Abnormally expressed JunB transactivated by IL-6/STAT3 signaling promotes uveal melanoma aggressiveness via epithelial–mesenchymal transition

Uveal melanoma (UM) is the most common primary ocular malignancy in adults, accounting for more than half of all eye cancers. Despite the rarity and proper treatment of the primary ocular tumor, approximately 50% of UM patients may develop lethal systemic metastases, in which the liver is the principal metastatic organ [1]. Unfortunately, once liver metastasis is diagnosed, the prognosis is poor, with a median survival time between 4 and 15 months [2]. Although previous studies have greatly increased the knowledge of UM development, the relevant factors that contribute to metastasis are still not well known. This knowledge is urgently needed for the early detection and treatment of metastatic UM.

Inflammation was recently described as a hallmark of cancer, in which cancer cells exploit immune evasion to survive and spread [3,4]. Although the eye is an immune-privileged organ, inflammation can be present within the established ocular tumor microenvironment [5]. Therefore, several proinflammatory cytokines released by infiltrating immune cells and other cells in the microenvironment are involved in UM development. Uveal melanoma-containing eyes often carry within the aqueous humor and vitreous fluid increased concentrations of several inflammatory cytokines and chemokines, such as interleukins (IL-1/6/8), MCP-1, TNF-α, IFN-γ, and IP-10, which correlate with clinicopathological features [6–9]. Of these, IL-6 appears to play a crucial role in the UM microenvironment, as indicated by the correlation between this cytokine and the presence of immune cell infiltrate [10]. However, to date, major studies have been restricted in describing intraocular cytokine levels in UM patients. Further investigation is needed to clarify the role of inflammatory cytokines in UM development.

Activating protein-1 (AP-1) is a transcription factor composed of homodimers between Jun (c-Jun, JunB, and JunD) proteins or heterodimers between the Jun and Fos (c-Fos, FosB, Fra1, and Fra2) families. These proteins regulate a magnitude of cellular processes, including cell proliferation, survival, differentiation, invasion, and carcinogenesis, depending on dimer composition [11]. AP-1 can induce epithelial–mesenchymal transition (EMT) and promote invasion in epithelial cell lines, disrupting epithelial cell polarity and cell–cell junctions [12]. As Jun family members, Jun and JunB share extensive homology within the leucine zipper and basic domains, and JunB can rescue the lethal phenotype in Jun-null mice [13]. Recently, JunB has been proposed as a driver of the EMT program induced by TGF-β [14,15]. Growing evidence has shown that aberrantly expressed JunB promotes migration, invasion, and metastasis during cancer progression [16,17].

Here, we found that JunB was transactivated by IL-6/STAT3 signaling and played a central role in IL-6-induced EMT in UM cells. As a result, the activation of the IL-6/STAT3/JunB axis promoted migration and invasion of UM cells. Thus, our findings suggest the existence of an overactivated IL-6/STAT3/JunB pathway through which the proinflammatory cytokine IL-6 in the tumor microenvironment promotes EMT and aggressiveness in UM.

C918 and OCM1A human UM cell lines were obtained from National Infrastructure of Cell Line Resource (Beijing, China) and cultured in RPMI-1640 medium containing 10% fetal bovine serum (Gibco, Waltham, MA, U.S.A.), 100 U/ml penicillin, and 100 mg/ml streptomycin. Both the two cell lines were derived from primary UM and established from biopsied specimens of choroidal melanomas of epithelioid and spindle type morphologies respectively. To establish IL-6-transformed UM cells, C918 and OCM1A cells were exposed to 2 ng/ml IL-6 for 8 weeks. The obtained transformed cells were named IL6/C918 and IL6/OCM1A, and they were then maintained in medium supplemented with 2 ng/ml IL-6. All of the cell lines were grown in a 5% CO₂ atmosphere at 37°C. Plasmids and siRNAs were transfected using Lipofectamine™ 3000 Transfection Reagent (Life Technologies, Carlsbad, CA, U.S.A.). JunB or EV stably transfected C918 cells were derived from the parental cells by G418 (Sigma, St. Louis, MO) selection. Recombinant human interleukin-6 (IL-6) was purchased from R&D Systems (Minneapolis, MN).

Antibodies against CDH1 (E-cadherin), Vimentin, FN1 (fibronectin), TJP1, TJP2, ICAM-1, and GAPDH were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The ZEB1, Snail, Slug, Twist1, STAT3, and p-STAT3 (Y705) antibodies were acquired from Cell Signaling Technology (Beverly, MA). JunB and c-Jun antibodies were bought from Abcam (Cambridge, U.K.). The siRNAs targeting STAT3 were synthesized by GenePharma (Shanghai, China). The p-STAT3 inhibitor, Stattic, was purchased from Calbiochem (San Diego, CA).

RNA from IL6/C918 or C918 cells was extracted and purified for quantification, RNA-seq library preparation, and sequencing. The libraries were sequenced on the Illumina HiSeq 2500 platform. The reads containing adapter or poly-N and reads of low quality were removed from raw data to generate clean reads for further analyses. Based on the clean reads, the Q20 (>90%), Q30 (>85%), and Error rate (<0.1%) of the clean data were required. Then, mapped reads were obtained by Tophat2 by aligning clean reads to the human genome reference (hg19). The number of mapped clean reads for each unigene was counted and normalized into reads/kb/million reads (RPKM) to calculate the expression level of the unigene.

Total RNA was isolated with RNAiso Plus (TaKaRa, Kyoto, Japan). RNA quality and concentration were evaluated using a Nanodrop ND-1000 Spectrophotometer. A total of 0.5 μg of RNA was reverse transcribed using the cDNA Reverse Transcription Kit (TaKaRa) according to manufacturer’s instructions. The acquired cDNA was analyzed in triplicate by real-time PCR on a Life Technologies Step One Plus Real-Time PCR System with SYBR Green Master Mix (Roche, Basel, Switzerland). Target gene expression was normalized to GAPDH levels in respective samples as an internal control. The sequences of the qRT-PCR primers are listed in Supplementary Table S1.

Cells were harvested and lysed using RIPA lysis buffer (50 mM Tris-HCl, pH 7.4, 100 mM 2-mercaptoethanol, 2% w/v SDS, 10% glycerol). Protein concentrations were determined with the Bradford method (Bio-Rad, Hercules, CA, U.S.A.). The proteins were separated by 10% SDS-PAGE and transferred to nitrocellulose membranes (Whatman, Maidstone, U.K.). The membranes were incubated with dilutions of primary antibodies followed by IRDye 800CW or IRDye 680-conjugated secondary antibodies, and then detected by an Odyssey Infrared Imaging System (LI-COR Biosciences, Lincoln, NE, U.S.A.).

IL6/C918 or C918 cells on coverslips were fixed and incubated with the described primary antibodies overnight at 4°C. After three washes, the coverslips were incubated with Alexa Fluor® 594 or Alexa Fluor® 488-conjugated secondary antibodies (CST, Beverly, MA) in a dark chamber for 60 min, followed by three washes with PBS. DAPI (Sigma, St Louis, MO, U.S.A.) was used to indicate the nuclei. Coverslips were mounted and visualized with a microscope using the appropriate filters.

The genomic regions surrounding the promoter of human JunB were amplified by PCR and inserted into the pGL3 vector. The reporter constructs containing various lengths of JunB promoter or mutated STAT3 binding sites were generated by subsequent PCR-based cloning. C918 cells were plated onto 24-well plates. The next day, the cells were cotransfected with firefly luciferase reporter constructs and pRL-SV40 Renilla luciferase reporter plasmids (Promega, Madison, WI, U.S.A.). The pRL-SV40 plasmid was used to normalize the transfection efficiency. At 24 h post-transfection, C918 cells were incubated with 20 ng/ml IL-6 for 24 h and luciferase activity was measured using a dual-luciferase reporter assay system (Promega) and a luminometer (LB 9507, Berthold, Bad Wildbad, Germany).

Chromatin from IL6/C918 or C918 cells was crosslinked with 1% formaldehyde and sonicated to obtain a DNA fragment of 200–500 bp. After centrifugation, the supernatants were subjected to immunoprecipitation overnight at 4°C with antibodies against STAT3 or normal IgG. The DNA–protein complexes were isolated using Protein A/G PLUS-Agarose (Santa Cruz). The crosslinking was reversed and released DNA fragments were purified and quantified by qRT-PCR using the following primer pairs for the JunB promoter. SBE1: CGTAGGATCCGAGTGACGG (Forward); CCCAACACCGTGTCGGCTCC (Reverse) / SBE2: TGCAGCCCCGCCGAGCCAC (Forward); TGCGCTCCGATTGGCCGTC (Reverse).

Cell viability was detected using a Cell Counting Kit-8 assay (Dojindo, Kumamoto, Japan). Cells were dispensed in triplicate into 96-well plates and incubated overnight at 37°C. After 96 h, 10 μl of CCK-8 kit solution was added to the cells, which were then incubated for 2.5 h at 37°C. Absorbance was then measured by a microplate reader at 450 nm. Data were obtained from at least three separate experiments carried out in triplicate.

Cell migration was determined by a scratch wound healing assay. Cells were allowed to reach confluence, and a wound was created in the monolayer by scraping with a sterile pipette tip across the entire diameter of the well. The culture was then washed with medium to remove free-floating cells and debris and cultured in serum-free medium for an additional 48 h. To monitor the wound closure, images of the wound area were captured in six fields.

The in vitro cell invasion assay was performed in 24-well Transwell plates (Corning, NY, U.S.A.) with 8 μm-pore inserts coated with Matrigel (BD Biosciences, San Jose, CA). Cells (1 × 10⁵) were applied to a culture insert in serum-free medium, whereas complete medium was applied to the lower compartment. After incubation for 48 h, cells on the upper surface of the filter were removed carefully with a cotton swab and the undersurface adherent cells that had invaded through the Matrigel were fixed in methanol and stained with 0.5% Crystal Violet. The air-dried filter membrane was viewed under a microscope and four random fields were selected for cell counting.

Statistical data analysis was performed with SPSS 22.0 and GraphPad Prism 5.0. Difference analysis was performed with the two-tailed Student’s t test and analysis of variance (ANOVA). Spearman’s correlation test and Pearson’s correlation coefficient were used to analyze correlation. Data were reported as the means ± S.E. A value of P<0.05 was considered statistically significant.

To characterize the function of IL-6 in the malignant phenotype of UM cells, two IL-6- transformed cell lines, IL6/C918 and IL6/OCM1A, were developed from UM cells (C918 and OCM1A) by continuous exposure to a low dose of IL-6 over a long period of time. Next, IL6/C918 and IL6/OCM1A cells were examined by wound healing, invasion, and viability assays. Compared with the parental cells, IL6/C918 and IL6/OCM1A cells exhibited increased migration and invasion (Figure 1A,B, relative wound closed: 2.11 (IL6/C918 vs C918); 1.43 (IL6/OCM1A vs OCM1A); relative invasive cells: 2.24 (IL6/C918 vs C918); 1.36 (IL6/OCM1A vs OCM1A)). It was notable that the acquired migration and invasion in the IL6/C918 cells was more obvious than in the IL6/OCM1A cells. In contrast, the IL6/C918 and IL6/OCM1A cells displayed a lower proliferative ability than the parental cells (Figure 1C, relative absorptivity: 0.71 (IL6/C918 vs C918); 0.74 (IL6/OCM1A vs OCM1A)). Therefore, long-term exposure to IL-6 induces migration and invasion but inhibits proliferation in UM cells.

To investigate which mechanisms are involved in IL-6-induced migration and invasion of UM cells, RNA-seq was performed to profile the transcriptome differences between IL6/C918 and its parental cell. In total, 5451 genes with significantly differential expression (P<0.05) were identified. Among these, 2747 genes were up-regulated and 2704 genes were down-regulated in IL6/C918 cells, compared with C918 cells (Figure 2A and Supplementary Table S2). A total of 846 genes (15.5%) had >2-fold change (Figure 2B). Furthermore, gene ontology (GO) enrichment analysis for the 5451 differentially expressed genes (DEGs) was carried out based on biological process, molecular function, and cellular component. The top ten terms with the most significant enrichment of DEGs are shown in Figure 2C, among which cell–cell adhesion and focal adhesion were focused on due to their involvement in malignant transformation. Then, we validated the expression levels of cell–cell adhesion molecules (TJP1, TJP2, and CDH1) and focal adhesion molecules (FN1 and ICAM-1) in IL6/C918 and C918 cells. Lower expression levels of TJP1, TJP2, and CDH1 along with higher levels of FN1 and ICAM-1 were detected in IL6/C918 cells than in C918 cells at both the mRNA and protein levels (Figure 2D,F,G). In addition, most integrin subunits, which function as FN1 receptors, were induced by long-term IL-6 exposure (Figure 2E). The above results suggest that IL-6 impairs cell–cell junction but enhances focal adhesion, which may contribute to the mobility and invasion of UM cells.

Considering the loss of membranous epithelial markers (TJP and CDH1) in IL-6-transformed UM cells, we set out to determine the potential role of IL-6 in EMT induction. Immunofluorescence showed an up-regulation of the cytoplasmic mesenchymal marker Vimentin, along with altered morphology after long-term IL-6 stimulation (Figure 3A). Western blotting results indicated that IL-6 exposure decreased the expression of the membranous epithelial marker TJP1 and increased the expression of Vimentin in a time-dependent manner (Figure 3B). Mechanistically, AP-1 family members were screened by RNA-seq analysis, among which JunB exhibited the most significant up-regulation in IL-6-transformed UM cells (Figure 3C). Furthermore, qRT-PCR and Western blot results confirmed the increase in mRNA and protein levels of JunB, c-Jun, and c-Fos (Figure 3D,E). However, despite the high homology between JunB and c-Jun, EMT characteristics of IL6/C918 cells (decreased TJP1 and induction of Vimentin) were blunted by JunB silencing, while the knockdown of c-Jun had no evident impact on the IL-6-induced EMT changes (Figure 3F). Thus, JunB is essential for IL-6-induced EMT in UM cells.

As IL-6/STAT3 signaling has been reported to drive the EMT process in various cancers, the STAT3 inhibitor Stattic was used to identify whether STAT3 mediated IL-6-induced JunB expression. As shown in Figure 4A, the induction of JunB in IL6/C918 cells was attenuated by treatment with Stattic. Similarly, the knockdown of STAT3 markedly undermined JunB expression in IL6/C918 cells, suggesting that IL-6/STAT3 signaling is critical for the up-regulation of JunB in IL-6-transformed UM cells (Figure 4B). The promoter sequence of the JunB gene was analyzed and two potential STAT binding elements (SBEs) were predicted (Figure 4C). Serial deletion constructs of the JunB promoter were examined by luciferase reporter assays to identify the transcriptional regulatory region responsive to IL-6/STAT3 signaling. The results indicated that the JunB promoter without the region between −200 and −100 lost the ability to respond to IL-6 stimulation (Figure 4D). Additionally, the region between −1500 and −1000 also appeared to be responsible for the transcriptional activity activated by IL-6 (Figure 4D). Further analysis revealed that the two regions (−1500 to −1000, −200 to −100) contained a predicted SBE (SBE1 and SBE2 respectively). A mutation in SBE1 significantly reduced the reporter activity induced by IL-6, while SBE2 mutation completely abolished the effect of IL-6 (Figure 4E). Importantly, ChIP-PCR analysis demonstrated that STAT3 directly bound to the JunB promoter region around SBE1 and SBE2 in IL6/C918 cells (Figure 4F). Consistent with the reporter assay, a stronger enrichment of STAT3 on SBE2 was observed than on SBE1. Altogether, these results indicate that IL-6/STAT3 signaling can transactivate JunB via a novel SBE in UM cells.

To investigate the function of IL-6/STAT3/JunB in EMT of IL-6-transformed UM cells, STAT3 silencing along with JunB overexpression was performed in IL6/C918 and IL6/OCM1A cells. It was shown that absence of STAT3 resulted in attenuated EMT markers (increased level of TJP1 and decreased level of Vimentin); however, the overexpression of JunB largely restored the molecular features in IL6/C918 and IL6/OCM1A cells (Figure 5A). In EMT-related functional assays, deprivation of STAT3 retarded the migration and invasion of IL6/C918 and IL6/OCM1A cells, while the abrogated migratory and invasive capacities were rescued by overexpressed JunB (Figure 5B,C). These data suggest that the IL-6/STAT3/JunB signaling axis facilitates a more aggressive phenotype of UM cells.

To further characterize the behavior of JunB in UM metastasis, migration and invasion were examined in C918 cells stably overexpressing JunB. As shown in Figure 6A,B, highly expressed JunB facilitated the migratory and invasive abilities of C918 cells. Then, based on RNA-seq data, the expression of EMT-related transcription factors (EMT-TFs) and matrix metalloproteinases (MMPs) were analyzed to explore the possible molecular mechanism. The results showed that overexpressed JunB significantly increased both the mRNA and protein levels of Snail, Twist1, MMP2, MMP14, and MMP19 (Figure 6C–F). Therefore, JunB may promote the migration and invasion of UM cells by regulating EMT-TFs and MMPs.

Data from the TCGA database (Uveal Melanoma: TCGA, Provisional) were analyzed to determine the correlation between JunB and IL-6/STAT3 in clinical UM specimens using the cBioPortal platform. A significant relationship between JunB and IL-6/STAT3 at the mRNA level was observed, as is shown on the dot plots (Figure 7A,B). With a threshold of > 0.3 or < −0.3 in either the Pearson or Spearman score, JunB was positively associated with IL-6 or STAT3, indicating the existence of the IL-6/STAT3/JunB pathway in UM tissues.

In the present study, two well-validated UM cell lines with different morphologies (epithelioid C918 and spindle OCM1A) were used to construct IL-6-transformed UM cells by long-term exposure to low-dose IL-6, which mimics the intraocular microenvironment of UM. IL-6-transformed UM cells were more aggressive than the parental cells, and stronger induction by IL-6 was observed in the IL6/C918 cells than in the IL6/OCM1A cells. Previous publications have suggested that epithelioid UM is more aggressive than spindle UM [18,19]. More recently, in vivo assays showed that C918-derived tumors completely disrupted the eye structure, while the structure of OCM1-grafted eyes remained intact [20]. Thus, the different behaviors of IL6-transformed C918 and OCM1A cells may arise from their different cellular contexts. Although IL-6 treatment enhanced the migratory and invasive capabilities of UM cells, proliferation was suppressed to some extent after IL-6 exposure. Consistently, RNA-seq results showed an evident down-regulation of cell-cycle drivers, such as E2F1, CDK4/6, CCNA2, and CDC6, in IL-6-transformed UM cells (Supplementary Table S2). Therefore, differing effects of IL-6 signaling on diverse biological processes in UM cells are interesting and warrant further investigation.

To understand the mechanisms underlying IL-6-induced aggressiveness in UM cells, RNA-seq was performed to analyze the gene expression profile of IL-6-transformed UM cells. The down-regulation of intercellular tight junction (TJ) molecules (TJP1/2) and up-regulation of focal adhesion (FA) molecules (FN1, ICAM-1, and integrins) was shown in IL-6-transformed UM cells, compared with the parental cells. Although the role of TJ in the progression of UM was rarely investigated, the expression of TJP1 is reduced or absent in several cancers, which is associated with poorly defined differentiation, higher metastatic frequency, and poor prognosis [21–23]. Mechanistically, decreased expression of TJ molecules leads to tumor dissociation and subsequent metastasis through the interruption of cell polarity and the promotion of the EMT program [24]. Importantly, inflammatory cytokines can disrupt TJ complexes by decreasing the expression of TJ molecules, including TJP1, which is consistent with our observations in UM [25]. Additionally, the importance of cell–ECM interactions, especially with FAs, in regulating and contributing to cancer migration and invasion has been increasingly accepted [26]. As the critical components of FA structures, both ECM molecules (FN1 and ICAM-1) and cell-surface receptors (integrin subunits) were induced by IL-6 in UM cells. Woodward et al. [27] found that invasive UM cells expressed higher levels of integrin subunits and adhered more strongly to FN1 than noninvasive cells. However, the role of ICAM-1 in UM is still controversial. A higher expression of ICAM-1 was found in UM tissues with a larger volume [28]. However, the loss of ICAM-1 expression was associated with an increased risk of UM metastasis [29]. Altogether, our study suggests that the proinflammatory cytokine IL-6 may facilitate UM metastasis through remodeling cell–cell and cell–ECM adhesions.

As a major mediator of IL-6 signaling, STAT3 is often constitutively activated in many solid tumors, and it transactivates the expression of genes responsible for cell growth, survival, angiogenesis, and invasion [30]. Here, we showed that phosphorylated STAT3 mediated IL-6-induced JunB up-regulation, which was required for the EMT and aggressiveness of UM cells. Apart from JunB, c-Fos was also induced by IL-6 stimulation. As a well-known STAT3 downstream target gene, c-Fos can trigger cell transformation and tumor metastasis through dimerization with Jun partners, and c-Fos has been shown to enhance the transactivating and transforming properties of JunB [31,32]. Hence, we hypothesized that IL-6-induced c-Fos may facilitate JunB in the acceleration of UM development. In addition to IL-6, STAT3 activation is also associated with the EMT programs induced by other cytokines, such as IL-8 and TGF-β, which are increased in UM-containing eyes [33,34]. Potential pathway crosstalk among these signals during UM EMT progression is interesting and warrant further investigation.

In the present study, we observed that the stable overexpression of JunB led to the induction of the EMT-TFs Snail and Twist1, as well as the secretion of extracellular matrix-degrading enzymes MMP2, MMP14, and MMP19, which may contribute to the migration and invasion of UM cells. Previous studies suggested that the down-regulation of EMT-TFs Zeb1, Twist1, and Snail reduced the invasive properties of UM cells, and the elevated mRNA levels of Zeb1 and Twist1 were associated with a more aggressive clinical phenotype in UM samples [20,35]. However, the induction of Zeb1 by IL-6 was not evident in our experiments. As the crucial ECM remodeler, MMP2 decreased when the migration and invasion of UM cells were suppressed, and the inhibition of MMP14 expression in UM cells impaired in vitro migration and in vivo invasiveness [36,37]. Although MMP19 had never been reported in UM, MMP19 was up-regulated during melanoma progression and increased the invasion of melanoma cells [38].

Collectively, our findings revealed that the EMT program in UM cells can be initiated by long-term stimulation with the proinflammatory cytokine IL-6. Further study showed that IL-6/STAT3 signaling induced the transactivation of JunB, resulting in EMT changes as indicated by the loss of cell–cell adhesion and the up-regulation of the mesenchymal marker Vimentin. Phenotype experiments further confirmed the driving effect of the IL-6/STAT3/JunB pathway on the migration and invasion of UM cells (Figure 7C). Therefore, investigating the role of IL-6 and its downstream pathway in UM EMT may help to deepen our understanding of UM metastasis.

This work was supported by Xuzhou Medical Outstanding Personnel Plan [XWJC 001].

The authors declare that there are no competing interests associated with the manuscript.

C.G. conceived and designed the study. C.G., Z.F., L.Q., and X.L. performed the experiments and wrote the manuscript. J.S., R.F., and S.L. analyzed and interpreted the data. C.G., Z.F., and S.L. contributed reagents and materials and helped to draft the manuscript. All authors read and approved the final manuscript.

EMT

epithelial–mesenchymal transition

IFN-γ

interferon γ

IL

interleukin

IP-10

interferon-inducible protein 10

MCP-1

monocyte chemotactic protein 1

MMP

matrix metalloproteinase

STAT3

signal transducer and activator of transcription 3

TNF-α

tumor necrosis factor α

TGF-β

transforming growth factor β

TJP1

tight junction protein 1

UM

uveal melanoma