Identification of 2-acylaminothiophene-3-carboxamides as potent inhibitors of FLT3

Article abstract of DOI:10.1016/j.bmcl.2006.03.032

A series of 2-acylaminothiophene-3-carboxamides has been identified which exhibit potent inhibitory activity against the FLT3 tyrosine kinase. Compound 44 inhibits the isolated enzyme (IC₅₀ = 0.027 μM) and blocks the proliferation of MV4-11 cel

Full text of DOI:10.1016/j.bmcl.2006.03.032

Bioorganic & Medicinal Chemistry Letters 16 (2006) 3282–3286
Identification of 2-acylaminothiophene-3-carboxamides
as potent inhibitors of FLT3
Raymond J. Patch, Christian A. Baumann, Jian Liu, Alan C. Gibbs,
Heidi Ott, Jennifer Lattanze and Mark R. Player*
Drug Discovery, Johnson & Johnson Pharmaceutical Research and Development, L.L.C., 8 Clarke Drive, Cranbury, NJ 08512, USA
Received 23 February 2006; revised 10 March 2006; accepted 10 March 2006
Available online 31 March 2006
Abstract—A series of 2-acylaminothiophene-3-carboxamides has been identified which exhibit potent inhibitory activity against the
FLT3 tyrosine kinase. Compound 44 inhibits the isolated enzyme (IC50 = 0.027 lM) and blocks the proliferation of MV4-11 cells
(IC50 = 0.41 lM). Structure–activity relationship studies within this series are described in the context of a proposed binding model
within the ATP binding site of the enzyme.
Ó 2006 Elsevier Ltd. All rights reserved.
Acute myeloid leukemia (AML) is the most common
form of adult leukemia and comprises approximately
20% of childhood leukemias.1 Current treatment for
AML typically involves treatment with general cytotoxic
agents such as Cytarabine (Ara-C) and Daunorubicin to
induce remission of disease. Induction therapy is fol-
lowed by a variety of post-induction treatments ranging
from high dose Cytarabine to bone marrow transplant.2
Such aggressive therapies are often poorly tolerated by
elderly patients. Additionally, since less than 20% of
diagnosed AML patients survive more than 4 years,
improved treatment paradigms are needed.
H
N
O
H
N
N
O
O
N
O
N
O
N
H
N
O
N
N
O
N
N
H
N
O
O
N
H
HO
OH
O
Midostaurin
Lestaurtinib
Tandutinib
Figure 1. FLT3 inhibitors in clinical development.
Activating mutations of the class III receptor tyrosine
kinase FLT3 have been detected in approximately 30%
of patients with acute myelogenous leukemia and a
small number of patients with acute lymphoblastic leu-
kemia or myelodysplastic syndrome.3 Patients with such
FLT3 mutations have a poor prognosis, with decreased
remission times and disease free survival.4 Furthermore,
>90% of blasts from AML patients express either wild-
type or mutant FLT3.5 This suggests that inhibition of
FLT3 may be of significant therapeutic benefit for
AML patients.
Reported FLT3 inhibitors under advanced clinical
development include the staurosporine structural ana-
logues, lestaurinib and midostaurin, the quinazoline,
tandutinib, and others.6 These exhibit varying degrees
of non-selectivity, particularly toward other receptor
tyrosine kinases, and as such, may present undesirable
side-effect profiles. Thus, the need still exists for further
development of potent and selective FLT3 inhibitors
(Fig. 1).
During the course of our screening efforts to discover
FLT3 inhibitors derived from alternative chemotypes,
we identified a class of 2-acylaminothiophene-3-carbox-
amides that exhibited potent inhibitory activities. Some
compounds of this class had been reported to potentially
inhibit a range of kinases in an earlier published patent
application; however, no data were provided to support
these claims, nor was FLT3 among the potentially
Keywords: FLT3; Acute myeloid leukemia; AML; Acylamino-3-
carboxamide.
*
Corresponding author. Tel.: +1 610 458 6980; e-mail: mplayer@
prdus.jnj.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.03.032

R. J. Patch et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3282–3286
3283
affected kinases cited.7 A more recent application cites
inhibition of a broad panel of kinases (including
FLT3) by various compounds within this class; howev-
er, data are presented for only a single inhibitory con-
centration (10 lM).8 Herein, we report on initial
structure–activity relationships for this series with the
goal of mapping out a general binding motif for the
inhibitory interaction with FLT3.9 This information
may serve in the development of FLT3 inhibitors exhib-
iting greater potencies and selectivity.
Table 2. FLT3 inhibitory activities for 2-phenacylaminothiophene-3-
carboxamides 11–29
R
O
S
NH
6, 11 - 29
O
H2N
Compound
R
FLT3 inhibition MV4-11
a
IC50 (lM)
proliferation
b
IC50 (lM)
6
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
H
0.36 ( 0.07)
0.052 ( 0.01)
0.12 ( 0.03)
0.31 ( 0.05)
0.31 ( 0.13)
0.59 ( 0.18)
0.63 ( 0.09)
>10
>10
0.45
1.3
A survey of simple 2-acylaminothiophene-3-carboxa-
mides (Table 1) revealed a minimal tolerance for steric
bulk at the carbon adjacent to the carbonyl of the sec-
ondary amide and a preference for sp2 hybridization at
this center. Thus, while acetamido and propionamido
analogues 1 and 2, respectively, inhibited FLT3 at
low-micromolar levels, larger alkyl amides resulted in re-
duced activities. Of particular interest are the contrast-
ing activities of the pseudo sp2 isopropylcarboxamide
3 and cyclopropylcarboxamide 4; the former exhibiting
greatly diminished activity. Cyclohexylcarboxamide 5
was inactive at a concentration of 10 lM; however, its
aromatic counterpart, 6 and 2-furoylamide 7, exhibited
significantly enhanced activities (IC50 = 0.36, 0.17 lM,
respectively).10 Pyridine carboxamides (8–10) exhibited
a range of inhibitory potencies against the isolated en-
zyme with p- > o- > m-isomer (10 > 8 > 9), an order
which may reflect important electronic ring effects as op-
posed to specific electrostatic interactions with the
enzyme.
4-Me
4-OMe
4-Cl
1.4
1.5
4-Br
4-Ac
4-t-Bu
4-Ph
1.8
1.6
4-OPh
3-Me
3-OMe
3-Cl
>10
0.14 ( 0.03)
0.11 ( 0.06)
0.63 ( 0.10)
3.0
3.2
1.5
2.2
2-Me
2-OMe
2-Cl
>10
>10
3,4-Di-OMe
3,4-Methylenedioxy 0.14 ( 0.01)
0.042 ( 0.03)
0.34
0.89
>10
3,5-Di-OMe
3,4,5-Tri-OMe
3,4-Di-Me
1.7 ( 0.62)
>10
0.54 ( 0.40)
0.71
a Values are means of three experiments, standard deviation is given in
parentheses.
b All dose–response data are calculated from an average of three
replicates per dose.
Substituent effects were evaluated on benzoylamide 6
(Table 2). Both steric and electronic factors affected
activity in this series. Small substituents (methyl, 11;
methoxyl, 12) at the 4-position afforded increased
FLT3 enzyme inhibition. Somewhat larger substituents
were tolerated (halo, 13, 14; acetyl, 15; t-butyl, 16),
but with increasing size activity dropped precipitously
(phenyl, 17; phenoxyl, 18). Similar effects on substitu-
tion of the 3-position were observed, with methyl and
methoxyl substituents providing increased activity (19,
20; IC50s = 0.14 and 0.11 lM, respectively). Ortho
substitution was uniformly detrimental to activity, likely
due to induced conformational effects away from
co-planarity.
Table 1. FLT3 inhibitory activities for 2-acylaminothiophene-3-carb-
oxamides 1–10
O
R
S
NH
Interestingly, while enhanced inhibitory activity was ob-
tained with 3,4-dimethoxybenzoylamide 25 (IC50
0.042 lM), 3,4,5-trimethoxybenzoylamide 28 was
devoid of activity.
1 - 10
=
O
H2N
Compound R
FLT3 inhibition MV4-11 proliferation
a
b
IC50 (lM)
IC50 (lM)
The introduction of spacer atoms between the aryl ring
and carbonyl carbon was also investigated (Table 3).
The results are consistent with a requirement of coplan-
arity between these two groups. Thus, saturated
homologues showed greatly reduced inhibitory activities
compared with vinyl analogues. Again, methoxyl
substitution was beneficial, particularly in the case of
3,4-dimethoxycinnamoyl analogue 39.
1
2
Methyl
Ethyl
2.0 ( 0.21)
2.1 ( 0.92)
>10
3
i-Propyl
c-Propyl
c-Hexyl
Phenyl
4
1.3 ( 0.37)
>10
0.36 ( 0.07)
1.2
5
6
>10
1.5
7
Furan-2-yl 0.17 ( 0.05)
Pyridin-2-yl 0.64 ( 0.10)
Pyridin-3-yl 1.1 ( 0.41)
Pyridin-4-yl 0.11 ( 0.05)
8
0.35
2.0
9
10
2.1
Structure–activity relationships of the bicyclic thio-
phene-3-carboxamide core were also investigated
(Table 4). Deletion of the fused cyclohexyl methylenes
(42) resulted in a 50-fold reduction in FLT3 inhibitory
activity, much of which could be recovered by
a Values are means of three experiments, standard deviation is given in
parentheses.
b All dose–response data are calculated from an average of three rep-
licates per dose.

3284
R. J. Patch et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3282–3286
Table 3. FLT3 inhibitory activities for homologated 2-acylaminothiophene-3-carboxamides 30–41
Ar
O
X
S
NH
30-41
O
H2N
a
b
Compound
X
Ar
FLT3 inhibition IC50 (lM)
MV4-11 proliferation IC50 (lM)
30
31
32
33
34
35
36
37
38
39
40
41
–CH2–
–CH2–O–
Ph
Ph
3.2 ( 1.1)
>10
>10
>10
–CH2–S–
–CH2–O–
Ph
4-MeOPh
Ph
3.8
ꢀ10
–CH2–CH2–
–CH2–CH2–
–CH@CH–
–CH@CH–
–CH@CH–
–CH@CH–
–CH@CH–
–CH@CH–
3.6 ( 0.76)
2.7
4-MeOPh
Ph
>10
0.53
1.8
0.25 ( 0.02)
0.25 ( 0.06)
0.49 ( 0.17)
0.13 ( 0.04)
0.56 ( 0.06)
1.0 ( 0.23)
4-MeOPh
4-i-PrPh
3,4-Di-OMe
Furan-2-yl
5-Methylfuran-2-yl
2.7
2.0
0.48
10
a Values are means of three experiments, standard deviation is given in parentheses.
b All dose–response data are calculated from an average of three replicates per dose.
of which exhibited greater anti-proliferative effects than
would have been anticipated from their FLT3 inhibitory
activities. This could be the result of enhanced cell perme-
abilities of these analogues; alternatively, it is possible
that additional anti-proliferative mechanisms were
involved in these cases.
Table 4. Effects of thiophene substitution upon FLT3 inhibitory
activities for 2-phenacylaminothiophene-3-carboxamides 42–46
OMe
O
R5
OMe
S
NH
R4
H2N
25, 42-46
O
In order to assess kinase selectivity, compound 25 was
tested against a standard panel of kinases (Table 5).
At a concentration of 3 lM, only the human Abl-related
gene kinase Arg(h) was inhibited at a level approaching
50%.
Compound R4, R5
FLT3 inhibition MV4-11 proliferation
a
b
IC50 (lM)
IC50 (lM)
42
43
44
25
45
46
H, H
2.1 ( 1.3)
2.4
Me, Me 0.14 ( 0.10)
–(CH2)3– 0.027 ( 0.01)
–(CH2)4– 0.042 ( 0.03)
–(CH2)5– 6.9 ( 2.9)
–(CH2)6– >100
0.33
0.41
0.34
Table 5. Kinase selectivity screen for compound 25
Kinase
% Inha
a Values are means of three experiments, standard deviation is given in
parentheses.
Abl(h)
21
À4
49
AMPK(r)
Arg(h)
Axl(h)
b All dose–response data are calculated from an average of three rep-
licates per dose.
À10
9
Bmx(h)
BTK(h)
14
incorporation of methyl groups at R4 and R5 (43). With-
in the series of fused alkylene ring homologues, activity
was inversely correlated with ring size, with optimal
activity residing with cyclopentyl-fused thiophene 44.
CDK2/cyclinA(h)
CDK7/cyclinH/MAT1(h)
CHK1(h)
c-RAF(h)
EphB2(h)
Fes(h)
À22
À1
À7
À1
À11
À12
33
In order to evaluate the functional activities within this
series, those inhibitors with approximate sub-micromolar
IC50s were assayed for antiproliferative effects on MV4-11
cells (a human acute myelogenous leukemia cell line
expressing a constitutively activated mutant FLT3).11
With only a few exceptions, activities between the two as-
says were well correlated. Thus, effective anti-prolifera-
tive concentrations (IC50s) were roughly 10-fold less
than corresponding inhibitory concentrations in the iso-
lated enzyme assay. Notable exceptions to this general
correlation were 2-pyridinylcarboxamide 8, dim-
ethylbenzoylamide 29, cinnamoylamide 36, furylacryla-
mide 40, and 4,5-dimethylthiophene analogue 43, each
Fyn(h)
GSK3ß(h)
JNK3(h)
Lyn(h)
À2
À6
21
Met(h)
13
3
PKCl(h)
PKD2(h)
Rsk1(h)
À7
3
TrkB(h)
Fes(h)
39
À12
a Compound 25 was tested at single concentration of 3 lM; ATP
concentration was 10 lM. Values are averages of duplicate
experiments.

R. J. Patch et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3282–3286
3285
The low level of inhibition of TrkB in this assay is also
noteworthy, given the reported potent inhibitory activity
of lestaurtinib against Trk isoenzymes A–C
(IC50 ꢀ 4 nM).12
To help rationalize the aforementioned structure–activi-
ty relationships and drive future syntheses, a homology
model of the FLT3 kinase domain was constructed and
compound docking studies were carried out. The kinase
domain of FGF receptor 1 (pdb Accession No. 2FGI)
was chosen as the template for FLT3 homology model-
ing because of the high homology and identity (66% and
50%, respectively) between the two enzymes.
Multiple conformations of representative compounds
from this series were generated, each with an internal
hydrogen bond between the primary amide CO and
the secondary amide NH fixed in place (Fig. 2).13 The
conformers were then subjected to rigid body docking
in order to determine a low-energy binding mode. (In
these studies, we assume that the intramolecular hydro-
gen bond is retained upon enzyme binding.) Figure 2
illustrates a common low-energy binding mode, which
correlates well with the SARs observed for this series
(compound 25). The figure shows a cut-away of the
active site between the N- and C-terminal lobes of the
kinase. The compounds are oriented so that the primary
amide engages in a hydrogen bonding interaction with
the backbone carbonyl of glutamate 692—the so-called
‘hinge region’ of receptor tyrosine kinases. The second-
ary amide projects its carboxylate-derived substituent
into a narrow cleft underneath phenylalanine 691 (in
the region of ATP binding). This is consistent with a
preference for aromatic or extended aromatic (vinylaryl)
systems that could participate in p-interactions with Phe
691 while at the same time avoiding steric clashes in this
narrow region. This binding mode also illustrates the
preference for (1) co-planarity of secondary amide
substituents and (2) a free primary amide capable of
hydrogen bonding to the hinge region of the backbone.
Figure 3. Cut-away illustration of the active site with superimpositions
of low-energy binding modes for lestaurtinib (yellow) and compound
25 (cyan). Hydrogen bonds between the hinge region of the backbone
and lestaurtinib are shown in green; that with compound 25 is shown
in blue. Only polar hydrogens are shown.
We were interested in comparing this hypothetical bind-
ing mode with one that might be predicted for the FLT3
inhibitor, lestaurtinib. Docking lestaurtinib into FLT3
(Fig. 3) revealed a preferred orientation similar to many
binding modes of staurosporine-like compounds found
in kinase co-crystal structures (e.g., pdb Accession
Nos. 1AQ1, 1BYG, and 1NVQ). Lestaurtinib binds in
a typical bi-dentate fashion to the protein backbone
(Glu 692, Tyr 693), whereas the acylaminothiophene-
carboxamides bind with a mono-dentate interaction
(Glu 692 only). However, the relative positions of the
molecules within the active site are very similar. Both
chemotypes are structurally quite flat and occupy the
same plane within the active site. As shown in Figure
3, the ring systems of both molecules overlap significant-
ly; however, the methoxyl substituents of the benzamide
allow for further extension into the binding region than
is accessed by lestaurtinib’s unsubstituted indolocarbaz-
ole core. Such substitution may confer a different kinase
selectivity profile on this chemotype than is observed
with staurosporine-based molecules. Finally, this mod-
eled overlap highlights potential regions where future
synthetic plans might be directed to afford FLT3 inhib-
itors with greater potencies and enhanced clinical
effectiveness.
References and notes
1. (a) Stone, R. M.; O’Donnell, M. R.; Sekeres, M. A.
Hematology (Am. Soc. Hematol. Educ. Program) 2004,
98; (b) Gilliland, D. G.; Jordan, C. T.; Felix, C. A.
Hematology (Am. Soc. Hematol. Educ. Program) 2004, 80.
2. (a) Tallman, M. S.; Gilliland, D. G.; Rowe, J. M. Blood
2005, 106, 1154; (b) Buchner, T.; Berdel, W. E.; Wormann,
B.; Schoch, C.; Haferlach, T.; Schnittger, S.; Kern, W.;
Aul, C.; Lengfelder, E.; Schumacher, A.; Reichle, A.;
Staib, P.; Balleisen, L.; Eimermacher, H.; Gruneisen, A.;
Rasche, H.; Sauerland, M. C.; Heinecke, A.; Mesters, R.
M.; Serve, H. L.; Kienast, J.; Hiddemann, W. Crit. Rev.
Oncol. Hematol. 2005, 56, 247.
Figure 2. Cut-away illustration of the active site between the N- and
C-terminal lobes (magenta). Predicted low-energy binding mode of
compound 25 (cyan). Intra-molecular (red dash) and inter-molecular
(blue dash) hydrogen bonds are shown. Only ligand polar hydrogens
are shown.

3286
R. J. Patch et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3282–3286
3. Stirewalt, D. L.; Radich, J. P. Nat. Rev. Cancer 2003, 3, 650.
Tyrosine Kinase Kit (Green) supplied by Invitrogen. The
FLT3 kinase reaction is incubated at room temperature
for 30 min under the following conditions: 10 nM FLT3
571-993, 20 lg/mL poly Glu4Tyr, 150 lM ATP, 5 mM
MgCl2, and 1% compound in DMSO. The kinase reaction
is stopped with the addition of EDTA. The fluorescein-
labeled phosphopeptide and the anti-phosphotyrosine
antibody are added and incubated for 30 min at room
temperature and polarization was read.
4. (a) Abu-Duhier, F. M.; Goodeve, A. C.; Wilson, G. A.;
Gari, M. A.; Peake, I. R.; Rees, D. C.; Vandenberghe, E.
A.; Winship, P. R.; Reilly, J. T. Br. J. Haematol. 2000,
111, 190; (b) Kiyoi, H.; Naoe, T.; Nakano, Y.; Yokota, S.;
Minami, S.; Miyawaki, S.; Asou, N.; Kuriyama, K.;
Jinnai, I.; Shimazaki, C.; Akiyama, H.; Saito, K.; Oh, H.;
Motoji, T.; Omoto, E.; Saito, H.; Ohno, R.; Ueda, R.
Blood 1999, 93, 3074.
5. (a) Rosnet, O.; Buhring, H. J.; Marchetto, S.; Rappold, I.;
Lavagna, C.; Sainty, D.; Arnoulet, C.; Chabannon, C.;
Kanz, L.; Hannum, C.; Birnbaum, D. Leukemia 1996, 10,
238; (b) Birg, F.; Courcoul, M.; Rosnet, O.; Bardin, F.;
Pebusque, M. J.; Marchetto, S.; Tabilio, A.; Mannoni, P.;
Birnbaum, D. Blood 1992, 80, 2584.
6. (a) Levis, M.; Small, D. Curr. Pharm. Des. 2004, 10, 1183;
(b) Advani, A. S. Curr. Pharm. Des. 2005, 11, 3449.
7. Fancelli, D.; Pevarello, P.;Varasi, M. Int. Patent Appl.
WO 01/98290, 2001.
11. MV4-11 cell-based assay. (a) Quentmeier, H.; Reinhardt,
J.; Zaborski, M.; Drexler, H. G. Leukemia 2003, 17, 120(b)
MV4-11 (ATCC Number: CRL-9591) cells were plated at
10,000 cells per well in 100 lL of in RPMI media
containing penn/strep, 10% FBS, and 0.2 ng/mL GM-
CSF. Compound dilutions or 0.1% DMSO (vehicle
control) is added to cells and the cells are allowed to
grow for 72 h at standard cell growth conditions (37 °C,
5% CO2). To measure total cell growth, an equal volume
of CellTiterGlo reagent (Promega, Madison, WI) was
added to each well, according to the manufacturer’s
instructions, and luminescence was quantified. Total cell
growth was quantified as the difference in luminescent
counts (relative light units, RLU) of cell number at Day 0
compared to total cell number at Day 3 (72 h of growth
and/or compound treatment). All IC50 values are calcu-
lated in GraphPadPrism using non-linear regression anal-
ysis with a multiparameter (variable slope) equation.
12. (a) Camoratto, A. M.; Jani, J. P.; Angeles, T. S.; Maroney,
A. C.; Sanders, C. Y.; Murakata, C.; Neff, N. T.; Vaught, J.
L.; Isaacs, J. T.; Dionne, C. A. Int. J. Cancer 1997, 72, 673;
(b) George, D. J.; Dionne, C. A.; Jani, J.; Angeles, T.;
Murakata, C.; Lamb, J.; Issacs, J. T. Cancer Res. 1999, 59,
2395.
8. Hodge, C. N.; Janzen, W. P.; Williams, K. P.; Cheatham,
L. A. Int. Patent Appl. WO 05/033102, 2005.;
9. Compounds for this study could be readily synthesized
from the corresponding 2-aminothiophene-3-carboxamide
by reaction with the appropriate acid chloride under basic
conditions as follows: To a solution of the acid chloride
(0.5 mmol) in THF (2 mL) was added the 2-aminothioph-
ene-3-carboxamide (0.5 mmol) followed by DIEA
(1 mmol), and the mixture was stirred at room tempera-
ture for 2 h. The mixture was then diluted with ether and
filtered. The resulting precipitate was suspended in DCM,
extracted with water (to remove DIEAÆHCl), and filtered
1
again to afford the pure product (LCMS; H NMR).
10. (FLT3 kinase assay protocol). To determine the activity of
the compounds of the present invention in an in vitro
kinase assay, inhibition of the isolated kinase domain of
the human FLT3 receptor (a.a. 571-993) was performed
using the following fluorescence polarization (FP) proto-
col. The FLT3 fluorescence polarization assay utilizes the
fluorescein-labeled phosphopeptide and the anti-phospho-
tyrosine antibody included in the Panvera Phospho-
13. Correlations between proton NMR chemical shifts and
calculated proton magnetic shielding tensors (via DFT
calculations at the B3LYP, 6-311+G[2d,p] level of theory)
support the existence of this internal hydrogen bond for
compounds within this class. This is manifested primarily
by extensive deshielding of the secondary amide NH
(d ꢀ 13 ppm in DMSO-d6).