A long-lived tyrosyl radical from the reaction between horse metmyoglobin and hydrogen peroxide

The reaction between metmyoglobin (metMb) and hydrogen peroxide has been known since the 1950s to produce globin-centered free radicals. The direct electron spin resonance spectrum of a solution of horse metMb and hydrogen peroxide at room temperature consists of a multilined signal that decays in minutes at room temperature. Comparison of the direct ESR spectra obtained from the system under N(2)- and O(2)-saturated conditions demonstrates the presence of a peroxyl radical, identified by its g-value of 2.014. Computer simulations of the spectra recorded 3 s after the mixture of metMb and H(2)O(2) were calculated using hyperfine coupling constants of a(H2,6) = 1.3 G and a(H3,5) = 7.0 G for the ring and a(beta)(H1) = 16.7 G and a(beta)(H2) = 14.2 G for the methylene protons, and are consistent with a highly constrained, conformationally unstable tyrosyl radical. Spectra obtained at later time points contained a mixture of the 3 s signal and another signal that was insufficiently resolved for simulation. Efficient spin trapping with 3, 5-dibromo-4-nitrosobenzenesulfonic acid was observed only when the spin trap was present at the time of H(2)O(2) addition. Spin trapping experiments with either 5,5-dimethyl-1-pyrroline N-oxide (DMPO) or perdeuterated 2-methyl-2-nitrosopropane (MNP-d(9)), which have been shown to trap tyrosyl radicals, were nearly equally effective when the spin trap was added before or 10 min after the addition of H(2)O(2). The superhyperfine structure of the ESR spectra obtained from Pronase-treated MNP-d(9)/*metMb confirmed the assignment to a tyrosyl radical. Delayed spin trapping experiments with site-directed mutant myoglobins in which either Tyr-103 or Tyr-146 was replaced by phenylalanine indicated that radical adduct formation with either DMPO or MNP-d(9) requires the presence of Tyr-103 at all time points, implicating that residue as the radical site.