Upregulated expression of miR-497 in HSCs is activated by C. sinensis ESPs and TGF-β1
To investigate the expression of miR-497 in the activated HSCs, the LX-2 cells were stimulated by TGF-β1 or ESPs of C. sinensis (CsESPs) for 6, 12, 24, and 48 h. The expression dynamics of miR-497 in LX-2 cells activated by TGF-β1 or CsESPs were detected by qRT-PCR (Fig. 1). The results showed that the expression of miR-497 in LX-2 cells stimulated by TGF-β1 was higher at 48 h than that in normal control cells (Fig. 1a, 48 h: t(5) = −5.42, P = 0.026). However, in CsESPs-stimulated HSCs, qRT-PCR data showed that miR-497 was significantly upregulated at 12 h compared with the normal control group (Fig. 1b, 12 h: t(5) = −4.706, P = 0.007). These data suggest that miR-497 was significantly upregulated in activated HSCs.
Smad7 is the target of miR-497
Bioinformatics analysis indicated that the 3′-UTR of Smad7 contained the binding site of miR-497, and the binding site is highly conserved across many different species (Fig. 2a). To further clarify whether miR-497 can inhibit the expression of Smad7 by binding to the 3′ UTR of Smad7, we constructed the luciferase reporter plasmid pmiRGLO with the 3′ UTR of Smad7 containing the miR-497 binding site on the 3′ UTR of Smad7 (WT plasmid), as well as luciferase reporter with Smad7 3′ UTR containing the miR-497 binding site mutation (MUT plasmid). The data showed that luciferase activity in the miR-497 mimic group was decreased compared with that in the no-load plasmid control group, with a statistically significant difference (Fig. 2b, NC vs. WT: F(3, 12) = 16.085, P = 0.0021). When MUT plasmid was co-transfected into 293T cells with miR-497 mimics, the activity of the reporter gene showed no difference from the non-load plasmid control group, but increased compared with the wild-type plasmid group, with a statistically significant difference (Fig. 2b, WT vs. Mut: F(3, 12) = 16.085, P < 0.001). To determine whether miR-497 has a regulatory role in the expression of Smad7, we transfected miR-497 mimic into LX-2 cells for 48 h, and a statistically significant increase in the relative expression level of miR-497 compared with the control group was observed (Fig. 2c, t(3) = 7.082, P = 0.0021), suggesting that miR-497 was successfully transfected into LX-2 cells. Furthermore, the levels of Smad7 mRNA transcript were decreased in LX-2 after treatment with the miR-497 mimic (Fig. 2d, t(3) = 4.437, P = 0.0091). Similarly, the level of Smad7 in the miR-497 mimic treated with LX-2 was lower than that in the scramble control-treated cells (Fig. 2e, t(3) = 3.192, P = 0.0332). Taken together, these data indicated that miR-497 directly inhibits the expression of Smad7 in HSCs.
Inhibition of miR-497 suppresses the TGF-β signaling pathway by targeting Smad7
To further verify whether miR-497 may be involved in TGF-β1-induced LX-2 cell activation by regulating the expression of Smad7, LX-2 cells were transfected with miR-497 inhibitor as well as its scramble and then stimulated by TGF-β1 for 48 h. The results showed that, compared with the scramble group, the relative expression of miR-497 was markedly reduced after transfection with miR-497 inhibitor, and the difference was statistically significant (Fig. 3a, t(3) = 5.766, P = 0.0045). Furthermore, the transfection of miR-497 inhibitor significantly increased the expression of Smad7, compared with the scrambled control (Fig. 3b, inhibitor + TGF-β1 vs. inhibitor NC + TGF-β1: F(3, 12) = 13.321, P = 0.002; inhibitor + TGF-β1 vs. control: F(3, 12) = 13.321, P = 0.017); subsequently, p-Smad2/3 protein in transfection of miR-497 inhibitor-HSCs was considerably decreased (Fig. 3b, control vs. inhibitor NC + TGF-β1: F(3, 12) = 13.321, P = 0.002; control vs. inhibitor + TGF-β1: F(3, 12) = 13.321, P = 0.017). Furthermore, the expression of collagen I (encoded by Col1a1) and α-SMA (encoded by Acta2) was also decreased in TGF-β1-stimulating LX-2 cells which were pretreated with miR-497 inhibitor, compared with that in non- or scramble-transfected LX-2 cells at both the mRNA and protein levels (Fig. 3c and Fig. 3d, Col1α1: F(3, 12) = 39.932 27.621, P = 0.001; Acta: F(3, 12) = 32 27.621, P = 0.001; α-SMA: F(3, 12) = 9.962, P = 0.006; C: F(3, 12) = 9.343, P = 0.013).
As our previous study showed that the expression of miR-497 was significantly increased in the liver of C. sinensis-infected mice and that CsESPs activated the TGF-β/Smad signaling pathway to promote activation of HSCs [7, 22], we further investigated whether miR-497 was involved in the CsESP activation of HSCs or not. The data showed that the specific miR-497 inhibitor significantly decreased the expression of α-SMA and COLI in the activation of HSCs induced by CsESPs compared with the scramble control-treated groups, indicating that miR-497 promoted the activation of HSCs induced by CsESPs (Fig. 3e, α-SMA: t(3) = 2.996, P = 0.041; COLI: t(3) = 3.227, P = 0.032). Collectively, these data demonstrate that the downregulation of miR-497 expression reduced the activation of HSCs via repressing the TGF-β/Smad signaling pathway.
Inhibition of miR-497 in mice reduced CCl4-induced liver damage
To investigate whether inhibition of miR-497 has a therapeutic effect on CCl4-induced liver damage in vivo, we injected carbon tetrachloride (CCl4) for 6 weeks to establish a mouse model of liver fibrosis, and the mice were treated with highly hepatotropic rAAV8 anti-miR-497 (anti-miR-497) or control rAAV8-scramble vectors (anti-SCR) with a single dose of 1 × 1012 virus or PBS by tail vein injection onset of CCl4 injection. Firstly, we found that the level of mature miR-497 in the liver of anti-miR-497 mice was remarkably decreased compared with that of the anti-SCR or PBS group (Fig. 4a, anti-miR-497 vs. anti-SCR: F(3, 18) = 7.959, P < 0.001; anti-miR-497 vs. PBS: P = 0.016). As demonstrated above, miR-497 can target Smad7 to regulate the expression of Smad7. Therefore, we further determined the level of Smad7 using qRT-PCR, which revealed a more than twofold increase in Smad7 expression in the liver of anti-miR-497-transfected mice compared with the anti-SCR mice (Fig. 4b, F(3, 18) = 7.008, P = 0.017).
H&E staining revealed disordered arrangement of the hepatic sinusoid, extensive hepatocellular degeneration and necrosis, and inflammatory cell infiltration in anti-SCR CCl4 mice; however, these histological changes were ameliorated after treatment with rAAV8-anti-miR-49 lentivirus (Fig. 4c), and pathology scores are statistically significant (Fig. 4c, control vs. anti-SCR: F(3, 18) = 30.892, P = 0.004; anti-SCR vs. anti-miR-497, F(3, 18) = 30.892, P = 0.013). The levels of the serum activities of ALT in the mice of the anti-miR-497 group were significantly reduced compared with those in the anti-SCR group, suggesting that hepatic damages caused by CCl4 were alleviated by inhibiting miR-497 (Fig. 4d, control vs. anti-SCR: F(3, 18) = 23.591; anti-SCR vs. anti-miR-497, F(3, 18) = 23.591, P < 0.001, P < 0.001).
Inhibition of miR-497 in mice reduced CCl4-induced liver fibrosis by targeting Smad7
Next, we investigated the protective roles of anti-miR-497 in liver fibrosis caused by CCl4. Masson staining showed obvious collagen deposition in the portal and sinusoidal areas in the CCl4 mouse model group; however, rAAV8 anti-miR-49 lentivirus resulted in a decrease in collagen deposition (Fig. 5a, control vs. anti-SCR: F(3, 12) = 21.489, P = 0.002; anti-SCR vs. anti-miR-497: F(3, 12) = 21.489, P = 0.048). We further detected collagen deposition in the liver using Sirius red staining. Similar to Masson staining, it showed that collagen deposition in the liver of rAAV8 anti-miR-49 lentivirus-treated mice was significantly decreased compared with that in scrambled lentivirus-treated (anti-SCR) mice (Fig. 5b, control vs. anti-SCR: F(3, 12) = 33.112, P < 0.001; anti-SCR vs. anti-miR-497: F(3, 12) = 33.112, P = 0.004). As hydroxyproline (HYP) content is a characteristic of liver fibrosis, we examined HYP levels in each group. The data showed a significant decrease in HYP content in the rAAV8-anti-miR-497-transfected mice that were subjected to CCl4 injection compared with the rAAV8-anti-SCR group (Fig. 5c, control vs. anti-SCR: F(3, 15) = 7.941, P = 0.002; anti-SCR vs. anti-miR-497: F(3, 15) = 7.941, P = 0.044). Furthermore, we examined the expression of an α-SMA-a marker of hepatic fibrosis by western blot; the data showed that the level of α-SMA was both significantly ameliorated in CCl4 mice treated with rAAV8-anti-miR-497, and Smad7 downstream p-Smad2/3 protein was also considerably decreased compared with rAAV8 anti-SCR (Fig. 5d, α-SMA: control vs. anti-SCR: F(3, 12) = 9.910, P = 0.002; anti-SCR vs. anti-miR-497: F(3, 12) = 9.910, P = 0.027; p-Smad2/3: control vs. anti-SCR: F(3, 12) = 9.910, P = 0.002; anti-SCR vs. anti-miR-497: F(3, 12) = 9.910, P = 0.004). We also examined other fibrosis markers including COLI and COLIII using western blot. The results showed that both were significantly decreased in CCl4 mice treated with rAAV8-anti-miR-497, compared with the mice that were treated with rAAV8 anti-SCR (Fig. 5d, COLI: control vs. anti-SCR: F(3, 12) = 11.922, P = 0.001; anti-SCR vs. anti-miR-497: F(3, 12) = 11.922, P = 0.018. COLIII: control vs. anti-SCR: F(3, 12) = 25.227, P = 0.001; anti-SCR vs. anti-miR-497: F(3, 12) = 25.227, P = 0.003). Collectively, these data suggest a protective role of inhibition of miR-497 in CCl4-induced liver fibrosis.