Unique Biradical Intermediate in the Mechanism of the Heme Enzyme Chlorite Dismutase

Abstract

The heme enzyme chlorite dismutase (Cld) catalyzes O-O bond formation as part of the conversion of the toxic chlorite (ClO₂ ^-) to chloride (Cl^-) and molecular oxygen (O₂). Enzymatic O-O bond formation is rare in nature, and therefore, the reaction mechanism of Cld is of great interest. Microsecond timescale pre-steady-state kinetic experiments employing Cld from Azospira oryzae (AoCld), the natural substrate chlorite, and the model substrate peracetic acid (PAA) reveal the formation of distinct intermediates. AoCld forms a complex with PAA rapidly, which is cleaved heterolytically to yield Compound I, which is sequentially converted to Compound II. In the presence of chlorite, AoCld forms an initial intermediate with spectroscopic characteristics of a 6-coordinate high-spin ferric substrate adduct, which subsequently transforms at k _obs = 2-5 × 10⁴ s^-1 to an intermediate 5-coordinated high-spin ferric species. Microsecond-timescale freeze-hyperquench experiments uncovered the presence of a transient low-spin ferric species and a triplet species attributed to two weakly coupled amino acid cation radicals. The intermediates of the chlorite reaction were not observed with the model substrate PAA. These findings demonstrate the nature of physiologically relevant catalytic intermediates and show that the commonly used model substrate may not behave as expected, which demands a revision of the currently proposed mechanism of Clds. The transient triplet-state biradical species that we designate as Compound T is, to the best of our knowledge, unique in heme enzymology. The results highlight electron paramagnetic resonance spectroscopic evidence for transient intermediate formation during the reaction of AoCld with its natural substrate chlorite. In the proposed mechanism, the heme iron remains ferric throughout the catalytic cycle, which may minimize the heme moiety's reorganization and thereby maximize the enzyme's catalytic efficiency.

Scheme 1. Currently Proposed Putative Mechanisms for Heterolytic (I) or Homolytic (II) Cl–O Bond Cleavage and O–O Bond Formation by the Enzyme Chlorite Dismutase

Figure 1

AoCld pre-steady-state kinetics with PAA and chlorite substrates observed with Nanospec continuous flow UV–vis spectroscopy. PAA substrate: (A,C) 350 μM AoCld and 1 mM PAA, both dissolved in 100 mM KPi, pH 7, were rapidly mixed with a flow rate of 7 mL min^–1. Spectra were obtained for 2.5 ms at 29 °C. Chlorite substrate: (B,D) 330 μM AoCld and 50 mM sodium chlorite, both dissolved in 100 mM KPi, pH 7, were rapidly mixed with a flow rate of 20 mL min^–1. Spectra were obtained for 700 μs at 29 °C. The reaction occurs so fast that the standard ferric resting state cannot be observed, but it is illustrated as a reference (unreacted sample). Arrows indicate the direction of absorption changes. (E) Reconstructed spectra of the catalytic intermediates of the reaction of PAA and AoCld after SVD analysis. EI, Early Intermediate; CpdI, Compound I; CpdII, Compound II. (F) Reconstructed spectra of the catalytic intermediates of the reaction of sodium chlorite and AoCld. I1, Intermediate 1; I2, Intermediate 2.

Figure 2

Time traces of the intermediate species of AoCld detected in the MHQ (solid lines) and Nanospec (dashed lines) experiments with PAA as the substrate. The traces from the Nanospec experiments have been recalculated to 9 °C. EI, early intermediate; R1, initially formed radical; and R2, second formed radical.

Figure 3

AoCld pre-steady-state kinetics with PAA and chlorite substrates observed with MHQ-EPR spectroscopy. PAA substrate: (A,C,E) 1.5 mM PAA and 500 μM AoCld, both dissolved in 100 mM KPi, pH 7, were mixed at 9 °C and frozen rapidly. The control experiment consisted of 500 μM AoCld and 100 mM KPi buffer, pH 7, mixed and quenched at 97 μs. The reaction between PAA and AoCld was quenched at 97 μs, 800 μs, 2 ms, and 16 ms after mixing. Chlorite substrate: (B,D,F) 75 mM sodium chlorite and 500 μM AoCld, both dissolved in 100 mM KPi, pH 7, were mixed at 9 °C and frozen rapidly. The reaction between sodium chlorite and AoCld was quenched at 97 μs, 300 μs, 2 ms, and 200 ms after mixing. EPR spectra in (A–D) were obtained with microwave frequency 9.41 GHz, microwave power 20 mW, modulation frequency 100 kHz, modulation amplitude 10 G, and temperature 20 K. EPR spectra in (E,F) were obtained with microwave frequency 9.406 GHz, microwave power 0.2 mW, modulation frequency 100 kHz, modulation amplitude 2 G, and temperature 20 K, and four spectra were averaged.

Figure 4

Time traces of the intermediate species of AoCld detected in the MHQ (solid lines) and Nanospec (dashed lines) experiments with chlorite as the substrate. The traces from the Nanospec experiments were recalculated to 9 °C. The vertical dotted line (black) indicates the time constant for a single turnover (τ_total = 1/k_cat) recalculated to 9 °C. HS, HS ferric intermediate; Cpd T, triplet-state intermediate Compound T; LS, LS ferric intermediate; radical, amino acid cation radical; I1, intermediate 1; and I2, intermediate 2.

Figure 5

Assignment of the Compound T EPR signal as a triplet-state biradical. (A) Cpd T EPR signal was recorded at two different microwave frequencies: 9.62054 and 9.40333 GHz. The AoCld MHQ sample at 97 μs reaction time with chlorite was measured at microwave frequencies of 9.40333 (blue or red line) and 9.62054 GHz (black line). The blue spectrum shows that the 9.40 GHz spectrum shifted to a field corresponding to that of the 9.62 GHz spectrum. The red spectrum shows that the 9.40 GHz spectrum transformed to the frequency 9.62 GHz. The blue and black spectra overlap, which shows that the spectral features are independent of the frequency; that is, they do not correspond to real g values. The red spectrum (transformed 9.40 GHz) slightly extends beyond the limits of the black spectrum, which again proves that the main spectral features are not real g values, that is, they cannot represent two independent S = 1/2 radicals. (B) Experimental and simulated Cpd T signal of AoCld reacted with chlorite as a triplet-state biradical (blue) or a chlorine-based radical (red). The top trace (black) is the AoCld MHQ 97 μs reaction mixture with chlorite recorded at 9.62054 GHz. The middle trace (blue) is a simulation assuming S = 1, g_iso = 2.0135, D = 0.0030 cm^–1, E = 0.00025 cm^–1, and linewidths w_z = 0.0045 and w_xy = 0.0015 (standard deviation of g strain). The lower trace (red) is a simulation assuming S = 1/2, I (³⁵Cl) = I (³⁷Cl) = 3/2, relative intensity ³⁵Cl/³⁷Cl component = 1:0.32, g_iso = 2.014, A_zyx(³⁵Cl) = 40, 4, and 4 gauss, linewidth W_zyx(³⁵Cl) = 3, 3, and 10 gauss, A_zyx(³⁷Cl) = 33, 3.3, and 3.3 gauss, and linewidth W_zyx(³⁷Cl) = 3, 3, and 10 gauss.

Scheme 2. Proposed Reaction Mechanism of the Reaction between AoCld and (I) Peracetic Acid (II) Chlorite Based on Direct Spectroscopic Evidence Presented in This Work

AA, amino acid.