a, Initial reaction of the cysteine thiolate with dioxygen and subsequent attack of the lysine amine onto the α-oxygen atom of the thio-(hydro)peroxy intermediate concomitantly with proton transfers and water release. b, Oxidation of both cysteine and lysine in either concerted fashion (top path) or independently (bottom path), followed by nucleophilic attack of the oxidized lysine as a O-nucleophile onto the cysteine sulfenic acid with concomitant water release. c, Initial attack of the lysine amine onto the sulfur atom of sulfenic acid or sulfinic acid to afford a sulfinamide species followed by [1,2] rearrangement driven by orbital steering. d, Depiction of selected reaction intermediates and relative free Gibbs energies (T = 298.15 K) computed at the B3LYP-D3(BJ)/def2-TZVPD//B3LYP-D3(BJ)/def2-SVPD level of theory (as described in ‘Computational details’ in Methods). The reference is given by the sulfinic acid state (the thermodynamically most stable intermediate in our investigations). Top, starting thio-(hydro)peroxy species, which bears a barrier of 12.2 kcal mol−1 for the heterolytic cleavage of the O–O bond and concurrent oxidation of the lysine residue. We compare two different pathways (using the same nomenclature as in a–c) through the sulfenic and sulfinic acids. The mechanism in a is not directly depicted, as we have not observed any concerted O–O cleavage with amino nucleophilic attack and NOS bridge formation. Only the mechanisms in b, c are depicted. A path that leads through the sulfinic acid or sulfinamide species is not viable, because the two species are far too stable. Instead, we suggest that the formal oxidation state of sulfur  is kept, with concurrent oxidation of the Lys8 (pathway from b). This would represent a reversible mechanism. For each intermediate, three different protonation states were investigated with a total cluster charge of −1, 0 and 1. Only the most stable species are shown.