Redox Cofactor Rotates during Its Stepwise Decarboxylation: Molecular Mechanism of Conversion of Coproheme to Heme b

Coproheme decarboxylase (ChdC) catalyzes the last step in the heme biosynthesis pathway of monoderm bacteria with coproheme acting both as redox cofactor and substrate. Hydrogen peroxide mediates the stepwise decarboxylation of propionates 2 and 4 of coproheme. Here we present the crystal structures of coproheme-loaded ChdC from Listeria monocytogenes (LmChdC) and the three-propionate intermediate, for which the propionate at position 2 (p2) has been converted to a vinyl group and is rotated by 90° compared to the coproheme complex structure. Single, double, and triple mutants of LmChdC, in which H-bonding interactions to propionates 2, 4, 6, and 7 were eliminated, allowed us to obtain the assignment of the coproheme propionates by resonance Raman spectroscopy and to follow the H2O2-mediated conversion of coproheme to heme b. Substitution of H2O2 by chlorite allowed us to monitor compound I formation in the inactive Y147H variant which lacks the catalytically essential Y147. This residue was demonstrated to be oxidized during turnover by using the spin-trap 2-methyl-2-nitrosopropane. Based on these findings and the data derived from molecular dynamics simulations of cofactor structures in distinct poses, we propose a reaction mechanism for the stepwise decarboxylation of coproheme that includes a 90° rotation of the intermediate three-propionate redox cofactor.


Generation of LmChdC variants
Site-directed mutagenesis to obtain LmChdC M149A, Q187A and M149A/Q187A variants using the QuikChange Lightning Kit (Agilent Technologies) has been described previously 1 .

Expression and purification of LmChdC wild-type and variants.
LmChdC wild-type and all variants were subcloned into a modified version of the pET21b(+) expression vector with an N-terminal StrepII-tag, cleavable TEV protease, or into a pETM11 expression vector with an N-terminal, TEV cleavable 6×His-tag, expressed in E. coli Tuner (DE3) cells (Merck/Novagen) and purified via a StrepTrap HP or a HisTrap HP 5 mL column (GE Healthcare) as described in detail previously 3 . The enzymes were reconstituted by the addition of equimolar concentrations of free Fe(III) coproporphyrin III chloride (coproheme; Frontiers Scientific) (in 50 mM phosphate buffer, pH 7.0, predissolved in 0.5 M NaOH, ε 390 = 128,800 M -1 cm -1 ) 4 to apo-LmChdC dissolved in 50 mM phosphate buffer, pH 7.0.

Coproheme decarboxylase activity
In order to test coproheme decarboxylase activity of the LmChdC Y147A variant, titrations were performed where small (sub-equimolar) aliquots of H 2 O 2 were added to coproheme-ChdCs. In a typical experiment H 2 O 2 was added in 1 µM aliquots to 5 µM coproheme bound to excess ChdC (20 µM). To ensure complete reaction of the H 2 O 2 , an interval of at least 15 min was set between each peroxide addition. The titrations were performed in a stirred 1 mL cuvette at room temperature and monitored by UV-vis absorption spectroscopy between 250 and 700 nm, using a scanning spectrophotometer (Cary60).

Coproheme decarboxylase activity triggered with chlorite
In order to determine the stoichiometry of chlorite to coproheme, titrations were performed where small (sub-equimolar) aliquots of chlorite were added to coproheme-LmChdC. In a typical experiment chlorite was added in 1 µM aliquots to 7 µM coproheme-LmChdC. The reaction was completed rapidly and, in contrast to titration with hydrogen peroxide, there was no need to wait for 15 min between additions of the oxidant. The titrations were performed in a stirred cuvette (1 mL sample volume) at room temperature and monitored by UV-vis absorption spectroscopy between 250 and 700 nm, using a scanning spectrophotometer (Cary60).

Pre-steady-state kinetics
Transient pre-steady-state kinetics were measured with a stopped-flow apparatus (SX-18MV or pi-star fitted equipped with diode array detector or a monochromator) from Applied Photophysics. The optical quartz cell has a path length of 10 mm and a volume of 20 µL. The fastest time for mixing was 0.68 ms. Measurements were performed at room temperature. For single wavelength measurements a minimum of three reactions were monitored for each substrate concentration. Hydrogen peroxide, hypochlorous acid, chlorite, peroxoacetic acid, dicumyl peroxide, and tert-butyl hydroperoxide were used as substrate in the concentration range from 1 µM to 1 mM. 1 -5 µM LmChdC variant Y147H was mixed with substrates at pH 7, pH 8.5 (50 mM HEPES) and pH 10 (50 mM borate buffer) and reactions were followed with the diode array detector for screening of the reactivity. For the determination of k app of Compound I formation of LmChdC Y147H, 5 µM enzyme were mixed with 250, 500, 750, 1000, 1250, 1500 µM chlorite at pH 7 and the reaction was followed at 388 nm. The rate was determined form the slope of the plot of k obs versus the chlorite concentration. Simulated spectra were obtained with Pro-Kineticists software (Applied Photophysics).

Crystallization
Crystallization trials of LmChdC wild-type were carried out in SWISSCI MRC three-well crystallization plates (Molecular Dimensions) using the vapor diffusion method. Crystallization drops were set using the mosquito LCP crystallization robot (TTP LabTech). The reservoir was filled with 35 µL of crystallant solution. In the sample wells drops containing proteins of 175 µM % (w/v) PEG 4000, 10 % (v/v) 2-propanol, which was a suitable condition for freezing the crystals without any further cryo-protectants. Addition of 2 µL of the mother liquor to the drop prior to freezing diluted the cyanide concentration significantly. 4.7 X-ray data collection, structure determination, and refinement Datasets were collected at beamline ID29 5 of the European Synchrotron Radiation Facility (ESRF, Grenoble, France) at 100 K using a DECTRIS PILATUS 6M detector. The data set was processed with XDS and symmetry equivalent reflections merged with XDSCONV 6 . Intensities were not converted to amplitudes. The high-resolution cutoff was based on a CC1/2* criterion 7 .
The phase problem was solved by molecular replacement using Phaser-MR 8 taking pdb structure 5LOQ, ChdC from Listeria monocytogenes 4 . The model was further improved by iterative cycles of manual model building using COOT 9 and maximum likelihood refinement Reevaluation and further refinement of the structure of LmChdC, originally deposited in the pdb-data bank as 5LOQ, was performed as above using COOT and PHENIX-Refine.
Restraints for monovinyl, monopropionate deuteroheme (Ligand ID VOV) were generated using eLBOW taking an sdf file as input and applying the final-geometry option.

Detection of tyrosyl radicals in LmChdC
Spin trapping experiments of LmChdC, reacting with H 2 O 2 , were performed using MNP 20 µg of each sample in 1×PBS were S-alkylated with iodoacetamide and further digested with sequencing grade modified trypsin (Promega). 5 µg of the peptide mixture were analyzed using a Dionex Ultimate 3000 HPLC-system (Thermo Scientific) directly linked to a QTOF instrument (maXis 4G ETD, Bruker) equipped with the standard ESI source in the positive ion, DDA mode (= switching to MSMS mode for eluting peaks). MS-scans were recorded (range: 150-2200 m/z, spectra rate: 1.0 Hz) and the 5 highest peaks were selected for fragmentation.
Instrument calibration was performed using ESI tuning mix (Agilent).
Peptides were separated using a Thermo BioBasic C18 separation column (5 µm particle size, 150×0.320 mm. A gradient from 96% solvent A and 4% solvent B (Solvent A: 65 mM ammonium formiate buffer, pH 3.0; Solvent B: 80% ACN and 20% A) to 40% B in 35 min was applied, followed by a 15 min gradient from 40% B to 94% B, at a flow rate of 6 µL/min at 32 °C.

Electronic absorption
Electronic absorption spectra were recorded using a 5 mm NMR tube (300 nm min -1 scan rate) or a 1 mm cuvette (600 nm min -1 scan rate) at 25 °C by means of a Cary 60 spectrophotometer (Agilent Technologies) with a resolution of 1.5 nm. For the differentiation process, the Savitzky-Golay method was applied using 15 data points (LabCalc, Galactic Industries, Salem, NH). No changes in the wavelength or in the bandwidth were observed when the number of points was increased or decreased.

Resonance Raman (RR)
The resonance Raman (RR) spectra were obtained at 25 °C using a 5 mm NMR tube by excitation with the 406.7 and 413.1 nm lines of a Kr + laser (Coherent, Innova300 C, Coherent, Santa Clara, CA, USA). Backscattered light from a slowly rotating NMR tube was collected and focused into a triple spectrometer (consisting of two Acton Research SpectraPro 2300i instruments and a SpectraPro 2500i instrument in the final stage with gratings of 3600 grooves/mm and 1800 grooves/mm) working in the subtractive mode, equipped with a liquid nitrogen-cooled CCD detector.
For the low temperature experiments, a 50 μL drop of the sample was put in a 1.5 cm diameter quartz crucible that was positioned in a THMS600 cryostat (Linkam Scientific Instruments, Surrey, UK) and frozen. After freezing the sample, the cryostat was positioned vertically in front of the triple spectrometer and the laser light was directed onto the quartz window. To avoid sample denaturation or photo-reduction, the laser position was changed frequently. The sample temperature was maintained at 80 K. With the low temperature setup, the plasma laser lines of the 406.7 nm excitation wavelength in the low frequency region are very intense. In particular those at 381 and 398 cm -1 overlap with the propionate bending modes, preventing their assignment. Hence, the RR spectra in the low frequency region of the coproheme complexes were obtained with the 413.1 nm excitation wavelength to avoid plasma lines in the propionate bending region. It should be noted that the spectra of the coproheme complexes obtained with λ exc 406.7 nm (in resonance with the Soret maximum) and λ exc 413. 1 nm are identical, as shown in Figure S5 where, as an example, the spectra of the coproheme-WT and -Y113A obtained with the two excitation wavelengths, are compared.
A spectral resolution of 1.2 cm −1 and spectral dispersion of 0.40 cm -1 /pixel were calculated theoretically on the basis of the optical properties of the spectrometer for the 3600 grating. The RR spectra were calibrated with indene and carbon tetrachloride as standards to an accuracy of 1 cm −1 for intense isolated bands. All RR measurements were repeated several times under the same conditions to ensure reproducibility. To improve the signal-to-noise ratio, a number of spectra were accumulated and summed only if no spectral differences were noted. Absorption spectra were measured both prior and after RR measurements to ensure that no degradation occurred under the experimental conditions used. All spectra were baseline-corrected. observed, the sample was transferred to the THMS600 cryostat (see above), frozen at 80 K and RR spectra taken. After the RR spectra, the sample was warmed to 25 °C and the UV-Vis spectrum recorded and compared with that obtained before freezing to ensure that no degradation had occurred. The titration with H 2 O 2 was then continued until new variations in the UV-Vis spectrum appeared and the sample was frozen again. This experiment was repeated several times to ensure reproducibility.

MD simulations
As a starting configuration for molecular dynamics simulations, we used the apo ChdC i.e. with their p2 towards Tyr147, were obtained by manual rotation of coproheme and monovinyl, monopropionate deuteroheme to resemble the binding pose of coproheme in Ref. 16 (PDB entry 5T2K, pose 0). All MD simulations were carried out using the GROMOS11 software simulation package 17 , employing the 54a8 forcefield 18 . Proteins were energy-minimized in vacuum using the steepest-descent algorithm and subsequently solvated in a rectangular, periodic and pre-equilibrated box of single point charge (SPC) water 19 . Minimum solute to boxwall distances were set to 0.75 in all dimensions. This led to systems containing about 104 thousand atoms for the pentamer. Another minimization in water was performed using the steepest descent algorithm. To achieve electroneutrality of the system, 65 sodium ions were added. For the equilibration, the following protocol was used: initial velocities were randomly assigned according to a Maxwell-Boltzmann distribution at 60 K. All solute atoms were positionally restrained with a harmonic potential using a force constant of 2.5×10 4 kJmol −1 nm −2 .
In each of the four subsequent 20 ps MD simulations, the force constant of the positional restraints was reduced by one order of magnitude and the temperature was increased by 60 K.   Figure S4. Comparison of the RR spectra in the low frequency region of the coproheme-LmChdC complexes of the WT and mutants obtained at 80 K (λ exc 413.1 nm). The bands tentatively assigned to the bending mode δ(C β C c C d ) of the propionate groups in positions 2, 4, 6 and 7 are reported in beige, green, brown and light blue, respectively. Accordingly, the mutant label colours indicate the position of the propionate group/s with which the mutated residues interact. In magenta is reported the band due to the 6cLS species. The spectrum of the coproheme-WT complex is reported in red. The spectra have been shifted along the ordinate axis to allow better visualization. Experimental conditions: laser power at the sample 5-10 mW, average of 6 spectra with a 120 min integration time (

S3
Supporting Tables   Table S1. Distances (Å) between the oxygen atoms of the propionate groups in positions 2, 4, 6 and 7 in the subunits A, B, C, D, E and the studied mutated residues of coproheme-LmChdC (6FXJ). Hydrogen bonding networks are marked in bold.