Experimental and Computational Elucidation of C( sp 3)–H Fluorination Barriers in an Iron(II)- and 2-Oxoglutarate-Dependent Halogenase

Abstract

Incorporation of fluorine into pharmaceuticals, agrochemicals, and molecular-imaging agents is of growing importance. Multiple synthetic fluorination methods have recently emerged, and metalloenzymes that are potentially capable of C(sp3)–H fluorination have been reported. Nevertheless, direct, regioselective fluorination of aliphatic carbon centers remains an unsolved problem. Here, we show for the iron(II) and 2-oxoglutarate-dependent (Fe/2OG) l-lysine 4-chlorinase, BesD, which might be envisaged to support C(sp3)–H fluorination by the direct cognate of its native chlorination mechanism, that the enzyme can (1) coordinate F– at its Fe(II) cofactor, (2) activate O2 to form a cis-FeIV(O)(F) (fluoroferryl) intermediate, and (3) use the intermediate to abstract hydrogen from its substrate. In what would be the key final step, fluorine (F•) coupling to the substrate radical is unable to compete with the hydroxyl-radical (HO•) “rebound” step characteristic of related hydroxylases. Electron paramagnetic resonance (EPR) and X-ray absorption spectroscopic (XAS) data establish that fluorine remains bonded to the iron cofactor through steps 1–3 and therefore available for transfer to the substrate radical. QM/MM calculations suggest that the F•-coupling step is associated with an activation barrier considerably higher than that of HO• rebound, consistent with the observed outcome. The findings experimentally verify prior proposals that the impediment to C(sp3)–H fluorination by the canonical mechanism of an Fe/2OG halogenase lies solely in the final radical-coupling step and set the stage for exploration of whether a potentially surmountable geometric barrier or an insurmountable electronic one is primarily responsible.