Classical structural identifiability methodology applied to low-dimensional dynamic systems in receptor theory.

Mathematical modelling has become a key tool in pharmacological analysis, towards understanding dynamics of cell signalling and quantifying ligand-receptor interactions. Ordinary differential equation (ODE) models in receptor theory may be used to parameterise such interactions using timecourse data, but attention needs to be paid to the theoretical identifiability of the parameters of interest. Identifiability analysis is an often overlooked step in many bio-modelling works. In this paper we introduce structural identifiability analysis (SIA) to the field of receptor theory by applying three classical SIA methods (transfer function, Taylor Series and similarity transformation) to ligand-receptor binding models of biological importance (single ligand and Motulsky-Mahan competition binding at monomers, and a recently presented model of a single ligand binding at receptor dimers). New results are obtained which indicate the identifiable parameters for a single timecourse for Motulsky-Mahan binding and dimerised receptor binding. Importantly, we further consider combinations of experiments which may be performed to overcome issues of non-identifiability, to ensure the practical applicability of the work. The three SIA methods are demonstrated through a tutorial-style approach, using detailed calculations, which show the methods to be tractable for the low-dimensional ODE models.

Insights into the dynamics of ligand-induced dimerisation via mathematical modelling and analysis.

The vascular endothelial growth factor (VEGF) receptor (VEGFR) system plays a role in cancer and many other diseases. It is widely accepted that VEGFR receptors dimerise in response to VEGF binding. However, analysis of these mechanisms and their implications for drug development still requires further exploration. In this paper, we present a mathematical model representing the binding of VEGF to VEGFR and the subsequent ligand-induced dimerisation. A key factor in this work is the qualitative and quantitative effect of binding cooperativity, which describes the effect that the binding of a ligand to a receptor has on the binding of that ligand to a second receptor, and the dimerisation of these receptors. We analyse the ordinary differential equation system at equilibrium, giving analytical solutions for the total amount of ligand bound. For time-course dynamics, we use numerical methods to explore possible behaviours under various parameter regimes, while perturbation analysis is used to understand the intricacies of these behaviours. Our simulation results show an excellent fit to experimental data, towards validating the model.

Exact solutions and equi-dosing regimen regions for multi-dose pharmacokinetics models with transit compartments.

Compartmental models which yield linear ordinary differential equations (ODEs) provide common tools for pharmacokinetics (PK) analysis, with exact solutions for drug levels or concentrations readily obtainable for low-dimensional compartment models. Exact solutions enable valuable insights and further analysis of these systems. Transit compartment models are a popular semi-mechanistic approach for generalising simple PK models to allow for delayed kinetics, but computing exact solutions for multi-dosing inputs to transit compartment systems leading to different final compartments is nontrivial. Here, we find exact solutions for drug levels as functions of time throughout a linear transit compartment cascade followed by an absorption compartment and a central blood compartment, for the general case of n transit compartments and M equi-bolus doses to the first compartment. We further show the utility of exact solutions to PK ODE models in finding constraints on equi-dosing regimen parameters imposed by a prescribed therapeutic range. This leads to the construction of equi-dosing regimen regions (EDRRs), providing new, novel visualisations which summarise the safe and effective dosing parameter space. EDRRs are computed for classical and transit compartment models with two- and three-dimensional parameter spaces, and are proposed as useful graphical tools for informing drug dosing regimen design.

Ligand Binding Dynamics for Pre-dimerised G Protein-Coupled Receptor Homodimers: Linear Models and Analytical Solutions.

Evidence suggests that many G protein-coupled receptors (GPCRs) are bound together forming dimers. The implications of dimerisation for cellular signalling outcomes, and ultimately drug discovery and therapeutics, remain unclear. Consideration of ligand binding and signalling via receptor dimers is therefore required as an addition to classical receptor theory, which is largely built on assumptions of monomeric receptors. A key factor in developing theoretical models of dimer signalling is cooperativity across the dimer, whereby binding of a ligand to one protomer affects the binding of a ligand to the other protomer. Here, we present and analyse linear models for one-ligand and two-ligand binding dynamics at homodimerised receptors, as an essential building block in the development of dimerised receptor theory. For systems at equilibrium, we compute analytical solutions for total bound labelled ligand and derive conditions on the cooperativity factors under which multiphasic log dose-response curves are expected. This could help explain data extracted from pharmacological experiments that do not fit to the standard Hill curves that are often used in this type of analysis. For the time-dependent problems, we also obtain analytical solutions. For the single-ligand case, the construction of the analytical solution is straightforward; it is bi-exponential in time, sharing a similar structure to the well-known monomeric competition dynamics of Motulsky-Mahan. We suggest that this model is therefore practically usable by the pharmacologist towards developing insights into the potential dynamics and consequences of dimerised receptors.

Modelling and simulation of biased agonism dynamics at a G protein-coupled receptor.

Theoretical models of G protein-coupled receptor (GPCR) concentration-response relationships often assume an agonist producing a single functional response via a single active state of the receptor. These models have largely been analysed assuming steady-state conditions. There is now much experimental evidence to suggest that many GPCRs can exist in multiple receptor conformations and elicit numerous functional responses, with ligands having the potential to activate different signalling pathways to varying extents-a concept referred to as biased agonism, functional selectivity or pluri-dimensional efficacy. Moreover, recent experimental results indicate a clear possibility for time-dependent bias, whereby an agonist's bias with respect to different pathways may vary dynamically. Efforts towards understanding the implications of temporal bias by characterising and quantifying ligand effects on multiple pathways will clearly be aided by extending current equilibrium binding and biased activation models to include G protein activation dynamics. Here, we present a new model of time-dependent biased agonism, based on ordinary differential equations for multiple cubic ternary complex activation models with G protein cycle dynamics. This model allows simulation and analysis of multi-pathway activation bias dynamics at a single receptor for the first time, at the level of active G protein (αGTP), towards the analysis of dynamic functional responses. The model is generally applicable to systems with NG G proteins and N* active receptor states. Numerical simulations for NG=N*=2 reveal new insights into the effects of system parameters (including cooperativities, and ligand and receptor concentrations) on bias dynamics, highlighting new phenomena including the dynamic inter-conversion of bias direction. Further, we fit this model to 'wet' experimental data for two competing G proteins (Gi and Gs) that become activated upon stimulation of the adenosine A1 receptor with adenosine derivative compounds. Finally, we show that our model can qualitatively describe the temporal dynamics of this competing G protein activation.

Model reduction in mathematical pharmacology : Integration, reduction and linking of PBPK and systems biology models.

In this paper we present a framework for the reduction and linking of physiologically based pharmacokinetic (PBPK) models with models of systems biology to describe the effects of drug administration across multiple scales. To address the issue of model complexity, we propose the reduction of each type of model separately prior to being linked. We highlight the use of balanced truncation in reducing the linear components of PBPK models, whilst proper lumping is shown to be efficient in reducing typically nonlinear systems biology type models. The overall methodology is demonstrated via two example systems; a model of bacterial chemotactic signalling in Escherichia coli and a model of extracellular regulatory kinase activation mediated via the extracellular growth factor and nerve growth factor receptor pathways. Each system is tested under the simulated administration of three hypothetical compounds; a strong base, a weak base, and an acid, mirroring the parameterisation of pindolol, midazolam, and thiopental, respectively. Our method can produce up to an 80% decrease in simulation time, allowing substantial speed-up for computationally intensive applications including parameter fitting or agent based modelling. The approach provides a straightforward means to construct simplified Quantitative Systems Pharmacology models that still provide significant insight into the mechanisms of drug action. Such a framework can potentially bridge pre-clinical and clinical modelling - providing an intermediate level of model granularity between classical, empirical approaches and mechanistic systems describing the molecular scale.

Structural identifiability for mathematical pharmacology: models of myelosuppression.

Structural identifiability is an often overlooked, but essential, prerequisite to the experiment design stage. The application of structural identifiability analysis to models of myelosuppression is used to demonstrate the importance of its considerations. It is shown that, under certain assumptions, these models are structurally identifiable and so drug and system specific parameters can truly be separated. Further it is shown via a meta-analysis of the literature that because of this the reported system parameter estimates for the "Friberg" or "Uppsala" model are consistent in the literature.

Transit and lifespan in neutrophil production: implications for drug intervention.

A comparison of the transit compartment ordinary differential equation modelling approach to distributed and discrete delay differential equation models is studied by focusing on Quartino's extension to the Friberg transit compartment model of myelosuppression, widely relied upon in the pharmaceutical sciences to predict the neutrophil response after chemotherapy, and on a QSP delay differential equation model of granulopoiesis. An extension to the Quartino model is provided by considering a general number of transit compartments and introducing an extra parameter that allows for the decoupling of the maturation time from the production rate of cells. An overview of the well established linear chain technique, used to reformulate transit compartment models with constant transit rates as distributed delay differential equations (DDEs), is then given. A state-dependent time rescaling of the Quartino model is performed to apply the linear chain technique and rewrite the Quartino model as a distributed DDE, yielding a discrete DDE model in a certain parameter limit. Next, stability and bifurcation analyses are undertaken in an effort to situate such studies in a mathematical pharmacology context. We show that both the original Friberg and the Quartino extension models incorrectly define the mean maturation time, essentially treating the proliferative pool as an additional maturation compartment. This misspecification can have far reaching consequences on the development of future models of myelosuppression in PK/PD.

Mathematical description of drug-target interactions: application to biologics that bind to targets with two binding sites.

The emerging discipline of mathematical pharmacology occupies the space between advanced pharmacometrics and systems biology. A characteristic feature of the approach is application of advance mathematical methods to study the behavior of biological systems as described by mathematical (most often differential) equations. One of the early application of mathematical pharmacology (that was not called this name at the time) was formulation and investigation of the target-mediated drug disposition (TMDD) model and its approximations. The model was shown to be remarkably successful, not only in describing the observed data for drug-target interactions, but also in advancing the qualitative and quantitative understanding of those interactions and their role in pharmacokinetic and pharmacodynamic properties of biologics. The TMDD model in its original formulation describes the interaction of the drug that has one binding site with the target that also has only one binding site. Following the framework developed earlier for drugs with one-to-one binding, this work aims to describe a rigorous approach for working with similar systems and to apply it to drugs that bind to targets with two binding sites. The quasi-steady-state, quasi-equilibrium, irreversible binding, and Michaelis-Menten approximations of the model are also derived. These equations can be used, in particular, to predict concentrations of the partially bound target (RC). This could be clinically important if RC remains active and has slow internalization rate. In this case, introduction of the drug aimed to suppress target activity may lead to the opposite effect due to RC accumulation.

Computational modelling of dynamic cAMP responses to GPCR agonists for exploration of GLP-1R ligand effects in pancreatic ß-cells and neurons.

The glucagon-like peptide-1 receptor (GLP-1R) is a class B G protein-coupled receptor (GPCR) which plays important physiological roles in insulin release and promoting fullness. GLP-1R agonists initiate cellular responses by cyclic AMP (cAMP) pathway signal transduction. Understanding of the potential of GLP-1R agonists in the treatment of type 2 diabetes may be advanced by considering the cAMP dynamics for agonists at GLP-1R in both pancreatic ß-cells (important in insulin release) and neurons (important in appetite regulation). Receptor desensitisation in the cAMP pathway is known to be an important regulatory mechanism, with different ligands differentially promoting G protein activation and desensitisation. Here, we use mathematical modelling to quantify and understand experimentally obtained cAMP timecourses for two GLP-1R agonists, exendin-F1 (ExF1) and exendin-D3 (ExD3), which give markedly different signals in ß-cells and neurons. We formulate an ordinary differential equation (ODE) model for the dynamics of cAMP signalling in response to G protein-coupled receptor (GPCR) ligands, encompassing ligand binding, receptor activation, G protein activation, desensitisation and second messenger generation. We validate our model initially by fitting to timecourse data for HEK293 cells, then proceed to parameterise the model for ß-cells and neurons. Through numerical simulation and sensitivity studies, our analysis adds support to the hypothesis that ExF1 offers more potential glucose regulation benefit than ExD3 over long timescales via signalling in pancreatic ß-cells, but that there is little difference between the two ligands in the potential appetite suppression effects offered via long-time signalling in neurons on the same timescales.

In vivo potency revisited - Keep the target in sight.

Potency is a central parameter in pharmacological and biochemical sciences, as well as in drug discovery and development endeavors. It is however typically defined in terms only of ligand to target binding affinity also in in vivo experimentation, thus in a manner analogous to in in vitro studies. As in vivo potency is in fact a conglomerate of events involving ligand, target, and target-ligand complex processes, overlooking some of the fundamental differences between in vivo and in vitro may result in serious mispredictions of in vivo efficacious dose and exposure. The analysis presented in this paper compares potency measures derived from three model situations. Model A represents the closed in vitro system, defining target binding of a ligand when total target and ligand concentrations remain static and constant. Model B describes an open in vivo system with ligand input and clearance (Cl(L)), adding in parallel to the turnover (ksyn, kdeg) of the target. Model C further adds to the open in vivo system in Model B also the elimination of the target-ligand complex (ke(RL)) via a first-order process. We formulate corresponding equations of the equilibrium (steady-state) relationships between target and ligand, and complex and ligand for each of the three model systems and graphically illustrate the resulting simulations. These equilibrium relationships demonstrate the relative impact of target and target-ligand complex turnover, and are easier to interpret than the more commonly used ligand-, target- and complex concentration-time courses. A new potency expression, labeled L50, is then derived. L50 is the ligand concentration at half-maximal target and complex concentrations and is an amalgamation of target turnover, target-ligand binding and complex elimination parameters estimated from concentration-time data. L50 is then compared to the dissociation constant Kd (target-ligand binding affinity), the conventional Black & Leff potency estimate EC50, and the derived Michaelis-Menten parameter Km (target-ligand binding and complex removal) across a set of literature data. It is evident from a comparison between parameters derived from in vitro vs. in vivo experiments that L50 can be either numerically greater or smaller than the Kd (or Km) parameter, primarily depending on the ratio of kdeg-to-ke(RL). Contrasting the limit values of target R and target-ligand complex RL for ligand concentrations approaching infinity demonstrates that the outcome of the three models differs to a great extent. Based on the analysis we propose that a better understanding of in vivo pharmacological potency requires simultaneous assessment of the impact of its underlying determinants in the open system setting. We propose that L50 will be a useful parameter guiding predictions of the effective concentration range, for translational purposes, and assessment of in vivo target occupancy/suppression by ligand, since it also encompasses target turnover - in turn also subject to influence by pathophysiology and drug treatment. Different compounds may have similar binding affinity for a target in vitro (same Kd), but vastly different potencies in vivo. L50 points to what parameters need to be taken into account, and particularly that closed-system (in vitro) parameters should not be first choice when ranking compounds in vivo (open system).