Travels with Carbon-Centered Radicals. 5′-Deoxyadenosine and 5′-Deoxyadenosine-5′-yl in Radical Enzymology

Abstract

As a graduate student under Professor R. H. Abeles, I began my journey with 5'-deoxyadenosine, studying the coenzyme B12 (adenosylcobalamin)-dependent dioldehydrase (DDH). I proved that suicide inactivation of dioldehydrase by glycolaldehyde proceeded with irreversible cleavage of adenosylcobalamin to 5'-deoxyadenosine. I further showed that suicide inactivation by [2-(3)H]glycolaldehyde produced 5'-deoxy[(3)H]adenosine, the first demonstration of hydrogen transfer to adenosyl-C5' of adenosylcobalamin. The tritium kinetic isotope effect (T)k was 15, which correlated well with the measurement (D)k = 12 for transformation of [1-(2)H]propane-1,2-diol to [2-(2)H]propionaldehyde by DDH. After establishing my own research program, I returned to the glycolaldehyde inactivation of DDH, showing by EPR that suicide inactivation produced glycolaldehyde-2-yl. In retrospect, suicide inactivation involved scission of adenosylcobalamin to 5'-deoxyadenosine-5'-yl, which abstracted a hydrogen from glycolaldehyde. Captodative-stabilized glycolaldehyde-2-yl could not react further, leading to suicide inactivation. In 1986, my colleagues and I took up the problem of the mechanism by which lysine 2,3-aminomutase (LAM) catalyzes S-adenosylmethionine (SAM) and pyridoxal-5'-phosphate (PLP)-dependent interconversion of l-lysine and l-β-lysine. Because the reaction followed the pattern of adenosylcobalamin-dependent rearrangements, I postulated that SAM might be an evolutionary predecessor to adenosylcobalamin. Testing this hypothesis, we traced hydrogen transfer from lysine through the adenosyl-C5' of SAM to β-lysine. Thus, the 5'-deoxyadenosyl of SAM mediated hydrogen transfer by LAM exactly as in adenosylcobalamin mediated hydrogen transfer in B12-dependent isomerizations. The mechanism postulated that SAM cleaves to form 5'-deoxyadenosine-5'-yl followed by abstraction of C3(H) from PLP-α-lysine aldimine to form PLP-α-lysine-3-yl. PLP-α-lysine-3-yl isomerizes to pyridoxal-β-lysine-2-yl, and a hydrogen abstraction from 5'-deoxyadenosine regenerates 5'-deoxyadenosine-5'-yl and releases β-lysine. Of four radicals in the postulated mechanism, three have been characterized by EPR spectroscopy as kinetically competent intermediates. The analysis of the role of iron allowed researchers to elucidate the mechanism by which SAM is cleaved to 5'-deoxyadenosine-5'-yl. LAM contains one [4Fe-4S] cluster ligated by three cysteine residues. As shown by ENDOR spectroscopy and X-ray crystallography, the fourth ligand to the cluster is SAM, through the methionyl carboxylate and amino groups. Inner sphere electron transfer within the [4Fe-4S](1+)-SAM complex leads to [4Fe-4S](2+)-Met and 5'-deoxyadenosine-5'-yl. The iron-binding motif in LAM, CxxxCxxC, found by other groups in four other SAM-dependent enzymes, is the founding motif for the radical SAM superfamily. These enzymes number in the tens of thousands and are responsible for highly diverse and chemically difficult transformations in the biosphere. Available information supports the hypothesis that this superfamily provides the chemical context from which the much more structurally complex adenosylcobalamin evolved.