Suzanne Tay

Discovery of Novel PI3-Kinase δ Specific Inhibitors for the Treatment of Rheumatoid Arthritis: Taming CYP3A4 Time-Dependent Inhibition

PI3Kδ is a lipid kinase and a member of a larger family of enzymes, PI3K class IA(α, β, δ) and IB (γ), which catalyze the phosphorylation of PIP2 to PIP3. PI3Kδ is mainly expressed in leukocytes, where it plays a critical, nonredundant role in B cell receptor mediated signaling and provides an attractive opportunity to treat diseases where B cell activity is essential, e.g., rheumatoid arthritis. We report the discovery of novel, potent, and selective PI3Kδ inhibitors and describe a structural hypothesis for isoform (α, β, γ) selectivity gained from interactions in the affinity pocket. The critical component of our initial pharmacophore for isoform selectivity was strongly associated with CYP3A4 time-dependent inhibition (TDI). We describe a variety of strategies and methods for monitoring and attenuating TDI. Ultimately, a structure-based design approach was employed to identify a suitable structural replacement for further optimization.

A Physiologically Based Pharmacokinetic Modeling Approach to Predict Disease–Drug Interactions: Suppression of CYP3A by IL‐6

Elevated cytokine levels are known to downregulate expression and suppress activity of cytochrome P450 enzymes (CYPs). Cytokine‐modulating therapeutic proteins (TPs) used in the treatment of inflammation or infection could reverse suppression, manifesting as TP‐drug–drug interactions (TP‐DDIs). A physiologically based pharmacokinetic model was used to quantitatively predict the impact of interleukin‐6 (IL‐6) on sensitive CYP3A4 substrates. Elevated simvastatin area under the plasma concentration–time curve (AUC) in virtual rheumatoid arthritis (RA) patients, following 100 pg/ml of IL‐6, was comparable to observed clinical data (59 vs. 58%). In virtual bone marrow transplant (BMT) patients, 500 pg/ml of IL‐6 resulted in an increase in cyclosporine AUC, also in good agreement with the observed data (45 vs. 39%). In a different group of BMT patients treated with cyclosporine, the magnitude of interaction with IL‐6 was underpredicted by threefold. The complexity of TP‐DDIs highlights underlying pathophysiological factors to be considered, but these simulations provide valuable first steps toward predicting TP‐DDI risk.

Discovery of GluN2A-Selective NMDA Receptor Positive Allosteric Modulators (PAMs): Tuning Deactivation Kinetics via Structure-Based Design.

The N-methyl-D-aspartate receptor (NMDAR) is a Na(+) and Ca(2+) permeable ionotropic glutamate receptor that is activated by the coagonists glycine and glutamate. NMDARs are critical to synaptic signaling and plasticity, and their dysfunction has been implicated in a number of neurological disorders, including schizophrenia, depression, and Alzheimers disease. Herein we describe the discovery of potent GluN2A-selective NMDAR positive allosteric modulators (PAMs) starting from a high-throughput screening hit. Using structure-based design, we sought to increase potency at the GluN2A subtype, while improving selectivity against related α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors (AMPARs). The structure-activity relationship of channel deactivation kinetics was studied using a combination of electrophysiology and protein crystallography. Effective incorporation of these strategies resulted in the discovery of GNE-0723 (46), a highly potent and brain penetrant GluN2A-selective NMDAR PAM suitable for in vivo characterization.

Prediction of Drug-Drug Interactions Arising from CYP3A induction Using a Physiologically Based Dynamic Model

Using physiologically based pharmacokinetic modeling, we predicted the magnitude of drug-drug interactions (DDIs) for studies with rifampicin and seven CYP3A4 probe substrates administered i.v. (10 studies) or orally (19 studies). The results showed a tendency to underpredict the DDI magnitude when the victim drug was administered orally. Possible sources of inaccuracy were investigated systematically to determine the most appropriate model refinement. When the maximal fold induction (Indmax) for rifampicin was increased (from 8 to 16) in both the liver and the gut, or when the Indmax was increased in the gut but not in liver, there was a decrease in bias and increased precision compared with the base model (Indmax = 8) [geometric mean fold error (GMFE) 2.12 vs. 1.48 and 1.77, respectively]. Induction parameters (mRNA and activity), determined for rifampicin, carbamazepine, phenytoin, and phenobarbital in hepatocytes from four donors, were then used to evaluate use of the refined rifampicin model for calibration. Calibration of mRNA and activity data for other inducers using the refined rifampicin model led to more accurate DDI predictions compared with the initial model (activity GMFE 1.49 vs. 1.68; mRNA GMFE 1.35 vs. 1.46), suggesting that robust in vivo reference values can be used to overcome interdonor and laboratory-to-laboratory variability. Use of uncalibrated data also performed well (GMFE 1.39 and 1.44 for activity and mRNA). As a result of experimental variability (i.e., in donors and protocols), it is prudent to fully characterize in vitro induction with prototypical inducers to give an understanding of how that particular system extrapolates to the in vivo situation when using an uncalibrated approach.

Discovery and Biological Profiling of Potent and Selective mTOR Inhibitor GDC-0349.

Aberrant activation of the PI3K-Akt-mTOR signaling pathway has been observed in human tumors and tumor cell lines, indicating that these protein kinases may be attractive therapeutic targets for treating cancer. Optimization of advanced lead 1 culminated in the discovery of clinical development candidate 8h, GDC-0349, a potent and selective ATP-competitive inhibitor of mTOR. GDC-0349 demonstrates pathway modulation and dose-dependent efficacy in mouse xenograft cancer models.

Evaluation of Time-Dependent Cytochrome P450 Inhibition in a High-Throughput, Automated Assay: Introducing a Novel Area Under the Curve Shift Approach

Early in the drug discovery process, the identification of cytochrome P450 (CYP) time-dependent inhibition (TDI) is an important step for compound optimization. Here we describe a high-throughput, automated method for the evaluation of TDI utilizing human liver microsomes and conventional CYP-specific mass spectrometer-based probes in a 384-well format. One of the key differences from other published TDI assays is the use of a shift in area the under curve of the percent activity remaining versus inhibitor concentration plot (AUC shift) rather than the traditional fold-shift in IC50, to determine the magnitude of TDI. An AUC shift of < 15% suggests negative TDI and > 15% suggests potential TDI. This AUC shift was used to achieve quantitative data reporting, even in the case of weak inhibitors for which IC50 values cannot be quantified. An Agilent Technologies BioCel 1200 System was programmed such that the TDI liability of up to 77 test compounds, incubated at four test concentrations, with and without NADPH in the pre-incubation, can be analyzed in a single run. The detailed automated methodology, assay validation, data reporting and the novel TDI AUC shift approach to describe magnitude of TDI are presented.

Evaluation of Time Dependent Inhibition Assays for Marketed Oncology Drugs: Comparison of Human Hepatocytes and Liver Microsomes in the Presence and Absence of Human Plasma

PurposeTo evaluate an alternative in vitro system which can provide more quantitatively accurate drug drug interaction (DDI) prediction for 10 protein kinase inhibitors for which DDI risk was over-predicted by inhibition data generated in human liver microsomes (HLM).MethodsThree cryopreserved human hepatocyte (hHEP) systems: 1) plated hHEPs; 2) hHEPs suspended in Dulbecco’s Modified Eagle Medium (DMEM) and 3) hHEPs suspended in human plasma (plasma hHEPs) were developed to detect CYP3A time dependent inhibition, and the static mechanistic model was used to predict clinical outcomes.ResultsA general trend was observed in the CYP3A inactivation potency (kinact/KI, app) as HLM > plated > DMEM ≥ plasma hHEPs. Using the static mechanistic model, DDIs predicted using parameters estimated from plated, DMEM and plasma hHEPs had 84, 74 and 95% accuracy (out of 19 clinical interaction studies) within 2-fold of the reported interaction, respectively. They demonstrated significant improvement compared to the DDIs predicted using parameters estimated from HLMs where 58% accuracy was obtained.ConclusionsBased on 19 DDIs, plasma hHEPs demonstrate a more reliable clinical DDI prediction for 10 protein kinase inhibitors and prototypical CYP3A time dependent inhibitors.

Discovery of GDC-0853: A Potent, Selective, and Noncovalent Bruton’s Tyrosine Kinase Inhibitor in Early Clinical Development

Brutons tyrosine kinase (Btk) is a nonreceptor cytoplasmic tyrosine kinase involved in B-cell and myeloid cell activation, downstream of B-cell and Fcγ receptors, respectively. Preclinical studies have indicated that inhibition of Btk activity might offer a potential therapy in autoimmune diseases such as rheumatoid arthritis and systemic lupus erythematosus. Here we disclose the discovery and preclinical characterization of a potent, selective, and noncovalent Btk inhibitor currently in clinical development. GDC-0853 (29) suppresses B cell- and myeloid cell-mediated components of disease and demonstrates dose-dependent activity in an in vivo rat model of inflammatory arthritis. It demonstrates highly favorable safety, pharmacokinetic (PK), and pharmacodynamic (PD) profiles in preclinical and Phase 2 studies ongoing in patients with rheumatoid arthritis, lupus, and chronic spontaneous urticaria. On the basis of its potency, selectivity, long target residence time, and noncovalent mode of inhibition, 29 has the potential to be a best-in-class Btk inhibitor for a wide range of immunological indications.

GluN2A-Selective Pyridopyrimidinone Series of NMDAR Positive Allosteric Modulators with an Improved in Vivo Profile

The N-methyl-d-aspartate receptor (NMDAR) is an ionotropic glutamate receptor, gated by the endogenous coagonists glutamate and glycine, permeable to Ca2+ and Na+. NMDAR dysfunction is associated with numerous neurological and psychiatric disorders, including schizophrenia, depression, and Alzheimers disease. Recently, we have disclosed GNE-0723 (1), a GluN2A subunit-selective and brain-penetrant positive allosteric modulator (PAM) of NMDARs. This work highlights the discovery of a related pyridopyrimidinone core with distinct structure-activity relationships, despite the structural similarity to GNE-0723. GNE-5729 (13), a pyridopyrimidinone-based NMDAR PAM, was identified with both an improved pharmacokinetic profile and increased selectivity against AMPARs. We also include X-ray structure analysis and modeling to propose hypotheses for the activity and selectivity differences.

Probing Mechanisms of CYP3A Time-Dependent Inhibition Using a Truncated Model System

Time-dependent inhibition (TDI) of cytochrome P450 (CYP) enzymes may incur serious undesirable drug-drug interactions and in rare cases drug-induced idiosyncratic toxicity. The reactive metabolites are often generated through multiple sequential biotransformations and form adducts with CYP enzymes to inactivate their function. The complexity of these processes makes addressing TDI liability very challenging. Strategies to mitigate TDI are therefore highly valuable in discovering safe therapies to benefit patients. In this Letter, we disclose our simplified approach toward addressing CYP3A TDI liabilities, guided by metabolic mechanism hypotheses. By adding a methyl group onto the α carbon of a basic amine, TDI activities of both the truncated and full molecules (7a and 11) were completely eliminated. We propose that truncated molecules, albeit with caveats, may be used as surrogates for full molecules to investigate TDI.