Ribonucleotide Reductase Large Subunit M1’s Role And Mechanistic Study In The Prognosis Prediction And Treatment Of Gastric Cancer

Gastric adenocarcinoma (gastric cancer. GC) is the second leading cause of cancer-associated mortality worldwide, particularly in our nation, where approximately half million cases are diagnosed annually. So far, few of the biomarkers are widely used clinically for GC. Therefore, biomarkers predicting survival and improving outcomes are urgently needed in GC patients. Ribonucleotide reductase (RNR) is considered a therapeutic target for cancer treatment. RNR is a time-limited enzyme that catalyzes the conversion of ribonucleoside diphosphates to deoxyribonucleoside diphosphates through de novo metabolism of endogenous nucleotides, plays important role in the cancer occurrence and metastasis. RRMl is the only regulatory subunit of RNR, which has been recognized as drug resistance marker in the gemcitabine treatment of advanced stage lung cancer. However, the role of RRM1in gastric cancer has never been explored yet, thus we aimed to investigate the prognostic and therapeutic value of RRM1in GC patients. First we enrolled a total of assessable389GC patients with clinicopathological and survival information were enrolled from City of Hope (COH, n=67) and Zhejiang University (ZJU, n=322). RRM1protein expression was determined by immunohistochemistry on FFPE tissue samples. Kaplan-Meier and Cox analyses were used to measure survival. We found that RRM1expression was significantly associated with lymph node involvement, tumor size, Ki67expression, histological subtype and histological grade in the GC tissue samples (p<0.05). Kaplan-Meier analysis illustrated that high RRM1expression predicted poor survival in GC patients in the COH and ZJU cohorts (log-rank p<0.01). In multivariate Cox analysis, the hazard ratios of RRM1for overall survival were2.55(95%CI1.27-5.15) and1.51(95%CI1.07-2.13) in the COH and ZJU sets, respectively. In particular, RRM1specifically predicted the outcome of advanced GCs with poor differentiation and high proliferative ability. Moreover, we explored the mechanism of RRM1in gastric cancer cell lines. Ras/Raf activity and invasion assays were used to evaluate the role of RRM1in GC cell lines. In vitro experiments demonstrated RRM1activated Ras/Raf/MAPK signal transduction and promoted GC cell proliferation. Meanwhile, Furthermore, inhibition of RRM1by siRNA significantly reduced the dNTP pool, Ras/Raf and MMP-9activities and the levels of p-MEK, p-ERK and NF-κB, resulting in growth retardation and reduced invasion in AGS and NCI-N87cells. Furthermore, we evaluated the role of RRM1in the chemotherapy of stage II GC patients. We found that patients with high RRM1expression benefit more from5-Fu treatment, so are the case in the5-Fu related metabolic genes TS (Thymidine synthatase) and dUTPase (dUTP Pyrophosphatase). Therefore, we formulated them together in a computer model, which is further validated and optimized using Naive Bayes classifier. The final model has been proved effective in the prediction of benefit from5-Fu treatment in the curable GC patients.