Protein-Ligand Interactions by NMR and EPR Spectroscopy

Abstract

Pulse electron paramagnetic resonance (EPR) distance measurements using double electron-electron resonance (DEER) experiments have been established as a powerful tool in structural biology. DEER experiments have the ability to measure the distance between two paramagnetic centres in biological macromolecules in the range of about 2 to 8 nm. The paramagnetic centres are usually introduced into proteins by site-directed spin labelling (SDSL) of cysteine residues. This thesis is based on the use of new lanthanide binding tags (LBTs) for paramagnetic nuclear magnetic resonance (NMR) spectroscopy (reported in papers 2 and 5), DEER distance measurements (reported in papers 1 and 3) and time-resolved luminescence resonance energy transfer (LRET) experiments (reported in paper 4). In particular, use of two complementary techniques, DEER experiments and paramagnetic NMR spectroscopy, was investigated for the study of conformational changes of proteins as a result of protein-ligand interactions. Two proteins were studied, the E. coli aspartate/glutamate binding protein (DEBP) and human calmodulin (CaM). Both proteins have different ligand binding characteristics: DEBP binds to small organic molecules, while CaM binds to specific peptide sequences. DEBP is a periplasmic binding protein responsible for the transport of aspartic acid and glutamic acid across the cell membrane and widely used in the design of biosensors of glutamate. The protein is composed of two domains, which bind one amino acid molecule at the domain interface. As DEBP contains a disulfide bond, an alternative cysteine-independent approach for site-specific protein tagging was used, which involved the use of genetically encoded unnatural amino acids that were site-specifically incorporated into proteins using orthogonal amber-suppressor tRNA/aminoacyl-tRNA synthetase systems. p-azido-L-phenylalanine (AzF) residues were incorporated into DEBP at different positions and paramagnetic lanthanide tags were attached to AzF via Cu(I)-catalyzed click chemistry (papers 1 and 2). Multiple Gd3+-Gd3+ distances measured by DEER experiments were used to define the metal positions, subsequently allowing deltachi-tensor determinations from sparse sets of pseudocontact shifts (PCSs). Both the DEER data and PCSs were in agreement with the closed conformation observed in the crystal structure of the homologue from S. flexneri. On the other hand, the PCSs indicated that the transition to the substrate-free protein involves a movement of the two domains as rigid entities relative to each other. CaM is a two-domain protein that acts as an intermediate messenger protein and intracellular calcium sensor, which responds to changes in Ca2+ concentrations by large conformational changes that enable binding to a range of different proteins involved in signalling pathways. The conformational changes of CaM upon binding of the myristoylated alanine-rich C-kinase substrate (MARCKS) peptide were studied using DEER experiments and paramagnetic NMR. MARCKS was chosen due to its unique binding mode compared to other CaM-target peptide complexes. The DEER results indicated that the binding of MARCKS peptide to CaM does not lock CaM in a single conformation. Deviations between the crystal and solution structure of the complex were also evident in the measured PCS data, highlighting the conformational flexibility of CaM that allows CaM to bind to diverse target proteins.