The sea is worth money. It contains 3.2 mg per tonne of uranium in the form of uranyl and no one owns it yet. But how to extract it? The competition is stiff. Seawater is (still) basic and abundant carbonate ions chelate well to uranium. Calcium ions are also abundant and compete with uranyl based on size. Lu Zhu et al. argue that complexation of uranyl with dedicated protein is the way to go and describe a novel strategy here. In step one a protein database was computationally screened for proteins that are able to accommodate a uranyl cation in a pocket. The assumption is that ligand substructures coordinate to uranium as a hexagonal bipyramid or a pentagonal bipyramid. This work yielded 12,000 candidates. Additional assumptions were made: a degree of hydrogen bonding thrown in, some residue mutations allowed and a ligand oxygen to uranium bond length preset. That filter eventually rounded up around 5000 proteins.

The ultimate super uranyl-binding protein (SUP) that came up in the search was at one time sourced from the otherwise very unremarkable methanobacterium thermoautotrophicum, an anaerobe that can be found in sewage sludge. The binding affinity is sufficient and can even improve with strategic mutations in the protein. The affinity for other ions such as calcium, copper or the vanadyl ion is far less. In synthetic seawater it was possible to retrieve 17% uranyl in 30 minutes from a 13 nM solution with SUP fused with a maltose binding protein on amylose resin.