Abstract
BMS-986094, a 2′-C-methylguanosine prodrug for the treatment of chronic hepatitis C virus infection, was withdrawn from phase 2 clinical trials because of unexpected cardiac and renal toxicities. To better understand these toxicities, the in vitro metabolism of BMS-986094 in human hepatocytes (HHs) and human cardiomyocytes (HCMs) and the measurement of BMS-986094 and selected metabolites in monkey plasma and tissues were assessed. BMS-986094 was extensively metabolized by HHs and HCMs, resulting in more efficient formation and accumulation of the active triphosphorylated metabolite, INX-09114, and less efficient efflux of metabolites in HCMs. The predominant metabolism pathway (hydrolysis) in HHs and HCMs was not associated with the formation of reactive metabolites or oxidative stress. In cynomolgus monkeys dosed with BMS-986094 of 15 or 30 mg/kg/d for 3 weeks, the nucleoside metabolite M2 was the major plasma analyte (66%-68% of the combined area under the curve). INX-09114 was the highest drug-related species in the heart and kidney (2,610-4,280 ng/mL [males]; ∼2-420× the concentration of other analytes). Other analytes increased dose dependently, with BMS-986094 highest in diaphragm (≤4,400 ng/mL) followed by M2 in liver and kidney (≤1,360 ng/mL), and M7 and M8 in other tissues (≤124 ng/mL). Three weeks after the last dose, INX-09114 remained high in the heart and kidney (≤1,870 ng/mL), with low M2 (≤37 ng/mL) in plasma and tissues. Persistent high concentrations of INX-09114 in the heart and kidney appeared to correlate with toxicities in these tissues in monkeys.
Introduction
Globally, an estimated 185 million people are infected with hepatitis C virus (HCV), including approximately 3.9 million people in the United States. 1,2 Chronic HCV infection is a major cause of liver cirrhosis and failure and hepatocellular carcinoma and is one of the most common hepatic causes of serious morbidity leading to death or liver transplantation. 3
Until 2011, pegylated interferon (PegIFN) and ribavirin (RBV) in combination for 24 or 48 weeks were the approved standard of care for HCV infection. With this regimen, sustained virologic response (SVR) rates were 40% to 50% with HCV genotype 1 but were higher for other genotypes. 4 However, the use of this combination is associated with a high incidence of adverse drug reactions including fatigue, flu-like symptoms, rash, anemia, and neutropenia. In 2011, the first direct acting antivirals (DAAs) that inhibit virally encoded enzymes critical for viral replication, and the serine protease inhibitors telaprevir and boceprevir, in combination with PegIFN and RBV were approved for use in HCV genotype 1 infection. This triple combination therapy achieved higher SVR rates but worsened some PegIFN/RBV-related side effects particularly anemia and neutropenia.
The large unmet medical need for chronic HCV infection resulted in extensive research efforts to identify more effective and better tolerated treatment modalities including second-generation DAA agents such as protease inhibitors of NS3/NS4a, NS5a, and nonnucleoside and nucleos(t)ide analogue inhibitors of NS5b. 5 -7 One of these was BMS-986094, a nucleotide analog HCV NS5B inhibitor (Figure 1), that bypasses the initial rate-limiting phosphorylation step of activation via prodrug delivery to the liver. Once in the liver, BMS-986094 is converted to the pharmacologically active triphosphate metabolite (INX-09114) with the other metabolites (eg, M2 and M7) converted to INX-09114 at a lower rate (Figure 2).
Structure of BMS-986094. BMS-986094 is an aryl-phosphoramidate of 6-O-methyl-2′-C-methyl guanosine (phosphoramidate prodrug) with significant potency in replicon assays and an ability to efficiently generate intracellular triphosphate in primary human hepatocytes. 5 The complete chemical structure is L-alanine, N-(2′-C-methyl-6-O-methyl-P-1-naphthalenyl-5′-guanylyl)-,2,2-dimethylpropyl ester.
Metabolism of BMS-986094 in human hepatocytes and cardiomyocytes. Chemical structures of BMS-986094 and its known metabolites and the biotransformation pathway leading to the pharmacologically active triphosphate metabolite INX-09114 along with alternate metabolic pathways. Note: aBased on related compoundsi. 18,19 Abbreviations: SULT = sulfotransferase; UGT = glucuronosyltransferase; ADA = adenosine deaminase; ADAL1 = adenosine deaminase-like protein 1; AMPD = adenosine monophosphate deaminase; CES1 = carboxylesterase 1; CatA = cathepsin A; and Hint1 = human phosphoramidase Hint 1.
In rat and monkey toxicology studies of BMS-986094 for up to 6 months in duration, skeletal muscle, heart, and kidney toxicities occurred at nontolerated doses, whereas minimal skeletal muscle degeneration (correlating with increased serum alanine and aspartate aminotransferases and creatine kinase) occurred at tolerated doses, suggesting that clinical dosing of BMS-986094 could proceed safely with appropriate clinical monitoring. In a phase 1 clinical study, 8 BMS-986094 alone or in combination with RBV for 7 days was well tolerated in treatment-naive HCV genotype 1-infected patients. However, dosing in a phase 2 study was immediately terminated after a patient who received BMS-986094 for 40 days experienced rapidly progressive heart failure characterized by decreased ventricular ejection fraction, electrocardiographic changes, and acute renal failure. Subsequently, other cases of unexpected clinical cardiac toxicity were identified. This was the first indication of clinical cardiac toxicity associated with the DAA treatment.
Based on this unexpected cardiac toxicity in clinical trials with BMS-986094, a series of nonclinical investigative studies were initiated to elucidate the potential mechanisms of toxicity. The primary objectives of the work described herein were to (1) evaluate in vitro metabolism and reactive metabolite formation of BMS-986094 in human hepatocytes (HHs), human cardiomyocytes (HCMs), and human liver microsomes and (2) assess tissue and plasma concentrations of BMS-986094 and selected metabolites in target and nontarget tissues in monkeys.
Materials and Methods
Materials
Cryopreserved HCM-induced pluripotent stem cells (iCell cardiomyocytes) were obtained from Cellular Dynamics International (Madison, Wisconsin). Cryopreserved HHs were purchased from Xenotech (Lenexa, Kansas) and In Vitro ADMET Laboratories (Columbia, Maryland). GSH-Glo glutathione assay and CellTiter-Glo luminescent cell viability assay kits were purchased from Promega (Madison, WI). 2′,7′- Dichlorofluorescein diacetate (DCFDA) cellular reactive oxygen species (ROS) detection assay kit was obtained from Abcam (Cambridge, Massachusetts). InvitroGro-HI was purchased from Bioreclamation IVT (Hicksville, New York). BMS-986094 (ie,
In Vivo Experiment
The protocol for a 3-week study in monkeys was approved by the Institutional Animal Care and Use Committee at BMS, and this study was conducted in accordance with the guidelines established by the American Association of Laboratory Animal Care. Cynomolgus monkeys (Macaca fascicularis) obtained from Charles River BRF, Inc (Houston, Texas) and Buckshire, Inc (Perkasie, PA) were approximately 2 to 6 years old on the first day of dosing. Monkeys were provided with certified primate diet (Harlan diet #2050C: certified 20% protein primate diet; Indianapolis, Indiana) and purified water ad libitum. Animal rooms were controlled for temperature and humidity and were maintained on an approximate 12-hour light/dark cycle. BMS-986094, corrected for purity and freebase content, was dissolved in the formulation vehicle that consisted of 95% Capmul medium chain monoglyceride and 5% Tween 80 (w/w) in water. BMS-986094 was then administered by oral gavage at 0 (vehicle), 15, or 30 mg/kg/d to groups of 5 cynomolgus monkeys/sex for 3 weeks followed by a 3-week recovery period. Control monkeys received the formulation vehicle alone at a 1 mL/kg dose volume. The high dose of 30 mg/kg/d was anticipated to produce multiorgan toxicity (heart, kidneys, and/or liver) based on the prior findings of 1- and 6-month oral toxicity studies in monkeys. Samples for BMS-986094 and metabolite analyses included plasma and peripheral blood monocytes (PBMCs) on days 1, 9, and 22; and tissues for drug/metabolite concentrations and biochemical analyses were collected at the end of dose (4 ± 1 hours after the last dose) and recovery necropsies. Prior to necropsy, all monkeys were euthanized by sedation with ketamine hydrochloride followed by anesthesia with intravenous propofol and subsequent exsanguination. After collection, tissue samples were weighed, homogenized with 4× (w/v) water containing 62.5 mM citric acid, snap frozen in liquid nitrogen, and stored in a freezer set to maintain a temperature ≤−70°C until analysis.
Cell Assays
In vitro incubations with HCMs
Cryopreserved HCMs were thawed according to the manufacturer’s recommendation and plated on 0.1% gelatin-coated 12-well plates at a seeding density of ∼0.5 × 106 cells/well. For 1-day incubations, BMS-986094 in fresh medium (10 µM, 1 mL) was added to each well and incubated in 7% CO2 at 37°C. Cells were harvested after 1, 3, 5, 7, 16, and 24 hours. For 10-day incubations, BMS-986094 in fresh medium (0.5 µM, 1 mL) was added to each well and incubated in 7% CO2 at 37°C. Freshly prepared culture media with BMS-986094 was replaced every 2 days (ie, 48, 96, 144, and 192 hours). Cells were harvested after 8, 24, 48, 72, 120, 168, and 216 hours. Six wells per time point were harvested and pooled for liquid chromatography tandem mass spectrophotometry (LC-MS/MS) analysis. For the long-term incubation, only 1 well per time point was collected for protein determination. The cells from 8 wells incubated with the medium (without BMS-986094) for 96 hours were used as blank for the standard curve. Cell culture medium was removed after centrifugation for 10 minutes at 13 000 rpm and stored at −70°C before LC-MS/MS analysis. The cells were briefly washed with 1 mL/well of ice-cold medium (without BMS-986094), 0.5 mL of 70% methanol with internal standard (IS) was added to each well, and the cells were removed by scraping and transferred to 2-mL centrifuge tubes. The wash step with 0.5 mL of 70% methanol with IS was repeated and the cells were vortexed and stored at −20°C overnight.
In vitro incubations with HHs
Cryopreserved HHs (lot number HC1-9A) were thawed according to the supplier’s recommended protocol and plated into each well of 0.1% gelatin-coated 12-well plates at a seeding density of ∼1.8 × 106 viable cells/well. The plates were placed in a 37°C, 5% CO2 humidified incubator for 2 to 4 hours to allow for attachment of cells. After attachment, media were removed from wells and replaced with cold hepatocyte culture media containing 0.25 mg/mL Matrigel (BD Biosciences, San Jose, California), and the plates were incubated for 2 days, with replacement of culture media every 24 hours. To initiate the assay, media were aspirated from wells, 0.65 mL of 10 μM BMS-986094 in fresh media was added to each well, and plates were returned to the incubator for up to 24 hours. At each time point, cell culture media were removed for LC/MS/MS analysis by gentle pipetting and hepatocytes were scraped into 70% ice-cold methanol (1 mL/well; 2 wells/time point), transferred to microtubes, and vortexed thoroughly. Cell lysates were then placed at −20°C overnight to facilitate the extraction of intracellular metabolites.
Glutathione Assay
Cryopreserved HHs (Lot number HH1020) and iPS-derived HCMs (Lot number 1912223) were thawed according to the manufacturer’s recommendation and plated on 96-well collagen-coated plates (Becton Dickinson, Franklin Lanes, New Jersey) 24 hours before compound exposure at a seeding density of 20,000 cells/well. Cells were then exposed to BMS-986094 (0.1, 0.25, 0.5, 1, 5, and 10 µM), M2 (0.1, 0.25, 0.5, 1, 5, and 10 µM), or 1-naphthol (0.1, 1, 10, 25, 50, and 100 µM) in culture medium for 3, 5, 24, 48, and 72 hours. At each time point, cellular glutathione (GSH) was measured using the GSH-Glo glutathione assay kit according to the supplier’s protocol. Buthionine sulfoximine (400 µM) was used as a positive control. Following treatment with the test compound, culture medium was removed and 100 µL of 1× GSH-Glo reagent was added to each well. Plates were mixed briefly on a plate shaker and allowed to incubate at room temperature for 30 minutes. After incubation, 100 µL of luciferin detection reagent was added to each well, mixed, and allowed to incubate at room temperature for 15 minutes. The assay mixture (100 µL) was transferred from each well to an opaque white plate, luminescence was recorded on an EnVision Plate Reader (PerkinElmer, Waltham, Massachusetts), and results (GSH levels) were expressed as a percentage of vehicle control. At the same time points, cellular adenosine triphosphate (ATP) content was measured as an indicator of cytotoxicity using the CellTiter-Glo luminescent cell viability assay kit, according to the supplier’s protocol. The cell plates were equilibrated to room temperature for 30 minutes. An equal volume of CellTiter-Glo reagent (100 µL) was added to each well, plates were rocked on a plate shaker to induce cell lysis, and allowed to incubate at room temperature for 15 minutes to stabilize the signal. The assay mixture (65 µL) was transferred from each well to an opaque white plate, luminescence was recorded on an EnVision Plate Reader (PerkinElmer), and the results (ATP levels) were expressed as a percentage of vehicle control.
Reactive Oxygen Species Assay
Generation of ROS in the hepatocytes was determined with the DCFDA cellular ROS detection assay kit, according to the supplier’s recommendation. Cryopreserved HHs (lot number HH1020) and HCMs (lot number 1912223) were thawed according to the manufacturer’s recommendation and were plated on 96-well collagen-coated plates (Becton Dickinson) for 24 hours before compound exposure at a seeding density of 40,000 cells/well. Cell culture medium was aspirated and the cells were washed with 100 µL of 1× phosphate-buffered saline (PBS)/well. The PBS was removed and 100 µL of DCFDA solution was added to each well, followed by a 45-minute incubation at 37°C. The DCFDA solution was removed and the plates were washed with 1× buffer. Growth medium (InvitroGro-HI + 10% fetal bovine serum, no phenol red) was added with BMS-986094 (0.1, 0.25, 0.5, 1, 5, and 10 µ M), M2 (0.1, 0.25, 0.5, 1, 5, and 10 µM), or 1-naphthol (0.1, 1, 10, 25, 50, and 100 µM). Growth medium with dimethyl sulfoxide was used as a vehicle control. The positive control, 2,5-dichloro-1,4-benzoquinone (0.01-100 µM), was used for the induction of ROS. Reactive oxygen species levels were measured after 1, 3, or 5 hours and 1, 2, or 3 days of exposure using an EnVision Plate Reader (Ex485 nm/Em535 nm), with results expressed as a percentage of vehicle control.
Total Protein Determination
Culture medium was removed from individual wells, cells were washed with 1× PBS, 1 mL of somatic cell ATP releasing agent (Sigma-Aldrich, St Louis, Missouri) was added to each well, and the plates were rocked for 10 minutes on a plate shaker to induce cell lysis. Total protein was determined using the bicinchoninic acid (BCA) protein assay (Cat No. BCA1; Sigma-Aldrich) for which the BCA working reagent (200 µL) was mixed with 25 µL of cell lysate and incubated for 30 minutes at 37°C, and absorbance (A562) was read on a SpectraMax i3 (Molecular Devices, Sunnyvale, California), with the protein concentration determined using a standard curve.
Bioanalytical Methods
The hepatocyte or cardiomyocyte cell extract samples were centrifuged for 10 to 30 minutes at 13,000 rpm and the supernatant was collected for LC-MS/MS analysis. Four different LC-MS/MS assays were used for the quantitation of BMS-986094 and its metabolites in cell medium and cell extracts from cardiomyocyte incubations. These assays were conducted with either a Shimadzu LC-30AD (Shimadzu Scientific Instruments, Somerset, New Jersey) high-performance liquid chromatography (HPLC) system or a Waters Acquity (Waters, Milford, Massachusetts) ultra performance liquid chromatography system. Mass spectroscopic analysis was performed with either an AB Sciex 4000 or 5500 Qtrap (Applied Biosystems, Toronto, Canada). Metabolite concentrations in cardiomyocyte cells were estimated based on 0.5 × 106 cells/well and a cell volume of 1 picoliter (pL) per cell. The cell volume was calculated using a measured cell radius of 6.25 μm. Metabolite concentrations in hepatocyte cells were estimated based on 1.8 × 106 cells/ well and a cell volume of 3 pL/cell.
Analyses of INX-09114 and M3 in Cell Extracts
Before HPLC injection, individual cell extracts (250 µL) were dried under a stream of nitrogen and reconstituted with 50 µL mobile phase A (10 mM N,N-dimethylhexylamine, 3 mM ammonium formate in water). The HPLC gradient elution was used with mobile phase A (10 mM N,N-dimethylhexylamine, 3 mM ammonium formate in water) and mobile phase B (20 mM N,N-dimethylhexylamine, 6 mM ammonium formate in acetonitrile) on a Xterra MS C18 50 × 2.1 mm, 3.5 µm column (Waters) at a flow rate of 0.3 mL/min and column temperature of 60°C. The initial gradient condition of 100% A was held for 0.1 minute, ramped to 100% B over a period of 7 minutes, held for 2.0 minutes, and then returned to initial conditions over 0.5 minute, where it remained for 2.5 minutes. Negative ion mode selected reaction monitoring (SRM) transitions were used to monitor for compounds INX-09114 (m/z 536.1 to m/z 158.7) and M3 (−m/z 456.0 to −m/z 158.8). INX-110037 was used as an IS.
Analyses of BMS-986094 in Cell Culture Medium and Cell Extracts
Before HPLC injection, cell medium (20 µL) was mixed with 50 µL of 70% methanol containing IS (INX-09146) and 130 µL of 30% acetonitrile in water. Cell extract (100 µL) was dried under a stream of nitrogen and reconstituted with 300 µL of 30% acetonitrile in water. The HPLC gradient elution was used with mobile phase A (0.1% formic acid in water) and mobile phase B (0.1% formic acid in acetonitrile) on a Synergy Polar reversed phase (50 × 2.0 mm, 5 µm) column (Phenomenex, Torrance, California) at a flow rate of 0.5 mL/min and column temperature of 60°C. The initial condition of 95% A was held for 0.1 minute, ramped to 100% B over a period of 1.9 minutes, held for 1.0 minute, and then returned to initial conditions over 0.5 minute, where it remained for another 0.5 minute. Positive ion mode SRM transitions were used to monitor for BMS-986094 concentrations (m/z 659.3 to m/z 166.0).
Analyses of M2, M8, and M4 in Cell Culture Medium and Cell Extracts
Before HPLC injection, cell culture medium (200 µL) was mixed with 20 μL of 70% methanol containing IS (INX-09146) and 30 µL of 0.1% formic acid in water. Cell extract (150 μL) was dried under a stream of nitrogen and reconstituted with 300 μL of 0.1% formic acid in water. The HPLC gradient elution was used with mobile phase A (0.1% formic acid in water) and mobile phase B (0.1% formic acid in acetonitrile) on a Synergy Polar reversed phase (50 × 2.0 mm, 5 µm) column (Phenomenex) at a flow rate of 0.5 mL/min and column temperature of 60°C. The initial condition of 98% A was held for 0.4 minute, ramped to 100% B over a period of 1.6 minutes, held for 1.0 minute, and then returned to initial conditions over 0.5 minute, where it remained for another 0.5 minute. Positive ion mode SRM transitions were used to monitor for M2 (m/z 298.1 to m/z 152.2), M8 (m/z 463.0 to m/z 166.0), and M4 (m/z 392.1 to m/z 166.1).
Analyses of M11 and M13 in Cell Culture Medium
Before HPLC injection, cell medium (200 µL) was mixed with 20 µL of 70% methanol containing IS (INX-09161) and 30 μL of 0.1% formic acid in water. The HPLC gradient elution was used with mobile phase A (0.1% formic acid in water) and mobile phase B (0.1% formic acid in acetonitrile) on a Synergy Polar reversed phase (50 × 2.0 mm, 5 µm) column (Phenomenex) at a flow rate of 0.5 mL/min and column temperature of 60°C. The initial condition of 95% A was held for 0.1 minute, ramped to 100% B over 1.9 minutes, held for 1.0 minute, and then returned to initial conditions over 0.5 minute. Negative ion mode SRM transitions were used to monitor for M11 (−m/z 319.0 to −m/z 142.8) and M13 (−m/z 222.9 to m/z 142.9).
Determination of Plasma and Tissue Exposures
In the 3-week study in monkeys, blood samples for determination of plasma concentration of BMS-986094 and M2 (the initial and primary nucleoside metabolite used as a surrogate marker for systemic exposure) and selected metabolites (M8, M7, and INX-09114) were obtained from the femoral vein of unanesthetized animals in weeks 1 and 3 of the dosing period and at the end of the 3-week recovery (week 6). At various time points throughout the study, approximately 1-mL blood samples were collected in prechilled tubes containing potassium ethylenediaminetetraacetic acid (K2EDTA) with 24 ± 2 μL of 2 M citric acid solution to stabilize the analytes. The tubes were inverted several times to ensure mixing and placed on ice. Within approximately 30 minutes after collection, blood samples were centrifuged refrigerated (2°C-8°C) at 1,000 to 1,300g for approximately 10 minutes to obtain acidified plasma samples.
At necropsy following the 3-week treatment and recovery periods, monkeys were euthanized by sedation with ketamine hydrochloride followed by anesthesia with intravenous propofol and subsequent exsanguination. Samples (approximately 6 g each) of heart, kidney, skeletal muscle (diaphragm), liver, and lung were collected and homogenized with 4× (w/v) water containing 62.5 mM citric acid and 2.25 mg/mL K2EDTA. Acidification with citric acid during tissue homogenization stabilizes BMS-986094, M8, and M7 in liver and heart homogenates for at least 2 hours in a water/ice bath. 9 All samples (plasma and tissue homogenates) were stored in a freezer set to maintain a temperature ≤−70°C until analysis. Plasma and tissue samples were analyzed for BMS-986094, M2, M8, M7, and INX-09114 using a validated LC-MS/MS detection method (lower limit of quantification [LLOQ] ranged from 1 to 20 ng/mL, for each respective analyte; except for INX-09114 in tissues where the LLOQ was 200 ng/mL). Individual and mean plasma and tissue concentrations were extracted from Watson (Version 7; Thermo Fisher Scientific, Waltham, MA) into an Excel (Version 97-2003) file, and the toxicokinetic parameter values were calculated from monkeys treated with test article only using noncompartmental methods by eToolbox version 2.7, Kinetica version 5.0 (Thermo Fisher Scientific, Philadelphia, Pennsylvania). Values below the LLOQ were not used in calculations, and area under the curve (AUC) was calculated using the linear trapezoidal rule.
Peripheral Blood Mononuclear Cell Processing
Blood samples (approximately 4 mL) in K2EDTA tubes were collected for isolation of PBMCs by centrifugal density gradient separation. Isolated PBMCs were resuspended in approximately 5 mL of RPMI-1640, cells were counted, and approximately 200 μL (∼2,500 cells/well) were transferred to a cryovial and stored frozen in a freezer set to maintain ≤−70°C for subsequent protein concentration determination by the Bradford method. The remaining PBMCs were pelleted by centrifugation, the supernatants were aspirated, and each cell pellet was resuspended in prechilled 100% methanol to attain a final concentration of approximately 70% methanol (v/v). To lyse cells, the cryovials containing methanol-suspended PBMCs were placed directly in a freezer set at ≤−70°C for approximately 2 hours, then thawed at room temperature for approximately 0.5 hour, and subsequently placed in a freezer at ≤−70°C until analyzed. Samples were analyzed for INX-09114 using a validated LC-MS/MS detection method (LLOQ was 20 ng/mL).
Results
The concentration–time curves of BMS-986094 and key intermediate metabolites that led to the formation of the pharmacologically active triphosphate metabolite, INX-09114 (Figure 2), in cells and culture medium following incubation of HHs and HCMs with BMS-986094 at 10 µM are shown in Figure 3. The disappearance rates of BMS-986094 in the culture media were estimated to be 0.023 hour−1 for HCMs at 0.5 × 106 cells/mL and 0.64 hour−1 for HHs at 1.8 × 106 cells/mL. The HHs to HCMs metabolism rate ratio was approximately 5:1 after normalization to cell densities. Three metabolites, M8, M4, and M2, were measured in the cell media and cell extracts/lysates. For these metabolites, intracellular concentrations were elevated above those determined in the culture medium, though the overall total metabolite amounts were higher in the culture medium (due to small cell volumes). While intracellular concentrations of M8, M4, and M2 were higher in HCMs compared to HHs after the same incubation time, concentrations in the culture medium were reversed. The diphosphate (M3) and INX-09114 metabolites were formed in both cell types, but at substantially higher concentrations in HCM cells (100-200 µM) compared to HH cells (20-30 µM). 1-Naphthol, the side product in the conversion of BMS-986094 to M8, was conjugated in HHs with sulfate and glucuronic acid to yield M13 and M11, respectively, as reflected by the presence of both metabolites in the HHs culture media. M13, but not M11, was detected in HCMs culture medium suggesting low or no glucuronidation activity in HCMs toward 1-napthol.
BMS-986094 and metabolite concentrations in human cardiomyocytes (HCMs) and hepatocytes (HHs). Time course for BMS-986094 metabolite formation as noted in media and cells (intracellular concentration) of HHs and HCMs incubated with 10 μM BMS-986094 for up to 24 hours. Concentrations of M4 and M11 in HCMs medium were below LLOQ (6.25 nM) and not shown in the figure. Intracellular concentrations of M3 in HCM cells were below LLOQ (80 µM) at 1, 3, and 16 hours.
Incubation of BMS-986094 (10 µM) up to 24 hours resulted in no apparent increase in the concentration of BMS-986094 or its metabolites in HHs and HCMs over time, except that intracellular concentrations of INX-09114 increased and plateaued after a 7-hour incubation in HCM. Incubations in HCMs with BMS-986094 at 10 µM beyond 24 hours were not feasible due to cytotoxicity. Therefore, a 10-day incubation in HCMs was performed with 0.5 µM BMS-986094 (replenished with freshly prepared BMS-986094-medium every 2 days) to evaluate possible accumulation of INX-09114 in HCM. However, data for 6 days or longer were deemed unreliable, as separate incubations of HCMs with BMS-986094 at 0.1 and 1 µM resulted in cytotoxicity at day 6 and beyond (data not shown). In addition, HCMs and HHs incubated with BMS-986094 for less than 6 days in the current study appeared to be affected by BMS-986094-related cytotoxicity as suggested by reductions in total cellular protein (Figure 4A). A reduction in BMS-986094 test concentrations from 10 to 0.5 µM substantially lowered the intracellular concentration of M2 but had less of an effect on INX-09114 concentrations. In light of total protein decreases, the intracellular concentration of INX-09114 increased with extended incubation up to 3 days but dropped substantially on day 5 (Figure 4A) likely due to excess cytotoxicity. Finally, the accumulation of INX-09114 in HCMs was more apparent when the metabolite concentrations were normalized with total protein (Figure 4B).
Intracellular concentrations of INX-09114 and M2 in human cardiomyocytes (HCMs) and hepatocytes (HHs). Time course for the formation of BMS-986094 metabolites, INX-09114 and M2, in HHs and HCMs treated with 0.5 μM BMS-986094 up to 120 hours. Intracellular metabolite concentrations were plotted against time and relative to total cellular protein (A). Graph depicting INX-09114 concentration ratio (prolonged vs 8-hour exposure) and normalized to total protein (B).
The potential for BMS-986094-related oxidative stress in HHs and HCMs was assessed by determining the levels of GSH and ROS in cells following incubation with BMS-986094, M2, or 1-naphthol as presented in Figures 5, 6, and 7. Cellular ATP content was also measured as an indicator of potential drug/metabolite-related cytotoxicity. Buthionine sulfoximine (a positive control for GSH depletion)-depleted GSH in both HHs and HCMs but did not deplete ATP at any time point up to 48 (HCMs) or 72 hours (HHs, Figure 8), indicating that the assays and test system were functional. BMS-986094 did not reduce GSH levels in either HHs or HCMs at the concentrations tested but induced time-related increases in cytotoxicity (ATP reduction) and increases in HCMs GSH at concentrations greater than 1 µM. M2 did not reduce GSH or ATP levels in HHs and HCMs in most incubations, except that a 30% ATP reduction in HHs were observed after a 72-hour incubation at 10 µM M2. Glutathione and ATP levels were not reduced by 1-naphthol in HCMs and HHs at concentrations up to 10 µM. At concentrations greater than 20 µM, 1-naphthol reduced hepatocyte GSH levels after only 1 hour of incubation, while increased GSH was evident in HCMs after 1- or 2-day incubations that resulted in cytotoxicity (reduced ATP). Decylubiquinone (positive control for ROS induction) increased ROS production in HHs after a 1 hour exposure by approximately 75% (Figure 9). After 5 hours of exposure, ROS production decreased with levels similar to controls after 72 hours. Decylubiquinone did not induce cytotoxicity (ATP depletion) in HHs for up to 48 hours (data not shown). Reactive oxygen species were not detected in HHs following exposure to INX-09114, M2, or 1-napthol at the time points evaluated (Figure 9).
Intracellular glutathione (GSH) and adenosine triphosphate (ATP) levels in human cardiomyocytes (HCMs) and hepatocytes (HHs) following treatment with BMS-986094. The HHs and HCMs were plated 24 hours before compound exposure at a seeding density of 20,000 and 40,000 cells/well for GSH and ATP assessments, respectively. Cells were exposed to BMS-986094 (0.1, 0.25, 0.5, 1, 5, and 10 µM) in culture medium for 3, 5, 24, 48, and 72 hours. At each time point, cellular GSH was determined and expressed as a percentage of vehicle control. At the same time points, cellular ATP content was measured using the CellTiter-Glo luminescent cell viability assay kit, and ATP levels were expressed as a percentage of vehicle control. Formation of intracellular GSH or ATP was plotted against increasing concentrations of BMS-986094.
Intracellular glutathione (GSH) and adenosine triphosphate (ATP) levels in human cardiomyocytes (HCMs) and hepatocytes (HHs) following treatment with M2. The HHs and HCMs were plated 24 hours before compound exposure at a seeding density of 20,000 cells/well. Cells were then exposed to M2 (0.1, 0.25, 0.5, 1, 5, and 10 µM) in culture medium for 3, 5, 24, 48, and 72 hours. At each time point, cellular GSH was determined and expressed as a percentage of vehicle control. At the same time points, cellular ATP content was measured using the CellTiter-Glo luminescent cell viability assay kit and ATP levels were expressed as a percentage of vehicle control. Formation of intracellular GSH or ATP was plotted against increasing concentrations of BMS-986094.
Intracellular glutathione (GSH) and adenosine triphosphate (ATP) levels in human cardiomyocytes (HCMs) and hepatocytes (HHs) following treatment with 1-naphthol. The HHs and HCMs were plated 24 hours before compound exposure at a seeding density of 20,000 cells/well. Cells were then exposed to 1-naphthol (0.1, 1, 10, 25, 50, and 100 µM) in culture medium for 3, 5, 24, 48, and 72 hours. At each time point, cellular GSH was determined and expressed as a percentage of vehicle control. At the same time points, cellular ATP content was measured using the CellTiter-Glo luminescent cell viability assay kit and ATP levels were expressed as a percentage of vehicle control. Formation of intracellular GSH or ATP was plotted against increasing concentrations of BMS-986094.
Intracellular glutathione (GSH) and adenosine triphosphate (ATP) levels in human cardiomyocytes (HCMs) and hepatocytes (HHs) following treatment with buthionine sulfoximine (BSO) as a positive control. The HHs and HCMs were plated 24 hours before compound exposure at a seeding density of 20,000 cells/well. Cells were then exposed to the positive control BSO at 400 µM in culture medium for 3, 5, 24, 48, and 72 hours. At each time point, cellular GSH was determined and expressed as a percentage of vehicle control. At the same time points, cellular ATP content was measured using the CellTiter-Glo luminescent cell viability assay kit and ATP levels were expressed as a percentage of vehicle control. Formation of intracellular GSH or ATP was plotted against increasing concentrations of BMS-986094.
Intracellular reactive oxygen species (ROS) levels in human hepatocytes (HHs) following treatment with BMS-986094, M2, 1-nathphol, and dichloro-1,4-benzoquinone (DBQ). The HHs were plated on 96-well collagen-coated plates 24 hours before compound exposure at a seeding density of 40,000 cells/well. Growth medium was added with BMS-986094 (0.1, 0.25, 0.5, 1, 5, and 10 µM), M2 (0.1, 0.25, 0.5, 1, 5, and 10 µM), 1-naphthol (0.1, 1, 10, 25, 50, and 100 µM), or dimethyl sulfoxide (DMSO, vehicle control). The positive control (DBQ; 0.01-100 µM) was used for the induction of ROS. The ROS levels were measured after 1, 3, 5, 24, 48, and 72 hours of exposure (Ex485 nm/Em535 nm). Results were expressed as a percentage of vehicle control against increasing concentrations of test substance.
To further characterize the possible relationship of BMS-986094, M2, and other metabolites to the cardiac and renal toxicities observed in vivo, plasma and tissue metabolite exposures were evaluated in monkeys dosed with BMS-986094 for 3 weeks at 15 or 30 mg/kg/d followed by a 3-week recovery period. The data are presented in Figure 10 and Tables 1 to 3. There were no substantial sex differences for the plasma exposures; thus, the plasma exposure values were reported as group (sex-combined) means. The AUC on day 1 was not calculated as there were only 2 concentration points at 4 and 24 hours postdose. BMS-986094 plasma concentrations were only detected at very early time points following dosing with most samples less than the LLOQ (1.00 ng/mL). Mean BMS-986094 concentrations were generally similar on days 9 and 22. Additionally, plasma concentration–time profiles were similar for the 3 sampling periods (days 1, 9, and 22) for M2, M8, and M7. After repeat dosing with BMS-986094 in monkeys, M2 values increased approximately dose proportionally from 15 to 30 mg/kg/d (AUC [0-24 hours] of 441 and 1,176 ng·h/mL, respectively), and mean M2 AUC values on day 22 were similar to those on day 9 but higher (1.3 to 4.0×) than on day 1, indicating that steady-state levels of M2 were achieved after 2 weeks of dosing. After repeat dosing, M7 and M8 AUC values generally increased proportionally from 15 to 30 mg/kg/d of BMS-986094. There was no apparent accumulation of these metabolites in monkey plasma. Concentrations of the active triphosphorylated nucleoside, INX-09114, were less than LLOQ in all plasma (1 ng/mL) and PBMC (20 ng/mL) samples.
Plasma concentration-time course of BMS-986094 and metabolites following repeat oral administration of BMS-986094 to monkeys. BMS-986094 and its M2, M7, and M8 metabolites were evaluated in monkeys treated with BMS-986094 for 3 weeks at 15 (A) or 30 (B) mg/kg/d.
Plasma exposure of BMS-986094 and Metabolites in Monkeys Following 3 Weeks of Daily Oral Administration of BMS-986094.
Abbreviation: AUC, area under the concentration curve.
a Three monkeys/sex/group.
Mean Tissue Concentration Data in Monkeys Following 3 Weeks of Daily Oral Administration of BMS-986094.
Abbreviations: LLOQ, lower limit of quantification; NA, not applicable.
a Three monkeys/sex/group; standard deviation was calculated when there were measurable concentrations for all monkeys. Data were NA as all concentrations were less than LLOQ of 10 ng/mL for tissue and 1 ng/mL for plasma.
Mean Tissue Concentration Data in Monkeys After 3 Weeks of Daily Oral Administration of BMS-986094 Followed by 3 Weeks of Recovery.
Abbreviations: LLOQ, lower limit of quantification; NA, not applicable.
a One or 2 monkeys/sex/group; Data were NA as all concentrations were less than the LLOQ of 10 ng/mL for tissue and 1 ng/mL for plasma.
Mean tissue concentration data at the end of dosing (day 22) are listed in Table 2. Tissue concentrations of BMS-986094 and most metabolites were similar in male and female monkeys, with the exception of greater concentrations of the active metabolite, INX-09114, in the heart of males (3.3× females) at 30 mg/kg/d. All monitored metabolites were detected in diaphragm, liver, lung, heart, kidney, and plasma with the exception of INX-09114 that was detected in heart and kidney only. For BMS-986094, the highest tissue concentrations generally were in the diaphragm followed by lung, liver, kidney, heart, and plasma with tissue/plasma ratios of 59 (diaphragm), 17 to 98 (lung), and 12 (liver) in the 15-mg/kg/d group and 3,610 (diaphragm), 2.6 to 58 (lung), 4.7 to 23 (liver), 1.9 to 10 (heart), and 3.3 to 3.9 (kidney) in the 30-mg/kg/d group. For M2, the highest concentrations were in the kidney followed by liver, heart, lung, diaphragm, and plasma with tissue/plasma ratios of 0.7 to 1.1 (diaphragm), 2.2 to 2.8 (heart), 19 to 37 (kidney), 10 to 12 (liver), and 1.1 to 1.5 (lung). M7 concentrations were similar in kidney and liver but slightly lower in diaphragm, heart, lung, and plasma with tissue/plasma ratios of 0.7 to 1.0 (diaphragm), 0.9 (heart), 1.6 to 3.5 (kidney), 3.0 to 4.9 (liver), and 0.7 to 1.1 (lung). Concentrations of INX-09054 were also similar in diaphragm, heart, kidney liver, and lung but slightly higher in plasma and with tissue/plasma ratios of 1.0 to 1.5 (diaphragm), 0.1 (heart), 0.3 to 0.4 (kidney), 0.2 to 0.8 (liver), and 0.3 to 0.5 (lung). M2, M7, and M8 were distributed to all tissues analyzed without substantial differences, except for notably higher M2 concentrations in kidney. Tissue concentrations of INX-09114 in heart and kidney were similar but substantially higher than for other metabolites. At the end of the 3-week recovery period (day 42), BMS-986094, M7, and M8 were not present in plasma or any tissue (data not shown); whereas M2 was still present in kidney, plasma, diaphragm, heart, and liver, though at levels much lower (0.02 to 0.2×) than at the end of dosing on day 22. INX-09114 remained remarkably high (≤1,870 ng/mL) in heart and kidney, although concentrations were less (0.1 to 0.5×) than those on day 22. In the 30-mg/kg/d group, INX-09114 concentrations were higher in males (3.5-3.7×) than in females.
Discussion
As part of the investigative studies aiming to define the mechanism(s) of the cardiovascular and renal toxicities seen following the administration of BMS-986094 to nonclinical species and humans, a targeted approach to understand reactive mechanisms (eg, potential ROS generation, covalent protein binding, and GSH depletion) and metabolite disposition in target tissues and their potential relationship to clinical toxicity was undertaken.
Data generated during this investigation indicate that the uptake of BMS-986094 into HHs and HCMs is extensive, with a higher metabolism rate in HHs than in HCMs. The compound was nearly depleted in cell culture medium after a 5-hour incubation with HHs in vitro, but in HCMs more than half of the initial drug concentration was maintained throughout a 24-hour incubation period. Compared to HHs, the reduced rate of BMS-986094 metabolism in HCMs was also reflected by the overall lower metabolite concentrations in culture media for M8, M4, M2, M11, and M13. In contrast, intracellular concentrations of all metabolites measured (M8, M4, M3, M2, and INX-09114) were higher in HCMs as compared to HHs. This finding likely reflected several factors including (1) the efflux transport of M8, M4, and M2 may be less efficient in HCM, (2) the fraction of BMS-986094 metabolism shunted the formation of M3 and INX-09114 may be elevated in HCM, and (3) elimination of M3 and INX-09114 may be reduced in HCMs, implying lower nucleotidase activities for dephosphorylation of these phosphate metabolites. When BMS-986094 concentrations in HCM incubations were reduced from 10 to 0.5 µM, the intracellular concentration of M2 was significantly reduced as anticipated; however, the intracellular concentration of INX-09114 was only slightly decreased. Additionally, there was an apparent accumulation over time of INX-09114 following the incubation of HCMs with BMS-986094 at 0.5 µM. As M2 is a final product of dephosphorylation of INX-09114, these data provide additional evidence of reduced dephosphorylation activity toward INX-09114 in HCMs, resulting in INX-09114 accumulation.
In addition to the hydrolytic pathway that leads to the formation of INX-09114, BMS-986094 was found to undergo oxidative metabolism in HHs but not in HCMs (unpublished data, Wenying Li, PhD, 2013). Therefore, it is plausible that competing metabolic pathways in HHs may reduce the fraction of BMS-986094 metabolism to INX-09114; thus, the nucleoside metabolite formation pathway appeared more efficient in HCMs. M11 was not detected in HCM incubations suggesting minimal or no glucuronidation activity toward 1-nathphol in HCMs. However, sulfate conjugation activity for 1-naphthol was present in HCMs.
Protein covalent binding indicative of reactive metabolite formation was found to be primarily associated with oxidative metabolism of BMS-986094 at the naphthyl moiety. However, no appreciable GSH reduction or ROS formation was detected following the exposure of intact HHs and HCMs to BMS-986094 or M2 at concentrations and incubation times that produced cytotoxicity as measured by ATP reduction. In contrast, while 1-naphthol did not reduce GSH or ATP in HCMs, GSH reduction preceded ATP reduction in HHs exposed to 1-naphthol at concentrations higher than 20 µM, which may be attributed to oxidative metabolism of 1-naphthol in HHs, leading to the formation of reactive metabolites. This oxidative metabolism pathway may be overshadowed by glucuronidation and sulfate conjugation pathways in HHs such that GSH reduction occurs or is apparent only at high drug concentrations. Due to lack of or low oxidative metabolism activity in HCM, the potential for reactive metabolite formation following oxidative metabolism of 1-naphthol or BMS-986094 is low, as confirmed by the GSH reduction data. Collectively, these results indicate that metabolism of BMS-986094 in hepatocytes and cardiomyocytes does not result in the formation of appreciable reactive metabolites, and thus, it is hypothesized that reactive metabolites likely do not play a substantial role in clinically identified cardiac toxicity in humans treated with BMS-986094. This hypothesis is further supported by data demonstrating that the predominant hydrolytic metabolism of BMS-986094 in intact HHs and HCMs is not generally associated with reactive metabolite formation. Finally, it should be noted that although important in HH, oxidative metabolism of BMS-986094 represents only a small portion of the overall metabolic pathways for this compound in HCM.
Following dosing of monkeys with BMS-986094 for 3 weeks, systemic exposures to M2 (initial nucleoside) were highest among BMS-986094 and the 4 select circulating metabolites analyzed and comprised 66% to 68% of the combined AUC. This is consistent with the results from a separate adsorption, distribution, metabolism, and excretion study in monkeys that defined M2 as the major circulating drug-derived species following a single 15 mg/kg oral dose of [14C]BMS-986094 (data not shown). Systemic exposure of BMS-986094 in the 3-week monkey was low throughout the study, suggesting that the prodrug (BMS-986094) undergoes rapid first-pass hepatic uptake and metabolism. As anticipated, the triphosphate metabolite, INX-09114, was not detected in plasma or PBMCs. The absence of INX-09114 in PBMCs was likely due to (1) little or no BMS-986094 uptake into PBMCs due to low plasma exposures and/or (2) the enzymatic pathway for converting BMS-986094 or M8 into INX-09114 is not fully present or functional in monkey PBMCs.
Tissue exposure data in the 3-week monkey study showed that BMS-986094 was present at low levels in all tissues analyzed after repeat oral dosing. While BMS-986094 was detected at notably high levels in the diaphragm (skeletal muscle) in 2 monkeys in the 30-mg/kg/d group, concentrations in the diaphragm in the remaining 4 monkeys in this group were below the detection limit. The cause of this wide variation in BMS-986094 diaphragm concentrations is unknown. The presence of M8, M7, and M2 in all tissues may be attributed to the uptake of these metabolites from plasma or metabolism of BMS-986094 or the metabolic precursors in the tissues. INX-09114 was below the detection limited in diaphragm, lung, and liver, but present at high concentrations in heart and kidney. Given the low circulating concentration of BMS-986094, it is possible that formation of INX-09114 in heart and kidney may have been the result of uptake and metabolism of M8. Phosphorylation of M2, which has been shown to be slower than phosphorylation of the monophosphate intermediates (such as M1; data not shown), may also contribute to the INX-09114 formation. As INX-09114 can be formed in hepatocytes, these data suggest a high rate of clearance of the triphosphate via excretion or metabolism in monkey liver and low dephosphorylation activity of INX-09114 in heart and kidney, which was consistent with the relatively high levels of INX-09114 in these tissues after a 3-week recovery period. Assuming no INX-09114 formation after the last dose in the monkey study (day 22), the half-life for INX-09114 clearance is estimated to be around 7 to 19 days in kidney and 13 to 18 days in the heart.
Several potential hypotheses/mechanisms have been proposed to explain the BMS-986094-related nonclinical and clinical cardiac toxicity including the formation of damaging ROS; mitochondrial toxicity (adverse effects on respiration and/or copy number); inhibition of cellular polymerases α, β, and γ; carnitine depletion associated with the administration of high molar doses of pivalic acid (seen with some pivalate containing prodrugs 10,11 ); and/or a combination of these mechanisms. It is widely accepted that the cardiac toxicity caused by some drugs is mediated by ROS produced during intracellular metabolism. 12,13 Cascades of ROS, including the potent hydroxyl radical (·OH), are capable of damaging DNA and proteins and initiating membrane lipid peroxidation leading to cell death and cardiac toxicity. Additionally, the cellular antioxidant defense system is an important factor in determining the fate of cells in response to oxidative injury. 14,15 Of all antioxidants found in cardiomyocytes, GSH plays the most important role of protecting against ROS. 16 In vitro studies demonstrate that GSH and N-acetyl cysteine, a precursor of GSH, can directly scavenge free radicals and reduce oxidant-induced cell damage and death. 17 Our in vitro results indicated minor oxidative metabolism of BMS-986094 in HCMs and HHs and that BMS-986094 was not associated with reduced intracellular GSH in HHs or HCMs or ROS induction in HCMs. Moreover, the increases in GSH in HCMs observed with longer incubations of high concentrations of BMS-986094 and 1-napthol may have been due to oxidative stress associated with cytotoxicity. Overall, these findings suggest that drug-induced changes in free radical generation or ROS may not be a primary contributing factor to the cardiac toxicity observed in nonclinical and clinical trials with BMS-986094. Moreover, although the mechanisms cannot be definitively explained at this time, it is hypothesized that the persistent high concentrations of INX-09114 in monkey heart and kidney may have contributed to target tissue toxicity.
In summary, in vitro and in vivo metabolism studies with BMS-986094 suggest the predominant metabolism pathway (hydrolysis) of BMS-986094 in HHs and HCMs was not associated with the formation of reactive metabolites or pronounced oxidative stress. Persistent high concentrations of the active triphosphorylated metabolite, INX-09114, in monkey heart and kidney appeared to correlate with toxicities in these tissues. There were no other in vivo correlations between other BMS-986094 metabolites evaluated and target organ toxicities.
Footnotes
Authors’ Note
W. Li and K. Trouba contributed to the conception and design, acquisition, analysis, and interpretation, drafted the manuscript, and critically revised the manuscript. W.G. Humphreys contributed to the conception, design, and interpretation and critically revised the manuscript. L. Ma, J. Kwagh, C. Storck, Y. Zhu, B. Wang, and A. Liu contributed to the design, data analysis and interpretation, and critical revised the manuscript. O. Flint contributed to the design, conception, analysis, and interpretation, and critically revised the manuscript. M. Graziano contributed to the conception and design and critically revised the manuscript. M. Davies contributed to the conception and design, interpretation, and critically revised the manuscript. T. Sanderson contributed to the conception and design, interpretation, and critically revised the manuscript. All authors gave final approval and agree to be accountable for all aspects of work ensuring integrity and accuracy.
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: When these data were generated, all authors were employed by Bristol-Myers Squibb Company that was developing BMS-986094.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Bristol-Myers Squibb Company funded this research.