Understanding Parahydrogen Hyperpolarized Urine Spectra: The Case of Adenosine Derivatives
Description
Abstract from publication: Parahydrogen hyperpolarization has emerged as a promising tool for sensitivity-enhanced NMR metabolomics. It allows for the resolution and quantification of NMR signals of certain classes of low-abundance metabolites that would otherwise be undetectable. Applications have been implemented in pharmacokinetics and doping drug detection, demonstrating the versatility of the technique. Yet, in order for the method to be adopted by the analytical community, certain limitations have to be understood and overcameovercome. One such question is NMR signal as-signment. At present, the only reliable way to establish the identity of an analyte that gives rise to certain parahydrogen hyperpolarized NMR signals is internal standard addition, which can be laborious. Herein we show that analogously to regular NMR metabolomics, generating libraries of hyperpolarized analyte signals is a viable way to address this limitation. We present hyperpolarized spectral data of adenosines and give an early example of identifying them from a urine sample with the small library. Doing so, we verify the detectability of a class of diagnos-tically valuable metabolites: adenosine and its derivatives, some of which are cancer biomarkers, and some are central to cellular energy management (e.g., ATP). Data from publication: Embedded data files contain raw NMR data acquired during preparation of this publication and that is presented in the publication. Files are named as follows: Fig2/21122303 - 2D dataset for adenosine 1a Fig2/21122323 - 2D dataset for 1-methyladenosine 1b Fig2/21122313 - 2D dataset for 6N-methyladenosine 1c Fig2/21122123 - 2D dataset for 2'-O-methyladenosine 1d Fig2/21122333 - 2D dataset for AMP 1e Fig2/21122204 - 2D dataset for ADP 1f Fig2/21122213 - 2D dataset for ATP 1g --- Fig3/19021921 - 2D dataset of a urine SPE extract. Can be compared to datasets from Fig2. Some spectra may display signals from impurities at intensity levels below what is displayed in the article.
Files
Steps to reproduce
Sample preparation is described in manuscript main text. Processing instructions: Embedded datasets can be processed with Bruker Topspin or Mestrenova processing softwares (and presumably other common packages). Mestrenova was used for the preparation of the figures in the publication for easier production of publication grade figures. However, convolution filtering capabilities of Mestrenova are inferior to Topspin. If convolution filtered spectra are desired in Mestrenova, we recommend applying convolution filtering to raw data in Topspin before processing in Mestrenova. The applied pulse sequence does not make use of the Bruker 'mc' macro to handle phase and evolution delay incrementation in the pulse sequence code. Consequently, Fourier transform parameters have to be set explicitly in Processing Parameters tab of Topspin or Fourier Transform window in Mestrenova. Relevant processing parameters are embedded into the datasets, but are also given here: f1 dimension: SI 2048 (or 4096 or urine and 1b) WDW QSINE SSB 2 PHC0 90 PH_mod pk BC_mod single FT_mod fsr (real FT in f1) MC2 TPPI f2 dimension: SI 16k WDW QSINE SSB 2 PH_mod pk FT_mod fqc For convolution filtering, set: BC_mod qfil COROFFS 186 BCFW 0.03