Modulated Calcium Homeostasis and Release Events Under Atrial Fibrillation and Its Risk Factors: A Meta-Analysis

Background: Atrial fibrillation (AF) is associated with calcium (Ca2+) handling remodeling and increased spontaneous calcium release events (SCaEs). Nevertheless, its exact mechanism remains unclear, resulting in suboptimal primary and secondary preventative strategies. Methods: We searched the PubMed database for studies that investigated the relationship between SCaEs and AF and/or its risk factors. Meta-analysis was used to examine the Ca2+ mechanisms involved in the primary and secondary AF preventative groups. Results: We included a total of 74 studies, out of the identified 446 publications from inception (1982) until March 31, 2020. Forty-five were primary and 29 were secondary prevention studies for AF. The main Ca2+ release events, calcium transient (standardized mean difference (SMD) = 0.49; I2 = 35%; confidence interval (CI) = 0.33–0.66; p < 0.0001), and spark amplitude (SMD = 0.48; I2 = 0%; CI = −0.98–1.93; p = 0.054) were enhanced in the primary diseased group, while calcium transient frequency was increased in the secondary group. Calcium spark frequency was elevated in both the primary diseased and secondary AF groups. One of the key cardiac currents, the L-type calcium current (ICaL) was significantly downregulated in primary diseased (SMD = −1.07; I2 = 88%; CI = −1.94 to −0.20; p < 0.0001) and secondary AF groups (SMD = −1.28; I2 = 91%; CI = −2.04 to −0.52; p < 0.0001). Furthermore, the sodium–calcium exchanger (INCX) and NCX1 protein expression were significantly enhanced in the primary diseased group, while only NCX1 protein expression was shown to increase in the secondary AF studies. The phosphorylation of the ryanodine receptor at S2808 (pRyR-S2808) was significantly elevated in both the primary and secondary groups. It was increased in the primary diseased and proarrhythmic subgroups (SMD = 0.95; I2 = 64%; CI = 0.12–1.79; p = 0.074) and secondary AF group (SMD = 0.66; I2 = 63%; CI = 0.01–1.31; p < 0.0001). Sarco/endoplasmic reticulum Ca2+-ATPase (SERCA) expression was elevated in the primary diseased and proarrhythmic drug subgroups but substantially reduced in the secondary paroxysmal AF subgroup. Conclusions: Our study identified that ICaL is reduced in both the primary and secondary diseased groups. Furthermore, pRyR-S2808 and NCX1 protein expression are enhanced. The remodeling leads to elevated Ca2+ functional activities, such as increased frequencies or amplitude of Ca2+ spark and Ca2+ transient. The main difference identified between the primary and secondary diseased groups is SERCA expression, which is elevated in the primary diseased group and substantially reduced in the secondary paroxysmal AF subgroup. We believe our study will add new evidence to AF mechanisms and treatment targets.

Background: Atrial fibrillation (AF) is associated with calcium (Ca 2+ ) handling remodeling and increased spontaneous calcium release events (SCaEs). Nevertheless, its exact mechanism remains unclear, resulting in suboptimal primary and secondary preventative strategies.
Methods: We searched the PubMed database for studies that investigated the relationship between SCaEs and AF and/or its risk factors. Meta-analysis was used to examine the Ca 2+ mechanisms involved in the primary and secondary AF preventative groups.
Results: We included a total of 74 studies, out of the identified 446 publications from inception (1982) until March 31, 2020. Forty-five were primary and 29 were secondary prevention studies for AF. The main Ca 2+ release events, calcium transient (standardized mean difference (SMD) = 0.49; I 2 = 35%; confidence interval (CI) = 0.33-0.66; p < 0.0001), and spark amplitude (SMD = 0.48; I 2 = 0%; CI = −0.98-1.93; p = 0.054) were enhanced in the primary diseased group, while calcium transient frequency was increased in the secondary group. Calcium spark frequency was elevated in both the primary diseased and secondary AF groups. One of the key cardiac currents, the L-type calcium current (I CaL ) was significantly downregulated in primary diseased (SMD = −1.07; I 2 = 88%; CI = −1.94 to −0.20; p < 0.0001) and secondary AF groups (SMD = −1.28; I 2 = 91%; CI = −2.04 to −0.52; p < 0.0001). Furthermore, the sodium-calcium exchanger (I NCX ) and NCX1 protein expression were significantly enhanced in the primary diseased group, while only NCX1 protein expression was shown to increase in the secondary AF studies. The phosphorylation of the ryanodine receptor at S2808 (pRyR-S2808) was significantly elevated in both the primary and secondary groups. It was increased in the primary diseased and proarrhythmic subgroups (SMD = 0.95; I 2 = 64%; CI = 0.12-1.79; p = 0.074) and secondary AF group (SMD = 0.66; I 2 = 63%; CI = 0.01-1.31; p < 0.0001). Sarco/endoplasmic reticulum Ca 2+ -ATPase (SERCA) expression was elevated in the primary diseased and proarrhythmic drug subgroups but substantially reduced in the secondary paroxysmal AF subgroup.

INTRODUCTION
Atrial fibrillation (AF) is the most common sustained arrhythmia, with markedly increasing prevalence (1,2). It is associated with significant mortality and morbidity and becomes more challenging to treat as it advances (3)(4)(5). AF is mainly managed by primary and/or secondary preventative therapies. Primary prevention includes early detection and intervention on risk factors before AF develops, while secondary prevention involves diagnosing and treating AF (4). However, current pharmacological strategies are often associated with limited efficacy and adverse consequences, mainly due to an incomplete understanding of underlying cellular mechanisms related to AF (4,6). In particular, calcium (Ca 2+ ) is one of the most crucial ions for cardiac excitation-contraction coupling and Ca 2+ -dependent signaling pathways for maintaining cardiac function (7,8).
Intracellular Ca 2+ release events are exclusively investigated in myocardial physiology and pathophysiology, as they hold the key to understanding how cardiomyocyte Ca 2+ signaling is regulated by ionic channels and Ca 2+ proteins (7,8). In a single cardiac cycle, the L-type calcium channels (LTCCs) localized on the sarcolemma and tubules are first activated (9). The opening of LTCCs results in the movement of Ca 2+ into the cytosol, which induces the cardiac type 2 ryanodine receptors (RyR2) located on the junctional sarcoplasmic reticulum (SR) to release Ca 2+ from its stores into the cytosol (9)(10)(11). This elementary Ca 2+ release event is observed as a form of a Ca 2+ spark, and the process is known as calcium-induced calcium release (12). Increases in highly localized, short-lived Ca 2+ signals raise intracellular Ca 2+ [Ca 2+ ] i , which contributes to global Ca 2+ waves or transients that propagate through the cell (10,13). [Ca 2+ ] i then binds to troponin to allow myosin adenosine triphosphatase (ATPase) to bind to actin in the sarcomere to initiate cardiac contraction (9,14). Ca 2+ is mainly recycled back into the SR via the SR Ca 2+ -ATPase (SERCA2a) pump or extruded across the cell membrane through the cardiac sodium-calcium exchanger (NCX1) (15). The reduction in [Ca 2+ ] i causes Ca 2+ to dissociate from troponin and terminate myofilament cross-bridge cycling for cardiac relaxation (9,16). SERCA2a activity is directly modulated by phospholamban (PLN). In its unphosphorylated state, PLN acts as an inhibitor to SERCA2a. When phosphorylated by protein kinase A (PKA), PLN dislodges from SERCA2a to enable the reuptake of Ca 2+ (9). Another important signaling protein besides PKA is the Ca 2+ /calmodulin-dependent protein kinase II (CAMKII), which is responsible for transducing cytosolic Ca 2+ , and calmodulin, a Ca 2+ -binding messenger protein that modulates RyR activity and transduces Ca 2+ signals to other protein kinases or phosphatases (17)(18)(19).
In diseased states, spontaneous Ca 2+ release events (SCaEs) are observed as spontaneous Ca 2+ sparks or arrhythmogenic Ca 2+ waves are substantially enhanced (20,21). Such defective Ca 2+ homeostasis often results from remodeled Ca 2+ -handling proteins (22)(23)(24). However, current studies reported conflicting results on how these Ca 2+ -handling proteins were remodeled in AF and its risk factors, which hinders the development of effective AF treatment and prevention. In this study, we aim to illustrate the precise mechanisms and targeted therapies for AF and AF prevention by investigating the pathophysiological role of Ca 2+ and its arrhythmogenicity. This systematic review has compared the different Ca 2+ mechanisms between the primary and secondary AF preventative groups in the existing studies to date.

Search Strategy and Eligibility
The systematic electronic search was performed using the terms "atrial fibrillation" AND "calcium wave"/"calcium transient"/"calcium spark" in all fields to identify articles in PubMed from inception through March 31, 2020. Based on their titles and abstracts, the searched articles were screened manually for inclusion. Screening criteria included publications that mapped SCaEs in atrial cardiomyocytes in sinus rhythm and/or AF. Publications that did not conduct any experimental studies on atrial cells, such as mathematical modeling, population-based or organ-level studies, review papers, and editorial reports, were excluded. We also excluded papers that focused on genes and/or miRNA, signaling pathways, tissue, and organelle calcium experimental studies. Two co-authors then reviewed the screened articles in full text for eligibility, and those that met the criteria were selected. Any discrepancies were resolved by a third author through discussion and consensus. Full details of the search terms were presented in Supplementary Table 1. The quality of the included studies was assessed according to the Newcastle-Ottawa Scale (Supplementary Table 2) (26). A study with a score of 5 and above was considered satisfactory.

Data Extraction
The atrial cellular activities were extracted from selected studies. It included SCaEs [Ca 2+ spark frequency (CaSpF), Ca 2+ transient frequency (CaTF), Ca 2+ spark amplitude (CaSpA), and Ca 2+ transient amplitude (CaTA)], and Ca 2+ load and leak. It also included atrial current densities, such as L-type calcium current (I CaL ), sodium-calcium exchanger current (I NCX ), late sodium current (I Na−Late ), and potassium current (I K ), and protein expressions, such as L-type alpha 1C subunit voltage-dependent calcium current (Ca v 1.2), NCX1, RyR2, phosphorylated ryanodine receptor 2 (pRyR2), SERCA2a, PLN, phosphorylated phospholamban (pPLN), CAMKII, phosphorylated Ca 2+ /calmodulin-dependent protein kinase II (pCAMKII), and PKA. We further categorized the above results into two main groups: primary and secondary preventative therapies for AF. Primary prevention was divided into three subclasses: the diseased group (risk factors for AF), the application of proarrhythmogenic agents or antiarrhythmogenic agents. Secondary prevention was classified into either the paroxysmal or chronic AF group. The analysis was conducted using R.

Data Synthesis and Statistical Analysis
Statistical analyses were performed using R (27). Dichotomous values were used to calculate 95% confidence interval (CI) of relative risk ratios, and continuous values to quantify standardized mean difference (SMD). Each study was given a weighting factor to determine its importance in the meta-analysis, which was represented by gray boxes in forest plots. When the boundaries of the CI were within the box, a white horizontal line was plotted; otherwise, it was illustrated by a black horizontal line. Studies with the CI not crossing zero were deemed to be statistically significant.
The overall SMD was interpreted using Cohen's guidelines (28), where a value of 0.2-0.49 was deemed to be small, 0.5-0.79 represented medium, and 0.8 and above was large. Statistical heterogeneity was calculated using I 2 for all studies (29). In general, heterogeneity was classified into three main categories, low, medium, and high, when I 2 values were ≤25%, between 25 and 50%, and ≥75%, respectively. Statistical significance was measured with p-values. We considered a result to be statistically significant when p-value was ≤0.05. We also FIGURE 2 | An overview of pharmacologically targeted ionic channels or proteins for AF treatment, grouped by primary (blue) and secondary (orange) prevention. The sunburst plot's core was divided into two groups, primary and secondary prevention for AF, where the middle section represented ion channels, and the outer segment symbolized various drug therapies. I NCX stands for the sodium-calcium exchanger current; I CaL , the L-type calcium current; RyR2, the ryanodine receptor 2; I K , the potassium current; I Na−Late , the late sodium current; CAMKII, the Ca 2+ /calmodulin-dependent protein kinase II; PLN, phospholamban; SERCA2a, the sarco/endoplasmic reticulum Ca 2+ -ATPase 2a pump; I Na , the sodium current. employed influence analysis and graphic display of heterogeneity (GOSH) plots to detect influential studies and remove outliers (27).

Study Characteristics
Our literature search identified a total of 446 publications from inception (1982) to March 31, 2020 (Figure 1). When screening the titles and abstracts, a total of 372 papers were excluded (Supplementary Table 3): 25 articles focused on other diseases instead of AF, 101 included AF but did not conduct experiments on SCaEs, 13 papers mentioned signaling pathways but not Ca 2+ , 23 were tissue or organelle studies, another 24 studied mRNA and genes, 27 indicated mathematical models, 23 articles were controlled trials or case reports, and 136 were review articles. Eventually, a total of 74 studies (12, 24, 30-100), consisting of 45 primary and 29 secondary prevention studies for AF, were eligible and included for this systematic review.
Based on the 74 selected studies, pharmacological targets were grouped by their mechanism of action on the ionic channel(s) or protein(s). We discovered that I CaL was the most widely studied current in both primary and secondary pharmacological therapy for AF, followed by I NCX and RyR2 channels (Figure 2). This coincides with the present targeted drug therapies available for AF, where LTCC antagonists are one of the most frequently prescribed drugs for the treatment of hypertension and AF. Figure 2 aids us in understanding and exploring other potential pathways for therapeutic drug discovery, such as the I NCX and RyR2. It is noteworthy that the late sodium current was only commonly studied for primary prevention.

SR Ca 2+ Leak-Load Relationship
SR Ca 2+ release is affected by the opening of RyR channels from its stores. In particular, SR Ca 2+ leak is a major contributor to cardiac arrhythmia. Ca 2+ load remained relatively unchanged in both the primary and secondary subgroups, except when antiarrhythmic drugs were applied in the primary group (SMD = −0.40; I 2 = 59%; CI = −0.62 to −0.17; p < 0.0001) ( Figure 5A). No change in Ca 2+ leak was observed in the secondary prevention group (Supplementary Figure 5), but it was significantly affected by pro-and antiarrhythmic drugs in the primary subgroups. Ca  Figure 6F).

Main Findings
This preclinical systematic review study analyzed 74 articles identified from 446 searched primary and secondary AF prevention articles. Forty-five publications were classified as primary AF prevention studies and 29 others as secondary prevention. To our knowledge, this is the largest study of this kind to explore the association between modulated calcium homeostasis and release events for primary and secondary prevention of AF. Our principal findings are summarized as follows ( Table 1).
With regard to the key Ca 2+ channels/proteins/mediators, our study found that I CaL was the most widely studied current in both primary and secondary AF prevention, followed by I NCX and RyR2 channels.
• We showed that I CaL was significantly downregulated in primary and secondary diseased groups, which were largely consistent with our results for Ca v 1.2 protein expression. Antiarrhythmic drugs in the primary group further reduced I CaL significantly. • Furthermore, the NCX1 protein expression was significantly enhanced in both the primary and secondary diseased groups, but I NCX was only elevated in the primary diseased group. • In addition, our study demonstrated that the key phosphorylation expression for RyR was enhanced at serine 2808 in both the primary diseased and secondary AF groups, and inhibited in the primary antiarrhythmic drug subgroup. On the other hand, the other key RyR phosphorylation expression at serine 2814 showed no significant changes in both the primary and secondary diseased groups. • SERCA expression was elevated in the primary diseased and proarrhythmic drug subgroups but substantially reduced in the secondary paroxysmal AF subgroup. tPLN remained relatively constant in all primary subgroups but was decreased in paroxysmal AF.
• Finally, the Ca 2+ signaling mediator CAMKII was increased in the secondary AF group. With its phosphorylation activity at threonine 286, pCAMII-T286 was significantly raised by proarrhythmic drugs and significantly reduced by antiarrhythmic therapies. It is noteworthy that there is a growing surge of interest for the late sodium current I Na−Late and its direct effect on arrhythmia. Our study identified many primary preventative publications that showed that I Na−Late was moderately enhanced in the primary diseased and proarrhythmic drug subgroups but significantly antagonized by antiarrhythmic agents, with low heterogeneities across all three subgroups. As a result of atrial remodeling in the ionic channels and protein/signaling expressions in diseased and AF conditions, we observed changed Ca 2+ functional activities, i.e., Ca 2+ spark, Ca 2+ transient, and Ca 2+ load/leak. In the primary prevention group, CaSpF, CaTA, and CaSpA were significantly enhanced in the diseased subgroup and decreased by antiarrhythmic drug agents. On the other hand, CaTF and CaSpF were significantly elevated in both the secondary paroxysmal and chronic AF subgroups. Interestingly, we discovered that SR Ca 2+ load and Ca 2+ leak remained relatively constant in the primary and secondary subgroups, except when SR Ca 2+ load was reduced when antiarrhythmic drugs were applied in the primary group. Furthermore, we found that Ca 2+ leak was raised by proarrhythmic agents and antagonized by antiarrhythmic agents in the primary group.

Potential Mechanisms for Primary AF Diseases
The pathophysiological mechanism that causes spontaneous sarcoplasmic calcium release in the primary group involves the downregulation of I CaL and dysfunction of the Ca 2+ -handling proteins, in particular pRyR at S2808. The enhanced pRyR-S2808 activity may increase the frequency of the Ca 2+ spark due to calcium-induced calcium releases. This would eventually lead to a high Ca 2+ level in the cytosol and enhanced trigger activities via the forward mode of I NCX , which was demonstrated in this review as the elevation of I NCX activity. Furthermore, the enhancement of I Na−Late was shown to potentially play a more significant role in the generation of arrhythmia in our review and recent studies (95)(96)(97). The reduction in I CaL was predicted to result in reduced SR load and diminished spontaneous activity. On the contrary, the downregulation of Ca v 1.2 current and protein expression has resulted in SCaEs, which could possibly be due to cell compensation (98,99). Overall, this interplay led to an overload of Ca 2+ in the cell, potentially causing AF.

Abnormal Ca 2+ Activity in AF
The electrophysiological remodeling induced in the fibrillating atria and its molecular basis were extensively reviewed. Recent and past AF studies (95)(96)(97) have suggested that I CaL density was downregulated, together with the reduction of its protein expression, Ca v 1.2. Strangely, reduced I CaL density did not diminish the SR load; it remained unchanged (96). In contrast to the primary AF disease, consistently reduced SERCA levels were identified, reducing the releasable SR Ca 2+ in the cytosol (98,99). On the other hand, some studies observed that the SR Ca 2+ leak and activity of RyR were consistently upregulated (97,100). Their observation justified the increase in SCaEs. The NCX expression was also increased, in contrast to what we saw earlier in the primary mechanism (100). The increased NCX expression could also account for the increase in frequencies of both the Ca 2+ sparks and Ca 2+ transients. The overextrusion of Ca 2+ explains the unchanged CaSpA and CaTA.
We have shown that I CaL and NCX1 protein are the primary remodeling targets identified, and this leads to spontaneous calcium activity due to its interrelationship with the SR proteins. The conclusion of the secondary group meta-analysis aligns perfectly with the AF mechanism provided by Madsen et al. (34). On the other hand, the proposed primary mechanism is the best agglomeration of the mechanisms acquired from individual modifiable/non-modifiable risk factors associated with AF.

Limitations
This review aims to provide a better understanding of the mechanisms involved in the [Ca 2+ ] i homeostasis within atrial cardiomyocytes and compare their activities among the primary and secondary AF subgroups. Although this review has comprehensively compiled the Ca 2+ activity from inception to date, it still presents several limitations. The activity of SERCA and PLN appears to be unclear. It is certain that I CaL was reduced in the primary diseased and secondary AF groups. Surprisingly, the SR Ca 2+ load-leak relationship was unaltered in the primary diseased and secondary AF groups with high heterogeneities. This could be influenced by a variety of non-controllable factors, such as the variability in the animal and human studies at various stages of AF or diseased states, the type and strength of pharmacological agents applied, and the different experimental settings and methodologies (101,102).

CONCLUSION
Our study identified that I CaL is reduced in both primary and secondary diseased groups. Furthermore, pRyR-S2808 and NCX1 protein expression are enhanced. The remodeling leads to elevated Ca 2+ functional activities, such as the frequencies or amplitude of Ca 2+ spark and Ca 2+ transient. The main difference identified between primary and secondary diseased groups is the SERCA expression, which is elevated in the primary diseased group and substantially reduced in the secondary paroxysmal AF subgroup. We believe our study will add new evidence to AF mechanisms and treatment targets.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.

AUTHOR CONTRIBUTIONS
SF and SA undertook data extraction, post-processing, and analysis, as well as drafting the manuscript. MG assisted in assessing the quality of the included studies. JZ guided the project and revised the manuscript. All authors contributed to the article and approved the submitted version.

FUNDING
This work was supported by the Health Research Council of New Zealand (16/385) and National Heart Foundation of New Zealand.