Proteomic analysis to identify cytokeratin 18 as a novel biomarker of nasopharyngeal carcinoma

In this study, we applied laser capture microdissection and a proteomic approach to identify novel nasopharyngeal carcinoma (NPC) biomarkers.

Proteins from pooled microdissected NPC and normal nasopharyngeal epithelial tissues (NNET) were separated by two-dimensional gel electrophoresis, and differential proteins were identified by mass spectrometry. Expression of the differential protein cytokeratin 18 in the above two tissues as well as 4 NPC cell lines was determined by Western blotting. Immunohistochemistry was also performed to detect the expression of cytokeratin 18 in 62 cases of primary NPC, 28 cases of NNET, and 20 cases of cervical lymph node metastases, and the correlation of its expression level with clinicopathologic features and clinical outcomes were evaluated. siRNA and in vitro cell invasion assay were used to check the correlation between the expression of cytokeratin 18 and invasive ability of NPC.

Thirty-six differential proteins between the NPC and NNET were identified. The expression level of cytokeratin 18 in the two types of tissues was confirmed by Western blotting and related to differentiation degree and metastatic potential of the NPC cell lines. Significant cytokeratin 18 down-regulation was observed in NPC versus NNET (P = 0.000), whereas significant cytokeratin 18 up-regulation was observed in lymph node metastasis versus primary NPC (P = 0.001). In addition, cytokeratin 18 down-regulation was significantly correlated with poor histological differentiation (P = 0.000), whereas cytokeratin 18 up-regulation was significantly correlated with advanced clinical stage (P = 0.019), recurrence (P = 0.000), and regional lymph node metastasis (P = 0.001), and distant metastasis (P = 0.000). And down-regulated cytokeratin 18 expression by siRNA significantly decreased in vitro invasive ability of 5-8F cells. Furthermore, survival curves showed that patients with cytokeratin 18 up-regulation had a poor prognosis (P = 0.000). Univariate analysis (Cox’s proportional hazards model) showed that WHO histologic type (P = 0.025), lymph node metastasis (P = 0.007), distant metastasis (P = 0.005), recurrence (P = 0.000), and cytokeratin 18 (P = 0.000) were significantly associated with the prognosis of NPC. Multivariate analysis confirmed that lymph node metastasis (P = 0.012), distant metastasis (P = 0.009), recurrence (P = 0.006), and cytokeratin 18 (P = 0.001) were independent prognostic indicators.

The data suggest that cytokeratin 18 is a potential biomarker for the differentiation and prognosis of NPC, and its dysregulation might play an important role in the pathogenesis of NPC.