Computational approaches to analyse and predict small molecule transport and distribution at cellular and subcellular levels

Physiologically-Based Pharmacokinetic Modeling Analysis for Quantitative Prediction of Renal Transporter-Mediated Interactions Between Metformin and Cimetidine

Metformin is an important antidiabetic drug and often used as a probe for drug–drug interactions (DDIs) mediated by renal transporters. Despite evidence supporting the inhibition of multidrug and toxin extrusion proteins as the likely DDI mechanism, the previously reported physiologically‐based pharmacokinetic (PBPK) model required the substantial lowering of the inhibition constant values of cimetidine for multidrug and toxin extrusion proteins from those obtained in vitro to capture the clinical DDI data between metformin and cimetidine.¹ We constructed new PBPK models in which the transporter‐mediated uptake of metformin is driven by a constant membrane potential. Our models successfully captured the clinical DDI data using in vitro inhibition constant values and supported the inhibition of multidrug and toxin extrusion proteins by cimetidine as the DDI mechanism upon sensitivity analysis and data fitting. Our refined PBPK models may facilitate prediction approaches for DDI involving metformin using in vitro inhibition constant values.

An Expandable Mechanopharmaceutical Device (1): Measuring the Cargo Capacity of Macrophages in a Living Organism

PURPOSE

Clofazimine (CFZ) is an FDA-approved, poorly soluble small molecule drug that precipitates as crystal-like drug inclusions (CLDIs) which accumulate in acidic cytoplasmic organelles of macrophages. In this study, we considered CLDIs as an expandable mechanopharmaceutical device, to study how macrophages respond to an increasingly massive load of endophagolysosomal cargo.

METHODS

First, we experimentally tested how the accumulation of CFZ in CLDIs impacted different immune cell subpopulations of different organs. Second, to further investigate the mechanism of CLDI formation, we asked whether specific accumulation of CFZ hydrochloride crystals in lysosomes could be explained as a passive, thermodynamic equilibrium phenomenon. A cellular pharmacokinetic model was constructed, simulating CFZ accumulation driven by pH-dependent ion trapping of the protonated drug in the acidic lysosomes, followed by the precipitation of CFZ hydrochloride salt via a common ion effect caused by high chloride concentrations.

RESULTS

While lower loads of CFZ were mostly accommodated in lung macrophages, increased CFZ loading was accompanied by organ-specific changes in macrophage numbers, size and intracellular membrane architecture, maximizing the cargo storage capabilities. With increasing loads, the total cargo mass and concentrations of CFZ in different organs diverged, while that of individual macrophages converged. The simulation results support the notion that the proton and chloride ion concentrations of macrophage lysosomes are sufficient to drive the massive, cell type-selective accumulation and growth of CFZ hydrochloride biocrystals.

CONCLUSION

CLDIs effectively function as an expandable mechanopharmaceutical device, revealing the coordinated response of the macrophage population to an increasingly massive, whole-organism endophagolysosomal cargo load.

Exosome is a mechanism of intercellular drug transfer: Application of quantitative pharmacology

PURPOSE

Exosomes are small membrane vesicles (30-100nm in diameter) secreted by cells into extracellular space. The present study evaluated the effect of chemotherapeutic agents on exosome production and/or release, and quantified the contribution of exosomes to intercellular drug transfer and pharmacodynamics.

METHODS

Human cancer cells (breast MCF7, breast-to-lung metastatic LM2, ovarian A2780 and OVCAR4) were treated with paclitaxel (PTX, 2-1000nM) or doxorubicin (DOX, 20-1000nM) for 24-48h. Exosomes were isolated from the culture medium of drug-treated donor cells (Donor cells) using ultra-centrifugation, and analyzed for acetylcholinesterase activity, total proteins, drug concentrations, and biological effects (cytotoxicity and anti-migration) on drug-naïve recipient cells (Recipient cells). These results were used to develop computational predictive quantitative pharmacology models.

RESULTS

Cells in exponential growth phase released ~220 exosomes/cell in culture medium. PTX and DOX significantly promoted exosome production and/or release in a dose- and time-dependent manner, with greater effects in ovarian cancer cells than in breast cancer cells. Exosomes isolated from Donor cells contained appreciable drug levels (2-7pmole/10⁶ cells after 24h treatment with 100-1000nM PTX), and caused cytotoxicity and inhibited migration of Recipient cells. Quantitative pharmacology models that integrated cellular PTX pharmacokinetics with PTX pharmacodynamics successfully predicted effects of exosomes on intercellular drug transfer, cytotoxicity of PTX on Donor cells and cytotoxicity of PTX-containing exosomes on Recipient cells. Additional model simulations indicate that within clinically achievable PTX concentrations, the contribution of exosomes to active drug efflux increased with drug concentration and exceeded the p-glycoprotein efflux when the latter was saturated.

CONCLUSIONS

Our results indicate (a) chemotherapeutic agents stimulate exosome production or release, and (b) exosome is a mechanism of intercellular drug transfer that contributes to pharmacodynamics of neighboring cells.

In silico prediction of chemical subcellular localization via multi-classification methods

Chemical subcellular localization is closely related to drug distribution in the body and hence important in drug discovery and design. Although many in vivo and in vitro methods have been developed, in silico methods play key roles in the prediction of chemical subcellular localization due to their low costs and high performance. For that purpose, machine learning-based methods were developed here. At first, 614 unique compounds localized in the lysosome, mitochondria, nucleus and plasma membrane were collected from the literature. 80% of the compounds were used to build the models and the rest as the external validation set. Both fingerprints and molecular descriptors were used to describe the molecules, and six machine learning methods were applied to build the multi-classification models. The performance of the models was measured by 5-fold cross-validation and external validation. We further detected key substructures for each localization and analyzed potential structure-localization relationships, which could be very helpful for molecular design and modification. The key substructures can also be used as features complementary to fingerprints to improve the performance of the models.

Establishing health benefits of bioactive food components: a basic research scientist's perspective

Bioactive food components or functional foods have recently received significant attention because of their widely touted positive effects on health beyond basic nutrition. However, a question continues to lurk: are these claims for 'super foods' backed by sound science or simply an exaggerated portrayal of very small 'benefits'? Efforts to establish health benefits by scientific means pose a real challenge in regards to defining what those benefits are, as well as how effective the foods are in justifying any health claim. This review discusses the pitfalls associated with the execution, interpretation, extrapolation of the results to humans and the challenges encountered in the dietary research arena from a basic scientist's perspective.

Predicting volume of distribution with decision tree-based regression methods using predicted tissue:plasma partition coefficients

BACKGROUND

Volume of distribution is an important pharmacokinetic property that indicates the extent of a drug's distribution in the body tissues. This paper addresses the problem of how to estimate the apparent volume of distribution at steady state (Vss) of chemical compounds in the human body using decision tree-based regression methods from the area of data mining (or machine learning). Hence, the pros and cons of several different types of decision tree-based regression methods have been discussed. The regression methods predict Vss using, as predictive features, both the compounds' molecular descriptors and the compounds' tissue:plasma partition coefficients (Kt:p) - often used in physiologically-based pharmacokinetics. Therefore, this work has assessed whether the data mining-based prediction of Vss can be made more accurate by using as input not only the compounds' molecular descriptors but also (a subset of) their predicted Kt:p values.

RESULTS

Comparison of the models that used only molecular descriptors, in particular, the Bagging decision tree (mean fold error of 2.33), with those employing predicted Kt:p values in addition to the molecular descriptors, such as the Bagging decision tree using adipose Kt:p (mean fold error of 2.29), indicated that the use of predicted Kt:p values as descriptors may be beneficial for accurate prediction of Vss using decision trees if prior feature selection is applied.

CONCLUSIONS

Decision tree based models presented in this work have an accuracy that is reasonable and similar to the accuracy of reported Vss inter-species extrapolations in the literature. The estimation of Vss for new compounds in drug discovery will benefit from methods that are able to integrate large and varied sources of data and flexible non-linear data mining methods such as decision trees, which can produce interpretable models. Graphical AbstractDecision trees for the prediction of tissue partition coefficient and volume of distribution of drugs.