cAMP signalling
Evaluating second messenger production following GPCR activation in live cells was once very challenging. The first RET-based biosensor to be developed was a multimolecular sensor, based on protein kinase A (PKA) subunits, sensitive to cAMP. PKA is a heterotetrameric complex, made up of two catalytic and two regulatory subunits that dissociate upon cAMP binding to the regulatory subunits (Kim et al., 2007). By monitoring FRET between the fluorescein labelled catalytic subunits and the rhodamine labelled regulatory subunits, changes in intracellular cAMP levels could be followed in real time. Pioneering work with this sensor, named FlCRhR, enabled the visualisation of cAMP compartmentalisation for the first time (Hempel et al., 1996, Bacskai et al., 1993, Adams et al., 1991).
Despite this, the use of FlCRhR was practically challenging, mainly due to the requirement to purify PKA subunits, label them ex vivo , and subsequently microinject them into live cells. Nearly ten years later, with the introduction of fluorescent proteins, a new generation of genetically encoded FRET sensors based on the FlCRhR concept were created (Zaccolo et al., 2000). This enabled direct imaging of cAMP and PKA activation in both single cells and multicellular preparations (Janetopoulos et al., 2001, Lissandron et al., 2005, Zaccolo et al., 2000).
However, multimolecular cAMP sensors have their own limitations. Firstly, to outcompete the binding of endogenous non-fluorescent PKA subunits, they must be expressed at a relatively high concentration, which may lead to cAMP buffering and distortion of cAMP dynamics (Paramonov et al., 2015). Secondly, they are catalytically active, which may affect downstream signalling (Paramonov et al., 2015). Thirdly, the fluorescently labelled regulatory and catalytic PKA subunits are usually expressed by separate vectors, which often makes achieving equal expression levels challenging (Zaccolo et al., 2000).
These potential limitations fostered the development of a new generation of unimolecular cAMP sensors. The majority of these sensors are based upon Epac, a cAMP-activated guanine-exchange factor (GEF) for the small GTPase Rap1 (Calebiro and Maiellaro, 2014). cAMP binding to Epac induces a conformational change that exposes an otherwise hidden catalytic domain to activate Rap1 (Calebiro and Maiellaro, 2014). Different cAMP FRET sensors based on Epac1 or, alternatively, a truncated form that expresses the isolated cAMP binding domain, were created in parallel by sandwiching Epac1 between two suitable fluorophores (Nikolaev et al., 2004, Ponsioen et al., 2004, DiPilato et al., 2004). Since cAMP binding results in a decrease in FRET, intracellular changes in cAMP can be detected (Nikolaev et al., 2004, Ponsioen et al., 2004, DiPilato et al., 2004). Considering these sensors are based on a single protein, they have a faster response to cAMP binding than multimolecular sensors, thus increasing temporal resolution (Paramonov et al., 2015). In addition, the use of a single encoding vector overcomes the problems associated with unequal expression levels as seen with previous multimolecular cAMP sensors (Paramonov et al., 2015). Importantly, some Epac sensors are also catalytically inactive, avoiding activation of downstream signalling pathways, and, thus, eliminating a potential source of experimental interference (Ponsioen et al., 2004, Nikolaev et al., 2004).
In addition to the previously described cAMP sensors, FRET-based sensors have also been designed to monitor endogenous PKA signalling (Calebiro and Maiellaro, 2014, Zhang et al., 2001). These sensors are called A-Kinase activity reporters (AKARs) and they contain a PKA substrate and a phosphoamino acid binding domain inserted between a FRET donor and acceptor fluorophore (e.g. CFP and YFP) (Zhang et al., 2001). Upon PKA-dependent phosphorylation of the PKA substrate, a conformational change occurs triggering the PKA substrate to interact with the phosphoamino acid binding domain, leading to an increase in FRET between donor and acceptor fluorophores (Calebiro and Maiellaro, 2014, Zhang et al., 2001).
Similar strategies have been followed to design BRET sensors for cAMP. The first BRET-based cAMP sensor was based on PKA (Prinz et al., 2006). By tagging the catalytic PKA subunit with GFP (GFP-C) and regulatory PKA subunits with RLuc (RLuc-RI and RLuc-RII), co-transfection of GFP-C with RLuc regulatory subunit induced a constitutive BRET signal that was reduced upon cAMP binding. As with their FRET counterparts, multimolecular BRET biosensors suffered from some potential limitations. This prompted the development of two unimolecular BRET sensors based on the design of Epac FRET sensors, where the donor fluorophore was replaced with RLuc (Jiang et al., 2007, Barak et al., 2008). One of these sensors, named CAMYEL (cAMP sensor using YFP-Epac-RLuc), uses circularly permuted citrine to enhance the BRET changes upon cAMP binding (Jiang et al., 2007).
More recently, a NLuc-based cAMP BRET sensor has been created. This is based upon a third generation FRET-based cAMP sensor (Klarenbeek et al., 2011) where the donor (mTurquoise) fluorophore has been replaced with NLuc (Masuho et al., 2015). The addition of NLuc, and, thus, the increased signal compared to RLuc, may enable the application of BRET-based cAMP sensors in harder to transfect cell types, and, with improvements to the solubility of bioluminescent substrates, in vivo (Su et al., 2020).