Transcription Factor SOX18 Promotes Clear Cell Renal Cell Carcinoma Progression and Alleviates Cabozantinib-Mediated Inhibitory Effects

Renal cell carcinoma (RCC), which originates in the proximal tubules, is one of the most lethal urologic malignancies and frequently spreads to distant organs, including lung, bone, and brain (1). Clear cell renal cell carcinoma (ccRCC) is the most common subtype of RCC and accounts for approximately 70% of the disease (2). Approximately two thirds of patients with ccRCC progress to suffer from metastatic diseases, of which half are initial metastases and the other half are metastatic relapses (3). The von Hippel–Lindau (VHL) gene alteration is the most frequent genetic event in ccRCC and has an 80% occurrence rate in sporadic cases (4). Notably, inactivation of the VHL gene as a result of mutation or methylation prevents the degradation of the hypoxia-inducible factor (HIF), and the accumulation of HIF in the nucleus alters the regulation of downstream target genes, including vascular endothelial growth factor (VEGF), which is implicated in angiogenesis (5). These causal events are the rationale for using angiogenesis inhibitors to treat advanced ccRCC. However, acquired resistance to small-molecule therapeutics inevitably occurs in many patients with advanced ccRCC and limits the clinical efficacy of treatments. An increased understanding of tumorigenesis and resistance mechanisms is necessary to identify novel therapeutic approaches to treat ccRCC.

The molecular events associated with blood and lymphatic vessel formation during embryonic development are also implicated in tumorigenesis and metastasis. One example is SOX18, which is a member of the SRY-related HMG box gene (SOX) family of transcriptional factors and functions as an activator of genetic transcription through selective binding to consensus DNA sequences (6). Initial studies reported that SOX18 was implicated in angiogenesis and lymphangiogenesis during embryogenesis, and that abnormal SOX18-mediated events during the embryonic period caused hypotrichosis–lymphedema–telangiectasia syndrome, which is characterized by an abnormal physical appearance in adults (7, 8). Later, reexpression of SOX18 during wound-healing and pathologic conditions, and a molecular switch role for SOX18 during tumor development were discovered (9, 10). Prior studies have revealed that SOX18 is overexpressed in various cancers, in which its effects on cell phenotype include promoting proliferation, migration, and invasion (11, 12). In addition, an in vivo study demonstrated that ablation of SOX18 had inhibitory effects on tumor lymphangiogenesis and metastasis of melanoma in mice (13). Importantly, Overman and colleagues demonstrated that targeting SOX18 with a small-molecule inhibitor was an effective approach to treat breast cancer in mice (14). These prior studies indicate a potential role for SOX18 as a novel biomarker or therapeutic target for cancer, including ccRCC. However, the significance and molecular mechanisms of SOX18 in ccRCC tumorigenesis and disease progression are unknown.

We investigated the expression of SOX18 in ccRCC and evaluated its potential as an indicator of tumor progression and patient prognosis. Cellular function experiments were completed to explore the role of SOX18 on cell proliferation, migration, and invasion. To evaluate the causal role of SOX18 on ccRCC progression, mechanistic studies were conducted. In addition, cells were treated with cabozantinib to investigate the effects of SOX18 on cabozantinib-mediated inhibition on ccRCC progression.

Two ccRCC tissue microarrays (TMA-1 and TMA-2) were obtained from Shanghai Outdo Biotech. TMA-1 consisted of 75 pairs of tumors and corresponding normal tissues. TMA-2 contained 150 tumor cases, of which 30 cases had corresponding normal tissues. In addition, 48 primary tumor samples, including 10 cases containing matched corresponding normal and bone metastatic tumor samples, and 11 cases containing other matched metastatic tumor tissues were obtained from the BioBank of Peking University People's Hospital and used for IHC or immunofluorescent (IF) staining. Collectively, the metastatic samples used for IHC and IF represented 14 cases in bone, 3 cases in the lungs, 1 case in the adrenal glands, and 3 cases in other organs.

In summary, 250 primary tumor cases, 103 normal tissue cases, and 21 metastatic tumor cases with valid information were used for the IHC analysis. Complete follow-up information about 156 patient cases was available. The clinicopathologic characteristics of the patients are presented in Supplementary Table S1. The clinical stage was classified in accordance with the American Joint Committee on Cancer (AJCC) criteria. Before primary tumor resection, none of the patients received treatment.

In addition, 28 pairs of fresh tumors and corresponding normal specimens were collected immediately following nephrectomy. The specimens were frozen using liquid nitrogen and stored at −80°C for protein and RNA extraction. All studies were performed with informed written consent signed by each participate, and approved by the Ethical Review Committee of Peking University People's Hospital.

The procedures for IHC staining and calculation of SOX18 staining scores were described previously (15). Low SOX18 expression was defined as a score of 0 to 3, and high expression was defined as a score of 4 to 9. IF staining was conducted according to standard procedures. Antibody information for IHC and IF is summarized in Supplementary Table S2.

The ccRCC cell lines ACHN, 7860, 769P, CAKI-1, the immortalized normal renal tubular epithelial cell line HK-2, and 293T cell line were purchased from ATCC in 2015. ACHN and 293T cells were cultured in Eagle's minimal essential medium (MEM) and Dulbecco's modified Eagle medium (DMEM; HyClone), respectively. All other cell lines were maintained in Roswell Park Memorial Institute (RPMI) 1640 medium (HyClone). All cell culture media were supplemented with 10% fetal bovine serum (Gibco) and 1% penicillin/streptomycin (Gibco). All cells were maintained in a humidified atmosphere with 5% CO₂ at 37°C and were authenticated via short tandem repeat profiling (last tested in July 2018).

Information regarding recombinant human hepatocyte growth factor (HGF; Lot# 0617S201) and recombinant human epidermal growth factor (EGF; Lot# 0517AFC05) can be obtained at www.peprotech.com. Two types of c-MET small-molecule inhibitors, cabozantinib (XL184, BMS-907351, www.selleck.cn) and capmatinib (INCB28060, www.selleck.cn), were also used (16).

Synthesis of SOX18, EGFR, and c-MET siRNA was completed by GenePharma. The small inhibiting RNA (siRNA) construct with the greatest SOX18 silencing efficiency and a negative control (siNC) were cloned into lentiviral vectors (shRNA and shNC) to create stable SOX18-silenced cells. Plasmid and lentiviruses carrying SOX18 (SOX18) and NC constructs were synthesized by the Shanghai GeneChem Co. Ltd. The siRNA transfection or viral infection was completed according to the manufacturer's instructions. Forty-eight hours after infection with a multiplicity of infection of 40, antibiotic selection with a puromycin concentration of 2 μg/mL (Calbiochem) was used to select for stable cell lines. The efficiency of siRNA transfection or viral infection was verified by Western blotting. The siRNA, short hairpin RNA (shRNA), and corresponding control sequences are summarized in Supplementary Table S3.

The effect of SOX18 on cell proliferation was evaluated using the cell counting kit-8 (CCK-8) assay (Dojinaodo). Briefly, 1,500 cells in 150 μL of medium were seeded onto 96-well plates. The absorbance of each well at 450 nm was measured at six different time points. To measure the effects of the c-MET inhibitors on cell proliferation, 5,000 cells in 200 μL of complete medium were plated onto 96-well plates and incubated for 72 hours, followed by absorbance measurements. Prior to all absorbance measurements, the medium in each well was replaced with 100 μL of complete medium supplemented with 10% CCK-8 solution, and the cells were incubated for 2 hours.

Cells were plated in 6-well plates at a density of 1,000 cells per well and cultured for 2 weeks. Then, the colonies were fixed with 4% paraformaldehyde for 20 minutes, stained with a 0.5% crystal violet solution for 20 minutes, and counted.

The animal studies were approved by the Institutional Animal Care and Use Committee of Peking University People's Hospital. Female BALB/c nude mice that were 4 to 6 weeks old were obtained from Beijing Vital River Laboratory Animal Technology Co., Ltd. The mice were randomly divided into two groups (n = 6), and 4 × 10⁶ ACHN cells transfected with shSOX18 or shNC were subcutaneously injected into the right forelimb armpit. On day 35, the mice were euthanized, and the tumors were weighed and preserved in liquid nitrogen for further analysis. The tumor length, width, and thickness were measured, and the tumor volume was calculated as follows: 0.5 × length × width × thickness.

Cells were stained with the Cell-Cycle Staining Kit (Liankebio). Then, the percentage of cells in G₁, S, and G₂–M phases was measured with flow cytometry (Beckman).

The Cell-Light EDU Apollo488 In Vitro Kit (RIBOBIO) was used to measure cell proliferation, according to the manufacturer's instructions.

293T cells at a density of 2 × 10⁴ cells in 100 μL of DMEM were plated in triplicate into a 96-well plate and cultured for 24 hours. The pGL-MYC-promoter (−3,500 kb to TSS) or pGL-control-promoter plasmids were cotransfected with SOX18 or pCDH plasmids (control) into cells using Lipofectamine 2000. Luciferase activity was measured 48 hours after transfection, using the Dual-Luciferase Reporter Assay System (Promega).

Cells were seeded at a density of 5,000 cells per well into 96-well plates and cultured for 48 hours. The TUNEL Apoptosis Assay Kit (Beyotime) was used to detect apoptotic cells.

Cell migration and invasion were evaluated using Transwell invasion assays with or without Matrigel. To assess the effect of SOX18 on cell migration and invasion, 4 × 10⁴ cells were plated into the upper chamber of a 24-well Transwell or Matrigel chamber with 8-μm pores (Corning). The same number of cells was also used to evaluate the effect of SOX18 on cell migration and invasion under c-MET inhibition. For cell migration assays, ACHN and 769P cells were incubated for 24 hours prior to the assay. For cell invasion assays, ACHN and 769P cells were incubated for 48 and 24 hours, respectively. Cell harvesting efficiency and cell calculations were estimated as described previously (17).

Total RNA was extracted from tissues and cells using the RNAsimple Total RNA Kit (Tiangen). FastQuant RT Kit (Tiangen) was used for cDNA synthesis. The quantitative polymerase chain reactions (qPCR) were performed using KAPA SYBR FAST Universal q-PCR Kit (KAPA). The relative mRNA levels of genes were calculated using cycle threshold (CT) methods, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as a control. The primer sequences are listed in Supplementary Table S3. The transcriptomic analysis was performed using the Illumina high-throughput sequencing platform (Beijing Genomics Institute, BGI).

Total protein from tissues and cells was extracted using RIPA lysis and extraction buffer and measured using a BCA kit (Solarbio). A total of 20 μg protein from each sample was separated using 10% or 12% separating gels and transferred to nitrocellulose membranes (Solarbio). Proteins were detected using a Fluorescence Imaging System (Sagecreation). Antibody information is summarized in Supplementary Table S2.

EGF and HGF secretion levels were detected using the Human EGF ELISA Kit and Human HGF ELISA Kit (Liankebio), respectively, according to the manufacturer's instructions.

Statistical analyses were conducted using the SPSS 19.0 package or GraphPad Prism 5.0. Data are presented as the mean ± SEM. The SOX18 expression between groups was analyzed with a χ² test. A Student t test was used for comparisons between two groups. Differences in cell-cycle and cell proliferation distributions were analyzed using a two-way analysis of variance test. The Cox proportional hazards model was applied in univariable or multivariable survival analysis to assess the role of SOX18 in the prognosis of ccRCC. Kaplan–Meier survival curves were used to plot overall patient survival and a log-rank test was used to analyze the differences between groups. P < 0.05 was considered statistically significant.

SOX18 expression was initially investigated in a panel of cell lines and paired patient samples using Western blot analysis. Four ccRCC cell lines showed a higher level of SOX18 protein when compared with HK-2 cells (Fig. 1A). Western blotting also indicated that increased SOX18 protein expression was present in six of the eight ccRCC tissues when compared with matched normal tissues (Fig. 1B). In addition, qPCR analysis revealed increased (>1.5-fold) SOX18 mRNA expression in 67.8% (19/28) of the ccRCC tissues when compared with corresponding normal tissues (Fig. 1C). Data from The Cancer Genome Atlas (TCGA) database demonstrated that SOX18 mRNA was elevated in most ccRCC tissues when compared with normal tissue samples (Fig. 1D), which is consistent with our results. Combined with the TCGA data, our results suggested that SOX18 is upregulated in ccRCC.

To further investigate SOX18 expression in ccRCC, we performed IHC in tissues. IHC analysis indicated that SOX18 staining in tumor tissues was mostly nuclear, with some cytoplasmic staining. Conversely, the staining in normal tissues exhibited an opposite pattern of SOX18 localization, which suggested that nuclear translocation of SOX18 was important for the development of ccRCC. SOX18 nuclear translocation was validated with IF staining (Fig. 1E). Based on these data, we specifically investigated nuclear SOX18 expression in the patient samples. High nuclear SOX18 expression occurred in 12.6% of the normal tissues, 49.6% of the primary tumors, and 85.7% of the metastatic tumors, indicating that a gradual increase in SOX18 expression occurs during ccRCC genesis and metastasis (Fig. 1F). As shown in Fig. 1G, SOX18 expression in primary tumor areas was higher than corresponding normal tissues and was less than metastatic tumors.

Statistical analysis revealed a strong positive correlation between SOX18 expression and pathologic grade and T, N, and M classifications (Fig. 1H). No association was observed between SOX18 expression and age, gender, or clinical stage (Supplementary Fig. S1A). Kaplan–Meier survival analysis revealed that patients with high SOX18 protein levels exhibited poor overall survival (OS) compared with patients with low SOX18 protein levels (Fig. 1I). Univariate Cox regression model analysis indicated that pathologic grade, TNM classification, clinical stage, and SOX18 expression were all prognostic factors for ccRCC. Furthermore, the multivariate Cox regression model analysis confirmed that pathologic grade and N and M classifications could act as independent prognostic factors (Supplementary Table S4). Collectively, these data corroborated the role of SOX18 as a biomarker in ccRCC progression and as a predictor for poor OS of patients.

To evaluate the role of SOX18 on ccRCC growth, we initially examined its effect on cell proliferation. Knockdown of SOX18 markedly inhibited cell proliferation, and the effect was more pronounced at later time points (Fig. 2A). Overexpression of SOX18 had the opposite effect and increased cell proliferation (Supplementary Fig. S1B). The role of SOX18 on cell proliferation was confirmed with colony formation assays (Fig. 2B; Supplementary Fig. S1C). Next, we performed animal studies to investigate the influence of SOX18 on in vivo tumor growth. Consistent with the cell studies, the tumor formation rate in the SOX18-silenced group was lesser than that in the control group. Tumors were observed in 83.3% (5/6) of the SOX18-silenced mice compared with 100% (6/6) of the controls. Moreover, tumor volume and weight in the SOX18-silenced group were lesser than those in the control group (Fig. 2C). Taken together, our data demonstrated that SOX18 promoted ccRCC growth.

Next, we investigated the effect of SOX18 on cell-cycle distribution. As shown in Fig. 2D, flow cytometry analysis demonstrated that SOX18 silencing increased the proportion of cells in G₁ phase and decreased the number of cells in S phase. Opposite distributions were measured when SOX18 was overexpressed (Supplementary Fig. S1D), suggesting that SOX18 promotes cell-cycle transition from G₁ to S phase. Similarly, using EDU incorporation as a metric, SOX18 silencing significantly reduced the proportion of cells in S phase, and an opposite trend was observed in SOX18-transducted cells (Fig. 2E; Supplementary Fig. S1E).

Results of the transcriptome analysis demonstrated that various genes were downregulated or upregulated in SOX18-silenced ACHN and 769P cells (Supplementary Fig. S2A, SRA accession number: PRJNA561456). The expression of MYC, CCND1, and other genes in SOX18-selenced cells was confirmed by qPCR analysis (Fig. 3A). In addition, Western blot analysis revealed that c-Myc (encoded by MYC) and cyclin D1 (encoded by CCND1) levels were decreased or increased in SOX18 silenced or overexpressed cells, respectively (Fig. 3B; Supplementary Fig. S2B). The effect of SOX18 on c-MYC was further validated by IF staining in xenograft models (Fig. 3C). Luciferase reporter assays also showed that cotransfection of pGL-c-MYC-promoter and SOX18 plasmids into 293T cells resulted in increased luciferase activity when compared with controls (Fig. 3D). These data indicated that SOX18 could increase the transcriptional activity of MYC. It was previously reported that c-Myc promoted cyclin D1 expression by directly binding to the promoter of CCND1, and consequently accelerated the cell-cycle transition from G₁ to S phase (18). Our data suggest that SOX18 accelerates ccRCC growth and cell-cycle progression by increasing transcriptional activity of MYC, which results in cyclin D1 expression.

To assess the effect of SOX18 on cell apoptosis, we performed apoptosis assays. As shown in Fig. 3E, SOX18 silencing significantly increased the proportion of apoptotic cells. However, ectopic overexpression of SOX18 had minimal effects on cell apoptosis (Supplementary Fig. S2C), which might be due to inherently low rates of cell apoptosis. The inhibitory effect of SOX18 on cell apoptosis may further contribute to ccRCC progression.

To investigate the implications of SOX18 expression on cell migration and invasion, we performed cell migration and invasion assays. Our data revealed that SOX18 silencing resulted in an obvious reduction in the migratory (Fig. 4A) and invasive (Fig. 4B) ability of cells. Conversely, SOX18 overexpression promoted increased cell migration and invasion (Supplementary Fig. S3A and S3B).

MMP-7, VIM, and SNAI1 mRNA levels were significantly decreased in SOX18-silenced cells, as assessed using qPCR (Fig. 3A). Western blot analysis corroborated that decreased MMP-7, vimentin, and snail (encoded by MMP-7, VIM, and SNAI1, respectively) expression occur in SOX18-silenced cells (Fig. 4C). Conversely, SOX18-induced cells exhibited increased MMP-7, vimentin, and snail expression (Supplementary Fig. S3C).

Because vimentin and snail are both vital EMT-related molecules, we speculated that SOX18 was linked to EMT in ccRCC. We used Western blot analysis to measure E-cadherin and N-cadherin expression. As shown in Fig. 4C and Supplementary Fig. S3C, N-cadherin levels decreased or increased in SOX18-silenced or -induced cells, respectively. Surprisingly, E-cadherin expression followed the same trend. These results were validated using IF staining analysis of E-cadherin, N-cadherin, and vimentin in ACHN xenograft tissues (Fig. 4D–F). Collectively, our data indicate that SOX18 may facilitate ccRCC cell migration and invasion by regulating both MMP-7 expression and the EMT process.

Western blot analysis showed that silencing or overexpressing SOX18 could decrease or increase the expression of EGFR, p-EGFR (Tyr1068), c-Met, p-Met (Tyr1349), p-Met (Tyr1234/5), p-Akt, p-MEK1/2, and p-Erk1/2 (Fig. 5A; Supplementary Fig. S3D). The Western blot results were confirmed by reduced IF staining for EGFR and p-MET (Tyr1234/5; Fig. 5B) in SOX18-silenced xenografts tissues. Previous studies have reported that AKT can be activated by the phosphorylation of EGFR and c-MET (19), which subsequently activates a series of downstream signaling molecules, including the phosphorylation of MEK1/2 (p-MEK1/2) and ERK1/2 (p-ERK1/2), to elicit biological effects (20). These suggested that SOX18 could promote activation of AKT and downstream signals via promoting the activation of EGFR and c-MET. The redundant interaction between growth factor EGF and HGF and their respective receptors EGFR and c-MET is responsible for activating the EGFR and c-MET (21, 22). We thus wanted to evaluate whether the effect of SOX18 on activation of EGFR and c-MET was EGF- and/or HGF-dependent. qPCR analysis revealed that EGF and not HGF mRNA was decreased in SOX18-silenced cells (Fig. 3A). However, silencing SOX18 inhibited the secretion of EGF and HGF (Fig. 5C). Thus, we next treated the SOX18-induced cells with EGFR (EGFR-si), c-MET siRNA (c-MET-si), or negative control (EGFR-ctrl or c-MET-ctrl) to inhibit the interaction of EGF or HGF with their cognate receptors EGFR or c-MET. As shown in Fig. 5D, EGF/EGFR inhibition reduced SOX18-induced the expression of EGFR, p-EGFR, and p-AKT. Similarly, HGF/c-MET inhibition reduced SOX18-induced expression of p-MET(1349) and p-AKT (Fig. 5E). Additionally, SOX18-induced cell migration and invasion was suppressed (Supplementary Fig. S3E). These data suggested that the effect of SOX18 activating AKT and downstream signals was EGF/EGFR or HGF/c-MET mediated.

Next, we treated the SOX18-silenced cells with 40 ng/mL of EGF or 5 ng/mL of HGF for different amounts of time (0, 10, 30, 60 minutes). Initially, we evaluated whether persistent exposure to EGF and HGF rescued the cellular functions inhibited by SOX18 knockdown. As shown in Fig. 5F and G, persistent treatment with EGF and HGF both rescued SOX18 knockdown-inhibited cell migration and invasion. Simultaneous treatment of HGF and EGF did not have an additive effect on cell migration and invasion (Fig. 5F and G), which suggests that the EGF/EGFR and HGF/c-MET signaling pathways have redundant roles in regulating cell migration and invasion. Next, we probed the EGF/EGFR and HGF/c-MET signaling pathways using Western blot analysis. As shown in Fig. 6A, EGF simultaneously activated the EGF/EGFR and HGF/c-MET signaling pathways with the increasing expression p-EGFR, p-MET (Tyr1349), p-MET (Tyr1234/5), and p-AKT, while HGF only activated the HGF/c-MET signaling pathway with increased p-MET (Tyr1349), p-MET (Tyr1234/5), and p-AKT expression. Interestingly, a dramatic activation of AKT was observed when cells were exposed to either HGF or EGF for 60 minutes, indicating that activation of AKT by EGF or HGF was time dependent. A similar phenomenon was also observed with respect to downstream p-MEK1/2 and p-ERK1/2. Collectively, our data demonstrate that SOX18 promotes ccRCC cell migration and invasion through the EGF/EGFR or HGF/c-MET signaling pathway.

Cabozantinib, a promising multitargeted c-MET inhibitor, is an approved therapeutic for patients with advanced ccRCC who failed to respond to sunitinib treatment. Importantly, cabozantinib significantly improved survival compared with current mechanistic target of rapamycin inhibitors (23, 24). Because cabozantinib partly suppresses ccRCC cellular function by inhibiting the activation of c-MET, we next evaluated the effect of SOX18 on cellular function under cabozantinib treatment.

Western blot analysis revealed that expression of c-MET, p-MET (Tyr1349), and p-MET (Tyr1234/5) were elevated in ccRCC tissues and cell lines (Supplementary Fig. S3F and S3G). Among the cell lines tested, 769p had the highest c-MET expression due to a VHL mutation (25). ACHN cells also exhibited substantial c-MET expression, although they do not have a VHL mutation. As shown in Fig. 6B, the inhibitory effects of cabozantinib on proliferation in SOX18-induced ACHN and 769P cells were limited when cabozantinib reached a certain concentration. The cells were then treated with another c-MET inhibitor, capmatinib, which is an adenosine triphosphate (ATP) competitive c-MET inhibitor. Consistent with the cabozantinib treatment results, ectopic expression of SOX18 suppressed the inhibitory effects of capmatinib on proliferation (Supplementary Fig. S4A). In contrast, silencing SOX18 augmented the inhibition of cell proliferation induced by both cabozantinib and capmatinib (Supplementary Fig. S4B and S4C).

Finally, we evaluated the influence of SOX18 on cell migration and invasion during cabozantinib treatment. We treated cells with cabozantinib at concentrations of 2 or 3 μmol/L. At these concentrations, cabozantinib has negligible effects on cell proliferation and the activation of c-MET could be effectively inhibited (Supplementary Fig. S4D). Ectopic expression of SOX18 significantly suppressed the drug-induced inhibition of cell migration and invasion (Fig. 6C and D; Supplementary Fig. S4E and S4F). In contrast, SOX18 silencing enhanced the inhibitory effects of cabozantinib on migration and invasion (Supplementary Fig. S5A and S5B). These data suggest that silencing SOX18 potentiates the suppression of cellular function induced by cabozantinib treatment and increases the efficacy of the drug on ccRCC cells.

To examine the effect of SOX18 on cabozantinib-mediated inhibition of activation of c-MET and downstream signals, we treated SOX18-induced cells with cabozantinib at concentrations of 2 or 3 μmol/L. As shown in Fig. 6E, the expression of c-MET, p-MET (Tyr1349), p-MET (Tyr1234/5), p-AKT, p-MEK, and p-ERK in SOX18-induced cells was higher than that of control cells. These data indicate that cabozantinib-mediated inhibition of c-MET activation is compromised by SOX18.

Despite considerable progress in the treatment of ccRCC, a lack of biomarkers that can be used for informed decision-making, combined with resistance to current therapeutics, necessitates further research on the molecular events that regulate ccRCC.

SOX18 is a member of the SOX family of transcriptional factors and is implicated in malignant phenotype, angiogenesis, and lymphangiogenesis in tumors. Pharmacologic targeting of SOX18 to treat breast cancer in mice indicated that SOX18 could be a potential therapeutic target (14). Here, we report that SOX18 is upregulated and localized to the nucleus during ccRCC development. Upregulation of SOX18 was associated with negative clinicopathologic characteristics and a poor prognosis. In addition, we demonstrated that SOX18 promoted ccRCC progression, accelerated malignant cellular behaviors, and attenuated the suppression of cellular functions in response to c-MET inhibitors. Furthermore, our data indicated that SOX18 promoted ccRCC progression via comprehensive molecular mechanisms (Fig. 6F).

The c-Myc oncogene is a modulator of proliferation, apoptosis, and chemosensitivity and is indispensable for cell-cycle regulation (26, 27). As a cell-cycle regulator, c-Myc induces the G₁–S phase transition by activating the transcription of cyclin D1 (18). Our data indicated that SOX18 increases the transcriptional activity of c-Myc and cyclin D1 expression to accelerate G₁–S transition; therefore, we postulate that SOX18 promotes ccRCC progression through cell-cycle regulation via c-Myc and cyclin D1 expression. Prior studies report that c-Myc and cyclin D1 are hyperactivated in ccRCC and are associated with poor patient prognosis (28), which may be partly explained by our data.

During malignant cancer progression, increased cancer cell invasion is a consequence of multiple molecular events, including the activation of AKT-related pathways, disrupted homeostasis of the extracellular matrix (ECM), and EMT (29, 30). EMT is responsible for increased invasion and dissemination of ccRCC cells in primary tumors. Importantly, EMT reversal in metastatic tumors, a process called mesenchymal–endothelial transition (MET), is also crucial for the tumorigenesis of metastatic cancer cells (29, 31). In the present study, we report a connection between SOX18 and EMT-inducing or -related molecules. The expression of N-cadherin, snail, and vimentin correlated with altered SOX18 levels. Despite altered E-cadherin levels, sustained changes caused by SOX18 are unlikely because of the important role of E-cadherin in cell–cell adhesion. The typical switch from E-cadherin to N-cadherin expression in the EMT process often occurs during the transition from normal tubular epithelial cells to cancer cells and MET occurs in metastatic tumors exhibiting depressed N-cadherin levels (32); therefore, we speculate that SOX18 may synchronize the processes of EMT and MET during ccRCC progression. Under this framework, SOX18 would promote ccRCC cell invasion by regulating N-cadherin, snail, and vimentin expression. Furthermore, it would simultaneously promote tumor formation by promoting E-cadherin expression, although these mechanisms need to be experimentally confirmed. In addition, our study revealed that SOX18 regulates MMP-7 expression, which indicates that SOX18 may promote cell invasion through ECM degradation. In vascular endothelial cells, MMP-7 is under the transcriptional regulation of SOX18, which was demonstrated in studies that targeted the MMP-7 promoter (33). Whether the regulation of MMP-7 by SOX18 in ccRCC cells occurs through a similar mechanism warrants further investigation. In addition, the association between SOX18 and other MMP family members is of interest.

In ccRCC, the receptor tyrosine kinases, EGFR and c-MET, are excessively hyperactivated and phosphorylated. In the current study, we discovered that SOX18 promoted cell migration and invasion through the EGF/EGFR and HGF/c-MET signaling pathways. Intriguingly, our data revealed that EGF could simultaneously activate the EGF/EGFR and HGF/c-MET signaling pathways, whereas HGF could only activate the HGF/c-MET signaling pathway. These data demonstrated that SOX18 activity occurs upstream of the EGF/EGFR and HGF/c-MET pathways and could activate the HGF/c-MET pathway even in the absence of HGF. Indeed, high EGF and HGF expression levels in blood samples from patients with ccRCC have been shown in previous studies (34, 35), and a redundant interaction between EGF and HGF and their respective receptors EGFR and c-MET is responsible for activating the EGF/EGFR and HGF/c-MET signaling pathway. Therefore, SOX18 is a potential therapeutic target by regulating EGF/EGFR and HGF/c-MET signaling pathways. Although individual pharmacologic inhibition of the activation of EGFR and c-MET in oncological treatment is of great clinical significance, activation of alternative pathways and signaling cross-talk impose drug-resistance limitations (36, 37). Therefore, simultaneous inhibition of the EGF/EGFR and HGF/c-MET pathways via SOX18 targeting has important clinical implications.

Although cabozantinib is a superior option in tumor treatment, the inevitability of resistance to cabozantinib has been validated (38). Studies have shown that further activation of c-MET is responsible for cabozantinib resistance (39). Our data revealed that SOX18 could overcome cabozantinib-mediated inhibition of cellular function. Moreover, inhibition of the activation of c-MET and downstream signals by cabozantinib was compromised in SOX18-induced cells. These data suggested that SOX18 could sustain activation of c-MET and downstream signals even under cabozantinib treatment, and thus may contribute to the resistance of ccRCC cells to cabozantinib treatment. In addition, the mechanisms of drug resistance to tyrosine kinases inhibitors (TKI) are being actively investigated, and recent studies reported that upregulation and activation of c-MET are involved in sunitinib-resistant ccRCC (40). Moreover, the EGFR and c-MET activation, as well as EMT, was also reported to cause acquired resistance in other TKIs that are used to treat ccRCC and lung cancer (41–43). Our data highlight the significance of SOX18 as a regulator of the EMT process and the EGF/EGFR and HGF/c-MET signaling pathways.

In conclusion, our findings, together with prior studies, highlight that SOX18 is a potential therapeutic target. Furthermore, therapeutics against SOX18 could target multiple dysregulated pathways to treat ccRCC and other cancers with increased efficacy.

No potential conflicts of interest were disclosed.

Conception and design: X. Tao

Development of methodology: T. Xu, Z. Xiaowei, L. Qing

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Y. Huaqi, D. Yiqing

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): D. Xiang

Writing, review, and/or revision of the manuscript: Y. Huaqi

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Y. Huaqi, Q. Caipeng, W. Qiang

Study supervision: L. Shijun, X. Tao

The authors thank the Department of Pathology at Peking University People's Hospital for their technical help with IHC staining and analysis. The present study was supported by the National Natural Science Foundation of China (NO. 81472393 and 31671469) from Xu Tao.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.