Towards a method for the simultaneous quantification of Gαq activation β-arrestin2 recruitment

In G protein-coupled receptor (GPCR) research, there is an increasing interest in the development of biased agonists as pharmacological tools and ultimately also as drugs, since adverse effects of certain pharmaceuticals are supposed to be associated with the activation of unfavourable signalling pathways. Additionally, for several receptors, bias of endogenous agonists has been discovered. Currently, the most common approach to determine biased agonism implies the application of two separate assays for detecting G protein-dependent and β-arrestin dependent signalling, respectively. Apart from the additional time needed to perform two assays instead of one, major shortcomings are associated with this approach. On one hand, the receptor sources in the employed assays can substantially differ in the extent of receptor and effector expression. On the other hand, the influence of signal amplification on the readouts of the two assays can vary tremendously. Taken together, this can lead to misinterpretations with respect to the signalling profile of the analysed agonist. Therefore, the aim of this thesis was the development of two techniques, applicable to live cells with high throughput. Firstly, for the proximal determination of Gαq protein activation and secondly, for assessing β-arrestin recruitment. The two probes were designed to be potentially compatible with a multiparametric assay format, affording information on both signalling pathways at the same time. Both probes were engineered at a similar, proximal level within each signalling cascade, to receive information unaltered by signal amplification. Methodologically, this was achieved by split luciferase complementation (SLC) using two luciferases emitting light spectra with substantially different emission maxima. The first assay was developed by applying SLC, using a red light-emitting luciferase, to probe the interaction of Gαq with phospholipase C-β3 (PLC-β3) proteins. By being independent on genetical receptor modifications and with its excellent Z’ value of 0.7, the sensor was proven to be very suitable for ligand characterization, which was shown for five different GPCRs. Furthermore, the sensor proved to be useful for imaging, as shown by live cell bioluminescence microscopy. Beyond these applications, the sensor might become a valuable tool for de-orphanisation and subsequent determination of signalling pathways of orphan GPCRs, the analysis of Gαq activation in cells endogenously expressing Gαq protein-coupled receptors and imaging in laboratory animals. Attempts to translate this assay principle to Gαs and Gαi proteins, interacting with adenylyl cyclases (AC), in a reproducible manner, failed. The most probable reason for this is poor membrane trafficking of the modified ACs, which could be improved by using artificial signalling sequences or generating a synthetic and soluble AC surrogate. For measuring β-arrestin2 recruitment, a blue light-emitting luciferase was used. During assay development it became apparent that solely focusing on signal-to-background (S/B) ratios in the initial phase, as often practiced, can be inappropriate. In fact, the development must also take the functional behaviour of the proteins into account, because otherwise, the resulting assays do not reflect the physiological behaviour of the analysed proteins anymore. For this reason, two different approaches to measure β-arrestin2 recruitment by SLC were compared: on one hand with respect to their impact on the recruitment process and, on the other hand, with respect to their influence on second messenger formation. Despite lower S/B ratios, the assay based on GPCR-NLucC/NLucN-β-arrestin2 was proven to be useful for determining ligand potencies and efficacies, and it was sensitive enough to identify iperoxo as a putative superagonist at the muscarinic acetylcholine receptors hM₁R and hM₅R with respect to β-arrestin2 recruitment. It was further demonstrated that temporal analyses of the receptor/β-arrestin2 interaction can be performed to e.g. discriminate between receptors that only transiently interact with arrestins after activation, presumably because they recycle to the plasma membrane very fast (class A), and those that show sustained interaction (class B). Furthermore, this assay principle should be broadly applicable to other GPCRs. Finally, both probes were expressed in combination in a single HEK293T cell population. This approach was applied successfully to the hM₁R, the hM₅R and to the neurotensin receptor hNTS₁R, and to a lesser extent also to the histamine receptor hH₁R, allowing the construction of concentration-response-curves of agonists and the determination of antagonistic activities for both, Gαq activation and β-arrestin2 recruitment at the same time. Results from the former three receptors suggest that both sensors do not interfere with each other when co-expressed. However, the results obtained at the hH₁R implicate that the expression levels of the sensors need to be optimised, ideally by restricting them to the endogenous expression level of their native counterparts.