Correlation between increased atrial expression of genes related to fatty acid metabolism and autophagy in patients with chronic atrial fibrillation

Atrial metabolic disturbance contributes to the onset and development of atrial fibrillation (AF). Autophagy plays a role in maintaining the cellular energy balance. We examined whether atrial gene expressions related to fatty acid metabolism and autophagy are altered in chronic AF and whether they are related to each other. Right atrial tissue was obtained during heart surgery from 51 patients with sinus rhythm (SR, n = 38) or chronic AF (n = 13). Preoperative fasting serum free-fatty-acid levels were significantly higher in the AF patients. The atrial gene expression of fatty acid binding protein 3 (FABP3), which is involved in the cells' fatty acid uptake and intracellular fatty acid transport, was significantly increased in AF patients compared to SR patients; in the SR patients it was positively correlated with the right atrial diameter and intra-atrial electromechanical delay (EMD), parameters of structural and electrical atrial remodeling that were evaluated by an echocardiography. In contrast, the two groups' atrial contents of diacylglycerol (DAG), a toxic fatty acid metabolite, were comparable. Importantly, the atrial gene expression of microtubule-associated protein light chain 3 (LC3) was significantly increased in AF patients, and autophagy-related genes including LC3 were positively correlated with the atrial expression of FABP3. In conclusion, in chronic AF patients, the atrial expression of FABP3 was upregulated in association with autophagy-related genes without altered atrial DAG content. Our findings may support the hypothesis that dysregulated cardiac fatty acid metabolism contributes to the progression of AF and induction of autophagy has a cardioprotective effect against cardiac lipotoxicity in chronic AF.


Introduction
incision site or the insertion point of a drainage cannula. The tissue was frozen and stored at −80˚C until analysis.
Type 2 diabetes was defined as a fasting glucose level �7.0 mmol/L and/or taking antidiabetic medications. Coronary artery disease was evaluated by coronary angiography, and stenosis �75% was defined as significant; a patient with a history of percutaneous coronary intervention was also regarded as having coronary artery disease.
The study protocol was approved by the Ethics Committees of Hokkaido University Hospital and Teine Keijinkai Hospital and performed according to the Declaration of Helsinki. Written informed consent was obtained from each patient before the surgery. This study was registered in the UMIN Clinical Trials Registry: UMIN000012405 and UMIN000018137.

Transthoracic echocardiography
A Vivid Seven system (GE/Vingmed, Milwaukee, WI) with an M3S (2.5-3.5 MHz) transducer, an Aplio system (Toshiba Medical Systems, Tokyo, Japan) with a 2.5-MHz transducer, or a Philips system (Philips Ultrasound, Bothell, WA) with a 2.5-MHz transducer was used for the pre-operative echocardiography. The left ventricular and atrial diameters were measured from the parasternal long-axis view. The right atrial diameter (the largest minor-axis diameter in the four-chamber view) was obtained at end-systole [17].
The electromechanical delay (EMD) of the right atrium was assessed in all but one of the SR patients as a time interval (T1) from the beginning of the P-wave on surface ECG to the beginning of the late diastolic wave (A') of the tricuspid annulus [17]. T2 was the time interval from the beginning of the P-wave on the surface ECG to the A' of the septal mitral annulus. The intra-atrial EMD was the interval from T1 to T2, expressed as T2-T1 (msec). The left ventricular ejection fraction (LVEF) was measured using the biplane method of disks. The mitral regurgitation grade was defined by the regurgitation jet area-to-left atrium ratio (mild, <20%; moderate, 20%-40%; severe, >40%).

Blood biochemistry
Blood was collected from each patient early in the morning after a 10-hr fast within 2 days before surgery. The following biochemical values were analysed by the methods shown in S1 Table: the blood glucose, hemoglobin A1c, insulin, FFA, total cholesterol, triglycerides, and Btype natriuretic peptide (BNP).

The atrial expression of genes related to fatty acid metabolism and autophagy
The atrial expression of the following genes related to fatty acid metabolism and autophagy was determined by reverse transcription-polymerase chain reaction (RT-PCR): cluster of differentiation 36/fatty acid translocase (CD36), carnitine palmitoyltransferase 1B (CPT1B), fatty acid-binding protein 3 (FABP3), autophagy-related gene 5 (ATG5), Unc-51-like kinase 1 (ULK1), Beclin-1 (BCLN1), and microtubule-associated protein light chain 3 (LC3). Myocardial total RNA was isolated from frozen tissue using a High Pure RNA Tissue Kit (Roche, Penzberg, Germany) and was then reversely transcribed into cDNA using a Transcriptor First Strand cDNA Synthesis Kit (Roche).
The RT-PCR was performed using FastStart Essential DNA Probes Master (Roche) and the Real-time Ready Assay (Roche Assay ID: CD36, 144833; CPT1B, 126501; FABP3, 118811; ATG5, 125999; ULK1, 109914; BCLN1, 100115; LC3, 144582; TBP, 143707). Detailed information about probes and primers is shown in S2 Table. PCR amplification was then performed with a reaction volume of 20 μL using a LightCycler Nano (Roche) under the conditions specified by the manufacturer. After the initial denaturation and activation of the enzyme for 10 min at 95˚C, 45 cycles of denaturation at 95˚C for 10 sec and annealing and extension at 60˚C for 30 sec were performed. TBP (TATA-binding protein) was used as a reference gene, as we confirmed that TBP is not influenced by the occurrence of AF.

The atrial protein expression level of autophagy
We measured atrial protein expression level of LC3-II, an autophagosome marker, by Western blot analysis. A semi-dry Western blot apparatus (Mini-PROTEAN Tetra Cell, BIO-RAD, CA) detected the conversion from LC3-I (cytosolic form) to LC3-II (membrane-bound lipidated form). The amount of LC3-II protein reflects the number of autophagosomes. After sodium dodecyl sulfate-polyacrylamide gel electrophoresis (12% Mini-PROTEANTGXTM, BIO-RAD, CA), the proteins were blotted to a polyvinyldine deflouride membrane and incubated with primary (Anti-LC3B, Abcam, Cambridge, UK) and secondary antibodies (Anti-rabbit IgG, Cell Signaling, MA). The bands were quantified by chemiluminescence using JustTLC (Sweday, Sodra Sandby, Sweden). The membranes were dyed with naphthol blue black solution to normalize the band intensity. Due to the lack of atrial muscle samples, the protein expression level of LC3-II was measured in three patients with SR and four patients with AF.

The atrial enzymatic activities of the mitochondrial TCA cycle and fatty acid β-oxidation
We spectrophotometrically determined the activity levels of both citrate synthase (CS), a key enzyme in the tricarboxylic acid (TCA) cycle, and β-hydroxyacyl CoA dehydrogenase (β-HAD), a key enzyme in fatty acid β-oxidation, in the myocardial samples as described [19]. Due to the lack of atrial muscle samples, these enzymatic activities were measured in 20 patients with SR and eight patients with AF.

The atrial DAG content
Heart tissue was homogenized in 1.5 mL of methanol, followed by mixing with 2.25 mL of 1 M NaCl and 2.5 mL of chloroform. The mixture was centrifuged at 1,500 g for 10 min at 4˚C, and the organic phase was dried. DAG was then detected with a DAG assay kit (Cell Biolabs, San Diego, CA) following the manufacturer's instructions. Due to the lack of atrial muscle samples, the atrial DAG content was measured in five patients with SR and six patients with AF.

Statistical analyses
Values are presented as the mean ± standard deviation (SD) or the median (interquartile range [IQR]) as appropriate. We used Student's t-test for continuous variables that are normally distributed and the Mann-Whitney U-test for other continuous variables. The chi-square test or Fisher's exact test were used for categorical variables. We conducted a Pearson's correlation analysis to determine linear relationships between continuous variables. The statistical analyses were performed using GraphPad Prism ver. 8 (GraphPad Software, San Diego, CA), and significance was defined as p<0.05.

Patient characteristics
The characteristics of the SR and AF patients are summarized in Table 1. The median duration of AF was 12 years (range 1-16 years). Except for the use of medications, all parameters including cardiovascular risk factors were comparable between the two groups. The surgical procedures performed after the excision of atrial specimens included aortic valve replacement in 30 cases, total arch replacement in eight cases, coronary artery bypass grafting in 10 cases, aortic root replacement in six cases, and mitral valve repair in 12 cases. Multiple procedures were performed in some patients.

Transthoracic echocardiography
As shown in Table 2, the diameters of the left and the right atria were significantly increased in the AF group. The left ventricular size (i.e., left ventricular end-diastolic diameter [LVDd]) and the left ventricular systolic function (i.e., the LVEF) were comparable between the two groups. Regarding valvular heart disease, the grade of mitral regurgitation was higher in the AF group than in the SR group.

The serum levels of FFA and the atrial expression of genes related to fatty acid metabolism
The serum FFA levels were significantly higher in the AF group than the SR group (719 ± 107 vs. 416 ± 37 μmol/L, p = 0.001, Fig 1).
The expressions of genes related to fatty acid metabolism in the right atrial muscle are shown in Fig 2. The gene expression of FABP3, which facilitates fatty acid uptake into the cell and intracellular fatty acid transport, was significantly increased in the AF group (Fig 2B), but there was no significant difference between the two groups in the atrial expression of CD36, which facilitates fatty acid uptake across the plasma membrane, and CPT1B, which is located on the outer mitochondrial membrane for fatty acid transport into the mitochondria (Fig 2A-2C).
In SR patients, FABP3 gene was positively correlated with the right atrial diameter (Fig 3A) and the intra-atrial EMD (Fig 3B), indicating that increased atrial gene expression related to intracellular fatty acid transport was associated with structural and electrical atrial remodeling.
The atrial enzymatic activities of the mitochondrial TCA cycle and fatty acid β-oxidation Fig 4 illustrates the enzymatic activities related to the mitochondrial fatty acid β-oxidation and TCA cycle in the right atrial muscle. The CS activity was comparable between the SR and AF groups (Fig 4A), as was the β-HAD activity (Fig 4B).

The atrial DAG content
Despite the higher serum levels of FFA and the upregulated atrial gene expression of FABP3 in the AF group, there was no significant difference in the atrial content of DAG, a major toxic fatty acid metabolite, between the SR and AF groups (Fig 5).

The atrial expression of autophagy-related genes
The gene expression of LC3 in the right atrial muscle was significantly higher in the AF group than in the SR group (Fig 6D), but there was no significant difference in other autophagyrelated genes including ATG5, ULK1, and BCLN1 between the SR and AF groups (Fig 6A-6C).

The linear relationship between the atrial expression of FABP3 gene and autophagy-related genes
The gene expression of FABP3 was positively correlated with autophagy-related genes including LC3 gene in all patients (Fig 7). Similar correlations were observed between FABP3 gene

The atrial protein expression level of autophagy
In consistent with the increased gene expression of LC3 in the AF group, the protein expression level of LC3-II in the right atrial muscle was higher in the AF group than in the SR group (Fig 8).

Discussion
Our findings demonstrated that the atrial expression of a gene involved in fatty acid metabolism, FABP3, was upregulated in patients with chronic AF compared to the SR patients. In the SR patients, the increased atrial expression of FABP3 was positively correlated with the right atrial diameter and the intra-atrial EMD, which are parameters of structural and electrical remodeling of the atrium. In contrast, there was no increase in the atrial content of DAG despite the increased atrial expression of FABP3 and higher serum levels of FFA in patients with chronic AF. Intriguingly, the atrial expression of a gene related to autophagosome maturation, LC3, was increased in chronic AF patients, and autophagy-related genes including LC3 were positively correlated with the atrial expression of FABP3. In consistent with the increased atrial gene expression of LC3, the atrial protein expression level of LC3-II, an autophagosome marker, was increased in patients with chronic AF. To the best of our knowledge, this is the first study showing that the atrial expression of FABP3 gene is increased in association with autophagy-related genes in patients with chronic AF.

The dysregulated fatty acid metabolism in the atrium of chronic AF patients
Compared to SR patients, chronic AF patients had higher serum FFA levels. The Cardiovascular Health Study has reported that an increase in the plasma levels of FFA by 200 μmol/L presents an 11% higher risk of AF occurrence even after adjustment for confounding risk factors including age, sex, race, physical activity, body mass index, coronary heart disease, congestive heart failure, smoking, alcohol use, log-C-reactive protein, diabetes mellitus, and hypertension in older adults [6]. Accordingly, elevated FFA levels can be an independent risk factor of AF. The fatty acid-binding proteins (FABPs) reversibly bind to fatty acid and other lipophilic molecules. FABPs, which are located on the plasma membrane, facilitate fatty acid uptake into the cells, and intracellular FABPs transport fatty acid to other locations such as the nucleus and mitochondrion [20]. Among the 10 isoforms of FABPs distributed in various tissues in mammals, FABP3 is most predominantly expressed in the heart [21]. Here, we observed that the gene expression of FABP3 in the right atrial muscle was enhanced in chronic AF patients, which may indicate increased fatty acid uptake into the cells and increased intracellular fatty acid transport in the atrial muscle in chronic AF. In contrast, the gene expression of CPT-1B (which facilitates fatty acid transport across the outer mitochondrial membrane), and β-HAD activity (an enzymatic activity of mitochondrial fatty acid β-oxidation) in the atrium were comparable between our AF and SR groups.

The association of the atrial expression of FABP3 with structural and electrical atrial remodeling in SR patients
The results of our analyses revealed that the expression of FABP3 in the right atrial muscle was positively correlated with the right atrium diameter and the intra-atrial EMD in SR patients. The intra-atrial EMD is the time delay from the electrical activation to the actual motion of the atrial myocardium, and a delayed intra-atrial EMD indicates excitation-contraction uncoupling in the atrium. Prolonged intra-atrial EMD after cardioversion was reported to predict AF recurrence in patients with persistent AF, and histopathological changes characterized by myocardial fibrosis in the atrium appear to be a major determinant of the prolonged intraatrial EMD [22].
Boldt et al. revealed that the atrial expression of collagen type I is enhanced in patients with lone AF, indicating that the occurrence of AF can directly increase the expression of collagen type I and cause myocardial fibrosis in the atrial muscle [23]. Although we did not conduct histopathological evaluations, previous reports and our present findings raise the possibility that the dysregulation of atrial fatty acid metabolism is linked to structural and electrical atrial remodeling, which may contribute to a future onset or recurrence of AF. Here, we cannot conclude that there is a causal relationship between dysregulation of atrial fatty acid metabolism and atrial remodeling in chronic AF. Further studies with a larger sample size of chronic AF patients are needed to examine whether chronic or persistent AF might contribute to atrial remodeling via abnormal fatty acid metabolism in the atrium.

Altered autophagy in the atrium in chronic AF
In 2012, Garcia et al. first reported impaired cardiac autophagy characterized by reduced LC3 processing (i.e., a reduced protein expression of LC3BⅡ) with an accumulation of lipofuscin deposit -a potential trigger of AF -in the atrial myocardium in patients with post-operative AF [15]. A pair of studies have shown that in chronic AF patients, the cardiac autophagy characterized by an increased protein expression of LC3BⅡ is induced in the atrial myocardium in association with AMPK or endoplasmic reticulum (ER) stress [13,14]. Our present findings demonstrated that the atrial gene expression of LC3 was upregulated and atrial protein expression of LC3-II was increased in patients with chronic AF. Taking these results together, we speculate that altered cardiac autophagy in the atrium may be involved in the progression of AF.
In addition, animal studies have shown that onset of ventricular fibrillation increases autophagy-associated proteins in the ventricle accompanied by myocardial damage [24][25][26], suggesting that arrhythmia itself may directly cause altered cardiac autophagy.

Implications of the association between the atrial expression of FABP3 and autophagy
The intracellular lipid content is generally deteremined by an imbalance between the uptake and the utilization of fatty acid, and thus the increased atrial expression of FABP3 that we observed might contribute to the accumulation of lipids including DAG in the atrial myocardium in chronic AF. However, we did not detect an accumulation of DAG in the atrium in our patients with chronic AF as was reported in another study [27]. Autophagy has been shown to play a role in the regulation of fatty acid metabolism via the degradation of excessive intracellular lipids, termed "lipophagy" [12]. Our findings of an association between the atrial expression of FABP3 gene and autophagy-related genes in chronic AF patients may support our hypothesis that in chronic AF, autophagy at least in part contributes to the prevention of the accumulation of toxic fatty acid metabolites via a degradation of intracellular lipids. Further research is necessary to clarify the mechanistic roles of cardiac autophagy in the atrium in AF progression.

The difference in the atrial expression of FABP3 between post-operative AF and chronic AF
We observed that the atrial gene expression of FABP3 was reduced in patients with post-operative AF in our prior study [16]; this is inconsistent with our present results regarding chronic AF patients. One of the possible explanations is a difference in pathophysiology between postoperative AF and chronic AF. It was demonstrated that in patients with metabolic syndrome, impairment in the mitochondrial respiratory capacity in the atrial tissues predicts the occurrence of post-operative AF [28]. In contrast, the mitochondrial respiratory capacity in the atrium was reported to be increased in chronic AF patients [29]. Taken together, these findings indicate that impaired energy metabolism (including reduced fatty acid utilization) in the atrial muscle might be a primary pathogenesis of post-operative AF, but in chronic AF, excessive fatty acid uptake into the atrial cells seems to play a crucial role in AF progression.

Study limitations
Several limitations of this study should be addressed. First, most of the patients had valvular heart diseases, and the results of this study thus cannot be directly applied to patients with lone AF. Second, the number of AF patients was smaller than that of SR patients. Third, we cannot completely exclude the possibility of effects of medicine including diuretics and beta-blockers on fatty acid metabolism in chronic AF patients. Finally, we cannot conclude that there is a causal relationship between fatty acid metabolism and autophagy in the atrium.

Conclusions
Our study is the first to demonstrate that compared to patients with SR, the atrial expression of FABP3 gene was upregulated in association with autophagy-related genes in patients with chronic AF. We also observed that the atrial gene expression of FABP3 was related to structural and electrical remodeling in SR patients. Despite the increased atrial expression of FABP3 with higher serum levels of FFA, atrial contents of DAG were not increased in patients with chronic AF. These findings provide new insights into the pathophysiology of chronic AF, and they suggest that dysregulated cardiac fatty acid metabolism might contribute to the progression of AF and induction of autophagy might have a cardioprotective effect against cardiac lipotoxicity in chronic AF.
Supporting information S1