Ablation of sarcolipin results in atrial remodeling

Abstract

Sarcolipin (SLN) is a key regulator of sarco(endo)plasmic reticulum (SR) Ca²⁺-ATPase (SERCA), and its expression is altered in diseased atrial myocardium. To determine the precise role of SLN in atrial Ca²⁺ homeostasis, we developed a SLN knockout (sln−/−) mouse model and demonstrated that ablation of SLN enhances atrial SERCA pump activity. The present study is designed to determine the long-term effects of enhanced SERCA activity on atrial remodeling in the sln−/− mice. Calcium transient measurements show an increase in atrial SR Ca²⁺ load and twitch Ca²⁺ transients. Patch-clamping experiments demonstrate activation of the forward mode of sodium/calcium exchanger, increased L-type Ca²⁺ channel activity, and prolongation of action potential duration at 90% repolarization in the atrial myocytes of sln−/− mice. Spontaneous Ca²⁺ waves, delayed afterdepolarization, and triggered activities are frequent in the atrial myocytes of sln−/− mice. Furthermore, loss of SLN in atria is associated with increased interstitial fibrosis and altered expression of genes encoding collagen and other extracellular matrix proteins. Our results also show that the sln−/− mice are susceptible to atrial arrhythmias upon aging. Together, these findings indicate that ablation of SLN results in increased SERCA activity and SR Ca²⁺ load, which, in turn, could cause abnormal intracellular Ca²⁺ handling and atrial remodeling.

sarcolipin (sln), a 31-amino acid sarco(endo)plasmic reticulum (SR) membrane protein, is predominantly expressed in atria (2, 49). Studies on transgenic mouse models overexpressing SLN in the heart suggest that SLN is a key regulator of cardiac SR Ca²⁺-ATPase (SERCA) and mediator of β-adrenergic responses (1, 3, 20). Altered expression of SLN has been shown in atria of a diseased myocardium. The mRNA and protein levels of SLN are significantly decreased in the atrial myocardium of patients with chronic atrial fibrillation (AF) (39, 44). In atria of pacing-induced heart failure dogs, SLN protein expression is significantly increased, whereas, in atria of hearts prone to myocardial ischemia, SLN protein level is decreased (2). The expression levels of SERCA2a and phospholamban (PLN) are not altered in atria of human AF and in the above-mentioned animal models. Together, these studies suggest that SLN levels could affect the SERCA activity and intracellular Ca²⁺ (Ca²⁺_i) handling in diseased atrial myocardium.

AF, the most common sustained arrhythmia, is associated with increased risk of stroke, symptoms of heart failure, and genetic disorders (29, 30, 36, 45). Several studies have suggested that abnormal Ca²⁺ handling is one of the major causes of atrial remodeling and AF (11, 14, 15, 52). The atrial tachycardia remodeling, which promotes AF, causes atrial contractile dysfunction, mainly via Ca²⁺-handling abnormalities (52). The reduced Ca²⁺ transients with slowed Ca²⁺ decay and reduced diastolic Ca²⁺ levels are suggested to cause atrial dilatation and AF (14, 42). Studies from a congestive heart failure dog model, which is susceptible to sustained AF, suggest that increased SR Ca²⁺ load may contribute to the generation of delayed afterdepolarizations (DAD) and triggered activities in the atrial myocytes (54). In mice expressing R176Q mutant ryanodine receptor 2 (RyR2), the Ca²⁺/calmodulin-dependent protein kinase II (CaMKII)-mediated SR Ca²⁺ leak promotes AF (9). Thus alterations in the SR Ca²⁺ load and/or Ca²⁺ release mechanisms can contribute to atrial remodeling and contractile dysfunction.

Given that SLN is a key regulator of the cardiac SERCA pump (1, 3, 20), the selective downregulation of SLN in AF (39, 44) could affect the Ca²⁺_i handling. In the present study, using a SLN knockout (sln−/−) mouse model (4), we determined if complete loss of SLN affects the Ca²⁺_i handling and causes atrial remodeling.

MATERIALS AND METHODS

All animal procedures were performed with the approval of Institute Animal Care and Use Committee of the University of Medicine and Dentistry New Jersey-Newark campus, in accordance with the provision of the animal welfare act, the Public Health Service policy on Humane Care and Use of Laboratory Animals. Five to six- and twelve-mo-old sln−/− and C57/BL6 wild-type (WT) mice were used in this study. The Ca²⁺ measurements, electrophysiological, histopathological, and gene expression studies were carried out on 5- to 6-mo-old sln−/− and WT mice hearts.

Electrophysiological recording.

Generation of sln−/− mice was described previously (4). Atrial or ventricular myocytes were enzymatically isolated from WT and sln−/− mice hearts, as described earlier (38). All patch-clamp experiments were carried out at 34–36°C, using the ruptured (21) patch whole cell recording technique. For action potential (AP) recordings, patch pipettes (resistance 1–3 MΩ) were filled with internal solution containing (in mmol/l): 110 potassium-aspartate, 30 KCl, 5 NaCl, 10 HEPES, 0.1 EGTA, 5 MgATP, 5 Na₂-phosphocreatine, 0.05 cAMP, pH 7.2 adjusted with KOH. The cells were superfused with Tyrode's solution containing (in mmol/l): 136 NaCl, 4.0 KCl, 0.33 Na₂PO₄, 1.8 CaCl₂, 1 MgCl₂, 10 glucose, and 10 HEPES, pH 7.4 adjusted with NaOH. APs were recorded under current clamp mode at pacing cycling length of 1 s. Whole cell currents were recorded under voltage-clamp configuration. For L-type Ca²⁺ current (I_Ca,L) recordings, patch pipettes were filled with internal solution containing (in mmol/l): 110 Cs-aspartate, 30 CsCl, 5 NaCl, 10 HEPES, 0.1 EGTA, 5 MgATP, 5 Na₂-phosphocreatine, 0.05 cAMP, pH 7.2 adjusted with CsOH. The cells were perfused with a modified Tyrode's solution containing (in mmol/l): 120 NaCl, 20 tetraethylammonium-Cl, 0.33 Na₂PO₄, 1.0 CaCl₂, 1 MgCl₂, 10 HEPES, pH 7.4 adjusted with NaOH. The myocytes were stimulated at a pacing cycling length of 6 s, with a double-pulse protocol. Following a 100-ms prepulse to −40 mV from the holding potential of −80 mV (to inactivate Na⁺ current and T-type Ca²⁺ current), I_Ca,L was elicited by a subsequent test depolarization step to 0 mV for 300 ms. For total outward K⁺ current (I_K) recording, the pipette and superfusion solutions were the same as those for AP recording. Tetrodotoxin (10 μmol/l) and CdCl₂ (0.5 mmol/l) were added into the Tyrode's solution to inhibit Na current and I_Ca,L. I_K were elicited from a holding potential of −80 mV by a series of 400-ms test pulses from −40 to +40 mV in 10-mV increments. The inward Na⁺/Ca²⁺ exchange (NCX) current (I_NCX) was recorded by a rapid application of caffeine (10 mmol/l) for ∼2-s duration. The application of caffeine was always preceded by a set of ten 200-ms conditioning depolarizations from −40 to 0 mV to ensure a consistent degree of SR Ca²⁺ loading. A holding potential of −40 mV was maintained during caffeine application. Membrane current and voltage were measured with an Axopatch 200B patch-clamp amplifier controlled by a personal computer using a Digidata 1200 acquisition board, driven by pCLAMP 10 software (Axon Instruments).

SR Ca²⁺ load and Ca²⁺ transient measurements in isolated myocytes.

SR Ca²⁺ load and Ca²⁺ transient measurements using fluo 4-AM were carried out at 34–36°C, as described earlier (38). Myocytes were field stimulated at 0.5 Hz to maintain consistent SR Ca²⁺ load. The height of the 10 mmol/l caffeine-induced Ca²⁺ transient was used as a measure of total SR Ca²⁺ content. Fractional SR Ca²⁺ release was calculated by dividing the height of the last twitch transient by the height of the caffeine transient.

ECG recordings.

Standard lead II ECG was recorded from anesthetized mice with a differential amplifier (Warner DP301). The ECG signals were collected for a total duration of 30 min for each mouse, and all analyses were performed blinded to genotype.

Histopathology.

Myocardial tissues were sectioned at 6 μm and stained with hematoxylin and eosin, Masson's trichrome or Picro-Sirius Red. Fibrosis was identified as accumulation of collagen between the myofibrils. Quantitation of fibrosis in Sirius red stained images was carried out using Image-pro software.

Western blot analyses.

Total protein extracts from the WT and sln−/− mice atria were used for Western blot analyses using protein-specific antibodies, as described earlier (4). Signals detected by Super Signal WestDura substrate (Thermo Scientific) were quantitated by densitometry and then normalized to calsequestrin levels. To determine the basal phosphorylation of PLN and RyR, the blots were immunoprobed with phospho-serine (S) 16 PLN (Cyclacel, Dundee, U.), phospho S2808 RyR (Abcam, cat. no. ab59225) or S2814(26) antibody (a kind gift from Dr. Xander H. T. Wehrens, Texas). After the signals were detected, the membranes were stripped and reprobed with regular PLN or RyR antibody.

Microarray analysis.

Gene expression profiling was carried out using Affymetrix GeneChip Mouse Gene 1.0 ST Array. Microarray data were submitted to Gene Expression Omnibus with the accession number GSE23403. Microarray data normalized by the robust multichip analysis method were subjected to significant analysis of microarrays. Significant genes were selected using fold change >1.2 and false discovery rate < 5%. Gene expression changes were analyzed by the hierarchical clustering method using the MultiExperiment Viewer program (http://www.tm4.org/mev.html). Regulated genes were analyzed by their Gene Ontology (GO) annotations. A hypergeometric test was used to assess the significance of association of a particular GO term with a set of up- or downregulated genes. To avoid redundant GO terms, we removed those with >70% of genes overlapped with a more significant GO term.

Quantitative real-time PCR.

First-strand cDNA was synthesized using total RNA prepared from WT and sln−/− mice atria with the high-capacity cDNA reverse transcription kit (Applied Biosystems). Quantitative real-time (qRT)-PCR was performed using Applied Biosystems 7500 RT-PCR system with Power SYBR Green Mastermix following the manufacturer's instructions. The data were analyzed by 7500 software version 2.0.2. Normalization of the signal was carried out based on the expression of cyclophilin D. The primers for qRT-PCR used were as follows: bone morphogenetic protein (BMP) 2 (forward 5′ GTTTGGCCTGAAGCAGAGAC 3′; reverse 5′ CGTCACTGGGGACAGAACTT 3′), BMP5 (forward 5′ TTCAAGGCAAGCGAGGTACT 3′; reverse 5′ GAAAAGAACATTCCCCGTCA 3′), cystatin 9 (forward 5′ CCGTGAACACATTCAACCAG 3′; reverse 5′ GGGTGACTCCCATCAGAGAA 3′), cathepsin K (forward 5′ AGGGAAGCAAGCACTGGATA 3′; reverse 5′ AGCACCAACGAGAGGAGAAA 3′), connexin 40 (forward 5′ GCCACCAAGTTCAGTGGAAT 3′; reverse 5′ TGCTATCACCACACCTCTCG 3′), matrix metalloproteinase 3 (MMP3) (forward 5′ CAGACTTGTCCCGTTTCCAT 3′; reverse 5′ GGAAGAGATGGCCAAAATGA 3′), S100A1 (forward 5′ GCCCTTCTGTCGAGAATCTG 3′; reverse 5′ TTGAAGTCCACTTCCCCATC 3′), and tissue inhibitor of MMP4 (forward 5′ ACCTCCGGAAGGAGTACGTT 3′; reverse 5′ CTGAGGGCAAGGCAGATAAG 3′).

Statistical analysis.

Data are presented as means ± SE. Statistical significance was assessed using paired, unpaired Student's t-tests, ANOVA analysis, or Fisher's exact test, with P < 0.05 considered significant.

RESULTS

Loss of SLN results in increased SR Ca²⁺ load in the atrial myocytes.

We first determined the effect of enhanced SERCA pump activity (4) on SR Ca²⁺ load and Ca²⁺ transients in the atrial myocytes of sln−/− mice. As shown in Fig. 1, A and B, the twitch Ca²⁺ transient was significantly increased in the sln−/− atrial myocytes (ratio of fluorescence over the basal diastolic fluorescence: WT = 2.2 ± 0.1, n = 13 vs. sln−/− = 3.1 ± 0.2, n = 11, P < 0.01). The fractional Ca²⁺ release was the same in WT and sln−/− atrial myocytes (Fig. 1D). The SR Ca²⁺ content as measured by caffeine-induced Ca²⁺ transient was significantly higher in the sln−/− atrial myocytes (Fig. 1, A and C; ratio of fluorescence over the basal diastolic fluorescence: WT = 4.7 ± 0.1, n = 10 vs. sln−/− = 6.1 ± 0.3, n = 11, P < 0.01). The Ca²⁺ transient (4) and SR Ca²⁺ load (data not shown) in the ventricular myocytes of sln−/− mice were not significantly different from that of WT mice.

Fig. 1.

Ca²⁺ transient amplitude and sarco(endo)plasmic reticulum (SR) Ca²⁺ content in sarcolipin knockout (sln−/−) atrial myocytes. A: twitch Ca²⁺ transients elicited by field stimulation at 0.5 Hz, and Ca²⁺ content measured as the height of the caffeine (Caff; 10 mmol/l)-induced Ca²⁺ transient in a representative atrial myocyte from wild-type (WT) and sln−/− mice. Ca²⁺ fluorescence intensity was recorded as the ratio F/F₀ of the fluorescence (F) over the basal diastolic fluorescence (F₀). B: summarized data for twitch Ca²⁺ transient in atrial myocytes. C: summarized data for SR Ca²⁺ contents in WT and sln−/− atrial myocytes. D: summarized data for fractional SR Ca²⁺ release, calculated by dividing the height of the last twitch transient by the height of the Caff transient. Values are means ± SE; no. of cells analyzed is indicated above each bar. **P < 0.05 compared with WT control myocytes. ns, Not significant.

Atrial myocytes from the sln−/− mice show prolonged AP durations.

To determine whether the increased Ca²⁺ transients and SR Ca²⁺ load affect the cellular electrical properties, we measured the AP in isolated atrial myocytes. Results in Fig. 2A indicate that there is a substantial difference in the morphology of APs between WT and sln−/− atrial myocytes. An apparent late plateau was observed after the rapid repolarization phase in the sln−/− atrial myocytes, causing prolongation of the AP duration (APD) at 90% repolarization (APD₉₀). There was no change in the APD at 50% repolarization (summarized in Fig. 2B).

Fig. 2.

Alternations of action potential (AP) morphology, occurrences of delayed afterdepolarizations (DADs), triggered activities, and spontaneous Ca²⁺ release in the atrial myocytes of sln−/− mice. A: representative AP recordings from the atrial myocytes of WT and sln−/− mice. Note the apparent late plateau phase in the AP recorded from sln−/− atrial myocyte. B: summary of AP duration (APD) values. The APD at 90% repolarization (APD₉₀) is prolonged in atrial myocytes of sln−/− mice. C: comparison of the effects of Na⁺-free, Li⁺-containing Tyrode's solution on the AP properties, especially the late plateau in sln−/− and WT mice atrial myocytes. D: DAD (*) and triggered activities (↓) appear in atrial myocytes from sln−/− mice. E: representative traces of intracellular Ca²⁺ concentration fluorescence from WT and sln−/− atrial myocyte showing different incidences of spontaneous Ca²⁺ release (4 cells for each). F: percentage of oscillating atrial cells with spontaneous Ca²⁺ release in the sln−/− and WT mice. Values are means ± SE; no. of cells analyzed is indicated above each bar. **Significantly different from the WT atrial myocytes, P < 0.05 by Fisher's exact test.

To investigate if the I_NCX contributes to the late plateau and prolongation of late APD, we inhibited the forward mode of I_NCX by replacing Na⁺ with Li⁺ in the perfusate Tyrode's solution. The Li⁺ solution suppresses the late plateau and prevents the prolongation of APD₉₀ (Fig. 2C). Together, these results suggest that the activation of forward mode of I_NCX can contribute to the late plateau and prolongation of APD₉₀ in the atrial myocytes of sln−/− mice.

Spontaneous Ca²⁺ waves and triggered activities in the atrial myocytes of sln−/− mice.

Afterdepolarizations and triggered activities are the major causes of cardiac arrhythmias, such as atrial and ventricular fibrillation (9, 41, 51, 53). Our results show that DADs and triggered activities were more frequent in the atrial myocytes of sln−/− mice (Fig. 2D). The SR Ca²⁺ overload and spontaneous Ca²⁺ release can contribute to the transient inward currents, which cause DADs. Therefore, we calculated the percentage of spontaneously beating cells by monitoring spontaneous Ca²⁺ release (without stimulation). Results (Fig. 2, E and F) demonstrate that the number of atrial myocytes showing spontaneous Ca²⁺ waves/oscillations was abundant in the sln−/− mice (P < 0.05 by Fisher's exact test).

Increased I_Ca,L density in the atrial myocytes of sln−/− mice.

The I_Ca,L (8, 10, 12, 47) and I_K (8, 12, 13, 16, 48) are the two major current components determining the APD. We therefore, evaluated the current density level of I_Ca,L and total I_K (I_K,total). The currents were recorded from 10–11 cells isolated from 4 mice in each group (WT vs. sln−/−). As shown in Fig. 3, A–C, the I_Ca,L density was increased in the atrial myocytes of sln−/− mice (pA/pF: WT = 2.8 ± 0.3 vs. sln−/− = 4.8 ± 0.3; P < 0.05). However, there was no alternation in current kinetics (e.g., current-voltage curve or inactivation). The outward I_K,total densities were also not altered in the sln−/− atrial myocytes (Fig. 3, D and E). It should be noted that the I_K,total showed less inactivation component, since the pipette solution had a low concentration of Ca²⁺ buffer. This phenomenon was consistent with a previous report on transient outward current modulation by CaMKII in human atrial cells (43).

Fig. 3.

L-type Ca²⁺ current (I_Ca,L), Na⁺/Ca²⁺ exchange current (I_NCX), and total outward K currents in the sln−/− atrial myocytes. A: representative current-voltage curve. B: I_Ca,L traces elicited at testing potential of 0 mV. C: summary of I_Ca,L densities. D: representative traces elicited with the voltage-clamp protocol shown. E: peak K currents at test potential of +40 mV in WT and sln−/− atrial myocytes. I_K,total, total K current. F: representative tracings of Ca²⁺ transients (top) and I_NCX (bottom) after rapid application of Caff in atrial myocytes isolated from sln−/− or WT control hearts. G: average values of peak Ca²⁺ transient (F/F₀, top) and peak I_NCX densities (bottom). Values are means ± SE; no. of cells analyzed is indicated above each bar. **P < 0.05; ***P < 0.01, sln−/− vs WT.

Increased I_NCX density in the atrial myocytes of sln−/− mice.

To obtain direct evidence for potential upregulation of I_NCX in the atrial myocytes of sln−/− mice, we measured I_NCX and SR Ca²⁺ release in response to caffeine (10 mmol/l) application. Results show that the amplitudes of both Ca²⁺ transients and peak I_NCX density were significantly increased in the atrial myocytes of sln−/− mice (Fig. 3, F and G). These results were consistent with AP morphology alteration, as shown in Fig. 2, A–C.

To determine cell size, we calculated the membrane capacitance, an index of cell surface area. The membrane capacitance of atrial myocytes isolated from the sln−/− mice atria were not different from that of atrial myocytes from WT mice (WT: 49.6 ± 2.1 pF vs. sln−/−: 54.5 ± 3.1 pF).

Increased RyR phosphorylation in the sln−/− atria.

Next, we examined if altered expression of NCX or L-type calcium channel contributes to the changes in I_NCX or I_Ca,L activity. In addition, we quantitated the SERCA2a, PLN, calsequestrin (CSQ), and RyR protein levels. Our results show that the protein levels of SERCA2a, PLN, RyR, NCX CSQ, and the L-type calcium channel subunit, dihydropyridine receptor-α, were not altered in the sln−/− mice atria compared with that of age- and sex-matched WT control mice (Fig. 4, A and B).

Fig. 4.

Western blot analysis of Ca²⁺ handling proteins in the sln−/− mice atria. A: Western blots showing the protein levels of SR Ca²⁺-ATPase (SERCA) 2a, calsequestrin (CSQ), NCX, dihydropyridine receptor (DHPR-α), phospholamban (PLN), and ryanodine receptor (RyR). B: the fold change in expression levels of these proteins. The expression levels are normalized to CSQ levels. C: the ratios of serine 16 phosphorylated PLN (PS16-PLN) to total PLN and serine 2808 or 2814 phosphorylated RyR to total RyR; n = 3. Values are means ± SE. *Significantly different from the WT mice atria, P < 0.01.

To determine the phosphorylation status of PLN and RyR in the sln−/− atria, we measured the S16 phosphorylation of PLN and S2808 and S2814 phosphorylation of RyR by Western blot analysis using phosphor-specific antibodies. Results in Fig. 4, A and C, show that the basal phosphorylation of PLN was not altered, whereas the basal phosphorylation of RyR both at S2808 (2.87 ± 0.2-fold, P < 0.05) and S2814 (6.29 ± 1.14-fold, P < 0.05) was significantly increased in atria of sln−/− mice.

Atrial fibrosis in the sln−/− mice.

Ablation of SLN caused no apparent hypertrophic effect. The sizes of the sln−/− atria and atrial myocytes were not significantly different from that of age- and sex-matched WT control mice (data not shown). Hematoxylin and eosin staining showed no apparent pathological changes in the sln−/− atria. Masson's trichrome staining, on the other hand, showed increased fibrosis (data not shown). The Sirius Red staining that marks collagen revealed that the sln−/− atria has increased amount of interstitial fibrosis, as indicated by thick strands between the fibers (WT = 13.4 ± 1.5% vs. sln−/− = 29.7 ± 1.3%; P < 0.0001; n = 5; Fig. 5). There was no significant difference in Sirius Red staining between the ventricles of WT and sln−/− mice (data not shown).

Fig. 5.

Increased interstitial fibrosis in the sln−/− atria. A: representative images showing the Picro Sirius Red staining of atrial tissues from 5-mo-old WT and sln−/− mice. Original magnifications ×400. Bar represents 20 μm. Arrow indicates the collagen accumulation between the fibers. B: quantitation of Sirius Red stained connective tissue in atria of WT and sln−/− mice. Values are means ± SE; n = 5. *P < 0.0001.

Gene expression changes in the sln−/− atria mimic AF phenotype.

Gene expression profiling was carried out using Affymetrix GeneChip Mouse Gene 1.0 ST Array. Among 1,415 genes, 400 genes were upregulated, and 1,011 genes were downregulated (significant analysis of microarrays fold change >1.2; q value <0.05) in atria of sln−/− mice. Significantly regulated genes were classified using the GO database. These genes were in three major ontology categories, such as biological process, cellular component, and molecular function (Table 1). The GO analysis showed that downregulated genes tend to be associated with immune responses, fatty acid metabolism, and G protein-coupled receptor protein signaling pathway, whereas those upregulated tend to be associated with ion channel activity.

Table 1.

Significant Gene Ontology terms associated with differentially expressed genes in the sln−/− mice atria

	Association With Upregulated Genes	Association With Downregulated Genes
Biological process
GO:0043933, macromolecular complex subunit organization	9.05E-03	9.57E-01
GO:0043085, positive regulation of catalytic activity	8.43E-04	9.18E-01
GO:0006457, protein folding	NA	4.74E-03
GO:0007186, G protein-coupled receptor protein signaling pathway	1.45E-01	3.79E-03
GO:0006629, lipid metabolic process	7.32E-01	2.22E-03
GO:0006633, fatty acid biosynthetic process	7.30E-01	8.65E-04
GO:0006952, defense response	9.94E-01	3.52E-04
GO:0006955, immune response	9.97E-01	9.95E-07
Cellular component
GO:0009986, cell surface	5.02E-01	1.38E-03
GO:0005783, endoplasmic reticulum	9.07E-01	1.63E-04
GO:0005615, extracellular space	1.18E-01	1.38E-04
Molecular function
GO:0005244, voltage-gated ion channel activity	9.11E-03	8.89E-01
GO:0005102, receptor binding	2.55E-01	1.07E-03
GO:0016860, intramolecular oxidoreductase activity	NA	8.84E-05
GO:0004930, G protein-coupled receptor activity	4.74E-01	4.62E-05

Three Gene Ontology (GO) categories are shown, i.e., biological process, cellular component, and molecular function. P values indicating significance of association with up- and downregulated genes were derived from hypergeometirc test. Redundant GO terms were removed. See materials and methods for details. NA indicates no genes associated with a GO term.

Next, we compared gene expression changes to the available human AF gene expression profiling (GEO accession no. GSE2240). Results show that ∼105 genes were significantly downregulated, and 25 genes were upregulated in common in the atria of human AF and sln−/− mice (Fig. 6A). We also found that the expression of procollagens type I, III, IV, V, VI, VIII, XIV, and XV; inflammatory genes like cathepsin K; MMPs such as MMP3, tissue inhibitor of MMP4; cytokines like BMP5; and S100A1 were significantly upregulated. On the other hand, the extracellular matrix protein connexin 40, and cystatin 9, an inhibitor of cathepsin, were selectively downregulated in the sln−/− mice atria (Table 2). To validate the microarray data, qRT-PCR analysis of selected genes was carried out. Results in Fig. 6B confirm and validate the microarray data. Together, these results indicate that gene expression changes in the sln−/− mice atria mimic gene expression profiling reported for human and various animal models of AF (27, 33, 37).

Fig. 6.

Genomewide analysis of gene expression regulation in sln−/− atria. A: comparative analysis of significantly regulated genes in atria of sln−/− mice (GEO accession no. GSE23403) vs. those in human patients with atrial fibrillation (AF) (GEO accession no. GSE2240). Significant genes were selected using cutoffs of fold change of 1.2 and 5% false discovery rate (significant analysis of microarrays). A P value (χ² test) indicating significance of the numbers of overlapped genes is shown in the graph. *Statistically overrepresented. B: quantitative RT-PCR analyses of selected genes. BMP, bone morphogenetic protein; Cyst9, cystatin 9; Cx40, connexin 40; Timp4, tissue inhibitor of MMP4; CathK, cathepsin K; MMP3, matrix metalloproteinase 3. Values are means ± SE; n = 5. *Significant difference vs. WT, P < 0.05.

Table 2.

Expression profiles of collagen isoforms and extracellular matrix protein genes in the sln−/− mice atria

Gene ID	GSB	Fold Change	P Value	Gene Description
12156	Bmp2	−1.64	1.86E-02	Bone morphogenetic protein 2
12160	Bmp5	3.03	3.11E-02	Bone morphogenetic protein 5
12818	Col14a1	1.49	2.92E-01	Collagen, type XIV, α1
12819	Col15a1	1.84	1.75E-01	Collagen, type XV, α1
12842	Col1a1	1.71	3.50E-01	Collagen, type I, α1
12843	Col1a2	1.55	3.31E-01	Collagen, type I, α2
12825	Col3a1	2.07	3.22E-01	Collagen, type III, α1
12826	Col4a1	2.16	1.22E-01	Collagen, type IV, α1
12827	Col4a2	2.06	1.62E-01	Collagen, type IV, α2
68018	Col4a3bp	−1.16	1.86E-01	Collagen, type IV, α3
12829	Col4a4	1.19	3.58E-02	Collagen, type IV, α4
12830	Col4a5	1.31	1.73E-01	Collagen, type IV, α5
94216	Col4a6	1.07	3.01E-01	Collagen, type IV, α6
12831	Col5a1	1.35	2.96E-01	Collagen, type V, α1
12833	Col6a1	1.77	2.82E-01	Collagen, type VI, α1
245026	Col6a6	−1.07	2.61E-01	Collagen, type VI, α6
245026	Col6a6	−1.16	1.86E-01	Collagen, type VI, α6
13013	Cst9	−3.43	7.06E-03	Cystatin 9
13038	Ctsk	4.14	2.77E-03	Cathepsin K
14613	Gja5	−1.82	7.80E-03	Gap junction protein, α5
17392	Mmp3	2.27	4.55E-02	Matrix metallopeptidase 3
20193	S100a1	1.40	4.50E-03	S100 calcium binding protein A1
110595	Timp4	1.96	9.07E-03	Tissue inhibitor of metalloproteinase 4

GSB, Gene Set Builder.

The sln−/− mice are susceptible to atrial arrhythmias upon aging.

At 5 mo of age, the WT and sln−/− mice showed normal ECG pattern (data not shown). ECG recordings from the 12-mo-old WT mice showed normal P waves and QRS complexes (Fig. 7A). There were no atrial flutters or fibrillations observed in the WT mice (n = 7). However, in some (4 out of 6) of the 12-mo-old sln−/− mice, the P waves were completely replaced with small oscillatory waves (Fig. 7B), similar to that of AF/flutters. Some recordings (Fig. 7C) showed typical atrial flutter behavior associated with rather regular R-R intervals, but are likened to a sawtooth pattern at baseline (18). The means of the R-R intervals for the two groups (WT and sln−/−) of mice were not significantly different (data not shown) under the anesthetic conditions used for ECG recordings.

Fig. 7.

ECG recordings of 12-mo-old sln−/− mice. Representative surface ECG lead II recordings from 12-mo-old WT (A) and sln−/− (B and C) mice are shown. Arrows indicate the P waves.

DISCUSSION

In this study, using the sln−/− mouse model (4), we investigated the effect of enhanced atrial SERCA pump activity on cellular electrophysiology and structural remodeling. The key findings demonstrate that ablation of SLN in atria results in 1) increased SR Ca²⁺ load, Ca²⁺ transients, and spontaneous Ca²⁺ oscillation; 2) prolongation of APD due to the activation of forward mode of I_NCX; 3) increased I_Ca,L channel activity; and 4) structural remodeling and gene expression changes. Furthermore, our findings suggest that the sln−/− mice are susceptible to spontaneous atrial arrhythmias upon aging.

It is well documented that SERCA pump plays a significant role in maintaining the intracellular Ca²⁺ concentration ([Ca²⁺]_i) in the heart (31, 32). SLN is a key regulator of the cardiac SERCA pump (1, 3, 5, 20), and its expression decreased in atria of patients with chronic AF (39, 44). To determine the consequences of SLN downregulation in SR Ca²⁺ handling, we generated a sln−/− mouse model (4). The data from the present study and an earlier publication (4) demonstrate that loss of SLN did not affect the Ca²⁺ transients and SR Ca²⁺ load in the ventricles, since it is a minor component. On the other hand, loss of SLN enhanced SERCA activity, caused SR Ca²⁺ overload, and increased both Ca²⁺ transients and I_NCX in atria. These observations, along with studies on transgenic mice overexpressing SLN (1, 3, 20), indicate that SLN levels can affect the SR Ca²⁺ cycling and SR Ca²⁺ load. Thus the downregulation of SLN in AF (39, 44) could contribute to the abnormal Ca²⁺_i handling through enhanced SERCA pump activity and SR Ca²⁺ overload.

In addition to SR Ca²⁺ overload, spontaneous Ca²⁺ release, DADs, and triggered activities were more frequent in the atrial myocytes of sln−/− mice. The SR Ca²⁺ leak due to hyperphosphorylation of RyR2 by protein kinase A (PKA) is suggested to play an important role in atrial arrhythmogenesis (24, 50). A recent study demonstrates that an increase in CaMKII phosphorylation of RyR at S2814 can also enhances the SR Ca²⁺ leak and contributes to AF initiation in FKBP12.6 knockout mice (26). The increased basal phosphorylation of RyR, both at S2808 (a PKA target site) and S2814 (a CaMKII site) in atria of the sln−/− mice, suggests that hyperphosphorylated RyR may contribute to increased SR Ca²⁺ leak and spontaneous Ca²⁺ waves. It has been shown that the enhanced SR Ca²⁺ leak is sufficient to induce atrial arrhythmias without altering the SR Ca²⁺ load in human AF (23) and in a FKBP12.6-deficient mouse model (40). However, in congestive heart failure dogs, the generation of DADs and triggered activities were associated with higher SR Ca²⁺ load in atria (54). Similarly, atrial arrhythmias observed during acidosis were reported to be associated with an increase in SR Ca²⁺ load (35). Thus the enhanced SR Ca²⁺ load and spontaneous Ca²⁺ release could contribute to the generation of DADs and triggered activities via upregulated I_NCX in the sln−/− atrial myocytes.

The activation of the forward mode of I_NCX is attributable to the late plateau and prolongation of APD₉₀ in the atrial myocytes of sln−/− mice. These observations, along with our laboratory's earlier studies showing enhanced SERCA pump activity and Ca²⁺ transients in the sln−/− mice atria (4), suggest that the high level of [Ca²⁺]_i may promote inward I_NCX. The NCX levels have been shown to affect the APD and the late plateau phase (22, 34). However, the NCX protein level was not altered in the sln−/− atria (Fig. 4), suggesting that the increased I_NCX may be due to the functional modifications of NCX protein (6). Together, our studies demonstrate that alterations in Ca²⁺_i cycling due to the enhanced SERCA pump activity and Ca²⁺ transients can cause abnormal APD dynamics via I_NCX.

The I_Ca,L activity was significantly increased in the sln−/− atrial myocytes. At present, we do not know the precise molecular mechanisms that enhance the I_Ca,L activity and its role in atrial remodeling in the sln−/− mice. PKA is shown to preassociate with I_Ca,L and RyRs, and thus the PKA-dependent phosphorylation can increases the I_Ca,L (55) and RyR (28) opening. However, a recent study demonstrated that PKA phosphorylation of S1928 is not functionally involved in β-adrenergic regulation of I_Ca,L (25). On the other hand, CaMKII-dependent phosphorylation of Cav1.2 at S1512 and S1570 has been shown to mediate the Ca²⁺ current facilitation and was suggested to contribute to arrhythmogenesis (7). Since the protein levels of dihydropyridine receptor-α were not altered (Fig. 4), we speculate that the upregulation of I_Ca,L activity could be through enhanced CaMKII-mediated phosphorylation. The increased I_Ca,L activity may be an adaptive mechanism to compensate the sustained activation of atrial SERCA pump in the absence of SLN and to maintain the cytosolic Ca²⁺ levels and contractile function of the heart. The increase in I_Ca,L activity could also trigger more RyR Ca²⁺ release and enhance the amplitude of Ca²⁺ transient, to overcome the shortening of Ca²⁺ transient duration.

In AF, electrical remodeling of atrial myocytes is characterized by shortening of APD, decreased SR Ca²⁺ content, I_Ca,L, and outward I_K (46). On the other hand, the atrial myocytes of sln−/− mice exhibited increased SR Ca²⁺ load, Ca²⁺ transient, and I_Ca,L activity. Furthermore, these mice are susceptible to spontaneous atrial arrhythmias upon aging. Consistent with our findings, a recent study demonstrated that the enhanced Ca²⁺ cycling, including increased SR Ca²⁺ load, Ca²⁺ transient, and I_Ca,L, can promote AF (54). Although we do not have a ready explanation for these discrepancies, atrial arrhythmias may correlate to either enhanced or decreased I_Ca,L activity.

In addition to ion channel remodeling, the abnormal Ca²⁺_i handling in the sln−/− atria induces gene expression changes. Some of the altered genes, in particular the key extracellular matrix protein genes, including collagens, fibrillin, and connexin 40, can contribute to the atrial structural remodeling in the sln−/− mice. The increased expression of transforming growth factor-β mRNA in the sln−/− atria suggests that the activation of transforming growth factor-β pathway could be involved in atrial fibrogenesis (17, 19) in the sln−/− mice.

Taken together, our studies demonstrate that structural and electrophysiological changes in the sln−/− mice atria mimic AF-associated atrial remodeling. It is also important to note that some (4 out of 6) of the 12-mo-old sln−/− mice developed spontaneous AFs/flutters, suggesting that the sln−/− mice may be susceptible to atrial arrhythmias. However, further studies using telemetry to detect AFs/flutters in both young and old sln−/− mice are warranted to validate these findings.

In conclusion, our studies indicate that SLN play a key role in maintaining atrial Ca²⁺ homeostasis. Loss of SLN function can enhance atrial SERCA pump activity and SR Ca²⁺ load, which, in turn, could cause abnormal Ca²⁺_i handling and subsequent atrial remodeling, characterized by changes in ion channel function and altered gene expression.

GRANTS

This work was supported by funds from the Department of Cell Biology and Molecular Medicine, University of Medicine and Dentistry of New Jersey (UMDNJ)-New Jersey Medical School, Newark, NJ, and a UMDNJ Foundation grant (to G. J. Babu). M. Periasamy is supported by a National Heart, Lung, and Blood Institute (NHLBI) Grant (HL 080551). L.-H. Xie is supported by a NHLBI Grant (HL097878).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

L.-H.X. and G.J.B. conception and design of research; L.-H.X., M.S., J.Y.P., Z.Z., H.W., and G.J.B. performed experiments; L.-H.X., M.S., J.Y.P., Z.Z., B.T., and G.J.B. analyzed data; L.-H.X., M.P., and G.J.B. interpreted results of experiments; L.-H.X. and G.J.B. prepared figures; L.-H.X. and G.J.B. drafted manuscript; L.-H.X., B.T., M.P., and G.J.B. edited and revised manuscript; L.-H.X., M.S., J.Y.P., Z.Z., H.W., B.T., M.P., and G.J.B. approved final version of manuscript.