ATP-Mediated Kinome Selectivity: The Missing ... - ACS Publications

May 12, 2014 - ABSTRACT: Kinases constitute an important class of therapeutic targets being explored both by academia and the pharmaceutical industry...
0 downloads 0 Views 1MB Size
Download PDF
Articles pubs.acs.org/acschemicalbiology

ATP-Mediated Kinome Selectivity: The Missing Link in Understanding the Contribution of Individual JAK Kinase Isoforms to Cellular Signaling Atli Thorarensen,* Mary Ellen Banker, Andrew Fensome, Jean-Baptiste Telliez, Brian Juba, Fabien Vincent, Robert M. Czerwinski, and Agustin Casimiro-Garcia Pfizer Worldwide Research, 200 Cambridgepark Drive, Cambridge, Massachusetts 02140, United States S Supporting Information *

ABSTRACT: Kinases constitute an important class of therapeutic targets being explored both by academia and the pharmaceutical industry. The major focus of this effort was directed toward the identification of ATP competitive inhibitors. Although it has long been recognized that the intracellular concentration of ATP is very different from the concentrations utilized in biochemical enzyme assays, little thought has been devoted to incorporating this discrepancy into our understanding of translation from enzyme inhibition to cellular function. Significant work has been dedicated to the discovery of JAK kinase inhibitors; however, a disconnect between enzyme and cellular function is prominently displayed in the literature for this class of inhibitors. Herein, we demonstrate utilizing the four JAK family members that the difference in the ATP Km of each individual kinase has a significant impact on the enzyme to cell inhibition translation. We evaluated a large number of JAK inhibitors in enzymatic assays utilizing either 1 mM ATP or Km ATP for the four isoforms as well as in primary cell assays. This data set provided the opportunity to examine individual kinase contributions to the heterodimeric kinase complexes mediating cellular signaling. In contrast to a recent study, we demonstrate that for IL-15 cytokine signaling it is sufficient to inhibit either JAK1 or JAK3 to fully inhibit downstream STAT5 phosphorylation. This additional data thus provides a critical piece of information explaining why JAK1 has incorrectly been thought to have a dominant role over JAK3. Beyond enabling a deeper understanding of JAK signaling, conducting similar analyses for other kinases by taking into account potency at high ATP rather than Km ATP may provide crucial insights into a compound’s activity and selectivity in cellular contexts.

K

number of compounds tested against a select panel of kinases7 or comprehensive selectivity assessments across the kinome for a more limited number of compounds have been published in recent years.8,9 This work resulted in the need to quantify selectivity to allow comparisons between compounds to be made, and it prompted the development of methods such as Gini or thermodynamic partition index.10,11 Janus kinases (JAK) are nonreceptor tyrosine kinases required for signaling through type I/II cytokine receptors.12 There are four JAK family members, JAK1, JAK2, JAK3, and TYK2. JAK3 and TYK2 are primarily involved in immunerelated functions, whereas JAK1 and JAK2 also play critical roles related to hematopoiesis, growth, and neuronal functions, among others. A notable feature of JAK signaling is the requirement for a dimer of JAK kinases at the cytokine receptor complex level. Cytokine receptors signaling through JAK1/ JAK3, JAK1/JAK2, JAK1/TYK2, JAK2/TYK2, and JAK2/JAK2 have been described.

inases constitute a large family within the human genome, with individual members playing key roles in numerous cellular signaling pathways. To date, it is estimated that only a fraction of the kinome has been mined as a source of therapeutic targets.1 Multiple kinases inhibitors have been approved by the U.S. FDA in recent years. Although a large majority of these has been directed at oncology indications, recent approval of Xeljanz for rheumatoid arthritis illustrates the potential importance of this class of targets for other disease indications.2 These successes led to explosive growth in the field of kinase drug discovery, and significant effort has been devoted toward the identification, optimization, and pharmacological characterization of novel inhibitors. While historically a majority of the work has focused on the design of ATP competitive inhibitors (type I), kinase inhibitor design started to include the discovery of alternative inhibition mechanisms (types II−IV) as the field matured.3 Significant effort is being devoted to understanding the structural requirements that place inhibitors in the appropriate mechanistic class. While structural elements likely to result in type I or II inhibitors are known, the other classes are less well understood.4,5 Selectivity is another key component in kinase inhibitor design, and great attention has been devoted to kinase selection for inclusion in selectivity panels.6 Substantial selectivity data sets containing either a large © XXXX American Chemical Society

Received: March 19, 2014 Accepted: May 12, 2014

A

dx.doi.org/10.1021/cb5002125 | ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology

Articles

Table 1. Characterization of the Four JAK Isoforms ATP Km assays

1 mM ATP assays

enzyme

active enzyme (%)

ATP (μM)

total enzyme (nM)

total enzyme (nM)

theoretical shift

experimental shift

JAK1 JAK2 JAK3 TYK2

12 60 40 40

40 4 4 12

30 2 2 2

20 1 1 1

13 125 125 42

∼10 ∼50 ∼100 ∼25

eq 2. This equation can be further simplified to eq 3 when one of the IC50 values is determined at [ATP] = Km.

The importance of the various individual JAKs in cytokine signaling has resulted in significant effort to identify selective inhibitors for each family member.13,14 Inhibitor design has been guided by the availability of crystal structures for each isoform.15 These inhibitors have served an important role in elucidating the role of JAKs in cell signaling.16,17 The selectivity characterization of these inhibitors has exclusively utilized enzymatic assays performed with an ATP concentration equal to enzyme Km, producing IC50,Km values, or by describing the intrinsic affinity of the inhibitors as a KD. This data was then supplemented by various cellular assays to confirm the selectivity profile.18,19 Discrepancies between enzyme and cellular data for JAK kinases has been recognized for some time, but potential explanations supported by experimentation have been published only recently to address these inconsistencies.20 It has long been recognized that the cellular concentration of ATP is in the 1 to 5 mM range. Therefore, neither a KD nor an IC50 measurement at ATP Km can be a good descriptor of cellular activity for ATP competitive inhibitors. The relationship of activity at various substrate concentrations was described by Cheng−Prusoff in the early 1970s (eq 1) for competitive compounds.21 The importance of that relationship did not escape the drug discovery community, and it has been invoked to explain disconnects between biochemical and cellular compound activity,22,23 as presented in a seminal article by Shokat et al.24 It should be noted that these concerns are relevant only in the context of ATP competitive, type I kinase inhibitors. Nonetheless, the literature is very sparse in case studies where this concern is clearly articulated or addressed experimentally. An example of incorporating this thinking into our understanding of kinase function was described for the identification of inhibitors for interleukin-2 inducible T cell kinase (ITK).25 In recent years, the importance of JAK3 in the JAK1/JAK3 heterodimeric pair has been the source of significant discussion and conflicting conclusions.26−28 In that context, the challenge for a dual JAK1/3 inhibitor is to determine correlations utilizing enzyme and cell data accounting for the individual kinase contribution to cellular signaling. In this article, we describe the relationships of enzyme inhibition measured at both Km and physiological ATP concentrations and how these relate to cellular activity. We find these considerations to be critical in understanding the role of individual JAK kinases in their heterodimeric signaling complexes. ⎛ [ATP] ⎞ ⎟⎟ IC50 = K i⎜⎜1 + [K m,ATP] ⎠ ⎝

(1 + (1 + (1 + =

IC shift = 50A = IC50B

IC50A shift = IC50K m

[ATPA ] [K m,ATP] [ATPB ] [K m,ATP]

) ) )

(2)

[ATPA ] [K m,ATP]

2

(3)

Experimentally, we measured the enzyme activity of the catalytic domains of the four JAK isoforms, determining both Km as well as the fraction of active protein through active site titration (Table 1). The assays were also optimized to measure enzyme inhibition in the presence of 1 mM ATP, with all isoforms producing linear correlations between IC50 values obtained in assays run at [ATP] = Km and 1 mM ATP. In order to understand the interplay of the various JAK heterodimers on cellular function in greater detail, we mined the Pfizer database for compounds that had been evaluated in both enzymatic and cellular JAK assays. We then evaluated these compounds against the four JAK isoforms, utilizing a 1 mM ATP concentration. The enzyme concentration coupled with its active fraction determines the lower limit of accuracy for enzymatic assays. Consequently, IC50 values above 3 nM were considered to be relevant for this discussion. Similarly, although the compound concentrations used allowed for IC50 measurements of up to 10 μM, we chose to restrict the range of IC50 values used in this analysis to those below 1 μM because of the lack of solubility at 10 μM of some compounds in this broad set. The entire range of determined IC50’s for all compounds is nonetheless illustrated in each graph. The measured shift was largest for JAK3 (approximately 100-fold, Figure 1), whereas it

(1)

■

RESULTS AND DISCUSSION Starting from the Cheng−Prusoff relationship (eq 1), the predicted potency shift for a compound tested at two different ATP concentrations (for example, [A] and [B]) is described by

Figure 1. Correlation of JAK3 compound potency at [ATP] = Km vs 1 mM. B

dx.doi.org/10.1021/cb5002125 | ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology

Articles

Figure 2. Relationship between IL-15-stimulated STAT5 phosphorylation and JAK3 enzymatic activity. (a) Enzyme assay with ATP concentration at Km and (b) enzyme assay with ATP concentration at 1 mM. Legend: color by binned RRCK passive permeability AB (red, 10 (high); gray, data not available). Shape by x = JAK1/JAK3 binned selectivity using 1 mM ATP assay IC50 value. Square, x < 1; circle, 1 < x < 50; triangle, 50 < x. Lines illustrate shifts between the assay (black, unity; blue, 10×; red, 100×).

Figure 3. Relationship between IL-15-stimulated STAT5 phosphorylation and JAK3/JAK1 enzymatic activity, considering only highly permeable compounds. (a) JAK3 enzyme assay at 1 mM ATP; data is illustrated only for permeable compounds with a JAK1/JAK3 ratio greater than 1. (b) JAK1 enzyme assay at 1 mM ATP; data is illustrated only for permeable compounds with a JAK1/JAK3 ratio lower than 1. Legend: shape by x = JAK1/JAK3 binned selectivity using 1 mM ATP assay IC50 values. Square, x 50, triangles) display a good correlation between enzymatic and cellular activities. In order to better understand the relationship between inhibitions of JAK1 and JAK3 enzymes to cellular function, we conducted a more refined analysis using the subset of compounds with good cellular permeability. The correlations between the individual JAK1 or JAK3 isoforms and the cellular heterodimeric readout are depicted in Figure 3a,b. It should be noted that these correlations are observed only when utilizing enzyme and selectivity data obtained at 1 mM ATP. This is due to a very limited overlapping dynamic range of the cell-based assay IC50’s and the Km-based JAK3 enzyme assay IC50’s that would be expected to translate into a measurable cellular inhibition. Here, compounds with selectivity greater than 1 for a specific isoform display an excellent correlation between enzyme inhibition of this individual JAK isoform at 1 mM ATP

was the smallest for JAK1 (approximately 10-fold, Figure S1). In the case of TYK2 and JAK2, the shifts were around 25- and 50-fold, respectively (data not shown). When compared to the shifts predicted by eq 3, there was good agreement between measured and theoretical shifts. Six cytokines (IL-2, -4, -7, -9, -15, and -21) signal through the JAK1/JAK3 heterodimeric pair. In this study, we elected to study IL-15 and IL-21 signaling through JAK1/JAK3 by measuring STAT5 phosphorylation in peripheral blood mononuclear cells (PBMCs) and in Kit225 cells. When we compared enzyme IC50 values obtained at [ATP] = Km for JAK1 and JAK3 isoforms and cellular activities, JAK3 enzyme potency displayed little correlation with Kit225 cellular activity, even when cellular permeability was taken into account (Figure S2a,b). Following this finding, we focused our analysis on data generated with human PBMCs to avoid any artificial bias introduced by using immortalized cells. Using this more physiologically relevant cellular system, a different picture emerges irrespective of whether the correlation is being C

dx.doi.org/10.1021/cb5002125 | ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology

Articles

selective inhibitors, NIBR304928 and WYE-151650,27 previously characterized in the literature. Previous work with NIBR3049 led to the conclusion that JAK1 is dominant over JAK3 in cellular signaling. In contrast, studies conducted with WYE-151650 indicated that inhibiting JAK3 is sufficient to fully inhibit JAK1/JAK3-dependent signaling pathways and to achieve efficacy in the mouse collagen-induced arthritis model.27 We profiled both compounds in our [ATP] = Km and 1 mM enzyme assays as well as in a series of cellular assays (Table 2). For NIBR3049, the half-maximal inhibition of JAK3 enzyme activity in the presence of 1 mM ATP was measured at 131 nM, as compared to 1.5 nM at Km (4 μM) ATP; an IC50 of 8 nM obtained with 18 μM ATP was reported in ref 28. Its IC50 for JAK1 was approximately 2 μM at 1 mM ATP. This compound therefore displays only 18-fold selectivity for JAK3 over JAK1 at high ATP. Importantly, the cellular inhibition measured downstream of IL-2 or IL-15 signaling is now comparable to the observed inhibition of JAK3 enzyme activity. In comparison, WYE-151650 showed a similar ratio of selectivity for JAK3 over JAK1 at 1 mM ATP, with the JAK3 inhibition at 1 mM ATP translating fairly well into the cellular inhibition observed downstream of IL-15. The additional JAK3 inhibition data obtained at 1 mM ATP thus bridges the contradiction within the literature on the importance of JAK3 function in γ-common chain cytokine receptor signaling. Importantly, the PBMC assays displayed little variability from batch to batch of PBMCs and from day to day. For example, in the IL-15 assay, IC50 = 535 nM is a geometric mean of five individual data points obtained on three different days with two batches of PBMCs (batch A2544 IC50 = 437, 439, 534, and 648 nM and batch 2687 IC50 = 659 nM). This clearly illustrates that cellular concentrations of ATP should be taken into account when interpreting enzyme to cell shifts for kinase targets. When this is done and with a larger data set, it becomes clear that JAK3 inhibition is in fact sufficient for blocking STAT5 phosphorylation in PBMCs stimulated with IL-15. In summary, we have illustrated that determining enzyme inhibition of the individual JAK1 and JAK3 isoforms under more physiologically relevant ATP conditions provides a better understanding of the role that each individual partner plays in the heterodimeric signaling pair. With a series of compounds displaying appropriate selectivity and high permeability, we demonstrated that inhibiting only one partner of the heterodimeric JAK1/JAK3 pair is sufficient to fully inhibit IL15 signaling. Accordingly, the differences in the ATP Km of the different JAK enzymes appear to be the major culprit behind the observed disconnects between enzymatic and cellular assays with regard to inhibition and selectivity. However, the significant differences observed in IC50 values for WYE151650 with different cytokines (e.g., IL-10 and IFNα) signaling through the JAK1/TYK2 heterodimer pair illustrates that additional work is still required to fully understand the role of JAKs in cytokine receptor signaling. Nonetheless, the approach described in this article can be applied to other cytokine receptors and can help decipher the interplay of JAK enzymes in cytokine signaling. Furthermore, this study illustrates the importance of utilizing biochemical IC50 values obtained with a physiologically relevant ATP concentration rather than at ATP Km in order to properly assess compound selectivity and to better predict cellular function.

and the cellular readout. This illustrates that activity of the more dominantly inhibited JAK isoform has good translation to cellular function. There was no correlation between inhibition of other JAK isoforms (JAK2 and TYK2) and cellular readout utilizing IL-15 as stimulant (data not shown). An alternative way of illustrating the relationship between JAK1 and JAK3 biochemical potency and IL-15-mediated cell signaling is to plot the cellular activity against the enzyme activity ratios for both JAK1 and JAK3 (Figure 4). Figure 4A depicts the

Figure 4. Relationship between (A) predicted and (B) experimental cellular and enzyme (1 mM ATP) potency for JAK1 and JAK3. Legend: color by binned IL-15 cellular IC50 (purple, 0.25 < x < 1 μM; blue, 0.05 < x < 0.25; yellow, 0.01 < x < 0.05; green, x < 0.01 μM). J1, JAK1; J3, JAK3.

theoretical correlation one could expect to observe depending on whether cellular activity would be controlled by JAK1 and/ or JAK3. This plotting approach illustrates that selective compounds for either isoform can lead to complete inhibition of cellular activity. In conclusion, inhibition of IL-15 signaling can be achieved by sufficiently inhibiting either one of the members of the JAK1/JAK3 heterodimer, and the potency in the enzyme assay determined with 1 mM ATP is predictive of the cellular readout. There are several reported examples of compounds utilized to investigate the role of the individual JAK partners in signaling events.16−20 We decided to evaluate two JAK3 D

dx.doi.org/10.1021/cb5002125 | ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology

Articles

Table 2. Characterization of NIBR3049 and WYE-151650

JAK1 IC50 (nM) Km/1 mM ATP JAK2 IC50 (nM) Km/1 mM ATP JAK3 IC50 (nM) Km/1 mM ATP TYK2 IC50 (nM) Km/1 mM ATP p-STAT5/IL-2 (nM) [J1,J3] p-STAT5/IL-15 (nM) [J1,J3] p-STAT3/IL-10 (nM) [J1,T2] p-STAT3/IFNα (nM) [J1,T2]

NIBR3049a

NIBR3049b

WYE-151650

1017 2550 8 8055 1294 525 NA NA

211/2387 (n = 4) 141/5453 (n = 4) 1.5/131 (n = 4) 2101/>10 000 (n = 4) 524 (n = 5) 535 (n = 5) ∼10 000 (n = 4) 536 (n = 6)

16/111(n = 2) 1/28 (n = 2) 0.4/12 (n = 2) 17/401(n = 2) NDc 29 (n = 4) 610 (n = 2) 41 (n = 1)a

a

Data from ref 28. bData generated for this article. IL-2-mediated signaling occurs through the JAK1/JAK3 heterodimer; IL-15-mediated signaling occurs through the JAK1/JAK3 heterodimer; IL-10-mediated signaling occurs through the JAK1/TYK2 heterodimer; IFNα-mediated signaling occurs through the JAK1/TYK2 heterodimer. Cellular assays were conducted using human PBMCs. cTesting limited because of compound availability.

■

and Km, the Michaelis−Menten constant, is the substrate concentration needed to achieve a half-maximal enzyme velocity. Caliper JAK Enzyme End-Point IC50 Assays. Test compounds were solubilized in dimethyl sulfoxide (DMSO) to a stock concentration of 30 mM. Compounds were diluted in DMSO to create an 11-point half-log dilution series with a top concentration of 600 μM. The test compound plate also contained positive control wells with a known inhibitor to define 100% inhibition and negative control wells with DMSO to define no inhibition. The compound plates were diluted 1 to 60 in the assay, resulting in a final assay compound concentration range of 10 μM to 100 pM and a final assay concentration of 1.7% DMSO. Two hundred fifty nanoliters of test compounds and controls solubilized in 100% DMSO were added to a 384-well polypropylene plate (Matrical) using an non-contact acoustic dispenser. Kinase assays were carried out at room temperature in a 15 μL reaction buffer containing 20 mM HEPES, pH 7.4, 10 mM magnesium chloride, 0.01% bovine serum albumin (BSA), 0.0005% Tween 20, and 1 mM DTT. Reaction mixtures contained 1 μM of a fluorescently labeled synthetic peptide, a concentration lower than the apparent Km (5FAM-KKSRGDYMTMQID for JAK1 and TYK2 and FITC-KGGEEEEYFELVKK for JAK2 and JAK3). Reaction mixtures contained ATP at either a level equal to the apparent Km for ATP (40 μM for JAK1, 4 μM for JAK2, 4 μM for JAK3 and 12 μM for TYK2) or at 1 mM ATP. The assays were stopped with 15 μL of a buffer containing 180 mM HEPES, pH 7.4, 20 mM EDTA, and 0.2% coating reagent, resulting in a final concentration of 10 mM EDTA, 0.1% coating reagent, and 100 mM HEPES, pH 7.4. Each assay reaction was then sampled to determine the level of phosphorylation. The data output used for calculations was percent product converted and was determined for each sample and control well based on peak height (percent product = product/(product + substrate)). The percent effect at each concentration of test compound was calculated on the basis of the positive and negative control well contained within each assay plate using the following formula: percent effect = 100((sample well − negative control)/(positive control − negative control)). The percent effect was plotted against the compound concentration compound. An unconstrained sigmoid curve was fitted using a four-parameter logistic model, and the concentration of test compound required for 50% inhibition (IC50) was determined for each test compound. Preparation of Peripheral Blood Mononuclear Cells (PBMC). Cryopreserved human PBMCs (no. PB005F), which were used in the

MATERIALS AND METHODS

JAK Enzymes. GST-tagged recombinant human kinase domains of JAK1, JAK2, and JAK3 were purchased from Invitrogen. His-tagged recombinant human TYK2 kinase domain was expressed in SF21/ baculovirus and purified using a two-step affinity (Ni-NTA) and sizeexclusion (SEC S200) purification method. See the Supporting Information for commercial and sequence information. JAK Caliper Assays. Human Janus kinase (JAK) activity was determined using a microfluidic assay to monitor phosphorylation of a synthetic peptide by the recombinant human kinase domain of each of the four members of the JAK family, JAK1, JAK2, JAK3, and TYK2. Reaction mixtures contained 1 μM of a fluorescently labeled synthetic peptide and ATP either at a level equal to the apparent Km for ATP or at 1 mM ATP. Each assay condition was optimized for enzyme concentration and room temperature incubation time to obtain a conversion rate of 20−30% phosphorylated peptide product. Reactions were terminated by the addition of stop buffer containing EDTA. Utilizing LabChip 3000 mobility shift technology (Caliper Life Science), each assay reaction was sampled to determine the level of phosphorylation. This technology is separation-based, allowing direct detection of fluorescently labeled substrates and products with separations controlled by a combination of vacuum pressure and electric field strength optimized for the peptide substrate. Determination of Apparent Km for ATP Using the Caliper JAK Enzyme Assays. Kinase assays were carried out at room temperature in a 384-well polypropylene plate in 80 μL of reaction buffer containing 20 mM HEPES, pH 7.4, 10 mM magnesium chloride, 0.01% bovine serum albumin (BSA), 0.0005% Tween 20, 1 mM DTT, and 2% DMSO. Reaction mixtures contained 1 μM of a fluorescently labeled synthetic peptide (5FAM-KKSRGDYMTMQID for JAK1 and TYK2 and FITC-KGGEEEEYFELVKK for JAK2 and JAK3) and various concentrations of ATP. The kinase reactions were initiated by the addition of JAK enzymes and were sampled at various time points to determine the level of peptide phosphorylation. The percent product converted was determined for each sample based on peak height (percent product = product/(product + substrate)). The enzyme velocities were determined for each concentration of ATP, and a Km for ATP was determined using the Michaelis−Menten model, Y = Vmax × X/(Km + X). Vmax is the maximum enzyme velocity E

dx.doi.org/10.1021/cb5002125 | ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology

Articles

Notes

IL-2, IL-10, IL-15, and IFNα assays, were purchased from Allcells (Emeryville, CA). Frozen PBMCs were thawed in a water bath (37 °C) and washed once with once with 5% FBS/RPMI-1640 medium (+Pen/Strep, +Sodium Pyruvate) for IFNα assay, and 10%FBS/ Dulbecco’s PBS (no. 14190-250; Life Technologies) for IL-2, IL-10 and IL-15 assay (no. 72400, Life Technologies). Cells were resuspended in RPMI-1640 Medium (+Pen/Strep, +Sodium Pyruvate) containing 5% fetal bovine serum (FBS) for IFNα assay, and 100% FBS for IL-2, IL-10 and IL-15 assays and incubated at 37 °C for 1.5 h for resting. Cytokine Stimulation and FACS Sample Preparation. Serum was removed from resting PBMCs by washing once with Dulbecco’s PBS. PBMCs were resuspended in RPMI-1640 medium (+ Pen/Strep, + Sodium Pyruvate). Ninety microliters of resuspended PBMCs (5 × 106 cells/mL) were aliquoted in 96-well, deep well, V-bottom plates (no. 82007-292; VWR, Radnor, PA) and incubated at 37 °C for 30 min. Cells were treated with compound (5 μL/well) at various concentrations (0.2% DMSO final) at 37 °C for 60 min for IFNα assay, and 75 min for IL-2, IL-10 and IL-15 assays followed by the challenge with cytokine (5 μL/well; final concentration of 41 ng/mL IL-2, 41 ng/mL IL-15, 5000 U/mL IFNα, 15 ng/mL IL-10) for 15 min. Cells were treated with warm 1% paraformaldehyde (300ul/well, prepared from 16% paraformaldehyde, Cat#15710-S Electron Miscroscopy Sciences) to terminate activation and were further incubated at 37 °C for 15 min for IFNα assay, and room temperature for 15 min for IL-2, IL-10 and IL-15 assays. Plates were centrifuged at 1200rpm for 5 min, supernatant was aspirated, and cells were washed with 800 μL per well of staining buffer (0.5% heat-inactivated FBS and 0.001% sodium azide in Dulbecco’s PBS). The washed cell pellets were resuspended with 350 μL per well of cold 90% methanol (−20 °C) and incubated 4 °C for 30 min. After the removal of 90% methanol, cells were washed once with staining buffer (800 μL/well). Cell pellets were resuspended in staining buffer containing anti-phospho-STATAlexaFluor647-conjugated antibodies (1 to 250 dilution, 210 μL/well) and incubated at rt in the dark overnight. Anti-phospho-STAT3AlexaFluor647 (no. 557815; BD Biosciences) was used for IL-10- and IFNα-stimulated cells, and anti-phospho-STAT5-AlexaFluor647 (no. 612599; BD Biosciences) was used for IL-2- and IL-15-stimulated cells. Flow Cytometry. Samples were transferred to 96-well U-bottom plates, and flow cytometric analysis was performed on a FACSCalibur, FACSCanto, or LSRFortessa equipped with a HTS plate loader (BD Biosciences). Lymphocyte population was gated for the pSTAT3 or pSTAT5 (APC channel) histogram analysis. Background fluorescence was defined using unstimulated cells, and a gate (M1) was placed at the foot of the peak to include ∼2% gated population. The histogram statistical analysis was performed using CellQuest Pro version 5.2.1 (BD Biosciences), FACSDiva version 6.2 (BD Biosciences), or FlowJo version 7.6.1 (Ashland, OR) software. Relative fluorescence unit (RFU), which measures the level of pSTAT, was calculated by multiplying the percent positive population (M1) and its mean fluorescence. IC50 values were determined using the Prism version 5 software (GraphPad, La Jolla, CA).

■

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We would like to thank the numerous members of Pfizer’s JAK research project team who designed, prepared, and evaluated JAK compounds, thus enabling the data mining and additional evaluation of these compounds.

■

ASSOCIATED CONTENT

S Supporting Information *

Correlation for JAK1 enzyme inhibition data generated at Km and 1 mM ATP concentration; relationship between JAK1 and JAK3 enzyme inhibition data obtained at Km ATP concentrations with cellular inhibition in Kit225 cells. Detailed information on the enzymes utilized in the assays. This material is available free of charge via the Internet at http:// pubs.acs.org.

■

REFERENCES

(1) Knapp, S., Arruda, P., Blagg, J., Burley, S., Drewry, D. H., Edwards, A., Fabbro, D., Gillespie, P., Gray, N. S., Kuster, B., Lackey, K. E., Mazzafera, P., Tomkinson, N. C., Willson, T. M., Workman, P., and Zuercher, W. J. (2013) A public−private partnership to unlock the untargeted kinome. Nat. Chem. Biol. 9, 3−6. (2) Cohen, P., and Alessi, D. R. (2013) Kinase Drug Discovery − What’s Next in the Field? ACS Chem. Biol. 18, 96−104. (3) Dar, A. C., and Shokat, K. M. (2011) The evolution of protein kinase inhibitors from antagonist to agonist of cellular signaling. Annu. Rev. Biochem. 80, 769−795. (4) Backes, A. C., Zech, B., Felber, B., Klebl, B., and Müller, G. (2008) Small-molecular inhibitors binding to protein kinases. Part I: Exceptions from the traditional pharmacophore approach of type I inhibition. Expert Opin. Drug Discovery 3, 1409−1425. (5) Backes, A. C., Zech, B., Felber, B., Klebl, B., and Müller, G. (2008) Small-molecular inhibitors binding to protein kinases. Part II: The novel pharmacophore approach of type II and type III inhibition. Expert Opin. Drug Discovery 3, 1427−1449. (6) Brandt, P., Jensen, A. J., and Nilsson, J. (2009) Small kinase assay panels can provide a measure of selectivity. Bioorg. Med. Chem. Lett. 19, 5861−5863. (7) Posy, S. L., Hermsmeier, M. A., Vaccaro, W., Ott, K.-H., Todderud, G., Lippy, J. S., Trainor, G. L., Loughney, D. A., and Johnson, S. R. (2011) Trends in kinase selectivity: insights for target class-focused library screening. J. Med. Chem. Lett. 54, 54−66. (8) Anastassiadis, T., Deacon, S. W., Devarajan, K., Ma, H., and Peterson, J. R. (2011) Comprehensive assay of kinase catalytic activity reveals features of kinase inhibitor selectivity. Nat. Biotechnol. 29, 1039−1046. (9) Davis, M. I., Hunt, J. P., Herrgard, S., Ciceri, P., Wodicka, L. M., Pallares, G., Hocker, M., Treiber, D. K., and Zarrnkar, P. P. (2011) Comprehensive analysis of kinase inhibitor selectivity. Nat. Biotechnol. 29, 1046−1052. (10) Graczyk, P. P. (2007) Gini coefficient: A new way to express selectivity of kinase inhibitors against a family of kinases. J. Med. Chem., 5773−5779. (11) Cheng, A. C., Eksterowicz, J., Geuns-Meyer, S., and Sun, Y. (2010) Analysis of kinase inhibitor selectivity using a thermodynamicsbased partition index. J. Med. Chem. 53, 4502−4510. (12) 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. (13) Norman, P. (2012) Selective JAK1 inhibitor and selective TYK2 inhibitor patents. Expert Opin. Ther. Pat. 22, 1233−1249. (14) Dymock, B. W., and See, C. S. (2013) Inhibitors of JAK2 and JAK3: an update on the patent literature 2010 − 2012. Expert Opin. Ther. Pat. 23, 449−501. (15) Alicea-Velázquez, N. L., and Boggon, T. J. (2011) The use of structural biology in Janus kinase targeted drug discovery. Curr. Drug Targets 12, 546−555. (16) Lee, J. E., Lee, A. S., Kim, D. H., Jung, Y. J., Lee, S., Park, B.-H., Lee, S. H., Park, S. K., Kim, W., and Kang, K. P. (2012) Janex-1, a JAK3 inhibitor, ameliorates tumor necrosis factor-α-induced expression of cell adhesion molecules and improves myocardial vascular permeability in endotoxemic mice. Int. J. Mol. Med. 29, 864−870. (17) Cetkovic-Cvrlje, M., Olson, M., and Ghate, K. (2012) Targeting Janus tyrosine kinase 3 (JAK3) with an inhibitor induces secretion of TGF-β by CD4+ T cells. Cell Mol. Immunol. 9, 350−60.

AUTHOR INFORMATION

Corresponding Author

*E-mail: Atli.Thorarensen@pfizer.com. F

dx.doi.org/10.1021/cb5002125 | ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology

Articles

(18) Zak, M., Mendonca, R., Balazs, M., Barrett, K., Bergeron, P., Blair, W. S., Chang, C., Deshmukh, G., DeVoss, J., Dragovich, P. S., Eigenbrot, C., Ghilardi, N., Gibbons, P., Gradl, S., Hamman, C., Hanan, E. J., Harsstad, E., Hewitt, P. R., Hurley, C. A., Jin, T., Johnson, A., Johnson, T., Kenny, J. R., Koehler, M. F. T., Kohli, P. B., Kulagowski, J. J., Labadie, S., Liao, J., Liimatta, M., Lin, Z., Lupardus, P. J., Maxey, R. J., Murray, J. M., Pulk, R., Rodriguez, M., Savage, S., Hia, S., Steffek, M., Ubhayakar, S., Ultsch, M., Abbema, A. V., Ward, S. I., Xiao, L., and Xiao, Y. (2012) Discovery and optimization of C-2 methyl imidazopyrrolopyridines as potent and orally bioavailable JAK1 inhibitors with selectivity over JAK2. J. Med. Chem. 55, 6176−6193. (19) Soth, M., Hermann, J. C., Yee, C., Alam, M., Barnett, J. W., Berry, P., Browner, M. F., Frank, K., Frauchiger, S., Harris, S., He, Y., Hekmat-Nejad, M., Hendricks, T., Henningsen, R., Hilgenkamp, R., Ho, H., Hoffman, A., Hsu, P.-Y., Hu, D.-Q., Itano, A., Jaime-Figueroa, S., Jahangir, A., Jin, S., Kuglstatter, A., Kutach, A. K., Liao, C., Lynch, S., Menke, J., Niu, L., Patel, V., Railkar, A., Roy, D., Shao, A., Shaw, D., Steiner, S., Sun, Y., Tan, S.-L., Wang, S., and Vu, M. D. (2013) 3Amido pyrrolopyrazine JAK kinase inhibitors: development of a JAK3 vs JAK1 selective inhibitor and evaluation in cellular and in in vivo models. J. Med. Chem. 56, 345−356. (20) Yu, V., Pistillo, J., Archibeque, I., Lee, J., Sun, B.-C., Schenkel, L. B., Geuns-Meyer, S., Liu, L., and Emkey, R. (2013) Differential selectivity of JAK2 inhibitors in enzymatic and cellular settings. Exp. Hematol. 41, 491−500. (21) Cheng, Y., and Prusoff, W. H. (1973) Relationship between the inhibition constant (K1) and the concentration of inhibitor which causes 50% inhibition (I50) of an enzymatic reaction. Biochem. Pharmacol. 22, 3099−108. (22) Frantz, B., Klatt, T., Pang, M., Parsons, J., Rolando, A., Williams, H., Toci, M. J., O’Keefe, S. J., and O’Neill, E. A. (1998) The activation state of p38 mitogen-activated protein kinase determines the efficiency of ATP competition for pyridinylimidazole inhibitor binding. Biochemistry 37, 13846−13853. (23) Anderson, D. R., Meyers, M. J., Vernier, W. F., Mahoney, M. W., Kurumbail, R. G., Caspers, N., Poda, G. I., Schindler, J. F., Reitz, D. B., and Mourey, R. J. (2007) Pyrrolopyridine inhibitors of mitogenactivated protein kinase-activated protein kinase 2 (MK-2). J. Med. Chem. 50, 2647−2654. (24) Knight, Z. A., and Shokat, K. M. (2005) Features of selective kinase inhibitors. Chem. Biol. 12, 621−637. (25) Zapf, C. W., Gerstenberger, B. S., Xing, L., Limburg, D. C., Anderson, D. R., Caspers, N., Han, S., Aulabaugh, A., Kurumbail, R., Shakya, S., Li, X., Spaulding, V., Czerwinski, R. M., Seth, N., and Medley, Q. G. (2012) Covalent inhibitors of interleukin-2 inducible T cell kinase (Itk) with nanomolar potency in a whole-blood assay. J. Med. Chem. 55, 10047−10063. (26) Thoma, G., Nuninger, F., Falchetto, R., Hermes, E., Tavares, G. A., Vangrevelinghe, E., and Zerwes, H. G. (2011) Identification of a potent Janus kinase 3 inhibitor with high selectivity within the Janus kinase family. J. Med. Chem. 54, 284−288. (27) Lin, T. H., Hegen, M., Quadros, E., Nickerson-Nutter, C. L., Appell, K. C., Cole, A. G., Shao, Y., Tam, S., Ohlmeyer, M., Wang, B., Goodwin, D. G., Kimble, E. F., Quintero, J., Gao, M., Symanowicz, P., Wrocklage, C., Lussier, J., Schelling, S. H., Hewet, A. G., Xuan, D., Krykbaev, R., Togias, J., Xu, X., Harrison, R., Mansour, T., Collins, M., Clark, J. D., Webb, M. L., and Seidl, K. J. (2010) Selective functional inhibition of JAK-3 is sufficient for efficacy in collagen-induced arthritis in mice. Arthritis Rheum. 62, 2283−2293. (28) Haan, C., Rolvering, C., Raulf, F., Kapp, M., Drückes, P., Thoma, G., Behrmann, I., and Zerwes, H. G. (2011) Jak1 has a dominant role over Jak3 in signal transduction through γc-containing cytokine receptors. Chem. Biol. 25, 314−323.

G

dx.doi.org/10.1021/cb5002125 | ACS Chem. Biol. XXXX, XXX, XXX−XXX