelson, S. E.; McGlade, J.; Olivier, P.; Pawson, T.;
Bustelo, X. R.; Barbacid, M.; Sabe, H.; Hanafusa, H.; Yi, T. Specific
motifs recognized by the SH2 domains of Csk, 3BP2, fps/fes, GRB-
2, HCP, SHC, Syk, and Vav. Mol. Cell. Biol. 1994, 14, 2777–2785.
(4) Tari, A. M.; Lopez-Berestein, G. GRB2: A pivotal protein in signal
transduction. Semin. Oncol. 2001, 28, 142–147.
(4R,5S,6S)-6-[5-(2-Oxohexahydrothieno[3,4-d]imidazol-6-yl)-
pentanoylamino]hexanoic Acid 5-{3-{2-({1-2(tert-butoxyoxyly-
lamino)-3-(4-tert-butoxyphenyl)propionylamino]cyclohexylcarbon-
yl}amino)-3-carbamoylpropionylamino]propyl}naphthalen-1-
yl Ester (14b). Compound 13b (80 mg, 0.10 mmol) was treated as
preparative procedure of 14a to provide 14b as a white solid (9
1
mg, 42% effective yield based on 65 mg of recovered 13b). H
NMR (CD3OD) δ 8.04 (d, 1H, J ) 8.6 Hz), 7.71 (dd, 1H, J ) 1.7
and 7.5 Hz), 7.51-7.40 (m, 3H), 7.19 (d, 1H, J ) 7.0 Hz),
7.11-7.08 (m, 2H), 6.89-6.85 (m, 2H), 4.68 (dd, 1H, J ) 6.8 and
8.6 Hz), 4.56 (dd, 1H, J ) 5.0 and 7.1 Hz), 4.32 (dd, 1H, J ) 4.7
and 8.2 Hz), 4.13 (dd, 1H, J ) 4.5 and 7.8 Hz), 3.40-3.21 (m,
4H), 3.17-3.10 (m, 4H), 2.99-2.61 (m, 7H), 2.18 (t, 2H, J ) 7.3
Hz), 2.02-1.24 (m, 24H), 1.45 (s, 9H), 1.30 (s, 9H). MS (FAB)
m/z 1127 (MH)+.
(5) Kumar, R.; Gururaj, A. E.; Barnes, C. J. p21-activated kinases in
cancer. Nat. ReV. Cancer 2006, 6, 459–471.
(6) Schlaepfer, D. D.; Hanks, S. K.; Hunter, T.; Van der Geer, P. Integrin-
mediated signal transduction linked to Ras pathway by GRB2 binding
to focal adhesion kinase. Nature 1994, 372, 786–791.
(7) Mitra, S. K.; Mikolon, D.; Molina, J. E.; Hsia, D. A.; Hanson, D. A.;
Chi, A.; Lim, S. T.; Bernard-Trifilo, J. A.; Ilic, D.; Stupack, D. G.;
Cheresh, D. A.; Schlaepfer, D. D. Intrinsic FAK activity and Y925

7468 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23
Giubellino et al.
phosphorylation facilitate an angiogenic switch in tumors. Oncogene
2006, 25, 5969–5984.
(8) McLean, G. W.; Carragher, N. O.; Avizienyte, E.; Evans, J.; Brunton,
V. G.; Frame, M. C. The role of focal-adhesion kinase in cancersa
new therapeutic opportunity. Nat. ReV. Cancer 2005, 5, 505–515.
(9) Dharmawardana, P. G.; Peruzzi, B.; Giubellino, A.; Burke, J.; Bottaro,
D. P. Molecular targeting of growth factor receptor-bound 2 (Grb2)
as an anti-cancer strategy. Anti-Cancer Drugs 2006, 17, 13–20.
(10) Burke, T. R. Development of Grb2 SH2 domain signaling antagonists:
a potential new class of antiproliferative agents. Int. J. Pept. Res. Ther.
2006, 12, 33–48.
(11) Atabey, N.; Gao, Y.; Yao, Z. J.; Breckenridge, D.; Soon, L.; Soriano,
J. V.; Burke, T. R.; Bottaro, D. P. Potent blockade of hepatocyte
growth factor-stimulated cell motility, matrix invasion and branching
morphogenesis by antagonists of Grb2 Src homology 2 domain
interactions. J. Biol. Chem. 2001, 276, 14308–14314.
(12) Soriano, J. V.; Liu, N.; Gao, Y.; Yao, Z. Y.; Ishibashi, T.; Underhill,
C.; Burke, T. R.; Bottaro, D. P. Inhibition of angiogenesis by growth
factor receptor bound protein 2-Src homology 2 domain binding
antagonists. Mol. Cancer Ther. 2004, 3, 1289–1299.
(19) Oishi, S.; Karki, R. G.; Shi, Z. D.; Worthy, K. M.; Bindu, L.; Chertov,
O.; Esposito, D.; Frank, P.; Gillette, W. K.; Maderia, M.; Hartley, J.;
Nicklaus, M. C.; Barchi, J.; Fisher, R. J.; Burke, T. R. Evaluation of
macrocyclic Grb2 SH2 domain-binding peptide mimetics prepared by
ring-closing metathesis of C-terminal allylglycines with an N-terminal
beta-vinyl-substituted phosphotyrosyl mimetic. Bioorg. Med. Chem.
2005, 13, 2431–2438.
(20) Puto, L. A.; Pestonjamasp, K.; King, C. C.; Bokoch, G. M. p21-
activated kinase 1 (PAK1) interacts with the Grb2 adapter protein to
couple to growth factor signaling. J. Biol. Chem. 2003, 278, 9388–
9393.
(21) Singh, R. R.; Song, C.; Yang, Z.; Kumar, R. Nuclear localization and
chromatin targets of p21-activated kinase 1. J. Biol. Chem. 2005, 280,
18130–18137.
(22) O’Sullivan, G. C.; Tangney, M.; Casey, G.; Ambrose, M.; Houston,
A.; Barry, O. P. Modulation of p21-activated kinase 1 alters the
behavior of renal cell carcinoma. Int. J. Cancer 2007, 121, 1930–
1940.
(23) Rayala, S. K.; Kumar, R. Sliding p21-activated kinase 1 to nucleus
impacts tamoxifen sensitivity. Biomed. Pharmacother. 2007, 61, 408–
411.
(13) Giubellino, A.; Gao, Y.; Lee, S.; Lee, M. J.; Vasselli, J. R.; Medepalli,
S.; Trepel, J. B.; Burke, T. R.; Bottaro, D. P. Inhibition of tumor
metastasis by a growth factor receptor bound protein 2 Src homology
2 domain-binding antagonist. Cancer Res. 2007, 67, 6012–6016.
(14) Shi, Z. D.; Liu, H.; Zhang, M.; Worthy, K. M.; Bindu, L.; Yang, D.;
Fisher, R. J.; Burke, T. R. Synthesis of a C-terminally biotinylated
macrocyclic peptide mimetic exhibiting high Grb2 SH2 domain-
binding affinity. Bioorg. Med. Chem. 2005, 13, 4200–4208.
(24) Machida, K.; Mayer, B. J. The SH2 domain: versatile signaling module
and pharmaceutical target. Biochim. Biophys. Actaica et Biophysica
Acta, Proteins Proteomics 2005, 1747, 1–25.
(25) Songyang, Z.; Shoelson, S. E.; Chaudhuri, M.; Gish, G.; Pawson, T.;
Haser, W. G.; King, F.; Roberts, T.; Ratnofsky, S.; Lechleider, R. J.;
Neel, B. G.; Birge, R. B.; Fajardo, J. E.; Chou, M. M.; Hanafusa, H.;
Schaffhausen, B.; Cantley, L. C. SH2 domains recognize specific
phosphopeptide sequences. Cell 1993, 72, 767–778.
(26) Gao, Y.; Luo, J.; Yao, Z.-J.; Guo, R.; Zou, H.; Kelley, J.; Voigt, J. H.;
Yang, D.; Burke, T. 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.
(27) Gay, B.; Suarez, S.; Caravatti, G.; Furet, P.; Meyer, T.; Schoepfer, J.
Selective Grb2 SH2 inhibitors as anti-Ras therapy. Int. J. Cancer 1999,
83, 235–241.
(28) Oishi, S.; Shi, Z. D.; Worthy, K. M.; Bindu, L. K.; Fisher, R. J.; Burke,
T. R. Ring-closing metathesis of C-terminal allylglycine residues with
an N-terminal beta-vinyl-substituted phosphotyrosyl mimetic as an
approach to novel Grb2 SH2 domain-binding macrocyles. ChemBio-
Chem 2005, 6, 668–674.
(15) Sasaki, H.; Irie, R.; Hamada, T.; Suzuki, K.; Katsuki, T. Rational design
of Mn-salen catalyst (2): highly enantioselective epoxidation of
conjugated cis-olefins. Tetrahedron 1994, 50, 11827–11838.
(16) Cabri, W.; Candiani, I.; Bedeschi, A.; Santi, R. 1,10-Phenanthroline
derivatives: a new ligand class in the Heck reaction. Mechanistic
aspects. J. Org. Chem. 1993, 58, 7421–7426.
(17) Gao, Y.; Burke, T. R. Stereoselective preparation of L-4-(2′-malo-
nyl)phenylalanine suitably protected for Fmoc-based synthesis of
potent signal transduction inhibitory ligands. Synlett 2000, 134–136.
(18) Mikol, V.; Baumann, G.; Zurini, M. G. M.; Hommel, U. Crystal
structure of the SH2 domain from the adaptor protein SHC: a model
for peptide binding based on X-ray and NMR data. J. Mol. Biol. 1995,
254, 86–95.
JM800523U