Notably this represents the first stable and functional CuHis3 site in aqueous solution. A type 1 copper

site has been designed within a four-stranded α-helical bundle (generated from a single peptide strand) with two His, one Cys and an exogenous fourth weakly interacting axial ligand. The nature of this fourth ligand is crucial in establishing a type 1 or 2 site, and so it was necessary to prevent water access. Like type 1 sites in native redox proteins, the mimic displayed fast electron reaction rates [20]. Various studies looking at the binding of heavy metals to thiol rich sites in the hydrophobic interior of coiled coils or helical bundles have been reported [21, 22 and 23], as these provide important insight into heavy metal biochemistry, and have allowed challenging and fundamental questions about metals in biology to be answered using these high throughput screening simplified scaffolds. For example, insight into metal exchange dynamics and the mechanism by which metal ions are sequestered into thiol sites [24]; whether the location of a metal site along a coiled coil alters its chemistry [17•• and 25]; the importance of ligand preorganisation for metal ion binding to symmetric

a or d substituted sites [ 26], or an asymmetric equivalent generated in a ABT-263 mouse single chain three-helix bundle [ 27]; and the importance of stereochemically active lone pairs (demonstrated for As(III) and Pb(II)) and the role second coordination sphere residues play in accommodating these, thereby dictating the binding mode [ 28]. The recent report of the Dolutegravir datasheet 207Pb NMR chemical shift of a water soluble 207PbCys3 site, is of huge significance considering the importance of these sites in lead toxicity and the wide chemical shift range. Intriguingly 207Pb

NMR was shown to be capable of discriminating between similar but not identical PbCys3 sites, and as such could be a very powerful tool in further understanding both metalloprotein design and lead toxicity [ 29•]. The design of multinuclear metal ion sites can be more challenging. However, an important success is the due ferri (two iron) family of designed proteins [30]. These have been redesigned to introduce O2-dependent phenol oxidase activity, by engineering an active site cavity in the interior of either a four-stranded heterotetrameric coiled coil [31] or a four-helix bundle (helix-loop-helix dimer) [32] (see Figure 3A). In addition to Fe, the latter was also able to bind Zn, Co or Mn [33]. The activity was then reprogrammed from the oxidation of hydroquinones to the N-hydroxylation of arylamines by four mutations, notably the addition of a His ligand in the active site (inspired by the active site of AurF) [ 34••]. A different dinuclear Fe complex, a mimic of the hydrogenase active site, has been linked to an α-helix through a non-natural residue.