Site-directed labelling of proteins for NMR and EPR studies

Abstract

Site-specific protein labelling presents an important tool for protein structural biology by spectroscopic techniques. This thesis focuses on the development of new spectroscopic labels and labelling strategies to improve the sensitivity, accuracy and scope of NMR and EPR spectroscopy experiments of proteins. Double electron–electron resonance (DEER) spectroscopy measures the distance between two paramagnetic metal centres introduced by site-specific attachment of suitable tag molecules. Good DEER tags possess rigid tethers to position their paramagnetic centres at a well-defined location relative to the protein and deliver narrow DEER distance distributions with high sensitivity, which can provide accurate information about protein flexibility. This thesis introduces new, cyclen-based Gd 3+ tags and small, Gd 3+ chelating tags designed to deliver narrow DEER distance distributions. Chapter 2 describes the development of two cyclen-based double-arm Gd 3+ tags designed for binding to the target protein at two and three points to obtain the narrowest Gd 3+ –Gd 3+ DEER distance distributions ever recorded with proteins. It also describes DEER distance measurements with the iminodiacetic acid tag attached to cysteine (Cys), where tags attached to two neighbouring Cys residues combine to chelate a single Gd 3+ ion. These results have been published in a journal article (Welegedara et al., Chem. Eur. J. 2017, 23, 11694−11702). Chapter 3 discusses two single-armed Gd 3+ tags, a cyclen-based Gd 3+ tag and a PyMTA tag that forms a heptadentate Gd 3+ binding motif. Both tags deliver the shortest possible tethers to cysteine residues and are shown to produce narrow DEER distance distributions in proteins. For applications in NMR spectroscopy, proteins can be labelled site-specifically with NMR probes such as trimethylsilyl (TMS) probes, which deliver readily detectable 1D 1 H-NMR signals. Introduction of paramagnetic tags and NMR probes by attachment to Cys residues requires mutations of native Cys residues to achieve site-selectivity, which is not possible with structurally and functionally important Cys residues. Chapter 4 demonstrates a solution to this limitation by introducing a selenocysteine (Sec) residue, which can be selectively reacted with probe molecules at slightly acidic pH without iiiaffecting naturally occuring Cys residues. To achieve this, a Sec residue was introduced as a photocaged unnatural amino acid (UAA), PSc, to bypass the otherwise unavoidable challenges associated with the natural Sec incorporation mechanism. UV illumination of PSc yielded Sec with no evidence for the formation of undesired dehydroalanine byproducts. Selective tagging of Sec residues with TMS tags was shown to deliver a useful tool for studies of ligand binding to proteins. These results have also been published in a journal article (Welegedara et al., Bioconjugate Chem. 2018, 19, 2257−2264). Site-selective incorporation of isotope-labelled PSc, photolysis and anaerobic deselenization opens an indirect route to labelling a single specific alanine residue in a protein with stable isotopes. Such samples would have important applications in heteronuclear NMR, as they allow the selective detection of the labelled alanine residue with maximal spectral resolution. As shown in Chapter 5, deselenization of selenoproteins into alanine is possible but requires extremely anaerobic conditions to eliminate serine as the main unwanted byproduct. A range of UAAs has been developed to serve as spectroscopic probes or facilitate the introduction of spectroscopic probes via biorthogonal reactions. The increasing demand for proteins with different UAAs and the significant cost of some of these UAAs has led to increasing popularity of cell-free protein synthesis (CFPS) systems, which use amino acids more sparingly than in vivo expression systems. Mutants of pyrrolysyl-tRNA synthetase (PylRS)/tRNA CUA pairs have been developed into a particularly versatile tool for the incorporation of many structurally different UAAs, but most produce disappointedly poor protein yields in vivo. Chapter 6 describes attempts to develop an in- house CFPS system with a PylRS/tRNA CUA pair. In addition, the polyspecific G2 synthetase has been reported to facilitate the incorporation of sterically demanding UAAs and Chapter 7 of this thesis describes attempts to develop a CFPS system with the G2 synthetase.