Downregulation of SRC-1 decreased cell invasion in HCC cells
We previously reported that SRC-1 facilitates HCC development via activating Wnt/β-catenin signaling [15]. However, it is unclear whether SRC-1 possesses function in cell metastasis in HCC. To evaluate the function of SRC-1 on cell invasiveness, we established SRC-1-knockdown in three HCC cell lines (HepG2, SK-Hep1 and MHCC97H) (Fig. 1A). Cell invasiveness was measured by Matrigel embedded trans-well assay. Comparing with cells that transfected with siCtrl RNA, the cell numbers that invaded through the Matrigel in SRC-1 transiently knockdown groups were decreased nearly 30% (Fig. 1B, 1C and 1D), suggesting that SRC-1 may involve in cell invasiveness.
Then we stably knocked down SRC-1 protein levels in HepG2 cells, as shown in Fig. 1E. Consistent with the result that SRC-1 transiently knockdown decreased cell invasiveness, SRC-1 stably knockdown cells showed a decreased invasion by nearly 40%, in comparison with control cells (Fig. 1F). To test whether SRC-1 could affect cell migration in HCC cells, cell migration capacity was detected via transwell assay without Matrigel embedded. As shown in Supplementary Fig. 1A-B, SRC-1 stable knockdown showed a lower migration rate than control cells, suggesting that SRC-1 downregulation decreased cell migration in HCC cells. These results indicated that SRC-1 can promote HCC cell metastasis.
Figure1. Downregulation of SRC-1 decreases HCC cells invasion. (A) Western blot of transient knockdown of SRC-1 in HepG2, SK-Hep-1 and MHCC97H cells. (B-D) Images and quantitation of SRC-1 transient knockdown and control cells infiltrate. All data are the means + sd. (E)the knockdown efficiency of SRC-1 was measured in HepG2 by Western Blot. (F) Images and quantitation of SRC-1 stable knockdown HepG2 cells penetrating through the matrigel-coated membrane.
Downregulation Of Src-1 Decreased Mmp-9 Expression
Since SRC-1 may promote HCC cell metastasis(invasion?), we were curious about the mechanism behind, previously our lab reported that SRC-3, another member of SRC family, can increase MMP-9 activity to enhance HCC cell migration and invasion(Xu, Chen et al. 2010, Liu, Tong et al. 2012). We wondered whether SRC-1 promotes HCC metastasis through enhancing MMP-9. To test this, we measured the mRNA levels of MMP-9 in SRC-1 transient knockdown cells and control cells. As shown in Fig. 2A, transient knockdown of SRC-1 significantly inhibited MMP-9 mRNA levels by nearly 40% in HepG2, SK-Hep1 and MHCC97H cells. Consistently, MMP-9 enzymatic activity, which was measured by zymography assay, was obviously reduced in SRC-1 transiently knockdown cells when treated with TPA. As shown in Fig. 2B, less MMP-9 protein (SK-Hep1 and MHCC97H) and even no MMP-9 protein (HepG2) from SRC-1 transiently knockdown cells can hardly digest the gelatin. These results showed that SRC-1 regulated MMP-9 mRNA levels and enzymatic activity. MMP-2, another important matrix metalloproteinase with a molecular weight of 72 kD, was not expressed in these three tested HCC cells (HepG2, SK-Hep1 and MHCC97H), as no 72 kD band could be seen from the zymography gel treated with or without TPA treatment (data not shown), suggesting that MMP-9 may be the main functional matrix metalloproteinase in these three tested HCC cells (HepG2, SK-Hep1 and MHCC97H). In SRC-1 stably knockdown HepG2 cell lines, MMP-9 mRNA levels and enzymatic activity were also significantly decreased compared with that in ctrl cells (Fig. 2C and 2D), in the presence or absence of TPA, confirming that SRC-1 can regulate MMP-9 expression.
To confirm that decreased invasion in SRC-1 knockdown cells were due to the loss of MMP-9, we restored the MMP-9 expression in HepG2 SRC-1 knockdown cells. As shown in Supplementary Fig. 2A, MMP-9 mRNA levels were significantly increased after stably transfected MMP-9 in SRC-1 knockdown HepG2 cells. Similarly, MMP-9 enzymatic activity was significantly increased after stably transfected MMP-9 in SRC-1 knockdown HepG2 cells (Supplementary Fig. 2B). Restoration of MMP-9 in SRC-1 knockdown HepG2 cells could rescued the invasiveness and migration rate (Fig. 2E and Supplementary Fig. 2C), indicating that SRC-1 increased HCC metastasis through enhancing MMP-9 expression. These results indicated that SRC-1 increased cell metastasis through inducing MMP-9 expression.
Our lab previously collected a set of human HCC patient specimens and measured the mRNA levels of SRC-1 and MMP-9 in the same set of human HCC specimens. As presented in Fig. 2F, a positive correlation was identified in human HCC specimens. In addition, In GEO database GSE20017, which contains a larger cohort number of HCC specimens, the positive correlation between SRC-1 and MMP-9 was also confirmed (Fig. 2G). These results indicated that SRC-1 may regulate MMP-9 expression in HCC cell lines and human HCC specimens.
Src-1 Regulated Mmp-9 Promoter Activity Through Coactivating Nf-κb And Ap-1
To test whether SRC-1 can regulate MMP-9 promoter activity, Luciferase assay was performed with the MMP-9 promoter reporter which is -670 bp to + 54bp away from transcriptional starting site. As shown in Fig. 3A, 12-O-tetradecanoylphorbol-13-Acetate (TPA) treatment, which is known to activate MMP-9 expression (Liang, Wu et al. 2009), can increase the MMP-9 promoter activity in HepG2 cells. Knockdown of SRC-1 observably reduced the MMP-9 promoter activity both in TPA treated and untreated group, suggesting that SRC-1 can regulate MMP-9 promoter activity.
SRC-1 has been reported to enhance MMP-9 transcription through coactivating PEA3[17] and Ets2[18]. To test the possibilities, PEA3, Ets2 and SRC-1 expression plasmids were co-transfected into the 293Tcells together with MMP-9 promoter reporter and Renilla luciferase plasmids. Co-overexpression of SRC-1 and PEA3 increased MMP-9 promoter activity by 3.9-fold, less than the additive effect of SRC-1 (2-fold) and PEA3 (2.8-fold) alone (Supplementary Fig. 3A). Co-overexpression of SRC-1 and Ets2 increased MMP-9 promoter activity by 5.6-fold, equals the additive effect of SRC-1 (2-fold) and PEA3 (3.6-fold) alone (Supplementary Fig. 3B). As co-overexpression of SRC-1, PEA3 and Ets2 did not dramatically induce MMP-9 promoter activity, suggesting that SRC-1 may regulate MMP-9 expression through coactivating other transcriptional factors in addition to PEA3 and Ets2.
NF-κB and AP1 are main transcriptional factors of MMP-9 (Sato and Seiki 1993). To test whether SRC-1 knockdown affected transcriptional activity of NF-κB and AP-1, NF-κB and AP-1 reporter assay were performed. SRC-1-knockdown significantly inhibited TPA-induced NF-κB and AP-1 transcriptional activity (Fig. 3B-C). And co-overexpression of SRC-1 and p65 or c-JUN synergistically increased the NF-κB and AP-1 transcriptional activity (Supplementary Fig. 3C-D). These results suggested that SRC-1 functioned as a coactivator for NF-κB and AP-1 in HCC cells.
To confirm that NF-κB and AP-1 can regulate MMP-9 promoter activity, we transiently knockdown NF-κB and AP-1 expression via transfecting siRNAs [24] that are specific for p65 and JunB into HepG2 cells and then detected MMP-9 mRNA levels and promoter activity. As shown in Fig. 3D-E, knockdown of NF-κB and JunB decreased the MMP-9 promoter activity both in TPA treated and untreated group. Consistently, the mRNA levels of MMP-9 were significantly decreased after knockdown of NF-κB (Fig. 3F) and AP-1 (Fig. 3G) when treated with TPA.
To further confirm that SRC-1 worked as a coactivator for NF-κB to enhance MMP-9 promoter, SRC-1 and NF-κB were overexpressed in 293T cells and MMP-9 promoter activity was measured. As shown in Fig. 3H, SRC-1-overexpression induced a 1.5-fold increase of MMP-9 promoter activity and overexpression of NF-κB induced a 3-fold increase of MMP-9 promoter activity. Synergistically, co-overexpression of SRC-1 and NF-κB significantly induced a 4.8-fold increase of MMP-9 promoter activity, indicating that SRC-1 coactivated NF-κB to enhance MMP-9 promoter activity. To further confirm that SRC-1 worked as a coactivator for AP-1 to enhance MMP-9 promoter, SRC-1 and AP-1 were overexpressed in 293T cells and MMP-9 promoter activity was measured. As shown in Fig. 3I, SRC-1-overexpression induced a 1.6-fold increase of MMP-9 promoter activity and overexpression of NF-κB induced a 5.3-fold increase of MMP-9 promoter activity. Synergistically, co-overexpression of SRC-1 and AP-1 significantly induced a 9.3-fold increase of MMP-9 promoter activity, indicating that SRC-1 coactivated AP-1 to enhance MMP-9 promoter activity. These results suggested that SRC-1 regulated MMP-9 promoter activity through coactivating NF-κB and AP-1.
MMP-9 gene expression is mainly regulated at the transcriptional level (Yan and Boyd 2007). The MMP-9 promoter contains multiple cis-elements that allow the regulation of MMP gene expression through a variety of trans-activators including PEA3, Ets2, NF-κB and AP-1. We summarized the binding motifs of PEA3, Ets2 and NF-κB and AP-1 in the core MMP-9 promoter region (-670 bp ~ + 54bp) according to the online tools JASPAR (Fornes, Castro-Mondragon et al. 2020) (http://jaspar.genereg.net/) and PROMO (Messeguer, Escudero et al. 2002, Farré, Roset et al. 2003) (http://alggen.lsi.upc.es/cgi-bin/promo_v3/promo/promoinit.cgi?dirDB=TF_8.3), as well as previous publications (Gum, Lengyel et al. 1996, Yan and Boyd 2007). At least one PEA3, Ets2, NF-kB and AP-1 binding motifs were identified in the core promoter region of MMP-9 (Supplementary Fig. 3E). SRC-1 coactivates PEA3, Ets2, NF-κB and AP-1 and simultaneously enhances MMP-9 promoter activity, suggesting that SRC-1 may enhance MMP-9 through coactivating PEA3, Ets2, NF-κB and AP-1 which are recruited to the promoter of MMP-9.
Src-1 Downregulation Decreased Cell Invasion In Hcc Cells In Vivo And Predict A Better Survival
Further, to inquiry the potential function of SRC-1 in HCC cell invasion in vivo, tumor metastasis was performed in BALB/c mice via tail vein injection of SRC-1-knockdown or control HepG2 cells. 21 days after injection, lungs were extracted for analysis. The number and size of tumors on lung surface in SRC-1-knockdown group were markedly less compared to those on control group (Figs. 4A-C).
Much less metastasis sites can be found in SRC-1 knockdown group (Fig. 4A). Large tumor nodule can be found in the ctrl group and only small tumor nodule can be observed in the SRC-1 knockdown group (Fig. 4B). The number of metastasis nodules was markedly reduced in SRC-1-knockdown group (Fig. 4C). These results suggested that SRC-1 knockdown decreased HCC cell invasiveness in vivo. To figure out the expression of SRC-1 in HCC primary tumor and circulating tumor cells (CTC), we analyses the GEO dataset GSE117623 with the online tool ctcRbase(Zhao, Wu et al. 2020) (http://www.origin-gene.cn/database/ctcRbase/index.html). As shown in Supplementary Fig. 4A, comparing with primary tumor, SRC-1 expression was significantly increased the circulating tumor cells. This result suggested that SRC-1 higher HCC cell have a higher ability to invade into the blood circulation system. In circulating tumor cells, MMP-9 expression was also significantly higher than that of primary tumor (Supplementary 4B), suggesting that MMP-9 higher HCC cell have a higher ability to invade into the blood circulation system.
To figure out the expression of SRC-1 in HCC primary tumor and HCC metastasis tumor, human HCC specimens were analyzed with the online tool HCMDB (Zheng, Ma et al. 2018) (http://hcmdb.i-sanger.com/index). In a cohort of HCC patients with lung metastasis or not, SRC-1 expression (Supplementary Fig. 4C) and MMP-9 expression (Supplementary Fig. 4D) expression were higher in lung metastasis tumors than that of primary HCC tumors. Similarly, in a cohort of HCC patients with adrenal gland metastasis or not, SRC-1 expression (Supplementary Fig. 4E) and MMP-9 expression (Supplementary Fig. 4F) were higher in adrenal gland metastasis tumors than that of primary HCC tumors. Due to the limited sample numbers in the above-mentioned human HCC cohort, there is no statistical significance in these comparisons. However, we observed a trend that comparing with primary tumor, SRC-1 and MMP-9 expression were increased in metastasis tumors. These results suggested that SRC-1 and MMP-9 higher HCC tumors may have more chance to metastasize to other organs.
To figure out whether SRC-1 expression correlated with HCC patient’s 5-year survival rate, we checked the database KM-plotter (Nagy, Lánczky et al. 2018) (http://kmplot.com/analysis/index.php?p=background), which supplies some survival data for some cancers including HCC. As shown in Fig. 4D, SRC-1 lower HCC patients have a better 5-year survival than SRC-1 higher HCC patients. Consistently, MMP-9 lower HCC patients have a better 5-year survival than MMP-9 higher HCC patents (Fig. 4E). These results suggested that SRC-1 downregulation correlated with a better survival.
Bufalin Treatment Inhibited Hcc Cell Invasiveness
As SRC-1 downregulation decreased HCC cell invasiveness both in vitro and in vivo, inhibition of SRC-1 should be an efficient means for controlling HCC metastasis. To test whether inhibition of SRC-1 reduced HCC cell invasiveness, we chose a well-known SRC-1 inhibitor, Bufalin, to treat the HCC cells and metastasis tumor mouse models. Inhibition of SRC-1 protein levels by Bufalin in HepG2 and SK-Hep-1 cells were confirmed by Western blot (Fig. 5A). And the results indicated that MMP-9 mRNA levels were significantly reduced after SRC-1 inhibitor treatment both in HepG2 (Fig. 5B) and SK-Hep-1 cells (Fig. 5C). Cell invasiveness was significantly decreased after Bufalin treatment in HepG2 cells (Fig. 5D), suggesting that SRC-1 inhibitor significantly decreased HCC invasion in vitro.
To test whether SRC-1 inhibitor decreased HCC cell metastasis in vivo, HepG2 cells were intravenously injected into nude mice and treated with Bufalin or vehicle. We can observe some metastasis sites in the lung of vehicle treated mice (Fig. 5E). However, much less numbers of metastasis sites can be seen from the lung of bufalin treated mice. Large tumor nodule can be found in the vehicle treated group and only small tumor nodule can be found in the Bufalin treated group (Fig. 5F). The total number of metastasis nodules were also significantly decreased in bufalin treated group (Fig. 5G). These results suggested that SRC-1 inhibitor treatment significantly decreased HCC cell metastasis in vivo.
Depletion Of Src-1 Decreases Mef Cells Invasion
To figure out whether SRC-1 affected cell invasiveness was specific to HCC cells or not, we isolated the MEF cells from SRC-1 wild type and SRC-1 knockout mice. SRC-1 knockout was confirmed by Western blot in MEF cells (Fig. 6A). Similarly, SRC-1 knockout in MEF cells obviously reduced MMP-9 mRNA levels (Fig. 6B). The cells that invade through Matrigel-embedded transwell were obviously reduced in SRC-1 knockout MEF cells (Fig. 6C). These results suggested that SRC-1 affected cell invasiveness was not only applied to HCC cells, but also can be effective in some other tissues.