To identify molecular differences between primary and metastatic OSCC, DNA copy number changes were correlated with gene expression changes using DNA and RNA arrays (see figure). A signature of 95 genes was identified in tumor cells from lymph node metastases (n=20) compared to non-metastatic primary tumors (n=17). These genes were associated with survival and prognosis of patients by looking at an independent set of 133 OSCC patients, where 65 patients were still alive at the end of the follow up time (mean: 5.4 years, range 4.0-7.3 years). Using the gene signature and hierarchical clustering based on the gene expression variances determined by principal component analysis, patients were divided into two clusters that correlated with the stage of disease and prognosis. The 5-year overall survival for patients in cluster 1, who had mainly primary tumors, was higher than patients in cluster 2, who were more likely to have nodal metastasis (69.1±6.98% versus 41.7±5.45%, P = 0.000858). Using multivariate analysis adjusting for sex, stage, age, and human papillomavirus (HPV) status, hazard ratios (HR) for cluster 2 patients were 3.45 (95% confidence interval: 1.84-6.50) for overall survival and 4.75 (95% confidence interval: 2.03-11.11) for OSCC-specific mortality.
Next, the authors examined if the genes unique to metastases also affected tumor growth. They established three OSCC cell lines from metastatic tumors, and two derived from early-stage primary tumors. The metastasis derived cell lines displayed a more aggressive phenotype in vitro and in vivo. Finally, 26 of the 95 transcripts with amplification and overexpression were silenced by RNA interference in the five OSCC cell lines. In at least one of the OSCC cell lines, 18 of the 26 genes (69%) reduced cell viability. On average, the early-stage OSCC cell lines were more sensitive to gene silencing. However, silencing expression of the gene GAP SH3 Binding Protein 1 (G3BP1) decreased cell viability and increased apoptosis in late-stage OSCC. Upon further characterization, the induction of cell death by G3BP1 silencing preferentially occurred in OSCC cell lines that also had a mutation in the tumor suppressor p53. The precise mechanism for why knockdown of G3BP1 expression induces cell death remains to be elucidated. Nevertheless, G3BP1 was identified as a novel therapeutic target previously not described for head and neck cancer.
According to Dr. Méndez, this study provides a roadmap showing which genes may be essential for the survival advantage inherent in metastatic tumor cells, and this can be exploited to selectively target these cells. Efforts are underway to develop the identified gene signature as a prognostic biomarker for OSCC, as well as to understand how these genes mediate metastasis and proliferation. Based of the current study, Dr. Méndez was awarded a $720,000 four-year Research Scholar Grant from the American Cancer Society to further study potential therapeutic targets in OSCC.
Xu C., Wang P., Liu Y., Zhang Y., Fan W., Upton M.P., Lohavanichbutr P., Houck J.R., Doody D.R., Futran N.D., Zhao L.P., Schwartz S.M., Chen C., and Méndez E. 2013. Integrative genomics in combination with RNA interference identifies prognostic and functionally relevant gene targets for oral squamous cell carcinoma. PLoS Genetics 9: e1003169.