Prediction of permeability across intestinal cell monolayers for 219 disparate chemicals using in vitro experimental coefficients in a pH gradient system and in silico analyses by trivariate linear regressions and machine learning

Image-based predictive modelling frameworks for personalised drug delivery in cancer therapy

Personalised drug delivery enables a tailored treatment plan for each patient compared to conventional drug delivery, where a generic strategy is commonly employed. It can not only achieve precise treatment to improve effectiveness but also reduce the risk of adverse effects to improve patients' quality of life. Drug delivery involves multiple interconnected physiological and physicochemical processes, which span a wide range of time and length scales. How to consider the impact of individual differences on these processes becomes critical. Multiphysics models are an open system that allows well-controlled studies on the individual and combined effects of influencing factors on drug delivery outcomes while accommodating the patient-specific in vivo environment, which is not economically feasible through experimental means. Extensive modelling frameworks have been developed to reveal the underlying mechanisms of drug delivery and optimise effective delivery plans. This review provides an overview of currently available models, their integration with advanced medical imaging modalities, and code packages for personalised drug delivery. The potential to incorporate new technologies (i.e., machine learning) in this field is also addressed for development.

Synthesis, antimicrobial properties and in silico evaluation of coumarin derivatives mediated by 1,4-dibromobutane

In this study, monobrominated coumarins (5-6) and bis-coumarins (7-9) were synthesized from 3-carboxylic coumarin and 7-hydroxy-4-methyl coumarin using 1,4-dibromobutane as a binding agent, according to the synthesis procedures described in the literature. Amongst these coumarins, three are new compounds: monobrominated coumarin 5 and bis-coumarins 7 and 9. The structures of the synthesized coumarins were confirmed by FTIR, NMR and HRMS-ESI. In vitro antimicrobial evaluation of these coumarins against strains of twelve bacteria and four fungi revealed their bactericidal and fungicidal properties, with increased antibacterial activity for monocoumarins and improved antifungal activity for bis-coumarins. It was also found that the antibacterial activity was enhanced by the etheric bond, Br atom and alkyl chain and reduced by the ester bonds at position 3 of the pyrone ring or an additional coumarin unit, while the antifungal activity was reinforced by ester bonds and deactivated by the Br atom. For the first time, the in silico investigations of such coumarins were carried out and it was observed that they are less toxic, suitable for oral administration with good permeability through cell membrane, are able to circulate freely in the bloodstream and cross Blood-Brain-Barriers. Moreover, their molecular docking in DNA indicated stable coumarin-DNA complexes with good scores. The results of molecular dynamics simulations performed for 200 ns revealed the rigidity and stability of bis-coumarins (7-9) in the DNA binding pocket and predict that they are potent binders.Communicated by Ramaswamy H. Sarma.

Physiologically based pharmacokinetic (PBPK) modeling of pitavastatin in relation to SLCO1B1 genetic polymorphism

Pitavastatin, a potent 3-hydroxymethylglutaryl coenzyme A reductase inhibitor, is indicated for the treatment of hypercholesterolemia and mixed dyslipidemia. Hepatic uptake of pitavastatin is predominantly occupied by the organic anion transporting polypeptide 1B1 (OATP1B1) and solute carrier organic anion transporter family member 1B1 (SLCO1B1) gene, which is a polymorphic gene that encodes OATP1B1. SLCO1B1 genetic polymorphism significantly alters the pharmacokinetics of pitavastatin. This study aimed to establish the physiologically based pharmacokinetic (PBPK) model to predict pitavastatin pharmacokinetics according to SLCO1B1 genetic polymorphism. PK-Sim® version 10.0 was used to establish the whole-body PBPK model of pitavastatin. Our pharmacogenomic data and a total of 27 clinical pharmacokinetic data with different dose administration and demographic properties were used to develop and validate the model, respectively. Physicochemical properties and disposition characteristics of pitavastatin were acquired from previously reported data or optimized to capture the plasma concentration-time profiles in different SLCO1B1 diplotypes. Model evaluation was performed by comparing the predicted pharmacokinetic parameters and profiles to the observed data. Predicted plasma concentration-time profiles were visually similar to the observed profiles in the non-genotyped populations and different SLCO1B1 diplotypes. All fold error values for AUC and Cmax were included in the two fold range of observed values. Thus, the PBPK model of pitavastatin in different SLCO1B1 diplotypes was properly established. The present study can be useful to individualize the dose administration strategy of pitavastatin in individuals with various ages, races, and SLCO1B1 diplotypes.

The Trends and Future Prospective of In Silico Models from the Viewpoint of ADME Evaluation in Drug Discovery

Drug discovery and development are aimed at identifying new chemical molecular entities (NCEs) with desirable pharmacokinetic profiles for high therapeutic efficacy. The plasma concentrations of NCEs are a biomarker of their efficacy and are governed by pharmacokinetic processes such as absorption, distribution, metabolism, and excretion (ADME). Poor ADME properties of NCEs are a major cause of attrition in drug development. ADME screening is used to identify and optimize lead compounds in the drug discovery process. Computational models predicting ADME properties have been developed with evolving model-building technologies from a simplified relationship between ADME endpoints and physicochemical properties to machine learning, including support vector machines, random forests, and convolution neural networks. Recently, in the field of in silico ADME research, there has been a shift toward evaluating the in vivo parameters or plasma concentrations of NCEs instead of using predictive results to guide chemical structure design. Another research hotspot is the establishment of a computational prediction platform to strengthen academic drug discovery. Bioinformatics projects have produced a series of in silico ADME models using free software and open-access databases. In this review, we introduce prediction models for various ADME parameters and discuss the currently available academic drug discovery platforms.

Caco-2 Cell Line Standardization with Pharmaceutical Requirements and In Vitro Model Suitability for Permeability Assays

The Caco-2 cell line derived from human colon carcinoma is commonly used to assess the permeability of compounds in in vitro conditions. Due to the significant increase in permeability studies using the Caco-2 cell line in recent years, the need to standardize this biological model seems necessary. The pharmaceutical requirements define only the acceptance criteria for the validation of the Caco-2 cell line and do not specify the protocol for its implementation. Therefore, the aim of this study is to review the conditions for permeability studies across the Caco-2 monolayer reported in the available literature concerning validation guidelines. We summarized the main aspects affecting the validation process of the Caco-2 cell line, including the culture conditions, cytotoxicity, cell differentiation process, and monolayer transport conditions, and the main conclusions may be useful in developing individual methods for preparing the cell line for validation purposes and further permeability research.

Modeled Rat Hepatic and Plasma Concentrations of Chemicals after Virtual Administrations Using Two Sets of in Silico Liver-to-Plasma Partition Coefficients

The hepatic elimination of chemical substances in pharmacokinetic models requires hepatic intrinsic clearance (CLh,int) parameters for unbound drug in the liver, and these are regulated by the liver-to-plasma partition coefficients (Kp,h). Both Poulin and Theil and Rodgers and Rowland have proposed in silico expressions for Kp,h for a variety of chemicals. In this study, two sets of in silico Kp,h values for 14 model substances were assessed using experimentally reported in vivo steady-state Kp,h data and time-dependent virtual internal exposures in the liver and plasma modeled by forward dosimetry in rats. The Kp,h values for 14 chemicals independently calculated using the primary Poulin and Theil method in this study were significantly correlated with those obtained using the updated Rodgers and Rowland method and with reported in vivo steady-state Kp,h data in rats. When pharmacokinetic parameters were derived based on individual in vivo time-dependent data for diazepam, phenytoin, and nicotine in rats, the modeled liver and plasma concentrations after intravenous administration of the selected substrates in rats using two sets of in silico Kp,h values were mostly similar to the reported time-dependent in vivo internal exposures. Similar results for modeled liver and plasma concentrations were observed with input parameters estimated by machine-learning systems for hexobarbital, fingolimod, and pentazocine, with no reference to experimental pharmacokinetic data. These results suggest that the output values from rat pharmacokinetic models based on in silico Kp,h values derived from the primary Poulin and Theil model would be applicable for estimating toxicokinetics or internal exposure to substances.

Machine learning and artificial intelligence in physiologically based pharmacokinetic modeling

Physiologically based pharmacokinetic (PBPK) models are useful tools in drug development and risk assessment of environmental chemicals. PBPK model development requires the collection of species-specific physiological, and chemical-specific absorption, distribution, metabolism, and excretion (ADME) parameters, which can be a time-consuming and expensive process. This raises a need to create computational models capable of predicting input parameter values for PBPK models, especially for new compounds. In this review, we summarize an emerging paradigm for integrating PBPK modeling with machine learning (ML) or artificial intelligence (AI)-based computational methods. This paradigm includes 3 steps (1) obtain time-concentration PK data and/or ADME parameters from publicly available databases, (2) develop ML/AI-based approaches to predict ADME parameters, and (3) incorporate the ML/AI models into PBPK models to predict PK summary statistics (eg, area under the curve and maximum plasma concentration). We also discuss a neural network architecture "neural ordinary differential equation (Neural-ODE)" that is capable of providing better predictive capabilities than other ML methods when used to directly predict time-series PK profiles. In order to support applications of ML/AI methods for PBPK model development, several challenges should be addressed (1) as more data become available, it is important to expand the training set by including the structural diversity of compounds to improve the prediction accuracy of ML/AI models; (2) due to the black box nature of many ML models, lack of sufficient interpretability is a limitation; (3) Neural-ODE has great potential to be used to generate time-series PK profiles for new compounds with limited ADME information, but its application remains to be explored. Despite existing challenges, ML/AI approaches will continue to facilitate the efficient development of robust PBPK models for a large number of chemicals.

Updated in Silico Prediction Methods for Fractions Absorbed and Key Input Parameters of 355 Disparate Chemicals for Physiologically Based Pharmacokinetic Models for Time-Dependent Plasma Concentrations after Virtual Oral Doses in Humans

Human metabolic profiles for substances such as toxic food-derived compounds are usually allometrically extrapolated from traditionally determined in vivo rat concentration profiles. To evaluate internal exposures in humans without any reference to experimental data, physiologically based pharmacokinetic (PBPK) modeling could be used if the model input parameters could be estimated in silico. This approach would simplify the use of PBPK models for forward dosimetry after oral doses. In this study, the in silico estimation of input parameters for PBPK models (i.e., fraction absorbed × intestinal availability, absorption rate constants, and volumes of the systemic circulation) was updated for an panel of 355 chemicals (212 previously analyzed and 143 additional substances) using a light gradient boosting machine learning algorithms (LightGBM) based on between 11 and 29 in silico-calculated chemical descriptors. Simplified human PBPK models were then used to calculate virtual maximum plasma concentrations (C_max) and areas under the concentration-time curve (AUC) based on two sets of input parameters, i.e., traditionally derived values from in vivo data and those calculated in silico using the current updated systems. Both sets of C_max and AUC data were well correlated (r = 0.87 and r = 0.73, respectively; p < 0.01, n = 355). Therefore, input parameters for human PBPK models for a diverse range of compounds could be successfully estimated using chemical descriptors and in silico tools. This approach to pharmacokinetic modeling has potential for application in computational toxicology and in the clinical setting for assessing the potential risk of general chemicals.

Low cerebrospinal fluid-to-plasma ratios of orally administered lenalidomide mediated by its low cell membrane permeability in patients with hematologic malignancies

Lenalidomide is a synthetic analog of thalidomide formed by the removal of one keto group (plus the addition of an amino group); it has anti-tumor activities beneficial for the treatment of hematologic malignancies. However, lenalidomide distribution to brain in animal models is reportedly low compared with that of thalidomide. The aim of this study was to evaluate plasma and cerebrospinal fluid concentrations of lenalidomide in three patients with malignant hematologic malignancies. Lenalidomide was detected in plasma from the three Japanese patients 1.5 h following oral administration of 20 mg lenalidomide using liquid chromatography/mass spectrometry, despite the in vitro gastrointestinal permeability of lenalidomide being low. Clinically observed cerebrospinal fluid-to-plasma ratios of lenalidomide were low (1.3-2.4%). Observed influx permeability values for lenalidomide in monkey blood-brain barrier model and human placental cell systems were one order of magnitude lower than those of thalidomide and another second-generation drug, pomalidomide along with a positive permeability control, caffeine. Because of the low cell-barrier permeability of lenalidomide demonstrated in in vitro assays, clinically relevant pharmacokinetic profiles of lenalidomide resulted in low penetrability from plasma into cerebrospinal fluid in patients with hematologic malignancies. Lenalidomide is conclusively suggested to expert its favorable immunomodulatory effects via systemic exposures in the patients.

Plasma concentration profiles for hepatotoxic pyrrolizidine alkaloid senkirkine in humans extrapolated from rat data sets using a simplified physiologically based pharmacokinetic model

AIM

The main aim of the current study was to obtain forward dosimetry assessments of pyrrolizidine alkaloid senkirkine plasma and liver concentrations by setting up a human physiologically based pharmacokinetic (PBPK) model based on the limited information available.

BACKGROUND

The risks associated with plant-derived pyrrolizidine alkaloids as natural toxins have been assessed.

OBJECTIVE

The pyrrolizidine alkaloid senkirkine was investigated because it was analyzed in a European transcriptomics study of natural hepatotoxins and in a study of the alkaloidal constituents of traditional Japanese food plants Petasites japonicus. The in silico human plasma and liver concentrations of senkirkine were modeled using doses reported for acute-term toxicity in humans.

METHODS

Using a simplified PBPK model established using rat pharmacokinetic data, forward dosimetry was conducted. Since in vitro rat and human intrinsic hepatic clearances were similar; an allometric scaling approach was applied to rat parameters to create a human PBPK model.

RESULTS

After oral administration of 1.0 mg/kg in rats in vivo, water-soluble senkirkine was absorbed and cleared from plasma to two orders of magnitude below the maximum concentration in 8 h. Human in silico senkirkine plasma concentration curves were generated after virtual daily oral administrations of 3.0 mg/kg senkirkine (the dose involved in an acute fatal hepatotoxicity case). A high concentration of senkirkine in the culture medium caused in vitro hepatotoxicity as evidenced by lactate dehydrogenase leakage from human hepatocyte-like HepaRG cells.

CONCLUSION

Higher virtual concentrations of senkirkine in human liver and plasma than those in rat plasma were estimated using the current rat and human PBPK models. Current simulations suggest that if P. japonicus (a water-soluble pyrrolizidine alkaloid-producing plant) is ingested daily as food, hepatotoxic senkirkine could be continuously present in human plasma and liver.

Machine Learning Prediction of the Three Main Input Parameters of a Simplified Physiologically Based Pharmacokinetic Model Subsequently Used to Generate Time-Dependent Plasma Concentration Data in Humans after Oral Doses of 212 Disparate Chemicals

Physiologically based pharmacokinetic (PBPK) modeling has the potential to play significant roles in estimating internal chemical exposures. The three major PBPK model input parameters (i.e., absorption rate constants, volumes of the systemic circulation, and hepatic intrinsic clearances) were generated in silico for 212 chemicals using machine learning algorithms. These input parameters were calculated based on sets of between 17 and 65 chemical properties that were generated by in silico prediction tools before being processed by machine learning algorithms. The resulting simplified PBPK models were used to estimate plasma concentrations after virtual oral administrations in humans. The estimated absorption rate constants, volumes of the systemic circulation, and hepatic intrinsic clearance values for the 212 test compounds determined traditionally (i.e., based on fitting to measured concentration profiles) and newly estimated had correlation coefficients of 0.65, 0.68, and 0.77 (p < 0.01, n = 212), respectively. When human plasma concentrations were modeled using traditionally determined input parameters and again using in silico estimated input parameters, the two sets of maximum plasma concentrations (r = 0.85, p < 0.01, n = 212) and areas under the curve (r = 0.80, p < 0.01, n = 212) were correlated. Virtual chemical exposure levels in liver and kidney were also estimated using these simplified PBPK models along with human plasma levels. These results indicate that the PBPK model input parameters for humans of a diverse set of compounds can be reliability estimated using chemical descriptors calculated using in silico tools.