Selenium is right below sulfur on the periodic table, and in many ways it's a "sulfur but even more so" case. For example, got a sulfur molecule that smells awful? The selenium analog will almost certainly smell worse. It's an element that balances on a toxicological tightrope. It's an essential trace nutrient - you will get into trouble if you don't have enough selenium in your diet. It is also toxic - you will get into trouble if you have too much selenium in your diet. The dose really does make the poison, as Paracelsus told us, and the dose makes the nutrient, too. The fact that it's essential tells you that it's almost certainly being incorporated into some biomolecules, and indeed, the Se analog of cysteine is an important amino acid in some key niches. There are over 100 human proteins that contain selenocysteine, and they are nonfunctional without it. Selenouridine is also known to show up in transfer RNA molecules, and it must be doing something interesting there, too,
But what other sulfur-containing biomolecules could have rare-but-important selenium analogs? This new paper has some surprising answers, obtained by rooting in a targeted manner through all sorts of microbial genomes. The authors searched for genes that colocalized with the (very few) known selenium-specific enzymes, on the theory that these might well be part of related biosynthetic clusters. A glycosyltranferase popped up rather quickly, one that no one has really paid any attention to (most genes are still in that category at this point, to be honest), along with a homologue of gene coding for a known carbon-sulfur bond-forming enzyme, also a strong candidate. And these two enzymes turn out to be part of a whole selenometabolome that has been unexplored until now - the selenoneine synthase (now designated SenA) and selenosugar enzyme (now designated SenB) take the number of selenium-specific enzymes from two to four, and produce odd-looking things like selenoglucose (with an SeH hanging off the anomeric center), selenoGlcNac, and the selenourea compound selenoneine. Since Se compounds like this are surely quite reactive, these intermediates will surely be going on to produce even more trace Se compounds that have yet to be worked out.