Structure-based design of selective janus kinase 2 imidazo[4,5- d ]pyrrolo[2,3- b ]pyridine inhibitors

Article abstract of DOI:10.1021/acsmedchemlett.5b00225

Early hit to lead work on a pyrrolopyridine chemotype provided access to compounds with biochemical and cellular potency against Janus kinase 2 (JAK2). Structure-based drug design along the extended hinge region of JAK2 led to the identification of an imp

Full text of DOI:10.1021/acsmedchemlett.5b00225

Letter
pubs.acs.org/acsmedchemlett
Structure-Based Design of Selective Janus Kinase 2
Imidazo[4,5‑d]pyrrolo[2,3‑b]pyridine Inhibitors
Amy C. Hart,* Gretchen M. Schroeder, Honghe Wan, James Grebinski,† Jennifer Inghrim,‡
James Kempson, Junqing Guo, William J. Pitts, John S. Tokarski, John S. Sack, Javed A. Khan,
Jonathan Lippy, Matthew V. Lorenzi,§ Dan You, Theresa McDevitt,§ Ragini Vuppugalla, Yueping Zhang,
Louis J. Lombardo, George L. Trainor,∥ and Ashok V. Purandare
Bristol-Myers Squibb Research & Development, Princeton, New Jersey 08543, United States
S
* Supporting Information
ABSTRACT: Early hit to lead work on a pyrrolopyridine
chemotype provided access to compounds with biochemical
and cellular potency against Janus kinase 2 (JAK2). Structure-
based drug design along the extended hinge region of JAK2 led
to the identification of an important H-bond interaction with
the side chain of Tyr 931, which improved JAK family
selectivity. The 4,5-dimethyl thiazole analogue 18 demon-
strated high levels of JAK family selectivity and was identified
as a promising lead for the program.
KEYWORDS: Kinase, Janus kinase 2 (JAK2), thiazole, myeloproliferative neoplasms, structure-based drug design (SBDD)
Screening of the Bristol-Myers Squibb compound collection
using a homogeneous time-resolved fluorescence (HTRF)
assay identified pyrrolopyridine 1 as a hit with good JAK2
potency but modest kinome selectivity (Figure 1). Replacement
of the tetrazole with a secondary amide (2) resulted in
ignal transduction cascades mediated by kinases have long
Sbeen viewed as promising targets for therapeutic
intervention. Janus kinase 2 (JAK2) is a nonreceptor
cytoplasmic tyrosine kinase involved in signal transducer and
activator of transcription (STAT) signaling. Imbalances in the
JAK-STAT pathway have recently been associated in a variety
of diseases, including inflammatory disease, various cancers, and
myeloproliferative neoplasms (MPNs).1 MPNs are a class of
myeloid malignancies that are represented by the uncontrolled
growth of a multipotent hematopoietic progenitor stem cell.
Identification of a mutation in the JH2 pseudokinase domain
(JAK2 V617F) in MPN patients was found to render the JAK2
kinase constitutively active.2−5 This mutation was observed in
high frequency in patients with MPNs (>95% in polycythemia
vera, >50% in essential thrombocytopenia, >50% in primary
myelofibrosis).
Success with kinase targeted therapies supported the
feasibility of targeting JAK2 for therapeutic intervention in
MPNs. Recent approval of the JAK1/JAK2 dual inhibitor
ruxolitinib6 for the treatment of primary myelofibrosis
reinforced the tractability of targeting the JAK family
specifically. A key challenge associated with design of a
selective JAK2 inhibitor is the high degree of homology in
the kinase domains of the JAK family.
Herein we report on initial hit to lead (HTL) efforts leading
to potent JAK2 inhibitors. Further structure-based drug design
(SBDD) efforts to explore and capitalize on several single
amino acid differences in the extended hinge region of the JAK
kinases led to the identification of 4,5-disubstituted thiazoles
(e.g., 18) as highly selective JAK2 inhibitors.
Figure 1. Hit to lead SAR.
Received: June 9, 2015
Accepted: July 10, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acsmedchemlett.5b00225
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
a b
,
significantly reduced potency, likely due to loss of the enforced
s-cis amide conformation. Tertiary amides that were unable to
position a cyclopropyl group in the optimal vector (4, 5), lost
potency and JAK3 selectivity as compared to 1. Dicyclopropyl
amide 3 regained the potency and selectivity of the initial hit,
reinforcing the importance of the vector for the cyclopropyl
group. Compound 3 demonstrated modest antiproliferative
activity in a JAK2 pathway-dependent cell line (SET-2).
X-ray crystallographic analysis of compound 37 bound to the
JAK2 kinase domain revealed that the tertiary amide is tucked
under the P-loop. The dicyclopropyl amide is well suited to
traverse the space between the floor of the binding pocket and
the P-loop, likely contributing to enhanced selectivity among
the JAK family. The diol side-chain on the azaindole nitrogen
extends into the ribose pocket. Hydrogen bond contacts with
the hinge occur from the imidazole nitrogen and 4-position
−NHMe, thereby projecting the methyl group toward the
extended hinge region of JAK2 (Figure 2).
Table 1. Azaindole SAR
a
Assay protocol details are provided in the Supporting Information.
Assay results are the average of at least two replicates. Not tested in
this assay. Assay results are N of 1.
b
c
d
a
Scheme 1. Synthesis of Early Lead Compounds
Figure 2. Crystal structure of compound 3 bound to the kinase
catalytic domain of JAK2. The carbons of 3 are colored in pink and the
carbons for JAK2 are colored in green. Oxygens are colored red,
nitrogens blue, and sulfurs yellow. Hydrogen bonds are indicated with
dashed lines.
a
Reagents and conditions: (a) NIS, acetonitrile, RT, 75%; (b) ethyl
pyruvate, Pd2(dba)3, Cy2NMe, DMA, 60 °C, 70%; (c) R-OTs or R-I,
Cs2CO3, DMF, 60 °C, 65−75%; (d) NaOH, EtOH, 60 °C, 80−90%;
(e) R1R2NH, EDCI, HOBt, i-Pr2NEt, MeCN, 50 °C, 70−75%; (f)
TFA-CH2Cl2 (10:1), RT, 90%.
Compound 3 demonstrated poor intrinsic permeability
(PAMPA) and high efflux (Caco-2). Sequential removal of
the hydroxyl groups maintained or improved JAK2 activity,
selectivity, and cellular activity (6, 7) (Table 1). The ethyl
analogue (8) was optimal with respect to permeability and
efflux and showed improved JAK3 selectivity (44×) and modest
selectivity against the other JAK family members JAK1 and
Tyk2 (15× and 16×, respectively).
Scheme 1 outlines the synthesis of compounds similar to 8
starting from amino imidazo[4,5-c]pyridine 9.8 The tricyclic
core 10 was assembled in a two-step protocol wherein a Heck
coupling-cyclization was preceded by iodination of 2-amino-
pyridine 9. Alkylation of the azaindole was followed by ester
hydrolysis and amide coupling with the appropriate amine to
yield the final compounds.
hydrogen bond acceptor with an appropriate chain length
appended at the C4 position may interact favorably with Tyr
931 in the hinge region. To probe this potential interaction,
alkyl and acyl linkers were attached at the C4 position of the
pyrrolopyridine core (Table 2).
Consistent with our prediction, the crystal structure of 12
bound to the JAK2 kinase domain showed a hydrogen bond
interaction of the methoxypropyl group with the phenolic
hydrogen of Tyr 931 (see Supporting Information). Molecular
modeling suggested that a trajectory from a sp2 center would
more effectively traverse the opening between the P-loop and
extended hinge region toward solvent. Appending the 2-
methoxyethyl side chain to the scaffold through a sp2
hybridized linker (13, 14) maintained planarity with the
azaindole core and further improved potency against JAK2
while preserving JAK family selectivity. Despite the improved
selectivity, these compounds (12, 13, 14) were less potent than
the methyl analogue (8). An amide isostere, thiazole (15),
With a goal of improving the broader JAK family selectivity
of 8, we began our SAR exploration around the C4 position
while holding positions N6 and C7 as ethyl and dicyclopro-
pylamide, respectively. Inspection of the X-ray cocrystal
structure of 3 with the JAK2 kinase domain indicated that a
B
DOI: 10.1021/acsmedchemlett.5b00225
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
ab
,
a
Table 2. C4 SAR
Scheme 2. Synthesis of C4 Analogues
a
Assay protocol details are provided in the Supporting Information.
Assay results are the average of at least two replicates. Assay results
b
c
are N of 1.
improved potency and cellular activity, while retaining
comparable levels of selectivity to the methoxyalkyl analogues.
In addition to Tyr 931, several amino acid differences among
JAK family members along the extended hinge region drove us
to postulate that substitution at the 4- or 5-position on the
thiazole may further enhance JAK family selectivity. Directing
hydrophobic groups toward Gln 853 may induce a repulsive
interaction associated with JAK1 (Arg 868), JAK3 (Ser 826),
and Tyk2 (Arg 901). Other nonconserved amino acids within
the JAK family, namely, JAK2/Met 865 (near Gln 853), and
along the extended hinge region (JAK2/Ser 936 and JAK2/Asp
939) may play a subtle role in influencing selectivity.
Selectivity of thiazole 15 was marginally enhanced through
incorporation of methyl groups at the 4- or 5-position of the
thiazole (16, 17) (Table 2). Incorporation of a methyl group at
both the 4- and 5- positions of the thiazole (18) was found to
significantly enhance JAK family selectivity while maintaining
cellular potency in the JAK2-dependent cell line. Compound 18
displayed superior JAK family selectivity as compared to
ruxolitinib9 and other JAK inhibitors in the clinic.10
a
Reagents and conditions: (a) TFA, CH2Cl2 (10:1), RT, 95%; (b) 3-
methoxypropanal, NaCNBH3, MeOH 32%; (c) 2-methoxyethyl
chloroformate, pyridine, CH2Cl2, 20%; (d) 2-methoxyethanamine,
DMF, phosgene, pyridine, CH2Cl2, 16%; (e) ethyl isocyanate,
pyridine, 23%; (f) 2-bromothiazole, Pd2(dba)3, BINAP, NaOtBu,
toluene, 110 °C, 11%; (g) benzoylisothiocyanate, acetone; then 1 N
NaOH, ethanol, 60 °C, 72%; (h) α-bromoketone, ethanol, 65 °C.
differences in the JAK family members, which may account for
the high selectivity found for compound 18 (Figure 3). The
thiazole ring nitrogen was observed to engage Tyr 931 in a
hydrogen bonding interaction similar to that observed with
methoxypropyl analogue 12. This interaction is not feasible in
JAK1 (Phe). The thiazole sulfur is positioned to have a positive
nonbonding interaction with the core pyridyl nitrogen, likely
reinforcing the desired conformation of the inhibitor in the
enzyme.13 Consistent with our prediction, the hydrophobic
methyl groups of 18 are directed toward Gln 853, presumably
contributing to the observed JAK family selectivity.
Compound 18 was the first compound from the pyrrolopyr-
idine series that demonstrated target potency (biochemical and
cellular) and significant JAK family selectivity in our enzymatic
assays (Table 3). The kinome selectivity of 18 was assessed in a
single point format at 1 μM against a panel of 412 kinases at
Ambit Biosciences (now DiscoverRx) and showed good
selectivity across the Ambit profile, based on less than 10%
control for 6 out of 412 kinases.14 Consistent with the
biochemical assays, no other JAK family members were hits in
Ambit profiling. Outside of the JAK family members, 18 also
demonstrated excellent selectivity against a panel of 22 kinases
tested in enzymatic assays. Although Src kinase was identified
The C4 analogues were prepared from the fully protected
aminopyridine11 19 (Scheme 2). Access to scaffold 19 was
available via a similar synthetic sequence to that described for
the C4-NHMe analogues. Deprotection provided the free
amine 20, which could undergo reductive amination (12),
chloroformate or isocyanate coupling (13, 14), or Buchwald−
Hartwig coupling (15).
The substituted thiazoles were also accessed from amino-
pyridine 20. Alkyl thiazoles (e.g., 16, 17, 18) were formed by
reaction of thiourea 21 with an appropriately substituted α-
bromoketone.
The crystal structure of 1812 bound to the JAK2 kinase
domain highlighted some of the extended hinge amino acid
C
DOI: 10.1021/acsmedchemlett.5b00225
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
enhanced JAK family selectivity. Incorporation of thiazoles
along the extended hinge region led to the identification of a
lead series, and compound 18 represented a significant advance
in our efforts to identify potent and selective JAK2 inhibitors
with promising pharmacokinetic properties. Additional SAR
around this lead series will be the subject of an additional
disclosure.15
ASSOCIATED CONTENT
■
S
* Supporting Information
Full experimental details for key compounds and biological
protocols. The Supporting Information is available free of
charge on the ACS Publications website at DOI: 10.1021/
acsmedchemlett.5b00225.
AUTHOR INFORMATION
■
Corresponding Author
*Tel: +1-609-252-3472. E-mail: amy.hart@bms.com.
Present Addresses
†Xenex, 121 Interpark, Suite 104, San Antonio, Texas 78216,
United States.
Figure 3. Crystal structure of 18 bound to the kinase catalytic domain
of JAK2. The carbons of 18 are colored in pink and the carbons for
JAK2 are colored in green except for the residues near the thiazole ring
that differ in the JAK family (carbons are colored cyan). Oxygens are
colored red and nitrogens blue. Hydrogen bonds are indicated with
dashed lines.
‡Immunomedics, 300 The American Road, Morris Plains, New
Jersey 07950, United States.
§Janssen Pharmaceutical Companies of Johnson and Johnson,
1400 McKean Road, Spring House, Pennsylvania 19002,
United States.
∥BioMotiv, 20600 Chagrin Boulevard, Suite 210, Cleveland,
Ohio 44122, United States.
Table 3. Compound 18 Selectivity
biochemical
Ambit
Author Contributions
All authors have given approval to the final version of the
manuscript.
kinase
IC50, μM
x-fold
kinase
JAK2
% control
JAK1
JAK3
Tyk2
LCK
DAPK1
Fyn
0.270
0.160
0.110
1.80
106
66
1.8
4.4
6.5
6.6
6.7
9.4
Fyn
Notes
46
Src
The authors declare no competing financial interest.
700
550
4
DAPK3
1.40
FLT3 N841I
ACK
ACKNOWLEDGMENTS
0.010
>2
■
Src
800
The authors would like to acknowledge our Department of
Discovery Synthesis and BBRC for scale-up synthetic work of
intermediates and Bethanne Warrack for analytical work.
as a hit in the Ambit profile, a biochemical assay indicated Src
was not significantly inhibited by 18.
Compound 18 was rapidly metabolized across species but
otherwise had an acceptable ADMET profile in in vitro assays
(Table 4). The good permeability associated with earlier
ABBREVIATIONS
■
JAK2, Janus kinase 2; JAK1, Janus kinase 1; JAK3, Janus kinase
3; STAT, signal transducer and activator of transcription; MPN,
myeloproliferative neoplasm; JH2, JAK homology domain 2;
HTL, hit to lead; SBDD, structure-based drug design; HTRF,
homogeneous time-resolved fluorescence; PAMPA, parallel
artificial membrane permeability assay; Tyk2, tyrosine kinase 2;
HLM, human liver microsome; CYP, cytochrome P450; BTK,
Burton’s tyrosine kinase; CDK2, cyclin-dependent kinase 2;
CK2, casein kinase; GSK, glycogen synthase kinase; IRAK4,
interleukin-1 receptor-associated kinase 4; AAK1, AP2-
associated protein kinase 1; Abl, Abelson leukemia oncogene
cellular homologue; DAPK1, death-associated protein kinase 1;
Fyn, a Src family tyrosine protein kinase; IKK-ε, IκB kinase-ε;
JNK1, Jun N-terminal kinase 1; LCK, lymphocyte-specific
protein tyrosine kinase; NEK2, NIMA-related kinase; PKC,
protein kinase C; Src, Rous sarcoma oncogene cellular
homologue; IGF1R, insulin-like growth factor receptor 1;
FLT3, FMS-like tyrosine kinase 3; ACK, activated CDC42
kinase; ADMET, absorption, distribution, metabolism, excre-
tion, and toxicity
Table 4. Compound 18 ADMET Profile
metabolic stability (T1/2 min)
PAMPA (pH 7.4)
HLM CYP IC50
16 (H), 14 (R), 36 (D)
532 nm/sec
all >5 μM; except CYP3A4 (3 μM)
>80 μM
hERG (flux)
PXR EC50
>16 μM
HEPG2 IC50
>50 μM
plasma protein binding (% bound)
99.1 (H), 99.0 (M), 99.0 (R)
analogues (8) was maintained in 18. Initial assessments of in
vitro safety (hERG flux, PXR, and HEPG2) for 18 were clean.
In an in vivo setting, compound 18 exhibited excellent oral
exposure (Rat, 5 mg/kg PO; Cmax 29 μM, AUC0−24h 131.4 μM·
h).
Early hit to lead efforts provided access to biochemical and
cellular active compound 8. Structure-based design efforts
around 8 identified extended hinge region modifications which
D
DOI: 10.1021/acsmedchemlett.5b00225
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
S.; Lippy, J.; Tokarski, J. S.; Blatt, Y.; Sack, J. S.; Khan, J. A.; Stefanski,
K.; Sleckza, B.; Arunachalam, P.; Pregaglathan, B.; Narayanan, S.;
Nanjundaswamy, K. C.; Kuppusamy, P.; Mathur, A.; Sun, J.-H.; Wong,
M. K.; Wu, D.-R.; Li, P.; Gupta, A.; Purandare, A. V. Discovery of a
highly selective JAK2 inhibitor, BMS-911543 for the treatment of
myeloproliferative neoplasms. ACS Med. Chem. Lett. 2015.
REFERENCES
■
(1) Morgan, K. J.; Gilliland, D. G. A role for JAK2 mutations in
myeloproliferative diseases. Annu. Rev. Med. 2008, 59, 213 and
references therein. .
(2) Levine, R. L.; Wadleigh, M.; Cools, J.; Ebert, B. L.; Wernig, G.;
Huntly, B. J.; Boggon, T. J.; Wlodarska, I.; Clark, J. J.; Moore, S.;
Adelsperger, J.; Koo, S.; Lee, J. C.; Gabriel, S.; Mercher, T.; D’Andrea,
A.; Frohling, S.; Dohner, K.; Marynen, P.; Vandenberghe, P.; Mesa, R.
̈
̈
A.; Tefferi, A.; Griffin, J. D.; Eck, M. J.; Sellers, W. R.; Meyerson, M.;
Golub, T. R.; Lee, S. J.; Gilliland, D. G. Activating mutation in the
tyrosine kinase JAK2 in polycythemia vera, essential thrombocythemia,
and myeloid metaplasia with myelofibrosis. Cancer Cell 2005, 7, 387.
(3) Baxter, E. J.; Scott, L. M.; Campbell, P. J.; East, C.; Fourouclas,
N.; Swanton, S.; Vassiliou, G. S.; Bench, A. J.; Boyd, E. M.; Curtin, N.;
Scott, M. A.; Erber, W. N.; Green, A. R. Acquired mutation of the
tyrosine kinase JAK2 in human myeloproliferative disorders. Lancet
2005, 365, 1054.
(4) Kralovics, R.; Passamonti, F.; Buser, A. S.; Teo, S.-S.; Tiedt, R.;
Passweg, J. R.; Tichelli, A.; Cazzola, M.; Skoda, R. C. A gain-of-
function mutation of JAK2 in myeloproliferative disorders. N. Engl. J.
Med. 2005, 352, 1779.
(5) James, C.; Ugo, V.; Le Coued
́
ic, J.-P.; Staerk, J.; Delhommeau, F.;
Lacout, C.; Garcon, L.; Raslova, H.; Berger, R.; Bennaceur-Griscelli,
̧
A.; Villeval, J. L.; Constantinescu, S. N.; Casadevall, N.; Vainchenker,
W. A unique clonal JAK2 mutation leading to constitutive signalling
causes polycythaemia vera. Nature 2005, 434, 1144.
(6) Cervantes, F.; Vannucchi, A.; Kiladjian, J.; Al-Ali, H.; Sirulnik, A.;
Stalbovskaya, V.; McQuitty, M.; Hunter, D.; Levy, R.; Passamonti, F.;
Barbui, T.; Barosi, G.; Harrison, C.; Knoops, L.; Gisslinger, H. Three-
year efficacy, safety, and survival findings from COMFORT-II, a phase
3 study comparing ruxolitinib with best available therapy for
myelofibrosis. Blood 2013, 122, 4047.
(7) PDB ID 5CF6.
(8) Kempson, J.; Junqing, Guo; Das, J.; Moquin, R.; Spergel, S.;
Watterson, S.; Langevine, C.; Dyckman, A.; Pattoli, M.; Burke, J.;
Yang, X.; Gillooly, K.; McIntyre, K.; Chen, L.; Dodd, J.; McKinnon,
M.; Barrish, J.; Pitts, W. Synthesis, initial SAR and biological evaluation
of 1,6-dihydroimidazo[4,5-d]pyrrolo[2,3-b]pyridin-4-amine derived
inhibitos of IκB kinase. Bioorg. Med. Chem. Lett. 2009, 19, 2646.
́
(9) Quintas-Cardama, A.; Vaddi, K.; Liu, P.; Manshouri, T.; Li, J.;
Scherle, P. A.; Caulder, E.; Wen, X.; Li, Y.; Waeltz, P.; Rupar, M.;
Burn, T.; Lo, Y.; Kelley, J.; Covington, M.; Shepard, S.; Rodgers, J. D.;
Haley, P.; Kantarjian, H.; Fridman, J. S.; Verstovsek, S. Preclinical
characterization of the selective JAK1/2 inhibitor INCB018424:
therapeutic implications for the treatment of myeloproliferative
neoplasms. Blood 2010, 115, 3109.
(10) Tam, C. S.; Verstovsek, S. Investigational Janus kinase
inhibitors. Expert Opin. Invest. Drugs 2013, 22, 687.
(11) Purandare, A.; Schroeder, G.; Wan, H.; Hart, A.; Grebinski, J.;
Inghrim, J. Preparation of dicyclopropyldihydroimidazopyrrolopyridi-
necarboxamide derivatives for use as JAK2 inhibitors. PCT Int. Appl.
WO 2011028864 A1 20110310, 2011. Purandare, A.; Grebinski, J.;
Hart, A.; Inghrim, J.; Schroeder, G.; Wan, H. JAK2 inhibitors ans their
use for the treatment of myeloproliferative diseases and cancer
U.S.Patent 8202881, 2012.
(12) PDB ID 5CF5.
(13) Beno, B. R.; Yeung, K.-S.; Bartberger, M. D.; Pennington, L. D.;
Meanwell, N. A. A survey of the role of noncovalent sulfur interactions
in drug design. J. Med. Chem. 2015, 58, 4383.
(14) Fabian, M. A.; Biggs, W. H., III; Treiber, D. K.; Atteridge, C. E.;
Azimioara, M. D.; Benedetti, M. G.; Carter, T. A.; Ciceri, P.; Edeen, P.
T.; Floyd, M.; Ford, J. M.; Galvin, M.; Gerlach, J. L.; Grotzfeld, R. M.;
́
Herrgard, S.; Insko, D. E.; Insko, M. A.; Lai, A. G.; Lelias, J.-M.; Mehta,
S. A.; Milanov, Z. V.; Velasco, A. M.; Wodicka, L. M.; Patel, H. K.;
Zarrinkar, P. P.; Lockhart, D. J. A small molecule-kinase interaction
map for clinical kinase inhibitors. Nat. Biotechnol. 2005, 23, 329.
(15) Wan, H.; Schroeder, G. M.; Hart, A. C.; Inghrim, J.; Grebinski,
J.; Lorenzi, M. V.; You, D.; McDevitt, T.; Penhallow, B.; Vuppugalla,
R.; Zhang, Y.; Gu, X.; Iyer, R.; Lombardo, L. J.; Trainor, G. L.; Ruepp,
E
DOI: 10.1021/acsmedchemlett.5b00225
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX