Preprint Article Version 2 Preserved in Portico This version is not peer-reviewed

Structural and Biochemical Characterization of a Dye Decolorizing Peroxidase from Dictyostelium discoideum

Version 1 : Received: 8 May 2021 / Approved: 10 May 2021 / Online: 10 May 2021 (10:12:43 CEST)
Version 2 : Received: 3 June 2021 / Approved: 3 June 2021 / Online: 3 June 2021 (12:12:22 CEST)

A peer-reviewed article of this Preprint also exists.

Rai, A.; Klare, J.P.; Reinke, P.Y.A.; Englmaier, F.; Fohrer, J.; Fedorov, R.; Taft, M.H.; Chizhov, I.; Curth, U.; Plettenburg, O.; Manstein, D.J. Structural and Biochemical Characterization of a Dye-Decolorizing Peroxidase from Dictyostelium discoideum. Int. J. Mol. Sci. 2021, 22, 6265. Rai, A.; Klare, J.P.; Reinke, P.Y.A.; Englmaier, F.; Fohrer, J.; Fedorov, R.; Taft, M.H.; Chizhov, I.; Curth, U.; Plettenburg, O.; Manstein, D.J. Structural and Biochemical Characterization of a Dye-Decolorizing Peroxidase from Dictyostelium discoideum. Int. J. Mol. Sci. 2021, 22, 6265.

Journal reference: Int. J. Mol. Sci. 2021, 22, 6265
DOI: 10.3390/ijms22126265


A novel cytoplasmic dye decolorizing peroxidase from Dictyostelium discoideum was investigated that oxidizes anthraquinone dyes, lignin model compounds and general peroxidase substrates like ABTS efficiently. Unlike related enzymes, an aspartate residue replaces the first glycine of the conserved GXXDG motif in Dictyostelium DyPA. In solution, Dictyostelium DyPA exists as a stable dimer with the side chain of Asp146 contributing to the stabilization of the dimer interface by extending the hydrogen bond network connecting two monomers. To gain mechanistic insights, we solved the Dictyostelium DyPA structures in the absence of substrate as well as in the presence of potassium cyanide and veratryl alcohol to 1.7, 1.85, and 1.6 Å resolution, respectively. The active site of Dictyostelium DyPA has a hexa-coordinated heme iron with a histidine residue at the proximal axial position and either an activated oxygen or CN- molecule at the distal axial position. Asp149 is in an optimal conformation to accept a proton from H2O2 during the formation of compound I. Two potential distal solvent channels and a conserved shallow pocket leading to the heme molecule were found in Dictyostelium DyPA. Further, we identified two substrate-binding pockets per monomer in Dictyostelium DyPA at the dimer interface. Long-range electron transfer pathways associated with a hydrogen-bonding network that connects the substrate-binding sites with the heme moiety are described.

Subject Areas

Dye decolorizing-type peroxidase; heme peroxidases; lignin degradation; Dictyostelium discoideum; B-type DyP; electron paramagnetic resonance (EPR) spectroscopy; compound I; enzyme kinetics; crystal structure; long-range electron transfer

Comments (1)

Comment 1
Received: 3 June 2021
Commenter: Dietmar Manstein
Commenter's Conflict of Interests: Author
Comment: Based on the comments of Reviewer I, we have introduced the following changes:
The text has been modified accordingly:Line 141-148:For biochemical characterization, recombinant Dictyostelium DyPA was over-produced and purified from Escherichia coli cells as described previously [15]. Purified Dictyostelium DyPA protein (Apo-form) was faintly yellow with a very small Soret peak at 410 nm and a Reinheitszahl (Rz value ASoret/Abs280) of 0.13, indicating the presence of a small, substoichiometric amount of heme. Heme reconstitution was performed by adding hemin chloride in a 2:1 molar excess, followed by size exclusion chromatography to remove any unbound heme. Heme reconstituted Dictyostelium DyPA was used throughout the study unless otherwise stated. Line 318. The authors provide an apparent “melting temperature” for DypA. However, Fig4D shows a single experimental data different to 100% or 0% activity (that at 50ºC). Thus, this thermal midpoint is unreliable. If no more data is included to calculate this apparent thermal midpoint, please rephrase/soften this result, even when the conclusion will be the same as that reported (lines 328-329) We appreciate this thoughtful comment. We agree and the statement has been rephrased.Line 337-338:Dictyostelium DyPA has comparable thermal stability to bacterial DyPs and its maximum activity is in the temperature range of 20-40 ˚C. The authors propose a dimeric lignin adduct (specie 2, Fig 5A), but the results do not totally support this specie (theoretical mass of "2"= 638.2 gr/mol; ESI-MS showed a m/z of 661.46 (Figure 5C). Whereas I mostly agree that a dimeric species could appear, since the results do not totally agree with this hypothesis, I suggest moving this part to the SI, indicating that a specie similar to “2” could be the product of the reaction, based on the NMR and ESI-MS results. The molecular weight of product (2) is 638.2 g/mol. Our ESI-MS data in figure 5C  show the expected m/z value 661.2 of the sodium adduct formed following ionization of (2) as is stated in the figure legend. To make this point clear, apart from the figure legend, we have now included this information also in the main text.Line 380-383:Consistent with the concept that radical recombination leads to the formation of a higher molecular weight species, further analysis of the second peak by ESI-MS showed a m/z of 661.2 (Figure 5C), which corresponds to the mass of the sodium adduct of the predicted product (2).  I have no access to the PDB structures, nor to the mtz-files reported in the paper. Thus, and without being able to check the binding-site fitting conducted by the authors, there are some concerns on the binding of VA to the surface/interace of DyPA. Firstly, the authors suggest the binding of VA to “pockets” 4 and 6, and based on previous works on different peroxidases (see below), they hypothesize on a plausible LRET mechanism. I find it necessary to show/compare the distances between the ligand-binding sites in DyPA and other peroxidases for which LRET mechanisms were previously proposed, since the distance to the heme group is larger than 20A in both cases; this distance is quite lower in some of the enzymes shown in Figure 8.The omit maps for the ligands are now included in the supplementary figures of the revised manuscript.  We also compared our binding sites with the other peroxidases  (LiP and VP) describing possible LRET. I partly agree with the proposed LRET hypothesis: it has been described elsewhere (a similar scenario was proposed for Auricularia auricula-judaeperoxidase, which might help the authors to further support their hypothesis; refs 43, 56, 58). Furthermore, the EPR support the formation of radical Trp and perhaps, Tyr residues. However, to ascertain the role of these residues, site-directed mutagenesis should be conducted. Since I do not find strictly necessary to include further experiments, as the LRET hypothesis is widely described elsewhere, I think that the authors could soften the statements on the plausible LRET mechanism, using further references where similar/counterpart Trp/Tyr residues have been proved to form radical species, or even theoretical studies on this mechanism (e.g., Acebes et al., 2017. J Phys Chem B. 121(16):3946-3954; Ruiz-Dueñas et al., 2009 J Exp Bot. 60(2):441-52; Romero et al., 2019. Comput Struct Biotechnol J. 17:1066-1074)We are thankful for this insight. Now, we have softened and described our results as a possible LRET pathway and also discussed our results with the published work. Line 526: please, specify a reference where the presence of a tyrosine residue for LRET mechanism has been proposed/shown (such as PDB 4UZI, ref 56).Reference included (Line 626).In the case of AauDyPI, a surface tyrosine and tryptophan-based radical center were reported [63]. Other minor concerns. - As I mentioned previously, I had no access to the PDB nor mtz files. Thus, it is difficult for me to evaluate the binding of the different ligands, and specifically for VA, which is the ligand providing the novelties of this work. If obtaining the different PDBs and structures factors is not possible for additional review, please include at least an OMIT map for the three ligands, representing it to 1.0-1.5 sigma, at least for review purpose.The omit maps for the ligands are now included in the supplementary figures of the revised manuscript. Figure S9B. - The authors could shorten the POCASA results (lines 496-514), and/or move them partially to the SI; this software supports the appearance of VA on the proposed pockets, but the experimental results are the key to support the LRET hypothesis. In this sense, I would also recommend to firstly introduce the appearance of the VA molecules based on the experimental X-Ray results, and later on, supporting substrate binding with this tool.POCASA analysis are descried briefly. Figure 7A shows the 6 predicted binding pockets per interface and panels B, C and D of Figure 7 are describing the solvent channels, leading to the heme center.- Change Dictosyleum to italics through the whole manuscript.Changed.- Please, specify whether the crystals obtained by co-crystallization with KCN and VA were obtained in the same condition as the free form.Changed.Line 812-814:CN--complexed/O2:VA-complexed crystals were prepared by adding 5 mM KCN/ 50 mM VA to the protein (10 mg/mL). They were grown in 2.4 M sodium malonate pH 7.0 in a hanging drop setup at 20°C. -Since formation of Compound I has been recently studied in detail for Streptomyces peroxidase, the authors should comment on this paper and the relationship with the results presented in their work (Lučić et al., 2020 Angew Chem Int Ed Engl. 59(48):21656-21662.)We are grateful for the reviewer's suggestion, we re-analyzed our DyPA structures in greater detail and it makes us realized that the dumble-shaped electron density belongs to an elongated O2 molecule. We compared our observation with the published result.  Now a section is added to the revised manuscript.Line 508-520Using serial femtosecond X-ray crystallography, Lucic et al., have determined DtpB structures in resting (FeIII) as well as in compound I state (FeIV=O and a porphyrin cation radical). Moreover using mutagenesis experiments they went on to show a catalytic role for the distal arginine residue in the formation of compound I [23]. However, compared to the DtpB resting state structure, FeIII is hexa-coordinated in our Dictyostelium DyPA:O2 complex structure and has an elongated O2 molecule at the distal face of heme. In the case of Dictyostelium DyPA, both the aspartate and arginine residue on the distal face are in an optimal position (Figure 6B-C) to take on a catalytic role during compound I formation upon H2O2 addition. Since discrepancies still exist regarding the mechanistic roles of the distal aspartate and arginine residues during compound I formation, further mutagenesis and biochemical experiments are required to assign to theses residues a definite role in the formation of Dictyostelium DyPA compound I. 

Based on the comments of Reviewer 2, we have made the following changes:
In the DyPA:KCN crystal structure, isn't the N atom of the CN molecule supposed to be bonded to the Fe of heme? The authors have to provide an electron density map for CN bonds and comment on the orientation of the CN molecule. Meanwhile, what is the B-factor or occuancy for heme or CN? The most negative charge, as well as the highest occupied molecular orbital of the CN(-) anion, are localized on the carbon atom, which interacts with the positively charged metal cations in the cyanide complexes, e.g., in KCN (K(+)-C(-)≡N). In the cyanide complexes of heme-containing enzymes, the iron atom directly interacts with carbon via the highest-occupied molecular orbital of CN(-), which is mostly carbon-localized. A similar mode of cyanide binding to the heme as observed for the DyPA:KCN complex structure was also observed in the CN-complexes of other heme-containing enzymes, for example, in the crystal structures of cyanide complexes of P450cam and Nitric Oxide Synthase by Fedorov R., Ghosh D., and Schlichting I.,  (2003) Archives of Biochemistry and Biophysics 409: 25–31. Our electron density clearly shows that the N atom is not binding to the Fe of heme. Our observation is further supported by the other DyP:CN- structures For example BadDyP:CN- complex (PDB: 3MM2).  I suggest that the authors add a conclusion and put together the highlights of this study's findings here. This would be good to remind readers of the important results of the manuscript. We added a conclusion paragraph as suggested by the referee. ConclusionsWe describe the comprehensive biochemical and structural characterization of a cytosolic dye decolorizing peroxidase from Dictyostelium discoideum. Dictyostelium DyPA is a dimer, with each monomer exhibiting a two-domain, α/β ferredoxin-like fold. The enzyme shows greater structural similarity to the “primitive” class P(B) DyP superfamily members produced by bacteria than to the “advanced” fungal DyPs of class V(C,D). UV-Vis and EPR spectroscopy identified the presence of a high-spin iron-containing heme that forms a protein-based radical upon H2O2 addition. Dictyostelium DyPA uses both Trp as well as a Tyr radical chemistry in the catalytic processing of its substrates. Lignin oxidation, dye decolorization, and general peroxidase activity were observed for Dictyostelium DyPA. The crystal structures of Dictyostelium DyPA in complex with either O2 or CN- show that Asp149 is in an optimal position to accept a proton from H2O2 during the formation of compound I. Moreover, we report a DyP structure with the lignin model compound veratryl alcohol and delineate a plausible LRET pathway from the substrate binding site to the heme center, which can now be validated by combining mutagenic and time-resolved spectroscopic studies.

Are the two molecules in the asymmetric unit similar in length of interaction at Heme or other major active sites? The author describes any similarity or differences between these two structures. The overall structures are quite similar for the two molecules in the asymmetric unit. Cα-rmsd between both the chains is now included in the main text. In addition, the structural alignment of both chains has been included in the new supplementary figures. Line 490-492:The asymmetric unit contains two copies of the complex. Both copies share nearly the same overall architecture, as indicated by RMSD of 0.128 Å (Figure S7A-C). Line 585-587The asymmetric unit contains two copies of the complex, having the same overall architecture (Cα-RMSD of 0.105 Å) (Figure S9).Minorline 44-50: Authors should add references.
References have been added. line 117-118:"Sequence alignment was performed using Clustal Omega." Author should move this sentence to the "Materials and methods" section.
Moved. line 125-126: "Sedimentation coefficient distributions were converted to 12 mm path length for better comparison." Author should move this sentence to the "Materials and methods" section.
Moved. line 131-132: "The buffer used contains 50 mM Tris-HCl pH 8.0 and 150 mM NaCl." Author should move this sentence to the "Materials and methods" section.
Moved. Go to line 133-134: "KCN. The buffer used contains 50 mM Tris-HCl pH 8.0 and 150 mM NaCl." Author should move this sentence to the "Materials and methods" section.
Moved. line 235-236: "The simulation parameters are given in Table II." Author should move this sentence to the text, not figure caption.
Moved. line 387: "The absorbance of substrate and product was detected at 254 nm." Author should move this sentence to the "Materials and methods" section.
Moved. Figure 6E: It is difficult to distinguish each lines. Author should increase the size of Figure 6E to see easily (maybe it is appropriate to move to supplementary figure).The size of the Figure 6E has been adjusted.

line 471: what is "(4)"

Removed “(4)”Figure S1: Author should add the sequence accession number and information about the strain.Thanks for the suggestion. The sequence accession number and information about the strains have been added to the figure legend.Figure S1: Sequence alignment of Dictyostelium DyPA with different bacterial and fungal DyPs. The conserved GXXDG motif is shown in a black box. Dictyostelium discoideum DyPA (UniProt ID: Q556V8), Escherichia coli (strain K12) YfeX (UniProt ID: P76536), Shewanella oneidensis (strain MR-1) TyrA (UniProt ID: Q8EIU4), Bacteroides thetaiotaomicron VPI-5482 BtDyP (UniProt ID: Q8A8E8), Rhodococcus jostii RHA1 RjDyPB (UniProt ID: Q0SE24), Auricularia auricula-judae AauDyPI (UniProt ID: I2DBY1), Bjerkandera adusta BadDyP (UniProt ID: Q8WZK8) Figure S4: It is suggested to concise the figure caption. (E.g., Structural alignment of Dictyostelium DyPA (gray) with (A) Escherichia coli O157 EfeB (blue; class A), (B) Escherichia coli O157 YfeX (blue; class B), (C) Klebsiella pneumoniae KpDyP (blue; class B) ....)Text has been modified as per the reviewer´s suggestion.Figure S4. Structural comparison of Dictyostelium DyPA with related bacterial and fungal dye decolorizing peroxidases. Structural alignment of Dictyostelium DyPA (gray) with (A) Escherichia coli O157 EfeB (blue; class A). (B) Escherichia coli O157 YfeX (blue; class B). (C) Klebsiella pneumoniae KpDyP (blue; class B). (D) Vibrio cholerae VcDyP (blue; class B). (E) Bacteroides thetaiotaomicron VPI-5482 BtDyP (blue; class B). (F) Rhodococcus jostii RHA1 RjDypB (blue; class B). (G) Shewanella oneidensis TyrA (blue; class B). (H) Amycolatopsis sp. 75iv2 DyP2 (blue; class C). (I) Bjerkandera adusta BadDyP (blue; class D). (J) Auricularia auricula-judae AauDyPI (blue; class D).Figure S5 and S6: If the two molecules in the asymmetric unit have different configurations, it should be added them.The overall structures are quite similar for the two molecules in the asymmetric unit. The structural alignment of both molecules in the asymmetric unit has been added to supplementary figures.Figure S9: Structure of Dictyostelium DyPA:O2:VA complex. (A) Structural alignment of both copies of the Dictyostelium DyPA:O2:VA complex in the asymmetric unit reveals that both copies are nearly identical (Cα-RMSD of 0.105 Å).Figure S8:  ….. The water molecules are shown as gray spheres. Key non-covalent bonds are represented as gray dashed lines and the bond length cutoff is 3.5 Å.  
+ Respond to this comment

We encourage comments and feedback from a broad range of readers. See criteria for comments and our diversity statement.

Leave a public comment
Send a private comment to the author(s)
Views 0
Downloads 0
Comments 1
Metrics 0

Notify me about updates to this article or when a peer-reviewed version is published.
We use cookies on our website to ensure you get the best experience.
Read more about our cookies here.