Discovery of 3-aminobenzyloxycarbonyl as an N-terminal group conferring high affinity to the minimal phosphopeptide sequence recognized by the Grb2- SH2 domain

Article abstract of DOI:10.1021/jm9702185

The observation that anthranilic acid as N-terminal group produces a dramatic increase of the binding affinity of the phosphopeptide sequence Glu- pTyr-Ile-Asn for the Grb2-SH2 domain was rationalized by molecular modeling. The model, which invokes a stac

Full text of DOI:10.1021/jm9702185

J . Med. Chem. 1997, 40, 3551-3556
3551
Discover y of 3-Am in oben zyloxyca r bon yl a s a n N-Ter m in a l Gr ou p Con fer r in g
High Affin ity to th e Min im a l P h osp h op ep tid e Sequ en ce Recogn ized by th e
Gr b2-SH2 Dom a in
Pascal Furet,* Brigitte Gay, Carlos Garc´ıa-Echeverr´ıa, J oseph Rahuel, Heinz Fretz, J oseph Schoepfer, and
Giorgio Caravatti
Novartis Pharma Inc., Oncology Research Department, CH-4002 Basel, Switzerland
Received March 31, 1997X
The observation that anthranilic acid as N-terminal group produces a dramatic increase of
the binding affinity of the phosphopeptide sequence Glu-pTyr-Ile-Asn for the Grb2-SH2 domain
was rationalized by molecular modeling. The model, which invokes a stacking interaction
between the N-terminal group and the SH2 domain residue Arg RA2, was subsequently used
to design the 3-aminobenzyloxycarbonyl N-terminal group. The latter confers high affinity
(IC50 ) 65 nM in an ELISA assay) to the minimal sequence pTyr-Ile-Asn recognized by the
Grb2-SH2 domain.
Ta ble 1. Structure-Activity Relationships of Reported
Phosphopeptides
In tr od u ction
Intervention in the signal transduction pathways of
tyrosine kinase growth factor receptors is a contempo-
rary theme in anticancer drug research.1-3 Many efforts
have focused on preventing tyrosine phosphorylation
and the consequent intracellular signal propagation by
chemical inhibition of the kinase enzymatic activity of
these receptors.4,5 Downstream in the signaling cas-
cade, blocking the interaction between the phosphoty-
rosine (pTyr)-containing activated receptors and the Src
homology 2 (SH2) domain of the growth factor receptor-
bound protein 2 (Grb2) constitutes another approach of
potential interest in the search for new antitumor
agents.6,7 This strategy, which targets a key component
of the ras activation pathway,8,9 is conceptually attrac-
tive. Starting from the minimal peptide sequence pTyr-
Ile-Asn recognized by the Grb2-SH2 domain,10 we have
initiated a medicinal chemistry project in this direction.
A first objective in the project was to improve the
binding affinity of the minimal sequence to progress
toward the identification of low molecular weight com-
pounds that can efficiently disrupt these protein-
protein interactions. As part of this effort, we report
here the rationalization by molecular modeling of the
unexpected beneficial effect on affinity observed upon
introduction of a particular fluorophore at the N-
terminus of a Grb2-SH2 phosphopeptide ligand. We
also describe how this finding has been exploited to
design an N-terminal capping group of phosphotyrosine
that increases the binding affinity of the unprotected
minimal sequence by 3 orders of magnitude.
no.
sequence
IC50 (µM)a
7.9 ( 0.46
1
2
3
4
5
6
H-Glu-pTyr-Ile-Asn-NH2
Abz-Glu-pTyr-Ile-Asn-NH2
Ac-pTyr-Ile-Asn-NH2
H-pTyr-Ile-Asn-NH2
(3-amino)Z-pTyr-Ile-Asn-NH2
Z-pTyr-Ile-Asn-NH2
0.022 ( 0.005
8.64 ( 1.56
56.9 ( 6.74
0.065 ( 0.0045
6.35 ( 0.59
a
IC50 concentration to inhibit the binding of the phosphorylated
C-terminal intracellular domain of EGFR to the Gbr2-SH2 domain.
purpose, reference phosphopeptides bearing the anthra-
nilic acid (Abz) fluorophore at the N-terminus were
synthesized. As illustrated by a comparison of the IC50
values of compounds 1 and 2 (Table 1), it turned out
that Abz, when appended at the N-terminus of the
amino acid preceding the phosphotyrosine, produced a
dramatic, unexpected increase of the binding affinity of
the phosphopeptides for the Grb2-SH2 domain. At the
time this result was obtained, we had not yet deter-
mined the X-ray crystal structure of the Grb2-SH2
domain in complex with a phosphopeptide.11 Still, using
the available coordinates of an homologous SH2 domain,
that of the protein p56lck (Lck) complexed with the high-
affinity phosphopeptide Glu-Pro-Gln-pTyr-Glu-Glu-Ile-
Pro-Ile-Tyr-Leu,12 we could derive a hypothesis for the
structural basis of this effect.
Wh y th e Lck Str u ctu r e Is a n Ap p r op r ia te Mod el.
Specificity in peptide ligands of SH2 domains is deter-
mined by the residues C-terminal to the phospho-
tyrosine.13 For instance, while for the Grb2-SH2 do-
main an asparagine at position pTyr+2 of the se-
quence14 is the dominant recognition feature, the Lck-
SH2 domain preferentially binds peptides presenting
the motif pTyr-Glu-Glu-Ile.15,16 As confirmed by X-ray
and NMR structural investigations,11,12,17-19 this sug-
gests that the three-dimensional structures of SH2
domains present significant differences in the regions
involved in binding the C-terminal part ot their phos-
phopeptide ligands. As a corollary, because they are not
critical for determining specificity, less variation is
expected in the N-terminal part binding regions as well
as in the phosphotyrosine recognition pockets. In
particular, the analysis detailed below gave us confi-
Resu lts a n d Discu ssion
An th r a n ilic Acid Effect. To assess the binding
affinity of ligands for the Grb2-SH2 domain, we have
set up an ELISA-type assay that measures their ability
to inhibit the binding of the phosphorylated C-terminal
intracellular domain of the epidermal growth factor
receptor (EGFR) to this SH2 domain.11 In the early
phase of the project, we envisaged to develop an
alternative binding test based on fluorescence. For that
* To whom correspondence should be addressed. E-mail: furet@
chbs.ciba.com.
X Abstract published in Advance ACS Abstracts, September 15, 1997.
S0022-2623(97)00218-5 CCC: $14.00 © 1997 American Chemical Society

3552 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 22
Furet et al.
Ra tion a liza tion of th e Effect. Thus, taking the
X-ray structure of the Lck-SH2 domain complex as a
model, we sought to understand the origin of the
anthranilic acid effect. The Lck structure shows that
amino acid Arg RA2 makes multiple contacts with the
phosphopeptide ligand. Its side chain forms hydrogen
bonds with the phosphate group and, as already men-
tioned, the backbone carbonyl of residue pTyr-1. A less
conventional amino-aromatic hydrogen bond, also seen
in the structures of other SH2 domains,22,23 is formed
between the η1 nitrogen and the phosphotyrosine ring.
In addition, a weak van der Waals contact exists
between the guanidinium moiety and the ring of the
proline residue in position pTyr-2 of the phosphopeptide
sequence. These different interactions are represented
in Figure 1. When the modeling experiment consisting
of replacing the pTyr-2 proline residue by the Abz
N-terminal group in the X-ray structure was performed,
it immediately suggested a reason why this structural
change could lead to a gain in binding affinity. As
depicted in Figure 2, constructing Abz at the location
of the proline ring in the X-ray structure followed by
energy minimization places the N-terminal group in an
optimal position to make a favorable planar π-π stacking
interaction with the guanidinium moiety of Arg RA2.
Stacking interactions between arginine side chains and
aromatic or heteroaromatic moieties have often been
observed in structural biology,24 and calculations on
model complexes suggest that they can produce sub-
stantial stabilization energy.25 In fact, statistical analy-
sis of protein structures indicates that arginine and
aromatic amino acid side chains interact more fre-
quently in the stacked geometry than in the perpen-
dicular (amino-aromatic hydrogen bond) one.25 Stack-
ing gives to an arginine guanidinium group that has
fulfilled its hydrogen bonding potential, as it is the case
for Arg RA2, an additional possibility of interaction.
Recently, this type of interaction between an electron-
rich aromatic moiety and the electron-poor guanidinium
group has also been invoked to rationalize the structure-
activity relationships of some matrix metalloproteinase
inhibitors.26 To check if other low-energy binding
geometries could exist, a local Monte Carlo search of
the conformations accessible to the N-terminal part of
the Abz-protected phosphopeptide ligand in complex
with the Lck-SH2 domain was carried out. The confor-
mation corresponding to the stacking interaction be-
tween Abz and Arg RA2 came out as the global mini-
mum, reinforcing the plausibility of our hypothesis
regarding the origin of the anthranilic acid effect.
F igu r e 1. Interactions between residue Arg RA2 (yellow) and
the phosphopeptide ligand in the Lck-SH2 domain X-ray
structure. Conventional hydrogen bonds are represented as
thin white lines. The amino-aromatic hydrogen bond formed
with the π system of pTyr is pictured in the form of pink dots
while the van der Waals contact with proline pTyr-2 as blue
dots.
dence that the Lck-SH2 X-ray structure provided an
appropriate frame to model the interactions between the
Grb2-SH2 domain and its ligands as far as phospho-
tyrosine and the N-terminal residues are concerned.
According to sequence alignments,18 the amino acids
of the Lck-SH2 domain seen to interact with the ligand
phosphotyrosine and its N-terminal residues in the
X-ray structure are absolutely conserved in the Grb2-
SH2 domain. These are Arg RA2, Arg âB5, Ser âC3,
Ser BC2, Ser âB7, Lys âD6, Glu BC1, and His âD4.21
Additional support to the notion that the three-
dimensional structures of the Lck and Grb2-SH2 do-
mains in complex with ligands are similar in this region
came from structure-activity relationships reported in
Table 1. In the crystal structure of the Lck-SH2
domain, the only contact existing between the glutamine
pTyr-1 phosphopeptide residue and the protein is a
bifurcated hydrogen bond interaction involving the
ligand backbone carbonyl and the terminal nitrogens
of the guanidinium moiety of Arg RA2 (see Figure 1).
The side chain of the pTyr-1 residue projects away from
the protein in the direction of solvent space. Consistent
with the idea that a similar situation exists in ligated
Grb2-SH2, comparison of the respective IC50 values of
compounds 1, 3, and 4 indicates that the pTyr-1 amino
acid contributes to affinity only through its backbone
carbonyl function. A simple acetyl N-protection of the
phosphotyrosine can replace the full pTyr-1 residue
without loss of affinity (peptide 3 compared to 1).
However, removal of the protecting group causes a 7-fold
increase of the IC50 value (peptide 4 compared to 3) by
abolishing, presumably, the hydrogen bond interaction
above mentioned.
As can be seen in Figure 2, in our model of the
complex in which Abz stacks on Arg RA2, the 2-amino
substituent of the protecting group is within hydrogen-
bonding distance of one of the pTyr phosphate oxgen
atoms. We speculated that this intramolecular hydro-
gen bond in the ligand, by locking its N-terminal part
in the conformation recognized by the SH2 domain, also
contributed to the exceptional increase in affinity con-
ferred by Abz.
Design of a 3-Am in oben zyloxyca r bon yl N-Ter -
m in a l Gr ou p . As discussed in the preceding sections,
the side chain of the ligand pTyr-1 amino acid does not
contribute to affinity. According to our model, the
glutamic acid pTyr-1 residue of 2, apart from its main
chain carbonyl group, acts as a neutral spacer connect-

3-Aminobenzyloxycarbonyl as an N-Terminal Group
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 22 3553
F igu r e 2. Model (stereoview) of the stacking interaction between the phosphopeptide Abz N-terminal group and the SH2 domain
residue Arg RA2 (yellow). The hydrogen bond between the 2-amino substituent of Abz and the phosphate group of pTyr is
represented as a dashed line.
ing the important pTyr and Abz moieties. This sug-
gested that it should be possible to design an appropri-
ate N-terminal group that could mimic the stacking
interaction of Abz and be directly attached to the
phosphotyrosine, thus eliminating one unnecessary
amino acid in potent compound 2.
Inspection of the model in Figure 2 revealed that the
part of the Abz group in contact with the guanidinium
moiety of Arg RA2 extends beyond the phenyl ring. It
also includes a surface patch belonging to the atoms
forming the pseudo-six-membered ring resulting from
the intramolecular hydrogen bond between the 2-amino
substituent and the carbonyl group (Abz was con-
structed in this conformation as described in the mo-
lecular modeling part of the Experimental Section). An
electrostatic potential map of the Abz fragment (Figure
3) clearly shows that the electron-rich region corre-
sponding to negative values of the molecular electro-
static potential and having affinity for the electron-poor
F igu r e 3. Electrostatic potential map of the Abz N-terminal
group. The electrostatic potential is displayed as isovalue
contour lines on the van der Waals surface of the molecule
(negative values, blue; positive values, red). The intramolecu-
guanidinium group is centered in the middle of the
bicyclic system formed by the phenyl and the pseudo-
six-membered rings. Interactive modeling work sug-
gested that the phenyl ring of a 3-amino-substituted
benzyloxycarbonyl (Z) group, appended at the N-termi-
nus of phosphotyrosine, could appropriately mimic the
bicycle-centered part of Abz in contact with the guani-
dinium moiety of Arg RA2 without compromising the
other ligand-SH2 domain interactions. The resulting
energy-minimized model is shown in Figure 4.
The synthetic realization of this idea, compound 5,
confirmed the validity of the design. With an IC50 value
of 65 nM (Table 1), 5 is almost equipotent to 2 while
being one amino acid residue shorter. Three orders of
magnitude in binding affinity were gained with respect
to the unprotected reference tripeptide sequence 4.
To assess the importance of the 3-amino substituent
in producing this dramatic increase in affinity, 6, the
unsubstituted Z-protected analogue of 5, was also
synthesized. As shown in Table 1, the amino substitu-
ent is essential. No gain in affinity was obtained with
6 compared to the acetyl-protected peptide 3. In the
model, the most obvious role of the amino substituent
lar hydrogen bond formed by the 2-amino substituent and the
carbonyl group appears as a dashed line.
is to lock the N-terminal part of the peptide in the
stacking conformation by forming a strong intramolecu-
lar hydrogen bond with the charged phosphate group
of pTyr. However, the electron-donating property of the
substituent, by increasing the electronic density on the
phenyl ring of Z with the result of strengthening the
interaction with the electron-deficient arginine guani-
dinium group, is very likely to also have a contribution
to the enhancement of affinity. Thus, the large increase
in potency induced by Abz or its (3-amino)Z mimetic
seems to depend on the synergy between an enthalpic
(π-π stacking strengthened by an electron-donating
substituent) and an entropic (conformational restriction
by an internal hydrogen bond) favorable effect.
Repeating the modeling work above described using
our recently solved X-ray crystal structure of the Grb2-
SH2 domain11 gave the same results in accordance with

3554 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 22
Furet et al.
F igu r e 4. Model (stereoview) of the stacking interaction between the designed phosphopeptide (3-amino)Z N-terminal group
and the SH2 domain residue Arg RA2 (yellow). The hydrogen bond between the 3-amino substituent of the N-terminal goup and
the phosphate moiety of pTyr is represented as a dashed line.
with petroleum ether (2 × 25 mL); the ether extracts were
the expected high degree of similarity observed between
the Lck and Grb2-SH2 structures in the region of the
phosphotyrosine binding pocket.
discarded. The aqueous phase was acidified with a 5% solution
of citric acid and extracted with ethyl acetate (3 × 30 mL).
The ethyl acetate extracts were pooled, washed with water (3
× 30 mL), dried over anhydrous Na2SO4, and evaporated in
Con clu sion
vacuo. Yield: 4.0 g (18 mmol, 90%). TLC: Rf ) 0.72 (chloro-
form:methanol:water:acetic acid ) 750:270:50:5, v/v/v/v). 1H
NMR (CD3OD): δ 1.50 (s, 9H, t-Bu), 4.56 (s, 2H), 6.98 (d, J )
7.5 Hz, 1H), 7.21 (t, J ) 7.5 Hz, 1H), 7.30 (d, J ) 7.5 Hz, 1H),
7.40 (s, 1H).
Molecular modeling based on the X-ray structure of
an SH2 domain homologue to that of Grb2 allowed us
to formulate a hypothesis regarding the structural basis
of the anthranilic acid effect discovered by serendipity.
The resulting model was successfully used to design
an N-terminal capping group of phosphotyrosine that
can impart high affinity to the minimal tripeptide
sequence recognized by the Grb2-SH2 domain. The
small size and the high potency of the phosphopeptide
thus obtained represent a major advance in the search
for low molecular weight compounds that can block a
crucial protein-protein interaction in the signal trans-
duction pathways of tyrosine kinase growth factor
receptors.
Since the anthranilic acid effect is assumed to origi-
nate in an interaction with the well-conserved Arg RA2
amino acid, it should also be observed for SH2 domains
other than Grb2. We are currently in the process of
checking this idea by applying the same N-terminal
modifications to phosphopeptide sequences specific to
various other SH2 domains
3-[N-(ter t-Bu toxyca r bon yl)a m in o]ben zyl 4-Nitr op h en -
yl Ca r bon a te. A solution of 3-[N-(tert-butoxycarbonyl)amino]-
benzyl alcohol (1.0 g, 4.5 mmol) in anhydrous pyridine (18 mL)
was cooled in an ice-water, bath and 4-nitrophenyl chloro-
formate (0.9 g, 4.5 mmol) was added with stirring. The
solution was stirred for 17 h at room temperature and then
added to a mixture of diisoporpyl ether/petroleum ether (1:1,
v/v; 200 mL). The ether mixture was extracted with brine (1
× 50 mL) and then with water (7 × 50 mL), dried over
anhydrous Na2SO4, and evaporated in vacuo. The crude
compound was purified by flash chromatography on silica gel,
using dichloromethane as eluent. Yield: 1.25 g (3.2 mmol,
71%). Mp: 93.3-94.3 °C. TLC: Rf ) 0.66 (chloroform/
methanol ) 95:5, v/v). MS: m/z 387 (M - H). 1H NMR
(CDCl3): δ 1.52 (s, 9H, t-Bu), 5.28 (s, 2H), 6.55 (s, 1H, NH),
7.10 (d, J ) 7.5 Hz, 1H), 7.30 (m, 2H), 7.40 (d, J ) 7.5 Hz,
2H), 7.58 (s, 1H), 8.38 (d, J ) 7.5 Hz, 2H).
P ep tid e Syn th esis. Peptides 1, 5, and 6 were synthesized
manually on
a 4-[[(2′,4′-dimethoxyphenyl)amino]methyl]-
phenoxy resin,28 employing the fluorenylmethoxycarbonyl
strategy. Fmoc-removal was with piperidine/DMA (1:4, v/v;
6 × 2 min), followed by washing with MeOH (3 × 1 min), NMP
(2 × 1 min), MeOH (3 × 1 min), and NMP (3 × 2 min). The
required Fmoc derivatives of asparagine, isoleucine, and
glutamic acid were incorporated using their 2,4,5-trichlorophe-
nyl esters (2 equiv) in the presence of HOBt (2 equiv) and
DIEA (0.75 equiv). Asparagine and glutamic acid side chains
were protected with the trityl and the tert-butyl group,
respectively. The incorporation of NR-Fmoc-Tyr(PO3H2)-OH
(3 equiv)27,29 was accomplished with BOP/HOBt (1:1; 3 equiv;
first coupling)30 and HATU (3 equiv; second coupling)31 in the
presence of DIEA (7 equiv). 3-[N-(tert-Butoxycarbonyl)amino]-
benzyl 4-nitrophenyl carbonate (3 equiv) was coupled to the
N-terminal residue of the peptide resin (peptide 5) in the
presence of an equimolar amount of DIEA in NMP during 17
h at room temperature. Benzyl chloroformate (3 equiv) was
coupled to the N-terminal residue of the peptide resin (peptide
6) in the presence of DIEA (6 equiv) in NMP during 17 h at
room temperature.
Exp er im en ta l Section
4-[[[(2′,4′-Dimethoxyphenyl)Fmoc]amino]methyl]phenoxy]-
polystyrene resin (1% DVB cross-linked, 0.4 mequiv/g, 100-
200 mesh) and BOP and HOBt reagents were purchased from
NovaBiochem (La¨ufelfingen, Switzerland). TPTU reagent was
from Senn Chemicals AG (Dielsdorf, Switzerland). HATU
reagent was from PerSeptive Biosystems (Hamburg, Ger-
many). The required Fmoc derivatives were obtained using
standard protocols. NR-Fmoc-Tyr(PO3H2)-OH was synthesized
as described previously.27 2-Aminobenzoic acid, trifluoroacetic
acid, diisopropylethylamine, and piperidine were from Fluka
(Buchs, Switzerland). N,N-Dimethylacetamide was from Mer-
ck (Hohenbrunn, Germany) and N-methylpyrrolidin-2-one was
from SDS (Peypin, France). All commerical chemicals used
were of the highest quality available (AR grade or higher).
3-[N-ter t-Bu toxyca r bon yl)a m in o]ben zyl Alcoh ol. Di-
tert-butyl dicarbonate (13.4 mL, 60 mmol) was added to a
solution of 3-aminobenzyl alcohol (2.46 g, 20 mmol) in THF:1
N NaOH (125 mL, 1:1, v/v), and stirring was continued until
completion of the reaction. The alkaline solution was extracted
Peptides 2-4 were synthesized on a Milligen 9050 auto-
mated peptide synthesizer, starting with an Fmoc-PAL-PEG-

3-Aminobenzyloxycarbonyl as an N-Terminal Group
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 22 3555
resin32 for establishing the C-terminal carboxamide, and using
chemical protocols based on the fluorenylmethoxycarbonyl
chemistry. The required Fmoc-amino acids (3 equiv) were
coupled using their 2,4,5-trichlorophenyl esters (single cou-
pling) with minimum reaction times of 30 min. The amino
acid side chains are protected as described above. Double
coupling with TPTU (3 equiv)33 was carried out for glutamic
acid (peptide 2). The incorporation of NR-Fmoc-Tyr(PO3H2)-
OH (3 equiv) was accomplished with BOP/HOBt (1:1; 3 equiv;
first coupling) and HATU (3 equiv; second coupling) in the
presence of DIEA (6 equiv). 2-Aminobenzoic acid (peptide 2)
was incorporated with BOP/HOBt as described above. 3-[N-
(tert-Butoxycarbonyl)amino]benzyl 4-nitrophenyl carbonate (3
equiv) was coupled manually to the N-terminal residue of the
peptide resin (peptide 5) in the presence of an equimolar
amount of DIEA in NMP during 17 h at room temperature
The complete peptide resins obtained after the last coupling
step were simultaneosly deprotected and cleaved by treatment
with TFA/H2O (95:5, v/v) for 3 h at room temperature. The
filtrate from the cleavage reaction was precipitated in diiso-
propyl ether/petroleum ether (1:1, v/v, 0 °C), and the precipi-
tate was collected by filtration. The crude peptides were
purified by medium-pressure liquid chromatography using a
C18-column (Merck LICHROPREP RP-18, 15-25 µm bead
diameter; detection at 215 nm) eluted with an acetonitrile-
water gradient containing 0.1% of TFA. Fractions shown by
HPLC to be >95% pure were pooled and lyophilized to provide
the title compounds as white powder TFA salts. Mass-spectral
analysis (MALDI-TOF) revealed a molecular mass within 0.1%
of the expected value (negative-ion mode): 615.9 (1, calcd
615.6, C24H36N6O11P1); 734.5 (2, calcd 734.7, C31H41N7O12P1);
528.8 (3, calcd 528.5, C21H31N5O9P1); 486.4 (4, calcd 486.5,
C19H29N5O8P1); 635.7 (5, calcd 635.6, C27H36N6O10P1); 620.8 (1,
calcd 620.6, C27H35N5O10P1). The purity of the peptide was
oxy ribonucleotides were prepared on an automated Applied
Biosystems Model 392 DNA synthesizer, using phosphora-
midite chemistry: (A) 5′ -tatagaattccagcgctaccttgctattcagggg-
3′; (B) 5′ -TATAAAGCTTTCATGCTCCAATAAATTCACTGC-
TTTG-3′. (A) + (B) were used as PCR primers to amplify
segments of the human EGFR cDNA sequence corresponding
to nucleotides 3112-3816. The PCR fragment was cleaved
with EcoRI and HindIII from Boehringer Mannheim before it
was ligated with EcoRI + HindIII-cleaved pMal-C2 DNA. The
ligation was used to transform E. coli SURE competent cells
(Stratagene) and transformants were selected at 37 °C on LB
agar plates supplemented with 100 g/mL ampicillin. Plasmid
DNA was isolated from individual ampicillin-resistant colonies
and analyzed by restriction endonuclease digestion to identify
the desired recombinants. Expression of the MBP-EGFR
fusion protein was performed essentially according to New
England Biolabs protocols.
EGF R Assa y. The assay has already been described
elsewhere.11 Briefly, phosphorylated MBP-EGFR immobilized
on a solid phase (polystyrene microtiter plates, NUNC MAX-
YSORB) was incubated with a GST/Grb2-SH2 fusion protein
capable of binding to it, in the presence of a phosphopeptide
or buffer. Bound SH2 was detected with polyclonal rabbit anti-
GST antibody. Following washing, horse raddish peroxidase-
conjugated mouse anti-rabbit antibody was added. Peroxydase
activity is monitored at 655 nm on a plate reader by adding
100 L/well of a solution of tetramethylbenzidine as substrate.
Da ta An a lysis. Peptide inhibitors effects were calculated
as a percentage of the reduction in absorbance in the presence
of each peptide inhibitor concentration compared to absorbance
obtained with GST/SH2 in the absence of peptide inhibitor.
Dose-response relationships were constructed by nonlinear
regression of the competition curves with Grafit (Erithacus
Software, London, U.K.). Fifty percent inhibitory (IC50) con-
centrations were calculated from the regression lines.
verified by reversed-phase analytical HPLC on a Nucleosil C18
-
column (250 × 4 mm, 5 µm, 100 Å): (A) linear gradient over
10 min of MeCN/0.09% TFA and H2O/0.1% TFA from 5:95 to
65:35; (B) linear gradient over 10 min of MeCN/0.09% TFA
and H2O/0.1% TFA from 1:49 to 3:2; (C) linear gradient over
10 min of MeCN/0.09% TFA and H2O/0.1% TFA from 1:49 to
2:3, flow rate 2.0 mL/min, detection at 215 nm; single peak at
tR ) 4.20 min (1, A); tR ) 5.38 min (2, B); tR ) 5.62 min (3, C);
tR ) 3.99 min (4, C); tR ) 5.54 min (5, B); tR ) 7.37 min (6, B).
Identical results were obtained by capillary electrophoresis
(Beckmann, P/ACE System 5000): normal capillary 75 µm ×
50 cm, sodium phosphate 0.05 M (pH 9.3) and/or sodium
tetraborate 0.05 M (pH 9.3), 10 min at 25 kV, detection at 214
nm, T ) 23 °C.
Abbr evia tion s. Abbreviations for amino acids and nomen-
clature of peptides structures follow the recommendations of
the IUPAC-IUB Commission on Biochemical Nomenclature
(Eur. J . Biochem. 1984, 138, 9). Other abbreviations are as
follows: Abz, anthranilic acid or anthranilamide; Fmoc, 9-fluo-
renylmethyloxycarbonyl; HATU, N-[[(dimethylamino)-1H-
1,2,3-triazolo[4,5-b]pyridin-1-yl]methylene]-N-methylmeth-
anaminium hexafluorophosphate N-oxide; Boc, tert-butyl-
oxycarbonyl; BOP, (1H-benzotriazol-1-yloxy)tris(dimethylami-
no)phosphonium hexafluorophospahte; DIEA, diisopropylethyl-
amine; DMA, N,N-dimethylacetamide; HOBt, 1-hydroxyben-
zotriazole; HPLC, high-performance liquid chromatography;
PAL, tris(alkoxy)benzylamide linker; PEG, poly(ethylene gly-
col); NMR, nuclear magnetic resonance; MALDI-TOF, matrix-
assisted laser-desorption ionization time-of-flight mass spec-
trometry; NMP, N-methylpyrrolidin-2-one; Rf, ratio of fronts
in thin layer chromatogrphy employing silica gel plates; TFA,
trifluoroacetic acid; THF, tetrahydrofuran; TLC, thin layer
chromatography; TPTU, 2-(2-oxo-1(2H)-pyridyl)-1,1,3,3-tet-
ramethyluronium tetrafluoroborate; pTyr, phosphotyrosine; tR,
retention time.
Molecu la r Mod elin g. The following methods as imple-
mented in MacroModel v.4.034 were used: (i) energy minimiza-
tion with the AMBER35 force field (dielectric constant of 4r
for the electrostatics), (ii) conformational search with the
Monte Carlo/energy minimization procedure36 (advanced pro-
tocol), (iii) calculation and display of electrostatic potential on
the molecular van der Waals surface37 employing MNDO38
(Mulliken population analysis) atomic charges computed with
the program MOPAC v.6.0.39
In the Figure 2 model, Abz was constructed in the confor-
mation presenting an intramolecular hydrogen bond between
the 2-amino substituent and the carbonyl group as observed
in the X-ray crystal structures of small organic molecules
containing this moiety (Cambridge crystallographic database40
codes J EGSOD and J IXCIC). In the Monte Carlo search and
minimizations, only the water molecules seen to mediate
hydrogen bonds between the ligand and the Lck-SH2 domain
in the X-ray structure were kept. The ligand, these water
molecules as well as the protein residues within a distance of
5 Å of the initial position of the ligand were allowed to move
freely upon energy minimization, while those at a distance
between 5 and 8 Å were constrained by application of a
parabolic force constant of 50 kJ /Å. Residues beyond 8 Å were
ignored. The Monte Carlo search consisted of 5000 steps
where the main chain angle of pTyr and all the torsion angles
of Abz and the pTyr-1 amino acid were taken as variables.
The conformation of the (3-amino)Z N-terminal group in the
model of Figure 4 was calculated to be of low energy as
confirmed by its occurrence in the X-ray crystal structure of a
Z-protected peptide molecule (Cambridge crystallographic
database40 code SIDWIL).
Figures 1, 2, and 4 were prepared in Insight II (MSI
corporation, San Diego, CA).
Clon in g a n d Exp r ession of Recom bin a n t P r otein s.
Glutathione S-transferase (GST) fusion proteins of Grb2 (GST/
Grb2-SH2) were from Santa Cruz Biotech. The pMAL-c2
expression vector (New England Biolabs) was used to express
the C-terminal domain of the epidermal growth factor receptor
intracellular domain (residues 976-1210) as a maltose-binding
fusion protein (MBP-EGFR) in Escherichia Coli. Two oligode-
Ack n ow led gm en t. The authors are grateful to R.
Wille, D. Arz, V. von Arx, and C. Stamm for their
technical assistance and to S. E. Shoelson for providing
the coordinates of the Lck-SH2 domain X-ray structure.

3556 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 22
Furet et al.
(22) Waksman, G.; Kominos, D.; Robertson, S. C.; Pant, N.; Balti-
more, D.; Birge, R. B.; Cowburn, D.; Hanafusa, H.; Mayer, B.
J .; Overduin, M.; Resh, M. D.; Rios, C. B.; Silverman, L.;
Kuriyan, J . Crystal Structure of the Phosphotyrosine Recogni-
tion Domain SH2 of v-src Complexed with Tyrosine-Phos-
phorylated Peptides. Nature 1992, 358, 646-653.
(23) Hatada, M. H.; Lu, X.; Laird, E. R.; Green, J .; Morgenstern, J .
P.; Lou, M.; Marr, C. S.; Phillips, T. B.; Ram, M. K.; Theriault,
K.; Zoller, M. j.; Karas, J . L. Molecular Basis for the Interaction
of the Protein Tyrosine Kinase ZAP-70 with the T-cell Receptor.
Nature 1995, 377, 32-38.
(24) Flocco, M. M.; Mowbray, S. L. Planar Stacking Interactions of
Arginine and Aromatic Side-Chains in Proteins. J . Mol. Biol.
1994, 235, 709-717.
(25) Mitchell, J . B. O.; Nandi, C. L.; McDonald, I. K.; Thornton, J .
M.; Price, S. L. Amino/Aromatic Interactions in Proteins: Is the
Evidence Stacked Against Hydrogen Bonding? J . Mol. Biol.
1994, 239, 315-331.
Refer en ces
(1) Gibbs, J . B.; Oliff, A. Pharmaceutical Research in Molecular
Oncology. Cell 1994, 79, 193-198.
(2) Saltiel, A. R; Sawyer, T. K. Targeting Signal Transduction in
the Discovery of Antiproliferative Drugs. Chem. Biol. 1996, 3,
887-893.
(3) Levitzki, A. Signal Transduction Therapy. A Novel Approach to
Disease Management. Eur. J . Biochem. 1994, 226, 1-13.
(4) Groundwater, P. W.; Solomons, K. R. H.; Drewe, J . A.; Munawar,
M. A. Protein Tyrosine Kinase Inhibitors. In Progress in
Medicinal Chemistry; Ellis, G. P., Luscombe, D. K., Eds.; Elsevier
Science B. V.: Amsterdam, 1996; Vol. 33, pp 233-329.
(5) Traxler, P.; Lydon, N. Recent Advances in Protein Tyrosine
Kinase Inhibitors. Drugs Future 1995, 20, 1261-1274.
(6) Gishizky, M. L. Tyrosine Kinase Induced Mitogenesis Breaking
the Link with Cancer. In Annual Reports in Medicinal Chem-
istry; Bristol, J . A., Ed.; Academic Press: San Diego, 1995; Vol.
30, pp 247-253.
(7) Smithgall, T. E. SH2 and SH3 Domains: Potential Targets for
Anti-Cancer Drug Design. J . Pharmacol. Toxicol. Methods 1995,
34, 125-132.
(8) Lowenstein, E. J .; Daly, R. J .; Batzer, W. L.; Margolis, B.;
Lammers, R.; Ullrich, A.; Skolnik, E. Y.; Bar-Sagi, D.; Schlessing-
er, J . The SH2 and SH3 Domain-Containing Protein Grb2 Links
Receptor Tyrosine Kinases to ras Signaling. Cell 1992, 70, 431-
442.
(9) Rojas, M.; Yao, S. Y.; Lin, Y. Z. Controlling Epidermal Growth
Factor (EGF)-stimulated Ras Activation in Intacts Cells by a
Cell-permeable Peptide Mimicking Phosphorylated EGF Recep-
tor. J . Biol. Chem. 1996, 271, 27456-27461.
(10) Peptide studies have established that the minimal sequence
retaining micromolar affinity for the Grb2-SH2 domain is the
tripeptide pTyr-Ile-Asn. Asparagine at position pTyr+2 is
absolutely required while the pTyr+1 position is more versatile,
valine, glutamine and glutamic acid being good substitutes for
isoleucine (Garcia-Echeverria, C.; et al. Novartis Pharma Inc.,
Oncology Research Department, unpublished results).
(11) Rahuel, J .; Gay, B.; Erdmann, D., Strauss, A.; Garcia-Echeverria,
C.; Furet, P.; Caravatti, G.; Fretz, H.; Schoepfer, J .; Gruetter,
M. Structural Basis for Specificity of Grb2-SH2 Revealed by a
Novel Ligand Binding Mode Nature Struct. Biol. 1996, 3, 586-
589.
(26) Gowravaram, M. R.; Tomczuk, B. E.; J ohnson, J . S.; Delecki,
D.; Cook, E. R.; Ghose, A. K.; Mathiowetz, A. M.; Spurlino, J .
C.; Rubin, B.; Smith, D. L.; Pulvino, T.; Wahl, R. C. Inhibition
of Matrix Metalloproteinase by Hydroxamates Containing Het-
′ Group. J . Med.Chem.
eroatom-Based Modifications of the P1
1995, 38, 2570-2581.
(27) Ottinger, E. A.; Shekels, L. L.; Bernlohr, D. A.; Barany, G.
Synthesis of Phosphotyrosine-Containing Peptides as Their Use
as Substrates for Protein Tyrosine Phosphatases. Biochemistry
1993, 32, 4354-4361.
(28) Rink, H. Solid-Phase Synthesis of Protected Peptide Fragments
Using a Trialkoxy-diphenyl-methylester resin. Tetrahedron Lett.
1987, 28, 3787-3790.
(29) For a study on the coupling of NR-Fmoc-Tyr(PO3H2)-OH, see:
Garc´ıa-Echeverr´ıa, C. Lett. Pept. Sci. 1995, 2, 369-373.
(30) (a) Castro, B.; Dormoy, J . R.; Evin, G.; Selve, C. Reactifs de
couplage peptidique IV (1) - L′ hexafluorophosphate de benzo-
triazolyl N-oxytrisdimethylamino phosphonium (B.O.P.). Tetra-
hedron Lett. 1975, 14, 1219-1922. (b) Ko¨nig, W.; Geiger, R. Eine
neue methode zur synthese von peptiden: aktivierung der
caboxylgruppe mit dicyclohexylcarbodiimid unter zusatz von
1-hydoxy-benzotriazolen. Chem. Ber. 1970, 103, 788-798.
(31) Carpino, L. A. 1-Hydroxy-7-azabenzotriazole. An Efficient Pep-
tide Coupling Additive. J . Am. Chem. Soc. 1993, 115, 4397-
4398.
(12) Eck, M. J .; Shoelson, S. E.; Harrison, S. C. Structure of the
Regulatory Domains of the Src-Family Tyrosine Kinase Lck.
Nature 1993, 362, 87-91.
(32) Albericio, F.; Kneib-Cordonier, N.; Biancalana, S.; Gera, L.;
Masada, R. I.; Hudson, D.; Barany, G. Preparation and Applica-
tion of the 5-(4-(9-fluorenylmethyloxycarbonyl)aminopmethyl-
3,5-dimethoxyphenoxy)-valeric Acid (PAL) Handle for the Solid-
Phase Synthesis of C-Terminal Peptide Amides under Mild
Conditions. J . Org. Chem. 1990, 55, 3730-3743.
(33) Knorr, R.; Trzeciak, A.; Bannwarth, W.; Gillessen, D. New
Coupling Reagents in Peptide Chemistry. Tetrahedron Lett.
1989, 30, 1927-1930
(34) Mohamadi, F.; Richards, N. G.; Guida, W. C.; Liskamp, R.;
Lipton, M.; Caufield, C.; Chang, G.; Hendrickson, T.; Still, W.
C. Macromodel- An Integrated Software System for Modeling
Organic and Bioorganic Molecules Using Molecular Mechanics.
J . Comput. Chem. 1990, 11, 440-467.
(35) Weiner, S. J .; Kollman, P.; Case, D. A.; Singh, U. C.; Ghio, C.;
Alagona, S.; Profeta, S.; Weiner, P. A Force Field for the
Simulation of Nucleic Acids and Proteins. J . Am. Chem. Soc.
1984, 106, 765-770.
(36) Chang, G.; Guida, W. C.; Still, W. C. An Internal Coordinate
Monte Carlo Method for Searching Conformational Space. J . Am.
Chem. Soc. 1989, 111, 4379-4386.
(37) “In-house” addition to MacroModel: (a) Bohacek, R. S.; Guida,
W. C. A Rapid Method for the Computation, Comparison and
display of Molecular Volumes. J . Mol. Graph. 1989, 7, 113-117.
(b) Eyraud, V.; Dietrich, A. Novartis Pharma Inc. Unpublished
results.
(38) Dewar, M. J . S.; Thiel, W. Ground States of Molecules. 38. The
MNDO Method. Approximation and Parameters. J . Am. Chem.
Soc. 1977, 99, 4899-4907.
(39) Stewart, J . J . P. MOPAC. A Semiempirical Molecular Orbital
Program. J . Comput.-Aided. Mol. Des. 1990, 1, 1-105.
(40) Allen, F. H.; Bellard, S.; Brice, M. D.; Cartwright, B. A.;
Doubleday, A.; Higgs, H.; Hummelink, T.; Hummelink-Peters,
B. G.; Kennar, O.; Motherwell, W. D. S.; Rodgers, J . R.; Watson,
D. G. The Cambridge Crystallographic Database. Acta Crystal-
logr. 1979, B39, 2331-2339.
(13) Birge, R. B.; Hanafusa, H. Closing in on SH2 Specificity. Science
1993, 262, 1522-1524.
(14) Ligand residues are numbered relative to the position of the
phosphotyrosine which is denoted pTyr 0.
(15) Songyang, Z.; Shoelson, S. E.; McGlade, J .; Olivier, P.; Pawson,
T.; Bustelo, X. R.; Barbacid, M.; Sabe, H.; Hanafusa, H.; Yi, T.;
Ren, R.; Baltimore, D.; Ratnofsky, S.; Feldman, R. A.; Cantley,
L. C. 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.
(16) Songyang, Z.; Shoelson, S. E.; Chaudhuri, M.; Gish, G.; Pawson,
T.; Hase, 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.
(17) Pascal, S. M.; Singer, A. U.; Gish, G; Yamazaki, T.; Shoelson, S.
E.; Pawson, T.; Kay, L. E.; Forman-Kay, J . D. Nuclear Magnetic
Resonance Structure of an SH2 Domain Of Phospholipase C-g1
Complexed with a High Affinity Binding Peptide. Cell 1994, 77,
461-472.
(18) Lee, C. H.; Kominos, D.; J acques, S.; Margolis, B.; Schlessinger,
J .; Shoelson, S. E.; Kuriyan, J . Crystal Structures of Peptide
Complexes of the Amino-Terminal SH2 Domain of the Syp
Tyrosine Phosphatase. Structure 1994, 2, 423-438.
(19) Narula, S. S.; Yuan, R. W.; Adams, S. E.; Green, O. M.; Green,
J .; Philips, T. B.; Zydowsky, L. D.; Botfield, M. C.; Hatada, M.;
Laird, E. R.; Zoller, M. J .; Karas, J . L.; Dalgarno, D. C. Solution
Structure of the C-terminal SH2 Domain of the Human Tyrosine
Kinase Syk Complexed with a Phosphotyrosine Pentapeptide.
Structure 1995, 3, 1061-1073.
(20) Breeze, A. L.; Kara, B. V.; Barratt, D. G.; Anderson, M.; Smith,
J . C.; Luke, R.W.; Best, J . R.; Cartlidge, S. A. Structure of a
Specific Peptide Complex of the Carboxy-Terminal SH2 Domain
from the p85a Subunit of Phosphatidylinositol 3-Kinase. EMBO
J . 1996, 15, 3579-3589.
(21) For the nomenclature of the SH2 domain residues, see ref 12 or
18.
J M9702185