A total of 9 cardiomyopathy datasets were analyzed, including six oligonucleotide and three cDNA microarray datasets. The first oligonucleotide microarray dataset (GDS411) (15) consisted of 53 samples, including 12 normal, 12 heart failure, 12 rescue heart failure and 17 other types of samples. The second dataset (GDS651) comprised 37 samples, including 11 normal, 15 idiopathic dilated and 11 ischemic cardiomyopathy (ICM) samples. The third dataset (GDS1264) (16) consisted of 23 samples, including 11 normal and 12 cardiomyopathy samples. The fourth dataset (GDS1362) (17) contained 30 samples, including 15 normal and 15 ICM samples. The fifth dataset (GDS3386) (18,19) contained 32 samples, including 16 normal and 16 myocardial hypertrophy samples. The sixth dataset (GDS2145) (20) consisted of 28 samples, including 15 normal and 13 dilated cardiomyopathy samples. Regarding the cDNA microarray datasets, the first dataset (GDS2206) (21) comprised 20 samples, including normal samples and those from patients with myocardial infarction (MI). The second dataset (GSE1616) (22) consisted of 37 samples, from six normal and 21 nonischemic patients, as well as 10 patients with ICM. The third dataset (GSE18224) (23) contained 24 samples, including 12 normal and 12 MI samples.

The cDNA data was log₂-transformed and each array was subsequently normalized as median zero and standard deviation (SD) one per array, as adopted from the Oncomine database (24). CloneIDs with a missing rate >20% were deleted. The remaining missing values were replaced using the k nearest neighbor imputation algorithm (k=15) (25). The Affymetrix GeneChip data were pre-processed using the robust multi-array analysis algorithm and then the between-array median was normalized (26).

The significance analysis of microarrays (SAM) method was performed using the samr R package (27) to select DEGs. Multiple statistical tests were controlled by false discovery rate (FDR), defined as the expected percentage of false positives among the claimed DEGs (28). The FDR estimation of SAM may be overly conservative (29,30); therefore, the FDR estimation method suggested by Zhang (30) and influenced by Xie et al (29), was also employed and referred to as the modified SAM method.

H9C2 cells (Chinese Academy of Sciences, Shanghai, China) were cultured on 96-well plates in Dulbecco's modified Eagle's Medium (DMEM) containing 1 g/l glucose supplemented with 10% heat-inactivated fetal bovine serum (Wisent Bioproducts, St. Bruno, QC, Canada), 100 U/ml penicillin and 100 μg/ml streptomycin (Invitrogen Life Technologies, Carlsbad, CA, USA) in a humidified atmosphere at 37°C in 5% CO₂. To prevent cell loss, cells were subcultured prior to confluence, at a ratio of ~1:2, every two days. Cells in the logarithmic growth phase were used in the experiments.

Following inoculation for 24 h, H9C2 cells were treated with the indicated concentrations of H₂O₂ (Sangon Biotech, Shanghai, China), which was added to the culture medium. Cells were then incubated further for the indicated times.

Following exposure to H₂O₂ for 4 h, H9C2 cells were treated with MTT (Nanjing KeyGen Biotech Co. Ltd., Nanjing, China) and then incubated for 4 h at 37°C in the dark. The supernatant was subsequently removed and 150 μl dimethylsulfoxide was added to each well. The optical density (OD) was measured at a wavelength of 550 nm once the crystals had dissolved. Cell viability was calculated as the percentage of the control OD.

The percentage of apoptotic cells was measured using the Annexin V-fluorescein isothiocyanate (FITC) Apoptosis Detection kit (Nanjing KeyGen Biotech Co. Ltd.) according to the manufacturer's instructions. Following exposure to H₂O₂ for 48 h, cells were harvested using 0.25% Trypsin without ethylenediaminetetraacetic acid, and then washed twice with phosphate-buffered saline (PBS), prior to resuspension in 500 μl binding buffer with 5 μl Annexin V-FITC and 5 μl propidium iodide (PI). Subsequent to incubation for 15 min in the dark, the samples were analyzed using flow cytometry (BD Biosciences, Franklin Lakes, NJ, USA). The number of cells in different quadrants represented different cell populations, with the lower left quadrant representing normal cells, the lower right representing viable apoptotic cells, the upper right representing non-viable apoptotic cells and the upper left representing necrotic cells.

Subsequent to the relevant treatment, cells were washed twice with ice-cold PBS and then lysed in cell lysis buffer containing 1% phenylmethylsulfonyl fluoride (PMSF) for 30 min on ice. The lysate was centrifuged at 15,249 × g at 4°C for 20 min to remove the insoluble materials, followed by collection of the supernatants. All samples were mixed with 5X loading buffer and boiled for 5 min. Samples were separated using 8% SDS-PAGE and then transferred to polyvinylidene fluoride (PVDF) membranes (Billerica, Millipore, MA, USA). Membranes were blocked for 1 h using 5% non-fat milk in Tris-buffered saline containing 1% Tween-20 (TBST) at room temperature and then incubated with anti-UBE3A (Cell Signaling Technology, Danvers, MA, USA), anti-p53 (BioWorld Products Inc., Visalia, CA, USA) and anti-GAPDH (Epitomics Inc., Burlingame, CA, USA) antibodies overnight at 4°C. Membranes were washed with TBST three times every 10 min, then incubated with a horseradish peroxidase-conjugated secondary antibody (Beyotime Institute of Technology, Haimen, China) for 1 h at room temperature. Following the secondary antibody incubation, membranes were further washed with TBST and the immunoreactive bands were visualized using the enhanced chemiluminescence method. The images were analyzed using the Quantity One^® software (Bio-Rad, Hercules, CA, USA).

RNA was prepared using TRIzol^® Reagent (Invitrogen Life Technologies) and RNA concentration and purity were then determined using a spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Total RNA was subsequently converted into cDNA by reverse transcription using the High-Capacity cDNA Reverse Transcription kit (Applied Biosystems, Carlsbad, CA, USA). qPCR was performed in triplicate using Power SYBR^® Green PCR Master mix and a 7500 Fast Real-Time PCR system (Applied Biosystems) according to the manufacturer's instructions. Analysis was performed using the software supplied with the instrument. Primer sequences were as follows: Forward: 5′-GAGGACTCG GAAAATTGAAGATG-3′ and reverse: 5′-CCGGAAGTA AAAGGACATTAAAGC-3′ for Ube3a; and forward: 5′-CCA TCAACGACCCCTTCATT-3′ and reverse: 5′-GACCAG CTTCCCATTCTCAG-3′ for GAPDH.

Statistical analyses were performed using the SPSS 14.0 statistical software (SPSS, Inc., Chicago, IL, USA). All data are expressed as the mean ± SEM from at least three independent experiments. Results were analyzed using one-way analysis of variance. P<0.05 was considered to indicate a statistically significant difference.

Each experiment was repeated at least three times with comparable results, unless indicated otherwise.

Current FDR control procedures, including that adopted in SAM (27), may be unstable in small samples particularly in the presence of correlated expression changes. Therefore, in the present study, the actual FDR of DEGs detected in simulated small samples was evaluated, according to predefined DEGs. Based on the simulated results, using SAM with 0.05% FDR control, the DEGs obtained from the full samples were defined as a nominal gold standard set (31). As shown in Table I, this procedure identified various DEGs. Although false positives were detected in the selected DEGs, this was a preliminary procedure to prepare for the subsequent analysis of various datasets.

Different datasets supply different information. The DEGs selected using the aforementioned method were associated with numerous cardiomyopathy-associated genes. A total of five DEGs were selected, all of which belong to the E3 ubiquitin ligase family, and were termed cardiomyopathy-associated E3 ubiquitin ligase genes. These DEGs were: Ube3a; WW domain containing E3 ubiquitin protein ligase 1 (Wwp1); itchy E3 ubiquitin protein ligase (Itch); HECT, C2 and WW domain containing E3 ubiquitin protein ligase 1 (Kiaa0322) and SMAD specific E3 ubiquitin protein ligase 2 (Smurf2). The mRNA levels of these five candidate genes were detected using qPCR analysis. Ube3a mRNA levels were observed to be significantly higher in the H₂O₂-treated group than in the control group; therefore, Ube3a was selected as a candidate myocyte apoptosis-associated gene for further investigation.

H9C2 cells were exposed to increasing concentrations of H₂O₂ ranging between 150 and 300 μM for 4 h. Cell viability was then assessed by MTT uptake. The results showed that in the H₂O₂-treated group, cell viability decreased in a dose-dependent manner compared with the PBS-treated control group (Fig. 1).

Reports have shown that H₂O₂ is capable of inducing myocardial hypertrophy at low concentrations and apoptosis at higher concentrations (32). Annexin V/PI staining and flow cytometry were used to determine the apoptotic response of H9C2 cells to high concentrations of H₂O₂. Cells were harvested following exposure to H₂O₂ (150–250 μM) for 4 h. As shown in Fig. 2A–D, the Annexin V/PI point diagram exhibited a significant dose-dependent increase in apoptotic cells in response to H₂O₂. Following treatment with 250 μM H₂O₂ for 4 h, the percentage of apoptotic H9C2 cells increased approximately five-fold compared with the PBS-treated control cells (Fig. 2E). These results suggest that H₂O₂ exposure is capable of inducing apoptosis in a dose-dependent manner in H9C2 cells.

To examine the effect of H₂O₂-induced apoptosis on Ube3a levels in H9C2 cells, cells were treated with H₂O₂ in accordance with the aforementioned method. Ube3a protein levels were observed to increase in the same manner as that of p53 (Fig. 3). An increase in Ube3a transcription was also observed following H₂O₂ treatment (Fig. 4).