The Popeye domain-containing (POPDC) protein family is strongly expressed in the heart and important for cardiac pacemaking and conduction. Knockout mutations of Popdc1 and Popdc2 were generated in mice by replacing the first coding exon with a LacZ reporter gene [11,12,14] [1][2][3]. No embryonic lethality was found in either of the homozygous mutants. LacZ staining for Popdc1 and Popdc2 revealed an exclusive expression in cardiac myocytes with an overlapping expression pattern of both genes [11,12,14] [1][2][3]. Popdc1 is expressed at higher levels in atrial than ventricular myocytes in both chicken and mouse hearts [4,11,14,98][4][1][3][5] while Popdc2 is expressed at equal levels in all heart chambers. There is significantly higher expression of both genes in the cardiac conduction system (CCS) of the mouse heart, which prompted an investigation of cardiac pacemaking and conduction by electrocardiography radiotelemetry in conscious animals [14][3]. Subjecting Popdc1 and Popdc2 null mutants to physical or emotional stress or isoproterenol injection induced a sinus node bradycardia that was absent at baseline. Both mutants developed extensive sinus pauses, and showed episodes of tachybradycardia, an increased heart rate variability, and an overall reduced mean heart rate [14] [3]. The phenotype developed in an age-dependent manner in homozygotes, while heterozygous mutants were indistinguishable from wild type animals.
Apart from sinus bradycardia, analysis of the
sinoatrial node (SAN
) morphology and histology revealed significant structural alterations in both mutants
[13]. SAN pacemaker cells have long, thin cell protrusions which warranted them the name spider or spindle and elongated spindle cells
[26]. These cells are typically electrically poorly coupled and represent a primitive embryonic-like form of cardiac myocytes that are embedded in a thick mesh of extracellular matrix and thereby are electrically isolated, preventing them from getting hyperpolarized by the much larger mass of atrial chamber myocytes
[37]. However, whole-mount immunohistochemical preparations, using HCN4 as a
n SAN marker, revealed that in
Popdc1 and
Popdc2 null mutants there were fewer spindle cells, a reduced number of cellular extensions in the remaining pacemaker myocytes, and an SAN structure that was more compact than in its wild type counterpart
[13]. These findings are significant in that a lack of cellular extensions on spindle cells may lead to an impaired electrical conduction from the SAN to the surrounding atrial cardiomyocytes. Moreover, the loss of nodal pacemaker myocytes could become a limiting factor for cardiac pacemaking under stress. The pathological phenotype and structural changes of the SAN are not present in young mice but were found in
Popdc1 and
Popdc2 null mutants that were five months and older
[13].
Similar to the cardiac arrhythmia phenotype observed in mice,
popdc1 and
popdc2 zebrafish morphants and the
popdc1S191F KI mutant also presented a cardiac arrhythmia phenotype, albeit in form of an
, atrioventricular (AV)AV-block as opposed to the stress-induced sinus bradycardia that was recorded in mice
[48][59].
In tThe
zebrafish, cardiac arrhythmia was already present at the embryonic stage and the severity of the phenotype increased in an age-dependent manner. Loss-of-function experiments showed that zebrafish larvae of three-to-four days post fertilization experienced AV-block type I, whilst larvae of older age developed complete heart block or occasionally a non-contracting heart
[48][59].
Morpholino-mediated knockdown of
popdc1 and
popdc2 also induced a severe muscular dystrophy phenotype, which was characterized by an impaired formation of the myotendinous junction (MTJ), which probably was the reason for the myofiber ruptures which were frequently present in the morphants
[48][59]. The MTJ is a specialized basement membrane present in skeletal muscle and is required for proper force transmission between the tendon and the muscle. It is a complex structure formed through the interaction of various extracellular matrix proteins and several membrane proteins
[610][711]. Electron microscopy of the
popdc1S191F KI mutant revealed a lack of the electron-dense matrix proteins within the MTJ
[48][59]. These findings collectively suggest that POPDC proteins play an important role in MTJ formation. A possible hint towards the discovery of the molecular pathway affected by loss of
popdc1 and
popdc2 is drawn from the recent discovery that POPDC1 interacts with dystrophin, which plays an important role in MTJ formation in zebrafish
[48][59].
SAN dysfunction in
Popdc1 and
Popdc2 null mutants i
n mice is reminiscent of sick sinus syndrome (SSS) in patients. SSS is the most frequent reason for pacemaker implantation in the elderly in the absence of any other heart disease
[812]. Therefore, it was initially hypothesized that POPDC proteins might have a modulatory function in the pacemaker current I
f [913]. To test this hypothesis, the I
f current density and activation time were measured in SAN myocytes that were isolated from
Popdc2 null mutant and wild-type mice
[13]. Recordings that were taken at basal conditions and after stimulation with the cAMP analogue 8-Br-cAMP showed no difference between genotypes. Therefore, it is likely that other proteins involved in cardiac pacemaking are regulated in their activity by POPDC proteins.
As outlined above, in SAN myocytes, I
f is an important current for pacemaking that is uniquely generated by HCN4. However, the oscillating pacemaker potential is produced through a collaboration of several sarcolemmal ion channels and pumps collectively called the membrane clock, of which one or more could possibly be modulated through interaction with POPDC proteins. For example, the sodium calcium exchanger NCX1 was recently identified as a POPDC2 interacting partner and the loss of NCX1 was shown to cause SAN dysfunction with a phenotype similar to the one observed in POPDC mutants
[1014][1115]. NCX1 is part of the membrane clock and functionally couples the membrane clock to the Ca
2+ clock
[1216].
TWIK-related K+ channel 1 (TREK-1) or KCNK is a background potassium channel, the main role of which, in the heart, is to control cell excitability and maintain the membrane potential below the threshold of depolarization
[1317]. It was one of the first POPDC-interacting proteins to be identified in the heart. The cardiac-specific knockout of
Kcnk2, which encodes TREK-1, produces a stress-induced sinus bradycardia with a phenotypical manifestation that largely resembles the one described for
Popdc1 and
Popdc2 null mutants
[1418]. Therefore, a hypothesis was put forward suggesting that an aberrant TREK-1 current causes the sinus bradycardia in POPDC null mutants. Co-expression of TREK-1 and the three POPDC isoforms in
Xenopus oocytes revealed a two-fold higher TREK-1 current and is probably the result of an increased membrane trafficking of TREK-1 in the presence of either POPDC isoform
[13]. When cAMP levels were raised in frog oocytes, the stimulatory effect of POPDC proteins on TREK-1 current was lost
[13]. It is therefore plausible that POPDC proteins modulate membrane trafficking of TREK-1, whilst cAMP, by binding to POPDC, regulates its interaction with TREK-1. The interaction of POPDC proteins with TREK-1 has been mapped to the Popeye domain by deletion analysis and based on that knowledge, a bi-molecular Förster-resonance energy transfer (FRET) sensor was constructed
[13][48]. The FRET ratio obtained at baseline decreased after the addition of isoproterenol or FSK, confirming that cyclic nucleotide binding affects the interaction of POPDC1 with TREK-1
[13]. These findings give support to the hypothesis that an aberrant TREK-1 current causes sinus bradycardia in
Popdc1 or
-2 null mutants. Loss of POPDC proteins at the plasma membrane should lead to a reduction in TREK-1 current, which in turn should make the cell more excitable, however, the opposite is true for the
Popdc1 and
-2 null mutants
[13]. However, one must be careful to extrapolate from
Xenopus oocytes to SAN pacemaker myocytes. There may be cell-type specific differences in the regulation of membrane trafficking or protein–protein interaction and thus a direct measurement of TREK-1 current in SAN pacemaker cells is required to rule out its functional involvement in the sinus bradycardia phenotype. It is likely that POPDC proteins are part of a complex network of proteins involved in modulating sinus node pacemaking at baseline and after ANS stimulation. Therefore, we currently favor the view that the pacemaker phenotype of POPDC null mutants is probably the result of many different feedback regulations and unlikely the result of a single aberrantly regulated ion channel.
A recent study by Tibbo et al. (2020) showed that POPDC1 proteins, similar to other cAMP effectors, also form a complex with PDE4s, particularly and preferably with the PDE4A isoform
[1519]. The study concluded that POPDC1–PDE4A complex formation serves as a protective measure, which prevents inappropriate cAMP binding to the Popeye domain under basal conditions. Experimentally orchestrated disruption of PDE4 and POPDC1 interaction resulted in a decreased interaction between POPDC1 and TREK1 as well as causing a prolongation of AP duration in isolated ventricular myocytes
[1519].
While the precise molecular interactions still need to be worked out, the protein complex in which POPDC proteins are probably working to modulate action potential duration is gradually emerging. TREK-1 binds to the proximal part of the Popeye domain and binding may be adjacent or overlapping with that of PDE4
[48][1519]. Thus, probably, these three proteins are in a common complex. TREK-1 also interacts with the AKAP protein AKAP79/150 and is subject to PKA-dependent phosphorylation, resulting in inactivation of TREK-1 current
[1620][1721][1822]. Thus, POPDC1 could be involved in balancing and fine-tuning the effect of βAR receptor stimulation on TREK-1 current. However, further work is required to identify the relevant molecular and functional interactions and address the question of whether POPDC proteins are part of the AKAP79/150-PKA complex or are located adjacent to it.
The interaction and complex formation between POPDC1 and PDE4A may also be relevant for the altered hippocampal synaptic plasticity that was recently described for
Popdc1 null mutants
[1923]. Activity-dependent modulation of synaptic plasticity is essential to learning and memory, and its many forms are associated with a complex set of molecular signaling pathways
[2024]. The cAMP–PKA-mediated pathways are of particular importance in persistent forms of synaptic plasticity characterized by long-term potentiation (LTP) or long-term depression (LTD)
[2125]. PKA activation functions like a gate between transient and persistent forms of LTP
[2125][2226]. POPDC1 is found in different subregions of the hippocampus including CA1 and seems to be particularly enriched in the synaptic membrane fraction, which supports a potential role in modulating synaptic plasticity
[1923]. Indeed, high-frequency electrical stimulation of hippocampal slices isolated from
Popdc1 null mutants showed enhanced LTP, especially in response to a single-train high-frequency electrical stimulation, which suggests that loss of
Popdc1 probably reduces the threshold for LTP induction
[1923]. Pharmacological treatment of hippocampal slices of
Popdc1 null mutants with FSK or FSK and IBMX (an unspecific PDE antagonist) differed in their effects on LTP formation. Enhancement of the LTP response was observed in
Popdc1 null mutant samples that were treated with FSK while FSK/IBMX treatment caused a decrease in LTP formation. It is currently unclear why these treatments triggered a different outcome; however, it can be hypothesized that in the absence of
Popdc1, the large surge in cAMP in response to FSK/IBMX treatment triggers a negative feedback loop, which might cause a reduction in the LTP response. Therefore, these results can be interpreted as evidence for a buffering function of POPDC proteins to prevent abnormal levels of cAMP accumulation in cells and thereby ensuring a graded response to different cAMP levels. PKA inhibition blocks enhanced LTP formation in
Popdc1 null mutants, which suggests that POPDC proteins similar to its function in cardiac myocytes are probably involved in fine tuning and limiting LTP formation by preventing PKA-driven phosphorylation of target proteins in response to sub
-threshold LTP-inducing stimuli
[1923].
The Popeye domain-containing genes have now been known for more than 20 years
[404][4131]. The association of POPDC gene mutations with striated muscle disease, however, has only recently (2016) been established. Model organisms, which carry loss-of-function or missense mutations found in patients, have been developed. The phenotypes presented in these animal models largely overlap with the pathologies found in patients carrying POPDC mutations. This is encouraging, as it reinforces these model organisms as being suitable to work out the underlying pathogenic mechanism of human POPDC mutations. Additionally, the overlapping nature of the aforementioned phenotypes suggests that POPDC proteins play a fundamental and essential role in cardiac electrophysiology. Recent work defining the role of POPDC1 in hippocampal long-term potentiation suggests that POPDC proteins’ role in membrane biology is not confined to the heart
[1923]. It is possible that the underlying molecular pathways in which POPDC proteins are involved in hippocampal neurons and cardiac myocytes may be similar or even identical. Therefore, it is possible, that the enhanced LTP and the stress-induced sinus bradycardia found in
Popdc1 null mutant mice may be based on the same or related molecular defects. However, these molecular defects are currently not fully understood and need to be investigated further.
The role of POPDC proteins as an important cAMP effector has been further confirmed by the findings that show complex formation between POPDC and PDE4 proteins. This interaction is essential and interference with the PDE4–POPDC complex formation has a direct, aberrant effect on cardiac action potential. All cAMP effector proteins are either directly associated with phosphodiesterases or both proteins are part of a complex. The PDE–effector protein interaction is essential for limiting cAMP effector protein activation by cAMP. This is a principle that also applies to POPDC proteins. Therefore, an important question for the near future is: how far-reaching is the analogy? That is, do other elements of the cAMP pathway also undergo complex formation with POPDC proteins? For example, do POPDC proteins interact with other effector proteins such as EPAC or PKA? Do POPDC proteins form a complex with AKAP proteins or adenylyl cyclases? Knowledge gained in this regard will lead to a better understanding of the role that POPDC proteins play in cAMP signaling.