Identifying the Reactive Metabolites of Tyrosine Kinase Inhibitor Pexidartinib In Vitro Using LC–MS-Based Metabolomic Approaches

Pexidartinib (PEX, TURALIO), a selective and potent inhibitor of the macrophage colony-stimulating factor-1 receptor, has been approved for the treatment of tenosynovial giant cell tumor. However, frequent and severe adverse effects have been reported in the clinic, resulting in a boxed warning on PEX for its risk of liver injury. The mechanisms underlying PEX-related hepatotoxicity, particularly metabolism-related toxicity, remain unknown. In the current study, the metabolic activation of PEX was investigated in human/mouse liver microsomes (HLM/MLM) and primary human hepatocytes (PHH) using glutathione (GSH) and methoxyamine (NH2OMe) as trapping reagents. A total of 11 PEX-GSH and 7 PEX-NH2OMe adducts were identified in HLM/MLM using an LC–MS-based metabolomics approach. Additionally, 4 PEX-GSH adducts were detected in the PHH. CYP3A4 and CYP3A5 were identified as the primary enzymes responsible for the formation of these adducts using recombinant human P450s and CYP3A chemical inhibitor ketoconazole. Overall, our studies suggested that PEX metabolism can produce reactive metabolites mediated by CYP3A, and the association of the reactive metabolites with PEX hepatotoxicity needs to be further studied.


INTRODUCTION
Pexidartinib (PEX, TURALIO), a novel small molecule tyrosine kinase inhibitor, has highly selective and potent inhibitory activity against macrophage colony-stimulating factor-1 receptor (CSF1R). 1,2 It has been approved by the Food and Drug Administration (FDA) for the treatment of tenosynovial giant cell tumor in adults who are not likely to benefit from surgery. 3−6 PEX is also under investigation for its potential as a monotherapy or combination therapy in various malignancies. 5,7,8 A recent study indicates that PEX also holds great promise as a treatment for anaplastic thyroid cancer. 9 Generally, PEX has shown efficacy and tolerability in patients. 7,10 However, serious adverse reactions caused by PEX were reported in 13% of patients, including lifethreatening hepatotoxicity occurring in 3.3% of patients. 4,11 Due to the potential for severe and even fatal liver injury, the FDA has issued a boxed warning for its hepatotoxicity. PEX is available to patients only through a restricted program under a Risk Evaluation and Mitigation Strategy, 4 and it is essential to monitor liver function prior to the initiation of PEX treatment. 10,12 To date, the mechanism(s) of PEX-related toxicity remains mostly unidentified. Drug metabolism is closely connected to both drug efficacy and adverse effects. 13,14 It is well appreciated that reactive metabolites play a critical role in idiosyncratic adverse drug reactions. 14,15 According to FDA documents, PEX is rapidly metabolized, primarily by the enzymes CYP3A and UGT1A4, which are involved in phase I and phase II metabolism, respectively. 12 PEX was recovered in the feces as the unchanged form (44%) and in the urine (27%) as metabolites (≈10% as N-glucuronide). 3,5 Our previous studies focused on the carbon−carbon bond cleavage in PEX metabolism, provided a comprehensive profile of the phase I metabolism of PEX in human liver microsomes (HLM), and identified its stable metabolites. 16 However, it is still unclear whether PEX metabolism is capable of generating reactive metabolites and what specific roles they may play in causing its severe hepatotoxicity.
In the present study, we employed LC−MS-based metabolomic approaches to investigate the metabolic bioactivation of PEX in in vitro systems. We used reduced glutathione (GSH) and methoxyamine (NH 2 OMe) as trapping reagents in HLM and mouse liver microsomes (MLM). 17−20 The drug metabolizing enzymes contributing to the generation of PEX reactive metabolites were determined using recombinant human P450s and a specific chemical inhibitor in HLM/MLM and primary human hepatocytes (PHH). Our studies revealed 11 PEX-GSH and 7 PEX-NH 2 OMe adducts in HLM/MLM. Four PEX-GSH adducts were observed in PHH treated with PEX. These findings from this study may facilitate the understanding of PEX hepatotoxicity from the perspective of drug metabolism.  pmol of each cDNA-expressed human P450 enzymes (control, CYP1A2, 2A6, 2B6, 2C8, 2C9, 2C19, 2D6, 2E1, 3A4, and 3A5) and 2.5 mM GSH or 2.5 mM NH 2 OMe in a final volume of 95 μL. After a 5 min preincubation at 37°C, 5 μL of 20 mM NADPH (final concentration: 1.0 mM) was added to initiate the reactions. The incubation continued for 40 min at 37°C with gentle shaking. Incubations without NADPH or GSH/NH 2 OMe were used as controls. Co-incubations of PEX (30 μM) and KCZ (human and mouse CYP3A inhibitor, 2 μM) in HLM and MLM with GSH or NH 2 OMe were performed to determine the role of CYP3A in the formation of GSH or NH 2 OMe adducts related to PEX. Reactions were quenched by adding 100 μL of ice-cold acetonitrile and vortexing for 30 s, and the mixtures were then centrifuged at 15,000 rcf for 15 min. The supernatant was transferred to an autosampler vial, and 3 μL was injected into an ultrahigh-performance liquid chromatography (UHPLC) Q Exactive MS system for analysis. Incubations were performed in duplicate for cDNA-expressed P450 enzymes and in triplicate for LM experiments. The PEX concentration (30 μM) used in liver microsomes is clinically relevant according to the C max in human subjects.

Materials and
The average plasma C max of PEX is around 4 μg/mL (∼9.6 μM) in human subjects at 400 mg per day. 21 Typically, the concentration of drugs in the liver is several times greater than that in the bloodstream.
2.3. Identifying the Role of CYP3A in the Formation of PEX-GSH Adducts in PHH. PHH (Xenotech, Kansas City, KS, 20 donors, Cat No. HPCH20-50, Batch No. 1910146) were thawed and plated in 12-well plates (Corning, Corning, NY) according to the protocol of the vendor with the density of 6.95 × 10 5 cells/well. The cells were cultured in complete HepatoZYME medium, containing GlutaMAX, Insulin-Transferrin-Selenium, and penicillin/streptomycin (all of the reagents were obtained from Thermo, San Jose, CA) at 37°C in a humidified atmosphere with 5% CO 2 for 24 h before PEX treatment. The PHH were treated with 20 μM PEX with or without 10 μM KCZ for 6 h. The medium was then transferred into Eppendorf tubes and  Table 1. (C) Relative abundance of GSH adducts associated with PEX in HLM and MLM. The relative quantification of trapped metabolites was conducted based on the peak area. The overall abundance of GSH adducts in each sample was set as 100%. The data are expressed as mean ± S.E.M (n = 3). (D) Relative abundance of PEX-GSH adducts in PHH lysate and medium. PHH seeded in 12-well plates were treated with 20 μM PEX for 6 h. The incubated samples, PHH medium, and cell lysates samples were all analyzed using the UHPLC-Q Exactive MS system to measure PEX-GSH adducts. The overall abundance of GSH adducts in each sample was set as 100%. The data are expressed as mean ± S.E.M (n = 4). PEX, pexidartinib; GSH, reduced glutathione; 1× PBS, 1× phosphate-buffered saline; HLM/MLM, human/mouse liver microsomes; OPLS-DA, orthogonal projection to latent structures-discriminant analysis; PHH, primary human hepatocytes; KCZ, ketoconazole.   Figure S1 for cleaved moieties.

UHPLC-MS Analyses.
PEX and its GSH or NH 2 OMe adducts in the samples were resolved, analyzed, and relatively quantified by a UHPLC-Q Exactive MS system (Thermo Fisher Scientific, San Jose, CA). The column for analyte separation was an Acquity 100 mm × 2.1 mm BEH C-18 column (1.7 μm, Waters, Milford, MA), whose temperature was maintained at 40°C. The flow rate was 0.3 mL/min with gradient ranging from 2 to 95% in the water-acetonitrile mobile phase system both containing 0.1% formic acid in a 15 min run. Q Exactive MS was operated in positive mode with electrospray ionization. Ultrapure nitrogen was applied as the sheath (45 arbitrary unit), auxiliary (10 arbitrary unit), sweep (1.0 arbitrary unit), and the collision gas. The capillary gas temperature was set at 350°C, and the capillary voltage was set at 4.3 kV. MS data were acquired from 80 to 1200 Da in the profile mode. The ion at m/ z 371.1012 was used as a reference for positive mode during acquisition. The MS/MS of GSH and NH 2 OMe adducts associated with PEX was performed with the normalized collision energy set at 10−35 arbitrary units with an isolation width of 2 m/z.

Data Analysis.
Chromatograms and mass spectra from m/z 80 to 1200 were acquired in profile format by Xcalibur software (Thermo Fisher Scientific, San Jose, CA). These data were processed by Compound Discoverer 3.1 software (Thermo Fisher Scientific, San Jose, CA) to generate a multivariate data matrix. Data matrices were exported into SIMCA14 (Umetrics, Kinnelon, NJ) for multivariate data analysis. Orthogonal projection to latent structures-discriminant analysis (OPLS-DA) was performed on Pareto-scaled data. 20 For chemometric analysis, matrix data were processed from m/z 80 to 1000.

Profiling PEX-GSH Adducts in Liver Microsomes and PHH.
The results of the chemometric analysis on the ions produced from the UHPLC-Q Exactive MS analysis of three groups of samples are presented in Figure 1A,B. These groups were generated by incubating PEX in HLM with or without NADPH or GSH. Two of the groups, which did not contain NADPH or GSH, were used as control groups. Principal component analysis conducted on the data revealed three distinguishable clusters, with two clusters corresponding to the control groups and the remaining one associated with the analyte group (as depicted in Figure 1A). These findings suggest differences in the chemical components among the three groups. The S-plot generated by OPLS-DA ( Figure 1B) illustrates the ion contributions to the separation of groups in HLM. The top-ranking ions, which were identified as PEX-GSH adducts, were labeled in the S-plot ( Figure 1B). The PEX-GSH adducts identified in HLM incubations were comparable to those detected in MLM incubations. Eleven GSH adducts associated with PEX (M1−M11) were determined in LM, and four of them (M1, M2, M9, and M11) were detected in both PHH lysate and medium (Table  1). Their relative abundances in LM and PHH are shown in Figure 1C,D, respectively. In HLM/MLM, M1 (16/5.3%), M2 (19/30%), and M6 (47/52%) were the relatively abundant adducts based on the peak areas ( Figure 1C). In PHH lysate, M9 (the GSH adduct of the cleaved left moiety of PEX, Figure  S1) was the most prevalent adduct, constituting 83% of the total PEX-GSH adducts. In contrast, M2 was the adduct with the largest peak in the PHH medium representing 51% of the total, while M9 accounted for 6.3% ( Figure 1D) based on the peak area.

Identification of PEX-GSH Adducts M1−M11.
Among 11 GSH adducts associated with PEX, 7 of them were formed by conjugating GSH with PEX (M1 and M2) and oxidized PEX (M3−M7) in the presence of NADPH and GSH in HLM and MLM. Their formation is NADPH-and GSHdependent, as they were not detected in control groups (representative trend plots of M1 and M8 shown in Figure S2). Protonated molecules with exact masses at m/z 723.1722 were observed for M1 and M2, eluting at 8.86 and 8.98 min, respectively (Figure 2A). M1 and M2 were identified as the GSH adducts of PEX. At a lower normalized collision energy (30 arbitrary unit), the MS/MS of both M1 and M2 produced two primary fragment ions at m/z 308.0903 and 416.0875, indicating that GSH was attached to PEX ( Figure 2B,C). At a higher normalized collision energy (45 arbitrary unit), fragment ions at m/z 160.0370 and 255.0435 were generated, as shown in the inlaid MS/MS spectra ( Figure 2B,C). The fragment ion at m/z 160.0370 was also observed in the MS/ MS of PEX ( Figure S3). Thus, it is likely that GSH in M1 and  Figure 2D and Table 1). Their molecular weight is 16 Da higher than those of M1 and M2, which were identified as GSH adducts of monohydroxylated PEX (PEX+O). The MS/ MS of M3 produced the fragment ions at m/z 179.0483, 266.0894, 308.0903, and 432.0824, suggesting GSH is linked to the pyrrolo[2,3-b]pyridinyl moiety ( Figure 2E). The MS/ MS of M5 only generated two fragment ions at m/z 308.0900 and 432.0827 ( Figure 2F). Their fragment ions were interpreted in the inlaid structural diagrams in Figure 2E,F,  Figure 2J), which were interpreted in the inlaid structural diagrams. Compared to the fragment ion at m/z 266.0894 from M6, the presence of the ion at m/z 282.0840 in M7 indicated that the second oxidation occurred on the right-framed moiety of PEX.
Interestingly, we also identified three GSH adducts (M8− M10) related to the carbon−carbon cleavage in PEX phase I metabolism, which was investigated in detail in our previous study. 16 M8 and M9 were generated by conjugating GSH with cleaved parts (Figures 2L and S1). M10 was formed by GSH reacting with the N-dealkylated PEX metabolite (amine, Figures 3A and S4A). M8 eluted at 6.27 min with a protonated m/z value of 559.1581 ( Figure 2K). The MS/MS of M8 produced fragment ions at m/z 284.0459, 327.0879, 430.1148, 484.1240, and 541.1441, which were interpreted in the inlaid structural diagram ( Figure 2L). M9 has a protonated molecule at m/z 472.1052, and it has been characterized in our previous studies. 16 The pathway of M9 formation is briefly described in Figure S1. M10, eluted at 3.88 min, has a protonated molecule at m/z 582.1532 ( Figure S4A and Table 1). Additionally, M11, a PEX + cysteine (Cys) + glycine (Gly) adduct, was observed in HLM/MLM and PHH, eluting at 8.53 min with a protonated molecule at m/z 594.1296 ( Figure S4B and Table 1). Unfortunately, we failed to obtain the MS/MS spectra of M10 and M11 with adequate quality for MS/MS analysis. Thus, M10 and M11 were identified as the GSH adduct of amine (amine + O + 2H + GSH) and PEX + Cys + Gly, respectively, based on their exact mass and predicted formulas (Table 1).

Profiling PEX-NH 2 OMe Adducts in Liver Microsomes.
Similar to the profiling of PEX-GSH adduct, we performed the chemometric analysis on the ions produced by the UHPLC-Q Exactive MS analysis of three groups of samples. These groups were generated by incubating PEX in HLM with or without NADPH or NH 2 OMe. The principal component analysis resulted in three distinct clusters, as shown in Figure S5A. The S-plot generated from OPLS-DA ( Figure  3A) illustrates the ions that contribute to the separation of groups in HLM, including PEX metabolites and PEX-MeONH 2 adducts, which were marked in the S-plot ( Figure  3A). Four NH 2 OMe adducts (M12, M13, M16, and M17) with larger peak areas associated with PEX in LM were determined in LM based on the metabolomic analysis (Table  1). Additionally, we manually identified 3 adducts (M14, M15, and M18) with smaller peak areas. The formation of PEX-NH 2 OMe adducts in HLM followed a similar trend to those in  Figure 3B.

Identification of PEX-NH 2 OMe Adducts M12− M18.
M12 is the dominant adduct in both HLM and MLM, followed by M13 ( Figure 3B). M12 has a protonated m/z value of 481.1361 and a retention time of 9.50 min ( Figure  4A). The MS/MS analysis of M12 produced fragment ions at m/z 107.0598, 160.0359, 266.0883, and 434.0971, which were elucidated in the inlaid structural diagram ( Figure 4B). M13, eluted at 10.82 min, has a protonated molecule at m/z 463.1261 ( Figure 4C). The MS/MS analysis of M13 produced fragment ions at m/z 257.0576, 272.0682, and 417.0940, which were elucidated in the inlaid structural diagram ( Figure  4D). The pathways of M12 and M13 formation are presented in Figure 5. The pyrrole ring of PEX can be first oxidized to an epoxide. Following hydrolysis and ring opening, the aldehyde was yielded and subsequently captured by NH 2 OMe to form M12. The elimination of H 2 O from M12 produced M13 ( Figure 5). M14 has a protonated m/z value of 322.1065 and is eluted at 4.74 min ( Figure 4E). The MS/MS analysis of M14 generated fragment ions at m/z 107.0599, 136.0611, 257.0576, and 273.0757, which were interpreted in the inlaid structural diagram ( Figure 4F). M15 eluted at 7.09 min, with a protonated molecule at m/z 304.0965 ( Figure S6A and Table 1). The putative structure of M15 was elucidated based on the exact mass and predicted formula as the MS/MS data was not available (Table 1). Our previous study demonstrated that PEX could produce the amine by releasing 4-(trifluoromethyl)-benzaldehyde via dealkylation ( Figure S7) in LM. 16 As proposed in Figure S7, the formation of amine-NH 2 OMe adducts M14 and M15 followed similar pathways to the formation of M12 and M13. Oxime M17 has the protonated m/z value of 205.0583 and was eluted at 12.41 min, which was identified as the NH 2 OMe adduct of 4-(trifluoromethyl)-benzaldehyde ( Figure S6D, an N-dealkylated product). The MS/MS analysis of M17 produced fragment ions at 147.0282 and 174.0390, which were elucidated in the inlaid structural diagram ( Figure S6E). M16 and M18 were formed by capturing the aldehydes produced from products of carbon−carbon cleavages of PEX with NH 2 OMe ( Figures  S6B,C,F). Their formation pathways have been extensively studied, and their structures were confirmed with standard compounds in our previous work. 16 3.5. Role of P450s in the Formation of GSH or NH2OMe Adducts Associated with PEX. We investigated the role of P450s enzymes in the formation of PEX-GSH and PEX-NH 2 OMe adducts using a panel of human recombinant P450 enzymes and the specific chemical inhibitor in HLM. Using human recombinant P450 enzymes, CYP3A4 and CYP3A5 were identified as the primary enzymes responsible for the formation of adducts (M1−M15, M17, and M18). Additionally, the CYP2A6 isoenzyme contributed to the formation of M4 (31%) ( Table 2). Several other isoenzymes also contributed to the formation of M16, although CYP3A4 was the major enzyme involved. In our chemical inhibitory experiments in HLM, the specific CYP3A inhibitor KCZ at a concentration of 2 μM suppressed the formation of M1-M18 by 87−100% ( Figure 6A,D). In MLM, KCZ was also found to significantly inhibit the formation of the adducts, although it was less effective compared to that in HLM. The observed suppression ranged from 65 to 98% (Figures S8A,B). In PHH lysate and medium, KCZ at a concentration of 10 μM inhibited Table 2

DISCUSSION
Metabolomics-based strategies have been widely adopted for studying both stable and reactive drug metabolites. [17][18][19][20]22 Reactive drug metabolites are well appreciated as significant contributors to drug adverse effects due to their potential for covalent modification of important biological macromolecules such as proteins and DNA. 14,17 Comprehensive understanding of the metabolic profile and characterizing reactive metabolites of drug benefit drug safety evaluation and improvement. In our previous work, we identified stable phase I metabolites of PEX, indicating extensive CYP3A-mediated metabolism of PEX in vitro. 16 In this study, we employed an LC−MS-based metabolomic approach to screen for trapped reactive metabolites related to PEX. A total of 11 PEX-GSH and 7 PEX-NH 2 OMe adducts were identified in HLM or MLM incubations. The formation of GSH adducts associated with PEX suggests the presence of active metabolites during in vitro metabolism, indicating the possibility of a similar reaction occurring between PEX and free sulfur groups in proteins, which could impair the functions of important macromolecules in vivo. In PHH lysate and medium, M1 and M2 were identified, showing a similar pattern to those in HLM/MLM, while the major adduct M6 observed in HLM/MLM was not detected in PHH ( Figure 1C,D). It is possible that certain enzymes in PHH directly convert M6 to M1 or M2 by eliminating H 2 O, resulting in the absence of M6 in PHH. Furthermore, the compositions of PEX-GSH adducts differ between PHH lysate and culture medium ( Figure 1D). M9 (Cleave-left-GSH adduct) has the largest peak area in PHH lysate, while it is smaller in PHH medium. Further investigation is necessary to establish the extent of the involvement of GSH-trapped reactive metabolites in PEX toxicity, as their roles are currently not well understood.
The NH 2 OMe is commonly used in vitro for trapping aldehydes in vitro, as they are often considered toxic metabolites. 17,19,23 By employing metabolomic strategies in conjunction with manual extraction to profile the PEX-NH 2 OMe adducts, we identified 7 oximes, indicating the formation of active aldehyde metabolites ( Figure 3A,B, Table  1). Previous studies have suggested that the interaction between aldehydes and the exocyclic amino groups of DNA and the ε-amino groups of lysine residues can lead to the formation of cross-links between deoxynucleotides and protein amino acids. 24,25 These interactions may cause toxicity by impairing the function of macromolecules. 26 Additional research is required to thoroughly examine the contribution of aldehyde in PEX toxicity and better understand its role.
The GSH adducts identified in both HLM and MLM were M1, M2, M6, and M9 ( Figure 1C). Species differences were observed in the formation of M1, M2, and M9 between HLM and MLM. HLM produced more M1 and M9 than MLM, while the percentage of M2 was higher in MLM than in HLM. As shown in Figure 2B,C, M1 and M2 represented PEX-GSH adducts by conjugating one molecule of GSH to the different sites on PEX. M3−M7 were formed by conjugating GSH to oxidized PEX metabolites, which were observed in PEX metabolism in HLM and MLM. 16 M11 (PEX-Cys-Gly) was likely produced from M1 or M2 by the loss of the glutamic acid motif. The exact mechanisms of formation for M1, M2, and M11 remain unclear. Adducts M8−M10 are three GSH adducts formed by conjugating GSH with cleaved parts of PEX ( Figures S1 and S4). Previous studies suggest that PEX can form phenol metabolites through carbon−carbon cleavage. 16 We suspected that the phenol metabolite could react with GSH, followed by the loss of H 2 O to form M8; however, incubation of the phenol metabolite with GSH in HLM did not yield M8, indicating that M8 formation from PEX occurred via alternative unknown pathways. We have determined the formation of M9 in the study of CYP3A-mediated carbon− carbon cleavage of PEX. 16 In brief, the formed cation The relative abundance from the control groups without KCZ was set as 100%. All data are expressed as mean ± S.E.M (n = 3 for HLM; n =4 for PHH). Statistical analysis was conducted using a two-tailed Student′s independent t-test. *P < 0.05, **P < 0.01, ***P < 0.001. ND, not detected.
conjugates with GSH to form M9 ( Figure S1 and Figure 7). PEX can undergo dealkylation to form the amine metabolite ( Figure S7). Following further oxidization and hydrogenation, the resulting metabolite can be modified by GSH to form M10  Figure 7). Unfortunately, for some minor metabolites like M4, M10, and M11 ( Figure S4), the MS/MS spectra of high quality were not obtained due to their low abundance. The tentative structures were speculated based on their exact masses. The formation of the oximes (M12−M18) in HLM and MLM indicated the generation of reactive aldehydes during PEX metabolism. The formation of M12 and M13 is proposed in Figure 5. The process involves the oxidation of the pyrrole ring of PEX to form an epoxide, which then undergoes hydrolysis to form an aldehyde intermediate. This aldehyde intermediate then reacts with NH 2 OMe to produce oxime M12. M13 can be formed from M12 by eliminating H 2 O. An alternative pathway involves the aldehyde losing H 2 O first and subsequently reacting with NH 2 OMe to produce M13. As shown in Figure S7, the formation of M14 and M15 followed a similar process when the amine served as a substrate. Among the oximes M16−M18, the formation of M16 has been detailed in our carbon−carbon cleavage paper. 16 The aldehyde M16 itself has also been detected in both HLM and MLM ( Figure S7). M17 is the product of the condensation of 4-(trifluoromethyl)benzaldehyde with NH 2 OMe. The formation of benzaldehyde in PEX metabolism was further confirmed by the detection of its corresponding acid and alcohol. The structures of these compounds were also validated using their standard compounds.
Among the tested P450 isoforms, CYP3A4 and CYP3A5 were identified as the primary enzymes responsible for the formation of the stable metabolites of PEX in HLM and the adducts of its reactive metabolites 16 (Table 2 and Figure 6A). Our findings are supported by the clinical data that inhibiting PEX metabolism should lead to higher PEX concentrations in vivo. In clinical practice, the potent CYP3A inhibitor itraconazole significantly increases PEX exposure in healthy human subjects. 27 It was observed that the oxime M16 can be produced by various recombinant human P450s, and even in the control group. Nevertheless, CYP3A enzymes remain the primary contributors as shown in Table 2, as the formation of M16 in HLM and MLM was significantly inhibited by the CYP3A inhibitor KCZ, with inhibition levels reaching up to 100% ( Figure 5A,B). These findings suggest that CYP3A isoforms are the primary enzymes involved in M16 formation in HLM, while some unknown components in the recombinant human P450s may also produce the aldehyde. Additional studies are required to investigate the roles of CYP3Amediated metabolism and reactive metabolites in PEX hepatotoxicity.
In summary, we conducted a comprehensive study on the reactive metabolites of PEX in HLM, MLM, and PHH using an LC−MS-based metabolomic approach and trapping agents (GSH and NH 2 OMe). A total of 11 GSH adducts and 7 NH 2 OMe adducts associated with PEX were identified ( Figure  7). Among these, 4 GSH adducts were also detected in PHH. Among the tested P450 isoforms, CYP3A4 and CYP3A5 were identified as the primary enzymes contributing to the formation of these adducts. The formation of these adducts suggested that PEX metabolism could produce reactive metabolites, which have the potential of modifying important biological macromolecules. Further studies are warranted to elucidate the potential association of these reactive metabolites with PEX-related adverse effects, especially liver injury.