Synthesis and evaluation of 4-anilino-6, 7-dialkoxy-3-quinolinecarbonitriles as inhibitors of kinases of the Ras-MAPK signaling cascade

Article abstract of DOI:10.1016/S0960-894X(03)00640-1

4-[3-Chloro-4-(1-methyl-1H-imidazol-2-ylsulfanyl)]anilino-6, 7-diethoxy-3-quinolinecarbonitrile (3) was identified as a MEK1 kinase inhibitor with exceptional activity against LoVo cells. The structure-activity relationships of the C-4 aniline substituents were explored, and water-solubilizing groups were added at the C-7 position to improve physical properties. Secondary cellular assays revealed that a compound possessing the appropriate aniline substituents inhibited MEK1 as well as MAPK phosphorylation, thereby acting as a dual inhibitor of the Ras-MAPK signaling cascade.

Full text of DOI:10.1016/S0960-894X(03)00640-1

Bioorganic & Medicinal Chemistry Letters 13 (2003) 3031–3034
Synthesis and Evaluation of 4-Anilino-6,7-dialkoxy-3-
quinolinecarbonitriles as Inhibitors of Kinases of the
Ras-MAPK Signaling Cascade
Dan Berger,a,* Minu Dutia,a Dennis Powell,a Biqi Wu,a Allan Wissner,a
Diane H. Boschelli,a M. Brawner Floyd,a Nan Zhang,a Nancy Torres,a Jeremy Levin,a
Xuemei Du,a Donald Wojciechowicz,b Carolyn Discafani,b Constance Kohler,b
Steven C. Kim,b Larry R. Feldberg,b Karen Collinsb and Robert Mallonb
aChemical Sciences, Wyeth Research, Pearl River, NY 10965, USA
bDiscovery Oncology, Wyeth Research, Pearl River, NY 10965, USA
Received 13 March 2003; accepted 8 April 2003
Abstract—4-[3-Chloro-4-(1-methyl-1H-imidazol-2-ylsulfanyl)]anilino-6,7-diethoxy-3-quinolinecarbonitrile (3) was identified as a
MEK1 kinase inhibitor with exceptional activity against LoVo cells. The structure–activity relationships of the C-4 aniline sub-
stituents were explored, and water-solubilizing groups were added at the C-7 position to improve physical properties. Secondary
cellular assays revealed that a compound possessing the appropriate aniline substituents inhibited MEK1 as well as MAPK phos-
phorylation, thereby acting as a dual inhibitor of the Ras-MAPK signaling cascade.
# 2003 Elsevier Ltd. All rights reserved.
The Ras-MAPK signaling cascade transmits mitogenic
stimuli from growth factor receptors and activated Ras
to the cell nucleus. Research on the Ras-MAPK signal-
ing pathway identified Raf as the downstream effector of
Ras.1 Activated Raf phosphorylates two serine residues
on MAP kinase kinase (MEK1), which in turn activates
Mitogen Activated Protein Kinases (MAPK) by phos-
phorylation of threonine and tyrosine residues.2,3
Activated MAPK translocates to the nucleus, mediating
cell growth and proliferation through the phosphoryla-
tion of various substrates.4 Inappropriate Ras activa-
tion is associated with ꢀ30% of all human cancers,
making the kinase components of the Ras-MAPK sig-
naling cascade attractive targets for pharmaceutical
intervention.5
Researchers at Pfizer (formerly Parke-Davis) identified
1 as a potent MEK1 kinase inhibitor with oral activity
in vivo against selected tumor lines.6 This compound,
currently known as CI-1040, has entered clinical trials.7
Recently, researchers at Wyeth have reported 4-anilino-
3-quinolinecarbonitrile analogues with potent activity
against MEK1 kinase, such as 2.8,9 The activity of these
inhibitors was determined by a direct MEK1-MAPK
fluorescent ELISA assay.8
We have developed a high throughput, non-radioactive
Raf/MEK1/MAPK cascade ELISA assay to detect Raf
and/or MEK1 kinase inhibitors.10 For this assay, the
end point was the measurement of MAPK phosphor-
ylation on Threonine (T) 202 and Tyrosine (Y) 204 by a
phospho-specific monoclonal antibody. The specific
*Corresponding author. Tel.: +1-845-602-3435; fax: +1-845-602-
5561; e-mail: bergerd@wyeth.com
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00640-1

3032
D. Berger et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3031–3034
target of Raf/MEK1/MAPK ELISA inhibition was
identified by enzymatic assays which directly measured
Raf phosphorylation of MEK1, or MEK1 phosphor-
ylation of MAPK.10 Using this system, 3 was identified
as a potent MEK1 inhibitor (IC50: 10 nM), with excep-
tional activity (IC50: 7 nM) against LoVo cells (a K-ras
mutant human colon tumor line).
Because of its potent cellular activity, compound 3
served as a particularly promising new lead structure.
We present here our exploration of the SAR of a series
of analogues of 3, as well as our efforts to improve the
aqueous solubility of this compound. Specifically, we
varied the substitution on the aniline ‘headpiece’ to
determine how this would affect both enzyme and cel-
lular activity. Based on the SAR of our previously
described MEK1 kinase inhibitors, it was anticipated
that the C-7 position of the 4-anilino-3-quinolinecarbo-
nitriles would be optimal for the addition of water-
solubilizing substituents. It was anticipated that such
substituents could improve the physical properties of 3,
which was highly insoluble in water (<0.1 mg/mL) at
pH 7.4, as well as a variety of organic solvents. Also of
interest was to determine why compound 3 was sig-
nificantly more active against LoVo cells than 2, while
possessing equivalent MEK1 kinase activity (Table 1).
Scheme 1. (a) HS-Het, NaH, DMF or NMP; (b) Fe/NH4Cl, MeOH,
H2O.
The synthesis of aniline 5i is outlined in Scheme 2.
Compound 612 was converted to intermediate 7 by
reaction with ethanolic HCl, followed by treatment with
2-aminoacetaldehyde dimethyl acetal and the sub-
sequent addition of acetic acid and HCl gas.13 Heating 7
in toluene with 1.5 equiv of dimethyl sulfate provided 8,
which was reduced with iron/ammonium chloride to
afford 5i.
Scheme 1 outlines the methods by which the majority of
the substituted anilines were constructed. The fluoride
substituents of 4a–d were readily displaced by a variety
of heterocyclic thiols to provide intermediates that were
subsequently reduced by iron/ammonium chloride to
provide anilines 5a–h. Compounds 5a,b and 5h have
been previously described.11
Scheme 2. (a) EtOH/HCl; (b) NH2CH2CH(OCH3)2; (c) HCl gas,
AcOH; (d) dimethylsulfate, toluene; (e) Fe/NH4Cl, MeOH, H2O.
Table 1. Inhibition in the Raf/MEK1 coupled assay and cellular activity for compounds 1–3, 10a–l
Compd
NRR
Het
X
Y
Raf/MEK1
IC50 (nM)a
LoVo
IC50 (nM)a
CaCo
IC50 (nM)a
10a
10b
10c
10d
10e
10f
10g
10h
10i
10j
10k
10l
1
N-Morpholine
N-Morpholine
N-Morpholine
N-Morpholine
N-Morpholine
N-Morpholine
N-Morpholine
N-Morpholine
N-Morpholine
1-(4-Ethylpiperazine)
1-(Thiomorpholine-4-oxide)
N-Pyrrolidine
—
2-(1-Methylimidazole)
2-(1-Methylimidazole)
2-(1-Methylimidazole)
2-(1-Methylimidazole)
2-(1,5-Dimethylimidazole)
2-(1,4,5-Trimethylimidazole)
2-Thiazole
Cl
H
S
S
S
S
S
S
S
S
CH2
S
S
S
—
O
—
2
5
28
2
1
2
9
8
212
3
2
5
43
36
5
325
342
56
200
380
7
36
10
131
40016
7
>1000
4700
3200
Me
Br
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
—
H
5200
>1000
900
2700
>1000
5500
>1000
>1000
>1000
>1000
5500
2-Pyridine
2-(1-Methylimidazole)
2-(1-Methylimidazole)
2-(1-Methylimidazole)
2-(1-Methylimidazole)
—
4
2
4
10
2
3
N-Morpholine
—
Phenyl
—
—
>1000
aThe IC50 values reported represent the means of at least two separate determinations with typical variations of less than 40% between replicate
values.

D. Berger et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3031–3034
3033
Target compounds 10a–l were synthesized by heating 98
with anilines 5a–i in the presence of pyridine hydro-
chloride, followed by amine displacement of the alkyl
chloride moiety (Scheme 3).14
LoVo cell proliferation, further mechanistic studies were
carried out with 10a and 2.17 LoVo cells were treated
with compounds 10a and 2 to determine their ability to
inhibit the phosphorylation of MEK1, MAPK and
other selected kinases. The results are summarized in
Table 2. Compound 10a inhibited the phosphorylation
of both MEK1 and MAPK with IC50s of 8 and 0.01
nM, respectively. Presumably, the potent action of 10a
on phospho-MAPK levels was due to the dual effect of
decreasing phospho-MEK1 levels, and inhibition of
MAPK phosphorylation by MEK1. In contrast, 2
inhibited the phosphorylation of MAPK with an IC50 of
380 nM, but did not affect phospho-MEK1. Neither
compound had an effect on the phosphorylation status
of Akt or EGFR kinases; nor did they affect overall cell
phospho-tyrosine levels at concentrations from 1 to
1000 nM. Furthermore, it was demonstrated that the
decrease in phospho-MEK1 levels by 10a was not due
to Raf kinase inhibition (Raf IC50: 850 nM). While the
source of this inhibition of phospho-Mek1 levels is still
under investigation, it suggests that this dual activity of
10a is responsible for its potent antiproliferative effects
on LoVo cells.
All compounds were tested in the high-throughput Raf/
MEK1/MAPK ELISA assay. Cellular proliferation
assays (sulforhodamine B staining) were carried out
using LoVo (K-ras mutant) and CaCo (K-ras wild-type)
human colon tumor cell lines.15 It was anticipated that
selective MEK1 inhibitors would be significantly more
active against the K-ras mutant LoVo cell line, and that
active, but non-selective or cytotoxic compounds, would
have similar activity against both cell lines. These results
are summarized in Table 1.
Compound 10a, with the same aniline substitution as 3,
but possessing a water-solubilizing group at C-7, was
particularly potent in the enzyme and cellular assays
(Raf/MEK1 IC50: 2 nM, LoVo: 5 nM). Replacement of
the 3-chloro substituent on the aniline ring with hydro-
gen or methyl resulted in a modest decrease in activity
(compounds 10b and 10c, respectively), while bromo
was equipotent (10d). Given the size difference between
the chloro and bromo substituents, it appears that these
groups provided a beneficial electronic effect, rather
than providing additional binding interactions to the
enzyme.
In order to establish the utility of our most potent
compounds, 10a and 10j were tested against other cell
lines. Both 10a and 10j showed good activity against
several cell lines, among them a pancreatic line: BXPC-3
(IC50s: 29 and 7 nM, respectively), a prostate line:
LnCAP (IC50s: 40 and 45 nM, respectively) and a blad-
der line: HT1197 (IC50s: 50 nM for both compounds).
Compounds 10e and 10f, which possess additional
methyl groups on the imidazole ring, were very active in
the enzyme assay (Raf/MEK1 IC50s: 1 and 2 nM,
respectively), but only modestly active against LoVo
cells (IC50s: 325 and 342 nM, respectively). Two hetero-
cyclic groups other than N-methylimidazole that pro-
vided moderate cellular activity were the 2-thiazole
substituted 10g (LoVo IC50: 56 nM) and the 2-pyridyl
substituted 10h (LoVo IC50: 200 nM). When the thioe-
ther linker to the N-methylimidazole moiety was replaced
by a methylene group (10i), activity was substantially
reduced (Raf/MEK1 IC50: 212 nM, LoVo: 380 nM).
Several compounds were tested against LoVo xeno-
grafts in nude mice. Compounds were dosed as suspen-
sions in 0.5% Methocel/2% Tween 80 solutions and
injected via an intraperitoneal route (ip) into groups of
five mice.18 One compound that produced a reasonably
consistent response was 10j (Fig. 1).
Table 2. Summary IC50 phospho-blot data from LoVo cells exposed
to compounds 10a and 2
Compd
P-MEK1
(nM)
P-MAPK
(nM)
P-Akt
(nM)
P-EGFR
(nM)
P-TYR
(nM)
Water-solubilizing groups other than morpholine pro-
vided analogues of 10a with similar potency in vitro
(10j–l). These compounds were more soluble in aqueous
media than 3 (for example, aqueous solubility of 10j=2
mg/mL, 10l=55 mg/mL, at pH 7.4). The Pfizer com-
pound 1 was comparable to our most potent analogues
in the Raf/MEK1 assay with an IC50 of 2 nM, while
being less active against LoVo cells (IC50 131 nM).
10a
2
8
>1000
0.01
380
>1000
>1000
>1000
>1000
>1000
>1000
To determine why compounds with similar MEK1
activity have significantly different inhibitory effects on
Scheme 3. (a) 5a–i, pyridine-HCl, 2-ethoxyethanol; (b) RRNH, KI,
DMF.
Figure 1. Effect of 10j in a LoVo xenograft model.

3034
D. Berger et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3031–3034
When dosed at 30 mg/kg twice daily, a statistically sig-
nificant (P<0.01) reduction in tumor growth was seen
at day 21 (T/C 44%).
11. Compounds 5a,b have been previously described: Gilman,
N. W; Chen, W. Y. US Patent 4,973,599, 1990; Chem. Abstr.
1990, 114, 185504. Compound 5h has been previously descri-
bed: Sutherland, B. D.; Morris, J. J.; Thomas, P. A. WO 96/
15118, 1996; Chem. Abstr. 1996, 125, 142741.
In summary, we have discovered a potent series of
MEK1 inhibitors with exceptional cellular activity
against several cell lines. These compounds possess addi-
tional activity that results in decreased phospho-MEK1
levels at low nanomolar concentrations. Compound 10j
showed activity in vivo against LoVo xenografts when
dosed ip. We are currently further optimizing this series
to improve the pharmacological properties and oral
absorption.
12. Makosza, M.; Winiarski, J. J. Org. Chem. 1984, 49, 1494.
13. Slavica, M.; Lei, L.; Patil, P. N.; Kerezy, A.; Feller, D. R.;
Miller, D. D. J. Med. Chem. 1994, 37, 1874.
14. All final products were characterized by 1H NMR, MS
and elemental analysis. A representative example, compound
1
10k: H NMR (DMSO-d6) d 1.98 (m, 2H); 2.55 (t, J=5.2 Hz,
2H); 2.70 (m, 4H); 2.87 (broad s, 4H); 3.60 (s, 3H); 3.91 (s,
3H); 4.21 (t, J=4.8 Hz, 2H); 6.57 (d, J=6.4 Hz, 1H); 7.12 (d,
J=7.8 Hz, 1H); 7.16 (s, 1H). 7.35 (s, 1H), 7.39 (s, 1H), 7.54 (s,
1H), 7.68 (s, 1H), 8.50 (s, 1H), 9.59 (s, 1H). MS (ES) m/z 597.2
+
.
(M+H) . Anal. for C28H29ClN6O3S2 H2O, calcd: C 54.66; H
5.08; N 13.66, found: C 54.62H 4.93; N 13.60.
Acknowledgements
15. Tasdemir, D.; Mallon, R.; Greenstein, M.; Feldberg, L. R.;
Kim, S. C.; Collins, K.; Wojciechowicz, D.; Mangalindan,
G. C.; Concepcion; Harper, M. K.; Ireland, C. M. J. Med.
Chem. 2002, 45, 529.
16. By our current cellular proliferation assay protocol, the
IC50 of compound 2 is higher than previously reported.8
17. Mallon, R.; Feldberg, L. R. Kim, S. C.; Collins, K.;
Kovacs, E. D.; Kohler, C.; Wojciechowicz, D.; Zhang, N.;
Wu, B.; Floyd, B.; Powell, D.; Berger, D. Mol. Cancer Ther.
Submitted for publication.
18. Human Colon Carcinoma LoVo cells (American Type
Culture Collection, Rockville, Maryland # CRL-155) were
grown in tissue culture in RPMI (Gibco/BRL, Gaithersburg,
MD) supplemented with 10% FBS (Gemini Bio-Products Inc.,
Calabasas, CA, USA). Athymic nu/nu female mice (Charles
River, Wilmington, MA, USA) were injected SC in the flank
area with 7Â106 LoVo cells plus Matrigel (BD Biosciences,
Billerica, MA, USA). When tumors attained a mass of
between 130 and 180 mg, the mice were randomized into
treatment groups (day zero), five animals per group. Animals
were treated IP twice a day on days 1 through 20 post staging
(day zero) with either 30, 10, or 3 mg/kg/dose of compound
prepared in 0.5% Methocel/2% Tween 80 as the vehicle
We thank members of the discovery analytical chemistry
group for spectral data, elemental analyses and aqueous
solubility determination of selected compounds.
References and Notes
1. McCormick, F.; Wittinghofer, A. Curr. Opin. Biotechnol.
1996, 7, 449.
2. Robinson, M. J.; Cobb, M. H. Curr. Opin. Cell Biol. 1997,
9, 180.
3. Lewis, T. S.; Shapiro, P. S.; Ahn, N. G. Adv. Cancer Res.
1998, 74, 49.
4. Kolch, W. Biochem. J. 2000, 351, 289.
5. Dancy, J. E. Curr. Pharm. Des. 2002, 8, 2259.
6. Sebolt-Leopold, J. S.; Dudley, D. T.; Herrera, R.; Van
Becelaere, K.; Wiland, A.; Gowan, R. C.; Tecle, H.; Barrett,
S. D.; Bridges, A.; Przybranowski, S.; Leopold, W. R.; Saltiel,
A. R. Nat. Med. 1999, 5, 810.
7. Sebolt-Leopold, J. S. Oncogene 2000, 19, 6594.
8. Zhang, N.; Wu, B.; Powell, D.; Wissner, A.; Floyd, M. B.;
Kovacs, E. D.; Toral-Barza, L.; Kohler, C. Bioorg. Med.
Chem. Lett. 2000, 10, 2825.
9. Zhang, N.; Wu, B.; Eudy, N.; Wang, Y.; Ye, F.; Powell, D.;
Wissner, A.; Feldberg, L. R.; Kim, S. C.; Mallon, R.; Kovacs,
E. D.; Toral-Barza, L.; Kohler, C. A. Bioorg. Med. Chem.
Lett. 2001, 11, 1407.
control. Tumor mass was determined every
7
days
[(lengthÂwidth2)/2] for 21 days post staging. Relative tumor
growth (Mean tumor mass on days 7, 14 and 21 divided by the
mean tumor mass on day zero) was determined for each
treatment group. Statistical analysis (Student t-test) of log
relative tumor growth was used to compare treated verses
control group in each experiment. A p value (pꢁ0.05) indi-
cates a statistically significant reduction in relative tumor
growth of treated group compared to the vehicle control.
10. Mallon, R.; Feldberg, L. R.; Kim, S. C.; Collins, K.;
Wojciechowicz, D.; Hollander, I.; Kovacs, E. D.; Kohler, C.
Anal. Biochem. 2001, 294, 48.