A divergent mitochondrial cytochrome c in malaria parasites with an anomalously low redox potential

Abstract

Eukaryotic cytochrome (cyt) c is a highly conserved mitochondrial protein central to cellular respiration, featuring a covalently attached heme with hexacoordinate iron whose redox potential is tuned by axial His/Met ligands and surrounding residues to support electron transport chain (ETC) function. We have identified a previously unrecognized lineage of eukaryotic cyt c homologs in Apicomplexa, a phylum of intracellular pathogens that includes Plasmodium falciparum malaria parasites. P. falciparum cyt c-2 (Pfcyt c-2) exemplifies this divergent lineage and displays heterogeneous heme coordination despite conservation of His/Met ligands. We determined that Pfcyt c-2 has a redox potential of –278 ± 6 mV that is over 500 mV lower than canonical mitochondrial cyt c homologs (+250 mV) and contradicts a conserved ETC role. This anomalous redox potential is lower than any natural monoheme c-type cyt and reveals exceptional evolutionary tuning for a eukaryotic cyt c. Although Pfcyt c-2 displays signs of structural heterogeneity, it retains low-level peroxidase activity and global protein thermostability comparable to canonical cyt c proteins. These results reveal a new clade of eukaryotic cyt c proteins with divergent biochemical properties and biological roles, opening new scaffolds for mechanistic discovery and redox engineering of natural and abiological functions.

Introduction

Heme is an ancient biological cofactor whose ability to bind diverse ligands and to gain or lose an electron has been exploited by nature to catalyze a range of key biochemical transformations, to sense and transport gases, and for biological signaling. The rich chemical properties of heme and their tunability by diverse ligands and protein scaffolds have also been exploited to engineer artificial catalysts that carry out a range of biochemical and abiological reactions, with applications in biomedicine, biotechnology, industrial chemistry, and energy conversion.

Cytochrome c has been a biochemical workhorse for understanding heme functionality. C-Type cytochromes (cyt) are defined by covalent heme attachment through their conserved CXXCH motif but span diverse structural classes [1, 2]. Eukaryotic mitochondrial cyt c represents the archetypal Class I c-type cyt [3]. Its hexacoordinate His/Met ligation enforces a low-spin heme with minimal electronic rearrangement between oxidation states, yielding a redox potential around 250 mV that enables electron transfer between Complex III and Complex IV in the mitochondrial ETC [4, 5]. Although cyt c also functions as a peroxidase in animals during cellular apoptosis, these roles are secondary to its role as an electron shuttle. Compared with other heme proteins, such as cytochrome P450s that catalyze diverse oxidative reactions, cyt c proteins are generally viewed as chemically conservative scaffolds [6, 7].

Plasmodium falciparum, the causative parasite of severe malaria, provides a striking contrast to other eukaryotic organisms. These parasites encode two cyt c paralogs, a canonical Pfcyt c (Uniprot Q8IM53) with conserved ETC function and a highly divergent Pfcyt c-2 (Uniprot Q8I6T6) that features extensive changes in the heme crevice residues proximal to the conserved axial Met (Met81, using bovine numbering) [8]. Pfcyt c-2 is hemylated on its CXXCH motif by the parasite holocytochrome c synthase (HCCS) and trafficked to mitochondria [8, 9]. Despite conservation of axial His/Met ligands, oxidized Pfcyt c-2 displays spectroscopic features (by UV-vis and electron paramagnetic resonance) that suggest heterogeneous heme coordination and spin states as well as an ability to bind external ligands, including imidazole and H₂O₂ [8]. In the reduced state, however, UV–vis spectra of Pfcyt c-2 resemble those of a low-spin hexacoordinate heme with undefined axial ligands [8].

These observations and X-ray structural studies, which revealed preferential crystallization of a domain-swapped dimer (DSD) over the monomeric form that predominates in solution, suggest that structural heterogeneity in the heme crevice region destabilizes axial coordination by Met81 [8]. The Pfcyt c-2 DSD demonstrated reciprocal swapping of C-terminal helices between dimerized Pfcyt c-2 subunits, perturbed Met81 positioning, and axial heme coordination by water (or hydroxide) in place of Met81. Although the relation of these crystallographic features to monomeric Pfcyt c-2 remains uncertain, the DSD structure confirmed covalent heme attachment, proximal His19 ligation, and retention of secondary structure elements typical of cyt c proteins (Fig. S1, PDB: 7TXE) [8]. This combination of structural and spectroscopic features makes Pfcyt c-2 highly unusual for a mitochondrial cyt c and suggests that it represents a distinct branch of cyt c variants with divergent chemical properties and cellular functions. We therefore set out to unravel the molecular phylogeny and determine the chemical redox properties of this exceptional cyt c.

Results

Divergent molecular phylogeny of Pfcyt c-2.

To trace the evolutionary origins of the parasite cyt c paralogs, we queried the canonical Pfcyt c and divergent Pfcyt c-2 across all major eukaryotic lineages. Cyt c-2 homologs were detected only for intracellular parasites in the phylum Apicomplexa and their closest free-living relative, Vitrella brassicaformis, suggesting that this molecular divergence accompanied the evolutionary appearance of parasitism [10]. Comparative analyses also revealed an unrecognized clade, which we term cyt c-1.5, that branches between cyt c and c-2 (Fig. S2). Sequence similarity networks resolved three distinct clusters (c, c-1.5, c-2), and phylogenetic reconstruction indicates that cyt c-1.5 and c-2 arose from two sequential duplication events (Fig. 1A, B) [11–13]. Many apicomplexan lineages, including Plasmodium parasites, appear to have lost cyt c-1.5 and retained only cyt c-2.

Figure 1.

Bioinformatic analysis of cyt c-2 homologs. Sequences of cyt c, c-1.5, and c-2 were analyzed by (A) sequence similarity network (54% identity threshold), (B) phylogenetic tree, and (C) heme crevice sequence logo. Blue and red depict positively and negatively charged residues, respectively. White stars indicate loss of structural proline/glycine, and the black star marks an inserted residue. (D) AlphaFold model of Pfcyt c-2 bound to heme c, highlighting possible mispositioning of Met residue with lone-pair electrons on S atom predicted to point away from heme Fe and unfavorably long S-Fe distance as labeled (N and C-termini truncated, full model provided in Fig. S5). Sequence similarity networks were generated with the EFI-EST webserver (https://efi.igb.illinois.edu/efi-est/) and visualized in Cytoscape (https://cytoscape.org/). A maximum-likelihood phylogeny was generated with IQ-TREE v2 and visualized using the Interactive Tree of Life (iTOL) webserver.

Multiple sequence alignments highlight insertions between α helices α1 and α2 (Fig. S2). The heme crevice is the most conserved region of cyt c proteins and shows striking variability in cyt c-2 homologs, including charge reversals, disruption of structural residues, and an unusual insertion of an additional amino acid (Fig. 1C) [14, 15]. Multiple residues that stabilize Met81 coordination have diverged, consistent with structural and spectroscopic studies suggesting heterogenous heme coordination in oxidized Pfcyt c-2. Despite alterations in the heme crevice region of Pfcyt c-2, hallmark features of canonical cyt c remain intact, including Met81 and residues required for folding, HCCS recognition, and porphyrin-ring contacts (Fig. S3) [3, 15].

Because a monomeric crystal structure was not obtainable, we modeled 16 unique cyt c-2 proteins with heme c using AlphaFold3 (Fig. S4) [16] [17]. The majority of cyt c-2 models maintained a general cyt c fold but lacked stable Met–Fe coordination and were predicted in one of three states: (i) Met81 proximal but unbound, (ii) Met81 partially displaced, or (iii) Met81 outside/adjacent to the heme crevice (Fig. S4 and S5). Pfcyt c-2 was predicted to adopt state (i), in which the Met81 sulfur lies 3.3 Å from the heme iron with its sulfur lone-pair electrons rotated away from the metal center and thus mispositioned to coordinate iron (Fig. 1D). These models are not definitive but provide hints of divergent active site features that enable testable hypotheses regarding altered redox properties.

Anomalous redox potential of Pfcyt c-2.

Perturbations to axial heme coordination and the local active site environment of Pfcyt c-2 suggested alterations to its redox properties. Using comparative dye-based reduction by UV–vis absorbance [18], we first determined midpoint potentials of 237 ± 12 mV for bovine cyt c and 216 ± 3 mV for Pfcyt c, which were similar to potentials of other mitochondrial cyt c homologs and consistent with their canonical roles in the ETC (Fig. S6) [19–21]. By contrast, Pfcyt c-2 was not reducible under the same conditions (Fig. S7), prompting us to test a panel of reductants (Fig. S8). Pfcyt c-2 could only be reduced by sodium dithionite (E° ≈ –660 mV) but not by other reductants of higher redox potential (Fig. S8), indicating an exceptionally low redox potential. Guided by this observation, we measured a midpoint potential of –278 ± 6 mV for Pfcyt c-2, a value independently confirmed using two distinct dyes (Fig. 2, Fig. S9) [18].

Figure 2.

Determination of the redox potential of Pfcyt c-2. (A) Representative UV–vis spectra of the oxidized (pink) to reduced (blue) transition during equilibrium titration of Pfcyt c-2 and safranin dye. Arrows mark wavelengths (420, 490 nM) used to calculate oxidized/reduced fractions for Nernst analysis. (B) Midpoint redox potentials of bovine cyt c, Pfcyt c, and Pfcyt c-2 (WT and mutants) shown as the mean of three independent determinations ± SD (versus standard hydrogen electrode). Data for all proteins are shown in supplementary material.

This anomalously low redox potential represents a >500 mV decrease relative to the conserved midpoint potential near 250 mV in canonical mitochondrial cyt c homologs. Such a low potential is incompatible with ETC electron shuttling and lies far below the previously lowest reported monoheme cyt c, Geobacter sulfurreducens PccH (–24 mV) [22]. To our knowledge, Pfcyt c-2 has the lowest redox potential described for any naturally occurring monoheme c-type cytochrome [22–24]. Comparable values have only been achieved in artificially engineered cyt c Met-to-Cys mutants, where axial replacement with Cys introduces an electron-donating thiolate ligand that stabilizes the Fe(III) state and drives the potential downward, or in multiheme cytochromes, where electronic communication between closely packed heme centers redistributes charge and lowers the midpoint potential through cooperative redox coupling [25–27].

Given the high sequence divergence of Pfcyt c-2 (Fig. 3A) and evidence that the distal Met81 does not stably coordinate heme in the oxidized state, we asked whether this residue might still contribute to Pfcyt c-2 hemylation by HCCS and/or the low redox potential as a basis for its conservation. An M81A Pfcyt c-2 mutant was efficiently recognized and hemylated by human HCCS (Fig. 3B and Fig. S10) [28]. UV–vis spectra of the oxidized and reduced forms of M81A showed only subtle differences (Fig. 3C, D), including mild broadening of the Soret peak, and the oxidized spectra of WT and M81A in the presence of imidazole were nearly superimposable (Fig. S11). The mutant displayed a small but reproducible 20 mV decrease in redox potential to −302 ± 5 mV (Fig. 2B and Fig. S12), in contrast to the >250 mV decrease previously observed for the corresponding mutant in human cyt c [28]. This nominal decrease in Pfcyt c-2 potential strongly suggests that Met81 likely modulates the heme environment indirectly rather than serving as a stable axial ligand. Indeed, a 695 nm peak indicative of Met-Fe coordination [3] was absent in UV-vis spectra of oxidized Pfcyt c-2, although such a peak was readily detected for oxidized bovine cyt c (Fig. 3C). These findings also argue against a role for the conserved Met81 as an axial ligand in the reduced state but instead suggest that it contributes structurally to shaping the electronic properties of the heme crevice. This suggestion is consistent with AlphaFold predictions, which place Met81 within or adjacent to that region but misposition it to coordinate the heme iron.

Figure 3.

Mutational analysis of Pf cyt c-2. (A) Partial sequence alignment of bovine and P. falciparum cyt c and cyt c-2 centered on conserved Met81 (purple) and showing sequence divergence in Pfcyt c-2, including His78 (red) and Lys84 (red) tested by mutation. (B) Expression and hemylation of Pfcyt c-2 (WT and mutants). Each protein was co-expressed in E. coli with human HCCS with N-terminal His₆ tag (cyt c) or GST tag (HCCS). Protein concentrations in lysates were quantified by Lowry method for equivalent loading and fractionation by SDS-PAGE and transfer to membrane. Pfcyt c-2 protein levels were visualized by anti-His-tag western blot (WT and M81A) or Coomassie stain (H78A and K84A). Bound heme was visualized by chemiluminescent heme stain. Uncropped gel images are shown in supplementary material. (C) UV-vis absorbance spectra of WT and mutant Pfcyt c-2 proteins oxidized by ammonium persulfate. Inset shows 695 peak indicative of Met-Fe ligation in WT bovine cyt c and its absence in Pfcyt c-2 in oxidized state. (D) UV-vis absorbance spectra of WT and mutant Pfcyt c-2 proteins after reduction by sodium dithionite. Spectra were normalized to the Soret peak for each protein and offset vertically for clarity. Samples contained 8–12 μM protein in 50 mM phosphate buffered saline at pH 7.4. Spectra for the inset in (C) were obtained using 100 μM protein.

We also tested possible roles for the active site His78 or Lys84 sidechains that are specific to Pfcyt c-2 and located three residues N- or C-terminal to Met81, respectively (Fig. 3A). Substituting the axial Met with His or Lys has been shown to strongly diminish redox potential in human bis-His cyt c (−70 mV) or in the Lys/His state engineered by Met-to-Lys mutation (−94 mV) or adopted in the alkaline transition in WT cyt c at elevated pH [3, 28, 29]. Both H78A and K84A mutants were hemylated by HCCS (Fig. 3B and Fig. S10), but neither mutation substantially altered UV-vis spectra of oxidized or reduced Pfcyt c-2 (Fig. 3C, D) or reduction potentials (H78A: −287 ± 3 mV; K84A: −281 ± 4 mV) compared to the WT protein (Fig. 2B and Fig. S13). These results argue against a primary role for these residues as axial ligands.

Finally, we acquired UV-vis spectra of WT Pfcyt c-2 and determined its redox potential at pH 5.5 to test a possible role for bound hydroxide as an axial ligand. Density to support an axial water or hydroxide was observed in the prior Pfcyt c-2 DSD structure [8], although its presence in monomeric Pfcyt c-2 is uncertain. Hydroxide has previously been shown to replace Met81 as axial ligand in the M81A mutant of human or bovine cyt c and to substantially lower the redox potential (−20 mV) in a pH-dependent manner [28, 30, 31]. UV-vis spectra of oxidized and reduced Pfcyt c-2 and redox potentials were nearly identical at pH 5.5 and 7.4 (Fig. S14), providing no evidence for pH-dependent changes in this range expected for a bound hydroxide. It remains possible that any hydroxide occupancy is governed by proton transfer outside this pH range or that water rather than hydroxide plays a role as axial ligand.

Pfcyt c-2 retains canonical peroxidase activity and thermostability.

In mammals, cyt c acquires pro-apoptotic functions when Met81 is displaced by proteins or lipids such as cardiolipin [3, 6]. This disruption of the native Met ligation exposes the heme to unlock low-level peroxidase activity, a property also reproduced in synthetic cyt c mutants [32]. Because Pfcyt c-2 lacks stable Met81 ligation and can natively bind exogenous ligands, we posited that it might display enhanced peroxidase activity. To test this hypothesis, we performed kinetic assays with H₂O₂ and guaiacol that revealed that bovine cyt c, Pfcyt c, and Pfcyt c-2 all had very similar k_cat values of 0.17 s⁻¹, 0.07 s⁻¹ and 0.10 s⁻¹, respectively (Fig. 4 and Fig. S15). Thus, despite evidence for enhanced heme accessibility and ligand binding, Pfcyt c-2 does not show evidence of increased peroxidase activity typical of higher eukaryotic cyt c proteins whose Met coordination is naturally or artificially disrupted [32, 33]. Heightened heme crevice flexibility and active-site accessibility [8] are therefore insufficient to enhance peroxidase function (with guaiacol) beyond that expected for a cyt c with axial His coordination [34, 35].

Fig. 4.

Determination of Pfcyt c-2 peroxidase activity and stability. Michaelis-Menton plot of peroxidase activity for bovine cyt c, Pfcyt c, and Pfcyt c-2 with varying guaiacol concentrations. Each point is the mean V_o of three independent replicates. Inset shows mean k_cat values determined from assays at three different enzyme concentrations (Fig. S15).

The extensive sequence insertions in Pfcyt c-2 and the inability to crystallize a physiological monomer raised the question of whether the protein might be destabilized relative to typical cyt c proteins. Thermal melt analysis by circular dichroism revealed a temperature-dependent unfolding transition in Pfcyt c-2 with a midpoint T_m of 75 °C, only slightly below the 82 °C observed for bovine cyt c, indicating that the Pfcyt c-2 fold remains thermostable (Fig. S16) despite evidence for local static or dynamic structural disorder in the heme crevice region [8, 28].

Discussion

Pfcyt c-2 has evolved sequence, structural, and biochemical features that are highly unusual for a mitochondrial cyt c. These features include loss of axial coordination by Met81, which leads to heterogeneous heme coordination and spin states, and an anomalously low reduction potential approaching −300 mV. We propose that these properties are distinguishing features of this newly identified cyt c-2 lineage present in other apicomplexan parasites, including the human pathogens Toxoplasma gondii and Babesia spp. Results in this and our prior study argue against a model that enhanced peroxidase activity or roles in oxidative protection underpin the cellular function of Pfcyt c-2 [8]. Instead, the open heme pocket and strongly shifted redox chemistry point toward alternative functions, potentially binding or chemically transforming small molecules or gases in a signaling or catalytic capacity [7, 24].

Although Pfcyt c-2 is dispensable during blood-stage growth of malaria parasites within humans [8], its expression is strongly upregulated during parasite development inside mosquitoes [36]. This enhanced expression suggests a stage-specific role in the insect vector that we are actively testing. The retention of Pfcyt c-2 in Plasmodium parasites, which have otherwise substantially reduced their heme proteome, underscores durable evolutionary pressure to preserve the atypical sequence and functional properties of this bioinorganic scaffold and its contribution to parasite mitochondrial metabolism [8, 37–39]. It remains a key future challenge to understand the divergent biological function of Pfcyt c-2 (and related homologs) and the selective pressure to retain Met81 in the absence of detectable heme coordination.

The physical features of Pfcyt c-2 that underpin its unusually low reduction potential remain to be elucidated, but loss of stable Met81 coordination alone is only expected to reduce the redox potential ~250 mV based on mutagenesis studies of canonical cyt c proteins [27, 28, 40, 41]. Substitution of Met with other axial ligands such as Lys, His, Cys, or hydroxide can further reduce E°. However, there are no Cys residues in Pfcyt c-2 beyond the CXXCH motif, and we found no evidence that these other groups play dominant roles as axial ligands in shaping redox potential. Evidence for structural flexibility in the heme crevice loop and observation of both high- and low-spin features in EPR spectra of oxidized Pfcyt c-2 suggest heterogeneity in distal heme coordination and spin states [8] that may reflect roles for distinct axial ligands in conformational exchange.

Beyond heme coordination geometry, several structural and electrostatic features of the Pfcyt c-2 heme environment also likely contribute to its anomalously low redox potential. AlphaFold models of cyt c-2 homologs suggest a shortening of the His19-Fe bond compared to canonical eukaryotic cyt c proteins, a feature also observed in the DSD Pfcyt c-2 dimer, which would stabilize the Fe(III) state (Fig. S17) and may contribute to lowering the redox potential [14, 42, 43]. This structural prediction can be tested in future experiments by solution methods such as EXAFS. Continuum electrostatics calculations also suggest enhanced (≥twofold) negative charge within 10 Å of the heme relative to canonical cyt c homologs, a shift predicted to further lower the E° (Fig. S18) [44–46]. These active-site features together with loss of Met81 ligation and resulting heterogenous heme coordination may collectively contribute to the >500 mV decrease in redox potential and can be dissected and more rigorously defined in future structural, spectroscopic, and computational studies.

Although canonical cyt c proteins are specialized for electron transfer, directed-evolution studies demonstrate that this scaffold can be reprogrammed for abiological catalysis. Mutations supporting C-Si bond-forming activity disrupt the native Met–Fe ligation, generating a pentacoordinate heme with increased solvent accessibility and anionic charge at the distal face [47–49]. Notably, Pfcyt c-2 naturally exhibits several of these same features, which may similarly impart a predisposition toward noncanonical promiscuous reactivity that may be further refined through protein engineering for new applications in synthetic biology and/or industrial chemistry. More broadly, properties of Pfcyt c-2 can guide ongoing efforts to understand and engineer novel heme protein functionalities.

Materials and Methods

Bioinformatic analysis.

Canonical Pfcyt c (Uniprot: Q8IM53) and Pfcyt c-2 (Uniprot: Q8I6T6) were used as bait sequences to identify homologs by PSI-BLAST across major eukaryotic clades. Up to ten representative sequences per clade were selected for phylogenetic and sequence similarity network analyses. Protein sequences were aligned using Clustal Omega [50], and the resulting multiple sequence alignment was used as input for phylogenetic inference. A maximum-likelihood phylogeny was generated with IQ-TREE v2 [13] using ModelFinder for best-fit substitution model selection and node-support estimation via 1,000 ultrafast bootstrap replicates and 1,000 SH-aLRT replicates. The final tree was visualized using the Interactive Tree of Life (iTOL) webserver [51]. Sequence similarity networks were generated with the EFI-EST webserver (https://efi.igb.illinois.edu/efi-est/) and visualized in Cytoscape (https://cytoscape.org/). Application of a 54% sequence-identity threshold resolved three distinct protein clusters [11].

Structural Modeling.

Structural models of cytochromes identified in the bioinformatic analysis were generated using the AlphaFold3 server, with heme c included in each prediction [16]. Predicted structures were compared to the solved crystal structure of bovine cytochrome c (PDB: 2B4Z) [17] by using PyMOL’s align function to calculate the RMSD of the Cα atoms. Coordination bond distances between the iron atom and the axial ligands His19 and Met81 were measured. Electrostatic surface potentials were calculated using the APBS plugin (pH 7.0, 150 mM ionic strength). The net charge within a 10 Å radius of the heme iron was determined by summing the partial charges of all atoms within this sphere.

Recombinant Protein Expression and Purification.

BL21/DE3 E. coli were co-transformed with the pET28 plasmid (encoding PfCyt c; PfCyt c-2; or mutant PfCyt c-2 M81A, H78A, or K84A; with N-terminal His-tag) and the pGEX plasmid (encoding the human HCCS with N-terminal GST), as previously reported [8]. Hemylated PfCyt c and PfCyt c-2 were purified from cell lysates by Ni-NTA affinity chromatography. The His-tag was removed by digestion with thrombin, followed by SP ion-exchange chromatography, and gel filtration chromatography, as previously published [8]. Hemylated PfCyt c-2 M81A, H78A, and K84A were purified similarly but without ion-exchange chromatography. Coomassie-stained SDS-PAGE and LC-MS/MS were used to validate the purity and identity of the purified cytochromes, as previously published [8]. We separately tested purification of several proteins without His-tag removal and obtained identical UV-vis absorbance spectra and redox potentials.

Reduction Potential Determination.

The reduction potentials of hemylated cytochromes were measured using a previously described method [18]. Reactions were prepared in 50 mM potassium phosphate buffer at pH 7.0. Anaerobic conditions were established by adding glucose (5 mM final) and glucose oxidase (50 μg/mL final) and providing xanthine (300 μM final) as an electron source. Once O₂-free conditions were achieved, freshly prepared catalase (5 μg/mL final) and purified cytochrome (4 μM final) were added, and the mixture was thoroughly mixed.

A redox dye with a midpoint potential close to that of the cytochrome (e.g., dichlorophenolindophenol (DCIP, 217 mV), Neutral Red (−325 mV), or Safranine T (−289 mV) was then titrated into the solution to produce an absorbance peak similar in height to the cytochrome Soret band. Reduction of the dye–protein mixture was initiated by adding xanthine oxidase (50 nM final). Absorbance spectra from 350–750 nm were recorded every 30 s on a ThermoFisher Evolution 350 spectrophotometer until no further spectral changes were observed (~40 min). Excess sodium dithionite was added at the end of each experiment to obtain fully reduced spectra.

Data were fitted to a Nernst plot using wavelengths corresponding to the dye peak and the cytochrome Soret peak. For Neutral Red experiments, the dye showed substantial spectral overlap with the cytochrome Soret region. To account for spectral overlap, stepwise UV–vis spectra of Neutral Red undergoing reduction were collected, and linear regression was used to define the relationship between absorbance at 480 nm and 417 nm. This regression equation was then applied to each Pfcyt c-2 titration to estimate the Neutral Red contribution to the Soret peak at 417 nm, which was subsequently subtracted during analysis based on the measured absorbance at 480 nm. All reported potentials (E°) are referenced to the standard hydrogen electrode.

Cloning.

The M81A Pfcyt c-2 mutant was generated by PCR amplification from WT Pf cyt c-2 pET28a plasmid [8] using mutagenic primer sets 1/2 and 3/4 (1: cgcggcagccatatgaat, 2: tttgaagttagctcctatgtgtgactcaacaaattgttgtgg 3: ataggagctaacttcaaaggtttatcgaatttgcaagaaagg 4: gtggtggtggtggtggtg) that introduced the M81A substitution. H78A and K84A Pfcyt c-2 were ordered as synthetic gBlocks from Integrated DNA Technologies (idtdna.com). The amplified or synthetic DNA inserts were cloned into pET28a using ligation-independent insertion (NEB E2621) between the NcoI and XhoI restriction sites, positioning the gene in-frame with the vector’s N-terminal 6x His tag. The final constructs were verified by whole plasmid sequencing (Plasmidsaurus).

SDS-PAGE and western blot analyses.

For analysis of Pfcyt c-2 hemylation in E. coli, BL21(DE3) bacteria co-expressing human HCCS-GST (encoded by pGEX vector with N-terminal GST tag) and either wild-type or mutant Pfcyt c-2 (encoded by pET28a vector with N-terminal His6 tag) were grown at 37°C in LB media until reaching an optical density at 600 nm of 0.5. Cultures were induced with 1 mM IPTG (GoldBio I2481C50) and 100 μM 5-aminolevulinic acid, then grown overnight at 20°C. Bacteria were harvested by centrifugation, washed, and resuspended in 1× PBS (137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, 1.8 mM KH₂PO₄, pH 7.4) supplemented with 1% Triton X-100 (v/v) and protease inhibitor cocktail (ThermoFisher J61852.XF). Bacteria were lysed by sonication (20 pulses, 50% duty cycle, 50% power) on a Branson sonicator equipped with a microtip, followed by incubation at 4°C on a rocker for 1 hour. Lysates were clarified by centrifugation (10 min at 14,000 rpm). Equal volumes of wild-type and mutant Pfcyt c-2 lysates were dissolved in 1× SDS sample buffer, heated at 95°C for 10 minutes, and fractionated by SDS-PAGE using 12.5% acrylamide gels run at 120 V in the Bio-Rad mini-PROTEAN electrophoresis system. Proteins were transferred to nitrocellulose membrane for 1 hour at 100 V using the Bio-Rad wet-transfer system. For evaluation of hemylation, membranes were developed for 5 min with Prometheus ProSignal Femto ECL reagent (Genesee Scientific 20–302) and imaged using an Invitrogen iBright 1500 Imaging System. Membranes were then stripped and reprobed for evaluation of cytochrome c and HCCS protein expression via western blot. Membranes were blocked with 5% (w/v) skim milk in 1× PBS for 1 hour at ambient temperature, probed at 1:1000 dilution with mouse monoclonal anti-GST-DyLight800 conjugated antibody (ThermoFisher MA4–004-D800) and mouse monoclonal anti-His6 tag-DyLight680 conjugated antibody (ThermoFisher MA1–21315-D680), washed in 1× PBS supplemented with 0.5% (v/v) Tween-20, and imaged on a LI-COR Odyssey CLx.

Circular dichroism.

All circular dichroism (CD) studies were completed on an Aviv Biomedical Inc. Model 410 CD spectrometer using a 1 cm quartz cuvette with a stir bar spinning at 30%. All samples were run at 2 μM[52] in PBS supplemented with 10 μM ammonium persulfate (to ensure oxidized protein state). Spectra for each sample were measured in triplicate from 250–210 nm, measuring every 2 nm with 10 seconds of averaging time at each point. Thermal melt studies measured CD signal at 222 nm, climbing every 2 °C from 25–95 °C. Each temperature was equilibrated for 1 minute before data collection over a 10 second averaging time. The first derivative of the thermal melt curve was calculated using the central difference method (algebraic calculation of slope using the point before and point after each temperature). T_m was calculated as the temperature at the maximum of each first derivative spectrum.

Guaiacol Assay of Peroxidase Activity.

Guaiacol assays were performed at 25 °C by monitoring the formation of tetraguaiacol at 470 nm under previously reported conditions [32, 53]. Stock solutions of H₂O₂ (100 mM, Fisher BP2633) and guaiacol (500 μM, Sigma Aldrich G5502) were verified spectrophotometrically using extinction coefficients of ε₂₄₀ = 41.5 M⁻¹ cm⁻¹ for H₂O₂ and ε₂₇₄ = 2150 M⁻¹ cm⁻¹ for guaiacol[53]. A 2× mixture containing purified cytochrome and varying concentrations of guaiacol was prepared in 50 mM potassium phosphate buffer, pH 7.0. This mixture was combined 1:1 with a 100 mM H₂O₂ stock solution to yield final concentrations of cytochrome (2, 4, or 6 μM), guaiacol (0, 2, 5, 7.5, 10, 25, 50, 60, 75, or 100 μM), and 50 mM H₂O₂. Immediately upon addition of H₂O₂, absorbance at 470 nm was recorded using a ThermoFisher Evolution 350 spectrophotometer until tetraguaiacol formation reached a plateau. Rates of guaiacol consumption (μM s⁻¹) were calculated from the change in absorbance at 470 nm over the initial linear phase using the extinction coefficient for tetraguaiacol (ε₄₇₀ = 26.6 mM⁻¹ cm⁻¹)[32]. Average initial rates from three replicates at each guaiacol concentration were fit to Michaelis–Menten curves in GraphPad Prism 10.6 to determine k_cat and K_m.

Supplementary Material

Supplementary Material: Includes AlphaFold3 models and analyses, UV-vis spectra and Nernst plots for all dyes and cytochromes, Michaelis-Menton curves, and circular dichroism (CD) data (PDF).