Coupling and regulation mechanisms of the flavin-dependent halogenase PyrH observed by infrared difference spectroscopy

Flavin-dependent halogenases are central enzymes in the production of halogenated secondary metabolites in various organisms and they constitute highly promising biocatalysts for regioselective halogenation. The mechanism of these monooxygenases includes formation of hypohalous acid from a reaction of fully reduced flavin with oxygen and halide. The hypohalous acid then diffuses via a tunnel to the substrate-binding site for halogenation of tryptophan and other substrates. Oxidized flavin needs to be reduced for regeneration of the enzyme, which can be performed in vitro by a photoreduction with blue light. Here, we employed this photoreduction to study characteristic structural changes associated with the transition from oxidized to fully reduced flavin in PyrH from Streptomyces rugosporus as a model for tryptophan-5-halogenases. The effect of the presence of bromide and chloride or the absence of any halides on the UV-vis spectrum of the enzyme demonstrated a halide-dependent structure of the flavin-binding pocket. Light-induced FTIR difference spectroscopy was applied and the signals assigned by selective isotope labeling of the protein moiety. The identified structural changes in α-helix and β-sheet elements were strongly dependent on the presence of bromide, chloride, the substrate tryptophan, and the product 5-chloro-tryptophan, respectively. We identified a clear allosteric coupling in solution at ambient conditions between cofactor-binding site and substrate-binding site that is active in both directions, despite their separation by a tunnel. We suggest that this coupling constitutes a fine-tuned mechanism for the promotion of the enzymatic reaction of flavin-dependent halogenases in dependence of halide and substrate availability.


S-2
Figure S1: The redox potential of FAD in PyrH was determined using the xanthine / xanthine oxidase system in conjunction with a reference dye as has been shown for other heme and FAD enzymes (1,2).PyrH (20 µM in 50 mM Tris-HCl, pH 8.0) was analyzed in a solution comprising xanthine (1.5 mL, saturated), glucose (200 µL, 100 mM), EDTA (200 µL, 10 mM), methyl viologen (10 µL, 400 µM), xanthine oxidase (5 µL, 20 mg/mL), glucose oxidase (20 µL, 5 mg/mL) and the reference dye 2-anthraquinone-sulfonate (35 µL, 2 mM).The solutions were purged with argon during the whole process and kept at 20 °C.Spectra were recorded in 15 s intervals and analyzed by a script written in Matlab R2012b (The Mathworks) using a least squares fit of reference spectra of each component.An example of the fitting procedure at the time point of 26 min is shown in (A), which was used to obtain the ratio of reduced and oxidized species at each time point accordingly, as shown in (B).

Figure S2 :
Figure S2: Nernst plot for determination of the redox potential from the ratios of oxidized and reduced species in Fig. S1 (2).The redox potential was obtained from the y-axis intercept of the linear fit with a given slope of one (shown in red).The redox potential of the reference dye 2-antraquinone-sulfonate of E0' = -225 mV vs. SHE at 20 °C was used (1) and corrected for pH 8.0 to E = -277 mV vs. SHE (3).Accordingly, the fit yields a redox potential of PyrH at pH 8 of E = -248 ± 4 mV vs. SHE from four separate experiments.

Figure S3 :
Figure S3: Purification and reconstitution of u-13 C-PyrH.(A) An SDS-PAGE of purified u-13 C-PyrH is shown in comparison to a protein ladder (Pierce Unstained Protein Molecular Weight Marker SM26610).The molecular weight of the purified protein agrees with the theoretical weight of 58 kDa.(B) UV-vis spectra of u-13 C-PyrH before and after reconstitution were recorded.Without reconstitution (red) an absorbance of FAD was not detected, whereas after reconstitution (black) the typical spectrum of FAD bound to the protein was obtained.Accordingly, all FAD in u-13 C-PyrH is at natural abundance of isotopes, because the FAD was incorporated to the protein only during reconstitution.

Figure S4 :
Figure S4: FT-IR spectra of u-13 C-PyrH (blue) in comparison to PyrH at natural isotope abundance (black).A downshift in the amide I band by 13 cm -1 and in the amide II band by 14 cm -1 were observed in agreement with band positions of other u-13 C-labeled proteins (4).

Figure S5 :
Figure S5: FT-IR difference spectra of the reduction of FAD at natural isotope abundance in u-13 C-PyrH (magenta) and PyrH (black) covering the whole spectral region.Bands marked with an asterisk do not shift and therefore originate from FAD.Small differences in wavenumber of FAD bands between both samples are caused by an overlap with bands of the protein moiety.

Figure S7 :
Figure S7: Crystal structures of PyrH with FADox in the absence and the presence of tryptophan.(A) In the absence of tryptophan, the binding loop is open and mostly dynamic, but some residual part of the loop forms an α-helical secondary structure.(B) In the presence of tryptophan, the binding loop is closed and partially forms an α-helical element, which differs from that in the absence of tryptophan (PDB entry 2WET) (7).

Table S1 :
(8)ignment of flavin bands in PyrH based on isotope labeling and comparison to the difference spectrum of reduced minus oxidized flavin mononucleotide (FMNH -/ FMNox) in solution(8).