Preclinical Safety Assessment of a Highly Selective and Potent Dual Small-Molecule Inhibitor of CBP/P300 in Rats and Dogs.

Cyclic adenosine monophosphate-response element (CREB)-binding protein (CBP) and EP300E1A-binding protein (p300) are members of the bromodomain and extraterminal motif (BET) family. These highly homologous proteins have a key role in modulating transcription, including altering the status of chromatin or through interactions with or posttranslational modifications of transcription factors. As CBP and p300 have known roles for stimulating c-Myc oncogenic activity, a small-molecule inhibitor, GNE-781, was developed to selectively and potently inhibit the CBP/p300 bromodomains (BRDs). Genetic models have been challenging to develop due to embryonic lethality arising from germline homozygous mutations in either CBP or P300. Hence, the purpose of this study was to characterize the role of dual inhibition of these proteins in adult rats and dogs. Repeat dose toxicity studies were conducted, and toxicologic and pathologic end points were assessed. GNE-781 was generally tolerated; however, marked effects on thrombopoiesis occurred in both species. Evidence of inhibition of erythroid, granulocytic, and lymphoid cell differentiation was also present, as well as deleterious changes in gastrointestinal and reproductive tissues. These findings are consistent with many preclinical (and clinical) effects reported with BET inhibitors targeting BRD proteins; thus, the current study findings indicate a likely important role for CBP/p300 in stem cell differentiation.

Assessment of OATP transporter-mediated drug-drug interaction using physiologically-based pharmacokinetic (PBPK) modeling - a case example.

GDC-0810 was under development as an oral anti-cancer drug for the treatment of estrogen receptor-positive breast cancer as a single agent or in combination. In vitro data indicated that GDC-0810 is a potent inhibitor of OATP1B1/1B3. To assess clinical risk, a PBPK model was developed to predict the transporter drug-drug interaction (tDDI) between GDC-0810 and pravastatin in human. The PBPK model was constructed in Simcyp® by integrating in vitro and in vivo data for GDC-0810. The prediction of human pharmacokinetics (PK) was verified using GDC-0810 phase I clinical PK data. The Simcyp transporter DDI model was verified using known OATP1B1/1B3 inhibitors (rifampicin, cyclosporine and gemfibrozil) and substrate (pravastatin), prior to using the model to predict GDC-0810 tDDI. The effect of GDC-0810 on pravastatin PK was then predicted based on the proposed clinical scenarios. Sensitivity analysis was conducted on the parameters with uncertainty. The developed PBPK model described the PK profile of GDC-0810 reasonably well. In the tDDI verification, the model reasonably predicted pravastatin tDDI caused by rifampicin and gemfibrozil OATP1B1/3 inhibition but under-predicted tDDI caused by cyclosporine. The effect of GDC-0810 on pravastatin PK was predicted to be low to moderate (pravastatin Cmax ratios 1.01-2.05 and AUC ratio 1.04-2.23). The observed tDDI (Cmax ratio 1.20 and AUC ratio 1.41) was within the range of the predicted values. This work demonstrates an approach using a PBPK model to prospectively assess tDDI caused by a new chemical entity as an OATP1B1/3 uptake transporter inhibitor to assess clinical risk and to support development strategy.

Assessment of the hepatic CYP reductase null mouse model and its potential application in drug discovery.

CYP Oxidoreductase (Por) is the essential electron donor for all CYP enzymes and is responsible for the activation of CYP. The Taconic Hepatic CYP Reductase Null (HRN) mouse model possesses a targeted mutation that results in liver-specific deletion of the Por gene thereby resulting in a disruption of CYP metabolism in the liver. The objectives of these studies were to further characterize the HRN mouse using probe drugs metabolized by CYP. In addition, tumor exposure in xenograft tumor bearing HRN immune-compromised (nude) mice was also determined. In HRN mice following intravenous (iv) administration of midazolam, clearance (CL) was reduced by ∼ 80% compared to wild-type mice (WT). After oral administration, the AUC of midazolam was increased by ∼ 20-fold in HRN mice compared to WT mice; this greater effect suggests that hepatic first pass plays a role in the oral CL of midazolam. A 50% and an 80% decrease in CL were also observed in HRN mice following iv administration of docetaxel and theophylline, respectively, compared to WT mice. In addition, a 2- to 3-fold increase in tumor concentrations of G4222, a tool compound, were observed in tumor bearing HRN nude mice compared to tumor bearing nude WT mice. The observations from these experiments demonstrate that, for compounds that are extensively metabolized by hepatic CYP, the HRN mouse model could potentially be valuable for evaluating in vivo efficacy of tool compounds in drug discovery where high hepatic CL and low exposure may prevent in vivo evaluation of a new chemical entity.

Preclinical assessment of novel BRAF inhibitors: integrating pharmacokinetic-pharmacodynamic modelling in the drug discovery process.

The objective of these studies were to determine the preclinical disposition of the two BRAF inhibitors, G-F and G-C, followed by pharmacokinetic (PK)-pharmacodynamic (PD) modelling to characterize the concentration-efficacy relationship of these compounds in the Colo205 mouse xenograft model. With G-F, the relationship of pERK inhibition to concentration was also characterized. Compounds G-F and G-C were administered to mice, rats and dogs and the pharmacokinetics of G-F and G-C was determined. In addition, using indirect response models the concentration-efficacy relationship was described. The clearance of G-F was low; 0.625 and 4.65 mL/min/kg in rat and dog respectively. Similarly, the clearance of G-C was low in rat and dog, 0.490 and 4.43 mL/min/kg, respectively. Both compounds displayed low volumes of distribution (0.140-0.267 L/kg), resulting in moderate half-lives across species (~2.5 to 4 h). Bioavailability was formulation dependent and decreased with increasing dose. Using the indirect response models, the KC(50) (50% K(max); maximal response) value for tumor growth inhibition for G-F and G-C were 84.5 and 19.2 µM, respectively. The IC(50) for pERK inhibition in Colo205 tumors by G-F was estimated to be 29.2 µM. High exposures of G-F and G-C were required for efficacy. Despite good PK properties of low CL and moderate half-life, limitations in obtaining exposures adequate for safety testing in rat and dog resulted in development challenges.