Metabolic Profiling Methodology Study On Bio-samples And Metabolomics Research Of Colorectal Cancer Based On Mass Spectrometry Coupled With Chromatography

In this dissertation, we developed robust and high throughput metabonomic methods on different bio-samples, including serum and human colorectal cancer (CRC) tissue, using Ultra Performance Liquid Chromatography and Quadruple/Time-of-flight mass spectrometry (UPLC-QTOFMS). Furthermore, we made a urinary metabonomic study on a cohort of CRC (n=101) and healthy subjects (n=103), based on gas chromatography time-of-flight mass spectrometry (GC-TOFMS) and UPLC-QTOFMS.Firstly, a robust and high throughput method was developed on UPLC-QTOFMS for performing global metabolic profiling analysis on a large batch of human serum, for example, over400bio-samples in a batch. Using38reference standards, we compared serum preparation methods for protein precipitation by different organic solvents. Different conditions for global metabolic profiling analysis were performed and optimized on UPLC-QTOFMS. We also assayed the reproducibility, precision, and stability of the method of profiling. The results indicated that methanol/acetonitrile (1/9) could effectively and reproducibly precipitated human serum proteins, performing the best extracting effect on the assayed reference standards. At least four fold of pre-cold organic solvents coupled with vortex for2min, ultrasonic treatment for1min and stayed at-20℃for10min were the most optimal for the extracting effect and protein precipitation. The results, according to the reproducibility, precision and stability, showed that the method was satisfactory in global metabolic profiling analysis of human serum. Furthermore, reproducibility of the method verified by PCA analysis was consistent with the CV analysis in the results. The method was adapted to metabolomics study, especially the large batch metabolomics study of over400samples, with a high throughput of over100samples a day (13min for each sample run). Secondly, a robust and high throughput metabolic profiling method on human CRC tissue was developed on UPLC-QTOFMS for a large batch, for example, over400bio-samples in a batch. We used a two-step extraction method for the preparation of the CRC tissue. Using38reference standards, we compared different reconstituted solvents for protein precipitation and extration effeciency. Different conditions for global metabolic profiling analysis were performed and optimized on UPLC-QTOFMS. We also assayed the reproducibility, precision, and stability of the method of profiling. The results indicated that reconstituted solvent of water/methanol/acetonitrile (1/2/7) could effectively and reproducibly precipitated residued proteins in the extraction of CRC tissue, performing the best extracting effect on the assayed reference standards. The results, according to the reproducibility, precision and stability, showed that the method was satisfactory in global metabolic profiling analysis of human CRC tissue extration. Furthermore, reproducibility of the method was verified by CV analysis in the results. The method was adapted to metabolomics study, especially for a large batch research of over400samples, with a high throughput of over100samples a day (13min for each sample run).Thirdly, we made a metabonomic research on urine samples from CRC patients and healthy control people. A full spectrum of metabolic aberrations that are directly linked to CRC at early curable stages is critical for developing and deploying molecular diagnostic and therapeutic approaches that will significantly improve patient survival. We reported previously a urinary metabonomic profiling study on CRC subjects (n=60) and health controls (n=63), in which a panel of urinary metabolite markers was identified. Here, we made a second urinary metabonomic study on a larger cohort of CRC (n=101) and healthy subjects (n=103), using GC-TOFMS and UPLC-QTOFMS. Consistent with our first metabonomic study, we discriminated the CRC subjects from the healthy controls, including24CRC cases of early pathological stage of TNM-I. Moreover, we observed a number of dysregulated metabolic pathways similar to our previous findings, such as glycolysis, TCA cycle, urea cycle, pyrimidine metabolism, polyamine metabolism as well as gut microbial-host co-metabolism in CRC subjects. Our findings confirm distinct urinary metabolic footprints of CRC patients characterized by altered levels of metabolites derived from gut microbial-host co-metabolism. A panel of metabolite markers composed of7metabolites (citrate, hippurate, p-cresol,2-aminobutyrate, myristate, putrescine, and kynurenate) was able to discriminate CRC subjects from their healthy counterparts. A receiver operating characteristic curve (ROC) analysis of these markers resulted in an area under the receiver operating characteristic curve (AUC) of0.993and0.998for the training set and the testing set, respectively. These metabolite markers provide a novel molecular diagnostic approach for the early detection of CRC.These studies in this dissertation covered the metabonomics methodology on different bio-samples and the metabonomic application in a certain malignant disease of colorectal cancer. The results in these studies enriched the understanding of metabonomics, and powered the potential early diagnositic research on colorectal cancer.