Combinatorial approach to identification of tyrphostin inhibitors of cytokine signaling

Article abstract of DOI:10.1016/j.bmc.2005.04.022

Aberrant or deregulated activity of certain cellular kinases has been shown to cause certain malignancies and other disorders. The tyrphostin molecule AG490 inhibits the action of the janus kinases JAK2 and JAK3. JAK2 is an indispensable molecule for tran

Full text of DOI:10.1016/j.bmc.2005.04.022

Bioorganic & Medicinal Chemistry 13 (2005) 4269–4278
Combinatorial approach to identification of tyrphostin
inhibitors of cytokine signaling
Ling Gu,a Hui Zhuang,b Brian Safina,c Xiao-yi Xiao,b
Wallace W. Bradfordd and Benjamin E. Richa,*
aHarvard Skin Disease Research Center, Department of Dermatology, Brigham and Women’s Hospital, Boston, MA 02115, USA
bChemRx, 9640 Towne Center Drive, San Diego, CA 92121, USA
cScripps Research Institute, 10550 North Torrey Pines Road, La Jolla, CA 92037, USA
dBiosciences Division, SRI International, 333 Ravenswood Ave., Menlo Park, CA 94025, USA
Received 1 September 2004; revised 10 April 2005; accepted 11 April 2005
Abstract—Aberrant or deregulated activity of certain cellular kinases has been shown to cause certain malignancies and other dis-
orders. The tyrphostin molecule AG490 inhibits the action of the janus kinases JAK2 and JAK3. JAK2 is an indispensable molecule
for transducing the signals conveyed by a large number of cytokines including IL-3 while JAK3 is essential for signaling by a smaller
number of cytokines including IL-7. A synthetic combinatorial chemical library containing 599 compounds was created and
screened for the ability to inhibit proliferation of IL3- and IL7-dependent cell lines to focus on molecules that interrupt those sig-
naling pathways. This screen identified a meta-trifluoromethyl derivative of AG490, 5H4, that is approximately twice as potent as
AG490 in cell-based assays. 5H4 blocked the factor-dependent proliferation of both of these cell lines, actively promoted cell death,
and diminished the JAK kinase activity. Administration of 5H4 to lymphoma-prone IL-7 transgenic mice reduced their spontaneous
lymphadenopathy. The improved characteristics of this novel compound bring this class of molecules closer to therapeutic utility.
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
processes, however, they can also contribute to patho-
logical and malignant cellular behavior. A body of evi-
dence has accumulated implicating abnormal
constitutive activation of JAK kinases in various can-
cers. This activation may be as a result of persistent
ligand expression, viral infection, chromosomal
rearrangement, or other unknown mechanisms. For
example, IL-7 and IL-15 convey potent mitogenic sig-
nals and enhance cell viability by activating JAK3,
which induces expression of an array of anti-apoptotic
genes including Bcl-2.6,7 Forced autocrine expression
of IL-7 or IL-15 in transgenic mice causes inflammatory
disorders and hematopoietic malignancies8–10 and per-
sistent IL-7 signaling has been implicated in a number
of human malignancies including cutaneous T cell lym-
phomas (CTCL),6 acute T cell leukemias,7,11 B cell lym-
The four members of the janus (JAK) kinase family,
JAK1, JAK2, JAK3, and TYK2 are non-receptor pro-
tein tyrosine kinases that play essential roles in the cellu-
lar responses to distinct cytokines. JAK kinases comprise
FERM, SH2, pseudokinase, and kinase domains. The
FERM domain mediates receptor interactions and both
the FERM and pseudokinase domains regulate the cata-
lytic activity of the kinase domain.1–4 SH2 domains bind
to certain protein domains containing phosphorylated
tyrosine residues.5 JAK1, JAK2, and TYK2 appear to
be expressed ubiquitously and are each activated by sev-
eral different cytokines while JAK3 expression is nor-
mally restricted to lymphoid cells.
The cellular responses to cytokines are important com-
ponents of normal developmental and immunological
phocytic
leukemia,12–14
Burkittꢀs
lymphoma,15
Hodgkinꢀs disease,16 and melanoma.17 Infection of cer-
tain cells with HTLV-I activates JAK kinases in T lym-
phocytes and leads to their transformation.18,19
Translocations between the TEL gene and JAK2 are
found in T cell20 and myeloid21 leukemias. These trans-
locations fuse the helix–loop–helix domains of the TEL
Keywords: Tyrphostin; JAK kinase; IL-7; Lymphoma.
*
Corresponding author. Tel.: +1 617 525 5555; fax: +1 617 525
5571; e-mail: brich@rics.bwh.harvard.edu
0968-0896/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2005.04.022

4270
L. Gu et al. / Bioorg. Med. Chem. 13 (2005) 4269–4278
gene with the kinase domain of JAK2. The TEL portion
of the resulting fusion protein causes it to oligomerize
spontaneously leading to constitutive activation of the
kinase in the JAK2 portion.
molecules have been shown to be purely competitive
inhibitors for the tyrosine-containing substrate and
non-competitive or mixed-competitive inhibitors with
respect to ATP.35 AG490 blocks the growth of a number
of different malignant cells including the K562 chronic
myelogenous leukemia cell line,36 and malignant cells
from acute lymphoblastic leukemia,31 mycosis fungo-
ides,37 myeloma,38 Sezary syndrome,39 large granular
lymphocytic leukemia,40 Hodgkinꢀs disease,41 and Phila-
delphia-positive B-lineage leukemia.42 JAK3 and the
v-abl protein interact with one another43 and AG490
has been shown to have an additive effect on the anti-
proliferative activity of STI571 on BCR-ABL driven
proliferation.44
Several types of small molecules that inhibit the activity
of JAK kinases and block cytokine signal transduction
have been identified and evaluated as potential thera-
peutic agents. These include the tyrphostin family of
tyrosine analogs,22,23 a natural product, cytovaricin
B,24 octylamino-undecyl-dimethylxanthine compounds,25
derivatives of dimethoxyquinazoline,26 some naphthyl
ketones,27 prodigiosin derivatives,28 and a pyridone-
containing-tetracycle29 as well as the recently described
CP-690,550.30 The best studied class of these inhibitors
is the tyrphostin family.31,32 These molecules are charac-
terized by an alkene substituted with a nitrile and a
hydroxylated phenyl group in a cis configuration (see
Fig. 1A). Tyrphostin AG490 (also known as B42) inhib-
its both JAK2 and JAK3 kinases and prevents phos-
phorylation of phosphatidyl inositol-3 kinase. AG490
blocks the signal transduction of a broad range of cyto-
kines that utilize JAK2 or JAK3 including IL-2, IL-3,
IL-7, IL-9, IL-12, IL-15, and GM-CSF.31,33,34 Although
as a potential Michael reaction acceptor AG490 might
be expected to react with cellular components, similar
In this study we have utilized cell-based assays to screen
for inhibitors for JAK2 and JAK3 among a combinato-
rial chemical library comprising 599 variants of
AG490.45 Two cytokine-dependent murine cell lines,
Baf/3 (an IL3-dependent pro-B cell line utilizing JAK2
kinase) and 2E8 (an IL7-dependent pre-B cell line utiliz-
ing JAK3 kinase), were used in the screening. We
identified several small molecules that inhibit the prolif-
eration of 2E8 and Baf/3 cells more potently than
AG490. The two most potent of these, containing trifluo-
romethyl groups, were obtained in larger quantities
Figure 1. Structures of compounds. (A) Tyrphostin AG490, also known as B42. (B) Variation of the combinatorial chemical library. (C) Compound
5H4. (D) Compound 8B9. (E) Synthetic scheme for individual compounds. Reagents and conditions: (a) 100 ꢁC, 6 h; 75% (3a), 73% (3b); (b)
piperidine, EtOH, reflux, 1 h; 82% (5H4), 75% (8B9).

L. Gu et al. / Bioorg. Med. Chem. 13 (2005) 4269–4278
4271
and characterized. These new inhibitors of JAK2 and
JAK3 may be useful as therapeutic agents to treat
kinase-mediated malignancies or other disorders.
of AG490 synthesized in parallel. Remarkably, these
two compounds 5H4 and 8B9 are similar meta- and
para-trifluoro-methyl substitutions of AG490, respec-
tively (see Fig. 1C and D).
2. Results
To facilitate further characterization of molecules 5H4
and 8B9, gram quantities of each were prepared. Figure
1E summarizes the syntheses of 5H4 (NSC D722757),
and 8B9 (NSC D722756). The procedures used were
based on those for similar compounds described in the
literature.46 By heating methyl cyanoacetate and the
appropriately substituted benzylamine 2a or 2b, the de-
sired a-cyanoamide 3a or 3b is obtained in good yield.
The a-cyanoamide is then condensed with 3,4-dihydr-
oxybenzaldehyde in the presence of piperidine to give
the desired tyrphostins 5H4 and 8B9 in good yield. Final
2.1. Sensitivity of cytokine-dependent cells to variants
of tyrphostin AG490
The IL7-dependent 2E8 cells that require activated
JAK3 for growth and the IL3-dependent Baf/3 cells that
require activated JAK2 were each used to screen the
library of AG490 variants. Cells were cultured in
the presence of the appropriate cytokine and each of
the 599 compounds from the chemical library as well
3
1
as AG490 and proliferation was measured as H-thymi-
compounds were fully characterized by H NMR, 13C
dine incorporation. Control cultures containing DMSO
only or no cytokine were also analyzed in parallel.
Screens of the compounds at 10 lM, in which AG490
inhibited proliferation by less than 50%, were most
informative. The two most potent candidate com-
pounds, 5H4 and 8B9, were substantially better inhibi-
tors of both Baf/3 and 2E8 cells than AG490. Samples
of these two molecules from the combinatorial library
were reproducibly 4–8 times more potent than samples
NMR, and MS.
These preparations were then evaluated for their abili-
ties to inhibit JAK-dependent proliferation of the 2E8
and Baf/3 cell lines as well as the Jurkat human acute
T cell leukemia cell line and the HL-60 human acute
promyelocytic leukemia cell line. The results of these
assays are depicted in Figure 2 and the concentrations
at which each of these preparations inhibit 50% of
Figure 2. Inhibition of cell proliferation by tyrphostins. Cell lines were cultured in the presence of increasing concentrations of the indicated
compounds and proliferation was measured as 3H-thymidine incorporation. Concentrations of compounds are indicated on the x-axis. 3H-thymidine
incorporation values are plotted on the y-axis as a percentage of the values obtained in the absence of added compound or the lowest dilution
(0.08 lM) and in the presence of an optimal concentration of the requisite cytokine for cytokine-dependent cells (IL-7 for 2E8 and IL-3 for Baf/3).
Each data point represents the mean of quadruplicate cultures. Error bars indicate the standard error. Isolated triangles on the right side of panels A
and B indicate values obtained in the absence (open) or presence (filled) of IL-7 or IL-3, respectively.

4272
L. Gu et al. / Bioorg. Med. Chem. 13 (2005) 4269–4278
Table 1. IC50 values (lM) for tyrphostins on different cell lines
cell proliferation. 2E8 and Baf/3 cells were treated with
AG490, 5H4, or 8B9 at the concentration of 10 or
20 lM for 48 h. Since active degradation of chromatin
is a hallmark of apoptosis, we measured the cellular
DNA content as an index of programmed cell death.
Cells were stained with propidium iodide and analyzed
by flow cytometry. The percentages of cells with less
DNA than diploid cells (sub-G1), representing dead
cells, are shown in Figure 3. Withdrawal of IL-7 from
2E8 cells induced 50% cell death, whereas treatment
with 20 lM of AG490, 8B9, or 5H4 induced 60%,
71%, and 83% cell death, respectively (Fig. 3A). Similar
data was obtained from Baf/3 cells. Withdrawal of IL-3
from Baf/3 cells induced 87% cell death, whereas treat-
ment with 20 lM of AG490, 8B9, 5H4 induced 49%,
75%, 74% of cell death, respectively (Fig. 3B). Consis-
tent with the cell proliferation results (Fig. 2), 5H4
inhibits more of the cell growth by inducing more cell
death. Control cultures contained 0.1% (14 mM)
DMSO, which is the same concentration as those con-
taining compounds at 20 lM. This is substantially lower
than the concentrations of DMSO that inhibit the pro-
liferation of 2E8 cells (IC50 = 150 mM) or Baf/3 cells
(IC50=260 mM) (data not shown).
Cell line
AG490
5H4
8B9
2E8
3.9 0.06
4.6 0.3
(37%)a
(15%)a
1.7 0.08
2.8 0.02
6.1 0.4
(40%)a
3.1 0.1
4.5 0.1
5.9 0.5
(29%)a
Baf/3
Jurkat
HL-60
a Inhibition at highest concentration tested (10 lM).
3H-thymidine incorporation (IC50) of each of the cell
lines are listed in Table 1.
Jurkat cells, which are reported not to express JAK3,33
were only moderately sensitive to all three compounds.
Interestingly, the Jurkat cells were significantly more
sensitive to 5H4 as well as 8B9 than they were to
AG490 (Fig. 2C). This is in contrast to a previous obser-
vation that Jurkat cells were relatively insensitive to
AG490.33 HL-60 cells exhibited very little sensitivity to
any of the compounds (Fig. 2D). Compounds 5H4
and 8B9 have also been screened for inhibition of
growth of the NCI-60 panel of cell lines by the Develop-
mental Therapeutics Program of NCI/NIH. The results
of these assays can be obtained at http://dtp.nci.nih.gov/
under the names NSC D722757 (5H4) and NSC
D722756 (8B9).
2.3. Inhibition of JAK3 kinases by novel tyrphostins
Because JAK kinases are the critical mediators for IL-7
and IL-3 signaling, we next examined the effects of the
compounds on kinase activities. 2E8 and Baf/3 cells
were collected after 48-h drug treatment and lysates
were prepared. JAK3 proteins were immunoprecipi-
tated from these lysates and the extent of their phos-
phorylation was determined by immunoblot with
anti-phosphotyrosine antibodies. As shown in Figure
2.2. Induction of programmed cell death by novel
tyrphostins
Since IL-7 supports the growth of CTCL by upregula-
ting the anti-apoptotic gene Bcl-26 and AG490 can
block cell growth by inducing apoptosis,31,47 we next
examined whether the candidates from the screening
use the same mechanism of inducing cell death to block
Figure 3. Induction of cell death by tyrphostins. 2E8 and Baf/3 cells were grown in the presence or absence of recombinant murine IL-7 or IL-3,
respectively, and 0.1% DMSO with or without the indicated concentrations of tyrphostins for 48 h. Cells were then stained with propidium iodide
and fluorescence was measured by flow cytometry to quantify DNA content. Representative histograms are shown in the upper panels. Bar graphs of
the percentages of cells with fluorescent intensities lower than cells in phase G1 (sub-G1) representing dead cells in each condition are plotted in the
lower panels.

L. Gu et al. / Bioorg. Med. Chem. 13 (2005) 4269–4278
4273
the effect of administration of compound 5H4 to IL7TG
mice. Magnetic resonance imaging (MRI) was used to
precisely measure the dimensions of lymph nodes of liv-
ing IL7TG mice before and after treatment. As shown in
Figures 5A and 6, IL7TG mice have greatly enlarged ax-
illary lymph nodes relative to normal wild-type FVB
mice and this can be clearly visualized by MRI.
Figure 4. Inhibition of JAK kinase activity by tyrphostins. IL-7-
dependent 2E8 cells were cultured without IL-7 or with IL-7 in the
presence of the indicated compounds for 48 h. Whole cell extracts
containing 500 lg of protein were immunoprecipated with anti-JAK3
antibody and the captured material was immunoblotted with anti-
phosphotyrosine antibody.
Two independent experiments were performed in which
the effects of 5H4 on IL7TG mice were evaluated. The
first experiment involved eight mice and the second
experiment involved four. In each experiment, 5–
6 month old heterozygous IL7TG mice were treated
with intraperitoneal injections of 0.1 mL of DMSO con-
taining 1 mg of 5H4 or DMSO alone for 14 days. In the
first experiment MRI images were collected of all mice
before treatment. One IL7TG mouse died after two
doses of 1 mg 5H4 injection on the first day of the first
experiment. Subsequently, all treatments were reduced
to one dose of 1 mg 5H4 per day. During the course
of the first experiment another 5H4-treated mouse and
one DMSO-treated mouse died. Each of the surviving
mice in the first experiment, two 5H4-treated mice and
three DMSO-treated mice, were analyzed by MRI and
histology. In the second experiment, one DMSO-treated
4, phosphorylation of JAK3 was inhibited to a greater
extent in lysates prepared from 5H4- and 8B9-treated
cell than in lysates from AG490-treated cells.
2.4. Regression of lymphadenopathy in IL-7 transgenic
mice by treatment with 5H4
Transgenic mice expressing IL-7 under the control of the
immunoglobulin heavy chain enhancer and promoter
(IL7TG) develop a lymphoproliferative disorder that
progresses to B and T cell lymphomas.8 Since JAK3 is
an essential mediator for IL-7 signaling, we examined
Figure 5. Treatment of IL7TG mice with tyrphostin 5H4. Groups of 5–6 month old heterozygous IL7TG mice were administered daily
intraperitoneal injections of 1.0 mg 5H4 in DMSO, or DMSO alone for 14 days. MRI imaging was performed before and after the treatment. (A)
Representative images of IL7TG mice (TG) treated with 5H4 in DMSO or DMSO alone as well as a matched wild-type mouse (WT) as indicated, are
presented. The axillary lymph nodes are indicated by arrows. (B) Axillary lymph node volumes before and after treatment were measured and
calculated as described in Experimental and plotted as white (before treatment) or shaded (after treatment) columns.

4274
L. Gu et al. / Bioorg. Med. Chem. 13 (2005) 4269–4278
Figure 6. Histology of lymph nodes and spleen. Hematoxylin and eosin stained paraffin sections of lymph nodes (A–F) and spleen (G–L) isolated
from wild-type or IL7TG mice, untreated, treated with DMSO, or treated with 5H4 as indicated. Representative low power (A–C, G–I) and high
power (D–F, J–L) fields are presented. One mm (low power) and 0.1 mm (high power) scale bars are embedded in the left image of each row. Lymph
node follicular structures are indicated by black triangles and paracortical regions are indicated by black · marks (A–C). Spleen white pulp (stained
blue) is indicated by black triangles and red pulp (stained red/purple) is indicated by black · marks (G–H).
mouse and one 5H4-treated mouse died on day 1 and
day 3 of the experiment. Those mice were replaced with
previously untreated IL7TG mice and given the same
amount of the treatment for the same period of time
by extending their courses by 1 and 3 days, respectively.
In the second experiment, only histology analysis was
performed.
14 days showed 30–50% reductions of lymph node vol-
umes, while similar changes were not seen in the mice
that received only DMSO. At necropsy, the lymph
nodes of the 5H4-treated mice appeared soft and de-
flated, in contrast to the turgid character of the lymph
nodes of normal and DMSO-treated transgenic mice.
Histological examination of lymph nodes of 5H4-treated
IL7TG mice revealed prominent spongy regions with
fewer lymphocytes and a lightly eosin-stained (pink)
stromal appearance (Fig. 6C and F), that is not observed
in the lymph nodes from DMSO-treated IL7TG mice
(Fig. 6B and E). Follicular structures were also reduced
in size and cellularity in the lymph nodes of 5H4-treated
IL7TG mice (Fig. 6C). The IL7TG mice also exhibit
splenomegaly with expanded regions of white pulp con-
taining abundant lymphocytes and lymphocytic accu-
mulations in the red pulp (Fig. 6G and J vs H and K).
These expanded regions of white pulp were diminished
in the spleens of 5H4-treated IL7TG mice (Fig. 6). In
The lymph nodes of wild-type mice housed in specific-
pathogen-free conditions are small and relatively quies-
cent (Figs. 5 and 6A and D). In contrast, the spontane-
ous lymphadenopathy of the IL7TG mice involves
characteristic expansion of the follicular and paracorti-
cal regions containing abundant tightly packed mono-
morphic lymphocytes, seen as round cells darkly
stained with hematoxylin (blue) and some effacement
of structures (Fig. 6B and E). As shown in Figure 5,
MRI measurements of the two analyzable mice that re-
ceived 1 mg/day of compound 5H4 in DMSO for

L. Gu et al. / Bioorg. Med. Chem. 13 (2005) 4269–4278
4275
summary, the lymphadenopathy of IL7TG mice was
markedly reduced by treatment with compound 5H4
and perturbations of the spleen architecture associated
with the IL7TG were partially reversed.
some cases improve, the biological function of drugs.
It is noteworthy that a para-chloro substituted deriva-
tive was also more active than AG490 (data not shown).
The trifluoromethyl group is lipophilic and thus it is
likely to increase the extent to which the molecule is ab-
sorbed into cell membrane and can enhance the ability
of the drug to pass through it.48 It may also increase
hydrophobic interactions between the drug and the tar-
get enzyme and might alter the metabolism of the mole-
cule and extend its biological half-life.49 While the cause
of the improved activity of compound 5H4 is not ad-
dressed by this study, it appears most likely that it is
due to an increased ability of the molecule to absorb
into and pass through the cell membrane. Although
AG490 and 5H4 are both likely to be susceptible to
nucleophilic attack, there is no evidence that their bio-
logical activities involve chemical reactivity. The biolog-
ical significance of the potential reactivity of these
molecules is unclear.
3. Discussion
This study identified two novel derivatives of tyrphostin
AG490, 5H4, and 8B9, which display significantly im-
proved inhibition of the proliferation of cell lines that
are dependent upon the continued activation of JAK2
or JAK3 kinases. Each of these molecules inhibit the
proliferation of cytokine-dependent 2E8 cells and Baf/
3 cells. They are substantially less effective as inhibitors
of the factor-independent Jurkat acute T cell leukemia
cell line and have minimal effects on the HL-60 acute
promyelocytic leukemia cell line. Both 5H4 and 8B9
are trifluoromethyl derivatives of the parent compound
tyrphostin AG490. These molecules each inhibit cyto-
kine-induced JAK kinase activities and induce pro-
grammed cell death. To examine the in vivo efficacy of
the most potent of these molecules, it was administered
to lymphoma-prone IL7TG mice. Fourteen daily doses
of 1 mg of 5H4 greatly reduced the volumes of the en-
larged lymph nodes of these mice. Histological examina-
tion revealed significant reductions in lymph node
cellularity.
It has recently been reported that a novel tyrosine kinase
inhibitor CR4 ((E,E)-2-(benzylaminocarbonyl)-3-(3,4-di-
hydroxystyryl)acrylonitrile) markedly inhibits JAK2
and Bcr-Abl kinase activity and the growth and survival
of acute lymphoblastic leukemia (ALL) and acute mye-
loid leukemia (AML).50 CR4 differs from AG490 only in
that the dihydroxystyryl group is displaced by two addi-
tional unsaturated carbon atoms in CR4. This study
suggests that similar trifluoromethyl modifications of
CR4 might also increase its potency.
The screening strategy employed in this study utilized
cytokine-dependent cell lines to focus on signal trans-
duction pathways. IL-7 signaling activates JAK3 and
JAK1 kinases, whereas IL-3 signaling activates JAK2
kinase. Cell lines dependent upon these two parallel
but distinct signal transduction pathways were used in
an attempt to identify compounds with improved activ-
ity as well as any that might display differential activity.
The trifluoromethyl derivatives 5H4 and 8B9 each inhi-
bit both of these pathways. The Jurkat T cell leukemia
cell line requires no exogenous factors for growth and
is reported not to express JAK3.33 Nonetheless, it is sen-
sitive to the tyrphostins, but at higher concentrations
than the factor-dependent lines. Even the non-lymphoid
promyelocytic leukemia HL-60 cell line has some sensi-
tivity to higher concentrations of 5H4. Thus it appears
that high concentrations of the tyrphostins may affect
important molecules other than the JAK kinases.
Reduction of the spontaneous lymphadenopathy of
IL7TG by treatment with tyrphostin 5H4 demonstrates
that the trifluoromethyl modification does not interfere
with the activity of the molecule in vivo. The ability of
AG490 and its derivatives to induce apoptosis and the
reductions in sizes of the lymph nodes that we observed
in 5H4-treated mice suggest the possibility of apoptosis.
However, we failed to detect the presence of apoptotic
cells using a TUNEL assay (data not shown). Nonethe-
less, it is possible that late stage apoptotic cells have
been cleared from the tissue, since the overall lymph
node volumes of 5H4-treated mice decreased by 30–
50%.
This study demonstrates a strategy of screening com-
pounds for bioactivity using factor-dependent cell lines
to focus on distinct signaling pathways. The improve-
ment in biological activity associated with the addition
of a trifluoromethyl group to the tyrphostin molecule
might reflect an improved interaction with the cell mem-
brane. The ability of 5H4 and other tyrphostins incorpo-
rating further modifications to impair JAK kinase
signaling in vitro and in vivo may ultimately lead to
their development as therapeutic agents.
AG490 has been shown to block the growth of acute
lymphoblastic leukemia (ALL) cells by inducing pro-
grammed cell death.31 Consistent with this, 5H4 and
8B9 also triggered programmed cell death in 2E8 and
Baf/3 cells.
Among the 599 derivatives of AG490 in the combinato-
rial library, 5H4 and 8B9 were the most potent inhibi-
tors of each of the factor-dependent cell lines and they
were more effective inhibitors of Jurkat cells and
HL-60 cells. They each have trifluoromethyl substitu-
tions of the benzylamino group at either meta or para
positions, respectively. The replacement of hydrogen
atoms or hydroxyl groups by halogens or halogen-con-
taining groups is a recognized strategy to alter, and in
4. Experimental
4.1. Cell cultures
The IL7-dependent murine pro-B cell line 2E8 (ATCC,
TIB-239,51) was maintained in cDMEM medium

4276
L. Gu et al. / Bioorg. Med. Chem. 13 (2005) 4269–4278
(Dulbeccoꢀs modified Eagle medium with 10% fetal bo-
vine serum, 2 mM L-glutamine, 100 IU/mL of penicillin,
and 100 lg/mL of streptomycin, 10 mM Hepes, 1 mM
sodium pyruvate, 1· non-essential amino acids, and
0.05 mM 2-mercaptoethanol), supplemented with
150 ng/mL of recombinant murine IL-7 (B. Rich and
L. Cosenza, unpublished). The IL3-dependent murine
pro-B cell line Baf/352,53 was maintained in RPMI-
1640 medium with 10% fetal bovine serum, 2 mM L-glu-
tamine, 100 IU/mL penicillin, and 100 lg/mL strepto-
mycin, supplemented with 0.5 ng/mL of recombinant
murine IL-3 (R&D Systems). Jurkat (ATCC, TIB-152)
and HL-60 (ATCC, CCL-240) cells were maintained in
RPMI-1640 medium with 10% fetal bovine serum,
2 mM L-glutamine, 100 IU/mL of penicillin, and
100 lg/mL of streptomycin. All cells were incubated at
37 ꢁC in 5% CO2 and 100% humidity.
low solid. 1H NMR (300 MHz, acetone-d6) d (ppm)
4.67 (d, 2H, J = 6.0 Hz), 6.97 (d, 1H, J = 8.4 Hz), 7.38
(d, 1H, J = 8.4 Hz), 7.59–7.73 (m, 6H), 8.10 (s, 1H). 13C
NMR (75 MHz, DMSO-d6) d (ppm) 42.75, 100.08,
115.87, 117.13, 123.18, 123.82, 124.23, 128.75, 129.16,
129.64, 131.35, 131.83, 140.63, 145.69, 150.98, 151.23,
161.92. ES/MS m/z 361 (M+H)+. Anal. Calcd for
C18H13F3N2O3: C, 59.67; H, 3.62; N, 7.73. Found: C,
59.59; H, 3.70; N, 7.76.
4.2.4. N-(Cyanoacetyl)-(4-trifluoromethyl)benzylamide (3b).
A mixture of methyl cyanoacetate (5.66 g, 57.1 mmol)
and 4-(trifluoromethyl)benzylamine (10.0 g, 57.1 mmol)
was heated at 100 ꢁC for 6 h with no condenser. The
resulting solid was triturated with 12 mL of 95% ethanol
and the product filtered as a yellow solid (11.0 g, 73%
1
yield). H NMR (300 MHz, CDCl3) d (ppm) 3.43 (2H,
s), 4.53 (2H, d, J = 6 Hz), 6.60 (1H, br s), 7.40 (2H, d,
J = 8.1 Hz), 7.62 (2H, d, J = 8.1 Hz).
4.2. Chemicals
4.2.1. Combinatorial library. A chemical library of vari-
ants of AG490 was constructed in radio frequency-
tagged reactors by a directed sorting method that
incorporated variability at three sites within the mole-
cule as described45 (see Fig. 1B). Twenty different R1
groups, 10 R2 groups, and 3 R3 groups were introduced
to make a total complexity of 600 variants (20 ·
10 · 3 = 600). One of the variants was lost because of
a failed synthesis leaving 599 different compounds.
Thirty of the compounds, selected at random were ana-
4.2.5. 2-Propenamide, 2-cyano-3-(3,4-dihydroxyphenyl)-
N-[[4-(trifluoromethyl)phenyl] methyl]-(E) (8B9) (NSC
D722756). The amide 3b from the previous reaction
(6.0 g,
24.77 mmol),
3,4-dihydroxybenzaldehyde
(3.31 g, 23.96 mmol), and piperidine (10 drops) were re-
fluxed in 95% ethanol (36 mL) for 1 h. The solution was
cooled and 72 mL of water was added to precipitate the
product. The solid was filtered, washed with 4 · 40 mL
of water and evaporated under high vacuum to give
6.70 g (75% yield) of the desired product 8B9, as a yel-
low solid. 1H NMR (300 MHz, acetone-d6) d (ppm)
4.67 (d, 2H, J = 6.0 Hz), 6.97 (d, 1H, J = 8.4 Hz), 7.38
(d, 1H, J = 8.4 Hz), 7.59–7.73 (m, 6H), 8.10 (s, 1H).
13C NMR (75 MHz, DMSO-d6) d (ppm) 42.75, 100.10,
115.88, 116.29, 117.18, 123.16, 124.93, 125.35, 125.64,
127.35, 127.74, 128.23, 144.02, 145.71, 151.02, 161.92.
ES/MS m/z 361 (M+H)+. Anal. Calcd for
C18H13F3N2O3Æ0.15ÆH2O: C, 59.23; H, 3.67; N, 7.67.
Found: C, 59.05; H, 3.70; N, 7.90.
1
lyzed by electron spray mass spectrometry, H NMR
and gravimetric analysis as described45 to confirm the
accuracy of the syntheses (data not shown). Each of
the 30 samples had molecular masses matching those
predicted by the synthetic scheme. Twenty-one of the
samples were greater than 80% pure; seven had purity
between 50% and 80%; and two were less than 50%
pure. The average yield of the syntheses was 15.2
4.8 lmol. The specific precursors used for the syntheses
are detailed in the Supplementary data. AG490 was
purchased from Sigma and from A. Gazit, Hebrew
University, Jerusalem.
4.3. Screening and cell proliferation assays
Each of the compounds of the combinatorial library or
AG490 were initially dissolved in DMSO at 1.5 mM.
Subsequent dilutions were made in culture media. Ten
microliters aliquots of appropriate dilutions were added
to 100 lL cultures of factor-dependent 2E8 cells
(100,000 per well) and Baf/3 cells (20,000 per well). After
4.2.2. N-(Cyanoacetyl)-(3-trifluoromethyl)benzylamide (3a).
A mixture of methyl cyanoacetate (8.49 g, 85.6 mmol)
and 3-(trifluoromethyl)benzylamine (15.0 g, 85.6 mmol)
was heated at 100 ꢁC for 6 h with no condenser. The
resulting solid was triturated with 20 mL of 95% ethanol
and the product filtered as a yellow solid (15.65 g, 75%
yield). H NMR (300 MHz, CDCl3) d (ppm) 3.43 (2H,
s), 4.52 (2H, d, J = 6 Hz), 6.60 (1H, br s), 7.40–7.60
(4H, m).
3
48 h, 0.5 lCi of H-thymidine (Perkin Elmer Life Sci-
1
ences) in 25 lL of media was added to each well and
the cultures were incubated for another 8 h. The cells
were then lysed by hypotonic shock and DNA was
captured onto glass fiber filters using an automated
harvester (Tomtec). Incorporated 3H-thymidine was
quantitated by scintillation counting (Wallac, Model
1450 MicroBeta TriLux scintillation and luminescence
counter). Statistical analysis was performed using
GraphPad Prism software.
4.2.3. 2-Propenamide, 2-cyano-3-(3,4-dihydroxyphenyl)-
N-[[3-(trifluoromethyl)phenyl] methyl]-(E) (5H4) (NSC
D722757). The amide 3a from the previous reaction
(5.0 g, 20.65 mmol), 3,4-dihydroxybenzaldehyde (2.76
g, 20.0 mmol), and piperidine (five drops) were refluxed
in 95% ethanol (30 mL) for 1 h. The solution was cooled
and 60 mL of water was added to precipitate the prod-
uct. The solid was filtered, washed with 4 · 30 mL of
water and evaporated under high vacuum to give
6.15 g (82% yield) of the desired product 5H4, as a yel-
4.4. Flow cytometry analysis
Cells (1 · 106) were resuspended in 0.5 mL solution A
(4 mM sodium citrate pH 7.8, 0.1% Triton X-100, 0.1

L. Gu et al. / Bioorg. Med. Chem. 13 (2005) 4269–4278
4277
mg/mL RNase A, 50 lg/mL propidium iodide). After
incubation at room temperature for 10 min, 0.5 mL of
solution B (0.4 M sodium chloride, 0.1% Triton X-100,
50 lg /mL propidium iodide) was added and mixed well.
Cells were stored at 4ꢁ for 2 h and subsequently
subjected to fluorescent flow cytometry analysis with a
Becton Dickinson FACScan machine. Data were
collected from 1 · 105 cells. Images were generated and
sub-G1 cell populations were counted using Cellquest
software.
toxylin and eosin by the Rodent Histopathology core
facility of Dana-Farber/Harvard Cancer Center. Micro-
scopic images were captured with a Spot camera (Diag-
nostic Instruments, Inc.) using a Nikon Eclipse E600
microscope (Nikon Corporation) and processed with
Spot 3.5.5 software (Diagnostic Instruments, Inc.).
Acknowledgments
We thank Ven Narayanan for assistance with chemistry,
Roderick Bronson for assistance with histology and Ma-
saya Takahashi for assistance with MRI analyses. This
work was supported by The Leukemia & Lymphoma
Society, The Dermatology Foundation, and the Drug
Synthesis and Chemistry Branch of the National Cancer
Institute under the Rapid Access to Intervention Devel-
opment (RAID) program.
4.5. Immunoprecipitation and immunoblot analysis
Cells (1 · 106) were lysed in 0.5 mL lysis buffer [10 mM
Tris–HCl pH 8.0, 5 mM EDTA, 150 mM NaCl, 0.5%
Nonidet P-40, 50 mM NaF, 1 mM sodium ortho-vana-
date, 1· protease inhibitor cocktail (Roche)] on ice,
sonicated, and centrifuged at 10,000g for 10 min at
4ꢁ. Supernatant containing 500 lg protein was
precleared with 20 lL of protein A/G plus-agarose
beads (Santa Cruz) and then incubated with 5 lg of
anti-JAK3 antibody (Upstate Biotechnology) and
20 lL of protein A/G plus-agarose beads at 4ꢁ over-
night. The beads were washed three times with lysis
buffer and separated on 4–20% SDS-PAGE and
transferred to PVDF membrane (Millipore). The mem-
brane was probed with anti-phosphotyrosine antibody
(4G10, Upstate Biotechnology), followed by goat
anti-mouse IgG-HRP secondary antibody (Upstate
Biotechnology). Complexes were detected by Super-
Signal West Pico chemiluminescent substrate (Pierce
Biotechnology).
Supplementary data
Three files showing the precursors used in the combina-
torial synthesis are provided as supplementary files. Sup-
plementary data associated with this article can be
found, in the online version at doi:10.1016/j.bmc.
2005.04.022.
References and notes
1. Hilkens, C. M.; Isꢀharc, H.; Lillemeier, B. F.; Strobl, B.;
Bates, P. A.; Behrmann, I.; Kerr, I. M. FEBS Lett. 2001,
505, 87.
2. Zhou, Y. J.; Chen, M.; Cusack, N. A.; Kimmel, L. H.;
Magnuson, K. S.; Boyd, J. G.; Lin, W.; Roberts, J. L.;
Lengi, A.; Buckley, R. H.; Geahlen, R. L.; Candotti, F.;
Gadina, M.; Changelian, P. S.; OꢀShea, J. J. Mol. Cell.
2001, 8, 959.
3. Saharinen, P.; Takaluoma, K.; Silvennoinen, O. Mol. Cell.
Biol. 2000, 20, 3387.
4. Yeh, T. C.; Dondi, E.; Uze, G.; Pellegrini, S. Proc. Natl.
Acad. Sci. U.S.A. 2000, 97, 8991.
4.6. Treatment of mice with 5H4
Age-matched groups of 12–16 week old IL7TG mice8
with palpable lymphadenopathy were given daily intra-
peritoneal injections of 5H4 dissolved in 0.1 mL of
DMSO or DMSO alone according to an institutional
review board approved protocol.
4.7. Magnetic resonance imaging
5. Schlessinger, J.; Lemmon, M. A. Sci. STKE 2003, 2003,
RE12.
6. Qin, J. Z.; Zhang, C. L.; Kamarashev, J.; Dummer, R.;
Burg, G.; Dobbeling, U. Blood 2001, 98, 2778.
7. Barata, J. T.; Cardoso, A. A.; Nadler, L. M.; Boussiotis,
V. A. Blood 2001, 98, 1524.
8. Rich, B. E.; Campos-Torres, J.; Tepper, R. I.; Moreadith,
R. W.; Leder, P. J. Exp. Med. 1993, 177, 305.
9. Fisher, A. G.; Burdet, C.; LeMeur, M.; Haasner, D.;
Gerber, P.; Ceredig, R. Leukemia 1993, 7(Suppl. 2), S66.
10. Fehniger, T. A.; Suzuki, K.; VanDeusen, J. B.; Cooper,
M. A.; Freud, A. G.; Caligiuri, M. A. Blood Cells Mol.
Dis. 2001, 27, 223.
11. Baba, H.; Yamada, Y.; Mori, N.; Hayashibara, T.;
Harasawa, H.; Tsuruda, K.; Sugahara, K.; Soda, H.;
Takasaki, Y.; Tawara, M.; Hirakata, Y.; Tomonaga, M.;
Kamihira, S. Eur. J. Haematol. 2002, 68, 362.
12. Plate, J. M.; Knospe, W. H.; Harris, J. E.; Gregory, S. A.
Hum. Immunol. 1993, 36, 249.
Magnetic resonance images of wild-type and IL7TG
mice were obtained by proton NMR scanning under
inhaled isofluorane anesthesia with respiratory gating
in an 8.5 T 9 cm vertical bore Bruker DRX magnet with
a 3 cm diameter surface coil at 0.5 mm resolution with
the support of the staff at the Center for Basic Magnetic
Resonance Research of the Beth Israel Deaconess
Medical Center.
MRI images were analyzed with Osiris imaging software
(Digital Imaging Unit, Switzerland). The volumes of
lymph nodes were calculated as the sums of the cross
sectional areas of the organ from each slice in which it
is visible, multiplied by the distance between slices
(0.5 mm).
13. Frishman, J.; Long, B.; Knospe, W.; Gregory, S.; Plate, J.
J. Exp. Med. 1993, 177, 955.
4.8. Histology
14. Long, B. W.; Witte, P. L.; Abraham, G. N.; Gregory, S.
A.; Plate, J. M. Proc. Natl. Acad. Sci. U.S.A. 1995, 92,
1416.
Tissue samples were fixed in 10% buffered formalin,
embedded in paraffin, sectioned, and stained with hema-

4278
L. Gu et al. / Bioorg. Med. Chem. 13 (2005) 4269–4278
15. Benjamin, D.; Sharma, V.; Knobloch, T. J.; Armitage, R.
J.; Dayton, M. A.; Goodwin, R. G. J. Immunol. 1994, 152,
4749.
16. Foss, H. D.; Hummel, M.; Gottstein, S.; Ziemann, K.;
Falini, B.; Herbst, H.; Stein, H. Am. J. Pathol. 1995, 146,
33.
17. Mattei, S.; Colombo, M. P.; Melani, C.; Silvani, A.;
Parmiani, G.; Herlyn, M. Int. J. Cancer. 1994, 56, 853.
18. Migone, T. S.; Lin, J. X.; Cereseto, A.; Mulloy, J. C.;
OꢀShea, J. J.; Franchini, G.; Leonard, W. J. Science 1995,
269, 79.
19. Xu, X.; Kang, S. H.; Heidenreich, O.; Okerholm, M.;
OꢀShea, J. J.; Nerenberg, M. I. J. Clin. Invest. 1995, 96,
1548.
Finco-Kent, D.; Baker, D. L.; Larson, M.; Si, M. S.;
Paniagua, R.; Higgins, J.; Holm, B.; Reitz, B.; Zhou, Y. J.;
Morris, R. E.; OꢀShea, J. J.; Borie, D. C. Science 2003,
302, 875.
31. Meydan, N.; Grunberger, T.; Dadi, H.; Shahar, M.;
Arpaia, E.; Lapidot, Z.; Leeder, J. S.; Freedman, M.;
Cohen, A.; Gazit, A.; Levitzki, A.; Roifman, C. M. Nature
1996, 379, 645.
32. Levitzki, A. Eur. J. Cancer 2002, 38, S11.
33. Kirken, R. A.; Erwin, R. A.; Taub, D.; Murphy, W. J.;
Behbod, F.; Wang, L.; Pericle, F.; Farrar, W. L. J.
Leukocyte Biol. 1999, 65, 891.
34. Wang, L. H.; Kirken, R. A.; Erwin, R. A.; Yu, C. R.;
Farrar, W. L. J. Immunol. 1999, 162, 3897.
20. Lacronique, V.; Boureux, A.; Valle, V. D.; Poirel, H.;
Quang, C. T.; Mauchauffe, M.; Berthou, C.; Lessard, M.;
Berger, R.; Ghysdael, J.; Bernard, O. A. Science 1997, 278,
1309.
21. Peeters, P.; Raynaud, S. D.; Cools, J.; Wlodarska, I.;
Grosgeorge, J.; Philip, P.; Monpoux, F.; Van Rompaey,
L.; Baens, M.; Van den Berghe, H.; Marynen, P. Blood
1997, 90, 2535.
22. Levitzki, A.; Gilon, C. Trends Pharmacol. Sci. 1991, 12,
171.
23. Levitzki, A. Pharmacol. Ther. 1999, 82, 231.
24. Yamashita, N.; Kazuo, S. Y.; Kitamura, M.; Wakao, H.;
Furihata, K.; Hayakawa, Y.; Miyajima, A.; Seto, H. J.
Antibiot. (Tokyo) 1997, 50, 440.
35. Posner, I.; Engel, M.; Gazit, A.; Levitzki, A. Mol.
Pharmacol. 1994, 45, 673.
36. Kaur, G.; Gazit, A.; Levitzki, A.; Stowe, E.; Cooney, D.
A.; Sausville, E. A. Anticancer Drugs 1994, 5, 213.
37. Nielsen, M.; Kaltoft, K.; Nordahl, M.; Ropke, C.; Geisler,
C.; Mustelin, T.; Dobson, P.; Svejgaard, A.; Odum, N.
Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 6764.
38. De Vos, J.; Jourdan, M.; Tarte, K.; Jasmin, C.; Klein, B.
Br. J. Haematol. 2000, 109, 823.
39. Eriksen, K. W.; Kaltoft, K.; Mikkelsen, G.; Nielsen, M.;
Zhang, Q.; Geisler, C.; Nissen, M. H.; Ropke, C.; Wasik,
M. A.; Odum, N. Leukemia 2001, 15, 787.
40. Epling-Burnette, P. K.; Liu, J. H.; Catlett-Falcone, R.;
Turkson, J.; Oshiro, M.; Kothapalli, R.; Li, Y.; Wang, J.
M.; Yang-Yen, H. F.; Karras, J.; Jove, R.; Loughran, T.
P., Jr. J. Clin. Invest. 2001, 107, 351.
25. Wasik, M. A.; Nowak, I.; Zhang, Q.; Shaw, L. M. Leuk.
Lymphoma 1998, 28, 551.
26. Sudbeck, E. A.; Liu, X. P.; Narla, R. K.; Mahajan, S.;
Ghosh, S.; Mao, C.; Uckun, F. M. Clin. Cancer Res. 1999,
5, 1569.
41. Kube, D.; Holtick, U.; Vockerodt, M.; Ahmadi, T.; Haier,
B.; Behrmann, I.; Heinrich, P. C.; Diehl, V.; Tesch, H.
Blood 2001, 98, 762.
27. Brown, G. R.; Bamford, A. M.; Bowyer, J.; James, D. S.;
Rankine, N.; Tang, E.; Torr, V.; Culbert, E. J. Bioorg.
Med. Chem. Lett. 2000, 10, 575.
42. Miyamoto, N.; Sugita, K.; Goi, K.; Inukai, T.; Lijima, K.;
Tezuka, T.; Kojika, S.; Nakamura, M.; Kagami, K.;
Nakazawa, S. Leukemia 2001, 15, 1758.
28. DꢀAlessio, R.; Bargiotti, A.; Carlini, O.; Colotta, F.;
Ferrari, M.; Gnocchi, P.; Isetta, A.; Mongelli, N.; Motta,
P.; Rossi, A.; Rossi, M.; Tibolla, M.; Vanotti, E. J. Med.
Chem. 2000, 43, 2557.
43. Danial, N. N.; Pernis, A.; Rothman, P. B. Science 1995,
269, 1875.
44. Sun, X.; Layton, J. E.; Elefanty, A.; Lieschke, G. J. Blood
2001, 97, 2008.
29. Thompson, J. E.; Cubbon, R. M.; Cummings, R. T.;
Wicker, L. S.; Frankshun, R.; Cunningham, B. R.;
Cameron, P. M.; Meinke, P. T.; Liverton, N.; Weng, Y.;
DeMartino, J. A. Bioorg. Med. Chem. Lett. 2002, 12,
1219.
45. Shi, S.; Xiao, X. Y.; Czarnik, A. W. Biotechnol. Bioeng.
1998, 61, 7.
46. Gazit, A.; Osherov, N.; Posner, I.; Yaish, P.; Poradosu, E.;
Gilon, C.; Levitzki, A. J. Med. Chem. 1991, 34, 1896.
47. Bright, J. J.; Du, C.; Sriram, S. J. Immunol. 1999, 162,
6255.
48. Gerebtzoff, G.; Li-Blatter, X.; Fischer, H.; Frentzel, A.;
Seelig, A. Chembiochem 2004, 5, 676.
49. Park, B. K.; Kitteringham, N. R.; OꢀNeill, P. M. Annu.
Rev. Pharmacol. Toxicol. 2001, 41, 443.
50. Grunberger, T.; Demin, P.; Rounova, O.; Sharfe, N.;
Cimpean, L.; Dadi, H.; Freywald, A.; Estrov, Z.; Roif-
man, C. M. Blood 2003, 102, 4153.
51. Pietrangeli, C. E.; Hayashi, S.; Kincade, P. W. Eur. J.
Immunol. 1988, 18, 863.
52. Palacios, R.; Henson, G.; Steinmetz, M.; McKearn, J. P.
Nature 1984, 309, 126.
53. Palacios, R.; Steinmetz, M. Cell 1985, 41, 727.
30. Changelian, P. S.; Flanagan, M. E.; Ball, D. J.; Kent, C.
R.; Magnuson, K. S.; Martin, W. H.; Rizzuti, B. J.;
Sawyer, P. S.; Perry, B. D.; Brissette, W. H.; McCurdy, S.
P.; Kudlacz, E. M.; Conklyn, M. J.; Elliott, E. A.;
Koslov, E. R.; Fisher, M. B.; Strelevitz, T. J.; Yoon, K.;
Whipple, D. A.; Sun, J.; Munchhof, M. J.; Doty, J. L.;
Casavant, J. M.; Blumenkopf, T. A.; Hines, M.; Brown,
M. F.; Lillie, B. M.; Subramanyam, C.; Shang-Poa, C.;
Milici, A. J.; Beckius, G. E.; Moyer, J. D.; Su, C.;
Woodworth, T. G.; Gaweco, A. S.; Beals, C. R.;
Littman, B. H.; Fisher, D. A.; Smith, J. F.; Zagouras,
P.; Magna, H. A.; Saltarelli, M. J.; Johnson, K. S.; Nelms,
L. F.; Des Etages, S. G.; Hayes, L. S.; Kawabata, T. T.;