corroborated. The inhibitor potently blocked JAK3-dependent STAT5 phosphorylation, e.g., after stimulation of murine CTLL-2 cells with IL-2 (IC 50 ¼ 70 nM) or
IL-7 stimulation of human PBMCs (IC 50 ¼ 280 nM). At the same time, an at least
35-fold selectivity over the JAK3-independent pathways triggered by IL-6,
GM-CSF, and EPO was observed. In agreement with the previously described
results from Pfizer employing a structurally related aromatic acrylamide (58,
Fig. 20b), a poor translation of the enzymatic and cellular assay data was observed
when moving to human whole blood. Despite remaining an effective suppressor of
JAK3-dependent pathways, the IC 50 values of 64 were significantly higher in the
latter case. This could be a result of the substantial intrinsic reactivity of N-aryl
acrylamides and their susceptibility to inactivation by promiscuous thiol binding.
In animal models, 64 was combined with the CYP inhibitor 1-aminobenzotriazole
(ABT) to prevent a rapid decrease in plasma concentration resulting from modest
stability against hepatic metabolism. In a modified CIA model, 64 was able to reduce
paw swelling in a dose-dependent manner. Nevertheless, even at the highest dose of
300 mg/kg (p.o., BID), compound 64 was still inferior to the positive control
dexamethasone. However, JAK3 selectivity could be substantiated in vivo, since
inhibition of T cell proliferation was observed, while hematopoiesis was not
affected.
In further experiments investigating the connection between PD and PK, it was
demonstrated that an extended effect of irreversible inhibition after clearance of 64
from plasma can only partially be achieved. This observation conforms to the
previously described high resynthesis rate of JAK3.
Kempson et al. from Bristol-Myers Squibb recently disclosed their efforts on the
development of nicotinamide-derived covalent JAK3 inhibitors [65]. In preliminary
studies, they identified compound 65 (Fig. 25) as a potent non-covalent JAK3
inhibitor with a reasonable selectivity profile (JAK3 IC 50 ¼ 4.5 nM and ca. 40-fold
to 180-fold selectivity against the other JAKs). This molecule served as the starting
point for further investigations. Guided by molecular modeling, the meta-position of
the benzylamine substituent was identified as promising attachment point for an
acrylamide warhead. Indeed, this modification leading to compound 66 boosted
JAK3 potency to reach subnanomolar IC 50 values and concomitantly increased
selectivity against the other isoforms (>1,100-fold, >1,300-fold, and >5,000-fold
over JAK1, JAK2, and TYK2, respectively). A successful covalent capture of
Cys909 can be assumed, since 66 demonstrated time-dependent inhibition of
JAK3, while the analogous propanamide did not. However, a final confirmation by
X-ray crystallography or mass spectrometry was not provided for this scaffold class.
Further characterization in cellular assays demonstrated 66 to potently inhibit
IL-2 driven T cell proliferation (IC 50 ¼ 22 nM) and to decrease IFNγ production
after IL-2 stimulation in human whole blood (IC 50 ¼ 490 nM). A favorable JAK3
selectivity was also confirmed in cellular models, since JAK2-dependent (EPO) and
JAK1/TYK2-driven (IFNα) pathways were blocked less efficiently (IC 50 values of
11 μM and 5 μM, respectively).
Good kinome-wide selectivity was proven for 66 in a panel of 350 kinases, where
only 3 kinases (JAK3, FMS, and BMPR2) were strongly affected (<10% of the
Covalent Janus Kinase 3 Inhibitors
249
IL-7 stimulation of human PBMCs (IC 50 ¼ 280 nM). At the same time, an at least
35-fold selectivity over the JAK3-independent pathways triggered by IL-6,
GM-CSF, and EPO was observed. In agreement with the previously described
results from Pfizer employing a structurally related aromatic acrylamide (58,
Fig. 20b), a poor translation of the enzymatic and cellular assay data was observed
when moving to human whole blood. Despite remaining an effective suppressor of
JAK3-dependent pathways, the IC 50 values of 64 were significantly higher in the
latter case. This could be a result of the substantial intrinsic reactivity of N-aryl
acrylamides and their susceptibility to inactivation by promiscuous thiol binding.
In animal models, 64 was combined with the CYP inhibitor 1-aminobenzotriazole
(ABT) to prevent a rapid decrease in plasma concentration resulting from modest
stability against hepatic metabolism. In a modified CIA model, 64 was able to reduce
paw swelling in a dose-dependent manner. Nevertheless, even at the highest dose of
300 mg/kg (p.o., BID), compound 64 was still inferior to the positive control
dexamethasone. However, JAK3 selectivity could be substantiated in vivo, since
inhibition of T cell proliferation was observed, while hematopoiesis was not
affected.
In further experiments investigating the connection between PD and PK, it was
demonstrated that an extended effect of irreversible inhibition after clearance of 64
from plasma can only partially be achieved. This observation conforms to the
previously described high resynthesis rate of JAK3.
Kempson et al. from Bristol-Myers Squibb recently disclosed their efforts on the
development of nicotinamide-derived covalent JAK3 inhibitors [65]. In preliminary
studies, they identified compound 65 (Fig. 25) as a potent non-covalent JAK3
inhibitor with a reasonable selectivity profile (JAK3 IC 50 ¼ 4.5 nM and ca. 40-fold
to 180-fold selectivity against the other JAKs). This molecule served as the starting
point for further investigations. Guided by molecular modeling, the meta-position of
the benzylamine substituent was identified as promising attachment point for an
acrylamide warhead. Indeed, this modification leading to compound 66 boosted
JAK3 potency to reach subnanomolar IC 50 values and concomitantly increased
selectivity against the other isoforms (>1,100-fold, >1,300-fold, and >5,000-fold
over JAK1, JAK2, and TYK2, respectively). A successful covalent capture of
Cys909 can be assumed, since 66 demonstrated time-dependent inhibition of
JAK3, while the analogous propanamide did not. However, a final confirmation by
X-ray crystallography or mass spectrometry was not provided for this scaffold class.
Further characterization in cellular assays demonstrated 66 to potently inhibit
IL-2 driven T cell proliferation (IC 50 ¼ 22 nM) and to decrease IFNγ production
after IL-2 stimulation in human whole blood (IC 50 ¼ 490 nM). A favorable JAK3
selectivity was also confirmed in cellular models, since JAK2-dependent (EPO) and
JAK1/TYK2-driven (IFNα) pathways were blocked less efficiently (IC 50 values of
11 μM and 5 μM, respectively).
Good kinome-wide selectivity was proven for 66 in a panel of 350 kinases, where
only 3 kinases (JAK3, FMS, and BMPR2) were strongly affected (<10% of the
Covalent Janus Kinase 3 Inhibitors
249