Improving In Vitro-In Vivo Extrapolation of Clearance Using Rat Liver Microsomes for Highly Plasma Protein-Bound Molecules.

It is common practice in drug discovery and development to predict in vivo hepatic clearance from in vitro incubations with liver microsomes or hepatocytes using the well-stirred model (WSM). When applying the WSM to a set of approximately 3000 Novartis research compounds, 73% of neutral and basic compounds (extended clearance classification system [ECCS] class 2) were well-predicted within 3-fold. In contrast, only 44% (ECCS class 1A) or 34% (ECCS class 1B) of acids were predicted within 3-fold. To explore the hypothesis whether the higher degree of plasma protein binding for acids contributes to the in vitro-in vivo correlation (IVIVC) disconnect, 68 proprietary compounds were incubated with rat liver microsomes in the presence and absence of 5% plasma. A minor impact of plasma on clearance IVIVC was found for moderately bound compounds (fraction unbound in plasma [fup] ≥1%). However, addition of plasma significantly improved the IVIVC for highly bound compounds (fup <1%) as indicated by an increase of the average fold error from 0.10 to 0.36. Correlating fup with the scaled unbound intrinsic clearance ratio in the presence or absence of plasma allowed the establishment of an empirical, nonlinear correction equation that depends on fup Taken together, estimation of the metabolic clearance of highly bound compounds was enhanced by the addition of plasma to microsomal incubations. For standard incubations in buffer only, application of an empirical correction provided improved clearance predictions. SIGNIFICANCE STATEMENT: Application of the well-stirred liver model for clearance in vitro-in vivo extrapolation (IVIVE) in rat generally underpredicts the clearance of acids and the strong protein binding of acids is suspected to be one responsible factor. Unbound intrinsic in vitro clearance (CLint,u) determinations using rat liver microsomes supplemented with 5% plasma resulted in an improved IVIVE. An empirical equation was derived that can be applied to correct CLint,u-values in dependance of fraction unbound in plasma (fup) and measured CLint in buffer.

Multispecies Machine Learning Predictions of In Vitro Intrinsic Clearance with Uncertainty Quantification Analyses.

In pharmaceutical research, compounds are optimized for metabolic stability to avoid a too fast elimination of the drug. Intrinsic clearance (CLint) measured in liver microsomes or hepatocytes is an important parameter during lead optimization. In this work, machine learning models were developed to relate the compound structure to microsomal metabolic stability and predict CLint for new compounds. A multitask (MT) learning architecture was introduced to model the CLint of six species simultaneously, giving as a result a multispecies machine learning model. MT graph neural network (MT-GNN) regression was identified as the top-performing method, and an ensemble of 10 MT-GNN models was evaluated prospectively. Geometric mean fold errors were consistently smaller than 2-fold. Moreover, high precision values were obtained in the prediction of "high" (>300 µL/min/mg) and "low" (<100 µL/min/mg) CLint compounds. Precision values ranged from 80 to 94% for low CLint predictions and from 75 to 97% for high CLint predictions, depending on the species. Uncertainty on experimental values and model predictions was systematically quantified. Experimental variability (aleatoric uncertainty) of all historical Novartis in vitro clearance experiments was analyzed. Interestingly, MT-GNN models' performance approached assays' experimental variability. Moreover, uncertainty estimation in predictions (epistemic uncertainty) enabled identifying predictions associated with lower and higher error. Taken together, our manuscript combines a multispecies deep learning model and large-scale uncertainty analyses to improve CLint predictions and facilitate early informed decisions for compound prioritization.

Time matters - in vitro cellular disposition kinetics help rationalizing cellular potency disconnects.

Loss in potency is commonly observed in early drug discovery when moving from biochemical to more complex cellular systems. Among other factors, low permeability is often considered to cause such potency disconnects.We developed a novel cellular disposition assay in MDCK cells to determine passive uptake clearance (PSinf), cell-to-medium ratios at steady-state (Kp) and the time to reach 90% steady-state (TTSS90) from a single experiment in a high-throughput format.The assay was validated using 40 marketed drugs, showing a wide distribution of PSinf and Kp values. The parameters generally correlated with transcellular permeability and lipophilicity, while PSinf data revealed better resolution in the high and low permeability ranges compared to traditional permeability data. A linear relationship between the Kp/PSinf ratio and TTSS90 was mathematically derived and experimentally validated, demonstrating the dependency of TTSS90 on the rate and extent of cellular accumulation.Cellular disposition parameters could explain potency (IC50) disconnects noted for seven Bruton's tyrosine kinase degrader compounds in a cellular potency assay. In contrast to transcellular permeability, PSinf data enabled identification of the compounds with IC50 disconnects based on their time to reach equilibrium. Overall, the novel assay offers the possibility to address potency disconnects in early drug discovery.

Prediction of Fraction Unbound in Microsomal and Hepatocyte Incubations: A Comparison of Methods across Industry Datasets.

The fraction unbound in the incubation, fu,inc, is an important parameter to consider in the evaluation of intrinsic clearance measurements performed in vitro in hepatocytes or microsomes. Reliable estimates of fu,inc based on a compound's structure have the potential to positively impact the screening timelines in drug discovery. Previous works suggested that fu,inc is primarily driven by passive processes and can be described using physicochemical properties such as lipophilicity and the protonation state of the molecule. While models based on these principles proved predictive in relatively small datasets that included marketed drugs, their applicability domain has not been extensively explored. The work presented here from the in silico ADME discussion group (part of the International Consortium for Innovation through Quality in Pharmaceutical Development, the IQ consortium) describes the accuracy of these models in large proprietary datasets that include several thousand of compounds across chemical space. Overall, the models do well for compounds with low lipophilicity. In other words, the equations correctly predict that fu,inc is, in general, above 0.5 for compounds with a calculated logP of less than 3. When applied to lipophilic compounds, the models failed to produce quantitatively accurate predictions of fu,inc, with a high risk of underestimating binding properties. These models can, therefore, be used quantitatively for less lipophilic compounds. On the other hand, internal machine-learning models using a company's own proprietary dataset also predict compounds with higher lipophilicity reasonably well. Additionally, the data shown indicate that microsomal binding is, in general, a good proxy for hepatocyte binding.

Benefit of Retraining pKa Models Studied Using Internally Measured Data.

The ionization state of drugs influences many pharmaceutical properties such as their solubility, permeability, and biological activity. It is therefore important to understand the structure property relationship for the acid-base dissociation constant pKa during the lead optimization process to make better-informed design decisions. Computational approaches, such as implemented in MoKa, can help with this; however, they often predict with too large error especially for proprietary compounds. In this contribution, we look at how retraining helps to greatly improve prediction error. Using a longitudinal study with data measured over 15 years in a drug discovery environment, we assess the impact of model training on prediction accuracy and look at model degradation over time. Using the MoKa software, we will demonstrate that regular retraining is required to address changes in chemical space leading to model degradation over six to nine months.

Passive lipoidal diffusion and carrier-mediated cell uptake are both important mechanisms of membrane permeation in drug disposition.

Recently, it has been proposed that drug permeation is essentially carrier-mediated only and that passive lipoidal diffusion is negligible. This opposes the prevailing hypothesis of drug permeation through biological membranes, which integrates the contribution of multiple permeation mechanisms, including both carrier-mediated and passive lipoidal diffusion, depending on the compound's properties, membrane properties, and solution properties. The prevailing hypothesis of drug permeation continues to be successful for application and prediction in drug development. Proponents of the carrier-mediated only concept argue against passive lipoidal diffusion. However, the arguments are not supported by broad pharmaceutics literature. The carrier-mediated only concept lacks substantial supporting evidence and successful applications in drug development.

CYP3A time-dependent inhibition risk assessment validated with 400 reference drugs.

Although reversible CYP3A inhibition testing is well established for predicting the drug-drug interaction potential of clinical candidates, time-dependent inhibition (TDI) has become the focus of drug designers only recently. Failure of several late-stage clinical candidates has been attributed to TDI, and this mechanism is also suspected to play a role in liver toxicities often observed in preclinical species. Measurement of enzyme inactivation rates (k(inact) and K(I)) is technically challenging, and a great deal of variability can be found in the literature. In this article, we have evaluated the TDI potential for 400 registered drugs using a high-throughput assay format based on determination of the inactivation rate (k(obs)) at a single concentration of test compound (10 µM). The advantages of this new assay format are highlighted by comparison with data generated using the IC50 shift assay, a current standard approach for preliminary assessment of TDI. With use of an empirically defined positive/negative k(obs) bin of 0.02 min⁻¹, only 4% of registered drugs were found to be positive. This proportion increased to more than 20% when in-house lead optimization molecules were considered, emphasizing the importance of identifying this property in selection of promising drug candidates. Finally, it is suggested that the data and technology described here may be a good basis for building structure-activity relationships and in silico modeling.

Predicting Oral Absorption for Compounds Outside the Rule of Five Property Space.

The estimation of the extent of absorption of drug candidates intended for oral drug delivery is an important selection criteria in drug discovery. The use of cell-based transwell assays examining flux across cell-monolayers (e.g., Caco-2 or MDCK cells) usually provide satisfactory predictions of the extent of absorption in vivo. These predictions often fall short of expection for molecules outside the traditional low molecular weight property space. In this manuscript the transwell permeability assay was modified to circumvent potential issues that can be encountered when evaluating the aforementioned drug molecules. Particularly, the addition of albumin in the acceptor compartment to reduce potential binding to cells and the acceptor compartment, improved the predictive power of the assay. Cellular binding and lysosomal trapping effects are significantly reduced for larger molecules, particularly lipophilic bases under these more physiological conditions, resulting in higher recovery values and a better prediction power. The data indicate that lysosomal trapping does not impact the rate of absorption of lipophilic bases in general but is rather an exception. Finally, compounds believed to permeate by passive mechanisms were used in a calibration curve for the effective prediction of the fraction absorbed of molecules of interest in current medicinal chemistry efforts.

Artificial membrane assays to assess permeability.

This paper reviews the development of artificial membrane assays in the last decade. Reasons why parallel artificial membrane assays (PAMPA) became widely used are discussed and the various PAMPA assays targeting gastro-intestinal absorption, blood brain barrier and skin penetration are presented. Improvements in the assay technology like the introduction of a paracellular component and the critical factors to get quality data are reviewed. The question how does PAMPA compare with Caco-2 monolayer permeability is being addressed. New dimensions in artificial membrane assays like octanol/water logP measurement, influence of excipients on solubility/permeability and binding constants measurements are introduced. Finally, similarity and differences between partition coefficients and permeability values are discussed.

New approach to measure protein binding based on a parallel artificial membrane assay and human serum albumin.

We report here a new, label-free approach to measure serum protein binding constants. The assay is able to measure HSA K d values in the milli-molar to micromolar range. The protein is not immobilized on any surface and the assay self-corrects for nonspecific adsorption. No mass balance is required to get accurate binding constants and it is not necessary to wait for equilibrium to extract the binding constant. The assay runs in a 96-well format using commercially available parts and is, therefore, relatively easy to implement and automate. As the chemical membranes used are not water permeable, there is no volume change due to the osmotic pressure and pretreatment (soaking) is not necessary. The concept can potentially be extended to other proteins and could thus serve as a label-free technique for general binding constant measurements.

Drug permeability profiling using cell-free permeation tools: Overview and applications.

Cell-free permeation systems are gaining interest in drug discovery and development as tools to obtain a reliable prediction of passive intestinal absorption without the disadvantages associated with cell- or tissue-based permeability profiling. Depending on the composition of the barrier, cell-free permeation systems are classified into two classes including (i) biomimetic barriers which are constructed from (phospho)lipids and (ii) non-biomimetic barriers containing dialysis membranes. This review provides an overview of the currently available cell-free permeation systems including Parallel Artificial Membrane Permeability Assay (PAMPA), Phospholipid Vesicle-based Permeation Assay (PVPA), Permeapad®, and artificial membrane based systems (e.g. the artificial membrane insert system (AMI-system)) in terms of their barrier composition as well as their predictive capacity in relation to well-characterized intestinal permeation systems. Given the potential loss of integrity of cell-based permeation barriers in the presence of food components or pharmaceutical excipients, the superior robustness of cell-free barriers makes them suitable for the combined dissolution/permeation evaluation of formulations. While cell-free permeation systems are mostly applied for exploring intestinal absorption, they can also be used to evaluate non-oral drug delivery by adjusting the composition of the membrane.

Computational approaches to determine drug solubility.

Water solubility is an important molecular property for successful drug development as it is a key factor governing drug access to biological membranes. There have been a number of review articles addressing computational models to predict water solubility with emphasis on the accuracy of the various prediction methods. This paper briefly reviews the available models and focuses on the value which can be extracted by comparing calculated and measured solubility, discusses the potential and limitations of the main computational approaches, and provides guidelines as to when to trust the computed value.

PAMPA--critical factors for better predictions of absorption.

PAMPA, log P(OCT), and Caco-2 are useful tools in drug discovery for the prediction of oral absorption, brain penetration and for the development of structure-permeability relationships. Each approach has its advantages and limitations. Selection criteria for methods are based on many different factors: predictability, throughput, cost and personal preferences (people factor). The PAMPA concerns raised by Galinis-Luciani et al. (Galinis-Luciani et al., 2007, J Pharm Sci, this issue) are answered by experienced PAMPA practitioners, inventors and developers from diverse research organizations. Guidelines on how to use PAMPA are discussed. PAMPA and PAMPA-BBB have much better predictivity for oral absorption and brain penetration than log P(OCT) for real-world drug discovery compounds. PAMPA and Caco-2 have similar predictivity for passive oral absorption. However, it is not advisable to use PAMPA to predict absorption involving transporter-mediated processes, such as active uptake or efflux. Measurement of PAMPA is much more rapid and cost effective than Caco-2 and log P(OCT). PAMPA assay conditions are critical in order to generate high quality and relevant data, including permeation time, assay pH, stirring, use of cosolvents and selection of detection techniques. The success of using PAMPA in drug discovery depends on careful data interpretation, use of optimal assay conditions, implementation and integration strategies, and education of users.

High-throughput in vitro profiling assays: lessons learnt from experiences at Novartis.

This article reviews the use of a selection of in vitro assays implemented at Novartis and intends to address exposure and safety in early drug discovery. The authors' own experience, based on a large number of 'real' drug discovery compounds, is described to reflect on what has worked, where improvement is needed and how to best use the data for decision making. Possible strategies are discussed, and guidelines are provided on how to organise assays, extract value and contribute knowledge from the data.

High-throughput lipophilicity measurement with immobilized artificial membranes.

We report on a new, high-throughput assay designed to measure octanol/water partition coefficients in early drug discovery. The assay is carried out in 96-well microtiterplates and measures the diffusion of compounds between two aqueous compartments separated by a thin octanol liquid layer. Octanol/water partition coefficients are derived from the apparent permeability (P(a)) values using a calibration curve. The assay can measure partition coefficients within the range -2 to + 8; thus, a dynamic range of 10 log units can be covered in one single run. Unlike chromatographic methods, the technology is not restricted to neutral and weakly basic compounds, and, as no stationary phase is involved, the data can be strictly compared with values obtained from traditional methods such as shake-flask/HPLC or dual-phase potentiometric titration.

Evidence-based approach to assess passive diffusion and carrier-mediated drug transport.

Evidence supporting the action of passive diffusion and carrier-mediated (CM) transport in drug bioavailability and disposition is discussed to refute the recently proposed theory that drug transport is CM-only and that new transporters will be discovered that possess transport characteristics ascribed to passive diffusion. Misconceptions and faulty speculations are addressed to provide reliable guidance on choosing appropriate tools for drug design and optimization.

Evolution of the physicochemical properties of marketed drugs: can history foretell the future?

A set of diverse bioactive molecules, relevant from a medicinal chemistry viewpoint, was assembled and used to navigate the physicochemical property space of new and old, or traditional drugs against a larger set of 12,000 diverse bioactive small molecules. Most drugs on the market only occupy a fraction of the property space of the bioactive molecules, whereas new molecular entities (NMEs) approved since 2002 are moving away from this traditional drug space. In this new territory, semi-empirical rules derived from knowledge accumulated from historic, older molecules are not necessarily valid and different liabilities become more prominent.

Structure-activity relationship and pharmacokinetic studies of sotrastaurin (AEB071), a promising novel medicine for prevention of graft rejection and treatment of psoriasis.

Protein kinase C (PKC) isotypes have emerged as key targets for the blockade of early T-cell activation. Herein, we report on the structure-activity relationship and the detailed physicochemical and in vivo pharmacokinetic properties of sotrastaurin (AEB071, 1), a novel maleimide-based PKC inhibitor currently in phase II clinical trials. Most notably, the preferred uptake of sotrastaurin into lymphoid tissues is an important feature, which is likely to contribute to its in vivo efficacy.

What is modulating solubility in simulated intestinal fluids?

The aim of this study was to understand which parameters are responsible for the selective modulation of compounds solubility in simulated intestinal fluids. The solubility of 25 chemically diverse reference compounds was measured in simulated intestinal fluid (FaSSIF-V2) and in aqueous phosphate and maleate buffers. Electrostatic interactions between compounds and the bio-relevant medium components seem to explain the different solubility behavior observed for acids and bases. The solubility of ionized acids is not increased in FaSSIF-V2 probably due to electrostatic repulsions with the media components. Lipophilicity plays an important role but mainly for charged bases with a logP>4 (or logD(6.5)>1.9). When the aqueous solubility is mainly driven by lipophilicity, the FaSSIF-V2 components seem to improve the solubility of basic compounds to a greater extent than for compounds whose solubility is limited by crystal packing. These results suggest that ionization, lipophilicity and crystal packing play important but peculiar roles in controlling solubility in FaSSIF-V2 compared to that in aqueous buffer and this information could be useful to guide medicinal chemists and formulation scientists.