miR-455-3p ameliorates pancreatic acinar cell injury by targeting Slc2a1

Data source

A miRNA-seq dataset (GSE81260) and two mRNA expression datasets (GSE109227 and GSE161945) were obtained from the GEO database (Dixit et al., 2016; Norberg et al., 2018; Wang et al., 2021). GSE81260 contains 15 samples of mice (three AP samples and three control samples were selected). GSE109227 contains 11 samples of mice, including five controls and six AP samples. GSE161945 contains six samples of mice, including three controls and three AP samples.

Acquisition of differentially expressed genes (DEGs)

The differentially expressed miRNAs (DEmiRs) of GSE81260 and the DEGs of GSE109227 and GSE161945 were analyzed by Limma, and then the DEGs of GSE109227 and GSE161945 were intersected (Ritchie et al., 2015).

miRNA-mRNA regulatory network and functional pathway enrichment analysis of target genes

Target genes were predicted by miRWalk 3.0 (Huai et al., 2020). The predicted results were intersected with DEGs to acquire the differentially expressed target genes. The pairs with opposite expression trends were selected to establish miRNA-mRNA regulatory network, and were visualized using Cytoscape. Then, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis of the target genes were conducted using clusterProfiler to obtain functional pathways (Chen et al., 2022).

PPI network construction

The Search Tool for the Retrieval of Interacting Genes (STRING) database was applied to conduct PPI network analysis of the target genes, and hub genes were picked out by means of Maximal Clique Centrality (MCC), Edge Percolated Component (EPC) and Degree through cytohubba plug-in (Xie et al., 2022). Then, the regulatory relationship between miRNAs and hub genes was further analyzed, and their expression was also detected.

Construction of mouse AP model

The Laboratory Animal Center of Hangzhou Medical College provided SPF grade C57BL/6 male mice for experiments, and the experimental protocols were permitted by the Institutional Animal Care and Use Committee (IACUC), Zhejiang Center of Laboratory Animals (ZJCLA) (Approval No. ZJCLA-IACUC-20040130). Healthy 8-week-old C57BL/6 mice (20–22 g) were kept in an air-conditioned room with suitable room temperature and humidness. Before animal modeling, the mice were adapted to the facility for 1 week, and fed with commercial rodent pellets.

The C57BL/6 mice were equally and randomly grouped into control and AP groups, with six mice in each group. Before modeling, mice in both groups were fasted for more than 8 h, during which their access to drinking water was not restricted. The treatment methods of mice were as follows: Mice in the AP group were injected intraperitoneally with caerulein, repeated 6 times, 1 h each time, and the dose was 100 µg/kg. Mice in the control group were injected with normal saline, repeated 6 times at an interval of 1 h (Zhu, Liu & Wang, 2020).

Pathological analysis of AP mice

The experimental mice were euthanized through administration of pentobarbital and cervical dislocation, and the serum was collected for subsequent testing. The Olympus Au2700 system, the MODULAR P800 automatic biochemical analyzer, and their companion reagents were applied to measure the contents of amylase and lipase. After staining the pancreatic tissues with HE, two pathologists applied a microscope to evaluate the pathological changes in a blinded manner. The pancreatic edema, immune cell infiltration and acinar necrosis were also graded (0-3), and the pathological grading standards are shown in Table 1. Mice with each pathological grade >1 were used for analysis (Miao et al., 2019).

Isolation of pancreatic acinar cells

Mice were euthanatized, and the abdomen was disinfected with 75% alcohol. Laparotomy was performed to remove the intestinal tract and expose the pancreas and spleen. Then, the pancreas was taken out, and the blood vessels and adipose tissues were removed in precooled DMEM/F12 medium. The pancreas was cut open, and subsequently, the tissue fragments underwent a 30-minute digestion period with the collagenase digestion solution at 37 °C. During this process, the mixture was intermittently agitated every 5 min to ensure thorough digestion. Following the completion of digestion, the resulting digestive juice was transferred into a centrifuge tube, centrifuged at 1,000 rpm for 3 min, and the supernatant was abandoned. The pancreatic acinar cells were resuspended with HEPES buffer, subjected to gentle agitation and then precipitated. After standing for 30s, the buffer was filtered by 150 mesh filter and transferred to 45 mL centrifuge tube. This process was repeated 3 times. Cells were centrifuged at 500 rpm for 3 min, and the supernatant was discarded. Finally, the pancreatic acinar cells were resuspended with DMEM/F12 supplemented with 10% fetal bovine serum (FBS) in preparation for subsequent experiments (Norberg et al., 2018).

Transfection assay of pancreatic acinar cells

Pancreatic acinar cells were transfected with miRNA-NC, miR-455-3p mimics, miR-455-3p inhibitor, siNC and siSlc2a1 (Shanghai GenePharma Co., Ltd., Shanghai, China) using 0.25% Lipofectamine 2000 (Invitrogen, Waltham, MA, USA). Their transfection concentration is 60nM, and the transfection time is 24 h (Zhang et al., 2019). The sequences of the oligonucleotides were showed as follows: miR-455-3p mimics: GCAGUCCACGGGCAUAUACAC; miR-455-3p inhibitor: CGUCAGGUGCCCGUAUAUGUG.

Detection of cytokines

The contents of TNF-α, IL-1 β, and IL-6 in the serum and pancreatic acinar cell supernatant were analyzed using the ELISA kits (R & D Systems, Minneapolis, MN, USA) as per the instructions. Then, the protein secretion was calculated based on the standard curve.

Detection of intracellular trypsin

The pancreatic acinar cells were grouped and treated as follows: Control: no treatment. Caerulein: pancreatic acinar cells were induced with caerulein (1 × 10⁻⁷M) for 5 h (Zhang et al., 2022). Caerulein + miRNA-NC group: pancreatic acinar cells were first treated with miRNA-NC transfection, and then stimulated with caerulein. Caerulein + miR-455-3p mimics group: pancreatic acinar cells were first treated with miR-455-3p mimics transfection, and then stimulated with caerulein. After the cells were digested by trypsin, the liquid was centrifuged, and the cells precipitates were collected. Then, the pancreatic acinar cells were resuspended with HEPES buffer containing 2mM EDTA (HBS, 5mM HEPES, 0.15M NaCl, PH 7.35). The cell density was controlled at 1 × 10⁶ cells/mL (Zhang et al., 2019), and 10 µM Rhodamine 110, bis-(p-tosyl-L-glycyl-L-prolyl-L-arginine amide) (R22124; Molecular Probes, Eugene, OR, USA) was added for 20 min of incubation. The fluorescence intensity was detected by laser confocal microscopy and the activity of trypsin was analyzed.

Western blot assay

The RIPA lysis buffer was used to extract proteins from pancreatic acinar cells. The proteins underwent SDS-PAGE, and were transferred onto polyvinylidene fluoride (PVDF) membranes. Then, the membranes were sealed in skim milk for 2 h. The primary antibodies (1:1000) was applied to incubate proteins overnight. Next day, the secondary antibody was used to incubate the membrane for 2 h. The ECL kit was used to evaluate the chemical luminescence intensity. The primary antibodies used were Slc2a1 (73015S; Cell Signaling Technology (CST), Danvers, MA, USA) and β-actin (4970S; CST, Danvers, MA, USA).

qRT-PCR

Total RNA of the samples was obtained using the RNA extraction kit (R1200; Solarbio, Beijing, China). The cDNA used for qPCR was obtained by reverse transcription. Subsequently, the qPCR experiments were performed by the use of SYBR Premix Ex Taq II (RR820A; TaKaRa, Shiga, Japan) on a real-time PCR system. The primer sequences are showed as follows: miR-455-3p forward primer (F-primer): CGGCAGTCCACGGGCAT; reverse primer (R-primer): AGTGCAGGGTCCGAGGTATT; the internal control gene U6 F-primer: CTCGCTTCGGCAGCACA; R-primer: AACGCTTCACGAATTTGCGT. The miR-455-3p expression was calculated by 2^−ΔΔCt method.

Assay of amylase activity

The amylase activity in the supernatant was detected based on the instructions of EnzChek Ultra Amylase Assay Kit (E33651; Molecular Probes, Eugene, OR, USA). The relative fluorescence intensity of each well was detected by a microplate detector. The maximum absorption wavelength was set at 505nm and the maximum emission wavelength was set at 512 nm.

Dual-luciferase reporter assay

The pGL3-Slc2a1-3′-UTR (WT) or pGL3-Slc2a1 mut- 3′-UTR (MT) and miRNA-NC or miR-455-3p mimics were transfected into 293T cells, respectively. Following the instructions, the dual-luciferase reporter gene detection kit (FR201-01; Beijing Trans Gen Biotech, Beijing, China) was applied to calculate the relative luciferase activity. The sequences are as follows: Slc2a1-3′ UTR: TGTGTGTGTGCTTACAGAGT; Slc2a1-mut-3′ UTR: TCACATATACGTTACAGAGT.

Statistical analysis

GraphPad Prism 8 (GraphPad Inc., San Diego, CA, USA) was used to analyze the experimental data. Comparisons between groups were assessed by Student’s t test, and one-way ANOVA, followed by Bonferroni test for the difference comparison among multiple groups. Data was showed as mean ± SD. At least three biological replicates were performed for each assay. P < 0.05 was judged to be statistically significant.

Acquisition of DEGs

DEGs analysis of Limma (|logFC| > 1, P < 0.05) identified 19 DEmiRs in GSE81260, and there were 8 up-regulated miRNAs and 11 down-regulated miRNAs (Fig. 1A). 983 DEGs were acquired in GSE161945, of which 402 DEGs were up-regulated and 581 DEGs were down-regulated (Fig. 1B). Among 1608 DEGs found in GSE109227, 1304 were up-regulated and 304 were down-regulated (Fig. 1C). Through Wayne diagram analysis, 132 DEGs were found in GSE161945 and GSE109227 datasets (Fig. 1D).

Functional enrichment analysis (GO and KEGG)

The target genes predicted by miRWalk 3.0 were crossed with 132 DEGs, and 88 DEGs were preserved. Only the opposite relationship between miRNA and mRNA expression was retained. Finally, 110 targeted relationships between 17 miRNAs and 64 mRNAs were obtained (Fig. 2A). GO and KEGG were then used to analyze the functional enrichment of the target genes. It was found that the functions of the target genes were primarily enriched in tissue remodeling, cytoplasmic membrane, nutrient interaction of cells, and non-membrane transmembrane tyrosine kinase activity (Figs. 2B–2E).

PPI network analysis

The target genes were subjected to PPI network analysis by STRING, and a network with 58 nodes and 197 edges was obtained (Fig. 3A). 5 hub genes (Fyn, Gadd45a, Sdc1, Slc2a1 and Src) were identified by MCC, EPC and degree algorithms (Table 2). The regulatory relationship between miRNAs and 5 hub genes (Fig. 3B) were revealed. Meanwhile, their expression was analyzed. miR-3064-3p, miR-455-3p, miR-129-1-3p, miR-214-5p, miR-382-3p, and miR-433-3p were down-regulated in AP (Fig. 3C). In GSE109227 and GSE161945, except Fyn, the expressions of Gadd45a, Sdc1, Slc2a1 and Src were up-regulated in AP (Figs. 3D–3E). In this study, miR-455-3p/Slc2a1 was further studied.

Successful construction of mouse AP model

In order to identify whether the mouse AP model was successfully built, the pancreas tissues of C57BL/6 mice were taken for HE staining. The results presented that the pancreatic lobules of control mice were clear, without edema, bleeding, inflammatory cell infiltration and necrosis, and its overall pathological score was 0 (Fig. 4A). For AP mice, the pancreatic lobule space widened, the acinar edema, flake necrosis, and a great quantity of immune cell infiltration were noticed in the parenchyma and stroma, indicating that red blood cells were scattered in the pancreatic tissues. The overall pathological score was 6.81 ± 0.67 (Fig. 4A). Furthermore, the serum amylase and lipase contents of AP mice increased appreciably as compared with control mice (Fig. 4B). The contents of TNF-α, IL-1 β and IL-6 in AP mice were significantly higher than those of control mice in the serum (Fig. 4C).

miR-455-3p and Slc2a1 in AP mice

qRT-PCR was applied to detect the level of miR-455-3p, and western blot was used to detect Slc2a1 in the mouse pancreatic tissues (Figs. 5A–5B). In contrast to control mice, miR-455-3p of AP mice was significantly down-regulated in the pancreatic tissues, while Slc2a1 protein was significantly increased, which was consistent with the findings of bioinformatics analysis.

miR-455-3p attenuates caerulein-induced pancreatic acinar cell injury

miR-455-3p and Slc2a1 were detected through qRT-PCR and western blot (Figs. 6A–6B). The cell morphology of each group was observed under an optical microscope (Fig. 6C). When compared with cells in the control, the density of caerulein-induced cells was obviously down-regulated, the number of bright and transparent normal cells was reduced, the cells were seriously damaged, the shape was irregular, and there were more suspended dead cells and cell debris. Compared with the caerulein + miRNA-NC, the cell damage was improved in the caerulein + miR-455-3p mimics, the cell shape was relatively regular, and the proportion of dead cells and cell debris were significantly reduced. The cells were stained with Rhodamine 110 staining kit, and a laser confocal microscope was applied for observation (Fig. 6D). The positive region of trypsin (green fluorescence) in the caerulein-induced cells was obviously increased compared with the control. Compared to the caerulein + miRNA-NC, the trypsin activity in caerulein + miR-455-3p mimics group decreased. Amylase is another common marker of pancreatic diseases. Therefore, the amylase activity detection kit was applied to analyze the amylase in the cell culture supernatant (Fig. 6E). Compared to the control, the amylase activity in the cell culture supernatant of the caerulein group was increased (Fig. 6E). The amylase activity in the caerulein + miR-455-3p mimics was down-regulated than the caerulein + miRNA-NC group (Fig. 6E). The secretion level of inflammatory cytokines (IL-1 β, IL-6 and TNF-α) was analyzed by ELISA (Fig. 6F). The secretion level of inflammatory cytokines in the cell culture supernatant of caerulein-induced cells were obviously increased than the control, while substantially reduced in the caerulein + miR-455-3p mimics by contrast to the caerulein + miRNA-NC group.

miR-455-3p directly regulates Slc2a1

Based on bioinformatics analysis, the binding sites to the miR-455-3p sequence exist in the Slc2a1 3′ UTR region (Fig. 7A). In primary mouse pancreatic acinar cells, Slc2a1 expression was significantly reduced by miR-455-3p mimics but markedly enhanced by miR-455-3p inhibitor (Figs. 7B–7C). In addition, the luciferase reporter vector was applied to further identify the combination between miR-455-3p and Slc2a1 3′ UTR region. After co-transfection of miR-455-3p mimics and pGL3-Slc2a1-3′-UTR (WT) plasmids in 293T cells, it was found that the luciferase activity was significantly decreased (Fig. 7D). There is no significant difference between pGL3-Slc2a1 mut-3′-UTR (MT)/miRNA-NC and pGL3-Slc2a1 mut-3′-UTR (MT)/miR-455-3p mimics (Fig. 7D).

miR-455-3p targets Slc2a1 and alleviates the damage of mouse pancreatic acinar cells induced by caerulein

Combined with the results of bioinformatics analysis and cell experiments, it was found that miR-455-3p bound to the 3′ UTR region in Slc2a1, thereby regulating the transcript level of Slc2a1 (Figs. 8A–8B). For further verification, siNC, siSlc2a1, miRNA-NC and miR-455-3p inhibitor were used to transfect into primary mouse pancreatic acinar cells. To test whether miR-455-3p alleviates the damage of caerulein-induced mouse pancreatic acinar cells by regulating Slc2a1, cells were grouped into control group, caerulein group, caerulein + siNC group, caerulein + siSlc2a1 group, caerulein + miRNA-NC + siSlc2a1 group, and caerulein + miR-455-3p inhibitor + siSlc2a1 group. It was shown that the caerulein-induced damage of pancreatic acinar cells was improved in the caerulein + siSlc2a1 compared to the caerulein + siNC, and the trypsin and amylase activity and the secretion levels of inflammatory cytokines were decreased (Figs. 8C–8F). Compared to the caerulein + miRNA-NC + siSlc2a1 group, the cell damage in the caerulein + miR-455-3p inhibitor + siSlc2a1 was aggravated, and the trypsin and amylase activity and the secretion levels of inflammatory cytokines were increased (Figs. 8C–8F).