Patients undergoing cardiac valve replacement (The cases of mitral valve replacement were excluded) at Shanghai Chest Hospital (Shanghai, China) were joined in our research and were split into the AF group (n = 15) and sinus rhythm (SR) group (n = 13), according to preoperative electrocardiogram examination and medical history. Patients with persistent atrial fibrillation were included in our study. Patients with previous coronary atherosclerotic heart disease, chronic pulmonary heart disease, infective endocarditis, hyperthyroidism, severe dysfunction of liver and kidney, and malignant tumors were excluded. All patients in the preoperative period of 6 months without applying any type of angiotensin II receptor blockers and angiotensin converting enzyme inhibitors. Superior vena cava intubation was placed in the right auricle, and the right auricle extracted during the procedure was collected for this study. The right atrium was carefully cleaned with normal saline to remove blood, and the adipose tissue was carefully pruned and removed for use in this study. This research protocol was permitted by the ethics committee of Shanghai Chest Hospital, and written informed consent was obtained from each patient.

Potential NEAT1 and miR-320 binding sites were predicted using starBase v2.0 (http://starbase.sysu.edu.cn/starbase2/mirLncRNA.php), and miR-320 and NPAS2 binding sites were predicted using Targetscan (http://www.targetscan.org/vert_72/).

Total RNA was obtained from indicated cells or tissues with Trizol (Invitrogen, Carlsbad, CA, United States). The ratio of the optical density of RNA at 280 and 260 nm was measured by ultraviolet spectrophotometer, and the determination value was between 1.8 and2.0. RNAs were converted into cDNA using the RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, Waltham, MA, United States). qRT–PCR was carried out tharough a LightCycler 480 real-time PCR system using SYBR® Premix Ex Taq™ (Takara Bio, Inc., Dalian, China). GAPDH or U6 was functioned as the internal control. The 2^−ΔΔCt method was performed to analyze the relative expression level of each gene. The sequence of primers is shown in Table 1.

Atrial muscle tissues and cardiac fibroblasts were homogenized or lyzed in radioimmunoprecipitation assay buffer containing a cocktail of protease inhibitors (Santa Cruz, CA, United States) in accordance with the manufacturer’s instruction. Protein was quantified using the Bicinchoninic Acid (BCA) Protein Assay Kit and boiled for 10 min at 95°C. Total protein was obtained and loaded into 10% SDS-PAGE and transferred onto PVDF membranes. The primary antibodies used in our study are as follows: anti-coll I (Proteintech Group, Wuhan, China), anti-coll III (Proteintech Group, Wuhan, China), anti-NPAS2 (Thermo Fisher Scientific), and anti-GAPDH (Proteintech). Quantification was performed by measuring the signal intensity using ImageJ (National Institute of Health, Rockville, MD, United States).

Mouse cardiac fibroblasts were obtained from Ginio Biotechnology (Guangzhou, China), cultured in DMEM containing 10% FBS, and were transfected with the plasmids and stimulated with Ang II (1 μM).

The miR-320 mimic, miR-320 inhibitor, and their negative controls (NC mimic and NC inhibitor) were obtained from GenePharma (Shanghai, China). The sequences as follows: miR-320 mimic (sense: 5′-AAA​AGC​UGG​GUU​GAG​AGG​A-3′) or NC mimic (sense: 5′-UUC​UCC​GAA​CGU​GUC​ACG​UTT-3′); miR-320 inhibitor (sense: 5′- UCC​UCU​CAA​CCC​AGC​UUU​U-3′) or NC inhibitor (sense: 5′-CAG​UAC​UUU​UGU​GUA​GUA​CAA-3′). The coding region of the NPAS2 mRNA was cloned into the pcDNA3.1 (+) vector. The lentiviral vector expressing shRNA targeting NEAT1 was obtained from HANBIO (Shanghai, China). Short-hairpin RNA directed against NEAT1 was constructed in pLKO.1-puro vector to generate NEAT1 shRNA expression constructs, the non-targeting sequence (negative control, shNC) were also synthesized. The sequences of shNEAT1 and shNC were as follows: shNEAT1-F: ccg​gCA​GGA​CTA​GGT​GCG​TAG​TGc​tcg​agC​ACT​ACG​CAC​CTA​GTC​CTG​ttt​ttg and shNEAT1-R: aat​tca​aaa​aCA​GGA​CTA​GGT​GCG​TAG​TGc tcg​agC​ACT​ACG​CAC​CTA​GTC​CTG. Cell transfection was carried out with validated vector and lentivirus packaging vectors (pMD2G and pSAX2) using Lipofectamine 2000 (Invitrogen).

For the CCK-8 assay, transfected cells were plated into 96-well plates, and the medium of each well was replaced with culture media containing 10% CCK-8 at 72 h. The absorbance was measured using a microplate reader at an optical density of 450 nm.

The transfected cells were stimulated with or without Ang II. After that, indicated cells were cultured in the upper chamber using serum-free DMEM. Twenty-four hours later, non-migratory cells on top of the membrane were taken out, and membranes containing cells on the bottom were fixed and stained. The migratory cells were counted under a microscope.

Human embryonic kidney 293 (HEK293) T-cells were co-transfected with reporter plasmids including either the wild-type NPAS2 3ʹUTR (NPAS2-WT) or wild-type NEAT1 containing miR-320 binding site (NEAT1-WT) or mutated NPAS2 3ʹUTR (NPAS2-MT) or mutated NEAT1 (NEAT1-MT) and with either the miR-320 mimic or NC mimic. After 48 h inductions, the luciferase assay was conducted, and the relative luciferase activity was determined.

C57BL/6J mice (6–7 weeks old) were subdivided into the control, Ang II (Ang II, 1200 ng/kg/min, was continuously perfused to mice through a micropump), Ang II/shNEAT1 (shNEAT1 was injected into the mice via the tail vein after 28 days of Ang II stimulation), and Ang II/shNC (shNC was injected into the mice via the tail vein after 28 days of Ang II treatment) groups. After 14 days, right atrial muscle tissues were obtained for research.

Atrial muscle tissues obtained from different groups were fixed in 4% paraformaldehyde, embedded in paraffin and then sectioned into slices (4 μm thick). The slices were conducted to HE and Masson’s trichrome stainings. The photographs of the stained tissues were captured, and analyze the pathological changes using Image-Pro Plus (version 6.0; Media Cybernetics, Inc., Rockville, MD, United States).

All data analyses were carried out using Prism 5.0 (GraphPad Software, San Diego, CA, United States), and all data are presented as mean ± SD. The significance of the differences was determined using Student’s t-test. Correlation between factors was analyzed using spearman’s correlation coefficient rank test. p values < 0.05 were recognized to be significant. Each in vitro experiment was performed a minimum of three times, and samples were measured in biological triplicates for each experiment.

To explore whether NEAT1 was involved in AF progression, we first determined NEAT1 expression in right atrial tissues of AF patients and SR patients. The results of qRT–PCR observed that NEAT1 expression in right atrial tissues was higher in patients with AF than in those with SR (Figure 1A). Additionaly, we confirmed that coll I and coll III levels were upregulated in the AF group compared to those in the SR group (Figures 1B,C). Additionally, coll I and coll III levels were positively related to NEAT1 expression in right atrial tissues of AF patients (Figure 1D). These observations suggest that NEAT1 participates in atrial fibrosis regulation.

To explore the effect of NEAT1 on atrial fibrosis, a specific shRNA against NEAT1 gene transcript was designed to knock down NEAT1 in cardiac fibroblasts, and qRT–PCR analysis showed that the expression of NEAT1 in cardiac fibroblasts was reduced by this shRNA (Figure 2A). We found that Ang II increased NEAT1 expression, but this effect was attenuated by NEAT1 shRNA (Figure 2B). The CCK-8 assay showed that NEAT1 downregulation significantly repressed Ang II-induced cell proliferation compared with the shNC-transfected group (Figure 2C). The Transwell migration assay results showed that Ang II treatment promoted the migration ability of cardiac fibroblasts, and this promotion effect was suppressed by shNEAT1 (Figure 2D). Moreover, NEAT1 knockdown attenuated the promoted coll I and coll III expressions after Ang II treatment (Figures 2E,F). These findings revealed that NEAT1 knockdown could inhibit Ang II caused cardiac fibroblast proliferation, migration, and collagen production.

To determine the mechanisms through which NEAT1 exerts its effects on atrial fibrosis, we predicted miR-320 using relevant binding sites of NEAT1, and miR-320 interacted with NPAS2 mRNA 3ʹUTR using bioinformatics databases (Figure 3A). Furthermore, bioinformatics analysis predicted that NEAT1 and NPAS2 mRNAs have the same binding site for miR-320 (Figure 3A). The data of luciferase reporter assay revealed that luciferase activity was inhibited in NEAT1-WT- and miR-320 mimic-co-transfected cells but was unaffected in NEAT1-MT-transfected cells, indicating that miR-320 is a NEAT1-targeting miRNA (Figure 3B). Co-transfection with NPAS2-WT and miR-320 mimic obviously suppressed luciferase activity, whereas the luciferase activity has no changed in co-transfection with NPAS2-MT and miR-320 mimic (Figure 3C). Furthermore, we observed that miR-320 was downregulated in Ang II-treated cardiac fibroblasts, but both parameters were enhanced by NEAT1 deletion (Figure 3D). Ang II-treated cardiac fibroblasts increased NPAS2 expression, but this increase was reduced by miR-320 (Figures 3E,F). Furthermore, NEAT1 knockdown significantly suppressed NPAS2 expression in Ang II stimulated cardiac fibroblasts, and this effect was reversed by miR-320 inhibition (Figures 3G,H). These findings revealed that NEAT1 positively regulates NPAS2 expression by sponging miR-320 in cardiac fibroblasts. NPAS2 overexpression reversed the effects of NEAT1 knockdown on Ang II-induced cardiac fibroblast proliferation, migration, and collagen production.

To determine whether NPAS2 contributes to the effect of NEAT1 on cardiac fibroblasts, Ang II treated cardiac fibroblasts were co-transfected with shNEAT1 and NPAS2. The cardiac fibroblasts viability, migratory ability, and collagen production ability in co-transfected shNEAT1 and NPAS2 cells were greater than those in cardiac fibroblasts transfected with shNEAT1 (Figures 4A–F). These findings indicated that NPAS2 is a functional target of NEAT1 and that NPAS2 eliminates the suppressive effect of NEAT1 inhibitors on cardiac fibroblasts.

We further affirmed the function of NEAT1 in atrial fibrosis through in vivo experiments. HE and Masson’s trichrome stainings indicated a disarray of myocardial fibers, expanded nuclear spacing, and increased atrial fibrosis in the Ang II induced group, while NEAT1 deletion suppressed the Ang II caused inflammatory cell infiltration and atrial fibrosis (Figure 5A). We found that NEAT1 and NPAS2 expressions were increased, whereas miR-320 expression was decreased in right atrial tissues from the Ang II group than those in the control group, and these expressions were reduced in the Ang II/shNEAT1 group (Figures 5B–E). Moreover, Ang II injection enhanced the protein expressions of coll I and coll III in right atrial tissues, whereas deletion of NEAT1 attenuated the Ang II caused collagen production (Figure 5E).