New insights into controlling radical migration pathways in heme enzymes gained from the study of a dye-decolorising peroxidase

In heme enzymes, such as members of the dye-decolorising peroxidase (DyP) family, the formation of the highly oxidising catalytic Fe(iv)-oxo intermediates following reaction with hydrogen peroxide can lead to free radical migration (hole hopping) from the heme to form cationic tyrosine and/or tryptophan radicals. These species are highly oxidising (∼1 V vs. NHE) and under certain circumstances can catalyse the oxidation of organic substrates. Factors that govern which specific tyrosine or tryptophan the free radical migrates to in heme enzymes are not well understood, although in the case of tyrosyl radical formation the nearby proximity of a proton acceptor is a recognised facilitating factor. By using an A-type member of the DyP family (DtpAa) as an exemplar, we combine protein engineering, X-ray crystallography, hole-hopping calculations, EPR spectroscopy and kinetic modelling to provide compelling new insights into the control of radical migration pathways following reaction of the heme with hydrogen peroxide. We demonstrate that the presence of a tryptophan/tyrosine dyad motif displaying a T-shaped orientation of aromatic rings on the proximal side of the heme dominates the radical migration landscape in wild-type DtpAa and continues to do so following the rational engineering into DtpAa of a previously identified radical migration pathway in an A-type homolog on the distal side of the heme. Only on disrupting the proximal dyad, through removal of an oxygen atom, does the radical migration pathway then switch to the engineered distal pathway to form the desired tyrosyl radical. Implications for protein design and biocatalysis are discussed.

1, the heme precursor 5-aminolevulinic acid (250 μM final concentration), an iron supplement, Fe(III) citrate (100 μM final concentration), and isopropyl -D-thiogalactopyranoside (500 μM final concentration) were added, followed by vigorous bubbling of carbon monoxide through the liquid culture for ~30 s before sealing each shake flask with a rubber bung.Shaking continued for 16 h at 100 rpm and 30 o C, followed by centrifugation and cell lysis and loading of the supernatant to a immobilised metal-affinity chromatography column (5 ml Ni-NTA; Cytiva).Fractions containing DtpAa were concentrated using centrifugal ultrafiltration devices and applied to a preparative size-exclusion chromatography (S200; Cytiva) equilibrated in 20 mM sodium phosphate, 150 mM NaCl, pH 7.0.The purity of all DtpAa proteins was determined using SDS-PAGE, with pure fractions from the S200 column combined and stored at -20 o C until required.

Stopped-flow absorption kinetics.
Transient kinetics of the reaction of H 2 O 2 with the ferric DtpAa variants was carried out using a SX20 stopped-flow spectrophotometer (Applied Photophysics) operating in a diode-array

The influence of the engineered hole hopping pathways on Compound I lifetime
Stoichiometric addition of H 2 O 2 to WT ferric DtpAa, reveals changes in the absorption spectrum, consistent with the formation of a Compound II species, the lifetime of which is > 5 min. 33On mixing WT ferric DtpAa with > 10-fold excess of H 2 O 2 by stopped-flow, rapid formation of an intermediate species possessing spectral features analogous to Compound I, before decay to Compound II, as indicated in Fig. 3C, has been reported. 33Thus, under stochiometric conditions Compound I must still form, but reduction to Compound II occurs before measurement using static absorption spectroscopy can be acquired.Stoichiometric addition of H 2 O 2 to the Y345F variant, revealed identical behaviour as WT DtpAa (Fig. S2A).
However, for the double variant, an absorption spectrum ~ 10 s after stoichiometric addition of H 2 O 2 was recorded that displayed wavelength features to suggest the presence of a mixed Compound I and II species, which after 30 s transformed into a spectrum with pure Compound II wavelength features (Fig. S2B).For the triple variant, a Compound I spectrum was clearly observed following stochiometric H 2 O 2 addition, which slowly decays over several minutes to a Compound II spectrum (Fig. S2C).
Figure S1: Microwave power / temperature saturation map of the SigA signal.Twelve MW power values (indicated on the right) have been used to measure the spectra at six different temperatures (indicated on the panes, in Kelvins).On each pane, two spectra measured at the lowest MW power values, 3.16 and 0.79 W, are shown at magnifications 5 and 10, respectively.Different vertical scale ranges in different panes effect in different overall magnification of the spectra indicated on top right of each pane with the × sign.Instrumental parameters used to measure the spectra were as in Figure 2C.

Figure S2 :
Figure S2: UV-visible absorbance spectra of DtpAa variants at pH 5.0.Spectral changes observed following stoichiometric addition of H 2 O 2 to the ferric heme of the Y345F variant A, the Y345F/F347Y variant B, and the Y345F/F347Y/Y389F variant C. Insets show a zoom-in of the Q-band region.

Figure S3 :
Figure S3: The EPR spectra of 40 M WT DtpAa before (Control) and different time after mixing with 400 M H 2 O 2 , pH 5. The instrumental conditions used to record the spectra were the same as in Figure 2A.

Figure S4 :
Figure S4: The EPR spectra of 40 M Y345F DtpAa before (Control) and different time after mixing with 400 M H 2 O 2 , pH 5. The instrumental conditions used to record the spectra were the same as in Figure 2A.

Figure S5 :
Figure S5: Optical time course simulation for the distal pathway (triple variant) using k 5 = 2 s -1 and k 7 = 1.5 x10 6 s -1 .The simulated data is overlayed with the experimental stopped-flow data for comparison.

Table S1 :
Hamiltonian parameters generated by TRSSA-Y for an input of  C1 = 0.36 and  = 56.5°(or the complimentary angle of 61.5°) and used to simulate EPR spectrum of the YO 