Knockdown of Myoferlin Suppresses Migration and Invasion in Clear-Cell Renal-Cell Carcinoma

Background/Aim: Myoferlin (MYOF) has emerged as an oncogenic protein in various human cancer types. This study was conducted to investigate comprehensively the expression and functional properties of MYOF in clear-cell renal-cell carcinoma (ccRCC) with respect to its value as diagnostic biomarker and therapeutic target. Materials and Methods: mRNA and protein expression of MYOF were assessed by quantitative polymerase chain reaction and immunohistochemistry. siRNA-mediated knockdown of MYOF was performed in the RCC cell line ACHN followed by proliferation, migration and invasion assays. Results: MYOF mRNA and protein expression were significantly up-regulated in ccRCC. Higher mRNA levels were measured in advanced tumors. MYOF protein expression was increased in tumors with higher histological grades, and those with positive lymph node and surgical margin status. MYOF knockdown led to reduction of migration and invasion in ACHN cells, whereas expression of angiogenesis-associated genes tyrosine-protein kinase receptor-2 (TIE2), angiopoietin 2 (ANG2) and caveolin-1 (CAV1) was up-regulated following knockdown. Conclusion: MYOF may serve as a diagnostic biomarker of tumor progression and a potential therapeutic target in ccRCC.

Renal cell carcinoma (RCC) is the most frequent solid renal malignancy and represents 2-3% of all cancer cases worldwide. Clear cell RCC (ccRCC) is the most common (70-80%) subtype, followed by papillary (10%), chromophobe (5%) and other subtypes (1). Up to 30% of all patients will experience tumor progression such as local relapse or distant metastasis after surgery (2). Implementation of tyrosine kinase receptor inhibitors or mechanistic target of rapamycin inhibitors has achieved improvement in therapy of metastatic RCC. In particular, the use of the immune checkpoint inhibitors nivolumab and ipilimumab marked the beginning of a promising era of immunotherapy. However, the response to therapy is still heterogeneous. Thus, it is important to identify novel biomarkers as progression or predictive factors, especially with regard to targeted therapy.
Ferlins are a family of six large (~200-240 kDa) transmembrane-anchored proteins noted for their role in membrane repair, endocytosis and vesicle trafficking (3). In recent years, ferlins have attracted interest as carriers of disease-causing mutations and the relevance of specific ferlins to carcinogenesis was identified. While mutations of dysferlin (DYSF; formerly FER1L1) cause limb-girdle muscular dystrophy type 2B (4), mutations of otoferlin (OTOF; formerly FER1L2;) were shown to be responsible for an autosomal-recessive form of non-syndromic deafness (5). To date, mutations of the other ferlins such as myoferlin (MYOF; formerly FER1L3), FER1L4, FER1L5 and FER1L6 have not been associated with any distinct disease but their expression seems to play an important role in carcinogenesis. For instance, we revealed that lncRNA FER1L4 is an independent prognostic factor in ccRCC (6). MYOF expression was up-regulated in breast cancer, non-small-cell lung cancer and pancreatic adenocarcinoma (7)(8)(9)(10). Recent studies also demonstrated MYOF overexpression in ccRCC (11,12). MYOF has emerging oncogenic impact and promotes cancer cell proliferation via different molecular mechanisms. In melanoma cells, down-regulation of MYOF suppressed vasculogenic mimicry via inhibition of matrix metalloproteinase-2 expression and induction of mesenchymal-to-epithelial transition (13). Furthermore, MYOF sustains the mitochondrial network in colon cancer. MYOF silencing was shown to lead to reduced oxidative phosphorylation in human colon cancer cells and as a consequence reduced cell growth and increased apoptosis (14). Additionally, MYOF was involved in angiogenesis and has been linked to angiogenesis-associated receptors and proteins such as vascular endothelial growth factor receptor 2 (VEGFR2) and tyrosine-protein kinase receptor-2 (TIE2) (15).
However, the function and molecular mechanisms of MYOF in tumorigenesis have not been fully understood. Little is known about the functional role of MYOF in ccRCC. In this study, we investigated its expression and functional properties in ccRCC cells to identify its value as a diagnostic biomarker and therapeutic target.

Materials and Methods
The Cancer Genome Atlas (TCGA) analysis. Relative mRNA expression and clinical data from patients with ccRCC were extracted from TCGA by TCGA browser (https://www.cancer.gov/tcga; v0.9.2) to investigate mRNA expression of MYOF. Analysis was carried out using SPSS Statistics v23 (IBM, Ehningen, Germany). The complete cohort comprised 603 samples of patients with available mRNA expression and clinical data (522 ccRCC and 81 benign).
Patients. We examined the expression profile of MYOF mRNA in 82 ccRCC and 44 benign renal tissue samples. Protein expression was assessed in paraffin-embedded RCC tissues on tissue microarrays (TMAs) from patients treated at University Hospital Bonn. The cohort included benign renal tissue and RCC samples (ccRCC n=142, benign renal tissue n=30). Fresh frozen renal tissue samples from patients who underwent radical or partial nephrectomy were collected from the Biobank of the Center for Integrated Oncology Cologne-Bonn. All samples were collected between 1997 and 2014 at the Department of Urology at the University Hospital Bonn. The clinicopathological parameters of the patients are shown in Table I.
Cell culture. The ACHN RCC cell line was obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). Cells were grown in RPMI1640 medium supplemented with 10% inactivated fetal bovine serum, 1% penicillin-streptomycin antibiotics and 1% glutamine (all Gibco, Life Technologies, Carlsbad, CA, USA) in an incubator with 5% CO 2 at 37˚C and 85% humidity.
SiRNA-mediated knockdown. Transfection of the ACHN cell line was performed with 100 nmol/l siRNAs (MYOF: HSS120115, Thermo Fisher Scientific, Darmstadt, Germany) using FuGENE ® HD Transfection Reagent (Promega, Mannheim, Germany) for 96 hours. The procedures were performed according to the guideline of the manufacturer. BLOCK-iT™ Fluorescent Oligo (Thermo Fisher Scientific) was used as negative control and indicator of transfection efficiency.
RNA isolation. The RNA isolation was performed as described in detail elsewhere (6). In brief, total RNA from fresh frozen tissues was isolated using mirVana™ miRNA Isolation Kit and treated with DNA-free™ Kit (both Ambion, Foster City, CA, USA). A NanoDrop 2000 spectrophotometer was used for measurement of RNA quantity (Thermo Fisher Scientific, Wilmington, DE, USA). RNA integrity was proven by agarose gel electrophoresis. cDNA was transcribed using 1 μg total RNA (PrimeScript™ RT reagent Kit with gDNA Eraser; Takara Bio, Saint-Germain-en-Laye, France).  Immunohistochemistry. A total of 172 specimens (30 benign renal tissue samples and 142 ccRCC samples) arranged on TMAs were used for this study. In brief, tissues were formalin-fixed and paraffin-embedded, thereupon cut into 4-μm-thick sections and placed on object plates. After staining with hematoxylin and eosin, representative areas of benign tissues and primary tumor were detected and circled by a pathologist. Three cores of the representative areas each measuring 0.6 mm in diameter were assembled into TMA blocks using a semiautomatic tissue arrayer (Beecher Instruments, Sun Prairie, WI, USA). TMA sections were assessed again to confirm the histology. The immunohistochemical staining was performed using the Ventana Benchmark automated staining system (Ventana Medical System, Tuscon, AZ, USA). The slides were incubated with primary antibodies against MYOF (dilution 1:50 with Ventana diluent). Signal was detected using an ultraView Universal 3-3'-diaminobenzidine detection kit (Ventana Medical System) conjugated with secondary antibody. Afterwards, the sections were counterstained with hematoxylin and bluing reagent, rehydrated and mounted. Staining was evaluated independently by two readers (AC and YT). Combined membranous and cytoplasmic expression was determined as MYOF is expressed both at the plasma membrane and in the membrane of cytoplasmic endosomes (18). To quantify MYOF protein expression, staining intensities were evaluated according to a score of 0 (no expression) to 3 (strong expression). The cut-off to determine high levels of expression was set at ≥2.
Proliferation assay. EZ4U cell proliferation assay kit (Biomedica Group, Vienna, Austria) was used according to the manufacturer's protocol. The siRNA transfections for proliferation assays were performed in 96-well plates. In each well of a 96-well plate, 3×10 4 ACHN cells (passage 18) were seeded in 200 μl cell culture medium. siRNA-mediated knockdown of MYOF was then performed. Afterwards, cells were incubated to adhere and grow for 96 hours. After incubation, 20 μl of EZ4U substrate solutions were added. Finally, the resultant solution was incubated for 2 hours until the color of the solution changed from yellow to orange. The absorbance was measured using a microplate reader (Spectra Thermo; Tecan Group, Crailsheim, Germany) at 450 nm wavelength. Each experiment was repeated at least three times.
Migration and invasion assays. siRNA transfections for migration and invasion assays were performed in 6-well plates. Cells were trypsinized and seeded into migration Boyden chambers 72 hours after transfection. A total of 5×10 4 cells were plated in the upper chamber of migration inserts (VWR, Darmstadt, Germany) containing fetal calf serum-free medium. The lower chamber was filled with medium containing 10% fetal calf serum for chemotactic attraction. After 24-hour incubation, cells on the upper surface were removed with the help of a cotton swab. Cells invading the lower surface of membrane were fixed with 4% paraformaldehyde (Merck, Darmstadt, Germany), stained with hematoxylin (Waldeck, Münster, Germany), and washed with water. Membranes were scanned and manually evaluated in four randomly selected fields by counting. Each experiment was repeated at least three times. The invasion analysis was carried out similarly to the migration experiment except for the use of Matrigel invasion chambers (VWR) and a cell number of 7.5×10 4 /well.

Statistical analyses. Statistical analyses (Mann-Whitney U-test and Cox
regression analyses) were carried out using SPSS Statistics v23 (IBM, Ehningen, Germany). Optimal cut-off values of mRNA expression data to determine dichotomization for survival analyses (based on consecutive evaluation of all available cut-offs using univariate Cox regression) were calculated using the survMisc package for R (the R Foundation for Statistical Computing, Vienna, Austria).

mRNA expression of MYOF is up-regulated in ccRCC tissue.
In the TCGA cohort, significantly higher levels of MYOF mRNA expression were found in ccRCC tumor tissues compared to benign samples (p<0.001; data not shown). However, mRNA expression did not correlate with clinicopathological parameters (pT-stage, histological grade, pN-stage, pM-stage) nor survival (all p>0.05).
In order to validate these findings, we investigated MYOF mRNA expression in 82 ccRCC and 44 benign renal tissue samples. In our study cohort, MYOF mRNA expression was also significantly up-regulated in tumor tissue (each p<0.001, Figure 1A). The receiver operating characteristics analysis identified MYOF as a useful tissue biomarker to discriminate between ccRCC and benign renal tissue: MYOF mRNA levels discriminated ccRCC from benign tissue with a high specificity of 93.2% and a mid-range sensitivity of 67.1% (area under the curve=0.834). Next, we correlated MYOF expression levels with clinicopathological parameters ( Figure  1). Higher MYOF mRNA levels were detected in advanced tumors (p<0.001; Figure 1B). MYOF mRNA levels did not correlate with overall, cancer-specific or progression-free survival (all p>0.05).
Protein expression of MYOF correlates with tumor aggressiveness. We first studied MYOF protein expression by western blot in four ccRCC and corresponding benign renal tissue samples. MYOF protein levels were increased in ccRCC ( Figure 2). Interestingly, we detected an additional band with lower molecular weight than the expected. According to Uniprot database there are eight isoforms of MYOF, with molecular weight ranging from 46.69 to 234.70 kDa. This likely represents isoform 5 (molecular weight 179 kDa) (19).
We next investigated the protein expression level using immunohistochemistry in tumor samples from a TMA cohort of 172 patients. Immunohistochemical staining of MYOF in tumor and benign tissues is shown in Figure 3. We confirmed up-regulation of the MYOF protein in ccRCC tissues (p<0.001, Figure 4A). Interestingly, increased MYOF expression was positively correlated with tumor grade p<0.001), as well as with the presence of lymph node metastasis (p=0.004) and positive surgical margins (p=0.012; Figures 3 and 4B-D). High levels of MYOF (staining score ≥2) showed a trend towards being associated with reduced cancer-specific survival as evaluated by Kaplan-Meier estimates and log-rank test (log-rank p=0.083, Figure 4E), however, this was not significant in multivariate analysis. MYOF protein levels were not correlated with overall or progression-free survival.

siRNA-mediated knockdown suppresses migration and invasion.
In order to investigate the functional role of MYOF in tumor progression, we performed siRNAmediated knockdown of MYOF in the ACHN cell line. MYOF mRNA and protein expression was significantly down-regulated following siRNA interference (p≤0.001; Figure 5A and B). We observed a significant reduction of migration (p=0.001) and invasion (p=0.045) in ACHN cells compared to the control following knockdown of MYOF ( Figure 5C and D).
As a recent study suggested a decrease in VEGF-mediated cellular proliferation following MYOF silencing in metastatic ccRCC (20), we additionally investigated the influence of MYOF knockdown on the angiogenesis-associated genes TIE2, ANG2 and CAV1: Knockdown of MYOF led to an up-  regulation of these genes ( Figure 6) indicating a role of MYOF in angiogenesis.

Discussion
As transmembrane-anchored vesicle-fusion proteins, ferlins have been described to play an important role in carcinogenesis. In particular, up-regulation of MYOF was found in various tumor entities as previously stated (7)(8)(9)(10). The aim of this study was to investigate MYOF expression and functional properties in ccRCC.
We observed significant up-regulation of MYOF in ccRCC tissues at the mRNA as well as the protein level. These findings confirm overexpression of MYOF in ccRCC, also recently described by Song et al. (11). Furthermore, increased mRNA and protein expression were correlated to adverse clinicopathological parameters, i.e. mRNA level and advanced stage, protein level and histological grade, lymph node metastasis and positive surgical margins. In summary, the overexpression of MYOF correlates with aspects of tumor aggressiveness. Our findings are therefore concordant with the results obtained for other cancer entities such as lung, breast and pancreatic cancer.   Involvement of MYOF in cellular signaling pathways has been described, especially an effect on the function of the epidermal growth factor receptor (EGFR) and the VEGFR2. In human breast cancer cells, MYOF regulates EGFR activity as MYOF silencing blocked EGF-induced migration and epithelial-to-mesenchymal transition (8). A recent study showed a positive relationship between MYOF and EGFR expression in ccRCC, thus indicating that EGFR is involved in the oncogenic properties of MYOF (12). Furthermore, in endothelial cells, MYOF forms a complex with dynamin-2 and VEGFR2 that prevents VEGFR2 polyubiquitination and proteasomal degradation. In MYOF knock-out mice, loss of protein expression inhibited proliferation and migration (15,21). Another study examined the expression of VEGFR as well as MYOF in ccRCC; although an inverse correlation of MYOF and VEGFR in ccRCC tissues was observed, knockdown of MYOF led to down-regulation of VEGFR in metastatic ccRCC cell lines. The loss of MYOF additionally reduced cell confluence, leading the authors to conclude that MYOF affects cell proliferation by regulating VEGFR degradation (20). MYOF also participates in angiogenesis through its interaction with other proteins. In pancreatic ductal adenocarcinoma cells, MYOF knockdown attenuated VEGFA secretion due to an impairment of VEGFA exocytosis, thereby reducing blood vessel density (22). Furthermore, an influence of MYOF on the well-known angiogenic TIE2 receptor was also described. In endothelial cells, MYOF silencing led to reduction of TIE2 receptor expression. (23). Besides these proteins, overexpression of the angiogenesis-associated proteins ANG2 and CAV1 in RCC with influence on tumor progression is well known (24)(25)(26)(27). Since the influence of MYOF on these proteins in ccRCC has not yet been investigated, we analyzed the expression of TIE2, ANG2 and CAV1 following knockdown of MYOF. We showed that TIE2, ANG2 and CAV1 expression increased following knockdown of MYOF in ACHN cells. This was accompanied by suppression of migration and invasion indicating that MYOF is able to exert its oncogenic properties in ccRCC by regulating these genes.

Conclusion
Our study shows that MYOF expression is associated with aggressive ccRCC and affects migration and invasion of tumor cells. Thus, MYOF plays an important role in the carcinogenesis of ccRCC. These findings highlight MYOF both as a biomarker of tumor progression and as a promising target for future treatment of ccRCC.

Conflicts of Interest
All Authors declare that they have no competing interests. They confirm that the research was conducted in the absence of any commercial or financial relationships.