SMYD3 is overexpressed in sorafenib-resistant tumors and cells.
Establishment of different sorafenib-resistant HCC cell lines has previously been successful in vitro and in vivo [4]. These cells were found to exhibit enhanced anti-apoptosis ability, slow cell-cycle and liver CSC properties, including enhanced tumorigenecity, self-renewal ability, and migration and invasion ability in vitro, by Annexin V staining, IP staining, colony and sphere formation assay, and migration and invasion assay (Fig. 1a) [4]. These cells exhibited enhanced metastatic potential, including liver and lung metastasis, upon long-term exposure to sorafenib in vivo (Fig. 1b) [4]. These results are in line with the finding that label-retaining liver cancer cells with resistance to sorafenib [5].
In order to study the role of SMYD3 in the treatment of sorafenib resistance in liver cancer, we firstly detected the expression of SMYD3 in sorafenib-resistant cell lines. Western blotting revealed that SMYD3 protein was highly expressed in sorafenib-resistant cell lines (Fig. 1c). Immunofluorescence assay further confirmed that the expression of SMYD3 in Huh7-derived sorafenib-resistant cells was significantly increased (Fig. 1c). Although it was also expressed in the nucleus, it was mainly enriched in the cytoplasm (Fig. 1c). The quantitative RT-PCR (qRT-PCR) method further confirmed the expression of SMYD3 at the transcription level was also significantly increased (Fig. 1c). Next, the SMYD3 protein level was detected by western blotting and immunohistochemistry in sorafenib-resistant tumors, and it was found that compared with control tumors, the SMYD3 protein level in sorafenib-resistant tumors was significantly increased (Fig. 1d). And qRT-PCR method further confirmed that the mRNA level of SMYD3 in sorafenib-resistant tissues was also significantly increased (Fig. 1d). Finally, we analyzed the liver cancer samples in the TCGA database and selected 3 groups: the group that did not receive chemotherapy drugs, the group that received sorafenib or other drug treatment groups, and compared the sequencing results of SMYD3 to find that compared with the untreated group, the expression of SMYD3 from drug-treatment groups was significantly higher, but there was no significant difference between sorafenib and other treatment groups (Fig. 1e). Therefore, SMYD3 is overexpressed in sorafenib-resistant cells and tumors.
Sorafenib-resistant cells exhibit widespread up-regulation of multiple cancer-promoting genes mediated by SMYD3.
In the progression of liver cancer, SMYD3 invades active chromatin domains via association with H3K4Me3 and RNA Pol-II, thereby stimulating the transcription of key genes in cancer-related pathways [24]. These genes include: JAK1, JAK2, SNAIL1, TWIST1, ZEB1, SOX4, MYC, IGFBP1, PCNA, CCNE1 and CCNA2 [24]. Therefore, as a first step toward understanding the significance of targeted SMYD3 therapy, we measured the expression of these genes in sorafenib-resistant cells and tumors by qRT-PCR. We found that a large majority of sorafenib-resistant samples exhibited a significant up-regulation the expression of these oncogenes compared with control samples (Fig. 2a). For example, except for IGFBP1 gene in Huh7-Res, JAK2, SOX4, SNAIL1, CCNE1, IGFBP1, and PCNA gene in 7404-Res cells, TWIST1 gene in 7721-Res cells, CCNA2 in SK-Hep1-Res cells, all other resistant cells expressed significantly higher levels of JAK1, JAK2, SNAIL1, TWIST1, ZEB1, SOX4, and MYC gene compared with parental cells (Fig. 2a). In sorafenib-resistant tumors derived from Huh7 cells, JAK1, SNAIL1, TWIST1, ZEB1, SOX4, and MYC gene were significantly higher compared with DMSO-treated tumors (Fig. 2b). However, JAK2, SOX4, IGFBP1, PCNA, CCNE1, and CCNA2 gene were not significantly up-regulated in sorafenib-resistant tumors, but some tended to be up-regulated (Fig. 2b). Therefore, sorafenib-resistant cells exhibit widespread up-regulation of multiple cancer-promoting genes in vitro and in vivo.
Genetic and pharmacological inhibition of SMYD3 suppresses the expression of multiple cancer-promoting genes and activation of their associated signaling pathways in sorafenib-resistant cells.
Next, we investigate whether SMYD3 regulates the expression of these cancer-promoting genes. We conducted loss-of-function analysis of SMYD3 in vitro, and achieved the stable knockdown of SMYD3 in Huh7-Res and 7721-Res cells with lentivirus-mediated-shRNA against SMYD3 (shSMYD3). Two different shSMYD3, labeled shSMYD3-#1 and shSMYD3-#2, effectively suppressed the expression of SMYD3 protein and mRNA levels (Fig. 3a and b). After silencing SMYD3 or inhibiting the activity of SMYD3 with a small molecule inhibitor BCL121 [29], qRT-PCR was used to detect the effect on the transcription level of these cancer-promoting genes, and it was found that both significantly reduced the expression of their mRNA (Fig. 3a and c). Western blotting assay was used to detect the effect on their proteins and found that SMYD3 depletion significantly reduced the protein contents of these genes in sorafenib-resistant cell lines, thereby inhibiting the phosphorylation of STAT3 and N-cadherin protein, as well as the increase of E-cadherin protein, suggesting that SMYD3 depletion inhibited cell stemness, activation of JAK/STAT3 and EMT pathway (Fig. 3b and d). However, the synergy of the combined use of BCL121 and sorafenib is not significant in the transcriptional level, but it is significant in the protein level (Fig. 3c and d). Therefore, these findings indicated that SMYD3 depletion inhibited the expression of these oncogenes and the activation of related signal pathways. Moreover, BCL121 also played a similar inhibitory effect on the expression of SMYD3 target genes, indicating that this effect requires the participation of SMYD3 enzyme activity.
SMYD3 epigenetically activated the expression of multiple cancer-promoting genes and is dependent on its HMT activity.
In order to further confirm the role of SMYD3 in the regulation of these genes, we conducted gain-of-function analysis of SMYD3 in vitro in sorafenib-resistant cells with SMYD3 knockdown. We constructed wild-type SMYD3 (SMYD3) and catalytic dead mutant SMYD3-Y239F (SMYD3-M) in the pcDNA3.1-flag vector, and transferred wild-type SMYD3 or mutant SMYD3 to sorafenib-resistant cells with SMYD3 knockdown, and studied the effect on the transcription level of target genes. Because knockdown of SMYD3 was more prominent with shSMYD3-#1 than shSMYD3-#2, we used shSMYD3-#1 for most of subsequent experiments. Quantitative RT-PCR showed that the expression of wild-type SMYD3 complemented down-regulated the expression of multiple cancer-promoting factors caused by SMYD3 silencing, but not mutant SMYD3 (Fig. 4a). This result suggested that SMYD3 was necessary for the regulation of multiple cancer-promoting genes and dependent on its enzyme activity. In order to further evaluate the mechanism of SMYD3 in the regulation of these genes, chromotin immunoprecipitation (ChIP) assay was performed to detect the recruitment of SMYD3 and RNA polymerase II in the promoter regions of these oncogenes and the modification of histone H3K4me3 in sorafenib-resistant cells with SMYD3 silence or inhibition. JAK1, SOX4, and ZEB1 gene were selected as representatives for further research. It was found that the recruitment of SMYD3 and RNA polymerase II in the promoter regions of JAK1, SOX4, and ZEB1 gene was reduced, and histone H3K4Me3 modification was also significantly reduced by SMYD3 depletion (Fig. 4b and c). Secondly, we transferred wild-type SMYD3 and enzyme-mutated SMYD3 into sorafenib-resistant cells with SMYD3 knockdown, and performed ChIP assay with flag antibody, and found that both wild-type and mutant SMYD3 can bind to the promoter regions of JAK1, SOX4, and ZEB1 gene (Fig. 4d). However, compared with the mutant SMYD3, the recruitment of RNA polymerase II and histone H3K4me3 modification on the detected promoters were significantly increased in cells expressing wild-type SMYD3 (Fig. 4d). These results showed that SMYD3 epigenetically regulated the expression of these cancer-promoting genes via histone H3K4me3 modification.
Genetic and pharmacological inhibition of SMYD3 suppresses tumorigenecity, self-renewal ability, and migration and invasion ability in HMT-dependent manner.
Drug-resistant cells have more cancer stem cells (CSCs), resulting in cancer relapse and metastasis. SMYD3-mediated cancer-promoting genes and their associated pathway directly or indirectly involved in liver CSCs. After confirming that both genetic and pharmacological inhibition of SMYD3 resulted in suppressing JAK/STAT3, SOX4, MYC, and EMT pathway, we tested if the treatment of sorafenib-resistant HCC cells with SMYD3 knockdown or inhibition result in decreasing these pathways mediated tumorigenecity, self-renewal, and migration and invasion ability. Regarding the relative self-renewal potential of sorafenib-resistant cells, silencing SMYD3 significantly decreased colony and sphere formation frequencies than controls (Fig. 5a). Consistent with this finding, limiting dilution analysis showed that sorafenib-resistant cells with shSMYD3 required more cells and longer incubation times to generate tumors in vivo as compared those with nonspecific shRNA, indicating that sorafenib-resistant cells with shSMYD3 were significantly less enriched in CSCs (Fig. 5b). To confirm the results obtained using SMYD3 shRNA, we treated sorafenib-resistant cells with SMYD3 inhibitor BCL121 and measured tumorigenecity and self-renewal ability. In agreement with the results observed with SMYD3 knockdown, SMYD3 inhibitor also resulted in decreasing colony and sphere formation frequencies, and the combination with sorafenib also has a significant synergistic effect (Fig. 5c). We further recovered SMYD3 expression to rescue the weakening of the stemness of sorafnieb-resistant cells caused by SMYD3 knockdown, and found that overexpression of wide-type SMYD3 significantly restored tumorigenecity and self-renewal ability of sorafenib-resistant cells, but not mutant SMYD3(Fig. 5d). Regarding the relative migration and invasion ability of sorafenib-resistant cells, we found that SMYD3 knockdown or SMYD3 inhibition by BCL121 significantly inhibited migration and invasion ability of sorafenib-resistant cells using transwell assay with or without matrigel (Fig. 6a-d). What’s more, overexpression of wild-type SMYD3 significantly restored the migration and invasion capabilities of sorafenib-resistant cells upon SMYD3 knockdown, but not mutant SMYD3 (Fig. 6e-f). These findings indicated that SMYD3 depletion could effectively inhibit the stemness, migration and invasion ability of sorafenib-resistant cells in HMT-dependent manner.
Genetic and pharmacological inhibition of SMYD3 inhibited the growth and metastasis of sorafenib-resistant tumors by inhibiting the expression of SMYD3-related cancer-promoting genes in vivo.
To evaluate to the impact of targeting SMYD3 in the treatment of sorafenib-resistant tumors, we used a xenograft model to test the effects of shSMYD3 or BCL121 on sorafenib-resistant tumor growth and metastasis. We subcutaneously injected Huh7-Res cells or Huh7-Res with stably expressing shSMYD3-#1 vector, and treated with sorafenib or not, or treated with BCL121 or not, and divided into 6 groups, including DMSO group, Sorafenib group, shSMYD3-#1 DMSO group, shSMYD3-#1 sorafenib group, BCL121 group and Sorafenib + BCL121 group (Fig. 7a). First, we found that compared with the DMSO treatment, sorafenib treatment continued to inhibit the growth of sorafenib-resistant tumors in vivo (Fig. 7b-c). Second, targeting SMYD3 by shSMYD3 or BCL121 both significantly inhibited the growth of sorafenib-resistant tumors, and silencing SMYD3 combined with sorafenib had a good synergistic effect on tumor growth, while BCL121 combined with sorafenib had not (Fig. 7b-c). Because Sorafenib-resistant cells acquire stem cell properties and metastatic properties, which induce cancer relapse and metastasis in multiple organs of nude mice [4, 5]. Finally, we studied the metastasis in liver, lung and spleen by H&E staining, and found that the DMSO and sorafenib treatment groups developed multiple obvious liver and lung metastases, with abnormal inflammatory cell enrichment in the spleen, while SMYD3 depletion by shSMYD3 or BCL121 had no obvious liver or lung metastasis was found in the group and its combination group with sorafenib, but there was still abnormal inflammatory cell enrichment in the lung and spleen (Fig. 7d). These findings directly show that SMYD3 depletion can effectively inhibit sorafenib-resistant tumor growth and distant metastasis in vivo.
To further confirm whether SMYD3 exerts a therapeutic effect by regulating the expression of related cancer-promoting genes in vivo, as a first step, we used qRT-PCR assay to detect the expression of SMYD3-related cancer-promoting genes at the transcriptional level (Fig. 8a). We firstly found that compared with the DMSO group, SMYD3 expression and its related multiple oncogene expressions were significantly increased in the sorafenib group (Fig. 8a). Secondly, compared with the sorafenib group, the expression of SMYD3 gene and its related multiple oncogenes, including SOX4, MYC, JAK1, ZEB1, and TWIST1 gene, were significantly reduced by SMYD3 depletion via shSMYD3 or BCL121, while their combination with sorafenib did not have a significant synergistic effect on these genes (Fig. 8a). In addition, for other genes such as CCNA2, CCNE1, and SNAIL1 genes, silencing SMYD3 significantly inhibited the expression, while BCL121 treatment caused a slight decrease in expression but not significant (Fig. 8a). However, regardless of silencing SMYD3 or BCL121 treatment, the JAK2 gene expression decreased slightly, but there was no statistical significance. What’s more, any treatment has no significant effect on the expression of PCNA gene (Fig. 8a). Next, we performed western blotting assay to detect the protein content of SMYD3-related cancer-promoting genes and related signaling pathways (Fig. 8b-c). The expression of five genes is consistent with qRT-PCR results, including SMYD3, SOX4, MYC, JAK1, and ZEB1 gene, which was increased significantly in sorafenib group compared with DMSO group, and was partially reversed by SMYD3 depletion via shSMYD3 or BCL121, but was not synergistically affect by combination with sorafenib (Fig. 8b-c). Besides, another cell stemness marker NANOG protein, and activated marker of JAK/STAT signaling pathway phosphorylated STAT3 (pSTAT3) were similar trends with them (Fig. 8b-c). However, only silencing SMYD3 exhibited molecular alterations as demonstrated by the decreased expression of epithelial marker E-cadherin and increased expression of mesenchymal markers N-cadherin, indicating that cancer cells was undergoing EMT. These findings showed that targeting SMYD3 inhibited the expression of multiple cancer-promoting genes, resulting in a significant reduction in protein production, thereby inhibiting the activation of corresponding signaling pathways such as stemness-related factors, JAK/STAT3 and EMT pathway.