Upregulation of sestrins protect atriums against oxidative damage and fibrosis in human and experimental atrial fibrillation

Atrial Fibrillation (AF) is common in the elderly and Sestrins (Sesns) have been suggested to prevent age-related pathologies. The aim of this study was to investigate the effects of Sesns in AF. Clinical data were collected and a small sample of atrial appendage and atrium was obtained from patients undergoing valve repairment. The expression of Sesn1, Sesn2, and Sesn3 was significantly higher in patients with permanent atrial fibrillation (PmAF) than that in sinus rhythm (SR), and further greater in the left atrium than the right in PmAF patients. Superoxide anion and malondialdehyde were enhanced and positively correlated to the protein expression of Sesn1/2/3. Reactive oxygen species (ROS) production and Ca2+ overload were significantly decreased and cell survival was enhanced by overexpression of Sesns 1/2/3 in cultured HL-1 cells. Conversely, knockdown of Sesn1/2/3 resulted in significantly increased ROS and Ca2+ overload. In addition, the overexpression of Sesn1/2 significantly reduced the proliferation of fibroblasts, as well as decreased the protein expression of collagen and fibronectin1 in angiotensin II-stimulated cardiac fibroblasts. Our study demonstrated for the first time that Sesns expression is significantly up-regulated in AF, which therefore may protect hearts against oxidative damage and atrial fibrosis.

. No significant difference was identified between the groups in terms of age, sex, ratio of rheumatic heart disease, left ventricular end-diastolic diameter, and ejection fraction. However, the left atrium dimension in PmAF group was larger than that in SR group (P < 0.05), and the use of warfarin and digoxin was more common in PmAF group than in SR group. All other medications were similar between the two groups. The mean AF duration was 4.6 ± 3.4 years in the PmAF group.

Upregulation of Sesns in PmAF Patients and Cells.
The location of Sesns proteins in the atrium was detected with immunofluorescence. Red fluorescence, indicating the expression of Sesns proteins, was extensively seen in the cytoplasm and nucleus of cells in PmAF group. Sesn1, Sesn2, and Sesn3 were expressed in a larger quantity in PmAF than in SR (Fig. 1A). As shown in Fig. 1A, the expression of Sesn1, Sesn2, and Sesn3 also increased in the cytoplasm and nucleus of paced HL-1 cells and AngII-stimulated cardiac fibroblasts. Furthermore, Sesns expression by western blots confirmed that PmAF had significantly elevated expression of Sesn1 (P < 0.01), Sesn2 (P < 0.01), and Sesn3 (P < 0.01) (Fig. 1B). In the PmAF group, the expression of Sesn1 (P < 0.01), Sesn2 (P < 0.01), and Sesn3 (P < 0.05) in the LA was significantly higher than in the RA. Additionally, the basal expression level of Sesn 2 in SR patients is much lower than Sesn 1 (0.39 ± 0.13 vs. 0.60 ± 0.15, P < 0.01) and Sesn 3 (0.39 ± 0.13 vs. 0.63 ± 0.24, P < 0.01). However, Sesn2 sharply increased by more than four times in AF patients, whereas Sesn 1 and Sesn 3 did not change so much.

Sesns Inhibit Generation of ROS in Paced HL-1 Cells. Previous reports indicated that ROS played
an important role in the development of AF 6 . To investigate the effects of Sesns on ROS in paced HL-1 cells, we expressed Sesns cDNA or Sesns siRNA in paced HL-1 cells (Fig. 4A) Sesns were documented by western blots (Fig. 4B). The amount of ROS was analyzed by flow cytometry (Fig. 5A).

Sesns Suppress AngII-induced Proliferation of Cardiac Fibroblasts and Collagen Synthesis.
To study the effect of Sesns on cardiac fibrosis, we used CCK-8 method to detect the proliferation of cardiac fibroblasts (CFs). As Fig. 8A shows that the overexpression of Sesn1 and Sesn2 in cardiac fibroblasts significantly reduced the proliferation of fibroblasts (P < 0.05). Next, we analyzed mRNA expression levels for COL I/III and FN1. As shown in Fig. 8B-D, Ang II induced increases of COL I/III and FN1 mRNA expression (P < 0.01). In contrast, increased the expression of Sesn1 and Sesn2 significantly attenuated the Ang II-induced COL I and FN1 mRNA expression (P < 0.05). To validate the mRNA findings, we identified the effect of Sesn1 and Sesn2 on protein expression of COL I/III and FN1. As shown in Fig. 8E-H, COL I/III and FN1 were severely increased induced by AngII (P < 0.05) and the effect were clearly inhibited by Sesn1/2-overexpressing (P < 0.05). These results demonstrated Sesn1 and Sesn2 negatively regulated AngII-mediated CFs proliferation.  Comparison with Previous Studies. Sesns are cytoplasmic stress proteins that accumulate in cells exposed to stress, hypoxia, and DNA damage 5,10 . Here, we identified significantly higher levels of Sesns are expressed in patients with PmAF than SR, which may be attributed to an increased oxidative stress in PmAF patients (Fig. 3).
The same increase was also seen in paced HL-1 cells (Fig. 5). Studies have demonstrated that AF itself induces substantial ROS and oxidative stress in fibrillating atrial tissue 18,19 , thus may induce Sesns upregulation in such condition. These results are consistent with a previous study which showed that any condition which leads to ROS accumulation may induce Sesns expression 10,20 . Interestingly, the expression of Sesns was higher in LA than RA in PmAF, which may partly due to AF cause greater changes in O 2 · − production in the LA or LAA than in the RA or RAA 18 . Also, there was a trend in the more increase of oxidative stress in LA than RA in PmAF suggested by of MDA in our study, although no statistical difference were observed (P = 0.07). An animal study of atrial tissue in AF shows a gradient of oxidative stress with greater oxidative stress on the left 18 , however, it is not the case in human 21,22 . Thus, besides oxidative stress, other mechanisms such as pressure, endocrine function and Ca 2+ homeostasis 19,23 of LA, by which the LA is distinguished from the RA, should be considered and require further investigation.
Noticeably, our study shows that there are significant basal expression level differences in Sesns. The expression of Sesn 2 in SR patients is much lower than Sesn 1 and Sesn 3, but increase more sharply than Sesn 1 and Sesn 3 in AF. This would suggest that Sesn 2 is the main subtype induced in oxidative stress in AF. Similarly, the same trend was noticed in the paced HL-1 cells in this study. Evidences were also shown in other studies that Sesn 2 is the main stress proteins induced in oxidative damage and provide significant protection in cardiovascular diseases 24,25 . Therefore, different potencies of Sesns are induced in AF, which may suggest Sesn 2 play more important role in the pathological process of AF, whereas Sesn 1 and Sesn 3 work both in physiological and pathological conditions.
Oxidative stress drives electrical and structural remodeling, which contributes to the development and progression of AF. Anti-inflammatory and antioxidant agents prevent atrial electrical remodeling in animal models of AF and reduce the incidence of postoperative AF in humans [6][7][8] . Heat shock proteins (HSPs), cytoplasmic stress proteins, are increased in AF and protect against atrial tachycardia-induced remodeling in cellular and animal models [26][27][28] . Treatment with geranylgeranylacetone (GGA), an HSP inducer, increases HSP expression, suppresses refractoriness, and prevents AF in dogs subjected to atrial tachypacing 26 . However, data are conflicting. The previous study suggested that GGA alone, without ischaemia, does not alter electrical conduction or AF duration 29 . Although the expression of Hsp 27 is significantly increased in paroxysmal AF compared with SR and persistent AF, no changes are observed in the expression of most HSPs, such as Hsp40, Hsc70, Hsp70, and Hsp90 28 . Like HSP, Sesns are also cytoplasmic stress proteins, which exhibit oxidoreductase activity and may function as antioxidants 12 . However, Sesns are widely expressed even in the absence of exogenous stress. In drosophila, expression of dSesn is increased upon maturation 10,20 . In humans, Sesn1, Sesn2, and Sesn3 are all expressed in atria without stress (shown in Fig. 1). Besides physiological conditions, we show for the first time that all kinds of Sesns are upregulated in human AF or paced HL-1 cells. These results have been consistent with previous studies, which show Sesns are cytoplasmic stress proteins accumulating in cells exposed to oxidative stress 5,12 . Here Sesns also display protection against oxidative damage and fibrosis in experimental AF, conforming to studies that Sesns has potential to limit liver damage and fibrosis in chronic ER stress and protect renal tubules again stress during acute kidney injury 30,31 . Sesns as a Protective Factor. The physiological functions of Sesns remain poorly defined. To our knowledge, this is the first study to report that all Sesns upregulate in human AF and protect paced cardiac myocytes against oxidative stress. Here we show that Sesns are positively correlated with oxidative stress in humans and paced HL-1 cells, which suggested oxidative stress may be the cause of Sesns upregulation in such condition. As a results, Sesns compensatory upregulate to antagonize injury induced by oxidative stress. However, this kind of compensatory mechanism is not enough to prevent the oxidative stress in AF, considering the fact that ROS accumulates in human AF and paced atrial cells in spite of theirs upregulation. Furthermore, the overexpression of Sesns decreases the amount of ROS, enhances cell survival and inhibits fibrosis in paced HL-1 cells and cardiac fibroblasts, suggesting that exogenetic Sesns may be protective and useful in AF. Some other studies also suggest that Sesns may play an important role in the regulation of cardiac pathophysiology and provide protection 24,32,33 . In Drosophila, heart-specific depletion of dSesn caused heart dilation, arrhythmias, and malfunction 10 . Sesn2 proteins were found to accumulate in the heart during ischemic conditions protecting the heart against ischemia and reperfusion injury 24 . Sesn1 inhibits angiotensin II-induced fibroblast proliferation and collagen production in cardiac cells 34 . Thus, the increase of Sesns in AF patients supports the notion that Sesns induced by ROS are an endogenous compensatory mechanism to prevent AF developing.
The mechanisms by which various types of Sesns provide cardioprotection in AF are not fully understood. The protective function of Sesns in AF is probably mediated by two pathways. In one pathway, the antioxidant activity of Sesns in vitro and in vivo alleviates the oxidative damage and protects in AF [10][11][12][13][14] . In the present study, Sesns overexpression decreased ROS production by 20-30% in paced HL-1 cells. In the second pathway, Sesns inhibit fibrosis, which is an important structural contributor to formation of AF substrate. We show that Sesn1 and Sesn2 inhibit cell proliferation and fibrosis in cardiac fibroblasts stimulated by AngII, which is in accordance with the findings of another study 34 . Target of rapamycin (TOR) and TGF-beta signals are two possible mechanisms associated in the anti-fibrosis function of Sesns [35][36][37] . Hyperactivity of TOR is related to activation of cardiac fibroblasts in stress 35 , and the inhibition of TOR can ameliorate the chronic pressure-induced left ventricular hypertrophy and cardiac fibrosis 36 . As an inhibitor of TOR signaling 10,11 , Sesns thereby presumably attenuate fibrosis in AF.
However, the protective function of Sesns may be independent of oxidative stress, since our data show the survival of paced HL-1 cells did not grow worse in Sesns knockdown with siRNA despite the fact that apparent effects of introducing Sesns on oxidative stress and survival. The previous study suggested that ROS can also be protective as signal preconditioning protection and induce stress responses that lead to survival 38 . Other mechanisms for protection are implied in the present study and should be further investigated in the future.

Potential Significance of Sesns Upregulation in Atrial Fibrillation. Sesns are induced in patients
with PmAF, which may serve as an endogenous protective mechanism to prevent AF. Oxidative stress has been implicated in the structural and electrophysiological remodeling of AF and antioxidants prevent atrial remodeling and AF incidence 7,8 . Thus, as antioxidants, Sesns probably attenuate the AF substrate and inhibit the remodeling of atria in AF [11][12][13] . Here we show the overexpression of Sesns decreases the amount of ROS, enhances cell survival in paced HL-1 cells, as well as inhibits proliferation and collagen production in cardiac fibroblasts induced by AngII. Sesns have been demonstrated to prevent age-related pathologies, such as cancer and type II diabetes 10,32,39 . AF is also an age-related disease and, therefore, may be prevented by Sesns. Thus, just like the effects of B-type natriuretic peptide in heart failure 40 , our results provide novel evidences that Sesns are induced and provide protection in AF, and may serve as a possible therapeutic target in the clinical management of AF.

Potential Limitations.
First, we demonstrated the protection of Sesns in HL-1 myocytes and cardiac fibroblasts, but not directly in AF patients. Although HL-1 cells beat spontaneously, their electrophysiological, functional, and metabolic properties are modified during culture. Second, most of the patients included in the present study had rheumatic heart disease. Although the groups were well matched for rheumatic heart disease, rheumatic fever influences oxidative stress and therefore may affect the Sesns level. Third, we did not establish a cause and effect relationship between ROS and Sesns, as well as clarify if the actions of Sesns are direct or indirect on oxidative stress. Fourth, transfection with empty vector or control RNA causes a significant increase in oxidative stress, which would influence the results of the study. Another means to introduce nucleotides which could not affect oxidative stress should be used. Fifth, the contradiction between ROS and survival indicating extra mechanisms of cardioprotection by Sesns, which is independent of oxidative stress, are implied in the present study and should be further investigated in the future.

Summary
Sesns are induced as an endogenous protective factor in AF, which protects the atria against oxidative damage and fibrosis. However, the compensatory increase of Sesns in AF is not enough to prevent this kind of disease, suggesting extra exogenetic Sesns are necessary and helpful for treatment of AF. This identification provides insight into the molecular basis of AF and suggests therapeutic targets for the common rhythm disturbances.

Materials and Methods
Patients. Tissue samples from the right atrium (RA) and left atrium (LA) appendage were obtained from 42 patients undergoing cardiac surgery for valvular replacement at the First Affiliated Hospital, and the Second Affiliated Hospital, Harbin Medical University, between 2011 and 2013. All patients had valvular heart disease and that a large number of patients had rheumatic heart disease. Patients were divided into two groups: permanent AF group (PmAF, n = 19) in which patients has an AF history of > 6 months and electrocardiographically documented AF at the time of surgery; and sinus rhythm group (SR, n = 23) in which patients had no documented AF and no AF history. All patients were excluded liver and renal function abnormality, neoplastic disease and inflammation before surgery. The size and function of hearts were measured by echocardiography (Philips iE33, Holland). One part of the atrium and appendage tissue was fixed in 4% formalin for histopathological examination, and the remaining tissue was frozen in liquid nitrogen and stored at − 80 °C for immunofluorescence and western blot and other analyses. The study was approved by the Ethics Committee of the First Clinical College of Harbin Medical University, and all patients gave informed consent. All the methods were carried out in accordance with the approved guidelines. Cardiomyocytes were isolated from adherent fibroblasts after incubation for 1.5 h. 0.25% trypson-EDTA and Phenol Red (Gibco, The Netherlands) was used in fibroblasts passage. The fibroblasts could be harvested while 2 times of the passages. All cells were incubated with above culture at 37 °C, 5% CO 2 . 1 μ M Ang II was used to stimulate cardiac fibroblasts. The proliferation of the cells was measured by CCK-8 assay.
Western Blot Analysis. The protein concentration in the supernatant was determined by BCA assay.
Histopathological Examination. Masson's trichrome staining for interstitial collagen deposition were analysed. Tissue specimens were fixed in 4% paraformaldehyde and embedded in paraffin. Serial sections about 5 μ m in thickness were cut and stained with Masson trichrome. Sections were photographed with an Olympus HPISA-1000 camera and the extent of fibrosis was quantified with Image-Pro Plus 6.0 (Media Cybernetics, Bethesda, Md). Sections of the heart were analysed for bright blue staining (collagen) and red staining (cardiacmyocyte). Collagen volume fraction (CVF) was defined as the sum of all stained interstitial collagen tissue areas divided by the whole tissue area.

Assessment of Oxidative Stress Markers.
The amount of superoxide anion (O 2 · − ) in the atrium was measured as described previously 18 . Electron spin resonance spectroscopy was used to examine intracellular O 2 · − production with the cell-permeable spin probe 1-hydroxy-3-methoxycarbonyl-2, 2, 5, 5-tetramethylpyrrolidine hydrochloride (CMH; Alexis Corp). Briefly, freshly isolated atrial tissues was incubated with deferoxamine-chelated Krebs-HEPES solution containing CMH (0.5 mmol/L), deferoxamine (25 μ mol/L), and DETC (5 μ mol/L) for 90 minutes at 37 °C. Then, samples were transferred into 1-mL syringes filled with Krebs-HEPES solution and frozen in liquid nitrogen. Samples were scanned with a Bruker EMX spectrometer. Analyses of the spectra peak height were used to quantify the amount of O 2 · − produced by the tissue. The amount of malondialdehyde (MDA) produced from the atrium was examined by thiobarbituric acid (TBA) assay with an MDA Detection Kit (Nanjing Jiancheng Bioengineering Institute, China). In brief, stored atrium samples were weighed and 5% tissue homogenate was obtained on ice in isotonic Na chloride. 0.1 ml tissue homogenate was processed according to the manufacturer's instructions and detected spectrophotometrically at 532 nm by an ultraviolet spectrophotometer (GeneQuant pro, Amersham Biosciences, England). Statistical Analysis. Quantitative data were expressed as mean ± SEM. Differences among quantitative data were analyzed by ANOVA. Multiple comparisons were made using SNK-q test. An unpaired student's t-test was used for comparisons between two groups. Categorical variables were analyzed by Chi-square test with continuity correction when 1 ≤ T < 5. Pearson's correlation coefficient (r) was used to measure the strength of association between quantitative parameters. A 2-tailed P < 0.05 was considered statistically significant.