Advertisement
Research ArticleArticle

Pharmacokinetics, Metabolism, and Excretion of the Glycogen Synthase Kinase-3 Inhibitor LY2090314 in Rats, Dogs, and Humans: A Case Study in Rapid Clearance by Extensive Metabolism with Low Circulating Metabolite Exposure

Maciej J. Zamek-Gliszczynski, Trent L. Abraham, Jeffrey J. Alberts, Palaniappan Kulanthaivel, Kimberley A. Jackson, Kay H. Chow, Denis J. McCann, Haitao Hu, Shelby Anderson, Nathan A. Furr, Robert J. Barbuch and Kenneth C. Cassidy
Drug Metabolism and Disposition April 2013, 41 (4) 714-726; DOI: https://doi.org/10.1124/dmd.112.048488

Abstract

LY2090314 (3-[9-fluoro-2-(piperidin-1-ylcarbonyl)-1,2,3,4-tetrahydro[1,4]diazepino[6,7,1-hi]indol-7-yl]-4-imidazo[1,2-a]pyridin-3-yl-1H-pyrrole-2,5-dione) is an intravenous glycogen synthase kinase-3 inhibitor in oncology trials. Drug disposition was characterized after intravenous infusion of [14C]LY2090314 to rats and dogs, and was related to available clinical data. LY2090314 exhibited high clearance (approximating hepatic blood flow) and a moderate volume of distribution (∼1–2 l/kg) resulting in rapid elimination (half-life ∼0.4, 0.7, and 1.8–3.4 hours in rats, dogs, and humans, respectively). Scaled clearance from liver microsomes accurately predicted perfusion-limited clearance across species. LY2090314 was cleared by extensive metabolism, and the numerous metabolites were rapidly excreted into feces via bile (69–97% of dose; 62–93% within 0–24 hours); urinary recovery of drug-related material was low (≤3% of dose). Despite extensive metabolism, in rats and humans the parent compound was the sole identifiable drug-related moiety in plasma. Even in Mdr1a-, Bcrp-, and Mrp2-knockout rats, LY2090314 metabolites did not appear in circulation, and their urinary excretion was not enhanced, because the hypothesized impaired biliary excretion of metabolites in the absence of these canalicular transporters was not observed. Canine metabolite disposition was generally similar, with the notable exception of dog-unique LY2090314 glucuronide. This conjugate was formed in the dog liver and was preferentially excreted into the blood, where it accounted for the majority of circulating radioactivity at later times, and was predominantly recovered in urine (16% of dose). In conclusion, LY2090314 was rapidly cleared by extensive metabolism with negligible circulating metabolite exposures due to biliary excretion of metabolites into feces with no apparent intestinal reabsorption.

Introduction

Glycogen synthase kinase-3 (GSK-3) is a serine/threonine kinase that regulates numerous pathways involved in protein synthesis, cell proliferation, differentiation, migration, and apoptosis (Doble and Woodgett, 2003). Due to its diverse physiological roles, GSK-3 has been explored as a therapeutic target for several diseases, including diabetes, inflammation, mood disorders, Alzheimer’s disease, and cancer (Jope et al., 2007). Glycogen synthase kinase inhibitors were initially studied for the treatment of diabetes (Kaidanovich and Eldar-Finkelman, 2002). GSK-3 also regulates Toll-like receptor production of inflammatory cytokines, and its inhibition demonstrated anti-inflammatory activity (Martin et al., 2005; Jope et al., 2007). The activity of lithium in mood disorders may be partially attributed to GSK-3 modulation, although the inhibition is not potent or extensive clinically (lithium GSK-3 Ki ∼2 mM; steady-state concentrations ∼1 mM) (Klein and Melton, 1996; Jope et al., 2007). GSK-3 has been implicated in Alzheimer’s disease pathology because it phosphorylates amyloid precursor and tau proteins, and its inhibition attenuates A-beta neurotoxicity (Grimes and Jope, 2001). GSK-3 constitutively suppresses pathways promoting neoplastic transformation, such as Wnt and PI3K, so its inhibition in the treatment of cancer is counterintuitive (Jope et al., 2007). However, GSK-3 inhibition alone is insufficient to elicit cellular transformation; to the contrary, more recent studies have demonstrated that GSK-3 inhibition potentiates apoptosis in solid-tumor chemotherapy and possesses single-agent efficacy in preclinical models of leukemia (Boissan et al., 2005; Tan et al., 2005; Wang et al., 2008).

The focus of our study is on an intravenous GSK-3 inhibitor synthesized by Eli Lilly and Company for the treatment of cancer. LY2090314 (3-[9-fluoro-2-(piperidin-1-ylcarbonyl)-1,2,3,4-tetrahydro[1,4]diazepino[6,7,1-hi]indol-7-yl]-4-imidazo[1,2-a]pyridin-3-yl-1H-pyrrole-2,5-dione) is a potent (GSK-3α IC50 = 1.5 nM; GSK-3β IC50= 0.9 nM) and selective ATP-competitive GSK-3 inhibitor currently in phase II studies for both hematological malignancies as a single agent and for solid tumors as a potentiator of platinum chemotherapy. GSK-3 supports proliferation and transformation of mixed-lineage leukemia cells, and its inhibition has demonstrated efficacy in preclinical leukemia models (Wang et al., 2008). Therefore, LY2090314 may have single-agent efficacy in leukemia patients. In contrast, GSK-3 inhibitors generally lack single-agent activity against solid tumor cells; however, they potentiate the activity of other oncolytics by promoting progression of chemotherapy-induced cell cycle arrest to apoptosis (Tan et al., 2005). Preclinically, LY3090314 enhanced the efficacy of cisplatin and carboplatin in solid tumor cancer cell lines in vitro and in vivo (Brail et al., 2011). Thus, LY2090314 may improve the efficacy of platinum-based chemotherapy.

We characterized the pharmacokinetics, metabolism, and excretion of LY2090314 in rats and dogs, and compared these findings to the clinical drug disposition observed in the phase I study in cancer patients (clinical study results expected to be reported separately in 2013). After intravenous infusion, LY2090314 was rapidly cleared by extensive metabolism. However, across species, the circulating metabolite exposure was negligible due to extensive biliary excretion of metabolites into feces, with the notable exception of the dog-specific LY2090314 glucuronide. LY2090314 presents an interesting case study in rapid metabolic clearance with negligible exposure to circulating metabolites.

Materials and Methods

Materials.

LY2090314 and [14C]LY2090314 were synthesized by Eli Lilly and Company. The preclinical vehicle for intravenous infusion was 20% Captisol (Ligand, La Jolla, CA) in 0.08N hydrochloric acid in sterile water for injection. For clinical trial use, the investigational drug was supplied as a lyophilized powder in glass vials (165 mg of free base/vial) containing LY2090314 ethanol solvate and the inactive ingredients Captisol and tartaric acid.

All microsomes for glucuronidation studies were purchased from Xenotech (Lenexa, KS), and microsomes for clearance studies were obtained from BD Gentest (Woburn, MA). Sprague-Dawley rats and Beagle dog blood (pool of ≥3 male donors) were purchased from Biochemed Services (Winchester, VA); human blood was obtained by Covance (Wisconsin, WI) from three male volunteers who had not taken any known medication during the previous 7 days. Plasma for protein-binding studies was obtained from Sprague-Dawley rats and Beagle dog blood (pool of ≥3 male donors) by Advinus Therapeutics (Bangalore, India), and human plasma was obtained from TTK Rotary Blood Bank (Bangalore, India).

Animals.

Male Sprague-Dawley rats (174–320 g) were purchased from Hilltop Laboratory Animals (Scottdale, PA). The supplier performed the femoral vein cannulation for drug infusion in all rats as well as the bile duct cannulation in the appropriate group of rats. For comparative metabolite profiling between wild-type and transporter knockout rats (290–370 g), male wild-type Sprague-Dawley rats were purchased from Harlan (Indianapolis, IN), and male Sprague-Dawley Mdr1a-, Bcrp (breast cancer resistance protein)-, and Mrp2-knockout rats were purchased from SAGE (St. Louis, MO); the femoral vein, artery, and bile duct cannulations were performed by Covance (Greenfield, IN). Male Beagle dogs (9.3–10.5 kg) were purchased from Covance Research Products (Kalamazoo, MI). The dogs were not precannulated, and the intravenous infusion was administered via a temporary cephalic vein catheter. CD-1 mice (20–30 g) of both sexes and female Cynomolgus monkeys (5–6 kg) were obtained from the colonies of Covance. The animal care and use committees at Hilltop Laboratory Animals and Covance approved all animal procedures.

Preclinical Pharmacokinetic and Excretion Studies.

All rats received a 30-minute intravenous infusion of [14C]LY2090314 (5 mg/kg; 50 µCi/kg; 5 ml/kg) via the femoral vein cannula. Pharmacokinetic samples were taken in three rats/time point (cardiac puncture) at 0.25 (middle of infusion), 0.5 (end of infusion), 0.75, 1, 1.5, 2, 3, 6, 12, 24, 72, and 120 hours after the start of infusion. In a separate group of rats housed in Nalgene metabolism cages (Nalgene Nunc, Penfield, NY) (n = 4), cumulative urine samples were collected between 0–12 and 12–24 hours in 24-hour intervals through 168 hours; feces were collected in 24-hour intervals through 168 hours. After each 24-hour excreta collection, the cages were rinsed with ethanol:water (1:1, v:v), and the wash fluid was collected. At the end of the study, rats were euthanized, and the carcasses and cage wash/wipe materials were retained for radioanalysis. The weights of all samples, except blood and plasma, were recorded.

In a separate group of bile duct-cannulated rats housed in Nalgene metabolism cages (n = 4), cumulative bile and urine samples were collected between 0–12 and 12–24 hours, and then at 24-hour intervals through 72 hours after start of infusion; feces were collected in toto at 24-hour intervals through 72 hours. At 24 and 48 hours in bile duct-cannulated rats, the cages were rinsed, and the wash fluid was retained for radioanalysis. After the last excreta collection, rats were euthanized, and the carcasses and cage wash/wipe material were retained for radioanalysis. The weights of all samples, except blood and plasma, were recorded.

In a separate study, the plasma, bile, and urine metabolite profile was compared between wild-type and Mdr1a-, Bcrp-, and Mrp2-knockout rats (n = 3–6). All rats received a 30-minute intravenous infusion of nonradiolabeled LY2090314 (5 mg/kg; 5 ml/kg) via the femoral vein cannula. Plasma samples were collected via the femoral artery cannula 0.5 (end of infusion), 1, 2, 3, and 6 hours after the start of infusion. Urine and bile (collected via the bile duct cannula) were collected in toto between 0 and 6 hours.

All dogs received a 30-minute intravenous infusion of [14C]LY2090314 (5 mg/kg; 50 µCi/kg; 2 ml/kg) via the cephalic vein. Dogs were housed in stainless steel metabolic cages during the duration of the study. Blood samples were collected via the jugular vein into K3EDTA tubes at 0 (preinfusion), 0.25 (middle of infusion), 0.5 (end of infusion), 0.75, 1, 1.5, 2, 3, 4, 6, 8, 12, 24, 48, 72, 96, and 120 hours after the start of infusion. Cumulative urine samples were collected between 0–12 and 12–24 hours, and at 24-hour intervals through 168 hours after the start of infusion. Feces were collected in toto at 24-hour intervals through 168 hours after the start of infusion. After each 24-hour excreta collection cages were rinsed, and the wash fluid was retained for radioanalysis. The weights of all samples, except blood and plasma, were recorded.

For the purposes of comparison with scaled clearance from liver microsomes, LY2090314 clearance was also determined in female Cynomolgus monkeys (n = 3, 1 mg/kg i.v. bolus) as well as in CD-1 mice of both sexes (n = 3 mice/time point, 1–20 mg/kg i.v. bolus dose range).

Clinical Pharmacokinetics.

The phase I clinical study was a multicenter, open-label, nonrandomized, dose-escalation clinical trial of intravenous LY2090314 in the treatment of patients with advanced and/or metastatic cancer for whom no treatment of higher priority existed. The institutional review boards at Moffitt Cancer Center (Tampa, FL), Sarah Cannon Research Institute (Nashville, TN), and Medical University of South Carolina (Charleston, SC) approved the clinical study protocol. In this study, LY2090314 was first given alone and then in combination with carboplatin and pemetrexed over a dose range of 10–120 mg, given as an approximate 60-minute intravenous infusion. Plasma pharmacokinetic samples for determination of LY2090314 pharmacokinetics were taken at the following times: 0 (predose), 0.5 (middle of infusion), 1 (end of infusion), 1.5, 2, 4, 6, 8, and 24 hours after the start of infusion. Note that additional samples were taken at other times during the conduct of the study for determination of LY2090314 as well as carboplatin (free and total) and pemetrexed pharmacokinetics, but those are beyond the scope of this article.

For determination of renal clearance in humans, the total urine output was collected for approximately 6 hours after the infusion (n = 8 measurements in 5 patients) at the 80 and 120 mg LY2090314 dose levels. Urine was only collected for the first 6 hours because this period on average accounted for 91% of LY2090314 area under the curve0–∞ (Table 1). Urine was pooled and refrigerated, the total volume was recorded at the end of the collection period, and an aliquot was frozen for later analysis. The corresponding LY2090314 systemic exposure was measured over the urine collection interval to enable calculation of renal clearance values.

View this table:
TABLE 1

Systemic pharmacokinetic (plasma) and excretion parameters for LY2090314 and total drug-related radioactivity in rats and dogs after a single dose of [14C]LY2090314 (5 mg/kg; 50 µCi/kg) administered as a 30-minute intravenous infusion

Plasma Protein Binding and Blood-to-Plasma Partitioning.

LY2090314 (0.05, 0.5, 5, and 50 μg/ml) rat, dog, and human plasma protein binding was determined in vitro using equilibrium dialysis, as described previously elsewhere (Zamek-Gliszczynski et al., 2011b). LY2090314 (0.05, 0.5, 5, 50, and 250 μg/ml) rat, dog, and human blood-to-plasma partitioning was determined in vitro by incubating whole blood with radiolabeled drug material for 60 minutes; the plasma was isolated by centrifugation, and the radioactive content of whole blood and plasma was quantified.

Metabolism in Microsomes.

LY2090314 metabolic stability was assessed in mouse, rat, monkey, dog, and human liver microsomes. LY2090314 was incubated in microsome suspension (2 μM; 0.5 mg/ml microsomal protein; 0.5 hours; 37°C). Incubations were stopped by addition of an equal volume of acetonitrile; parent concentrations were determined by liquid chromatography-tandem mass spectrometry (LC-MS/MS), and, in addition, samples were profiled for parent-related metabolite(s) using LC-MS/MS. Based on the clearance properties of LY2090314 (high clearance, high binding), the direct scaling method using the well-stirred model of hepatic clearance was used to project in vivo clearance (Poulin et al., 2012).

Formation of the N-glucuronide metabolite (M39) was studied across rat, dog, and human liver and kidney microsomes (37°C, 60 minute, 10 µM LY2090314, 1 mg/ml microsomal protein concentration, 0.5 mM UDP-glucuronic acid, 1 mM MgCl2, and 50 µg alamethicin/mg protein) using 10 µM 4-methylumbelliferone as the positive control. To further rule out the formation of M39 by dog kidney microsomes, incubations were repeated as described earlier by using a 2-hour incubation with 1.8 mg/ml of microsomal protein concentration, 18 mM UDP-glucuronic acid, and 5 mM saccharolactone. M39 generated by dog liver microsomes was isolated by semipreparative liquid chromatography-mass spectrometry (LC-MS), and fractions were collected manually upon detection of the 689 m/z (mass-to-charge ratio) ion. Collected fractions were combined, freeze dried, dissolved in 150 µl of deuteromethanol, and analyzed by NMR (proton, two-dimensional homonuclear gradient-selected correlation spectroscopy, direct-correlation multiplicity-edited heteronuclear single quantum coherence, and long-range gradient Heteronuclear Mulitiple Bond Correlation Adiabatic, 500 MHz Varian NMR System, Palo Alto, CA).

Quantification of Radioactivity and Parent.

Radioactivity levels in plasma, blood, urine, and bile were determined by liquid scintillation counting; all other matrices were analyzed by combustion followed by liquid scintillation counting of the trapped 14CO2. Parent concentrations in plasma were determined with a validated LC-MS/MS assay. LY2090314 and its internal standard, D10-LY2090314, were extracted from rat, dog, or human plasma by liquid-liquid extraction; the extraction recovery was 34, 50, and 95% for rat, dog, and human plasma, respectively. After evaporation of the organic layer under nitrogen, the residue was reconstituted and analyzed. LY2090314 and its internal standard were eluted from a Kromasil Silica 50 × 3 mm, 5 μm column (Thermo Fisher Scientific, Inc., Waltham, MA) with a mobile phase gradient. Analytes were detected in positive ion mode using multiple reaction monitoring [Sciex API 3000 triple quadrupole mass spectrometer equipped with a TurboIonSpray interface (Applied Biosystems/MDS; Foster City, CA)]: LY2090314: 513.5 → 112.2 and D10-internal standard: 523.5 → 122.2 m/z. The dynamic range of the assays was 5–5000 ng/ml for LY2090314 in rat and dog plasma, and 0.5–250.0 ng/ml in human plasma.

Parent concentrations in human urine were determined with a validated LC-MS/MS assay. Samples were mixed with an organic internal standard solution and centrifuged, and the resulting supernatants were directly analyzed. LY2090314 and its internal standard were eluted from a Betasil C18 Dash HTS 20 × 2.1 mm, 5 μm column (Thermo Fisher Scientific) with a mobile phase gradient. Analytes were detected in positive ion mode using multiple reaction monitoring (Sciex API 4000 triple quadrupole mass spectrometer equipped with a TurboIonSpray interface): LY2090314: 513.0 → 112.0 and D10-internal standard: 523.3 → 122.1 m/z. The dynamic range of the assays was 1–5000 ng/ml.

Samples with LY2090314 concentrations above the upper limit of quantification were diluted with matrix to within the assay range; concentrations below the lower limit of quantification were reported as such.

Extraction of Preclinical Matrices for Metabolite Profiling.

Time-pooled samples were prepared for metabolite profiling by mixing equal volumes of plasma from animals at each time point (0.5, 1, 2, 3 and 6 hours for rats, 0.5 and 2 hours for dogs). Plasma samples underwent three extractions as follows: 1) 0.1% formic acid in acetonitrile; 2) and 3) 0.1% formic acid in 9:1 (v:v) acetonitrile:water. Extracts from each step were combined and centrifuged, and the resulting supernatant was evaporated. The dried residue was reconstituted in 60 µl of 42.5:18.5:38.5 (v:v:v) acetonitrile:formic acid:10 mM ammonium formate.

Radioprofiling of rat urine was not performed due to the low recovery of radioactivity in this matrix (< 2% of dose). A pooled 0–12-hour dog urine sample was prepared by mixing equal urine output fractions from the three dogs. An aliquot of pooled urine was diluted 1:1 (v:v) with 0.5% formic acid in 10 mM ammonium formate and centrifuged, and the supernatant used for analysis.

A pooled 0–12-hour rat bile sample was prepared by mixing equal bile output fractions (n = 3; one rat was excluded as an outlier by Dixon’s Q test due to low dose recovery). The pooled bile was diluted 1:4 (v:v) with 0.5% formic acid in 10 mM ammonium formate and centrifuged, and the supernatant used for radioprofiling.

The 0–24-hour rat fecal slurry was pooled from bile duct–intact rats by mixing equal fecal output fractions. Dog feces were pooled for the time intervals of 0–24 and 24–48 hours separately by mixing equal fecal output fractions. Fecal homogenates were sequentially extracted with 1) 0.2% formic acid in 9:1 (v/v) acetonitrile:water, 2) 0.1% formic acid in 9:1 (v/v) acetonitrile:water, 3) 0.1% formic acid in 9:1 (v/v) methanol:water, and 4) 0.1% formic acid in methanol. Solvents from extracts from each step were combined, then evaporated completely. Rat fecal extract residue was reconstituted with 300 μl of 0.5% formic acid in methanol and further diluted with 3 ml of 0.5% formic acid in 10 mM ammonium formate. Dog fecal extract residue was reconstituted with 0.35 ml of 0.1% formic acid in 1:1 acetontrile/methanol (v/v) and further diluted in 3.5 ml with 0.5% formic acid in 10 mM ammonium formate. Reconstituted fecal extracts were centrifuged, and the supernatant was used for metabolite profiling.

Metabolite Profiling in Preclinical Species.

We eluted 14C-analytes from a Zorbax SB-C18 4.6 × 150 mm column (5 µm particle size, 40°C; Agilent Technologies, Santa Clara, CA) with a mobile phase gradient. The aqueous mobile phase (A) was 0.5% formic acid in 10 mM ammonium formate in water, and the organic mobile phase (B) was 0.5% formic acid in methanol. The mobile phase flow rate was 1.0 ml/min, and the gradient was as follows [time (min)/B (%)]: 0/10, 2/10, 5/26, 48/26, 63/52, 68/7069/90, 81/90, 81.1/10, and 85/10%; except for rat plasma, for which the gradient was 0/10, 5/27, 40/27, 60/52, 65/70, 66/90, 70/90, 70.1/ 10, and 75/10%. The column eluate was split between the MS source and fraction collector. Radioprofiles were generated by scintillation counting of high-performance liquid chromatography fractions collected at 12-second intervals. Peaks in resulting radiochromatograms were expressed as the percentage of region of interest, with the sum of all integrated peaks defined as 100%. Recovery of radioactivity from the analytical column was 105.3% for plasma extract, 97.5% for bile, and 96.0% for urine. Metabolite identification was performed using MS, MS/MS, and Fourier-transformed MS. A ThermoFinnigan OrbiTrap LTQ mass spectrometer (Thermo Fisher Scientific) was used to perform the Fourier-transformed MS to acquire accurate mass data in electrospray ionization mode with positive ion detection (source voltage = 4.5 kV, 350°C, sheath gas = 60, auxiliary gas = 20, collision energy = 50%, and resolution = 30,000).

Metabolite Profiling in Transporter Knockout Rats.

Plasma was pooled across time points within each of the four groups (n = 4–6 rats/group), and a single 0–6-hour plasma Hamilton pool was prepared for each rat group (Hamilton et al., 1981). Plasma protein was precipitated with acetonitrile (1:4, v:v), and the resulting supernatant was dried and reconstituted in 1:9 (v:v) acetonitrile:water for LC-MS/MS analysis or in 1:9 (v:v) deuterated methanol:methanol for 19F-NMR. The pooled 0–6-hour bile and urine sample was prepared for LC-MS/MS analysis as described earlier (Metabolite Profiling in Preclinical Species). Profiling for LY2090314 metabolites in plasma, bile, and urine was conducted using the 75-minute gradient chromatography with detection as described earlier (Metabolite Profiling in Preclinical Species).

All NMR analyses were performed on an Avance III 600 MHz spectrometer (Bruker, Billerica, MA) equipped with a QCI HFCN quadrupole resonance CryoProbe. A standard 1D 19F pulse sequence with broadband 1H decoupling was used for all data acquisition. All data were processed identically using a 10-Hz line broadening.

Metabolite Profiling in Human Plasma and Urine.

Plasma was pooled across patients for the 1- and 4-hour time points separately at the 80- and 120-mg dose levels (separate pooling at each dose) by mixing equal volumes (n = 3–6 patients). Plasma proteins were precipitated with acetonitrile, and the supernatants were dried and reconstituted in 1:9 (v:v) acetonitrile:water. Urine was pooled across patients between 0–6 hours at the 80- and 120-mg dose levels (separate pooling at each dose) by mixing equal volumes (n = 3–5 patients). The pooled urine was concentrated 2-fold with a centrifugal evaporator and was analyzed without further processing. Profiling for LY2090314 metabolites in plasma and urine was conducted using the 75-minute gradient chromatography with detection as described earlier (Metabolite Profiling in Preclinical Species).

Data Analysis.

Rat and dog noncompartment pharmacokinetic parameters were calculated using WinNonlin v. 5.2 (Pharsight, Cary, NC). Clinical noncompartmental pharmacokinetic parameters were computed using WinNonlin Enterprise v. 5.3. All data are reported as mean ± S.D. unless otherwise noted.

Results

Rat Pharmacokinetics, Metabolism, and Excretion.

Rat pharmacokinetics and recovery of LY2090314-related radioactivity are presented in Fig. 1 and Table 1. The Tmax of both parent and radioactivity coincided with the end of the 30-minute drug infusion. LY2090314 concentrations declined rapidly in a monoexponential manner with a half-life of 21 minutes and were below the limit of quantification by the 4-hour time point. High clearance approximating hepatic blood flow (3300 ml/h/kg; Davies and Morris, 1993) and moderate volume of distribution resulted in rapid elimination of LY2090314. Hepatic blood flow-limited clearance despite the high extent of plasma protein binding (98–99%) highlights the magnitude of the high rat intrinsic clearance of LY2090314, consistent with rapid turnover in rat liver microsomes (Fig. 2; Pang and Rowland, 1977). Total radioactivity followed a biexponential decline with a steep alpha phase during the first 3 hours and a relatively longer terminal elimination half-life of 7.25 hours for the residual low levels of radioactivity, which were below the limit of quantification after 24 hours. Radioactivity was eliminated rapidly, with 87% dose recovery in excreta in the first 12 hours and 93% recovery in the first 24 hours. Elimination of radioactivity was predominantly by biliary excretion (86%) into the feces (97%), with little drug-related material recovered in urine (<2% of the dose).

Fig. 1.
Fig. 1.

Concentrations of total radioactivity in blood (▲) and plasma (Δ) and concentrations of the parent compound (○) in plasma after a single dose of [14C]LY2090314 (5 mg/kg; 50 µCi/kg) administered as a 30-minute intravenous infusion to male Sprague-Dawley rats (A and B) or Beagle dogs (C and D); mean ± S.D., n = 3. Parent plasma concentrations in cancer patients by dose group on linear scale after the first LY2090314 infusion (E), and semi-log plot of the pharmacokinetic profile after a 60-minute infusion of 80 mg of LY2090314 (F); mean ± S.D., n = 3–11.

Fig. 2.
Fig. 2.

Correlation between LY2090314 in vivo clearance and scaled in vitro clearance from liver microsomes. Note that in addition to the strong correlation, the in vitro clearance is quantitatively in agreement with the observed in vivo clearance, as indicated by the slope of approximately unity (slope = 1.29).

View this table:
TABLE 2

Parent LY2090314 systemic (plasma) pharmacokinetic and urinary excretion parameters in cancer patients following dose administration as a 60-minute intravenous infusion

At doses of 10 and 20 mg, LY2090314 was coadministered with pemetrexed/carboplatin on cycle 2, day 1; for all subsequent dose levels, LY2090314 was coadministered with pemetrexed/carboplatin on cycle 1, day 8.

Parent LY2090314 accounted for the majority of the total drug-related radioactivity systemic exposure: 70.4% in plasma and 83.4% in blood (in vitro LY2090314 blood-to-plasma partitioning ratio = 0.83 ± 0.04; in vivo total drug-related radioactivity blood-to-plasma ratio = 0.81 ± 0.26). Metabolite identification studies confirmed that parent LY2090314 was the sole identifiable circulating entity in rat plasma between 0.5 and 6 hours [Fig. 3; Supplemental Table 3 (see Supplemental Figs. 1 and 2 and Supplemental Tables 1 and 2 for a detailed summary of the metabolite identification data)]. Two radiolabeled peaks eluted after LY2090314, but they could not be identified due to matrix effects. Unknown metabolites 1 and 2 accounted for 2.2 and 4.1% of the systemic radioactivity exposure (2.5 and 4.8% of the parent exposure), respectively. The sum of other minor drug-related radioactivity peaks accounted for 6.7% of the radioactivity exposure (7.8% of the parent exposure).

Fig. 3.
Fig. 3.

Reconstructed radioprofile of pooled rat or dog plasma collected at 0.5 (A and B) and 2 hours (E and F) after the start of the 30-minute [14C]LY2090314 infusion, and accurate mass summed positive ion chromatogram of pooled human plasma collected at 1 (C and D) and 4 hours (G and H) after the start of the 60-minute of LY2090314 infusion. Note that the first time point in each species corresponds to the Tmax (end of infusion), and the second time point corresponds to at least one half-life after Tmax. See Supplemental Figures 1 and 2 and Supplemental Tables 1 and 2 for a detailed summary of the metabolite identification data.

LY2090314 was extensively metabolized in rats, with metabolites excreted into bile and recovered in feces [Supplemental Fig. 3; Tables 3 and 4 (see Supplemental Figs. 1 and 2 and Supplemental Tables 1 and 2 for a detailed summary of the metabolite identification data)]. In the 0–12-hour bile, which contained 85.2% of the dosed radioactivity, 34 metabolites were identified, accounting for 79.4% of the dose. The most abundant were M22 (P + 2O on the piperidine; 18.6% of the dose), M24 (P + O on the piperidine; 10.4% of the dose) and M5 (taurine conjugate of the piperidine ring-opened beta-oxidized acid; 6.8% of the dose). Additionally, 31 biliary metabolites individually accounted for <5% of the dose, and parent in bile accounted for <2%.

View this table:
TABLE 3

Biliary recovery of dose as parent and metabolites in bile duct-cannulated rats

See Supplemental Figs. 1 and 2 and Supplemental Tables 1 and 2 for detailed summary of metabolite identification data.

View this table:
TABLE 4

Fecal recovery of dose as parent and metabolites in bile-intact rats and dogs

See Supplemental Figs. 1 and 2 and Supplemental Tables 1 and 2 for a detailed summary of the metabolite identification data.

In the 0–24-hour feces from bile-intact rats, which contained 91.8% of the dosed radioactivity, the 24 identified metabolites largely paralleled those found in the bile and accounted for 74.9% of the dose [Supplemental Fig. 3; Table 4 (see Supplemental Figs. 1 and 2 and Supplemental Tables 1 and 2 for a detailed summary of the metabolite identification data)]. As in the bile, the most abundant were M24 (P + O on the piperidine; 21.7% of the dose), M22 (P + 2O on the piperidine; 13.1% of the dose), M5 (taurine conjugate of the piperidine ring-opened beta-oxidized acid, co-eluting with M4 in the feces; 6.1% of the dose). Additionally, 21 fecal metabolites each accounted for ≤5% of the dose, with parent in the feces accounting for <3%. Based on the comparable time course of biliary and fecal elimination of radioactivity as well as the negligible systemic metabolite exposure, the overall rat data indicate that LY2090314 metabolites are eliminated via biliary excretion into the feces without any appreciable reabsorption of metabolites from the intestine into systemic circulation.

Interestingly, in bile duct–cannulated rats 6.75% of the intravenous radioactive dose was recovered in the feces (Table 1). Although this was a minor route of elimination, it may be indicative of intestinal metabolism with lumenal secretion of the formed metabolites. Direct intestinal secretion of the parent is unlikely because fecal recovery of the parent in bile-intact rats (2.4%) was consistent with the extent of parent biliary excretion (1.8%). Intestinal metabolism/secretion may be associated with unique or more abundant fecal versus biliary metabolites. In fact, M7 (P + O + H2O; 5.1% of the dose in feces and 0.4% of the dose in bile) was more abundant in the feces, and M33 (P + O; 0.7% of dose) was a unique fecal metabolite (Tables 3 and 4). Although these observations are conceptually consistent with intestinal metabolism, they do not directly prove that M7 and M33 are formed in the intestine and are then secreted into the feces.

Comparative metabolite profiling studies were conducted across wild-type, Mdr1a-, Bcrp-, and Mrp2-knockout rats to determine whether the biliary excretion of metabolites may be attenuated in the absence of these canalicular transporters, resulting in exposure to circulating metabolites (Fig. 4). Across all the three knockout rats tested, parent remained the sole identifiable circulating drug-related moiety, and no additional metabolites were detected in plasma relative to wild-type controls by either mass spectrometry or 19F-NMR. Likewise, the urinary metabolite profile was unchanged. Biliary excretion of metabolites was not impaired in the knockouts; however, a general trend of shift toward greater abundance of more polar metabolites was observed in the bile. Most notably M24 (P + O on the piperidine; 10–22% of dose recovery in wild-type rats) was far less abundant in the bile of all three knockouts; instead, M21 (P + O–2H on the piperidine), M20 (P + O + glucuronide on the piperidine), M8 (P + O + H2O on the piperidine), and novel bis and triple oxidation products were more abundant.

Fig. 4.
Fig. 4.

Accurate mass summed positive ion chromatogram (A, C, D) and 19F-NMR spectra (B) of 0–6-hour Hamilton pooled plasma (A, B), 0–6-hour urine (C), and 0–6-hour bile (D) from wild-type (red), Mrp2 (green), Bcrp (purple), and Mdr1a (black) knockout male Sprague-Dawley rats following a single 5 mg/kg dose of LY2090314 administered as a 30-minute infusion.

Dog Pharmacokinetics, Metabolism, and Excretion.

Dog pharmacokinetics and recovery of LY2090314-related radioactivity are presented in Fig. 1 and Table 1. The Tmax of both parent and radioactivity coincided with the end of the 30-minute drug infusion. Parent LY2090314 concentrations declined rapidly in a monoexponential manner with a half-life of 40 minutes, and they were below the limit of quantification by the 8-hour time point. High clearance approximating hepatic blood flow in dogs (1854 ml/h/kg; Davies and Morris, 1993) and a moderate volume of distribution resulted in rapid systemic elimination of LY2090314. Hepatic blood flow-limited clearance despite the high extent of plasma protein binding (99%) highlights the magnitude of the high dog intrinsic clearance of LY2090314, consistent with rapid turnover in dog liver microsomes (Fig. 2; Pang and Rowland, 1977). Radioactivity was excreted rapidly, with 62% dose recovery in the first 24 hours and 88% in the first 48 hours. Total radioactivity followed a biexponential decline, with a steep alpha phase consistent with predominant recovery of radioactivity in the first 24 hours, and a long terminal elimination half-life of 73.6 ± 12 hours for the residual low levels of radioactivity, which were quantifiable through 120 hours. The intravenous radioactive dose was primarily recovered in the feces (69.2% ± 7.2% of the dose), and to a lesser extent in the urine (20.6% ± 6.6% of the dose).

Parent LY2090314 accounted for 21.4% (plasma) and 25.5% (blood) of the total drug-related radioactivity exposure (in vitro LY2090314 blood-to-plasma partitioning ratio = 0.83 ± 0.04; in vivo total drug-related radioactivity blood-to-plasma ratio = 0.64 ± 0.08). At Tmax, circulating radioactivity in the plasma was composed of parent (82.8%) and M39 (direct N-glucuronide; 12.7%), with nine other minor metabolites each accounting for <0.8% of the circulating radioactivity [Fig. 2; Table 5 (see Supplemental Figs. 1 and 2 and Supplemental Tables 1 and 2 for a detailed summary of the metabolite identification data)]. The radioactivity in plasma at 2 hours (2.3 parent half-lives after Tmax) was composed of parent (36.1%) and LY2090314 glucuronide (59.3% overall, with M38 40.5% and M39 18.8%), with five other minor metabolites each accounting for <1.8% of circulating radioactivity. Thus, at later times LY2090314 glucuronide was the major component of total circulating radioactivity, consistent with the relatively low contribution of parent to the total systemic radioactivity exposure. Dog urine contained 19.2% ± 6.5% of the radioactive dose, primarily due to excretion of the circulating M39 [15.6% of the dose; Supplemental Fig. 4; Table 6 (see Supplemental Figs. 1 and 2 and Supplemental Tables 1 and 2 for a detailed summary of the metabolite identification data)]. The other eight metabolites identified in the urine each accounted for ≤1% of the dose; only 0.4% of the administered LY2090314 was excreted unchanged in the urine. M39 was isolated from dog liver microsomes, and the site of direct N-glucuronidation was confirmed by NMR (Supplemental Table 2). Glucuronidation studies conducted in liver and kidney microsomes across species (see Metabolism in Microsomes) provided further support that N-glucuronidation of LY2090314 is a dog-specific hepatic pathway.

View this table:
TABLE 5

Dog plasma abundance of LY2090314 parent and metabolites

See Supplemental Figs. 1 and 2 and Supplemental Tables 1 and 2 for a detailed summary of the metabolite identification data.

View this table:
TABLE 6

Dog urine recovery of dose as parent and metabolites

See Supplemental Figs. 1 and 2 and Supplemental Tables 1 and 2 for a detailed summary of the metabolite identification data.

Following intravenous infusion, LY2090314-related radioactivity was primarily recovered in feces (69.2% ± 7.2% of the dose). In the 0–48-hour dog feces, which contained 65.8% of the radioactive dose, 21 metabolites were identified, accounting for 58.4% of the dose [Supplemental Fig. 3; Table 4 (see Supplemental Figs. 1 and 2 and Supplemental Tables 1 and 2 for a detailed summary of the metabolite identification data)]. The most abundant metabolites were M22 (P + 2O on the piperidine; 23.1% of the dose), M39 (direct N-glucuronide; 5.4% of the dose), and M34 (P + O + H2O, 5.2%). Dog feces also contained M40 (P + 3O + H2O; 4.8% of the dose), M4 (P − C5H8, 2.9%), M23 (P + H2O, 2.9%), 15 minor metabolites individually accounting for <2% of the dose, and parent LY2090314 (4.1% of the dose).

Cancer Patient Pharmacokinetics, Metabolism, and Excretion.

The pharmacokinetics in cancer patients after an approximately 60-minute intravenous infusion of LY2090314 (10–120 mg) are summarized in Fig. 1 and Table 2. As expected, Tmax coincided with the end of the 60-minute infusion in the majority of patients. After the infusion was stopped, LY2090314 concentrations declined very rapidly in the first hour, relatively less rapidly up to the 8-hour time point, and then more slowly up to the 24-hour time point in those patients where LY2090314 could still be measured at 24 hours (Fig. 1, E and F). LY2090314 concentrations were detectable up to the 8-hour sampling time point in the majority of patients, with approximately half the patients at different dose levels exhibiting quantifiable concentrations at 24 hours. The pharmacokinetics of LY2090314 were similar when LY2090314 was given alone or in combination with pemetrexed/carboplatin.

In cancer patients, the mean terminal elimination half-life estimates ranged from 1.84 to 3.35 hours (Table 1). High clearance and a moderate volume of distribution, both of which appeared independent of dose, resulted in rapid systemic elimination of LY2090314. The intravenous clearance values based on plasma exposure (on average 36–52 l/h) are similar to the hepatic plasma flow rate (49 l/h) and are 70–100% of the hepatic blood flow after correcting for the 0.60 ± 0.01 LY2090314 blood-to-plasma partitioning ratio (Davies and Morris, 1993). Perfusion-limited clearance despite the high extent of plasma protein binding (97–98% bound) highlights the magnitude of the high human intrinsic clearance of LY2090314, consistent with rapid turnover in human liver microsomes (Fig. 2; Pang and Rowland, 1977).

Metabolite profiling in cancer patients who received 80 or 120 mg of LY2090314 as an approximately 60-minute intravenous infusion showed that parent LY2090314 was the major circulating and sole identifiable species in plasma at both Tmax (end of infusion) and approximately one half-life after Tmax (Fig. 2). At the 80- and 120-mg dose levels, on average 1.6% ± 0.8% of the intravenous dose was recovered in the urine as the parent drug over 6 hours, a period representing on average 91% of the LY2090314 total systemic exposure (Table 2). Despite the low extent of LY2090314 urinary excretion, parent was the major peak in human urine (based on LC-MS ion intensity), in addition to 13 identified metabolites with a lesser relative abundance [Supplemental Fig. 4 (see Supplemental Figs. 1 and 2 and Supplemental Tables 1 and 2 for a detailed summary of the metabolite identification data)]. Of the detected metabolites, M10 (P + 2O – C2H6 with piperidine ring opening to aminopropionic acid), M4 (piperidine N-dealkylation), and M22 (P + 2O on the piperidine) were relatively most abundant, although less than parent which accounted for only 1.6% of the dose. Other metabolites in human urine included M61 as well as nine other metabolites, which did not produce a strong enough signal to register a visible peak on the LC-MS extracted ion chromatogram.

Metabolism in Microsomes.

LY2090314 was rapidly metabolized in mouse, rat, dog, monkey, and human liver microsomes (Fig. 2). Scaled microsomal clearance approached hepatic blood flow across species (scaled microsomal clearance as percentage of hepatic flow: mouse = 81.0% ± 4.3%, rat = 81.2% ± 7.5%, monkey = 90.8% ± 3.2%, dog = 81.1% ± 6.4%, human = 65.3% ± 3.7%). In addition to the strong correlation (r2 = 0.97) between in vivo and scaled in vitro clearance, the in vitro values were quantitatively in agreement with the observed in vivo values, as indicated by the slope of approximately unity (slope = 1.29).

LY2090314 glucuronidation was studied across rat, dog, and human liver and kidney microsomes. Direct N-glucuronidation was extensive in dog liver microsomes but was not observed in the other species or in kidney microsomes. To further rule out canine renal glucuronidation, a longer incubation was conducted at higher microsomal protein and cofactor concentrations in dog kidney microsomes, but LY2090314 glucuronidation was not observed even under these enhanced incubation conditions. The positive control, 4-methylumbelliferone, was glucuronidated by liver and kidney microsomes from all species.

Discussion

LY2090314, an intravenous GSK-3 inhibitor in clinical trials for the treatment of cancer, presents an interesting case study in rapid clearance by extensive metabolism with negligible circulating metabolite exposure. Across species, clearance approximated hepatic blood flow, despite extensive plasma protein binding (97–99%), highlighting the high magnitude of LY2090314 intrinsic clearance also observed in liver microsomes (Pang and Rowland, 1977). In conjunction with a moderate volume of distribution (approximately 1–2 l/kg across species), the high clearance resulted in rapid systemic elimination of LY2090314 with a half-life of 0.4 hours in rats, 0.7 hours in dogs, and a mean half-life of 1.8 to 3.4 hours in humans.

LY2090314 metabolites were extensively excreted into feces (69–97% of dose) via bile (86% of dose in rats). The comparable extent and time course of biliary and fecal metabolite recovery, as well as the negligible presence of circulating metabolites (except dog-specific LY2090314 glucuronide), leads to the conclusion that metabolites are not reabsorbed from the intestine. This apparent absence of reabsorption cannot be discussed in the context of high permeability and low solubility of the parent, because oxidative metabolites in feces are more polar, as evidenced by their earlier elution in reverse-phase chromatography. The solubilities of LY2090314 metabolites are almost certainly higher and permeabilities lower. Although direct measurements are lacking, the fact that metabolites are cleared by excretion is consistent with low permeability (Benet, 2010). In silico predictions for the parent and eight of the more prominent metabolites (M24, M22, M4, M5, M10, M40, M34, and M39) predicted high permeability/low solubility for LY2090314 (in agreement with experimental data), while all eight metabolites were predicted to have high solubilities, and six of the eight were predicted to have low permeability (M24 and M4 were intermediate). Furthermore, in silico predictions for P-glycoprotein (P-gp) transport predicted LY2090314 to be a substrate (in agreement with experimental data), and metabolites also were predicted to be P-gp substrates with high confidence, except M5 and M39, as expected of molecules with a sulfate or glucuronide functionality (Zamek-Gliszczynski et al., 2005, 2006a). Overall, the likelihood of metabolites having low permeability and efflux by P-gp is consistent with the absence of their reabsorption.

Biliary excretion of metabolites is unlikely to be passive, because 1) diffusion would also occur across the sinusoidal membrane, resulting in the presence of circulating metabolites and also likely leading to greater urinary recovery; 2) metabolites were predicted to have low passive permeability; and 3) the parent is a known, and the metabolites are predicted substrates for hepatic canalicular transporters, namely, P-gp and Bcrp. Because the biliary excretion of metabolites is conceptually consistent with an active process, we hypothesized that in Mdr1a-, Bcrp-, and/or Mrp2-knockout rats, biliary excretion of LY2090314 metabolites may be impaired, resulting in excretion of these metabolites across the sinusoidal membrane into the circulation, with potentially increased urinary recovery (Zamek-Gliszczynski et al., 2006b, 2009; Tian et al., 2007). However, across all three knockouts, the metabolites did not appear in circulation, and their urinary excretion was not enhanced, because the hypothesized impairment of biliary excretion did not materialize. In the knockout rats, a general trend of shift toward greater abundance of more polar metabolites in the bile was observed. Most notably, biliary excretion of M24 (P + O on the piperidine; 10–22% of dose recovery in wild-type rats) was markedly reduced, and instead a greater abundance of more polar secondary metabolites was observed in all three knockouts. These data suggest that when canalicular efflux is impaired, M24 is metabolized into more polar metabolites that are subsequently excreted into the bile instead of undergoing direct sinusoidal excretion into the circulation for subsequent renal clearance. Metabolites whose biliary excretion was unchanged in the knockouts may be explained by the presence of more than one canalicular transporter with high intrinsic clearance, such that knockout of any one transporter individually does not markedly alter metabolite disposition. This proposed explanation is consistent with the high efflux activity of the parent compound by both P-gp and Bcrp (asymmetry ratios = 108 and 46 in MDR1- and bcrp-MDCKII monolayers, respectively; unpublished data). The phenomenon of redundant canalicular excretion is not unprecedented; the biliary excretion of the dual P-gp/Mrp2 substrate [D-penicillamine-2,5]enkephalin is only sensitive to P-gp inhibition/knockout when Mrp2 is also impaired and P-gp modulation alone is insufficient to attenuate canalicular secretion (Chen and Pollack, 1998, 1999; Hoffmaster et al., 2004). Finally, it merits noting that the knockout rats used in this study did not demonstrate unexpected compensatory transport changes on a functional level (Zamek-Gliszczynski et al., 2012), and subsequent gene expression analysis of 137 absorption, distribution, metabolism, and excretion-relevant genes revealed that in the liver only catechol-O-methyltransferase was induced in Bcrp-knockout rats and that Mrp3 was induced in Mrp2-knockout rats (Zamek-Gliszczynski, expected publication in 2013).

Surprisingly, the rapid and extensive metabolic clearance of LY2090314 was not associated with appreciable exposure to circulating metabolites. Although dogs were exposed to both the parent and LY2090314 glucuronide in circulation, extensive direct glucuronidation of LY2090314 was unique to this species. In dogs, LY2090314 glucuronide was preferentially secreted across the hepatic sinusoidal membrane into the circulation for ultimate urinary excretion, as demonstrated by the 3:1 urine:feces recovery ratio (16 versus 5% of the dose). The recovery of dose-related radioactivity in dog urine was an order of magnitude greater than in the other species (21% versus ≤2% of the dose). However without LY2090314 glucuronide, the parent and the other eight metabolites together accounted for only 3.2% of the dose, bringing dog urinary recovery of radioactivity in line with that of rats and humans. Conjugation with glucuronic acid greatly increases polarity such that glucuronide conjugates exhibit poor passive membrane permeability and instead rely on active transport for excretion from their sites of formation (Zamek-Gliszczynski et al., 2006b). In the liver, glucuronide metabolites formed in hepatocytes can be either excreted across the canalicular membrane into bile by MRP2 and/or BCRP, or they can be excreted across the hepatic sinusoidal membrane into blood via MRP3 for ultimate renal clearance (Belinsky et al., 2005; van de Wetering et al., 2007; Zamek-Gliszczynski et al., 2008; Lagas et al., 2010; Zamek-Gliszczynski et al., 2011a). Predominant urinary excretion of LY2090314 glucuronide in dogs suggests greater hepatic sinusoidal versus canalicular efflux clearance for this metabolite.

LY2090314 was extensively metabolized via numerous pathways (Supplemental Fig. 2). The piperidine ring was a major site of metabolism across species, yielding mono- and di-oxygenated metabolites, ring opening and dealkylation products, and taurine conjugates. Direct N-glucuronidation of LY2090314 was a major and unique pathway in dogs. Other metabolic pathways included net addition of H2O, hydrolysis and/or reduction of the maleimide ring, and oxidation of the imidazopyridine and diazepine rings.

Although several compounds have advanced to clinical testing in the maleimide kinase inhibitor class, including bisarylmaleimides (“BAMs,” e.g., sotrastaurin) and bisindolylmaleimides (“BIMs,” e.g., enzastaurin and ruboxistaurin), drug disposition and metabolism data are only available for ruboxistaurin in the literature (Burkey et al., 2002; 2006; Barbuch et al., 2006). LY2090314 is similar to ruboxistaurin in its clearance by extensive metabolism and predominant biliary/fecal elimination. Across mice, rats, dogs, and humans, 83–90% of the ruboxistaurin oral dose was recovered in feces, with ≤4% recovery in urine; in bile-cannulated rats, 59–66% of the oral dose was recovered in the bile as metabolites (Burkey et al., 2002; 2006; Barbuch et al., 2006). However, unlike LY2090314, ruboxistaurin metabolites collectively circulate at high systemic exposures, with the N-desmethyl metabolite circulating at 1.2- to 6.4-fold the parent exposure across species (Burkey et al., 2002; Barbuch et al., 2006). As a result of the structural differences, ruboxistaurin produces more N-dealkylation metabolites and exclusively undergoes secondary glucuronidation once nitrogens are exposed or oxygens are added by phase I metabolism (Barbuch et al., 2006; Burkey et al., 2006). Furthermore, ruboxistaurin lacks the piperidine ring of LY2090314, so all the various piperidine oxidation metabolites, ring opening, and taurine conjugation are not observed for this molecule. Ruboxistaurin is nonetheless extensively metabolized and forms 29 unique metabolites, but due to the absence of the piperidine ring, the number of metabolites is approximately half of those identified for LY2090314 (Barbuch et al., 2006; Burkey et al., 2006).

The present study validates the choice of rats and dogs as toxicology species with respect to safety coverage of circulating drug-related moieties in humans (EMA, 2012; http://www.ema.europa.eu/docs/en_GB/document_library/Other/2011/07/WC500109298.pdf). Besides the parent compound, no metabolites were detected in human plasma that approached the >10% of total drug-related material threshold, which would trigger confirmation of exposure in nonclinical safety testing as well as toxicology studies for unique major circulating human metabolites. In contrast, the circulating LY2090314 glucuronide unique to dogs does not disqualify this nonrodent species for safety testing because dogs demonstrated sufficient parent-compound coverage.

In conclusion, LY2090314 is rapidly cleared by extensive metabolism, with biliary excretion of the metabolites into the feces. The parent was the predominant circulating moiety, except for LY2090314 glucuronide, which was unique to dogs. Negligible exposure to circulating metabolites is a result of the rapid biliary excretion of metabolites, which was not fundamentally affected by individual knockout of Mdr1a, Bcrp, or Mrp2, as well as the apparent lack of metabolite reabsorption from the intestine. Our study validates the choice of rats and dogs as toxicology species with respect to coverage of circulating drug-related moieties in humans.

Acknowledgments

The authors thank Elizabeth M. Peck for human renal clearance calculations, Richard D. Moulton for conducting the microsomal glucuronidation studies, Dr. Chad E. Hadden for NMR analysis of M39, Geri A. Sawada for examining LY2090314 flux in MDR1- and bcrp-MDCKII monolayers, Dr. Lian Wen (ABC Laboratories, Columbia, MO) for work in establishing the initial analytical methodologies that facilitated the characterization of LY2090314 metabolism, and Dr. Richard Burton (presently with Advion Bioanalytical Laboratories, a Quintiles Company, Indianapolis, IN) for critically reviewing the rat and dog metabolite identification reports.

Authorship Contributions

Participated in research design: Zamek-Gliszczynski, Abraham, Alberts, Kulanthaivel, Jackson, McCann, Anderson, Cassidy.

Conducted experiments: Abraham, Alberts, Kulanthaivel, Furr, Barbuch.

Contributed new reagents or analytic tools: Hu.

Performed data analysis: Zamek-Gliszczynski, Abraham, Alberts, Kulanthaivel, Jackson, Chow, McCann, Furr, Barbuch, Cassidy.

Wrote or contributed to the writing of the manuscript: Zamek-Gliszczynski, Abraham, Alberts, Kulanthaivel, Jackson, Chow, Anderson, Furr, Barbuch, Cassidy.

Footnotes

Abbreviations

Bcrp
breast cancer resistance protein
GSK-3
glycogen synthase kinase-3
LY2090314
3-[9-fluoro-2-(piperidin-1-ylcarbonyl)-1,2,3,4-tetrahydro[1,4]diazepino[6,7,1-hi]indol-7-yl]-4-imidazo[1,2-a]pyridin-3-yl-1H-pyrrole-2,5-dione
LC-MS
liquid chromatography–mass spectrometry
LC-MS/MS
liquid chromatography–tandem mass spectrometry
MS
mass spectrometry
m/z
mass-to-charge ratio
P-gp
P-glycoprotein
  • Received August 10, 2012.
  • Accepted January 10, 2013.

References

  1. ↵
    1. Barbuch RJ,
    2. Campanale K,
    3. Hadden CE,
    4. Zmijewski M,
    5. Yi P,
    6. O’Bannon DD,
    7. Burkey JL,
    8. Kulanthaivel P
    (2006) In vivo metabolism of [14C]ruboxistaurin in dogs, mice, and rats following oral administration and the structure determination of its metabolites by liquid chromatography/mass spectrometry and NMR spectroscopy. Drug Metab Dispos 34:213–224.
    OpenUrlAbstract/FREE Full Text
  2. ↵
    1. Belinsky MG,
    2. Dawson PA,
    3. Shchaveleva I,
    4. Bain LJ,
    5. Wang R,
    6. Ling V,
    7. Chen ZS,
    8. Grinberg A,
    9. Westphal H,
    10. Klein-Szanto A,
    11. et al.
    (2005) Analysis of the in vivo functions of Mrp3. Mol Pharmacol 68:160–168.
    OpenUrlAbstract/FREE Full Text
  3. ↵
    1. Benet LZ
    (2010) Predicting drug disposition via application of a Biopharmaceutics Drug Disposition Classification System. Basic Clin Pharmacol Toxicol 106:162–167.
    OpenUrlCrossRefPubMed
  4. ↵
    1. Boissan M,
    2. Beurel E,
    3. Wendum D,
    4. Rey C,
    5. Lécluse Y,
    6. Housset C,
    7. Lacombe ML,
    8. Desbois-Mouthon C
    (2005) Overexpression of insulin receptor substrate-2 in human and murine hepatocellular carcinoma. Am J Pathol 167:869–877.
    OpenUrlCrossRefPubMed
  5. ↵
    1. Brail LH,
    2. Gray JE,
    3. Burris HA,
    4. Simon GR,
    5. Cooksey J,
    6. Jones SF,
    7. Farrington DL,
    8. Lam T,
    9. Jackson KC,
    10. Brandt JT,
    11. Infante JR
    (2011) A phase I dose-escalation, pharmacokinetic (PK), and pharmacodynamic (PD) evaluation of intravenous LY2090314 a GSK3 inhibitor administered in combination with pemetrexed and carboplatin [abstr 3030]. J Clin Oncol 29(Suppl):15S.
    OpenUrlCrossRef
  6. ↵
    1. Burkey JL,
    2. Campanale KM,
    3. Barbuch R,
    4. O’Bannon D,
    5. Rash J,
    6. Benson C,
    7. Small D
    (2006) Disposition of [14C]ruboxistaurin in humans. Drug Metab Dispos 34:1909–1917.
    OpenUrlAbstract/FREE Full Text
  7. ↵
    1. Burkey JL,
    2. Campanale KM,
    3. O’Bannon DD,
    4. Cramer JW,
    5. Farid NA
    (2002) Disposition of LY333531, a selective protein kinase C beta inhibitor, in the Fischer 344 rat and beagle dog. Xenobiotica 32:1045–1052.
    OpenUrlCrossRefPubMed
  8. ↵
    1. Chen C,
    2. Pollack GM
    (1998) Altered disposition and antinociception of [D-penicillamine(2,5)] enkephalin in mdr1a-gene-deficient mice. J Pharmacol Exp Ther 287:545–552.
    OpenUrlAbstract/FREE Full Text
  9. ↵
    1. Chen C,
    2. Pollack GM
    (1999) Enhanced antinociception of the model opioid peptide [D-penicillamine] enkephalin by P-glycoprotein modulation. Pharm Res 16:296–301.
    OpenUrlCrossRefPubMed
  10. ↵
    1. Davies B,
    2. Morris T
    (1993) Physiological parameters in laboratory animals and humans. Pharm Res 10:1093–1095.
    OpenUrlCrossRefPubMed
  11. ↵
    1. Doble BW,
    2. Woodgett JR
    (2003) GSK-3: tricks of the trade for a multi-tasking kinase. J Cell Sci 116:1175–1186.
    OpenUrlAbstract/FREE Full Text
  12. ↵
    1. Grimes CA,
    2. Jope RS
    (2001) The multifaceted roles of glycogen synthase kinase 3beta in cellular signaling. Prog Neurobiol 65:391–426.
    OpenUrlCrossRefPubMed
  13. ↵
    1. Hamilton RA,
    2. Garnett WR,
    3. Kline BJ
    (1981) Determination of mean valproic acid serum level by assay of a single pooled sample. Clin Pharmacol Ther 29:408–413.
    OpenUrlCrossRefPubMed
  14. ↵
    1. Hoffmaster KA,
    2. Zamek-Gliszczynski MJ,
    3. Pollack GM,
    4. Brouwer KL
    (2004) Hepatobiliary disposition of the metabolically stable opioid peptide [D-Pen2, D-Pen5]-enkephalin (DPDPE): pharmacokinetic consequences of the interplay between multiple transport systems. J Pharmacol Exp Ther 311:1203–1210.
    OpenUrlAbstract/FREE Full Text
  15. ↵
    1. Jope RS,
    2. Yuskaitis CJ,
    3. Beurel E
    (2007) Glycogen synthase kinase-3 (GSK3): inflammation, diseases, and therapeutics. Neurochem Res 32:577–595.
    OpenUrlCrossRefPubMed
  16. ↵
    1. Kaidanovich O,
    2. Eldar-Finkelman H
    (2002) The role of glycogen synthase kinase-3 in insulin resistance and type 2 diabetes. Expert Opin Ther Targets 6:555–561.
    OpenUrlCrossRefPubMed
  17. ↵
    1. Klein PS,
    2. Melton DA
    (1996) A molecular mechanism for the effect of lithium on development. Proc Natl Acad Sci USA 93:8455–8459.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Lagas JS,
    2. Sparidans RW,
    3. Wagenaar E,
    4. Beijnen JH,
    5. Schinkel AH
    (2010) Hepatic clearance of reactive glucuronide metabolites of diclofenac in the mouse is dependent on multiple ATP-binding cassette efflux transporters. Mol Pharmacol 77:687–694.
    OpenUrlAbstract/FREE Full Text
  19. ↵
    1. Martin M,
    2. Rehani K,
    3. Jope RS,
    4. Michalek SM
    (2005) Toll-like receptor-mediated cytokine production is differentially regulated by glycogen synthase kinase 3. Nat Immunol 6:777–784.
    OpenUrlCrossRefPubMed
  20. ↵
    1. Pang KS,
    2. Rowland M
    (1977) Hepatic clearance of drugs. I. Theoretical considerations of a “well-stirred” model and a “parallel tube” model. Influence of hepatic blood flow, plasma and blood cell binding, and the hepatocellular enzymatic activity on hepatic drug clearance. J Pharmacokinet Biopharm 5:625–653.
    OpenUrlCrossRefPubMed
  21. ↵
    1. Poulin P,
    2. Hop CE,
    3. Ho Q,
    4. Halladay JS,
    5. Haddad S,
    6. Kenny JR
    (2012) Comparative assessment of In Vitro-In Vivo extrapolation methods used for predicting hepatic metabolic clearance of drugs. J Pharm Sci 101:4308–4326.
    OpenUrlCrossRefPubMed
  22. ↵
    1. Tan J,
    2. Zhuang L,
    3. Leong HS,
    4. Iyer NG,
    5. Liu ET,
    6. Yu Q
    (2005) Pharmacologic modulation of glycogen synthase kinase-3beta promotes p53-dependent apoptosis through a direct Bax-mediated mitochondrial pathway in colorectal cancer cells. Cancer Res 65:9012–9020.
    OpenUrlAbstract/FREE Full Text
  23. ↵
    1. Tian X,
    2. Li J,
    3. Zamek-Gliszczynski MJ,
    4. Bridges AS,
    5. Zhang P,
    6. Patel NJ,
    7. Raub TJ,
    8. Pollack GM,
    9. Brouwer KLR
    (2007) The role of P-gp, Bcrp and Mrp2 in the biliary excretion of spiramycin. Antimicrob Agents Chemother 51:3230–3234.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. van de Wetering K,
    2. Zelcer N,
    3. Kuil A,
    4. Feddema W,
    5. Hillebrand M,
    6. Vlaming ML,
    7. Schinkel AH,
    8. Beijnen JH,
    9. Borst P
    (2007) Multidrug resistance proteins 2 and 3 provide alternative routes for hepatic excretion of morphine-glucuronides. Mol Pharmacol 72:387–394.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    1. Wang Z,
    2. Smith KS,
    3. Murphy M,
    4. Piloto O,
    5. Somervaille TC,
    6. Cleary ML
    (2008) Glycogen synthase kinase 3 in MLL leukaemia maintenance and targeted therapy. Nature 455:1205–1209.
    OpenUrlCrossRefPubMed
  26. ↵
    1. Zamek-Gliszczynski MJ,
    2. Bedwell DW,
    3. Bao JQ,
    4. Higgins JW
    (2012) Characterization of SAGE Mdr1a (P-gp), Bcrp, and Mrp2 knockout rats using loperamide, paclitaxel, sulfasalazine, and carboxydichlorofluorescein pharmacokinetics. Drug Metab Dispos 40:1825–1833.
    OpenUrlAbstract/FREE Full Text
  27. ↵
    1. Zamek-Gliszczynski MJ,
    2. Day JS,
    3. Hillgren KM,
    4. Phillips DL
    (2011a) Efflux transport is an important determinant of ethinylestradiol glucuronide and ethinylestradiol sulfate pharmacokinetics. Drug Metab Dispos 39:1794–1800.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    1. Zamek-Gliszczynski MJ,
    2. Hoffmaster KA,
    3. Humphreys JE,
    4. Tian X,
    5. Nezasa K,
    6. Brouwer KL
    (2006a) Differential involvement of Mrp2 (Abcc2) and Bcrp (Abcg2) in biliary excretion of 4-methylumbelliferyl glucuronide and sulfate in the rat. J Pharmacol Exp Ther 319:459–467.
    OpenUrlAbstract/FREE Full Text
  29. ↵
    1. Zamek-Gliszczynski MJ,
    2. Hoffmaster KA,
    3. Nezasa K,
    4. Brouwer KL
    (2008) Apparent differences in mechanisms of harmol sulfate biliary excretion in mice and rats. Drug Metab Dispos 36:2156–2158.
    OpenUrlAbstract/FREE Full Text
  30. ↵
    1. Zamek-Gliszczynski MJ,
    2. Hoffmaster KA,
    3. Nezasa K,
    4. Tallman MN,
    5. Brouwer KL
    (2006b) Integration of hepatic drug transporters and phase II metabolizing enzymes: mechanisms of hepatic excretion of sulfate, glucuronide, and glutathione metabolites. Eur J Pharm Sci 27:447–486.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Zamek-Gliszczynski MJ,
    2. Hoffmaster KA,
    3. Tian X,
    4. Zhao R,
    5. Polli JW,
    6. Humphreys JE,
    7. Webster LO,
    8. Bridges AS,
    9. Kalvass JC,
    10. Brouwer KL
    (2005) Multiple mechanisms are involved in the biliary excretion of acetaminophen sulfate in the rat: role of Mrp2 and Bcrp1. Drug Metab Dispos 33:1158–1165.
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Zamek-Gliszczynski MJ,
    2. Kalvass JC,
    3. Pollack GM,
    4. Brouwer KL
    (2009) Relationship between drug/metabolite exposure and impairment of excretory transport function. Drug Metab Dispos 37:386–390.
    OpenUrlAbstract/FREE Full Text
  33. ↵
    1. Zamek-Gliszczynski MJ,
    2. Ruterbories KJ,
    3. Ajamie RT,
    4. Wickremsinhe ER,
    5. Pothuri L,
    6. Rao MV,
    7. Basavanakatti VN,
    8. Pinjari J,
    9. Ramanathan VK,
    10. Chaudhary AK
    (2011b) Validation of 96-well equilibrium dialysis with non-radiolabeled drug for definitive measurement of protein binding and application to clinical development of highly-bound drugs. J Pharm Sci 100:2498–2507.
    OpenUrlCrossRefPubMed
Back to top

In this issue

Download PDF
Article Alerts
Email Article
Citation Tools

Research ArticleArticle

PK, Metabolism, and Excretion of LY2090314

Maciej J. Zamek-Gliszczynski, Trent L. Abraham, Jeffrey J. Alberts, Palaniappan Kulanthaivel, Kimberley A. Jackson, Kay H. Chow, Denis J. McCann, Haitao Hu, Shelby Anderson, Nathan A. Furr, Robert J. Barbuch and Kenneth C. Cassidy
Drug Metabolism and Disposition April 1, 2013, 41 (4) 714-726; DOI: https://doi.org/10.1124/dmd.112.048488
Reddit logo Twitter logo Facebook logo Mendeley logo

Jump to section

Advertisement