Comparison of the Plasma Metabolome Profiles Between the Internal Thoracic Artery and Ascending Aorta in Patients Undergoing Coronary Artery Bypass Graft Surgery Using Gas Chromatography Time-of-Flight Mass Spectrometry

Abstract

Background The left internal thoracic artery (LITA) has been used as the first conduit of choice in coronary artery bypass grafting (CABG) because of excellent long-term patency and outcomes. However, no studies have examined substances other than nitric oxide that could be beneficial for the bypass conduit, native coronary artery or ischemic myocardium. This study was conducted to evaluate differences in metabolic profiles between the LITA and ascending aorta using gas chromatography-time of flight-mass spectrometry (GC-TOF-MS). Methods Twenty patients who underwent CABG using the LITA were prospectively enrolled. Plasma samples were collected simultaneously from the LITA and ascending aorta. GC-TOF-MS based untargeted metabolomic analyses were performed and a 2-step volcano plot analysis was used to identify distinguishable markers from two plasma metabolome profiles. Semi-quantitative and quantitative analyses were performed using GC-TOF-MS and enzyme-linked immunosorbent assay, respectively, after selecting target metabolites based on the metabolite set enrichment analysis. Results Initial volcano plot analysis demonstrated 5 possible markers among 851 peaks detected. The final analysis demonstrated that the L-cysteine peak was significantly higher in the LITA than in the ascending aorta (fold change = 1.86). The concentrations of intermediate metabolites such as L-cysteine, L-methionine and L-cystine in the ‘cysteine and methionine metabolism pathway' were significantly higher in the LITA than in the ascending aorta (2.0-, 1.4- and 1.2-fold, respectively). Quantitative analysis showed that the concentration of hydrogen sulfide (H 2 S) was significantly higher in the LITA. Conclusion The plasma metabolome profiles of the LITA and ascending aorta were different, particularly higher plasma concentrations of L-cysteine and H 2 S in the LITA.

How searching against multiple libraries can lead to biased results in GC/MS‐based metabolomics

“…[15][16][17][18][19][20][21][22][23][24] The NIST and FiehnLib libraries are often used simultaneously. [15][16][17][18][19][20][21][22][23][24] It may seem that the difference in the scan range of library spectra should not influence the results of the library search. However, we assumed that such a difference can lead to biased results under certain circumstances.…”

“…In untargeted metabolomics, it is a common practice to search against multiple libraries. [15][16][17][18][19][20][21][22][23][24] The NIST and FiehnLib libraries are often used simultaneously. [15][16][17][18][19][20][21][22][23][24] It may seem that the difference in the scan range of library spectra should not influence the results of the library search.…”

“…Despite the apparent complexity, the first three conditions are satisfied when the NIST and FiehnLib libraries are used simultaneously and the scan range of an instrument starts from m/z 50. One can find such metabolomic studies in the literature 17–24 . As presented later, the last condition takes place when AMDIS or ChromaTOF software is used to perform the library search.…”

“…In contrast, according to our estimation (performed during the preparation of data sets for this work), about 90% of TMS derivatives from the NIST20 library ( mainlib ) contain at least one peak with m/z ≤ 41. In untargeted metabolomics, it is a common practice to search against multiple libraries 15–24 . The NIST and FiehnLib libraries are often used simultaneously 15–24 .…”

“…One can find such metabolomic studies in the literature. [17][18][19][20][21][22][23][24] As presented later, the last condition takes place when AMDIS or ChromaTOF software is used to perform the library search.…”