Metopimazine is primarily metabolized by a liver amidase in humans

Abstract Metopimazine (MPZ) is a peripherally restricted, dopamine D2 receptor antagonist used for four decades to treat acute nausea and vomiting. MPZ is currently under clinical investigation for the treatment of gastroparesis (GP). MPZ undergoes high first‐pass metabolism that produces metopimazine acid (MPZA), the major circulating metabolite in humans. Despite a long history of use, the enzymes involved in the metabolism of MPZ have not been identified. Here we report a series of studies designed to identify potential MPZ metabolites in vitro, determine their clinical relevance in humans, and elucidate the enzymes responsible for their formation. The findings demonstrated that the formation of MPZA was primarily catalyzed by human liver microsomal amidase. Additionally, human liver cytosolic aldehyde oxidase (AO) catalyzes the formation of MPZA, in vitro, although to a much lesser extent. Neither cytochrome P450 enzymes nor flavin‐monooxygenases (FMO) were involved in the formation MPZA, although two minor oxidative pathways were catalyzed by CYP3A4 and CYP2D6 in vitro. Analysis of plasma samples from subjects dosed 60 mg of MPZ verified that these oxidative pathways are very minor and that CYP enzyme involvement was negligible compared to microsomal amidase/hydrolase in overall MPZ metabolism in humans. The metabolism by liver amidase, an enzyme family not well defined in small molecule drug metabolism, with minimal metabolism by CYPs, differentiates this drug from current D2 antagonists used or in development for the treatment of GP.


Metopimazine (MPZ) is an approved drug in France under the brand
name Vogalene ® (metopimazine, free base) that has been used for many years for the short-term treatment of nausea and vomiting. A new salt formulation of MPZ is in clinical development in the US for the treatment of gastroparesis, a chronic disorder of the stomach characterized by delayed gastric emptying without evidence of mechanical obstruction. 1-3 MPZ, a phenothiazine derivative, is a potent D2/D3 selective, dopamine receptor antagonist that is peripherally restricted. 2 The bioavailability of MPZ in humans is low. A 10 mg dose of MPZ was reported to have an absolute bioavailability under 20%. 4 MPZ does not penetrate the blood-brain barrier and is subject to high first pass metabolism in humans. 4,5 While much is known about the pharmacology of MPZ, the metabolism of this drug is not fully elucidated. MPZ is deaminated to form metopimazine acid (MPZA), the major circulating metabolite, present at much higher plasma concentrations than the parent. At clinical doses of 20 and 50 mg of MPZ, MPZA accounted for 80% or more of the circulating drug-related material in plasma following a single dose taken either preprandially or postprandially. 4 This biotransformation is clearly the predominant metabolic pathway and yet the enzyme(s) responsible for this deamination reaction have not been determined. This metabolite retains some pharmacologic activity but is about 200-fold less potent at the human D2 receptor compared to the parent. 2 This underscores the importance of understanding this particular metopimazine metabolic transformation. Drugs or polymorphisms that could affect the enzymes involved in this biotransformation could significantly affect circulating metopimazine concentrations, which could affect both safety and efficacy. For example, phenothiazines at high enough doses, have been linked to drug-induced orthostatic hypotension. 6 Given the renewed interest in exploring the potential of MPZ for new clinical indications, identifying the enzyme(s) responsible for the MPZ to MPZA conversion is important to understanding any potential impact of metabolic variability on efficacy and safety risks and in assessing drug-drug interaction potential.
In this paper, we describe the multidisciplinary approach designed to determine the main human metabolites and elucidate the enzymes involved in MPZ metabolism. This entailed a thorough in vitro characterization of metopimazine metabolites derived from cellular and in vitro incubations as well as exploring the clinical relevance by looking for these metabolites in samples of healthy volunteers who were administered a 60 mg single dose of metopimazine.
The limited role of cytochrome P450s in the metabolism of metopimazine has been more thoroughly defined. Most importantly, using a series of recombinant enzymes, inhibitor phenotyping and a variety of conditions, we identified the enzymes involved in the first pass metabolism of MPZ to form MPZA.

| MATERIAL S AND ME THODS
The following experiments were performed in an exploratory manner. MPZ

| Clinical study
The clinical study was conducted at Biotrial Rennes, in accordance with the Declaration of Helsinki. The study (NG100-101) was a double-blind, placebo-controlled, ascending single-and multipledose study of the safety, tolerability, and pharmacokinetics of metopimazine administered orally to healthy adult subjects. The study protocol, informed consent form, and appropriate related documents were approved by the institutional review board, and all subjects provided written informed consent prior to participation in the study.
In this study, sequential ascending single doses were administered to 4 cohorts of 8 participants (6 actives and 2 placebos per cohort, in cohorts 1, 2 and 3 following a randomization 6:2 and in cohort 4 with a 1:1 randomization for the two sentinel participants).
In the morning, after at least 10 h of fasting overnight, eight healthy adult subjects (male or female aged 18 to 45), were orally adminis- plasma samples were obtained by centrifugation of the blood samples. A comprehensive safety and pharmacokinetic analysis was performed for each cohort for the SAD and MAD parts of the study and will be presented elsewhere. A fit-for-purpose discovery method was used to analyze 4 subjects from the highest single dosed cohort for the presence of metabolites.
Plasma concentrations of metopimazine, metopimazine acid and metopimazine sulfoxide were quantified using LC-MS/MS methods.

| Rat PK study
The pharmacokinetics of metopimazine (parent) and metopima-

| Incubations with HLC fractions for AO activity
The metabolism of MPZ was evaluated in the human liver cytosolic incubation. A metabolic incubation was performed with HLC (1 mg/ml protein) and 10 µM MPZ at 37°C for on an orbital shaker for 60 min. A parallel set of incubations were also conducted 10 µM MPZ in the presence of 100 µM menadione, a known inhibitor of human aldehyde oxidase. Another set of incubation was also con-

| Data analysis
The degree of disappearance of each analyte was evaluated by compar-  were available for quantification, while MPZH could only be semiquantified using peak area ratio due to lack of a synthetic standard.  Table 2. As expected, the major circulating metabolite was MPZA, with Tmax ranging from 1.25-3 h and this analyte was present at all time points. MPZS was also found in all 4 subjects, though at much lower concentrations. It appeared at 30-45 min and was not quantifiable by 12 h. A representative chromatogram is shown in Figure 1. There was a very small isobaric peak that eluted just after 2 min (see arrow in Figure 1) in some of the samples. While it is speculative, this could potentially be the MPZH metabolite, albeit present at only trace amounts and unlikely to be quantifiable even if a standard could be made. A rough estimate of percent circulating drug-related material of the verified in vivo metabolites based on AUClast is shown in Table 3. Based on this, MPZA accounts for approximately 90% of circulating material whereas the parent is less than 10%. Also, MPZS is a minor metabolite present at less than 1% of total circulating drug-related material.

| Nonclinical pharmacokinetics
Given that MPZA was the predominant circulating metabolite in humans, bioanalytical methods were developed to quantify both parent and the major human metabolite. To determine the pharmacokinetic properties of metopimazine and its acid metabolite in rats and dogs, single doses of metopimazine (free base) were administered orally to each species and plasma samples were taken over time and analyzed. A noncompartmental analysis of the bioanalytical data was performed and the pharmacokinetic parameters determined. Table 4 shows the mean PK exposure parameters in rats and

dogs.
Interestingly, the MPZA metabolite to parent ratio is far lower in dogs and particularly lower in rats than in humans, indicating MPZA is disproportionately produced in humans.

| Cytochrome P450 involvement
To determine the involvement of cytochrome P450 enzymes, MPZ was incubated in HLM at 1 and 10 μM in the presence and absence of 1-ABT, a known non-specific CYP inhibitor, and with and without NADPH cofactor addition to the incubation mixture. As shown Figure 2A, the presence of 1-ABT in the incubation inhibited the formation of MPZS but had no effect on MPZA. In addition, the absence of NADPH also did not reduce the formation of MPZA, but    (Figure 6). In addition, bis-pnitrophenyl phosphate (BNPP), an amidase inhibitor, was also added to some of the incubations. When this amidase inhibitor was present, at least 97% of all activity was inhibited at 2 μM, even at the highest and most active pH conditions. When this inhibitor concentration was increased to 16 μM, >99% of MPZA formation was inhibited at all three pH values. Taken together, these results are consistent with MPZA being formed by a human microsomal liver amidase.

| Enzymes involved in formation of MPZA
The formation of MPZA was also observed in the incubation of 10 μM MPZ in human liver cytosol (HLC). HLC catalyzed the formation of some MPZA over 60 min of incubation, which was inhibited (48%) when 100 μM menadione, a known inhibitor of human AO, was included in the incubation (Table 5). These findings suggests that human liver cytosolic aldehyde oxidase is a potential contributor to the MPZA pathway.

| DISCUSS ION
Clinical pharmacokinetic studies with MPZ have demonstrated that MPZA is the major in vivo metabolite in humans and based on AUC exposure, it accounts for over 90% of circulating drug-related material. 9 MPZA is a disproportionate metabolite, produced in far lower Importantly, no CYPs are involved with MPZA formation.
In this study, MPZ was shown to undergo hydrolysis in human liver microsomes to form MPZA. This conversion of MPZ to MPZA is not inhibited by any individual CYP inhibitor, including ABT, a nonspecific inhibitor of CYPs. This activity is not dependent on NADPH and increases under conditions above physiologic pH. Amidase/hydrolases, such as fatty acid amide hydrolase (FAAH) are known to have optimal activity in the pH range of 8.5-10. [12][13][14] In HLM, the microsomal conversion of MPZ to MPZA was maximal at the highest pH tested, pH 9.4. These characteristics clearly point to a liver amidase as the enzyme responsible for MPZA formation. Consistent with these findings, the addition of BNPP, a known amidase inhibitor, 15,16 dramatically inhibited the formation of MPZA in HLM.
There are over 100 human genes that encode hydrolases that include esterases, amidases, peptidases and other enzymes. 17 For prodrugs, hydrolytic activation by non-CYP enzymes is the norm. 18,19 In an examination of 22 pro-drugs approved by the FDA between 2006 and 2015, 19 underwent metabolic bioactivation via a "hydrolase," which encompassed the broad category of esterases, amidases and other hydrolytic activities. 18 When prodrugs are excluded, however, non-CYP metabolism is not as common as biotransformation by CYP enzymes. Looking at this same time period but excluding prodrugs, 30% of the 125 FDA-approved drugs were metabolized by non-P450 enzymes compared to over half that were. 18 Not surprisingly, conjugation by uridine 5'-diphospho-glucuronosyltransferases (UGTs) was the top non-CYP activity but it was followed surprisingly closely by hydrolase activity (11.7% vs. 10.8%, for UGTs and hydrolases, respectively). Of the non-prodrugs that were metabolically hydrolyzed, amidase activity accounted for 84%. 18 Amidases, a class of hydrolase enzymes, are ubiquitous and their functions vary widely. 20,21 They hydrolyze a wide variety of amides including short-and mid-chain aliphatic amides, arylamides, α-aminoamides, and α-hydroxyamides. The catalytic mechanism of amidases involves the formation and hydrolysis of a covalent intermediate and adds water to the substrate without requiring additional cofactors. Amidases characteristically resist denaturation at higher pHs and temperatures, in stark contrast to CYP enzymes, likely due to the compact multimeric structure of amidases. Amidases such as Fatty Acid Amide Hydrolase (FAAH) are known to hydrolyze endogenous substrates such as endocannabinoid anandamide and oleamide. 22 In vitro, FAAH can also cleave a variety of lipid amides 23 and inhibition of FAAH-1 has been used as a therapeutic strategy. 24 However, amidases are not well defined in drug metabolism, despite the many amide-containing drugs that undergo hydrolysis. 21 While liver amidase conversion appears to be the primary biotransformation enzyme involved in MPZ metabolism, another enzyme that may contribute to the formation of MPZA is the cytosolic enzyme AO. We observed the conversion of MPZ to MPZA in HLC, though to a much lesser extent than HLM. This activity was moderately inhibited by menadione, a known inhibitor of AO, suggesting the possible contribution of AO to MPZ metabolism. AO is a metalloenzyme that contains molybdenum cofactor and typically shows hydroxylase activity. 25,26 It is most extensively expressed in the liver but also in the GI tract as well as the kidney, lungs, and skin. 27 Abbreviation: BQL, below quantifiable limit.

TA B L E 5
Concentrations of metopimazine acid in pooled human cytosolic incubations without and with menadione (AO Inhibitor) azaheterocycles to lactams. 29 AO is involved in the metabolism of several clinically significant drugs such as famciclovir, zaleplon, zonisamide, and ziprasidone. 17 Association of AO with pathophysiology of a number of clinical disorders such as amyotrophic lateral sclerosis and alcohol induced liver injury, has been suggested. 30 There is at least one published case of AO having amide hydrolase activity.
Amide hydrolysis is the primary human metabolic clearance pathway for GDC-0834, a Bruton's tyrosine kinase inhibitor; however, it is produced in much lower quantities in preclinical species. 31 In humans, this amide hydrolase is catalyzed by AO. 32 Although the hydrolysis of the amide bond in metopimazine appears to be catalyzed to a greater degree by a microsomal liver amidase, cytosolic AO can catalyze this reaction in vitro. While AO is known to catalyze a wide variety of biotransformations that include both oxidase activity and reductase activity, 30 this is the second example in recent years of an atypical AO-catalyzed amide hydrolysis of a small molecule and may indicate an expanding role for AO in xenobiotic metabolism.
We have demonstrated that the formation of MPZA from MPZ is primarily catalyzed by human liver amidase, with the potential for contribution from human liver cytosolic aldehyde oxidase, however to a much

ACK N OWLED G EM ENTS
None.

D I SCLOS U R E
Dr. Busby has received consulting fees from Neurogastrx. Drs. De Colle and Wax have stock options in Neurogastrx.

E TH I C S S TATEM ENT
This paper is the original and authentic work of the authors. All authors read and approved the final manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.