Coupling of Cell Surface Biotinylation and SILAC-Based Quantitative Proteomics Identified Myoferlin as a Potential Therapeutic Target for Nasopharyngeal Carcinoma Metastasis

Abstract

Distant metastasis is a major cause of treatment failure in nasopharyngeal carcinoma (NPC) patients. Cell surface proteins represent attractive targets for cancer diagnosis or therapy. However, the cell surface proteins associated with NPC metastasis are poorly understood. To identify potential therapeutic targets for NPC metastasis, we isolated cell surface proteins from two isogenic NPC cell lines, 6-10B (low metastatic) and 5-8F (highly metastatic), through cell surface biotinylation. Stable isotope labeling by amino acids in cell culture (SILAC) based proteomics was applied to comprehensively characterize the cell surface proteins related with the metastatic phenotype. We identified 294 differentially expressed cell surface proteins, including the most upregulated protein myoferlin (MYOF), two receptor tyrosine kinases(RTKs) epidermal growth factor receptor (EGFR) and ephrin type-A receptor 2 (EPHA2) and several integrin family molecules. These differentially expressed proteins are enriched in multiple biological pathways such as the FAK-PI3K-mTOR pathway, focal adhesions, and integrin-mediated cell adhesion. The knockdown of MYOF effectively suppresses the proliferation, migration and invasion of NPC cells. Immunohistochemistry analysis also showed that MYOF is associated with NPC metastasis. We experimentally confirmed, for the first time, that MYOF can interact with EGFR and EPHA2. Moreover, MYOF knockdown could influence not only EGFR activity and its downstream epithelial-mesenchymal transition (EMT), but also EPHA2 ligand-independent activity. These findings suggest that MYOF might be an attractive potential therapeutic target that has double effects of simultaneously influencing EGFR and EPHA2 signaling pathway. In conclusion, this is the first study to profile the cell surface proteins associated with NPC metastasis and provide valuable resource for future researches.

Keywords: SILAC; biotinylation; cell surface proteins; metastasis; nasopharyngeal carcinoma.

FIGURE 1

Experimental workflow for cell surface labeling and preparation for mass spectrometry. (A) Workflow of SILAC mass spectrometry-based analysis of the cell surface proteome, where 6-10b and 5-8F cells were labeled with “light” and “heavy” amino acids, respectively. (B) Scatter plot of protein ratio (5-8F vs. 6-10B) between the biological and technical replicates of the cell surface proteome. The Pearson correlation coefficient is presented in blue at the top of each plot. (C) Venn diagram showing the number of identified proteins from three biological replicate experiments. (D) Identified proteins grouped by their annotated subcellular localization (UniProtKB). Surface-exposed proteins represented by their annotated detailed categories are shown in the Venn diagram.

FIGURE 2

Identification and evaluation of significantly altered proteins between the 6-10B and 5-8F cell lines. (A) Volcano plot indicating significantly altered surface-exposed proteins identified in the datasets. Log-transformed P-values (t-test) against log-transformed fold change in abundance between 6-10B and 5-8F cells. (B) Western blot analysis evaluating the expression of proteins that reside in the plasma membrane. The density of the bands was analyzed by using NIH ImageJ software and normalized by the arbitrary units of b-actin. Data are the means ± SDs of 3 experiments. **p < 0.05 and **p < 0.01. (C) Representative immunohistochemistry (IHC) results showing increased detection of the indicated proteins in NPC with metastasis compared with NPC without metastasis.

FIGURE 3

Signaling pathway enrichment analysis and interaction network analysis of DEPs (A) Enriched Wikipathway for differentially expressed surface-exposed proteins in 5-8F cell lines vs. 6-10B cell lines. (B) Overlapping of MYOF-interacting proteins and differentially expressed signaling receptors in 5-8F vs. 6-10B cells. MYOF-interacting proteins were retrieved using Harmonizome. (C) Partial interaction network of differentially expressed proteins focused on the interaction of MYOF, EGFR and EPHA2. (D) The interactions of MYOF with EGFR and EPHA2 were confirmed by Co-IP assay. Co-IP using anti-MYOF and negative control antibodies were performed in 5-8F cells. Western blots for all three proteins were performed. The density of the bands was analyzed by using NIH ImageJ software and normalized by the arbitrary units of the band of the input of EphA2. Data are the means ± SDs of 3 experiments. **p < 0.05 and **p < 0.01.

FIGURE 4

MYOF silencing reduces the malignant phenotype of 5-8F cells. (A) Cellular proliferation of control and MYOF knockdown 5-8F cells when passaged for the indicated days. Independent experiments performed in triplicate (**p < 0.01 and ***p < 0.001, respectively, Students t-test). (B) Wound-healing assay performed with Vector- or shMYOF-transfected 5-8F cells. Representative images acquired at the indicated time points are shown. The unhealed area was measured in three independent experiments (**p-value < 0.01). (C) Control or MYOF knockdown 5-8F cells were subjected to Transwell migration assays in three independent experiments. Migratory cells were counted under a microscope (**p < 0.01). (D) Control or MYOF knockdown 5-8F cells were subjected to Transwell invasion assays (with Matrigel). Invasive cells were counted under a microscope (**p < 0.01).

FIGURE 5

Myoferlin knockdown influences membrane receptor activity and EMT in NPC cells. (A) Western blot analysis of time-dependent expression and phosphorylation level of EGFR and EPHA2 following EGF stimulation and myoferlin knockdown. (B) MYOF knockdown inhibited EGF-induced VIM expression. The density of the bands was analyzed by using NIH ImageJ software and normalized by the arbitrary units of b-actin. Data are the means ± SDs of 3 experiments. *p < 0.05 and **p < 0.01.