Application of integrative metabolomics to evaluate the quality and safety of maize

다국어 초록 (Multilingual Abstract)

Safety assessment of genetically modified (GM) maize, based on ‘substantial equivalence’, is crucial for consumer trust and commercialization. Quality assessment could also be conducted to ensure commercial competitiveness. The current safety assessment is limited to the compositional components proposed under the concept of substantial equivalence. Furthermore, the quality assessment is limited to known antioxidants. To address these limitations, untargeted metabolite analysis could be conducted alongside targeted metabolite analysis, providing a more comprehensive evaluation of safety and quality. Therefore, this study presents a more advanced workflow for assessing the safety and quality of GM maize using integrative metabolomics. The two GM maize lines (GM4 and GM6) were developed by hybridizing Hi ⅡA with B73 to enhance yield stability after Hi ⅡA was being transformed with herbicide resistance via Agrobacterium-mediated transformation. The two parental lines (Hi ⅡA and B73) originated in Iowa, the United States, and the four commercial cultivars (KPO, Kwangpyeongok; GDO, Godangok; IMC, Ilmichal; DO3, Danok3ho) originated in South Korea. All samples were grown simultaneously in Jeonju, South Korea. This study was conducted following a five-step workflow proposed for evaluating the safety and quality of GM maize. Step 1 involved acquiring metabolite data. Untargeted metabolite profiling using UPLC-qTOFMS detected 18,387 peaks in positive ion mode and 12,978 peaks in negative ion mode. A total of 132 metabolites were identified by comparison with the NIST20 Tandem Library. Targeted metabolite profiling using GC-TOFMS identified 47 metabolites. Step 2–4 involved safety assessment. In Step 2, principal component analysis revealed no substantial variation between the GM lines and their parental lines. Orthogonal partial least squares−discriminant analysis showed significant and clear distinctions between the GM lines with their parental lines, and the commercial cultivars according to geographic origin. In Step 3, using Venn diagrams, nine biomarkers— seven lipids and two phenolic compounds—satisfying all three statistical criteria (variable importance in projection, receiver operating characteristic curve, and log2 fold change) were selected. In Step 4, these biomarkers were validated through correlation analysis (r > 0.6208, p < 0.0012). The selected biomarkers clearly grouped the GM lines with their parental lines, and the commercial cultivars, suggesting their applicability for determining the geographical origin of GM maize. Step 5 involved quality assessment. Antioxidant assays indicated higher activity in the commercial cultivars than in the GM lines with their parental lines. Based on this, feature-based molecular networking revealed additional unknown metabolites, such as phenolamides and phenolic aldehydes, contributing to antioxidant activity. These results demonstrate that integrated untargeted and targeted metabolomics could serve as a comprehensive platform for the safety and quality assessment of GM maize.

목차 (Table of Contents)

• Abstract i
• Table of Contents iv
• List of Tables vi
• List of Figures vii
• Chapter 1. Introduction 1
• Chapter 2. Materials and Methods 5
• 2.1. Sample preparation 5
• 2.2. Extraction, analysis and data processing of untargeted metabolites 7
• 2.3. Extraction, analysis and data processing of targeted polar metabolites 11
• 2.4. Determination of antioxidant activity 14
• 2.5. Statistical analysis 15
• 2.6. Feature-based molecular networking 16
• Chapter 3. Results & Discussion 17
• 3.1. Overview of the overall experimental workflow 16
• 3.2. Step 1: Untargeted and targeted metabolites analysis 20
• 3.3. Step 2: Detection of unintended metabolite variations through multivariate analysis 31
• 3.4. Step 3: Biomarkers searching based on factors selected in Step 2 37
• 3.5. Step 4: Validation of biomarkers selected in Step 3 through correlation analysis 43
• 3.6. Step 5: Visualization of the antioxidant activity through FBMN 46
• Chapter 4. Conclusion 51
• Reference 54
• 국문 초록 62