Abstract
The cannabinoid type 1 receptor (CB1) is among the most abundant G-protein-coupled receptors (GPCRs) in the central nervous system (CNS). CB1 plays a vital role in CNS physiology and pathology, however, CB1 ligands are limited in clinical use due to on-target adverse effects caused by simultaneously elicited multiple pathways. The allosteric modulation of CB1, which elicits its effects through distinct sites to orthosteric binding, increases additional complexity of the signalling. The overall aim of the thesis is to gain a comprehensive and quantitative understanding of the mechanisms of CB1 orthosteric and allosteric modulation through pharmacometric approaches.
Potential advantages of kinetic models over steady-state models were investigated (Chapter 2). The kinetic model evaluates the full longitudinal time-course of the data, whereas the steady-state model only evaluates the data at a snapshot in time. The kinetic model provided equally or more precise parameter estimates compared to steady-state model. In addition, kinetic methods avoid misleading results when non-equilibrium (or non-steady-state) processes exist.
A unified mathematical model was constructed to describe the multiple types of interaction effects of CB1 allosteric ligand ORG27569 (ORG) and orthosteric agonist CP55940 (CP) (Chapter 3). The allosteric model incorporated kinetic properties to describe the time-course of effects of ORG and CP reported in literature. A hypothetical transitional state of CP-CB1-ORG, which can internalise but cannot inhibit cAMP, was shown to be necessary and sufficient to describe the allosteric modulation prior to the receptor adopting an inactive conformation. The model indicated that both the transitional and final inactive states of CP-CB1-ORG contribute to the enhancing CP binding. The inactive CP-CB1-ORG leads to reduced internalisation and cessation of cAMP inhibition.
The allosteric model of Chapter 3 was externally validated and used to explore mechanisms of delayed onset and probe-dependence of CB1 allosteric modulation (Chapter 4). The model identified the delayed-onset of allosteric modulation was caused by the abundance of the receptor which exceeded that required for inhibition of cAMP (i.e., cAMP inhibition was saturated). This mechanism was examined through model-designed experiments (performed by collaborators) which were found to be consistent with the model-proposed hypothesis. In addition, the model was used to design an experiment to distinguish orthosteric and orthosteric-allosteric interaction effects of probe-dependence by estimating corresponding parameters. Probe-dependence of allosteric modulation refers to that allosteric effects can differ based on the presence of different orthosteric ligands. The model identified orthosteric effect alone can generate responses of probe-dependence therefore the commonly accepted interaction effects are not necessary.
In Chapter 5 the seeming paradox that CB1 preferentially couples with Gi and inhibits cAMP under normal conditions but CB1 stimulates cAMP through Gs under other conditions was explored. Two conditions, developed as hypotheses, where stimulation of cAMP were considered: (1) concurrent stimulation with type 2 dopamine receptors (D2); and (2) high-level expression of CB1. Hypothesis 1 explains the phenomenon by depicting that Gi is consumed by coupling to activated D2 or extra CB1, so the remaining CB1 binds to Gs. Hypothesis 2 assumes that CB1 forms homo- and heterodimers, CB1-CB1 or CB1-D2, which prefer binding to Gs. In this Chapter a quantitative systems pharmacology (QSP) model was built based on Hypothesis 1. The successful description of multiple types of data (cAMP, Gi dissociation and receptor internalisation) by the model showed that Hypothesis 1 is consistent with observations but did not disprove Hypothesis 2. Simulations from the model indicated that the cAMP signalling switch is caused due to pre-coupling of G proteins with receptors. The model was then applied to design experiments to distinguish Hypothesis 1 and Hypothesis 2 by increasing Gi & Gs by the same fold, where the cAMP response is distinct according to the two hypotheses.
The QSP model developed in Chapter 5 was reduced to for a minimal model that could be used for parameter estimation (Chapter 6). The minimal model was used to analyse data arising from six different CB1 orthosteric agonists. The results indicate that although the CB1 agonists have different extents of Gi & Gs activation they have similar ratios of Gi/Gs activation. Hence suggesting that Gi/Gs signalling preference is more likely to be a system effect rather than a ligand-specific effect.
In conclusion, pharmacometric methods have been applied to enable a comprehensive and quantitative understanding of CB1 orthosteric and allosteric modulation. Kinetic analysis showed advantages over equilibrium analysis and was applied in the models of the following chapters. The models developed in this thesis provided new insights into the complicated mechanisms of CB1 allosteric modulation and Gi/Gs pathways under different scenarios and ligands. An important future work is to test model-inspired hypotheses and perform subsequent model-designed experiments.