VLDL from Metabolic Syndrome Individuals Enhanced Lipid Accumulation in Atria with Association of Susceptibility to Atrial Fibrillation

Metabolic syndrome (MetS) represents a cluster of metabolic derangements. Dyslipidemia is an important factor in MetS and is related to atrial fibrillation (AF). We hypothesized that very low density lipoproteins (VLDL) in MetS (MetS-VLDL) may induce atrial dilatation and vulnerability to AF. VLDL was therefore separated from normal (normal-VLDL) and MetS individuals. Wild type C57BL/6 male mice were divided into control, normal-VLDL (nVLDL), and MetS-VLDL (msVLDL) groups. VLDL (15 µg/g) and equivalent volumes of saline were injected via tail vein three times a week for six consecutive weeks. Cardiac chamber size and function were measured by echocardiography. MetS-VLDL significantly caused left atrial dilation (control, n = 10, 1.64 ± 0.23 mm; nVLDL, n = 7, 1.84 ± 0.13 mm; msVLDL, n = 10, 2.18 ± 0.24 mm; p < 0.0001) at week 6, associated with decreased ejection fraction (control, n = 10, 62.5% ± 7.7%, vs. msVLDL, n = 10, 52.9% ± 9.6%; p < 0.05). Isoproterenol-challenge experiment resulted in AF in young msVLDL mice. Unprovoked AF occurred only in elderly msVLDL mice. Immunohistochemistry showed excess lipid accumulation and apoptosis in msVLDL mice atria. These findings suggest a pivotal role of VLDL in AF pathogenesis for MetS individuals.


Introduction
The incidence of atrial fibrillation (AF), the most common arrhythmia, is rising as the population ages [1]. Metabolic syndrome (MetS) represents a cluster of cardiovascular and metabolic derangements

Very Low Density Lipoproteins (VLDL) in MetS (MetS-VLDL) Increased Oxidative Stress and Cytotoxicity to HL-1 Atrial Myocytes
To identify the cytotoxic effect of VLDLs on cardiomyocytes, we exposed cultured HL-1 cells to different concentrations of normal-VLDL and MetS-VLDL, using a CCK-8 assay to determine cell viability. Optical density (OD) values decreased at 450 nm indicating reduction of cell viability. Normal-VLDL did not affect HL-1 cell viability at concentrations from 3.125 to 25 mg/dL after 24 h. MetS-VLDL, however, significantly reduced cell viability at concentrations of 25 mg/dL after 24 h (p < 0.01; Figure 1A,B). Reactive oxygen species (ROS) activity was significantly increased by 25 mg/dL MetS-VLDL incubation compared with control and normal-VLDL (n = 4, p < 0.01 and p < 0.05, respectively; Figure 1C), suggesting that only MetS-VLDL caused oxidative stress to HL-1 cells.

Isoproterenol Challenge Induced Atrial Fibrillation (AF) in msVLDL Mice
Isoproterenol infusion is used for arrhythmia induction in clinical electrophysiology. For mice, we tested a single dose of intraperitoneal isoproterenol for inducing arrhythmias [19,20]. We performed electrocardiography after three, four, five and six weeks of VLDLs or saline injection for young mice. Two msVLDL mice had frequent premature atrial complexes (PACs at the 3rd and 6th week). The msVLDL mouse with PACs at week 3 died at week 4. After week 6, remaining mice were used for isoproterenol challenge. After a single injection of isoproterenol (100 ng/kg), the heart rate (HR) increased dramatically. HR responses following isoproterenol injection were similar among groups ( Figure 3E). After injection of isoproterenol, one control mouse (n = 5), and one nVLDL mouse (n = 7) had PACs ( Figure 3B). Two msVLDL mice had premature ventricular complexes (PVCs) and atrial fibrillation (n = 5) ( Figure 3C,D).

Isoproterenol Challenge Induced Atrial Fibrillation (AF) in msVLDL Mice
Isoproterenol infusion is used for arrhythmia induction in clinical electrophysiology. For mice, we tested a single dose of intraperitoneal isoproterenol for inducing arrhythmias [19,20]. We performed electrocardiography after three, four, five and six weeks of VLDLs or saline injection for young mice. Two msVLDL mice had frequent premature atrial complexes (PACs at the 3rd and 6th week). The msVLDL mouse with PACs at week 3 died at week 4. After week 6, remaining mice were used for isoproterenol challenge. After a single injection of isoproterenol (100 ng/kg), the heart rate (HR) increased dramatically. HR responses following isoproterenol injection were similar among groups ( Figure 3E). After injection of isoproterenol, one control mouse (n = 5), and one nVLDL mouse (n = 7) had PACs ( Figure 3B). Two msVLDL mice had premature ventricular complexes (PVCs) and atrial fibrillation (n = 5) ( Figure 3C,D). in the nVLDL group, premature ventricular complex (PVC *) and AF (absence of clear P waves and irregular RR intervals) in the msVLDL group (n = 5); (E) Heart rate responses after isoproterenol injection were not different among groups; (F) For elderly mice, spontaneous, unprovoked AF was noted in the msVLDL group (n = 6) with an incidence of 50%. PAC was observed in one mouse in the nVLDL group (n = 5). All control mice (n = 5) had sinus rhythm. $ p < 0.001 for msVLDL vs. control and # p < 0.001 for msVLDL vs. nVLDL.

Unprovoked Electrocardiography Showed AF in msVLDL Mice
We tested the hypothesis that spontaneous, unprovoked AF might develop if we injected msVLDL in elderly mice. Therefore, we performed the same VLDLs or saline injection experiment using 9-month-old mice. The groups, breeding and injection fashion were all the same as that used in the young mice. After 6 weeks of injection, AF was observed in 50% of msVLDL mice (n = 6). Sinus rhythm was noted for all nVLDL (n = 5) and control mice (n = 5) ( Figure 3F). The incidence of documented AF was significantly higher in msVLDL compared to nVLDL and control mice (p < 0.001).

MetS-VLDL Caused Atrial Tissue Apoptosis
Histological studies determined that 6 weeks' injection of VLDL had cytotoxic effects in vivo. In in situ terminal deoxynucleotidyl transferase (TUNEL) staining, normal nuclei appear blue with DAPI staining while bright green condensed or fragmented nuclei indicate apoptosis. Apoptotic cells were observed only in msVLDL mice (n = 3 for each group; Figure 4). , premature atrial complex (PAC) in the nVLDL group, premature ventricular complex (PVC *) and AF (absence of clear P waves and irregular RR intervals) in the msVLDL group (n = 5); (E) Heart rate responses after isoproterenol injection were not different among groups; (F) For elderly mice, spontaneous, unprovoked AF was noted in the msVLDL group (n = 6) with an incidence of 50%. PAC was observed in one mouse in the nVLDL group (n = 5). All control mice (n = 5) had sinus rhythm. $ p < 0.001 for msVLDL vs. control and # p < 0.001 for msVLDL vs. nVLDL.

Unprovoked Electrocardiography Showed AF in msVLDL Mice
We tested the hypothesis that spontaneous, unprovoked AF might develop if we injected msVLDL in elderly mice. Therefore, we performed the same VLDLs or saline injection experiment using 9-month-old mice. The groups, breeding and injection fashion were all the same as that used in the young mice. After 6 weeks of injection, AF was observed in 50% of msVLDL mice (n = 6). Sinus rhythm was noted for all nVLDL (n = 5) and control mice (n = 5) ( Figure 3F). The incidence of documented AF was significantly higher in msVLDL compared to nVLDL and control mice (p < 0.001).

MetS-VLDL Caused Atrial Tissue Apoptosis
Histological studies determined that 6 weeks' injection of VLDL had cytotoxic effects in vivo. In in situ terminal deoxynucleotidyl transferase (TUNEL) staining, normal nuclei appear blue with

Increased Lipid Accumulation in Atrial Tissue of msVLDL Mice
To determine if normal-VLDL and MetS-VLDL were internalized into atrial tissues differently, we performed Oil-Red-O staining of atrial tissues ( Figure 5). A few tiny red lipid droplets were seen in control atrial tissue. In nVLDL and msVLDL atrial tissue, lipid droplets were significantly increased (n = 3 for each, p < 0.001 vs. control) and lipid droplets were larger and more numerous in msVLDL atrial tissues compared to nVLDL (p < 0.01 for msVLDL vs. nVLDL), suggesting that MetS-VLDL increased lipid accumulation in atrial tissues.

Discussion
This study showed the distinctive effects of MetS-VLDL on the heart, specifically atrial myocyte apoptosis, left atrial dilation, and AF vulnerability and incidence. Physiological concentrations of MetS-VLDL caused greater lipid accumulation in atrial cells and tissue than normal-VLDL, partially

Increased Lipid Accumulation in Atrial Tissue of msVLDL Mice
To determine if normal-VLDL and MetS-VLDL were internalized into atrial tissues differently, we performed Oil-Red-O staining of atrial tissues ( Figure 5). A few tiny red lipid droplets were seen in control atrial tissue. In nVLDL and msVLDL atrial tissue, lipid droplets were significantly increased (n = 3 for each, p < 0.001 vs. control) and lipid droplets were larger and more numerous in msVLDL atrial tissues compared to nVLDL (p < 0.01 for msVLDL vs. nVLDL), suggesting that MetS-VLDL increased lipid accumulation in atrial tissues.

Increased Lipid Accumulation in Atrial Tissue of msVLDL Mice
To determine if normal-VLDL and MetS-VLDL were internalized into atrial tissues differently, we performed Oil-Red-O staining of atrial tissues ( Figure 5). A few tiny red lipid droplets were seen in control atrial tissue. In nVLDL and msVLDL atrial tissue, lipid droplets were significantly increased (n = 3 for each, p < 0.001 vs. control) and lipid droplets were larger and more numerous in msVLDL atrial tissues compared to nVLDL (p < 0.01 for msVLDL vs. nVLDL), suggesting that MetS-VLDL increased lipid accumulation in atrial tissues. . Each red rectangle indicates the area to be magnified (20×). Tiny red lipid droplets in controls were few but the number increased in nVLDL and msVLDL atria. Some lipid droplets increased in size in the msVLDL; (B) Lipid droplets were significantly increased in the VLDL groups, especially in the msVLDL group (** p < 0.01, *** p < 0.001; n = 3 for each group).

Discussion
This study showed the distinctive effects of MetS-VLDL on the heart, specifically atrial myocyte apoptosis, left atrial dilation, and AF vulnerability and incidence. Physiological concentrations of MetS-VLDL caused greater lipid accumulation in atrial cells and tissue than normal-VLDL, partially Each red rectangle indicates the area to be magnified (20ˆ). Tiny red lipid droplets in controls were few but the number increased in nVLDL and msVLDL atria. Some lipid droplets increased in size in the msVLDL; (B) Lipid droplets were significantly increased in the VLDL groups, especially in the msVLDL group (** p < 0.01, *** p < 0.001; n = 3 for each group).

Discussion
This study showed the distinctive effects of MetS-VLDL on the heart, specifically atrial myocyte apoptosis, left atrial dilation, and AF vulnerability and incidence. Physiological concentrations of MetS-VLDL caused greater lipid accumulation in atrial cells and tissue than normal-VLDL, partially via VLDL receptors.

Increased Lipid Accumulation and in Vivo and in Vitro Cytotoxicity of MetS-VLDL to Atrium
Greater lipid accumulation in hearts of MetS patients is associated with decreased ventricular function and cardiomyopathy [21,22]. Studies have suggested that increased triglyceride accumulation could lead to reduced energy efficiency by inducing mitochondrial damage and uncoupling, thus increasing cellular ROS, impairment of mitochondrial calcium handling, and lipoapoptosis [23]. Our study showed in vitro evidence that compared to normal-VLDL, MetS-VLDLs caused greater lipid uptake and cytotoxicity in parallel to increased cellular ROS in atrial cells (Figures 1 and 4). Consistently, MetS-VLDL induced greater lipid accumulation and apoptosis in vivo (Figures 1 and 5). In addition to direct cytotoxicity, we suggest that in vivo MetS-VLDL-induced apoptosis in atrial tissue may be also related to excessive lipid accumulation.
In normal healthy hearts, cardiac energy primarily relies on oxidation of fatty acids and to a lesser extent on glucose metabolism [24]. VLDL and chylomicrons are major sources of triglyceride for the heart. Triglyceride may be taken up through lipoprotein-lipase (LPL)-mediated lipolysis and lipoprotein receptor-mediated endocytosis. The VLDL uptake can be increased in hypoxic conditions mediated by HIF-1 α through low-density lipoprotein-related protein (LRP1) [25]. In hypoxic HL-1 cells and ischemic myocardium, VLDLR plays a major role in VLDL uptake by the heart, and closely interacts with LPL [26,27]. We found that MetS-VLDL uptake, at least partially, occurred through VLDLR ( Figure 1D,E). Further experiments are needed to establish the precise contribution of LPL, VLDLR, and LRP1 in MetS-VLDL and normal-VLDL uptake in atrial and ventricular cardiomyocytes.
Normal VLDL particles are considered non-toxic to vascular cells, but apolipoprotein CIII (apoCIII)-rich VLDL exhibits atherogenicity by enhancing monocyte-endothelial cell (EC) adhesion [28,29]. ApoCIII also inhibits the uptake of VLDL by the liver. Our chemical analyses suggest that MetS-VLDL is an apoCIII-and apoE-rich lipoprotein (data not shown). Our past study showed that higher negative charged subfraction of VLDL in MetS compared to healthy subjects [17]. The negatively charged property of MetS-VLDL may be caused by post-translational modification of apolipoproteins. Further studies are mandatory to determine the biochemical changes with MetS-VLDL and the mechanism behind its negative charge.

MetS-VLDL Causes Cardiac Remodeling
An elegant study by Asai et al. [30] provided evidence that intracellular lipid moieties mediate metabolic signals leading to cardiomyocyte growth, and that marked myocardial triglyceride accumulation was associated with left ventricular dysfunction. In agreement with these findings, our study showed that ventricular dilatation with increased LV mass was caused by both normal-VLDL and MetS-VLDL injection, with worse left ventricular dysfunction after MetS-VLDL injection ( Table 1). The important role of lipids in mediating left ventricular structure change has been well established, but little is known about how atrial myocytes and tissue are affected by lipid accumulation. In our animal model, the temporal change in atrial and ventricular sizes is illustrated in Figure 2. Although it is generally agreed that ventricular dysfunction produces atrial remodeling, atrial remodeling in our msVLDL mice seemed to precede the ventricular remodeling ( Figure 2). The atrial remodeling in msVLDL mice was significant as early as four weeks after injection, but significant ventricular remodeling was not observed until six weeks. Therefore, we suggested that MetS-VLDL-caused atrial dilatation is not due to ventricular remodeling. Although the immunological reactions may not be avoided using human VLDLs in our mouse model, this study provides direct evidence that the in vivo effects of MetS-VLDL and normal-VLDL on atrial remodeling and AF vulnerability are different.
Two major remodeling mechanisms have been proposed for AF development; structural remodeling characterized by atrial fibrosis, and electrical remodeling characterized by ion channels' modulation [31]. We performed Masson's Trichrome staining of the young mouse atrial tissue but did not observe significant atrial fibrosis in our VLDL injection mice (data not shown). To further elucidate the cellular mechanism in MetS-VLDL-induced atrial dilation, we performed a messenger RNA microarray study of control, nVLDL, and msVLDL mice atrial tissues (data not shown). The preliminary microarray data showed that some genes involving the lipid metabolism, contractile proteins, and calcium regulation proteins were dysregulated in msVLDL atrium, suggesting that electrical remodeling may be associated with atrial dilatation in msVLDL mice. Castellano et al. [32] reported that in neonatal rat cardiomyocytes, high concentrations of VLDL can induce cholesteryl ester and triglyceride accumulation, and reduce sarcoplasmic reticulum Ca ATPase-2 expression, calcium transient amplitude and sarcoplasmic reticulum calcium loading. Myocardial lipid and fatty acid compositions in atrial tissues were not determined in this study. It would be interesting to study the related electrical remodeling in msVLDL mice in terms of atrial effective refractory period, calcium transient and/or electrical propagation and its relation to lipid accumulation. Although the molecular mechanisms by which MetS-VLDL leads to cardiac remodeling in parallel with lipid accumulation remain undetermined, the current study provides evidence on the scope of VLDL roles in AF pathogenesis.

Clinical Implications
In view of the debate on the relationship between dyslipidemia and AF in recent years, this study offers new insights addressing the discrepancy seen among diverse clinical studies. The strikingly different in vivo and in vitro effects of normal and MetS-VLDL shown in this study suggested a positive correlation for dyslipidemia and AF, especially in MetS.
The prevalence of AF increased with aging. In this study, we observed spontaneous AF in nine-month-old mice but not in young mice after receiving six-week MetS-VLDL injection. The development of cardiac fibrosis, one of the important changes in aging hearts, may lead to a better substrate for AF development [33]. Nevertheless, it is still unknown if VLDL could interact with fibrosis or any aging process of the heart, especially in the atrium.
Our results show that msVLDL mice is more vulnerable to AF and suggest a pivotal role of VLDL in AF pathogenesis for MetS ( Figure 6). In MetS, the biochemical properties of VLDL are changed and MetS-VLDL can induce cellular reactive oxygen species, atrial myocyte cytotoxicity, and excess lipid accumulation resulting in subsequent gene dysregulation corresponding to metabolic derangement. We suggested that both structural and electrical remodeling initiated by MetS-VLDL contribute in concert to AF. After the underlying mechanisms of the biochemical property changes of VLDL in MetS are determined, we may be able to revert bad VLDL to a normal state and thus prevent the development of AF in MetS individuals, especially in the aging population. and excess lipid accumulation resulting in subsequent gene dysregulation corresponding to metabolic derangement. We suggested that both structural and electrical remodeling initiated by MetS-VLDL contribute in concert to AF. After the underlying mechanisms of the biochemical property changes of VLDL in MetS are determined, we may be able to revert bad VLDL to a normal state and thus prevent the development of AF in MetS individuals, especially in the aging population.

VLDL Isolation
Human normal-VLDL and MetS-VLDL were isolated from pooled blood of healthy volunteers (two males and two females, age 36˘8) and individuals meeting criteria for MetS based on National Cholesterol Education Program-Adult Treatment Panel III guidelines (five males, age 48˘5) [34]. All participants gave informed consent and the study followed Helsinki Declaration principles and was approved by the Kaohsiung Medical University Hospital Ethics Review Board (the project identification code KMUH-IRB-20130351, 24 January 2014. Total VLDL (density = 0.930-1.006 g/mL) was isolated by sequential ultracentrifugation [17]. VLDL protein concentration was determined by the bicinchoninic acid method.

HL-1 Atrial Myocyte Culture
A murine HL-1 atrial myocyte cell line was maintained with fresh Claycomb medium in pre-coated culture flasks at 37˝C in a humidified atmosphere containing 5% CO 2 . When the cells reached confluence, splitting was performed by recommended passaging procedures. Culture medium was supplemented with 87% Claycomb medium, 2 mM/L L-glutamine, 10% fetal bovine serum, 100 U/mL penicillin, 100 µg/mL streptomycin, and 0.1 mM/L norepinephrine.

Cytotoxicity of VLDL to HL-1 Cells
Cell viability was evaluated by CCK-8 assay (Sigma-Adrich, St. Louis, MO, USA). HL-1 cells were treated with either normal-VLDL or MetS-VLDL in 96-well plates (5ˆ10 3 cells per well) at different test concentrations (3.125, 6.25, 12.5 and 25 mg/dL). Controls were incubated with ordinary medium. After 24 h, cells were washed with D-Hanks buffer. Two hundred microlitres of CCK-8 solution was added to each well and incubated for 3 h at 37˝C. Optical density (OD) at 450 nm was read on a Microplate Reader (Thermo, Waltham, MA, USA). Cell viability (% of control) is expressed as the percentage of (ODtest´ODblank)/(ODcontrol´ODblank), where ODtest is the optical density of the cells exposed to VLDL, ODcontrol is the optical density of the control sample and ODblank is the optical density of the wells without HL-1 cells.

Reactive Oxygen Species (ROS)
To measure effects of VLDLs on cytosolic oxidant activity, HL-1 cells were seeded at 8ˆ10 4 cells/well overnight and then treated with normal-VLDL or MetS-VLDL at a concentration of 25 µg/mL for 16 h (n = 4 for each group). The culture medium was replaced with pre-warmed D-PBS and 2 1 ,7 1 -dichlorofluorescein diacetate (DCFH-DA 20 µM, Molecular Probes, Eugene, OR, USA) for 20 min of incubation. After incubation, cells were washed twice with D-PBS. A Bio-Tek FL-800 reader (BioTek, Winooski, VT, USA) was used to measure DCF fluorescence (excitation: 480 nm, emission 520 nm).

VLDL Uptake by HL-1 Cells
VLDLs were labeled with DiI by a modification of the method of Beisiegel et al. [35], incubating VLDL (1 mg/mL) in PBS-0.5% BSA with 80 µmol/L DiI in DMSO (3 mg/mL) for 8 h at 37˝C. Cells were pre-incubated in the presence or absence of anti-VLDL antibody (Novus Biologicals, Littleton, CO, USA) for 24 h and then incubated with DiI-normal-VLDL and DiI-MetS-VLDL at a final concentration of 25 µg/mL for 16 h before staining with 10 µg/mL Hoechst 33258 for fluorescent microscopic observation (n = 3 for each group).

Mice and Diet
To test in vivo arrhythmogenic effects of MetS-VLDL, we gave MetS-VLDL or normal-VLDL by intravenous injection in mice tail veins at a dose of 15 µg/g, three times a week for 6 consecutive weeks. The 15 µg/g dose was chosen to match the human normal VLDL concentration [34], which is between 2 and 30 mg/dL. A single 15 µg/g VLDL injection will result in a 15 mg/dL plasma level for mice. Five-week-old male C57BL/6 male mice from the National Laboratory Animal Center (Taipei, Taiwan) were maintained in a temperature-controlled facility (21-22˝C) with a 12-h light/dark cycle, free access to water and a standard chow diet. After a two-week acclimatization period, mice were randomly separated into three groups: control (n = 10), nVLDL (n = 7), and msVLDL (n = 11), and an equivalent 50 µL volume of phosphate buffered saline, 15 µg/g normal-VLDL or MetS-VLDL was injected respectively. Body weights were recorded weekly. To examine if AF could spontaneously occur in elderly mice, we used a batch of 9-month-old mice to perform the same VLDL injection for 6 weeks and afterwards performed electrocardiography without any provoking (control, n = 5; nVLDL, n = 5; msVLDL, n = 6; see below). All applicable institutional and governmental regulations concerning ethical use of animals conformed to the NIH guidelines and all animal procedures were approved by the Institutional Animal Care and Use Committee of Kaohsiung Medical University.

Unprovoked Electrocardiography for Elderly Mice
After 6 weeks of VLDL or saline injection, platinum electrodes were inserted subcutaneously in the limbs and connected to a custom-built ECG amplifier under anesthesia with intraperitoneal injection of pentobarbital 0.5-1.0 µg/g. The recording was initiated after the tracing was stable and lasted for 5 min for each test.

Oil Red O (ORO) Staining of Atrial Tissue and Quantification
Lipid accumulation was assessed by ORO staining of 10 µm paraffin sections of atrial tissues fixed in phosphate-buffered 4% paraformaldehyde. Color images were acquired using a micropublisher 3.3 RTV camera, saved as TIFF files, and analyzed using Image J software. A 100-µm 2 grid overlay was generated over each image, and the area fraction (%), defined as (points over ORO)/(points over image)/100, was determined. The values were derived from the average over the entire area of atrial tissue in multiple animals for each experimental group (n = 3 for each).

Data Analysis and Statistics
Data were expressed as means˘SD; n indicates the number of cell samples or mice. One-way ANOVA and Tukey's multiple comparisons test were used to compare values among groups. Chi-square test was used to determine the difference of AF incidence among groups. Statistical significance was considered as p value ď0.05.

Conclusions
In this study, we showed for the first time that physiological concentrations of MetS-VLDL cause atrial myocyte cytotoxicity, excess lipid accumulation and apoptosis in atria, resulting in left atrial enlargement, and that these changes are associated with increased incidence of AF. These findings have potential clinical impact. Our results suggest that VLDL may serve as a potential mediator of MetS-related AF and thus be promising a therapeutic target for AF prevention in MetS.