Discovery of thienopyridines as Src-family selective Lck inhibitors

Article abstract of DOI:10.1016/j.bmcl.2006.12.035

We describe the identification, SAR, and in vivo pharmacology of a new series of Src-family selective Lck inhibitors. These thienopyridines were designed based on a desire to access the unique residues in the extended hinge region of Lck.

Full text of DOI:10.1016/j.bmcl.2006.12.035

Bioorganic & Medicinal Chemistry Letters 17 (2007) 1167–1171
Discovery of thienopyridines as Src-family selective Lck inhibitors
Lily Abbott, Patrick Betschmann, Andrew Burchat, David J. Calderwood, Heather Davis,
*
Peter Hrnciar, Gavin C. Hirst, Biqin Li, Michael Morytko, Kelly Mullen and Bryant Yang
Abbott Bioresearch Center, 100 Research Drive, Worcester, MA 01605-5314, USA
Received 14 November 2006; revised 8 December 2006; accepted 8 December 2006
Available online 15 December 2006
Abstract—We describe the identification, SAR, and in vivo pharmacology of a new series of Src-family selective Lck inhibitors.
These thienopyridines were designed based on a desire to access the unique residues in the extended hinge region of Lck.
Ó 2006 Elsevier Ltd. All rights reserved.
The Src-family of tyrosine kinases comprises eight high-
ly homologous proteins that are primarily expressed in
hematopoietic tissues, two of which, Lck and Fyn, are
expressed in T cells. Lck plays a critical function during
grafts in models of solid organ transplant rejection after
4
oral administration. Although A-420983 provides a
useful tool, it exhibits little selectivity within the Src-
family. Herein we describe our efforts that result in the
identification of a new class of Lck inhibitors with high
selectivity within the Src-family.
1
the initial steps of T-cell receptor signaling resulting in
a cascade of downstream signaling pathways leading to
T-cell activation and the production of cytokines such as
2
,3
IL-2 and IFNc. A selective inhibitor of Lck should
prevent T-cell activation and thus have broad applica-
tion for the treatment of T-cell dependent processes such
as autoimmune and inflammatory diseases as well as or-
gan transplant rejection.
The crystal structure of A-420983 in highly homologous
Hck (residues 60–505, pY -EEI) as a Lck surrogate has
5
01
4
been determined and the binding mode is represented in
Figure 1 (Lck residues in red). The 4-amino group makes
a key H-bond donor contact to the backbone C@O of
Glu317 (Lck numbering) whilst the N5 pyrimidine nitro-
gen contacts the backbone NH of Met319 in the hinge re-
gion. The indole extends into the hydrophobic pocket
and is surrounded by residues Leu303, Ile314, Met292,
and Leu385. The methoxy group makes a van der Waals
contact with the side chain of Thr316 whilst the amide
carbonyl contacts the backbone NH of Asp382 in the
conserved DFG-motif. The trans-cyclohexyl piperazine
moiety extends into the solvent exposed region where
the terminal piperazine nitrogen makes a charge-rein-
forced hydrogen-bond to the side chain of the conserved
N
HN
O
O
1
NH2
1; R =
N
N
NMe A-420983
NAc A-770041
N
N
1
R =
N
N
R1
5
Asp326. As previously demonstrated with A-770041 ,
making this contact with basic amines was detrimental
for selectivity, hence our current approach was to focus
on breaking this latter interaction and probe alternative
extended hinge interactions. In particular, we felt that
making productive contacts with the side chains of
Tyr318 and the unique Glu320 in the extended hinge re-
gion of Lck would drive selectivity within the Src-family.
We previously reported the identification of a novel pyr-
azolopyrimidine A-420983, a potent Lck inhibitor that
prevents TCR-mediated T-cell activation and prolongs
survival of major histocompatibility mismatched allo-
Keywords: Lck
Thienopyridines.
inhibitors;
Src-family;
Kinase
inhibitors;
*
Corresponding author at present address: SGX Pharmaceuticals,
0505 Roselle St., San Diego, CA 92121, USA. Tel.: +1 858 228 1630;
fax: +1 858 558 0642; e-mail: gavin_hirst@sgxpharma.com
1
Inhibitors were screened against a non-phosphorylated
construct of human Lck, Lck (64–509) in HTRF format
0
960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.12.035

1168
L. Abbott et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1167–1171
Phe318
Tyr318
Glu317
Ala320
Glu320
O
O
Thr316
Met 319
H
N
H
N
H
N
N
H
N
HN
O
H
O
O
O
5
H
N
H
7
N
N
4
H
OMe
N
N
N
NH
O
N
Asp326
N
Leu303
Ile314
Met 292
Leu385
Asp382
Figure 1. Binding mode of A-420983 in Hck (Black). Lck residues in red.
at physiologically relevant 1 mM ATP concentration
6
and biotinylated Lck peptide as substrate. The closely
significant reduction in hinge binding efficiency in this
core compared to the pyrimidine counterpart 2. Our ini-
tial investigation of the impact of branching at C-7 is
shown in Table 1.
related Src-family kinase Hck served as our initial selec-
tivity counterscreen. This was predicated on the fact that
Hck differentiates itself from Lck at the residues of cur-
rent focus, Tyr318 and Glu320, as shown in Figure 1.
Selected compounds were progressed for wider kinase
profiling and for activity in cellular settings and ulti-
mately in vivo. Data are presented here for the inhibi-
tion of anti-CD3 mAb induced IL-2 production in
It is clear that substitution at C-7 can provide a signifi-
cant potency increase. Comparison of the Lck potency
of the two regioisomeric amides 5 (20.8 lM) and 11
2
(0.24 lM) suggests a requirement for an Sp carbon at
the C-7 branch point. This observation is supported by
data for the vinyl amide 12 and by the improved potency
of the allylic alcohol 14 compared to its saturated coun-
terpart 15. Cyclization of amide 9 to the piperazine
amide 10 results in greater than 100-fold reduction in
potency. Despite increases in potency, the only com-
pound in Table 1 that exhibited more than 14-fold selec-
tivity against Hck was the allylic amine 16. Although
definitive proof awaits the identification of Lck: ligand
cocrystal structures, these data are consistent with the
allylic amine making a productive interaction with the
distinguishing Tyr318 in Lck and that the carboxamides
such as 9 and 11 are unable to access this side chain.
6
human whole blood and for an acute in vivo assay of
the inhibition of TCR stimulated (anti-CD3 mAb)
IL-2 production in mice after oral dosing. Resulting
IL-2 levels were measured by an Enzyme Linked Immu-
nosorbent Assay (ELISA) method.
N
O
N
O
HN
HN
O
O
NH2
NH2
Having established that our initial area of opportunity
to identify compounds with improved Src-family selec-
tivity resided in the allylic amines, we investigated this
in more detail. These results are shown in Table 2. Ini-
tially noteworthy is the decreased Lck potency of the
cis-allylic amine 17 compared to the trans diastereomer
N
N
N
N
S
N
N-7
CH3
C-7
Lck IC50
2 = 0.031 μM
Lck IC50 = 5.2 μM
3
1
8. In addition, we observed an increased selectivity for
Lck against Hck going from the amide 13 (10-fold) to
the corresponding amine 25 (220-fold) primarily due to
an 18-fold reduction in Hck potency. This suggests a
detrimental impact of either a basic residue or increased
entropy for Hck but not for Lck. This observation mir-
rored the analogous selectivity found in Table 1 for the
amide 12 (4-fold) and amine 16 (38-fold). Non-basic sul-
fonamide, urea and amide termini represented by 19, 20,
and 21, respectively, lost selectivity and/or potency com-
pared to the amine parent 16. Further extending the dis-
tance from C-7 to the basic terminus as seen in 22, 25,
We were very encouraged by the fact that the truncated
analog 2, despite being devoid of the cyclohexyl pipera-
zine, was equipotent to A-420983. In order to access the
extended hinge region, our analysis of the crystal struc-
ture of A-420983 suggested branching at N-7 was desir-
able, a site excluded by valency requirements. We
therefore elected to probe thienopyridines as potential
Lck inhibitors, an approach that would allow branching
at C-7. The drop in potency of the thienopyridine bench-
mark compound 3, a 5.2 lM Lck inhibitor, suggests a

L. Abbott et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1167–1171
Table 1. Inhibition of Lck and Hck (IC50, lM) by compounds 4–16 Table 3. Kinase inhibition (IC50 lM) by compounds 18, 25–28
1169
Compound Lck Src
Fyn Fgr Hck Lyn
Tie-2 Kdr
N
1
R
NH2
18
25
26
0.29
9.05 44.1 14.1 13.6
1.18 >50
>50
>50
18.5
>50
>50
OMe
S
0.19 14.4
0.11 19.7
0.48 27.1
0.21 33.9
26.2 26.5 24.2
3.7
3.7
1.9
NH
O
25.9 17.9 24.1 18.5
26.3 21.7 20.1 12.9
N
2
2
7
8
0.57
10.1
35.1 37.9 10.0
5.3
1
R
Compound
Lck
Hck
2
6, and 27 maintains both Lck potency and selectivity
–
NH
2
4
5
9.65
20.8
>50
45.29
against Hck. Of note, the trans amine 26 is the more po-
tent of the two diastereomers. We selected a subset of
compounds for wider kinase profiling both within the
Src-family and against two receptor tyrosine kinases,
Kdr and Tie-2. These data are shown in Table 3.
-
NHCOCH CH N
2
2
-NHCONHCH CH N
6
6.85
>50
2
2
–
–
–
COOH
CONH
CONHCH
7
8
9
>50
0.38
0.14
>50
1.41
1.97
2
Although many of these analogs portrayed very high
selectivity across all members of the Src-family, the ami-
no alcohol 28 was the only example with >20-fold selec-
tivity against all the kinases. It is of note that this
selectivity profile is increased compared to the previous-
ly published pyrazolopyrimidine A-770041. In particu-
lar, A-770041 exhibited only 8-fold selectivity against
Hck and Lyn.
2
CH
2
NMe
2
O
N
NMe
10
11
12
17.6
0.24
0.541
0.13
>50
2.55
2.26
1.29
-
CONHCH CH N
2
2
5
trans-CH@CHCONH
2
O
1
3
N
H
NH
OH
4,5
In accord with our previous publications
all com-
pounds profiled in the human whole blood assay (data
not shown) showed alignment with the Lck in vitro
potency suggesting that the compounds achieved effec-
tive cellular penetration and that plasma protein binding
was not effecting potency. In mice, compound 28 inhib-
ited TCR stimulated (anti-CD3 mAb) IL-2 production
in a dose-responsive manner with an ED50 of 5 mg/kg
after oral administration. The pharmacokinetic profile
in rats is shown in Table 4.
trans-CH@CHCH
2
14
15
16
0.77
4.81
0.23
2.36
>50
8.8
–
CH
2 2 2
CH CH OH
trans-CH@CHCH
2
NH
2
Table 2. Inhibition of Lck and Hck (IC50, lM) by compounds 16–26
N
S
NH2 OMe
H
N
N
R1
Despite exhibiting a long half-life, compound 28 had a
pharmacokinetic profile that excluded it from advance-
ment into chronic disease models. In particular, the high
clearance and moderate oral bioavailability were areas
of concern. Efforts to marry the unprecedented Src-fam-
ily selectivity of this series of compounds with improved
pharmacokinetic profile are ongoing.
O
1
R
Compound
Lck
Hck
cis-CH
2
NEt
NEt
NHSO
2
17
18
19
20
21
4.47
0.29
1.37
6.7
>50
13..6
4.3
–CH
–CH
–CH
–CH
2
2
2
2
2
2
Me
NHCONHPh
NHCOMe
21.9
2.36
0.72
Access to the compounds in Tables 1 and 2 was
achieved from the orthogonally functionalized thieno-
pyridine 32. Many examples are captured in Schemes
CH N
NMe
22
0.29
2.95
5.97
2
1
and 2. Electrophilic bromination of thiophene 29
CH N
NPh
O
23
>50
2
followed by a Knoevenagel reaction with malonic acid
and acyl azide formation gave the cyclization precur-
sor 30 in 86% yield over three steps. Curtius rear-
rangement followed by intramolecular cyclization at
high temperature gave the pyridone 31 in 90% yield.
Chlorination and subsequent displacement followed
by regioselective iodination gave the key building
2
4
1.52
12.9
CH N
NMe
2
CH NH
NH
2
2
2
2
5
6
7
8
0.19
0.11
0.48
0.21
24.2
24.1
20.1
10.0
2
CH NH
NH2
NH2
2
Table 4. Rat pharmacokinetic parameters for compound 28
CH NH
2
T
max (h)
Clp (l/h/kg)
6.3
T
1/2 (h)
F (%)
5.3
8.1
21
CH N
OH
2
5
mg/kg iv, 10 mg/kg po.

1170
L. Abbott et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1167–1171
O
Br
Br
d
a,b,c
e,f,g
HN
N OC
3
OHC
S
S
S
2
9
31
3
0
O
H
N
NH2
Br
O
NH2
Br
N
N
h
i,j,k
N
NH2
S
O
S
H
N
N
O
I
N
S
tBuO
O
B
O
O
3
2
33
12, 13
3
5
NHR
O
O
O
B
l
37
O
H
N
CH(OEt)2
O
NH2
Br
N
m,n,o
N
NH2
S
N
S
CH(OEt)2
1
6, 18, 22 - 28
3
6
NHR
Cl
Scheme 1. Reagents and conditions: (a) Br
DMF (cat), 14 h, rt; ii—NaN , rt, 3 h, 100%; (d) Bu
8 h, 94%; (g) NIS, DMF, rt, 75%; (h) tert-butyl acrylate, Pd(OAc)
reflux, 75%; (j) TFA, CH Cl , 100%; (k) HBTU, HOBT, DIEA, DMF, amine, 55–76%; (l) 37, DME/H
DME/H O, Na CO , Pd(PPh
2
, AlCl
3
, CH
N, PhOPh, 225 °C, 45 min, 86%; (e) POCl
, PPh , Na CO
, DMF, 80 °C, 68%; (i)—35, DME/H
2
2
, 90%; (b) CH
2
(COOH)
2
, piperidine(cat) pyridine, 100 °C, 22 h, 96%; (c) i—(COCl)
, 135 °C, 2.5 h, 95%; (f) NH OH, dioxane, 150 °C,
O, Na CO , Pd(PPh
, Pd(PPh , reflux, 75%; (m) 35,
2
,
3
3
3
4
1
2
3
2
3
2
2
3
3 4
) ,
2
2
2
O, Na
2
CO
3
3 4
)
2
2
3
3
)
4
, reflux, 76%; (n) p-TSA, acetone, water, 100%; (o) amine, NaB(OAc)
3
, DCE, 67–88%.
O
N
H
N
NH2
Br
O
NH2
Br
N
a
b,c d
N
NH2
S
S
N
I
S
O
R
32
38
7
8
(R = OH)
-11
e
Scheme 2. Reagents and conditions: (a) Bu ) , reflux, 90%; (b) 35, DME/H
c) OsO /NaIO , dioxane/water, 22%; (d) TPAP, 49%; (e) HOBt/HBTU, DMF, RNH , 85%.
3
SnCH@CH
2
, KF/toluene, Pd(PPh
3
4
2 2 3 3 4
O, Na CO , Pd(PPh ) , reflux, 76%;
(
4
4
2
block 32. Heck reaction with the corresponding acrylic
ester presented ester 33. Suzuki coupling with the pre-
viously published boronate ester 35 installed the
northern domain which gave the a–b unsaturated
amides 12 and 13 after standard deprotection and
coupling protocols.
gave amines 16, 18, and 22–28 after appropriate depro-
tection where required.
Scheme 2 portrays our route to the amides 8–11. Stille
coupling with vinyl-tributyl tin gave 38 which was func-
tionalized to the aforementioned amides through stan-
dard oxidative and coupling procedures.
Alternatively, regioselective Suzuki reaction of 32 with
the vinyl boronate 37 gave acetal 36. A second Suzuki
reaction with boronate 35 and unmasking to the a–b
unsaturated aldehyde followed by reductive amination
In summary, by designing molecules to probe the unique
residues in the hinge region of Lck, we have identified a
series of potent and highly selective Lck inhibitors that

L. Abbott et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1167–1171
1171
exhibit oral activity in acute in vivo models of T-cell acti-
vation. This represents an advancement in the under-
standing of features important for selectivity within the
Src-family of tyrosine kinases.
3. Van Oers, N. S.; Kileen, N.; Weiss, A. J. Exp. Med. 1996,
83, 1053.
1
4
. Borhani, D. W.; Calderwood, D. J.; Freidman, M. M.;
Hirst, G. C.; Li, B.; Leung, A. K. W.; McRae, B.;
Ratnofsky, S.; Ritter, K.; Waegell, W. Bioorg. Med. Chem.
Lett. 2004, 14, 2613.
References and notes
5. Burchat, A. F.; Borhani, D. W.; Calderwood, D. J.; Hirst,
G. C.; Li, B.; Stachlewitz, R. F. Bioorg. Med. Chem. Lett.
2006, 16, 118.
6. Mean of two experiments performed with seven concentra-
tions of test compound.
1
2
. Straus, D. B.; Weiss, A. Cell 1992, 70, 585.
. Weil, R.; Veillette, A. Curr. Top. Microbiol. Immunol. 1996,
2
05, 63.