We previously reported that various metabolites were detected in mice, monkeys, and humans after oral administration of radiolabeled DS-1971a. Among these metabolites, mono- (M6) and di-oxidized (M12 and M13) forms were recognized as being specific to monkeys. Here, we report the definitive structures and formation mechanisms of these metabolites to elucidate the monkey-specific metabolism of DS-1971a. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis indicated that mass fragments of standard compounds for M6, M12 and M13, which were biosynthesized by a microorganism, were identical to those of in vivo samples from monkeys, but their full structures could not be identified. Therefore, M6, M12, and M13 were additionally subjected to UV, nuclear magnetic resonance (NMR) and X-ray crystal structure analyses, revealing that these metabolites possess 4-hydroxy (M6), and 4-hydroxy and 5-oxo pyrazole ring moieties (M12 and M13). Notably, M12 and M13 were rearranged stereoisomers of each other. Incubation study with recombinant monkey cytochrome P450s (P450s) suggested that monkey CYP2C8 is responsible for producing not only M6 from DS-1971a but also M12 and M13, both of which are further oxidized metabolites of M6. These results demonstrate novel and unusual metabolism of the pyrazole ring moiety of DS-1971a via monkey CYP2C8-mediated oxidation and rearrangement.
Antibody-drug conjugates (ADCs) have emerged as promising cancer therapeutics; however, predicting their pharmacokinetics (PK) in humans remains challenging, particularly during early development. Therefore, we aimed to develop a practical translational approach to estimate the human PK profiles of ADCs and their cytotoxic payloads using nonclinical data. PK data from nine FDA and PMDA approved ADCs were analyzed. Allometric scaling (exponent = 1.0) was applied to the PK parameters in monkeys to predict the clearance and volume of distribution in humans, which had a high accuracy (within two-fold of the observed clinical values). Subsequently, compartment models were constructed for four ADCs, and the payload release rate constant (krel) was estimated by fitting the monkey ADC and payload concentration-time profiles. The estimated krel values were then directly applied to simulate human payload concentration-time profiles. The predicted human payload exposures were mostly within three-fold of the clinical data, regardless of differences in the linker type. In conclusion, this study demonstrates the use of a simple, single species scaling method using direct antibody measurements for the robust prediction of ADC and payload PK in humans. Our approach provides novel and efficient dose selection and risk assessment with minimal preclinical data.
Interstitial fluid (ISF) is a promising alternative matrix for therapeutic drug monitoring; however, practical and quantitative ISF-based therapeutic drug monitoring has not been established. The aim of this study was to develop a method to quantify vancomycin concentrations in dermal ISF using novel porous poly(glycolic acid) (PGA) microneedles and characterize vancomycin pharmacokinetics in rat ISF by applying a sodium-based correction method to address evaporation-related variability in the ISF sampling volume. Porous PGA microneedles fabricated using a non-solvent-induced phase separation process passively absorbed ISF within a short insertion time (e.g., 5 min). Vancomycin extracted from MNs was quantified via LC–MS/MS, showing linearity over 0.5–100 μg/mL (R2 = 0.994), precision of <8%, and accuracy within 92–107%. Following intravenous vancomycin administration (30 mg/kg) to rats, ISF concentrations mirrored those in the plasma at the terminal elimination phase, indicating that ISF may serve as a surrogate for plasma kinetics. Sodium-based correction reduced the variability in ISF vancomycin concentrations and improved their correlation with plasma concentrations (R2 = 0.834) relative to weight-based method results. These findings indicate that ISF sampling using PGA-based porous microneedles, combined with sodium-based volume correction, provides a minimally invasive and quantitatively reliable platform for ISF-based therapeutic drug monitoring.
The efficacy and safety of brexpiprazole in patients with agitation in Alzheimer’s dementia (AAD) were demonstrated in Japanese study with 1 and 2 mg/day doses. The purpose of this study was to investigate need for dose adjustment in Japanese patients with AAD using population pharmacokinetics (PopPK) and exposure–response (ER) analysis, given that this population was older and renally impaired.First, using PopPK analysis, applicability of the previous PopPK model for Japanese healthy volunteers and patients with schizophrenia, established in a previous study, was confirmed to predict brexpiprazole concentrations in Japanese patients with AAD. Then, using exposure data from the PopPK model, ER analysis demonstrated that ER relationship was adequately described by sigmoid Emax model, although some parameters (EC50 and Hill) had RSEs > 60%. Covariate analysis using PopPK model indicated that older age and/or reduced renal function were associated with increased exposure. However, these changes were considered to remain within the plateau region of ER relationship, suggesting limited clinical impact on efficacy.Therefore, brexpiprazole dose adjustment according to patient characteristics was not considered necessary in Japanese patients with AAD. Given the limited data and residual model instability in ER analysis, further investigation may be warranted to confirm the findings.
The poor bioavailability of the bioactive coumarin Mammea A/AA (MA) due to its hydrophobicity necessitates delivery-enhancing strategies. Cyclodextrins (βCD) and serum albumin (BSA) are known solubilizing and transport agents, respectively, but their combined allosteric effect on MA binding kinetics remains unexplored. This study elucidates the supramolecular interactions in the MA-BSA-βCD ternary system. UV-visible spectroscopy indicated MA relocation into hydrophobic microenvironments. Ratiometric analysis quantified a binding constant (Kb = 3.51 x 104 M-1), revealing a 68.4% reduction in BSA-MA affinity upon βCD inclusion, demonstrating clear negative allosteric modulation. Job’s plot confirmed a 1:1:1 stoichiometry. Thermodynamic profiling (ΔH = -12.32 kcal/mol; ΔS = -70.59 J/mol K) indicated an enthalpy-driven process stabilized by specific interactions. Molecular dynamics simulations validated the experimental binding free energy and identified βCD-induced conformational changes in BSA’s subdomain IB, altering key residue interactions and reducing ligand flexibility. A comparative analysis positions βCD as a generalized allosteric modulator for albumin-bound ligands. These findings provide fundamental insights into competitive binding in complex biological fluids, with direct implications for the rational design of drug delivery systems where fine-tuning protein-ligand affinity is critical.
The analysis of oligonucleotide therapeutics, including antisense oligonucleotides and siRNAs, using existing bioanalytical methods presents unique analytical challenges owing to their distinctive physicochemical properties. Although the ICH M10 guideline provides general bioanalytical validation frameworks, specific considerations for oligonucleotide therapeutics remain undefined, highlighting an urgent need for targeted points to be considered. This white paper presents comprehensive recommendations developed through expert consensus and a literature review by the Japan Agency for Medical Research and Development research groups, addressing the critical technical challenges in liquid chromatography-mass spectrometry (LC-MS)-based bioanalytical methods for oligonucleotide therapeutics. It focuses on ion-pair reversed-phase LC-MS methodologies and emerging techniques, such as hybridization extraction, addressing key issues, including mobile phase stability and degradation, metal adsorption mitigation, carryover prevention, internal standard selection, and matrix effects unique to nucleic acids. The comprehensive coverage includes sample preparation strategies, chromatographic optimization, mass spectrometric detection approaches, and validation parameters adapted from ICH M10, with fit-for-purpose modifications on a case-by-case basis when conventional acceptance criteria cannot be applied. This white paper provides practical recommendations for the development and validation of analytical methods specific to oligonucleotide therapeutics, supports regulatory submissions, and promotes the international harmonization of bioanalytical standards in this rapidly expanding therapeutic modality.
Poorly permeable compounds in the beyond rule of 5 (bRo5) space often demonstrate acceptor well concentrations below the lower limit of quantification or limited resolution between apparent permeability (Papp) values. To address the acceptor well concentration limitation and better resolve differences between poorly permeable molecules, a straightforward solution was proposed, extend the incubation time from 2-h to 24-h to enable acceptor well concentrations to achieve quantifiable values. MDCK cells with knock out of the endogenous canine Mdr1 gene (MDCK-KO) were seeded on 96-Tanswell plates and grown to confluence over three days at which point the 24-h assay was initiated. The longer incubation time enabled receiver compartment concentrations to exceed limits of quantification enabling Papp determination in a test set of approximately 30 molecules. The approximately 10-fold increase in incubation time enabled Papp value determination of up to an order of magnitude lower compared to a 2-h incubation. The assay demonstrates good reproducibility and resolution between poorly permeable compounds. In summary, a straightforward modification of a traditional Transwell permeability assay has produced an assay suitable for the measurement and resolution of very poorly permeable compounds.
Plasma stability assays are routinely applied in early drug discovery to evaluate enzymatic liability and to support lead optimization efforts. Although these assays are generally regarded as robust, the extent to which pre-analytical variables influence specific plasma hydrolase pathways has not been systematically examined. In this study, the impact of plasma anticoagulant selection on hydrolysis was investigated using procaine and pilocarpine as prototypical substrates of the serine hydrolase butyrylcholinesterase (BChE) and the calcium-dependent hydrolase paraoxonase 1 (PON1), respectively. Plasma stability was assessed in dipotassium ethylenediaminetetraacetic acid (K2-EDTA) and lithium heparin plasma collected from the same donors. Procaine exhibited comparable stability across both plasma matrices, consistent with BChE-mediated hydrolysis. In contrast, pilocarpine underwent substantial hydrolysis in lithium heparin plasma but appeared stable in K2-EDTA plasma. This effect is most consistent with calcium chelation by EDTA, as evidenced by complete suppression of PON1-mediated pilocarpine hydrolysis in lithium heparin plasma following exogenous EDTA supplementation. Collectively, these findings demonstrate that K2-EDTA anticoagulation can selectively mask calcium-dependent PON1-mediated plasma hydrolytic pathways, leading to underestimation of plasma lability for susceptible compounds. Anticoagulant choice therefore represents an underappreciated determinant in plasma stability assessment, with direct implications for mechanistic interpretation of plasma-mediated clearance pathways during early DMPK evaluation.
Bicyclo[1.1.1]pentane (BCP) is widely used as a bioisostere of 1,4-disubstituted benzene to prevent oxidative metabolism of the benzene ring by cytochrome P450s (P450s). In this study, we investigated its effectiveness in reducing metabolism-dependent toxicity using acetaminophen (APAP) as a model compound, which causes hepatotoxicity by forming the reactive metabolite N-acetyl-p-benzoquinone imine (NAPQI). In cultured rat hepatocytes, APAP markedly reduced intracellular glutathione levels, whereas N-{3-hydroxybicyclo[1.1.1]pentan-1-yl}acetamide (BCP-APAP), in which the benzene ring of APAP is replaced with BCP, had only a minor effect, suggesting that BCP-APAP does not form NAPQI-like reactive metabolites. We then investigated the interactions of these compounds with human CYP1A2, CYP2E1, and CYP3A4, the major P450s involved in NAPQI formation, using inhibition assays. While both compounds similarly inhibited CYP1A2 and CYP3A4, only APAP inhibited CYP2E1, suggesting that BCP-APAP does not bind to CYP2E1. Docking simulations with 3D crystal structures of these P450s revealed that both compounds can bind to CYP1A2 and CYP3A4 in a similar orientation for heme-mediated metabolism, whereas BCP-APAP did not adopt an APAP-like docking pose in CYP2E1. These findings indicate that replacing the benzene ring in APAP with BCP prevents the formation of reactive metabolites and may subtly alter P450 binding properties in an isoform-dependent manner.
Coproporphyrin I (CP-I), an endogenous substrate of OATP1B, is gaining attention as a biomarker for predicting drug–drug interactions (DDIs) mediated by hepatic organic anion transporting polypeptide 1B (OATP1B). Its application for bridging preclinical and clinical DDI predictions has been widely studied in cynomolgus monkeys. This study aimed to construct a physiologically based pharmacokinetic (PBPK) model of CP-I in monkeys based on our human CP-I model. First, reported DDIs involving an OATP1B substrate pitavastatin and an OATP1B inhibitor rifampicin at several doses were simultaneously analyzed using PBPK coupled with a cluster Gauss–Newton method (CGNM), estimating the in vivo inhibition constant (Ki,u,OATP1B,PTV) in the cynomolgus monkeys. Considering substrate-dependent difference in Ki,u,OATP1B (CP-I vs. pitavastatin) examined in plated monkey hepatocytes, the in vivo Ki was converted to that for CP-I (Ki,u,OATP1B,CP-I). A middle-out approach enabled simultaneous fitting using plasma CP-I and rifampicin. The synthesis rate and hepatic overall intrinsic clearance of CP-I were found to be identifiable by the CGNM. Both parameters were several-fold higher in monkeys than in humans, suggesting the difference in the DDI sensitivity. In conclusion, this study demonstrates that the integrated PBPK-CGNM approach enables quantitative assessment of OATP1B-mediated DDIs in preclinical species while accounting for interspecies differences.
Various functionalization methods have been explored to enhance the targeted delivery of small extracellular vesicles (sEVs). We recently developed a microfluidic post-insertion method that enables the simultaneous incorporation of polyethylene glycol (PEG) lipids and high-functionality and high-quality (HFQ) lipids into bovine milk-derived sEVs. Since sEVs from different sources differ in surface composition and properties that significantly impact modifications, it is crucial to determine whether this method is applicable to sEVs from other origins. In this study, we used a PEG-lipid and a peptide consisting of arginine-glycine-aspartic acid (RGD) linked to a serine-glycine repeat (SG)5 to modify sEVs derived from mouse macrophage cells (RAW264.7) and compared the results with those obtained using the conventional bulk mixing method in vitro. The RGD-PEG sEVs prepared via microfluidic post-insertion and bulk mixing exhibited similar physicochemical properties and cellular binding functions. Successful incorporation of both lipids was confirmed by changes in physicochemical properties, cellular binding capacity, and endocytosis pathways. The results indicated that PEG reduced overall uptake, while RGD facilitated cellular association of RGD-PEG sEVs compared to PEG sEVs alone. Our approach for the simultaneous incorporation of PEG and HFQ lipids may open new avenues for the functional application of sEVs derived from diverse origins.
Hydralazine is commonly employed as a time-dependent inhibitor of aldehyde oxidase (AO) in human hepatocytes for reaction phenotyping yet profound inter-lot variability in its inhibitory effects has been reported. To date, the exact mechanistic basis for this observation remains unclarified. Recently, it was reported that the time-dependent inhibition of AO by hydralazine could be reversed by glutathione (GSH). The present study integrates these previously independent observations for the first time by investigating whether GSH-dependent reversibility of hydralazine-mediated AO inactivation contributes to this observed variability. The time-dependent inhibition of human liver cytosolic AO elicited by hydralazine was fully reversed by GSH, whereas no protection was observed for erlotinib, a structurally distinct AO time-dependent inhibitor. Consistent with these findings, the apparent inhibitory potency (IC50) of hydralazine in pooled human hepatocytes was intermediate between GSH-free and GSH-supplemented cytosolic systems, while erlotinib exhibited comparable inhibition across all conditions. These results indicate that GSH-reversible AO inactivation by hydralazine contributes to inter-lot inhibition variability in human hepatocytes and highlight a mechanistic limitation of hydralazine as a chemical inhibitor for AO reaction phenotyping in human hepatocytes.
In the drug development process, quantitative prediction of drug-drug interaction (DDI) risk must be a vital evaluation item to ensure drug safety in real clinical settings. Therefore, according to the ICH M12 regulatory guideline, it is strongly recommended to clarify the risk of DDIs for new chemical entities based on preclinical in vitro experiments and to further perform clinical DDI studies if predicted DDI risk is suspected, prior to drug approval. Although various in vitro assays have been established, these in vitro kinetic parameters have not always been identical to in vivo ones, which may lead to the misunderstanding of the severity of DDI risks. In this review, we introduced the reported cases of discrepancies between in vitro and in vivo kinetic parameters and provided evidence to consider how to overcome such problems when considering reversible inhibition of metabolizing enzymes and transporters, since many researchers have thought that various methodologies for the evaluation of DDI risks have already been established theoretically and seemed to have no discussion points.
Advances in machine learning and artificial intelligence have recently extended to the quantitative prediction of drug-drug interaction (DDI). Because DDIs arise from diverse mechanisms and the required level of predictive accuracy varies with both the endpoint and the stage of drug development, evaluating their significance and deciding what is needed demand unusually broad expertise-ranging from fundamental biology all the way to state-of-the-art machine-learning methods. In this review, DMPK scientists with expertise in machine learning survey and critique the most recent literature covering the following DDI categories: Cytochrome P450 (CYP) substrates, CYP competitive and time-dependent inhibition, CYP induction, non-CYP substrates, non-CYP inhibition, transporter substrates, transporter inhibition, and cutting-edge predictive algorithms based on deep learning applied for the task of DDIs. For each category we summarize current in silico methodologies and their performance, and we provide expert opinions on how these tools can be optimally incorporated into contemporary drug-discovery workflows.
G-Quadruplexes (G4s) are non-canonical DNA structures formed in guanine-rich sequences and recognized as crucial regulators of gene expression. We identified potential G4-forming sequences within the 5'-flanking region of the human carboxylesterase 2 (CES2) gene, which encodes an enzyme that hydrolyses clinical drugs and endogenous lipids, including triacylglycerols. The aim of this study is to elucidate the impact of G4 formation on CES2 expression and metabolic function. qPCR stop assays and chromatin immunoprecipitation (ChIP) assays using an anti-G4 antibody demonstrated that two guanine-rich sequences in the 5'-flanking region of the CES2 form G4 structures. Treatment with the G4-stabilizing ligand pyridostatin (PDS) reduced CES2 expression and hydrolase activity in HepG2, Huh-1, and HepaSH cells. Luciferase and ChIP assays revealed that PDS suppressed CES2 transcription by inhibiting hepatocyte nuclear factor 4α (HNF4α) binding to the 5'-flanking region of the CES2. PDS also reduced HNF4α expression, through G4 formation within the HNF4α promoter. Notably, PDS decreased CES2 mRNA levels even under HNF4α-overexpressing conditions, indicating that CES2 repression involves both impaired HNF4α binding and reduced HNF4α expression. PDS treatment accumulated triacylglycerol in hepatic cells. Collectively, these findings indicate that G4 formation suppresses CES2 transcription and function, highlighting G4 structures as potential therapeutic avenues for modulating lipid homeostasis.
Cholestatic liver injury is primarily caused by the accumulation of bile acids (BA) in hepatocytes. Based on this mechanism, an in vitro evaluation system for BA-dependent toxicity (BAtox) has been previously developed, using cryopreserved primary human hepatocytes (PHHs) as standard hepatocytes; however, lot-to-lot variability hindered detailed mechanistic elucidation. Therefore, we evaluated the utility of consistently available HepaSH cells, fresh PHHs derived from humanized liver chimeric mice, for assessing BAtox. HepaSH cells exhibited BSEP and BA conjugation activity. These activities were inhibited by known inhibitors resulting in BA accumulation via multiple pathways. BAtox assays using 22 compounds previously tested in PHHs demonstrated that 9-14 compounds induced BAtox in HepaSH cells. Among them, cyclosporin A in the presence of BAs strongly induced LDH release, reaching 81.3%. We further investigated its mechanism of action focusing on BA accumulation, and identified four BAs (chenodeoxycholic acid, deoxycholic acid, and their glycine-conjugated forms) as key contributors to toxicity. Significant intracellular accumulation of these four BAs was observed in the presence of cyclosporin A. These findings demonstrate that HepaSH cells can detect BAtox induced by drug-mediated BA accumulation. This in vitro system is a valuable tool for preclinical hepatotoxicity assays during drug development.
Ketoconazole (KC), a potent cytochrome P450 (CYP) 3A inhibitor, is hydrolyzed by arylacetamide deacetylase (AADAC) in the liver to N-deacetylketoconazole (DAK). Both KC and DAK are known to inhibit CYP3A in human liver microsomes, and DAK accumulates extensively in the liver. We investigate the impact of AADAC on KC-induced drug-drug interactions (DDI) in vivo. DAK inhibited Cyp3a-mediated triazolam hydroxylation in mouse liver microsomes, whose potency was lower than KC (Ki was 0.11 ± 0.03 μM vs 0.008 ± 0.002 μM). Upon KC administration (20 mg/kg, p.o.) to wild-type mice, KC was cleared from the liver within 8 h, whereas DAK remained after 24 h. These findings suggested a smaller impact on DDI in Aadac knockout mice than in wild-type mice. However, the plasma AUC of triazolam (0.5 mg/kg, p.o.) following KC preadministration was significantly higher in Aadac knockout mice than in wild-type mice, indicating that accumulated hepatic DAK in wild-type mice does not affect triazolam hydroxylation. This can be explained by preferential DAK accumulation in lipid-rich organelles, limiting its effective cytosolic concentration to inhibit Cyp3a. In conclusion, Aadac attenuated KC-related DDIs by increasing DAK, which preferentially accumulated in intracellular organelles and thus had a limited impact on cytosolic Cyp3a inhibition.
Uridine 5'-diphospho-glucuronosyltransferases (UGT) are the best known and most well characterized sugar conjugation biotransformation enzymes. Their main function is to catalyze conjugation of glucuronic acid (sugar) to an exposed (or created) hydroxyl, amine or other nucleophilic functional group on a substrate, known as an aglycone. Since these enzymes are subject to inhibition and induction, there is considerable interest in characterizing their role in mediating drug-drug interactions. Glucuronidation reactions are readily recreated in vitro, although most researchers agree rates of catalysis continue to be underestimated compared to those in vivo, thus hampering the prediction of human systemic clearance. Similarly, the magnitude of UGT induction observed in vitro appears to be smaller than those reported in clinical trials. Despite these shortcomings, drug candidates and marketed drugs rarely encounter clinically relevant pharmacokinetic changes attributable to UGT enzyme inhibition or induction. This is likely due to a combination of enzyme properties (high capacity/low affinity, high promiscuity, and minimal inducibility). By contrast, genetic polymorphisms resulting in poor metabolizer or null phenotypes have been demonstrated to have considerable impact on drug disposition, underscoring the need to characterize the fraction of drug metabolized by polymorphic enzymes, especially when glucuronidation is a major clearance mechanism. Recent developments in identifying specific chemical inhibitors and substrate probes, as well as enzyme induction test systems, are helping to further investigate UGT function and role in drug disposition. The aim of this review is to provide an overview of these recent advances and an industry perspective on in vitro testing of UGT-mediated drug-drug interaction potential.
ADCs represent a rapidly growing class of cancer therapeutics and hold significant promise. As of December 2025, fourteen ADCs have received FDA approval, with more than 100 additional ADCs currently in clinical development. Assessment of ADC DDI risk requires consideration of both mAb and cytotoxic payload components of ADCs. The mAb-related DDIs may occur under certain circumstances typically as a pharmacodynamic target mediated interaction, but no clinically meaningful impact has been reported. Whereas, the payload, once released from the ADC, is expected to behave as a small molecule and may pose an enzyme- or transporter-mediated DDI risk. Although the risk of payloads as precipitant of DDI is low and likely of little if any clinical relevance, the potential of circulating payloads as a DDI object still exists depending on its disposition pathways and pharmacokinetic characteristics. In this manuscript, fifteen approved ADCs with eight distinct payloads—MMAE, MMAF, DXd, ozogamicin, DM1, DM4, SN-38, and PBD SG3199 were reviewed for its DDI characterization and risk assessment. Different approaches, including in vitro and in vivo characterization, especially the use of PBPK modeling, were summarized. Insight from the Pharm industry on ADC DDI risk assessment as well as regulatory impact and future directions were discussed.
