Prediction Of Drug Absorption Research Articles

Integrating (deep) machine learning and cheminformatics for predicting human intestinal absorption of small molecules

The oral route is the most preferred route for drug delivery, due to which the largest share of the pharmaceutical market is represented by oral drugs. Human intestinal absorption (HIA) is closely related to oral bioavailability making it an important factor in predicting drug absorption. In this study, we focus on predicting drug permeability at HIA as a marker for oral bioavailability. A set of 2648 compounds were collected from some early as well as recent works and curated to build a robust dataset. Five machine learning (ML) algorithms have been trained with a set of molecular descriptors of these compounds which have been selected after rigorous feature engineering. Additionally, two deep learning models - graph convolution neural network (GCNN) and graph attention network (GAT) based model were developed using the same set of compounds to exploit the predictability with automated extracted features. The numerical analyses show that out the five ML models, Random forest and LightGBM could predict with an accuracy of 87.71 % and 86.04 % on the test set and 81.43 % and 77.30 % with the external validation set respectively. Whereas with the GCNN and GAT based models, the final accuracy achieved was 77.69 % and 78.58 % on test set and 79.29 % and 79.42 % on the external validation set respectively. We believe deployment of these models for screening oral drugs can provide promising results and therefore deposited the dataset and models on the GitHub platform (https://github.com/hridoy69/HIA).

Generation and application of CES1-knockout Tet-Off-regulated CYP3A4 and UGT1A1-expressing Caco-2 cells

Caco-2 cells, a human colorectal adenocarcinoma cell line, are widely used to model small intestinal epithelial cells in the drug development process because they can predict drug absorption with high accuracy. However, Caco-2 cells have several issues. First, Caco-2 cells have little expression of cytochrome P450 3A4 (CYP3A4), which is a major drug-metabolizing enzyme in the human intestine. We previously developed Caco-2 cells whose expression of CYP3A4 can be controlled using doxycycline (Dox) (CYP3A4-Caco-2 cells) (Ichikawa et al., Sci. Rep, 2021). However, since the Tet-On system was used to regulate CYP3A4 expression in these cells, there was concern about drug-drug interactions. The second issue is that in the human small intestine, carboxylesterase 2 (CES2) is more highly expressed than carboxylesterase 1 (CES1), while in Caco-2 cells CES1 is more highly expressed. The third issue is the low level expression of uridine diphosphate glucuronosyltransferase 1A1 (UGT1A1), a phase II drug-metabolizing enzyme. In this study, we used genome-editing technology to establish CES1-knockout Caco-2 cells whose CYP3A4 and UGT1A1 expression can be regulated by the Tet-Off system. These cell lines would be useful in pharmaceutical researches, including intestinal toxicological studies, as an in vitro model for orally administered drugs.

Improving ADMET Prediction Accuracy for Candidate Drugs: Factors to Consider in QSPR Modeling Approaches.

Quantitative Structure-Property Relationship (QSPR) employs mathematical and statistical methods to reveal quantitative correlations between the pharmacokinetics of compounds and their molecular structures, as well as their physical and chemical properties. QSPR models have been widely applied in the prediction of drug absorption, distribution, metabolism, excretion, and toxicity (ADMET). However, the accuracy of QSPR models for predicting drug ADMET properties still needs improvement. Therefore, this paper comprehensively reviews the tools employed in various stages of QSPR predictions for drug ADMET. It summarizes commonly used approaches to building QSPR models, systematically analyzing the advantages and limitations of each modeling method to ensure their judicious application. We provide an overview of recent advancements in the application of QSPR models for predicting drug ADMET properties. Furthermore, this review explores the inherent challenges in QSPR modeling while also proposing a range of considerations aimed at enhancing model prediction accuracy. The objective is to enhance the predictive capabilities of QSPR models in the field of drug development and provide valuable reference and guidance for researchers in this domain.

SubQ-Sim: A Subcutaneous Physiologically Based Biopharmaceutics Model. Part 1: The Injection and System Parameters.

To construct a detailed mechanistic and physiologically based biopharmaceutics model capable of predicting 1) device-formulation-tissue interaction during the injection process and 2) binding, degradation, local distribution, diffusion, and drug absorption, following subcutaneous injection. This paper is part of a series and focusses on the first aspect. A mathematical model, SubQ-Sim, was developed incorporating the details of the various substructures within the subcutaneous environment together with the calculation of dynamic drug disposition towards the lymph ducts and venous capillaries. Literature was searched to derive key model parameters in healthy and diseased subjects. External factors such as body temperature, exercise, body position, food or stress provide a means to calculate the impact of "life events" on the pharmacokinetics of subcutaneously administered drugs. The model predicts the tissue backpressure time profile during the injection as a function of injection rate, volume injected, solution viscosity, and interstitial fluid viscosity. The shape of the depot and the concentrations of the formulation and proteins in the depot are described. The model enables prediction of formulation backflow following premature needle removal and the resulting formulation losses. Finally, the effect of disease (type 2 diabetes) or the presence of recombinant human hyaluronidase in the formulation on the injection pressure, are explored. This novel model can successfully predict tissue back pressure, depot dimensions, drug and protein concentration and formulation losses due to incorrect injection, which are all important starting conditions for predicting drug absorption from a subcutaneous dose. The next article will describe the absorption model and validation against clinical data.

Artificial Intelligence-Based Quantitative Structure-Property Relationship Model for Predicting Human Intestinal Absorption of Compounds with Serotonergic Activity.

Oral medicines represent the largest pharmaceutical market area. To achieve a therapeutic effect, a drug must penetrate the intestinal walls, the main absorption site for orally delivered active pharmaceutical ingredients (APIs). Indeed, predicting drug absorption can facilitate candidate screening and reduce time to market. Algorithms are available with good prediction accuracy that however focus only on solubility. In this work, we focused on drug permeability looking at human intestinal absorption as a marker for intestinal bioavailability. Being of considerable therapeutic relevance, APIs with serotonergic activity were selected as a dataset. Due to process complexity, experimental data scarcity, and variability, we turned toward an artificial intelligence (AI)-based system, which is a hierarchical combination of classification and regression models. This combination of seemingly two models into a single system widens the space of molecules classified as highly permeable with high accuracy. The specialized and optimized system enables in silico and structure-based prediction with a high degree of certainty. Predictions in external validation allowed correct selection of the 38% of highly permeable molecules without any false positives. The proposed system based on AI represents a promising tool useful for oral drug screening at an early stage of drug discovery and development. Datasets and the obtained models are available on the GitHub platform (https://github.com/nczub/HIA_5-HT).

Control and Prediction of Drug Absorption at Mucosal Tissues

The mucosal drug delivery system (mDDS) is one of the promising approaches to control the pharmacokinetic behavior of drugs. In this approach, surface properties of drug nanoparticles are key determinants to provide particles with mucoadhesive and mucopenetrating properties for prolonged retention at mucosal tissue and rapid mucosal absorption, respectively. In this paper, we would like to discuss the preparation of mDDS formulations by flash nanoprecipitation using a four-inlet multi-inlet vortex mixer, in vitro and ex vivo evaluation of mucopenetrating and mucoadhesive properties of polymeric nanoparticles as well as the application of mDDS to the pharmacokinetic control of cyclosporine A after oral administration to rats. We also share our current research on in silico modeling and prediction of the pharmacokinetic behavior of drugs after intratracheal administration to rats.

Construction of the small intestine on molecular dynamics simulation and preliminary exploration of drug intestinal absorption prediction

In this study, molecular dynamics simulation was applied to the construction of the small intestinal epithelial cell membrane and prediction of drug absorption. First, we constructed a system of a small intestinal epithelial cell membrane that was close to the real proportion and investigated the effects of temperature, water layer thickness, and ionic strength on membrane properties to optimize environmental parameters. Next, three drugs with different absorptivity, including Ephedrine (EPH), Quercetin (QUE), and Baicalin (BAI), were selected as model drugs to study the ability of drugs through the membrane by the free diffusion and umbrella sampling simulation, and the drug permeation ability was characterized by the free diffusion coefficient D and free energy barrier (△G) in the processes. The results showed that the free diffusion coefficient D and △G orders of the three drugs were consistent with the classical experimental absorption order, indicating that these two parameters could be used to jointly characterize the membrane permeability of the drugs.

Optimized In Silico Modeling of Drug Absorption after Gastric Bypass: The Case of Metformin.

Bariatric surgery is an effective treatment for severe obesity and related comorbidities, such as type II diabetes. Gastric bypass surgery shortens the length of the intestine, possibly leading to altered drug absorption. Metformin, a first-line treatment for type II diabetes, has permeability-dependent drug absorption, which may be sensitive to intestinal anatomic changes during bypass surgery, including Roux-en-Y gastric bypass (RYGB). Previous computer simulation data indicate increased metformin absorption after RYGB. In this study, we experimentally determined the region-dependent permeability of metformin, using the rat single-pass intestinal perfusion method (SPIP), which we then implemented into GastroPlusTM to assess the contribution of our SPIP data to post-RYGB metformin absorption modeling. Previous simulations allowed a good fit with in vivo literature data on healthy and obese control subjects. However, it was revealed that for post-RYGB drug absorption predictions, simply excluding the duodenum/jejunum is insufficient, as the software underestimates the observed plasma concentrations post-RYGB. By implementing experimentally determined segmental-dependent permeabilities for metformin in the remaining segments post-surgery, GastroPlusTM proved to fit the observed plasma concentration profile, making it a useful tool for predicting drug absorption after gastric bypass. Reliable evaluation of the parameters dictating drug absorption is required for the accurate prediction of overall absorption after bariatric surgery.

Does Chemotherapy-Induced Gastrointestinal Mucositis Affect the Bioavailability and Efficacy of Anti-Infective Drugs?

Antimicrobial prophylaxis is increasingly being used in patients with hematological malignancies receiving high-dose chemotherapy and hematopoietic stem cell transplantation (HSCT). However, few studies have focused on the potential impact of gastrointestinal mucositis (GI-M), a frequently observed side effect of chemotherapy in patients with cancer that affects the gastrointestinal microenvironment, on drug absorption. In this review, we discuss how chemotherapy leads to an overall loss of mucosal surface area and consequently to uncontrolled transport across the barrier. The barrier function is depending on intestinal luminal pH, intestinal motility, and diet. Another factor contributing to drug absorption is the gut microbiota, as it modulates the bioavailability of orally administrated drugs by altering the gastrointestinal properties. To better understand the complex interplay of factors in GI-M and drug absorption we suggest: (i) the longitudinal characterization of the impact of GI-M severity on drug exposure in patients, (ii) the development of tools to predict drug absorption, and (iii) strategies that allow the support of the gut microbiota. These studies will provide relevant data to better design strategies to reduce the severity and impact of GI-M in patients with cancer.

Pooled Analysis of Gastric Emptying in Patients With Obesity: Implications for Oral Absorption Projection

PurposeGastric emptying time is one of limiting factors that determines the pharmacokinetic properties of drugs administered by mouth. Despite the high prevalence of obesity worldwide, modifications in gastric emptying time have not been systematically addressed in this set of patients. The current analysis aims to quantitatively address obesity-related changes in gastric emptying time of solids, semisolids, and liquids compared with lean individuals, highlighting the relevant pharmacokinetic implications of oral drug absorption in patients with obesity. MethodsWe searched the Cochrane Library, PubMed, Web of Science, and Embase for all relevant articles published until November 1, 2020. Differences in gastrointestinal variables in relation to gastric emptying between obese and lean individuals were quantified by weighted mean difference (WMD) and ratio of means (RoM). Robustness of the analyses was evaluated by subgroup analysis and publication bias test. FindingsA total of 17 studies with 906 participants were included. The gastric half-emptying time of solids (WMD, −10.4 minutes; P = 0.001; RoM, 0.90; P = 0.01) and liquids (WMD, −6.14 minutes; P < 0.001; RoM, 0.83, P = 0.03) was significantly shorter in individuals with obesity compared with lean individuals. These findings were confirmed by the subgroup analyses and publication bias tests. ImplicationsOur pooled analysis systemically quantifies the differences in gastric half-emptying time between individuals with obesity and lean individuals, facilitating better understanding and prediction of drug absorption in individuals with obesity through physiologically based pharmacokinetic approaches. Obesity is associated with a faster transit of both solids and liquids through the stomach.

Investigation of Self-Emulsifying Drug-Delivery System Interaction with a Biomimetic Membrane under Conditions Relevant to the Small Intestine.

Self-emulsifying drug-delivery systems (SEDDS) have been extensively shown to increase oral absorption of solvation-limited compounds. However, there has been little clinical and commercial use of these formulations, in large part because the demonstrated advantages of SEDDS have been outweighed by our inability to precisely predict drug absorption from SEDDS using current in vitro assays. To overcome this limitation and increase the biological relevancy of in vitro assays, an absorption function can be incorporated using biomimetic membranes. However, the effects that SEDDS have on the integrity of a biomimetic membrane are not known. In this study, a quartz crystal microbalance with dissipation monitoring and total internal reflection fluorescence microscopy were employed as complementary methods to in vitro lipolysis-permeation assays to characterize the interaction of various actively digested SEDDS with a liquescent artificial membrane comprising lecithin in dodecane (LiDo). Observations from surface analysis showed that interactions between the digesting SEDDS and LiDo membrane coincided with inflection points in the digestion profiles. Importantly, no indications of membrane damage could be observed, which was supported by flux profiles of the lipophilic model drug felodipine (FEL) and impermeable marker Lucifer yellow on the basal side of the membrane. There was a correlation between the digestion kinetics of the SEDDS and the flux of FEL, but no clear correlation between solubilization and absorption profiles. Membrane interactions were dependent on the composition of lipids within each SEDDS, with the more digestible lipids leading to more pronounced interactions, but in all cases, the integrity of the membrane was maintained. These insights demonstrate that LiDo membranes are compatible with in vitro lipolysis assays for improving predictions of drug absorption from lipid-based formulations.

Generation of tetracycline-controllable CYP3A4-expressing Caco-2 cells by the piggyBac transposon system

Caco-2 cells are widely used as an in vitro intestinal epithelial cell model because they can form a monolayer and predict drug absorption with high accuracy. However, Caco-2 cells hardly express cytochrome P450 (CYP), a drug-metabolizing enzyme. It is known that CYP3A4 is the dominant drug-metabolizing enzyme in human small intestine. In this study, we generated CYP3A4-expressing Caco-2 (CYP3A4-Caco-2) cells and attempted to establish a model that can simultaneously evaluate drug absorption and metabolism. CYP3A4-Caco-2 cells were generated by piggyBac transposon vectors. A tetracycline-controllable CYP3A4 expression cassette (tet-on system) was stably transduced into Caco-2 cells, thus regulating the levels of CYP3A4 expression depending on the doxycycline concentration. The CYP3A4 expression levels in CYP3A4-Caco-2 cells cultured in the presence of doxycycline were similar to or higher than those of adult small intestine. The CYP3A4-Caco-2 cells had enough ability to metabolize midazolam, a substrate of CYP3A4. CYP3A4 overexpression had no negative effects on cell proliferation, barrier function, and P-glycoprotein activity in Caco-2 cells. Thus, we succeeded in establishing Caco-2 cells with CYP3A4 metabolizing activity comparable to in vivo human intestinal tissue. This cell line would be useful in pharmaceutical studies as a model that can simultaneously evaluate drug absorption and metabolism.

In vitro - in vivo - in silico approach in the development of inhaled drug products: Nanocrystal-based formulations with budesonide as a model drug

This study aims to understand the absorption patterns of three different kinds of inhaled formulations via in silico modeling using budesonide (BUD) as a model drug. The formulations investigated in this study are: (i) commercially available micronized BUD mixed with lactose (BUD-PT), (ii) BUD nanocrystal suspension (BUD-NC), (iii) BUD nanocrystals embedded hyaluronic acid microparticles (BUD-NEM). The deposition patterns of the three inhaled formulations in the rats’ lungs were determined in vivo and in silico predicted, which were used as inputs in GastroPlus™ software to predict drug absorption following aerosolization of the tested formulations. BUD pharmacokinetics, estimated based on intravenous data in rats, was used to establish a drug-specific in silico absorption model. The BUD-specific in silico model revealed that drug pulmonary solubility and absorption rate constant were the key factors affecting pulmonary absorption of BUD-NC and BUD-NEM, respectively. In the case of BUD-PT, the in silico model revealed significant gastrointestinal absorption of BUD, which could be overlooked by traditional in vivo experimental observation. This study demonstrated that in vitro-in vivo-in silico approach was able to identify the key factors that influence the absorption of different inhaled formulations, which may facilitate the development of orally inhaled formulations with different drug release/absorption rates.

3D Printed Stackable Titer Plate Inserts Supporting Three Interconnected Tissue Models for Drug Transport Studies.

Current in vitro drug screening methods often rely on single-cell models and are therefore imprecise in predicting drug absorption, distribution, metabolism, excretion, and toxicity. This study presents a method to fabricate 3D printed inserts that are compatible with commercially available titer plates. Hydrogels can be casted into the inserts and cells can be cultured either in or on the hydrogels. Once individual cell cultures are fully differentiated, the three different cell cultures are stacked on top of each other for biological experiments. To show the possibilities of this approach, three tissue models representing the first pass metabolism is used. The three tissue models are based on gelatin hydrogels and Caco-2, HUVEC, and HepG2 cells to simulate the small intestine, vascular endothelium, and liver, respectively. The device is simple to fabricate, user friendly, and an alternative to microfluidic-based organ on a chip systems. The presented first pass metabolism study allows for gaining information on drug absorption, distribution, metabolism, and, in the future, excretion in one compact device complying the micro titer plate format.

Robotically handled whole-tissue culture system for the screening of oral drug formulations.

Monolayers of cancer-derived cell lines are widely used in the modelling of the gastrointestinal (GI) absorption of drugs and in oral drug development. However, they do not generally predict drug absorption in vivo. Here, we report a robotically handled system that uses large porcine GI tissue explants that are functionally maintained for an extended period in culture for the high-throughput interrogation (several thousand samples per day) of whole segments of the GI tract. The automated culture system provided higher predictability of drug absorption in the human GI tract than a Caco-2 Transwell system (Spearman's correlation coefficients of 0.906 and 0.302, respectively). By using the culture system to analyse the intestinal absorption of 2,930 formulations of the peptide drug oxytocin, we discovered an absorption enhancer that resulted in a 11.3-fold increase in the oral bioavailability of oxytocin in pigs in the absence of cellular disruption of the intestinal tissue. The robotically handled whole-tissue culture system should help advance the development of oral drug formulations and might also be useful for drug screening applications.

Quantitative prediction of human pharmacokinetic responses to drugs via fluidically coupled vascularized organ chips.

Analyses of drug pharmacokinetics (PKs) and pharmacodynamics (PDs) performed in animals are often not predictive of drug PKs and PDs in humans, and in vitro PK and PD modelling does not provide quantitative PK parameters. Here, we show that physiological PK modelling of first-pass drug absorption, metabolism and excretion in humans-using computationally scaled data from multiple fluidically linked two-channel organ chips-predicts PK parameters for orally administered nicotine (using gut, liver and kidney chips) and for intravenously injected cisplatin (using coupled bone marrow, liver and kidney chips). The chips are linked through sequential robotic liquid transfers of a common blood substitute by their endothelium-lined channels (as reported by Novak et al. in an associated Article) and share an arteriovenous fluid-mixing reservoir. We also show that predictions of cisplatin PDs match previously reported patient data. The quantitative in-vitro-to-in-vivo translation of PK and PD parameters and the prediction of drug absorption, distribution, metabolism, excretion and toxicity through fluidically coupled organ chips may improve the design of drug-administration regimens for phase-I clinical trials.

Exploring Multicompartment Plug Flow–Based Model Approach in Biopharmaceutics: Impact of Stomach Setting and the Estimation of the Fraction Absorbed of Orally Administered Basic Drugs

There has been an increasing interest in accurate prediction of human pharmacokinetics of drug candidates to reduce cost and increase productivity during research and development. Modeling efforts have primarily focused on predicting drug absorption after oral administration because it is the most desired route for small molecule drug delivery. Despite significant progress in the field, the fraction of dose absorbed (Fa) is still considered to be a challenging parameter to predict. In recent years, compartment and transit models have become increasingly popular because of their effectiveness. A multicompartment plug flow–based model in which the stomach is assigned as a single compartment has been built. However, this model was found to be less accurate in estimating stomach solubilization of basic drugs under certain conditions and leads to false-negative results. Therefore, a modified multicompartment approach was developed by dividing the stomach into 4 compartments allowing the model to better mimic the physiological conditions. This approach was found to be more precise in estimating Fa for basic drugs compared to the previous approach. Based on this finding, it is believed the aforementioned approach should also be applicable to all compartmental-based models when similar issues are encountered.

Erratum: Scalable Fabrication of Stretchable, Dual Channel, Microfluidic Organ Chips.

Erratum: Scalable Fabrication of Stretchable, Dual Channel, Microfluidic Organ Chips.

Quantitative Estimation of the Effect of Nasal Mucociliary Function on in Vivo Absorption of Norfloxacin after Intranasal Administration to Rats.

Nasal drug delivery has attracted significant attention as an alternative route to deliver drugs having poor bioavailability. Large-molecule drugs, such as peptides and central nervous system drugs, would benefit from intranasal delivery. Drug absorption after intranasal application depends on the nasal retention of the drug, which is determined by the nasal mucociliary clearance. Mucociliary clearance (MC) is an important determinant of the rate and extent of nasal drug absorption. The aim of the present study was to clarify the effect of the changes in MC on in vivo drug absorption after nasal application, and to justify the pharmacokinetic model to which the MC parameter was introduced, to enable prediction of bioavailability after intranasal administration. The pharmacokinetics of norfloxacin (NFX) after intranasal administration were evaluated following the modification of nasal MC by pretreatment with the MC inhibitors propranolol and atropine and the MC enhancers terbutaline and acetylcholine chloride. From the relationship between nasal MC and bioavailability after nasal application, prediction of drug absorption was attempted on the basis of our pharmacokinetic model. Propranolol and atropine enhanced the bioavailability of NFX by 90 and 40%, respectively, while the bioavailability decreased by 30% following terbutaline and 40% following acetylcholine chloride. As a result of changes in the MC function, nasal drug absorption was changed depending on the nasal residence time of the drug. On the basis of our pharmacokinetic model, the nasal drug absorption can be precisely predicted, even when the MC is changed. This prediction system allows the quantitative evaluation of changes in drug absorption due to changes in nasal MC and is expected to contribute greatly to the development of nasal formulations.

Application of Intestinal Epithelial Cells Differentiated from Human Induced Pluripotent Stem Cells for Studies of Prodrug Hydrolysis and Drug Absorption in the Small Intestine

Cell models to investigate intestinal absorption functions, such as those of transporters and metabolic enzymes, are essential for oral drug discovery and development. The purpose of this study was to generate intestinal epithelial cells from human induced pluripotent stem cells (hiPSC-IECs) and then clarify whether the functions of hydrolase and transporters in them reflect oral drug absorption in the small intestine. The hiPSC-IECs showed the transport activities of P-glycoprotein (P-gp), breast cancer resistance protein (BCRP), and peptide transporter 1 (PEPT1), revealed by using their probe substrates ([3H]digoxin, sulfasalazine, and [14C]glycylsarcosine), and the metabolic activities of CYP3A4, CES2, and CES1, which were clarified using their probe substrates (midazolam, irinotecan, and temocapril). The intrinsic clearance by hydrolysis of six ester prodrugs into the active form in hiPSC-IECs was correlated with the plasma exposure (Cmax , AUC, and bioavailability) of the active form after oral administration of these prodrugs to rats. Also, the permeability coefficients of 14 drugs, containing two substrates of P-gp (doxorubicin and [3H]digoxin), one substrate of BCRP (sulfasalazine), and 11 nonsubstrates of transporters (ganciclovir, [14C]mannitol, famotidine, sulpiride, atenolol, furosemide, ranitidine, hydrochlorothiazide, acetaminophen, propranolol, and antipyrine) in hiPSC-IECs were correlated with their values of the fraction of intestinal absorption (Fa) in human clinical studies. These findings suggest that hiPSC-IECs would be a useful cell model to investigate the hydrolysis of ester prodrugs and to predict drug absorption in the small intestine.