Bone marrow-derived mesenchymal stem cells (BMSCs) repair acute necrotized pancreatitis by secreting microRNA-9 to target the NF-κB1/p50 gene in rats

Abstract

Acute pancreatitis (AP) is a common acute abdominal disease, 10-20% of which can evolve into severe AP (SAP) causing significant morbidity and mortality. Bone marrow-derived mesenchymal stem cells (BMSCs) have the potential of repairing SAP, but the detailed mechanism remains unknown. We demonstrate here that microRNA-9 (miR-9) modified BMSCs (pri-miR-9-BMSCs) can significantly reduce the pancreatic edema, infiltration, hemorrhage, necrosis, the release of amylase and lipase. Meanwhile, decreased local/systemic inflammatory response (TNF-α↓, IL-1β↓, IL-6↓, HMGB1↓, MPO↓, CD68↓, IL-4↑, IL-10↑, and TGF-β↑) and enhanced regeneration of damaged pancreas (Reg4↑, PTF1↑, and PDX1↑) are also promoted. But these effects diminish or disappear after antagonizing miR-9 (TuD). Besides, we find that miR-9 is negatively correlated with AP and miR-9 agomir which can mimic the effects of pri-miR-9-BMSCs and protect injured pancreas. Furthermore, we investigate that BMSCs deliver miR-9 to the injured pancreas or peripheral blood mononuclear cell (PBMC), which can target the NF-κB1/p50 gene and inhibit the NF-κB signaling pathway (p-P65↓, NF-κB1/p50↓, IκBα↑, IκBβ↑). Taken together, these results show that miR-9 is a key paracrine factor of BMSCs attenuating SAP targeting the NF-κB1/p50 gene and suppressing the NF-κB signaling pathway.

Figure 1

pri-miR-9-BMSCs could attenuate SAP. (A) BMSCs infected by pri-miR-9- and Empty- virus were expressing the Green Fluorescent Protein (GFP). (B) The identification of recombinant pri-miR-9-1-PCDH-CMV-MSCs-EF1-GFP-T2A-Puro plasmid (pri-miR-9-1-PCDH) by double enzyme digestion and a 368 bp DNA fragment (pri-miR-9-1) was shown as red arrow. (C) pri-miR-9-1 was amplified from pri-miR-9-1-PCDH by applying the special primers. (D and E) The expression of mature miR-9 in pri-miR-9-BMSCs was higher than that in Empty virus-BMSCs by gPCR and qRT-PCR. Data are shown as mean ± SD for at least 3 separate experiments. ^###p < 0.001, compared with Empty virus-BMSCs by paired t test. (F,H,K,G) pri-miR-9-BMSCs could markedly reduce the pancreatic edema, infiltration, hemorrhage and necrosis, decrease the levels of serum amylase, lipase and pro-inflammatory cytokines (IL-1β, IL-6, TNF-α, HMBG1, and MPO) and increase the levels of serum anti-inflammatory cytokines (IL-4, IL-10, and TGF-β), compared with SAP, SAP+PBS, BMSCs, Empty virus-BMSCs, or TuD-BMSCs groups. (M,W) The cell apoptosis was significantly reduced by pri-miR-9-BMSCs, compared with SAP, SAP+PBS, BMSCs, Empty virus-BMSCs, or TuD-BMSCs groups. (P and Q) The expression of miR-9 in damaged pancreas was significantly increased by pri-miR-9-BMSCs, compared with NC, SAP, SAP+PBS, BMSCs, or Empty virus-BMSCs groups. Data are shown as mean ± SD for at least 3 separate experiments. ^%P < 0.05, ^%%p < 0.01, ^%%%p < 0.001, compared with NC, ^**p < 0.01 and ^***p < 0.001, compared with NC, ^@@p < 0.01 and ^@@@p < 0.001, compared with SAP, ^&&p < 0.01 and ^&&&p < 0.001, compared with PBS treatment (SAP+PBS), ^#p < 0.05, ^##p < 0.01 and ^###p < 0.001, compared with BMSCs, ^$p < 0.05, ^$$p < 0.01 and ^$$$p < 0.001, compared with Empty virus-BMSCs, ⁺p < 0.05 and ⁺⁺⁺p < 0.001, compared with TuD-BMSCs by using two-tailed t test. (X) GFP-BMSCs could deliver exogenous Cy3-miR-9a-5p to the liver, spleen, lung and pancreas of SAP rats, of which the number was observed more in the liver and spleen. gPCR, General PCR, qRT-PCR, quantitative Real-time PCR, SAP, severe acute pancreatitis.

Figure 2

miR-9 could alleviate SAP. (A,B,C) miR-9a-5p agomir could significantly reduce the pancreatic edema, infiltration, hemorrhage and necrosis, the release of serum amylase, lipase and pro-inflammatory cytokines (IL-1β, IL-6, TNF-α, HMBG1, and MPO) and elevate the levels of anti-inflammatory cytokines (IL-4, IL-10, and TGF-β), compared with SAP+PBS or miR-9a-5p control groups. (D,E) The cell apoptosis was significantly decreased by miR-9a-5p agomir, compared with SAP+PBS or miR-9a-5p control groups. (F,G) The expression of miR-9 in damaged pancreatic tissues was significantly up-regulated by miR-9a-5p agomir, compared with SAP+PBS or miR-9a-5p control groups. Data are shown as mean ± SD for at least 3 separate experiments. ^%P < 0.05, ^%%P < 0.01, ^§§p < 0.01 and ^%%%P < 0.001, compared with NC, ^***p < 0.001, compared with NC, ^@@p < 0.01 and ^@@@p < 0.001, compared with SAP+PBS group, ^##p < 0.01 and ^###p < 0.001, compared with miR-9a-5p control group by using two-tailed t test. (H,K) The results of in-situ hybridization showed that the expressions of miR-9 in damaged pancreas were significantly increased by pri-miR-9-BMSCs or miR-9a-5p agomir, compared with SAP, SAP+PBS, BMSCs, TuD-BMSCs, or miR-9a-5p control groups. Data are shown as mean ± SD for at least 3 separate experiments. ^**p < 0.01, compared with NC, ^&p < 0.05, compared with SAP, ^@@p < 0.01, compared with SAP+PBS, ^†p < 0.05, compared with BMSCs, ⁺⁺p < 0.01, compared with Empty virus-BMSCs, ^$p < 0.05, compared with TuD-BMSCs, ^##p < 0.01, compared with miR-9a-5p control by using two tailed t test.

Figure 3

pri-miR-9-BMSCs could inhibit the local inflammatory response. The expressions of pro-inflammatory cytokines(IL-1β, IL-6, TNF-α, and MPO), NF-κB proteins (p-P65, P50) and CD68 were significantly decreased by pir-miR-9-BMSCs. On the contrary, the expressions of superoxide dismutase (SOD₁ and SOD₂) and anti-inflammatory proteins (IκBα and IκBβ) were significantly increased by pir-miR-9-BMSCs, compared with SAP, SAP+PBS, BMSCs, or TuD-BMSCs groups by gPCR (A,B), Western-blot (C,D) and IHC (E,F). Data are shown as mean ± SD for at least 3 separate experiments. ^%%P < 0.01, compared with NC, ^#p < 0.05, ^##p < 0.01 and ^###p < 0.001, compared with NC, ⁺p < 0.05, ⁺⁺p < 0.01 and ⁺⁺⁺p < 0.001, compared with SAP, ^*p < 0.05 and ^**p < 0.01, compared with SAP+PBS, ^@p < 0.05 and ^@@p < 0.01, compared with BMSCs, ^$p < 0.05 and ^$$p < 0.01, compared with Empty-virus BMSCs, ^†p < 0.05, ^††p < 0.01 and ^†††p < 0.001, compared with TuD-BMSCs by using paired t test. gPCR, General PCR, IHC, immunohistochemistry.

Figure 4

miR-9 could antagonize the local inflammatory response. The expressions of IL-1β, IL-6, TNF-α, MPO, p-P65, P50, and CD68 were significantly decreased by miR-9a-5p agomir. Inversely, the expressions of SOD₁, SOD₂, IκBα, and IκBβ were significantly increased by miR-9a-5p agomir, compared with SAP+PBS or miR-9a-5p control group, by gPCR (A,B), Western-blot (C,D) and IHC (E,F). Data are shown as mean ± SD for at least 3 separate experiments. ^%P < 0.05, compared with NC, ^#p < 0.05 and ^###p < 0.001, compared with NC, ^*p < 0.05, ^**p < 0.01 and ^***p < 0.001, compared with SAP+PBS, ^&p < 0.05, ^&&p < 0.01, and ^&&&p < 0.001, compared with miR-9a-5p control by using paired t test. SOD, Superoxide Dismutase, gPCR, General PCR, IHC, immunohistochemistry.

Figure 5

pri-miR-9-BMSCs and miR-9a-5p agomir promote the regeneration of damaged pancreas mainly depending on the paracrine. The expressions of pancreatic regenerative proteins (Reg4, PTF1, and PDX1) were significantly promoted by pri-miR-9-BMSCs (A,B) or miR-9a-5p agomir (C,D), compared with SAP, SAP+PBS, BMSCs, TuD-BMSCs, or miR-9a-5p control groups. Data are shown as mean ± SD for at least 3 separate experiments. ^#p < 0.05 and ^##p < 0.01, compared with NC, ⁺p < 0.05, compared with SAP, ^*p < 0.05 and **p < 0.01, compared with SAP+PBS, ^@p < 0.01, compared with BMSCs, ^$p < 0.05, compared with Empty virus-BMSCs, ^†p < 0.05 and ^††p < 0.01, compared with TuD-BMSCs, ^&p < 0.05, compared with miR-9a-5p control group. (E,F) The distributions of CM-Dil- and SPION- labeled BMSCs in vivo were observed by the fluorescence microscope and Prussian blue staining respectively. The number of cells migrating to the damaged pancreas had no difference among BMSCs, pri-miR-9-BMSCs, Empty-virus-BMSCs and TuD-BMSCs groups, but less than those of migrating to lung at day 3 after transplantation. Data are shown as mean ± SD. ^***p < 0.001, compared with pancreas by using two-tailed t test. SPION, Superparamagnetic Iron Oxide Nanoparticles.

Figure 6

BMSCs could transfer exogenous miR-9 to PBMC, which inhibited the expression of NF-κB1/p50 gene. (A) Eight paired base between miR-9 and NF-κB1/p50. (B) The structure of dual luciferase reporter vector-psiCHECK-2 (Promega, Beijing, China) and the wide-type NF-κB1 3′UTR (wtUTR) harboring the predicted binding sites of miR-9 and the mutant NF-κB1 3′UTR (CAAAG → TGCGA) (mutUTR) were cloned into psiCHECK-2 to generate the recombinant vectors of wtUTR- and mutUTR- psiCHECK-2. (C) The recombinant plasmids of wtUTR-/mutUTR-psiCHECK-2 were identified by double enzyme digestion. (D) The sequence of NF-κB1 3′UTR (319 bp) was amplified by gPCR from wtUTR- and mutUTR- psiCHECK-2. (E) miR-9a-5p mimics could reduce the activity of firefly luciferase with a dose-effect relationship. Data are shown as mean ± SD for at least 3 separate experiments. ^*p < 0.05, compared with control (con), ⁺p < 0.05, compared with 50 nM, ^#p < 0.05, compared with con, ^%p < 0.05, compared with 100 nM, ^&p < 0.05, compared with con by using paired t test. (F) The activity of firefly luciferase could be significantly repressed by miR-9a-5p mimics, but rescued by the mutation of NF-κB1 3′UTR. ^#p < 0.05 and ^@p > 0.05, compared with miR-9a-5p control by using paired t test. (G) The miR-9’s inhibitory effect on the activity of firefly luciferase disappeared after co-transfection with TuD plasmid. Data are shown as mean ± SD for at least 3 separate experiments. ^##p < 0.01, compared with miR-9a-5p control. ^$$p < 0.01, compared with miR-9a-5p mimics + TuD plasmid, ^@@p < 0.01, compared with mutUTR plasmid by using paired t test. (L) The GFP-BMSCs (Green) of Cy3-miR-9a-5p mimics (Red) transfection were co-cultured with PBMC (Gray) and 16 hr later, the exogenous Cy3-miR-9a-5p could be observed to be transferred into PBMC as indicated by the red arrow. (M,N,P,Q,K) miR-9a-5p mimics could inhibit the expression of NF-κB1/p50 and reduce the activity of firefly luciferase in LPS-activated PBMC. Data are shown as mean ± SD for at least 3 separate experiments. ^%P < 0.05 and ^%%P < 0.01, compared with NC, ^#p < 0.05, ^##p < 0.01 and ^##p < 0.01, compared with NC, ^*p < 0.05 and ^**p < 0.01, compared with LPS, ^@p < 0.05, ^@@p < 0.01 and ^@@@p < 0.001, compared with miR-9a-5p control by using paired t test.

Figure 7

The expression of miR-9 in injured pancreas was negatively correlated with the severity of AP. Interestingly, compared with Caerulein, the expression of miR-9 was significantly up-regulated by 3% NaT. (A,B) Rat AP models were induced by Caerulein and 3% NaT and the H&E staining showed that the injury of pancreas in 3% NaT group was more severe than that in Caerulein group. Data are shown as mean ± SD for at least 3 separate experiments. ^***p < 0.001 and ^###p < 0.001, compared with NC, ^$$$p < 0.001, compared with Caerulein by using two tailed t test. (C,D,E) The expressions of miR-9 in damaged pancreatic tissues and serum in 3% NaT group were significantly higher than that in Caerulein group, but lower than that in NC or Sham groups. Data are shown as mean ± SD. ^&&&P < 0.001 and ^##p < 0.01, compared with NC, ^*p < 0.05, compared with NC, ^%%%p < 0.001, compared with Sham, ^$$p < 0.01, compared with Caerulein group by using paired t test. (F) The expressions of miR-9 in damaged pancreas and serum were negatively correlated with pathological scores of AP by Pearson correlation analysis (p < 0.05). AP, acute pancreatitis, H&E, hematoxylin eosin, NC, normal control, NaT, sodium taurocholate, SD, standard deviation, miR-9, microRNA-9. (G) miR-9, produced by infused BMSCs, can target NF-κB1/p50 gene and suppress the activation of NF-κB signaling pathway in PBMC/Macrophage to reduce the release of the pro-inflammatory cytokines and prevent the occurrence of SIRS and MODS, which can promote the repair and regeneration of necrotized pancreatic tissues.