Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Dec 10.
Published in final edited form as: J Am Chem Soc. 2025 Oct 6;147(41):37688–37700. doi: 10.1021/jacs.5c12710

A post-translational histidine-histidine cross-link enhances enzymatic oxygen reduction activity with greater pH adaptability

Yiwei Liu 1,3, Avery C Vilbert 2, Barshali Ghosh 1, Robert P Young 2, Eric D Merkley 2, Arnab Mukherjee 3, Lisa Phan 1, Casey Van Stappen 1, Ambika Baghi-Damodaran 3, Kyle D Miner 3, Joshua N Adkins 2, John R Cort 2, Yi Lu 1,2,3
PMCID: PMC12687290  NIHMSID: NIHMS2118589  PMID: 41047894

Abstract

Cross-linked protein residues exist as enzyme cofactors to enable or enhance catalytic activities. Despite their importance in nature, the chemical identity of the cross-links is limited to certain amino acid combinations, whose function and the formation mechanism remain insufficiently understood due to the difficulty in isolating native enzymes without the cross-links. Herein, we report the formation and characterization of both His-Tyr and His-His cross-links under oxidative enzymatic turnover conditions in L29H/F33Y/F43H Mb, a structural and functional model of heme-copper oxidase (HCO). The connectivity of the cross-link was characterized as Nε2(His29)-Cδ2(His43) by mass spectrometry (LC-MS/MS) and nuclear magnetic resonance (NMR). Interestingly, formation of the cross-link significantly enhances the oxygen reduction activity of the enzyme at neutral or basic pH with higher product specificity. X-ray crystallography has identified a novel Tyr-His cross-link through Tyr-O-His linkage. Our mechanistic studies indicate the involvement of high-valent heme-iron and the neighboring tyrosine in an oxidative self-processing pathway to generate the cross-link. This work serves as a new example while providing insights into the enzyme cross-link formation, allowing the design of artificial biocatalysts containing these novel cross-links with higher activity and pH adaptability.

Graphical Abstract

graphic file with name nihms-2118589-f0001.jpg

INTRODUCTION

Cross-links between protein residues, formed as post-translational modifications (PTMs), play critical roles in modulating the functional properties of proteins (Figure 1a).1–3 One classic example is the Cys-Tyr cross-link. In the copper enzyme galactose oxidase, the cross-link modulates reduction potential while maintaining structural stability of the active site.4–6 A similar cross-link exists in thiol dioxygenases, an important class of sulfur-metabolizing enzymes. The biogenesis of the cross-link was probed using a delicate halide cleavage reaction and was found to be an iron-mediated oxidation process,7–9 yet the catalytic role of the cross-link remains under debate.10 With continued discovery of cross-links in redox enzymes,11–16 our understanding of cross-links in enzyme catalysis has been constrained by three primary challenges. First, while many amino acids are theoretically capable of forming cross-links, only a few types of non-disulfide cross-links are known to date in a limited number of enzymes. Second, although many cross-links are known to be essential for enzyme activities, their exact roles have not been fully understood. Third, the formation mechanisms of these cross-links, which must be precisely controlled oxidation processes, remain to be fully illustrated. These challenges are further exacerbated by the structural complexity of many of the native enzymes and the difficulty in obtaining them in pre-cross-linked states, which often makes structural, functional, and mechanistic study of enzyme cross-links impractical.

Figure 1.

Figure 1.

(a) Scheme of cross-link formation and its role in regulating enzyme activity. (b) Overall scaffold and the active site for oxygen reduction in heme-copper oxidase (PDB 2OCC).4 (c) Scaffold overlay of swMb (grey, PDB 5YCE)18 and the designed F33Y-CuBMb (cyan, PDB 9DTI).19

A notable example is the highly conserved His-Tyr cross-link found in the active site of heme-copper oxidase (HCO, Figure 1b).4 HCO is a terminal oxidase catalyzing the four-electron reduction of oxygen to water, providing energy for ATP synthesis that sustains most biological processes.5–7 While this His-Tyr cross-link is proposed to be essential for HCO’s oxidase activity,8,9 its formation mechanism and the specific functional roles remained elusive since its discovery over 30 years ago. A major barrier is the inability to isolate HCO prior to the cross-link formation, which precludes direct comparisons with its cross-linked form. Synthetic model compounds have been used to probe the roles of the cross-link in HCO, which indicated higher pKa and reduction potential,10–17 yet it remained challenging to investigate the cross-link formation process within a protein environment under enzymatically relevant conditions. Additionally, the limited knowledge of existing cross-links highlights a considerable potential to discover and utilize new-to-nature cross-links beyond His-Tyr to modulate enzyme functions. Such advancement would significantly broaden our understanding of the chemical diversity and catalytic capability of cross-links, but is largely untenable in complex native systems like HCO.

To address these issues, we turned our attention to the biosynthetic model of HCO using sperm whale myoglobin (swMb), which is an O2 transport protein in nature but has been engineered for different reactivities.20–29 Through L29H/F33Y/F43H mutations (called F33Y-CuBMb, Figure 1c),30 the designed F33Y-CuBMb displayed oxygen reduction reaction (ORR) activity, albeit having a simpler scaffold and higher solubility (~ 17 kDa) compared to the native, membrane-bound HCO (~ 200 kDa). These features not only simplify the expression and characterization of the protein but also dissect the core components essential for catalysis and cross-link formation from other parts of the native system. Moreover, unlike HCO and many other proteins with cross-linked motifs, the engineered Mb can be isolated without any protein residues being cross-linked,31 which allows us to explore cross-link formation and compare the ORR activity before and after the cross-link formation.

Building on this design, we herein present the discovery of a novel His-His cross-link and a Tyr-His cross-link in F33Y-CuBMb, together with the characterization, formation mechanism and resulting impact on ORR activity. Treatment of Fe(III)-heme-F33Y-CuBMb with H2O2 in a “peroxide shunt” pathway in the ORR cycle (Scheme 1)32,33 produced a His29-His43 cross-link as observed by mass spectrometry. Nuclear magnetic resonance (NMR) of the purified cross-linked peptide showed a linkage between the Cδ2 of His43 and Nε2 of His29. Interestingly, in crystallo reaction with H2O2 resulted in an OH(Tyr33)-Cδ2(His43) species in the secondary coordination sphere (SCS) of the heme center, which is a possible precursor to the His29-His43 cross-link. Mechanistic study also revealed a compound II ferryl species and the Tyr33 radical leading to cross-link formation. Notably, upon formation of the His29-His43 cross-link, F33Y-CuBMb displayed nearly tripled four-electron ORR activity at pH 7.0 and partially retained up to pH 9.5, whereas the pre-cross-linked F33Y-CuBMb became inactivated above pH 8.0. Based on these results, we propose a self-processing mechanism for cross-link formation, including Tyr-His and His-His. Through the discovery of the enzymatically unprecedented His-His cross-link in an artificial oxidase model, this study contributes to understanding the formation mechanism of active site cross-links in natural oxidoreductases, and presents a facile but unexplored approach to fine-tuning enzymatic activity.

Scheme 1.

Scheme 1.

General mechanism of heme-catalyzed oxygen reduction reaction (ORR). The reaction between Fe(III) and H2O2 (the “peroxide shunt”, middle dashed line) bypasses the first half ORR cycle, creating a two-electron oxidant available for the cross-linking reaction.32,33

RESULT AND DISCUSSION

Formation and MS Identification of a His-His Cross-link.

To investigate whether F33Y-CuBMb can form any cross-link, we expressed and purified Fe(III)-F33Y-CuBMb that contains Fe(III)-heme but no Cu ion in the CuB site (referred to as “F33Y-CuBMb” for simplicity). This form has been used in oxidase catalysis with high ORR activity.30,34 We then reacted F33Y-CuBMb with different equivalents of H2O2 at varying reaction times. As the cross-link formation would lead to a small overall mass change of - 2 Da due to the loss of two hydrogen atoms, which might not be accurately determined by MS of the whole protein (~ 17300 Da), we performed trypsin digestion on the reacted protein (Figure 2a). The selective digestion at arginine or lysine residues by trypsin resulted in a predicted list of peptides (Table S1). From the matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) of the digested sample, a peak with moderate intensity was observed at m/z = 2962.71 (Figure 2b), which did not correspond to the mass of any predicted peptide but matched that of two peptides combined minus 2 Da. The two peptides are VEADVAGHGQDIHIR (17–31, peptide A) and SHPETLEKHDR (35–45, peptide B), with molecular masses of 1615.81 and 1347.65 Da, respectively. Considering the dehydrogenation and a proton needed to ionize such species, the expected mass of the combined peptides (peptide C) is 2962.45 Da, which fell within a reasonable error range of the 2962.71 Da species observed on MALDI-TOF MS. The isotope distribution also matched the calculation (Figure 2b inset) based on the chemical formula. These results suggested that the peptides A and B were cross-linked during the reaction to form a higher-mass species.

Figure 2.

Figure 2.

MS characterization of the cross-link. (a) Scheme of cross-link formation and trypsin digestion to produce the cross-linked peptide. (b) The MALDI-TOF MS spectrum of the digested sample. The inset shows the isotope pattern of the 2962.71 species (dark cyan) and the expected pattern calculated from C124H197N42O43 (red). (c) The +2~+6 charged states of the cross-linked peptide eluted at 33 min on the LC-MS chromatogram (inset). (d) An example MS/MS spectrum from the m/z = 741.3693 (+4) precursor. Two of the cross-linked fragments are highlighted.

Interestingly, peptide A contains His29 while peptide B contains His43, the two His residues in the active site, which prompted us to investigate the connectivity of the cross-link by liquid chromatography-tandem mass spectrometry (LC-MS/MS). The +2~+6 charged states were observed at ~ 33 min in the MS1 spectra from the untargeted mode of the LC-MS chromatogram (Figure 2c), with the deconvoluted mass matching that from MALDI-TOF MS and calculation with a smaller error of ~ 2 ppm. To examine peptide fragmentation and localize the cross-link, MS2 spectra were extracted and analyzed by pLINK2,35 a software to identify cross-linked species from LC-MS/MS profiles, set to search for His-His cross-links. The cross-link was located by pLink2 between His29 and His43 (Figure S1). Figure 2d shows the b and y-type species generated by higher-energy collisional dissociation (HCD). Importantly, we identified fragment ions resulting from fragmented peptide A combined with full-length peptide B, such as A(y32+) at 885.9559, and fragmented peptide B combined with full-length peptide A, such as B(y72+) at 1256.6310. These higher-mass fragments could only be generated when two peptides are cross-linked. In addition, the fragments always contained His29 and His43 but not His36 or His24 in their sequences, supporting the cross-link between His29 and His43 and excluding possible cross-links between other histidine residues in these two peptides.

NMR Characterization of the His-His Cross-link.

To support the His29-His43 cross-link identified from LC-MS/MS, we performed NMR analyses of the cross-linked peptide. Milligram scale of peptide C was prepared using an optimized protocol and separated by FPLC and then HPLC with > 95% purity (Figure S2, S3, see Experimental Section) for the NMR experiment. Analysis of the 1H-TOCSY indicated that all of the expected amino acid residues comprising the proposed peptide C, namely the combination of Val17-Arg31 (peptide A) and Ser35-Arg45 (peptide B), were present. The NOESY experiment confirmed their expected sequential positions within the respective A and B chains of the excised fragment (see Figures S4–S6 in the SI for assigned 15N-HSQC and 1H-TOCSY, and 1H-NOESY spectra). These results validated the findings from the MS studies of the cross-link between peptides A and B. The proposed N(His29)-C(His43) cross-link was further supported by the fact that the major cross-peptide interactions observed in the 1H-NOESY spectrum were between the His29 residue of peptide A and His43 of peptide B. In the NOESY spectra, both the His29 Hε1 and Hδ2 showed roughly equally intense NOE crosspeaks to the His43 Hβ2/Hβ3protons (Figure 3a), and both also had correlations to the Hα protons on His43. The Hδ2 of His29 had a correlation to HN of His43, and Hε1 of His29 showed one to the Hε1’ of His4336. Minor interactions were observed between His29 and the Hα of Lys42, due to its immediate adjacency to His43 on the peptide B. Combining the information from the NOESY spectra, an analysis on the chemical structure of His29 and His43 indicated that only Nε2(His29)-Cε1(His43) or Nε2(His29)-Cδ2(His43) is consistent with the NMR data.

Figure 3.

Figure 3.

NMR analysis of the cross-link. (a) 1H-NOESY spectra focusing on the imidazole hydrogen atoms of His29. The arrows in blue and orange colors correspond to the NOE interactions highlighted in the NOESY spectra. (b) Overlay of 1H-13C-HSQC (red) and HMBC (grey) spectra focusing on the imidazole ring resonances. Note that His43 shows only one C-H and two quaternary carbon atoms, indicating a missing hydrogen atom on the imidazole ring. (c) 1H-15N-HMBC spectrum focusing on the imidazole rings. The signals from His29 and His43 are highlighted by red frames.

Furthermore, the 1H-13C-HSQC spectrum revealed only one imidazole ring C-H correlation present on the His43 sidechain, and its chemical shift value of 137.8 ppm led to its tentative assignment as Cε1, as it fell within the typical range of that carbon regardless of tautomers.37 The 1H-13C-HMBC spectrum indicated another quaternary carbon, in addition to the Cγ, correlated to the Hε1 proton. Its chemical shift value of 121.5 ppm, close to the typically protonated Cδ2 value, could also be similar if the proton was replaced by a nitrogen. Meanwhile, the non-cross-linked His24 and His36 are overlapped slightly at the H/Cε1 position and have values typical for a fully protonated imidazole ring, whereas His29 is perturbed to higher chemical shift values for both carbon and hydrogen resonances at both positions, indicating modifications on His29. These observations lend further support that the cross-link is between Nε2(His29)-Cδ2(His43) (red circle, Figure 3b). On the other hand, protons directly bound to either Nδ1 or Nε2 were not observed in the 1D 1H or 2D 15N-HSQC, indicating they were in fast exchange with water. The imidazole ring nitrogen atoms were observed through correlations to the Hδ2 and Hε1 protons in the 1H-15N HMBC spectrum. His24 and His36 ring nitrogen chemical shifts were consistent with the fully protonated tautomer. The Hε1 of His43 correlated to 15N signals consistent with the Nε2H tautomer (171.6 and 240.3 ppm).37 For the His29 imidazole ring, only the Hε1 showed correlations to the ring nitrogen atoms, which had 15N chemical shifts of 176.1 and 193.1 ppm. The former falls in the range of a protonated ring nitrogen, whereas the latter is shifted downfield by over 10 ppm. Based on the evidence that only the Nε2 nitrogen could participate in the cross-link bond to His43 (red circle, Figure 3c), the 193.1 ppm value was assigned as Nε2, and 176.1 ppm was assigned as a protonated Nδ1. Taken together, these results indicated that the cross-link was formed between Nε2(His29)-Cδ2(His43).

Crystallographic Observation of Tyr-His Cross-link:

To better understand the structure of the cross-link, we crystallized F33Y-CuBMb and performed in crystallo reaction with H2O2. The reacted crystals gained a dark red color in contrast to the brown color of the non-reacted crystals (Figure S7), from which we obtained diffraction data and resolved the crystal structure at 1.8 Å resolution. Surprisingly, instead of a His-His cross-link, we observed the phenol oxygen of Tyr33 becoming spatially close enough to the imidazole side chain of His43 (~1.4 Å) to warrant a covalent bond (Figure 4a). Removal of the OH(Tyr33) atom left a peak in the difference map between the phenyl side chain and His43 (Figure 4c), supporting the placement of the phenol oxygen atom between Tyr33 and His43 and the formation of an OH(Tyr33)-Cδ2(His43) cross-link. The connectivity is different from the C3(Tyr)-Nε(His) cross-link in heme-copper oxidases or the Cβ(Tyr)-Nδ(His) cross-link in E coli catalase HPII,4,38 possibly due to the shorter distance and more favored conformations of the cross-linked atoms in the F33Y-CuBMb pocket (Figure 4b). Comparison between the structures before and after H2O2 treatment revealed a change of one active site water molecule (W3). In the unreacted structure, this water molecule formed a hydrogen bond to His43, but was excluded from the active site in the Tyr33-His43 cross-linked structure.

Figure 4.

Figure 4.

(a) Crystal structure overlay of the Tyr33(OH)-His43(Cδ2) cross-linked intermediate (coral, PDB: 9DTH) with the unreacted crystal (cyan, PDB 9DTI). The three water molecules in the non-cross-linked F33Y-CuBMb are set transparent. (b) The crystal structure of the H2O2-soaked F33Y-CuBMb focusing on the heme pocket. The hydrogen bond interactions are presented with dashed lines with the lengths labeled. Note the short 1.4 Å distance between Tyr33(OH) and His43(Cδ2). (c) Comparison of the Tyr33(OH)-His43(Cδ2) linked and the unreacted F33Y-CuBMb structures with electron density maps. The |2Fo-Fc| map is shown in blue at 1.0 RMSD, and the Fo-Fc omit map is in green at 3.0 RMSD.

Upon formation of the Tyr33-His43 cross-link, the two connected atoms in the His29-His43 cross-link (Cδ(His43) and Nε(His29)) become spatially closer compared to the unreacted state, although not within the range of a covalent bond. To explore whether the His29-His43 cross-link could be formed under crystallographic conditions, we extended the soaking time (for example, beyond 3 hours at room temperature) with or without greater amounts of H2O2 (for example, 10 mM H2O2). All attempts resulted in deformed crystals (Figure S7), from which no diffraction data could be obtained.

Mechanistic Considerations of His-His and Tyr-His Cross-links:

Given the fact that the Tyr33-His43 and His29-His43 cross-links formed in crystals and solution, respectively, we continued to explore whether these two cross-links are related. By dissolving one of the Tyr33-His43 cross-linked crystals and digesting with trypsin, we observed a small peak at m/z = 1768, matching the mass of a Tyr33-His43 cross-linked peptide (LYK-SHPETLEKHDR) (Figure S8). Interestingly, the His29-His43 cross-linked peptide was observed at m/z = 2963 from the same dissolved crystal, matching the MS result from the solution phase H2O2 reaction. The His29-His43 peak displayed 6-fold greater intensity, even though it is completely absent in the crystal structure, while the Tyr33-His43 cross-link was the sole cross-linked species with nearly 100% occupancy.

To probe the role of Tyr33, we replaced it with Phe so that it would not be able to form the Tyr33-His43 cross-link.39 An LC-MS analysis of the reaction of this mutant with H2O2 revealed no His29-His43 cross-link formation (Figure S9). This result, combined with the observation of both Tyr33-His43 and His29-His43 cross-links in the dissolved crystals, suggested both necessity and sufficiency of the Tyr33-His43 cross-link toward the formation of the His29-His43 cross-link, in which the Tyr33-His43 cross-link acted as a precursor. This hypothesis of cross-link conversion could be rationalized by a more stable N-C type His29-His43 cross-link than the O-C type Tyr33-His43 cross-link as per the bond-dissociation energy prediction (by approximately 15 kcal/mol).40,41 Although the exact cause remains to be investigated, one possible explanation of such difference is that the Tyr33(OH)-His43 cross-link was captured as an intermediate in the crystal structure favored by its spatial adjacency, whereas the higher protein flexibility in the solution phase led to the formation of the thermodynamically favored His29-His43 cross-link as the final product observed in MS and NMR. This hypothesis corroborates our mutagenesis results that indicate the involvement of Tyr33 in forming the His-His cross-link.

Regardless of Tyr-His or His-His cross-link, their formation must involve a two-electron oxidation process. To investigate the earlier intermediates in the F33Y-CuBMb-H2O2 reaction, we recorded the stopped-flow UV-Vis spectra for the first 140 seconds. As shown in Figure 5a, a 424 nm peak was observed from the difference spectra, with an isosbestic point at 416 nm correlated with the 405 nm heme Soret peak. The 424 nm peak and the 17 nm red shift from the resting state Soret band were similar to the reported spectra of the Fe(IV)=O heme species in HCO (~430 nm)42,43 and other heme enzymes such as cytochrome c peroxidase (420 nm)44,45 and ascorbate peroxidase (418 nm).46 The observed ferryl compound II most likely originated from a higher valence transient Fe(IV)=O(heme·+) compound I species, which is a typical “peroxide shunt” process in the ferric heme-H2O2 reaction (Scheme 1).47–49 The color of the solution was also consistent with that of the H2O2-reacted crystal, indicating a similar chemical process in the soaking experiment.

Figure 5.

Figure 5.

(a) Stopped-flow UV-Vis spectroscopy of Fe(III)-F33Y-CuBMb + H2O2. (b) Q-band EPR spectrum focusing on the g = 2 region.

Meanwhile, when F33Y-CuBMb was reacted with H2O2, a g = 2 species was observed on the X-band (Figure S10) and Q-band CW EPR (Figure 5b). Being separated from the Fe spins, the Q-band peak at 1210 mT could be simulated with an S = 1/2 spin matching that of a tyrosyl radical in a previous report on F33Y-CuBMb (Table S3). The simultaneous presence of the compound II and the tyrosyl radical could be explained by an internal redox reaction in which the reactive compound I was reduced to compound II by the tyrosyl radical. Interestingly, the relative intensity of the radical species compared to the Fe(III) spin was significantly lower than what was observed in the previous report,50 which used a lower equivalence of H2O2 and shorter reaction time. We postulate that the decrease in the radical signal was due to its conversion toward the cross-link generation.

Combining the crystallography spectroscopy results, we propose that the formation of the His29-His43 cross-link in F33Y-CuBMb involves a Fe(IV) ferryl species and the crucial Tyr33 (Scheme 2). A His/Tyr substitution was proposed as the basis of cross-link conversion, as we observed MS signals of His43-imidazole or His43-histidine cross-links when excess imidazole and histidine were added as nucleophiles competing with His29 (Figure S11). The mechanism resembles one of the mechanisms proposed for the His-Tyr cross-link formation in HCO (Figure S12a),51 while being different from another proposed mechanism using Fe(III) and Cu(II) as the electron acceptors (Figure S12b). A recent report suggests a Fe(III)-porphyrin di-cation oxidizing a histidine into a His-porphyrin adduct, followed by the nucleophilic attack by tyrosine (Figure S12c).52 While we did not observe the porphyrin di-cation due to its transient nature similar to the isoelectronic compound I or His-adduct species, we cannot rule out a partial contribution from such species to the formation of the His-His cross-link in the initial steps.

Scheme 2.

Scheme 2.

Proposed mechanism of His29(Nε2)-His43(Cδ2) cross-link formation.

Oxidase Activity in the Absence and Presence of the Cross-link.

One of the most intriguing aspects of the cross-link is its ability to fine-tune the function of enzymes. To examine the effect of the His29-His43 cross-link on the oxidase activity, we performed ORR assays of F33Y-CuBMb before and after cross-link formation. (Figure 6a). Due to the minimal change to the charge and size-exclusion properties caused by the cross-link, we were unable to separate the intact, cross-linked protein from the non-cross-linked species. Therefore, the cross-linked protein sample was directly applied in the ORR assay after H2O2 treatment.

Figure 6.

Figure 6.

Comparison of ORR activity between the cross-linked (orange) and non-cross-linked F33Y-CuBMb (grey). (a) O2 concentration curves recorded by a Clark-type electrode in an ORR assay at pH 7.0. (b) The apparent ORR rate monitored by O2 consumption from pH 5.0 to 10.0. (c) The rate of four-electron ORR generating H2O. (d) The ratio of ROS side products. (e) X-ray crystal structure of the active site of the resting state F33Y-CuBMb (PDB 9DTI). W1-W3 denote the three water molecules in the heme pocket. The hydrogen bond interactions are labeled as dashed lines with numbers showing distances in Å. (f) Chemical structures of the cross-linked and non-linked histidine, and their respective predicted pKa.41

The non-cross-linked Mb showed an optimal pH of 6.0 with a pseudo-zero order apparent ORR rate constant of 0.050 s−1, similar to previous reports.16,34,53 Rapid decrease of ORR activity was observed with pH changes as small as 0.5 units, which is consistent with those observed in native HCOs.54,55 The activity dropped further at neutral or alkaline pH, and completely diminished at pH 8.0. In contrast, the cross-linked F33Y-CuBMb remained catalytically active up to pH 9.5, while retaining over 75% of its initial activity at pH 8.0 (Figure 6b). Such activity enhancement became more apparent in the complete four-electron ORR, which is the native reactivity of HCO. The cross-linked F33Y-CuBMb displayed equal or higher water generation rate than the non-cross-linked form at all pH conditions (Figure 6c). In particular, we saw the water generation rate reaching 2.9-fold at pH 7.0 after the cross-linking. In case the modification was incomplete, the observed ORR would be contributed by the cross-linked protein and the non-cross-linked portion, although the activity enhancement is attributable solely to the cross-linked form. Accordingly, the actual enhancement effect of the His-His cross-link would be (1+1.9/yield), which can be higher than the observed 2.9-fold if the yield of the cross-link is not 100%.

Since both Tyr33-His43 and His29-His43 cross-links were observed, we carried out a mutagenesis study to identify the functional cofactor. Specifically, in the H29Q and H29F mutants, which eliminated the possibility of the His29-His43 cross-link while still allowing the Tyr33-His43 cross-link, the activity enhancement effect by H2O2 treatment could no longer be observed (Figure S13). The result demonstrated that both His29 and His43 are required for the higher ORR activity at pH 7 and broader working pH as described earlier. In addition, while it is conceivable that the His-His cross-link could arise from a rearrangement of the Tyr-His cross-link upon protein denaturation and digestion, this evidence obtained without protein denaturation and digestion suggests that the His-His cross-link is the functional cofactor causing the higher activity.

The enzymatic ORR requires protonation of the reactive oxygen species (ROS), such as superoxide or peroxide anions, to proceed with further reduction. Based on the crystal structure of F33Y-CuBMb, such functionality is fulfilled by a local water network as well as the Tyr33 residue.56,53,34 Formation of the cross-link, on the other hand, will likely replace the water molecules such as W3 near His29 and His43, and affect ROS generation. As shown in Figure 6d, the His-His cross-link reduced ROS generation (9–23%) compared to the non-cross-linked form (35–47%), which aligns with a previous report having Glu with a similar configuration.53 To probe such an effect in the cross-linked F33Y-CuBMb, we mutated His43 to Glu (fully deprotonated as hydrogen bond receptor), Gln (hydrogen bond donor), and Phe (not hydrogen bonding). The ORR activity is shown in Figure 7 with a trend Glu43>Phe43>Gln43. Notably, the F33Y-CuBMb without cross-link showed an activity slightly higher than the Gln43 mutant, but lower than the Phe and Glu mutants. After the cross-link, the activity becomes similar to that of Glu43. Meanwhile, the increased acidity of His43 was implied by the chemical shifts in the 1H-15N HMBC (Figure 3c), and a predicted pKa of 4.641, similar to that of glutamate (4.2). We therefore propose increased acidity and potential alteration in proton-donating ability as the mechanistic basis for the ORR-enhancing effect of His29-His43 cross-link.

Figure 7.

Figure 7.

ORR activities of His43 mutants and cross-linked F33Y-CuBMb.

Bioinformatics Search for His-His Cross-link in Nature.

There have been reports on His-His cross-link involving 2-oxo histidine in immunoglobulin G (IgG) antibodies under intended photo irradiation and/or oxidative damage conditions.57–59 These cross-links are likely unrelated or even detrimental to the natural functions of the antibodies due to protein aggregation and denaturation.58,59 His-His cross-links also form by the biochemical processing of peptide-like natural products,60,61 such as the oxygen-bridged Cε-O-Cε link in gristide 834 and the Cδ-Nε link of oxo-His in Histatin-5, but they do not function as enzyme cofactors.62,63 Meanwhile, unlike many of the previous reports, the His-His cross-link discovered in F33Y-CuBMb does not have oxygenation on either histidine. To investigate whether His-His cross-link exists as natural, structurally characterized cofactors with similar connectivity to our findings, we performed a bioinformatic search among over 200000 structures from the Protein Data Bank (PDB). If such a His-His cross-link exists, the distance between Cδ2 and Nε2 atoms should fall into the range of a covalent bond. With two distance filters applied, one for His(Cα)-His(Cα) < 10 Å and the other for Cδ2-Nε2 < 2.0 Å, only 4 PDB entries were found (PDB 5X41, 5CH9, 5J8V, and 3TYW, Table S4). The structures were inspected in detail with the electron density map, from which all candidates showed either possible misassignment indicated by the negative difference map, or low resolution, causing widespread steric clashes and lessening the validity of a His-His cross-link (Figure S14). Therefore, a similar His-His cross-link is rare in nature, if any precedence exists, and this study serves as the first known structurally and functionally characterized His-His cross-link in an enzyme.

CONCLUSION

In this work, we have demonstrated a new-to-nature Cδ2(His43)-Nε2(His29) cross-link in the structural and functional HCO model, F33Y-CuBMb, as characterized by MS and NMR. The cross-link conferred enhanced oxidase activity at neutral or basic pH while simultaneously generating less ROS side product compared to the non-cross-linked enzyme. Further mechanistic studies suggested that the His-His cross-link is formed through a heme-mediated peroxidation reaction via high-valence heme-Fe species, a tyrosyl radical, and a Tyr-O-His cross-link observed by X-ray crystallography and proposed as an intermediate. Although different residues are cross-linked compared to HCO, this study has demonstrated that the more challenging oxidation and cross-link of histidine64 can be feasibly accomplished in an oxidase system, and that such a cross-link enhances ORR activity. Being a new member of the cross-link family as enzyme cofactors, the facile formation of the His-His cross-link by enzyme self-processing provides a new perspective for generating otherwise challenging histidine modifications and tuning of enzymatic activities in both natural and artificial systems.

EXPERIMENTAL SECTION

Cloning, Expression, and Purification of F33Y-CuBMb.

The gene of L29H/F33Y/F43H sperm whale myoglobin (F33Y-CuBMb) was cloned into the pET-17b vector. Transformation of chemically competent BL21-DE3 Escherichia coli (E. coli) cells was performed by heat-shock at 42 °C for 30 s. The transformed cells were incubated in 1 mL of Luria–Bertani (LB) medium at 37 °C. A 100 μL portion of the culture was spread and incubated on a 100 mg/L ampicillin LB-agar plate at 37 °C overnight. A single colony was selected and incubated in 4 mL of LB media with 100 mg/L ampicillin at 37 °C for 6 h until cloudy as the starting culture. A 1 mL amount of the starting culture was inoculated into 1 L of 2XYT media in a 4 L flask. The culture was incubated at 37 °C while shaking at 200 rpm for 14 h, reaching an optical density (O.D.) of 6–8. The protein expression was induced by 0.25 mM isopropyl β-d-1-thiogalactopyranoside (IPTG) for 4 h at 37 °C with continued shaking. The cells were harvested by centrifugation at 11500g for 10 min. For a total of 4 L of culture, the pellet was resuspended in 100 mL of 12 mM Tris-HCl pH 8.0 buffer containing 40 mg of Na2EDTA, 15.4 mg of DTT, and 1 μL of universal nuclease (Thermo Fisher). The cells were lysed by sonication at 70% amplitude for a 12 min pulse time, with 6 s on and 12 s off pulses. The lysed cells were centrifuged at 25000g for 30 min. The F33Y-CuBMb and variants were found mostly in the inclusion body (pellet) without the heme cofactor. A refolding and heme reconstitution protocol was performed as follows.

The inclusion body was mixed with 33.3 g of guanidine hydrochloride and mixed into a slurry. The volume was adjusted to 50 mL by Milli-Q water at 4 °C. The slurry was shaken at 200 rpm at 4 °C for 2 hours. Meanwhile, 14 mg of heme b was dissolved in 10 mL water (add 1 M NaOH and shake until dissolved) and transferred into 750 mL water at 4 °C. The denatured inclusion body slurry was added dropwise (~ 3 mL/min) to the 750 mL heme b solution under vigorous stirring. After being 1-hour of stirring, the solution was centrifuged at 11500 × g for 30 minutes. The supernatant was transferred into 10 kDa cutoff dialysis tubings and dialyzed against 11 L of 5 mM pH 8.0 KH2PO4-K2HPO4 buffer for no less than 4 hours and 2 times. The dilute, dialyzed F33Y-CuBMb solution was centrifuged at 11500 × g for 10 minutes, with the supernatant concentrated by a 10kDa cutoff Amicon stir cell unit. The protein was then purified by size-exclusion chromatography using Superdex-100 beads and 10 mM pH 8.0 KH2PO4-K2HPO4 buffer. Fractions with 407nm/280nm absorbance ratio greater than 4.0 were collected, given an approximate 407 nm absorbance of 1.0, concentrated, dialyzed against 10 × volume of 30 mM pH 8.0 EDTA solution, followed by three dialyses against 10 × volume of 5 mM metal-free KH2PO4-K2HPO4. The protein was concentrated to 1 mM, frozen in liquid nitrogen, and stored at - 80 °C.

Cross-link Formation in F33Y-CuBMb.

The ferric F33Y-CuBMb was exchanged into 10 mM pH 7 potassium phosphate buffer at a final concentration of 200 μM. The reaction was performed by adding 2.0 equivalents of H2O2 to the protein and reacting for 5 minutes at 25 °C. Interestingly, the yield of His-His cross-link increases with increasing equivalents of H2O2 (Figure S15). This result suggests that an excess of H2O2 used in the reaction may drive the C-N bond formation and generate the His-His cross-link. However, in the presence of 3 equivalents of H2O2, the yield decreased, probably because of oxidative damage to the protein.

Trypsin Digestion.

The cross-linked F33Y-CuBMb sample was mixed with urea to reach a final concentration of 8 M (37 mg urea per 50 μL of F33Y-CuBMb sample), with 1 mM DTT added. The sample was incubated at 37 °C for 1 hour, then diluted by 7 × volume of 100 mM NH4HCO3 solution. 1 mM CaCl2 and 1/20 (w/w of F33Y-CuBMb) trypsin (Bovine pancreas, Sigma) were added. The sample was incubated at 37 °C for 4 hours.

Desalting and MALDI-TOF MS.

The digested sample was dialyzed against 100 × volume of milli-Q water using 0.1–0.5 kDa dialysis tubing (SpectrPor) for 12 hours and repeated 3 times, to remove the urea and phosphate buffer. 1 μL of the sample was mixed with 1 μL of 30 mg/mL 2,5-dihydroxybenzoic acid (DHB) matrix dissolved in 30% acetonitrile/H2O, cast on a stainless-steel target plate and dried in air. MALDI-TOF MS was acquired on a Bruker Ultraflextreme or Autoflex mass spectrometer.

LC-MS/MS Characterization.

The sample of the cross-linked peptide was desalted and diluted to approximately 50 μg peptide in 50 μL LC-MS grade water. The F33Y-CuBMb sample was analyzed on a Thermo Orbitrap Fusion hybrid mass spectrometer in conjunction with an Ultimate 3000 RSLCnano ultra-performance liquid chromatography (UPLC) system using a water/acetonitrile gradient. The ion trap MS/MS (untargeted) or Fourier transform MS/MS (FT-MS/MS, targeted) and HCD fragmentation were used for analysis. The initial search of the cross-link was performed by pLINK235. Data extraction and analysis of the LC-MS/MS chromatogram and mass spectra were performed on Mzmine 3.65 The expected fragment list of the cross-linked peptide was generated by the Cross-link Transition Calculator module in Skyline 24.166 to identify relevant peaks in tandem mass spectra within errors of 5 ppm.

Separation of the Cross-linked Peptide.

The digested sample was thawed, concentrated to 1 mL by lyophilization, and syringe-filtered to remove any precipitate. The sample was first injected into an ÄKTAxpress FPLC equipped with a gel filtration column (HiLoad 16/600 Superdex-75 pg, Cytiva) pre-equilibrated with 10 mM pH = 5.0 NH4OAc/HOAc buffer. Size-exclusion purification was performed with an isocratic flow at 0.5 mL/min. Fractions containing the cross-linked peptide were identified by MALDI-TOF MS, combined, and lyophilized to approximately 1 mL. The mixture was then injected into a semi-preparative loop on a Delta 600 HPLC (Waters) and separated on a C-18 reverse-phase column (XBridge Prep C18 5 μm OBD, Waters) using a water/acetonitrile gradient. When separating the His-His cross-linked peptide, the fraction size was reduced to 5 mL from 10 mL around its retention time to offer better separation. The purity was determined by MALDI-TOF MS. By comparing the area of the cross-linked peak (~2963 Da) versus the peptide A (~1617 Da) as part of the cross-link (Figure 2b), we estimate the yield of the His-His cross-link to be ~ 38%. Another estimate based on the LC-MS peaks of the cross-linked and non-cross-linked peptides gives a yield of 18% (Figure S16). After digestion, FPLC, HPLC, and lyophilization, the isolation yield of the cross-linked peptide was approximately 20%. This yield was estimated by the 205 nm absorbance in pure water (no background absorption) and the extinction coefficient (93400 M−1cm−1) calculated from its sequence, which was compared to the original amount of protein used to generate the cross-linked peptide sample.

NMR Characterization.

NMR experiments were performed on Bruker Avance Neo spectrometer operating at a field strength of 18.8 T (800.3 MHz/201.3 MHz/81.1 MHz for 1H/13C/15N) equipped with a 5mm TCI H&F/C/N-D CryoProbe with Z-gradient and operated using Bruker’s TopSpin 4.0.9 software. 1H pulse width calibration, water suppression optimization, and experimental selection and setup for most experiments were performed through the BioTop interface in TopSpin. Experiments were conducted at 278 K. The experiments, along with relevant parameters, are provided in Table S2. Relaxation times (d1) employed ranged between 1 – 2 s. For experiments utilizing non-uniform sampling, default settings in TopSpin were used to generate the sampling points/schedules, and only the sampling percentage parameter was modified (see Table S2 for values). Final post-acquisition processing was performed in TopSpin 4.0.9. All spectra were zero-filled 2–4 times the number of real points. Apodization for phase-sensitive 2D experiments included the use of a cosine squared function in both dimensions (QSINE option with SSB = 2) and non-phase sensitive (HMBC and COSY) spectra used unshifted sinebell squared apodization function (QSINE option with SSB = 0) in both dimensions. Non-uniformly sampled spectra were processed with the following flags/options selected: mdd_mod = cs, mdd_csalg = irls, and mddcexp = true with the remaining options set to false or zero. The 1H chemical shift dimension was referenced to 0.0 ppm using 2,2-Dimethyl-2-silapentane-5-sulfonate-d6 (DSS-d6, sodium salt) as an internal standard (0.25mM) and all other nuclear chemical shift dimensions were referenced indirectly67 from it using ΞDSS = 0.25144953 for 13C and ΞDSS (NH3(l)) = 0.10132912 for 15N. Analysis was performed using MNova 14 software (Mestrelab Research S.L.).

Oxidase (ORR) Activity Assays.

The oxidase activity was measured by the O2 consumption rate monitored by a Clark-type electrode (Hansatech Instruments). The signal response to the O2 concentration was calibrated at 25 °C by an air-saturated solution and an O2-depleted solution by Na2S2O4. 20 μM F33Y-CuBMb before or after cross-link, 20 mM sodium ascorbate, and 2 mM N,N,N′,N′-tetramethyl-1,4-phenylenediamine (TMPD, from a 20× stock with ascorbate) were added to 50 mM NaOAc-HOAc (pH 5.0–5.5), 50 mM KH2PO4-K2HPO4 (pH 6.0–8.0), or 50 mM CHES-Na (pH 8.5–10.0) buffer to make a final volume of 1 mL. Pseudo-zero order reaction rates at the initial stage of the reaction (~ 30 seconds) were acquired by linear fitting.

The rates of four-electron reduction and ROS were measured by a reported method.30 Specifically, In ORR assays, three species may contribute to O2 consumption depending on the fate of oxygen atoms, which are partially reduced superoxide (HO2), peroxide (H2O2), and fully reduced water (H2O). The H2O generation rate was measured by the apparent O2 consumption rate minus the rate of ROS generation. The ROS generation rate was quantified by the difference of O2 consumption rates with or without ROS-scavenging enzymes superoxide dismutase (HO2 → 0.5 H2O2 + 0.5 O2) and catalase (H2O2 → H2O + 0.5 O2). Either enzyme generates 0.5 equivalent of O2 from each equivalent of ROS, which causes reduced apparent O2 consumption rates. In this study, we observed no superoxide generation from ORR, so the ROS was solely contributed by H2O2. Let the apparent ORR rate without catalase be a, and that with catalase be b, then the H2O2 generation rate is 2(a-b), and the rate of H2O is [a - 2(a-b)] = 2b-a.

In Crystallo Reaction with H2O2 and Structure Determination.

Crystals of F33Y-CuBMb were obtained by the hanging-drop method. The quality of the crystals was improved by streak-seeding to increase the resistance to H2O2 treatment. A previous crystal structure has been reported for F33Y-CuBMb (PDB 4FWX)30 yet with different crystallization conditions. We acquired new crystal structures for the unreacted F33Y-CuBMb (PDB: 9DTI) to ensure consistency with the structure after H2O2 treatment. Nevertheless, the new F33Y-CuBMb structure displayed accurate alignment with the reported structure (Figure S17).

For the H2O2-soaking experiment, the droplets containing the protein crystals (approximately 1.1 mM of protein based on the protein concentration before crystallization) were mixed with the same volume of 2.0 molar equivalent of H2O2 regarding the amount of protein crystals in 100 mM pH 7.0 KH2PO4-K2HPO4 buffer. The drop was kept at room temperature for 3 hours or at 4 °C for 16 hours with aluminum foil to block the light. Crystals with reasonable shape were harvested and cryoprotected with 30% glycerol and screened for diffraction. It is worth noting that some crystals with shorter soaking time also showed the Tyr-O-His cross-link (for instance, 4 °C for 4 hours), while longer soaking time or higher H2O2 equivalent resulted in deformed or partially dissolved crystals, indicating oxidative damage or drastic change of the protein conformation in the crystals. Crystal diffraction data was acquired at the beamlines 8.2.1 and 5.0.2 of Advanced Light Source (ALS) at Lawrence Berkeley National Laboratory (LBNL). Diffraction images were processed and reduced with Xia2-DIALS68,69 (or XDS70) and AIMLESS.71 The molecular replacement was performed with the non-cross-linked CuBMb (PDB: 4FWZ) as the search model using Phaser in the Phenix software suite.72,73 The refinement was performed in the Phenix.refine-WinCoot interface.74 The final models were evaluated using Molprobity75 and ProCheck.76 All structures were visualized and presented by the ChimeraX software suite.77 The crystallography information is listed in Table S5.

Stopped-flow Reaction with H2O2.

10 μM F33Y-CuBMb in 50 mM KH2PO4-K2HPO4 pH=7.0 buffer was loaded into a syringe and attached to a stopped-flow UV-Vis spectrometer (SX-20, Applied Photophysics) and cooled down to 4 °C by a water bath. The protein was mixed with 20 μM H2O2 of the same volume. The UV-Vis spectra were recorded through a time course of 200 seconds (or longer for observing the decay of Fe=O). The difference spectra were calculated by subtracting the first UV-Vis plot from the following traces.

EPR of H2O2 Reacted Samples.

250 μL 800 μM F33Y-CuBMb in 50 mM KH2PO4-K2HPO4 pH=7.0 buffer was reacted with 2.0 eq. H2O2 was added from a 100 mM stock with continuous stirring. The reaction was kept at room temperature for 5 minutes. The reacted protein was quickly mixed with 60 mg of glycerol and transferred to a 4 mm quartz EPR tube, flash-frozen in liquid nitrogen. Another F33Y-CuBMb of the same volume and concentration was prepared as the resting state sample. The spectra were obtained on a Bruker EMX Plus spectrometer. The X-band spectra were recorded at 10 K, with frequency of 9.375 GHz, modulation frequency of 100.0 kHz, modulation amplitude of 7.46 G, microwave power of 0.2 mW, conversion time of 20.00 ms, time constant of 20.48 ms. The Q-band spectra were recorded at 40 K, with frequency of 33.905 GHz, modulation frequency of 100.0 kHz, modulation amplitude of 1.00 G, microwave power of 1.45 mW, conversion time of 40.00 ms, time constant of 20.48 ms. The EPR spectra were simulated using the EasySpin package78 in the MATLAB R2020b interface.

Cross-link structure search.

PDB files were sourced from the PDB archive during August 2023 with approximately 204k entries. Structures with spatially close histidine-histidine pairs were queried using an exhaustive search with a custom Python script and the Bio.PDB package.79,80 Two distance filters were applied on every two histidine residues in each PDB entry. The first was a 10.0 Å distance filter between the Cα atoms and the second was a 3.5 Å distance between the Cδ2 and Nε2 atoms. The results of histidine pairs were sorted by the Cδ2-Nε2 distance from smallest to largest and those with distances < 2.0 Å were selected for visual inspection of their structures. Structure analysis and visualization, including the electron density map, were conducted in Wincoot.74

Supplementary Material

Supplementary Material

Supplementary Information includes NMR, EPR, mass spectrometry spectra, microscope pictures, activity assay results, mechanism schemes, and parameter tables for data acquisition. The figures and tables are available in the PDF document.

ACKNOWLEDGMENTS

This work is supported by the National Institutes of Health NIH grant 5R35GM141931-03. A portion of this research was performed on a project award, 50231 (DOI 10.46936/sarr.proj.2018.50231/60000029), from the Environmental Molecular Sciences Laboratory, a DOE Office of Science User Facility sponsored by the Biological and Environmental Research program under Contract No. DE-AC05-76RL01830. We also thank the Robert A. Welch Foundation (grant F-0020) for supporting the Lu group's research programs at the University of Texas at Austin. We thank the experimental work and discussion in mass spectrometry with Dr. Justin Arrington, Dr. Brian Imai, Dr. Peter Yau at the University of Illinois at Urbana-Champaign, Dr. Maria Person and Ms. Michelle Gadush at the University of Texas at Austin. We acknowledge the help in crystallography data acquisition from Ms. Stacey Ortega and Dr. Yang Ha at LBNL under the proposal ALS-11565. We thank Dr. Aaron Ledray, Prof. Yan Zhang, Prof. Jonathan Sessler, Dr. Wantae Kim, Prof. Wilfred van der Donk, and Prof. Anabella Ivancich for the discussions to improve this manuscript, and Mr. Aritra Roy for proofreading this manuscript.

ABBREVIATIONS

PTM

post-translational modification

HCO

heme-copper oxidase

Mb

myoglobin

GAO

galactose oxidase

SCS

secondary coordination sphere

ORR

oxygen reduction reaction

ROS

reactive oxygen species

E. coli

Escherichia coli

LC

liquid chromatography

MS

mass spectrometry

NMR

nuclear magnetic resonance

EPR

electron paramagnetic resonance

FPLC

fast protein liquid chromatography

HPLC

high-performance liquid chromatography

UPLC

ultra-performance liquid chromatography SEC, size-exclusion chromatography

MALDI

matrix-assisted laser desorption/ionization

TOF

time-of-flight

HCD

Higher-energy collisional dissociation

KPi

potassium phosphate (buffer)

HSQC

heteronuclear single-quantum correlation spectroscopy

HMBC

heteronuclear multiple bond correlation

TOCSY

total correlation spectroscopy

NOESY

nuclear Overhauser Effect spectroscopy

RMSD

root mean square deviation

IgG

immunoglobin G

TMPD

N,N,N′,N′-tetramethyl-1,4-phenylenediamine

EDTA

ethylenediaminetetraacetic acid

PDB

Protein Data Bank

REFERENCES

  • (1).Walsh CT; Garneau-Tsodikova S; Gatto GJ Jr Protein Posttranslational Modifications: The Chemistry of Proteome Diversifications. Angew. Chem. Int. Ed 2005, 44 (45), 7342–7372. [DOI] [PubMed] [Google Scholar]
  • (2).Fuentes-Lemus E; Hägglund P; López-Alarcón C; Davies MJ Oxidative Cross-linking of Peptides and Proteins: Mechanisms of Formation, Detection, Characterization and Quantification. Molecules 2021, 27 (1), 15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (3).Graciano A; Liu A Protein-Derived Cofactors: Chemical Innovations Expanding Enzyme Catalysis. Chem. Soc. Rev 2025, 54, 4502–4530. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (4).Tsukihara T; Aoyama H; Yamashita E; Tomizaki T; Yamaguchi H; Shinzawa-Itoh K; Nakashima R; Yaono R; Yoshikawa S Structures of Metal Sites of Oxidized Bovine Heart Cytochrome c Oxidase at 2.8 Å. Science 1995, 269 (5227), 1069–1074. [DOI] [PubMed] [Google Scholar]
  • (5).Ferguson-Miller S; Babcock GT Heme/Copper Terminal Oxidases. Chem. Rev 1996, 96 (7), 2889–2908. [DOI] [PubMed] [Google Scholar]
  • (6).Wikström M; Krab K; Sharma V Oxygen Activation and Energy Conservation by Cytochrome c Oxidase. Chem. Rev 2018, 118 (5), 2469–2490. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (7).Rousseau DL; Ishigami I; Yeh S-R Structural and Functional Mechanisms of Cytochrome c Oxidase. J. Inorg. Biochem 2024, 112730. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (8).Das TK; Pecoraro C; Tomson FL; Gennis RB; Rousseau DL The Post-Translational Modification in Cytochrome c Oxidase Is Required to Establish a Functional Environment of the Catalytic Site. Biochemistry 1998, 37 (41), 14471–14476. [DOI] [PubMed] [Google Scholar]
  • (9).Rauhamäki V; Baumann M; Soliymani R; Puustinen A; Wikström M Identification of a Histidine-Tyrosine Cross-Link in the Active Site of the Cbb 3-Type Cytochrome c Oxidase from Rhodobacter Sphaeroides. Proc. Natl. Acad. Sci 2006, 103 (44), 16135–16140. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (10).Pratt DA; Pesavento RP; Van Der Donk WA Model Studies of the Histidine-Tyrosine Cross-Link in Cytochrome c Oxidase Reveal the Flexible Substituent Effect of the Imidazole Moiety. Org. Lett 2005, 7 (13), 2735–2738. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (11).McCauley KM; Vrtis JM; Dupont J; Van Der Donk WA Insights into the Functional Role of the Tyrosine- Histidine Linkage in Cytochrome c Oxidase. J. Am. Chem. Soc 2000, 122 (10), 2403–2404. [Google Scholar]
  • (12).Himo F; Noodleman L; Blomberg MRA; Siegbahn PEM Relative Acidities of Ortho-Substituted Phenols, as Models for Modified Tyrosines in Proteins. J. Phys. Chem. A 2002, 106 (37), 8757–8761. [Google Scholar]
  • (13).Liu J-G; Naruta Y; Tani F; Chishiro T; Tachi Y Formation and Spectroscopic Characterization of the Dioxygen Adduct of a Heme–Cu Complex Possessing a Cross-Linked Tyrosine–Histidine Mimic: Modeling the Active Site of Cytochrome c Oxidase. Chem. Commun 2004, No. 1, 120–121. [DOI] [PubMed] [Google Scholar]
  • (14).Kim E; Kamaraj K; Galliker B; Rubie ND; Moënne-Loccoz P; Kaderli S; Zuberbühler AD; Karlin KD Dioxygen Reactivity of Copper and Heme-Copper Complexes Possessing an Imidazole-Phenol Cross-Link. Inorg. Chem 2005, 44 (5), 1238–1247. [DOI] [PubMed] [Google Scholar]
  • (15).Cappuccio JA; Ayala I; Elliott GI; Szundi I; Lewis J; Konopelski JP; Barry BA; Einarsdóttir Ó Modeling the Active Site of Cytochrome Oxidase: Synthesis and Characterization of a Cross-Linked Histidine- Phenol. J. Am. Chem. Soc 2002, 124 (8), 1750–1760. [DOI] [PubMed] [Google Scholar]
  • (16).Liu J-G; Naruta Y; Tani F A Functional Model of the Cytochrome c Oxidase Active Site: Unique Conversion of a Heme–μ-Peroxo–CuII Intermediate into Heme–Superoxo/CuI. Angew. Chem 2005, 117 (12), 1870–1874. [DOI] [PubMed] [Google Scholar]
  • (17).Liu J-G; Naruta Y; Tani F Synthetic Models of the Active Site of Cytochrome c Oxidase: Influence of Tridentate or Tetradentate Copper Chelates Bearing a His–Tyr Linkage Mimic on Dioxygen Adduct Formation by Heme/Cu Complexes. Chem. Eur. J 2007, 13 (22), 6365–6378. [DOI] [PubMed] [Google Scholar]
  • (18).Isogai Y; Imamura H; Nakae S; Sumi T; Takahashi K; Nakagawa T; Tsuneshige A; Shirai T Tracing Whale Myoglobin Evolution by Resurrecting Ancient Proteins. Sci. Rep 2018, 8 (1), 16883. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (19).Jensen LMR; Sanishvili R; Davidson VL; Wilmot CM In Crystallo Posttranslational Modification Within a MauG/Pre–Methylamine Dehydrogenase Complex. Science 2010, 327 (5971), 1392–1394. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (20).Zhang J-L; Garner DK; Liang L; Barrios DA; Lu Y Noncovalent Modulation of pH-Dependent Reactivity of a Mn–Salen Cofactor in Myoglobin with Hydrogen Peroxide. Chem. Eur. J 2009, 15 (30), 7481–7489. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (21).Yeung N; Lin Y-W; Gao Y-G; Zhao X; Russell BS; Lei L; Miner KD; Robinson H; Lu Y Rational Design of a Structural and Functional Nitric Oxide Reductase. Nature 2009, 462 (7276), 1079–1082. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (22).Yu Y; Cui C; Liu X; Petrik ID; Wang J; Lu Y A Designed Metalloenzyme Achieving the Catalytic Rate of a Native Enzyme. J. Am. Chem. Soc 2015, 137 (36), 11570–11573. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (23).Mukherjee S; Mukherjee A; Bhagi-Damodaran A; Mukherjee M; Lu Y; Dey A A Biosynthetic Model of Cytochrome c Oxidase as an Electrocatalyst for Oxygen Reduction. Nat. Commun 2015, 6 (1), 8467. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (24).Vargas DA; Tinoco A; Tyagi V; Fasan R Myoglobin-Catalyzed C-H Functionalization of Unprotected Indoles. Angew. Chem. Int. Ed 2018, 57 (31), 9911–9915. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (25).Pott M; Hayashi T; Mori T; Mittl PR; Green AP; Hilvert D A Noncanonical Proximal Heme Ligand Affords an Efficient Peroxidase in a Globin Fold. J. Am. Chem. Soc 2018, 140 (4), 1535–1543. [DOI] [PubMed] [Google Scholar]
  • (26).Schaus L; Das A; Knight AM; Jimenez-Osés G; Houk K; Garcia-Borràs M; Arnold FH; Huang X Protoglobin-Catalyzed Formation of Cis-Trifluoromethyl-Substituted Cyclopropanes by Carbene Transfer. Angew. Chem 2023, 135 (4), e202208936. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (27).Bhattacharya S; Margheritis EG; Takahashi K; Kulesha A; D’Souza A; Kim I; Yoon JH; Tame JR; Volkov AN; Makhlynets OV; others. NMR-Guided Directed Evolution. Nature 2022, 610 (7931), 389–393. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (28).Deng Y; Dwaraknath S; Ouyang WO; Matsumoto CJ; Ouchida S; Lu Y Engineering an Oxygen-Binding Protein for Photocatalytic CO2 Reductions in Water. Angew. Chem. Int. Ed 2023, 62 (20), e202215719. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (29).Nie L-S; Liu X-C; Yu L; Liu A-K; Sun L-J; Gao S-Q; Lin Y-W Rational Design of an Artificial Metalloenzyme by Constructing a Metal-Binding Site Close to the Heme Cofactor in Myoglobin. Inorg. Chem 2024, 63(40), 18531–18535. [DOI] [PubMed] [Google Scholar]
  • (30).Miner KD; Mukherjee A; Gao Y-G; Null EL; Petrik ID; Zhao X; Yeung N; Robinson H; Lu Y A Designed Functional Metalloenzyme That Reduces O2 to H2O with over One Thousand Turnovers. Angew. Chem. Int. Ed 2012, 51 (23), 5589–5592. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (31).Liu X; Yu Y; Hu C; Zhang W; Lu Y; Wang J Significant Increase of Oxidase Activity through the Genetic Incorporation of a Tyrosine-Histidine Cross-Link in a Myoglobin Model of Heme-Copper Oxidase. Angew. Chem. Int. Ed 2012, 51 (18), 4312–4316. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (32).Huang X; Groves JT Oxygen Activation and Radical Transformations in Heme Proteins and Metalloporphyrins. Chem. Rev 2017, 118 (5), 2491–2553. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (33).Blomberg MR. Activation of O 2 and NO in Heme-Copper Oxidases–Mechanistic Insights from Computational Modelling. Chem. Soc. Rev 2020, 49 (20), 7301–7330. [DOI] [PubMed] [Google Scholar]
  • (34).Ledray AP; Dwaraknath S; Chakarawet K; Sponholtz MR; Merchen C; Van Stappen C; Rao G; Britt RD; Lu Y Tryptophan Can Promote Oxygen Reduction to Water in a Biosynthetic Model of Heme Copper Oxidases. Biochemistry 2022, 62 (2), 388–395. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (35).Chen Z-L; Meng J-M; Cao Y; Yin J-L; Fang R-Q; Fan S-B; Liu C; Zeng W-F; Ding Y-H; Tan D; others. A High-Speed Search Engine pLink 2 with Systematic Evaluation for Proteome-Scale Identification of Cross-Linked Peptides. Nat. Commun 2019, 10 (1), 3404. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (36).Neuhaus D; Williamson MP The Nuclear Overhauser Effect in Structural and Conformational Analysis, 2nd ed.; Wiley-VCH, 2000. [Google Scholar]
  • (37).Platzer G; Okon M; McIntosh LP pH-Dependent Random Coil 1H, 13C, and 15N Chemical Shifts of the Ionizable Amino Acids: A Guide for Protein p K a Measurements. J. Biomol. NMR 2014, 60, 109–129. [DOI] [PubMed] [Google Scholar]
  • (38).Bravo J; Fita I; Ferrer JC; Ens W; Hillar A; Switala J; Loewen PC Identification of a Novel Bond between a Histidine and the Essential Tyrosine in Catalase HPII of Escherichia Coli. Protein Sci. 1997, 6 (5), 1016–1023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (39).Sigman JA; Kwok BC; Lu Y From Myoglobin to Heme-Copper Oxidase: Design and Engineering of a CuB Center into Sperm Whale Myoglobin. J. Am. Chem. Soc 2000, 122 (34), 8192–8196. [Google Scholar]
  • (40).St. John PC; Guan Y; Kim Y; Kim S; Paton RS Prediction of Organic Homolytic Bond Dissociation Enthalpies at near Chemical Accuracy with Sub-Second Computational Cost. Nat. Commun 2020, 11 (1), 2328. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (41).Yang Q; Li Y; Yang J-D; Liu Y; Zhang L; Luo S; Cheng J-P Holistic Prediction of the pKa in Diverse Solvents Based on a Machine-Learning Approach. Angew. Chem 2020, 132 (43), 19444–19453. [DOI] [PubMed] [Google Scholar]
  • (42).Yoshikawa S; Shimada A Reaction Mechanism of Cytochrome c Oxidase. Chem. Rev 2015, 115 (4), 1936–1989. [DOI] [PubMed] [Google Scholar]
  • (43).Wikström M; Krab K; Sharma V Oxygen Activation and Energy Conservation by Cytochrome c Oxidase. Chem. Rev 2018, 118 (5), 2469–2490. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (44).Ortmayer M; Fisher K; Basran J; Wolde-Michael EM; Heyes DJ; Levy C; Lovelock SL; Anderson JR; Raven EL; Hay S; others. Rewiring the “Push-Pull” Catalytic Machinery of a Heme Enzyme Using an Expanded Genetic Code. ACS Catal. 2020, 10 (4), 2735–2746. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (45).Kwon H; Basran J; Pathak C; Hussain M; Freeman SL; Fielding AJ; Bailey AJ; Stefanou N; Sparkes HA; Tosha T; others. XFEL Crystal Structures of Peroxidase Compound II. Angew. Chem. Int. Ed 2021, 60 (26), 14578–14585. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (46).Ledray AP; Krest CM; Yosca TH; Mittra K; Green MT Ascorbate Peroxidase Compound II Is an Iron (IV) Oxo Species. J. Am. Chem. Soc 2020, 142 (48), 20419–20425. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (47).Denisov IG; Makris TM; Sligar SG; Schlichting I Structure and Chemistry of Cytochrome P450. Chem. Rev 2005, 105 (6), 2253–2278. [DOI] [PubMed] [Google Scholar]
  • (48).Veitch NC. Horseradish Peroxidase: A Modern View of a Classic Enzyme. Phytochemistry 2004, 65 (3), 249–259. [DOI] [PubMed] [Google Scholar]
  • (49).Moody PC; Raven EL The Nature and Reactivity of Ferryl Heme in Compounds I and II. Acc. Chem. Res 2018, 51 (2), 427–435. [DOI] [PubMed] [Google Scholar]
  • (50).Yu Y; Mukherjee A; Nilges MJ; Hosseinzadeh P; Miner KD; Lu Y Direct EPR Observation of a Tyrosyl Radical in a Functional Oxidase Model in Myoglobin during Both H2O2 and O2 Reactions. J. Am. Chem. Soc 2014, 136 (4), 1174–1177. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (51).Adam SM; Wijeratne GB; Rogler PJ; Diaz DE; Quist DA; Liu JJ; Karlin KD Synthetic Fe/Cu Complexes: Toward Understanding Heme-Copper Oxidase Structure and Function. Chem. Rev 2018, 118 (22), 10840–11022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (52).Ehudin MA; Senft L; Franke A; Ivanović-Burmazović I; Karlin KD Formation and Reactivity of New Isoporphyrins: Implications for Understanding the Tyr-His Cross-Link Cofactor Biogenesis in Cytochrome c Oxidase. J. Am. Chem. Soc 2019, 141 (27), 10632–10643. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (53).Petrik ID; Davydov R; Kahle M; Sandoval B; Dwaraknath S; Adelroth P; Hoffman B; Lu Y An Engineered Glutamate in Biosynthetic Models of Heme-Copper Oxidases Drives Complete Product Selectivity by Tuning the Hydrogen-Bonding Network. Biochemistry 2021, 60 (4), 346–355. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (54).Kornblatt JA. pH-Induced Alterations in the Structure and Activity of Cytochrome c Oxidase. Can. J. Biochem 1980, 58 (10), 840–850. [DOI] [PubMed] [Google Scholar]
  • (55).Maurel P; Douzou P; Waldmann J; Yonetani T Enzyme Behaviour and Molecular Environment. The Effects of Ionic Strength, Detergents, Linear Polyanions and Phospholipid on the pH Profile of Soluble Cytochrome Oxidase. Biochim. Biophys. Acta BBA-Enzymol 1978, 525 (2), 314–324. [DOI] [PubMed] [Google Scholar]
  • (56).Petrik ID; Davydov R; Ross M; Zhao X; Hoffman B; Lu Y Spectroscopic and Crystallographic Evidence for the Role of a Water-Containing H-Bond Network in Oxidase Activity of an Engineered Myoglobin. J. Am. Chem. Soc 2016, 138 (4), 1134–1137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (57).Liu M; Zhang Z; Cheetham J; Ren D; Zhou ZS Discovery and Characterization of a Photo-Oxidative Histidine-Histidine Cross-Link in IgG1 Antibody Utilizing 18O-Labeling and Mass Spectrometry. Anal. Chem 2014, 86 (10), 4940–4948. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (58).Xu C-F; Chen Y; Yi L; Brantley T; Stanley B; Sosic Z; Zang L Discovery and Characterization of Histidine Oxidation Initiated Cross-Links in an IgG1 Monoclonal Antibody. Anal. Chem 2017, 89 (15), 7915–7923. [DOI] [PubMed] [Google Scholar]
  • (59).Powell T; Knight MJ; O’Hara J; Burkitt W Discovery of a Photoinduced Histidine-Histidine Cross-Link in an IgG4 Antibody. J. Am. Soc. Mass Spectrom 2020, 31 (6), 1233–1240. [DOI] [PubMed] [Google Scholar]
  • (60).Kandy SK; Pasquale MA; Chekan JR Aromatic Side-Chain Cross-linking in RiPP Biosynthesis. Nat. Chem. Biol 2025, 21 (2), 168–181. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (61).Shi Y; Xia Y; Gao W; Wang J; Shi B; Wang H Enzymatic Cross-linking of Histidine Side Chains in Peptide Natural Products. Nat. Prod. Rep 2025, 42 (5), 763–773. [DOI] [PubMed] [Google Scholar]
  • (62).He B-B; Liu J; Cheng Z; Liu R; Zhong Z; Gao Y; Liu H; Song Z-M; Tian Y; Li Y-X Bacterial Cytochrome P450 Catalyzed Post-Translational Macrocyclization of Ribosomal Peptides. Angew. Chem. Int. Ed 2023, 62 (46), e202311533. [DOI] [PubMed] [Google Scholar]
  • (63).Bontreger LJ; Gallo AD; Moon J; Silinski P; Monson EE; Franz KJ Intramolecular Histidine Cross-Links Formed via Copper-Catalyzed Oxidation of Histatin Peptides. J. Am. Chem. Soc 2025, 147 (15), 12749–12765. [DOI] [PubMed] [Google Scholar]
  • (64).Close DM; Wardman P Calculation of Standard Reduction Potentials of Amino Acid Radicals and the Effects of Water and Incorporation into Peptides. J. Phys. Chem. A 2018, 122 (1), 439–445. [DOI] [PubMed] [Google Scholar]
  • (65).Schmid R; Heuckeroth S; Korf A; Smirnov A; Myers O; Dyrlund TS; Bushuiev R; Murray KJ; Hoffmann N; Lu M; others. Integrative Analysis of Multimodal Mass Spectrometry Data in MZmine 3. Nat. Biotechnol 2023, 41 (4), 447–449. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (66).Pino LK; Searle BC; Bollinger JG; Nunn B; MacLean B; MacCoss MJ The Skyline Ecosystem: Informatics for Quantitative Mass Spectrometry Proteomics. Mass Spectrom. Rev 2020, 39 (3), 229–244. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (67).Harris RK; Becker ED; Cabral De Menezes SM; Granger P; Hoffman RE; Zilm KW Further Conventions for NMR Shielding and Chemical Shifts (IUPAC Recommendations 2008). Pure Appl. Chem 2008, 80 (1), 59–84. [Google Scholar]
  • (68).Winter G. Xia2: An Expert System for Macromolecular Crystallography Data Reduction. J. Appl. Crystallogr 2010, 43 (1), 186–190. [Google Scholar]
  • (69).Winter G; Waterman DG; Parkhurst JM; Brewster AS; Gildea RJ; Gerstel M; Fuentes-Montero L; Vollmar M; Michels-Clark T; Young ID; others. DIALS: Implementation and Evaluation of a New Integration Package. Acta Crystallogr. Sect. Struct. Biol 2018, 74 (2), 85–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (70).Kabsch W. XDS. Acta Crystallogr. D Biol. Crystallogr 2010, 66 (2), 125–132. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (71).Evans PR; Murshudov GN How Good Are My Data and What Is the Resolution? Acta Crystallogr. D Biol. Crystallogr 2013, 69 (7), 1204–1214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (72).McCoy AJ; Grosse-Kunstleve RW; Adams PD; Winn MD; Storoni LC; Read RJ Phaser Crystallographic Software. J. Appl. Crystallogr 2007, 40 (4), 658–674. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (73).Adams PD; Afonine PV; Bunkóczi G; Chen VB; Echols N; Headd JJ; Hung L-W; Jain S; Kapral GJ; Kunstleve RWG; others. The Phenix Software for Automated Determination of Macromolecular Structures. Methods 2011, 55 (1), 94–106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (74).Emsley P; Lohkamp B; Scott WG; Cowtan K Features and Development of Coot. Acta Crystallogr. D Biol. Crystallogr 2010, 66 (4), 486–501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (75).Williams CJ; Headd JJ; Moriarty NW; Prisant MG; Videau LL; Deis LN; Verma V; Keedy DA; Hintze BJ; Chen VB; others. MolProbity: More and Better Reference Data for Improved All-Atom Structure Validation. Protein Sci. 2018, 27 (1), 293–315. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (76).Laskowski RA; MacArthur MW; Moss DS; Thornton JM PROCHECK: A Program to Check the Stereochemical Quality of Protein Structures. J. Appl. Crystallogr 1993, 26 (2), 283–291. [Google Scholar]
  • (77).Meng EC; Goddard TD; Pettersen EF; Couch GS; Pearson ZJ; Morris JH; Ferrin TE UCSF ChimeraX: Tools for Structure Building and Analysis. Protein Sci. 2023, 32 (11), e4792. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (78).Stoll S; Schweiger A EasySpin, a Comprehensive Software Package for Spectral Simulation and Analysis in EPR. J. Magn. Reson 2006, 178 (1), 42–55. [DOI] [PubMed] [Google Scholar]
  • (79).Hamelryck T; Manderick B PDB File Parser and Structure Class Implemented in Python. Bioinformatics 2003, 19 (17), 2308–2310. [DOI] [PubMed] [Google Scholar]
  • (80).Cock PJ; Antao T; Chang JT; Chapman BA; Cox CJ; Dalke A; Friedberg I; Hamelryck T; Kauff F; Wilczynski B; others. Biopython: Freely Available Python Tools for Computational Molecular Biology and Bioinformatics. Bioinformatics 2009, 25 (11), 1422–1423. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material

RESOURCES