LncRNA SNHG1 promotes sepsis‐induced myocardial injury by inhibiting Bcl‐2 expression via DNMT1

Abstract Myocardial injury is a frequently occurring complication of sepsis. This study aims to investigate the molecular mechanism of long noncoding RNA (lncRNA) small nucleolar RNA host gene 1 (SNHG1)‐mediated DNA methyltransferase 1/B‐cell lymphoma‐2 (DNMT1/Bcl‐2) axis in sepsis‐induced myocardial injury. Mice and HL‐1 cells were treated with lipopolysaccharide (LPS) to establish animal and cellular models simulating sepsis and inflammation. LncRNA SNHG1 was screened out as a differentially expressed lncRNA in sepsis samples through microarray profiling, and the upregulated expression of lncRNA SNHG1 was confirmed in myocardial tissues of LPS‐induced septic mice and HL‐1 cells. Further experiments suggested that silencing of lncRNA SNHG1 reduced the inflammation and apoptotic rate of LPS‐induced HL‐1 cells. LncRNA SNHG1 inhibited Bcl‐2 expression by recruiting DNMT1 to Bcl‐2 promoter region to cause methylation. Inhibition of Bcl‐2 promoter methylation reduced the inflammation and apoptotic rate of LPS‐induced HL‐1 cells. In vivo experiments substantiated that lncRNA SNHG1 silencing alleviated sepsis‐induced myocardial injury in mice. Taken together, lncRNA SNHG1 promotes LPS‐induced myocardial injury in septic mice by downregulating Bcl‐2 through DNMT1‐mediated Bcl‐2 methylation.

lncRNA in sepsis. Notably, lncRNA SNHG1 has been involved in multiple pathologies including colorectal cancer cell growth, 8 neurotoxicity in sevoflurane 9 and oxygen-glucose deprivation/ reperfusion-induced cell death. 10 Emerging evidence has supported the protective role exerted by lncRNA SNHG1 in cardiomyocytes against hypertrophy, 11 drug toxicity 12 and apoptosis 13 ; and it has further been suggested by Luo et al., that SNHG1 may alleviate sepsisinduced cardiomyopathy through a microRNA (miRNA)-dependent regulatory mechanism. 14 On this basis, this study sought to further probe the role and mechanistic actions of lncRNA SNHG1 in sepsisinduced cardiac injury.
As previously reported, lncRNA SNHG1 promoted gastric cancer cell proliferation by elevating the expression of DNA methyltransferase 1 (DNMT1). 15 DNMT1 is regarded as a crucial gene being capable of regulating DNA methylation patterns in the process of genomic DNA replication. 16 Intriguingly, increased total DNMT messenger RNAs (mRNAs) were observed in extracellular vesicles (EVs) from patients with septic shock compared with those in EVs from controls or patients with sepsis. 17 Further, in a mouse model of emphysema, DNMT1 has been highlighted to induce hypermethylation of the Bcl-2 promoter and downregulate Bcl-2, which is an anti-apoptotic factor to repress cell apoptosis. 18,19 Of note, cardiomyocyte apoptosis has been indicated to result in continuous increase of ventricular dysfunction in heart failure, including cardiomyopathy, myocardial infarction and ischaemia-reperfusion injury. 20 Therefore, we made a hypothesis in the present study that ln-cRNA SNHG1 may orchestrate the DNMT1/Bcl-2 axis to affect the sepsis-induced myocardial injury.

| Ethical approval
This study was performed under the approval of the laboratory animal care committee of Xi'an Jiaotong University. The animal experiments were conducted in accordance with the guidelines for the care and use of laboratory animals.

| Bioinformatics analysis
Sepsis-related microarrays GSE9667 (containing 6 myocardium samples of septic mice and 3 of normal controls) and GSE4479 (containing 5 myocardium samples of septic mice and 5 of normal controls) were retrieved from the GEO database. The GEO2R online analysis tool was then utilized for the identification of differentially expressed lncRNAs with p < 0.05 as the screening threshold.

| Cell treatment
HL-1 mouse cardiac muscle cell line was purchased from Procell and resuspended and incubated in 10% FBS-contained DMEM medium under 37°C, 5% CO 2 . Upon reaching 80%-90% confluence, cells were subcultured. The cells of the third passage (P3) were treated with 1 μg/ml lipopolysaccharide (LPS) to induce inflammatory reaction.

| RNA quantification and RT-qPCR
The Trizol reagent (Invitrogen) was utilized to extract total RNA from tissues or cells, and the RNA concentration was determined. mRNA was reversely transcribed into complementary DNA (cDNA) by the ImProm-II™ RT system kit (Promega). The reversely transcribed cDNA was diluted to 50 ng/μl, followed by fluorescence qPCR using SYBR Premixed ExTaq II Kit (Takara). Primers were synthesized by Shanghai Genechem, as shown in Table S2. With GAPDH as the housekeeping gene, the relative mRNA expression was calculated with the 2 −ΔΔC T method. The grey value of protein bands was analysed by Image J software.

| CCK-8 assay
Cells were collected, detached, resuspended (1 × 10 5 cells/ml) and seeded into a 96-well plate with 100 μl/well. Cells were routinely cultured overnight and treated with CCK-8 solution (Beyotime). The cell viability was detected by CCK-8 method at 24, 48 and 72 h after seeding. Next, 10 μl of CCK-8 detection solution was added to the cells in each test, followed by 4-h incubation. The absorbance at 450 nm was detected using a microplate reader (Invitrogen; Thermo Fisher Scientific, Inc.).

| Establishment of a mouse model of LPSinduced sepsis
C57BL/6 male mice aged 6-10 weeks were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. and raised in a specific-pathogen-free animal room at constant temperature (25-27°C) under constant humidity (45%-50%). Eight mice as controls (referred to as the Normal group) were intraperitoneally injected with of PBS (10 mg/kg). Then, 48 mice were intraperitoneally injected with LPS (10 mg/kg, Sigma-Aldrich) for the model establishment and classified into three groups (n = 16), which were injected with LPS alone, or pre-treated with sh-NC or sh-SNHG1 48 h before LPS induction, referred to as LPS, LPS + sh-NC and LPS + sh-SNHG1 groups.
Five hours after LPS or PBS injection, the mice were analysed by high-resolution imaging system VisualSonics Vevo2100

| Echocardiography
The heart rate (HR) and ventricular function were evaluated by tran-

| Nuclear-cytoplasmic fractionation
The cytoplasmic and nuclear RNA purification kits (Norgen) were used for isolation and purification of cytoplasmic and nuclear RNAs.

| Methylation-specific PCR
Genomic DNA was extracted by a genomic DNA extraction kit (Beijing Tiangen Biochemical Technology). Concentration and purity of DNA were quantified by ultraviolet spectrophotometry, followed by storage in a refrigerator at −80°C. DNA was treated with sodium sulphite using EZ DNA methylation kits (Zymo Research Corp.), desulphurized and purified for subsequent PCR. Methylated and unmethylated primers were designed for the CpG island enrichment region of Bcl-2 gene promoter (Table S3). The products were subjected to agarose gel electrophoresis. Image analysis was performed using a gel electrophoresis imaging analysis system. If Following IP, the supernatant was removed through centrifugation at 4°C and 3000 ×g for 5 min. Protein A/G magnetic beads were precipitated with 1 ml lysate, followed by centrifugation at 4°C and 1000 ×g for 1 min. The RNA was isolated and purified from the precipitate by Trizol, and lncRNA SNHG1 expression was determined by RT-qPCR.

| RNA pull-down assay
HL-1 cells were harvested and lysed for 30 min with 100 μl of precooled lysis buffer containing protease inhibitor and RNAase inhibitor on ice for 30 min, followed by 30 min centrifugation (12,000 ×g, 4°C). The supernatant added with biotin-labelled LncRNA SNHG1 probes (Ribobio) and incubated overnight with protein lysis buffer, followed by another 3-h incubation at 4°C with M-280 streptavidin magnetic beads (Sigma-Aldrich). After that, the mixture was successively washed with cold lysis buffer, low salt buffer and high salt buffer to extract proteins, and DNMT1 expression was then determined using Western blot analysis. At the same time, an equal amount of cells were selected for lysis and the protein was extracted, and the expression of GAPDH was detected by Western blot as an internal reference protein. The grey value of protein expression in Western blot was determined with ImageJ software (NIH free software), the grey value of DNMT1 protein in each group was normalized with the grey value of GAPDH internal reference protein, and then, Input normalize to 1.0 to calculate the other two groups of DNMT1 protein grey value.

| Chromatin immunoprecipitation
When the confluence of HL-1 cells reached 70%-80%, 1% formaldehyde was added for fixation at room temperature for 10 min The specific primers for Bcl-2 gene promoter were purchased from Ribobio.

| Haematoxylin and eosin staining
The specimens fixed by formaldehyde were dehydrated, embedded into paraffin at 60°C, cut into 5μm slices and then dewaxed with xylene. After that, the slices were stained with HE reagents (Beijing Solarbio Biotechnology Co., Ltd.), dehydrated, permeabilized and observed using a fluorescent microscope (BXF-200; Bingyu). Five different fields of view were selected for photographing, and then, ImageJ software was used to calculate the percentage of necrotic area in each field of view: necrotic area/ total area × 100%.

| Picrosirius red staining
The heart was isolated from euthanized mice, perfused with PBS, fixed in 4% paraformaldehyde for 8 h, dehydrated with 20% sucrose, embedded with paraffin and sliced into 5μm sections, followed by Picrosirius red staining to detect collagen deposition. The myocardial collagen deposition area fraction was calculated as the ratio of stained collagen deposition area to total myocardial area.

| Statistical analysis
All the data in this study were analysed using the SPSS 21.0 statistical software (IBM). The quantitative data were expressed by mean ± standard deviation. Data between two groups were compared by independent sample t test, and those among multiple groups by one-way analysis of variance (ANOVA) with Tukey's post-hoc test.
Variables were analysed at different time points using Bonferronicorrected repeated measures ANOVA. p < 0.05 indicated statistically significant difference.

| LncRNA SNHG1 was highly expressed in myocardial tissues of LPS-induced septic mice and cell model
Sepsis-related datasets GSE9667 and GSE4479 were obtained from GEO database. Through microarray profiling, ten upregulated lncRNAs and four downregulated lncRNAs were identified in GSE9667 ( Figure 1A). Meanwhile, 49 upregulated lncRNAs and 41 downregulated lncRNAs were screened from GSE4479 ( Figure 1B).
Among them, lncRNA SNHG1 was upregulated in both the two datasets ( Figure 1C, D) and was thus selected for subsequent experiments.
In terms of the in vitro experiments, results of CCK-8 assay showed that LPS-treated cells had diminished cell viability ( Figure 1E).

RT-qPCR results demonstrated that LPS-treated cells presented
with an increase in the expression of lncRNA SNHG1 ( Figure 1F).
Next, a notably increased apoptotic rate was found in LPS-treated cells, as detected by flow cytometry and TUNEL staining ( Figure 1G, H). ELISA results showed that LPS-treated cells displayed increased production of inflammatory factors TNFα, IL-1β and IL-6 in LPSinduced HL-1 cells ( Figure 1I).
The expression of lncRNA SNHG1 was subsequently determined by RT-qPCR to be upregulated in the myocardium of LPS-induced septic mice relative to that of control mice ( Figure 1J). These results demonstrate that lncRNA SNHG1 is upregulated in myocardial tissue of LPS-induced septic mice and cell model.

| Silencing lncRNA SNHG1 reduced LPSinduced inflammation and apoptosis in HL-1 cells
To investigate the effect of lncRNA SNHG1 on LPS-induced myocardial injury, we further infected LPS-treated cells with lentiviral sh-SNHG1/oe-SNHG1, and the knockdown/overexpression efficiency was verified by RT-qPCR (Figure 2A). The cell viability upon sh-SNHG1 treatment was increased, while oe-SNHG1 markedly diminished the cell viability, as determined by CCK-8 assay ( Figure 2B). In terms of the effect of lncRNA SNHG1 on apoptosis in myocardial injury, flow cytometry and TUNEL staining displayed that sh-SNHG1 diminished apoptotic rate, while oe-SNHG1 increased the apoptotic rate ( Figure 2C, D). ELISA results showed that silencing lncRNA SNHG1 led to marked decreases in the production of inflammatory factors TNFα, IL-1β and IL-6, while lncRNA SNHG1 overexpression increased the production of these inflammatory factors ( Figure 2E). These results suggest that silencing lncRNA SNHG1 reduces LPS-induced inflammation and apoptosis in HL-1 cells.

| LncRNA SNHG1 inhibited Bcl-2 expression through DNMT1
Subsequently, we investigated whether lncRNA SNHG1 can regulate the expression of DNMT1 in HL-1 cells. FISH results unravelled that lncRNA SNHG1 was distributed in the nucleus and cytoplasm, but mainly in the nucleus ( Figure 3A), which was further confirmed by nuclear-cytoplasmic fractionation, similar to nuclear positive control U1 and different from cytoplasmic positive control GAPDH ( Figure 3B). It was predicted by RPISeq online tool that lncRNA SNHG1 might bind to DNMT1 ( Figure 3C).
Hence, we speculated that lncRNA SNHG1 may interact with DNMT1.
To verify the speculation, we detected the interaction between lncRNA SNHG1 and DNMT1 by RNA pull-down and RIP experiments in HL-1 cells. The results of RNA pull-down assay ( Figure 3D) showed that lncRNA SNHG1 enriched DNMT1 protein, while GAPDH and magnetic beads alone could not enrich it. RIP experiment further confirmed the interaction between lncRNA SNHG1 and DNMT1. As shown in Figure 3E, overexpression of lncRNA SNHG1 resulted in increased binding between DNMT1 and lncRNA SNHG1, while silencing lncRNA SNHG1 led to opposite effects.
These data confirmed the interaction between lncRNA SNHG1 and DNMT1.
We further explored the interplay among lncRNA SNHG1, Bcl-2 methylation and DNMT1. ChIP assay results revealed that DNMT1 enrichment on Bcl-2 promoter was increased in HL-1 cells as compared with IgG, while silencing of lncRNA SNHG1 reduced the DNMT1 enrichment on Bcl-2 promoter, which was reversed by overexpressing lncRNA SNHG1 ( Figure 3F). In addition, the MSP experiments showed that the methylation of Bcl-2 promoter in the presence of lncRNA SNHG1 knockdown was lower than that upon lncRNA SNHG1 overexpression ( Figure 3G). Western blot analysis displayed that sh-SNHG1 treatment led to increased Bcl-2 protein expression and oe-SNHG1 treatment led to the opposite results ( Figure 3H, Figure S1A).
Further, the MSP experiments suggested that the methylation of Bcl-2 upon 5-aza-dc treatment was diminished, while the methylation of Bcl-2 upon SNHG1 + DMSO displayed a marked increase.
The investigation focus was then shifted to the effect of Bcl-2 promoter methylation on cardiomyocyte functions. CCK-8 results displayed that the cell viability upon 5-aza-dC was increased, and that upon SNHG1 + DMSO was diminished. Compared with SNHG1 alone, its combination with 5-aza-dC resulted in increased cell viability ( Figure 4D). Flow cytometric results showed that 5-aza-dC alone notably diminished the apoptotic rate, while additional SNHG1 overexpression increased apoptosis. Relative to SNHG1 overexpression alone, additional 5-aza-dC treatment diminished the apoptotic rate ( Figure 4E). TUNEL staining results were consistent with flow cytometric results ( Figure 4F). As shown by ELISA results, 5-aza-dC led to diminished production of TNFα, IL-1β and IL-6, which was reversed when lncRNA SNHG1 was simultaneously overexpressed. Relative to lncRNA SNHG1 overexpression alone, its combination with 5-aza-dC resulted in lower TNFα, IL-1β and IL-6 ( Figure 4G).
These results suggest that inhibition of Bcl-2 promoter methylation is capable of inhibiting HL-1 cell inflammation and apoptosis.

| Silencing of lncRNA SNHG1 in vivo alleviated sepsis-induced myocardial injury in mice
To further explore the relationship between lncRNA SNHG1 and myocardial injury in septic mice, lncRNA SNHG1 was silenced in  Figure 5C). In the presence of sh-SNHG1, HR, LVEDD and LVESD were decreased, while E/A ratio was elevated ( Figure 5D). These results suggest that silencing lncRNA SNHG1 can improve myocardial function in LPS-induced septic mice.
The results from Picrosirius red staining displayed that LPS induction led to larger collagen deposition area in left ventricular myocardium, while lncRNA SNHG1 knockdown reduced the myocardial collagen deposition area of LPS-induced septic mice ( Figure 5E). The results of HE staining revealed that in sh-NC-treated septic mice, myocardial tissues presented obvious degeneration and necrosis, swelling and breaking of partial nuclei, rupture of some myocardial fibres, blurring of myocardial cross stripes, oedema and increase of lipid in interstitial tissues. The aforementioned symptoms were alleviated, and the necrotic area in myocardial tissues was reduced in the presence of lncRNA SNHG1 silencing ( Figure 5F). Moreover, lncRNA SNHG1 knockdown further led to diminished protein expression of Bax and cleaved caspase 3/caspase 3, while increasing that of Bcl-2 in myocardial tissues of LPS-induced septic mice ( Figure 5G, Figure S1C).
These results suggest that silenced lncRNA SNHG1 in vivo can alleviate sepsis-induced myocardial injury.

| DISCUSS ION
Myocardial injury is considered to be a frequently occurring symptom in sepsis. 22 Unfortunately, sepsis-induced myocardial injury presents a highly severe disorder for intensive care medicine because of its high morbidity and mortality. 23 In this study, we investigated the role of lncRNA SNHG1 in sepsis-induced myocardial injury and illustrated that lncRNA SNHG1 promoted sepsis-induced myocardial injury by regulating the DNMT1/Bcl-2 axis.
Our initial finding revealed that lncRNA SNHG1 was highly expressed in the LPS-induced mouse and cell models, and that Moreover, the use of DNMT inhibitors including procainamide and hydralazine contributed to alleviated hypotension, hypoglycaemia, as well as multiple organ dysfunction in rats with endotoxic shock induced by LPS. 27 Importantly, the interaction between DNMT1 and Bcl-2 has been increasingly reported. For example, DNMT1 contributed to hypermethylation of the Bcl-2 promoter, which diminished the protein expression of Bcl-2, thereby impairing lung function in a mouse model of emphysema. 18 Moreover, E2-mediated DNMT1 led to methylation of the Bcl-2 promoter in Japanese eel ovarian follicle, thereby decreasing the expression of Bcl-2. 28 Of note, a prior study revealed that increased Bcl-2 expression in mouse liver tissues following pretreatment with SDF-1 aided in the therapeutic functions of endometrial regenerative cells on ameliorating sepsis symptoms. 29 Upregulation of Bcl-2 in rat lung tissues brought about reduced production of inflammatory cytokines and cell apoptosis in mice with sepsis-induced lung injury. 30 Overall, lncRNA SNHG1 regulated the DNMT1/Bcl-2 axis to promote sepsis-induced myocardial injury.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that supports the findings of this study are available in this article and its supporting information files. Necrotic area/total area (%)