基于磁共振的代谢组成像新方法及其在活体组织中的应用

With the characteristics of globality and systematicness, the nuclear magnetic resonance (NMR)-based metabolomics approach is widely used in the early diagnosis of diseases and pathological study. Metabolomics study was mainly targeted at the biological fluids and tissues in vitro while the in vivo detection of metabolic profilings of tissues and their dynamic variations with the course of diseases is still rare. On the basis of our previous favourable work in NMR metabolomics, this project plans to extend the static and in vitro tissue metabolomics to the dynamic and in vivo metabolomic imaging through in vivo magnetic resonance spectroscopy (MRS) technology. A new metabolomic imaging approach suitable for living tissues will be established. The pulse sequence is designed for acquiring high resolution MRS in vivo, and with the combination of high resolution magic angle spinning (HR-MAS) spectroscopy and its corresponding feature space, the metabolic patterns of living tissues will be tracked. This approach is applied to non-alcoholic fatty liver disease (NAFLD) animal model. The metabolic phenotypes of target organs will be real-time monitored during the pathological process and the characteristic metabolites and related metabolic pathways in tissues will be identified. Furthermore, the variations of metabolism correlated with the development of NAFLD could be elucidated and metabolic sub-network model will be constructed. This NAFLD experiment will provide a practical foundation for clinical application of NMR-based metabolomics imaging.

基于核磁共振(NMR)的代谢组学方法以其整体性和系统性等特点广泛应用于疾病的早期诊断和病理研究。代谢组学的研究对象主要是生物体液和离体组织，对活体组织代谢成分的检测及其随病程发展的动态变化规律尚缺乏研究。本项目拟在前期较好的NMR代谢组学工作基础上，借助活体磁共振定域谱(MRS)，将静态的、离体的组织代谢组学研究扩展到动态的、活体的代谢组成像研究。我们将建立一套适用于活体组织的代谢组成像新方法，设计在活体条件下获得高分辨MRS谱的脉冲序列，结合离体组织魔角旋转谱(HR-MAS)并建立其对应的特征空间，跟踪监测活体组织的代谢模式。我们将这种方法应用于非酒精性脂肪肝(NAFLD)动物模型，实时监测病理进程中靶标器官的代谢表型，辨识组织中的特征代谢物及代谢途径变化，阐明与NAFLD发生发展密切相关的代谢模式变化规律，构建代谢子网络模型，为基于磁共振的代谢组成像新方法的临床应用提供实践基础。

代谢组学方法以其系统性和整体性为突出特点，已被广泛应用于生命科学研究的诸多领域。常用的代谢组学数据分析方法在处理高维数据降维、数据尺度缩放、核磁共振波谱的谱峰对齐、抑制干扰因素影响以及多源数据融合等方面还存在着许多值得探索的问题。本课题通过完善和发展基于核磁共振的代谢组学谱图预处理方法，包括提出了一种基于高斯平滑的谱峰对齐算法，开发一种模型自适应的数据归一化方法以及基于K-L散度的核磁共振波谱数据尺度缩放方法，设计一种基于聚类的代谢组学数据去干扰方法，探索多源数据融合算法等，提高数据模型对特征变量的筛选和识别能力，为后续的代谢通路分析提供支持。我们将这些新方法应用于高脂饲养的肥胖大鼠、肝豆状核变性、家族遗传性甲亢等常见代谢性疾病的研究，相关结果陆续发表在Applied Biochem Biotechnol，PLoS One，Evid-based Complement Alternat Med等刊物上。这一系列成果将有助于拓展代谢组学方法中的多源数据融合和代谢组学数据建模方法，挖掘与疾病相关的通路信息，为个性化治疗提供可能的治疗新靶点。