Discovery of TAK-041: a Potent and Selective GPR139 Agonist Explored for the Treatment of Negative Symptoms Associated with Schizophrenia.

The orphan G-protein-coupled receptor GPR139 is highly expressed in the habenula, a small brain nucleus that has been linked to depression, schizophrenia (SCZ), and substance-use disorder. High-throughput screening and a medicinal chemistry structure-activity relationship strategy identified a novel series of potent and selective benzotriazinone-based GPR139 agonists. Herein, we describe the chemistry optimization that led to the discovery and validation of multiple potent and selective in vivo GPR139 agonist tool compounds, including our clinical candidate TAK-041, also known as NBI-1065846 (compound 56). The pharmacological characterization of these GPR139 agonists in vivo demonstrated GPR139-agonist-dependent modulation of habenula cell activity and revealed consistent in vivo efficacy to rescue social interaction deficits in the BALB/c mouse strain. The clinical GPR139 agonist TAK-041 is being explored as a novel drug to treat negative symptoms in SCZ.

Dose-related reduction in bupropion plasma concentrations by ritonavir.

The effect of repeat oral doses of ritonavir, at high (600 mg twice daily) and low (100 mg twice daily) doses, on the pharmacokinetics of a single dose of bupropion was evaluated in healthy volunteers. Subjects received a single dose of 150 mg of bupropion on day 1 and twice-daily ritonavir from day 8 through day 30. Ritonavir was up-titrated from 300 mg twice daily to 600 mg twice daily in the high-dose ritonavir study, whereas subjects remained on 100 mg twice-daily ritonavir in low-dose ritonavir study. Subjects received a second single dose of bupropion on day 24. Serial blood samples were obtained to evaluate the pharmacokinetics of bupropion and its metabolites on days 1 and 24. Steady-state ritonavir led to a decrease of area under the curve and maximum plasma concentration of bupropion by 62% to 67% in the high-dose study and by 21% to 22% in the low-dose study, indicating a drug interaction of statistical and clinical significance, particularly at high doses of ritonavir. These studies demonstrate that the reduction of bupropion exposure by ritonavir is dose-related. Dosage adjustment of bupropion may be needed when administered with ritonavir. However, the maximum recommended daily dose of bupropion should not be exceeded.

An in vitro mechanistic study to elucidate the desipramine/bupropion clinical drug-drug interaction.

There are documented clinical drug-drug interactions between bupropion and the CYP2D6-metabolized drug desipramine resulting in marked (5-fold) increases in desipramine exposure. This finding was unexpected as CYP2D6 does not play a significant role in bupropion clearance, and bupropion and its major active metabolite, hydroxybupropion, are not strong CYP2D6 inhibitors in vitro. The aims of this study were to investigate whether bupropion's reductive metabolites, threohydrobupropion and erythrohydrobupropion, contribute to the drug interaction with desipramine. In human liver microsomes using the CYP2D6 probe substrate bufuralol, erythrohydrobupropion and threohydrobupropion were more potent inhibitors of CYP2D6 activity (K(i) = 1.7 and 5.4 microM, respectively) than hydroxybupropion (K(i) = 13 microM) or bupropion (K(i) = 21 microM). Furthermore, neither bupropion nor its metabolites were metabolism-dependent CYP2D6 inhibitors. Using the in vitro kinetic constants and estimated liver concentrations of bupropion and its metabolites, modeling was able to predict within 2-fold the increase in desipramine exposure observed when coadministered with bupropion. This work indicates that the reductive metabolites of bupropion are potent competitive CYP2D6 inhibitors in vivo and provides a mechanistic explanation for the clinical drug-drug interaction between bupropion and desipramine.

Lysosomes contribute to anomalous pharmacokinetic behavior of melanocortin-4 receptor agonists.

PURPOSE: A series of melanocortin-4 receptor (MC4R) agonists, developed for use as anti-obesity agents, were found to have unusual pharmacokinetic behavior arising from excessive retention in the liver, with nearly undetectable levels in plasma following oral administration in mice. This work investigates the molecular basis of the prolonged liver retention that provided a rational basis for the design of an analog with improved behavior. MATERIALS AND METHODS: The livers of mice were harvested and techniques were utilized to fractionate them into pools differentially enriched in organelles. The distribution of organelles in the fractions was determined using organelle-specific enzymatic assays. Livers from mice dosed with drug were fractionated and comparisons with organelle distributions assisted in determining the subcellular localization of the drug. Further analysis in cell culture systems was used to confirm results from liver fractionation studies and also allowed for more extensive evaluations to examine the mechanism for organelle compartmentalization RESULTS: Fractionation of livers following oral administration of the agonist showed sequestration in lysosomes. Subsequent evaluations in a cell culture system confirmed this finding. Agents used to disrupt acidification of lysosomes led to decreased lysosomal accumulation of the drug, which implicated a pH-partitioning type sequestration mechanism. These findings led to the rational synthesis of an analog of the parent compound with properties that reduced lysosomal sequestration. When this compound was examined in mice, the liver retention was found to be greatly reduced and plasma levels were significantly elevated relative to the parent compound. CONCLUSIONS: Weakly basic drugs with optimal physicochemical properties can be extensively sequestered into lysosomes according to a pH-partitioning type mechanism. When administered orally in animals, this particular sequestration event can manifest itself in long term retention in the liver and negligible levels in blood. This work revealed the mechanism for liver retention and provided a rational platform for the design of a new analog with decreased liver accumulation and better opportunity for pharmacokinetic analysis and therapeutic activity.

Differences in the inhibition of cytochromes P450 3A4 and 3A5 by metabolite-inhibitor complex-forming drugs.

The objectives of this study were to characterize and compare the reversible inhibition and time-dependent inactivation of cytochromes P450 3A4 and 3A5 (CYP3A4 and CYP3A5) by erythromycin, diltiazem, and nicardipine. In the following experiments, we used cDNA-expressed CYP3A Supersomes and CYP3A-phenotyped human liver microsomes. We estimated the apparent constants for reversible inhibition (Ki(app) and IC50) and the irreversible kinetic constants (KI and kinact) for time-dependent inhibition. Based on an aggregate of Ki(app) and IC50 measurements, all inhibitors showed a greater inhibitory potency for CYP3A4 compared with CYP3A5. In addition, for each inhibitor, the kinact for CYP3A4 was approximately 4-fold higher than that for CYP3A5, indicating a greater propensity for time-dependent loss of CYP3A4 activity than of CYP3A5. Difference spectra experiments revealed an NADPH-dependent peak at approximately 455 nm [metabolite-inhibitor (MI) complex] following incubation of all three drugs with CYP3A4. There was no discernable MI complex formation following CYP3A5 incubation with any of the inhibitors. However, when CYP3A4 and CYP3A5 were incubated simultaneously with erythromycin, both enzymes appeared to contribute to the formation of a MI complex. Additional experiments revealed that erythromycin caused a comparable type I spectral change when bound to CYP3A5 and CYP3A4 (Ks=48 microM and 52 microM, respectively). Moreover, CYP3A5 exhibited only a moderately slower rate for the initial N-demethylation than did CYP3A4 (intrinsic clearance=41 versus 99 microl/min/nmol, respectively). In conclusion, erythromycin, diltiazem, and nicardipine were weaker inhibitors of CYP3A5 and inactivated the enzyme at a slower rate than their respective effects on CYP3A4. With respect to erythromycin, the failure of CYP3A5 to form a MI complex appears to be the result of slowed or impaired metabolic events downstream from the initial catalytic step, possibly due to a different orientation of the substrate molecule in the active site.

Evidence of significant contribution from CYP3A5 to hepatic drug metabolism.

CYP3A4 and CYP3A5 exhibit significant overlap in substrate specificity but can differ in product regioselectivity and formation activity. To further explore this issue, we compared the kinetics of product formation for eight different substrates, using heterologously expressed CYP3A4 and CYP3A5 and phenotyped human liver microsomes. Both enzymes displayed allosteric behavior toward six of the substrates. When it occurred, the "maximal" intrinsic clearance was used for quantitative comparisons. Based on this parameter, CYP3A5 was more active than CYP3A4 in catalyzing total midazolam hydroxylation (3-fold) and lidocaine demethylation (1.4-fold). CYP3A5 exhibited comparable metabolic activity as CYP3A4 (90-110%) toward dextromethorphan N-demethylation and carbamazepine epoxidation. CYP3A5-catalyzed erythromycin N-demethylation, total flunitrazepam hydroxylation, testosterone 6beta-hydroxylation, and terfenadine alcohol formation occurred with an intrinsic clearance that was less than 65% that of CYP3A4. Using two sets of human liver microsomes with equivalent CYP3A4-specific content but markedly different CYP3A5 content (group 1, predominantly CYP3A4; group 2, CYP3A4 + CYP3A5), we assessed the contribution of CYP3A5 to product formation rates determined at low substrate concentrations (< or = Km). Mean product formation rates for group 2 microsomes were 1.4- to 2.2-fold higher than those of group 1 (p < 0.05 for 5 of 8 substrates). After adjusting for CYP3A4 activity (itraconazole hydroxylation), mean product formation rates for group 2 microsomes were still significantly higher than those of group 1 (p < 0.05 for 3 substrates). We suggest that, under conditions when CYP3A5 content represents a significant fraction of the total hepatic CYP3A pool, the contribution of CYP3A5 to the clearance of some drugs may be an important source of interindividual variability.

Integrating in vitro kinetic data from compounds exhibiting induction, reversible inhibition and mechanism-based inactivation: in vitro study design.

Drug:drug interactions continue to be an obstacle for the pharmaceutical industry in the development of potential drug candidates. Considering the number of compounds that have been withdrawn from the market due to drug:drug interactions (e.g. cisapride, terfenadine and mibefradil), more pressure is placed on the pharmaceutical industry to investigate potential interactions prior to regulatory submission. In particular, induction and inhibition of drug metabolizing enzymes can profoundly alter the pharmacological and toxicological effects observed during monotherapy. However, due to differences in the expression and regulation of both metabolic enzymes and nuclear receptors responsible for induction, in vivo studies with pre-clinical species are not predictive of the human clinical situation. Although in vitro kinetic data also have limitations when extrapolating in vivo, in vitro testing has become more commonplace due to reduced cost and higher throughput. However, in the in vitro setting, complex enzyme kinetics can alter the estimation of kinetic parameters. Time-dependent or non-Michaelis-Menten kinetics can alter parameter estimates if experimental conditions are not optimal, and can therefore confound clinical predictions. Furthermore, mechanism-based inactivation (MBI) will reduce the active enzyme pool, both in vitro and in vivo, and thus complicate any parameter estimates. To further complicate matters, some compounds (e.g., ritonavir) inhibit, induce, as well as cause mechanism-based enzyme inactivation. For compounds such as ritonavir, the accurate estimation of kinetic parameters requires optimal experimental design at a minimum. This review will highlight the challenges in estimating enzyme kinetic parameters when both inhibition and induction are present, and will offer experimental viewpoints for the optimization of the experimental conditions.