Overexpression of RCC2 Enhances Cell Motility and Promotes Tumor Metastasis in Lung Adenocarcinoma by Inducing Epithelial–Mesenchymal Transition

This article is featured in Highlights of This Issue, p. 5323

Lung cancer, the foremost cause of global cancer-related mortality (1), is responsible for over a million deaths each year worldwide, with adenocarcinoma being the most common histologic type (2). Lung adenocarcinoma (LUAD) comprises approximately 40% of all lung cancer with increasing incidence in both Asian and Western countries (3). Notably, the average 5-year survival rate for LUAD is about 15% (4) and has not improved in the past few decades (5), mostly since patients eventually succumb to metastatic relapse after surgical resection, who are almost always incurable (6). Tumor metastasis is the major cause of LUAD-related death; moreover, somatic mutations of many oncogenes and perturbation of key pathways have been frequently discovered in LUAD (7). Therefore, it is strongly needed to reveal the molecular mechanisms underlying LUAD genesis and development and to identify effective prognosis biomarkers for high-risk patients with LUAD.

Regulator of chromosome condensation 2 (RCC2), also known as TD-60, was originally identified using human autoimmune antiserum at the spindle midzone in anaphase and telophase (8). RCC2 was reported as a component of the chromosomal passenger complex (CPC) together with Aurora B kinase (9), INCENP (10), and Survivin (11) involved in regulating chromosomes and the spindle assembly in mitosis and cytokinesis processes (12, 13), and recent studies showed that RCC2 was associated with the integrin adhesion complexes and involved in cell migration during interphase, suggesting its role in cell signaling and interphase cell-cycle progression (14–17). Growing evidence indicated that RCC2 was related to RCC1 (regulator of chromatin condensation 1), a guanine exchange factor (GEF) for Ran and also exhibited GEF activity for the small GTPases Rac1 (12) and RalA (13). The gene RCC2 was also reported to play a potential role in tumorigenesis. Matsuo and colleagues showed that RCC2, as the target of tumor suppressor miR-29c, was upregulated in gastric carcinoma tissues and contributed to the proliferation of gastric carcinoma cells by regulating cell-cycle progression (18). Two genome-wide SNP association analysis showed the SNP rs7538876, which mapped in the vicinity of RCC2, conferred risk of melanoma (19) and cutaneous basal cell carcinoma (BCC; ref. 20) progression, and a series of experiments confirmed the potential association of rs7538876 with the expression of RCC2. Feng and colleagues found that ENST00000439577, a long noncoding RNA (lncRNA), was positively correlated with the expression of RCC2. Furthermore, high expression of ENST00000439577 was associated with metastasis status and poor overall survival in non–small cell lung cancer (NSCLC; ref. 21). Although Bruun and colleagues demonstrated that the 5′UTR mutation of RCC2 reduced protein expression, it was significantly associated with poor prognosis in microsatellite-stable (MSS) tumors (22). However, the potential correlation between RCC2 and LUAD progression is still unclear.

Herein, we reported that overexpression of RCC2 in LUAD was important in the acquisition of an aggressive and poor prognostic phenotype. Overexpression of RCC2 promoted LUAD cell proliferation and metastatic capacity both in vitro and in vivo. More importantly, we demonstrated, for the first time, that RCC2 regulated LUAD cell epithelial–mesenchymal transition (EMT) and extracellular matrix remodeling via JNK activation, ultimately driving the aggressiveness of cancer cells. Our results provided functional and mechanistic links between RCC2 and LUAD.

Twelve fresh tissues including LUAD (n = 6) and paired adjacent normal lung (n = 6) were obtained from the Second Affiliated Clinical Hospital (Harbin Medical University, Harbin, China). No patient had received chemotherapy or radiotherapy before surgery. Samples used in this study were approved by the Institutional Review Board of Harbin Medical University (HMUIRB20160025). All procedures were conducted in accordance with the Declaration of Helsinki and International Ethical Guidelines for Biomedical Research Involving Human Subjects (CIOMS). Written informed consents were obtained from all patients. In addition, the tissue microarrays (TMAs) consisted of 122 pairs of human LUAD and their adjacent non-neoplastic lung tissues and TMA containing 120 human LUAD tissues were purchased from Outdo Biotech Co. Ltd.

We downloaded the RNA-seq data (level 3) and corresponding clinical information of 488 patients with lung adenocarcinoma and 58 normal tissues from The Cancer Genome Atlas (TCGA) data portal (https://portal.gdc.cancer.gov/legacy-archive/search/f). The expression levels for each gene were calculated as TPM values, which were defined by RSEM (23).

Human LUAD cell lines A549, NCI-H1650, and NCI-H23 were obtained from the ATCC. Other LUAD cell lines including NCI-H1975 and HCC827 were purchased from the Xiangf Biological Technology Co. Ltd. (Shanghai, China). A549 cell lines were routinely cultured in Ham F12K media. NCI-H1650, NCI-H1975, HCC827, and NCI-H23 were cultured in RPMI1640 media. All cell lines were routinely cultured in media supplemented with 10% FBS (PAA Laboraties GmbH) at 37°C in a humidified atmosphere containing 5% CO₂.

A549 and NCI-H1650 cells were transfected with RCC2-eYFP by Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. RCC2 constructs were kindly gifted from Dr. Sheng Xiao (Department of Pathology, Brigham and Women's Hospital, Harvard Medical School). For the establishment of the A549-RCC2 cell line stably expressing RCC2, 24 hours after transfection, cells were maintained in Ham F12K media containing 1,000 μg/mL of G418 (Calbiochem). After 3 weeks, resistant colonies stably transfected with RCC2 were pooled. Short interfering RNAs (siRNAs) specifically against RCC2 and corresponding scrambled siRNA were purchased from RiboBio and then transiently transfected into A549-RCC2 and HCC827 cell lines using Lipofectamine 2000 reagent (Invitrogen) according to the manufacturer's protocol. The gene silencing effect was measured by Western blotting 48 hours posttransfection.

Western blot analysis was performed according to the standard protocol. Proteins were extracted from cells and tissues with RIPA buffer (Thermo Fisher Scientific) complemented with protease inhibitors (Roche). For Western blot analysis, protein extracts were subjected to electrophoresis on 10% SDS-PAGE and transferred onto polyvinylidene fluoride membranes (Millipore). Immunoblots were blocked with 5% BSA in TBS/Tween 20 and incubated with primary antibodies (Supplementary Table S1) overnight at 4°C. The bands were visualized by the ChemiDoc MP Imaging System (Bio-Rad). The specificity of anti-RCC2 antibody used in this study was verified according to the proposal for validation of antibody (ref. 24; Supplementary Fig. S1).

Total RNA was extracted using the High Pure RNA Isolation Kit (Roche), and reverse transcription was performed with the PrimeScript RT Reagent Kit Perfect Real Time (Takara). The cDNA was subjected to quantitative RT-PCR using the LightCycler 480SYBR Green I Master (Roche) and the assay was performed on a CFX96 Real-Time System (Bio-Rad) following the manufacturer's instructions. GAPDH was used to normalize the amount of cDNA between different samples. Primer sequences were listed as follows: RCC2 forward 5′-TTCCTTTGGGTGCCCTGAA-3′, reverse 5′-GGCAGAATCTGTCCATCTTTCG-3′, GAPDH forward 5′-ATCACTGCCACCCAGAAGAC-3′, reverse 5′-TTTCTAGACGGCAGGTCAGG-3′.

Cells were seeded in triplicate in 24-well plates. Cell numbers were counted every 24 hours after cell adhesion over a 5-day period. The experiment was independently repeated three times.

For the wound-healing assay, cells were grown on a 6-well plate until 95% confluence and monolayers were scratched with a pipette tip. Cell migration was recorded at 0, 24, and 48 hours after wound scratch. Three independent experiments were performed.

Invasion and migration assays were performed using Corning chambers (Corning) with or without Matrigel following the manufacturer's protocol. Briefly, cell suspensions prepared in media containing 2% FBS were added into top chambers, while media containing 20% FBS was placed in the bottom chambers as a chemoattractant stimulus. The cells were incubated for 24 hours or 48 hours at 37°C and allowed to migrate or invade through the membrane filter. Afterward, the nonmigratory and noninvasive cells on the top surface were gently removed by a cotton swab. Then, cells that had invaded or migrated to the bottom surface of the membrane were fixed with methanol and stained with hematoxylin and eosin. Cells in four randomly microscopic fields (at 100× magnification) were counted and photographed. Both experiments were independently repeated in triplicate.

Cells were transfected with RCC2-eYFP and cultured in media. After fixing with 4% paraformaldehyde (PFA) for 15 minutes, cells were then permeabilized with 0.5% Triton X-100 for 20 minutes. Primary antibody against RCC2 (1:100) was incubated overnight at 4°C. Sequentially, the cells were incubated with secondary antibodies anti-rabbit at 37°C for 30 minutes. DAB-H₂O₂ was used for the color reaction and all of the cells were counterstained with hematoxylin.

Immunohistochemical staining was performed using PowerVision Two-Step Histostaining Reagent (Zhongshan Golden Bridge) according to the manufacturer's instructions. After deparaffinization and rehydration, tissue slides were routinely treated with 3% H₂O₂ at room temperature for 10 minutes. EDTA (10 mmol/L) was used for antigen retrieval. Then, rabbit primary antibodies against human RCC2 (1:100 for TMAs, 1:400 for mouse tissues), N-cadherin (1:100 for TMAs), α-SMA (1:200 for TMAs), β-catenin (1:200 for TMAs), E-cadherin (1:200 for TMAs), and p-JNK (1:50 for TMAs) were added to the tissues and incubated overnight at 4°C in a moist chamber. Sequentially, the slides were incubated with goat anti-rabbit IgG at 37°C for 30 minutes. After washing in PBS, DAB-H₂O₂ was used for the color reaction. Finally, all of the slides were counterstained with hematoxylin, dehydrated, and mounted. For negative controls, the processing was performed as described above, except that they were incubated overnight without the primary antibody at 4°C. Immunoreactivity on TMAs was assessed independently by three researchers. The specificity of anti-RCC2 antibody was verified according to the proposal for validation of antibody (ref. 24; Supplementary Fig. S1).

Cells were fixed with cold ethanol for 20 minutes and permeabilized with 0.1% Triton X-100 for 5 minutes. The samples were washed with PBS and blocked with 5% BSA in PBS. Subsequently, cells were incubated with primary antibodies against β-catenin, E-cadherin, N-cadherin (ProteinTech, 1: 50 dilution) overnight at 4°C and then incubated with fluorescence-labeled rabbit antibody. The nuclei were counterstained with 4′-6-diamidino-2-phenylindole (DAPI) for 5 minutes.

All animal experiments were conducted in accordance with the ethical standards and national guidelines and approved by the Institutional Review Board of Harbin Medical University (HMUIRB20160025). Female BALB/c-nude mice (4–5 weeks old), purchased from Vital River Laboratory (Beijing, China), were kept under standard conditions. Xenograft tumor growth assay was established by subcutaneous injection of A549-RCC2 cells and control cells (7 × 10⁶, suspended in PBS) into the right dorsal flank. To preparation of xenograft model for RCC2 knockdown study on tumor growth, HCC827 cells (1 × 10⁷) were subcutaneously injected into the right flanks (four mice per group). Ten days after tumor inoculation, cholesterol-conjugated RCC2 siRNAs (siRNA1 and siRNA2) and their negative controls were from RiboBio for in vivo siRNA delivery. siRNA (10 nmol) in 0.1 mL saline buffer was injected into tumor mass once every 3 days for 3 weeks (25, 26). Tumor volume (TV) was measured by calipers and calculated as formula: TV (mm³) = (L × W²)/2 (L, long diameter; W, wide diameter). After sacrifice, tumors were excised for following assays. For the intravenous mouse model, two groups of four mice each were given intravenous tail-vein injections of 6 × 10⁶ A549-RCC2 cells and control cells, respectively. After 8 weeks, the mice were sacrificed, and the tumor nodules formed on the lung surfaces were counted. Lungs were collected and embedded in paraffin for further hematoxylin and eosin (H&E) staining and IHC analysis.

All statistical analysis were performed using R software (version 3.2.2; http://www.R-project.org/) or SPSS 17.0 (IBM). The differential expression level of RCC2 between LUAD and adjacent normal tissues was evaluated by McNemar χ² test. The correlation between RCC2 expression and clinicopathologic features of patients with LUAD was analyzed by the χ² test or Fisher exact test. Survival curves were obtained using the Kaplan–Meier method, and statistical analysis was performed using the log-rank test. Cox proportional hazards regression model was performed for univariate and multivariate survival analyses. Other data were expressed as the means ± SD, and the independent Student t test was applied to analyze the statistical significance between two groups. P < 0.05 was considered statistically significant.

To investigate the expression status of RCC2 in LUAD, TMAs, which contained 122 pairs of LUAD specimens and corresponding adjacent noncancerous tissues and detailed pathologic information with survival prognosis of patients were examined by IHC staining. The frequency of positive staining of RCC2 was significantly higher in tumors (88/122; 72.13%) than that in nontumor tissues (11/122; 9.02%; P < 0.001). IHC staining of RCC2 in representative samples of LUAD and normal lung tissues was shown in Fig. 1A and B. To further confirm our observation, we expanded the analysis to a larger cohort of TCGA RNA-seq expression data that contained 488 lung adenocarcinoma and 58 normal lung tissues. Again, RCC2 expression in adenocarcinoma was statistically (P < 0.001) higher than that in the normal tissues (Fig. 1C). In addition, RCC2 protein expression between six pairs of fresh LUAD tissues (T) with their adjacent normal lung tissues (N) was compared by Western blotting. As shown in Fig. 1D, the result revealed a similar trend that RCC2 overexpression was detected in 4/6 (66.67%) of LUAD tissues compared with the matched nontumor tissues.

We next evaluated the association of RCC2 overexpression with clinicopathologic features in 122 LUAD samples with informative IHC data. The results showed that RCC2 overexpression was significantly associated with T status of tumor (P = 0.002), lymph node metastasis (P = 0.004) and a more advanced clinical stage (P = 0.001; Table 1). Kaplan–Meier survival analysis showed that patients with LUAD with higher expression of RCC2 had significantly poorer overall survival using TMA data (P = 0.003, log-rank test; Fig. 1E), TCGA dataset (P = 0.032, log-rank test; Fig. 1F) of LUAD and the Gene Expression Omnibus (GEO) dataset (GSE31210) from PrognoScan (P = 0.002, log-rank test; Supplementary Fig. S2). Further multivariate Cox regression analysis demonstrated that RCC2 overexpression was an independent prognostic factor for poor survival of patients with LUAD in TCGA dataset (HR, 1.687; 95% CI, 1.098–2.594; P = 0.017; Table 2), although there was no significance in TMA data (Supplementary Table S2). In conclusion, these findings suggested that overexpression of RCC2 was frequently detected in LUAD, and its overexpression was associated with tumor metastasis and poor prognosis in patients with LUAD.

To explore the impact of RCC2 on the biological characteristics of LUAD cells, we first assessed expression levels of endogenous RCC2 in a panel of five LUAD cell lines (A549, NCI-H1650, HCC827, NCI-H1975, NCI-H23) by Western blotting (Fig. 2A). A549 and NCI-H1650 cell lines showed low endogenous RCC2 expression in this group and therefore were selected for an overexpression study. A RCC2 expression construct was transfected into A549 and NCI-H1650 cells, and then the expression of exogenous RCC2 was demonstrated by Western blotting (Fig. 2B). These cells were subjected to a cell growth assay, which showed significantly increased cell proliferation in RCC2-transfected cells as compared with the control cells (Fig. 2C). Because our clinical correlation analysis revealed association of RCC2 overexpression with LUAD metastasis, we investigated the impact of RCC2 on LUAD cell motility and invasiveness. Wound-healing assay showed that overexpression of RCC2 enhanced A549 cell migration at the edge of exposed regions (Fig. 2D). Furthermore, migration and Matrigel invasion assays in vitro were also performed in A549 and NCI-H1650 cells. The results showed that the expression of exogenous RCC2 in LUAD cells significantly increased cell motility (P < 0.001; Fig. 2E and F) and invasion potential (P < 0.001; Fig. 2G and H) as compared with control cells. These results indicated that RCC2 overexpression enhanced LUAD cell proliferation, migration, and invasion in vitro.

To further confirm whether RCC2 silencing could inhibit cell growth and invasiveness, the RCC2 gene was silenced by RNA interference (RNAi) using two targeted siRNAs (siRNA1 and siRNA2) in A549 cells with stable RCC2 overexpression and HCC827 cells with relatively high endogenous RCC2 expression. Both siRNAs efficiently knocked down RCC2 in LUAD cells (Fig. 3A). RCC2 depletion led to decrease cell growth (Fig. 3B) and wound-healing ability as compared with control cells (Fig. 3C and D). Migration and Matrigel invasion assays also demonstrated that ablation of RCC2 markedly reduced the migration and invasion capacity of both A549-RCC2 and HCC827 cell lines (P < 0.01; Fig. 3E and F). Collectively, these results provided evidence that RCC2 silencing could inhibit LUAD cell tumorigenicity in vitro.

EMT and extracellular matrix remodeling are the key events in driving tumor invasion and metastasis. To investigate the effects of RCC2 on EMT and extracellular matrix remodeling, we analyzed the expression levels of EMT markers, EMT-related transcription factors and MMPs. RCC2-transfected A549 and NCI-H1650 cells showed increased expression of mesenchymal markers (N-cadherin and α-SMA), EMT-related transcription factors (Snail and Slug), and two MMPs (MMP-2 and MMP-9), while the expression of β-catenin and E-cadherin was decreased (Fig. 4A). Immunofluorescent staining assays confirmed the increased expression of N-cadherin and decreased expression of β-catenin in the A549-RCC2 cells (Fig. 4C). On the contrary, after the silence of RCC2 in A549-RCC2 and HCC827 cells, the levels of β-catenin and E-cadherin were upregulated, whereas the levels of N-cadherin, α-SMA, Snail, Slug, MMP-2, and MMP-9 were downregulated obviously (Fig. 4B). To determine the clinical relevance, TMAs containing 120 human LUAD specimens (119 specimens available) were subjected to IHC staining for RCC2 and EMT markers (N-cadherin, α-SMA, β-catenin, and E-cadherin) in serial sections (Supplementary Fig. S3). Consistently, IHC analysis confirmed that RCC2 expression level was positively correlated with the expression levels of N-cadherin and α-SMA, and inversely correlated with the expression levels of β-catenin and E-cadherin (Supplementary Table S3). The correlations of the RCC2 expression level with N-cadherin, Snail, MMP-2, and MMP-9 were further studied in the same cohort of LUAD samples of TCGA dataset. The results confirmed that RCC2 expression was significantly correlated with the expression of N-cadherin (R = 0.284; P < 0.001), Snail (R = 0.238; P < 0.001), MMP-2 (R = 0.171; P < 0.001), and MMP-9 (R = 0.176; P < 0.001; Fig. 4D). Taken together, these results suggested that the RCC2 is associated with EMT and extracellular matrix remodeling, which likely contribute to tumor metastasis in LUAD.

To assess whether overexpression of RCC2 could enhance the tumorigenic capacity in vivo, the xenograft tumor mouse model was established by subcutaneously injecting A549-Vec and A549-RCC2 into the right dorsal flanks of nude mice. During the course of 9 weeks, tumor volume was monitored by a caliper, which showed significantly larger tumors in the group with RCC2 overexpression as compared with control mice (P < 0.05; Fig. 3A). At the end of the experiments, xenograft tumors were isolated, and the weights were measured, which again showed increased tumor size and weight in mice with RCC2 overexpression (Fig. 3B and C). Moreover, these tumors appeared multinodular in the A549-RCC2 group compared with the control group. Results from the hematoxylin and eosin (H&E) staining confirmed the tumor formation (Fig. 3D). Furthermore, the expression of RCC2 in xenograft tumors developed from the RCC2-transfected cells was confirmed by IHC staining and qRT-PCR (Fig. 3D and E). We further verified several EMT markers (N-cadherin and E-cadherin) by Western blotting using tumor tissues from xenograft mice. Compared with the A549-Vec group, N-cadherin protein level significantly increased while the expression of E-cadherin decreased in A549-RCC2 group (Fig. 3F).

To demonstrate whether RCC2 silencing could inhibit tumorigenicity in vivo, HCC827 cells that showed relatively high endogenous RCC2 expression were subcutaneously injected into the right flanks (four mice per group). Ten days after tumor inoculation, intratumoral injection of cholesterol-conjugated RCC2 siRNAs (siRNA1 and siRNA2), and their negative controls (10 nmol siRNA in 0.1 mL saline buffer) were used for in vivo siRNA delivery once every 3 days for 3 weeks (25, 26). RCC2 siRNA treatment significantly decreased tumor growth, as shown by significantly reduced tumor size and weight compared with control group (Supplementary Fig. S4A–S4C). To confirm that RCC2 was knocked down in vivo by siRNA, we measured RCC2 expression by IHC staining and Western blotting and found that it was significantly decreased in RCC2 siRNA–treated xenografts (Supplementary Fig. S4D and S4E).

To evaluate whether RCC2 could promote tumor metastasis in vivo, we injected A549-Vec and A549-RCC2 in the tail vein intravenously of nude mice (four mice/group), respectively. Eight weeks after injection, mice were sacrificed and the metastatic nodules formed on the lung surfaces were examined. As showed in Fig. 5G, the mice injected with A549-Vec cells formed fewer nodules on the lung surfaces than the mice injected with A549-RCC2 cells (P < 0.05, independent Student t test). Metastatic nodules on the surfaces of mice lungs were confirmed by H&E staining, whereas the expression level of RCC2 in the nodules was also verified by the IHC staining (Fig. 5H). Therefore, our data demonstrated that high RCC2 expression enhanced tumor growth and metastasis in vivo, which was consistent with our in vitro and clinical findings.

JNK pathway is critical for the progression and maintenance of phenotypic and cellular changes associated with EMT (27). To further explore the molecular mechanisms of RCC2, we evaluated the possibility of RCC2 contributing to EMT and extracellular matrix remodeling by regulating JNK signaling. We observed that increased expression of activation JNK1/2 in RCC2-transfected cells by Western blot analysis (Fig. 6A). As expected, opposite expression pattern of JNK1/2 was examined in RCC2-silenced cells (Fig. 6B). Nevertheless, there was no obvious influence of RCC2 on ERK and p38 phosphorylation. In addition, a significant positive correlation between the overexpression of RCC2 and p-JNK1/2 was evaluated in the same cohort of TMAs containing 120 human LUAD tissues (119 specimens available; Supplementary Fig. S5). The LUAD samples had high RCC2 expression, which correlated with high JNK1/2 activation (representative case 2). Accordingly, LUAD samples with low RCC2 expression showed reduced JNK1/2 activation (representative case 1; Supplementary Table S3). These results suggested the involvement of JNK pathways in the effects of RCC2 on LUAD. To confirm that the aggressive effect of RCC2 was mediated by the activation of the JNK1/2, the JNK inhibitor SP600125 was used in A549-RCC2 cells. The results showed that SP600125 could effectively decrease expression levels of phosphorylated JNK1/2 induced by RCC2 in A549-RCC2 cells (Fig. 6C). Wound-healing, migration assays, and invasion Transwell assay demonstrated that migratory and invasive ability of A549-RCC2 cells were dramatically inhibited after SP600125 treatment (Fig. 6D and E). In addition, SP600125 also inhibited the RCC2-induced EMT and the expression of MMP-2 and MMP-9 (Fig. 6F). Immunofluorescent staining also confirmed the changes of EMT markers (E-cadherin and N-cadherin) after SP600125 treatment in the A549-RCC2 cells (Fig. 6G). These data, taken together, provided evidence that MAPK–JNK signaling was responsible for the RCC2-induced metastasis in LUAD cells (Fig. 6H).

LUAD is considered to be a multistep process, involving a premalignant lesion to adenocarcinoma in situ, followed by invasive adenocarcinoma (28). Notably, the average 5-year survival rate for LUAD is poor, and tumor metastasis is the major cause of LUAD-related death. A better understanding of the underlying molecular mechanisms of LUAD development and metastasis is highly desirable for improvement of clinical management. The gene RCC2, located on chromosome 1p36, has been reported to play a potential role in tumorigenesis. To our knowledge, the abnormalities of RCC2 was associated with gastric carcinoma (18), colorectal cancer (22) directly, and basal cell carcinoma (BCC; ref. 20), melanoma (19), NSCLC (21) indirectly. Recent study demonstrated that RCC2 formed a complex with the regulator of actin filament assembly cortactin (29), which had been linked to invasive cancers including colorectal cancer, melanoma, and glioblastoma, making cortactin as an invasive cancers biomarker (30). RCC2 also acted as a dual activity-limiter of Rac1 and Arf6 to regulate adhesion complex formation, cell spreading, and directional migration (14). Although there was no direct evidence of RCC2 involving in tumor metastasis, all these RCC2-interacting proteins, including cortactin, Rac1 and Arf6, had been reported to have direct association with tumor invasiveness and metastasis already (31–34). It was noteworthy that these findings gave the potential links between RCC2 and cancer metastasis. In this study, we first demonstrated that RCC2 was frequently overexpressed in LUAD and RCC2 overexpression was an independent prognostic factor for poor survival of patients with LUAD.

In the current study, Western blotting results showed that the majority of the LUAD tissues examined had high levels of RCC2 expression. Subsequently, IHC staining on TMAs that contained 122 pairs of LUAD specimens and matched nontumor tissues was performed to validate the frequent overexpression of RCC2 in 88 of 122 (72.13%) of LUADs, which was significantly correlated with lymph node metastasis and a more advanced clinical stage, suggesting that upregulated expression of RCC2 in LUAD might facilitate the metastatic phenotype. Furthermore, the analysis of a larger cohort of TCGA RNA-seq expression data that contained 488 lung adenocarcinoma and 58 normal lung tissues confirmed that RCC2 was frequently overexpressed in LUAD tissues versus normal tissues. Moreover, survival analysis showed its important role in the cancer development that patients with RCC2 overexpression had poorer overall survival using multiple datasets. Furthermore, univariate and multivariate Cox regression analysis demonstrated that RCC2 was an independent prognostic factor for poorer outcome in patients with LUAD using TCGA expression dataset, suggesting that RCC2 expression could be used as a potential predictor as it is intimately involved in LUAD progression. These findings underscore a potentially important role of RCC2 as an underlying biological mechanism in the development and progression of LUAD.

To clarify the biological function of RCC2 in regulating LUAD cell motility and invasiveness, a series of in vitro and in vivo assays were conducted. The results showed that the ectopic overexpression of RCC2 could substantially promote cell proliferation, migration, and invasiveness. On the contrary, gene silencing of RCC2 by siRNA inhibited its effect on cell growth, motility, and invasion. In subcutaneous injection and the tail-vein injection mouse models, overexpression of RCC2 resulted in a significant increase in the tumor growth and in the number of lesions of lung metastasis, while knockdown of RCC2 by intratumor injection of cholesterol-conjugated RCC2 siRNA significantly inhibited tumorigenicity in vivo. All these findings strongly supported that overexpression of RCC2 played an important role in LUAD invasion and metastasis.

Metastasis represented a multistep cell biological process termed the invasion–metastasis cascade including two key events, EMT and extracellular matrix remodeling (35). EMT, involved in normal embryonic morphogenesis initially, played a pivotal role in tumor invasion and metastasis during tumor progression by suppressing expression of epithelial markers and inducing expression of mesenchymal markers, as well as transcription factors (36). In our studies, we found that overexpression of RCC2 had a significant impact on EMT in vitro and in vivo and upregulated the expression of EMT transcription factors Snail and Slug. In contrast, opposite results were found in RCC2 silencing LUAD cells in vitro. Clinical relevance analysis indicated that there were significant correlations between the expression of RCC2 and EMT markers in LUAD samples. In addition, recent studies showed that RCC2-depleted colon cancer cells appeared more distinctly epithelial-like (22), which was consistent with our results that RCC2 could induce EMT. Besides, matrix metalloproteinases (MMP), as modulators of the tumor microenvironment, were essential for extracellular matrix remodeling associated with tumorigenesis and tumor metastasis (37, 38). As expected, our studies showed that overexpression of RCC2 could stimulate the expression of MMP-2 and MMP-9. More importantly, using the LUAD samples of TCGA dataset, the results that RCC2 expression was significantly associated with the expression of N-cadherin, Snail, MMP-2, and MMP-9 confirmed our experimental data again. Thus, LUAD cells overexpressing RCC2 probably undergo EMT and extracellular matrix remodeling to achieve higher motility and invasiveness.

To date, however, the molecular mechanisms by which RCC2 regulated EMT and expression of MMP-2 and MMP-9 to promote cancer cell migration and invasiveness have not been elucidated. Previous studies found that the activation of MAPK–JNK pathway was closely associated with lung tumorigenesis (39, 40). JNK not only could regulate critical EMT genes through inducing key transcription factors to influence the progression of phenotypic changes but also regulated MMPs (such as MMP-1, MMP-2, and MMP-9), which played important roles in cell outgrowth, tissue destruction, and inflammation (27, 41–43). However, the detail regulatory mechanisms remain unclear. Our results indicated that the expression of phosphorylated JNK increased or decreased according to overexpression or silence of RCC2. Importantly, RCC2 overexpression was positively correlated with JNK activation in human LUAD samples. Moreover, the enhancement of RCC2 on LUAD cell motility and invasiveness and the RCC2-induced EMT and expression of MMP-2 and MMP-9 were dramatically prevented when JNK was inhibited. Therefore, the role of RCC2 in promoting EMT and extracellular matrix remodeling might be carried out through the activation of the MAPK–JNK signaling pathway, ultimately driving LUAD cells motility and invasiveness. Further work is needed to clarify the mechanisms of how RCC2 activates JNK in detail.

In conclusion, we first demonstrated that RCC2 played an important role in the development and progression of LUAD. Overexpression of RCC2 in LUAD was significantly associated with tumor metastasis and was an independent prognostic factor for poorer outcomes in patients with LUAD. Furthermore, functional and mechanistic studies revealed that RCC2 regulated EMT and extracellular matrix remodeling to promote cancer cell motility and tumor metastasis through the activation of the MAPK–JNK signaling. Therefore, our studies suggested that RCC2 may serve as a clinically useful diagnostic and prognostic biomarker, and a potentially therapeutic target in LUAD.

No potential conflicts of interest were disclosed.

Conception and design: S. Xiao, S. Fu, Y. Jin

Development of methodology: B. Pang, S. Fu, Y. Jin

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): B. Pang, N. Wu, R. Guan, L. Pang, X. Li, S. Li, L. Tang, J. Dai

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): B. Pang, N. Wu, L. Pang, D. Sun, H. Sun, S. Fu, Y. Jin

Writing, review, and/or revision of the manuscript: B. Pang, N. Wu, L. Pang, S. Xiao, S. Fu, Y. Jin

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): X. Li, L. Tang, Y. Guo, J. Chen, D. Sun, H. Sun, J. Bai, G. Ji, P. Liu, A. Liu, Q. Wang

Study supervision: S. Fu, Y. Jin

This work was supported by the National Natural Science Foundation of China (grant no. 81372784, to Songbin Fu; and grant no. 81371617, for Yan Jin), the Program for Changjiang Scholars and Innovative Research Team in University of China (grant no. IRT1230; to Yan Jin), and the National Key Research and Development Program of China (grant no. 2016YFC1000504; to Songbin Fu), the Outstanding Youth foundation of Heilongjiang Province of China (grant no. JC201215; for Yan Jin).

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.