New kids on the block: The Popeye domain containing (POPDC) protein family acting as a novel class of cAMP effector proteins in striated muscle.

The cyclic 3',5'-adenosine monophosphate (cAMP) signalling pathway constitutes an ancient signal transduction pathway present in prokaryotes and eukaryotes. Previously, it was thought that in eukaryotes three effector proteins mediate cAMP signalling, namely protein kinase A (PKA), exchange factor directly activated by cAMP (EPAC) and the cyclic-nucleotide gated channels. However, recently a novel family of cAMP effector proteins emerged and was termed the Popeye domain containing (POPDC) family, which consists of three members POPDC1, POPDC2 and POPDC3. POPDC proteins are transmembrane proteins, which are abundantly present in striated and smooth muscle cells. POPDC proteins bind cAMP with high affinity comparable to PKA. Presently, their biochemical activity is poorly understood. However, mutational analysis in animal models as well as the disease phenotype observed in patients carrying missense mutations suggests that POPDC proteins are acting by modulating membrane trafficking of interacting proteins. In this review, we will describe the current knowledge about this gene family and also outline the apparent gaps in our understanding of their role in cAMP signalling and beyond.


Exchange protein directly activated by cAMP
It was long thought that PKA and cyclic nucleotide-gated ion channels such as the hyperpolarization-activated cyclic nucleotide-gated (HCN) channels would be the only effector proteins involved in cAMP signalling.

Discovery of the Popeye domain containing (POPDC) genes
With the goal to clone novel genes involved in heart development, two groups independently identified the first member of the Popeye domain containing In the adult heart, Popdc1 is expressed at higher levels in the atria than the ventricles and expression is also elevated in the cardiac conduction system (CCS) which includes the SAN, the atrioventricular node (AVN), the His bundle, the bundle branches and the Purkinje fibres. Popdc2 on the other hand is expressed at equal levels in atria and ventricles but also displays a higher expression level in the CCS { ADDIN EN.CITE { ADDIN EN.CITE.DATA }}. A detailed expression pattern of Popdc3 is currently not available.

Drosophila
Loss-of-function experiments have been performed in a number of organisms. In Drosophila, a single POPDC gene (bves) is present in the genome. Due to the complexity of the genomic region, P-element mobilization did not result in the identification of a null mutant for bves { ADDIN EN.CITE { ADDIN EN.CITE.DATA }}. As an alternative approach, heat shock-induced expression of an antisense transcript caused early lethality with embryos displaying abnormal gastrulation movements and aberrant gut closure. No further analysis of organ formation and in particular no analysis of heart and skeletal muscle structure and function has been performed in the transgenic flies.

Zebrafish
Morpholino-mediated knockdown of popdc2 in zebrafish caused a severe muscular dystrophy phenotype, which is characterized by the disruption of the myotendinous junction POPDC proteins have an important function in organizing tight junction formation and the lack of popdc1 in zebrafish by morpholino-mediated knockdown causes a defective barrier formation of the skin leading to a fragility of the embryo when challenged by hyperosmotic conditions. The inability to maintain osmotic homeostasis is probably the reason for the pericardial effusion phenotype, as it was rescued by raising the osmolality of the culture medium { ADDIN EN.CITE <EndNote><Cite><Author>Wu</Author><Year>2012</Year><RecNum>9392< /RecNum><DisplayText>[57]</DisplayText><record><rec-number>9392</rec-number><foreign-keys><key app="EN" db-id="rt59fwpfsdprwves59gpr5fwafzf2rxaa0ft" timestamp="0">9392</key></foreign-keys><ref-type name="Journal

{PAGE }
resource-num></record></Cite></EndNote> } . Loss of Popdc1 causes an increase in the expression of the proapoptotic BCL2 interacting protein 3 (Bnip3), which may explain the increased vulnerability when null mutant hearts are subjected to ischemia/reperfusion injury. As mentioned before, Popdc1 and Popdc2 display high-level expression in the CCS. Implantation of telemetric ECG devices revealed normal heart rate and a normal electrocardiogram at baseline. However, when mouse mutants were subjected to physical exercise, mental stress or isoproterenol injection, the heart rate became highly variable and the SAN pacemaker was pausing for different lengths of time { ADDIN EN.CITE { ADDIN EN.CITE.DATA }}. Interestingly, the pathological phenotype, which is present in both Popdc1 and Popdc2 null mutants was not present in young mice, however, at 5-8 months of age, mutants displayed a severe stress-induced bradycardia with episodes of sinus node dysfunction. The age-dependency of phenotype development in the mutants is reminiscent of the sick sinus syndrome (SSS), which is the most frequent reason for pacemaker implantation and most prevalent in the elderly in the absence of any

Association of POPDC1 mutations with heart disease
While the association of heart and muscle phenotypes in animal models carrying loss-of-function mutations was demonstrated in several species, the question, whether POPDC mutations are also disease-causing in patients remained unanswered until recently.

Popdc1 interacts with the two-pore potassium channel TREK-1
The plasma membrane localization of POPDC proteins, the pacemaker phenotype in Popdc1 and Popdc2 null mutants, the AV-block in patients and in zebrafish mutants and the high expression levels of POPDC proteins in the cardiac conduction system all point to a role of POPDC proteins in cardiac action potential generation and/or cardiac conduction. The interaction of POPDC proteins with ion channels or other electrogenic proteins was recently tested, which resulted in the identification of the potassium channel TWIK-related K + channel 1 (TREK-1)  This effect was modulated by cAMP and was lost in the presence of the general PDE inhibitor theophylline. The interaction of POPDC proteins with TREK-1 is mediated by the cytoplasmic part of the protein. Based on the interaction of POPDC and TREK-1, a bi-molecular FRET sensor was constructed. The FRET ratio obtained at baseline decreased after the addition of isoproterenol or forskolin, suggesting that cyclic nucleotide-binding affects the interaction of POPDC1 with TREK-1 { ADDIN EN.CITE { ADDIN EN.CITE.DATA }}. A cardiac specific knockout of Kcnk2, which encodes TREK-1, displays a stress-induced sinus bradycardia similar to the one observed in Popdc1 and Popdc2 null mutants { ADDIN EN.CITE { ADDIN EN.CITE.DATA }}, suggesting that the sinus bradycardia in POPDC mutants may in part be due to an impaired TREK-1 current.

{PAGE }
num>10.1126/scisignal.2003506</electronic-resource-num></record></Cite></EndNote>}. It can be envisioned that cAMP binding and phosphorylation of POPDC proteins may lead to conformational changes that affect access of PKA or other kinases to their substrates. Support for such a hypothetical mechanism stems from the electrophysiological measurement of TREK1 current in Xenopus oocytes, which is inhibited by cAMP and therefore blunting the effect of co-expression of POPDC proteins { ADDIN EN.CITE { ADDIN EN.CITE.DATA }}. However, TREK-1 current maintained an increased level when the same experiment was performed in the presence of a cAMP binding-deficient POPDC1 mutant. These results may indicate that POPDC proteins form a complex with TREK-1, which may prevent PKA from gaining access and being able to phosphorylate TREK-1

Outlook
As in any other novel scientific field, many open questions remain to be addressed and it is the firm belief of the authors of this review that we currently are just at the beginning of our understanding of the role of POPDC proteins in cAMP signalling. An important question that needs to be answered is, whether POPDC proteins are directly interacting with other proteins of the cAMP signalling pathway and whether POPDC proteins modulate the enzymatic activity of these proteins. Thus, we need to address the question, whether POPDC proteins do physically interact with adenylate cyclases or phosphodiesterases. Equally important will be to answer the question, whether there are AKAP proteins, which bind to POPDC proteins. A number of AKAP proteins are localized to the same the ability of muscle regeneration { ADDIN EN.CITE <EndNote><Cite><Author>Andrée</Author><Year>2002</Year><RecNum>26 86</RecNum><DisplayText>[60]</DisplayText><record><rec-number>2686</rec-number><foreign-keys><key app="EN" db-id="rt59fwpfsdprwves59gpr5fwafzf2rxaa0ft" timestamp="0">2686</key></foreign-keys><ref-type name="Journal Article">17</ref-type><contributors><authors><author>Andrée, B.</author><author>Fleige, A.</author><author>Arnold, H. H.</author><author>Brand, T.</author></authors></contributors><titles><title>Mouse Pop1 is required for muscle regeneration in adult skeletal muscle</title><secondary-title>Mol Cell Biol</secondary-title></titles><periodical><full-title>Mol Cell Biol</full-title><abbr-1>Molecular and cellular biology</abbr-1></periodical><pages>1504-12.</pages><volume>22</volume><number>5</number><dates><year>2002 </year></dates><label>21829279</label><urls></urls></record></Cite></En dNote>}. However, currently nothing is known about the role of POPDC genes in cardiac regeneration. Likewise, we do not understand how POPDC proteins mechanistically affect membrane trafficking of interacting proteins. Thus, the interaction of POPDC proteins with proteins involved in vesicle transport and membrane protein recycling will be an important avenue to be investigated. Electrophysiological defects have been observed in the heart of patients carrying POPDC1 mutations, in contrast structural defects were observed in case of skeletal muscle. It will be interesting to find out, whether mutations in POPDC genes are also able to cause structural defects in the heart such as dilated cardiomyopathy and likewise, whether skeletal muscle could also be functionally affected. We will also need to screen patient populations for the presence of mutations in any of the other POPDC isoforms, given their overlapping expression and function. It is likely that we might find mutations also in these family members. Currently, no specific drugs are available to activate or inhibit POPDC protein function. Thus, functional analysis is solely dependent on the genetic approach of generating gain-or loss-of-function mutations. Over the years several specific agonists and antagonists have been characterized for PKA, EPAC and HCN channels { ADDIN EN.CITE { ADDIN EN.CITE.DATA }}. None of these compounds have made it into the clinics, however, they are very useful as scientific tools allowing isoform-specific modulation of different effector proteins. At present, similar tools are not available in case of the POPDC protein family. Given the unique sequence of the Popeye domain and in particular of the cyclic nucleotide binding cassette { ADDIN EN.CITE { ADDIN EN.CITE.DATA }}, which does not resemble any other cAMP binding domains, it is very likely that specific small molecules could be defined, which will help to analyse the functional involvement of POPDC proteins in different aspects of cAMP signalling. Many more problems will need to be addressed to find out what the new kids on the block are up to. Figure 1. The cAMP signalling pathway in cardiac myocytes. In response to adrenergic stimulation via b1or b2adrenergic receptors (b1AR, b2AR), cAMP is synthesized by adenylyl cyclases (AC) and degraded by phosphodiesterases (PDEs). The effector proteins protein kinase A (PKA) and exchange protein activated by cAMP (EPAC) are part of signalosomes, which are organized by A protein kinase associated proteins (AKAPs). Target proteins, which are getting phosphorylated by PKA are often part of AKAP-dependent signalosomes. Popeye domain containing (POPDC) proteins are localized to many different subcellular compartments, some of which are also shared by other cAMP binding proteins. It is unknown whether POPDC proteins are part of the same signalosome complex as other cAMP effector proteins. Like PKA and EPAC proteins, POPDC proteins are also found in the nucleoplasm probably lacking the transmembrane domains (DTM-POPDC). Secondary structure prediction of POPDC proteins using Psipred [108]. In each isoform, a short (26-47 residues) extracellular domain (ECD) is present, which harbors one or two N-glycosylation motifs (asterisks) followed by three transmembrane domains (TM1-3). A large part of the cytoplasmic domain consists of the Popeye domain, which is close to 50% similar between family members. The carboxy-terminal domain (CTD) is most variable part of the POPDC proteins and distinct between isoforms. (B) 3-D models of POPDC1, POPDC2, and POPDC3. The models were produced with the help of the Phyre 2 web portal, which predicts 3-D protein structures on the basis of homology modelling { ADDIN EN.CITE <EndNote><Cite><Author>Kelley</Author><Year>2015</Year><RecNum>192 58</RecNum><DisplayText>[109]</DisplayText><record><rec-number>19258</rec-number><foreign-keys><key app="EN" db-id="rt59fwpfsdprwves59gpr5fwafzf2rxaa0ft" timestamp="1499941431">19258</key></foreign-keys><ref-type name="Journal Article">17</ref