Hit-to-lead studies on benzimidazole inhibitors of ITK: Discovery of a novel class of kinase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2007.04.045

Benzimidazole 1 was identified as a selective inhibitor of ITK by high throughput screening. Hit-to-lead studies defined the SAR at all three substituents. Reversing the amide linkage at C6 led to 16, with a fivefold improvement of potency. This enhanceme

Full text of DOI:10.1016/j.bmcl.2007.04.045

Bioorganic & Medicinal Chemistry Letters 17 (2007) 3660–3665
Hit-to-lead studies on benzimidazole inhibitors of ITK:
Discovery of a novel class of kinase inhibitors
Roger J. Snow,a,* Asitha Abeywardane,a Scot Campbell,c John Lord,a
Mohammed A. Kashem,a Hnin Hnin Khine,b Josephine King,b Jennifer A. Kowalski,a
Steven S. Pullen,b Teresa Roma,b Gregory P. Roth,a Christopher R. Sarko,a
Noel S. Wilson,a Michael P. Winters,d John P. Wolaka and Charles L. Cywina
aMedicinal Chemistry, Boehringer Ingelheim Pharmaceuticals Inc., 900 Ridgebury Road, Ridgefield, CT 06877, USA
bInflammation and Immunology, Boehringer Ingelheim Pharmaceuticals Inc., 900 Ridgebury Road, Ridgefield, CT 06877, USA
cAnalytical Sciences, Boehringer Ingelheim Pharmaceuticals Inc., 900 Ridgebury Road, Ridgefield, CT 06877, USA
dJohnson and Johnson PRD, 665 Stockton Drive, Exton, PA 19341, USA
Received 9 February 2007; revised 6 April 2007; accepted 16 April 2007
Available online 25 April 2007
Abstract—Benzimidazole 1 was identified as a selective inhibitor of ITK by high throughput screening. Hit-to-lead studies defined
the SAR at all three substituents. Reversing the amide linkage at C6 led to 16, with a fivefold improvement of potency. This
enhancement is rationalized by the conformational preference of the substituent. A model for the binding of the benzimidazoles
to the ATP-binding site of ITK is proposed.
ꢀ 2007 Elsevier Ltd. All rights reserved.
Interleukin-2-inducible T cell kinase (ITK), also known
as T cell-specific kinase (TSK) and, expressed mainly in
T-lymphocytes (EMT) is a member of the Tec family of
protein tyrosine kinases.1–3 ITK is expressed in T cells,
mast cells, and natural killer cells. It is activated in T
cells upon stimulation of the T cell receptor (TCR),
and in mast cells upon activation of the high affinity
IgE receptor. Following stimulation of the TCR, ITK
is phosphorylated in the kinase domain activation loop
by the kinase LCK.4 Activated ITK is required for phos-
phorylation of PLC-c5 which in turn leads to calcium
mobilization and activation of downstream pathways.6
CD4+-T cells from ITK knockout mice have a dimin-
ished proliferative response in a mixed lymphocyte reac-
tion7 or upon anti-CD3 stimulation. These T cells
produce little IL-2 upon TCR stimulation, resulting in
reduced proliferation. Diminished production of IL-4,
IL-5, and IL-13 upon stimulation of the TCR has also
been observed in ITK deficient CD4+-T cells.8 ITK defi-
cient mice show a greatly diminished inflammatory
response in a model of allergic asthma induced by
OVA.9 These studies support a key role for ITK in the
activation of T cells, thus inhibitors of ITK should be
useful as immunosuppressive or anti-inflammatory
agents.
Recently Das et al.10,11 disclosed a series of thiazole
inhibitors of ITK. We now report our initial studies in
this area which led to the discovery of a novel class of
kinase inhibitors active against ITK. Ultra high
throughput screening (UHTS) of the BI sample collec-
tion against ITK was carried out using a DELFIA
assay.12 Initial hits were retested in dose response
against ITK, and also against BTK and LYN, both of
which are highly homologous with ITK, to focus on
selective inhibitors. One class of hits that did not inhibit
in either of the counterscreens was a series of benzimi-
dazoles from a general screening library. The initial hit
1 had ITK IC50 0.11 lM and showed activity in a
cellular assay, at 3.0 lM. This series was particularly
interesting because its structure was not related to any
known kinase inhibitor.
Keywords: ITK; Kinase; Kinase inhibitor; Benzimidazole; Inducible
T cell kinase.
The original screening library contained 1760 members
(11R1 · 20R2 · 8R3) (Fig. 1). At the R1 amide position,
active hits included the thiophene, benzene, and
*
Corresponding author. Tel.: +1 203 798 4060; fax: +1 203 798
5297; e-mail: rsnow@rdg.boehringer-ingelheim.com
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.04.045

R. J. Snow et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3660–3665
3661
Initial exploration of other side chains at R3 was carried
out using a solution-phase method (Scheme 2). For this
study R1 was kept constant as the thiophene and R2 as
N-cyclohexyl-N-methyl amine. The same general ap-
proach was followed, but in this case the R2 amide
was introduced first from 4-fluoro-3-nitro benzoic acid,
yielding 9. This intermediate was reacted with a variety
of R3 amines to give 10. For the solution-phase synthe-
sis, reduction of the nitro group by transfer hydrogena-
tion was preferred over tin chloride reduction. The
2-aminobenzimidazole 12 was formed with cyanogen
bromide, and acylation yielded 1. In the final step impu-
rities presumably resulting from acylation on the ring
nitrogen and double acylation were observed. Pure
product was obtained by chromatography on silica.
An analogous route was used to obtain the cyclohexyl
ester 50 at R2. The analog 47 where the amide has been
moved to the 6-position of the benzimidazole was
prepared in a similar way from 3-fluoro-4-nitrobenzoic
acid. The reverse amide analogs of 1 were obtained as
shown in Scheme 3, by acylation of 4-fluoro-3-nitroani-
line with cyclohexylcarbonyl chloride. For the tertiary
amides, the product was methylated on nitrogen to give
14 before performing the SNAr reaction. Completion of
the synthesis followed the same steps of reduction,
benzimidazole formation, and acylation as before to
give reverse amide 16.
O
O
O
O
R1
R2
N
N
N
N
N
S
N
H
N
H
R3
NH2
1
O
Figure 1. General structure of the screening library.
4-methoxyphenyl. At R2 the preferred substituents were
the N-cyclohexyl-N-methyl and N-benzyl-N-methyl
amides. For the R3 group on the benzimidazole nitro-
gen, the amide side chain derived from b-alanine was
the only member of the library that showed appreciable
activity. Many of the compounds in the library had R3
groups derived from a-amino acid amides, none of
which hit in the screen.
To further define the SAR, exploration of the hit was
carried out by a combination of solid-phase library syn-
thesis and single point synthesis using solution-phase
chemistry. To this end two focused libraries were synthe-
sized, in one case varying R1, and keeping R2 constant
as N-cyclohexyl-N-methyl, (77 compounds) and in the
other varying R2 and keeping R1 constant as phenyl
(99 compounds). In both cases the b-alanine amide side
chain at R3 was kept constant. The libraries were syn-
thesized using solid-phase methodology reported by
Lee13 (Scheme 1). Sieber amide resin loaded with b-ala-
nine 2 was treated with the 4-fluoro-3-nitro benzoic acid
to effect the SNAr reaction. The R2 amide was then
introduced, by coupling the acid with N-methyl cyclo-
hexylamine in the presence of TBTU to give 4. Reduc-
tion of the nitro group with tin chloride worked well
on solid phase, to form the diamine 5, which was con-
verted to the 2-aminobenzimidazole with cyanogen bro-
mide. The amine was acylated with the appropriate acid
chloride, and the product was cleaved from the resin
with TFA to leave the primary amide at the terminus
of the R3 side chain. Compounds that showed activity
were resynthesized and purified.
Compounds were assayed for ITK inhibitory activity in
a DELFIA format.12 Cellular activity was assessed by
measuring calcium influx following B cell receptor stim-
ulation of BTK deficient DT40 cells14 stably transfected
to express human ITK, using a FLIPR instrument. Dur-
ing the hit-to-lead phase the cellular assay was used to
rule out inactive compounds, rather than to rank activ-
ity. Work to improve cellular potency will be detailed in
subsequent papers. Results for selected compounds illus-
trating the SAR at R1 and R2 are shown in Table 1. At
R1 an aromatic amide is required. No inhibition was ob-
served with aliphatic amides (data not shown). Phenyl
can take the place of thiophene with only a slight loss
in potency (17). Electron withdrawing groups at C3 or
O
O
O
NO2
NO2
NH
NH2
NH
HO
N
N
NH2
NH
a
b
c
O
NH
Pol
O
NH
Pol
O
NH
Pol
O
NH
Pol
2
3
4
5
Pol = Sieber amide resin
O
O
N
O
O
O
N
N
N
N
N
N
N
N
S
N
H
S
NH2
N
H
f
d
e
O
NH2
O
NH
Pol
O
NH
Pol
6
7
1
Scheme 1. Solid-phase synthesis of compound 1. Reagents: (a) 4-fluoro-3-nitrobenzoic acid, DMF, DIPEA; (b) N-methyl cyclohexylamine, TBTU,
DMF/DCM; (c) SnCl2Æ2H2O; (d) BrCN, EtOH; (e) thiophene-2-carbonylchloride, pyridine; (f) 20% TFA, DCM.

3662
R. J. Snow et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3660–3665
O
O
O
O
a
NO2
NH
NH2
NH
NO2
F
NO2
F
b
c
N
N
HO
N
8
9
10
11
O
NH2
O
NH2
O
O
O
N
N
N
N
N
S
d
NH2
NH2
e
N
H
N
O
O
NH2
12
1
Scheme 2. Solution-phase synthesis of compound 1. Reagents and condition: (a) N-methyl cyclohexylamine, EDC, HOBT, DCM; (b)
b-alanineamide, Et3N, DMSO, 90 ꢁC; (c) NH4HCO2, Pd/C, EtOH; (d) BrCN, EtOH; (e) thiophene-2-carbonylchloride, pyridine.
O
H2N
NO2
F
N
NO2
F
N
NO2
NH
N
N
N
a,b
d,e,f
c
S
N
H
O
O
O
13
14
O
NH2
O
NH2
15
16
Scheme 3. Solution-phase synthesis of compound 16. Reagents and conditions: (a) cyclohexanecarbonylchloride THF, reflux; (b) NaHMDS, MeI,
THF; (c) b-alanineamide, Et3N, DMSO, 90 ꢁC; (d) NH4 HCO2, Pd/C, EtOH; (e) BrCN, EtOH; (f) thiophene-2-carbonylchloride, pyridine.
Table 1. Representative SAR of the R1 and R2 groups
O
O
R1
R2
N
N
H
N
NH2
O
R1
R2
ITK IC50 (lM)
DT40 cell IC50 (lM)
a
b
Compound
1
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
2-Thienyl
Ph
c-HexNMe
c-HexNMe
c-HexNMe
c-HexNMe
c-HexNMe
c-HexNMe
c-HexNMe
c-HexNMe
c-HexNMe
c-HexNMe
c-HexNMe
c-HexNH
c-HexNH
(c-Hex)2N
i-PrNMe
0.11
0.37
1.1
3.0
11
4-MeOPh
4-BrPh
4-FPh
4-NO2Ph
3-FPh
3-CNPh
2-MePh
5-Isoxazolyl
3-Pyridyl
2-Thienyl
Ph
17
0.26
0.25
0.076
0.11
0.20
2.7
1.2
4.1
6.9
7.4
7.2
nt
0.062
0.069
>20
>20
>20
11
4.1
2.7
nt
nt
Ph
Ph
nt
nt
Ph
Ph
c-HexNEt
c-HexNallyl
c-HexNEt
c-HexNallyl
PhNMe
0.39
0.73
0.081
0.23
0.33
8.1
7.2
1.2
1.2
2.6
4-BrPh
4-BrPh
Ph
nt, not tested.
a Values are geometric means of at least two experiments. In most cases the individual results vary by less than a factor of 2.
b Mean of at least two experiments.

R. J. Snow et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3660–3665
3663
C4 of the phenyl ring (20–23) are tolerated, the 4-nitro
analog 21 showing a fourfold improvement, whereas
electron donating substituents such as methoxy (18)
decrease activity. Substitution at C2 in 24 is detrimental.
Isoxazole 25 and pyridine 26 show somewhat improved
potency in the enzyme assay, though this is not reflected
in the cellular assay. At R2 there is a requirement for a
tertiary amide with one cyclic substituent and one small
alkyl substituent. Removing the N-methyl to give a sec-
ondary amide in 27 or 28 greatly reduces activity. The
methyl can be replaced with small alkyl groups 31–34,
but not with a more bulky group such as a second
cyclohexyl ring 29. The cyclohexyl ring can be replaced
with phenyl 35, but potency is lost on replacing it with a
smaller alkyl group in 30. This intriguing SAR
observation is discussed in more detail below.
viewed as isosteric with the corresponding amide. The
most exciting result was observed with the reverse amide
16, IC50 25 nM, a significant improvement over the
screening hit. Removal of the N-methyl from 16 again
gives a substantial loss in affinity, though in this case
the activity of the secondary amide 51 is measurable.
The SAR at this position clearly involved a hydrophobic
interaction with the cyclohexyl ring. The large difference
in potency between the secondary and tertiary amides
seemed to argue against a hydrogen bonding interaction
with the carbonyl. We hypothesized that this difference
in affinity is due to changes in the conformation of the
side chain. Tertiary amides can more readily adopt
the cis amide conformation than secondary ones, and
the activity may be linked to the ability to achieve the
cis amide conformer. To determine the conformational
preference, molecular modeling was carried out on a
simplified fragment of the molecule, with methyl groups
at R1 and R3. In the cis amide conformation, the cyclo-
hexyl ring is forced out of the plane of the benzimid-
azole. The energy difference between the two
conformers is shown in Table 4. In all cases the trans
conformation is preferred, but for the reverse amide 16
the energy difference is much lower (0.3 kcal/mol) than
for the other linkers. The NH version of the reverse
amide 51 and the original tertiary amide 1 are in a sim-
ilar range of 2–3 kcal/mol, while the inactive secondary
amide 27 and the ester 50 have values over 3 kcal/mol,
corresponding to <1% cis conformer at equilibrium.
This analysis provides support for the hypothesis that
the R2 side chain binds in a cis amide conformation,
which would allow the cyclohexyl ring to access a puta-
tive hydrophobic interaction out of the plane of the
benzimidazole ring.
At the R3 position more variation is possible (Table 2).
While the unsubstituted benzimidazole 41 is only weakly
active, a simple alkyl substituent 42 restores much of the
potency. Polar substituents, either as the amide or alco-
hol, are preferred, with 40 and 45 showing the best po-
tency. Substitution of the amide (37,38) maintains
binding, but diminishes cellular activity. Replacing the
hydroxyl in 45 with an amine (46) resulted in a 10-fold
decrease in enzyme activity, and loss of cellular activity.
Substitution or branching at the carbon next to the
benzimidazole in the R3 side chain is not tolerated (data
not shown).
To further investigate the requirements of the R2 substi-
tuent, additional variations at the 5- and 6-positions of
the benzimidazole were investigated (Table 3). Moving
the amide to the 6-position (47) causes a sixfold decrease
in activity, and removing the amide altogether (48) still
retains micromolar potency. Replacing the amide with
a methyl (49) or a cyclohexyl ester (50) greatly reduces
potency, even though the cyclohexyl ester might be
With the initial SAR established, we sought to under-
stand how these novel inhibitors bound to the ITK
active site. The majority of kinase inhibitors bind to
the ATP site via an essential H-bonding interaction. In
many cases, including ATP itself, this involves a biden-
tate interaction with a donor–acceptor pair. Compound
1 contains several possible H-bonding sites, but appar-
ently none that would give a bidentate interaction. A
clue to this came during structure confirmation by
NMR, when an NOE was observed between the
exchangeable NH and the proton at the 4-position of
the benzimidazole. This can only be consistent with
the exocyclic imine tautomer, 1b, where the proton is
on the ring nitrogen. This tautomerism has been
observed previously in 2-acylaminobenzimidazoles.15
Table 2. Representative R3 groups
O
O
N
N
S
N
H
N
R3
Compound
R3
ITK
DT40 cell
b
IC50 (lM)
a
IC50 (lM)
1
36
37
38
39
40
41
42
43
44
45
46
(CH2)2CONH2
(CH2)2CO2Et
(CH2)2CONHMe
(CH2)2CONMe2
(CH2)3CONH2
(CH2)4CONH2
H
0.11
0.27
0.23
0.16
0.14
0.06
4.8
2.7
1.1
5.4
24
O
O
O
O
N
H
N
4.6
3.1
nt
N
N
N
N
S
S
N
H
N
n-Pr
(CH2)2Ph
0.76
1.3
1.6
1.8
4.6
2.0
>20
NH2
NH2
1a
1b
O
O
(CH2)2OH
(CH2)3OH
(CH2)3NH2
0.19
0.092
0.95
In this tautomer a potential H-bonding interaction is
evident, and this led us to propose a binding interaction
with ITK outlined in Figure 2. By analogy with the
structure of HCK,16 ATP binds to ITK by way of a
donor interaction with the carbonyl of Glu436, and an
nt, not tested.
a Values are geometric means of at least two experiments. In most cases
the individual results vary by less than a factor of 2.
b Mean of at least two experiments.

3664
R. J. Snow et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3660–3665
Table 3. Variation of C5 and C6 substitution
O
R5
R6
N
N
S
N
H
NH2
O
Compound
R5
R6
ITK
DT40 cell
b
IC50 (lM)
a
IC50 (lM)
1
27
47
48
49
50
51
16
c-HexN(Me)CO
c-HexNHCO
H
H
H
0.11
>20
0.71
1.5
2.7
nt
c-HexN(Me)CO
20
12
nt
H
MeO2C
H
H
H
H
H
11
c-HexO2C
c-HexCONH
c-HexCONMe
>20
2.5
nt
1.7
2.4
0.025
nt, not tested.
a Values are geometric means of at least two experiments. In most cases the individual results vary by less than a factor of 2.
b Mean of at least two experiments.
Table 4. Energy difference between the cis and trans conformers
corresponding to compounds in Table 3
Table 5. Kinase selectivity of compound 16
a
Kinase
IC50 (lM)
Kinaseb
% inhibitionc
O
BMX
BTK
ECK
>22
>22
>10
>6
CDK2
CDK5
7
O
Y
N
N
Y
X
À4
8
N
N
X
N
N
H
CHK1
GSK3b
JAK3
H
trans
Compound
cis
EGFR
FGFR3
HGFR
IRK
À18
À7
À2
1
>6
>10
0.2
MKK1
MAPKAPk2
PDGFR
PKCb
X
Y
Energy differencea
(kcal/mol)
LYN
TEC
>10
22
10
21
À1
6
1
27
50
16
51
N(Me)
NH
C@O
C@O
C@O
2.8
3.8
9.3
0.3
2.2
TXK
VEGFR1
1.4
>10
ROCKII
SYK
O
C@O
C@O
N(Me)
NH
a Assays run in dose response.
b Screening at a single concentration.
a Difference between the minimum energy conformers of the cis and
trans isomers of the above structure, calculated using Cerius 2
(Accelerys Inc.).
c % inhibition at 10 lM.
Glu436
CO2H
To provide support for the involvement of tautomer 1b,
an analog of 16 was subjected to methylation, which oc-
curred on the ring nitrogen, and resulted in loss of
activity.20
S
Met438
O
H
O
H
N
N
N
N
H
H
H
N
O
O
O
H
N
O
N
The improved potency of the reverse amide 16 resulted
in this compound becoming the program lead. To con-
firm that these compounds are ATP-competitive the en-
zyme assay was run at four ATP concentrations: 3, 20,
100, and 500 lM. IC50 values of 0.11, 0.32, 1.6, and
>5 lM, respectively, were observed for 1. The corre-
sponding results for 16 were 0.025, 0.08, 0.24, and
1.1 lM. 16 was profiled against more than 100 other ki-
nases. Selected data are shown in Table 5. The main
cross-reactivity was with insulin receptor kinase (IRK),
and to a lesser extent with TXK, also a member of the
Tec family.
Phe435
N
S
N
H2N
Lys391
NH2
O
Figure 2. Proposed binding interaction of compound 1, as the
exocyclic imine tautomer, with the hinge region of ITK.
acceptor interaction with the NH of Met438. This inter-
action is observed with many kinase inhibitors, but a
second binding mode has also been observed in which
both donor and acceptor interactions are with a single
residue,17–19 corresponding to Met438 in ITK. This
seems much more likely for these benzimidazoles than
an alternative with the ATP mode, which would reverse
the orientation of the core.
Optimization of all three substituents of compound 16
to improve molecular and cellular potency and evalu-
ate in vivo activity will be reported in subsequent
papers.

R. J. Snow et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3660–3665
3665
Kamath, A.; Zhang, H.; Marathe, P.; Kanner, S. B.;
Lin, T. A.; Dodd, J. H.; Barrish, J. C.; Wityak, J. Bioorg.
Med. Chem. Lett. 2006, 16, 3706.
References and notes
1. Gibson, S.; Leung, B.; Squire, J. A.; Hill, M.; Arima, N.;
Goss, P.; Hogg, D.; Mills, G. B. Blood 1993, 82, 1561.
2. Siliciano, J. D.; Morrow, T. A.; Desiderio, S. V. Proc.
Natl. Acad. Sci. U.S.A. 1992, 89, 11194.
3. Yamada, N.; Kawakami, Y.; Kimura, H.; Fukamachi, H.;
Baier, G.; Altman, A.; Kato, T.; Inagaki, Y.; Kawakami,
T. Biochem. Biophys. Res. Commun. 1993, 192, 231.
4. Heyeck, S. D.; Wilcox, H. M.; Bunnell, S. C.; Berg, L. J.
J. Biol. Chem. 1997, 272, 25401.
5. Bunnell, S. C.; Diehn, M.; Yaffe, M. B.; Findell, P. R.;
Cantley, L. C.; Berg, L. J. J. Biol. Chem. 2000, 275, 2219.
6. Kawakami, Y.; Kitaura, J.; Hata, D.; Yao, L.; Kawakami,
T. J. Leukocyte Biol. 1999, 65, 286.
7. Liao, X. C.; Littman, D. R. Immunity 1995, 3, 757.
8. Fowell, D. J.; Shinkai, K.; Liao, X. C.; Beebe, A. M.;
Coffman, R. L.; Littman, D. R.; Locksley, R. M.
Immunity 1999, 11, 399.
9. Mueller, C.; August, A. J. Immunol. 2003, 170, 5056.
10. Das, J.; Liu, C.; Moquin, R. V.; Lin, J.; Furch, J. A.;
Spergel, S. H.; McIntyre, K. W.; Shuster, D. J.; O’Day, K.
D.; Penhallow, B.; Hung, C. Y.; Kanner, S. B.; Lin, T. A.;
Dodd, J. H.; Barrish, J. C.; Wityak, J. Bioorg. Med. Chem.
Lett. 2006, 16, 2411.
11. Das, J.; Furch, J. A.; Liu, C.; Moquin, R. V.; Lin, J.;
Spergel, S. H.; McIntyre, K. W.; Shuster, D. J.; O’Day, K.
D.; Penhallow, B.; Hung, C. Y.; Doweyko, A. M.;
12. Kashem, M. A.; Nelson, R. M.; Yingling, J. D.; Pullen, S.
S., ; Prokopowicz, A. S., III; Jones, J. W.; Wolak, J. P.;
Rogers, G. R.; Morelock, M. M.; Snow, R. J.; Homon, C.
A.; Jakes, S. J. Biomol. Screen. 2007, 12, 70.
13. Lee, J.; Doucette, A.; Wilson, N. S.; Lord, J. Tetrahedron
Lett. 2001, 42, 2635.
14. Takata, R.; Kurosaki, T. J. Exp. Med. 1996, 184, 31.
15. Grundemann, E.; Graubaum, H.; Martin, D.; Schiewald,
E. Magn. Reson. Chem. 1986, 24, 21.
16. Sicheri, F.; Moarefi, I.; Kuriyan, J. Nature 1997, 385,
602.
17. Schulze-Gahmen, U.; Brandsen, J.; Jones, H. D.; Morgan,
D. O.; Meijer, L.; Vesely, J.; Kim, S. H. Proteins 1995, 22,
378.
18. Mohammadi, M.; Froum, S.; Hamby, J. M.; Schroeder,
M. C.; Panek, R. L.; Lu, G. H.; Eliseenkova, A. V.; Green,
D.; Schlessinger, J.; Hubbard, S. R. EMBO J. 1998, 17,
5896.
19. Snow, R. J.; Cardozo, M. G.; Morwick, T. M.; Busacca,
C. A.; Dong, Y.; Eckner, R. J.; Jacober, S.; Jakes, S.;
Kapadia, S.; Lukas, S.; Panzenbeck, M.; Peet, G. W.;
Peterson, J. D.; Prokopowicz, A. S., III; Sellati, R.;
Tolbert, R. M.; Tschantz, M. A.; Moss, N. J. Med. Chem
2002, 45, 3394.
20. Nemoto, P. unpublished observation.