Osteopontin promotes hepatocellular carcinoma progression through inducing JAK2/STAT3/NOX1-mediated ROS production

Abstract

Osteopontin (OPN) is a multifunctional cytokine that can impact cancer progression. Therefore, it is crucial to determine the key factors involved in the biological role of OPN for the development of treatment. Here, we investigated that OPN promoted hepatocellular carcinoma (HCC) cell proliferation and migration by increasing Reactive oxygen species (ROS) production and disclosed the underlying mechanism. Knockdown of OPN suppressed ROS production in vitro and in vivo, whereas treatment with human recombinant OPN produced the opposite effect. N-Acetyl-L-cysteine (NAC, ROS scavenger) partially blocked HCC cell proliferation and migration induced by OPN. Mechanistically, OPN induced ROS production in HCC cells by upregulating the expression of NADPH oxidase 1 (NOX1). NOX1 knockdown in HCC cells partially abrogated the cell proliferation and migration induced by OPN. Moreover, inhibition of JAK2/STAT3 phosphorylation effectively decreased the transcription of NOX1, upregulated by OPN. In addition, NOX1 overexpression increased JAK2 and STAT3 phosphorylation by increasing ROS production, creating a positive feedback loop for stimulating JAK2/STAT3 signaling induced by OPN. This study for the first time demonstrated that HCC cells utilized OPN to generate ROS for tumor progression, and disruption of OPN/NOX1 axis might be a promising therapeutic strategy for HCC.

Fig. 1. OPN-induced ROS production promotes the cell proliferation and migration of HCC cells.

The results of oxidoreductase activity and cell redox homeostasis in the tumor tissues of HCC patients (data from GEO data-set GSE76427) (n = 115) (A). The level of ROS in SMMC7721 cells after treatment with different concentrations of OPN recombinant protein (10 ng/mL, 100 ng/mL, 1 μg/mL) (B). CCK8 and transwell assays were performed to detect the viability and migration of SMMC7721 cells (C, E). The effects on cell viability and migration were analyzed in hOPN treated SMMC7721 cells after treating them with NAC (5 mM) for 2 h (D, F). The level of ROS in HCCLM3 cells, 72 h after infecting them with siOPN (G). The transwell assay and growth curve were performed to detect the proliferation and migration of HCCLM3 cells (H, I). Original magnification, ×100 (scale bars: 100 μm). Data indicated mean ± SD. n = 3, *P < 0.05, **P < 0.01, ***P < 0.001, # Compared with hOPN treatment.

Fig. 2. NOX1 is upregulated by OPN in HCC cells.

qRT-PCR was performed to examine the effect of hOPN treatment (1 μg/mL) and OPN knockdown on the mRNA levels of NOX family (A, B). Western blot was performed to examine the effect of hOPN treatment (10 ng/mL, 100 ng/mL, 1 μg/mL) and OPN knockdown on the level of NOX1 expression (C, D). Immunohistochemical analysis of 24 human HCC samples stained with OPN and NOX1 antibodies (E). Original magnification, ×200 (scale bar: 50 μm). The correlation between the IHC scores of OPN and NOX1 was analyzed (F). Data indicated mean ± SD. n = 3, *P < 0.05, **P < 0.01, ***P < 0.001, NS not significant.

Fig. 3. OPN stimulates ROS production and promotes cell proliferation and migration through upregulating NOX1 in HCC cells.

A, B DCFH-DA fluorescence was used to evaluate total ROS production in control HCC cell, hOPN treated (1 μg/mL) HCC cells, siNOX1 HCC cells and siNOX1 HCC cells treated with hOPN. Representative images of cell viability (C), and cell migration (E) assays using SMMC7721 control, hOPN treated, siNOX1 treated, and siNOX1+OPN treated SMMC7721 cells. All the experimental conditions in (C, E) were the same as in D, F, except SMMC7721 cells were used instead of HCCLM3 cells. Original magnification, ×100. Data indicated mean ± SD. n = 3, *P < 0.05, **P < 0.01, ***P < 0.001, # Compared with hOPN treatment, NS not significant.

Fig. 4. JAK2/STAT3 signaling plays a crucial role in OPN-induced NOX1 upregulation.

Western blot was performed to examine the effect of hOPN treatment (10 ng/mL, 100 ng/mL, 1 μg/mL) and OPN knockdown on the levels of pJAK2 and pSTAT3 expression (A, B). SMMC7721 cells were incubated with PBS or 1 μg/mL OPN and fluorescence-stained with STAT3 antibody and DAPI (C). The NOX1 promoter construct pGL3-NOX1 (−200/+1) was transiently co-transfected with pRL-SV40 into SMMC7721 cells treated with OPN or HCCLM3 cells transfected with siOPN. The dual-luciferase activity was determined (D, E). CHIP assays in SMMC7721 cells incubated with hOPN or PBS. Chromatin was immunoprecipitated with anti-STAT3 or anti-IgG, and the amounts of precipitated NOX1 promoter fragments were determined by qPCR using the specific primer sets (F). The levels of NOX1 protein were analyzed in hOPN treated SMMC7721 cells and hOPN treated SMMC7721 cells after treating them with FLLL32 (10 μM, 8 h) (H). Data indicated mean ± SD. n = 3, *P < 0.05, **P < 0.01, ***P < 0.001, # Compared with hOPN treatment.

Fig. 5. NOX1-induced ROS activity promotes the proliferation and migration of HCC cells.

Western blot analysis of NOX1 expression in 4 HCC cancer cell lines (A). SMMC7721 and HCCLM3 cells were transfected with siNOX1 or the negative control, as well as NOX1 overexpression plasmid or vector control. B, C growth curve and D, E transwell assays were performed to detect the proliferation and migration of SMMC7721 and HCCLM3 cells. DCFH-DA fluorescence was used to evaluate the total ROS production in two cell lines mentioned above (F–I). The levels of cell viability and migration were analyzed in NOX1 overexpressed HCC cells and NOX1 overexpressed HCC cells with NAC (5 mM) pretreatment for 2 h (J, K). Original magnification, ×100 (scale bars: 100 μm). Data indicated mean ± SD. n = 3, *P < 0.05, **P < 0.01, ***P < 0.001, # Compared with NOX1 overexpression.

Fig. 6. NOX1-induced ROS stimulates JAK2/STAT3 pathway and amplifies the effect of JAK2/STAT3 by OPN in HCCLM3 cells.

HCCLM3 cells were transfected with siNOX1 or negative control. A Western blot analysis of NOX1, pJAK2, and pSTAT3 expressions was performed after transfection. The levels of NOX1, pJAK2, and pSTAT3 proteins were analyzed after transfection with NOX1 overexpression plasmid or vector control (B). The levels of pJAK2 and pSTAT3 proteins were analyzed in NOX1 overexpressed HCCLM3 cells and NAC treated NOX1 overexpressed HCCLM3 cells (C). Western blot analysis was carried out for the levels of pJAK2 and pSTAT3 proteins of control HCCLM3 cell, hOPN treated HCCLM3 cells, siNOX1 HCCLM3 cells, and siNOX1 HCCLM3 cells treated with hOPN (D). Data indicated mean ± SD. n = 3, *P < 0.05, **P < 0.01, ***P < 0.001, # Compared with NOX1 overexpression or hOPN treatment.

Fig. 7. OPN depletion suppresses ROS production, NOX1 expression, and xenograft growth in vivo.

A–D Nude mice (ten per group) were injected with negative control or shOPN HCCLM3 cells. The mice were killed after 23 days of injection, and the tumors were excised and weighed (B). The volume of the tumors (C) were measured every other day for an indicated 23-day period. D The effect of OPN knockdown on the body weight of mice. E The level of ROS in the tumor tissue was assayed by DHE. F Western blot analyses of NOX1, pJAK2, and pSTAT3 expressions in the subcutaneous tumor (n = 6). G Immunohistochemistry of tumor tissues stained with OPN, NOX1, and Ki67 antibodies (n = 6). Quantification is on the right side. Original magnification, ×200 (scale bars: 50 μm). Data indicated mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. H Schematic representation of the role and mechanism of OPN-induced ROS in HCC.