Design of peptidomimetics that inhibit the association of phosphatidylinositol 3-kinase with platelet-derived growth factor-β receptor and possess cellular activity

Article abstract of DOI:10.1021/jm9802766

Phosphorylated tyrosine residues of growth factor receptors that associate with intracellular proteins containing src-homology 2 (SH2) domains are integral components in several signal transduction pathways related to proliferative diseases such as cancer, atherosclerosis, and restenosis. In particular, a phosphorylated pentapeptide [pTyr⁷⁵¹-Val-Pro-Met⁷⁵⁴-Leu (pTyr = phosphotyrosine)] derived from the primary sequence of platelet- derived growth factor-β (PDGF-β) receptor blocks the association of the C- terminal SH2 domain of the p85 subunit of phosphatidylinositol 3-kinase (PI 3-kinase) to PDGF-β receptor with an IC₅₀ of 0.445 ± 0.047 μM. Further evaluation of the structure-activity relationships for pTyr⁷⁵¹-Val-Pro- Met-Leu resulted in the design of smaller peptidomimetics with enhanced affinity including Ac-pTyr Val-Ala-N(C₆H₁₃)₂ (IC₅₀ = 0.076 ± 0.010 μM). In addition, the phosphotyrosine residue was replaced with a difluorophosphonate derivative [4-phosphono(difluoromethyl)phenylalanine (CF₂Pmp)] which has been shown to be stable to cellular phosphatases. The extracellular administration of either CF₂Pmp-Val-Pro-Met-Leu or Ac-CF₂Pmp- Val-Pro-Met-NH₂ in a whole cell assay resulted in specific inhibition of the PDGF-stimulated association from the C-terminal SH2 domain of the p85 subunit of PI 3-kinase to the PDGF-β receptor in a dose-dependent manner. These compounds were also effective in inhibiting GLUT4 translocation, c-fos expression, and cell membrane ruffling in single-cell microinjection assay.

Full text of DOI:10.1021/jm9802766

J . Med. Chem. 1998, 41, 4329-4342
4329
Design of P ep tid om im etics Th a t In h ibit th e Associa tion of P h osp h a tid ylin ositol
3-Kin a se w ith P la telet-Der ived Gr ow th F a ctor -â Recep tor a n d P ossess Cellu la r
Activity
Scott R. Eaton, Wayne L. Cody,* Annette M. Doherty, Debra R. Holland, Robert L. Panek,† Gina H. Lu,†
Tawny K. Dahring,† and David R. Rose‡
Departments of Chemistry and Vascular and Cardiac Diseases, Parke-Davis Pharmaceutical Research, Division of
Warner-Lambert Company, Ann Arbor, Michigan 48105, and Department of Medicine, University of California at San Diego,
Whittier Diabetes Program, Veteran’s Administration Research Foundation, 9500 Gilman Drive, La J olla, California 92093
Received May 4, 1998
Phosphorylated tyrosine residues of growth factor receptors that associate with intracellular
proteins containing src-homology 2 (SH2) domains are integral components in several signal
transduction pathways related to proliferative diseases such as cancer, atherosclerosis, and
restenosis. In particular, a phosphorylated pentapeptide [pTyr751-Val-Pro-Met754-Leu (pTyr )
phosphotyrosine)] derived from the primary sequence of platelet-derived growth factor-â (PDGF-
â) receptor blocks the association of the C-terminal SH2 domain of the p85 subunit of
phosphatidylinositol 3-kinase (PI 3-kinase) to PDGF-â receptor with an IC50 of 0.445 ( 0.047
µM. Further evaluation of the structure-activity relationships for pTyr751-Val-Pro-Met-Leu
resulted in the design of smaller peptidomimetics with enhanced affinity including Ac-pTyr-
Val-Ala-N(C6H13)2 (IC50 ) 0.076 ( 0.010 µM). In addition, the phosphotyrosine residue was
replaced with a difluorophosphonate derivative [4-phosphono(difluoromethyl)phenylalanine
(CF2Pmp)] which has been shown to be stable to cellular phosphatases. The extracellular
administration of either CF2Pmp-Val-Pro-Met-Leu or Ac-CF2Pmp-Val-Pro-Met-NH2 in a whole
cell assay resulted in specific inhibition of the PDGF-stimulated association from the C-terminal
SH2 domain of the p85 subunit of PI 3-kinase to the PDGF-â receptor in a dose-dependent
manner. These compounds were also effective in inhibiting GLUT4 translocation, c-fos
expression, and cell membrane ruffling in single-cell microinjection assay.
In tr od u ction
carotid arteries an upregulation of PDGF and FGF
receptor isoforms has been observed. In this model, the
infusion of PDGF following arterial damage resulted in
an increase of VSMC migration along with a corre-
sponding increase in thickening of the neointimal
layer.8,9 However, antibodies raised to PDGF signifi-
cantly inhibited this process.10
The proliferation, differentiation, and survival of
normal cells are regulated by a variety of extracellular
signaling mitogens known as growth factors. These
polypeptidic growth factors include epidermal growth
factor (EGF), fibroblast growth factors (FGFs), and
platelet-derived growth factor (PDGF). Growth factors
have been implicated in numerous disease states in-
volving uncontrolled cellular proliferation and differ-
entiation, including cancer, atherosclerosis, and resteno-
sis.1-3 In clinical studies, restenosis [i.e., the angio-
graphic reocclusion of the arterial wall following per-
cutaneous transluminal coronary angioplasty (PTCA, or
better know as balloon angioplasty)] occurs in 20-45%
of patients within the first 3-6 months following
treatment.4-6 In this report, a unique approach to the
prevention of restenosis utilizing peptidomimetics de-
signed to block intracellular signaling pathways medi-
ating cellular proliferation and differentiation will be
described.
Upon the binding of PDGF to its extracellular recep-
tor, dimerization of the receptor occurs followed by
autophosphorylation of several tyrosines on intracellular
protein substrates and/or the receptor, itself. These
phosphorylated tyrosines act as high-affinity binding
sites for several cellular proteins involved in signal
transduction pathways which are linked to cellular
proliferation.2,11-13 One such protein is phosphatidyli-
nositol 3-kinase (PI 3-kinase). PI 3-kinase is 1 of more
than 20 cytosolic proteins involved in intracellular
signaling that contain src-homology (SH2) domains.
SH2 domains consist of approximately 100 amino acids
that are highly conserved across other such cytosolic
proteins and possess high-affinity binding sites for
specific phosphorylated tyrosines of growth factor
receptors.2,11-13 In particular, it has been shown that
a significant increase in DNA synthesis occurs in
NMuMG cells when PDGF receptors are specifically
associated with PI 3-kinase.14 Thus, the intracellular
blockade of the association of PDGF with PI 3-kinase
should have therapeutic implications for the modulation
of restenosis and other proliferative diseases mediated
by growth factor receptors.
Both the FGFs and PDGF are potent vascular smooth
muscle cell (VSMC) mitogens and chemoattractants.7
The migration and proliferation of VSMCs are an initial
and critical step in the onset of restenosis following
balloon angioplasty. In fact, after balloon injury to rat
* To whom correspondence should be addressed. Phone: (734) 622-
7729. Fax: (734) 622-3107. E-mail: Wayne.Cody@wl.com.
† Vascular and Cardiac Diseases, Parke-Davis Pharmaceutical
Research.
‡ University of California at San Diego.
10.1021/jm9802766 CCC: $15.00 © 1998 American Chemical Society
Published on Web 09/26/1998

4330 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22
Eaton et al.
PI 3-kinase is a heterodimeric enzyme that contains
a noncatalytic 85-kDa (p85) subunit and a catalytic 110-
kDa (p110) subunit.15,16 The p85 subunit consists of one
SH3 and two SH2 domains (N- and C-terminal) that
bind specifically to phosphorylated tyrosines on acti-
vated growth factor receptors.17 Of the two SH2 do-
mains the C-terminal SH2 domain, like the full-length
p85, distinguishes between the wild-type and a mutant
PDGF receptor that lacks the PI 3-kinase binding site.
Thus, the C-terminal SH2 domain of the p85 subunit
of PI 3-kinase (p85 C-SH2) accounts for the high-affinity
binding of PI 3-kinase to PDGF-â receptor.18 This
report will concentrate on the inhibition of this interac-
tion.
In addition, the X-ray crystal structures of phospho-
undecapeptides encompassing pTyr740 and pTyr751 of
PDGF-â receptor bound to the N-terminal SH2 domain
of p85 of PI 3-kinase (p85 N-SH2) have been reported.31
Both of these structures were highly similar to those
reported for phosphopeptides binding to the src29 and
lck30 SH2 domains, further supporting a two-pronged
binding model for phosphopeptides to SH2 domains.31
Recently, a crystal has been obtained for a phosphory-
lated pentapeptide bound to p85 C-SH2 of human PI
3-kinase which diffracted to 1.7 Å, although specific
structural data was not provided.32
The proton NMR structure of Ac-pTyr751-Val-Pro-Met-
Leu bound to p85 C-SH2 has been reported and further
highlights the molecular interactions observed in the
X-ray crystallographic studies.33 Basically, the pY and
pY + 3 pockets were seen to be occupied by the
phosphotyrosine and methionine of the above substrate,
respectively, with the remainder of the residues lying
across the surface of the protein. In addition, a hetero-
nuclear NMR study of the association of a synthetic 12-
amino acid, phosphopeptide that encompasses pTyr751
to p85 N-SH2 further supports the importance of these
interactions. This solution-phase NMR study revealed
that few conformational changes were observed in the
protein except for those residues that were directly
associated with the pY and pY + 3 binding pockets.34
Therefore, the X-ray crystallographic and homonuclear
and heteronuclear NMR results were consistent with a
two-pronged binding model for phosphopeptides to SH2
domains. Herein, we will discuss how these observa-
tions, together with homology molecular modeling, were
exploited in the design of peptidomimetic inhibitors of
the association of p85 C-SH2 with PI 3-kinase.
It has been shown that the inhibition of the associa-
tion of PDGF-â receptor with the p85 C-SH2 can be
achieved with synthetic phosphopeptides and that this
may provide a novel approach to the treatment of
restenosis; however, several concerns need to be ad-
dressed. These include the following: (a) the stability
of peptide/peptidomimetic compounds to exo- and/or
endopeptidases, (b) the stability of phosphorylated ty-
rosine-containing compounds to intracellular phos-
phatases, (c) the ability to deliver the drug candidate
to the point of injury, (d) the ability of peptides contain-
ing a highly charged phosphate group to be permeable
to the cellular membrane, and (e) the ability to specif-
ically inhibit the association of PDGF-â receptor with
PI 3-kinase given the number of cytosolic proteins
containing SH2 domains. The potential to address and
circumvent these concerns with the peptidomimetics
reported herein will be discussed and evaluated.
As previously discussed, several cytosolic proteins
contain SH2 domains. Several high-affinity phospho-
rylated small peptides and peptidomimetic inhibitors
of src19-24 and, more recently, grb225 SH2 domains
have been discovered or designed through library ap-
proaches19,20 and systematic rational drug design
strategies.21-25 Previous successes in the systematic
rational design of high-affinity and selective peptides/
peptidomimetics for src21-24 prompted us to initiate a
similar approach for inhibition of the association of p85
C-SH2 with PDGF-â receptor.
In particular, it has been shown that PI 3-kinase
specifically associates with pTyr740 (pTyr ) phosphoty-
rosine) and pTyr751 of the PDGF-â receptor. In fact, it
has been shown that pentapeptides corresponding to
these phosphotyrosines [pTyr740-Met-Asp-Met-Ser (1)
and pTyr751-Val-Pro-Met-Leu (2)] were able to inhibit
90% of the in vitro association of PDGF-â receptor with
p85 C-SH2 at a concentration of 100 µM.14 In subse-
quent studies, it was determined that compound 2 has
an IC50 of 0.445 ( 0.04 µM for inhibiting this associa-
tion.26-28 [In our hands, pTyr740-Met-Asp-Met-Ser (com-
pound 1) possessed an IC50 of 0.522 ( 0.12 µM.
However, further structure-activity relationship (SAR)
studies utilizing rational drug design strategies were
performed on compound 2, since this compound was
conformationally constrained with a proline residue in
its primary sequence and additional structural data was
also available.26] Further analysis of compound 2
revealed that several modifications were well-tolerated
for the Val (pY + 1), Pro (pY + 2), and Leu (pY + 4)
residues. However, the pTyr (pY) and Met (pY + 3)
residues were intolerant of even minimal modification.26
The X-ray crystal structures of phosphorylated pep-
tides containing the pTyr-Glu-Glu-Ile motif bound to
both the src29 and lck30 SH2 domains have been
reported. Both of these structures were highly similar
and further emphasized the importance of the molecular
interactions of pTyr (pY) and Ile (pY + 3) for high-
affinity binding. In both cases, an arginine residue
was found at the base of the phosphotyrosine binding
pocket (pY) that formed a bidentate ion-pairing inter-
action with the phosphate group, and the isoleucine
was completely buried in a hydrophobic binding pocket
(pY + 3). The remaining residues were extended across
the surface of the protein and through a network of
hydrogen bonds that held the pY and pY + 3 residues
in their respective binding pockets. Essentially, the
mode of binding is best described as a two-pronged
model.
Resu lts a n d Discu ssion
A phosphorylated pentapeptide corresponding to the
PDGF-â receptor sequence of pTyr751-Val-Pro-Met754
-
Leu (compound 2) has been shown to inhibit the
association of PDGF-â receptor with p85 C-SH2 of PI
3-kinase in vitro.14 Specifically, an IC50 of 0.445 ( 0.047
µM has been determined for this pentapeptide for the
specific blockade of the association of the phosphorylated
PDGF-â receptor tyrosine kinase with [35S]GST p85
C-SH2 fusion protein (Table 1).26-28 In addition, com-
pound 2 has been shown to be selective for the p85
C-SH2 domain over that of src and abl (Table 1).

Design of Peptidomimetics
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22 4331
Ta ble 1. Binding Affinities of Synthetic Compounds 1-23 (R1-pTyr-R2-R3-R4) to p85, src, and abl SH2 Domains
compd no.
R1
R2
R3
R4
Met-Ser
Met-Leu
Met
p85 C-SH2 ( SEMa (n)b
srca
abla
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
H
H
H
Ac
H
Ac
H
Ac
H
Ac
Ac
Ac
Ac
Ac
Ac
Ac
Ac
Ac
Ac
Ac
Ac
Ac
Ac
Met
Val
Val
Val
Val
Val
Val
Val
Val
Val
Val
Val
Val
Val
Val
Val
Val
Val
Val
Val
Val
Val
Val
Asp
Pro
Pro
Pro
Pro
Pro
Pro
Pro
Pro
0.522 ( 0.12 (2)
0.445 ( 0.047 (16)
6.85 ( 3.26 (2)
0.18
0.40
0.16
0.485 ( 0.225 (3)
0.264 ( 0.029 (14)
1.36 ( 0.22 (16)
2.5
16.1 ( 3.4 (3)
4.56 ( 1.64 (3)
3.32 ( 0.99 (4)
0.22
ND
12.7
ND
5.9
ND
ND
ND
ND
>10
ND
ND
ND
2.7
1.4
12.5
5.7
3.5
1.0
6.1
8.6
17.8
5.0
18.4
ND
60%c
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
25.7
ND
7.4
ND
6.0
5.0
95.9
46.2
72.0
26
Met-Leu
Met-Leu-NH2
Met-Leu-NH2
Met-NH2
Met-NH2
Nle-Leu
N-BuGly-NH2
NHC5H11
NHC5H11
Ala
Ala
D-Ala
D-Ala
Ala
D-Ala
Ala
D-Ala
Ala
D-Ala
Ala
N(C5H11
N(C5H11
)
)
2
2
N(C4H9)2
N(C4H9)2
5.7
0.478 ( 0.129 (3)
2.5
N(C6H13
N(C6H13
N(C8H17
)
2
)
2
)
2
0.076 ( 0.010 (3)
3.47 ( 0.48 (4)
37.6 ( 8.6 (3)
>100
19.0 ( 1.2 (3)
>100
NH(CH2)2C6H5
NH(CH2)2C6H5
NH(CH2)2C6H11
NH(CH2)2C6H11
D-Ala
Ala
D-Ala
ND
a
b
IC50 values in micromolar concentrations ( standard error of the mean. ND, not determined. n ) number of determinations if
greater than 1 determination. c Percent inhibition at a concentration of 100 µM.
Initial SAR studies were directed at determining the
minimal pharmacophore that would inhibit this associa-
tion. Thus, simple truncation of compound 2 by remov-
ing the C-terminal Leu resulted in a greater than 15-
fold loss of affinity (compound 3).26-28 Interestingly, it
was observed that individual or combined neutralization
of the amino and carboxy termini was shown to enhance
the affinity of compound 2. For example, N-terminal
acetylation (compound 4) resulted in a 4-fold enhance-
ment of affinity for p85 SH2, while C-terminal amida-
tion (compound 5) enhanced the affinity ∼1.5-fold with
respect to compound 2. Although these modifications
were not shown to be additive, N-terminal acetylation
and C-terminal amidation of the parent pentapeptide
resulted in a very potent analogue (compound 6) with
an IC50 of 0.16 µM.26-28 Interestingly, the neutraliza-
tion of the amino and carbox termini seems to be a
general theme for the design of highly potent inhibitors
of several SH2 domains, including src,19-24 grb2,25 and
p85.26-28
The SAR observed in this pentapeptide series was
directly translatable to the truncated tetrapeptide se-
ries, such that C-terminal amidation led to a 14-fold
enhancement in affinity (compound 7 versus compound
3). Likewise, N-terminal acetylation and C-terminal
amidation resulted in a further 2-fold enhancement of
affinity (compound 8, IC50 ) 0.264 ( 0.029 µM). Thus,
the minimal pharmacophore required for inhibiting the
association of PDGF-â receptor with p85 C-SH2 was
defined by Ac-pTyr-Val-Pro-Met-NH2.26-28
(compound 9) with respect to the parent molecule
(compound 2).26,27,35
Since neither the proton NMR or X-ray crystal-
lographic structures were available at the time of this
study, homology-based molecular modeling was intiated
to prepare a theoretical model of p85 C-SH2 domain
(Figure 4). As observed for other related SH2 domain
structures,36 the model suggests that the ligand binding
surface of p85 C-SH2 is characterized by two distinct
pockets: the so-called pY and pY + 3 pockets. The pY
pocket is highly electropositive, comprised of two argi-
nine and two serine residues which coordinate with the
phosphate oxygens. One of these two arginines, located
on the surface of the pY pocket, is also available to
interact with an N-terminal group, such as an acetyl.
The model also suggested a wide, shallow hydrophobic
pY + 3 pocket into which both the pY + 1 and pY + 3
inhibitor residues bind. This characteristic correlates
with the reported preference of p85 C-SH2 for peptide
sequences of the type pTyr-Hyd-Xxx-Hyd,19 where Hyd
represents a hydrophobic residue and Xxx represents
any residue. The pY + 3 pocket is comprised of a
phenylalanine, two valines, a leucine, and a cysteine
residue. In addition, hydrophobic residues (proline,
phenylalanine, valine, and the aliphatic side chain of
glutamine) form a shallow channel on the surface of the
protein which extends the hydrophobic pY + 3 pocket
in a direction away from the pY pocket (Figure 4).
In retrospect, a comparison of this homology model
with the later published proton NMR solution structure
of p85 C-SH2 complexed to the peptide Ac-pTyr-Val-
Pro-Met-Leu33 does reveal some differences in the
protein backbone and side-chain orientations. However,
the key features of the model’s ligand binding region
used for the inhibitor design work described herein are
shared by the proton NMR structure: (1) the pY pocket
is comprised of the same electrostatically charged
residues, (2) the pY + 3 pocket is comprised of the same
hydrophobic residues and is approximately the same
size, (3) the distance between pY and pY + 3 pockets is
As previously discussed, the SAR,19,20,26-28 X-ray
crystallographic,29-32 and homonuclear33 and hetero-
nuclear NMR34 studies of phosphorylated peptides
encompassing Tyr751 thru Met754 of the PDGF-â receptor
sequence have implicated the importance of both resi-
dues for high-affinity association with p85 C-SH2 in a
two-pronged binding mode. In fact, the only substitu-
tion that is readily tolerated for methionine was nor-
leucine.26,27,35 In particular, this pseudoisosteric sub-
stitution resulted in only a 3-fold loss of affinity

4332 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22
Eaton et al.
F igu r e 1. Design of the peptidomimetic dialkylamide series from the peptoid (compound 10) and the monoalkylamide (compound
12).
F igu r e 2. Structures of phosphotyrosine (pTyr), 4-phospho-
nomethylphenylalanine (Pmp), and 4-phosphono(difluorom-
ethyl)phenylalanine (CF2Pmp).
approximately the same (∼12 Å), and (4) a significant
hydrophobic channel does exist in the proton NMR
structure as predicted by the model, although this
channel as seen in the proton NMR structure is much
less solvent-exposed than in our model. This observa-
tion might help explain the increased potency of the
dialkyl inhibitors described below (Figure 4b). Thus,
overall, this homology model, although somewhat dif-
ferent from the later published experimental structure,
did predict the key features critical to successful inhibi-
tor design.
Utilizing this homology-based model for p85 C-
SH2, it was suggested that transposition of the side
chain of methionine and/or norleucine to the amide
F igu r e 3. Immunoblot analysis showing the effects of com-
pounds 25 and 26 on the association of the 85-kDa subunit of
PI 3-kinase with immunoprecipitated PDGF-â receptors from
nitrogen should be a well-tolerated modification while
still effectively utilizing the pY + 3 pocket. This
“peptoid”37 (N-alkylated glycine) was efficiently pre-
pared by a solid-phase approach. Initially, bromoacetic
acid was coupled to a Rink amide [4-(2′,4′-dimethox-
yph en yl-F m oc-a m in om et h yl)ph en oxya cet a m ido-
norleucylaminomethyl]resin,38,39 followed by nucleo-
philic displacement of the bromine with N-butylamine,
and normal solid-phase peptide synthetic strategies (see
the Experimental Section). The peptoid corresponding
to compound 2 [Ac-pTyr-Val-Pro-N-BuGly-NH2 (N-Bu-
Gly ) N-butylglycine)] was not prepared since this
peptoid would be prone to synthetic difficulties such as
diketopiperazine formation, premature resin cleavage,
and chain termination due to the proline residue
adjacent to the N-BuGly residue. It was previously
reported that pTyr-Val-Ala-Met-Leu possessed an IC50
of 2.7 µM,26 which was only approximately 5-fold less
rat aortic smooth muscle cells. After a 24-h incubation of cells
with various concentrations of compound, PDGF-BB (30 ng/
mL) was added to cells to stimulate PDGF receptor phospho-
rylation and the subsequent association of PI 3-kinase. PDGF
receptors were immunoprecipitated with PDGF receptor poly-
clonal antibodies directed against the â-receptor isoform and
Western blotted with anti-rat PI 3-kinase polyclonal antibody.
potent than its corresponding proline-containing parent
peptide (compound 2). Therefore, the analogous peptoid
corresponding to this sequence was prepared (Ac-pTyr-
Val-Ala-N-BuGly-NH2). In addition, molecular model-
ing also suggested the Ala for Pro substitution, since
proline might not allow the necessary binding confor-
mation for the ligand. In fact, the peptoid (compound
10, Figure 1) possessed an IC50 of 2.5 µM. This IC50
was within a factor of 10 with respect to the parent
tetrapeptide (compound 8). Thus, it can be postulated
that the majority of affinity lost for compound 10 may

Design of Peptidomimetics
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22 4333
F igu r e 4. Comparison of the solid molecular surfaces for (a) the p85 C-SH2 homology-based model and (b) the published p85
C-SH2 NMR solution structure.33 Color coding for the surfaces represents electrostatic potential: red is negative, and blue is
positive. In both panels, peptide 14 is shown as a stick figure in the conformation docked into the homolgy model (a). The
electropositive pY pocket, hydrophobic pY + 3 pocket, and hydrophobic channel are identified for each molecular surface. Atoms
of the inhibitor are colored as red, oxygen; blue, nitrogen; yellow, phosphorus; white, carbon. This figure was prepared using the
program GRASP.71
be the result of the alanine for proline substitution due
to the increased flexibility in the molecule and that, in
fact, the peptoid modification was very well-tolerated.
In an effort to design first-generation peptidomimet-
ics, it was observed that a C-terminal carboxylate was
detrimental to activity in the tetrapeptide series, while
norleucine was well-tolerated as a substitution for
methionine. Therefore, it was postulated and supported
by modeling that the C-terminal amino acid was re-
placed with an alkyl chain, such as N-pentylamine
(Table 1). This modification could be simply incorpo-
rated by the preparation of Ac-Tyr(Bzl)-Val-Ala on a
Sasrin resin [2-methoxy-4-alkoxybenzyl alcohol (a resin
highly susceptible to acidic cleavage conditions)],40,41
which was cleaved with 1.0% trifluoroacetic acid (TFA)
in dichloromethane (DCM) to yield the free carboxylate.
The remaining solution-phase synthesis of the target
compound consisted of coupling of the amine to the
C-terminal carboxylate, deprotection of the tyrosine,
phosphorylation, oxidation, deprotection of the phos-
phorylated tyrosine, and final purification. In principle,
this approach should have been effective. Unfortu-
nately, the coupling of N-pentylamine to Ac-Tyr(Bzl)-
Val-Ala in the presence of 1-(3-dimethylamiopropyl)-3-
ethylcarbodiimide hydrochloride (EDCI) and N-hydroxy-
benzotriazole (HOBt) resulted in extensive racemization
(∼40-50%) of the C-terminal alanine. However, these
diastereoisomers were separable by reversed-phase
high-pressure liquid chromatography (RP-HPLC). The
major diastereoisomer corresponded to the L-enantiomer
of Ala. Ac-pTyr-Val-Ala-NHC5H11 (compound 11, Fig-
ure 1) showed approximately 65-fold less affinity (IC50
) 16.1 ( 3.4 µM) than the parent tetrapeptide. How-
ever, interestingly the D-Ala-containing analogue (com-
pound 12) exhibited an IC50 of 4.56 ( 1.64 µM. Once
again, this binding would be comparable to the parent
peptide if the 4-5-fold loss of affinity was a result of
the proline to alanine substitution. Despite this loss of

4334 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22
Eaton et al.
potency, this modification is significant in that it should
impart stability against exopeptidases, such as carbox-
ypeptidase, to the molecule.
lamine resulted in a significant loss of affinity (com-
pound 19, IC50 ) 3.47 ( 0.48 µM). Thus, the dihexyl
chain is optimal for maximum affinity, presumably since
each side chain makes optimal contacts with the pY +
3 pocket and the surface channel (Figure 4).
The design of a second-generation peptidomimetic
series resulted from molecular modeling suggesting that
the pY + 3 substitution of the peptoid and the C-
terminus of the peptidomimetic (compounds 10 and 12,
Figure 1) could be combined into one molecule via a
dialkylamide. It was envisioned that this modification
would allow interaction with the pY + 3 pocket of p85
C-SH2 via one alkyl chain which is normally occupied
by methionine or norleucine, while the second alkyl
chain would have further interactions with the hydro-
phobic channel of the protein surface (Figure 4). [In
retrospect, the published NMR structure of p85 C-SH2
suggests that the second alkyl chain would indeed
interact with such a hydrophobic surface channel but
more extensively than would have been predicted by the
homology model since that channel is much less solvent-
exposed in the NMR structure (Figure 4b).33] Thus, Ac-
pTyr-Val-D-Ala-N(C5H11)2 (compound 13) was prepared
using the same synthetic strategy used to prepare
compounds 11 and 12. This compound had an IC50 of
3.32 ( 0.99 µM and was essentially equipotent to the
parent peptoid and peptidomimetic. However, as seen
for the preparation of compound 11, significant racem-
ization of the C-terminal alanine had occurred. Sur-
prisingly, upon separation by RP-HPLC and evaluation
of the L-amino acid-containing isomer (compound 14,
Figure 1), this analogue exhibited an IC50 of 0.22 µM
for inhibition of the association of the PDGF-â receptor
with p85 C-SH2.
(Interestingly, during the course of this study, it was
reported that C-terminal alkylamides are effective
substitutions in ligands that bind to src21-24 and grb225
SH2 domains. It was originally suggested in an X-ray
crystallography paper23 and confirmed in a later pub-
lication24 highlighting the SAR of a series of dipeptido-
mimetic inhibitors of the src SH2 domain that a
dipentylamide represented the optimal C-terminal group.
In this case, the dipentylamide was slightly better than
the dihexylamides and significantly better than the
diheptylamides. These results further point to the great
similarity between SH2 domains of different proteins
and the potential problems in developing compounds
selective for a specific SH2 domain.)
Further attempts to optimize the C-terminal amides
by the incorporation of phenethyl and cyclohexylethyl
resulted in compounds that were essentially inactive
(compounds 20-23) which is consistent with the p85
C-SH2 model. However, in both cases, the L-alanine-
containing analogues were still more potent than the
corresponding D-alanine-containing analogues. Thus,
the SAR observed for these compounds is consistent
with the dialkylamide series.
Selectivity of Syn th etic Com p ou n d s for Va r iou s
SH2 Dom a in s. To develop inhibitors of SH2 domains
that may have therapeutic utility for the treatment of
restenosis and proliferative diseases, it is critical to
design compounds that are selective for p85 C-SH2 over
other SH2 domains. In this regard, several of the
synthetic analogues were evaluated for their affinity to
inhibit the association of the SH2 domains of src and
abl. For inhibitors of src, it has been shown that an
acidic residue is preferred in the position pY + 1,19-24
while for the p85 C-SH2 domain a hydrophobic residue,
such as valine, is preferred in this position.14,19,20,26-28
In addition, the native peptide binding sequence for the
src SH2 domain contains an isoleucine residue in the
pY + 3 position,19-24 while ligands for the p85 C-SH2
domain contain a methionine residue.14,19,20,26-28
For compound 2, moderate selectivity for the p85
C-SH2 domain over both the src and abl SH2 domains
(> 28-fold) was observed. Compound 4 showed greater
selectivity for p85 C-SH2 of 33-fold versus the src SH2
domain. Interestingly, the pseudoisosteric replacement
of Nle for Met in compound 2 (compound 9) led to a
modest loss of affinity for p85 C-SH2, but this ligand
completely lacked affinity for src. Perhaps this is not
that surprising since src can accommodate branched
amino acids in the pY + 3 pocket.
It was very interesting and unexpected that the SAR
differed between the monoalkyl- and dialkylamides with
respect to the stereochemistry of the pY + 2 residue.
Previously, it has been shown that utilization of O-(7-
azabenzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexaflu-
orophosphate (HATU) as a coupling agent, in the
presence of 1H-hydroxy-7-azabenzotriazole (HOAt),
greatly reduced the amount of racemization of the
C-terminal amino acid.42,43 Even though, compounds
17 and 18 were initially prepared racemic and then
separated by RP-HPLC (method B in the Experimental
Section); they were also synthesized using the HATU/
HOAt coupling strategy to verify the stereochemical
assignments (method C in the Experimental Section).
Indeed, the independent preparation of compounds 17
and 18 by method C confirmed that the stereochemical
assignment of the C-terminal alanine was correct and
that a reverse in the SAR between the monoalkyl- and
dialkylamide series was observed.
To further optimize the activity of these amides, the
ideal length of the dialkyl chain was determined.
Shortening the chain by one carbon (dibutyl) with both
D- and L-alanine in the pY + 2 position (compounds 15
and 16) led to a modest loss in affinity in both cases
(∼1.5-2-fold) with respect to the parent dipentylamides.
However, increasing the length of the dialkyl chain by
one carbon (dihexyl) led to an enhancement of affinity
for both the D- and L-alanine cases (compounds 17 and
18). In fact, compound 18, is the most potent compound
reported to date for inhibiting the association p85 C-SH2
to PDGF-â receptor (IC50 of 0.076 ( 0.010 µM). Further
extension of the C-terminal dialkyl chain with diocty-
In the peptidomimetic series, some of the selectivity
for src appeared to be diminished. However, compounds
containing an L-Ala residue in the pY + 2 position
maintained selectivity over those containing the D-Ala
residue. For example, compound 13 (D-Ala) showed
little difference in preference between p85 C-SH2 and
src, but in compound 14 (L-Ala) some selectivity was
observed (6-fold). Interestingly, compound 13 did ex-
hibit 8-fold selectivity versus abl. In addition, com-
pound 15 exhibited minimal selectivity for p85 C-SH2

Design of Peptidomimetics
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22 4335
Ta ble 2. Binding Affinities of Synthetic Compounds
Stabilized to Cellular Phosphatases
medium was removed, the cells were lysed, and the
supernatant was incubated with anti-human PDGF-â
receptor monoclonal antibody which only recognized the
â-isoform of the PDGF receptor. Following electro-
phoresis, the separated proteins were immunoblotted
with anti-rat PI 3-kinase antiserum which recognizes
the 85 kDa subunit of PI 3-kinase. Thus, at this point
the only PI 3-kinase present must have been associated
with the PDGF-â receptor. As can be seen in Figure 3,
increasing concentrations of both compounds (25 and
26) were able to inhibit the association of PI 3-kinase
with PDGF-â receptor in a dose-dependent manner. The
IC50’s for these compounds are estimated to be 40 and
38 µM, respectively. Since this is an intracellular event,
compounds 25 and 26 must be permeable to the cellular
membrane and inhibiting a specific cellular response.
This observation is quite remarkable considering the
highly charged nature of the molecules.
compd
no.
p85 C-SH2
structure
( SEMa (n)b
2
8
pTyr-Val-Pro-Met-Leu
Ac-pTyr-Val-Pro-Met-NH2
Pmp-Val-Pro-Met-Leu
CF2Pmp-Val-Pro-Met-Leu
Ac-CF2Pmp-Val-Pro-Met-NH2
Ac-CF2Pmp-Val-Ala-N(C5H11
Ac-CF2Pmp-Val-Ala-N(C6H13
0.445 ( 0.047 (16)
0.264 ( 0.029 (14)
8.7 (1)
0.945 ( 0.212 (4)
1.45 ( 0.26 (9)
0.372 ( 0.053 (8)
0.446 ( 0.037 (4)
24
25
26
27
28
)
)
2
2
a
IC50 values in micromolar concentrations ( standard error of
b
the mean. n ) number of determinations.
over src or abl, but compound 16 exhibited a 12-fold
selectivity versus src. Similarly, compound 17 was not
selective, while compound 18 was 13-fold selective
versus src and 65-fold selective versus abl. Interest-
ingly, compounds 20-23 were relatively weak inhibitors
of both p85 C-SH2 and abl but had enhanced activity
versus src.
An a logu es Sta bilized to Cellu la r P h osp h a ta ses.
To evaluate the potential therapeutic utility of this
series of phosphorylated peptides, it is critical to dem-
onstrate activity in a whole cell-based model. Unfor-
tunately, it has been shown that phosphorylated ty-
rosines are rapidly metabolized by intracellular phos-
phatases.44 Replacement of phosphotyrosine in com-
pound 2 with the corresponding methylenephosphonate
derivative (Pmp, Figure 2, compound 24) resulted in a
10-fold loss of affinity with respect to compound 2 (Table
2). This loss of affinity has routinely been seen with
this substitution in other inhibitors45-47 and is most
likely due to the different pKa2 values for the Pmp
residue (7.1) versus phosphotyrosine (5.7).44
However, it has been reported that the difluorophos-
phonate derivative (CF2Pmp, Figure 2) is an acceptable
substitution for tyrosine maintaining high-affinity while
imparting stability to intracellular phosphatases. In
particular, the CF2Pmp substitution for phosphotyrosine
has been successfully incorporated in inhibitors of the
src,36,44,45,48 grb2,44 and p8535,46 SH2 domains and
inhibitors of the protein tyrosine phosphatases PTP 1
and PTP 1B.47 In addition, CF2Pmp has a pKa2 value
that is much closer to that of phosphotyrosine.44 Sub-
stitution of the CF2Pmp into the pentapeptide (com-
pound 25 versus 2) and the tetrapeptide amide (com-
pound 26 versus 8) resulted in analogues that maintained
good affinity. The CF2Pmp residue was also incorpo-
rated into the peptidomimetic dialkylamide series. For
compounds 27 and 28, good affinity versus the parent
phosphopeptides (compounds 14 and 18, respectively)
was maintained with IC50’s of 0.372 ( 0.053 µM for the
dipentylamide (27) and 0.446 ( 0.037 µM for the
dihexylamide (28). In all cases, the phosphotyrosine-
containing peptides were more potent than the corre-
sponding CF2Pmp analogues.
Since analogues 25 and 26 possessed good affinity for
the inhibition of the association of PDGF-â receptor with
p85 C-SH2, these compounds were further evaluated in
a cellular assay. In this cellular assay, the test com-
pounds were incubated in the presence of rat aortic
smooth muscle cells for 5 min. At this point, the
medium was treated with rat PDGF-BB (â-homodimer)
for an additional 24 h to induce autophosphorylation of
the PDGF receptor and association of PI 3-kinase. The
These compounds were then further evaluated for
biological activity in a series of single-cell microinjection
assays. These assays allowed the direct evaluation of
the intracellular biological consequences of these com-
pounds, as well as providing an opportunity to test their
inherent stability in living cells. Thus, the effect of
microinjection for compounds 25 and 26 on four distinct
insulin-stimulated and PI 3-kinase-dependent biological
phenotypes which can be quantitatively evaluated was
measured.49-52 These assays were selected based upon
the previous observation that the optimum time for
measurement of the four endpoints varied from minutes
to hours after microinjection, allowing an assessment
of intracellular stability of the injected compounds.
Thus, the results of microinjection upon insulin-
stimulated membrane ruffling (3 min), c-fos expression
(60 min), GLUT4 translocation (2 h), and cell cycle
progression from G1 to S phase (14 h) was examined.
Experiments to evaluate the effects of compounds 25
and 26 upon lamellipodia formation (membrane ruffling)
are illustrated in Figure 5A,C. Stimulation of the cells
with insulin resulted in a rearrangement of the actin
network from stress fibers in resting cells to membrane-
associated ruffling structures (shown as concentrated
areas of rhodamine staining in Figure 5A). Ruffles were
found in less than 2% of cells which were not stimulated
with insulin and in 94% of stimulated cells (Figure 5C).
Microinjection of either compound 25 or compound 26
resulted in the complete inhibition of stress fiber rear-
rangement. This phenotype was also seen in cells
adjacent to cells which were microinjected with either
compound, presumably due to membrane permeability
(Figure 5, top right).
The effect of each compound was then evaluated upon
the expression of the immediate early gene, c-fos.
Detection of c-fos expression by immunofluorescence
staining is optimal at 60 min following insulin stimula-
tion.51 Again, microinjection of either compound 25 or
26 completely inhibited the nuclear detection of c-fos
expression (Figure 5C). This inhibitory phenotype was
also seen in NIH 3T3 cells after stimulation with
fibroblast growth factor (FGF; data not shown).
These compounds were also tested for their effect
upon insulin-stimulated translocation of the GLUT4
glucose transport protein in 3T3-L1 adipocytes. The
movement of the GLUT4 protein from intracellular

4336 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22
Eaton et al.
F igu r e 5. Single-cell microinjection assays. A. Immunofluorescence photomicrographs showing cells injected with compound 25
(top row) or preimmune carrier IgG alone (bottom row). Shown are pictures demonstrating injected cells (green staining) and the
same field of cells stained with rhodamine-conjugated phalloidin to show distribution of actin network (red). Arrows are to assist
in identification of corresponding cells in each photograph. 63× photographs (left) demonstrate the inhibition of actin rearrangement
from stress fibers to membrane-associated lamellipodia following insulin stimulation and injection of compound 25. 20× photographs
(right) demonstrate that this inhibition of membrane ruffling is also detected in adjacent uninjected cells due to the membrane
permeability of the compound. B. Photomicrographs showing typical results of DNA synthesis assay. Injected cells possess green
fluorescein-stained cytoplasm in contrast to red nuclear staining found in cells which have incorporated BrdU into newly synthesized
DNA. Recombinant GST-p85 SH2 domain and anti-p85 IgG are shown as controls. C. Combined results of all four single-cell
assays used. Results are expressed as the percentage of cells demonstrating the specific immunofluorescence endpoint ( standard
error.
stores in the Golgi apparatus to the plasma membrane
as detected by immunofluorescence staining has been
used as a single-cell assay for the PI 3-kinase-dependent
metabolic effects of insulin in adipocytes.51,53 The
GLUT4 protein is found at the plasma membrane in
approximately 5% of cells in the resting state and

Design of Peptidomimetics
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22 4337
translocates in response to physiological insulin con-
centrations within 2 h in almost 70% of treated cells
(Figure 5C). Microinjection of either compound 25 or
26 completely inhibited this insulin-stimulated pheno-
type.
receptor with the C-terminal SH2 domain of p85 PI
3-kinase when administered extracellularly to rat aortic
smooth muscle cells with IC50’s of 40 and 38 µM,
respectively. In addition, these compounds were shown
to have specific cellular effects upon microinjection. The
potency of these analogues was further enhanced in the
dialkylamide series. For example, compounds 27 and
28 possessed IC50’s of 0.372 ( 0.053 and 0.446 ( 0.037
µM, respectively, but unfortunately these compounds
proved to be cytotoxic in the whole cell assay.
The concerns raised in our introduction regarding the
utility of peptides/phosphopeptides and/or phosphopep-
tidomimetics for the treatment of restenosis can, at least
in part, be addressed via design of the appropriate
analogue. In particular, the N-terminal acetylation and
C-terminal amidation impart stability to exopeptidases.
The CF2Pmp substitution for pTyr incorporates stability
to intracellular phosphatases while maintaining good
affinity. In addition, results from the whole cellular
experiments with compounds 25 and 26 demonstrate
that a specific intracellular event can be modulated via
extracellular administration of a therapeutic agent even
if highly charged. However, the concern of drug delivery
has yet to be addressed. It is envisioned that the
resulting drug could be delivered directly to the point
of injury (i.e., lesion resulting from PTCA) via the
catheter during the balloon angioplasty procedure.
Finally, both compounds were tested in an assay
which requires 14 h of incubation following microinjec-
tion to complete. In these experiments, quiescent
fibroblasts were injected with each compound, or with
GST-p85 SH2 or anti-p85 IgG as controls, and then
stimulated with insulin. The cells were labeled with
3-bromo-5′-deoxyuridine (BrdU) to measure progression
through G1 phase of the cell cycle to DNA synthesis. In
this case, microinjection of either compound was without
any inhibitory effect, despite the complete inhibition
seen in cells injected with either of the control proteins.
This negative result may be due to the relatively long
(14-h) incubation time in these experiments following
injection of the compounds. Thus, this lack of inhibition
may be a reflection of intracellular instability, but in
light of the membrane permeable nature of the com-
pounds it is likely to be a reflection of the loss of
concentration in injected cells over the course of the
incubation.
The more potent analogues (27 and 28) were also
evaluated for their ability to block the association of
PDGF-â receptor with p85 C-SH2 in rat aortic smooth
muscle cells via extracellular administration (as de-
scribed above). Unfortunately, these compounds re-
sulted in cell detachment and death at the concentra-
tions tested (>25 µM) for reasons that are unknown and
were not further investigated. Interestingly, it has been
reported that dialkylamides coupled to amino acids
possessed good antimicrobial activities but did not
induce hemolysis of red blood cells (i.e., were not
cytotoxic against a eucaryotic cell line).54 In addition,
it has been reported that peptides rich in R,R-dialky-
lated amino acids display potent antibacterial activity.55
Perhaps, the cytotoxicity of compounds 27 and 28 is
related to the highly hydrophobic nature of these
molecules.
Exp er im en ta l Section
Ma ter ia ls a n d Meth od s. The orthogonally protected NR-
t-Boc and NR-Fmoc amino acids and the resins Wang (HMP,
p-benzyloxybenzyl alcohol),56 Rink amide,38,39 and Sasrin40,41
were purchased from either Bachem Bioscience, Advanced
Chemtech, or Peninsula Laboratories, Inc. The unnatural
amino acids NR-Fmoc-diethyl-4-phosphonomethylphenylala-
nine [NR-Fmoc-Pmp(OEt)2] and NR-t-Boc-diethyl-4-phosphono-
(difluoromethyl)phenylalanine [NR-t-Boc-CF2Pmp(OEt)2] were
prepared by previously reported methods.57 All amino acids
were of the L-configuration unless otherwise indicated.
Trifluoroacetic acid (TFA) was purchased from Halocarbon.
N,N′-Dicyclohexylcarbodiimide (DCC), diisopropylethylamine
(DIEA), and HOBt were purchased from Applied Biosystems
Inc. (ABI). N,N-Dimethylformamide (DMF), dichloromethane
(DCM), N-methylpyrrolidone (NMP), and methanol (MeOH)
were purchased from Burdick & J ackson (reagent grade or
better). Acetonitrile (AcCN) and water (HPLC grade) were
purchased from Mallinckrondt and EM Science, respectively.
Trimethylsilyl trifluoromethanesulfonate (TMSOTf), dimethyl
sulfide (DMS), ethanedithiol (EDT), phenol, thioanisole, m-
cresol, bromoacetic acid, diisopropylcarbodiimide (DIC), 1H-
tetrazole, di-tert-butyl diethylphosphoramidite, N-methylmor-
pholine (NMM), 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide
hydrochloride (EDCI), and tert-butyl hydroperoxide were
purchased from Aldrich Chemical Co., Inc. O-(7-Azabenzot-
riazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate
(HATU) and 1H-hydroxy-7-azabenzotriazole (HOAt) were
purchased from PerSeptive Biosystems.
The peptides were prepared on an ABI model 431A auto-
mated peptide synthesizer using software version 1.00 for
Fmoc synthesis. RP-HPLC was performed on a Waters HPLC
system from Millipore Corp. equipped with a model 600E
system controller, a model 600 solvent delivery system, a model
490 variable wavelength detector at 214 and 280 nm, and a
model 717 plus autosampler. Vydac analytical and prepara-
tive C18 RP-HPLC columns were purchased from Resolution
Systems. Preparative RP-HPLC was performed using a C18
preparative scale Vydac column (218TP1022, 2.2 × 25.0 cm,
10-20-µm particle size) eluting with a linear gradient of 0.1%
aqueous TFA with increasing concentrations of AcCN at 15
mL/min. Analytical RP-HPLC analyses were performed on a
Con clu sion s
A phosphorylated pentapeptide (pTyr-Val-Pro-Met-
Leu, compound 2) from the kinase insert region of
PDGF-â receptor inhibits the complexation of PDGF
â-receptor with the C-terminal SH2 domain of PI
3-kinase with an IC50 of 0.445 ( 0.04µM. Structural
modifications via classical medicinal chemistry ap-
proaches resulted in a tetrapeptide (Ac-pTyr-Val-Pro-
Met-NH2, compound 8) with enhanced activity (IC50
)
0.264 ( 0.029 µM). Dialkylamides substituted for the
C-terminal methionine residue were shown to enhance
affinity. In fact, the dihexylamide represents the most
potent analogue reported to date with an IC50 of 0.076
( 0.010 µM (compound 18). In addition, compound 18
was shown to possess good selectivity for p85 C-SH2
over src and abl SH2 domains. Analogues stabilized
against intracellular phosphatases were also prepared
by the incorporation of the CF2Pmp moiety for pTyr
(compounds 25 and 26). These compounds maintained
good affinity with IC50’s of 0.945 ( 0.212 and 1.45 (
0.26 µM for p85 C-SH2, respectively. Compounds 25
and 26 were able to block the association of PDGF-â

4338 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22
Eaton et al.
Vydac column (218TP54, 0.46 × 25.0 cm, 5-µm particle size).
The conditions for analytical RP-HPLC were as follows: 100:0
to 34:66 0.1% aqueous TFA-0.1% TFA in AcCN, linear
gradient over 22 min at 1.5 mL/min (λ ) 214 and 280 nm).
Electrospray mass spectra (ESMS) were determined with
a Fisons VG Trio 2000 quadrupole mass spectrometer using a
1.0% 50:50 water-methanol in acetic acid solution. Fast atom
bombardment mass spectra (FABMS) were obtained on a VG
analytical 7070E/HF mass spectrometer in a matrix of either
thioglycerol or 3-nitrobenzyl alcohol using xenon as the target
gas.
Syn th etic Ch em istr y. Meth od A: The peptides (com-
pounds 1-10) were synthesized by solid-phase peptide syn-
thetic techniques on an ABI 431A automated peptide synthe-
sizer on a 0.25-mmol scale. The C-terminal carboxylates were
prepared on a Wang resin,56 while the C-terminal carboxam-
ides were assembled on a Rink amide38,39 resin using an NR-
Fmoc protection strategy.58,59 Individual NR-Fmoc amino acids
were coupled as their preformed symmetrical anhydrides with
DCC in NMP. At each step the NR-Fmoc group was removed
with 20% piperidine in NMP. Acetylation of the N-terminus
was performed on the resin with 50% acetic anhydride in DCM.
The peptides were cleaved from the resin using TFA-water-
phenol-thioanisole-ethanedithiol (82.5:5.0:5.0:5.0:2.5).60 The
crude peptides were then solubilized in AcCN/water and
lyophilized. The resulting crude peptide was purified by RP-
HPLC.
Meth od B: For the peptidomimetic tripeptides (compounds
11-16), the tripeptide Ac-Tyr(Bzl)-Val-Ala was prepared on
a Sasrin resin40,41 (5.0 mmol of free alcohol total) in a manual
shaker employing the same solid-phase peptide synthesis
techniques as above. After resin cleavage with 1% TFA in
DCM and lyophilization the appropriate amine (5.0 equiv) was
coupled using EDCI (5 equiv), HOBt (5.0 equiv), and NMM
(5.0 equiv) in tetrahydrofuran (THF; 6 h). Debenzylation of
tyrosine was achieved with hydrogen under pressure (50-55
psi) with 20% Pd/C in methanol. The tripeptides were then
phosphitylated with 2.63 g (37.5 mmol) of 1H-tetrazole and
3.12 g (12.5 mmol) of di-tert-butyl diethylphosphoramidite.61
The peptide resin was then oxidized with 0.68 mL (5.0 mmol)
of 70% aqueous tert-butyl hydroperoxide. The compounds were
cleaved with 95% TFA/5% water, concentrated under reduced
pressure, resuspended in water (∼100 mL), and lyophilized.
The resulting crude diastereomeric peptidomimetics were
purified and separated by RP-HPLC.
Meth od C: For the peptidomimetic tripeptides (compounds
17-23), the tripeptide Ac-Tyr(Bzl)-Val-Ala or Ac-Tyr(Bzl)-Val-
D-Ala was prepared on a Sasrin resin40,41 (5.0 mmol of free
alcohol total) in a manual shaker employing the same solid-
phase peptide synthesis techniques as above on a 5.0-mmol
scale. After resin cleavage with 1% TFA in DCM and lyo-
philization, the appropriate amine (5.0 equiv) was coupled
using HATU (5.0 equiv), HOAt (5.0 equiv), and NMM (5.0
equiv) in THF (6 h). Debenzylation of tyrosine was achieved
with hydrogen under pressure (50-55 psi) with 20% Pd/C in
methanol. The tripeptides were then phosphitylated with 2.63
g (37.5 mmol) of 1H-tetrazole and 3.12 g (12.5 mmol) of di-
tert-butyl diethylphosphoramidite.61 The compound was then
oxidized with 0.68 mL (5.0 mmol) of 70% aqueous tert-butyl
hydroperoxide. The compounds were cleaved with 95% TFA/
5% water, concentrated under reduced pressure, resuspended
in water (∼100 mL), and lyophilized. The resulting crude
diastereomeric peptidomimetics were purified and separated
by RP-HPLC.
and then purified by preparative RP-HPLC as described above.
Fractions determined to be pure by analytical RP-HPLC were
combined, concentrated under reduced pressure, and lyophi-
lized.
P ep tid e P u r ity a n d Ch a r a cter iza tion . The purified
peptides/peptidomimetics were assessed for purity by analyti-
cal RP-HPLC and characterized by ESMS or FABMS.
P r ep a r a tion of p Tyr -Met-Asp -Met-Ser (1). The peptide
synthesis, cleavage of the resin, and deprotection were per-
formed as described in the aforementioned general procedure
(method A) to afford 35 mg of the title compound: HPLC tR
)
10.0 min (>99%); FABMS (m/z)+ calcd 725.74, found 726.2 (M
+ 1).
P r ep a r a tion of p Tyr -Va l-P r o-Met-Leu (2). The peptide
synthesis, cleavage of the resin, and deprotection were per-
formed as described in the aforementioned general procedure
(method A) to afford 87 mg of the title compound: HPLC tR
)
14.0 min (>99%); ESMS (m/z)+ calcd 701.78, found 702.2 (M).
P r ep a r a tion of p Tyr -Va l-P r o-Met (3). The peptide syn-
thesis, cleavage of the resin, and deprotection were performed
as described in the aforementioned general procedure (method
A) to afford 6 mg of the title compound: HPLC tR ) 13.2 min
(>99%); ESMS (m/z)+ calcd 588.20, found 590.0 (M).
P r ep a r a tion of Ac-p Tyr -Va l-P r o-Met-Leu (4). The pep-
tide synthesis, cleavage of the resin, and deprotection were
performed as described in the aforementioned general proce-
dure (method A) to afford 37 mg of the title compound: HPLC
tR ) 17.7 min (>98%); ESMS (m/z)+ calcd 743.82, found 744.0
(M).
P r ep a r a tion of p Tyr -Va l-P r o-Met-Leu -NH2 (5). The
peptide synthesis, cleavage of the resin, and deprotection were
performed as described in the aforementioned general proce-
dure (method A) to afford 99 mg of the title compound: HPLC
tR ) 15.8 min (>98%); ESMS (m/z)+ calcd 700.80, found 700.5
(M).
P r ep a r a tion of Ac-p Tyr -Va l-P r o-Met-Leu -NH 2 (6). The
peptide synthesis, cleavage of the resin, and deprotection were
performed as described in the aforementioned general proce-
dure (method A) to afford 18 mg of the title compound: HPLC
tR ) 17.1 min (>98%); ESMS (m/z)+ calcd 742.86, found 742.4
(M).
P r ep a r a tion of p Tyr -Va l-P r o-Met-NH2 (7). The peptide
synthesis, cleavage of the resin, and deprotection were per-
formed as described in the aforementioned general procedure
(method A) to afford 43 mg of the title compound: HPLC tR
)
10.4 min (>99%); ESMS (m/z)+ calcd 587.64, found 589.1 (M).
P r ep a r a tion of Ac-p Tyr -Va l-P r o-Met-NH2 (8). The pep-
tide synthesis, cleavage of the resin, and deprotection were
performed as described in the aforementioned general proce-
dure (method A) to afford 35 mg of the title compound: HPLC
tR ) 11.5 min (>99%); ESMS (m/z)+ calcd 629.64, found 630.3
(M).
P r ep a r a tion of p Tyr -Va l-P r o-Nle-NH2 (9). The peptide
synthesis, cleavage of the resin, and deprotection were per-
formed as described in the aforementioned general procedure
(method A) to afford 8 mg of the title compound: HPLC tR
)
18.6 min (>97%); ESMS (m/z)+ calcd 683.75, found 684.6 (M).
P r ep a r a t ion of Ac-p Tyr -Va l-Ala -N-Bu Gly-NH 2 (10).
Initially, 0.42 g (3.0 mmol) of bromoacetic acid was coupled to
a Rink amide38,39 resin (0.25 mmol of free amine total) with
0.52 mL (3.3 mmol) of diisopropylcarbodiimide for 30 min in
a manual shaker,37 the resin was washed with DMSO (2×),
and N-butylamine (0.99 mL, 10.0 mmol) was added and
allowed to mix for 2.5 h. The remaining peptide synthesis,
cleavage of the resin, and deprotection were performed as
described in the aforementioned general procedure (method
A) to afford 34 mg of the title compound: HPLC tR ) 11.9 min
(>99%); ESMS (m/z)- calcd 585.60, found 584.5 (M).
P r ep a r a tion of Ac-p Tyr -Va l-Ala -NHC5H11 (11). The
peptide synthesis, cleavage of the resin, and deprotection were
performed as described in the aforementioned general proce-
dure (method B) to afford 4 mg of the title compound: HPLC
tR ) 13.6 min (>99%); ESMS (m/z)- calcd 542.57, found 541.4
(M).
Meth od D: The peptides (compounds 24-26) were syn-
thesized as in method A with the exception that TFA-
TMSOTf-DMS-EDT-m-cresol (65:19:15:0.8:0.2) was utilized
for the resin cleavage and deprotection.62
Meth od E: The peptidomimetic tripeptides (compounds 27
and 28) were synthesized as in method C with the exception
that TFA-TMSOTf-DMS-EDT-m-cresol (65:19:15:0.8:0.2)
was utilized for the resin cleavage and deprotection.
P ep tid e P u r ifica tion . The crude peptides/peptidomimet-
ics were dissolved in a mixture of aqueous 0.1% TFA and AcCN

Design of Peptidomimetics
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22 4339
P r ep a r a tion of Ac-p Tyr -Va l-D-Ala -NHC5H11 (12). The
peptide synthesis, cleavage of the resin, and deprotection were
performed as described in the aforementioned general proce-
dure (method B) to afford 2 mg of the title compound: HPLC
tR ) 14.0 min (>99%); ESMS (m/z)- calcd 542.57, found 541.4
(M).
P r ep a r a tion of Ac-p Tyr -Va l-D-Ala -N(C5H11)2 (13). The
peptide synthesis, cleavage of the resin, and deprotection were
performed as described in the aforementioned general proce-
dure (method B) to afford 17 mg of the title compound: HPLC
tR ) 13.4 min (>99%); ESMS (m/z)- calcd 612.71, found 611.5
(M).
formed as described in the aforementioned general procedure
(method D) to afford 9 mg of the title compound: HPLC tR
)
17.6 min (>98%); ESMS (m/z)- calcd 699.81, found 699.3 (M).
P r ep a r a tion of CF 2P m p -Va l-P r o-Met-Leu (25). The
peptide synthesis, cleavage of the resin, and deprotection were
performed as described in the aforementioned general proce-
dure (method D) to afford 57 mg of the title compound: HPLC
tR ) 14.2 min (>99%); ESMS (m/z)- calcd 735.79, found 735.3
(M).
P r ep a r a tion of CF 2P m p -Va l-P r o-Met-NH2 (26). The
peptide synthesis, cleavage of the resin, and deprotection were
performed as described in the aforementioned general proce-
dure (method D) to afford 44 mg of the title compound: HPLC
tR ) 11.0 min (>99%); ESMS (m/z)- calcd 622.65, found 622.3
(M).
P r ep a r a tion of Ac-CF2P m p -Va l-Ala -N(C5H11)2 (27). The
peptide synthesis, cleavage of the resin, and deprotection were
performed as described in the aforementioned general proce-
dure (method E) to afford 5 mg of the title compound: HPLC
tR ) 16.4 min (>99%); ESMS (m/z)- calcd 646.72, found 645.4
(M).
P r ep a r a tion of Ac-CF2P m p -Va l-Ala -N(C6H13)2 (28). The
peptide synthesis, cleavage of the resin, and deprotection were
performed as described in the aforementioned general proce-
dure (method E) to afford 16 mg of the title compound: HPLC
tR ) 19.9 min (>85%); ESMS (m/z)- calcd 674.77, found 673.6
(M).
P r ep a r a tion of Ac-p Tyr -Va l-Ala -N(C5H11
) (14). The
2
peptide synthesis, cleavage of the resin, and deprotection were
performed as described in the aforementioned general proce-
dure (method B) to afford 6 mg of the title compound: HPLC
tR ) 13.1 min (>96%); ESMS (m/z)- calcd 612.71, found 611.6
(M).
P r ep a r a tion of Ac-p Tyr -Va l-D-Ala -N(C4H9)2 (15). The
peptide synthesis, cleavage of the resin, and deprotection were
performed as described in the aforementioned general proce-
dure (method B) to afford 22 mg of the title compound: HPLC
tR ) 10.7 min (>97%); ESMS (m/z)- calcd 584.66, found 583.6
(M).
P r ep a r a tion of Ac-p Tyr -Va l-Ala -N(C4H9)2 (16). The
peptide synthesis, cleavage of the resin, and deprotection were
performed as described in the aforementioned general proce-
dure (method B) to afford 17 mg of the title compound: HPLC
tR ) 10.3 min (>93%); ESMS (m/z)- calcd 584.66, found 583.6
(M).
P r ep a r a tion of Ac-p Tyr -Va l-D-Ala -N(C6H13)2 (17). The
peptide synthesis, cleavage of the resin, and deprotection were
performed as described in the aforementioned general proce-
dure (methods B and C) to afford 32 mg of the title compound
via method C: HPLC tR ) 15.8 min (>99%); ESMS (m/z)- calcd
640.76, found 639.6 (M).
Molecu la r Mod elin g. The homology modeling was per-
formed using SYBYL software package (version 6.0)63 and a
Silicon Graphics workstation. Due to the overall similarity
of the sequences and loop sizes of syp N-terminal SH2 domain
and p85 C-SH2, the 2.05 Å X-ray structure of syp N-terminal
SH2 domain bound to the high-affinity PDGF receptor se-
quence of Val-Leu-pTyr-Thr-Ala-Val-Gln-Pro64 was used as a
basis coordinate set in designing a three-dimensional homology
model of p85 C-SH2. The sequence alignment used was that
published by Lee et al.64 Subsequently, the NMR structures
of p85 N-SH265 and PLCγ-SH2 domains66 were used to model
short portions of the p85 C-SH2 sequence that showed greater
local similarity. (Note: at the time that this work was
performed, the NMR structure of p85 SH2 had not been
P r ep a r a tion of Ac-p Tyr -Va l-Ala -N(C6H13
) (18). The
2
peptide synthesis, cleavage of the resin, and deprotection were
performed as described in the aforementioned general proce-
dure (methods B and C) to afford 11 mg of the title compound
via method C: HPLC tR ) 15.7 min (>99%); ESMS (m/z)- calcd
640.76, found 639.3 (M).
reported.33
) The starting model for the parent molecule
P r ep a r a tion of Ac-p Tyr -Va l-Ala -N(C8H17
) (19). The
2
(compound 2) was based upon the structure of the high-affinity
PDGF receptor sequence of Val-Leu-pTyr-Thr-Ala-Val-Gln-Pro
starting from residue 740 bound to syp N-terminal SH2
domain64 and was modified to the various peptide structures
which were each manually docked into the p85 C-SH2 domain
model and minimized. In particular, the aliphatic groups at
the p + 3 site were modeled (1) into the pY + 3 pocket based
on the +3 Val conformation in the syp-bound PDGF sequence
and (2) into the nearby hydrophobic channel based on the
conformation of the +3/+4 PDGF backbone sequence. Mini-
mizations of the inhibitor models and protein model were
carried out using the Tripos force field.63
peptide synthesis, cleavage of the resin, and deprotection were
performed as described in the aforementioned general proce-
dure (method C) to afford 68 mg of the title compound: HPLC
tR ) 15.5 min (>99%); ESMS (m/z)- calcd 696.87, found 696.4
(M).
P r ep a r a tion of Ac-p Tyr -Va l-D-Ala -NH(CH2)2-C6H5 (20).
The peptide synthesis, cleavage of the resin, and deprotection
were performed as described in the aforementioned general
procedure (method C) to afford 21 mg of the title compound:
HPLC tR ) 10.7 min (>99%); ESMS (m/z)- calcd 576.59, found
575.5 (M).
P r ep a r a t ion of Ac-p Tyr -Va l-Ala -NH(CH 2)2-C6H5 (21).
The peptide synthesis, cleavage of the resin, and deprotection
were performed as described in the aforementioned general
procedure (method C) to afford 23 mg of the title compound:
HPLC tR ) 14.2 min (>99%); ESMS (m/z)- calcd 576.59, found
575.3 (M).
P r ep a r a tion of Ac-p Tyr -Va l-D-Ala -NH(CH2)2-C6H11 (22).
The peptide synthesis, cleavage of the resin, and deprotection
were performed as described in the aforementioned general
procedure (method C) to afford 12 mg of the title compound:
HPLC tR ) 9.4 min (>99%); ESMS (m/z)- calcd 582.64, found
581.6 (M).
P r ep a r a tion of Ac-p Tyr -Va l-Ala -NH(CH2)2-C6H11 (23).
The peptide synthesis, cleavage of the resin, and deprotection
were performed as described in the aforementioned general
procedure (method C) to afford 9 mg of the title compound:
HPLC tR ) 9.8 min (>99%); ESMS (m/z)- calcd 582.64, found
581.4 (M).
Exp r ession a n d P u r ifica tion of th e C-Ter m in a l SH2
Dom a in of th e p 85 Su bu n it of P I 3-Kin a se-GST F u sion
P r otein (GST p 85 C-SH2). The pGEX plasmid expressing
the GST p85 C-SH2 fusion protein was used in these studies.
Both the expression and purification of the fusion protein were
performed as previously described.67 To prepare the [35S]GST
p85 C-SH2 fusion protein, a 75-mL overnight culture of
Escherichia coli expressing the GST p85 fusion protein was
added to 1.0 L of LB broth containing 100 µg/mL ampicillin.
The cultures were incubated at 37 °C until reaching a density
of A600 ) 1.0. Isopropyl thio-â-D-galactoside (1 mM) was added;
15 min later, 10 mCi of trans-35S-label was added and cultures
were incubated for an additional 3 h at 37 °C. The cells were
then lysed by sonication and fusion proteins purified by affinity
chromatography using glutathione-agarose (CL-4B) beads.
P r ep a r a tion of th e In tr a cellu la r Tyr osin e Kin a se
Dom a in of th e P DGF -â Recep tor . Lysates from SF9 insect
cells expressing the PDGF-â receptor tyrosine kinase (PDGFR-
TK) were incubated with M2 affinity beads, and the complexes
were washed several times with Tris buffer containing pro-
P r ep a r a tion of P m p -Va l-P r o-Met-Leu (24). The peptide
synthesis, cleavage of the resin, and deprotection were per-

4340 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22
Eaton et al.
tease inhibitors and sodium orthovanadate. Complexes were
centrifuged and resuspended in HEPES buffer containing 1
mM ATP, 10 mM MnCl2, and 5 mM MgCl2 to stimulate
phosphorylation of the PDGFR-TK.
concentration of 5 mg/mL to allow the unambiguous identifica-
tion of microinjected cells. Each experiment was conducted a
minimum of three separate times, with at least 250 cells
injected on each occasion. After injection, cells were stimu-
lated and stained for immunofluorescence microscopy as
described below. Results were analyzed on a Zeiss Axiophot
epifluorescence microscope, and digital photography was ac-
complished with a Photometrix CCD camera.
Sin gle-Cell Im m u n oflu or escen ce Assa ys. 1. DNA Syn -
th esis. Two hours following injection, BrdU was added to the
medium, and the cells were incubated at 37 °C for 16 h.69 Cells
were then fixed in 3.7% formaldehyde and stained with a
monoclonal anti-BrdU antibody (Accurate Scientific). Incor-
poration of BrdU into newly synthesized DNA (an indication
of progression of the cells into S phase) was detected by
simultaneous staining with rhodamine anti-rat IgG and
fluorescein anti-rabbit IgG.70 This staining method results in
red nuclear staining in cells which have incorporated BrdU
and green cytoplasmic staining in cells which were microin-
jected.
2. GLUT4 Tr a n sloca tion . One hour after injection,
differentiated 3T3-L1 cells were treated with or without 10
ng/mL insulin for 2 h and then fixed, permeabilized, and
stained50 using affinity-purified rabbit polyclonal antibody
F349 (a gift from Mike Mueckler) raised against GLUT4.
Detection of GLUT4 distribution in the cells was detected with
fluorescein anti-rabbit IgG, and injected cells were identified
by staining with rhodamine anti-sheep IgG. This single-cell
assay for insulin-stimulated glucose uptake was then quan-
titated by assessment of the translocation of the GLUT4
transporter from its intracellular stores in the trans Golgi
apparatus to the plasma membrane.53
3. c-fos Exp r ession . One hour after injection, cells were
stimulated with 100 ng/mL insulin where indicated and fixed
in 3.7% formaldehyde after 60 min of stimulation.51 Nuclear
c-fos expression was detected in single cells by staining with
polyclonal anti-fos-specific rabbit IgG (Oncogene Science),51
followed by rhodamine anti-rabbit IgG and fluorescein anti-
sheep IgG to detect injected cells. Results were quantitated
by assessment of nuclear rhodamine staining.
GST p 85 C-SH2-P DGF R-TK Bin d in g Assa y. Binding
of [35S]GST p85 C-SH2 fusion proteins to the phosphorylated
PDGFR-TK was assayed in 20 mM HEPES buffer containing
10 µg/mL each of the protease inhibitors, phenylmethanesulfo-
nyl fluoride (PMSF), pepstatin, leupeptin, and aprotinin with
0.5 mM EDTA and 0.1% NP-40. The binding assays were
performed in 96-well Millipore filter plates in a final volume
of 250 µL of HEPES buffer containing 135 µL of phosphory-
lated PDGFR-TK beads complex (1 µg of receptor/well), 10 µL
of [35S]GST p85 C-SH2 fusion protein (30 000 cpm/well), and
5 µL of peptide inhibitor as indicated. Samples were incubated
at 25 °C for 20 min with continuous rocking. Binding was
terminated by filtration through the filter plates using a
Millipore multiscreen filtration manifold. Filter plates were
washed four times with 150 µL of HEPES buffer followed by
the addition of 30 µL of Hi-load scintillant. The radioactivity
that was retained on the filters was counted in a Wallac 1450
Microbeta counter. Total binding was defined as [35S]GST p85
C-SH2 fusion protein bound to the PDGFR-TK bead complex
retained on the filter plates. Specific binding was defined as
total binding minus nonspecific binding. IC50 values were
calculated by weighted nonlinear regression curve fitting.
GST-sr c a n d -a b l SH2 Dom a in Bin d in g Assa ys.
These assays were performed as previously described.21,22
P DGF Recep tor /P I 3-Kin a se Cell Associa tion Assa y.
Rat aortic smooth muscle cells were grown to confluency in
100-mm dishes. Growth medium was removed and replaced
with serum-free medium, and cells were incubated at 37 °C
for an additional 24 h. Test compounds were then added
directly to the medium and cells incubated for an additional
24 h. At this point, rat PDGF-BB (â-homodimer) was added
at a final concentration of 30 ng/mL for 5 min at 37 °C to
stimulate autophosphorylation of the PDGF-â receptor and
association of PI 3-kinase to the phosphorylated receptors.
Following growth factor treatment, the medium was removed,
and cells were washed with cold phosphate-buffered saline and
immediately lysed with 1 mL of lysis buffer [50 mM HEPES
(pH 7.5), 150 mM NaCl, 10% glycerol, 1.0% Triton-X 100, 1
mM EDTA, 1 mM EGTA, 50 mM NaF, 1 mM sodium ortho-
vanadate, 30 mM p-nitrophenyl phosphate, 10 mM sodium
pyrophosphate, 1 mM PMSF, 10 µg/mL aprotinin, and 10 µg/
mL leupeptin]. Lysates were centrifuged at 10000g for 10 min.
Supernatants were incubated for 2 h with 10 µL of anti-human
PDGF receptor polyclonal antibody (1:1000) which recognizes
the PDGF receptor â-isoform. Following the incubation,
protein-A-sepharose beads were added for 2 h with continuous
mixing and immune complexes bound to the beads washed four
times with 1 mL of lysis wash buffer. Immune complexes were
solubilized in 30 µL of Laemmli sample buffer and electro-
phoresed in 4-20% SDS-polyacrylamide gels. Following
electrophoresis, separated proteins were transferred to nitro-
cellulose and immunoblotted with anti-rat PI 3-kinase anti-
serum which recognizes the 85-kDa subunit of PI 3-kinase.
Following incubation with [125I]protein-A, p85 protein levels
were detected by phosphorimage analysis and protein bands
quantitated via densitometry. IC50 values were generated
from the densitometric data.
4. Mem br a n e Ru fflin g. Thirty minutes after injection,
cells were stimulated with 100 ng/mL insulin where indicated
for 3 min and then fixed and stained.52 Ruffling was detected
by staining with rhodamine-conjugated phalloidin, and in-
jected cells were detected with fluorescein anti-rabbit IgG.
Lamellipodia (membrane ruffles) were quantitated by visual-
ization of actin rearrangement from organized stress fibers in
unstimulated cells to concentrated areas of actin-containing
ruffles at the plasma membrane.
Ackn owledgm en t. The authors wish to thank Darin
R. Kent for the preparation of compound 19. The
authors also thank J ames H. Fergus and Susan Hubbell
for providing the selectivity data.
Refer en ces
(1) Lechleider, R. J .; Sugimoto, S.; Bennett, A. M.; Kashishian, A.
S.; Cooper, J . A.; Shoelson, S. E.; Walsh, C. T.; Neel, B. G.
Activation of the SH2-containing phosphotyrosine phosphatase
SH-PTP2 by its binding site, phosphotyrosine 1009, on the
human platelet-derived growth factor receptor â. J . Biol. Chem.
1993, 268, 21478-21481.
(2) Fry, M. J .; Panayotou, G.; Booker, G. W.; Waterfield, M. D. New
insights into protein-tyrosine kinase receptor signaling com-
plexes. Protein Sci. 1993, 2, 1785-1797.
(3) Beattie, J . SH2 domain protein interaction and possibilities for
pharmacological intervention. Cell. Signal. 1996, 8, 75-86.
(4) Capron, L.; Heudes, D.; Chajara A.; Bruneval, P. Effect of
ramipril, an inhibitor of angiotensin converting enzyme, on the
response of rat thoracic aorta to injury with a balloon catheter.
J . Cardiovasc. Pharmacol. 1991, 18, 207-211.
(5) Bourassa, M. Silent myocardial ischemia after coronary angio-
plasty: distinguishing the shadow from the substance. J . Am.
Coll. Cardiol. 1992, 19, 1410-1411.
(6) Herrman, J . R.; Hermans, W. R. M.; Vos, J .; Serruys, P. W.
Pharmacological approaches to the prevention of restenosis
following angioplasty. Drugs 1993, 46, 18-52.
Micr oin jection Exp er im en ts. The procedures used for
the microinjection of purified antibodies and fusion proteins
has been previously described.49 Insulin-responsive rat 1
fibroblasts were maintained68 and stimulated where indicated
with 100 ng/mL insulin. 3T3-L1 adipocytes were cultured and
differentiated using established methods and injected.53 Cells
were plated on 12-mm glass coverslips and were rendered
quiescent by incubation in serum-free medium for 24-36 h.
Cells were microinjected with either compound 25 or 26 (50
µM), recombinant GST-p85 SH2 domain fusion protein (12
mg/mL), or a polyclonal affinity-purified anti-p85 IgG (3.5 mg/
mL) (a gift from Lou Cantley). Preimmune IgG of the
appropriate species was added to all samples to give a final

Design of Peptidomimetics
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22 4341
(7) Gordon, D.; Reidy, M. A.; Benditt, E. P.; Schwartz, S. M. Cell
proliferation in human coronary arteries. Proc. Natl. Acad. Sci.
U.S.A. 1990, 87, 4600-4604.
(8) Majesky, M. W.; Reidy, M. A.; Bowen-Pope, D. F.; Hart, C. E.;
Wilcox, J . N.; Schwartz, S. M.; PDGF ligand and receptor gene
expression during repair of arterial injury. J . Cell Biol. 1990,
111, 2149-2158.
istry, Structure and Biology. Proceedings of the Fourteenth
American Peptide Symposium; Kaumaya, P. T. P., Hodges, R.
S., Eds.; Mayflower Scientific Ltd.: West Midlands, England,
1996; pp 414-415.
(28) Cody, W. L.; Eaton, S. R.; Kent, D. R.; Panek, R. L.; Lu, G. H.;
Dahring, T. K.; Doherty, A. M. The design of peptidomimetic
inhibitors of the association of PDGF-â receptor and PI 3-kinase
with cellular activity. In Peptides 1996: Proceedings of the
Twenty-Fourth European Peptide Symposium; Ramage, R.,
Epton, R., Eds.; Mayflower Scientific Ltd.: West Midlands,
England, 1998; pp 307-308.
(29) Wasksman, G.; Shoelson, S. E.; Pant, N.; Cowburn, D.; Kuriyan,
J . Binding of a high affinity phosphotyrosyl peptide to the src
SH2 domain: Crystal structures of the complexed and peptide-
free forms. Cell 1993, 72, 779-790.
(30) Eck, M. J .; Shoelson, S. E.; Harrison, S. C. Recognition of a high-
affinity phosphotyrosyl peptde by the src homology-2 domain of
p56lck. Nature 1993, 362, 87-91.
(9) J awien, A.; Bowen-Pope, D. F.; Lindner, V.; Schwartz, S. M.;
Clowes, A. W. Platelet-derived growth factor promotes smooth
muscle migration and initimal thickening in a rat model of
balloon angioplasty. J . Clin. Invest. 1992, 89, 507-511.
(10) Ferns, G. A. A.; Raines, E. W.; Sprugel, K. H.; Motani, A. S.;
Reidy, M. A.; Ross, R. Inhibition of neointimal smooth muscle
accumulation after angioplasty by an antibody to PDGF. Science
1991, 253, 1129-1132.
(11) Birge, R. B.; Hanafusa, H. Closing in on SH2 specificity. Science
1993, 262, 1522-1524.
(12) Pawson, T. Protein modules and signaling networks. Nature
1995, 373, 573-580.
(31) Nolte, R. T.; Eck, M. J .; Schlessinger, J .; Shoelson, S. E.;
Harrison, S. C. Crystal structure of the PI 3-kinase p85 amino-
terminal SH2 domain and its phosphopeptide complexes. Nature
Struct. Biol. 1996, 3, 364-373.
(13) Marengere, L. E. M.; Pawson, T. Structure and function of SH2
domains. J . Cell Sci. 1994, 18, 97-104.
(14) Fantl, W. J .; Escobedo, J . A.; Martin, G. A.; Turck, C. W.; del
Rosario, M.; McCormick, F.; Williams, L. T. Distinct phospho-
tyrosines on a growth factor receptor bind to specific molecules
that mediate different signaling pathways. Cell 1992, 69, 413-
423.
(32) Weston, S.; Derbyshire, D.; Breeze, A.; Pauptit, R. 1.7
Å
Structure of the C-terminal SH2 domain of the p85 subunit of
human phosphatidylinositol 3-kinase. Abstracts of Papers; In-
ternational Union of Crystallography, IUCr XVII: Seattle, WA,
1996; Abstract PS04.10.26.
(15) Kaplan, D. R.; Morrison, D. K.; Wong, G.; McCormick, F.;
Williams, L. T. PDGF â-receptor stimulates tyrosine phospho-
rylation of GAP and association of GAP with a signaling complex.
Cell 1990, 61, 125-133.
(16) Koch, A.; Anderson, D.; Moran, M.; Ellis, C.; Pawson, T. SH2
and SH3 domains: elements that control interactions of cyto-
plasmic signaling proteins. Science 1991, 252, 668-674.
(17) Panayotou, G.; Gish, G.; End, P.; Troung, O.; Gout, I.; Dhand,
R.; Fry, M. J .; Hiles, I.; Pawson, T.; Waterfield, M. D. Interac-
tions between SH2 domains and tyrosine-phosphorylated platelet-
derived growth factor â-receptor sequences: Analysis of kinetic
parameters by a novel biosensor-based approach. Mol. Cell. Biol.
1993, 13, 3567-3576.
(18) Klippel, A.; Escobedo, J . A.; Fantl, W. J .; Williams, L. T. The
C-terminal SH2 domain of p85 accounts for the high affinity and
specificity of the binding of phosphatidylinositol 3-kinase to
phosphorylated platelet-derived growth factor â receptor. Mol.
Cell. Biol. 1992, 12, 1451-1459.
(19) 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, Cantley, L. C. SH2 domains
recognize specific phosphopeptide sequences. Cell 1993, 72, 767-
778.
(20) Cantley, L. C.; Songyang, Z. Specificity in recognition of phos-
phopeptides by src-homology 2 domains. J . Cell Sci. 1994, 18,
121-126.
(21) Plummer, M. S.; Lunney, E. A.; Para, K. S.; Shahripour, A.;
Stankovic, C. J .; Humblet, C.; Fergus, J . H.; Marks, J . S.;
Herrera, R.; Hubbell, S.; Saltiel, A.; Sawyer, T. K. Design of
peptidomimetic lignads for the pp60src SH2 domain. Bioorg. Med.
Chem. 1997, 5, 41-47.
(22) Plummer, M. S.; Holland, D. R.; Shahripour, A.; Lunney, E. A.;
Fergus, J . H.; Marks, J . S.; McConnell, P.; Mueller, W. T.;
Sawyer, T. K. Design, Synthesis, and Cocrystal Structure of a
Nonpeptide Src SH2 Domain Ligand. J . Med. Chem. 1997, 40,
3719-3725.
(23) Charifson, P. S.; Shewchuk, L. M.; Rocque, W.; Hummel, C. W.;
J ordan, S. R.; Mohr, C.; Pacofsky, G. J .; Peel, M. R.; Rodriguez,
M.; Sternbach, D. D.; Consler, T. G. Peptide ligands of pp60c-src
SH2 domains: A thermodynamic and structural study. Bio-
chemistry 1997, 36, 6283-6293.
(24) Pacofsky, G. J .; Lackey, K.; Alligood, K. J .; Berman, J .; Charifson,
P. S.; Crosby, R. M.; Dorsey, G. F.; Feldman, P. L.; Gilmer, T.
M.; Hummel, C. W.; J ordan, S. R.; Mohr, C.; Rodriguez, M.;
Shewchuk, L. M.; Sternbach, D. D. Potent dipeptide inhibitors
of the pp60c-src SH2 domain. J . Med. Chem. 1998, 41, 1894-
1908.
(25) Garcia-Echeverria, C.; Furet, P.; Gay, B.; Fretz, H.; Rahuel, J .;
Schoepfer, J .; Caravatti, G. Potent Antagonists of the SH2
Domain of Grb2: Optimization of the X+1 Position of 3-Amino-
Z-Tyr(PO3H2)-X+1-Asn-NH2. J . Med. Chem. 1998, 41, 1741-
1744.
(26) Ramalingam, K.; Eaton, S. R.; Cody, W. L.; Lu, G. H.; Panek, R.
L.; Waite, L. A.; Decker, S. J .; Keiser, J . A.; Doherty, A. M.
Structure-activity studies of phosphorylated peptide inhibitors
of the association of phosphatidylinositol 3-kinase with PDGF-â
receptor. Bioorg. Med. Chem. 1995, 3, 1263-1272.
(27) Eaton, S. R.; Ramalingam, K.; Cody, W. L.; Holland, D. R.;
Panek, R. L.; Lu, G. H.; Doherty, A. M. Structure-activity
relationships of peptides that block the association of PDGF
â-receptor with phosphatidylinositol 3-kinase. In Peptides: Chem-
(33) 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 p85 alpha subunit of phosphatidylinositol 3-kinase.
EMBO J . 1996, 15, 3579-3589.
(34) Hensmann, M.; Booker, G. W.; Panayotou, G.; Boyd, J .; Linacre,
J .; Waterfield, M.; Campbell, I. D. Phosphopeptide binding to
the N-terminal SH2 domain of the p85R subunit pf PI 3-kinase:
A heteronuclear NMR study. Protein Sci. 1994, 3, 1020-1030.
(35) Roller, P. P.; Otaka, A.; Nomizu, M.; Smyth, M. S.; Barchi, J . J .;
Burke, T. R.; Case, R. D.; Wolf, G.; Shoelson, S. E. Norleucine
as a replacement for methionine in phosphatase-resistant linear
and cyclic peptides which bind to p85 SH2 domains. Bioorg. Med.
Chem. Lett. 1994, 4, 1879-1872.
(36) Stankovic, C. J .; Plummer, M. S.; Sawyer, T. K. Peptidomimetic
ligands for src homolgy-2 domains. In Advances in Amino Acid
Mimetics and Peptidomimetics; Abell, A., Ed.; J ai Press: Green-
wich, CT, 1997; Vol. 1, pp 127-163.
(37) Zuckermann, R. N.; Kerr, J . M.; Kent, S. B. H.; Moos, W. H.
Efficient method for the preparation of peptoids [oligo(N-
substituted glycines)] by submonomer solid-phase synthesis. J .
Am. Chem. Soc. 1992, 114, 10646-10647.
(38) Rink, H. Solid-phase synthesis of protected peptide fragments
using a trialkoxy-diphenyl methyl ester resin. Tetrahedron Lett.
1987, 28, 3787-3790.
(39) Bernatowicz, M. S.; Daniels, S. B.; Koster, H. A comparison of
acid-labile linkage agents for the synthesis of peptide C-terminal
amides. Tetrahedron Lett. 1989, 30, 4645-4648.
(40) Mergler, M.; Tanner, R.; Gosteli, J .; Grogg, P. Peptide synthesis
by a combination of solid-phase and solution methods. I. A new
very acid-labile anchor group for the solid phase synthesis of
fully protected fragments. Tetrahedron Lett. 1988, 29, 4005-
4008.
(41) Mergler, M.; Nyfeler, R.; Tanner, R.; Gosteli, J .; Grogg, P. Peptide
synthesis by a combination of solid-phase and solution methods.
II. Synthesis of fully protected peptide fragments on 2-methoxy-
4-alkoxybenzyl alcohol resin. Tetrahedron Lett. 1988, 29, 4009-
4012.
(42) Carpino, L. A.; El-Faham, A.; Albericio, F. Racemization studies
during solid-phase peptide synthesis using azabenzotriazole-
based coupling reagents. Tetrahedron Lett. 1994, 35, 2279-2282.
(43) S. R. Eaton and W. L. Cody, Parke-Davis Pharmaceutical
Research, Division of Warner-Lambert Co., unpublished results.
(44) Burke, T. R.; Smyth, M. S.; Otaka, A.; Nomizu, M.; Roller, P.
P.; Wolf, G.; Case, R.; Shoelson, S. E. Nonhydrolyzable phos-
photyrosyl mimetics for the preparation of phosphatase-resistant
SH2 domain inhibitors. Biochemistry 1994, 33, 6490-6494.
(45) Gilmer, T.; Rodriguez, M.; J ordan, S.; Crosby, R.; Alligood, K.;
Green, M.; Kimery, M.; Wagner, C.; Kinder, D.; Charifson, P.;
Hassell, A. M.; Willard, D.; Luther, M.; Rusnak, D.; Sternbach,
D. D.; Mehrotra, M.; Peel, M.; Shampine, L.; Davis, R.; Robbins,
J .; Patel, I. R.; Kassel, D.; Burkhart, W.; Moyer, M.; Bradshaw,
T.; Berman, J . Peptide inhibitors of src SH3-SH2-phosphoprotein
interactions. J . Biol. Chem. 1994, 269, 31711-31719.
(46) Burke, T. R.; Nomizu, M.; Otaka, A.; Smyth, M. S.; Roller, P.
P.; Case, R. D.; Wolf, G.; Shoelson, S. E. Cyclic peptide inhibitors
of phosphatidylinositol 3-kinase p85 SH2 domain binding.
Biochem. Biophys. Res. Commun. 1994, 201, 1148-1153.
(47) Burke, T. R.; Kole, H. K.; Roller, P. P. Potent inhibition of insulin
receptor dephosphorylation by a hexamer peptide containing the
phosphotyrosyl mimetic F2Pmp. Biochem. Biophys. Res. Com-
mun. 1994, 204, 129-134.

4342 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 22
Eaton et al.
(48) Stankovic, C. J .; Surendran, N.; Lunney, E. A.; Plummer, M.
S.; Para, K. S.; Shahripour, A.; Fergus, J . H.; Marks, J . S.;
Herrera, R.; Hubbell, S. E.; Humblet, C.; Saltiel, A. R.; Stewart,
B. H.; Sawyer, T. K. The role of 4-phosphonodifluoromethyl- and
4-phosphono-phenylalanine in the selectivity and cellular uptake
of SH2 domain ligands. Bioorg. Med. Chem. Lett. 1997, 7, 1909-
1914.
(49) Rose, D. W.; Saltiel, A. L.; Majumdar, M.; Olefsky, J . M. IRS-1
is required for insulin-mediated mitogenic signal transduction.
Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 797-801.
(50) Morris, A. J .; Haruta, T.; Martin, S. S.; Ricketts, W. R.;
Gustafson, T.; Rose, D. W.; Olefsky, J . M. Evidence for an insulin
receptor substrate 1 independent insulin signaling pathway that
mediates insulin-responsive glucose transporter (GLUT4) trans-
location. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 8401-8406.
(51) J hun, B. H.; Rose, D. W.; Seely, B. L.; Saltiel, A. L.; Cantley, L.;
Olefsky, J . M. Microinjection of the SH2 domain of the 85 kDa
subunit of phosphatidylinositol 3-kinase inhibits insulin-induced
DNA synthesis and c-fos expression. Mol. Cell. Biol. 1994, 14,
7466-7475.
(52) Martin, S. S.; Rose, D. W.; Klippel, A.; Williams, L. T.; Olefsky,
J . M. Phosphatidylinositol 3-kinase is necessary and sufficient
for insulin-stimulated stress fiber breakdown. Endocrinology
1996, 137, 5045-5054.
(53) Haruta, T.; Morris, A. J .; Rose, D. W.; Nelson, J . G.; Mueckler,
M.; Olefsky, J . M. Insulin stimulated GLUT4 translocation is
mediated by a divergent intracellular signaling pathway. J . Biol.
Chem. 1995, 270, 27991-27994.
(54) Kari, U. P.; Maddukuri, S. R.; MacDonald, M. T.; J ones, L. M.;
MacDonald, D. L.; Maloy, W. L. Structure-activity relationship
studies of dialkylamine derivatives exhibiting wide spectrum of
antimicrobial activities. Program & Abstracts, Fifteenth Ameri-
can Peptide Symposium, Nashville, TN; American Peptide
Society: Minneapolis, MN, 1997; P0005.
(55) Yokum, T. S.; Elzer, P. H.; McLaughlin, M. L. Antimicrobial R,R-
dialkylated amino acid rich peptides with in-vivo activity against
an intracellular pathogen. J . Med. Chem. 1996, 39, 3603-3605.
(56) Wang, S. S. p-Alkoxybenzyl alcohol resin and p-alkoxybenzyl-
carbonylhydrazide resin for solid-phase synthesis of protected
peptide fragments. J . Am. Chem. Soc. 1973, 95, 1328-1333.
(57) Smyth, M. S.; Burke, T. R. Enantioselective synthesis of N-Boc
and N-Fmoc protected diethyl 4-phosphono(difluoromethyl)-L-
phenylalanine; agents suitable of the solid-phase synthesis of
peptides containing nonhydrolyzable analogues of O-phospho-
tyrosine. Tetrahedron Lett. 1994, 35, 551-554.
(60) King, D. S.; Fields, C. G.; Fields, G. B. A cleavage method which
minimizes side reactions following Fmoc solid-phase peptide
synthesis. Int. J . Pept. Protein Res. 1990, 36, 255-266.
(61) Kitas, E. A.; Knorr, R.; Trzeciak, A.; Bannwarth, W. Alternative
strategies for the Fmoc solid-phase synthesis of O-4-phospho-
L-tyrosine-containing peptides. Helv. Chim. Acta 1991, 74, 1314-
1328.
(62) Otaka, A.; Burke, T. R.; Smyth, M. S.; Nomizu, M.; Roller, P. P.
Deprotection and cleavage methods for protected peptide resins
containing 4-[(diethylphosphono)difluoromethyl]-DL-phenylala-
nine residues. Tetrahedron Lett. 1993, 34, 7039-7042.
(63) Sybyl, molecular modeling software version 6.0 ver. 1.0; Tripos
Inc., St. Louis, MO; 1994.
(64) 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, 423-425.
(65) Booker, G. W.; Breeze, A. L.; Downing, A. K.; Panayotou, G.;
Gout, I.; Waterfield, M. D.; Campbell, I. D. Structure of an SH2
domain of the p85 R subunit of phosphatidylinositol-3-OH
kinase. Nature 1992, 358, 684-687.
(66) 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 phospho-
liphase C-g1 complexed with a high affinity binding peptide. Cell
1994, 77, 461-472.
(67) Zhu, G.; Decker, S. J .; Mayer, B. J .; Saltiel, A. R. Direct analysis
of the binding of the abl src homology 2 domain to the activated
epidermal growth factor receptor. J . Biol. Chem. 1993, 268,
1775-1779.
(68) McClain, D. A.; Maegawa, H.; Lee, J .; Dull, T. J .; Ullrich, A.;
Olefsky, J . M. A mutant insulin receptor with defective tyrosine
kinase displays no biologic activity and does not undergo
endocytosis. J . Biol. Chem. 1987, 262, 14663-14671.
(69) Xiao, S.; Rose, D. W.; Sasaoka, T.; Maegawa, H.; Burke, T. R.;
Roller, P.; Shoelson, S. E.; Olefsky, J . M. Syp (SH-PTP2) is a
positive mediator of growth factor stimulated mitogenic signal
transduction. J . Biol. Chem. 1994, 269, 21244-21248.
(70) Kolch, W.; Philipp, A.; Mischak, H.; Dutil, E. M.; Mullen, T. M.;
Meinkoth, J . M.; Feramisco, J . R.; Rose, D. W. Inhibition of Raf-1
signaling by a monoclonal antibody which interferes with Raf-1
activation and with Mek substrate binding. Oncogene 1996, 13,
1305-1314.
(58) Stewart, J . M.; Young, J . D. Solid-Phase Peptide Synthesis, 2nd
ed.; Pierce Chemical Co.: Rockford, IL, 1984.
(59) Fields, G. B.; Noble, R. L. Solid-phase peptide synthesis utilizing
9-fluorenylmethoxycarbonyl amino acids. Int. J . Pept. Protein
Res. 1990, 35, 161-214.
(71) Nicholls, A.; Sharp, K. A.; Honig, B. Protein folding and
association: Insights from the interfactial and thermodynamic
properties of hydrocarbons. Proteins 1991, 11, 281-296.
J M9802766