Rejigging Electron and Proton Transfer to Transition between Dioxygenase, Monooxygenase, Peroxygenase, and Oxygen Reduction Activity: Insights from Bioinspired Constructs of Heme Enzymes

Abstract

Nature has employed heme proteins to execute a diverse set of vital life processes. Years of research have been devoted to understanding the factors which bias these heme enzymes, with all having a heme cofactor, toward distinct catalytic activity. Among them, axial ligation, distal super structure, and substrate binding pockets are few very vividly recognized ones. Detailed mechanistic investigation of these heme enzymes suggested that several of these enzymes, while functionally divergent, use similar intermediates. Furthermore, the formation and decay of these intermediates depend on proton and electron transfer processes in the enzyme active site. Over the past decade, work in this group, using in situ surface enhanced resonance Raman spectroscopy of synthetic and biosynthetic analogues of heme enzymes, a general idea of how proton and electron transfer rates relate to the lifetime of different O₂ derived intermediates has been developed. These findings suggest that the enzymatic activities of all these heme enzymes can be integrated into one general cycle which can be branched out to different catalytic pathways by regulating the lifetime and population of each of these intermediates. This regulation can further be achieved by tuning the electron and proton transfer steps. By strategically populating one of these intermediates during oxygen reduction, one can navigate through different catalytic processes to a desired direction by altering proton and electron transfer steps.

Naturally occurring metalloenzymes catalyze crucial chemical steps in nature and have inspired chemists to design small molecule mimics of these enzymes.¹⁻¹³ Among the known metallo-enzymes, the heme super family of enzymes comprises a large number of members which exhibit versatile activity.¹⁴ All these metallo-proteins contain heme as a prosthetic group covalently or noncovalently bound to the protein backbone. Heme enzymes are known to catalyze electron transfer (cytochrome c; Cytc, Figure 1A) between proteins¹⁵ and reduce O₂ by four electrons to H₂O (cytochrome c oxidase; CcO, Figure 1B) during mitochondrial respiration.¹⁶ It also transports and stores O₂, e.g., hemoglobin (Hb) and myoglobin (Mb, Figure 1C).^17,18 Additionally, the versatility of heme enzymes in different catalytic processes involving small molecule activation, helping to achieve useful chemical transformations, is well established.¹⁹ After the historical discovery of Mason and Hayashi, it took a few years to realize that heme enzymes can take part in oxygenation reaction, i.e., incorporation of oxygen atoms from molecular oxygen into inert organic substrates (monooxygenase and dioxygenase).^20,21 The diverse reactivity of heme enzymes, essentially containing the same heme cofactor, have been attributed to the differences in their proximal ligand and distal super structure. Years of research have been invested to investigate the effect of these two factors to rationalize the diverse functional activity of these enzymes.^14,22−26 The variation in these distal superstructures (e.g., hydrogen bonding, local electrostatics, proton transfer channels, etc.) may or may not affect the electronic structure of the metal center in the resting state substantially unlike the proximal axial ligation. However, the fact that both axial ligands and distal residues affect the reactivity of these active sites dramatically by tuning the transition states of different possible reaction pathways has been established beyond doubt.

Figure 1.

Active site structures of (A) cytochrome c (PDB ID: 1HRC), (B) cytochrome c oxidase (PDB ID: 1OCC), (C) myoglobin (PDB ID: 2 V1J), (D) cytochrome P450 (PDB ID: 1AKD), and (E) heme peroxidase (PDB ID: 2YLJ).

Apart from their different structural attributes, such as different proximal ligation and distal superstructure, heme enzymes involved in oxidative chemistry can be grouped in two broad classes based on the oxidants used, namely, oxygenases (e.g., cytochrome P450; Cyt P450, Figure 1D) and oxidases that use O₂ and peroxidases (Figure 1E) and peroxygenases that use H₂O₂ to oxidize substrates.^14,27 Molecular O₂, while being a green and sustainable oxidant, is also difficult to activate. The spin forbidden electron transfer to the triplet ground state of O₂ from the diamagnetic ground states of organic molecules/substrates makes this process kinetically uphill. To overcome this, nature has evolved to use transition metals like iron in the active site of the enzymes. Apart from the spin barrier, these reactions are challenging owing to the inertness of these organic substrates; and to activate the strong covalent bonds in the substrates, highly reactive oxidants are generated from O₂ and H₂O₂ (Figure 2). The electron rich porphyrin systems as the ligand, provide reasonably stable ligand framework for this chemistry. Uncontrolled reaction between H₂O₂/O₂ and transition metals results in reactive hydroxyl radicals, and the metallo-enzymes have evolved to circumvent such reactions in the enzymatic active site.²⁸⁻³¹

Figure 2.

Oxidation of substrates by (A) peroxidase using H₂O₂ and (B) Cyt P450 using molecular O₂

The diverse reactivity derived from iron porphyrin and dioxygen makes heme enzymes versatile catalysts, and decades of investigation into their mechanism has led to the identification of the reactive intermediates involved in these different reactions.^1,14,32−36 Interestingly, several of these enzymes, while functionally divergent, go through similar intermediates. In fact, the O₂ reduction process in CcO involves most of these intermediates (Figure 3, black line).¹⁶ In CcO, O₂ reduction is initiated by the binding of O₂ to a ferrous heme, formed after one electron reduction of the resting ferric state (E₁⁰, Figure 3), to afford a ferric-superoxo (Fe^III–O₂^•–) species. The Fe^III–O₂^•– species is also the reactive intermediate in dioxygenase.³⁷ This Fe^IIIO₂^•– species after a one electron (E₂⁰) reduction and a single protonation generates a ferric–hydroperoxide species PFe^III–OOH. The O–O bond of this peroxide intermediate is heterolytically cleaved. One of these two electrons needed to heterolyze the O–O bond is provided by the ferric heme which gets oxidized to PFe^IV=O, and the other electron is derived from another redox active site nearby (e.g., Cu_B or tyrosine residue).^16,38 Stepwise reduction of these high-valent species leads to the elimination of a second molecule of water and regeneration of the resting ferric state.

Figure 3.

Consolidated reactivity scheme for heme enzymes involved in O₂ activation. O₂ binding proteins (orange arrow) use O₂ to form Fe^IIIO₂^•– to carry and store O₂. Heme dioxygenase (purple arrow) uses the same intermediate to incorporate both the oxygen atoms of the molecular O₂ into organic substrates. CcO (black arrow) and Cyt P450 (golden arrow) enter into the cycle through the Fe^IIIO₂^•– intermediate but move forward in the cycle to form high valent intermediates and take part in the O₂ reduction and oxygenase mechanism, respectively. The prominent bifurcation of the reactivity happens after the formation of the high valent intermediate, and this branching is schematically represented in the figure using color coded arrows. On the other hand, heme peroxidase (red arrow) enters cycle using H₂O₂ by forming the Fe^III–OOH intermediate and then proceeds through the cycle to oxidize substrates. The formation of high valent intermediates with peroxides was also achieved using a similar technique for oxygen carrying enzymes and dioxygenase enzymes (dashed purple and orange arrows). Note that this does not always lead to any native enzymatic reaction.

The PFe^III–OOH species formed in CcO is also an intermediate in peroxidases and peroxygenases (Figure 3, red lines). In peroxidases, the formation of the reactive species, compound I, does not require additional electron but requires a protons for efficient O–O heterolysis.³⁴ In peroxidases and monooxygenases, the two electrons needed for the heterolysis can be provided by the ferric porphyrin which, in turn, gets oxidized to compound I (described as P^•+Fe^IV=O, i.e., a ferryl unit bound to a porphyrin cation-radical).^32,39 The compound I species is the reactive intermediate in peroxidases which can oxidize 2 equiv of substrate by one electron oxidation to regenerate the resting ferric state. In monooxygenases, the compound I generated from O₂ in a mechanism similar to CcO (i.e., via sequential ET and PT to a PFe^III–O₂^– species; Figure 3 brown line) oxidizes a substrate by two electrons to generate the resting ferric state which then requires an electron to be reduced to the active ferrous state.

The occurrence of similar intermediates in heme enzymes which are functionally different suggests that the enzymatic activities of all these heme enzymes can perhaps be integrated into one general reactivity cycle which can be branched out to different catalytic pathways by regulating the lifetime and population of each of these intermediates. Such a proposal was originally forwarded by Sligar et al..⁴⁰ This regulation can further be achieved by tuning the electron and proton transfer steps as has been clearly demonstrated in peroxidases and cytP450 monooxygenases as has been discussed elsewhere.^14,40 This idea gained support from the fact that several of these enzymes allows differences in their reactivity when the reaction conditions are changed (Figure 3, dashed lines).⁴¹⁻⁴⁵ Good examples of these are the monooxygenases which can act as peroxygenasse when H₂O₂ is used as an oxidant⁴² or oxygen reductases when sufficient reducing equivalents are provided in the absence of specific substrates.⁴¹ In bioinspired analogues, strategically populating one of these intermediates during the oxygen reduction, via catalyst design, one may be able to navigate a catalytic process to a desired direction by controlling the electron and proton transfer steps.

In enzymatic systems, there are several factors which decisively affect electron and proton transfer steps. The proton transfer (PT) steps in the active site are determined by the nature and number of proton transfer residues. Generally, proton transfer channels shuttle protons from protein surface to the catalytic center which is often harbored deep inside a protein.^44,46 The electron transfer (ET) rate, on the other hand, depends on factors like free energy of electron transfer (ΔG°_ET), reorganization energy (λ), and electronic coupling between the donor and acceptor (H_AB) centers.^47,48 The axial ligand to the heme affects both the ET and PT steps. The thermodynamic potential of the heme center, the pK_a of the transaxial ligands, as well as the H_AB all depend on the axial ligand.^22,49−53 The same have also been established in synthetic systems, and in general a thiolate ligand has been known to lower reduction potentials of the heme and raise the pK_a of any transaxial ligand due to its strong covalent donation.⁵⁴⁻⁵⁹

The heme enzymes discussed above that react with O₂ can be classified with respect to the number of electrons and protons it requires as well as the rates of ET and PT. As illustrated in Figure 3, hemoglobin and dioxygenases require no electrons or protons although their reactivities are different from each other (brown and purple lines, respectively).⁶⁰ Peroxidases require only protons as the two oxidizing equivalents of peroxide are used to oxidize a substrate (red lines).³⁹ Monooxygenase oxidizes the organic substrate but requires two electrons and two protons to do so (yellow lines).⁶¹ The rate of electron transfer (two steps, E₁° and E₂°) to the monooxygenase active site varies between 10 and 500 s^–1.⁶² The catalytic cycle during the oxygen reduction (e.g., in CcO, black line) requires four electrons, having four different reduction potentials (E₁⁰, E₂⁰, E₃⁰, and E₄⁰, Figure 3) to be rapidly transferred to the reaction center (k_ET > 10⁵ s^–1).^63,64 The Fe^III–O₂^•– species is common to all O₂ binding enzymes, the Fe^III–OOH (compound 0), P^•+Fe^IV=O (compound I), is common to peroxidases and monooxygenases, and Fe^IV=O (compound II, Fe^IV–OH in cytP450) is common to peroxidase and CcO. Due to similarities in the intermediates involved, it is possible for some of these heme enzymes to access intermediates that belong to the reaction cycles of different enzymes (dashed lines); e.g., myoglobin can react with H₂O₂ to form compound I,⁴⁵ and Cyt P450 can react with H₂O₂ (peroxide shunt) like peroxidases.⁶⁵ However, the reactivities of these species are different as they are strongly dependent on the axial ligand and/or distal environment of the heme and thus the ability to form these intermediates did not automatically result in facile reactivity.

Years of research have resulted in impressive structural models of several heme enzymes and helped to understand their chemistry.^{5,54,61,66−69} In addition to synthetic models, biosynthetic models using myoglobin and cytochrome c peroxidase scaffolds have provided a bridge between synthetic models and natural enzymes.⁷⁰⁻⁷⁹ However, creating functional synthetic models of these O₂ activating enzymes in solutions is associated with several challenging limitations. In some cases, stable models are difficult to generate; for example, assembling thiolate bound heme complexes are synthetically challenging.^68,80−82 In other cases, reaction of these complexes with the oxidant H₂O₂, for example, leads to self-degradation.^83,84 Attempting to use O₂ as an oxidant by activating it with ferrous porphyrin requires a chemical reducing agent to reinstate the reduced ferrous heme for catalysis.⁸⁵ However, that inevitably leads to inefficient catalytic conditions owing to the adventitious partial reduction of O₂ resulting in the formation of reactive oxygen species including radicals.⁸⁶ At the same time, these reactivities are complicated by a bimolecular reaction between freely diffusing reactive iron porphyrin oxidant species. Proteins manage to avoid this by selectively controlling the electron transfer and proton transfer pathways and at the same time avoid direct reaction of O₂ with the reductant which is generally an ET protein or an ET active site in the same protein. Recent developments have focused on including ET and PT processes in the design of biomimetic systems, relying substantially on the insight from in situ spectroscopy, which has resulted in several functional analogues of the heme proteins involved in O₂ and H₂O₂ activation. These are discussed in this perspective.

Electrocatalytic O₂ reduction by synthetic analogues of CcO has been extensively investigated where the electrode provides the electrons necessary for the reduction of O₂ via ET.^66,86−89 A reasonably detailed understanding of the species involved in this process has evolved from spectro-electrochemical investigations of iron porphyrins involved in O₂ reduction both under heterogeneous and homogeneous electrocatalytic conditions.^86,87 The mechanism evolving from these investigations points to distinct similarities with that of the naturally occurring metallo-enzymes involved in O₂ reduction (Figure 3, CcO). A ferric porphyrin needs to be reduced to its ferrous state (Figure 4, E₁), under both homogeneous and heterogeneous conditions, which in the presence of O₂ should produce Fe^III–O₂^•– (Figure 4). At E₂, this Fe^III–O₂^•– gets reduced by an electron to produce a ferric peroxo (Figure 4, Fe^III–O₂^2–) or hydroperoxo (Figure 4, Fe^III–O₂H) intermediate depending on the availability of protons. Nonetheless, for most cases, the potential range between E₁ and E₂ will be most likely be populated with the Fe^III–O₂^•– intermediate during electrochemical O₂ reduction, assuring O₂ binding to ferrous porphyrin is fast. Fe^III–O₂^•– can undergo different chemical transformations depending on the reaction conditions. The progress of the reaction in the forward direction depends on the proton sources that have been used. Depending on the relative pK_a of the proton source used and the basicity of the Fe^III–O₂^•– species (depends on axial ligand and distal hydrogen bonding interactions), it can undergo either proton transfer followed by electron transfer (PTET),⁹⁰ electron transfer followed by proton transfer (ETPT),^68,88 or proton coupled electron transfer (PCET)⁹¹ to form Fe^III–O₂H (Figure 4). Using in situ surface enhanced resonance Raman spectroscopy and electrochemistry, Chatterjee et al. demonstrated that for heterogeneous conditions a proton coupled electron transfer to this Fe^III–O₂^•– species may be the rate-determining step for thiolate ligated iron porphyrins.⁶⁷ This Fe^III–O₂H undergoes O–O bond scission to form compound I during the distal oxygen protonation followed by an electron transfer to form compound II, and each of these intermediates should exist in potential region E₃. Alternatively, the protonation in the proximal oxygen of the Fe^III–O₂H leads to PROS generation.⁹²

Figure 4.

Schematic representation of the branching pathways during the ORR by iron porphyrin systems. Here, L represents different axial ligands.

Several groups have been investigating the roles of proton transfer residues in O₂ reduction by these heme enzymes. Functionalities like carboxylic acids,^93,94 triazoles,^95,96 pyridines,⁹⁷ amines,^98,99 phenols,^100,101 guanidines,¹⁰² etc. have been appended to the porphyrin skeleton to facilitate proton transfer to the porphyrin active site. These modifications have produced enhancements in rate as well as selectivity in O₂ reduction under both homogeneous and heterogeneous conditions and mimics the PT pathways present in heme enzymes involved in O₂ activation. Alternatively, the ET process in nature has been modeled by providing electrons from the electrodes at controlled potentials using heterogeneous electrocatalysis (Figure 5B).¹⁰³ In heterogeneous catalysis, reducing equivalents (electron) are provided from the electrode and substrates (O₂, small molecules, and protons) are obtained from the solution, making the pathways of these components to the catalyst mutually exclusive (Figure 5A). The electrodes used for this purpose need to be modified with self-assembled monolayers (SAMs) of nonconducting organic molecules.¹⁰⁴ ET through a SAM follows a tunnelling mechanism, and its rate can be tuned by altering the chain length of the thiols comprising the SAM.¹⁰⁵ The most widely used form of SAMs is thiols absorbed on Au surfaces. Using these constructs, both the potential (driving force of ET) as well as the rate of ET can be independently addressed.

Figure 5.

Pictorial representation of the (A) the mutually exclusive delivery nature of the protons and electrons in the Cyt P450 active site and (B) electrodes used for heterogeneous electrocatalytic purposes where ‘L’ depicts the axial ligands to the Iron porphyrin.

Duly functionalized iron porphyrins mimicking the heme cofactor can be attached to the SAM impregnated with terminally functionalized thiols and can act as axial ligands to the iron porphyrin (Figure 5B). This results in electrodes with site-isolated bioinspired catalysts. Such constructs allow specific control of electron and proton transfer residues and allow the assembly of iron porphyrin active sites bound to specific axial ligands of choice. At the same time, thinly dispersing these “active sites” on the electrode (Figure 5B) rescinds all bimolecular pathways that lead to catalyst degradation in homogeneous solutions. Based on the pattern of reactivity of these heme enzymes described above (Figure 3), the reaction of their synthetic analogues with O₂ can be classified in four categories: (i) fast proton and fast electron transfer, (ii) fast proton and slow electron transfer, (iii) no electron transfer and no proton transfer, and finally (iv) fast proton transfer and no electron transfer.

Fast Proton and Fast Electron Transfer: The Case of CcO

More than 90% of O₂ consumption by living organisms is catalyzed by a superfamily of enzymes, the heme–copper oxidases, among which CcO is the most significant one.^16,106−108 CcO is a large membrane bound enzyme (active site structure of CcO is depicted in Figure 6A and B) in the respiratory chain which catalyzes the 4e^–/4H⁺ reduction of oxygen to water, and the free energy generated from this reaction is used to pump protons across the mitochondrial membranes. The resulting proton gradient drives the synthesis of adenosine 5′-triphosphate (ATP), an organic molecule that provides energy necessary to drive all processes in a living cell. However, these codependent processes which supply energy for the survival of all aerobic living organisms rely on efficient electron and proton transfer to the active site of CcO. Intraprotein electron transfer from the Cu_A and heme b ET sites to the binuclear O₂ reduction site of CcO occurs at a very fast rate of ∼10⁵–10⁶ s^–1.³⁸ On the other hand, proton transfer happens via separate proton channels which are mediated by conserved residues, e.g., D91, E242. Apart from facile intramolecular proton and electron transfer, the 4e^–/4H⁺ reduction of water in these proteins is also facilitated by conserved tyrosine residue.⁹¹

Figure 6.

(A) Active site structure of CcO in an enzyme environment (1OCC)¹⁰⁹ and (B) synthetic model of the CcO active site. (C) Computer model of G65YCu_BMb showing its catalytic center containing the distal Cu_B bound to three histidine and a tyrosine residue. Crystal structures of (D) V68ECuBMb and (E) I107E/V68ECuBMb showing the glutamate residues (yellow).

Myoglobin based biosynthetic models of CcO were constructed to investigate the role of proton transfer rate in the oxygen reduction reaction (ORR).¹¹⁰ In these models, the distal phenylalanine and leucine residues are mutated to histidine to create the Cu_B binding site^111,112 and a glycine residue is mutated to tyrosine to emulate the Tyr244 residue in CcO (this construct is labeled G65YCu_BMb, Figure 6C).¹¹³ In other constructs, glutamate residues were systematically introduced (V68ECu_BMb and I107E V68ECu_BMb; Figure 6D and E, respectively) by similar site directed mutagenesis of Mb to help facile proton transfer.^74,114 The introduction of the tyrosine and glutamate residues helped achieve faster ORR rates in solution; however, ET from the reductant limited the rate of catalysis.^71,113,115

To achieve facile catalysis not limited by ET from solution species, these biosynthetic models were directly attached to SAMs covered Au electrodes via 1,3-cycloaddition of the alkyne functionalized heme cofactor to azide terminated thiols thinly interspersed on the SAM surface (Figure 7).¹¹⁶ The short chain length of the SAM ensures fast ET rates from the electrode to the protein. All these biochemical electrodes demonstrated efficient 4e^–/4H⁺ ORR with turnover rates in the order of 10⁷ M^–1 s^–1. The introduction of tyrosine residue (G65YCu_BMb) displayed a higher rate of ORR in heterogeneous conditions relative to the Cu_BMb (biosynthetic heme-Cu active site without tyrosine residue) and a substantial increase in the selectivity for 4e^–/4H⁺ reduction of O₂ to water. Systematic incorporation of proton transfer glutamate residues near the protein active site, via site-directed mutagenesis, resulted in an ORR rate that rivals that of native CcO. Importantly, while the rates of O₂ reduction were all 10⁷ M^–1 s^–1, the solvent kinetic isotope effect (KSIE) for O₂ reduction changed from 16 in the G65YCu_BMb mutant to 4.3 in V68ECu_BMb and 2.4 in I107EV68ECu_BMb (Table 1, rows 4); i.e., the KSIE decreases with increasing number of local proton transfer residues.¹¹⁷ In situ surface enhanced resonance Raman spectroscopy coupled to rotating disc electrochemistry (SERRS-RDE), a technique developed to probe electrocatalysts in operando, reveals that the O₂ reduction mechanism was substantially perturbed on swapping the tyrosine with glutamate. Specifically, a ferric peroxide intermediate that accumulate under steady state catalytic conditions in the G65YCu_BMb does not accumulate in either of the glutamate containing mutants. Avoiding the accumulation of the peroxide species also avoided the generation of the partially reduced product (2e^–/2H⁺) H₂O₂ (resulting from hydrolysis of peroxide intermediate), resulting in increases selectivity of O₂ reduction to water for these mutants. Similar enhancement in the rate and 4e^–/4H⁺ selectivity for O₂ reduction has been reported for iron porphyrin complexes appended with pendant basic residues like pyridine, guanidine, and amines (FeL2, Fe-MARG and FeL₃ in Figure 8) that stay protonated at neutral pH.^98,102,118 These complexes facilitate the rate determining O–O bond cleavage step by disposing of the Fe^III–OOH intermediate rapidly, via O–O heterolysis, resulting in diffusion limited catalysis with high selectivity. Indeed, rapid kinetics experiments in organic solvents show that the pendant amine can catalyze heterolysis of the O–O bond of peracids forming compound I species having activation energies lower than that reported for native horseradish peroxidase (HRP). The hydrogen bonding between a protonated pendant amine and the distal oxygen of a bound −OOH group strongly polarizes the electron density of the peroxide, resulting in weakening of the O–O bond. This weakening lowers the energy of the O–O σ* orbital which in turn enhances the back bonding into it from the occupied t₂ electrons in the low-spin Fe^III center; all of these effects lead to a facile O–O bond heterolysis.⁹⁸ In situ SERRS-RDE data on these systems indicated that these synthetic complexes show little or no accumulation of Fe^III–OOH under catalytic condition as is observed for the V68ECu_BMb and I107EV68ECu_BMb and were thus, successful in emulating that protons transfer mechanism observed in the biochemical constructs.¹¹⁸

Figure 7.

Schematic representation of the fabrication of the bioinspired electrodes.¹¹⁹

Table 1. Rate and Selectivity of Different Biosynthetic and Synthetic Models of CcO toward Electrochemical O₂ Reduction Reaction.

models	synthetic model	ET rate (s–1)	proton transfer channel (numbers)	second order rate (M–1 s–1)	selectivity (4e–/4H+)	KSIE	ref
biosynthetic models	G65YCuBMb	∼3.1 × 104	no	1.98 × 107	94%	16.0 ± 1.0	(116)
	V68ECuBMb	∼3.1 × 104	yes (1)	2.20 × 107	96%	4.3 ± 0.5	(117)
	I107E/V68E CuBMb	∼3.1 × 104	yes (2)	8.6 × 106	93%	2.4 ± 0.07	(117)
synthetic models	Fe–Cu heme Cu oxidase model	∼106	no	1.2 × 105	96%		(121)
		∼1–10	no	ND	87%		(120)
	6L-FeCu	∼106	no	4.49× 106	96%		(67)
		∼1–10	no	ND	87%
	FeL2	∼106	yes	1.80 × 107	99%	N.D	(98)
		∼1–10		ND	95%
	FeL3	∼106	yes	1.35 × 107	98%	N.D	(98)
		∼1–10		ND	96%
	Fe-Marg	∼106	yes	4.2 × 106	98%	N.D	(118)
		∼106		ND	82%

Figure 8.

(A) L⁶-FeCu and (B) Fe–Cu heme Cu oxidase model. Porphyrins with basic residues in the distal super structure: (C) FeL₂, (D) FeL₃, and (E) Fe-Marg.

Under electrocatalytic conditions at neutral pH’s, a Fe^III–OOH intermediate formed during O₂ reduction needs to be reduced for the O–O bond to be cleaved.^120,121 Failing to provide an electron rapidly results in its hydrolysis to generate H₂O₂.¹²¹ Alternatively, rapid heterolysis of the O–O bond would result in the formation of Compound I and avoid the release of H₂O₂. The effectiveness of these pendant groups in assisting the protonation and the ensuing heterolysis of the O–O bond in a Fe^III–OOH intermediate, thus, can be judged better by slowing down the ET to the catalytic active site. ET rate constants for standard EPG electrodes are as high as ∼10⁶ s^–1, whereas electrodes modified with an octanethiol SAM (C₈SH) and hexadecanethiol SAM (C₁₆SH) show rate constants of 10³ and 6–10 s^–1, respectively.^122−125 Under these slow ET fluxes, simple iron porphyrins, including ones having a distal hydrogen bonding cavity, which exhibit substantial selectivity for 4e^–/4H⁺ ORR under rapid ET flux, show very high H₂O₂ production.^67,120,126 Alternatively, these porphyrins with pendant basic residues show only a negligible decrease in selectivity even when the electron transfer was slowed down by several orders of magnitude, as they could heterolyze the O–O bond of this Fe^III–OOH intermediate by rapid protonation, avoiding its hydrolysis as has been experimentally demonstrated in organic solution.⁹⁸ The role of distal hydrogen bonding residues is also well established in binuclear synthetic models of CcO and mononuclear hangman porphyrins during the selective 4e^–/4H⁺ reduction of O₂.^{93,94,101,127}

Slow Electron Transfer and Fast Proton Transfer: The Case of Cyt P450

The Cyt P450 superfamily of enzymes requires O₂ binding and subsequent O–O bond scission during the ORR cycle but avoids reduction of these high valent compound I species.^61,128 Cyt P450 (Figure 9A) is one of the largest known family of metallo-enzymes, which takes part in several catabolic and anabolic processes including but not limited to xenobiotic detoxification, steroids and hormone biosynthesis. Due to this diverse reactivity, Cyt P450 has always been a subject of interest to the scientific community, striving toward better understanding of the factors that govern their reactivity with an eventual goal of developing artificial catalysts mimicking the reactivity of these enzymes. The generation of the high valent intermediate compound I,³² responsible for the monooxygenation via a hydrogen atom transfer and rebound mechanism,¹²⁹ from ferrous heme and O₂ requires two protons and two electrons. Importantly further reduction of that compound I via ET would direct the reaction to the oxygen reduction pathway (Figure 9B, red arrow). The electrons required for generation of compound I from O₂ are derived from a reductase protein or an ET active site in the same protein and, importantly, the ET rates in Cyt P450 are substantially lower than those of CcO. The reductase components of Cyt P450 systems involve various combinations of flavin and iron–sulfur clusters and these transfer electron to the Cyt P450 active site at rates of around 10–500 s^–1.⁶² This is in contrast to the ET rates at the oxygen reduction site of CcO which are orders of magnitude greater (>10⁵ s^–1).

Figure 9.

(A) Active site of Cyt P450 (PDB ID: 1AKD)¹³⁰ with thiolate bound heme. (B) Schematic representation of the competition between monooxygenation (blue arrow) and oxygen reduction (red arrow) by CytP450 family of enzymes.

A key toward achieving Cyt P450 reactivity is, thus, attenuating the rate of ET from the reductase component to the catalytic center to allow monooxygenase activity in preference to reductase activity (Figure 9B). Keeping this in mind, site isolated thiolate ligated Iron porphyrin active site which mimics of the active site of Cyt P450, were fabricated atop a SAM covered Au electrode and electrochemical O₂ reduction was performed with that.^69,131 An electrocatalytic O₂ reduction (reductase activity) involves the same intermediates (Figure 9B) involved in substrate oxidation (oxidase reactivity) even in artificial systems as well. The only difference is that during electrocatalytic O₂ reduction the high valent oxidants are reduced by ET from the electrode which acts as a mimic of the reductase component in the enzymatic systems. This raises the possibility of harnessing the oxidase activity of these high valent intermediates to oxidize organic molecules if their rates of formation via O–O heterolysis during O₂ activation can be made much faster than their rates of decay (by electron transfer from the electrode), offering them a substantial lifetime to oxidize substrates present in the medium. The rate of electron transfer from the electrode to the iron porphyrin site was attenuated using SAMs of different chain lengths to derive monooxygenase reactivity from these constructs which otherwise show ORR activity at fast (∼10⁶ s^–1) ET rates.

Catalytic hydroxylation of strong C–H bonds (BDE > 100) to alcohol and epoxidation of alkenes, using molecular O₂, with this electrochemical construct were demonstrated with turnover numbers > 10⁴. Maximum yield was achieved when the gold electrode was modified with an octanethiol SAM which shows moderate electron transfer rate (10³ s^–1). Faster electron transfer than that with alkane thiols of shorter chain length resulted in reduced monooxygenase activity which was evident from the decreased yield of the oxidized product. Thus, tuning the electron transfer rate during the electrochemical ORR, molecular O₂ can be efficiently harnessed for useful chemical transformation with synthetic molecules. Several Cyt P450 enzymes are known to exhibit k_H/k_D as high as 10 for HAT by compound I during the enzymatic process. The high H/D isotope effect (∼10) for cyclohexane (sp³ C–H) oxidation by this bioinspired electrocatalyst falls in the observed range expected for rate determining hydrogen atom transfer (HAT) reactions (Figure 10) by a compound I intermediate.^132−136 Since these reactions were performed in an aqueous medium, it provided a unique opportunity to explore the H₂O/D₂O KSIE in these reactions in a synthetic system. The oxidation of cyclohexane by oxygen catalyzed by this electrochemical analogue has a H₂O/D₂O isotope effect of 12.8 ± 0.1. The large KSIE was associated with the heterolytic cleavage of the O–O bond in a Fe^III–OOH intermediate owing to a lack of organized proton transfer channel mentioned in the previous section.⁶⁷ Similar observation during the electrochemical ORR was also found in a previous investigation. For a combination of SERRS-RDE and H/D isotope effects on rate determining steps in that electrochemical O₂ reduction, a KSIE as high as 18 was observed for an iron porphyrin complex with a covalently bound thiolate without any preorganized proton channel in the distal geometry.⁶⁷ This suggests that the rate of O–O bond heterolysis, which depends on the availability of protons, becomes a decisive factor during the electrochemical ORR as well. Similarly, high KSIE of 15 was observed when the highly conserved aspartate residue asp251 in P450cam was mutated.¹³⁷

Figure 10.

Schematic representation of the monooxygenase activity depicting both the kinetic solvent isotope effect (KSIE) and kinetic isotope effect (KIE) during the oxygenation of C–H bond.

Apart from that, this electrochemical construct was seen to exhibit an unusual selectivity. An apparent disparity with the erstwhile reported synthetic systems regarding the preferential oxidation of 2° C–H over 3° C–H and lack of overoxidation of the sterically demanding 2° alcohol produced was noted (Table 2).^138−140 This was rationalized by the steric hindrance provided by the distal pocket as decorated by flanking, peripheral substituents of the synthetic porphyrin electrocatalyst. The four bulky pivaloyl groups in the distal structure provides a very narrow opening for the substrates to access the high valent intermediate (Figure 11). This narrow opening restricts the access of sterically demanding 3° C–H bonds over lesser sterically demanding 2° C–H bonds accounting for the preferential 2° C–H bond oxidation over 3° C–H bonds. Decreasing the steric hindrance by decreasing the number of peripheral substituents diminished the unusual selectivity of this electrocatalytic analogue of Cyt P450, supporting the steric control of selectivity proposed (Figure 11D).¹³¹

Table 2. Difference in Regioselectivity.

Figure 11.

(A) Top view of iron-picket fence porphyrin (FePf) showing a substrate access cavity with a diameter of 6 Å. (B) The access of tertiary H of adamantane demands a conelike geometry resulting in steric hindrance. This accounts for the lesser selectivity toward tertiary C–H bond oxidation. (C) Secondary H accesses the cavity with a lesser amount of steric hindrance because of its trigonal architecture, resulting in more selectivity toward secondary C–H oxidation despite the higher BDE. (D) Wider substrate access cavity of iron-half picket fence (FehPf) depicting the lesser amount of bulky pivolyl groups in the secondary geometry.

No Electron Transfer and No Proton Transfer: The Case of Heme Dioxygenase

Heme dioxygenases promote incorporation of both the oxygen atoms of molecular oxygen into organic substrates.^37,141 Indoleamine 2,3-dioxygenase (IDO) and tryptophan 2,3-dioxygenase (TDO) are two heme based dioxygenases which play a crucial role in the physiological regulation of tryptophan (l-Trp) flux by catalyzing the initial and bottleneck step of l-Trp metabolism in the kynurenine (Kyn) pathway. This step involves the oxidative cleavage of the 2,3-double bond in the indole moiety of l-Trp, resulting in the formation of N-formyl Kyn. Both of the active sites of IDO and TDO contain the heme cofactor, proximally ligated to an imidazole residue and the active site is harbored inside a hydrophobic cavity (Figure 12A). Spectroscopic and analytical characterization of intermediates suggests that the mechanism of dioxygenation begins with the electrophilic attack of heme superoxide (Fe^III–O₂^•–) to the C² or the C³ position of the substrate, producing epoxides and a ferryl intermediate. This ferryl intermediate then continues further to produce a 4e^– oxidized product (Figure 12B). Apart from the initial reduction from ferric heme to its ferrous state, unlike the monooxygenases, this reaction proceeds without any proton and electron as it regenerates the ferrous heme species in the end of the catalytic process. The active site is hydrophobic as indicated by ENDOR (electron–nuclear double resonance) data.¹⁴² The physiological reductase component for the dioxygenases is not known with certainty, but for IDO there is some evidence of it being cytochrome b5.^143,144

Figure 12.

(A) O₂ bound active site of heme dioxygenase (PDB ID: 5TI9).¹⁴⁵ (B) Schematic representation of the formation of NFK by heme dioxygenase enzyme.

Replicating the reactivity of heme dioxygenase by generating Fe^III–O₂^•– to achieve 4e^– oxidation of organic substrates can be a lucrative chemical reaction. A major complication involved is that most reactions of ferrous porphyrins with O₂ result in ferric porphyrin, and attempts to reduce the ferric porphyrin in the presence of O₂ result in uncontrolled production of partially reduced O₂ species (PROS) like O₂^–- and H₂O₂. However, in a recent report by Singha et al,. a Fe^III–O₂^•– porphyrin was demonstrated to oxidize pendant quinols by 4 e^– oxidized quinone molecules in homogeneous conditions (Figure 13A).¹⁴⁶ The presence of Fe^III–O₂^•–, Fe^III–OOH, and Fe^IV=O intermediates is identified by employing EPR and resonance Raman spectroscopy during the course of reaction. A series of ferric heme superoxo oxidants was reported to oxidize an array of indole substrates into their corresponding 2,3-dioxygenated products in solution (Figure 13B).¹⁴⁷ For the reasons described above, these transformations were not catalytic.

Figure 13.

Schematic representation of (A) reaction of a pendant quinol molecule with a heme superoxide complex and (B) reaction of a synthetic heme superoxide adduct with an array of indole substrates in solution.

Achieving the 4 e^– oxidation of substrates using Fe^III–O₂^•– in homogeneous conditions does not often ensure multiple turnovers, as the unreactive ferric porphyrin species is easily produced via the solvolysis of the Fe^III–O₂^•– species.¹⁴⁶ This can be remedied by performing the catalytic process in heterogeneous conditions. As stated earlier, an Fe^III–O₂^•– species, the putative reactive intermediate of the heme dioxygenase enzyme, is an intermediate in heterogeneous electrochemical O₂ reduction by iron porphyrins too and its population, under electrochemical conditions, can be tuned by controlling the applied potential. Recently, iron porphyrins having different axial ligands have been constructed on a self-assembled monolayer of thiols on a gold electrode, and electrochemical oxygen reduction reaction was performed with those constructs.¹⁴⁸ The potential applied was held between E₁ and E₂ (Figure 3) to allow maximum population of the Fe^III–O₂^•– species. During that process, dioxygenation of 3-methylindole and 4e^– oxidation of a series of organic compounds having C–H bond dissociation energies (BDE) between 80 and 90 kcal/mol was achieved, with very high turnover, at potentials where Fe^III–O₂^•– species are likely to be formed on the electrode (Figure 14A, Table 3). In contrast to the monooxygenation reaction, the involvement of the high valent compound-I-like species during this dioxygenation process was rescinded as evident from the fact that these electrochemical constructs were unable to oxidize substrates having BDE > 90 kcal/mol at the same potential. Mechanistic investigations through isotope effect calculations suggests that hydrogen atom transfer (HAT) by Fe^III–O₂^•– is likely to be the rate determining step (Figure 14B). The lack of KSIE and insensitivity to ET rates (observed during monooxygenase activity) advocates for a 4e^– oxidation catalyzed by these constructs using Fe^III–O₂^•–. These results highlight a new reaction engineering approach to harness O₂ as a green oxidant for efficient chemical oxidation.

Figure 14.

(A) Pictorial depiction of the electrochemical construct performing the dioxygenation of indole. (B) Schematic representation of the mechanism involved in the process of 4e- oxidation.

Table 3. Oxidation of the Substrates with Low BDE: Turnover Numbers (TON), Turnover Frequencies (TOF), Maximum TON Possible for Monooxygenase Reaction (TON_max), and ratio of TON and TON_max (μ) for the Products Obtained after the Oxidation of Substrates by Differently Ligated FePf Complexes.

^aL is the axial ligand.

^bTON implies turnover number.

^cTurnover frequency averaged over the entire duration of the experiment.

^dMaximum turnover expected from monooxygenase reactivity and.

^eThe ratio of b and d. The >1 μ value clearly indicates a reaction mechanism other than monooxygenation.

Fast Proton Transfer and No Electron Transfer: The Case of Heme Peroxidase

Unlike heme oxygenases, heme peroxidases engage themselves in useful chemical transformations through oxidation and oxygenation by wielding the oxidizing power of peroxides.^39,149 Among the various activities, biosynthesis of prostanoids,¹⁵⁰ defending plants and animals against pathogens,¹⁵¹ and bioregulation of free peroxides to protect organisms from oxidative damage¹⁵² are a few worth mentioning. The prolific reactivity of peroxidase is due to the generation of the versatile high valent compound I intermediate upon the reaction of peroxides with resting ferric heme active site. Generally, reaction of heme enzymes with hydrogen peroxide generates a Fe^III–OOH species which, after the heterolytic cleavage of the O–O bond, forms compound I.¹⁵³ The facile generation of compound I in this process depends on the participation of the distal residues such as Arg and His (Figure 15).^154,155 These residues contribute to the facile heterolytic cleavage of the O–O bond by an acid–base catalysis mechanism. This compound I oxidizes substrates by one electron, forming compound II which in turn participates in a second one electron oxidation process and regenerates the resting ferric species. Apart from accelerating the facile formation of compound I, the distal Arg residue is also proposed to stabilize both the compound I and II intermediates by hydrogen bonding to the ferryl oxygen. Thus, peroxidases participate in the oxidation procedure via the same intermediate as Cyt P450 without the requirement of any reducing equivalent. Note that a group of Cyt P450 enzymes, known as peroxygenases, use peroxides as a surrogate of O₂ not only in the bacterial system but also in several developed organisms like human beings primarily if the electron transfer to the active site becomes unfavorable. Though the use of peroxides can lead to high inefficiency and inactivation of enzymes, there are many examples where peroxides have been used effectively in α–β hydroxylation, decarboxylation, and oxidation of the cyano groups.^42,156,157

Figure 15.

(A) Active site of heme peroxidase (PDB ID: 2CYP).¹⁴⁹ (B) Interaction of the distal Arg and His with Fe^III–OOH.

The critical role of the Arg residue in the enzymatic activity of heme peroxidase enzymes has always been a matter of interest to the scientific community.^14,155 Recently, a synthetic iron porphyrin (Fe-MARG) was reported where a pendant guanidine residue, mimicking the Arg residue in the enzymatic site, was strategically placed on top of the active site.¹⁰² This synthetic design achieves the goal of incorporating a distal guanidinium residue in an iron porphyrin mimicking the Arg in the active site of HRP. The crystal structure of the free ligand indicates a suitable orientation of the guanidinium group with respect to the metal binding site in the porphyrin. The Fe-MARG forms compound I (Figure 16), verified by mechanistic and spectroscopic analysis, using organic peroxide which rapidly oxidizes phenolic and amine substrates with high rates. Unsubstituted iron TPP does not show either rapid peroxidase activity or formation of compound I in the same time scales, illustrating the advantage of the pendant guanidinium (protonated at neutral pH) group. The guanidinium group is a strong hydrogen bond donor and can hydrogen bond to the substrates used. This strong H-bonding interaction is responsible for enzyme-like Michalis-Menten kinetics exhibited by this complex. However, the activation of peroxide is not limited to a guanidium group in the distal pocket, any pendant basic residue that are protonated at neutral pH and potent to form second sphere hydrogen bonding interactions, results in facile formation of compound I from peracids. Short-lived Compound I intermediates are formed from peroxides and characterized at cryogenic temperatures using UV–vis electronic absorption spectroscopy and EPR spectroscopy with synthetic iron porphyrins having pendant aliphatic amine in the distal position (FeL₃, Figure 16B) having pK_a close to that of guanidine.⁹⁹ The advantage of distal basic residue which remain protonated in the neutral pH was assessed with Density Functional Theory (DFT) calculations. DFT calculations indicate that these protonated residues result in the polarization of the electron density of the O–O σ-bond and thus results in facile O–O bond heterolysis. These artificial constructs could catalyze the oxidation of Ferrocene, 2,4,6-Tri tertiarybutyl phenol (TBPH), o-Phenylenediamine, 3,3′,5,5′-Tetramethylbenzidine with TOF of 2.7, 1.4, 0.85, and 25 s^–1 respectively.^99,102

Figure 16.

Peroxidase activity of synthetic heme (A) Fe-MARG and (B) Fe-L₃.

Formation of compound I was envisaged during the H₂O₂ disproportionation reaction with phenolate ligated iron porphyrin complexes.¹⁵⁸ Compound I has been detected using in situ SERRS during the H₂O₂ disproportionation (HPD) reaction by phenolate ligated iron porphyrin immobilized on self-assembled monolayer modified Au electrodes where Fe–O vibration of the compound I species were detected at 796 and 803 cm^–1 for porphyrins containing hydrogen bonding (Fe-tetraester porphyrin, FeEs₄) and non-hydrogen bonding (Fe-“picket fence” porphyrin, FePf) distal structures, respectively. Compound I, so generated, was used to oxidize range organic substrates in aqueous medium.¹⁵⁹ The in situ generated compound I, with each of the axially ligated constructs, via a peroxide shunt, abstracts a hydrogen atom from the C–H bond as is evident from the recorded KIEs. Variation of the axial ligand from thiolate to phenolate to imidazole during the oxidation not only result in the decrease in the yield but also result in significantly higher isotope effect than thiolate ligation. These higher KIE values in phenolate and imidazole ligated constructs likely reflect the involvement of HAT transition states (TS) having different polarity.¹⁶⁰ For example, a higher KIE for FePf with an anionic axial ligand (phenolate) than for that with a neutral axial ligand (imidazole) is consistent with the hypothesis of Shaik et al.¹⁶⁰ The barrier of hydrogen atom tunneling depends on the electrostatic stabilization of the H^δ+ as it moves between the C^δ− and O^δ− (Figure 17B) moieties in the transition state (TS).¹⁶⁰ The neutral imidazole axial ligand results in a lesser electrostatic stabilization in the TS than the anionic phenolate axial ligand in the rate-limiting step. Thus, the lesser electrostatically stabilized transition state for imidazole bound compound I goes through a wider tunneling barrier and thus results in lower KIE values than the more electrostatically stabilized TS of phenolate bound compound I.

Figure 17.

(A) Tunneling lowers the observed barrier relative to the semi classical TS. (B) Schematic representation of the TS where L represents the different axial ligations.

Electron Transfer and No Proton Transfer: The Case of Cytochrome c

Heme enzymes that only have a high electron transfer rate to the active site are the electron transfer proteins like cytochrome c.¹⁵ The Fe-porphyrin active site on top of a SAM modified gold electrode was demonstrated by Bandyopadhyay et al. where spectroscopic characterization coupled to electrochemical and theoretical calculation provided a detailed understanding of the factors affecting the ET kinetics in the synthetic Fe-porphyrin systems.¹⁶¹ Detailed discussion on this topic is mindfully avoided here as it is not the focus of this Article.

Outlook

The ability to control the ET and PT pathways to iron porphyrins which mimic the active site of different heme enzymes has led to the realization that by manipulating the ET and PT supply the function of the same iron porphyrin site may be controlled. Such control of branching of different reactive pathways has been demonstrated in heme enzymes, with monooxygenase vs peroxygenase being a great example of that.^40,162 In these heterogeneous constructs, SAM chain length and porphyrin ligand design allow attenuation of ET and PT pathways, respectively. The branching or reactivity by controlling the ET and PT is also enacted in these constructs. For example, a thiolate bound iron porphyrin can reduce O₂ to H₂O when ET and PT are fast. The same system acts as a monooxygenase when the ET is slowed down and dioxygenase when ET is utilized only to reduce the ferric state to the ferrous state. Similarly, iron porphyrins with active proton shuttles can reduce O₂ to H₂O when ET is fast while they act as peroxidases when H₂O₂ is used and ET is blocked. Thus, in addition to the generally held convictions regarding the role of axial ligands and distal environment, biomimetic chemistry should be cognizant of the fact that the ET and PT steps are factors that contribute to the diversity of reactivity exhibited by heme active sites as well. While the individual oxygenase, oxidase, peroxidase, and peroxygenase activities are optimal only when proper combinations of axial ligands and distal residues are used, the ability to use these heterogeneous bioinspired constructs to catalyze O₂ reduction, monooxygenase, dioxygenase, catalase, and peroxidase (all with TON > 10³ at room temperature and in aqueous solvent) simply by controlling the ET rate, nonetheless, opens up a new versatile reaction engineering approach toward O₂ activation for useful chemical transformations.

The authors declare no competing financial interest.