Abstract
Phosphoinositide 3-kinases (PI3K) are key signaling enzymes regulating cellular survival, development, and function. Expression of the PI3Kδ isoform is largely restricted to leukocytes and it plays a key role in immune cell development and function. Seletalisib is a novel small-molecule inhibitor of PI3Kδ that was evaluated in biochemical assays, cellular assays of adaptive and innate immunity, and an in vivo rat model of inflammation. Our findings show that seletalisib is a potent, ATP-competitive, and selective PI3Kδ inhibitor able to block protein kinase B (AKT) phosphorylation following activation of the B-cell receptor in a B-cell line. Moreover, seletalisib inhibited N-formyl peptide–stimulated but not phorbol myristate acetate–stimulated superoxide release from human neutrophils, consistent with a PI3Kδ-specific activity. No indications of cytotoxicity were observed in peripheral blood mononuclear cells (PBMCs) or other cell types treated with seletalisib. Findings from cellular assays of adaptive immunity demonstrated that seletalisib blocks human T-cell production of several cytokines from activated T-cells. Additionally, seletalisib inhibited B-cell proliferation and cytokine release. In human whole blood assays, seletalisib inhibited CD69 expression upon B-cell activation and anti-IgE-mediated basophil degranulation. Seletalisib showed dose-dependent inhibition in an in vivo rat model of anti-CD3-antibody-induced interleukin 2 release. Collectively, these data characterize seletalisib as a selective PI3Kδ inhibitor and potential therapeutic candidate for the treatment of B-cell malignancies and autoimmune diseases driven by dysregulated proinflammatory cytokine secretion.
Introduction
PI3Kδ as a Therapeutic Target for Inhibition.
Phosphoinositide 3-kinases (PI3Ks) are a family of enzymes that phosphorylate phosphoinositides and regulate key cellular processes that include survival, proliferation, and differentiation. Class I PI3Ks have been shown to participate in the orchestration of signaling events that lead to immune cell development, with specific functional effects in individual cell-lineage subsets (Okkenhaug, 2013a). The PI3Kα and PI3Kβ isoforms are ubiquitously expressed, whereas expression of PI3Kγ and PI3Kδ is enriched in leukocytic cells compared with other cell types, implying a more specific and important role in leukocyte development and function (Ji et al., 2007; Banham-Hall et al., 2012; Okkenhaug, 2013a). Although PI3Kγ and PI3Kδ have similar cellular distributions, they are differentially coupled in specific cells. PI3Kγ is generally coupled to G protein-coupled receptors (GPCRs), including leukotriene and chemokine receptors, and plays a particularly important role in chemokine-mediated recruitment and the activation of innate immune cells at sites of inflammation (Rommel et al., 2007; Okkenhaug, 2013a; Hawkins and Stephens, 2015). In contrast, PI3Kδ signals predominantly downstream of several immune receptor tyrosine kinases, including the T-cell receptor (TCR), the B-cell receptor (BCR), and the high-affinity IgE receptor Fc epsilon receptor 1 (FcεR1), thereby playing a greater role in the acquired immune response (Bilancio et al., 2006; Al-Alwan et al., 2007; Okkenhaug et al., 2007, 2014; Costa et al., 2011; Bartok et al., 2012; Hawkins and Stephens, 2015).
Selective expression of PI3Kδ in leukocytes provides a rationale for investigating this isoform as a therapeutic target for diseases featuring pathologic activation of the PI3Kδ pathway in hematopoietic cells, including B-cell malignancies and immune and inflammatory disorders such as rheumatoid arthritis, asthma, psoriasis, and systemic lupus erythematosus (Foster et al., 2012; Somoza et al., 2015; Stark et al., 2015). Lymphocyte developmental defects, attenuated signaling by antigen receptors, B-cell functional defects, and decreased T-cell-dependent responses are observed in mice that lack or express mutant p110δ, the catalytic subunit of PI3Kδ (Clayton et al., 2002; Jou et al., 2002; Okkenhaug et al., 2002; Hebeis et al., 2004). Moreover, it has been demonstrated that lipopolysaccharide-induced functions of B-cells in mice require PI3K activity mediated primarily via the p110δ catalytic subunit (Hebeis et al., 2004).
There is evidence that B-cell-mediated presentation of antigens to T-cells is a key process underlying the initiation and amplification of autoimmune diseases (Puri and Gold, 2012). Inhibition of PI3Kδ activity prevents B-cell migration, adhesion, survival, activation, and proliferation, which diminishes the capacity for B-cell-mediated presentation to autoreactive T-cells (Puri and Gold, 2012). This results in nonactivation of autoreactive T-cells and subsequent reduced secretion of autoantibodies and proinflammatory cytokines.
Although PI3Kδ activity has been shown to be required for normal immune cell development and function, inherited gain-of-function mutations in the p110δ protein are associated with recently described primary immunodeficiency conditions in humans that can increase susceptibility to bacterial and viral infections (Angulo et al., 2013; Deau et al., 2014; Lucas et al., 2014). The apparently paradoxical outcome of both loss-of-function and gain-of function mutations in PI3Kδ leading to immune cell defects and immunosuppression has been proposed to result from a requirement for specific cellular and temporal regulation of PI3Kδ activity for normal immune cell development (Lucas et al., 2016). Enhanced activation of PI3K has been identified in peripheral blood T-cells from patients with systematic lupus erythematosus (Suarez-Fueyo et al., 2011) and is believed to account for vascular dysfunction observed in a mouse model of type-1 diabetes (Pinho et al., 2010). This raises the possibility that selective PI3Kδ inhibitors may provide a viable alternative therapeutic approach to treating immune-inflammatory diseases. Indeed, the potential for such a role has been enhanced by evidence that PI3Kδ inhibition can be achieved not only in naïve T-cells but also in previously activated memory T-cells from healthy and allergic donors and from patients with spondyloarthritis (Soond et al., 2010). The potential for inhibition of PI3Kδ to result in undesirable on-target pharmacology, including gut inflammation (Okkenhaug et al., 2002) and defective responses to infection (Gracias et al., 2016), came from studies with kinase-inactive p110δ knock-in animals. Data from clinical studies with PI3Kδ inhibitors are becoming available and a better defined picture of the effects of PI3Kδ blockade will emerge in the coming years. Interestingly, topline results from two phase II studies of duvelisib, a selective PI3Kδ/γ inhibitor administered across different low-dose ranges, suggest poor efficacy in the immune-inflammatory setting (Infinity Pharmaceuticals, 2014, 2015). However, in the oncology setting, phase I studies evaluating higher-dose duvelisib in patients with advanced hematologic malignancies (including relapsed/refractory indolent non-Hodgkin lymphoma) or treatment-naïve chronic lymphocytic leukemia have observed clinical activity (Flinn et al., 2014; Patel et al., 2015) .
Seletalisib, N-[(1R)-1-[8-chloro-2-(1-oxidopyridin-1-ium-3-yl)-3-quinolyl]-2,2,2-trifluoro-ethyl]pyrido[3,2-day]pyrimidin-4-amine, is a novel small-molecule quinolone-based selective PI3Kδ inhibitor with potential as an oral agent for the treatment of a range of immune-inflammatory diseases. The medicinal chemistry program leading to the discovery of seletalisib will be described elsewhere in due course. In this article, we present the studies of the biochemical and biologic profile of seletalisib to demonstrate its potency and selectivity as a PI3Kδ inhibitor.
Materials And Methods
Compound Handling, Blood Donation, and Data Analysis.
Figure 1 shows the structure of seletalisib. For in vitro assays, compounds were made up from solid as a 10 mM stock in dimethyl sulfoxide (DMSO; Sigma-Aldrich, Poole, UK). Serial dilutions were carried out in DMSO. Compounds were prediluted into assay buffer (biochemical assays) or media (cellular assays) prior to addition to the final assay wells or tubes. Human blood samples were obtained under a license (#12504) granted to UCB under Section 16 (2) (e) (ii) of the Human Tissue Act (UK) and all donors provided their written consent to participate. Donated samples were anonymized and came from working-age individuals from the in-house UCB blood donation service. The concentration of DMSO was maintained in all treated and control wells at 0.1% (v/v) unless otherwise stated. Concentration-response curves were analyzed using four-parameter logistic curve using XLfit or GraphPad Prism (version 6.01) software.
In Vitro Kinase Activity Assays and Off-Target Profiling.
The activities of seletalisib and the pan-PI3K inhibitor UCB1370037 (Hutchinson et al., 2008) were tested in biochemical kinase assays of the PI3Kα, PI3Kβ, PI3Kδ (EMD Millipore, Darmstadt, Germany), and PI3Kγ (manufactured in-house) isoforms. Competitive time-resolved (TR) fluorescence resonance energy transfer (FRET) assays using a four-step PI3K assay reagent kit (EMD Millipore) were performed according to manufacturer instructions. In this assay, PI3K activity is measured by the detection of phosphatidylinositol triphosphate (PIP3) as follows: A complex is formed between biotinylated PIP3 and a glutathione S-transferase (GST)-tagged pleckstrin homology–domain protein, which are detected by streptavidin-allophycocyanin (APC) and europium-labeled anti-GST antibody, respectively. Proximity of the two fluorophores results in a stable TR-FRET signal. The production of unlabeled PIP3 by PI3K in the kinase reaction results in competitive inhibition of the complex and therefore the FRET signal. Briefly, 2 μl of compound was added to wells containing 8 μl of a mixture containing phosphatidylinositol bisphosphate (PIP2) substrate and ATP (Roche, Burgess Hill, UK), before addition of 10 μl of PI3K (prediluted to the required concentration). The final assay concentrations of ATP and PIP2 substrate were 2 μM and 10 μM, respectively. The final assay concentration of PI3K enzymes was 1–8 nM depending on batch activity. The final assay concentration of DMSO was 2.0%. The plate was incubated at room temperature for 30 minutes before addition of stop buffer and detection buffer according to manufacturer instructions, and the plate was then incubated for a further 6 hours at room temperature. TR-FRET measurements were performed on an Analyst GT plate reader (330 nm excitation filter, 615/665-nm emission filters; Molecular Devices, Wokingham, UK). The ATP concentration dependence of the activity of seletalisib against PI3Kδ was established by running the assay in the presence of varying ATP concentrations (2, 40, 200, and 1000 μM). The final assay concentrations of PI3Kδ and PIP2 substrate in these assays were 1 nM and 25 μM, respectively.
The kinase selectivity of seletalisib was assessed using the SelectScreen FRET-based biochemical profiling service provided by Life Technologies, Paisley, UK. A screening concentration of 10 μM was selected to assess a concentration several orders of magnitude above the active concentration of seletalisib against its target in biochemical and cellular systems. This concentration is commonly used in secondary pharmacology studies to understand potential off-target activity (Bowes et al., 2012). Seletalisib was screened against a panel of kinases using the Z’-LYTE FRET and the Adapta Universal Kinase Assays (Thermo Fisher Scientific, Waltham, MA) according to manufacturer’s protocols. In the Z’-LYTE FRET primary reaction, the γ-phosphate of ATP is transferred to a synthetic FRET peptide labeled with coumarin (donor) and fluorescein (acceptor). Subsequently, nonphosphorylated peptide is cleaved by a site-specific protease, disrupting FRET between donor and acceptor, and allowing measurement of the kinase reaction. Kinase activity is measured in Adapta assays by the production of ADP. Following the enzyme reaction, europium-labeled anti-ADP antibody and an Alexa-Fluor 647-labeled ADP tracer are added to the assay well, generating a TR-FRET signal. ADP formed in the kinase reaction will displace the labeled ADP tracer, allowing kinase activity to be determined.
The binding of seletalisib to a range of nonkinase enzymes, receptors and ion channels was assessed at Cerep (Celle L’Evescault, France). In each experiment the reference compound was tested concurrently with seletalisib and the data were compared with historical values determined at Cerep. The assays were undertaken in accordance with Cerep’s standard operating procedures. In brief, binding activity to receptors and ion channels was assessed by competition in radiolabeled ligand binding assays. The enzyme assays used a range of assay technologies depending on the enzyme class and activity, including radiometric and colorimetric assays. Information on specific assays can be found at http://www.cerep.fr/cerep/users/pages/catalog/profiles/catalog.asp.
BioMAP Assays.
The BioMAP assays (Supplemental Table 1) were performed at BioSeek (San Francisco, CA) with standardized protocols and methods described in recent publications (Kunkel et al., 2004a,b; Berg et al., 2010). Seletalisib and UCB1370037 were supplied on dry ice as a 1 mM solution in DMSO and tested in a concentration response (seletalisib) or at a single concentration of 1 μM (UCB1370037) to explore the effects of PI3Kδ-specific inhibition compared with complete inhibition of class I PI3K signaling. In addition, seletalisib was tested in the BioMAP BT cell system at concentrations of 1000, 100, 10, and 1 nM, and UCB1370037 at 1000 nM. An activity profile was generated on the basis of the effect of the compounds on the levels of cellular readouts, including cytokines, growth factors, adhesion molecules, and proliferation endpoints. Activity was considered significant when it fell outside the 99% confidence interval (CI) calculated for the log ratio for each assay and endpoint.
Anti-IgM-Mediated Ramos Phosphorylated AKT Assay.
Ramos cells (LGC Standards, Teddington, UK) were plated at 1 × 105 cells/well in serum-free RPMI 1640 (Life Technologies) in a 96-well plate to which serially diluted seletalisib was added. Anti-IgM (goat anti-human F(ab)2 IgM; Jackson ImmunoResearch Laboratories, West Grove, PA) in serum-free RPMI 1640 was added to all wells, with the exception of the negative control wells, at a final concentration of 20 μg/ml. The plate was incubated (10 minutes, 37°C with shaking), chilled on ice, and the cells pelleted by centrifugation. Cellular phosphorylated protein kinase B (AKT) was detected following the Meso Scale Diagnostics (MSD) pAKT assay protocol (Rockville, MD). Briefly, the cell pellet was lysed (30 minutes) in MSD lysis buffer then stored at –80°C overnight. The lysate was thawed, transferred to the MSD assay plate, and incubated (room temperature, 1 hour). Following washing, the plate was incubated with detection antibody, washed, and read on an MSD plate reader.
fMLP- and PMA-Activated Human Neutrophil Assays.
The enriched neutrophil fraction from human blood was prepared from granulocytes using Ficoll-Paque (GE Healthcare, Little Chalfont, UK). Separation followed by red blood cell lysis (with ammonium chloride) of the granulocyte/erythrocyte pellet. Granulocytes were incubated (37°C, 30 minutes) with diluted seletalisib and further diluted in phosphate-buffered saline (PBS) supplemented as follows: 0.8 μg/ml dihydrocytochalasin B (Sigma-Aldrich), 0.5 IU/ml horseradish peroxidase type IV, 1 mM sodium azide (Sigma-Aldrich), and 25 μM Amplex Red (Invitrogen/Thermo Fisher Scientific) containing either fMLP (20 nM) or PMA (2 nM). Following incubation at 37°C for a further 30 minutes, fluorescence intensity was measured using a Synergy 2 plate reader (Biotek, Swindon, UK) with a 535-nm excitation filter and a 580-nm emission filter. Percent inhibition was calculated on the basis of the minimum signal generated in the absence of stimulus and the maximum signal generated in the presence of fMLP or PMA.
T-Cell Activation Assay.
Section 3.07 PBMCs were prepared from human donor blood using Leucosep tubes (Greiner Bio One, Stonehouse, UK) according to the manufacturer instructions. Briefly, blood diluted 1:1 (v/v) with RPMI 1640 was added to 50-ml Leucosep tubes and centrifuged (room temperature, 1000g, 10 minutes, no brake). The cells were washed with RPMI 1640, spun (250g, 10 minutes), and resuspended in RPMI 1640 [containing l-glutamine 5 mM and 10% fetal bovine serum (FBS)]. Black clear-bottomed plates (Corning Costar, Corning, NY) were coated with 5 μg/ml mouse anti-human CD3 (UCHT1; R&D Systems, Abingdon, UK) in PBS (Life Technologies), incubated overnight at 4°C, and then washed three times in PBS. PBMCs were added to the anti-CD3-coated plates along with dilutions of seletalisib made in RPMI 1640. The plates were incubated (48 hours, 37°C, 5% CO2) and centrifuged (300g, 5 minutes) before 25 μl of supernatant was removed for cytokine analysis using the MSD cytokine multiplex assay (IFN-γ, IL-17, TNF-α; Meso Scale Diagnostics), performed according to the manufacturer. Rat PBMCs were prepared from lithium-heparinized whole blood from healthy Wistar Han male rats (Harlan Laboratories, Indianapolis, IN) using Lympholyte-mammal cell separation media (VH Bio, Gateshead, UK). Briefly, 9 ml blood diluted 1:2 (v/v) with PBS was added to 12 ml Lympholyte-mammal cell separation media in Universal tubes and centrifuged (room temperature, 800g, 20 minutes, no break). The cells were washed with PBS, spun (1200 rpm, 4 minutes), and resuspended in assay media (RPMI 1640 containing l-glutamine 2 mM, 10% FBS, 50 IU/ml penicillin, 50 μg/ml streptomycin, and 0.5 M 2-mercaptoethanol). U-bottomed tissue culture plates (Corning Costar) coated with 1 μg/ml mouse anti-rat CD3 (G4.18) in PBS, were incubated overnight at 4°C, and then washed once in assay media. PBMC were added to the anti-CD3-coated plates along with dilutions of seletalisib made in RPMI 1640. The plates were incubated (48 hours, 37°C, 5% CO2) and centrifuged (300g, 5 minutes) before supernatant was removed for rat TNF-α detection using the R&D Systems DuoSet antibodies (dy510) and MSD plates, according to the manufacturer’s instructions.
House Dust Mite Assay.
PBMCs were prepared (as detailed above) from blood drawn from human donors allergic to house dust mite (HDM). PBMCs (2–4 × 105 cells) were plated in a 96-well round-bottomed plate to which prediluted seletalisib and HDM extract (Dermatophagoides pteronyssinus; Greer Laboratories, Lenoir, NC) at 100 μg/ml was added (to give a final concentration of 25 μg/m). The plates were incubated (6 days, 37°C, 5% CO2) and then centrifuged (300g, 5 minutes). The supernatant was harvested (50 μl each) and IL-5 (Human IL-5 ELISA Assay Kit; R&D Systems) and IL-13 enzyme-linked immunosorbent assays (ELISA) (Human IL-13 ELISA Assay Kit; Life Technologies) were performed according to manufacturer instructions. Percent inhibition was calculated on the basis of the minimum signal generated in the absence of HDM and the maximum signal generated in the presence of HDM.
B-Cell Proliferation.
B-cells were isolated from PBMCs, cryopreserved tonsil suspensions, or buffy coats that had undergone plateletpheresis by CD19-selection (B-cell Isolation Kit II; Miltenyi Biotec, Bisley, UK). Purity of CD19+ cells was determined to be ≥98% by staining with anti-CD19-APC and anti-CD45-FITC antibodies (BD Biosciences, Oxford, UK) and analyzing by flow cytometry. Non-CD19 cells, collected as the APC fraction, were counted, treated with mitomycin C (25 μg/ml, 40 minutes, 37°C), centrifuged (300g, 5 minutes), and washed twice with RPMI 1640 (containing, l-glutamine 5 mM and 10% FBS) to remove all traces of mitomycin C. The cells were then resuspended in RPMI 1640.
CD19+ cells were washed with PBS and resuspended in PBS at room temperature. The same volume of CFSE (Life Technologies) solution was added to the cells (final concentration of CFSE: 0.25 μM). The CD19+ cells were incubated (room temperature, 10 minutes, dark) before at least 10 ml of RPMI 1640 with 20% fetal calf serum was added to inactivate the CFSE. The cells were centrifuged (300g, 5 minutes), the supernatant discarded, and the PBMCs washed (RPMI 1640, l-glutamine 5 mM, 10% FBS). The centrifugation and washing of the PBMCs was repeated, and the cells resuspended in RPMI 1640.
The CD19+ cells were plated in a 96-well U-bottom plate to which the APCs and prediluted seletalisib were added. The plates were incubated (1 hour, 37°C, 5% CO2) and stimulation media was added so that the final culture media was RPMI 1640, l-glutamine 5 mM, 10% FBS, 5 μg/ml anti-IgM, F(ab)2 (Jackson ImmunoResearch Laboratories), 20 ng/ml IL-2, and 20 ng/ml IL-10 (R&D Systems).
The cells were incubated (6 days, 37°C, 5% CO2) and then analyzed by flow cytometry. CFSE was detected in the FL1 channel and B-cells were detected using APC-labeled CD19 antibody. Dilution of the CFSE signal in gated CD19+ cells was detected and used as a measure of proliferation. The effect of seletalisib was expressed as percentage inhibition of the maximal response (DMSO-stimulated control) compared with the unstimulated minimal response.
CpG-Induced IL-6 and IL-10 Cytokine Secretion by Purified B-Cells.
CD19+ B-cells were isolated from PBMCs by negative selection, as described above, and resuspended in RPMI 1640 (containing l-glutamine 5 mM and 10% FBS). In a 96-well U-bottom plate, 1 × 105 CD19+ cells were incubated with serially diluted seletalisib (1 hour, 37°C, 5% CO2) prior to stimulation with 2.5 μg/ml CpG oligodeoxynucleotides (ODN 2006; Invivogen, San Diego, CA). After 48 hours, culture supernatants were harvested and IL-6 and IL-10 levels were determined by human ELISA assays (R&D Systems). The effect of seletalisib was expressed as percentage inhibition of the maximal response (DMSO, CpG-stimulated control) compared with the unstimulated control.
Anti-IgM-Mediated CD69 B-Cell Activation Assay.
Whole blood was collected in sodium heparin vacutainers from human donors. Whole blood aliquots (100 μl) were incubated with anti-human IgM (50 μg/ml; Stratech, Newmarket, UK) to activate B-cells or PBS in a 96-well plate in assay medium (ratio of whole blood to media, 90%) consisting of RPMI 1640 with 10% human AB serum, 50 IU/ml penicillin, 50 μg/ml streptomycin, and 2 mM glutamine. The anti-human IgM-stimulated samples were further incubated for 20 hours at 37°C with PBS, DMSO or fixed concentrations (10–12 at 0.5 log unit increments) of seletalisib over the concentration range of 0.03 nM to 10 μM. For detection of the level of B-cell activation, samples were stained for CD19 (1:300 dilution of anti-human CD19 APC; BD Biosciences) and CD69 (1:300 dilution of anti-human CD69 phycoerythrin (PE); BD Biosciences) and/or the corresponding isotype control antibodies (IgGk1 Isotype Control PE; BD Biosciences) using a 30-minute incubation at 37°C. The isotype control antibodies were used to determine the level of background staining. Red blood cells were then lysed using FACS Lysing Solution (BD Biosciences) and the samples were centrifuged at 600g for 6 minutes and the pellets were resuspended in 100 μl of fluorescence-activated cell sorting (FACS) buffer prior to analysis. The mean fluorescence intensity of CD69 staining on CD19+ cells was determined on BD FACSCanto (BD Biosciences) and data analysis was performed using Flowjo Software.
Anti-IgE-Mediated Basophil Degranulation Assay.
One hundred microliters of whole blood drawn from human donors was added to a 96-well plate at 37°C for 1 hour. Seletalisib prepared in basophil stimulation buffer (BSB; BD Biosciences) was added to the blood and incubated (37°C, 45 minutes). Degranulation was stimulated with the addition of 2 μg/ml anti-IgE (Dako, Glostrup, Denmark) diluted in BSB to the plate. The plate was incubated (37°C, 12 minutes) and the degranulation stopped by placing the plate on ice (10 minutes). Basophil activation was detected using flow cytometry and an antibody cocktail (CD123/HLA-DR/CD63; BD Biosciences) or single stains (human blood: CD123 clone 9F5; BD Biosciences) for compensation controls. Antibodies were added, the blood mixed, and the 96-well plates then incubated on ice (45 minutes, dark). The whole blood samples were lysed with FACS Lysing Solution (BD Biosciences) and centrifuged (1000g, 5 minutes). Supernatant was aspirated with the exception of the last 300–400 μl that contained cells, and samples were analyzed on a BD FACSCanto flow cytometer (BD Biosciences). At least 500 basophil events were acquired per sample. Cells with a side scatter low (SSClo), CD123+ phenotype were gated. Basophils were identified as the SSClo, CD123+ HLA-DR– population. The cells that were CD123+ and HLA-DR– were further analyzed for their CD63 cell surface expression as a measure of degranulation. The BSB control was used to set the baseline CD63 level and the anti-IgE control was used to establish the maximal response.
Anti-CD3 Antibody-Induced IL-2 Release in the Lewis Rat.
Adult Lewis rats (male, 6–8 weeks of age) were used in accordance with the Animals (Scientific Procedures) Act 1986 and were purchased from Charles River (Margate, UK). Animals were kept under a light/dark cycle of 12/12 hours and had access to food and water ad libitum.
Rats were dosed with seletalisib (0.1–10 mg/kg in 500 μl volume) or vehicle via oral gavage (75 mm, 16-guage curved stainless steel cannula; Vet-Tech Solutions, Cheshire, UK) 30 minutes prior to intravenous administration (16 mm 26-gauge needle; Becton Dickinson Medical, Oxford, UK) of anti-CD3 antibody (100 μg/kg; BD Biosciences) administered in a 200-μl dose volume. The vehicle was methylcellulose (0.5%, 400 cps; Sigma Aldrich, Gillingham, Dorset, UK) or saline (Polyfusor; Fresenius Kabi Ltd, Runcorn, UK) for oral and intravenous administration, respectively. To assess basal IL-2 levels, a group of animals received vehicle both orally and intravenously. Animals were anesthetized with isoflurane (Henry Schein Animal Health, Dumfries, UK) for the intravenous administration. At 90 minutes post-anti-CD3 antibody, the rats were anesthetized and blood was removed by terminal cardiac puncture into EDTA or heparin-coated tubes (Multivette 600 μl lithium heparin or Multivette 600 μl K3 EDTA; Sarstedt AG & Co., Nümbrecht, Germany) and then killed by cervical dislocation. Two samples were prepared from the blood, enabling pharmacokinetic analysis of seletalisib levels and IL-2 levels.
IL-2 levels were measured from plasma using the Rat IL-2 Quantikine ELISA Kit (R&D Systems) per manufacturer instructions and with the plasma samples being diluted 1:2 prior to analysis. Dunnett’s multiple comparison tests (GraphPad Prism version 6.01) were used to test for statistical difference between the drug treatment groups and the positive control group.
For the bioanalysis of seletalisib, blood samples were diluted 50:50 with water and measured against calibration standard prepared in matrix akin to study samples. Samples and calibration standards were precipitated in acetonitrile containing internal standard (100 ng/ml), mixed, and centrifuged at 4000 rpm for 10 minutes. The supernatant was reconstituted in 20% acetonitrile and injected into a liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) system. LC-MS/MS analysis was carried out by Agilent 1100 series binary pump (Agilent Technologies, Stockport, UK) coupled to electrospray ionization–triple quadruple mass spectrometer (Micromass Quattro Utima; Waters Ltd, Elstree, UK). The extracted sample was injected by CTC Analytics HTS Pal autosampler (Presearch, Hitchin, UK) onto a reversed-phase high-performance liquid chromatography column (Phenomenex Luna 5-μm C18 100Å LC column, 50 × 2.0 mm; Phenomenex, Macclesfield, UK) maintained at 40°C. A linear gradient of 5–95% acetonitrile in 0.1% formic acid was applied for 1.5 minutes at a flow rate of 1 ml/min with a split ratio of 5:1 to the mass spectrometer. The mass spectrometer was set up to run multiple reaction monitoring analysis to detect multiple transitions at a dwell time of 100 milliseconds per transition.
Results
Seletalisib Is a Potent, Selective, and ATP-Competitive Inhibitor of PI3Kδ
FRET-based protein kinase activity assays showed that seletalisib is a potent and selective inhibitor of the PI3Kδ isoform. The potency and selectivity of seletalisib against the four PI3K isoforms generated are shown in Table 1. Seletalisib had an IC50 value for PI3Kδ of 12 nM [geometric mean, 95% CI, 8.4–17.1 (n = 10)] and showed significant selectivity to PI3Kδ with respect to the other class I PI3K isoforms (between 24- and 303-fold). The pan-PI3K inhibitor UCB1370037 showed particularly poor selectivity against the ubiquitously expressed PI3Kα isoform. Increasing the concentration of ATP in the PI3Kδ TR-FRET assay resulted in a reduction in the potency of seletalisib, indicating that its inhibitory activity is competitive with respect to ATP (Fig. 2).
In vitro receptor binding and enzyme assays across a broad range of target classes showed that seletalisib is selective for PI3Kδ (Supplemental Table 2). From 239 kinases screened, seletalisib at a concentration of 10 μM showed no inhibitory activity greater than 47% (MAP4K4) against non-PI3K kinase enzymes (Supplemental Table 2A). Against nonkinase enzymes, seletalisib showed weak activities against phosphodiesterase (PDE)3A, PDE2A1, and PDE4D2, with inhibition varying between 32 and 74% at 10 μM (Supplemental Table 2B). When screened at a concentration of 10 μM against 55 receptors and ion channels, the highest inhibitory activity of seletalisib observed was 20% (Supplemental Table 2C). One receptor, neuropeptide Y receptor (Y1) showed 54% activation. In vitro receptor binding and enzyme assays across a broad range of target classes showed that seletalisib is selective for PI3Kδ.
Seletalisib Potently Inhibits Signaling through the B-Cell Receptor.
PI3Kδ has been shown to be a requirement for signaling downstream of the BCR (Al-Alwan et al., 2007). Activation of this receptor results in PI3Kδ-dependent phosphorylation of AKT at serine 473 (Bilancio et al., 2006). The ability of seletalisib to inhibit AKT phosphorylation on B-cell ligation with anti-IgM crosslinking was determined in Ramos cells (a human B-cell line derived from a patient with Burkitt’s lymphoma). Seletalisib potently inhibited the phosphorylation of AKT following anti-IgM stimulation of the BCR on Ramos cells with an IC50 of 15 nM [geometric mean, 95% CI, 9.3–23.5 (n = 6)] (Fig. 3A).
Seletalisib Inhibits fMLP-Stimulated Human Neutrophil Superoxide Release Assay.
PI3Kδ has been shown to be a requirement for superoxide generation by neutrophils, cells of the innate immune system, stimulated with the bacterial peptide ligand fMLP (Condliffe et al., 2005). Generation of superoxide in human neutrophils stimulated with an alternative ligand, PMA, is signaled through protein kinase C and is not dependent on PI3Kδ. When the activity of seletalisib was tested in fMLP-stimulated and in PMA-stimulated human neutrophil superoxide release assays, the fMLP-stimulated superoxide was inhibited by seletalisib with an IC50 of 16 nM (Fig. 3B; geometric mean, 95% CI, 11–24 [n = 20]). The PMA-activated superoxide release assay was not inhibited by seletalisib.
BioMAP Profiling Illustrates the Selective PI3Kδ Inhibition by Seletalisib and Characterizes Specific Effects on B- and T-Cell Activation.
The BioMAP profile for seletalisib is shown and compared with a pan-PI3K inhibitor, UCB1370037 (Fig. 3C and Supplemental Table 1). Seletalisib showed little or no activity in the PBMC cytotoxicity assays (SRB) but was shown to have selective effects on T-cell-dependent B-cell activation readouts, including B-cell proliferation and antibody class-switching. Furthermore, seletalisib inhibited the release of cytokines involved in T-cell activation responses through TCR stimulation. Seletalisib was shown to have a comparatively selective BioMAP profile with activities confined to the SAg cell system (PBMC+endothelial cells stimulated via the TCR) and, most profoundly, in the BT coculture cell system [B-cells+PBMC stimulated with anti-IgM + staphylococcal enterotoxin B (SEB) + toxic shock syndrome toxin (TSST)]. In the BT cell system, seletalisib was shown to strongly inhibit B-cell proliferation and cytokine production (IL-2, TNF-α, IL17A, and IL17F) and also inhibit the production of secreted IgG. In addition, seletalisib was also shown to inhibit E-selectin, IL-8, and proliferation responses in the BioMap SAg cell system. Seletalisib had a selective effect on these specific immune cell BioMAP systems and had no significant effects on other systems. Conversely, in addition to effects on immune cell systems, UCB1370037 also modulated responses in systems containing endothelial cells (3C and 4H), epithelial cells (BE3C), smooth muscle cells (CASM3C), fibroblasts (HDF3CGF), and myofibroblasts (MyoF).
Seletalisib Shows Anti-Inflammatory Activity in Cellular Assays of Adaptive Immunity
Seletalisib Inhibits T-Cell Cytokine Production.
PI3Kδ activity is required for T-cell differentiation and function, as demonstrated in studies of genetic ablation of PI3Kδ in mouse cells and PI3Kδ inhibitors in human cells (Okkenhaug et al., 2006; Soond et al., 2010). The activity of seletalisib was characterized in primary T-cell assays. Secretion of inflammatory cytokines was measured 48 hours post-anti-CD3 stimulation of PBMCs in the presence and absence of seletalisib. The compound inhibited IFNγ, IL-17, and TNF-α secretion from PBMCs with IC50 values of 54, 21, and 31 nM, respectively (5–6 donors; Fig. 4A). As seletalisib was to be assessed in an in vivo rat model, we wished to characterize the functional potency of seletalisib in a rat cellular assay similar to those established in human systems. Activation of isolated rat PBMCs with anti-CD3 resulted in the release of TNF-α, which was inhibited by seletalisib with a potency of 22 nM (n = 2), similar to the potency of seletalisib in the equivalent anti-CD3-stimulated TNF-α release from human PBMCs (31 nM). Seletalisib also inhibited HDM-induced IL-5 and IL-13 secretion from HDM-allergic human donor PBMCs (mean IC50 values of 15 and 2 nM, respectively, 2 donors; Fig. 4B).
Seletalisib Inhibits B-Cell Function.
The activity of seletalisib was assessed in a functional B-cell assay by stimulating human PBMCs with anti-IgM and measuring the proliferation of B-cells. In this system, seletalisib inhibited B-cell proliferation (IC50 16 nM, range 10–19 nM, n = 3 donors) (Fig. 4C). Furthermore, seletalisib inhibited IL-6 and IL-10 release in human B-cells stimulated with CpG ODN 2006, an agonist that signals selectively via TLR9 (Dil and Marshall, 2009), with IC50 values of 57 nM (95% CI, 33–101 nM, n = 4 donors) and 18.1 nM (95% CI, 10–34 nM, n = 3 donors), respectively (Fig. 4D).
Seletalisib Shows Anti-Inflammatory Activity in Whole Blood Assays
In addition to assays on isolated human cells, the activity of seletalisib was evaluated in a more physiologic environment, human whole blood. In whole blood activation assays, seletalisib inhibited anti-IgM-induced CD69 expression on B-cells in a concentration-dependent manner across five donors (Fig. 5A). An IC50 value of 57 nM (geometric mean, 95% CI, 22–143 [n = 5]) was obtained.
Basophils are circulating granulocytes that play a role in allergic hypersensitivity reactions. FcεR1/IgE-induced degranulation has been shown to be PI3Kδ dependent and therefore provided a mechanistic readout of the activity of seletalisib (Ali et al., 2004, 2008). In a degranulation assay developed with human whole blood, seletalisib blocked anti-IgE-triggered CD63 cell surface expression, a marker of basophil degranulation. This was achieved with an IC50 of 35 nM (geometric mean, 95% CI, 27–45 [n = 13]) (Fig. 5B).
Seletalisib Shows In Vivo Activity in a Rodent Model of Cytokine Release following T-Cell Activation through TCR Stimulation
Activation of T cells through TCR stimulation following anti-CD3 administration results in the release of cytokines, of which IL-2 is one of the earliest cytokines released. Seletalisib significantly inhibited IL-2 release following TCR stimulation in the rat. The inhibition was observed at all tested doses of seletalisib with almost complete inhibition reached at dose levels ≥1 mg/kg. A representative experiment is shown in Fig. 6A. Analysis of the relationship between inhibition of IL-2 release and seletalisib blood concentration, using combined data across experiments, demonstrated that seletalisib has potent in vivo effects with an estimated IC50 value of <10 nM (Fig. 6B).
Discussion
Seletalisib is a novel PI3Kδ inhibitor with a high degree of biochemical selectivity to PI3Kδ compared with the other class I PI3K enzymes, and to non-PI3K kinases, nonkinase enzymes, and a panel of receptors and ion channels. Compared with data from published sources for idelalisib [IC50 values of 2.5, 89, 565, and 820 nM for PI3Kδ, α, β, and γ, respectively (Lannutti et al., 2011)] and duvelisib [IC50 values of 2.5, 27, 85, 1602 nM for PI3Kδ, α, β, and γ, respectively (Winkler et al., 2013)], seletalisib is in a range of selectivity similar to the other class I PI3K enzymes but with greater selectivity toward the PI3Kγ and β isoforms compared with duvelisib (Table 1). Seletalisib demonstrates good cellular activity, potently inhibiting phosphorylation of AKT following activation of the BCR in a B-cell line (Fig. 3A). In addition to its role downstream of tyrosine kinase receptors, it has emerged in recent years that PI3Kδ is also able to signal downstream of GPCRs (Okkenhaug, 2013b). It has been demonstrated previously that fMLP is able to activate neutrophils in a PI3Kδ-dependent manner through the GPCR formyl peptide receptor 1 (Condliffe et al., 2005). Seletalisib potently and fully inhibited fMLP-activated superoxide release from human neutrophils. To demonstrate that this effect was not through inhibition of other targets on the superoxide pathway, superoxide release was activated by PMA, which was not inhibited by seletalisib.
When profiled across a broad range of human primary cell systems (BioMAP), seletalisib showed activity in systems restricted to the adaptive immune system, consistent with its PI3Kδ-selective specificity. In the PBMC-containing BioMAP assays, seletalisib showed inhibition of proliferation and cytokine and IgG release with little or no signs of cytotoxicity. The lack of activity of seletalisib in non–leukocyte-containing BioMAP systems contrasted with the activities obtained with the pan-PI3K isoform inhibitor UCB1370037, which showed activity in all of the systems tested, reflecting the critical cellular roles of the ubiquitously expressed PI3Kα and β isoforms. The inhibition of fMLP-induced neutrophil superoxide release by seletalisib contrasted with lack of activity observed in macrophage readouts in the BioMAP profile. These data indicate that PI3Kδ plays a restricted role in cells of the innate immune system and its relevance in these cell types to immune disease awaits elucidation. The selective functional activity of seletalisib in the broad range of cellular assays in the BioMAP profile is in accordance with our current understanding of the role of PI3Kδ from genetic and pharmacological studies (Stark et al., 2015). Together with the biochemical data, these findings indicate that seletalisib is a potent and selective inhibitor of PI3Kδ activity.
Studies using genetic ablation of PI3Kδ in mouse cells have shown that PI3Kδ activity is required for the differentiation and function of T-cells (Okkenhaug et al., 2006). These findings have been supported by the use of PI3Kδ-specific inhibitors, such as IC87114 (Soond et al., 2010). Our findings show that seletalisib inhibits secretion of cytokines (IFNγ, IL-17, and TNF-α) from human-donor PBMCs following TCR activation by crosslinking of the receptor. As an in vivo rat model was used to assess the systemic action of seletalisib, the functional activity of the compound was determined in rat PBMCs by measuring TNF-α levels following activation of the TCR. Seletalisib showed similar potency in the rat and human TNF-α release assays (22 and 31 nM, respectively), supporting the use of rat as the species for an in vivo mechanistic study. Furthermore, antigen-specific activation of T-cells from HDM-allergic donors, measured by release of IL-5 and IL-13, was inhibited by seletalisib. The higher potency of seletalisib in the HDM-activated system compared with the anti-CD3 crosslinking system may reflect the more selective activation of T-cells caused by HDM allergen compared with anti-CD3. T-helper (Th) cells types 1 and 17 have been implicated in the pathogenesis of autoimmune diseases and chronic inflammatory disorders, and Th2 is acknowledged to have a pivotal role in human allergies [as reviewed by Cosmi et al. (2014)]. In a T cell-driven cytokine release model in the rat, seletalisib potently inhibited IL-2 release, demonstrating its activity in an in vivo setting. These data are consistent with a PI3Kδ-selective mechanism in accordance with previous findings in both mouse and human systems using genetic and pharmacological approaches (Han et al., 2012).
PI3Kδ plays a central role in B-cell differentiation, expansion, and function. It is recruited upon activation of the BCR and the coreceptor CD19 and is required for full functioning of activating receptors such as CD40 and Toll-like receptors (TLRs), and cytokine receptors for B-cell activation factor and IL-6, for example (Limon and Fruman, 2012). Anti-IgM-driven cultures showed that seletalisib inhibits B-cell proliferation with a potency similar to the Ramos pAKT assay (26 and 15 nM, respectively). CpG signals through TLR9 and the potent activity of seletalisib in the CpG-activated cytokine release assay confirms the importance of PI3Kδ in TLR signaling. Interestingly, the inhibition of cytokine release in this case is incomplete, indicating that TLR9 signaling is not solely dependent on PI3Kδ. B-cells have been shown to proliferate in the absence of PI3Kδ, so despite this central role, there are compensating mechanisms that can be activated under certain circumstances (Fruman et al., 1999; Omori et al., 2006). To assess seletalisib activity in a more physiologic environment, a whole blood assay using CD69 as a marker of B-cell activation was developed. The potency of the compound in the assay (57 nM) was similar to that achieved in the other in vitro assays carried out in which only fetal calf serum is present to provide a nonspecific binding component. There was sufficient unbound compound in human blood to allow potent inhibition of B-cell activation.
The ability of seletalisib to act in the whole blood compartment was further demonstrated by inhibition of basophil degranulation measured via CD63 expression. Degranulation was triggered by anti-IgE-mediated crosslinking of FcεR1, which requires PI3Kδ to signal (Okkenhaug et al., 2007). Seletalisib showed potent and complete inhibition of this system consistent with the absolute requirement for PI3Kδ for FcεR1-activated degranulation.
The selective inhibition of PI3Kδ by seletalisib infers a potent effect specifically directed at the adaptive immune system by virtue of PI3Kδ’s receptor coupling (Hawkins and Stephens, 2015). In particular, inhibition of PI3Kδ by seletalisib could provide benefit to patients with autoimmune diseases that are driven by a dysregulated proinflammatory cytokine secretion, including rheumatoid arthritis, systematic lupus erythematosus, type-1 diabetes, psoriasis, and multiple sclerosis (Haylock-Jacobs et al., 2011; Puri and Gold, 2012). The recent discovery of activating mutations in the p110δ catalytic domain and in the p85α regulatory subunit in a group of primary immune deficiency patients provides a further rationale for the therapeutic use of PI3Kδ inhibitors (Angulo et al., 2013; Deau et al., 2014; Lucas et al., 2014; Hawkins and Stephens, 2015). The high degree of selectivity achieved with seletalisib may allow sparing of the other PI3K isoforms, at the same time fully blocking PI3Kδ activity, thereby avoiding potentially undesirable effects on immune cell functions unrelated to disease pathology.
Seletalisib has undergone clinical testing in healthy donors to assess its tolerability and pharmacokinetic and pharmacodynamic profile in healthy volunteers and patients with psoriasis (NCT02303509). It is currently being tested in patients with primary Sjögren’s syndrome (NCT02610543). The utility of isoform-selective PI3K inhibitors for patients with autoimmune and inflammatory disease will emerge in the coming years as more data from clinical studies become available.
In conclusion, seletalisib is a new chemical entity that is a potent and selective inhibitor of PI3Kδ with potential for the treatment of a range of immune-inflammatory and immunodeficiency diseases. The effects of seletalisib in more complex pharmacological disease models will be reported in future publications.
Acknowledgments
We thank Eugene Healy (University of Southampton) for his critical reading of the manuscript. Jen Timoshanko (UCB Pharma) provided editorial and publication support. Medical writing and editorial support was provided by Mark O’Connor, Stephen Paterson, and Emma Donadieu from, and on behalf of, iMed Comms, an Ashfield Company, part of UDG Healthcare plc.
Authorship Contributions
Conducted experiments: Allen, Brookings, Crosby, Cutler, Davies, McCluskey, Merriman, Powell, Shuttleworth, Tewari, Twomey, Watt, Payne.
Performed data analysis: Allen, Powell, Crosby, Cutler, Delgado, Fahy, Silva, Healy, Davies, McCluskey, Merriman, Kotian, Tewari, Watt, Payne.
Wrote or contributed to the writing of the manuscript: Allen, Brookings, Crosby, Cutler, Davies, Delgado, Fahy, Healy, Kotian, McCluskey, Merriman, Payne, Powell, Shuttleworth, Silva, Tewari, Twomey, Watt.
Footnotes
- Received August 22, 2016.
- Accepted March 21, 2017.
↵1 Current affiliation: Immunocore Limited, Abingdon, UK.
↵2 Current affiliation: Beckman Coulter, High Wycombe, UK.
↵3 Current affiliation: Aurelia Bioscience, Nottingham, UK
The studies were funded by UCB Pharma. Medical writing and editorial support was funded by UCB Pharma. All authors are current or former employees of UCB Pharma and may hold stock options in UCB Pharma.
This article has supplemental material available at jpet.aspetjournals.org.
Abbreviations
- APC
- allophycocyanin
- BCR
- B-cell receptor
- BSB
- basophil stimulation buffer
- CFSE
- carboxyfluorescein N-succinimidyl ester
- CI
- confidence interval
- DMSO
- dimethyl sulfoxide
- ELISA
- enzyme-linked immunosorbent assay
- FBS
- fetal bovine serum
- FcεR1
- Fc epsilon receptor 1
- fMLP
- N-formyl-peptides
- FRET
- fluorescence resonance energy transfer
- GPCR
- G protein-coupled receptors
- HDM
- house dust mite
- MSD
- Meso Scale Diagnostics
- PBMC
- peripheral blood mononuclear cells
- PBS
- phosphate-buffered saline
- PI3K
- phosphoinositide 3-kinases
- PMA
- phorbol myristate acetate
- Copyright © 2017 by The American Society for Pharmacology and Experimental Therapeutics