Biotransformation research advances - 2022 year in review.

The phytocannabinoid cannabigerol (CBG) is a central biosynthetic precursor to many cannabinoids including Delta 9-tetrahydrocannabinol and cannabidiol (Figure 10; Gagne et al. 2012). CBG is monocyclic, contains a linear prenyl side chain, and is unique amongst phytocannabinoids (e.g. Delta 9-tetrahydrocannabinol and cannabidiol) that are polycyclic. Phytocannabinoids such as Delta 9-tetrahydrocannabinol and cannabidiol have demonstrated beneficial effects in the treatment of neurodegenerative diseases, asthma, pain, arthritis, cancer, etc. (Pacher et al. 2020). More recently, CBG has also gained widespread medical use due to a lack of psychoactivity and a wide range of pharmacologic effects demonstrated in animals and humans (Nachnani et al. 2021). Roy et al. (2022) disclosed the first report on the characterization of CBG metabolites in in vitro studies. Cyclo-CBG (a recently isolated phytocannabinoid) was the major metabolite of CBG in NADPH-supplemented human liver microsomes; incubations with a panel of human recombinant cytochrome P450 (CYP) enzymes revealed a major role for CYP3A4 and CYP2C9 in its formation. An initial CYP-catalyzed epoxidation of the 2 ',3 '-double bond in the prenyl side chain in CBG, followed by spontaneous cyclization with the pendant phenolic group is a likely mechanism yielding cyclo-CBG (Figure 10). A minor metabolite, 6 ',7 '-epoxy-CBG (Figure 10), was also formed in human liver microsomal incubations with CBG; its formation was principally mediated by CYP3A4 and CYP2J2. CBG metabolites were unambiguously identified by comparing their bioanalytical and mass spectral properties with authentic standards prepared via independent synthesis. Intravenous or intraperitoneal administration of CBG to mice also revealed the presence of both cyclo-CBG and 6 ',7 '-epoxy-CBG as circulating metabolites; the amount of cyclo-CBG formed was similar to 100-fold greater than 6 ',7 '-epoxy-CBG, which was consistent with data from human liver microsomal incubations. The pentyl side chain in CBG was resistant toward oxidative metabolism. Importantly, CBG and its oxidized metabolites demonstrated anti-inflammatory properties in BV2 microglial cells stimulated with lipopolysaccharide. In developing a new generation of PET imaging compounds targeting the mutant huntingtin protein (mHTT) aggregates, Liu et al. overcame several metabolic liabilities (Figure 21). The introduction of the pyrimidine ring on the right side introduced a metabolic soft spot for aldehyde oxidase (AO). Interestingly, the active site of AO seemed to discriminate between the two sites near the nitrogen, preferentially oxidizing position 4 rather 2. This directed attack could be mitigated by inserting a methyl group at that specific carbon. In addition, selective methylation of the pyrimidine ring proved valuable to mitigate aldehyde oxidase metabolism (Manevski et al. 2019). In moving away from the [11C] radiolabel atom with a short half-life, the authors developed a synthetic route to insert the more advantageous and longer lived [18F] isotope. This substitution was achieved by introducing a fluoro-ethoxy moiety on the pyridine ring, replacing the previous methoxy group. However, as it is often the case with alkyl fluoro compounds, oxidative defluorination was a concern as it leads to accumulation of [18F]-signal in skull and bones. Deuteration of the ethylene linker close to the labile C-F bond abolished the oxidative defluorination and decreased the O-dealkylation-sulfation observed with the protonated analog (4.6% in human and 23.6% in rhesus hepatocytes, respectively). Overall, the ADME data provided evidence for species-dependent metabolic stability of the lead compound in rodents and higher species. Bansal et al. (2022) quantified the inhibition of cytochrome P450 enzymes by cannabidiol (CBD), delta-9-tetrahydrocannabinol (THC) and their P450 metabolites. The authors then predicted the potential in vivo drug interactions for CBD and THC with P450 probe drugs. CBD and THC are metabolized similarly by CYP enzymes: first via methyl hydroxylation to form 7-OH CBD and 11-OH THC, and then further metabolized to carboxylic acids 7-COOH CBD and 11-COOH THC (Figure 11). This work, in combination with previously published work by Bansal et al. in 2020, quantified the reversible inhibition by THC, CBD, and their metabolites of CYP1A2, -2A6, -2B6, -2C8, -2C9, -2C19, -2D6, and -3 A in human liver microsomes incubated with probe P450 substrates. Protein binding was also determined for THC, CBD, and metabolites, and unbound IC50 values reported (IC50,u). P450s with quantifiable in vitro inhibition by THC, CBD, and metabolites are summarized in Figure 11. CBD, THC, and their monooxygenated metabolites reversibly inhibit multiple P450s, while the carboxylic acid metabolites exhibited little to no inhibition. Time dependent inhibition (TDI) was also demonstrated for CYP2C19 and CYP3A by CBD and 7-OH CBD, with Ki,u and kinact values reported. Four different static models, derived from previous work by Tseng et al. 2021 with varied inputs for hepatic and gut inhibitor concentrations ([I]H and [I]G, respectively), were used along with the measured in vitro IC50,u, Ki,u,. and kinact values to predict potential cannabinoid-drug interactions. The magnitude of the interactions varied across the four models, with some predicting interactions and others not. The model in which [I]H was estimated as the unbound maximum hepatic inlet concentration and [I]G was estimated as the maximum intestinal fluid solubility gave noticeably higher values than the other three models. Interactions were predicted (descending order of magnitude) with oral CBD for CYP3A, -2C9, -2C19, -1A2, -2B6, -2D6, and -2C8 probe substrates, and with oral or inhaled THC with CYP2C9, -3 A, and -1A2 probe substrates. The authors disclose a series of highly selective inhibitors of CDK2, a promising oncology target. Extensive SAR on the scaffold shown in Figure 22 highlighted factors that confer selectivity within the CDK family and the general kinome. One of the key factors was methylation of the sulfonamide that resolved a whole blood issue since compounds bearing a primary aminobenzene sulfonamide have tendencies to bind red blood cells. Unfortunately, the methylated lead compound displayed poor in vivo stability in higher species. The metabolic softspot was revealed to be the sulfonamide methyl group after MetID profiling in higher species. Introduction of polar or bulky substituents did not substantially affect microsomal stability while negatively impacting potency and selectivity. The metabolic liability could eventually be overcome by deuterating the methyl substituent. Demethylation was reduced from 26% to 10% in human microsomes with 85% parent compound remaining after 30 min incubation. Similar effects were noted in cyno microsomes in cynomolgus monkey albeit with lower parent drug remaining (20%). Overall, these efforts resulted in several compounds with submicromolar potency, great selectivity within the CDK family and improved PK profiles. In humans, linezolid morpholine 2-hydroxylation metabolism was the major route of its clearance in in vitro systems as well as in vivo (Figure 12). The investigators, oeren et al. present a reactivity-accessibility model for predicting isoform-specific metabolism of major non-CYP drug metabolizing enzymes for human and general CYP metabolism for preclinical species (oeren et al. 2022; Figure 13). The human enzyme models included in this report are aldehyde oxidase (AO), flavin-containing monooxygenase (FMO), and select uridine 5 '-diphospho-glucuronosyltransferase (UGT) trained in an isoform specific manner including AO1, FMO1, FMO3, UGT1A1, UGT1A4, UGT1A9, and UGT2B7 isoforms. The non-isoform specific preclinical species CYP models incorporate CYP metabolism trends for Sprague-Dawley rat, mouse, and beagle dog for generalized predictions of preclinical species. The ligand-based model uses semi-empirical quantum mechanical (QM) simulations trained with machine learning (ML) methods for predicting sites of metabolism for small molecule drugs. It is validated with empirically determined sites of metabolism for various substrates and their enzymes, as well as with more expensive density functional theory (DFT) QM calculations. Model performance is generally described using several statistical methods (ROC-AUC, confusion matrix, specificity, selectivity, balanced accuracy, and Cohen's kappa). All computational parameters are included in the supporting information. This resource includes transition state cartesian coordinates for AO, FMO, and UGT enzyme reactions, semi empirical scalars, useful references, and input files for NWChem and CP2K program packages which are open source computational chemistry softwares. This review by Le et al. discusses the commonly employed estrogen- and testosterone-based treatment options for transgender patients and the proposed effects of these treatments on drug metabolism (Le et al. 2022). The authors first provide a general overview of transgender medicine: highlighting the diversity of transgender individuals, providing a glossary of commonly used terms, and reviewing the treatment regimens for transgender individuals. These treatment options have been demonstrated to improve quality of life for transgender adults and can be divided into two basic categories: estrogen and testosterone. There have been no studies to examine the effects of these regimens on drug metabolism directly in this patient population, so that the authors leverage the available in vitro and in vivo animal data to propose potential effects of estrogen or testosterone treatment on drug metabolizing enzymes, which are summarized in Table 14. Typically, testosterone regimens consist of weekly injections, or daily topical administration via gel or patch. To date, very little work has been done to examine the effects of testosterone treatment on drug metabolism, with the exception being CYP3A, for which no effect has been observed in vitro or in vivo. Typically estrogen treatment consists of topical, injectable, or oral administration of 17 beta-estradiol, which may be combined with other antiandrogenic agents. More work has been done to understand the effects of estrogenic compounds (as compared to testosterone) on drug metabolism, though much of it in vitro. Based on the results, the authors propose estrogenic compounds may increase CYP3A, CYP2B6, CYP2D6, and UGT1A activity, decrease CYP1A2, and have no effect on CYP2C9 and CYP2C19. The results for CYP2E1 were varied and thus inconclusive. The authors provide a list of 'outstanding questions' with respect to how exogenous testosterone and estrogen treatments regulate drug metabolizing enzymes, prompting extensive further study. Liu et al. reported for the first time the structure of cytochrome P450 (CYP) 8B1 and characterized its active site features (Liu et al. 2022). CYP8B1 (also known as sterol 12alpha-hydroxylase) catalyzes the 12-hydroxylation of 7alpha-hydroxycholest-4-en-3-one leading to formation of the bile acid cholic acid (Staels and Fonseca 2009; Figure 15). CYP8B1 has gained interest as a potential target for treating nonalcoholic fatty liver disease (NAFLD) and type II diabetes. To determine the structure of CYP8B1, Liu et al. expressed and purified human CYP8B1. The authors noted that the active site tryptophan 281 may be important for CYP8B1 function. Therefore, binding assays and kinetic assays were performed with wild-type CYP8B1 and a tryptophan 281 to phenylalanine (W281F) mutant to evaluate the role of tryptophan 281 in CYP8B1 function. The W281F CYP8B1 mutant showed lower binding affinity for the native substrate 7alpha-hydroxycholest-4-en-3-one and reduced catalytic efficiency compared to wild-type CYP8B1 (Liu et al. 2022). Several azole inhibitors were tested for CYP8B1 binding and inhibition; miconazole, econazole, and tioconazole showed high-affinity binding to CYP8B1 and potent inhibition of CYP8B1 activity (Liu et al. 2022). CYP8B1 was co-crystallized with tioconazole, and the structure of CYP8B1 bound to tioconazole was determined by X-ray crystallography at 2.6 angstrom resolution (Liu et al. 2022). The structure and active site of CYP8B1 were compared with previously reported structures for CYP8A1 and CYP7A1. Structural features were identified that may provide selectivity for CYP8B1 compared to CYP8A1 and CYP7A1. This report by Ueda et al. details the investigation of six single nucleotide polymorphisms (SNPs) of the human AOX1 gene [Q314R(rs58185012), I598N (rs143935618), T755I (rs35217482), A1083G (rs139092129), N1135S (rs55754655), and H1297R (rs3731722)]. The SNPs evaluated were identified from a search of the 1000 Genome Project and genome aggregation databases. Criteria for selection included 1) location of the SNP within the AOX1 coding region and 2) a minor allele frequency greater than 1% within a given population. Overexpression of the variants in HEK293 cells revealed no differences in total protein expression relative to the wild-type (WT) enzyme. However, using native PAGE analysis, a 30% reduction in dimer formation relative to WT AOX1 was identified with a variant prevalent in the East Asian population, T755I (rs35217482). In addition, T755I was the only SNP evaluated that resulted in a statistically significant reduction in AOX1 catalytic activity vs WT, as measured by oxidation of phthalazine in S9 fractions of HEK293 cells expressing each variant (Table 16). The reduction in catalytic activity was determined to result from a decrease in maximal product formation rates (Vmax) with no change in affinity (Km). Bamfo et al. report a series of in vitro studies that define the stereoselective hepatic and intestinal metabolism of bupropion (Figure 17). Key to this work was the development of a high-throughput (10 min run-time per sample) chiral LC-MS/MS method that provided chromatographic separation and quantification of bupropion (BUP), hydroxybupropion (OHBUP), erythrohydrobuproprion (EHBUP), and threohydrobupropion (THBUP). Bupropion is a racemate and is metabolized to diastereomeric pairs of metabolites, thus carefully developed chiral analytics are necessary to individually quantify the enantiomers of parent and each metabolite to properly define the kinetics of the stereoselective metabolism. They determine that in vitro clearance of S-BUP with human liver fractions (microsomes, S9, and cytosol) is higher than that of R-BUP. This difference in in vitro clearance is accounted for by ketone reduction of S-BUP to SS-THBUP (8.3-to-42-fold higher than R-BUP to RR-THBUP) and in forming SS-OHBUP and RS-EHBUP. In contrast, when human intestinal fractions (microsomes, S9, and cytosol) were tested, ketone reduction of R-BUP to RR-THBUP dominated. Thus, the metabolism of bupropion is demonstrated as tissue and cellular fraction-dependent with stereoselectivity. Milani et al. investigated the contribution of stereoselective UGT enzymes to the metabolism of four enantiomer or epimer compound pairs (Milani et al. 2023). Using optimized in vitro experiments with human liver microsomes (HLM), they confirmed that levomedetomidine, RO5263397, S-propranolol, and epitestosterone all showed significant differences in glucuronidation rates compared to their respective enantiomers (or epimer, in the case of epitestosterone) (Table 18). By further assessing the contribution of in vitro CYP metabolism and evaluating plasma protein binding and blood plasma partitioning for each of the compound pairs, in vivo clearance values were predicted. As CYP-mediated intrinsic clearance was relatively low compared to UGT-mediated clearance for levomedetomidine, RO5263397, and epitestosterone, the predicted enantioselective clearance was aligned with the UGT selectivity for these three pairs. However, for propranolol, CYP metabolism was the driver of enantioselectivity with R-propranolol predicted to have a similar to 2-fold higher hepatic clearance than the S-enantiomer. Relatively minor enantiomeric differences in plasma binding were observed with the tested compounds though the propranolol enantiomers showed some difference in their respective blood to plasma ratios. The authors also assessed species differences in metabolism stereoselectivity by performing experiments with mouse, rat, dog, minipig, and cynomolgus monkey liver microsomes, and they found many instances in which the enantioselectivity and/or the relative contribution of UGT vs. CYP metabolism was reversed compared to HLM. This annual review is the eighth of its kind since 2016 (Baillie et al. 2016, Khojasteh et al. 2017, Khojasteh et al. 2018, Khojasteh et al. 2019, Khojasteh et al. 2020, Khojasteh et al. 2021, Khojasteh et al. 2022). Our objective is to explore and share articles which we deem influential and significant in the field of biotransformation. The investigators provided an aggregated analysis of absorption, distribution, metabolism, and excretion (ADME) data generated from preclinical studies of 22 N-acetylgalactosamine-conjugated small interfering RNA (GalNAc-siRNA) (McDougall et al. 2022). The majority of the siRNAs currently in development are conjugated to GalNAc ligands to achieve targeted liver delivery. The metabolic stability of the linker-GalNAc was investigated by in vitro incubation in hepatocytes from rat, monkey and human. The results showed similar cleavage and amidase activity between monkey and human, which was slower compared to that in rats. The instability is due to loss of GalNAc moieties followed by cleavage of amide bonds as highlighted in Figure 1. Similar to the in vitro incubation, the in vivo hepatic clearance of monkey was slower than that of rat. For the duplex siRNA, the metabolism was mainly via exonucleases and endonucleases ubiquitously present in plasma and tissues rather than via cytochrome P450s (Ramsden et al. 2019). The exonucleases cleaved nucleic acids from the end of the strand, resulting in release of mononucleotides, while endonucleases worked on the internal regions, resulting in fragmented strands of varying lengths. Migliorati et al. selected 10 FDA-approved antisense oligonucleotide (ASO) drugs to collectively review their formulation, dosage, sites of administration, local and systematic distribution, metabolism, degradation, and excretion (Migliorati et al. 2022). The ASO drugs were fomivirsen, pegaptanib, mipomersen, nusinersen, inotersen, defibrotide, eteplirsen, golodirsen, viltolarsen, and casimersen. Their major ADME features are membrane permeabilization through endocytosis and nucleolytic degradation by endonucleases and exonucleases, which differ from small molecule drugs. This summary of information for the FDA-approved ASO drugs supports a better understanding of their therapeutic efficacy and potential adverse drug reactions (ADRs). Several significant knowledge gaps, particularly on cellular uptake and subcellular trafficking and distribution, are identified, and future perspectives are discussed. This comprehensive review by Peter Dragovich outlines an emerging delivery strategy for protein degraders (also broadly known as proteolysis targeting chimeras, PROTACs)-degrader-antibody conjugates (DACs) (Dragovich 2022). Many PROTACs are large and flexible molecules and their physicochemical properties include high molecular weight, high polar surface area, high number of rotatable bonds and poor solubility, which renders poor DMPK properties and the delivery of these molecules to target tissues can be challenging. DACs can potentially provide an alternate means for efficient delivery in vivo. DAC is an analog of antibody-drug conjugates (ADCs), which connects the degraders to a monoclonal antibody (mAb) via various chemical linkers. Similar to ADCs, the linker chemistry design of DACs can be complex and involves the optimization of several components, including an antibody spacer, a trigger and an optional drug spacer. A variety of DAC linker design and their cleavage mechanisms are summarized in this review, and two selected examples are presented in Figure 3. After the initial internalization and mAb catabolism, the linker drug in Figure 3A underwent disulfide reduction to form the corresponding thiol, followed by self-immolation to release the degrader, while the linker drug in Figure 3B cleaved its peptidomimetic linker to form a para-aminobenzyl intermediate, which self-immolates to release the payload. In both cases, there are multiple steps involved in the conversion of conjugate to the final payload, underlining the complexity of DAC design. In addition to the self-immolated linkers, there are also other linkers used in DAC design, such as an enzyme-cleavable pyrophosphate-containing linker (Dragovich et al. 2020). The outcomes from the summarized studies have demonstrated the ability of DACs to deliver protein degraders to targets, and this emerging technique is well-positioned for future development. Here, Wolfson et al. investigated non-enzymatic mechanisms via redox active H2S for reduction of azo based dyes and drugs. The investigators carried out systematic evaluation of the mechanisms by which H2S is formed. They also studied the alterations in the amounts of H2S due to presence or absence of flavin mononucleotide (FMNs), azo compounds and dietary cysteine/sulfur (Figure 4a). The in vitro experiments consisted of incubations in buffer with FMNs and sodium sulfide, cysteine supplemented Escherichia coli and human fecal matter in live and heat-deactivated cultures. Higher amounts of azo reduction was observed in cysteine supplemented samples in a concentration dependent manner. The cysteine reduction hypothesis was further tested in vivo in mice fed with and without cysteine-rich diets and found that azo reduction was higher in the former case. The degradation of cysteines by sulfur reducing bacteria was not universal and potential inter-species/individual variations were presumed. To understand this variability, the authors utilized bioinformatics tools and found multiple sulfidogenic cysteine degrading enzymes in genomes of diverse gut microorganisms and published data on human gut metagenomic samples. However, they concluded that cysteine metabolism is a core biochemical output in the human gut regardless of individual community compositions. An anticancer drug, capecitabine (CAP) is metabolized to 5-fluorouracil (5-FU). The pharmacokinetics (PK) of CAP exhibited high inter-individual variability which cannot be explained by the known host risk factors. The investigators elucidated the potential role of gut microbial metabolism in the variability of CAP PK and efficacy (Spanogiannopoulos et al. 2022). The authors screened 47 human gut bacteria strains for sensitivity to CAP and 5-FU and found a large variation in the minimal inhibitory concentration (MIC) against each strain. They also found that E. coli was able to convert CAP to 5-FU (Figure 5). By further screening 5-FU resistant strains, they discovered that a few strains almost completely depleted 5-FU to an inactive metabolite, dihydrofluorouracil (DHFU), and the bacterial preTA operon was responsible for the metabolism of 5-FU to DHFU (Figure 5). In mammalian cells, dihydropyrimidine dehydrogenase (DPYD) is responsible for the biotransformation of 5-FU to DHFU. Mice colonized with preTA expressing E. coli showed decreased anticancer efficacy, AUC and Cmax of 5-FU following oral administration of CAP compared to mice colonized with nonfunctional mutant preTA expressing E. coli. Lastly, the authors showed that the abundance of preTA and the 5-FU inactivation activity in the stools from humans exhibited significant inter-individual variability. The investigators highlighted interspecies differences in protein conjugation, absorption, and metabolism of sotorasib in rats and dogs. In rat and dog ADME studies of [14C]-sotorasib, Dahal et al. observed the presence of long-lived blood and plasma radioactivity in rats, but not in dogs (Figure 6). The blood-to-plasma ratio of radioactivity in rats was less than 1 for the first hour after dosing, but increased slowly over time, reaching a ratio of 32 at 144 h after dosing. In dogs, however, the long-lived radioactivity was not observed in blood and plasma profiles, with radioactivity below measurable levels after 8 h post-dose. The investigators hypothesized that the prolonged levels of circulating radioactivity, and the increase in blood-to-plasma ratio over time in rats, is likely to be due to the acrylamide warhead and its covalent adducts to proteins. Based on proteomic analysis of samples from incubations of sotorasib in rat blood and plasma, it was determined that the observed extended radioactivity profile in rats could be due to covalent binding of sotorasib with a cysteine residue in hemoglobin and a lysine residue in serum albumin. According to the results from the sequence alignment of rat and dog hemoglobin, the reactive cysteine identified in rat hemoglobin was not present in the dog, providing a potential explanation as to why long-lived radioactivity in dogs was not observed. Another interesting interspecies difference was the low oral absorption of sotorasib in dogs compared to rats. The investigators stated that this observation may be due to a higher physiologic pH within the GI tract in dogs, as the solubility of sotorasib in aqueous solution decreases at a higher pH. In addition, unchanged sotorasib accounted for approximately 100% of the radioactive dose recovered in dog feces. The major metabolic pathways of sotorasib in rats and dogs were glutathione (GSH) conjugation, downstream metabolism of the GSH conjugate, and oxidation. Based on in vitro data with or without ethacrynic acid, a selective glutathione S-transferases (GSTs) inhibitor, non-enzymatic conjugation (Michael addition) was the primary mechanism of the glutathione reaction, with a minimal enzymatic contribution from GSTs. Additionally, the authors investigated the downstream metabolism of sotorasib GSH metabolite (M12) to cysteinylated metabolite (M10) via gamma-glutamyl-transferase (GGT)-mediated hydrolysis. Kidney S9 fractions were used due to a known high level of GGT expression in kidneys. Interestingly, the production of M10 from M12 was approximately 4-fold higher in rat kidney S9 incubations than that in dog. Finally, the formation of M10 from M12 in rat kidney S9 incubations was almost completely inhibited by a known irreversible GGT inhibitor, acivicin, suggesting that the cysteine conjugate was converted from the glutathione conjugate via the GST pathway. The peptidomimetic agent PF-07321332 (nirmatrelvir) (Figure 7), in combination with the pharmacokinetics booster ritonavir, is the first oral severe acute respiratory syndrome coronavirus 2 main protease inhibitor (Owen et al. 2021; Hammond et al. 2022) to receive Emergency Use Authorization (EUA) for the treatment of coronavirus disease 2019. Singh et al. (2022) leveraged quantitative 19F nuclear magnetic resonance (NMR) spectroscopy to characterize nirmatrelvir disposition in humans, taking into consideration the presence of the three fluorine atoms (part of the trifluoroacetamide group) in the nirmatrelvir structure. Thus, the ADME portion could be accomplished by dosing nirmatrelvir in the phase 1 clinical study without a need for a time-consuming [14C]-nirmatrelvir radiosynthesis in preparation. Following oral administration of a single dose nirmatrelvir (300 mg) in the presence of ritonavir to healthy participants, a mass balance of 84.9% was achieved, which included 80.7% of the material measured by quantitative 19F NMR and an additional 4.2% measured as M8 (an 19F-silent metabolite derived from the hydrolysis of the trifluoroacetamide moiety) detected by HPLC-MS. The vast majority of nirmatrelvir-related material was excreted as unchanged parent (82.5% of the administered dose), with 55.0% in urine and 27.5% in feces. Hydrolysis products M5 and M8 (Figure 7), derived from cleavage across the P2 peptide backbone and the trifluoroacetamide capping group, respectively, were the principal metabolites observed. The high extent of disposition proceeding via excretion of unchanged parent and hydrolysis was due to the coadministration of the selective CYP3A4 inhibitor and pharmacokinetics booster ritonavir, since in vitro studies had revealed that cytochrome P450 (CYP)3A4 primarily contributed toward nirmatrelvir metabolism (Eng et al. 2022). The 19F NMR study was also supplemented with a validated liquid chromatography tandem mass spectrometry (LC-MS/MS) assay to quantitatively measure nirmatrelvir and its hydrolytic metabolite M5. The mass balance (similar to 75.6%) using the LC-MS/MS method further corroborated the data obtained using 19F NMR spectroscopy. Unchanged nirmatrelvir was the only component in circulation, which mitigated potential metabolite in safety testing concerns. Uno et al. investigated the role of various dog CYP3A isoforms in liver, small intestine and other tissues (Uno et al. 2023). They complemented what was known about CYP3A12 and 3A26 with new understanding about CYP3A98 and 3A99. The enzyme's sequence identity compared to human CYP3A was between 72-80%. The significant finding was the importance of CYP3A98 in the intestine (in both jejunum and ileum) compared to the other CYP3As and yet in the liver, it is CYP3A12 followed by CYP3A26 (Table 8). In the kidney, CYP3A12 had the highest expression followed by CYP3A98 and CYP3A99 and in lung, it was CYP3A98 followed by CYP3A26 and CYP3A12. These dog CYP3A enzymes were expressed and analyzed for their kinetics properties using human CYP3A probe substrates (testosterone, estradiol, alprazolam and midazolam). One take home message was that all of these dog CYP3A enzymes were capable of metabolizing the human probes with similar human isoform specificity, which suggests similar active site cavity. Finally, based on the expression and enzyme kinetics, the investigators concluded that most of the activity in the liver is by CYP3A12 and in the intestine, by CYP3A98. Previous reports by Rendic and Guengerich, such as the most recent one in 2021 (Rendic and Guengerich 2021), have highlighted the overwhelming contribution of cytochrome P450 to the metabolism of compounds. This report provides an extensive review of the human non-P450 oxidoreductive enzymes in the metabolism of drugs, general chemicals, natural products and physiological compounds. The enzymes reviewed comprise microsomal flavin-containing monooxygenase (FMO), monoamine oxidase (MAO), NAD(P)H quinone oxidoreductase (NQO) and molybdenum-containing hydroxylases (AOX and xanthine oxidoreductase (XOR) enzymes) with each enzyme contributing to no more than 2% of overall metabolism, compared to the similar to 95% by the P450s. Paludetto et al. comprehensively investigated the metabolism and inhibitory effects of hydroxychloroquine (HCQ) on cytochrome P450 (CYP) metabolism (Paludetto et al. 2023). In this work, they used human liver microsomes and recombinant cytochrome P450 (CYP) to explore the metabolism of HCQ. Additionally, the investigators determined the reversible and time-dependent inhibitory effects of HCQ and its major metabolites using an automated substrate cocktail method. The major finding was that CYP2D6, CYP3A4, and CYP2C8 are involved in the metabolism of HCQ in vitro. In addition, they found that HCQ and its metabolites were reversible competitive inhibitors of CYP2D6. Importantly, the results from this work can be used to improve physiologically-based pharmacokinetic models, as well as to understand risk estimations of drug-drug interactions for HCQ. The in vitro and in vivo biotransformation pathways of an oral selective estrogen receptor degrader, LSZ102 [(E)-3-(4-((2-(2-(1,1-difluoroethyl)-4-fluorophenyl)-6-hydroxybenzo[b]thiophen-3-yl)oxy)phenyl)acrylic acid], revealed remarkable species differences (Pearson et al. 2022). The metabolites in human hepatocytes, intestinal S9 fractions, and plasma were predominantly sulfate conjugates. However, metabolites in the corresponding rat matrices were majorly glucuronide conjugates. The identified conjugative metabolites of LSZ102 comprised of a sulfonyl conjugate (M5), an acyl glucuronide (M6), a glucuronide-sulfate diconjugate (M12), and a diglucuronide (M14). While the circulating metabolites were identified in plasma from patients dosed with cold LSZ102, the other reported studies were carried out with 14C-LSZ102. The predominance of metabolites was reported using radiochromatography and/or LC-UV chromatography, as is the current best practice. This is because in the absence of authentic metabolite reference standards, equimolar mass spectrometric responses are known to be highly variable (Hatsis et al. 2017). Interesting di-conjugate metabolites i.e. a glucuronide-sulfate conjugate and diglucuronides as shown in Figure 20, were also reported, and are discussed further in the commentary.