Replacement of pyrazol-3-yl amine hinge binder with thiazol-2-yl amine: Discovery of potent and selective JAK2 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2010.01.091

Thiazol-2-yl amine was identified as an isosteric replacement for pyrazol-3-yl amine during our efforts to identify potent and selective JAK2 inhibitors. The rationale, synthesis and biological evaluation of several analogs is reported, along with the in vivo evaluation of the lead compounds.

Full text of DOI:10.1016/j.bmcl.2010.01.091

Bioorganic & Medicinal Chemistry Letters 20 (2010) 1669–1673
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Replacement of pyrazol-3-yl amine hinge binder with thiazol-2-yl amine:
Discovery of potent and selective JAK2 inhibitors
a
a
a
a
Stephanos Ioannidis a, , Michelle L. Lamb , Lynsie Almeida , Huiping Guan , Bo Peng ,
*
Geraldine Bebernitz b, Kirsten Bell b, Marat Alimzhanov b, Michael Zinda b
a Department of Cancer Chemistry, AstraZeneca R&D Boston, 35 Gatehouse Drive, Waltham, MA 02451, USA
b Department of Cancer Bioscience, AstraZeneca R&D Boston, 35 Gatehouse Drive, Waltham, MA 02451, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Thiazol-2-yl amine was identified as an isosteric replacement for pyrazol-3-yl amine during our efforts to
identify potent and selective JAK2 inhibitors. The rationale, synthesis and biological evaluation of several
analogs is reported, along with the in vivo evaluation of the lead compounds.
Ó 2010 Elsevier Ltd. All rights reserved.
Received 10 December 2009
Revised 7 January 2010
Accepted 11 January 2010
Available online 25 January 2010
Keywords:
Pyrazol-3-yl amine
Thiazol-2-yl amine
JAK2 inhibitors
AZ960
As part of our efforts to identify potent and selective JAK2 (Janus
Kinase 2) inhibitors, we have recently described pyrazol-3-yl ami-
no nicotinonitrile (AZ960, Fig. 1) and pyrazol-3-yl amino pyrazines
as ATP-competitive compounds.1,2 Based on past experience with
kinase inhibitors,3 we anticipated that for both scaffolds, the pyra-
zol-3-yl amine group occupies the ATP binding site and interacts
via hydrogen bonds with the JAK2 protein backbone (hinge). More
specifically, three hydrogen bond interactions in a cis-donor/
acceptor/donor motif were expected with the pyrazol-3-yl amine.
Based upon this binding mode, any substituents at the C5 of the
pyrazole ring would be stacked against the selectivity pocket gate-
keeper, which for JAK2 is a methionine residue. Earlier work by
Pierce et al.4 with quinazoline-containing kinase inhibitors sug-
gested for this binding motif that a further intra-molecular hydro-
gen bond between the C4–H of the pyrazole and the adjacent
nitrogen of the pyridine/pyrazine might be crucial for activity
(see Fig. 1). This interaction would lock the two rings into a co-pla-
nar conformation and permit efficient interaction of the pyrazole
with the hinge.
teric replacement of the pyrazol-3-yl amine group with thiazol-2-
yl amine group should result in potent JAK2 inhibitors with our pyr-
azine scaffold.2 The thiazol-2-yl amine moiety is present in several
reported kinase inhibitors such as dasatinib.6The hypothesis was
the cis-donor motif would be retained by the C2–NH and C4–H of
the thiazole with the predicted increase in the basic character of
thiazole nitrogen enhancing the acceptor potential with the back-
bone, and balancing the weaker acidity of the C–H bond.
Hinge
F
H
N
Solvent
N
4
NH
N
F
H
N
HN
Selectivity
Pocket
Interestingly, these authors found that a thiazole could replace
the pyrazole and retain potency for their quinazoline analogs. This
was explained by the aromatic C–H adjacent to a heteroatom par-
ticipating in a hydrogen bond as donor4,5 and a favorable electro-
static interaction between the sulfur and the nitrogen of the
adjacent ring.4 With this background, we considered that bioisos-
AZ960
* Corresponding author. Tel.: +1 781 839 4556; fax: +1 781 839 4630.
E-mail address: Stephanos.Ioannidis@astrazeneca.com (S. Ioannidis).
Figure 1. Intramolecular hydrogen bond between C-4 and nitrogen of the pyridine
locks the two rings in co-planar conformation in AZ960.
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.01.091

1670
S. Ioannidis et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1669–1673
Table 1
B-ring
N
H
N
Comparison of thiazol-2-yl pyrazine and pyrazol-3-yl pyrazine 3 and 4
N
A-ring
H
N
H
N
S
N
F
2
Solvent
Tail
N
N
δ
R
δ
N
N
R1
NH
HN
Y
S
N
N
HN
HN
N
N
N
N
N
C-ring
F
F
Figure 2. Proposed electrostatic interaction between the N (dÀ) of the pyrazine and
the sulfur (d+) of the thiazole.
3
4
Compd
JAK2 IC50
(lM)
JAK2 (5 mM ATP) IC50
,
lM
Solubilityb
(lM)
a
3
4
0.004
<0.003
0.049
0.003
>1000
>1000
Likewise, the co-planarity between the thiazole and the B-ring
(Fig. 2) would be sustained via a favorable electrostatic interaction
between the nitrogen (dÀ) of the B-ring and the sulfur of the thia-
zole (d+) as was seen with the pyrazol-3-yl quinazoline inhibitors.
We envisioned that prompt confirmation of whether the thia-
zole replacement is tolerated within the pyrazine series could be
determined by the synthesis of compound 3 (Scheme 1).
Compound 3 was readily accessible via nucleophilic aromatic
substitution of 2,6-dichloropyrazine with the chiral amine 1 under
microwave conditions. The thiazole moiety was installed by a stan-
dard palladium-catalyzed amination reaction. We were gratified to
find that compound 3 showed excellent JAK2 inhibition, compara-
ble to compound 4.
As shown in Table 1, both molecules were potent in the JAK2
enzyme assay at Km ATP,1 indicating that the thiazol-2-yl group
binds sufficiently to the hinge. The thiazole showed decreased po-
tency in JAK2 inhibitory activity at high concentrations of ATP in
comparison with the corresponding pyrazole analog. However, it
was not possible to estimate the exact drop off in potency since
JAK2 inhibition at Km was at the detection limit of the assay. It
would be expected that for a tightly-bound, ATP-competitive
inhibitor, the drop off in JAK2 potency between Km ATP and high
ATP (5 mM) concentrations would be small. The use of a high con-
centration of ATP in JAK2 enzyme screening was anticipated to mi-
mic the cellular environment and allow more accurate
differentiation between the compounds tested. Thus, we envisaged
that reduced inhibitory activity in the high ATP JAK2 enzyme assay
might translate into weaker cellular activity. Hence we were not
surprised when testing compound 3 in a TEL-JAK2 proliferation as-
a
At Km ATP concentration.
Obtained from solid after drying its DMSO solution.
b
an attempt to improve these properties in the thiazole series, we
considered the installation of several groups (such as morpholine)
at the solvent tail position R2. Despite our efforts, a viable synthetic
route allowing attachment of a solvent tail to the pyrazine scaffold
was not identified; therefore, we turned our attention to thiazol-2-
yl pyrimidines in order to test our hypothesis.
We anticipated that introduction of the solvent tail first, fol-
lowed by the addition of the C-ring and subsequent palladium-cat-
alyzed amination would furnish the desired thiazol-2-ylamines as
shown in Scheme 2. Reaction of amine 1 with intermediate 5 re-
sulted in the formation of two regio-isomers with the chiral amine
attached either at the 2- or 4-position of the pyrimidine (only the
desired regio-isomer is depicted in Scheme 2). Separation of the re-
gio-isomers was readily accomplished by column chromatography.
Table 2 illustrates the JAK2 enzymatic and cellular (TEL-JAK2)
potencies together with a measure of JAK3 selectivity, as deter-
mined in a TEL-JAK3 proliferation assay in Ba/F3 cells.1 This mea-
surement was triggered because of the potential for
immunosuppression due to attenuated JAK3 activity.
Both 6a and 6b showed excellent JAK2 activity at high ATP con-
centrations, implying tight binding of these analogs with the JAK2
protein. In addition, the nature of the substituent at the solvent tail
position appeared important in determining the level of selectivity
versus JAK3. The morpholine analog 6b exhibited excellent po-
tency against JAK2 along with exquisite selectivity versus JAK3.
In contrast, the JAK2 potency of 6a did not appear to translate to
the cellular level.
say revealed only moderate activity (GI50 ꢀ0.37
lM). In vitro
screening in rat microsomes showed the compound to have mod-
erate intrinsic clearance (CLint = 52 lL/min/mg). It was known from
previous studies that the substituents at R2 (Fig. 2) could effec-
tively improve cellular potency as well as metabolic stability. In
We extended our efforts by replacing the (S)-5-fluoropyrimidin-
2-yl-ethanamine C-ring with (S)-5-fluoropyridin-2-yl-ethanamine.
H2N
HCl
H
N
N
N
N
N
H2N
N
H
N
S
N
N
S
Cl
F
N
HN
N
N
N
2
b
1
a
N
Cl
N
N
Cl
F
F
3
Scheme 1. Reagents and conditions: (a) NMP, Hünig’s base, 180 °C, microwave; (b) Cs2CO3, Pd2(dba)3, XantphosÒ, dioxane, 100 °C.

S. Ioannidis et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1669–1673
1671
H
N
R2
Cl
R2
N
N
N
F
R2
S
N
N
F
Cl
Cl
Cl
HN
HN
N
N
N
N
a
b
c
Cl
R3=solvent tail
Cl
5
N
N
N
N
6a-b
Scheme 2. Reagents and conditions: (a) for 6a: NaOMe, MeOH, 0–25 °C; for 6b: morpholine, EtOH, À20 °C to 0 °C; (b) (S)-5-fluoropyrimidin-2-yl-ethanamine hydrochloride
1, Hünig’s base, n-BuOH, 120 °C; (c) 5-methyl-1,3-thiazol-2-amine, Cs2CO3, Pd2(dba)3, BINAP, dioxane, 95 °C.
Table 2
Interestingly, while the morpholine analog 7b showed excellent
selectivity vs JAK3, the 1-methylpiperazine, although very potent,
seemed to be less selective (Table 3).
Enzymatic and cellular evaluation of (S)-N2-(1-(5-fluoropyrimidin-2-yl)ethyl)-N4-(5-
methylthiazol-2-yl)pyrimidines
The methyl group of the thiazole is presumed to occupy the
selectivity pocket. It was speculated that only small groups could
be accommodated in the selectivity pocket due to steric interac-
tions with the Met gatekeeper in JAK2. Hence, we investigated
replacement of 5-methylthiazol-2-amine with 2-aminothiazole-
5-carbonitrile.7 The morpholine solvent tail was retained due to
the excellent potency (6b and 7b) and selectivity (see 6b vs 6a
and 7b vs 7a) against JAK3 demonstrated by the analogs described
above. The effect of the two different C-rings was also tested
(Table 4).
H
R2
N
N
S
N
N
HN
N
N
F
To our surprise, incorporation of the cyano group appeared to
result in both a reduction in JAK2 activity, and a loss of selectivity
against JAK3.
Compd R3
JAK2a
TEL-JAK2
GI50 M)
TEL-JAK3
JAK3 fold
IC50
,
l
M
(l
GI50
(lM)
selectivityb
6a
6b
OMe
Morpholine
<0.003
0.003
0.042
0.005
1.5
1.1
35
237
The thiazol-2-yl pyrimidines were screened in an in vitro
metabolic stability assay in rat microsomes and their potential to
inhibit the hERG channel was assessed. In addition, the pharmaco-
kinetic (PK) profile of the most potent and selective JAK2 inhibitors
(6b, 7a and 7b) was determined following iv dosing in rats (Ta-
ble 5). All compounds containing the 5-methylthiazol-2-amine
hinge binder displayed high intrinsic clearance, while the 2-amino-
thiazole-5-carbonitrile analogs (8a and 8b) appeared to be more
stable. This is possibly because the methyl group is prone to oxida-
tion by CYP450 enzymes while such biotransformation is not pos-
sible for the cyano group.8 Likewise, comparing the rat microsomal
stability between 3 and 4 (Table 5) suggests the metabolic instabil-
ity of 5-methylthiazol-2-amine.
a
At 5 mM ATP concentration.
b
Calculated from TEL-JAK3 to TEL-JAK2 ratio.
Table 3
Enzymatic and cellular evaluation of (S)-N2-(1-(5-fluoropyridin-2-yl)ethyl)-N4-(5-
methylthiazol-2-yl)pyrimidines
H
R2
N
N
N
N
S
Table 4
HN
The effect of 2-aminothiazole-5-carbonitrile hinge binder
N
O
H
N
N
N
F
N
N
S
Compd R3
JAK2a
IC50
TEL-JAK2
GI50 M)
TEL-JAK3
GI50 (lM)
JAK3 fold
CN
selectivityb
HN
(
l
(lM)
Y
N
7a
7b
1-Methyl
piperazine
Morpholine
<0.003
0.005
0.005
0.62
1.2
129
243
0.005
F
a
At 5 mM ATP concentration.
Calculated from TEL-JAK3 to TEL-JAK2 ratio.
b
Compd
Y
JAK2a IC50
M)
TEL-JAK2 GI50 TEL-JAK3 GI50 JAK3 fold
(
l
(
lM)
(
lM)
selectivityb
8a
8b
CH <0.003
0.079
0.036
0.053
0.17
0.52
5
10
N
Analogs 7a–b were prepared according to the procedures de-
scribed in Scheme 2 for the corresponding (S)-5-fluoropyrimidin-
2-yl-ethanamine derivatives.
a
At 5 mM ATP concentration.
Calculated from TEL-JAK3 to TEL-JAK2 ratio.
b

1672
S. Ioannidis et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1669–1673
Table 5
Despite the less than ideal PK profile of these analogs in rats,
b
Rat microsomal stability,a PK parameters,
thiazol-2-yl pyrimidines
hERG and lipophilicity evaluation of
compounds 6b, 7b, and 9 were administered orally (q.d.) at a dose
of 10 mg/kg in nude mice bearing TEL-JAK2 transfected Ba/F3 cells.
The percentage (%) inhibition of STAT5 phosphorylation (signal
transducers and activators of transcription, PSTAT5) caused by
the compounds was determined post-dose at 2 and 6 h. Compound
7b showed sustained inhibition of PSTAT5 (ꢀ50%) 6 h after dosing
(Fig. 3). In contrast 6b only showed substantial inhibition for 2 h
after dosing while at the same dose, compound 9 did not appear
to prevent the phosphorylation of STAT5 to any appreciable level.
The total plasma concentrations of each compound are illustrated
in Figure 3. From the data it can be deduced that compound 6b
has a more a profound effect on PSTAT5 inhibition at 2 h (>80%)
due to the higher plasma concentration than the other two com-
pounds. For 7b, the PK/PD correlation likely reflects the excellent
potency of this molecule in the TEL-JAK2 proliferation assay. In
contrast, 9 did not show any substantial PSTAT5 inhibition pre-
sumably by being about sixfold weaker than 7b in TEL-JAK2 assay
although the total plasma concentrations looked similar.
Compd Rat CLint
Rat CL
Rat T1/2 Vdss
L/min/mg) (ml/min/Kg) (h)
hERG IC50 c log P
M)
(l
(L/Kg)
(l
3
4
52
<4
NT
14
NT
32
80
38
NT
NT
NT
1.4
NT
1.3
1.3
1.2
NT
NT
NT
1.1
NT
2.3
8.5
2.8
NT
NT
NT
1.75
1.25
2.68
1.94
3.39
2.95
1.93
0.94
>32
>32
>32
13
12
>32
NT
6a
6b
7a
7b
8a
8b
>100
84
220
>120
31
41
a
The time-dependent disappearance of compound (2 lM initial concentration)
incubated with microsomes is measured by LC–MS/MS.
b
Han Wistar rat male; 3 mg/kg iv.
Results from the in vivo PK screening revealed a lack of correla-
tion between the in vitro and in vivo determinations for the thia-
zole analogs. Compounds 6b and 7b showed moderate plasma
stability while compound 7a was highly cleared from the plasma.
The only difference between 7a and 7b is the 1-methylpiperazine
versus the morpholino solvent tail, suggesting that the 1-methylpi-
perazine may be more susceptible to metabolism.
From the results shown in Table 5, it also appeared to us that
incorporation of the (S)-5-fluoropyrimidin-2-yl-ethanamine C-ring
resulted in negligible inhibition of the hERG potassium channel. In
contrast, the more basic pyridine C-ring (7a–7b) along with in-
creased lipophilicity as indicated by calculated log P (c log P) led
to compounds having hERG channel activity. Interestingly, chang-
ing the C-5 substituent on the thiazole ring from Me (7a) to CN (8a)
reduces the hERG liability presumably due to the reduced lipo-
philic nature of the cyano group.9
120
100
80
60
40
20
0
800
700
600
500
400
300
200
100
0
In an attempt to further improve metabolic stability while
retaining JAK2 activity and JAK3 selectivity, we sought to prepare
analog 9 (Scheme 3) as it was speculated that the ‘naked’ C-5 posi-
tion of the pyrimidine B-ring might constitute another metabolic
liability.
The pyrimidine 10 was readily prepared via condensation of
urea with methyl fluoromalonate followed by subsequent chlori-
nation of the intermediate 5-fluoropyrimidine-2,4,6(1H,3H,5H)-tri-
one with POCl3.
Figure 3. Pharmacodynamic effect of compounds 6b, 7b, and 9 on PSTAT5 in the
Ba/F3 TEL-JAK2 mouse model and relationship with plasma pharmacokinetics. The
bars represent the average percentage (%) inhibition in phosphorylation of STAT5
( SD, n = 3 for each dose and each time point) and are calculated using vehicle and
control inhibitor for maximum and minimum value estimation. Nude mice (female)
were implanted intravenously with Ba/F3 TEL-JAK2 cells on day 0 and were dosed
on day 10. The squares with numbers represent the corresponding average total
plasma drug concentration. Mice used in these studies were maintained under
specific pathogen-free conditions and were used in compliance with protocols
approved by the Institutional Animal Care and Use Committees of AstraZeneca,
which conform to institutional and national regulatory standards on experimental
animal usage. The pharmacokinetic samples were run on an Applied Biosystems
API4000 QTrap using AnalystÒ software version 1.4.2. The sample analyses were
Compound
9 exhibited good cellular activity (TEL-JAK2
GI50 = 0.03 M) and excellent JAK3 selectivity (TEL-JAK3/TEL-
l
JAK2 = 161-fold). However, rat pharmacokinetic analysis after iv
dosing indicated no improvement in the observed clearance
(CL = 41 mL/min/kg) compared to 6b. We therefore reached the
conclusion that in rats, the methyl group of the thiazole hinge bin-
der most likely appears to constitute the major metabolic liability.
carried out in Sciex#5 LC–MS, using an ACE 3C18 3
lm 20 Â 2.1 mm column at
ambient temperature and the following eluents: Mobile Phase A: water + 0.1%
formic acid + 10 mM ammonium formate B: MeCN + 0.1% formic acid.
O
F
O
c
F
H
N
N
Cl
N
N
N
F
O
F
N
N
F
S
N
N
Cl
Cl
N
Cl
HN
N
HN
N
N
N
N
N
a
b
Cl
Cl
N
10
F
9
Scheme 3. Reagents and conditions: (a) morpholine, EtOH, À20 °C to 0 °C; (b) (S)-5-fluoropyrimidin-2-yl-ethanamine hydrochloride (1), Hünig’s base, n-BuOH, 120 °C; (c) 5-
methyl-1,3-thiazol-2-amine, Cs2CO3, Pd2(dba)3, BINAP, dioxane, 95 °C.

S. Ioannidis et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1669–1673
1673
3. Wang, T.; Lamb, M. L.; Scott, D. A.; Wang, H.; Block, M. H.; Lyne, P. D.; Lee, J. W.;
Davies, A. M.; Zhang, H.; Zhu, Y.; Gu, F.; Han, Y.; Wang, B.; Mohr, P. J.; Kaus, R. J.;
Josey, J. A.; Hoffmann, E.; Thress, K.; MacIntyre, T.; Wang, H.; Omer, C. A.; Yu, D. J.
Med. Chem. 2008, 51, 4672.
4. Pierce, A. C.; ter Haar, E.; Binch, A. M.; Kay, D. P.; Patel, S. R.; Li, P. J. Med. Chem.
2005, 48, 1278.
5. Pierce, A. C.; Sandretto, K. L.; Bemis, G. W. Proteins: Structure, Function, and
Genetics 2002, 49, 567.
6. Das, J.; Chen, P.; Norris, D.; Padmanabha, R.; Lin, J.; Moquin, R. V.; Shen, Z.; Cook,
L. S.; Doweyko, A. M.; Pitt, S.; Pang, S.; Shen, D. R.; Fang, O.; de Fex, H. F.;
McIntyre, K. W.; Shuster, D. J.; Gillooly, K. M.; Behnia, K.; Schieven, G. L.; Wityak,
J.; Barrish, J. C. J. Med. Chem. 2006, 49, 6819.
7. Procedure: To a solution of a mixture of 2-(E and Z)-3-methoxyacrylonitrile
(0.18 mmol, 1.51 mL) in MeCN (3 ml) was added bromine at 0 °C (ꢀ1 mL,
0.18 mmol). The resulting mixture was stirred at this temperature for 30 min,
whereupon cold water (12 ml) was added. After vigorous stirring for 1 h,
NaOAcÁ3H2O (16 mmol) was added and stirred for 15 min. Thiourea (19.8 mmol)
was added and the solution was stirred for 2 h at 0 °C, additional NaOAcÁ3H2O
(10.8 mmol) was added and the resulting mixture was heated to 60 °C for 1 h.
The mixture was allowed to cool to room temperature and stirred overnight at
this temperature. 10 N NaOH (aq) was added slowly into the reaction mixture
until the pH ꢀ4. 2-Aminothiazole-5-carbonitrile was collected by filtration after
being washed with cold H2O. LC–MS: 126 [M+H]+. 1H NMR (300 MHz, DMSO-d6)
d 8.10 (s, 2H), 7.81 (s, 1H).
In conclusion, we have demonstrated that thiazol-2-yl amine
can effectively replace the pyrazol-3-yl amine, leading to potent
JAK2 ATP-competitive inhibitors such as 6b, 7a–b, and 9. The intro-
duction of solvent tail substituents, such as morpholine, resulted in
an improvement in cellular activity. Likewise significant selectivity
against JAK3 at the cellular level was observed in some compounds
containing solvent tails. However, the moderate pharmacokinetic
properties of these compounds in rodents has limited their
in vivo activity as shown by the results from testing in the reported
pharmacodynamic model.
Acknowledgments
The authors would like to thank Ethan Hoffmann for PK studies
and Jing Dai for PK samples analysis. Special thanks to Susan Ash-
well and John Haberman for useful recommendations during the
preparation of this Letter.
References and notes
8. Schmid, J.; Busch, U.; Trummlitz, G.; Prox, A.; Kaschke, S.; Waschsmuth, H.
Xenobiotica 1995, 25, 1219.
9. Bilodeau, M. T.; Balitza, A. E.; Koester, T. J.; Manley, P. J.; Rodman, L. D.; Buser-
Doepner, C.; Coll, K. E.; Fernandes, C.; Gibbs, J. B.; Heimbrook, D. C.; Huckle, W.
R.; Kohl, N.; Lynch, J. J.; Mao, X.; McFall, R. C.; McLoughlin, D.; Miller-Stein, C. M.;
Rickert, K. W.; Sepp-Lorenzino, L.; Shipman, J. M.; Subramanian, R.; Thomas, K.
A.; Wong, B. K.; Yu, S.; Hartman, G. D. J. Med. Chem. 2004, 47, 6819.
1. Gozgit, J. M.; Bebernitz, G.; Patil, P.; Ye, M.; Wu, J.; Su, N.; Wang, T.; Ioannidis, S.;
Davies, A. M.; Huszar, D.; Zinda, M. J. Biol. Chem. 2008, 283, 32334.
2. Ioannidis, S.; Lamb, M. L.; Davies, A. M.; Almeida, L.; Su, M.; Bebernitz, G.;
Ye, M.; Bell, K.; Alimzhanov, M.; Zinda, M. Bioorg. Med. Chem. Lett. 2009, 19,
6524.