S-Adenosyl-L-methionine methylene radical-mediated cyclopropanation in nature

A HemN-like radical S-adenosyl-L-methionine (SAM) enzyme adds a SAM-based methylene radical generated via a radical relay to an sp2 carbon of substrate to produce a SAM-substrate covalent adduct, which is converted to the cyclopropane ring by a methyltransferase in the biosynthesis of natural product CC-1065.

The paper in Nature Communications is here: go.nature.com/2JSnn9N

Cyclopropane-containing complex natural products present significant challenges for both chemical synthesis and biosynthesis because of the highly strained cyclopropane ring. Olefin cyclopropanation via addition of a reactive one-carbon species (e.g., carbenoids) is well-developed in synthetic chemistry. Moreover, cytochrome P450 enzymes were recently engineered to catalyze olefin cyclopropanation via a carbene transfer mechanism. However, natural cyclopropane biosynthesis employing a one-carbon donor strategy is very limited. So far, cyclopropane fatty acid/mycolic acid synthases are the known enzymes that catalyze the cyclopropanation of unsaturated fatty acid/mycolic acid using S-adenosyl-L-methionine (SAM) as the methyl donor.

CC-1065, gilvusmycin, yatakemycin (YTM), duocarmycin SA, and duocarmycin A are members of the spirocyclopropylcyclohexadienone family of antitumor antibiotics. In light of their remarkably potent cytotoxicity using a unique shape-dependent DNA alkylation mechanism, these natural products have attracted increasing concerns for chemical, biological and pharmaceutical studies. Our group has focused on the biosynthetic studies of CC-1065 and YTM, and we have identified their biosynthetic gene clusters, and proposed their biosynthetic pathways, respectively. Previously, a HemN-like radical SAM enzyme C10P was revealed essential for the formation of the cyclopropane ring in CC-1065, but how the cyclopropanation proceeds is elusive. In this work, C10P and a SAM-dependent methyltransferase C10Q are verified as a two-component cyclopropanase system capable of catalyzing a chemically challenging cyclopropanation reaction in CC-1065 biosynthesis.

Systematic gene deletions led us to the identification of c10Q within the CC-1065 biosynthetic gene cluster that is also required for the cyclopropane construction. Next, in vitro enzymatic assays unraveled that the two-component system (reconstituted C10P and C10Q) is sufficient to catalyze cyclopropane ring formation in the present of SAM and sodium dithionite under strictly anaerobic conditions. Based on subsequent product analysis and labeling experiments, especially the observation of key SAM-substrate adduct intermediate, we have raised a proposal as follows. Reductive cleavage of the first SAM in C10P yields a highly reactive 5’-deoxyadenosyl (dAdo) radical, which abstracts a hydrogen atom from the activated methyl group of the second SAM in C10P. A SAM-based methylene radical is thus produced and then adds to an sp2 carbon of substrate to form a SAM-substrate covalent adduct. C10Q converts this adduct to CC-1065 via an intramolecular SN2 cyclization with elimination of S-adenosylhomocysteine.

Combined with the findings from the methyltransferases TbtI, NosN, and ChuW, our results from C10P likely suggest a catalytic mechanism for these HemN-like proteins, that is, they utilize the two bound SAM molecules to successively generate a dAdo radical and a SAM-based methylene radical. Hence, this work also raises an open question of what is the function of the second SAM in the prototype HemN from E. coli. In this prototype enzyme, the dAdo radical was proposed to abstract a hydrogen atom from coproporphyrinogen-III, thereby initiating the oxidative decarboxylation. It would be thus intriguing to address this exact mechanism that may unravel the powerful catalysis by the members of these radical SAM proteins. In addition, it will be interesting to investigate the enzymatic mechanism of C10Q, a methyltransferase catalyzing cyclopropanation via an intramolecular SN2 cyclization process.