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.

DruMAP: A Novel Drug Metabolism and Pharmacokinetics Analysis Platform.

We developed a novel drug metabolism and pharmacokinetics (DMPK) analysis platform named DruMAP. This platform consists of a database for DMPK parameters and programs that can predict many DMPK parameters based on the chemical structure of a compound. The DruMAP database includes curated DMPK parameters from public sources and in-house experimental data obtained under standardized conditions; it also stores predicted DMPK parameters produced by our prediction programs. Users can predict several DMPK parameters simultaneously for novel compounds not found in the database. Furthermore, the highly flexible search system enables users to search for compounds as they desire. The current version of DruMAP comprises more than 30,000 chemical compounds, about 40,000 activity values (collected from public databases and in-house data), and about 600,000 predicted values. Our platform provides a simple tool for searching and predicting DMPK parameters and is expected to contribute to the acceleration of new drug development. DruMAP can be freely accessed at: https://drumap.nibiohn.go.jp/.

Utilizing public and private sector data to build better machine learning models for the prediction of pharmacokinetic parameters.

One solution to compensate for the shortage of publicly available data is to collect more quality-controlled data from the private sector through public-private partnerships. However, several issues must be resolved before implementing such a system. Here, we review the technical aspects of public-private partnerships using our initiative in Japan as an example. In particular, we focus on the procedure for collecting data from multiple private sector companies and building prediction models and discuss how merging public and private sector datasets will help to improve the chemical space coverage and prediction performance. Teaser: Japan's first public-private consortium in pharmacokinetics has incorporated data from multiple pharmaceutical companies to create useful predictive models.

Development of an In Silico Prediction Model for P-glycoprotein Efflux Potential in Brain Capillary Endothelial Cells toward the Prediction of Brain Penetration.

Developing in silico models to predict the brain penetration of drugs remains a challenge owing to the intricate involvement of multiple transport systems in the blood brain barrier, and the necessity to consider a combination of multiple pharmacokinetic parameters. P-glycoprotein (P-gp) is one of the most important transporters affecting the brain penetration of drugs. Here, we developed an in silico prediction model for P-gp efflux potential in brain capillary endothelial cells (BCEC). Using the representative values of P-gp net efflux ratio in BCEC, we proposed a novel prediction system for brain-to-plasma concentration ratio (Kp,brain) and unbound brain-to-plasma concentration ratio (Kp,uu,brain) of P-gp substrates. We validated the proposed prediction system using newly acquired experimental brain penetration data of 28 P-gp substrates. Our system improved the predictive accuracy of brain penetration of drugs using only chemical structure information compared with that of previous studies.

A public-private partnership to enrich the development of in silico predictive models for pharmacokinetic and cardiotoxic properties.

A novel framework for a public-private (PP) partnership was established by a national initiative of the Development of a Drug Discovery Informatics System, supported by the Japan Agency for Medical Research and Development (AMED). This informatics PP partnership consortium comprised private and public sectors. A database of pharmacokinetic (PK) and cardiotoxic properties was developed, with considerable expansion after integrating proprietary data from private-sector members. This database led to robust in silico prediction models with higher performance than those from the original database. This partnership is a unique example worldwide and could substantially strengthen drug discovery capabilities in both sectors.

Constructing an In Silico Three-Class Predictor of Human Intestinal Absorption With Caco-2 Permeability and Dried-DMSO Solubility.

Absorption of drugs is the first step after dosing, and it largely affects drug bioavailability. Hence, estimating the fraction of absorption (Fa) in humans is important in the early stages of drug discovery. To achieve correct exclusion of low Fa compounds and retention of potential compounds, we developed a freely available model to classify compounds into 3 levels of Fa capacity using only the chemical structure. To improve Fa prediction, we added predicted binary classification results of membrane permeability measured using Caco-2 cell line (Papp) and dried-dimethyl sulfoxide solubility (accuracy, 0.836; kappa, 0.560). The constructed models can be accessed via a web application.

Prediction of Hepatic Clearance of Stereoselective Metabolism of Carvedilol in Liver Microsomes and Hepatocytes of Sprague-Dawley and Cytochrome P450 2D-Deficient Dark Agouti Rats.

Hepatic clearance (CLh) of carvedilol (CAR), which is eliminated via stereoselective metabolism by the CYP2D subfamily of cytochromes P450 (CYPs), was predicted using liver microsomes and hepatocytes from Sprague-Dawley (SD) rats and CYP2D-deficient Dark Agouti (DA) rats to determine the usefulness of prediction method. Plasma concentrations of CAR following intravenous injection to DA rats were higher than those in SD rats. The volume of distribution at steady state and total clearance (CLtot) of S-CAR were approximately two times greater than those of R-CAR in both strains. CLh predicted from in vitro studies using DA rat liver microsomes was different from that obtained from in vivo studies. In contrast, in vitro CLh prediction using DA rat hepatocytes was nearly identical to the CLh observed in DA rats in vivo, and was lower than that in SD rats. The predicted CLh in vitro using hepatocytes correlated well with the observed CLtot in vivo, which is expected to be nearly the same as CLh. These results suggest that in vitro metabolic studies using hepatocytes are more relevant with regard to stereoselectively predicting CLh of CAR than those using liver microsomes.

Relative contribution of rat CYP isoforms responsible for stereoselective metabolism of carvedilol.

The relative contribution of cytochrome P450 (CYP) isoforms responsible for carvedilol (CAR) oxidation in rats were evaluated in order to compare with that of reported human CYPs responsible for the metabolism of CAR enantiomers. The depletion of CAR enantiomers by recombinant CYPs and the effects of CYP-selective inhibitors on the depletion catalyzed by rat liver microsomes (RLM) was determined. Quinine (rat CYP2D inhibitor) markedly inhibited the metabolism of both R- and S-CAR by RLM. The metabolism of S-CAR was inhibited more than that of R-CAR by furafylline, (a CYP1A2 inhibitor, 53.5% vs 11.3%), α-naphthoflavone (a CYP1A2 inhibitor, 64.5% vs 33.6%), and ketoconazole (a CYP3A inhibitor, 87.1% vs 51.2%). Among the CYPs examined, CYP2D2 showed the highest metabolic activities against both the enantiomers. R-CAR was mainly metabolized by CYP2D2 and CYP3A2. CYP2C11 and CYP3A1, in addition to CYP2D2 and CYP3A2 showed higher metabolic activities against S-CAR than that against R-CAR. These results suggest that CYP2D2 predominantly catalyzed R-CAR metabolism, whereas CYP2D2 and CYP3A1/2 catalyzed S-CAR metabolism in rats.

Application of substrate depletion assay to evaluation of CYP isoforms responsible for stereoselective metabolism of carvedilol.

To evaluate the relative contribution of cytochrome P450 (CYP) isoforms responsible for carvedilol (CAR) oxidation, enantioselective metabolism of CAR was investigated in human liver microsomes (HLMs) and recombinant human CYPs by using the substrate depletion assay. CYP2D6 exhibited the highest contribution to the metabolism of R-CAR, followed by CYP3A4, CYP1A2, and CYP2C9, whereas the metabolism of the S-enantiomer was mainly mediated by CYP1A2, followed by CYP2D6 and CYP3A4. In HLMs, metabolism of R- and S-CAR was markedly inhibited by quinidine; R-CAR metabolism (57-61% decrease) was more inhibited than S-CAR metabolism (37-43% decrease), and furafylline and ketoconazole almost equally inhibited metabolism of both enantiomers by 25-32% and 30-50%, respectively. The absence of CYP2D6 in a mixture of five major recombinant CYP isoforms at the approximate ratio as in HLMs resulted in a 42% and 25% decrease in the metabolic activities for R- and S-CAR, respectively. Moreover, the absence of CYP1A2 in the mixture resulted in a 16% and 39% decrease in the metabolic activities for R- and S-CAR, respectively. Our results suggest the stereoselective metabolism of CAR is determined by not only the activity of CYP2D6 but also of CYP1A2 and CYP3A4.

Plasma levels of natriuretic peptides and development of chronic kidney disease.

BACKGROUND: Plasma levels of atrial and brain natriuretic peptides (ANP and BNP) are increased in patients with chronic kidney disease (CKD) complicated with deteriorated kidney function, but the relationship between the plasma level of ANP or BNP and the future development of CKD is unclear. METHODS: We measured the plasma ANP and BNP levels of 294 local residents without CKD in a Japanese community (56.5 ± 10.4 years, mean ± S.D.), who were followed up for the development of CKD over the next 7 years. RESULTS: Sixty-three residents developed CKD during the follow-up period, and the baseline level of plasma ANP of these residents was significantly higher than in those without CKD development. Kaplan-Meier analysis showed that the residents with higher ANP than the median value developed CKD more frequently than those with lower ANP. The association between plasma ANP level and CKD development was found to be independent of baseline estimated glomerular filtration rate by a Cox proportional hazards model, while this association became insignificant when adjusted by age; plasma ANP was significantly correlated with age. Compared with ANP, the relationship between plasma BNP and CKD development was unclear in these analyses. CONCLUSIONS: Age-related elevation of plasma ANP levels preceded the development of CKD in the general population of Japan, raising a possibility for ANP being involved in the development of CKD.

Application of a physiologically based pharmacokinetic model informed by a top-down approach for the prediction of pharmacokinetics in chronic kidney disease patients.

Quantitative prediction of the impact of chronic kidney disease (CKD) on drug disposition has become important for the optimal design of clinical studies in patients. In this study, clinical data of 151 compounds under CKD conditions were extensively surveyed, and alterations in pharmacokinetic parameters were evaluated. In CKD patients, the unbound hepatic intrinsic clearance decreased to a similar extent for drugs eliminated via hepatic metabolism by cytochrome P450, UDP-glucuronosyltransferase, and other mechanisms. Renal clearance showed a similar decrease to glomerular filtration rate, irrespective of the contribution of tubular secretion. The scaling factor (SF) obtained from the interquartile range of the relative change in each parameter was applied to the well-stirred model to predict clearance in patients. Hepatic and renal clearance could be successfully predicted for approximately half and two-thirds, respectively, of the applied compounds, showing the high utility of SFs. SFs were also introduced to a physiologically based pharmacokinetic (PBPK) model, and the plasma concentration profiles of 12 model compounds with different elimination pathways were predicted for CKD patients. The PBPK model combined with SFs provided good predictability for plasma concentration. The developed PBPK model with information on SFs would accelerate translational research in drug development by predicting pharmacokinetics in CKD patients.

Pharmacokinetics and disposition of dalcetrapib in rats and monkeys.

1. The pharmacokinetics and metabolism of dalcetrapib (JTT-705/RO4607381), a novel cholesteryl ester transfer protein inhibitor, were investigated in rats and monkeys. 2. In in vitro stability studies, dalcetrapib was extremely unstable in plasma, liver S9 and small intestinal mucosa, and the pharmacologically active form (dalcetrapib thiol) was detected as major component. Most of the active form in plasma was covalently bound to plasma proteins via mixed disulfide bond formation. 3. Following oral administration of (14)C-dalcetrapib to rats and monkeys, active form was detected in plasma. The active form was mainly metabolized to the glucuronide conjugate and the methyl conjugate at the thiol group. Several minor metabolites including mono- and di-oxidized forms of the glucuronide are also detected in the plasma and urine. 4. The administered radioactivity was widely distributed to all tissues and mainly excreted into the feces (85.7 and 62.7% of the dose in rats and monkeys, respectively). Most of the radioactivity was recovered by 168 h. Although the absorbed dalcetrapib was hydrolyzed to the active form and was bound to endogenous thiol via formation of disulfide bond, it was relatively rapidly eliminated from the body and was not retained.

Characterization of gastrointestinal absorption of digoxin involving influx and efflux transporter in rats: application of mdr1a knockout (-/-) rats into absorption study of multiple transporter substrate.

1. This study was aimed to characterize gastrointestinal absorption of digoxin using wild-type (WT) and multidrug resistance protein 1a [mdr1a; P-glycoprotein (P-gp)] knockout (-/-) rats. 2. In WT rats, the area under the plasma concentration-time curve (AUC) of oral digoxin increased after oral pretreatment with quinidine at 30 mg/kg compared with non-treatment, but the increasing ratio tended to decrease at a high dose of 100 mg/kg. In mdr1a (-/-) rats, however, quinidine pretreatment caused a dose-dependent decrease in the AUC. 3. Quinidine pretreatment did not alter the hepatic availability of digoxin, indicating that the changes in the digoxin AUC were attributable to inhibition of the absorption process by quinidine; i.e. inhibition of influx by quinidine in mdr1a (-/-) rats and inhibition of efflux and influx by quinidine in WT rats. 4. An in situ rat intestinal closed loop study using naringin implied that organic anion transporting peptide (Oatp) 1a5 may be a responsible transporter in the absorption of digoxin. 5. These findings imply that the rat absorption behavior of digoxin is possibly governed by Oatp1a5-mediated influx and P-gp-mediated efflux. The mdr1a (-/-) rat is therefore a useful in vivo tool to investigate drug absorption associated with multiple transporters including P-gp.

Gender difference in relationship between body mass index and development of chronic kidney disease.

BACKGROUND: An epidemiological approach to preventing the development or progression of chronic kidney disease (CKD) is necessary, while few effective preventive measures are currently available. We conducted a community-based, cohort study to identify the factors associated with the development of CKD in the general population. METHODS: We examined 1876 local residents of a Japanese community who had an annual health check-up and, of those, 1506 residents judged not to have CKD (473 men and 1033 women) were followed for the development of CKD over 10 years. RESULTS: The numbers of male and female residents who developed CKD during the follow-up period were 167 (35.3%) and 299 (28.9%), respectively. As compared to those without CKD development, the residents who developed CKD were older, and had a higher body mass index (BMI), systolic blood pressure, and creatinine in both genders. The rate of CKD development in obese female residents was higher than in non-obese women, but such a difference was not noted in male residents. In addition to age and serum creatinine, we identified BMI as an independently significant factor for the development of CKD in women, but not in men. CONCLUSIONS: Increased BMI is a significant risk factor for the development of CKD in women, and there seems to be a gender difference in the association between increased BMI and the development of CKD in the general population.

Development of a hybrid physiologically based pharmacokinetic model with drug-specific scaling factors in rat to improve prediction of human pharmacokinetics.

Accurate prediction of pharmacokinetics (PK) in humans has been a vital part of drug discovery. The aims of this study are to verify the usefulness of scaling factors for clearance (CL) and apparent volume of distribution at the steady state (Vss ) estimated from the difference between observed and predicted PK profiles in rats for human PK prediction, and to develop a novel hybrid physiologically based pharmacokinetic (PBPK) model with the two scaling factors. The human prediction accuracies for CL with in vitro-in vivo extrapolation and Vss with a tissue composition model were improved by using rat-scaling factors. This improvement was explainable by data that the scaling factors for CL and Vss in rats were correlated with those in humans. The predictability of plasma concentration-time profiles by the hybrid PBPK model incorporating two scaling factors was compared mainly with that by the conventional PBPK model. The hybrid PBPK model yielded higher prediction accuracy for plasma concentrations than the conventional method. Furthermore, we proposed a tiered approach using the three prediction methods, including the hybrid Dedrick approach, that were previously reported (Sayama H, Komura H, Kogayu M. 2013. Drug Metab Dispos 41:498-507), taking the available information in the individual stages of drug discovery and development into consideration.

Application of hybrid approach based on empirical and physiological concept for predicting pharmacokinetics in humans--usefulness of exponent on prospective evaluation of predictability.

We developed a hybrid method for predicting plasma concentration-time curves in humans by integrating species differences in in vitro intrinsic clearance (CL(int)) into the Dedrick approach based on the allometry concept. With prediction of clearance (CL) by allometric scaling, taking in vitro CL(int) into consideration improved the accuracy and reduced the average fold error from 2.72 to 1.99. With the hybrid approach of applying the same concept to the Dedrick approach, the predictability of plasma concentration profiles was compared with the results of the conventional Dedrick approach and the physiologically based pharmacokinetic model using 15 compounds with widely ranging physicochemical and pharmacokinetic profiles. The hybrid approach showed the highest predictability among the examined methods. For CL and the apparent volume of distribution at the steady state (V(ss)), the relationship between the exponent of allometric equation and fold error was also evaluated with the hybrid approach. The relationship appeared to be a horseshoe curve. Six compounds with exponents ranging from 0.7 to 1.1 for both CL and V(ss) [antipyrine, caffeine, epiroprim, propafenone, theophylline, and verapamil] displayed higher predictability. Three compounds with an exponent ranging from 0.7 to 1.1 for CL showed better predictability for CL, and the other four compounds appeared to display similar relationship between the exponent and predictability for V(ss). These findings indicated that the exponent becomes a preliminary index to speculate on predictability. Combination of the hybrid approach and exponent allows us to prospectively draw human plasma concentration-time curves, with the implication of possible prediction accuracy prior to clinical studies.

In vitro and in vivo small intestinal metabolism of CYP3A and UGT substrates in preclinical animals species and humans: species differences.

Intestinal first-pass metabolism has a great impact on the bioavailability of cytochrome P450 3A4 (CYP3A) and/or uridine 5'-diphosphate (UDP)-glucoronosyltranferase (UGT) substrates in humans. In vitro and in vivo intestinal metabolism studies are essential for clarifying pharmacokinetics in animal species and for predicting the effects of human intestinal metabolism. We review species differences in intestinal metabolism both in vitro and in vivo. Based on mRNA expression levels, the major intestinal CYP3A isoform is CYP3A4 for humans, CYP3A4 (3A8) for monkeys, CYP3A9 for rats, cyp3a13 for mice, and CYP3A12 for dogs. Additionally, the intestinal-specific UGT would be UGT1A10 for humans, UGT1A8 for monkeys, and UGT1A7 for rats. In vitro and in vivo intestinal metabolism of CYP3A substrates were larger in monkeys than in humans, although a correlation in intestinal availability between monkeys and humans has been reported. Little information is available regarding species differences in in vitro and in vivo UGT activities; however, UGT-mediated in vivo intestinal metabolism has been demonstrated for raloxifene in humans and for baicalein in rats. Further assessment of intestinal metabolism, particularly for UGT substrates, is required to clarify the entire picture of species differences.

Comparative assessment of empirical and physiological approaches on predicting human clearances.

The empirical and physiological predictive approaches to human clearance were evaluated using preclinical in vitro and in vivo data of various datasets to establish a methodology for the prediction of clearance. Among the examined empirical approaches, an allometric scaling method with the rule of exponent (ROE), based on the exponent in simple allometry, provided better prediction. The effect of lipophilicity (clog P) and clearance on the predictivity was investigated using the ROE method. High predictivity was found for a low lipophilic compound with clog P < 0 and for a compound with moderate or high clearance. As a physiological approach, the in vitro-in vivo scaling method using metabolic stability in liver microsomes and hepatocytes was evaluated, and the predictivity taking the plasma protein binding and the nonspecific binding in incubation into consideration was compared with the ROE method. The two methods appeared to show comparable predictivity, although the in vitro-in vivo scaling was conducted under limited conditions like the use of physiological scaling factor and lipophilicity-derived nonspecific binding data. The ROE method could be an alternative predictor of the human clearance of compounds to which a physiological approach cannot be applied, in addition to low lipophilic compounds, with acceptable accuracy.

Differences in cytochrome P450 and nuclear receptor mRNA levels in liver and small intestines between SD and DA rats.

This study aimed to clarify the differences in mRNA levels of cytochrome P450 (CYP) isoforms and nuclear receptors between Dark Agouti (DA) and Sprague-Dawley (SD) rats which are animal models for poor metabolizers and extensive metabolizers for CYP2D6, respectively. Using liver and small intestine tissues of both rat strains, we investigated the mRNA levels of CYP1A, 2A, 2B, 2C, 2D, 2E, and 3A subfamilies and nuclear receptors which regulate the transcription of CYP isoforms. In the liver, male DA rats showed a low CYP2D2 mRNA level but high mRNA levels of CYP3A1, 3A2, and 1A1 compared to SD rats. No significant difference was noted in other CYP isoforms. The mRNA levels of CAR were higher in DA rats than those in SD rats. In small intestine, the mRNA levels of CYP isoforms and nuclear receptors exhibited no significant strain differences. In addition, the activity of CYP3A in small intestinal microsome did not differ between SD and DA rats. Female DA rats exhibited higher mRNA levels of CYP3A1, 3A2, and 2B1 in the liver than female SD rats. In conclusion, the mRNA levels of CYP3A1 and 3A2 isoforms and CAR in the liver but not in the small intestines were different between DA and SD rats in both sexes.

Species differences in in vitro and in vivo small intestinal metabolism of CYP3A substrates.

Intestinal first-pass metabolism has a great impact on the bioavailability of CYP3A substrates in humans, and the in vivo impact has quantitatively been evaluated using CYP3A inhibitors or inducers. In vitro and in vivo preclinical investigations for intestinal metabolism are essential in clarifying pharmacokinetic behavior in animal species and predicting the effect of intestinal metabolism in the human. In this review, we will discuss species differences in intestinal CYP3A enzymes, and CYP3A-mdediated intestinal elimination. Identical CYP3A4 enzyme is expressed in human intestine and liver, but different CYP3A enzymes in both tissues of the mouse and rat are found, that is, respective intestinal enzyme is considered as cyp3a13 and CYP3A62. There is little information on CYP3A enzymes in the monkey and dog intestine, unlike the liver. In vitro metabolic activities of midazolam and nisoldipine are higher in the human and monkey than in the rat. In vivo assessment of cyclosporine, midazolam, nifedipine, tacrolimus, and verapamil has been reported in various species (monkey, rat, mouse, and/or dog) including the human. For midazolam, the monkey shows significant in vivo intestinal metabolism, as evidenced in the human. The monkey might be an appropriate animal model for evaluating small intestinal first-pass metabolism of CYP3A substrates.