Research Project
10276342
PROJECT ABSTRACT: The ability to measure and quantify the composition and abundance of various metabolites in biological samples, also referred to as metabolomics, provides a unique window into the complex biological processes at different scales. So far, the field of metabolomics has mainly been driven by technologies based on mass spectrometry (MS) and nuclear magnetic resonance (NMR) spectroscopy. These technologies, although powerful, only measure metabolite profiles in homogenized biological extracts, e.g., biofluids or dissected tissues, thus losing the spatial information of the underlying metabolic processes. As spatial heterogeneity is a hallmark of metabolism, especially in complex biological systems such as animals and humans, obtaining spatially resolved metabolomics has been a dream of many biomedical scientists and engineers. In recent years, MS imaging (MSI) has emerged as a tool of choice for imaging metabolomics, which allows for the generation of spatially localized metabolite profiles from tissue sections. One major limitation of MSI is that it requires post-mortem or invasive tissue sampling, thus unable to probe metabolism at the most physiologically relevant states. This has limited its translation to human studies. MR spectroscopic imaging (MRSI) is another alternative for imaging metabolomics. It combines the powers of MRI and NMR spectroscopy to produce spatially resolved tissue metabolite profiles, noninvasively. However, MRSI is highly limited in its poor spatial resolutions. Furthermore, most MRSI studies only target a single nucleus (e.g., 1H), thus limited in the number of molecular species measured. The overall goal of the proposed research is to develop a research program that will pave a path towards in vivo imaging of tissue metabolomics. Specifically, we aim to develop an unprecedented high-resolution multinuclear MRSI technology that can simultaneously map a large number of metabolites in vivo, synergizing advancements in ultrahigh- field MRI instrumentation, fast data acquisition, and machine learning driven computational imaging techniques. We also propose a novel multimodal MRSI and MSI imaging framework for validating our multinuclear MRSI technology and integrating two complementary biochemical imaging modalities for tissue metabolic profiling. Novel computational approaches will be developed to analyze the high- dimensional metabolomic data. Success of the proposed research will establish a new paradigm for generating and analyzing imaging metabolomics data. This paradigm will transform metabolomics into a powerful noninvasive and tissue specific technology (from an invasive and nonspatial-specific one) for studying metabolism in living animals and humans. These advances will enable new means to unravel the metabolic basis of normal physiological functions and different diseases, inspiring developments of new biomarkers, novel treatments, disease prognosis and management strategies.
