1. Introduction 168

1.1 Genomic annotation of metalloproteins 168

1.2 Why NMR structures? 168

1.3 Why paramagnetic metalloproteins? 169

2. General theory 170

2.1 Nuclear and electron spins 170

2.2 Hyperfine coupling 171

2.3 The effect of the hyperfine coupling on the NMR shift: the hyperfine shift 173

2.4 The effect of the hyperfine coupling on nuclear relaxation 174

2.5 Interplay between electron spin properties and features of the NMR spectra 178

3. Paramagnetism-based structural restraints 180

3.1 Contact shifts and relaxation rates as restraints 181

3.2 Locating the metal ion within the protein frame: pseudocontact shifts 184

3.3 Cross-correlation rates 186

3.4 Residual dipolar couplings 188

3.5 Interplay between different restraints 190

4. NMR without1H detection 191

4.1 The protocol for 13C-detected protonless assignment of backbone and side-chains 194

4.2 Protonless heteronuclear NMR experiments tailored to paramagnetic systems 196

5. The use of lanthanides as paramagnetic probes 198

6. The case of Cu(II) proteins 202

7. Perspectives 208

8. Acknowledgments 209

9. References 209

Metalloproteins represent a large share of the proteome and many of them contain paramagnetic metal ions. The knowledge, at atomic resolution, of their structure in solution is important to understand processes in which they are involved, such as electron transfer mechanisms, enzymatic reactions, metal homeostasis and metal trafficking, as well as interactions with their partners. Formerly considered as unfeasible, the first structure in solution by nuclear magnetic resonance (NMR) of a paramagnetic protein was obtained in 1994. Methodological and instrumental advancements pursued over the last decade are such that NMR structure of paramagnetic proteins may be now routinely obtained. We focus here on approaches and problems related to the structure determination of paramagnetic proteins in solution through NMR spectroscopy. After a survey of the background theory, we show how the effects produced by the presence of a paramagnetic metal ion on the NMR parameters, which are in many cases deleterious for the detection of NMR spectra, can be overcome and turned into an additional source of structural restraints. We also briefly address features and perspectives given by the use of 13C-detected protonless NMR spectroscopy for proteins in solution. The structural information obtained through the exploitation of a paramagnetic center are discussed for some Cu2+-binding proteins and for Ca2+-binding proteins, where the replacement of a diamagnetic metal ion with suitable paramagnetic metal ions suggests novel approaches to the structural characterization of proteins containing diamagnetic and NMR-silent metal ions.