Novel synthesis and structural characterization of a high-affinity paramagnetic kinase probe for the identification of non-ATP site binders by nuclear magnetic resonance

Article abstract of DOI:10.1021/jm901525b

To aid in the pursuit of selective kinase inhibitors, we have developed a unique ATP site binder tool for the detection of binders outside the ATP site by nuclear magnetic resonance (NMR).We report here the novel synthesis that led to this paramagnetic spin-labeled pyrazolopyrimidine probe (1), which exhibits nanomolar inhibitory activity against multiple kinases. We demonstrate the application of this probe by performing NMR binding experiments with Lck and Src kinases and utilize it to detect the binding of two compounds proximal to the ATP site. The complex structure of the probe with Lck is also presented, revealing how the probe fits in the ATP site and the specific interactions it has with the protein. We believe that this spinlabeled probe is a valuable tool that holds broad applicability in a screen for non-ATP site binders. 2009 American Chemical Society.

Full text of DOI:10.1021/jm901525b

1238 J. Med. Chem. 2010, 53, 1238–1249
DOI: 10.1021/jm901525b
Novel Synthesis and Structural Characterization of a High-Affinity Paramagnetic Kinase Probe for the
Identification of Non-ATP Site Binders by Nuclear Magnetic Resonance
Franklin J. Moy,† Arthur Lee,‡ Lori Krim Gavrin,‡ Zhang Bao Xu,† Annette Sievers,† Elizabeth Kieras,† Wayne Stochaj,†
†
Lidia Mosyak, John McKew, and Desiree H. H. Tsao*
‡
,†
ꢀ
ꢀ
†Structural Biology and Computational Chemistry and ‡Chemical Sciences, Wyeth Research, 200 CambridgePark Drive, Cambridge,
Massachusetts 02140
Received October 14, 2009
To aid in the pursuit of selective kinase inhibitors, we have developed a unique ATP site binder tool for the
detection of binders outside the ATP site by nuclear magnetic resonance (NMR). We report here the novel
synthesis that led to this paramagnetic spin-labeled pyrazolopyrimidine probe (1), which exhibits nanomolar
inhibitory activity against multiple kinases. We demonstrate the application of this probe by performing
NMR binding experiments with Lck and Src kinases and utilize it to detect the binding of two compounds
proximal to the ATP site. The complex structure of the probe with Lck is also presented, revealing how
the probe fits in the ATP site and the specific interactions it has with the protein. We believe that this spin-
labeled probe is a valuable tool that holds broad applicability in a screen for non-ATP site binders.
Introduction
competitive scaffolds have been disclosed. To develop com-
pounds with better selectivity among kinases, inhibitors that
bind outside the ATP site show great promise and are
currently being explored by many groups. Inhibitors that
can target kinases in sites nearby the substrate binding loca-
tions have the potential to exploit the features that are unique
to each kinase. For example, 4-[(4-methylpiperazin-1-yl)-
methyl]-N-{4-methyl-3-[(4-pyridin-3-ylpyrimidin-2-yl)amino]-
phenyl]benzamide, also known as Imatinib, binds to Abl, a
platelet-derived growth factor receptor and stem cell factor
receptor, and has been used to successfully treatcancer.5,6 This
drug binds in the substrate pocket and further extends into the
ATP site.2 Structural work further reveals that it stabilizes an
inactive conformation of Abl kinase,7 and this binding mode has
been observed in other known kinase inhibitors, such as 4-{4-[3-(4-
chloro-3-trifluoromethylphenyl)ureido]phenoxy}pyridine-2-car-
boxylic acid methylamide (Sorafenib, BAY-439006) bind-
ing to B-Raf8 and 1-(5-tert-butyl-2-p-tolyl-2H-pyrazol-3-yl)-
3-[4-(2-morpholin-4-ylethoxy)naphthalen-1-yl]urea (BIRB796)
to p38.9 These inhibitors are called type II, as they move the
activation loop (DFG loop) to an “out” conformation, creating
a pocket next to the ATP site. Type I inhibitors, on the other
hand, bind in the ATP site with the DFG loop in the “in”
conformation. Occupation of the pocket next to ATP induced
by type II inhibitors provides another possibility for kinase
selectivity, as this region is much less conserved than the ATP
binding region. The success of type II inhibitors has broadened
the kinase inhibitor field tremendously. Non-ATP site (NAS)a
Kinases belong to a large family of proteins known to
regulate many cellular processes that are implicated in patho-
logies encompassing cancer and autoimmune, metabolic,
inflammatory, neurological, and infectious diseases. They
play important roles in signal transduction and coordination
of complex cell cycle function and apoptosis, transferring a
phosphate group from ATP to the hydroxyl group of a
substrate, which includes, but is not limited to, serine, threo-
nine, or tyrosine residues. Thus, regulation of kinase activity
has the potential to treat many disorders. A large number of
kinase inhibitor programs are currently active in the pharma-
ceutical industry, targeting a wide range of the more than 500
known human kinases. To date, this effort has led to 11
approved kinase inhibitors that target the highly conserved
ATP site for oncology indications.1,2 Additionally, many kinase
inhibitors are in clinical trials, including the first non-oncology
kinase inhibitor, targeting Jak-3, 3-{4-methyl-3-[methyl-(7H-
pyrrolo[2,3-d]pyrimidin-4-yl)amino]piperidin-1-yl}-3-oxo-
propionitrile, CP-690550,3 which is currently in phase II/III.
The catalytic domain of kinases shows a high degree of
sequence homology, especially for kinases that belong to the
same family. They share a common ATP binding site with a
conserved activation loop and similar three-dimensional
structure.4 Consequently, a major challenge in kinase research
exists in achieving selectivity among the >500 family mem-
bers, since they all process the same substrate. In addition to
requiring selectivity against other kinases and ATP-binding
proteins, these ATP site inhibitors must also bind tightly to
overcome the high physiological concentration of ATP in the
cell. Currently, the development of novel ATP site inhibitors
is also becoming increasingly challenging, as many ATP
a Abbreviations: Abl, Abelson tyrosine kinase; AMP-PNP, adenylyl-
imidodiphosphate; B-Raf, V-raf murine sarcoma viral oncogene homo-
logue B1; Jnk, C-Jun protein kinase; Jak-3, Janus kinase 3; LANCE,
homogeneous time-resolved fluorescence quenching assay; Lck, lym-
phocyte-specific protein tyrosine kinase; KD, kinase domain; Mek-1,
meiosis-specific serine/threonine protein kinase; NAS, non-ATP site;
PRE, paramagnetic relaxation enhancement; TEMPO, 2,2,6,6-tetra-
methylpiperidine 1-oxyl; Src, proto-oncogenic tyrosine kinase; STD,
saturation transfer difference; RT, room temperature.
*To whom correspondence should be addressed: 200 CambridgePark
Dr., Cambridge, MA 02140. Phone: (617) 665-5667. Fax: (617) 665-
5682. E-mail: dtsao@wyeth.com.
r
pubs.acs.org/jmc
Published on Web 12/28/2009
2009 American Chemical Society

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3 1239
Chart 1a
a1: Paramagnetic kinase inhibitor probe; 2: 1 in reduced state; 3: KX-01; 4: N-phenylanthranilic acid.
binders have the potential to be exploited in conjunction with
common ATP binding motifs to create more potent and
selective drugs.
We present here the design, synthesis, application, and
potential implications of a nanomolar-affinity, paramagnetic
kinase inhibitor [1, reduced form, 2 (Chart 1)] as a tool to aid
in the discovery of new NAS binders using nuclear magnetic
resonance (NMR) spectroscopy. NMR is a well-known tech-
nique used to characterize intermolecular interactions and has
been widely used in the drug discovery process for finding hits,
validatingleads, and subsequentoptimization.10 Anattractive
feature of NMR is that it is a sensitive method that can detect
weak interactions up to the millimolar range.11 It is a method
uniquely suited to screening and detecting NAS binders,
where the main site of a target is occupied by a ligand during
a screen.12 NAS binders have been identified using paramag-
netic probes, such as organic nitroxide radicals coupled
to known ATP site binders.10,13-15 Groups such as 2,2,6,6-
tetramethylpiperidine-1-oxyl (TEMPO) radicals are widely
used as spin probes due to their electron paramagnetic
resonance, slow electronic relaxation time, and prolonged
radical stablility.16 The unpaired electron exerts a dramatic
relaxation effect on the neighboring protons as the electron
has a much higher gyromagnetic ratio (658 times) compared
to that of a proton,17 the effects of which are easily detected by
NMR relaxation experiments. Protons from the protein and
Figure 1
binders, they bind to their kinase targets with relatively low
affinity and have not been broadly profiled. For weak binding
probes, additional control experiments are necessary to vali-
date that the probe binds exclusively at the ATP-binding site.
If nonspecific binding is observed, false positive results will be
obtained.
Here, we discuss the advantages and utility of our low-
nanomolar affinity TEMPO-containing probe 1 and its ap-
plication to the kinase domains (KDs) of Lck and Src kinases
by NMR to identify NAS binders. Using 1, we confirmed the
binding of a knownnon-ATPcompetitive selective Srcbinder,
N-benzyl-2-{5-[4-(2-morpholin-4-ylethoxy)phenyl]pyridin-
2-yl}acetamide, also known as KX-0119-21 (3, Chart 1), and
also established that N-phenylanthranilic acid (4, Chart 1)
binds proximal to the ATP site in Lck. In addition, we present
a novel synthetic route used to obtain the previously unre-
ported 3-bromo-1H-pyrazol-5-amine (17). This useful inter-
mediate (17) was central in the synthesis of 1. We will discuss
the synthesis, features, and applications of 1. Probe 1 has the
potential to be used as a screening tool for a variety of kinases
in the identification of NAS binders by NMR. Lastly, the
crystal structure of 1 with Lck KD will also be presented,
revealing how the probe is oriented in the ATP binding site of
Lck and the specific interactions it makes with the protein.
˚
ligands that are within 15-25 A of the paramagnetic center
are broadened, and their signals are easily detected in an
NMR experiment.
Previously reported work13 of McCoy et al. describes
MnATP as a probe, where Mn2þ was chelated to ATP and
this complex is then used to detect binding of a diphenylamine
inhibitor with Mek-1 kinase. This probe can be applied to
nucleotide binding proteins in general and therefore provides
a versatile and especially useful way to find NAS binders to
kinases. However, this probe binds weakly to most proteins
and, as a result, requires an excess to ensure full binding
occupancy. Jahnke and co-workers have also used spin-
labeled ligands as a screening tool for several biological
systems,14,15,17 including a spin-labeled adenine analogue with
TEMPO, developed for screening NAS inhibitors of kinases.
More recently, an indazole analogue containing TEMPO
was synthesized as a probe for NAS screening against Jnk.18
Although these probes are efficient in detecting nearby
Results and Discussion
Rationale for the Design and Synthesis of the Paramagnetic
Probe. Despite the initial success in previously synthesized
paramagnetic probes, it would be of clear interest to obtain
an ATP competitive probe with higher affinity for a broad
range of kinases, which could be used as a screening tool for
NAS binders. Our efforts have identified the pyrazolo[1,5-
R]pyrimidine core (5, Figure 1) as a kinase inhibitor. Via
modification of the head- and/or tailpiece, high affinity was

1240 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3
Moy et al.
achieved against multiple kinases. Our efforts to prepare the
spin-labeled probe required a flexible synthesis of these types
of molecules. Syntheses of pyrazolo[1,5-R]pyrimidines with
aryl substitution at the “tailpiece” and aniline or aryl sub-
stitution at the “headpiece” of the molecule, such as com-
pounds 622 and 7,23 have been previously conducted, as
shown in Figure 1.
Previous efforts in our laboratory have shown that cycli-
zation of various aminopyrazoles with the sodium salt of
ethyl formylacetate provides the pyrazolo[1,5-R]pyrimidine
core, 5.24 In addition, our work also showed that these cores
could be synthesized starting from more readily available
reagents such as R-bromoketones.24 Displacement of the
R-bromo group with cyanide followed by condensation with
hydrazine afforded pyrazole 9. Cyclization with diethyl-
ethoxymethylene malonate and subsequent saponification/
decarboxylation yielded the pyrazolo[1,5-R]pyrimidin-7-
one, 10, in five steps and in 32% overall yield (Scheme 1).
Standard chemistry25 would allow for elaboration on the
pyrimidone core, as shown in Scheme 2, to afford a variety of
compounds with diverse aniline headpieces but with limited
diversity in the tail region, 13 (Scheme 2).
In our quest to synthesize the best probe for NMR studies,
we examined a collection of uniquely substituted pyrazolo-
pyrimidines. However, the substitution of the tailpiece moi-
ety was limited to accessible aminopyrazoles. Ideally, the
starting point would be 3-bromo-1H-pyrazol-5-amine that
has a handle for diverse functionalization. Despite many
recent developments in aminopyrazole chemistry,26 this
compound has not been reported. Therefore, our synthetic
efforts became focused on synthesizing this key intermediate
whereby a convergent and flexible synthesis allowing for late
stage diversity at both the headpiece and the tailpiece of the
pyrazolo[1,5-R]pyrimidines could be achieved. These efforts
resulted in the first report of the synthesis of 3-bromo-1H-
pyrazol-5-amine, 17, obtained from the concomitant debro-
mination and nitro reduction of the previously reported 16,27
shown in Scheme 3. The regiochemistry of the reduction
product was confirmed by spectroscopic comparison of 17
with commercially available 4-bromo-1H-pyrazol-5-amine
(Scheme 3).
Aminopyrazole, 17, was elaborated as in the above-men-
tioned methodology to provide the common intermediate, 20
(Scheme 4), that is uniquely suited for rapid functionaliza-
tion. From this common intermediate, synthesis of a diverse
set of analogues can be realized. After amine installation in
the headpiece region, assorted tailpiece moieties may be
installed using Buchwald, Heck, Sonogashira, or Suzuki
couplings. Since the tailpiece is built in at a late stage of the
synthesis, the choice of tailpiece functionality is not limited.
Additionally, compared to the linear route shown in
Schemes 1 and 2, a route involving intermediate 20 enables
Scheme 1a
a Reagents and conditions: (a) KCN, MeOH, H2O, RT, 0.5 h, 65%;
(b) hydrazine hydrate, EtOH, 100 °C, 2 h, 69%; (c) diethylethoxymethy-
lene malonate, AcOH, reflux, 16 h, 85%; (d) NaOH, EtOH, 100 °C, 3 h,
91%; (e) Dowtherm A, 250 °C, 2 h, 91%.
Scheme 2a
a Reagents and conditions: (a) POCl3, DIPEA, 130 °C, 16 h, 83%; (b) aniline, acetic acid, dioxane, 150 °C, 10 min, microwave, 80-100%.
Scheme 3a
a Reagents and conditions: (a) nitric acid, acetic anhydride, acetic acid; (b) 3,5-dimethyl-1H-pyrazole, toluene, reflux, 20 min; (c) SnCl2 2H2O, 42%
over three steps.
3
Scheme 4a
a Reagents and conditions: (a) diethylethoxymethylene malonate, reflux, 4 h, 77%; (b) NaOH, MeOH, H2O, 100 °C, 2 h, 92%; (c) Dowtherm, 240 °C,
2.5 h, 98%; (d) POCl3, DIPEA, 130 °C, 16 h, 77%.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3 1241
Scheme 5a
a Reagents and conditions: (a) 3-aminophenol, acetic acid, dioxane, 150 °C, 20 min, microwave, 92%; (b) 3-methoxycarbonylphenylboronic
acid, Pd(dppf)Cl2, 2 M K3PO4, dioxane, 90 °C, 1.5 h, 83%; (c) LiOH, THF, H2O, 25 °C, 45 min, 90%; (d) 4-amino-TEMPO, BOP, Et3N, CH2Cl2, 25 °C,
1.5 h, 36%.
Table 1. IC50 Values of 1 in Kinases from the Selectivity Panel
kinase
IC50 (μM)
kinase
IC50 (μM)
ABL1
Aurora B
0.005
2.9
p38alpha (MAPK14)*
MAPKAPK2 (MK2)
PKA (PRKACA)**
PKC-β (PRKCB1)
PRKCA (PKC-R)
CSNK1G1 (CK1 γ1)
MAP4K2 (GCK)
PDGFRA
>50
>50
>50
>60
1.3
CDK1/cyclin B
CDK2/cyclin A
Fyn
>50
>50
0.005
0.016
0.007
0.005
HCK
Lyn A
>50
0.024
0.038
>50
SRC
KDR (VEGFR2)
MAPK1 (ERK2)
0.0033 RPS6KA1 (RSK1)
>50
a wider array of products to be prepared due to the abun-
dance of commercially available boronic acids.
On the basis of previously determined complex structures
of various pyrazolo[1,5-R]pyrimidines with Lck (not shown),
we knew that a large group such as TEMPO could be
accommodated without an effect on the important interac-
tions between the protein-ligand complex. Hence, we
decided to couple 4-amino-TEMPO to a carboxylic acid.
The synthesis of acid, 23, and probe compound, 1, from the
versatile common intermediate, 20, follows standard chem-
istry and is outlined below (Scheme 5).
Figure 2. Titration shows 2 binds tightly to Lck KD. 2 (10 μΜ) with
(A) 0, (B) 2.5, (C) 5, and (D) 10 μΜ Lck. The resonance peak at 8.51
ppm comes from the buffer.
Compound 1 Binds with High Affinity to Multiple Kinases.
This compound was submitted to a Caliper kinase selectivity
panel to evaluate its affinity for different kinases. Table 1
summarizes the results, showing that 1 had an IC50 of
<0.040 μM in eight kinases, most of them belonging to the
tyrosine kinase (TK) branch of the human kinome. The IC50
of 1 with Lck was determined via a LANCE fluorescence
assay, and it was shown to be 0.035 μM. This initial testing
indicates the potential of this inhibitor to have a high affinity
for many kinases, as our panel is comprised of only a small
subset of kinases from the human kinome. Compound 1 was
shown to be competitive with the stable ATP analogue,
AMP-PNP (adenylyl imidodiphosphate), by NMR, indicat-
ing binding in the ATP binding site, subsequently confirmed
by the complex structure of 1 with Lck KD presented here.
NMR Spectroscopy and Detection of NAS Binders. As
reported in Table 1, our probe shows low nanomolar activity
for several kinases. Several NMR experiments were con-
ducted to evaluate 1 and its interaction with two kinases, Lck
and Src kinase. Since 1 itself exhibits broad lines in the NMR
spectrum, ascorbic acid was added to reduce the nitroso to
the hydroxyl amine in situ (2, Chart 1), which resulted in
sharp lines necessary for the protein titration experiment.
The addition of ascorbic acid (5-fold over 1) did not affect the
binding of the test compound, as evidenced by control NMR
experiments. To confirm the high affinity of 2 for Lck KD,
an NMR protein titration was performed, where the line
shapes of 2 in the presence and absence of the kinase were
compared. The signals for 2 were monitored as Lck was
added (Figure 2), and as expected for a tight inhibitor, a
decrease in signal intensity with addition of Lck in a stoichio-
metric fashion was observed. Low nanomolar inhibitors are
in slow exchange with the protein target, and the bound
ligand NMR signals will broaden, since the ligand’s line
width is inversely proportional to its transverse relaxation
rates; in this case, it becomes that of the kinase. Only the free
ligand that remains in solution is observed, displaying sharp
NMR signals,28 and at equimolar amounts of 2 and Lck

1242 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3
Moy et al.
Figure 3. STD spectra of Lck and AMP-PNP in the presence and
absence of 1 showing that 1 competes with AMP-PNP. (A and B)
One-dimensional reference and STD spectra, respectively, of 7.5 μΜ
Lck and 125 μΜ AMP-PNP. (C and D) One-dimensional reference
and STD spectra, respectively, of 7.5 μΜ Lck, 125 μΜ AMP-PNP,
and 10 μΜ 1.
(Figure 2D), the ligand signals are no longer visible, con-
firming that 2, and presumably 1, bind with high affinity to
Lck.
Figure 4. T1F spectra at 10 (blue) and 200 ms (red) of (A) 125 μΜ
AMP-PNP, (B) 125 μΜ AMP-PNP and 7.5 μΜ Lck, (C) 125 μΜ
AMP-PNP, 7.5 μΜ Lck, and 10 μΜ 1, and (D) 125 μΜ AMP-PNP,
7.5 μΜ Lck, and 10 μΜ 2. Comparison of relaxation intensity ratios
of T1F spectra from panels A and B indicates that AMP-PNP binds
to Lck with a reduction of 7-9% units. Relaxation intensity ratio
comparison of T1F spectra between panels C and D shows no PRE of
AMP-PNP. Compared to panel A, this shows that AMP-PNP is free
in solution.
Binding of 1 was also monitored by the saturation transfer
difference (STD) experiment with Lck. In this difference
experiment, small amounts of protein are present with a large
excess of ligand, the protein is saturated on and off resonance,
and the saturation propagates throughout the protein via
spin diffusion.29 If the compound binds to the protein, the
saturation will be transferred to the compound, which upon
exchange back into solution is detected from the difference
spectrum between the off and on resonance spectra. The
subtracted spectrum shows only resonances that have been
saturated, arising mainly from the bound compound.29 We
confirmed that 1 is an ATP site binder, by adding it as a
competitor to the mixture of Lck and AMP-PNP. AMP-
PNP binds to Lck, as shown in Figure 3, giving an STD
signal of 28% (Figure 3B). Addition of 1 at a similar
concentration to Lck completely abolishes binding of
AMP-PNP, confirming that 1 binds in the ATP binding site
and prevents AMP-PNP binding (Figure 3D). Similar results
were also achieved with Src kinase with complete inhibi-
tion of AMP-PNP binding upon addition of 1 (spectra not
shown).
and a long relaxation period of 200 ms (red). In the longer
delay spectrum, the resonance intensity will be reduced
accordingly by binding to the protein and/or enhancement
by 1. The effects of the test compound binding and para-
magnetic relaxation enhancement (PRE) are evaluated in
two ways: (1) by measuring the changes in the relaxation
intensity ratio of the T1F spectra acquired at 10 ms versus 200
ms delays (i.e., comparing red and blue traces in Figure 4)
and (2) by noting changes in spectra acquired at 200 ms
delays (i.e., comparing red and red traces in panels A and B
and panels C and D of Figure 4). In Figures 4-6, the
relaxation intensity ratio is reported as the percentage ratio
of the peak at 200 ms to the peak intensity at 10 ms. Panels A
and B of Figure 4 display the T1F spectra obtained for AMP-
PNP and AMP-PNP with Lck, respectively. As anticipated,
a decrease in the magnitude of the AMP-PNP signal at both
10 and 200 ms is observed in Figure 4B (relative to free AMP-
PNP) due to binding to Lck. Upon addition of 1 and its
reduced form, 2 (panels C and D of Figure 4, respectively),
there is a significant increase in the ratios (panel B vs panels C
and D of Figure 4), indicating that AMP-PNP no longer
binds to Lck and is competed off by 1. This is in complete
agreement with the STD experiments presented in Figure 3.
In addition, as expected, the AMP-PNP resonance relaxa-
tion intensity ratios are unchanged in Figure 4C,D.
The application of this paramagnetic probe was also
demonstrated with the T1F experiment, in which the spin-
lattice relaxation rates in the rotating frame of the ligand in
the presence and absence of protein were compared.30,31 The
free ligand relaxes slowly and upon binding to the protein
acquires characteristics of the large molecule and will then
relax rapidly. The relaxation rate observed will be a weighted
average of the relaxation rates between the free and bound
states for fast exchanging systems. In the presence of a
paramagnetic center, the relaxation rate of the ligand is
further enhanced in the bound state due to the unpaired
electron. Compound 1 enhances the relaxation of nearby
˚
protons significantly (15-25 A radius), both for the protein
We sought a test compound that could be used to validate
1 as a screening tool. Kinex Pharmaceuticals has reported on
3 (Chart 1), an inhibitor of Src that is not competitive with
ATP and is postulated to bind to the substrate binding
and for the potential binders. Figure 4 shows the T1F control
experiment with Lck and AMP-PNP. All T1F experiments
were conducted with a short relaxation period of 10 ms (blue)

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3 1243
Figure 6. T1F partial aromatic spectra at 10 (blue) and 200 ms (red)
of (A) 100 μΜ 4, (B) 100 μΜ 4 and 7.5 μΜ Lck, (C) 100 μΜ 4, 7.5 μΜ
Lck, and 10 μΜ 1, and (D) 100 μΜ 4, 7.5 μΜ Lck, and 10 μΜ 2.
Comparison of relaxation intensity ratios of T1F spectra of panels A
and B shows that 4 binds to Lck giving an 8% unit reduction. The
10% unit reduction in relaxation intensity ratios of T1F spectra of
panels C and D is due to PRE.
Figure 5. T1F partial aromatic spectra at 10 (blue) and 200 ms (red)
of (A) 100 μΜ 3, (B) 100 μΜ 3 and 6 μΜ Src, (C) 100 μΜ 3, 6 μΜ Src,
and 10 μΜ 1, and (D) 100 μΜ 3, 6 μΜ Src, and 10 μΜ 2. Comparison
of relaxation intensity ratios of T1F spectra of panels A and B shows
that 3 binds to Src giving a 8-14% unit reduction. The 5-10% unit
reduction in relaxation intensity ratios of T1F spectra of panels C
and D is due to PRE.
performed, which provides evidence that 4 binds in a site
adjacent to the ATP site. Panels A and B of Figure 6 confirm
the binding of 4 to Lck. Upon addition of 1, there is a
reduction in the relaxation intensity ratio (Figure 6C,D),
similar to the results obtained for 3 and Src with 1, showing
simultaneous binding of 1 and 4 to Lck. In both cases,
experiments displayed in Figures 5 and 6 show that the test
compounds may bind differently to the kinase targets when 2
(and, by inference, 1) is bound to the kinases. This is not
unexpected, since the presence of 1 (and presumably 2) keeps
the kinase in a closed conformation. Control experi-
ments conducted with all three test compounds (3, 4, and
AMP-PNP) and 1 show that their relaxation rate did not
change significantly in the absence of protein.
Structure of 1 with Lck. The ribbon diagram of Lck and
positioning of 1 in the ATP-binding site are shown in Figure
7. The diagram is oriented such that the TEMPO moiety
points toward the bulk solvent, the phenol is buried in a deep
hydrophobic pocket with the hydroxyl group hydrogen-
bonded to the catalytic Lys 273, and the pyrazolopyrimidine
core binds along the hinge making a hydrogen bond to the
NH group of Met 319. The amide between the core and the
probe tail interacts indirectly with hinge Met 319 through a
water-mediated hydrogen bond. There are also a large
number of van der Waals interactions with the Lck ATP-
binding domain that stabilize the conformation of the bound
inhibitor. Figure 7B shows these specific interactions with 1.
The changes observed in T1F relaxation experiments for 3
and 4 are clearly due to binding to the kinases, and PRE is
observed in the presence of the spin-labeled probe. The
amount of PRE observed depends on factors such as the
domain.19-21 It has been reported by Kinex that in cells 3
shows nanomolar inhibition of Src-dependent growth,
whereas in assays that examine the functional activity of the
Src KD, it appears to be weakly potent (IC50 ∼ 40 μM).19-21
The unique binding mode and weak affinity for the Src KD
made 3 an ideal test compound for demonstrating the potential
of 1 to detect NAS binders via NMR relaxation experiments.
We confirmed the activity of 3 to Src KD in house (IC50
>
20 μM) and verified that 3 binds to Src kinase by the STD
experiment and did not compete with either AMP-PNP or
staurosporine, a known high-affinity pan kinase inhibitor
(data not shown). In addition, a titration similar to the one
reported in Figure 2 was performed, and the results were in
agreement with 3 binding weakly to Src KD. Figure 5 shows
the T1F binding results for Src with 3, where a decrease in peak
intensity for the 200 ms delay (Figure 5A, red trace, vs
Figure 5B, red trace) and a reduced relaxation intensity ratio
(change of 8-14 units between panels A and B) are observed,
consistent with 3 binding to Src KD. In addition, we observe
PRE via a lower ratio observed in the peak at 8.74 ppm
(Figure 5C) compared to that in Figure 5D, indicating that 3
and 1 are binding simultaneously to Src.
The utility of 1 was also demonstrated in the characteriza-
tion of 4 (Chart 1) binding to Lck. Compound 4 was chosen
to be tested as a potential binder since it is a common scaffold
in known drugs32 and, due to its simple core, would likely be
a weak binder. We verified that 4 binds to Lck KD by the
STD experiment, and it did not compete with staurosporine
(spectra not shown). Figure 6 displays the T1F experiments

1244 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3
Moy et al.
Figure 7. (A) Ribbon diagram of the catalytic domain of Lck kinase in complex with 1. The DFG loop is in the closed (“in”) conformation. 1
binds in the ATP site, and the nitroxide extends to the solvent. (B) Close-up of the complex, showing specific interactions in the complex.
Several hydrogen bonds contribute to the tight binding of 1 to Lck. Lys 273 and Asp 382 make hydrogen bonds to the oxygen of the phenol
˚
moiety and Met 319, and N4 of pyrazolopyrimidine also forms a hydrogen bond. (C) Electrostatic surface area showing a 20 A radius from the
nitroxide end of 1, colored yellow, to illustrate the range in which binders would be affected by 1. Lck is in the closed conformation, and there is
a cavity next to the probe that could potentially accommodate other binders. Blue represents positively charged residues and red negatively
charged residues.
concentration of the protein, compound 1, and the test
compound, their affinities, and the distance between the test
compound and the paramagnetic center. Under our experi-
mental conditions, Lck (7.5 μΜ) is completely saturated with
1 (10 μΜ) because of its high affinity. We observe relatively
weak paramagnetic effects on 4 when it is bound to Lck in the
presence of 1 (Figure 6), which could be consistent with the
compound being positioned far from the nitroxide end of 1
adjacent to the ATP binding site, proximal to the peptide
binding site, which could accommodate a small molecule
such as 4. The relatively weak affinity of 4 for Lck would also
contribute to a weaker paramagnetic effect. Crystal structure
elucidation of this ternary complex is underway to confirm
the binding site of 4 relative to 1 and its specific interaction
with Lck.
A comparison between the intensity ratios observed for 3
with Src in the presence of 2 (Figure 5D) and 1 (Figure 5C)
shows a weak PRE. This weak effect could be explained by
˚
(up to 25 A). Figure 7C shows the surface of Lck with 1 with a
˚
20 A radius from the nitroxide. There is a deep pocket

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3 1245
the weak interaction of 3 with the KD of Src, as well as 3
possibly binding in the peptide site of Src. Compound 3 was
designed to inhibit Src by binding on the peptide side, which
equipped with an actively shielded 7 T superconducting magnet
and an electrospray ionization (ESI) source. Purity in two
solvent systems (H2O/CH3CN and H2O/MeOH) was deter-
mined using a HPLC system (Agilent) under the following
conditions: UV-DAD; Agilent Zorbax SB-C18 column (3 μm,
2.1 mm ꢀ 30 mm); condition 1, solvent A (H2O with 0.1%
formic acid), solvent B (acetonitrile with 0.1% formic acid),
gradient, from 95% A and 5% B to 0% A and 100% B from 0 to
7 min and to 0% A and 100% B from 7 to 9 min; condition 2,
solvent A (H2O with 0.1% formic acid), solvent B (methanol
with 0.1% formic acid), gradient, from 95% A and 5% B to 0%
A and 100% B from 0 to 7 min and to 0% A and 100% B from 7
to 9 min; flow rate, 0.8 mL/min; temperature, 35 °C. All
˚
would place it ∼20-25 A from the paramagnetic center of 1.
To obtain more specific information about the NASs,
further experiments are required, such as cocrystallization
of the ternary complex or competition-displacement
studies. Theoretical calculations could also be performed
to estimate how far the binders are from the paramagnetic
center. By simulating the T2 relaxation equation for the
protein-ligand system in fast exchange, where at least
10% of the ligand is bound, we could obtain quantitative
1
compounds were determined to be >95%. H and 13C NMR
˚
distances of up to 15 A. This method can also detect NAS
tabulations as well as purity information are available in the text
and Supporting Information.
˚
binders up to 25 A from the paramagnetic center of 1.
˚
However, quantitative measurement beyond 15 A is difficult
3-Bromo-1H-pyrazol-5-amine (17). To a cooled (10 °C) solu-
tion of 3,4,5-tribromo-1H-pyrazole (20 g, 67 mmol) in acetic
acid (300 mL) was added nitric acid (90%, 7 mL). Acetic
anhydride (100 mL) was added over 20 min. After being stirred
for 3 h at room temperature, the mixture was poured over ice
and diluted with water. The white precipitate was washed with
water and dissolved in toluene (250 mL). The organic solution
was washed with water and brine, dried (Na2SO4), and filtered
to yield a toluene solution (200 mL) of 1-nitro-3,4,5-tribromo-
1H-pyrazole in toluene. To the toluene solution was added 3,5-
dimethyl-1H-pyrazole (6.6 g, 67 mmol), and the mixture was
heated to reflux for 20 min and cooled and the solvent removed
in vacuo. The crude mixture was triturated with hexanes to give
a beige solid containing 4,5-dibromo-3-nitro-1H-pyrazole and
other products (23 g). The crude 4,5-dibromo-3-nitro-1H-pyr-
azole was reduced by refluxing with tin(II) chloride dihydrate
(45 g, 200 mmol) in ethyl acetate (200 mL) and ethanol (100 mL).
After 45 min, the reaction mixture was cooled and slowly poured
over a vigorously stirring mixture of sodium bicarbonate (33 g,
400 mmol) in water (200 mL) and ethyl acetate (800 mL). To the
resultant slurry was added Celite (30 g), and this slurry was
filtered through a bed of Celite. The filter cake was washed with
additional ethyl acetate (600 mL). The organic solution was
washed with brine, dried (Na2SO4), and concentrated in vacuo.
The mixture was purified on silica gel eluting with a 3% and then
a 6% ethanol/dichloromethane mixture to yield a tan solid (4.42
(see the Supporting Information for simulations and
calculations).
Since 1 binds specifically and with high affinity to multiple
kinases, it is a robust tool that can be used to screen for NAS
binders. NMR is a very sensitive technique that will detect
weak binders, and via adjustment of the concentration of
ligand and protein used, even millimolar binders could be
detected. Compound 1 has a high solubility in its reduced
form in aqueous buffer of at least 50 μM, as estimated by
NMR; therefore, solubility is not a limiting factor. This high
level of solubility allows one to adjust the ratios of 1 to target
kinases as needed. An NAS binder could be a good starting
point for the development of a type II kinase inhibitor for
which selectivity and potency could be further enhanced by
chemically expanding the initial lead. This starting point
coupled with robust high-throughput crystallography would
allow for rapid structure-based optimization of these leads.
In summary, we report here a novel synthesis, character-
ization, and application of a low nanomolar pyrazolopyr-
imidine paramagnetic kinase inhibitor. This first report of
3-bromo-1H-pyrazole-5-amine allows access to a previously
inaccessible common intermediate, 20. This common inter-
mediate provides entry into a wide variety of analogues
where modifications can be made at either the headpiece or
the tailpiece of the molecule late in the synthesis. This
synthesis led to the discovery of 1, a potent kinase inhibitor
that we were able to use as a spin-labeled probe in NMR
experiments. Our in-house kinase assays show that 1 binds
with low nanomolar affinity to multiple kinases. We demon-
strated by NMR that 1 binds to Lck and Src kinases, and it
easily detects NAS binders in T1F experiments. Furthermore,
we believe that high-affinity compounds such as 1 will have
broad applicability when used to screen for NAS binders.
These leads could be important for kinase inhibitor pro-
grams targeting NAS inhibition to achieve the goal of
enhanced kinase selectivity.
1
g, 42% over three steps): H NMR (400 MHz, DMSO-d6) δ
11.66 (br s, 1H), 4.98-5.48 (m, 3H); 13C NMR (400 MHz,
DMSO-d6) δ 149.61, 125.15, 89.15; HRMS ESIþ m/z 161.9664
[M þ H].
Ethyl 2-Bromo-7-oxo-4,7-dihydropyrazolo[1,5-r]pyrimidine-
6-carboxylate (18). To a solution of 3-bromo-1H-pyrazol-5-
amine (4.14 g, 25.5 mmol) in acetic acid (150 mL) was added
diethylethoxymethylene malonate (5.7 mL, 28 mmol) at room
temperature. The mixture was refluxed for 4 h. The reaction
mixture was then cooled to room temperature, and the acetic
acid was removed in vacuo. The crude solid was suspended in
cold ethanol (100 mL), filtered, and washed with cold ethanol to
1
give 5.60 g (77%) of a pure white solid: H NMR (400 MHz,
DMSO-d6) δ 13.25 (br s, 1H), 8.59 (s, 1H), 6.49 (s, 1H), 4.23 (q,
J = 7.07 Hz, 2H), 1.28 (t, J = 7.07 Hz, 3H); HRMS ESIþ m/z
285.9822 [M þ H].
Experimental Section
2-Bromo-7-oxo-4,7-dihydropyrazolo[1,5-r]pyrimidine-6-car-
boxylic Acid (18b). To ethyl 2-bromo-7-oxo-4,7-dihydropy-
razolo[1,5-R]pyrimidine-6-carboxylate (4.72 g, 18 mmol) in
ethanol (40 mL) was added 2.5 N NaOH (20 mL) at room
temperature. The reaction mixture was heated to 100 °C and
stirred at this temperature for 2 h. The mixture was then cooled
to room temperature and diluted with water (200 mL). Citric
acid (10.5 g, 50 mmol) was added to the mixture and the mixture
stirred for 45 min. The resultant solid was filtered and washed
with water. The wet solid was azeotroped to dryness with
toluene (200 mL) using a Dean-Stark apparatus to give 3.90 g
(92%) of an off-white solid: 1H NMR (400 MHz, DMSO-d6) δ
General Chemistry. All solvents and reagents were used as
obtained. All reactions, except those involving water as a
solvent, were performed under a nitrogen atmosphere. All
reaction mixtures were stirred using a magnetic stir bar. All
reactions were followed by thin-layer chromatography (TLC)
using 250 μm prescored silica gel 60 F254 plates. Flash chromato-
graphy was performed using 230-400 mesh silica gel or flash
columns packed with KP-SIL 60 Angstrom silica gel. Proton
NMR spectra were recorded on a 400 MHz spectrometer in
DMSO-d6 solvent using TMS (δ 0.0) as a reference. High-
resolution mass spectra were recorded using a Fourier trans-
form ion cyclotron resonance (FT-ICR) mass spectrometer

1246 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3
Moy et al.
12.98 (br s, 2H), 8.58 (s, 1H), 6.49 (s, 1H); 13C NMR (400 MHz,
DMSO-d6) δ 165.25, 156.19, 147.92, 145.49, 132.98, 97.45,
95.12; HRMS ESIþ m/z 255.9367 [M - H].
8.84 (m, 1H), 8.39 (dq, J = 1.01, 7.83 Hz, 1H), 8.19 (d, J = 5.31
Hz, 1H), 8.02 (dq, J = 0.97, 7.70 Hz, 1H), 7.50-7.77 (m, 1H),
7.28 (t, J = 7.96 Hz, 1H), 7.11 (s, 1H), 6.81-7.02 (m, 2H),
6.56-6.79 (m, 1H), 6.34 (d, J = 5.05 Hz, 1H); 13C NMR (400
MHz, DMSO-d6) δ 166.05, 158.13, 152.75, 150.52, 150.04,
145.13, 138.14, 133.33, 131.11, 130.18, 130.03, 129.31, 129.18,
126.52, 114.75, 112.84, 111.18, 92.27, 87.38, 52.19; HRMS ESIþ
m/z 361.1296 [M þ H].
2-Bromo-4,7-dihydropyrazolo[1,5-r]pyrimidin-7-one (19). A
mixture of 2-bromo-7-oxo-4,7-dihydropyrazolo[1,5-R]pyrimi-
dine-6-carboxylic acid (3.83 g, 16.4 mmol) in Dowtherm
(60 mL) was heated to 240 °C for 2.5 h. At that point, the
mixture was cooled to room temperature and diluted with
hexanes (200 mL). The tan precipitate was filtered, resuspended
and stirred in hexanes (200 mL), filtered, and washed with
hexanes (200 mL) to give 3.10 g (98%) of an off-white solid.
The product can be further purified by triturating in hot ethanol
to yield the highly pure product (>98%): 1H NMR (400 MHz,
DMSO-d6) δ 12.52 (br s, 1H), 7.87 (d, J = 7.58 Hz, 1H), 6.36 (s,
1H), 5.74 (d, J = 7.58 Hz, 1H); 13C NMR (400 MHz, DMSO-d6)
δ 155.06, 142.92, 139.68, 132.00, 96.08, 91.56; HRMS ESIþ m/z
213.9612 [M þ H].
3-{[2-(3-Carboxyphenyl)pyrazolo[1,5-r]pyrimidin-7-yl]amino}-
phenol (23). To a solution of 3-{[2-(3-methoxycarbonylphenyl)-
pyrazolo[1,5-R]pyrimidin-7-yl]amino}phenol (180 mg, 0.5 mmol)
dissolved in THF (8 mL) was added a 1 N LiOH solution (2 mL).
Additional water (6 mL) was added to make the mixture homo-
geneous, and then the solution was stirred for 45 min at room
temperature. The reaction mixture was diluted with water (20
mL), and then formic acid (139 mg, 0.114 mL) was added. The
precipitate was filtered and washed with water. The solid was
triturated in ethanol and dried to yield 155 mg of a white solid
(90%): 1H NMR (400 MHz, DMSO-d6) δ 13.17 (br s, 1H), 10.15
(br s, 1H), 9.77 (br s, 1H), 8.54-8.91 (m, 1H), 8.37 (dq, J = 1.01,
7.83Hz, 1H), 8.22 (d, J =5.56 Hz, 1H), 7.92-8.10 (m,1H), 7.54-
7.73 (m, 1H), 7.29 (t, J = 8.08 Hz, 1H), 7.13 (s, 1H), 6.83-7.01
(m, 2H), 6.73 (dd, J = 0.76, 2.27 Hz, 1H), 6.33 (d, J = 5.56 Hz,
1H); 13C NMR (400 MHz, DMSO-d6) δ 167.19, 158.26, 153.40,
149.10, 148.91, 145.85, 137.87, 132.93, 131.42, 130.88, 130.17,
129.75, 129.10, 126.94, 115.08, 113.31, 111.55, 91.76, 87.60;
HRMS ESIþ m/z 347.1145 [M þ H].
2-Bromo-7-chloropyrazolo[1,5-r]pyrimidine (20). To 2-bro-
mo-4,7-dihydropyrazolo[1,5-R]pyrimidin-7-one (3.62 g, 16.9
mmol) were added POCl3 (70.0 mL) and N,N-diisopropylethy-
lamine (6.5 mL, 37 mmol) slowly at room temperature. The
reaction mixture was heated to reflux for 16 h. The mixture was
cooled to room temperature, and most of the solvent was
evaporated. The resultant residue was diluted with ethyl acetate
(100 mL) and poured slowly into a vigorously stirring mixture of
solid sodium bicarbonate (50 g), water (100 mL), and ethyl
acetate (300 mL). The solution was checked for basicity (pH 8),
and then the biphasic mixture was filtered through Celite. The
organic layer was washed with brine (200 mL), dried (Na2SO4),
filtered, and concentrated. The crude product was dissolved in
dichloromethane (100 mL) and filtered through a silica plug
with a 20% ethyl acetate/hexane mixture to remove polar
impurities. Concentration followed by trituration in hexanes
gave 3.01 g of a white solid (77%): 1H NMR (400 MHz, DMSO-
d6) δ 8.55 (d, J = 4.55 Hz, 1H), 7.47 (d, J = 4.80 Hz, 1H), 7.14 (s,
1H); 13C NMR (400 MHz, DMSO-d6) δ 150.45, 150.07, 137.41,
134.21, 109.33, 100.29.
3-[(2-Bromopyrazolo[1,5-r]pyrimidin-7-yl)amino]-2-methyl-
phenol (21). To 2-bromo-7-chloropyrazolo[1,5-R]pyrimidine
(2.35 g, 10 mmol) and 3-aminophenol (1.15 g, 10.5 mmol) were
added dioxane (20 mL) and acetic acid (0.9 mL). The reaction
vessel was sealed and was heated in the microwave reactor at
150 °C for 10 min. The mixture was diluted with ethyl acetate
(100 mL) and washed with concentrated aqueous sodium bicar-
bonate (100 mL) and then brine (100 mL), dried with Na2SO4,
filtered, and concentrated. The resultant solid was triturated
with ethanol to give 2.85 g of a tan solid after filtration and
drying (92%): 1H NMR (400 MHz, DMSO-d6) δ 10.03 (s, 1H),
9.69 (s, 1H), 8.18 (d, J = 5.56 Hz, 1H), 7.24 (t, J = 7.96 Hz, 1H),
6.79-6.95 (m, 2H), 6.63-6.74 (m, 2H), 6.32 (d, J = 5.56 Hz,
1H); 13C NMR (400 MHz, DMSO-d6) δ 158.10, 150.69, 150.20,
144.82, 137.87, 132.67, 129.99, 114.72, 112.94, 111.20, 97.12,
87.85; HRMS ESIþ m/z 305.0035 [M þ H].
3-{7-[(3-Hydroxyphenyl)amino]pyrazolo[1,5-r]pyrimidin-2-yl}-
N-(1-oxy-2,2,6,6-tetramethylpiperidin-4-yl)benzamide (1). To a
solution of 3-{[2-(3-carboxyphenyl)pyrazolo[1,5-R]pyrimidin-7-
yl]amino}phenol (87.0 mg, 0.25 mmol) and triethylamine (38 mg,
0.052 mL) in dichloromethane (4 mL) was added (benzotriazol-1-
yloxy)tris(dimethylamino)phosphonium hexafluorophosphate
(133 mg, 0.30 mmol), and the mixture was stirred at room
temperature for 15 min. A solution of 4-amino-TEMPO (51 mg,
0.30 mmol) in dichloromethane (1 mL) was added. After the
mixture had been stirred for 90 min, the solvent was removed.
The crude material was purified on reverse phase (20 to 70%
CH3OH/H2O) to give 44.3 mg (36%) of an off-white solid after
removal of the solvent: HRMS ESIþ m/z 500.2533 [M þ H]; 1H
NMR after treatment with 5 equiv of ascorbic acid to give 2 (400
MHz, DMSO-d6) δ 9.77 (br s, 1H), 8.53 (s, 1H), 8.29 (d, J = 7.58
Hz, 2H), 8.13-8.23 (m, 2H), 7.86 (d, J = 8.08 Hz, 1H), 7.59 (t, J =
7.71 Hz, 1H), 7.20-7.42 (m, 2H), 7.08 (s, 1H), 6.87-6.99 (m, 2H),
6.71 (dd, J= 1.52, 8.08 Hz, 1H), 6.34 (d, J= 5.31 Hz, 1H), 4.26 (dt,
J = 3.95, 8.02 Hz, 1H), 1.79 (dd, J = 3.66, 12.51 Hz, 2H), 1.56 (t,
J = 12.38 Hz, 2H), 1.13 (s, 6H), 1.11 (s, 6H); 13C NMR after
treatment with 5 equiv of ascorbic acid to give 2(400MHz, DMSO-
d6) δ 165.40, 163.20, 158.15, 153.33, 150.42, 149.99, 145.01, 138.13,
135.34, 132.77, 130.04, 128.59, 128.53, 127.53, 125.07, 114.62,
112.80, 111.06, 92.20, 87.27, 58.04, 48.49, 44.66, 40.78, 32.50, 19.58.
Synthesis of Kinex Compound (3). This compound was synthe-
sized at Wyeth according to the published procedures.33
3-{[2-(3-Methylcarboxyphenyl)pyrazolo[1,5-r]pyrimidin-7-yl]-
amino}phenol (22). To a solution of 3-[(2-bromopyrazolo[1,5-
R]pyrimidin-7-yl)amino]phenol (0.305 g, 1 mmol) in dioxane
(6 mL) were added 3-methoxycarbonylphenylboronic acid
(0.271 g, 1.5 mmol) and then 2 M potassium carbonate (1.6
mL, 3.2 mmol). Nitrogen was bubbled through the solution. Pd-
(dppf)Cl2 (73 mg, 0.1 mmol) was added, and then the mixture
was heated at 90 °C for 1.5 h. After the mixture was cooled, it was
diluted with ethyl acetate (50 mL) and shaken vigorously with
brine (50 mL). The biphasic mixture wasfilteredthrough a pad of
Celite; the layers were separated, and the organic layer was
washed again with brine. The organic layer was dried
(Na2SO4), filtered, and concentrated in vacuo. Eluting through
a small silica plug with a 1:1 EtOAc/hexane mixture followed by
trituration of the resultant solid with a 2:1 EtOAc/hexane
mixture yielded 300 mg (83%) of an off-white powder: 1H
NMR (400 MHz, DMSO-d6) δ 9.85 (s, 1H), 9.71 (s, 1H), 8.60-
Protein Expression and Purification with Lck and Src. The Lck
KD construct that encodes residues Q225-P509 was expressed
in insect cells using the Baculovirus expression system. This
construct was designed with an N-terminal His6 purification tag
followed by a thrombin protease cleavage site (MHHHHH-
HSSGLVPRGS). The kinase domain was PCR amplified from
a full-length Lck cDNA and subsequently cloned into pBac-
PAK9 (Clontech, Moutain View, CA) using Clontech’s In-
fusion protocol. A DNA sequence-confirmed clonal isolate
was recombined and transfected into insect cells with Baculo-
Gold (BD Biosciences, San Diego, CA) following the protocol
of the manufacturer. High titer virus (HTV) was prepared from
pooled P1 virus and amplified in Spodoptera frugiperda 9 (Sf9)
cells at a density of 5 ꢀ 105 cells/mL and a multiplicity of
infection (MOI) of 0.1 plaque forming unit (pfu)/cell. The HTV
was then used to infect S. frugiperda 21 (Sf21) cells at a density of
1.9 ꢀ 106 cells/mL with an MOI of 1.5 pfu/cell. The cells were

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3 1247
harvested by centrifugation 72 h postinfection, resuspended in 5
mL of PBS, flash-frozen dropwise in liquid nitrogen, and then
stored at -80 °C. Purification followed by lysing typically
produced 100 g of cells in 800 mL of 25 mM Tris (pH 7.5),
5 mM BME, Complete Protease inhibitor Cocktail [ethyl-
enediaminetetraacetic acid (EDTA)-free, Roche Applied Science,
Indianapolis, IN], 1 mM sodium orthovanadate, 50 mM sodium
fluoride, 0.25 mM AEBSF, and 5 mM β-glycerophosphate. The
lysate was cleared by centrifugation and filtered through a 0.45 μm
membrane. The His-tagged protein was captured on a 5 mL
HisTrap HP column (GE) and eluted using a 30 to 500 mM
imidazole gradient in 20 mM Tris (pH 8), 400 mM NaCl, and
5 mM 2-mercaptoethanol. The His tag was removed by thrombin
cleavage during dialysis overnight against 50 mM Tris (pH 8),
20 mM NaCl, and 5 mM dithiothreitol [20 units of thrombin
(Amersham)/mg of protein]. The dialyzed sample was passed
through a 3 mL Poros 50 HS column. The unphosphorylated
Lck protein and monophosphorylated Lck protein from the flow
through were separated on a Mono Q 10/100 GL column using a
NaCl gradient from 100 to 260 mM over 15 column volumes. The
unphosphorylated Lck was used in the NMR experiments without
further purification. The monophosphorylated Lck was concen-
trated using an Amicon (10000 MWCO) centrifugal concentrator
and subsequently loaded onto a HiLoad 16/60 Superdex 200
column equilibrated with TBS and 5 mM DTT. The eluent which
came off as a single peak was used in Lck X-ray data collection.
The human Src KD construct that encodes residues T250-
L536 was expressed in Sf21 insect cells using the Baculovirus
expression system. Src KD has a C-terminal His tag preceded by
a thrombin cleavage site (LVPRGSHHHHHH) where the
leucine is L536. The construct was generated by PCR amplifica-
tion using full-length cDNA and cloned into pBacPAK 9
(Clontech) using Clontech’s In-fusion protocol. The plasmids
were transformed, and the DNA sequences of clonal isolates
were confirmed. The Src KD construct was recombined and
transfected into insect cells with BaculoGold (BD Biosciences)
following the manufacturer’s recommendations. High titer virus
(HTV) was prepared from pooled P1 virus amplified in Sf9 cells
at a density of 5 ꢀ 105 cells/mL and an MOI of 0.1 pfu/cell. The
HTV was then used to infect Sf21 cells at a density of 1.5 ꢀ 106
cells/mL with an MOI of 2.0 pfu/cell. The cells were harvested by
centrifugation 72 h postinfection, resuspended in 5 mL of PBS-
CMF/L, flash-frozen dropwise in liquid nitrogen, and stored at
-80 °C. Purification of Src KD was conducted using a mod-
ification of the method of Dalgarno.34 All procedures were
performed at 4 °C. Cells from 1 L of culture were lysed in 200
mL of buffer A [25 mM Tris (pH 8), 150 mM NaCl, 10%
glycerol, 5 mM imidazole, and 5 mM 2-mercaptoethanol]
supplemented with 10 mM sodium fluoride, Complete Protease
Inhibitor Cocktail (EDTA-free, Roche Applied Science), 25 μg/
mL RNase, and 50 μg/mL DNase. Lysate was centrifuged at
25000g for 30 min, and the supernatant was brought to 400 mM
NaCl. Lysate was subsequently batch bound to 7 mL of Ni-
NTA superflow resin (Qiagen, Valencia, CA) for 1.5 h at 4 °C.
The resin was packed into an Applied Biosystems column and
washed with 7 column volumes of buffer A. The resin was then
washed with 7 column volumes of buffer B [50 mM Tris (pH 8),
500 mM NaCl, 5% glycerol, and 1 mM DTT] supplemented
with 10 mM imidazole, and then 7 column volumes of buffer B
supplemented with 30 mM imidazole. The His6-tagged Src KD
was eluted with 7 column volumes of buffer B supplemented
with 300 mM imidazole. The pooled peak from the Ni-NTA
column was concentrated to a final volume of 5 mL using a
Vivaspin 20 (10000 MWCO) centrifugal concentrator (Sartorius
Biolabs, Goettingen, Germany). This concentrated sample was
purified to >95% homogeneity via gel filtration at a flow rate of
1.5 mL/min on a HiLoad 1.6 ꢀ 60 Superdex 75 column (GE
Healthcare, Piscataway, NJ) equilibrated with 50 mM Tris-HCl
(pH 7.5), 10% glycerol, 1 mM tris(2-carboxyethyl)phosphine
hydrochloride (TCEP) (Pierce Chemical Co., Rockford, IL),
and 1 mM EDTA.
Enzymatic Assays: Kinase Panel. For the kinase selectivity
profiling assay, all buffer salts and reagents were purchased
from Sigma and were of the highest purity available. Kinases
were purchased from Invitrogen and were used without further
purification. Fluorescently labeled peptide substrates were pur-
chased from Anaspec, and the peptide sequence was optimized
for each kinase. Compound plates were prepared by pipetting an
11-point dose-response titration in DMSO with final assay
concentrations ranging from 10 μM to 5 nM. Assay plates were
prepared with a 150 nL addition of compound, followed by
7.5 μL of 50 mM assay buffer (10 mM MgCl2, 0.005% Brij, and
5 mM 2-mercaptoethanol) containing substrate and ATP, follo-
wed by 7.5 μL of buffer containing enzyme. All reactions were
performed at 1.5 μM peptide substrate ([S] , Km) and at the
experimentally determined Km of ATP for each kinase. All final
enzyme concentrations were below 50 nM and in most cases
were e10 nM. The reaction mixtures were incubated at room
temperature for 1-2 h depending on kinase activity but in all
cases exhibited between 20 and 30% substrate phosphorylation.
The reactions were quenched by the addition of 15 μL of stop
buffer (100 mM HEPES, 20 mM EDTA, 0.005% Brij, and 0.2%
Caliper coating reagent). Substrates and products were sepa-
rated on the Caliper LC3000 instrument using standard separa-
tion protocols. The percent inhibition was calculated for each
compound concentration, and the IC50 was determined using
the following equation:
0
@
1
B
B
C
C
max -min
B
C
C
A
% inhibition ¼ max þ
B
ꢀ
ꢁ
Hill
Conc
IC50
1 þ
LANCE TR-FRET Lck. Inhibitory effects of compounds on
Lck were determined using a homogeneous time-resolved fluor-
escence (LANCE) assay. Human full-length Lck protein was
produced with a C-terminal His6 tag using a Baculovirus
expression system. The purified Lck protein was confirmed to
be more than 95% pure by SDS-PAGE. Reagents are pipetted
and dispensed in polystyrene V-bottom 384-well microtiter
plates using Matrix PlateMate 4 ꢀ 4 liquid handling equipment.
Kinase activity was determined in 20 mM HEPES buffer (pH
7.4) containing 10 mM MgCl2, 1 mM 2-mercaptoethanol,
0.0025% Brij 35, 10 μg/mL BSA, 50 pM Lck, 50 μM adenosine
triphosphate (ATP), and 400 nM synthetic peptide biotin-
EGPWLEEEEEAYGWMDF-NH2. The reaction mixture was
incubated at room temperature for 2 h for the kinase reaction.
The reactions were terminated by the addition of buffer contain-
ing 45 mM EDTA, 20 mM HEPES (pH 7.4), and 10 μg/mL
BSA, containing a streptavidin-allophycocyanin (SA-APC) and
a monoclonal phosphospecific antibody conjugated to a euro-
pium cryptate. The plate was incubated at room temperature
for 30 min with shaking according to the manufacturer’s
instructions. The time-resolved fluorescence signal of SA-
APC, produced only when the synthetic peptide is phosphory-
lated by Lck, was measured on a PerkinElmer EnVision micro-
plate reader. Test compounds were prediluted using 100%
dimethyl sulfoxide solutions to produce dose-response curves
(11-point curves from final concentrations of 30 μM to ∼0.5
nM). Inhibition of Lck activity with compounds of the invention
was expressed as percentage inhibition of control activity
exhibited in the absence of test compounds. The concentration
of test compounds resulting in 50% inhibition of the Lck activity
(IC50) was determined from the dose-response curves.
NMR Samples and Binding Experiments. NMR samples were
prepared with Lck KD at 25 mM HEPES in 200 mM NaCl,
10 mM MgCl2, 1 mM DTT (pH 7.5), and a 5% DMSO/95%
D2O mixture and for Src KD at 20 mM HEPES in 200 mM NaCl,
5 mM MgCl2, 2 mM DTT (pH 7.5), and a 2% DMSO/98% D2O

1248 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3
Moy et al.
mixture. Titration with the reduced probe, 2, was performed at a
compound concentration of 10 μΜ and with addition of Lck
KD at 2.5, 5, and 10 μΜ. Excess ascorbic acid (5ꢀ) was added to
1 to produce its reduced form, 2. The NMR sample used for the
saturation transfer difference (STD) and T1F experiments con-
tained 7.5 μΜ Lck, 125 μΜ AMP-PNP, and 10 μΜ 1. All spectra
were recorded at 25 °C with a Bruker Avance 600 MHz spectro-
meter equipped with a triple-resonance 5 mm inverse cryoprobe.
For the STD experiments, nearly complete saturation of the
protein was achieved with irradiation at 0.78 ppm for 3 s, using a
train of 60 G4 Gaussian cascade shape pulses35 of 50 ms, each
separated by a 1 ms delay. Off-resonance irradiation was applied
at -4 ppm, void of any protein resonances. The subtraction
of STD spectra was performed internally with on- and off-
resonance protein excitation using phase cycling.36 All STD
experiments were performed with 400 scans with a total time of
35 min along with a one-dimensional 1H reference spectrum
with 200 scans. The STD % was reported as the ratio of the peak
intensities in the STD spectra and the one-dimensional 1H
reference spectra. T1F experiments were conducted at 10 and
200 ms spin-lock times with 128 scans. In the T1F experiments,
we report the relaxation intensity ratio as a ratio of resonance
signals collected with a 200 ms time over the 10 ms spin-lock
spectra. A lower percentage qualitatively indicates more en-
hanced relaxation when comparing experiments with the probe,
1, and with the reduced probe, 2, for fast exchanging systems.
X-ray Crystallography. The crystals of the Lck KD-1 com-
plex were grown by the vapor diffusion method at room
temperature. The molar ratio of the complex solution is 1:1.5
(protein:inhibitor). Equal volumes of protein complex solution
[10 mg/mL protein, TBS (pH 7.5), 100 mM NaCl, and 5 mM
DTT] and well solution [25% PEG 4000, 80 mM HEPES (pH
7.5), 150 mM ammonium sulfate, and 5% 2-propanol] were
mixed and suspended over 1 mL of well solution. Crystals were
obtained 2 days following the setup. The crystals were flash-
frozen in liquid nitrogen by using a cryosolution composed of
(4) Noble, M. E. M.; Endicott, J. A.; Johnson, L. N. Protein kinase
inhibitors: Insights into drug design from structure. Science 2004,
303 (5665), 1800–1805.
(5) Druker, B. J. Imatinib as a paradigm of targeted therapies. Adv.
Cancer Res. 2004, 91, 1–30.
(6) Wong, S.; Witte, O. N. The BCR-ABL story: Bench to bedside and
back. Annu. Rev. Immunol. 2004, 22, 247–306.
(7) Nagar, B.; Bornmann, W. G.; Pellicena, P.; Schindler, T.; Veach,
D. R.; Miller, W. T.; Clarkson, B.; Kuriyan, J. Crystal structures of
the kinase domain of c-Abl in complex with the small molecule
inhibitors PD173955 and Imatinib (STI-571). Cancer Res. 2004, 62,
4236–4243.
(8) Wan, P. T. C.; Garnett, M. J.; Roe, S. M.; Lee, S.; Niculescu-Duvaz,
D.; Good, V. M.; Project, C. G.; Jones, C. M.; Marshall, C. J.;
Springer, C. J.; Barford, D.; Marais, R. Mechanism of activation of
the RAF-ERK signaling pathway by oncogenic mutations of B-RAF.
Cell 2004, 116 (6), 855–867.
(9) Pargellis, C.; Tong, L.; Churchill, L.; Cirillo, P. F.; Gilmore, T.;
Graham, A. G.; Grob, P. M.; Hickey, E. R.; Moss, N.; Pav, S.;
Regan, J. Inhibition of p38 MAP kinase by utilizing a novel
allosteric binding site. Nat. Struct. Biol. 2002, 9 (4), 268–272.
(10) Pellechia, M. Solution nuclear magnetic spectroscopy techniques
for probing intermolecular interactions. Chem. Biol. 2005, 12 (9),
961–971.
(11) Meyer, B.; Peters, T. NMR spectroscopy techniques for screening
and identifying ligand binding to protein receptors. Angew. Chem.,
Int. Ed. 2003, 42 (8), 864–890.
(12) Jahnke, W.; Florsheimer, A.; Blommers, M. J. J.; Paris, C. G.;
Heim, J.; Nalin, C. M.; Perez, L. B. Second-site NMR screening
and linker design. Curr. Top. Med. Chem. 2003, 3 (1), 69–80.
(13) McCoy, M. A.; Senior, M. M.; Wyss, D. F. Screening of protein
kinases by ATP-STD NMR spectroscopy. J. Am. Chem. Soc. 2005,
127 (22), 7978–7979.
(14) Jahnke, W.; Perez, L. B.; Paris, C. G.; Strauss, A.; Fendrich, G.;
Nalin, C. M. Second-site NMR screening with a spin-labeled first
ligand. J. Am. Chem. Soc. 2000, 122 (30), 7394–7395.
(15) Jahnke, W.; Ruedisser, S.; Zurini, M. Spin label enhanced NMR
screening. J. Am. Chem. Soc. 2001, 123 (13), 3149–3150.
(16) Deschamps, M. L.; Pilka, E. S.; Potts, J. R.; Campbell, I. D.; Boyd,
J. Probing protein-peptide binding surfaces using charged stable
free radicals and transverse paramagnetic relaxation enhancement
(PRE). J. Biomol. NMR 2005, 31 (2), 155–160.
˚
(17) Jahnke, W. Spin labels as a tool to identify and characterize
protein-ligand interactions by NMR spectroscopy. ChemBio-
Chem 2002, 3 (2-3), 167–173.
the well solution with 25% (v/v) glycerol. The 2.36 A diffraction
data were collected by using synchrotron radiation at Brookhaven
National Laboratories, beamline NSLS X-29, and processed
and scaled with HKL2000.37 The crystals belong to space
group P212121, with the following unit cell parameters: a =
(18) Vazquez, J.; De, S. K.; Chen, L.-H.; Riel-Mehan, M.; Emdadi, A.;
Cellitti, J.; Stebbins, J. L.; Rega, M. F.; Pellecchia, M. Develop-
ment of paramagnetic probes for molecular recognition studies in
protein kinases. J. Med. Chem. 2008, 51 (12), 3460–3465.
(19) Marsilje, T. H.; Milkiewicz, K. L.; Hangauer, D. G. The design,
synthesis and activity of non-ATP competitive inhibitors of pp60c-
src tyrosine kinase. Part 1. Hydroxynaphthalene derivatives.
Bioorg. Med. Chem. Lett. 2000, 10 (5), 477–481.
(20) Milkiewicz, K. L.; Marsilje, T. H.; Woodworth, R. P., Jr.; Bifulco,
N., Jr.; Hangauer, M. J.; Hangauer, D. G. The design, synthesis
and activity of non-ATP competitive inhibitors of pp60c-src
tyrosine kinase. Part 2. Hydroxyindole derivatives. Bioorg. Med.
Chem. Lett. 2000, 10 (5), 483–486.
(21) Hangauer, D. G. The Discovery of KX01: A Highly Selective, Non-
ATP Competitive, Src Inhibitor. CHI’s 5th Annual Protein Kinase
Targets Meeting, Boston, June 4-6, 2007.
(22) Sofan, M. A.-M.; Abdel-Aziz El-Taweel, F. M.; El-Maati, T. A.;
Ali El-Agamey, A.-G. Acylation reactions of 5-amino-3-phenyl-
pyrazole: Routes to pyrazolopyrimidines, pyrazolodiazepine, pyr-
azoloquinazoline and pyrazolopyrrolopyrimidine derivatives.
Indian J. Chem., Sect. B: Org. Chem. Incl. Med. Chem. 1994, 33B
(8), 738–741.
˚
˚
˚
42.0 A, b = 73.8 A, and c = 84.7 A. The protein and inhibitor
model were built using COOT38 and refined using PHENIX39
to a final R factor of 0.192 and an Rfree of 0.26. The structure
of Lck with 1 has been deposited in the Protein Data Bank as
entry 3KXZ.
Acknowledgment. We thank Nelson Huang for providing
mass spectrometry data for 1 and Karen Moreira, Shashi
Mohan, Jenny Tobias, and Richard Harrison for their help in
the assays of 1 with Src, Lck, and the kinase panel. We also
thank Weixin Xu for input provided on X-ray crystallo-
graphy.
Supporting Information Available: Purity data for all novel
compounds and simulations of the T2 relaxation equation for
the protein/ligand system in fast exchange. This material is
available free of charge via the Internet at http://pubs.acs.org.
(23) Elmaati, T. M. A.; El-Taweel, F. M. New trends in the chem-
istry of 5-aminopyrazoles. J. Heterocycl. Chem. 2004, 41 (2),
109–134.
(24) Gavrin, L. K.; Lee, A.; Provencher, B. A.; Massefski, W. W.; Huhn,
S. D.; Ciszewski, G. M.; Cole, D. C.; McKew, J. C. Synthesis of
pyrazolo[1,5-R]pyrimidinone regioisomers. J. Org. Chem. 2007, 72
(3), 1043–1046.
(25) Springer, R. H.; Scholten, M. B.; O’Brien, D. E.; Novinson, T.;
Miller, J. P.; Robins, R. K. Synthesis and enzymic activity of
6-carbethoxy- and 6-ethoxy-3,7-disubstituted pyrazolo[1,5-a]-
pyrimidines and related derivatives as adenosine cyclic 30,50-phos-
phate phosphodiesterase inhibitors. J. Med. Chem. 1982, 25 (3),
235–242.
References
(1) Grant, S. K. Therapeutic protein kinase inhibitors. Cell. Mol. Life
Sci. 2009, 66 (7), 1163–1177.
(2) Johnson, L. N. Protein kinase inhibitors: Contributions from
structure to clinical compounds. Q. Rev. Biophys. 2009, 42 (1),
1–40.
(3) West, K. CP-690550, a JAK3 inhibitor as an immunosuppressant
for the treatment of rheumatoid arthritis, transplant rejection,
psoriasis and other immune-mediated disorders. Curr. Opin. Invest.
Drugs 2009, 10 (5), 491–504.
(26) Anwar, H. F.; Elnagdi, M. H. Recent developments in aminopyr-
azole chemistry. ARKIVOC 2009, 1, 198–250.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 3 1249
(27) Juffermans, J. P. H.; Habraken, C. L. Pyrazoles. 19. Selective
thermolysis reactions of bromo-1-nitro-1H-pyrazoles. Formation
of 3-nitro-1H- vs. 4-nitro-1H-pyrazoles. J. Org. Chem. 1986, 51
(24), 4656–4660.
(28) Lepre, C. A.; Moore, J. M.; Peng, J. W. Theory and applications of
NMR-based screening in pharmaceutical research. Chem. Rev.
2004, 104 (8), 3641–3675.
(29) Mayer, M.; Meyer, B. Characterization of ligand binding by
saturation transfer difference NMR spectroscopy. Angew. Chem.,
Int. Ed. 1999, 38 (12), 1784–1788.
(30) Berger, S. 200 and More NMR Experiments; Wiley-VCH: Weinheim,
Germany, 2004.
(31) Leipert, T. K.; Noggle, J. H.; Freeman, W. J.; Dalrymple, D. L.
Rotating frame nuclear relaxation of PBr3: Off-resonance studies
by Fourier transform NMR. J. Magn. Reson. 1975, 19, 208–221.
(32) Bemis, G. W.; Murcko, M. A. The properties of known drugs. 1.
Molecular frameworks. J. Med. Chem. 1996, 39 (15), 2887–2893.
(33) Hangauer, D. G. Biaryl compositions and methods for modulating
a kinase cascade, preparation, pharmaceutical compositions, and
use in the treatment of diseases. US 2007/0015752 A1, 2007.
(34) Dalgarno, D.; Stehle, T.; Narula, S.; Schelling, P.; van Schraven-
dijk, M. R.; Adams, S.; Andrade, L.; Keats, J.; Ram, M.; Jin, L.;
Grossman, T.; MacNeil, I.; Metcalf, C., III; Shakespeare, W.;
Wang, Y.; Keenan, T.; Sundaramoorthi, R.; Bohacek, R.; Weigele,
M.; Sawyer, T. Structural basis of Src tyrosine kinase inhibition
with a new class of potent and selective trisubstituted purine-based
compounds. Chem. Biol. Drug Des. 2006, 67, 46–57.
(35) Emsley, L.; Bodenhausen, G. Gaussian pulse cascades: New ana-
lytical functions for rectangular selective inversion and in-phase
excitation in NMR. Chem. Phys. Lett. 1990, 165 (6), 469–476.
(36) Ernst, R. R.; Bodenhausen, G.; Wokaun, A. Principles of Nuclear
Magnetic Resonance in One and Two Dimensions; Clarendon Press:
Oxford, U.K., 1987; Vol. 14.
(37) Otwinowski, Z.; Minor, W. Processing of X-ray diffraction
data collected in oscillation mode. Methods Enzymol. 1997, 276,
307–326.
(38) Emsley, P.; Cowtan., K. Coot: Model-building tools for molecular
graphics. Acta Crystallogr. 2004, D60, 2126–2132.
(39) Adams, P. D.; Gopal, K.; Grosse-Kunstleve, R. W.; Hung, L.-W.;
Ioerger, T. R.; McCoy, A. J.; Moriarty, N. W.; Pai, R. K.; Read,
R. J.; Romo, T. D.; Sacchettini, J. C.; Sauter, N. K.; Storoni, L. C.;
Terwilliger, T. C. Recent developments in the PHENIX software
for automated crystallographic structure determination. J. Syn-
chrotron Radiat. 2004, 11, 53–55.