One metabolite to bind them – using ligand-detected NMR to fish out protein-metabolite interactions

In this work researchers from the institutes of Molecular Systems Biology and Molecular Biology & Biophysics at ETH Zurich teamed up to develop a NMR-based approach which directly detects protein-metabolite binding events within complex metabolite mixtures in vitro. As a proof-of-concept, they selected four well-characterized bacterial and mammalian and showed that this approach recovers known protein binders among mixtures of up to >30 metabolites, and also identifies novel, functionally relevant, interactions. This work, published in Biochemistry, provides a new arrow in the quiver of researchers who seek to systematically map out contact points between metabolism and the proteome.

Abstract

Protein-metabolite interactions play a vital role in the regulation of numerous cellular processes. Consequently, identifying such interactions is a key prerequisite for understanding cellular regulation. However, the non-covalent nature of the binding between proteins and metabolites has so far hampered the development of methods to systematically map protein-metabolite interactions. The few available, largely mass-spectrometry based, approaches are restricted to specific metabolite classes, such as lipids. In this study, we address this issue and show the potential of ligand-detected nuclear magnetic resonance (NMR) spectroscopy, which is routinely used in drug development, to systematically identify protein-metabolite interactions. As a proof-of-concept, we selected four well-characterized bacterial and mammalian proteins (AroG, Eno, PfkA, BSA) and identified metabolite binders in complex mixes of up to 33 metabolites. Ligand-detected NMR captured all of the reported protein-metabolite interactions, spanning full range of physiologically relevant K_ds (low-μM to low-mM). We also detected a number of novel interactions, such as promiscuous binding of the negatively charged metabolites citrate, AMP, and ATP, as well as binding of aromatic amino acids to AroG protein. Using in vitro enzyme activity assays, we assessed the functional relevance of these novel interactions in the case of AroG and show that L-tryptophan, L-tyrosine and L-histidine act as novel inhibitors of AroG activity. Thus, we conclude that ligand-detected NMR is suitable for the systematic identification of functionally relevant protein-metabolite interactions.

Reference

Yaroslav V. Nikolaev, Karl Kochanowski, Hannes Link, Uwe Sauer and Frederic H.-T. Allain
external pageSystematic identification of protein-metabolite interactions in complex metabolite mixtures by ligand-detected NMR spectroscopycall_made. (Biochemistry, 2016 April 11, DOI: 10.1021/acs.biochem.5b01291)