Compartmentalized cAMP signaling via the PDE4-Popeye protein complex

The Popeye domain containing (POPDC) protein family are a unique family of transmembrane proteins with several proposed functions that are not fully understood. POPDC proteins are abundantly expressed in cardiac and skeletal muscle. Within the heart, POPDC1 has been shown to be highly expressed in the pace making centres and at moderate to low levels in the atria and ventricles. Given this localisation, a role for POPDC1 was hypothesised to be in the maintenance of normal heartbeat rhythm. Studies involving POPDC1 mutant mice and zebrafish provided evidence for this proposed function as the genetically modified model animals displayed cardiac arrhythmias as the predominant phenotype. Since POPDC proteins have been shown to be cAMP effectors, I set out to characterise their functions that that were regulated by cyclic nucleotide levels. As all the other known cAMP effector proteins such as PKA and EPAC form signalling complexes with phosphodiesterase (PDE) enzymes to limit cAMP concentrations in the vicinity of the cAMP effector, thus hindering their activation under basal conditions, the work presented in this thesis aimed to discover whether such a complex existed that contained POPDC and PDE4. This thesis begins with characterisation of the molecular interaction between POPDC1 and PDE4A. It was hypothesised that a signalling complex containing POPDC1 and PDE4A would be formed as a regulatory mechanism to control cAMP concentrations in microdomains close to POPDC1 so modulating the activity of the protein. Co-immunoprecipitation studies along with proximity ligation assays confirmed the presence of this interaction in transiently transfected HEK293 cells, endogenously expressing neonatal rat ventricular myocytes (NRVM) and adult rabbit septal myocytes (ARSM). Fine mapping of the binding sites on each respective protein was carried out using peptide array technology. This allowed identification of key docking sites that mediate the interaction between POPDC1 and PDE4. Using the binding sequence for PDE4A on POPDC1, a cell penetrating disruptor peptide was created. It was proposed that the disruptor peptide would ‘unhook’ the POPDC1-PDE4A interaction, allowing for enhanced cAMP dynamics around POPDC1 therefore, modulating its interactions with other proteins such as the potassium channel TREK1. Using fluorescence resonance energy transfer (FRET) it was shown that the disruptor peptide led to a reduction in the interaction between POPDC1 and TREK1, mimicking the effect of PDE4 inhibition with rolipram (4-[3-(cytopentyloxy)-4 methophenyl]-2-pyrrolidinone). Furthermore, treatment with the disruptor peptide in adult rabbit ventricular myocytes created a marked elongation of the cardiac repolarisation phase. It could be suggested that by blocking PDE4A binding to POPDC1, the subsequent increase in cAMP in its vicinity may lead to a reduction in the interaction between POPDC1 and TREK1. Given that the interaction with POPDC1 increases the current through the channel two-fold, the elongated repolarisation phase may be due to a decrease in K+ efflux caused by the lack of POPDC1-TREK1 complex formation. This conclusion represents the first instance where PDE activity has been shown to influence POPDC function. POPDC1 has previously been shown to be downregulated in human heart failure. Using a porcine model of myocardial infarction, I have demonstrated that there is an initial disease-induced reduction in POPDC1 expression levels that is lost after 3 months. In human patients suffering from heart failure, this loss in expression was not replicated. This data suggests that the initial reduction in POPDC1 creates a protective effect that helps the injury affected area of the heart in animal models. A secondary aspect of this work was to investigate whether POPDC1 was subjected to any post translational modification (PTMs). PTMs are known to be able to modulate interactions between proteins through several mechanisms. In silico analysis of POPDC1 sequence revealed a high probability SUMOylation site (K119) and a PKA-dependent phosphorylation (T236) site which were confirmed using peptide array. Forced SUMOylation in NRVM further confirmed that Popdc1 is a SUMO substrate. It remains unclear at this time what the functional relevance of these PTMs are, but I hypothesize that either or both contribute towards the ability of POPDC1 to bind to different interaction partners by inducing a conformational change. This is the first study to show that POPDC1 is subjected to SUMOylation and phosphorylation. The final part of this thesis reports work undertaken to determine the threedimensional structure model of POPDC1. Currently, only indirect structural information about the protein is available, generated by homology modelling. Protein structure can often provide clues about putative protein function as well as confirming that binding sites such as those identified in this work are surface exposed. Four expression constructs containing different fragments of POPDC1 were produced and the solubility of the proteins produced analysed. A construct containing only the Popeye domain was taken forward and highly purified samples were produced. Despite the quantity of recombinant protein, no structural analysis could take place as the protein aggregated in solution. Further testing of other constructs will be required to develop a construct that can be used to determine the structure of POPDC1. In combination, the work described here provides a major contribution to the field showing, for the first time, that POPDC1 forms an interaction with PDE4A in the heart. I have also described novel methods and tools to investigate the functional correlates of the complex. In addition, this interaction underpins the first regulatory mechanism of POPDC1 to have been identified. Understanding the manner in which the POPDC1-PDE4A interaction affects cardiac pace making is necessary to pinpoint how POPDC1 contributes towards disease. This may lead to new therapeutic avenues that are able to target the POPDC1-PDE4A interaction.