Discovery of a JAK3-Selective Inhibitor: Functional Differentiation of

Oct 28, 2016 - (1, 3) The JAK family of nonreceptor tyrosine kinases is composed of four .... The percent inhibition is color coded and ranges from 0%...
0 downloads 0 Views 4MB Size
Download PDF
Articles pubs.acs.org/acschemicalbiology

Discovery of a JAK3-Selective Inhibitor: Functional Differentiation of JAK3-Selective Inhibition over pan-JAK or JAK1-Selective Inhibition Jean-Baptiste Telliez,*,† Martin E. Dowty,§ Lu Wang,† Jason Jussif,† Tsung Lin,† Li Li,† Erick Moy,† Paul Balbo,† Wei Li,† Yajuan Zhao,† Kimberly Crouse,† Caitlyn Dickinson,† Peter Symanowicz,† Martin Hegen,† Mary Ellen Banker,∥ Fabien Vincent,∥ Ray Unwalla,‡ Sidney Liang,⊥ Adam M. Gilbert,⊥ Matthew F. Brown,⊥ Matthew Hayward,⊥ Justin Montgomery,⊥ Xin Yang,§ Jonathan Bauman,§ John I. Trujillo,⊥ Agustin Casimiro-Garcia,‡ Felix F. Vajdos,⊥ Louis Leung,§ Kieran F. Geoghegan,⊥ Amira Quazi,† Dejun Xuan,† Lyn Jones,‡ Erik Hett,‡ Katherine Wright,§ James D. Clark,† and Atli Thorarensen*,‡ †

Inflammation and Immunology, Pfizer Worldwide R&D, 610 Main Street, Cambridge, Massachusetts 02139, United States Worldwide Medicinal Chemistry, Pfizer Worldwide R&D, 610 Main Street, Cambridge, Massachusetts 02139, United States § Pharmacokinetics, Dynamics, and Metabolism, Pfizer Worldwide R&D, Eastern Point Road, Groton, Connecticut 06340, United States ∥ Primary Pharmacology Group, Pfizer Worldwide R&D, Eastern Point Road, Groton, Connecticut 06340, United States ⊥ Worldwide Medicinal Chemistry, Pfizer Worldwide R&D, Eastern Point Road, Groton, Connecticut 06340, United States ‡

S Supporting Information *

ABSTRACT: PF-06651600, a newly discovered potent JAK3-selective inhibitor, is highly efficacious at inhibiting γc cytokine signaling, which is dependent on both JAK1 and JAK3. PF-06651600 allowed the comparison of JAK3-selective inhibition to panJAK or JAK1-selective inhibition, in relevant immune cells to a level that could not be achieved previously without such potency and selectivity. In vitro, PF-06651600 inhibits Th1 and Th17 cell differentiation and function, and in vivo it reduces disease pathology in rat adjuvant-induced arthritis as well as in mouse experimental autoimmune encephalomyelitis models. Importantly, by sparing JAK1 function, PF-06651600 selectively targets γc cytokine pathways while preserving JAK1-dependent antiinflammatory signaling such as the IL-10 suppressive functions following LPS treatment in macrophages and the suppression of TNFα and IL-1β production in IL-27-primed macrophages. Thus, JAK3-selective inhibition differentiates from pan-JAK or JAK1 inhibition in various immune cellular responses, which could potentially translate to advantageous clinical outcomes in inflammatory and autoimmune diseases.

C

JAK1, JAK2, JAK3, and TYK2. Functioning in pairs, these enzymes drive signaling through type I/II cytokine receptors.4,5 Signaling through the six γc cytokines mentioned earlier requires both JAK1, associated with various beta chains (βc), and JAK3 associated with a unique γc shared by all six receptors.2 One important functional difference between JAK1 and JAK3 is the fact that JAK3 is exclusively associated with the γc receptor chain while JAK1 is associated with a wider array of receptor chains, beyond the βc, signaling through a number of receptors for other cytokines such as IFNs, IL-6, IL-10, IL-27, and IL-35 among others. This functional difference implies that

ytokines play an important role in the induction and regulation of inflammatory and autoimmune diseases such as systemic lupus erythematosus, rheumatoid arthritis, psoriasis, multiple sclerosis, systemic sclerosis, diabetes, and inflammatory bowel disease.1 Many of the cytokines involved in autoimmune diseases signal via the Janus kinase (JAK)/Signal Transducer and Activator of Transcription (STAT) signaling pathway. JAK/STAT-dependent cytokines such as gamma chain (γc) cytokines (IL-2, IL-4, IL-7, IL-9, IL-15 and IL-21), IL-6, IL-12, IL-23, and IFNγ are essential for the development, proliferation, and function of T-cells,2 including the proinflammatory subsets of Th1, Th2, and Th17 T cells, which have been shown to be important players in the development of autoimmune and inflammatory pathologies.1,3 The JAK family of nonreceptor tyrosine kinases is composed of four isoforms: © 2016 American Chemical Society

Received: August 5, 2016 Accepted: October 28, 2016 Published: October 28, 2016 3442

DOI: 10.1021/acschembio.6b00677 ACS Chem. Biol. 2016, 11, 3442−3451

Articles

ACS Chemical Biology

Figure 1. PF-06651600, a JAK3-selective inhibitor. (A) Structure of PF-06651600. (B) Crystal structure of PF-06651600 (magenta ball and stick) bound to JAK3 kinase domain (green; PDB: 5TOZ). The hinge residues (JAK3 residues 903−908) are shown as green thick sticks, with Cys909 as white sticks. Key hydrogen bonds are illustrated as dashed lines with heavy atom distances indicated. Two well-ordered water molecules which complete the hydrogen bonding capacity of PF-06651600 are shown as red spheres. (C) In vitro relationship between JAK3 target occupancy and functional inhibition by PF-06651600. The curves and IC50 are representative of the average obtained with PBMCs from two donors. (D) Schematic representation of PF-06651600 kinome selectivity superimposed on the kinome tree. PF-06651600 was tested at 1 μM against a panel of 304 kinases at Invitrogen. The percent inhibition is color coded and ranges from 0% inhibition (darker green) to 100% inhibition (darker red) with 50% inhibition displayed in yellow.

■

RESULTS AND DISCUSSION PF-06651600 Inhibits JAK3 Selectively in Vitro. In the presence of a physiologically relevant ATP concentration (1 mM), PF-06651600 (Figure 1A) inhibited JAK3 kinase activity with an IC50 of 33.1 nM but without activity (IC50 > 10 000 nM) against JAK1, JAK2, and TYK2 (Table 1). PF-06651600 inhibits JAK3 in an irreversible fashion, as demonstrated using TR-FRET competition assays (Table S1), forming a covalent bond with Cys909 as demonstrated in the JAK3/PF-06651600 cocrystal structure (PDB: 5TOZ; Figure 1B and Table S2). The

a JAK1 inhibitor would have a wider array of effects as compared to a JAK3-selective inhibitor. Improved understanding of the role of Janus kinases (JAKs) in cytokine signaling and immune functions has led to the discovery of multiple JAK inhibitors currently in clinical trials or used in clinical practice.6 However, no truly JAK3-selective inhibitor has reached clinical stages to date.6 While JAK inhibitors such as tofacitinib (CP-690550) and decernotinib are JAK3 inhibitors, and have been described as JAK3-selective in the literature,7−9 they also potently inhibit JAK1 in biochemical assays as well as in cellular assays.6 This fact hinders data interpretation regarding specific JAK isoform biology as JAK3 signals in tandem with JAK1 in the context of the γ-common chain receptors.2 Only recently have “truly” JAK3-selective inhibitors been described. They owe their selectivity to covalent binding to Cys909 in JAK3, a residue replaced by a serine in the other three JAK isoforms.10−12 However, the poor pharmacokinetic properties of these compounds have made them unsuitable for in vivo studies or further clinical development. Here, we describe PF-06651600 (made available at SigmaAldrich, cat # PZ0316), a newly discovered irreversible covalent JAK3-selective inhibitor that exhibits overall properties suitable for both preclinical assessment and advancement to human clinical trial (Clin.gov ref NCT02309827). Importantly, it allows studying biological functions that are differentially affected by JAK3- versus JAK1-specific inhibition.

Table 1. PF-06651600 Inhibition of JAK Isoforms in Biochemical Assaysa JAK isoform

ATP [μM]

IC50 [nM]

SEM (n)

ATP [μM]

IC50 [nM]

SEM (n)

JAK1 JAK2 JAK3 TYK2

40* 4* 4* 12*

1638 1507 0.346 3779

43 (7) 88 (7) 0.025 (7) 464 (7)

1000 1000 1000 1000

>10 000 >10 000 33.1 >10 000

(16) (15) 3.1 (16) (16)

a

Enzymatic assays were performed in the presence of ATP concentrations close to the Km for ATP for each JAK isoform (*) or at 1 mM ATP to approximate cellular concentrations. IC50 values represent the geometric mean of “n” independent experiments. SEM: standard error of the mean. (n) Denotes the number of independent experiments. 3443

DOI: 10.1021/acschembio.6b00677 ACS Chem. Biol. 2016, 11, 3442−3451

Articles

ACS Chemical Biology

IFNγ, G-CSF, or EPO in CD34+ cells (Table 2). Additionally, human whole blood studies with PF-06651600 showed a linear correlation between JAK3 covalent occupancy and inhibition of STAT5 phosphorylation downstream of γ-common cytokines signaling (Figure 1C). This linear correlation was demonstrated in experiments including two arms where, from the same sample, one arm assessed the inhibition of a mixture of γcommon cytokines (IL-2, IL-4, IL-7, IL-15 and IL-21) induced p-STAT5 by FACS, and the other arm quantified by mass spectrometry the amount of free LVM peptide (LVMEYLPSGCLR) containing Cys909 (i.e., unbound to PF06651600). Under these conditions, the inhibition of p-STAT5 is inversely proportional to the JAK3 occupancy by PF06651600 (Figure 1C). The broader kinome selectivity profile of PF-06651600 at 1 μM was assessed against a panel of 304 kinases (Figure 1D, Table S9), demonstrating excellent overall kinome selectivity. The broad kinome selectivity screen revealed that PF-06651600 crossed over into a few kinases that belong to the TEC family of kinases. These kinases share a common feature with JAK3 which is a Cys residue at the same position as Cys909 in JAK3. We therefore assessed the activity of PF-06651600, in the presence of 1 mM ATP, against the 10 kinases in the kinome that possess a Cys residue at the equivalent position of Cys909 in JAK3 (Table S3). Interestingly, the five members of the TEC kinases family were inhibited by PF-06651600 with IC50 values ranging from 194 nM to 8510 nM. Further mechanistic evaluation of inhibition of the kinases that were inhibited by PF-06651600 revealed that in every case the inhibition involved covalent irreversible binding to the kinase with JAK3 being the most potently inhibited due to the faster rate of inactivation (kinact; Table S1). PF-06651600 Has a Low Propensity to Bind Covalently to Other Protein than JAK3. With a potentially reactive chemical moiety, it is important to monitor the

potency and selectivity of PF-06651600 in a cellular context was assessed in total lymphocytes in human whole blood by FACS where it inhibited the phosphorylation of STAT5 elicited by IL-2, IL-4, IL-7, and IL-15 with IC50 values of 244, 340, 407, and 266 nM, respectively, and it inhibited the phosphorylation of STAT3 elicited by IL-21 with an IC50 of 355 nM (Table 2). Table 2. Cellular Potency of PF-06651600 in Total Lymphocytes in Human Whole Blooda cytokine

JAK pairing

p-STAT measured

IC50 [nM]

SEM (n)

IL-2 IL-4 IL-7 IL-15 IL-21 IL-6 IL-6 IL-12 IL-10 IL-27 IFNγ IFNα IL-23 G-CSF EPOb

JAK1/JAK3 JAK1/JAK3 JAK1/JAK3 JAK1/JAK3 JAK1/JAK3 JAK1/JAK2 JAK1/JAK2 JAK2/TYK2 JAK1/TYK2 JAK1/JAK2 JAK1/JAK2 JAK1/TYK2 JAK2/TYK2 JAK1/JAK2 JAK2/JAK2

p-STAT5 p-STAT6 p-STAT5 p-STAT5 p-STAT3 p-STAT1 p-STAT3 p-STAT4 p-STAT3 p-STAT3 p-STAT1 p-STAT3 p-STAT3 p-STAT3 p-STAT5

244 340 407 266 355 >20 000 >20 000 >20 000 >60 000 >60 000 >20 000 >60 000 >20 000 >20 000 >20 000

16 (6) 49 (6) 24 (6) 24 (21) 38 (12) (3) (3) (2) (6) (6) (2) (3) (1) (1) (1)

a

PF-06651600 cellular IC50 values for the phosphorylation of STAT proteins in response to various cytokine treatments. bFor EPO, pSTAT5 was measured in CD34+ cells that were spiked in whole blood.

Selectivity against JAK1, JAK2, and TYK2 was also confirmed in total lymphocytes as exemplified by a lack of inhibition of any STAT phosphorylation downstream of JAK3-independent cytokines including IL-6, IL-10, IL-12, IL-23, IL-27, IFNα,

Figure 2. PF-06651600 inhibition of Th1 and Th17 differentiation and function. (A) PF-06651600 inhibits Th1 differentiation as measured by IFNγ production and (B) Th17 differentiation as measured by IL-17 production. (C) PF-06651600 inhibits in vitro differentiated Th1 function as measure by IFNγ production and (D) in vitro differentiated Th17 function as measured by inhibition of IL-17 production. IC50 values represent the geometric mean of n independent experiments ± SEM. Error bars represent the standard deviation of technical duplicates or triplicates. 3444

DOI: 10.1021/acschembio.6b00677 ACS Chem. Biol. 2016, 11, 3442−3451

Articles

ACS Chemical Biology potential of nonspecific off-target covalent bond formation to monitor the potential generation of haptens that could lead to idiosyncratic reactions.13 To that end, the nonspecific chemical reactivity of PF-06651600 against the proteome was assessed using human serum albumin (HSA) as a surrogate protein. In this system, PF-06651600 clearly showed a lower propensity for covalent binding to HSA compared to CI-1033, a discontinued EGFR inhibitor clinical candidate14 (Figure S1). Furthermore, following incubation with metabolically viable human hepatocytes, 98.4% of radiolabeled PF-06651600 remained nonprotein bound,15 indicating that a small fraction of the daily dose will react covalently with proteins in hepatocytes (Figure S2). JAK3-Selective Inhibition with PF-06651600 Inhibits Th1 and Th17 T Cell Differentiation and Function. T-cell differentiation in Th1 or Th17 is dependent on cytokines, some of which are JAK3-dependent like IL-2 for Th1 cells and IL-21 for Th17 cells, and some are JAK3-independent such as IL-6 and IL-23 for Th17 cells. We therefore assessed the effect of JAK3- versus JAK1-selective inhibition on T helper differentiation. Functional assessment in T-cell differentiation assays demonstrated that PF-06651600 suppressed Th1 and Th17 differentiation as measured by IFNγ, after 5 days under Th1 conditions, and IL-17 production, after 6 days under Th17 conditions, with IC50 values of 30 nM and 167 nM, respectively (Figure 2A and B). PF-06651600 also suppressed Th1 and Th17 function as measured by the inhibition of IFNγ production (IC50 = 48 nM; Figure 2C) and IL-17 production (IC50 = 269 nM; Figure 2D) in cells that had been previously differentiated and rested before being treated with PF06651600. JAK3-Selective Inhibition with PF-06651600 Preserves IL-10-Dependent Suppression of Proinflammatory Cytokines Production Following LPS Treatment in Macrophages. In macrophages, LPS-induced production of IL-6, TNFα, and IL-1β can be suppressed via an IL-10mediated feedback inhibition.16 Since JAK3-selective inhibition spares the signaling of anti-inflammatory cytokines such as IL10 and IL-27 in proximal (p-STAT) cellular assays (Table 2), we performed further functional assessment JAK3-selective inhibition by PF-06651600 in macrophages and compared it to pan-JAK inhibition with CP-690550 (Tables S4 and S5) and JAK1-selective inhibition with PF-02384554 (Tables S6 and S7). Both CP-690550 and PF-02384554 are ATP-competitive reversible inhibitors. In macrophages in vitro, CP-690550 and PF-02384554 inhibited the IL-10-dependent inhibition of LPSinduced proinflammatory cytokine production for IL-6, TNFα, and IL-1β (Figure 3), in agreement with published data for CP690550 and ruxolitinib,16 a JAK1/2 dual inhibitor. In contrast, PF-06651600 preserved the inhibitory effect of IL-10 on LPSinduced production of IL-6, TNFα, and IL-1β (Figure 3). The IL-10 dependence for the LPS-induced production of IL-6, TNFα, and IL-1β is demonstrated by the restoration of a maximal induction of each cytokine following LPS treatment in the presence of a neutralizing IL-10 antibody, regardless of the treatment with JAK inhibitors (Figure 3). JAK3-Selective Inhibition with PF-06651600 Preserves IL-27-Dependent Suppression of TNFα and IL1β Signaling in Macrophages. Kallolias et al. have previously shown that IL-27 can suppress macrophage responses to TNFα and IL-1β.17 Since IL-27 signaling is JAK1-dependent but JAK3-independent, we hypothesized that a JAK3-selective inhibitor would not affect IL-27 signaling while

Figure 3. JAK3-selective inhibition with PF-06651600 preserving IL10-dependent suppression of pro-inflammatory cytokines production following LPS treatment. LPS induced IL-6 secretion (A) or TNFα (B) and IL-1β (C) is enhanced in the presence of pan-JAK inhibition with CP-690550 or with JAK1-selective inhibitor PF-02384554, but not in the presence of JAK3-selective inhibitor PF-06651600, in human monocyte derived macrophages (MDM). These effects are IL10 dependent. Human MDM were stimulated with 20 ng/mL of LPS in the presence or absence of CP-690550, PF-02384554, or PF06651600 ± 4 μg/mL anti-IL-10 antibody. Error bars indicate standard deviation of the technical replicates. Data are representative of two or more independent experiments with different donors.

a JAK1 inhibitor would block IL-27-dependent signaling. Indeed, JAK3-selective inhibition with PF-06651600 maintained the IL-27-dependent suppression of TNFα and IL-1β responses in macrophages as demonstrated by its lack of effect on IL-8 mRNA levels (Figure 4A and B). Conversely, the IL-27 suppressive effect on TNFα and IL-1β responses was inhibited by pan-JAK inhibition with CP-690550 (data not shown) and JAK1-selective inhibition with PF-02384554 as evidenced by the restoration of the IL-8 mRNA to the same level measured in the absence of IL-27 pretreatment (Figure 4A and B). Additionally, JAK3-selective inhibition with PF-06651600 had no effect on IL-27-induced induction of SOCS1 or PD-L1 mRNAs (Figure 4). Conversely, JAK1-selective inhibition with PF-02384554 inhibited SOCS1 and PD-L1 mRNA expression induced by IL-27 pretreatment (Figure 4). The same effects on IL-8, SOCS1, and PD-L1 mRNA levels, as with PF-02384554, 3445

DOI: 10.1021/acschembio.6b00677 ACS Chem. Biol. 2016, 11, 3442−3451

Articles

ACS Chemical Biology

Figure 4. JAK3-selective inhibition with PF-06651600 preserving IL-27-dependent suppression of TNFα and IL-1β signaling in macrophages. IL-27 modulates responses of human MDM to (A) IL-1β-induced IL-8 and SOCS1 mRNA or (B) TNFα-induced IL-8 and SOCS1 mRNA. These responses can be reversed by JAK1-selective inhibition with PF-02384554 but not by JAK3-selective inhibition with PF-06651600. (C) PD-L1 mRNA or cell surface expression induced by TNFα is inhibited by JAK1-selective inhibition with PF-02384554 but not by JAK3-selective inhibition with PF-06651600. Monocytes were pretreated 24 h with 10 ng/mL M-CSF ± 100 ng/mL IL-27 in the presence or absence of JAK inhibitors. Following incubation cells were stimulated with 10 ng/mL of either TNFα or IL-1β for 3 h; cells were then lysed and IL-8 and SOCS1 transcript determined by Q-PCR. Error bars indicate standard deviation on technical replicates. Data are representative of two or more independent experiments with different donors.

were made with the pan-JAK inhibitor CP-690550 (data not shown). The effects of JAK1- and JAK3-selective inhibition on PD-L1 mRNA expression were also observed when cells were stained and analyzed for the surface expression of PD-L1 by FACS (Figure 4C) Effects of JAK3-Selective Inhibition with PF-06651600 on Polarized Macrophages. Macrophages are highly plastic cells, and in response to the microenvironment they can be polarized to fulfill specific functional programs.18 Notably, macrophages can be polarized into pro- (M1) or antiinflammatory (M2) macrophages that can be associated with

beneficial or detrimental roles in various diseases. The polarization toward a classically activated macrophage M1 phenotype is dependent on IFNγ as polarization toward an alternatively activated M2a phenotype is dependent on IL-4 and IL-13.18 We assessed the effect of JAK3-selective inhibition with PF-06651600 on M1 and M2a polarized macrophages and compared it to the effect of JAK1-selective inhibition with PF02384554. In M1 macrophages, both JAK3-selective inhibition with PF-06651600 and JAK1-selective inhibition with PF02384554 inhibited the expression of TNF, IL-6, and CCL3, albeit PF-02384554 did so more potently (Figure 5A), but only 3446

DOI: 10.1021/acschembio.6b00677 ACS Chem. Biol. 2016, 11, 3442−3451

Articles

ACS Chemical Biology

Figure 5. Effects of JAK3-selective inhibition with PF-06651600 on polarized macrophages. (A) RT-PCR analysis of TNFα, IL-6, and CCL3 mRNA in M1 polarized macrophages treated with DMSO (vehicle control) or different doses of JAK inhibitors for 2 days. Results are presented as mean ± SD of triplicate wells normalized relative to GAPDH mRNA. Both JAK1-selective inhibition with PF-02384554 and JAK3-selective inhibition with PF-06651600 inhibit TNF, IL-6, and CCL-3 expression. (B) RT-PCR analysis of CXCL9 and CXCL10 mRNA in M1 macrophages. JAK1-selective inhibitor inhibits CXCL9 and CXCL10 expression while JAK3-selective inhibitor does not affect their expression. (C) Cell surface expression analysis by FACS of CD23 and CD206 in M2a polarized macrophages treated with DMSO (vehicle control) or different doses of JAK inhibitors for 2 days. PF-02384554 inhibits CD23 and CD206 expression while PF-06651600 does not affect their expression. (D) RT-PCR analysis of CCL13, CCL18, and IL-10 mRNA from M2a polarized macrophages treated with DMSO (vehicle control) or different doses of JAK inhibitors for 2 days. PF-02384554 inhibits CCL13 and CCL18 expression while PF-06651600 does not affect their expression. Neither PF-02384554 nor PF-06651600 affects IL-10 expression. Data are representative of three to six independent experiments with different donors.

obtained with pan-JAK inhibition with CP-690550 in M2a macrophages were similar to the data obtained with JAK1selective inhibition with PF-02384554 (data not shown). PF-06651600 Has Suitable Pharmacokinetics and Pharmacodynamics Properties for Preclinical and Clinical Evaluations. A key feature distinguishing PF06651600 from other JAK3 covalent irreversible inhibitors in the same chemical series or published in the literature are favorable pharmacokinetic properties making it suitable for oral dosing and study in preclinical models of inflammation (Table S8). This is critical because the half-life for JAK3 enzyme turnover was in the 3−4 h range in human primary CD4+ T cells (Figure S3), indicating that irreversible covalent inhibition of JAK3 would not lead to a significantly extended effect. The

PF-02384554 inhibited the expression of CXCL9 and CXCL10 (Figure 5B). Data obtained with pan-JAK inhibition with CP690550 in M1 macrophages were similar to the data obtained with JAK1-selective inhibition with PF-02384554 (data not shown). Under M2a macrophage polarizing conditions, PF02384554 inhibited surface expression of the M2a macrophage markers CD23 and CD206, while PF-06651600 had no effect on the expression of these two molecules (Figure 5C). In M2a macrophages, JAK1-selective inhibition with PF-02384554 also inhibited the expression of CCL13 and CCL18 while JAK3selective inhibition with PF-06651600 had no effect on the expression of these chemokines (Figure 5D). Interestingly, IL10 mRNA expression in M2a macrophages was not affected by either JAK1- or JAK3-selective inhibition (Figure 5D). Data 3447

DOI: 10.1021/acschembio.6b00677 ACS Chem. Biol. 2016, 11, 3442−3451

Articles

ACS Chemical Biology

Figure 6. JAK3-selective inhibition with PF-06651600 ameliorating rat AIA. (A) Immunized rats were enrolled into groups and dosed with either: 30, 10, or 3 mg/kg PF-06651600 or vehicle control (1 mEQ/methylcellulose) PO, QD. Compared to vehicle, treatment with PF-06651600 significantly reduced disease severity measured by paw swelling by plethysmograph in the 30 mg/kg (day 2 to 7), 10 mg/kg (day 2 to 7), and 3 mg/ kg (day 2 to 7) dosed groups (p ≤ 0.05, t test). (B) Percent effect was assessed by evaluating the difference in area under the curve (AUC) between vehicle and drug treatment arms normalized to the vehicle AUC. Percent effect versus unbound average concentration (Cav) data was fitted to an Emax model yielding an EC50 of 169 nM (95% CI 39−299 nM), which matched well with the in vitro unbound rat whole blood IL-21-induced pSTAT3 IC50 of 181 nM.

Conclusions. The discovery and characterization of PF06651600 described in this work allow for the first time to differentiate the biological effects of JAK3 from JAK1 inhibition. Since JAK3 always signals in pair with JAK1 in the context of the γc cytokine receptors, it has been challenging to characterize the relative contribution of both JAK isoforms to transduce signaling from these receptors in the absence of JAK isoform-selective inhibitors. Although there has been significant effort to develop such isoform-selective inhibitors for the treatment of autoimmune diseases, so far only JAK1-selective or pan-JAK inhibitors have reach clinical development stages,6 while developing a JAK3-selective inhibitor suitable for preclinical or clinical development has eluded the scientific community for the past two decades. Although, JAK3 inhibitors such as CP-690550 and decernotinib have been developed,7−9 they are not JAK3-selective as they have been shown to inhibit JAK1 comparably to JAK3 under 1 mM ATP concentration, which is physiologically relevant, and they also both inhibit JAK1-dependent but JAK3-independent cytokine signaling, such as IFNα in primary human cells.6,23 One noticeable difference between JAK1 and JAK3 is the affinity for ATP, with JAK3 having about a 10-fold higher affinity for ATP,23 with a Km for ATP around 4 μM for JAK3 and about 40 μM for JAK1.23 This enzymatic difference between JAK1 and JAK3 results in a higher biochemical hurdle to inhibit JAK3 with ATP competitive inhibitors, when compared to JAK1, in a cellular context where the ATP concentration is in the millimolar range. This is one of the reasons for the development of a truly JAK3selective ATP competitive and reversible inhibitor being unfruitful. This higher biochemical hurdle to inhibit JAK3 has also led to the notion that JAK1 is dominant over JAK3 to transduce a signal through γc containing receptors.24 Our current work demonstrates that JAK3-selective inhibition with PF-06651600 is sufficient to abrogate signaling through γc containing receptors, thus disproving the notion of a dominant role of JAK1 over JAK3 in γc receptor signaling. In the work of Haan et al., we attribute the lack of efficiency of the JAK3-selective inhibitor NIBR3049 to inhibit STAT5 phosphorylation to insufficient potency in cellular settings.23,24 Our work, as well as the work of others,10−12 demonstrates that potent JAK3-selective inhibition over other JAK isoforms can be achieved by capitalizing on the fact that JAK3 contains a

metabolism of PF-06651600 was determined to be mediated by both CYP450 and glutathione-S-transferase (GST) enzymes. To monitor PF-06651600 stability in human whole blood, novel methods were used to scale in vitro to in vivo clearance through scaling hepatocyte intrinsic clearance by standard wellstirred methods and empirically scaling intrinsic clearance by blood GST as a surrogate for extra-hepatic GST-mediated clearance. Human pharmacokinetics of PF-06651600 were predicted to show approximate values of blood clearance of 5.6 mL/min/kg, a steady state volume of distribution of 1.3 L/kg, oral bioavailability of 90%, and a systemic half-life of 2 h. An initial pharmacokinetic and pharmacodynamic analysis of PF06651600 showed good translation between whole blood (correcting for blood/plasma distribution and plasma protein binding) and isolated PBMC with IL-21-induced P-STAT3 IC50 values of 71 nM and 54 nM, respectively (data not shown). These data indicated that PF-06651600 has good cell availability consistent with its high passive permeability and that free drug concentration primarily drives pharmacology. In fact, the in vivo dynamic IL-21 pSTAT3 modulation potency (unbound IC50 = 190 nM) in rats was in agreement with static in vitro assessments in rats (unbound IC50 = 181 nM) without evidence of hysteresis (Figure S4). Downstream pharmacodynamic anti-inflammatory effects of PF-06651600 in vivo were shown in several preclinical models.19 JAK3-Selective Inhibition with PF-06651600 Is Efficacious in the Rat Adjuvant-Induced Arthritis (AIA) and the Mouse Experimental Autoimmune Encephalomyelitis (EAE) Model. The favorable rodent PK properties of PF06651600 (Table S8) allowed for the investigation of the effects of JAK3-selective inhibition in vivo in two rodent models of arthritis and encephalomyelitis. In the rat adjuvant-induced arthritis20 (AIA) model, PF-06651600 reduced paw swelling (Figure 6A) with an unbound EC50 of 169 nM (Figure 6B). Similarly, PF-06651600 significantly reduced disease severity in the experimental autoimmune encephalomyelitis21,22 (EAE) mouse model when dosed either therapeutically at 30 or 100 mg/kg or prophylactically at 20 and 60 mg/kg (Figure S5). The efficacy of PF-06651600 in these two rodent models of inflammatory and autoimmune diseases illustrate that JAK3selective inhibition could be sufficient to have disease modifying effects in human diseases. 3448

DOI: 10.1021/acschembio.6b00677 ACS Chem. Biol. 2016, 11, 3442−3451

Articles

ACS Chemical Biology

not with JAK3-selective inhibition with PF-06651600 (Figure 5C and D). Interestingly, IL-10 mRNA expression was not affected by either PF-02384554 or PF-06651600 (Figure 5D). Further functional studies, in vitro and in vivo, in various macrophages populations will be required to further assess the differential effects of JAK1- versus JAK3-selective inhibition in these cells. The selectivity of PF-06651600 for the inhibition of JAK3 over the other three JAK isoforms is achieved by irreversible covalent binding to Cys909 in JAK3, which is replaced by a Ser residue in the other three JAKs. Importantly, PF-06651600 displayed excellent overall kinome selectivity when tested against a panel of 304 kinases in vitro. Interestingly, the five TEC kinases members,29 which all contain a Cys residue at the same position as Cys909 in JAK3, were also inhibited to various extents by PF-06651600, albeit less potently than JAK3 (Table S1). Further in vitro characterization of PF-06651600 in cellular assays dependent on BTK kinase activity measuring anti-IgDinduced CD69 expression in primary human B cells or antiIgM-induced IP1 accumulation in Ramos cells showed no to little inhibition (data not shown). Further in vitro and in vivo characterization will be required to further assess the potential functional effects of PF-06651600 crossing over into the TEC kinases. One important characteristic of PF-06651600 over previously described JAK3 covalent inhibitors10−12 is its favorable pharmacokinetics and pharmacodynamics properties. Indeed, PF-06651600 displayed a low propensity to form a nonspecific covalent adduct with HSA. Additionally, PF06651600 showed suitable oral bioavailability and clearance parameters (Table S8) for further preclinical and clinical testing (Clin.gov ref NCT02309827). Based on these favorable properties, we evaluated JAK3selective inhibition with PF-06651600 for its anti-inflammatory properties in the EAE mouse model on encephalomyelitis and in the AIA rat model of arthritis. In both disease models, JAK3selective inhibition was shown to be efficacious at reducing disease severity (Figures 6 and S5). The fact that JAK3-selective inhibition is efficacious in two rodent models of rheumatoid arthritis and multiple sclerosis in humans suggests that JAK3selective inhibition could be considered as a target for intervention in these diseases as well as other inflammatory and autoimmunes diseases. Here, we have shown that PF-06651600 is an irreversible covalent and selective JAK3 inhibitor with potency and properties that make it suitable for further clinical development. We have also illustrated in various cellular systems that JAK3selective inhibition differentiates from JAK1-selective or panJAK inhibition and that it is sufficient to inhibit signaling from γc cytokine receptors in vitro and achieve efficacy in vivo. This mechanistic differentiation suggests that in disease settings, JAK3-selective inhibition could not only preserve the immunoregulatory function of cytokines such as IL-10 and IL-27 but also preserve M2a differentiation/function, which is inhibited by JAK1 inhibitors. Inhibiting signaling via the γcommon chain receptors with a JAK3-selective inhibitor while preserving signaling via IL-10, IL-27, and other immunoregulatory cytokines such as IL-35 could potentially offer better anti-inflammatory properties as compared to inhibiting signaling of a broader spectrum of cytokines with JAK1 inhibitors, particularly in diseases such as IBD or MS where preserving signaling of such immunoregulatory cytokines is beneficial in preclinical models.25−28 Further preclinical investigations and

unique Cys residue at position 909, which is replaced by a Ser residue in the other three JAKs, allowing for covalent irreversible inhibition of JAK3. Importantly, PF-06651600 is the first JAK3-selective inhibitor with suitable PK/PD properties allowing for in vivo characterization of JAK3-selective inhibition. Having developed a JAK3-selective inhibitor, we sought to characterize JAK3-selective inhibition in comparison to JAK1-selective or pan-JAK inhibition. To do so, we compared the properties of PF-06651600 to a JAK1-selective inhibitor (PF-02384554) developed at Pfizer as well as to panJAK inhibition with CP-690550 (Figure S6). In cellular settings, JAK3-selective inhibition with PF06651600 inhibited STATs phosphorylation downstream of γc cytokine receptors signaling but left signaling of JAK3independent cytokine receptors intact (Table 2). The inhibition of γc cytokine receptor signaling translated functionally in the inhibition of IFNγ and IL-17 production in Th1 and Th17 cells, respectively. These observations are consistent with a role for γc cytokines in Th1 and Th17 differentiation and function including IL-2 for Th1 and IL-21 for Th17. Although both JAK1- and JAK3-selective inhibition block signaling through the γc cytokine receptors, one significant difference between JAK1- and JAK3-selective inhibitors is the spectrum of cytokines inhibited, limited to the γc cytokines with JAK3-selective inhibition but including many other cytokines with JAK1-selective inhibition. Importantly, a number of JAK1-dependent cytokines, like IL-10, IL-27, or IL-35, have immunoregulatory or anti-inflammatory functions.25−28 It has previously been shown that JAK inhibition with CP690550 or ruxolitinib increases LPS-induced cytokine production at later time points by blocking the IL-10-mediated feedback.16 Since IL-10 signaling is JAK1-dependent and JAK3independent, we hypothesized that JAK1-selective inhibition would block this feedback loop as shown with CP-690550 but would be spared by JAK3-selective inhibition. Under such conditions, JAK1-selective inhibition restored the full induction of IL-6, TNFα, and IL-1β while JAK3-selective inhibition maintained the IL-10 feedback suppression of IL-6, TNFα, and IL-1β following LPS treatment (Figure 3). Similarly, IL-27 has been shown to suppress macrophage responses to TNFα and IL-1β.17 IL-27 signaling, like IL-10, being JAK1-dependent and JAK3-independent, was inhibited by JAK1-selective inhibition but was not affected by JAK3-selective inhibition as measured by their effects on IL-8, SOCS1, and PD-L1 mRNA expression (Figure 4). Further preclinical and clinical investigations will be required to further evaluate the potential differential effects of inhibiting immunoregulatory cytokines such as IL-10 and IL-27 with JAK1 inhibitors while sparing them with JAK3-selective inhibition. We also investigated the effects of JAK3- and JAK1-selective inhibition on M1 and M2a macrophage polarization. In M1 macrophages, both JAK3-selective and JAK1-selective inhibition with PF-06651600 and PF-02384554, respectively, inhibited the expression of TNFα, IL-6, and CCL3, albeit PF-02384554 was more potent than PF-06651600 (Figure 5). Interestingly, two other chemokines, CCL9 and CCL10, which are expressed in M1 macrophages but not in unpolarized macrophages or M2a and M2c polarized macrophages (data not shown), were clearly inhibited by JAK1-selective inhibition but not with JAK3-selective inhibition (Figure 5B). In M2a macrophages, expression of specific markers of M2a polarization such as CD23, CD206, CCL13, and CCL18 were all inhibited by JAK1-selective inhibition with PF-02384554 but 3449

DOI: 10.1021/acschembio.6b00677 ACS Chem. Biol. 2016, 11, 3442−3451

Articles

ACS Chemical Biology

Discovery of VX-509 (Decernotinib): A Potent and Selective Janus Kinase 3 Inhibitor for the Treatment of Autoimmune Diseases. J. Med. Chem. 58, 7195−7216. (9) Changelian, P. S., Flanagan, M. E., Ball, D. J., Kent, C. R., Magnuson, K. S., Martin, W. H., Rizzuti, B. J., Sawyer, P. S., Perry, B. D., Brissette, W. H., McCurdy, S. P., Kudlacz, E. M., Conklyn, M. J., Elliott, E. A., Koslov, E. R., Fisher, M. B., Strelevitz, T. J., Yoon, K., Whipple, D. A., Sun, J., Munchhof, M. J., Doty, J. L., Casavant, J. M., Blumenkopf, T. A., Hines, M., Brown, M. F., Lillie, B. M., Subramanyam, C., Shang-Poa, C., Milici, A. J., Beckius, G. E., Moyer, J. D., Su, C., Woodworth, T. G., Gaweco, A. S., Beals, C. R., Littman, B. H., Fisher, D. A., Smith, J. F., Zagouras, P., Magna, H. A., Saltarelli, M. J., Johnson, K. S., Nelms, L. F., Des Etages, S. G., Hayes, L. S., Kawabata, T. T., Finco-Kent, D., Baker, D. L., Larson, M., Si, M. S., Paniagua, R., Higgins, J., Holm, B., Reitz, B., Zhou, Y. J., Morris, R. E., O’Shea, J. J., and Borie, D. C. (2003) Prevention of organ allograft rejection by a specific Janus kinase 3 inhibitor. Science 302, 875−878. (10) Gehringer, M., Pfaffenrot, E., Bauer, S., and Laufer, S. A. (2014) Design and synthesis of tricyclic JAK3 inhibitors with picomolar affinities as novel molecular probes. ChemMedChem 9, 277−281. (11) Goedken, E. R., Argiriadi, M. A., Banach, D. L., Fiamengo, B. A., Foley, S. E., Frank, K. E., George, J. S., Harris, C. M., Hobson, A. D., Ihle, D. C., Marcotte, D., Merta, P. J., Michalak, M. E., Murdock, S. E., Tomlinson, M. J., and Voss, J. W. (2015) Tricyclic covalent inhibitors selectively target Jak3 through an active site thiol. J. Biol. Chem. 290, 4573−4589. (12) Tan, L., Akahane, K., McNally, R., Reyskens, K. M., Ficarro, S. B., Liu, S., Herter-Sprie, G. S., Koyama, S., Pattison, M. J., Labella, K., Johannessen, L., Akbay, E. A., Wong, K. K., Frank, D. A., Marto, J. A., Look, T. A., Arthur, J. S., Eck, M. J., and Gray, N. S. (2015) Development of Selective Covalent Janus Kinase 3 Inhibitors. J. Med. Chem. 58, 6589−6606. (13) Singh, J., Petter, R. C., Baillie, T. A., and Whitty, A. (2011) The resurgence of covalent drugs. Nat. Rev. Drug Discovery 10, 307−317. (14) Slichenmyer, W. J., Elliott, W. L., and Fry, D. W. (2001) CI1033, a pan-erbB tyrosine kinase inhibitor. Semin. Oncol. 28, 80−85. (15) Bauman, J. N., Kelly, J. M., Tripathy, S., Zhao, S. X., Lam, W. W., Kalgutkar, A. S., and Obach, R. S. (2009) Can in vitro metabolismdependent covalent binding data distinguish hepatotoxic from nonhepatotoxic drugs? An analysis using human hepatocytes and liver S-9 fraction. Chem. Res. Toxicol. 22, 332−340. (16) Pattison, M. J., Mackenzie, K. F., and Arthur, J. S. (2012) Inhibition of JAKs in macrophages increases lipopolysaccharideinduced cytokine production by blocking IL-10-mediated feedback. J. Immunol. 189, 2784−2792. (17) Kalliolias, G. D., Gordon, R. A., and Ivashkiv, L. B. (2010) Suppression of TNF-alpha and IL-1 signaling identifies a mechanism of homeostatic regulation of macrophages by IL-27. J. Immunol. 185, 7047−7056. (18) Martinez, F. O., Sica, A., Mantovani, A., and Locati, M. (2008) Macrophage activation and polarization. Front. Biosci. 13, 453−461. (19) Leung, L., Yang, X., Strelevitz, T. J., Montgomery, J., Brown, M. F., Zientek, M. A., Banfield, C., Gilbert, A. M., Thorarensen, A., and Dowty, M. E. (2016) Clearance Prediction of Targeted Covalent inhibitors by in vitro-in vivo extrapolation of hepatic and extrahepatic clearance mechanisms. Drug Metab. Dispos., DOI: 10.1124/ dmd.116.072983. (20) Langman, C. B., Ford, K. K., Pachman, L. M., and Glorieux, F. (1990) Vitamin D metabolism in rats with adjuvant-induced arthritis. J. Bone Miner. Res. 5, 905−913. (21) Stromnes, I. M., and Goverman, J. M. (2006) Active induction of experimental allergic encephalomyelitis. Nat. Protoc. 1, 1810−1819. (22) McCarthy, D. P., Richards, M. H., and Miller, S. D. (2012) Mouse models of multiple sclerosis: experimental autoimmune encephalomyelitis and Theiler’s virus-induced demyelinating disease. Methods Mol. Biol. 900, 381−401. (23) Thorarensen, A., Banker, M. E., Fensome, A., Telliez, J. B., Juba, B., Vincent, F., Czerwinski, R. M., and Casimiro-Garcia, A. (2014) ATP-mediated kinome selectivity: the missing link in understanding

clinical development of PF-06651600 will advance assessment of the biological impact and clinical benefit of JAK3-selective inhibition.

■ ■

EXPERIMENTAL PROCEDURES

Methods and procedures are provided in the Supporting Information.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acschembio.6b00677. Methods, Tables S1−S9, Figures S1−S6 (PDF)

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: Jean-Baptiste.Telliez@pfizer.com. *E-mail: Atli.Thorarensen@pfizer.com. Notes

The authors declare the following competing financial interest(s): All authors are or were Pfizer employees.

■

ACKNOWLEDGMENTS All authors of this manuscript are or were employees of Pfizer. Use of the IMCA-CAT beamline 17-ID at the Advanced Photon Source was supported by the companies of the Industrial Macromolecular Crystallography Association through a contract with Hauptman-Woodward Medical Research Institute. Use of the Advanced Photon Source was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. DE-AC0206CH11357.

■

REFERENCES

(1) Yang, J., Sundrud, M. S., Skepner, J., and Yamagata, T. (2014) Targeting Th17 cells in autoimmune diseases. Trends Pharmacol. Sci. 35, 493−500. (2) Rochman, Y., Spolski, R., and Leonard, W. J. (2009) New insights into the regulation of T cells by gamma(c) family cytokines. Nat. Rev. Immunol. 9, 480−490. (3) Kimura, A., and Kishimoto, T. (2010) IL-6: regulator of Treg/ Th17 balance. Eur. J. Immunol. 40, 1830−1835. (4) Ghoreschi, K., Laurence, A., and O’Shea, J. J. (2009) Janus kinases in immune cell signaling. Immunol. Rev. 228, 273−287. (5) Laurence, A., Pesu, M., Silvennoinen, O., and O’Shea, J. (2012) JAK Kinases in Health and Disease: An Update. Open Rheumatol. J. 6, 232−244. (6) Clark, J. D., Flanagan, M. E., and Telliez, J. B. (2014) Discovery and development of Janus kinase (JAK) inhibitors for inflammatory diseases. J. Med. Chem. 57, 5023−5038. (7) Mahajan, S., Hogan, J. K., Shlyakhter, D., Oh, L., Salituro, F. G., Farmer, L., and Hoock, T. C. (2015) VX-509 (decernotinib) is a potent and selective janus kinase 3 inhibitor that attenuates inflammation in animal models of autoimmune disease. J. Pharmacol. Exp. Ther. 353, 405−414. (8) Farmer, L. J., Ledeboer, M. W., Hoock, T., Arnost, M. J., Bethiel, R. S., Bennani, Y. L., Black, J. J., Brummel, C. L., Chakilam, A., Dorsch, W. A., Fan, B., Cochran, J. E., Halas, S., Harrington, E. M., Hogan, J. K., Howe, D., Huang, H., Jacobs, D. H., Laitinen, L. M., Liao, S., Mahajan, S., Marone, V., Martinez-Botella, G., McCarthy, P., Messersmith, D., Namchuk, M., Oh, L., Penney, M. S., Pierce, A. C., Raybuck, S. A., Rugg, A., Salituro, F. G., Saxena, K., Shannon, D., Shlyakter, D., Swenson, L., Tian, S. K., Town, C., Wang, J., Wang, T., Wannamaker, M. W., Winquist, R. J., and Zuccola, H. J. (2015) 3450

DOI: 10.1021/acschembio.6b00677 ACS Chem. Biol. 2016, 11, 3442−3451

Articles

ACS Chemical Biology the contribution of individual JAK Kinase isoforms to cellular signaling. ACS Chem. Biol. 9, 1552−1558. (24) Haan, C., Rolvering, C., Raulf, F., Kapp, M., Druckes, P., Thoma, G., Behrmann, I., and Zerwes, H. G. (2011) Jak1 has a dominant role over Jak3 in signal transduction through gammaccontaining cytokine receptors. Chem. Biol. 18, 314−323. (25) Collison, L. W., Chaturvedi, V., Henderson, A. L., Giacomin, P. R., Guy, C., Bankoti, J., Finkelstein, D., Forbes, K., Workman, C. J., Brown, S. A., Rehg, J. E., Jones, M. L., Ni, H. T., Artis, D., Turk, M. J., and Vignali, D. A. (2010) IL-35-mediated induction of a potent regulatory T cell population. Nat. Immunol. 11, 1093−1101. (26) Fitzgerald, D. C., Ciric, B., Touil, T., Harle, H., Grammatikopolou, J., Das Sarma, J., Gran, B., Zhang, G. X., and Rostami, A. (2007) Suppressive effect of IL-27 on encephalitogenic Th17 cells and the effector phase of experimental autoimmune encephalomyelitis. J. Immunol. 179, 3268−3275. (27) Fitzgerald, D. C., Zhang, G. X., El-Behi, M., Fonseca-Kelly, Z., Li, H., Yu, S., Saris, C. J., Gran, B., Ciric, B., and Rostami, A. (2007) Suppression of autoimmune inflammation of the central nervous system by interleukin 10 secreted by interleukin 27-stimulated T cells. Nat. Immunol. 8, 1372−1379. (28) Zhang, X., Koldzic, D. N., Izikson, L., Reddy, J., Nazareno, R. F., Sakaguchi, S., Kuchroo, V. K., and Weiner, H. L. (2004) IL-10 is involved in the suppression of experimental autoimmune encephalomyelitis by CD25+CD4+ regulatory T cells. Int. Immunol. 16, 249− 256. (29) Horwood, N. J., Urbaniak, A. M., and Danks, L. (2012) Tec family kinases in inflammation and disease. Int. Rev. Immunol. 31, 87− 103.

3451

DOI: 10.1021/acschembio.6b00677 ACS Chem. Biol. 2016, 11, 3442−3451