Ultrasensitive Simultaneous Detection of Multiple Rare Modified Nucleosides as Promising Biomarkers in Low-Put Breast Cancer DNA Samples for Clinical Multi-Dimensional Diagnosis.

Abstract

Early cancer diagnosis is essential for successful treatment and prognosis, and modified nucleosides have attracted widespread attention as a promising group of cancer biomarkers. However, analyzing these modified nucleosides with an extremely low abundance is a great challenge, especially analyzing multiple modified nucleosides with a different abundance simultaneously. In this work, an ultrasensitive quantification method based on chemical labeling, coupled with LC-MS/MS analysis, was established for the simultaneous quantification of 5hmdC, 5fdC, 5hmdU and 5fdU. Additionally, the contents of 5mdC and canonical nucleosides could be obtained at the same time. Upon derivatization, the detection sensitivities of 5hmdC, 5fdC, 5hmdU and 5fdU were dramatically enhanced by several hundred times. The established method was further applied to the simultaneous detection of nine nucleosides with different abundances in about 2 μg genomic DNA of breast tissues from 20 breast cancer patients. The DNA consumption was less than other overall reported quantification methods, thereby providing an opportunity to monitor rare, modified nucleosides in precious samples and biology processes that could not be investigated before. The contents of 5hmdC, 5hmdU and 5fdU in tumor tissues and normal tissues adjacent to the tumor were significantly changed, indicating that these three modified nucleosides may play certain roles in the formation and development of tumors and be potential cancer biomarkers. While the detection rates of 5hmdC, 5hmdU and 5fdU alone as a biomarker for breast cancer samples were 95%, 75% and 85%, respectively, by detecting these three cancer biomarkers simultaneously, two of the three were 100% consistent with the overall trend. Therefore, simultaneous detection of multiple cancer biomarkers in clinical samples greatly improved the accuracy of cancer diagnosis, indicating that our method has great application potential in clinical multidimensional diagnosis.