Discovery of novel Jak2-Stat pathway inhibitors with extended residence time on target

Article abstract of DOI:10.1016/j.bmcl.2013.02.111

The discovery of the activating mutation V617F in the JH2 domain of Jak2 and the modulation of oncogenic Stat3 by Jak2 inhibitors have spurred a great interest in the inhibition of the Jak2/Stat pathway in oncology. In this Letter, we communicate the discovery of novel inhibitors of the Jak2/Stat5 axis, the N-(1H-pyrazol-3-yl)pyrimidin-2-amino derivatives. The rationale, synthesis and biological evaluation of these derivatives are reported. Two lead analogs from this series, 6 and 9, displayed prolonged residence time on Jak2, at enzymatic level. Although 6 and 9 exhibited moderate selectivity in a selected kinase panel, we chose to test these inhibitors in vivo as a consequence to their long residence time. However, extended inhibition of Jak2 due to the long residence time, in the form of inhibiting phosphorylation of downstream Stat5, was not recapitulated in an in vivo setting.

Full text of DOI:10.1016/j.bmcl.2013.02.111

Bioorganic & Medicinal Chemistry Letters 23 (2013) 3105–3110
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of novel Jak2–Stat pathway inhibitors with extended
residence time on target
a
a
a
a
a
b
b
Huiping Guan , Michelle L. Lamb , Bo Peng , Shan Huang , Nancy DeGrace , Jon Read , Syeed Hussain ,
c
d
c
c
c
c
Jiaquan Wu , Caroline Rivard , Marat Alimzhanov , Geraldine Bebernitz , Kirsten Bell , Minwei Ye ,
c
a,
⇑
Michael Zinda , Stephanos Ioannidis
a
Department of Cancer Chemistry, Oncology Innovative Medicines Unit, AstraZeneca R&D Boston, 35 Gatehouse Drive, Waltham, MA 02451, USA
Discovery Enabling Capabilities and Sciences-Cells, Protein & Structural Sciences AstraZeneca, Alderley Park, Cheshire SK10 4TG, UK
Department of Cancer Bioscience, Oncology Innovative Medicines Unit, AstraZeneca R&D Boston, 35 Gatehouse Drive, Waltham, MA 02451, USA
DMPK Department, Oncology Innovative Medicines Unit, AstraZeneca R&D Boston, 35 Gatehouse Drive, Waltham, MA 02451, USA
b
c
d
a r t i c l e i n f o
a b s t r a c t
Article history:
The discovery of the activating mutation V617F in the JH2 domain of Jak2 and the modulation of onco-
genic Stat3 by Jak2 inhibitors have spurred a great interest in the inhibition of the Jak2/Stat pathway in
oncology. In this Letter, we communicate the discovery of novel inhibitors of the Jak2/Stat5 axis, the
N-(1H-pyrazol-3-yl)pyrimidin-2-amino derivatives. The rationale, synthesis and biological evaluation
of these derivatives are reported. Two lead analogs from this series, 6 and 9, displayed prolonged
residence time on Jak2, at enzymatic level. Although 6 and 9 exhibited moderate selectivity in a selected
kinase panel, we chose to test these inhibitors in vivo as a consequence to their long residence time. How-
ever, extended inhibition of Jak2 due to the long residence time, in the form of inhibiting phosphorylation
of downstream Stat5, was not recapitulated in an in vivo setting.
Received 11 January 2013
Revised 18 February 2013
Accepted 25 February 2013
Available online 7 March 2013
Keywords:
Jak
Myeloproliferative neoplasms
N-(1H-Pyrazol-3-yl)pyrimidin-2-amines
V617F mutation
Ó 2013 Elsevier Ltd. All rights reserved.
Residence time
The Jak (Janus kinase)-Stat (Signal transducers and activators of
transcription) pathway has attracted great attention in cancer
therapy, after the identification of the mutant Jak2 V617F protein,
O
O
F
H
N
H
N
N
F
N
N
N
N
S
1
N
F
S
N
N
which displayed a pivotal role in myeloproliferative neoplasms.
Likewise, the demonstration of modulation of oncogenic Stat3 by
Jak2 inhibitors, such as AZD1480, in solid tumors has expanded
the potential utility of such agents.2
HN
N
HN
N
,3
N
N
Our laboratories have disclosed several Jak-Stat pathway inhib-
itors.4 Recently, we described the discovery of a series of N -
–7
4
F
(
thiazol-2-yl)pyrimidine-2,4-diamines as ATP-competitive Jak2
6
1
2
inhibitors. These inhibitors displayed good activity against Jak2,
both in in vitro and in vivo settings, and sufficient selectivity
against Jak3. To further expand our SAR efforts, we envisioned
Comparing compound 1 versus 2 in the Jak2 enzymatic assay
revealed that indeed the change in orientation of the B-ring was
tolerated and this analog displayed selectivity against Jak3 at both
biochemical and cellular levels as shown in Table 1. We decided to
proceed with in vivo screening of compound 1 in nude mice bear-
ing TEL-Jak2 transfected Ba/F3 cells via oral administration (q.d.) at
a dose of 10 mg/kg. The percentage (%) downstream inhibition of
Stat5 phosphorylation (pStat5) caused by 1 was determined at 2
and 6 h post-dose. In this particular case, some inhibition of pStat5
4
the modification by design of the N -(thiazol-2-yl)pyrimidine-
2
,4-diamine scaffold by moving the thiazol-2-yl amino hinge tem-
plate from the C to the C of the pyrimidine B-ring (Fig. 1). We
4
2
anticipated that this change would be well tolerated since the
interaction of the thiazoly-3-yl amino group with the kinase hinge
residues would be retained.
(
ꢀ50%) was observed at 2 h while no appreciable inhibition of the
phosphorylation was evident after 6 h. Although compound 1 had
exhibited significant growth inhibition of TEL-Jak2 cells in vitro
(Table 1), its in vivo activity was considered sub-optimal with
⇑
Corresponding author. Tel.: +1 781 839 4556; fax: +1 781 839 4120.
E-mail address: stephanos.ioannidis@astrazeneca.com (S. Ioannidis).
0
960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.02.111

3
106
H. Guan et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3105–3110
R2
R2
B-ring
N
the pyrimidine B-ring. Addition of the C-ring proceeded at elevated
temperature, resulting in mixtures of C and C adducts.
X
H
N
H
N
2
6
N
N
N
N
The two regioisomers were readily separated by either liquid
chromatography or Super Critical Fluid Chromatography (SFC).
The desired regio-isomer could be assigned unambiguously based
Solvent
Tail
Solvent
Tail
S
N
N
F
S
N
X
R1
A-ring
R1
A-ring
B-ring
HN
Y
HN
1
9
upon H NMR. With the desired regio-isomer in hand, the synthe-
sis was completed by the attachment of the Boc-protected pyrazole
N
Y
N
1
0
via Buchwald–Hartwig amination conditions, under microwave
irradiation. It is noteworthy that the protecting group was re-
C-ring
C-ring
F
11
moved under the employed experimental conditions. We priori-
tized incorporation of morpholine as the solvent tail group
because, in the past, this ring had resulted in favorable cellular
activity. On this basis, several 4-morpholino-N-(1H-pyrazol-3-
yl)pyrimidin-2-amine derivatives were prepared and screened in
both enzymatic and primary cellular assays (Table 2). As seen in
Figure 1. The thiazol-2-yl amino hinge binder.
Table 1
Biochemical and cellular comparison of regioisomeric compounds 1 and 2, in the Jak2
and Jak3 assays
1
Table 2, the presence of the Me substituent at R seems to secure
optimal cellular activity in the TEL-Jak2 assay. Comparison of close
analogs 5 (X = N) and 6 (X = O) seemed to indicate discrimination
in Jak3 enzymatic activity with preference to compound 5. A sim-
ilar behavior was observed for the other matching pair of 7 and 8,
where the oxygen analog 8 exhibited poorer selectivity for Jak3 in
the biochemical assay. However, this trend did not seem to trans-
late into the cellular context, and similar selectivities were ob-
served for both X = N and X = O. Obviously there was no
significant difference in Jak2 activity in examples 5–10, while the
Compd
Jak2
IC50a
Jak3
TEL-Jak2
TEL-Jak3
a
b
GI50b
(lM)
(lM)
IC50
16.4
>30
(
lM)
GI50
(
lM)
1
2
0.02
0.025
0.03
0.03
2.7
4.9
a
See Ref. 5 for experimental details. Testing was performed at 5 mM concen-
tration of ATP.
b
See Ref. 4 for experimental details.
respect to what has been previously observed for 2 in a similar set-
6
3
ting. The poor in vivo performance could be attributed to the high
nature of R substituent (e.g., 11 vs 5 and 12 vs 7) did not seem
metabolic liability of this compound in rodents and high plasma
protein binding.
to affect the enzymatic activities of either Jak2 or Jak3. Interest-
8
ingly, the cellular potency of analog 11, in the TEL-Jak2 assay,
3
Encouraged by the favorable cellular activity of 1 in the TEL-
seemed to be affected due to the lack of a halogen at the R posi-
Jak2 assay, however, we sought to explore the effect of a pyrazol-
tion. At this stage we considered that the cellular activity versus
Jak2 was favorable, however we were still concerned with the
Jak3 activity being in the nanomolar range.
3
-yl amine as a hinge binder motif. We anticipated attenuation
of the metabolic liability due to the thiazole ring with the incorpo-
6
2
2
ration of pyrazole. The synthesis of analogs with the general struc-
Comparison of derivative 5 with previously reported N -(1-(5-
4
ture 3 is depicted in Scheme 1. The synthetic route commenced
fluoropyrimidin-2-yl)ethyl)-N -(5-methyl-1H-pyrazol-3-yl)-6-
from 2,4,6-trichloro pyrimidine, which was either commercially
morpholinopyrimidine-2,4-diamines 13 and 14 suggests that the
lack of substituent a-to the pyrazol-3-yl amine can have a signifi-
cant effect on Jak3 versus Jak2 selectivity on a biochemical level
3
3
available (for R = Cl) or readily accessible (for R = F). First, the sol-
vent tail was attached chemoselectively at the C position of the
4
pyrimidine B-ring at low temperature. Executing the initial step
at cooler temperatures as well as ensuring the slow addition of
the amine, we secured the mono-addition of the solvent tail onto
(Fig. 2).
Further testing revealed that compounds 6 and 9 were compet-
itive inhibitors of Jak2 with K
i
’s of 0.72 ± 0.005 and 1.41 ± 0.18 nM.
Solvent tail
R2
Cl
R3
N
Cl
2 Cl
N
N
i
ii
4
N
N
R3
Cl
Cl
R2
R2
R2
R3
H
N
N
N
Cl
N
N
N
N
Cl
N
N
NH
N
F
R3
N
F
N
N
F
R3
iii
R1
X
Y
X
Y
X
Y
N
Desired regioisomer
3
Scheme 1. Preparation of analogs 3. Reagents and conditions: (i) Amine (such as morpholine), EtOH, ꢁ20 °C; (ii) for X = N: amine such as (S)-1-(5-fluoropyridin-2-
yl)ethanamine, hydrochloride DIPEA, n-BuOH, 130 °C, o/n. For X = O: alcohol such as 1-(5-fluoropyridin-2-yl)ethanol t-BuOH, t-BuOK, 50 °C, o/n. and (iii) Boc-protected
pyrazole, Pd (dba) , BINAP, Cs CO , 180 °C.
2 3 2 3

H. Guan et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3105–3110
3107
Table 2
2
Biochemical and cellular evaluation of 6-morpholino-N -(1H-pyrazol-3-yl)pyrimidin-2-amine derivatives 4–12
O
H
N
N
N
N
NH
R1
N
F
R3
X
Y
N
Compd
R1/R3
X/Y
Jak2 IC50a
(lM)
Jak3 IC50a
(lM)
Tel-Jak2 GI50b
(l
M)
Tel-Jak3 GI50b
(lM)
4
5
6
7
8
9
0
1
2
OMe/F
Me/F
Me/F
Me/F
Me/F
Me/F
Me/F
Me/H
Me/Cl
NH/N
NH/N
O/N
NH/CH
O/CH
NH/CF
O/CF
0.037
0.003
<0.003
0.006
0.004
0.003
0.004
0.005
0.006
11.23
0.229
0.020
0.262
0.087
0.008
0.045
0.095
0.262
0.16
9.4
0.015
0.006
0.019
0.016
0.009
0.010
0.058
n.d
0.39
0.69
0.85
0.24
0.30
0.43
n.d
1
1
1
NH/N
NH/CH
n.d
a
See Ref. 5 for experimental details. Testing was performed at 5 mM concentration of ATP.
See Ref. 4 for experimental details.
b
C-5 substituent
No substituent
H
No substituent
O
O
H
O
F
H
H
N
N
N
N
N
N
N
N
N
N
NH
Me
NH
Me
NH
Me
N
N
N
N
N
F
HN
N
HN
HN
N
N
N
N
N
5
13
14
F
F
F
Jak3/Jak2 selectivity:
~76-fold
>10-fold
>10000-fold
Figure 2. Impact of the presence or absence of a substituent next to the pyrazol-3-yl amino hinge binder on Jak3 versus Jak2 selectivity.
Confirmation of compound 6 being ATP-competitive was obtained
after co-crystallization with the ATP-binding domain of Jak2. The
X-ray of the interaction of the pyrazol-3-yl amine with the hinge
is shown in Figure 3. The heteroaryl C-ring occupies the hydropho-
bic pocket created by a leucine, stacking the 5-fluoro substituent
against the glycine prior to the activation loop. The morpholine
ring is positioned towards the solvent channel however there
was not any distinguishable interaction with the protein
1
2
backbone.
Our goal was to evaluate the selectivity of this series outside the
Jak family. Compound 9 was selected and assessed for selectivity in
a panel of 82 different kinases. This compound displayed greater
than 500-fold selectivity against 46 of the kinases tested at a con-
centration of 0.1 lM. The remaining set of kinases exhibited either
strong or moderate inhibition at the same concentration. The over-
all selectivity was determined by estimating the Gini co-efficient
1
3
for this molecule (Gini co-efficient = 0.54) , which indicated a
moderate selectivity in this particular panel (Fig. 4).
Figure 3. X-ray crystal structure of compound 6 in complex with human JAK2
kinase domain (PDB code: 3zmm). The structure is oriented with the N-terminal
lobe towards the top left, the C-terminal lobe towards the bottom right and the
hinge region on the bottom left. Selected atoms from the protein are represented as
sticks and cartoon in cyan. Compound 6 is shown in green. Hydrogen bonds are
Lately, the time of engagement between an inhibitor and its
biological target has gained attention in the drug discovery field,
as long residence time may enhance the observed in vivo efficacy
of a compound. According to this paradigm, the pharmacokinetic
properties of an inhibitor have less of an impact on the observed
pharmacodynamic effect the longer the complex between inhibitor
shown as dotted yellow lines. Electron density (2F
as a wire mesh. Main chain for the P-loop above the plane of the figure has been
o
ꢁ F
c
), contoured at 1r, is shown
omitted for the purposes of clarity.

3
108
H. Guan et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3105–3110
7
exhibited less of a difference between residence times for Jak2
Compound 9
and Jak3. This kinetic observation is opposite to the thermody-
namic effect (IC50) that was observed for Jak2 and Jak3. For this
reason, compounds 6 and 9 were selected to investigate further
in our cascade, despite their moderate selectivity against Jak3
Selectivity profile
1
0
0
0
0
.00
.75
.50
.25
.00
Gini Coefficient = 0.54
(Fig. 3).
Next, we sought to further explore the SAR in this series by
looking into solvent tail replacements for morpholine. Our scope
was to identify other solvent tails with improved Jak3 selectivity
and possibly longer residence time on Jak2. Table 3 shows exam-
ples of the morpholine replacements that were investigated. Com-
pounds 13–16 displayed good activity against Jak2 in both
enzymatic and cellular settings (data not shown). Examining clo-
sely the Jak3 activity, we could discern that morpholine was opti-
mal at the biochemical level for securing selectivity versus Jak3.
Particularly in 15 and 16, a complete loss of selectivity was evident,
which, however, seemed to be restored when both of these com-
pounds were evaluated in Tel-Jak2 cellular assay (for both example
0
.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Cumulative fraction of kinases
Figure 4. Gini co-efficient of compound 9 in a Millipore panel of 82 kinases,
screened at concentration of 0.1 M. Gini co-efficient can be used as a tool to assess
quickly the selectivity of a kinase inhibitor against a large set of kinases and it is
independent of ATP concentrations. Particularly, a Gini co-efficient number close to
zero would reflect poor selectivity while a number closer to 1 denotes high degree
of selectivity versus the set of kinases evaluated.
l
1
5 and example 16 TEL-Jak2 GI50 = 0.01 lM). Finally, residence
time on Jak2 for these examples was not determined due to the
lack of improvement in selectivity versus Jak3, which made them
less attractive than the morpholine analogs.
Unable to further improve the selectivity versus Jak3, we eval-
uated 9 in the Ba/F3 Tel-Jak2 cell line to establish whether the
anti-proliferative effect observed in these cells upon treatment
would correlate with inhibition of pStat5. Indeed, it was found that
the GI50 obtained was closely related to the resulting pStat5 IC50
and target lasts.14 Thus a fast clearing drug may still achieve appre-
ciable inhibition of the target due to its long residence time. In
addition, the combination of long residence time on-target and
the simultaneous elimination of the inhibitor from the plasma
could allow the amelioration of potential off-target issues. We
determined the residence time (t
Jak2 and Jak3 kinase domains, using a rapid dilution enzymatic as-
say. From Table 6, compounds 6 and 9 displayed much longer
residence times for Jak2 than Jak3. In contrast, compounds 5 and
R
) of four compounds with both
value (pStat5 IC50 = 0.007 lM). Similarly, the inhibition of Stat5
1
5
Table 4
V617F data for 6, 9 and 10
Compd
TEL-Jak2 GI50a
(l
M)
SET-2 GI50a
(
lM)
UKE1 GI50a
(lM)
Table 3
4
6
9
0.006
0.008
0.010
0.078
0.014
0.018
0.050
0.022
0.030
Biochemical and cellular evaluation of (S)-5-fluoro-N -(1-(5-fluoropyridin-2-
yl)ethyl)-N -(5-methyl-1H-pyrazol-3-yl)pyrimidine-2,4-diamine
derivatives 13–16
2
6
N -substituted
10
a
R2
For experimental conditions see Ref. 4.
H
N
F
N
N
N
N
NH
N
Table 5
PK parameters for compounds 6 and 9
HN
Compd Species %Fd AUC
M ꢂ h) min/kg) (L/kg) (mL/min/mg) min/10 cells)
CL (mL/ Vdss Hu Mics
Hu Heps (mL/
6
(l
6
Rata
Mouse 64 14.7
74 27.5
11
17
13
12
18
14
1.6
1.4
2.4
3.1
2.9
2.5
<4
<2
b
c
F
Dog
Rat
56 8.5
51 17.7
a
9
<4
5.2
b
Mouse 64 13.3
R2
Jak2 IC50a
M)
Jak3 IC50a
M)
Tel-Jak3 GI50b
M)
c
e
Dog
n.d n.d
Compd
N
N
(l
(l
(l
a
Han Wistar rat male; 10 mg/kg po (0.1% HPMC); 3 mg/kg iv (DMA/PEG/sal-
ine = 40/40/20).
O
b
7
0.006
0.004
0.262
0.030
0.85
0.48
CD-1 mouse female; 10 mg/kg po (0.1% HPMC); 3 mg/kg iv (DMA/PEG/sal-
ine = 40/40/20).
c
HN
Male Beagle; 5 mg/kg po (0.1% HPMC); 2.5 mg/kg iv (HP-b-CD).
Bioavailability was calculated according to the equation %F = (AUC po ꢂ Dose IV
1
1
3
4
d
/
N
AUC IV ꢂ Dose po) ꢂ 100.
O
e
Denotes not tested.
S
O
0.003
0.020
0.05
N
N
Table 6
O
Residence time (t
Jak3 kinases
R
) for compounds 5, 6, 7 and 9 in the ATP-binding site of Jak2 and
1
1
5
6
O
S
0.003
0.017
0.004
0.015
0.27
0.38
N
a
Jak3 t a
(h)
R
Compd
Jak2 IC50
(l
M)
Jak3 IC50
(
lM)
Jak2 t
R
(h)
N
5
6
7
9
0.003
<0.003
0.006
0.229
0.020
0.262
0.008
2.68
6.65
3.20
0.98
0.78
1.30
3.15
N
a
See Ref. 5 for experimental. Testing was performed at 5 mM concentration of
ATP.
0.003
11.07
b
a
See Ref. 4 for experimental.
See Ref. 15 for experimental procedure.

H. Guan et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3105–3110
3109
Compound9
3000 1000
300 100
30
10
3
1
0
nM
TEL JAK2
pStat5
GI50 (μM)
0.008
IC50 (μM)
T Stat5
pStat5
0.007
TEL JAK3
GI50 (μM)
IC50 (μM)
T Stat5
0.30
0.37
Figure 5. Correlation of anti-proliferative effect with inhibition of Stat5 phosphorylation in TEL-Jak2 and TEL-Jak3 cells, for compound 9. T = Total, p = phosphorylated.
1
1
20
00
600
500
400
300
200
100
0
mouse for both 6 (fu = 0.16) and 9 (fu = 0.06) indicated that the
available quantities of both 6 and 9 were higher than the observed
GI50’s. Monitoring the inhibition of pStat5 at later time points (8 h)
for compound 9 revealed significant inhibition at 10 mg/kg (ꢀ70%
pStat5 inhibition with total plasma concentration of 123 ng/ml) as
well as at 5 mg/kg (ꢀ50% pStat5 inhibition with total plasma con-
centration of 55 ng/ml). Given the PK properties of both 6 and 9 in
mice, in combination with the long residence time estimated by
in vitro experiments, we expected sustained and significant inhibi-
tion of phosphorylation of the downstream Stat5. At this point we
were reluctant to attribute the observed pharmacodynamic effect
to the long residence time particularly when there seemed to be
a rough correlation between the observed plasma concentration
and observed pStat5 inhibition (Fig. 6). We would have expected
that inhibition of Stat5 phosphorylation would not fade with
declining plasma concentration, if there were slow-off kinetics ob-
served for these inhibitors when occupying the Jak2 kinase. How-
ever, the good pharmacokinetic properties of 6 and 9 may possibly
mask the long residence occupancy of the target. Finally, tolerabil-
ity studies of both 6 and 9 in rodents revealed a narrow therapeutic
index, possibly due to the moderate overall kinase panel selectiv-
ity. As a consequence of the tolerability data, progression of these
inhibitors through our cascade was halted.
8
6
4
2
0
0
0
0
0
Fig. 6. Pharmacodynamic effect of compounds 6 and 9 on pStat5 in the Ba/F3 TEL-
JAK2 mouse model17 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 estimations. The yellow bars corre-
spond to measurement of percentage (%) pStat5 inhibition at 2 h post-dose. The
green bars correspond to measurement of percentage (%) pStat5 inhibition at 6 h
post dose.
18
2
In conclusion, we have described a novel series of N -(1H-pyra-
zol-3-yl)pyrimidines that display activity against Jak2. Compounds
6
and 9 from this series were extensively investigated and showed
phosphorylation (IC50 ꢀ0.37
lM) in the TEL-Jak3 cells indicated a
activity in cells harboring the V617F mutation as well as displaying
downstream inhibition of Stat5 phosphorylation in TEL-Jak2 cells.
These compounds demonstrated selectivity against JAK3 at the cel-
lular level in the nanomolar range, and they were found to have
slow-off kinetics for Jak2 compared to Jak3. Testing in vivo in a
murine Ba/F3 TEL-Jak2 model did not differentiate between the
thermodynamic and kinetic activities of these compounds.
strong correlation with the anti-proliferative effect (Fig. 4). In the
more clinically relevant cells, SET-2 and UKE-1, compounds 6, 9
and 10 inhibited effectively the growth of cells bearing the
V617F mutation (Table 4).16
To evaluate the pharmacokinetic properties of this series, com-
pound 6 was dosed both orally and intravenously in rodents and
dogs and exhibited high bioavailability (%F) and oral exposure, as
determined by the area under the curve (AUC). The observed clear-
ance (CL) was low in both rat and mouse, while it was rather mod-
erate in dog. Likewise, the volume of distribution in all three
species was on the moderate to low side. A similar profile was ob-
served for compound 9, as depicted in Table 5. In addition, com-
pound 6 appeared to be metabolically stable in both human
microsomes and hepatocytes, while compound 9 was marginally
less stable in human hepatocytes.
Acknowledgments
The authors would like to thank Jeffrey Varnes and Melissa Vas-
binder for useful recommendations during the preparation of this
manuscript. Likewise, Jing Dai for PK sample analysis and Nancy
Su for ATP competition experiments and Isabelle Green and Han-
nah Pollard for X-ray crystallography.
After administration of compounds 6 and 9 in the allograft Ba/
F3 TEL-Jak2 murine model we observed significant inhibition of
pStat5 (>80%) 6 h post 10 mg/kg dose (Fig. 5, denoted by the green
color bars). Inhibition of Stat5 phosphorylation was still evident
after 6 h post a single oral 5 mg/kg dose. Looking at the total plas-
ma concentrations (denoted as squares in Fig. 5) of both com-
pounds at the two different doses, we concluded that there was a
substantial quantity still present in plasma even after 6 h. In addi-
tion, correcting the plasma concentration with the free fraction in
References and notes
1
2
.
.
Quintas-Cardama, A.; Vertovsek, S. Expert Opin. Invest. Drugs 2011, 20, 961.
Ioannidis, S.; Lamb, M. L.; Wang, T.; Almeida, L.; Block, M. H.; Davies, A. M.;
Peng, B.; Su, M.; Zhang, H.-J.; Hoffmann, E.; Rivard, C.; Green, I.; Howard, T.;
Pollard, H.; Read, J.; Alimzhanov, M.; Bebernitz, G.; Bell, K.; Ye, M.; Huszar, D.;
Zinda, M. J. Med. Chem. 2011, 54, 262.
3.
Hedvat, M.; Huszar, D.; Herrmann, A.; Gozgit, J. M.; Schroeder, A.; Sheehy, A.;
Buettner, R.; Proia, D.; Kowolik, C. M.; Xin, H.; Armstrong, B.; Bebernitz, G.;
Weng, S.; Wang, L.; Ye, M.; McEachern, K.; Chen, H.; Morosini, D.; Bell, K.;

3
110
H. Guan et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3105–3110
Alimzhanov, M.; Ioannidis, S.; McCoon, P.; Cao, Z. A.; Yu, H.; Jove, R.; Zinda, M.
Cancer Cell 2009, 16, 487.
structure factors for the Human JAK2 complex24 with compound 6 have been
deposited in the Protein Data Bank (3zmm) together with structure factors and
detailed experimental conditions.
4
.
.
.
.
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.
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.
Ioannidis, S.; Lamb, M. L.; Almeida, L.; Guan, H.; Peng, B.; Bebernitz, G.; Bell, K.;
Alimzhanov, M.; Zinda, M. Bioorg. Med. Chem. Lett. 2010, 20, 1669.
Wang, T.; Ioannidis, S.; Almeida, L.; Block, M. H.; Davies, A. M.; Lamb, M. L.;
Scott, D. A.; Su, M.; Zhang, H.-J.; Alimzhanov, M.; Bebernitz, G.; Bell, K.; Zinda,
M. Bioorg. Med. Chem. Lett. 2011, 21, 2958.
13. Craczyk, P. P. J. Med. Chem. 2007, 50, 5773.
14. Lu, H.; Tonge, P. J. Curr. Opin. Chem. Biol. 2010, 14, 467.
15. Biochemical evaluation of drug–target residence time: The drug–target
residence time of JAK2 inhibitors was evaluated using a mobility shift based
biochemical assay. Specifically, JAK2 or JAK3 kinase (aa 808–1132, Millipore,
5
6
7
catalog
#
14–640), peptide substrate (FL-Ahx-IPTSPITTTY FFFKKK-COOH,
particular inhibitor of interest were
Primm Biotech, MA, USA), and
a
incubated at room temperature for 30 min. Kinase reactions were initiated
by diluting small aliquots of the mixture into solution of ATP in HEPES buffer
(75 mM, pH 7.3), and reaction progress was monitored by a Caliper LC3000
system. Initial and steady-state reaction velocities of individual were then
derived from each reaction progression curve, and the data were fitted to
8
.
.
Rat microsomal CLint = 82 lL/min/mg. Compound 1 exhibited reasonable total
plasma concentrations in mice 2 and 6 h post-dose (average of 463 and 185 ng/
ml, respectively). The percentage (%) free of compound 1 in humans was
measured to be around 3%.
9
The two regioisomers could be distinguished by the use of 1H NMR as shown
equation, p ¼
v
; t þ
v
0
ꢁ
v
s
ð1 ꢁ e Þ, where p is the reaction product amount, v
s
ꢁkt
k
below.
0
is the steady-state reaction velocity, t is the reaction time, v is the initial
reaction velocity, and k is the rate of transition from
residence time (t ) is calculated by the following equation t
R3
0 s
v to v . The drug–target
0
:693
k
R2
R2
R3
R
R
¼
.
Cl
N
N
Cl
N
1
1
6. Quentmeier, H.; MacLeod, R. A.; Zaborski, M.; Drexler, H. G. Leukemia 2006, 20, 471.
7. 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.
N
N
N
HN
H
Y
HN
H
Y
1
8. The pharmacokinetic samples were run on an Applied Biosystems API4000
Ò
QTrap using Analyst software version 1.4.2. The sample analyses were carried
F
F
out in Sciex#5 LC–MS, using an ACE 3C18 3
ambient temperature and the following eluents: Mobile Phase A: water + 0.1%
formic acid + 10 mM ammonium formate B: MeCN + 0.1% formic acid.
lm 20 ꢂ 2.1 mm column at
N-H δ~5.81-5.85
H δ ~5.3-5.5
N-H δ ~6.11-6.19
H δ ~5.5-5.7
19. DeLano, W. L. The PyMOL Molecular Graphics System; DeLano Scientific: San
Carlos, CA, USA, 2002. http://www.pymol.org.
2
2
2
0. CCP4 Acta Crystallogr., Sect. D 1994, 50, 760.
1. Coot; Emsley, P.; Cowtan, K. Acta Crystallogr., Sect. D 2004, 60, 2126.
2. AFITT: OpenEye Scientific Software, Inc., Santa Fe, NM, USA, http://
www.eyesopen.com, 2010.
1
1
0. For a recent review see: Surry, D. S.; Buchwald, S. L. Chem. Sci. 2011, 21, 2958.
1. For synthetic experimental see: Guan, Huiping; Hayter, B. R.; Huang, S.;
Ioannidis, S.; Johannes, J.; Lamb, M.; Peng, B.; Yang, B. WO 2008132502, 2008.
2. Protein expression and purification, crystallization and cryoprotection
23. Refmac5: Murshudov, G. N.; Vagin, A. A.; Dodson, E. J. Acta Crystallogr., Sect. D
1
2
1997, 53, 240.
procedures have been described previously. Picture of the crystal structure
1
9
24. Crystallographic statistics for the JAK2 bound to 6 are as follows. Numbers in
parenthesis characterize the higher resolution shell. Space group C2, unit cell a,
b, c dimensions 44.5, 126.7, 135.6 Å, b = 97.2° resolution 17.57 2.50 Å (2.64–
of 6 with Jak2 kinase has been generated by PYMOL. Diffraction data were
collected ‘in-house’ using a Rigaku MM007 X-ray generator equipped with a
MAR345 mage plate X-ray detector, using a CuK
a radiation at a wavelength of
20.5 Å), 24425 unique reflections with an overall redundancy of 6.6 (5.9), 95.0%
1
.5418 Å focused using Rigaku VariMax HF mirrors. Data were processed using
2
0
(83.2%) completeness with Rmerge of 6.5% (30.3%) and mean I/ (I) of 21.0 (5.0).
The final model containing 4455 protein, 140 solvent, and 60 compound atoms
has an R-factor of 19.3% (Rfree using 5% of the data 24.6%). Mean temperature
factors for the protein and the ligand are 43.0 and 50.1 Å , respectively.
r
MOSFLM and SCALA and reduced using CCP4 software. The structures were
solved by molecular replacement using coordinates of the JAK2 kinase domain
(
PDB entry 2XA4) as a trial model using CCP4 software. Protein and inhibitor
2
21 22
were modeled into the electron density using, COOT
respectively. The model was refined using Refmac. Atomic coordinates and
and AFITT ,
2
3