SH2-directed ligands of the Lck tyrosine kinase

Article abstract of DOI:10.1021/jm990462r

Two separate libraries, prepared via parallel synthesis, were employed to identify low-molecular-weight SH2-targeted ligands for the Lck tyrosine protein kinase. These libraries were constructed to furnish non-amino acid analogues of the (1) Glu-Glu and (2) Ile residues of the Lck SH2 domain peptide ligand Ac-pTyr-Glu-Glu-Ile-amide. The lead compound acquired in this study exhibits a dissociation constant for the Lck SH2 domain that is comparable to that displayed by Ac-pTyr-Glu-Glu-Ile-amide. These results demonstrate that the standard amino acid residues Glu-Glu-Ile can be completely replaced with non-amino acid moieties without loss of SH2 affinity.

Full text of DOI:10.1021/jm990462r

J . Med. Chem. 2000, 43, 1173-1179
1173
SH2-Dir ected Liga n d s of th e Lck Tyr osin e Kin a se
Tae Ryong Lee and David S. Lawrence*
Department of Biochemistry, The Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, New York 10461
Received September 10, 1999
Two separate libraries, prepared via parallel synthesis, were employed to identify low-molecular-
weight SH2-targeted ligands for the Lck tyrosine protein kinase. These libraries were
constructed to furnish non-amino acid analogues of the (1) Glu-Glu and (2) Ile residues of the
Lck SH2 domain peptide ligand Ac-pTyr-Glu-Glu-Ile-amide. The lead compound acquired in
this study exhibits a dissociation constant for the Lck SH2 domain that is comparable to that
displayed by Ac-pTyr-Glu-Glu-Ile-amide. These results demonstrate that the standard amino
acid residues Glu-Glu-Ile can be completely replaced with non-amino acid moieties without
loss of SH2 affinity.
In tr od u ction
10-7-10-6 M.4 In addition, we have recently found that
the latter affinities can be significantly enhanced by
appending non-amino acid substituents off the N-
terminus of peptide ligands (i.e. RCO-NH-pTyr-Glu-Glu-
Ile-amide).4a
Tyrosine-specific protein kinases are composed of two
subfamilies: (1) the receptor tyrosine kinases, which are
integral membrane proteins, and (2) their nonreceptor
cytoplasmic counterparts. The former, upon coordina-
tion of specific extracellular ligands, forms aggregates
and subsequently suffers phosphorylation of key tyro-
sine residues. Cytoplasmic signaling proteins, including
nonreceptor tyrosine kinases, coordinate to these phos-
photyrosine (pTyr) residues through Src homology 2
(SH2) domains.1 This binding event triggers the activa-
tion of specific intracellular signaling pathways ulti-
mately leading to a cellular response in reaction to the
extracellular stimulus. For example, antigen presenta-
tion to T cells results in the clustering of T cell receptors
(TcR) and subsequent TcR tyrosine phosphorylation by
the Src family member Lck (Lymphoid T cell tyrosine
kinase).2 These TcR pTyr moieties serve as high-affinity
binding sites for the SH2 domains of cytoplasmic
signaling proteins, which generate complexes that are
required for ensuing downstream events such as an
increase in cytoplasmic Ca2+ levels and the ultimate
production of interleukin-2. In short, SH2 domains play
a critical role in organizing coherent signal transducing
complexes that are essential for the appropriate cellular
response to extracellular stimuli.
Although moderately high-affinity peptide-based
ligands for SH2 domains have been reported, their
ultimate utility as therapeutic agents is suspect. In
general, peptide-based species exhibit poor bioavailabil-
ity and are prone to hydrolysis by the presence of intra-
and intercellular proteases. Furthermore, the pTyr
moiety, which is essential for SH2 recognition, is hydro-
lytically unstable due to the presence of protein tyrosine
phosphatases. The latter difficulty can be resolved via
the introduction of nonhydrolyzable pTyr mimetics, such
as the difluorophosphonate analogue of pTyr.5 Peptide-
based species that contain this nonhydrolyzable deriva-
tive exhibit affinities comparable to their pTyr-contain-
ing counterparts.5 Nevertheless, the acquisition of
nonpeptidic ligands that display high-affinity binding
for SH2 domains has not yet been fully realized.6,7 We
report herein that the three C-terminal amino acids in
the peptide-based SH2 ligand Ac-pTyr-Glu-Glu-Ile-
amide can be replaced by non-amino acid moieties
without loss of SH2 affinity.
Resu lts a n d Discu ssion
Constituitively active signal transduction pathways
have been identified in a variety of disease states.
Ligands that are able to disrupt these inappropriately
hyperstimulated pathways, by blocking SH2 domain-
dependent interactions, may ultimately find utility as
therapeutic agents. For example, ligands directed against
the Lck SH2 domain could serve in various capacities,
such as for the treatment of autoimmune diseases and
T cell-based leukemias and lymphomas. The Lck SH2
domain exhibits a marked preference for the sequence
-pTyr-Glu-Glu-Ile-, and short peptides bearing this
sequence exhibit a reasonably high affinity for the SH2
domain of Lck.3 Dissociation constants for the Src family
SH2 ligands, Ac-pTyr-Glu-Glu-Ile and Ac-pTyr-Glu-Glu-
Ile-amide, have been reported to be in the range of
The discovery of the role of SH2 domains in organiz-
ing coherent signaling pathways created an obvious
target for the design of potential therapeutic agents.8
An early report that small peptide ligands of SH2
domains exhibit dissociation constants that approach
the picomolar range provided additional incentive to
explore the possibility of creating nonpeptidic ana-
logues.9 Unfortunately, these KD values overestimated
true affinity by 2-3 orders of magnitude.10 Neverthe-
less, the acquisition of nonpeptidic SH2 ligands has been
abetted by several key recent contributions. First, a
potent nonhydrolyzable pTyr mimetic not only has been
described5 but also has been rendered membrane per-
meable.11 Second, SH2 domains that belong to different
protein families exhibit significantly different amino
acid sequence specificities, an observation that could
lead to the eventual acquisition of nonpeptidic analogues
* To whom correspondence should be addressed. Tel: (718)430-8641.
Fax (718)430-8565. E-mail: dlawrenc@aecom.yu.edu.
10.1021/jm990462r CCC: $19.00 © 2000 American Chemical Society
Published on Web 02/29/2000

1174 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 6
Lee and Lawrence
Ta ble 1. Diamines (A-G) and Carboxylic Acids (7-18) Used in the Preparation of the 84-Member Library Outlined in Scheme 1
displaying analogous specificities.3 Finally, nonpeptide/
peptide conjugates have recently been reported whose
SH2 affinities dramatically exceed those displayed by
their peptidic counterparts.4a These relatively recent
innovations suggest that there is good cause for opti-
mism in using SH2 domains as possible therapeutic
targets. Given these facts, we set out to determine
whether the three C-terminal amino acids in the pep-
tide-based species Ac-pTyr-Glu-Glu-Ile-amide can be
replaced without loss of affinity for the Lck SH2 domain.
Our strategy is based on previously reported struc-
tural and biochemical studies on the interaction of
peptide ligands with the SH2 domains of the Src family
of protein kinases, including the Src family member
Lck.1 The 3-dimensional structure of Ac-pTyr-Glu-Glu-
Ile bound to the Lck SH2 domain has been solved.12 Two
residues on the peptide ligand are engaged in key
interactions with the SH2 domain. This includes the
pTyr moiety, which is embedded within a cavity on the
SH2 surface. The negatively charged phosphate group
participates in an electrostatic interaction with an Arg
moiety. The other key peptide residue is the P+4 Ile,
whose hydrophobic side chain is buried within a second
cavity located on the SH2 surface. The double-cavity
motif has been described as a “two-holed socket” that
engages the “two-pronged plug” (i.e. pTyr and Ile) of the
peptide-based ligand.1b In contrast, the two bridging
glu ta m ic a cid residues of the Ac-pTyr-Glu -Glu -Ile-
amide ligand do not appear to be strongly bound.13 With
these features in mind, we developed a two-step strategy
to identify analogues of the SH2 tetrapeptide ligand by
posing the following questions. Can the Glu-Glu dyad,
which serves to link the pTyr and Ile prongs of the “two-
pronged plug”, be replaced by a simple non-amino acid
moiety that functions in an analogous capacity? Can the
C-terminal Ile “prong”, which is embedded within a
lipophilic cavity of the SH2 domain, be replaced with a
non-amino acid analogue? We addressed these ques-
tions, in a stepwise fashion, by first preparing an 84-
member library to identify a Glu-Glu dyad surrogate.
We subsequently synthesized a 900-member library to
acquire non-amino acid mimetics for the P+4 Ile moiety.
We examined the ability of seven different diamines
(A-G) to serve as Glu-Glu dyad replacements in the
context of twelve (7-18) different aliphatic and aromatic
acyl moieties (i.e. Ile replacements) (Table 1). The 84
members of this initial library have the general struc-
ture 6 (Scheme 1). The seven diamine tethers (A-G)
were chosen using the known 3-dimensional structure
of the Lck SH2/Ac-pTyr-Glu-Glu-Ile complex as a guide.12
Molecular modeling studies suggested that diamines of
the general structure H2N(CH2)nNH2 (where n ) 2-5)
would be ideal as Glu-Glu analogues, depending upon
the nature of the appended acyl moiety. Consequently,
A-C were examined as possible Glu-Glu surrogates.
Modeling also suggested that hydrophobic substituents
on the diamine chain could augment SH2 affinity.
Therefore, we included D and E as part of this prelimi-
nary evaluation of Glu-Glu dyad replacements. Finally,
substituents that could hydrogen bond with residues on
the SH2 domain might also promote SH2 affinity. We
employed F and G to address the latter issue.
The disulfide-containing Tentagel-based resin 2 was
prepared via modification of commercially available
Tentagel resin (1) with 3,3′-dithiodipropionic acid in the
presence of benzotriazol-1-yloxytris(dimethylamino)-
phosphonium hexafluorophosphate (BOP), hydroxy-
benzotriazole (HOBt), and N-methymorpholine (NMM)
in DMF (Scheme 1). The Kaiser ninhydrin test was used
to monitor completion of the reaction. The free carboxyl-
ate moiety in 2 was activated with O-(N-succinimidyl)-
1,1,3,3-tetramethyluronium tetrafluoroborate (TSTU)

SH2-Directed Ligands of Lck TK
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 6 1175
Sch em e 1. Preparation of the 84-Member Library 6a
a
(a) Tentagel resin (1), dithiodipropionic acid, BOP, HOBt, NMM, DMF, 2 h at rt; (b) TSTU, NMM, pTyr, DMF/H2O; (c) BOP, HOBt,
NMM, CH2Cl2/DMF, Boc-protected diamine (A-G), 2 h at rt; (d) i. TFA/CH2Cl2, 1 h at rt, ii. BOP, HOBt, NMM, DMF, RCO2H (7-18);
(e) 5 washes of 10 mM dithiothreitol, 50 mM Tris, pH 7.5, 1 h at rt.
and NMM in DMF and subsequently condensed with
pTyr to furnish 3. The free C-terminus of the latter
was activated, split into seven equal portions, and
condensed with the array of mono-Boc-protected di-
amines displayed in Table 1. The resultant products 4
were deprotected with 25% trifluoroacetic acid (TFA) in
CH2Cl2 and then further split into twelve equal portions
and condensed with a variety of representative aliphatic
and aromatic carboxylic acids, to yield the library of 84
nonpeptidic compounds 5 in a 96-well microfiltration
plate format. Each member of this library was individu-
ally treated with five washings (1 h duration/washing)
of a 10 mM dithiothreitol (DTT)-based buffer (pH 7.6,
50 mM Tris). We found that the first two of these five
DTT washings cleave greater than 90% of the disulfide-
appended compounds from the resin. The final three
washings were employed to remove any residual resin-
bound material. Three members of this library were
assessed for purity by HPLC and characterized by mass
spectrometry following removal of the Tris buffer and
DTT by HPLC (see Experimental Section). The resulting
individual solutions of 6 were then vacuum-filtered into
a 96-well receiving plate using a vacuum filter manifold.
The final synthetic step (e) depicted in Scheme 1 also
represents the first step of the subsequent library
assay since cleavage from the resin is effected by
treatment with assay buffer. This furnishes a library
of compounds in an assay-ready form. The highest-
affinity SH2-targeted ligands in this library were iden-
tified via an enzyme-linked immunosorbent assay
(ELISA).14 ELISAs were performed using a 96-well
strepavidin-coated plate that had been pretreated with
the biotinylated SH2-directed ligand, biotinyl-ꢀ-amino-
caproyl-EPQpYEEIPIYL. The GST-SH2 fusion protein
of Lck was introduced and the structural integrity of
the GST-SH2/biotinyl-ꢀ-aminocaproyl-EPQpYEEIPIYL
complex separately challenged with each component
(@ 100 µM) of the 84-member nonpeptide SH2 library.
The presence of GST-SH2 bound to the strepavidin-
coated plate was assessed via the standard sequential
treatment of each well with rabbit anti-GST antibody
followed by horseradish peroxidase-conjugated mouse
anti-rabbit antibody.
In general, all seven of the diamine tethers furnish
effective SH2 ligands (data not shown). Although some
differences in SH2 affinity between the seven different
diamine tethers do exist, this limited survey revealed
that it is the appended acyl moiety, and not the diamine,
that has the most pronounced influence on SH2 affinity.
For example, all of the best SH2 ligands from this
library contain an aromatic acyl group, with the naph-
thalene-containing species 18 exerting the greatest
beneficial influence on SH2 affinity, an effect that is
almost independent of the actual diamine to which it is
appended. The roughly comparable influence that A-G
have on SH2 affinity is not too surprising given the fact
that we employed the Lck SH2 crystal structure to
assist in the selection of the diamines (Table 1). We
decided to use diamine B as the Glu-Glu replacement
in the subsequent 900-member library due to its struc-
tural simplicity and the fact that two of the twelve
diamine B-containing derivatives are among the most
potent SH2 ligands in the 84-member library.
We subsequently focused our efforts on obtaining
a non-amino acid replacement for the Ile moiety of
Ac-pTyr-Glu-Glu-Ile-amide. A secondary library of the
general structure 20 was prepared using a protocol
analogous to that depicted in Scheme 1. Intermediate
19 was separately condensed with 900 different carbox-
ylic acids and then cleaved from the resin with DTT, as
described above, to furnish the library of compounds 20
(Scheme 2). Two lead SH2 ligands were identified from
the ELISA screen, and these were subsequently resyn-
thesized (Scheme 3) as the coumarin derivatives 21 and

1176 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 6
Sch em e 2. Preparation of the 900-Member Library 20a
Lee and Lawrence
Ta ble 2. IC50 and KD Values for the Lck SH2 Domain
Complexed with Ligands 21-24 and
Ac-pTyr-Glu-Glu-Ile-amidea
Sch em e 3. Preparation of Compounds 21-24a
SH2 ligand
IC50 (µM)
KD (µM)
Ac-PYEEI-NH2
0.66 ( 0.02
1.4 ( 0.1
2.4 ( 0.1
53 ( 5
1.3 ( 0.2
21
22
23
24
2.9 ( 1.0
-
-
-
22 ( 4
a
IC50 values were obtained via the ELISA assay, and KD values
were acquired by equilibrium dialysis (see Experimental Section).
these IC50s vary considerably from plate-to-plate and
from day-to-day. Given the uncertainty of these values,
we employed an alternate method to assess absolute
ligand affinity for the SH2 domain. The fluorescence
associated with the coumarin moiety in 21 is not altered
when the ligand coordinates to the SH2 domain. Con-
sequently, we were able to obtain a KD for 21 via
equilibrium dialysis using 10K cutoff Slide-A-Lyzer
cassettes (Table 2).4a The dissociation constant for the
tetrapeptide Ac-pTyr-Glu-Glu-Ile-amide was obtained
via competitive displacement of a coumarin-substituted
peptide ligand. The equilibrium dialysis assay confirms
that 21 exhibits an affinity for the Lck SH2 domain
comparable to that displayed by Ac-pTyr-Glu-Glu-Ile-
amide.
The aromatic substituents identified from the 900-
member library 20 display surprisingly little structural
homology with the P+4 Ile side chain (cf. 21-22). As
noted above, the side chain of Ile is embedded within a
narrow hydrophobic pocket of the Lck SH2 domain. One
possible explanation for the selection of the aromatic
substituents in 21 and 22 is that the P+4 Ile binding
pocket is conformationally flexible and can therefore
accommodate an array of structurally diverse, yet
compatible, moieties. Alternatively, these substituents
may be simply interacting with sites other than the Ile
binding pocket. This particular issue awaits future
resolution.
a
(a) RCO2H, BOP, HOBt, NMM, DMF added to the diamino-
propane trityl resin, overnight at rt; (b) 10% TFA/CH2Cl2; (c)
Ac-pTyr, BOP, HOBt, NMM, DMF, 2 h at rt; (d) TSTU, NMM,
DMF and added to pTyr/H2O; (e) 25, BOP, HOBt, NMM, DMF,
2 h at rt.
22. The coumarin analogue that is appended to the
amine of pTyr in 21 and 22 has been previously shown
to enhance the SH2 affinity of pTyr-Glu-Glu-Ile-amide.4a
In addition to 21 and 22, we also prepared the acetyl-
ated compounds 23 and 24, which allowed us to assess
the influence that the coumarin- and pyrazoline-deriva-
tized benzoyl substituents in 21 have on SH2 affinity.
For comparison, the Ac-pTyr-Glu-Glu-Ile-amide tetra-
peptide was synthesized as well.
IC50 values for the five compounds listed in Table 2
were obtained using the ELISA assay. These values
were acquired (in duplicate) for all five ligands using a
single 96-well plate. Compound 21 is nearly as potent
an Lck SH2 ligand as its tetrapeptide counterpart. The
acridine-containing species 22 is a likewise proficient
SH2 ligand. The coumarin-containing species 21 is a 37-
fold more potent SH2 ligand than the corresponding
acetylated derivative 23. Furthermore, the pyrazoline-
derivatized benzoyl substituent furnishes a 16-fold
enhancement relative to a simple acetyl moiety (24).
Unfortunately, in our hands, the absolute values of
We have identified two species that exhibit affinities
for the Lck SH2 domain that are comparable to that
displayed by the standard peptide ligand for this
protein. Recently, a Boehringer Ingelheim group has

SH2-Directed Ligands of Lck TK
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 6 1177
achieved similar success with dipeptide derivatives.7
Nevertheless, from the therapeutic point of view, it is
clearly desirable to identify compounds that exhibit
significantly greater affinities for SH2 domains than
conventional peptides. Naturally occurring amino acids
are limited to only 20 different side chains. In contrast,
the array of structural motifs available with non-amino
acid substituents is virtually limitless, allowing access
to interactions with a target protein that are simply not
possible using common peptides. The high-throughput
parallel synthesis strategy described herein takes ad-
vantage of the ready commercial availability of the wide
structural array of carboxylic acid derivatives. Analo-
gous studies are now underway to improve SH2 affinity
via the identification of Glu-Glu dyad analogues that
are both structurally less flexible than the diamine
tether B yet able to productively interact with adjacent
subsites on the SH2 domain.
preparative HPLC using three Waters radial compression C18
reversed-phased column modules (25 × 10 cm) connected in
series. A linear gradient of 100% solvent A (0.1% TFA in water)
to 50% solvent A/50% solvent B (0.1% TFA in CH3CN) was
applied over the course of 50 min. Purity was assessed by
HPLC using an analytical C18 column and two different
solvent systems: a 100% A to 50%A/50%B linear gradient over
25 min, and a 100% solution of 20 mM aqueous Na2HPO4 (pH
7.4) to 75% (20 mM aqueous Na2HPO4)/25% CH3CN linear
gradient over 25 min. 1H NMR spectra were obtained on a
Bruker spectrometer (300 MHz) using DMSO-d6 as solvent.
Data are reported as follows: chemical shift (ppm relative to
the solvent), multiplicity (s ) singlet, d ) doublet, t ) triplet,
q ) quartet, qn ) quintet, br ) broad, m ) multiplet), coupling
constant (J , reported to the nearest 0.5 Hz), and integration.
Compounds were further characterized by electrospray mass
spectrometry.
P ep tid e Syn th esis. All peptides, except biotinyl-ꢀ-amino-
caproyl-EPQpYEEIPIYL, were synthesized on an automated
peptide synthesizer (Advanced Chemtech model ACT90) using
a standard Fmoc solid-phase peptide synthesis protocol. Rink
amide and Wang acid resins were employed for the preparation
of peptides containing amide and free acid C-termini, respec-
tively. Crude peptides were purified on a preparative HPLC
using three Waters radial compression modules (25 × 10 cm)
connected in series. Purified peptides were further character-
ized by ESI mass spectrometry.
Con clu sion
SH2 domains have been identified as potential thera-
peutic targets for a variety of medical disorders. Con-
sequently, ligands that are able to interfere with the
biochemical role of SH2 domains, such as peptido-
mimetics of the Lck SH2 domain ligand Ac-pTyr-Glu-
Glu-Ile-amide, are a much sought after commodity. Key
residues of the tetrapeptide include the pTyr and P+4
Ile moieties, whose side chains interact with cavities
that are embedded within the SH2 surface. In contrast,
the Glu-Glu component appears to play a less critical
role in terms of promoting SH2 affinity. We initially
examined the ability of seven structurally simple di-
amines to serve as Glu-Glu dyad mimetics in the context
of an 84-member library. We found that diamines A-G
(Table 1) are all able to furnish effective SH2 ligands.
The diamine B-based 900-member library (20) was
subsequently prepared and assayed using an ELISA
screen. Two lead compounds were identified and resyn-
thesized as the coumarin derivatives 21 and 22. Both
compounds exhibit affinities for the target SH2 protein
that are comparable to that of the tetrapeptide Ac-pTyr-
Glu-Glu-Ile-amide. These results demonstrate that the
three C-terminal amino acid residues of the standard
SH2 ligands can be replaced with non-amino acid
substituents without loss of affinity. Future studies will
focus on the acquisition of SH2-targeted species that
exhibit significantly better affinity for SH2 domains
than conventional peptide ligands.
Syn th esis of th e 84-Mem ber Com bin a tor ia l Liga n d
Libr a r y. 3,3′-Dithiodipropionic acid (15 mmol, 3.15 g) was
added to a mixture of Tentagel S NH2 resin (90 µm, 5 g, 0.3
mmol/g), BOP (30 mmol, 13.2 g), HOBt (30 mmol, 4.59 g), and
NMM (45 mmol, 4.5 g) in 60 mL DMF and subsequently
shaken for 2 h at room temperature (Scheme 1). Completion
of the reaction was monitored using the Kaiser ninhydrin test.
After a series of wash steps (3 × 50 mL DMF, 3 × 50 mL
MeOH, 3 × 50 mL CH2Cl2), the free carboxylate moiety in 2
was activated with TSTU (3 mmol, 903 mg) and NMM (4.5
mmol, 455 mg) in 60 mL DMF for 10 min, washed with DMF
(2 × 50 mL), and then condensed with pTyr (3 mmol, 784 mg)
in NMM (15 mmol, 1.52 g) and 60 mL DMF/water (1:1) to
furnish 3. The resin (3) was divided into 7 equal portions (200
mg each) and activated with BOP (0.15 mmol, 67 mg), HOBt
(0.15, 23 mg) and NMM (1 mmol, 101 mg) in CH2Cl2/DMF -
(1:1) and condensed with the array of mono-Boc-protected
diamines (displayed in Table 1) for 2 h at room temperature.
After a series of wash steps (3 × 10 mL DMF, 3 × 10 mL
MeOH, 3 × 10 mL CH2Cl2), the Boc protecting group was
removed via treatment with 10 mL TFA/CH2Cl2 (1:3) for 1 h.
The products were extensively washed as described above and
subsequently dried in vacuum. The substitution level for the
free amine, determined via the Kaiser ninhydrin test, was 0.05
mmol/g. The 7 free amine-containing resins were further split
into 12 equal portions (4-mg quantities into each well of
solvent-resistant 96-well filter plates) and condensed with 12
representative alipatic and aromatic acids (40 µmol) using BOP
(20 µmol), HOBt (20 µmol), and NMM (100 µmol) in 100 µL
DMF. The plates were shaken overnight and then each well
was subjected to a series of wash steps (3 × 200 µL DMF, 3 ×
200 µL water, 3 × 200 µL DMF, 3 × 200 µL CH2Cl2, 200 µL
10% NMM in CH2Cl2, 2 × 200 µL 50 mM Tris, pH 7.5) using
a 96-well filter plate vacuum manifold. The products were
cleaved from the disulfide-containing resin with 10 mM DTT
in Tris buffer (5 × 200 µL for 1 h each) and filtered into a
receiving set of 96-well plates using the vacuum manifold (final
volume: 1 mL). The efficiency of acid coupling, product
cleavage from the resin with DTT solution, and purity of the
products were assessed with several ligands (B-18, D-10, and
F -12 from Table 1). No free N-terminus peptide was detected
and over 90% of total ligand was cleaved from resin with first
two DTT wash steps. The final three DTT washings removed
the residual resin-bound material. Compound purity was
greater than 60% as assessed by HPLC and the HPLC-purified
compounds (i.e. removal of Tris buffer and DTT) were char-
acterized by ESI mass spectrometry.
Exp er im en ta l Section
Gen er a l. Chemicals were obtained from Aldrich, except for
piperidine (Advanced Chemtech); protected amino acids, amino
acid derivatives, HOBt, BOP, and TSTU (Advanced Chemtech
and Bachem California); diaminopropane trityl resin (Nova
biochem); and Rink resin, Wang resin, and Tentagel resin
(Advanced Chemtech). Biotinyl-ꢀ-aminocaproyl-EPQpYEEIP-
IYL was purchased from Bachem California. Lck SH2-GST
fusion proteins and polyclonal rabbit anti-GST antibody were
purchased from Santa Cruz Biotechnology. Horseradish per-
oxidase-conjugated mouse anti-rabbit antibody, peroxidase
substrate (1-Step Turbo TMB-ELISA, trimethylbenzidine),
steptavidin-coated 96-well plates, and Slide-A-Lyzer 10K
MWCO dialysis slide cassettes were purchased from Pierce.
Solvent-resistant multiscreen 96-well filter plates and the
multiscreen 96-well filter plate vacuum manifold were pur-
chased from Millipore. All final products were purified via

1178 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 6
Lee and Lawrence
Syn th esis of th e 900-Mem ber Liga n d Libr a r y. The 900-
member ligand library (20) was prepared as described above
except for the following modifications: After 2 was condensed
with pTyr, 3 was activated with BOP (3 mmol, 1.33 g), HOBt
(3 mmol, 459 mg), and NMM (15 mmol, 1.52 g) in CH2Cl2/
DMF (1:1) and condensed with mono-Boc-protected diamino-
propane (7.5 mmol, 1.31 g) for 2 h at room temperature. After
deprotection of Boc group, the free amine-containing resin (19)
was distributed in 4-mg quantities into each well of solvent-
resistant 96-well filter plates. In addition, each well contained
a carboxylic acid-containing compound (40 µmol), BOP (20
µmol), HOBt (20 µmol), and NMM (100 µmol) in 100 µL DMF.
A total of 900 different carboxylic acids were employed.
3H), 7.98 (d, J ) 8.5 Hz, 1H), 7.82 (br, 3H), 7.70 (br, 1H), 6.91
(d, J ) 2 Hz, 1H), 6.67 (d, J ) 2 Hz, 1H), 3.5-3.3 (m, 2H),
2.94 (br, 2H), 1.91 (qn, J ) 7 Hz, 2H); ESIMS m/z 312.2 (MH+).
Syn th esis of N-(7-Hyd r oxycou m a r in -4-a cetyl)-O-p h os-
p h o-L-tyr osin e (26). 7-Hydroxycoumarin-4-acetic acid (10
mmol, 2.2 g), TSTU (10 mmol, 3.0 g), and NMM (30 mmol,
3 g) were added to 3 mL DMF and the mixture was shaken
for 5 min at room temperature and subsequently added to a
solution of O-phospho-L-tyrosine (2 mmol, 522 mg) in 3 mL
water. The mixture was shaken for 2 h at room temperature
and directly purified using preparative HPLC without further
manipulation: yield 840 mg, 91% based on O-phospho-L-
1
tyrosine; H NMR (300 MHz, DMSO-d6) δ 8.66 (d, J ) 8 Hz,
1H), 7.42 (d, J ) 8 Hz, 1H), 7.19 (d, J ) 8.5 Hz, 2H), 7.06 (d,
J ) 8.5 Hz, 2H), 6.72-6.67 (m, 2H), 6.15 (s, 1H), 4.41 (br, 1H),
3.65 (s, 2H), 3.1-3.0 (m, 1H), 2.9-2.7 (m, 1H); ESIMS m/z
464.1 (MH+).
Scr een in g of th e Liga n d Libr a r y. An ELISA assay was
employed to screen the library for SH2 affinity. 100 µL biotinyl-
ꢀ-aminocaproyl-EPQpYEEIPIYL (10 ng/mL in 50 mM Tris, 150
mM NaCl, pH 7.5) was added to each well of streptavidin-
coated 96-well microtiter plates. The plates were shaken
Syn th esis of Com p ou n d s 21 a n d 22. 25 (0.05 mmol) was
added to a mixture of 26 (0.25 mmol, 116 mg), BOP (0.25
mmol, 111 mg), HOBt (0.25 mmol, 38 mg), and NMM (1 mmol,
101 mg) in 1 mL DMF and subsequently shaken for 2 h at
room temperature. The mixture was purified on preparative
HPLC without further manipulation. Compound 21: analyti-
cal HPLC >95% pure; yield 10 mg, 28% based on 25a ; 1H NMR
(300 MHz, DMSO-d6) δ 8.65 (br, 1H), 8.45 (br, 1H), 8.11 (br,
1H), 7.99-7.26 (m, 5H), 7.13 (br, 4H), 6.70 (br, 2H), 6.09 (s,
1H) 5.37 (s, 1H), 4.39 (br, 1H), 3.94-2.72 (m, 8H), 2.11 (s, 3H),
1.62 (br, 2H); ESIMS m/z 720.2 (MH+). Compound 22: ana-
o
overnight at 4 C and rinsed with TBS (50 mM Tris, 150 mM
NaCl, pH 7.5, 2 × 200 µL) and then rinsed with 2 × 200 µL of
a standard BSA-T-TBS solution (0.2% BSA, 0.1% Tween 20,
TBS). A 50-µL solution of the nonpeptidic compounds (200 µM,
in BSA-T-TBS) from the library and a 50-µL solution of the
Lck SH2-GST fusion protein (6.4 ng/mL, in BSA-T-TBS) were
added to each well of a 96-well plate and the plate was then
shaken for 1 h at room temperature. The solutions were
removed and each well was rinsed with 4 × 200 µL BSA-T-
TBS. 100 µL polyclonal rabbit anti-GST antibody (100 ng/mL
in BSA-T-TBS) was then added to each well and incubated
for 1 h at room temperature. Following subsequent washing
steps with BSA-T-TBS (4 × 200 µL), 100 µL horseradish
peroxidase-conjugated mouse anti-rabbit antibody (200 ng/mL
in BSA-T-TBS) was added to each well and subsequently
incubated for 1 h at room temperature. After a series of final
wash steps (4 × 200 µL BSA-T-TBS, 2 × 300 µL TBS), 100 µL
peroxidase substrate (1-Step Turbo TMB-ELISA, trimethyl-
benzidine) was added to each well and incubated for 5-15 min.
100 µL 1 M sulfuric acid solution was introduced to stop the
peroxidase reaction and the absorbance was measured at 450
nm with a plate reader.
Deter m in a tion of IC50 Va lu es. IC50 values were deter-
mined using the ELISA screening method described above
except that, instead of a fixed 100 µM concentration for each
library member, a 200-fold range of concentrations were
employed around the apparent IC50. Since the IC50 values
determined in this way appear to be extraordinarily sensitive
to even subtle changes in experimental conditions, the IC50
values for all compounds were determined at the same time
using a single 96-well plate.
1
lytical HPLC >90% pure; yield 14 mg, 37% based on 25b; H
NMR (300 MHz, DMSO-d6) δ 8.71 (br, 2H), 8.20 (br, 1H), 8.02-
7.30 (m, 6H), 7.20-7.10 (m, 4H), 6.88 (s, 1H), 6.70-6.63 (m,
3H), 6.08 (s, 1H), 4.43 (br, 1H), 3.94-2.64 (m, 8H), 1.74 (br,
2H); ESIMS m/z 757.2 (MH+).
Syn th esis of Com p ou n d 23. 25a (0.07 mmol, 20 mg) was
added to a mixture of N-acetyl-O-phospho-L-tyrosine (0.35
mmol, 106 mg), BOP (0.35 mmol, 155 mg), HOBt (0.35 mmol,
54 mg), and NMM (1 mmol, 101 mg) in 1 mL DMF and
subsequently shaken for 2 h. The mixture was purified on
preparative HPLC without further manipulation and analyzed
via analytical HPLC (>90% pure): yield 25 mg, 61% based on
1
25a ; H NMR (300 MHz, DMSO-d6) δ 8.41 (br, 1H), 8.10 (d,
J ) 8.5 Hz, 1H), 8.00 (br, 1H), 7.89-7.79 (m, 4H), 7.19-7.02
(m 4H), 6.52 (br, 3H), 5.36 (s, 1H), 4.37 (br, 1H), 3.72-3.09
(m, 4H), 2.95-2.67 (m, 2H), 2.11 (s, 3H), 1.77 (s, 3H), 1.60 (br,
2H); ESIMS m/z 560.0 (MH+).
Syn th esis of Com p ou n d 24. 3-Acetamido-1-propylamine
(0.1 mmol, 11 mg) was added to a mixture of 26 (0.5 mmol,
232 mg), BOP (0.5 mmol, 221 mg), HOBt (0.5 mmol, 77 mg),
and NMM (2 mmol, 202 mg) in 1 mL DMF and subsequently
shaken for 2 h. The mixture was purified on preparative HPLC
without further manipulation and analyzed on analytical
HPLC (>95% pure): yield 27 mg, 48% based on 3-acetamido-
Det er m in a t ion of KD Va lu es: E q u ilib r iu m Dia lysis
Meth od . The KD values were determined using a previously
described equilibrium dialysis assay,4a except that the con-
centrations employed for Lck GST-SH2 fusion protein and
compound 21 were 2 µM each. The final volume in the Slide-
A-Lyzer dialysis slide cassettes was 200 µL.
1
1-propylamine; H NMR (300 MHz, DMSO-d6) δ 8.65 (d, J )
7.5 Hz, 1H), 8.06 (br, 1H), 7.80 (br, 1H), 7.18 (d, J ) 8 Hz,
2H), 7.06 (d, J ) 8 Hz, 2H), 6.66-6.52 (m, 3H), 6.08 (s, 1H),
4.41 (br, 1H), 3.72-2.91 (m, 5H), 2.77-2.68 (m, 1H), 1.78 (s,
3H), 1.48 (br, 2H); ESIMS m/z 561.8 (MH+).
Syn th esis of NH2-(CH2)3-NHCOR (25). A mixture of
RCOOH (2 mmol), BOP (2 mmol, 885 mg), HOBt (2 mmol, 306
mg), and NMM (6 mmol, 607 mg) in 10 mL DMF was added
to a diaminopropane trityl resin (0.5 g, 0.7 mmol/g) and shaken
overnight at room temperature. After a series of wash steps
(3 × 10 mL DMF, 3 × 10 mL MeOH, 3 × 10 mL CH2Cl2), the
compounds were cleaved from the resin with 10% TFA in CH2-
Cl2 (3 × 10 mL × 20 min). After evaporation of solvent, crude
compounds were purified via preparative HPLC and subse-
quently found to be over 98% pure based on two different
analytical HPLC solvent systems. Yields were over 70% based
on substitution level of diaminopropane trityl resin. N-(3-
Aminopropyl)-4-(3-methyl-5-oxo-2-pyrazolin-1-yl)benzamide
(25a ): 1H NMR (300 MHz, DMSO-d6) δ 8.59 (t, J ) 5.5 Hz,
1H), 7.89 (d, J ) 8.5 Hz, 2H), 7.83 (d, J ) 8.5 Hz, 2H), 7.69
(br, 3H), 5.37 (s, 1H), 3.5-3.3 (m, 2H), 2.83 (br, 2H), 2.11 (s,
3H), 1.79 (qn, J ) 7 Hz, 2H); ESIMS m/z 275.2 (MH+). N-(3-
Aminopropyl)-1,3-dihydroxy-9-acridinecarboxamide (25b): 1H
NMR (300 MHz, DMSO-d6) δ 8.85 (t, J ) 5.5 Hz, 1H), 8.06 (s,
Ack n ow led gm en t. We thank the National Insti-
tutes of Health for generous financial support.
Refer en ces
(1) (a) Mayer, B. J .; Gupta, R. Functions of SH2 and SH3 Domains.
Curr. Top. Microbiol. Immunol. 1998, 222, 1-22. (b) Kuriyan,
J .; Cowburn, D. Modular Peptide Recognition Domains in
Eukaryotic Signaling. Annu. Rev. Biophys. Biomol. Struct. 1997,
26, 259-88. (c) Schaffhausen B. SH2 domain structure and
function. Biochim. Biophys. Acta 1995, 1242, 61-75.
(2) Lewis, L. A.; Chung, C. D.; Chen, J .-y.; Parnes, J . R.; Moran,
M.; Patel, V. P.; Miceli, M. C. The Lck SH2 Phosphotyrosine
Binding Site is Critical for Efficient TCR-Induced Processive
Tyrosine Phosphorylation of the Zeta-Chain and IL-2 Production.
J . Immunol. 1997, 159, 2292-300 and references therein.
(3) (a) Songyang, Z.; Shoelson, S. E.; McGlade, J .; Olivier, P.;
Pawson, T.; Bustelo, X. R.; Barbacid, M.; Sabe, H.; Hanafusa,
H.; Yi, T.; et al. Specific motifs recognized by the SH2 domains
of Csk, 3BP2, fps/fes, GRB-2, HCP, SHC, Syk, and Vav. Mol.

SH2-Directed Ligands of Lck TK
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 6 1179
Cell. Biol. 1994, 14, 2777-85. (b) Songyang, Z.; Shoelson, S. E.;
Chaudhuri, M.; Gish, G.; Pawson, T.; Haser, W. G.; King, F.;
Roberts, T.; Ratnofsky, S.; Lechleider, R. J .; Neel, B. G.; Birge,
R. B.; Fajardo, J . E.; Chou, M. M.; Hanafusa, H.; Schaffhausen,
B.; Cantley, L. C. SH2 Domains Recognize Specific Phospho-
peptide Sequences. Cell 1993, 72, 767-78.
J .; Sharma, R.; Tong, L.; Moss, N. Ligands for the Tyrosine
Kinase p56lck SH2 Domain: Discovery of Potent Dipeptide
Derivatives with Monocharged, Nonhydrolyzable Phosphate
Replacements. J . Med. Chem. 1999, 42, 1757-66. (b) Llinas-
Brunet, M.; Beaulieu, P. L.; Cameron, D. R.; Ferland, J . M.;
Gauthier, J .; Ghiro, E.; Gillard, J .; Gorys, V.; Poirier, M.;
Rancourt, J .; Wernic, D.; Betageri, R.; Cardozo, M.; J akes, S.;
Lukas, S.; Patel, U.; Proudfoot, J .; Moss, N. Phosphotyrosine-
Containing Dipeptides as High-Affinity Ligands for the p56lck
SH2 Domain. J . Med. Chem. 1999, 42, 722-9.
(4) (a) Lee, T. R.; Lawrence, D. S. Aquisition of High-Affinity, SH2-
Targeted Ligands via a Spatially Focused Library. J . Med. Chem.
1999, 42, 784-7. (b) Lunney, E. A.; Para, K. S.; Rubin, J . R.;
Humblet, C.; Fergus, J . H.; Marks, J . S.; Sawyer, T. K. Structure-
Based Design of a Novel Series of Nonpeptide Ligands that Bind
to the pp60src SH2 Domain. J . Am. Chem. Soc. 1997, 119, 12471-
6. (c) Morelock, M. M.; Ingraham, R. H.; Betageri, R.; J akes, S.
Determination of receptor-ligand kinetic and equilibrium bind-
ing constants using surface plasmon resonance: application to
the lck SH2 domain and phosphotyrosyl peptides. J . Med. Chem.
1995, 38, 1309-18.
(5) (a) Burke, T. R., J r.; Smyth, M. S.; Otaka, A.; Nomizu, M.; Roller,
P. P.; Wolf, G.; Case, R.; Shoelson, S. E. Nonhydrolyzable
Phosphotyrosyl Mimetics for the Preparation of Phosphatase-
Resistant SH2 Domain Inhibitors. Biochemistry 1994, 33, 6490-
4. (b) Wange, R. L.; Isakov, N.; Burke, T. R., J r.; Otaka, A.;
Roller, P. P.; Watts, J . D.; Aebersold, R.; Samelson, L. E.
F2(Pmp)2-TAM zeta 3, A Novel Competitive Inhibitor of the
Binding of ZAP-70 to the T Cell Antigen Receptor, Blocks Early
T cell Signaling. J . Biol. Chem. 1995, 270, 944-8.
(6) Nonpeptidic SH2-targeted compounds have been described, but
the affinity of these species for SH2 domains is typically 1-3
orders of magnitude less than that displayed by their peptidic
counterparts. For an exception see ref 5b: (a) Burke, T. R., J r.;
Barchi, J . J ., J r.; George, C.; Wolf, G.; Shoelson, S. E.; Yan, X.
Conformationally Constrained Phosphotyrosyl Mimetics De-
signed as Monomeric Src Homology 2 Domain Inhibitors. J . Med.
Chem. 1995, 38, 1386-96. (b) 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, 3719-25. (c) 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 Ligands
for the pp60src SH2 Domain. Bioorg. Med. Chem. 1997, 5, 41-7.
(7) A series of dipeptide ligands for the Lck SH2 domain have
recently been described that exhibit KD values comparable
to that of Ac-pTyr-Glu-Glu-Ile: (a) Beaulieu, P. L.; Cameron,
D. R.; Ferland, J . M.; Gauthier, J .; Ghiro, E.; Gillard, J .; Gorys,
V.; Poirier, M.; Rancourt, J .; Wernic, D.; Llinas-Brunet, M.;
Betageri, R.; Cardozo, M.; Hickey, E. R.; Ingraham, R.; J akes,
S.; Kabcenell, A.; Kirrane, T.; Lukas, S.; Patel, U.; Proudfoot,
(8) Smithgall, T. E. SH2 and SH3 domains: Potential Targets for
Anti-Cancer Drug Design. J . Pharmacol. Toxicol. Methods 1995,
34, 125-32.
(9) Payne, G.; Shoelson, S. E.; Gish, G. D.; Pawson, T.; Walsh, C.
T. Kinetics of p56lck and p60src Src Homology 2 Domain Binding
to Tyrosine-Phosphorylated Peptides Determined by a Competi-
tion Assay or Surface Plasmon Resonance. Proc. Natl. Acad. Sci.
U.S.A. 1993, 90, 4902-6.
(10) Ladbury, J . E.; Lemmon, M. A.; Zhou, M.; Green, J .; Botfield,
M. C.; Schlessinger, J . Measurement of the Binding of Tyrosyl
Phosphopeptides to SH2 domains: A Reappraisal. Proc. Natl.
Acad. Sci. U.S.A. 1995, 92, 3199-203.
(11) Stankovic, C. J .; Surendran, N.; Lunney, E.; Plummer, M. S.;
Para, K. S.; Shahripour, A.; Humblet, C.; Fergus, J . H.; Marks,
J . S.; Herrera, R.; Hubbell, S. E.; Saltiel, A. R.; Stewart, B. H.;
Sawyer, T. K. The Role of 4-Phosphonodifluoromethyl- and
4-phosphonophenylalanine in the Selectivity and Cellular
Uptake of SH2 Domain Ligands. Bioorg. Med. Chem. Lett. 1997,
7, 1909-14.
(12) Tong, L.; Warren, T. C.; King, J .; Betageri, R.; Rose, J .; J akes,
S. Crystal Structures Of The Human p56Lck SH2 Domain In
Complex With Two Short Phosphotyrosyl Peptides At 1.0 Å And
1.8 Å Resolution. J . Mol. Biol. 1996, 256, 601-10.
(13) A recent study that assessed the thermodynamic binding
parameters of various pTyr-Xaa-Xaa-Xaa tetrapeptides with the
SH2 domain of c-Src suggests that the Glu residue at +2 may
play a role in promoting a high-affinity interaction with the Src
SH2 domain; see: Bradshaw, J . M.; Waksman, G. Calorimetric
Examination of High-Affinity Src SH2 Domain-Tyrosyl Phos-
phopeptide Binding: Dissection of the Phosphopeptide Sequence
Specificity and Coupling Energetics. Biochemistry 1999, 38,
3147-54.
(14) Gay, B.; Furet, P.; Garcia-Echeverria, C.; Rahuel, J .; Chene, P.;
Fretz, H.; Schoepfer, J .; Caravatti, G. Dual Specificity Of Src
Homology 2 Domains For Phosphotyrosine Peptide Ligands.
Biochemistry 1997, 36, 5712-18.
J M990462R