Cytoskeletal Protein Variants Driving Atrial Fibrillation: Potential Mechanisms of Action

The most common clinical tachyarrhythmia, atrial fibrillation (AF), is present in 1–2% of the population. Although common risk factors, including hypertension, diabetes, and obesity, frequently underlie AF onset, it has been recognized that in 15% of the AF population, AF is familial. In these families, genome and exome sequencing techniques identified variants in the non-coding genome (i.e., variant regulatory elements), genes encoding ion channels, as well as genes encoding cytoskeletal (-associated) proteins. Cytoskeletal protein variants include variants in desmin, lamin A/C, titin, myosin heavy and light chain, junctophilin, nucleoporin, nesprin, and filamin C. These cytoskeletal protein variants have a strong association with the development of cardiomyopathy. Interestingly, AF onset is often represented as the initial manifestation of cardiac disease, sometimes even preceding cardiomyopathy by several years. Although emerging research findings reveal cytoskeletal protein variants to disrupt the cardiomyocyte structure and trigger DNA damage, exploration of the pathophysiological mechanisms of genetic AF is still in its infancy. In this review, we provide an overview of cytoskeletal (-associated) gene variants that relate to genetic AF and highlight potential pathophysiological pathways that drive this arrhythmia.


Introduction
Atrial fibrillation (AF) is the most common age-related cardiac arrhythmia in Western society [1]. AF is characterized by irregular and often very fast contractions of the atrial cardiomyocytes, resulting in an irregular heart rate, palpitations, dizziness, shortness of breath, and tiredness in the patient. AF can occur when abnormal electrical impulses suddenly start firing in the atria and override the heart's natural pacemaker, which can no longer control the rhythm of the heart [1]. Importantly, AF is associated with severe complications, such as thromboembolic events, heart failure, cognitive impairment, and increased mortality [1]. Although in most cases, AF initially presents as short, self-terminating episodes, it often progresses into long-lasting episodes that are more difficult to reverse to sinus rhythm [1]. The progressive stages of AF are associated with structural changes that promote contractile dysfunction and the impairment of electrical conduction in the atrial myocardium [2][3][4][5][6][7]. Thus, the early detection of AF and the identification of patients at risk is of utmost importance in order to treat this arrhythmia and prevent its progression. Therefore, knowledge on the root causes of AF is essential.
Recent research investigated the potential root causes for AF. These include environmentallyinduced 'wear and tear' AF, congenital AF, and genetic AF [1]. 'Wear and tear' AF is associated with the sequela of aging as well as Western diet and lifestyle-related risk factors, such as hypertension, diabetes, obesity, and coronary artery diseases, as well as various non-cardiovascular diseases, including chronic kidney disease [1]. Furthermore, an estimated prevalence of~5% Figure 1. The variety of root causes that drive AF. In most AF patients, environmentally-induced 'wear and tear' related to aging or the Western lifestyle triggers AF. In addition, flaws in genetics (variants in ion channels and cytoskeletal proteins) and embryogenesis contribute to AF susceptibility.
Emerging research findings indicate a prominent role of genetic variations mainly in ion channel and cytoskeletal (-associated) genes in driving AF [15]. Although a research paper described the role of ion channel gene variants to underlie arrhythmias, including AF [16], insights into the role of cytoskeletal (-associated) protein variants as triggers for AF are in their infancy. In this review, we provide an overview of variants in cytoskeletal proteins associated with AF promotion and highlight potential pathophysiological pathways.

Key Role of the Cytoskeletal Network in Cardiomyocyte Function
It was only recently that the crucial importance of the cytoskeleton function to maintain balanced protein (i.e., proteostasis [3,6,17,18]) and cardiomyocyte function has been recognized. In cardiomyocytes, the cytoskeleton not only provides a communication highway by transporting proteins throughout the cell, but it also capacitates its contractile function [19][20][21]. In cardiomyocytes, the cytoskeleton is highly specialized, consisting of actin filaments, desmin (intermediate) filaments, and microtubules, interacting with membrane-associated proteins, sarcomeric and nuclear proteins, and proteins of the intercalated disk. This complex network, which also interacts with various organelles, including sarcoplasmic reticulum (SR), mitochondria, and the nucleus, plays an important role in the transmission of signals and the transport of (ubiquitinated) proteins within the proteostasis network [6,19,22] (Figure 2). Within the proteostasis network, especially the The variety of root causes that drive AF. In most AF patients, environmentally-induced 'wear and tear' related to aging or the Western lifestyle triggers AF. In addition, flaws in genetics (variants in ion channels and cytoskeletal proteins) and embryogenesis contribute to AF susceptibility.
Emerging research findings indicate a prominent role of genetic variations mainly in ion channel and cytoskeletal (-associated) genes in driving AF [15]. Although a research paper described the role of ion channel gene variants to underlie arrhythmias, including AF [16], insights into the role of cytoskeletal (-associated) protein variants as triggers for AF are in their infancy. In this review, we provide an overview of variants in cytoskeletal proteins associated with AF promotion and highlight potential pathophysiological pathways.

Key Role of the Cytoskeletal Network in Cardiomyocyte Function
It was only recently that the crucial importance of the cytoskeleton function to maintain balanced protein (i.e., proteostasis [3,6,17,18]) and cardiomyocyte function has been recognized. In cardiomyocytes, the cytoskeleton not only provides a communication highway by transporting proteins throughout the cell, but it also capacitates its contractile function [19][20][21]. In cardiomyocytes, the cytoskeleton is highly specialized, consisting of actin filaments, desmin (intermediate) filaments, and microtubules, interacting with membrane-associated proteins, sarcomeric and nuclear proteins, and proteins of the intercalated disk. This complex network, which also interacts with various organelles, including sarcoplasmic reticulum (SR), mitochondria, and the nucleus, plays an important role in the transmission of signals and the transport of (ubiquitinated) proteins within the proteostasis network [6,19,22] (Figure 2). Within the proteostasis network, especially the microtubules are of vital importance. As the microtubules, SR/endoplasmic reticulum (ER), sarcomeres, and mitochondria are in contact with each other, loss in contact results in Ca 2+ overload in the organelles, unfolded protein response in the ER, and, consequently, excessive autophagic protein degradation and mitochondrial dysfunction [5,6,[23][24][25]. Contact microtubules are of vital importance. As the microtubules, SR/endoplasmic reticulum (ER), sarcomeres, and mitochondria are in contact with each other, loss in contact results in Ca 2+ overload in the organelles, unfolded protein response in the ER, and, consequently, excessive autophagic protein degradation and mitochondrial dysfunction [5,6,[23][24][25]. Contact between organelles and trafficking through the cells is mediated by acetylated microtubules [3,6,26,27]. Consequently, AF-induced histone deacetylase 6 (HDAC6) activation and the subsequent deacetylation and degradation of the microtubule network have detrimental effects on the trafficking of proteins as well as on the contractile function of the atrial cardiomyocytes [3,28,29]. Therefore, a functional cytoskeletal network underlies a balanced communication within the proteostasis network and ensures proper contractile function. The conservation of this network is of utmost importance to ensure proper cardiac function. As such, attenuation of the disruption of the cytoskeleton protects against cardiac diseases such as AF and heart failure [3,26,30].

Cytoskeletal (-Associated) Variants Associated with Clinical AF
In approximately 15% of AF patients, AF occurs in the absence of common risk factors and at a younger age [1]. In these patients, AF may be familial, suggesting a heritable genetic predisposition. To explore the role of gene variations in AF, the emergence of exome and genome sequencing data has provided extensive new data and revealed a previously unsuspected link between AF and several cytoskeletal (-associated) proteins (Table  1).

Cytoskeletal (-Associated) Variants Associated with Clinical AF
In approximately 15% of AF patients, AF occurs in the absence of common risk factors and at a younger age [1]. In these patients, AF may be familial, suggesting a heritable genetic predisposition. To explore the role of gene variations in AF, the emergence of exome and genome sequencing data has provided extensive new data and revealed a previously unsuspected link between AF and several cytoskeletal (-associated) proteins (Table 1).
Recently, several AF families have been identified to carry a mutation in genes encoding the intermediate filament proteins lamin A/C (LMNA), desmin (DES), and titin (TTN) [31][32][33][34] (Table 1). Intermediate filament proteins integrate the outer cell membrane via desmoplakin (DSP), with sarcomeric proteins such as titin, Z-disk, and the nuclear membrane, thereby regulating the sarcomere architecture and function as well as the nu-  [35][36][37]. Variants in these cytoskeletal proteins are known to be associated with the development of dilated cardiomyopathy (DCM), hypertrophic cardiomyopathy (HCM), and peripartum cardiomyopathy (PPCM) [38][39][40][41]. Of note, various studies have revealed that early-onset AF often represents the initial manifestation of the cardiac phenotype, sometimes even preceding cardiomyopathy by several years, and in the absence of common risk factors as well as gross structural changes in the heart. This observation indicates that AF is a direct consequence of the mutation rather than being caused by structural abnormalities in the ventricle [42,43]. Over 50% of LMNA [31,42,44], 60% of DES [32] and 30-60% of TTN [34] mutant carriers develop cardiac conduction disorders, arrhythmias, and atrial tachycardia, including AF. Furthermore, these patients reveal a more progressive form of cardiac disease with poor outcomes. These findings are further supported by other studies. In a Chinese family, four subjects revealed an AV block, and three of them suffered from AF, which was related to a new frameshift insertion in the LMNA gene (c.825_826insCAGG) [45]. In addition, in a study in Italian families with LMNA variants [46], a total of 30 subjects were included, of which 19 were positive and 11 were negative for LMNA variants. Of the 19 positive study subjects, 11 showed early AF versus none of the 11 negative subjects. These studies reveal a clear association between LMNA variants and the development of AF.
Interestingly, a recent study including early-onset familial AF patients uncovered the phosphodiesterase-4D-interacting-protein (PDE4DIP) p.A123T mutation as a genetic modifier of DES p.S13F [47]. PDE4DIP p.A123T increases the penetrance of cardiac arrest and early-onset AF in the DES mutation carriers. These findings suggest an epistatic interaction of DPE4DIP with the DES gene, leading to increased penetrance of both traits and AF promotion [47].
Interestingly, a recent study identified multiple index patients with DCM carried an identical mutation (c.59926 + 1G>A) in the TTN gene, encoding the giant titin protein, which is the major gene underlying inherited DCM. The identified variant likely leads to truncated titan (TTNtv), which is associated with AF onset [34,53]. In an important subset (53%) of index patients and their family members carrying this founder mutation, atrial tachyarrhythmias, in particular AF, were demonstrated. In three patients, AF preceded the development of TTNtv-associated DCM by 11-14 years [34]. In patients with TTNtvassociated DCM and AF, left atrial enlargement was not noticed in half of them. In addition, the traditional risk factors for AF were absent, suggesting a potential intrinsic effect of TTNtv [34], which is in line with the study of Ahlberg et al. [53]. Together, these data strongly suggest that (paroxysmal) AF is an important part of the clinical disease spectrum caused by TTNtv, even if major structural heart abnormalities are still absent, and the individual does not present traditional risk factors for AF. However, the underlying mechanism of TTNtv causing cardiomyocyte structural and contractile remodeling and, consequently, AF remains to be elucidated.
Thus, emerging evidence has identified novel variants in cytoskeletal proteins to underlie early AF (Table 1). Currently, no detailed mechanistic explanation, treatment modalities, or diagnostic screening tools for genetic AF are available. To overcome this, we first need to understand the mutation-specific molecular pathways that drive genetic AF. This knowledge may ultimately lead to the identification of novel drug targets and therapeutic and diagnostic strategies. Arrhythmogenic, not related to connexin redistribution. [48] Lamin A/C LMNA ↓ I Na PQC, HSP, myolysis, nuclear blebbing, nuclear protein aggregation, altered nucleocytoplasmatic transport.
Truncated titin variants associated with AF. [53] Loss of function variants in titin is associated with early-onset AF. [54] Loss of function mutation in titin is related to early-onset AF, including in ethnic minority probands. [55] Myosin heavy chain

Pathophysiological Mechanisms Driving Cytoskeletal Protein Mutation-Induced AF
As mentioned previously, all cytoskeletal proteins are linked via the microtubule network with the outer cell membrane, the sarcomeric proteins, and the nuclear membrane, and thereby regulate sarcomere architecture and function as well as nuclear morphology, DNA stability, and gene expression ( Figure 2) [19,67,68]. The malfunction of cytoskeletal proteins has been associated with cardiac manifestations such as compromised conduction disorders and arrhythmogenesis and, as such, contributed to clinical phenotypes compatible with DCM as well as AF [69]. Knowledge of the molecular mechanisms underlying cytoskeletal protein mutation-induced AF is needed to develop therapeutic and diagnostic tools for these patients. Although research findings provide some preliminary insights into underlying mechanisms, as, for example, TTNtv, as described above, most findings are still in their infancy (Figure 3).

Desmin Variant-Induced Remodeling
Desmin is one of the intermediate filaments that integrates the sarcolemma, Z disk, and nuclear membrane in sarcomeres and regulates sarcomere architecture. Desmin is encoded by the DES gene and highly expressed in the heart, the conduction system of the heart, and the pulmonary vein myocardial sleeve, which is the original focus of AF.
A previous study in desmin knockout mice showed a reduction in the atrial refractory period and increased susceptibility for AF [48], suggesting a key role of desmin dysfunction as a molecular substrate for the onset of AF. Isolated sinoatrial nodes from DES −/− mice showed a significant increase in pacemaker potentials and diastolic depolarizations. Furthermore, these mice showed increased PR intervals and P wave durations combined with supraventricular premature beats, which may trigger AF onset [70].
The role of desmin in AF promotion has been confirmed by various additional studies. In iPSC-derived cardiomyocytes expressing the homozygous mutation DES p.Y122H [71] and the myocardial tissue of patients carrying DES p.A120D [72], impaired assembly of the intermediate filament structure and the presence of cytosolic desmin aggregates were observed. Importantly, variants in DES near the c-terminus of the protein tend not to hamper filament formation; however, they interfere with the mechanical properties of the desmin network [73,74].
Moreover, in a mouse model for desmin-related cardiomyopathy, it was shown that

Desmin Variant-Induced Remodeling
Desmin is one of the intermediate filaments that integrates the sarcolemma, Z disk, and nuclear membrane in sarcomeres and regulates sarcomere architecture. Desmin is encoded by the DES gene and highly expressed in the heart, the conduction system of the heart, and the pulmonary vein myocardial sleeve, which is the original focus of AF.
A previous study in desmin knockout mice showed a reduction in the atrial refractory period and increased susceptibility for AF [48], suggesting a key role of desmin dysfunction as a molecular substrate for the onset of AF. Isolated sinoatrial nodes from DES −/− mice showed a significant increase in pacemaker potentials and diastolic depolarizations. Furthermore, these mice showed increased PR intervals and P wave durations combined with supraventricular premature beats, which may trigger AF onset [70].
The role of desmin in AF promotion has been confirmed by various additional studies. In iPSC-derived cardiomyocytes expressing the homozygous mutation DES p.Y122H [71] and the myocardial tissue of patients carrying DES p.A120D [72], impaired assembly of the intermediate filament structure and the presence of cytosolic desmin aggregates were observed. Importantly, variants in DES near the c-terminus of the protein tend not to hamper filament formation; however, they interfere with the mechanical properties of the desmin network [73,74].
Moreover, in a mouse model for desmin-related cardiomyopathy, it was shown that mitochondrial dysfunction preceded the onset of cardiac remodeling and pathological heart function [75]. Here, desmin misfolding suppresses mitochondrial respiration and membrane potential and increases the ADP/ATP ratio and mitochondrial DNA (mtDNA) release, which was accompanied by the disruption of the cardiomyocyte cytoskeletal network [76]. Interestingly, cardiac remodeling was shown to be subject to mitochondrial dysfunction in both experimental models for AF and clinical AF [5].
Furthermore, desmin dysfunction is also linked to cardiomyopathy without being directly linked to a specific desmin mutation. In a cardiomyopathy Lmna H222P/H222P mouse model, desmin organization was disrupted and accumulated in the cytoplasm. Moreover, desmoplakin, plakoglobin, and connexin-43 were mislocalized, and mitochondrial function was impaired. Interestingly, all the effects were ameliorated by the cardiac-specific overexpression of αB-crystallin [77].
The depletion of desmin in ventricular cardiomyocytes derived from a rat leads to nuclear involution, the infolding of the nuclear lamina, and DNA damage driven by the microtubule network, as well as loss in contractile function [78]. A similar phenotype is seen when depleting the desmin binding partner nesprin 3, which links desmin to the nucleus [78]. This points to desmin as an important protein in the linker of nucleoskeleton and cytoskeleton (LINC) complex and for the maintenance of a proper nuclear shape in cardiomyocytes. This notion was further supported by the rescue of nuclear shape in desmin knockout rats by disrupting the connection between nesprins and Sad1p/UNC84 (SUN) domain-containing proteins via dominant-negative KASH (Klarsicht, ANC1, syne homology) peptides (DN-KASH) [78].
Thus, the findings indicate a key role of desmin dysfunction in AF onset. Desmin dysfunction may result in desmin aggregate formation, mitochondrial dysfunction, and disruption of the cytoskeletal network.

Mutations in Lamin A/C
The lamin A/C are type V intermediate filament proteins expressed in terminally differentiated somatic cells, encoded by the LMNA gene [79,80]. As lamin A/C are expressed in the nuclear envelope, they provide structural function and transcriptional regulation in the cell nucleus. How LMNA variants cause AF onset is unknown, but experimental data from cardiomyopathy and myopathy studies may identify potential mechanisms.
A recent study utilizing induced pluripotent stem cell-derived cardiomyocytes (iPSC-CM) from DCM patients with an LMNA p.E342K variant showed a decrease in the spontaneous contraction rate and I f density, a prolonged APD and increased I Ca,L , and an increase in arrhythmias [81]. Ventricular cardiomyocytes isolated from LMNA N195K/N195K mice also showed a prolonged APD and exhibited a significant increase in both the peak and late I Na [82]. These experimental findings indicate that LMNA variants result in electrophysiological alterations, which may drive AF.
The nuclear lamin A/C regulates transcription factors either by sequestering them to the nuclear periphery or by a mechanism involved in activating specific signaling pathways [83,84]. Furthermore, lamin A/C interacts with chromatin by binding to specific genomic regions defined as lamin-associated domains (LADs) [85]. In a recent study on patient iPSC-derived LMNA p.K219T cardiomyocytes, mutant lamin A/C suppressed sodium channel SCN5A expression, resulting in diminished I Na and impaired signal propagation [86].
So far, studies in ventricular cardiomyocytes revealed LMNA variants to result in cytoskeletal and microtubule disruption [35,37,77], dysmorphology of the nuclei [37], activation of the DNA damage response [87], and PARP1 activation [39,88]. PARP1 activation results in the consumption of mitochondrial NAD + levels [88], which drive cardiomyocyte dysfunction and cardiomyopathy onset [39,89]. In DCM, all these effects were ameliorated by supplementation with a precursor of NAD + , nicotinamide [39], or conservation of the cytoskeletal network with geranylgeranylacetone (GGA), a heat shock protein (HSP)inducer [77]. Importantly, the activation of the DNA damage-PARP1-NAD + depetion axis has also been found to drive 'wear and tear' AF [90,91], indicating a potential key pathway in LMNA mutation-induced AF.
Furthermore, minor deviations on lamin A/C are linked to morphological abnormities of the nucleus that cause increased nuclear fragility and mechanosensitivity [92]. Emerging research findings uncovered that the expressions of the various lamin subtypes are dependent on the degree of mechanical stress. A-type lamins are viscoelastic and, therefore, provide structural stiffness, whereas the elastic B-type lamins allow nuclear deformability [93,94]. Mechanotransduction between the structural components of the cardiomyocyte requires the interaction of lamin A/C with the proteins involved in the LINC complex [95]. Interestingly, the disruption of the LINC complex significantly ameliorates DCM progression, prolongs longevity, and helps to retain nuclear morphology. The findings indicate that the LINC complex plays a detrimental role in nuclear morphology and cardiac disease progression in LMNA-compromised mice [96]. Whether LMNA variants drive AF due to mechanostress transduction on the nucleus via the LINC complex, to changes in transcription and chromosomal organization, or due to an interplay between both remains to be elucidated.

Variants in Myosin Heavy Chain
The two isoforms of cardiac myosin heavy chain (MYH), α and β, are coded by the MYH6 and MYH7 genes, respectively. The atrial myocardium expresses αMYH, and the ventricular myocardium expresses βMYH. MYH7 gene variants are associated with a variety of cardiomyopathies in humans, including HCM [97] and DCM [98]. Emerging evidence also indicates that MYH6 variants may result in AF. A large whole-genome sequencing study in Icelanders showed that a missense variant in MYH6 p.R721W is related to several cardiac diseases, particularly conduction disorders, including AF and sick sinus syndrome [99].
Furthermore, miR-208a is encoded by the intron of MYH6, which is a cardiac-specific highly conserved microRNA (miRNA). In mice, miR-208a regulates normal cardiac conduction and is required for the expression of GJA5 (Cx40), indicating a link between MYH6 and GJA5 expression and cardiomyocyte function via miR-208a [100]. Recent experimental findings in MYH6 knockout zebrafish show the activation of the ER stress pathway and enlargement of the cardiomyocytes [101]. In HL-1 and isolated rat atrial cardiomyocytes, the overexpression of MYH6 resulted in sarcomere impairment, electrophysiological abnormalities, and a slower conduction velocity [102]. As AF is also associated with all these endpoints, comparable pathways may play a role in mutant MYH6-induced AF.

Mutation in Myosin Light Chain
Myosin light-chain 4 (MYL4) is important for the contractile function of the cardiomyocytes. Variants in the MYL4 gene have been associated with clinical AF and have been further explored in experimental studies [58]. In zebrafish, mutant MYL4 leads to the disruption of the sarcomeric structure, atrial enlargement, and electrical abnormalities. These changes are also observed in clinical AF. In addition, electron microscopic studies revealed myofibrillar disarray and the disappearance of Z-disks, which were related to the destabilization of the F-actin-Z-disk complex, impaired Ca 2+ signaling, and the onset of atrial myopathy and atrial arrhythmia [58]. Comparable molecular and structural adaptations in the cardiomyocytes may also drive MYL4 mutation-induced AF.

Mutations in Connexin-40 and Connexin-43
Connexins are channel proteins that form hemichannels specialized in cell-to-cell communication at the intercalated disks [103]. Connexins enable the passive diffusion of compounds up to 1 kDa between two connected cells [104]. Depending on the connexin isoform, they transport water, ions, and second messengers, such as IP3, ADP, ATP, cAMP, and miRNAs [104]. Because of this role, connexins play a vital role in the heart by coordinating the depolarization of cardiac muscle and ensuring cardiac muscle function. The primary atrial connexin proteins are connexin-43 (Cx43) and connexin-40 (Cx40), encoded by GJA1 and GJA5, respectively [105]. Abnormalities in connexin levels, function, or localization impair electrical propagation and can lead to the initiation and/or stabilization of reentry circuits that support AF [105]. Moreover, Cx40 and Cx43 get mislocalized in case of LMNA variant-induced cytoskeletal changes [36,77,106,107] and DES variants [108]. This observation indicates that impairment in the cytoskeletal network at a specific location affects the network as a whole.
In 'wear and tear' AF, reduced Cx40 but not Cx43 levels were found in the atrial tissue of patients with AF compared to controls [109]. As such, several Cx40 mutants have been associated with clinical AF onset [110]. HL-1 atrial cardiomyocytes overexpressing functional Cx40 mutants G38D, V85I, or L229M, resulted in protein instability and accelerated degradation by the proteasome. Similarly, Cx40 mutants cause the gain of the function of hemichannels (half of the gap junction channels) [111,112], resulting in altered conductive and permeability properties [113], which may drive Cx40 mutant-induced AF.

Variants in Junctophilin 2
Junctophilin 2 (JPH2) is a member of the junctophilin gene family, which plays an important role in SR Ca 2+ handling and the modulation of ryanodine receptors (RyRs) [114]. Beavers et al. screened 203 unrelated HCM patients and uncovered the JPH2 p.E169K variant in two patients with juvenile-onset paroxysmal AF [63]. Follow-up studies in mouse models demonstrated that mice expressing JPH2 p.E169K exhibited a higher incidence of inducible AF compared to wild-type mice [63]. These changes were attributed to reduced binding of JPH2 p.E169K to RyR2, resulting in abnormal Ca 2+ release events, which are associated with AF [63]. In addition, atrial JPH2 protein levels in mice correlated negatively with the incidence of pacing-induced AF, indicating a crucial role of JPH2 in atrial cardiomyocyte function and AF promotion [115].

Variants in Nucleoporins
Nucleoporins (NUPs) are the critical molecular components for the assembly and function of the nuclear pore complex (NPC). The NPC is composed of 30 different nucleoporins that are each present in multiple copies [116] and contain cytoplasmic and nucleoplasmic filaments [117]. The filaments consist of microfilaments, intermediate filaments, and myosin filaments, which are all part of the cytoskeletal network. In addition, NPCs interact with the LINC-complex [118]. Due to its localization in the cardiomyocytes, NUPs are cytoskeletal-associated proteins [119].
Alterations in the NPC have been related to a variety of cardiac diseases, such as heart failure [120,121] and AF [64]. The first observation describing the relation between NUP and AF was a study including a large consanguineous AF family from Uruguay and South America, with 57 family members carrying the NUP155 p.R391H variation [122]. Follow-up studies in mouse models demonstrated that the homozygous R391H variation in NUP155 resulted in impaired nuclear localization of NUP155, reduced nuclear envelope permeability for HSP70, shortened action potential duration, and AF onset [64]. In addition, NUP155 is an essential gene, as NUP155 −/− knockout mice died during embryogenesis, whereas heterozygous NUP155 +/− mice were viable and showed reduced action potential durations, which may relate to AF onset [64]. Similarly, a study expressing truncated NUP155 in mouse embryonic stem cell lines showed impaired genome integrity compared to controls, which may also underlie AF [123].

Variants in Nesprin-1/2
Nesprins (SYNE) are a family of multi-isomeric scaffolding proteins. Nesprin 1/2 are located at the nuclear envelope and interact with SUN domain-containing proteins, lamin A/C and emerin, to form the LINC complex ( Figure 2) [124]. The LINC complex provides a stable physical connection between the nucleus and the cytoskeleton [124]. As mentioned above, the LINC complex participates in many cellular activities, including the maintenance of nuclear morphology, nuclear positioning, the mediation of mechanical cell signaling, and downstream gene regulation. Nesprin 1/2 are highly expressed in skeletal and cardiac muscles. Recent studies showed that variants in SYNE1 are linked to nuclear morphological changes and the mislocalization of and perturbed interaction with binding partners lamin A/C, emerin, and SUN2 [124,125]. Furthermore, variants in the SYNE1 and SYNE2 isoforms are also linked to abnormal mechanotransduction and altered gene expression [126]. These findings indicate that variants in nesprin may result in the derailment of the cytoskeletal network comparable to other mutant LINC-associated proteins, including lamin A/C and, as such, may drive AF.

Variants in Filamin C
Filamin C (FLNC) is one of three filamin protein that participates in sarcomere stability [127]. Filamin C is most commonly found in skeletal and cardiac muscles and is localized to the Z-disc, myotendinous junctions, the sarcolemma, and intercalated discs. FLNC variants may lead to proteostasis derailment, such as protein misfolding and aggregation and, consequently, the saturation of the ubiquitin-proteasome and autophagy pathways [128]. Since 'wear and tear' AF is also associated with proteostasis derailment and autophagic protein degradation, a comparable pathway may drive mutant FLNC-induced AF.

Conclusions
In 15% of AF patients, AF is caused by gene variants. Although emerging research findings indicate that variants in cytoskeletal (-associated) proteins may trigger early AF onset, knowledge on the underlying pathophysiological mechanisms is still in its infancy. Recent findings suggest that cytoskeletal (-associated) protein variants cause disruption of the structural network, mitochondrial dysfunction, electrophysiological changes, DNA damage-induced PARP1 activation, and NAD + depletion. As several of these mechanisms are also found to play a key role in 'wear and tear' AF, they may represent canonical pathophysiological pathways driving AF. Future research on cytoskeletal (-associated) protein variants should be directed at the dissection of the molecular root causes driving AF and the identification of novel druggable targets for therapy. As no treatments are available for variant carriers, these studies are clinically highly relevant. Funding: This research was funded by DHF, DZKH (DnAFix project, 2020B003, 2020) and the Atrial Fibrillation Innovation Platform (www.AFIPonline.org, accessed on 9 December 2021).

Conflicts of Interest:
The authors declare no conflict of interest.