Detailed NMR fingerprinting of proteins

The current technology is increasingly employing so-called “fingerprints”. The “fingerprint” means a characteristic feature of an individual, it can be either spiral marks on a surface of a finger or preferences of an internet browser (font, resolution and other). The method is applied extensively in many different sciences, including chemistry, biology, physics, and social sciences. A particular “fingerprint” can vary in its specificity, depending on what kind of information it carries. Nuclear magnetic resonance spectroscopy (NMR) is one of the techniques that employ “fingerprints”.

NMR is an important research tool in chemical and biological sciences. Among many instrumental techniques, NMR is the only one that enables for investigation of proteins in solution with atomic resolution. Moreover, the technique is highly repeatable, reproducible and non-invasive. The idea of a “fingerprint” can be utilized for protein stability studies, ligand binding studies or screening studies. In the above-mentioned examples, the key point is to determine the peak resonant frequencies. This is usually done with HSQC experiment, which yields a two-dimensional (2D) NMR spectrum correlating ¹H and ¹⁵N nuclei connected through one chemical bond. The obtained spectrum can be easily used to confirm whether protein underwent any changes, but the information provided is non-specific for amino acid residue type, i.e. the peak cannot be assigned to the amino acid type. An alternative experiment is HSQC–TOCSY, which besides of ¹H–¹⁵N pairs correlates whole ¹H spin systems. Although this experiment serves a kind of amino acid-specific information, it does not allow for assigning a signal with a high probability.

The goal of the project is to develop a method for quick and amino-acid specific NMR measurement of proteins. The method will utilize a modified HSQC-TOCSY experiment, in which besides of ¹H and ¹⁵N dimensions, a new TOCSY–transfer dimension will be introduced. The additional dimension may serve as an extra unique variable that will increase the confidence of guess on amino acid residue type. The proposed experiment alone would not be sufficient to provide assignment guess. A dedicated library of amino acid TOCSY-transfer profiles will be created using quantum mechanical NMR simulations taking input data from protein databases. Along with available chemical shifts in databases, simulated TOCSY-profiles will be utilized for the creation of an algorithm that will match experimental and library data to give a reasonable guess on amino acid residue type for the observed peak.

The method proposed in the project has a chance to become an important tool for research on protein using NMR spectroscopy, in particular in a field of protein stability, ligand binding, and high-throughput studies. The method would deliver extra information on amino acid residue type in a time comparable to routinely employed ¹⁵N HSQC experiment. Thanks to that, the method would serve a new and more detailed fingerprint for proteins.