Modelling of the cancer cell cycle as a tool for rational drug development: A systems pharmacology approach to cyclotherapy.

The dynamic of cancer is intimately linked to a dysregulation of the cell cycle and signalling pathways. It has been argued that selectivity of treatments could exploit loss of checkpoint function in cancer cells, a concept termed "cyclotherapy". Quantitative approaches that describe these dysregulations can provide guidance in the design of novel or existing cancer therapies. We describe and illustrate this strategy via a mathematical model of the cell cycle that includes descriptions of the G1-S checkpoint and the spindle assembly checkpoint (SAC), the EGF signalling pathway and apoptosis. We incorporated sites of action of four drugs (palbociclib, gemcitabine, paclitaxel and actinomycin D) to illustrate potential applications of this approach. We show how drug effects on multiple cell populations can be simulated, facilitating simultaneous prediction of effects on normal and transformed cells. The consequences of aberrant signalling pathways or of altered expression of pro- or anti-apoptotic proteins can thus be compared. We suggest that this approach, particularly if used in conjunction with pharmacokinetic modelling, could be used to predict effects of specific oncogene expression patterns on drug response. The strategy could be used to search for synthetic lethality and optimise combination protocol designs.

Combenefit: an interactive platform for the analysis and visualization of drug combinations.

MOTIVATION: Many drug combinations are routinely assessed to identify synergistic interactions in the attempt to develop novel treatment strategies. Appropriate software is required to analyze the results of these studies. RESULTS: We present Combenefit, new free software tool that enables the visualization, analysis and quantification of drug combination effects in terms of synergy and/or antagonism. Data from combinations assays can be processed using classical Synergy models (Loewe, Bliss, HSA), as single experiments or in batch for High Throughput Screens. This user-friendly tool provides laboratory scientists with an easy and systematic way to analyze their data. The companion package provides bioinformaticians with critical implementations of routines enabling the processing of combination data. AVAILABILITY AND IMPLEMENTATION: Combenefit is provided as a Matlab package but also as standalone software for Windows (http://sourceforge.net/projects/combenefit/). CONTACT: Giovanni.DiVeroli@cruk.cam.ac.uk SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

An automated fitting procedure and software for dose-response curves with multiphasic features.

In cancer pharmacology (and many other areas), most dose-response curves are satisfactorily described by a classical Hill equation (i.e. 4 parameters logistical). Nevertheless, there are instances where the marked presence of more than one point of inflection, or the presence of combined agonist and antagonist effects, prevents straight-forward modelling of the data via a standard Hill equation. Here we propose a modified model and automated fitting procedure to describe dose-response curves with multiphasic features. The resulting general model enables interpreting each phase of the dose-response as an independent dose-dependent process. We developed an algorithm which automatically generates and ranks dose-response models with varying degrees of multiphasic features. The algorithm was implemented in new freely available Dr Fit software (sourceforge.net/projects/drfit/). We show how our approach is successful in describing dose-response curves with multiphasic features. Additionally, we analysed a large cancer cell viability screen involving 11650 dose-response curves. Based on our algorithm, we found that 28% of cases were better described by a multiphasic model than by the Hill model. We thus provide a robust approach to fit dose-response curves with various degrees of complexity, which, together with the provided software implementation, should enable a wide audience to easily process their own data.

A new classifier-based strategy for in-silico ion-channel cardiac drug safety assessment.

There is currently a strong interest in using high-throughput in-vitro ion-channel screening data to make predictions regarding the cardiac toxicity potential of a new compound in both animal and human studies. A recent FDA think tank encourages the use of biophysical mathematical models of cardiac myocytes for this prediction task. However, it remains unclear whether this approach is the most appropriate. Here we examine five literature data-sets that have been used to support the use of four different biophysical models and one statistical model for predicting cardiac toxicity in numerous species using various endpoints. We propose a simple model that represents the balance between repolarisation and depolarisation forces and compare the predictive power of the model against the original results (leave-one-out cross-validation). Our model showed equivalent performance when compared to the four biophysical models and one statistical model. We therefore conclude that this approach should be further investigated in the context of early cardiac safety screening when in-vitro potency data is generated.

hERG inhibitors with similar potency but different binding kinetics do not pose the same proarrhythmic risk: implications for drug safety assessment.

INTRODUCTION: Since the discovery of the link that exists between drug-induced hERG inhibition and Torsade de Pointes (TdP), extreme attention has been given to avoid new drugs inhibiting this channel. hERG inhibition is routinely screened for in new drugs and, typically, IC50 values are compared to projected plasma concentrations to define a safety margin. METHODS AND RESULTS: We aimed to show that drugs with similar hERG potency are not uniformly pro-arrhythmic-this depends on the drug binding kinetics and mode of action (trapped or not) rather than the IC50 value only. We used a mathematical model of hERG and its related encoded current IKr to simulate drug binding in different configurations. Expression systems mimicking the screening process were first investigated. hERG model was then incorporated into a canine action potential (AP) and tissue model to study the impact of drug binding configurations on AP and pseudo-ECG (QT interval prolongation). Our data show that: (1) trapped and not trapped configurations and different binding kinetics could be identified during hERG screening; (2) slow binding, not trapped drugs, induced less AP prolongation and minimal QT interval prolongation (4.7%) at a concentration equal to the IC50 whereas maximal pro-arrhythmic risk was observed for trapped drugs at the same concentration (QT interval prolongation, 23.1%). CONCLUSION: Our study demonstrates the need for screening for hERG binding configurations rather than potency alone. It also demonstrates the potential link between hERG, drug mode of action and TdP, and the need to question the current regulatory guidance.

High-throughput screening of drug-binding dynamics to HERG improves early drug safety assessment.

The use of computational models to predict drug-induced changes in the action potential (AP) is a promising approach to reduce drug safety attrition but requires a better representation of more complex drug-target interactions to improve the quantitative prediction. The blockade of the human ether-a-go-go-related gene (HERG) channel is a major concern for QT prolongation and Torsade de Pointes risk. We aim to develop quantitative in-silico AP predictions based on a new electrophysiological protocol (suitable for high-throughput HERG screening) and mathematical modeling of ionic currents. Electrophysiological recordings using the IonWorks device were made from HERG channels stably expressed in Chinese hamster ovary cells. A new protocol that delineates inhibition over time was applied to assess dofetilide, cisapride, and almokalant effects. Dynamic effects displayed distinct profiles for these drugs compared with concentration-effects curves. Binding kinetics to specific states were identified using a new HERG Markov model. The model was then modified to represent the canine rapid delayed rectifier K(+) current at 37°C and carry out AP predictions. Predictions were compared with a simpler model based on conductance reduction and were found to be much closer to experimental data. Improved sensitivity to concentration and pacing frequency variables was obtained when including binding kinetics. Our new electrophysiological protocol is suitable for high-throughput screening and is able to distinguish drug-binding kinetics. The association of this protocol with our modeling approach indicates that quantitative predictions of AP modulation can be obtained, which is a significant improvement compared with traditional conductance reduction methods.