The R26A and N33A mutations were introduced into the pCPO3.I–AmdS (a PUC19-based 13.8-kb plasmid),33 via a polymerase chain reaction (PCR) amplification using the QuikChange II XL site-directed mutagenesis kit. The PCR reaction was carried out in an Eppendorf thermal cycler (Mastercycler gradient) using PfuUltra HF DNA polymerase. The primers used to introduce the R26A and N33A mutations are listed below:

Coding strain – R26A primer 1 5’- CTACCGACTCTGCTGCTCCTTGCCCAGCTCTGAACG - 3’

Non-coding strain – R26A primer 2 5’- GAGCTGGGCAAGGAGCAGCAGAGTCGGTAGGACCTG - 3’

Coding strain – N33A primer 1 5’- CTTGCCCAGCTCTGGCCGCTCTTGCCAACCACGGTT - 3’

Non-coding strain – N33A primer 2 5’- GTTGGCAAGAGCGGCCAGAGCTGGGCAAGGAGCACGAG - 3’

Coding strain – R26A/N33A primer 1 5’- GTCCTACCGACTCTGCTGCTCCTTGCCCAGCTCTGG-3’

Non-coding strain – R26A/N33A primer 2 5’- CAAGGAGCAGCAGAGTCGGTAGGACCTGGGGCGAC -3’

The plasmid encoding the double mutant was prepared from the plasmid already containing the N33A mutation by introducing the second R26A mutation. The DNA product was digested with Dpn I restriction endonuclease to eliminate the native plasmid. Mutated plasmids were propagated in E. Coli (DH5-α strain) with the R26A and N33A mutations confirmed by DNA sequencing (Figure S1).

The mutant pCPO3.I-AmdS was co-transformed with pAB4-1 (Figure S2) into the pyrG mutant (uridine-requiring) protease deficient MGG029 strain of A. niger (prtTgla::fleorpyrG) using procedures described previously3334 with the following modifications. Brown conidiospores were inoculated in 50 mL minimal medium (MM) containing 0.5% yeast extract, 0.5% Casamino acids, 5% maltose, 1% glucose, 70 mM NaNO₃, 7 mM KCl, 6 mM KH₂PO₄, 6 mM K₂HPO₄, 2 mM MgSO₄, 10 mM uridine, 25 µL of carbenicillin (100 mg/mL stock), 50 µL of the Hutner trace elements solution.35 The mini-culture was incubated at 250 rpm and 37°C for 1 day and then at 30°C until white-yellow mycelium was observed (2-5 days). After disruption of the mycelium with a homogenizer, the culture was transferred to 1 L MM and incubated at 30°C and 250 rpm for 16-20 hours. Mycelium was collected by filtration through sterile Myracloth followed by protoplast preparation (collected by centrifugation at 3000 rpm, 10 min) and DNA transformation using polyethylene glycol as described previously.34 Successful transformants were selected on the polyacrylamide plates lacking uridine supplementation and used to express the mutant CPO. Polyacrylamide was used as a source of carbon and nitrogen for growth. Each clone from the acrylamide plates was inoculated in 50 ml MM lacking uridine and yeast extract and incubated at 250 rpm and 22°C for 1 week. The clone with the highest peroxidation activity as determined by 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) assay, was used to set up large-scale protein expression.

The A. niger spores were inoculated into several 50-mL cultures and grown for 3-4 days at 250 rpm, 25°C. After mycelium homogenization, small cultures were transferred into 4-L flasks containing MM lacking uridine and yeast extract in the presence of 10% fructose, carbenicillin (50 mg/L), and 100 µM δ-aminolevulinic acid (72 L total). Cultures were incubated at 250 rpm and 25°C for 3 days and then at 22°C for 4 more days. Then, the medium was filtered through the glass fiber filter circles G6 and concentrated down to 50 mL in the Amicon stirring pressure cells. The concentrated sample was dialyzed against buffer A (25 mM phosphate buffer, pH 5.9) and filtered through a 0.45 µm membrane.Mutant CPO was purified using an AKTA prime chromatography system. All buffers used during purification were filtered through a 0.45 µm membrane and de-gassed for 15 minutes. The protein sample in buffer A was applied onto a 100-mL column packed with 50 mL of diethylaminoethanol (DEAE) sepharose resin equilibrated with buffer A and eluted with buffer A while gradually increasing NaCl concentration. The R26A and R26A/N33A CPO proteins eluted at approximately 20% 0.5 M NaCl while N33A CPO eluted at 30% 0.5 M NaCl. Fractions with Reinheitzahl (RZ) values (A₄₂₀/A₂₈₀) above 0.5 were combined and concentrated down to about 1 mL using a Millipore 30,000 centrifugal filter unit. The sample was filtered through a 0.45 µm membrane and applied onto a gel filtration column containing 500 mL of Sephadex G75 resin and eluted with buffer A. Fractions with RZ values between 1.0 – 1.1 were used in the subsequent studies.

Structural characterization of R26A, N33A, and R26A/N33A CPO was performed using a JASCO CD spectrometer and VARIAN UV-Vis spectrophotometer (Cary 300 Bio). The CD spectra of 1 µM mutant and WT CPO in buffer A were collected in the range of 190-260 nm at room temperature using 5-mm quartz cuvettes. Three spectra were averaged for each protein sample. The secondary structures were estimated using the CDSSTR method and reference therein.36 To evaluate the tertiary structural information of the mutant protein, CD spectra of 16 µM R26A, 35 µM N33A, 11 µM R26A/N33A, and corresponding concentrations of WT CPO in phosphate buffer, pH 5.9 were collected in the range of 250-350 nm at room temperature using 1-cm quartz cuvettes. Three spectra were averaged for each protein sample.

The UV-Vis absorption spectra of the mutant and WT CPO proteins in buffer A were collected from 250-700 nm at room temperature using 1-cm quartz cuvettes. Three spectra were averaged for each protein. To study the effect of pH, the spectra were collected in either 100 mM phosphate-citrate buffer or 100 mM phosphate buffer (pH 2.4 – 10.0) using 1-cm quartz cuvettes at room temperature. Buffer exchange was performed using a PD-10 column from GE Healthcare Life Sciences.

Cyanide (CN^-) and carbon monoxide (CO) are known ligands for most heme-containing proteins including CPO.3738 The UV-Vis absorption spectra of ferric and ferrous mutant and WT CPO in 25 mM phosphate buffer, pH 6.0 in the absence and presence of the ligands were collected using 1-cm quartz cuvettes at room temperature. For CO binding, WT and mutant CPO were reduced with sodium dithionite and purged with CO for 1 minute.

In the CN^- binding study, the final concentration of cyanide in the samples was 100 mM. The spectra were collected upon initial addition of the anion ligand (0 min) and after 1-hour incubation (60 min) of the sample at room temperature. Titrations of CPO (3-8 μM) with CN^- (10-800 μM for WT and 1-10 mM for mutants) were performed in a phosphate buffer, pH 6.0. The cyanide dissociation constant (Kd), which is an inverse of a binding constant (K), was calculated using equation 1,39 where ∆A is the maximum absorbance difference, ∆A∞ is the change in absorbance of CPO-CN complete formation, K is a binding affinity constant, and ^S is the concentration of free CN^-:

1/ΔA = (1/(S)) × (1/(KΔA∞)) + (1/ΔA∞) Eq. 1

The Hill equation40 was used to determine the stoichiometry of CN^- binding to the mutants, where h is the number of the binding sites:

log(ΔA/(ΔA∞−ΔA))=h×log(S)+log(K) Eq. 2

All reactions were carried out in 1-cm quartz cuvettes containing a total of 3 ml of the reaction mixture composed of the appropriate concentrations of the buffer and substrates as described in each specific activity assay. All assays were initiated by the addition of hydrogen peroxide (except for the catalase assay) and monitored for 180 seconds with a UV/Vis spectrophotometer at room temperature. The initial rates of reactions were calculated from the linear portions of the reaction curves.

The peroxidase activity of CPO was measured by the oxidation of ABTS commonly used to estimate the peroxidation activity of heme peroxidases. The reaction system consisted of either WT (0.03-5.0 μM) or mutant CPO (0.03-0.2 μM) in 100 mM phosphate-citrate buffer at different pH (2.75, 3.0, 3.5, 4.0, 4.5, 5.0, 5.5, 6.0) containing 36.8 μM ABTS. After H₂O₂ (3.6 mM, final concentration) addition, the progress of the reaction was monitored for the formation of oxidized ABTS at 405 nm.41

The monochlorodimedone (MCD) assay was used to determine the chlorination activity of CPO that catalyzes the conversion of MCD into dichlorodimedone (DCD).1 The reaction mixture consisted of either 0.005-0.05 μM WT or 0.6 μM mutant CPO in 100 mM phosphate-citrate buffer at different pH (2.75, 3.0, 3.5, 4.0, 4.5, 5.0) containing 20 mM KCl and 0.17 mM MCD. After H₂O₂ (2.2 mM, final concentration) addition, the reaction was monitored at 278 nm.1

The catalatic (dismutation) activity was measured by monitoring decomposition of hydrogen peroxide. The reaction system consisted of 0.05% hydrogen peroxide in 100 mM phosphate-citrate buffer (pH 3.0, 3.5, 4.0, 4.5, 5.0, 5.5). After either 0.5 µM WT or mutant CPO was added, the reaction was monitored at 240 nm.42

The styrene epoxidation assay43 was performed in 100 mM phosphate-citrate buffer (pH 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0) containing 0.3 mM styrene and either 0.3 µM WT or mutant CPO. After addition of 2.2 mM H₂O₂, the reaction was monitored at 262 nm at which styrene absorbs. The extinction coefficient of styrene at 262 nm was determined to be 0.6225 ± 0.0781 mM^-1cm^-1 from the slope of a plot of absorbance versus concentration (Figure S3). Styrene was dissolved in 100 mM phosphate-citrate buffer, pH 3.25. Since styrene absorbance decreases with time due to its poor solubility in aqueous solutions, all solutions with styrene were freshly prepared for each trial.

Specific activities (SA) were calculated using equation 3 while the change in the amount of a substrate or product over time was calculated using equation 4, where A is absorbance, ε is the extinction coefficient at a specific wavelength, b is the length of the light pathway, and c is the concentration of the substrate or product. The extinction coefficient values for the substrates are listed in Table S3.

SA = (μmol substrate or product/sec) / (μmol enzyme) Eq. 3

A = εbc Eq. 4

The expression levels of R26A CPO in A. niger were quite low, which necessitated large volumes of culture in order to obtain sufficient amount of the mutant CPO for proper characterization studies (approximately 6 mg of pure protein was obtained from 72-L cultures). After purification, the mutant protein contained both halo- and apoprotein. To improve incorporation of the heme into the R26A mutant, hemin or its precursor (δ-aminolevulinic acid) was added to some cultures grown at 25°C for three days to a final concentration of either 25 µM or 50 µM. However, addition of either hemin or its precursor failed to improve heme incorporation into R26A as determined from similar Rz values of the protein expressed with and without hemin or its precursor. Furthermore, R26A had lower overall stability than the WT or the other two mutant CPO proteins (N33A and R26A/N33A) as indicated by the significant decrease in the Soret band intensity of R26A within 1-2 months of protein expression (not shown).

Circular dichroism spectroscopy was used to verify whether a disruption of the hydrogen bonds between the Arg26-Asn37 and Ala27-Asn33 pairs individually and simultaneously had a significant effect on the secondary and tertiary structures of the enzyme, which in turn could affect catalysis and the ligand binding properties.

The CD spectral feature from 190 to 260 nm revealed the secondary structural information of the proteins. The essentially identical CD spectral pattern of WT, R26A, and N33A CPO shown in Figure 3A suggests that disrupting the hydrogen bonds between Arg26-Asn37 and Ala27-Asn33 individually have negligible effect on the secondary structure of CPO. However, the CD spectrum of R26A/N33A CPO is significantly different from that of the WT and the other two mutants. Using the CDSSTR method 33 and reference therein, the alpha-helical content was estimated for WT and all mutants (Figure 3A). The 1-2% difference between the secondary structures of the WT and single mutants is unlikely to significantly influence the tertiary structure and ligand binding or catalytic properties of CPO. However, removal of both pairs of H-bonds resulted in a noticeable alteration of the alpha-helix (20%-increase).

The CD spectra between 250 and 600 nm were used to evaluate the tertiary structure of the proteins (Figure 3B and Table S1). The signals in the 250-350 nm region arise from both the aromatic residues and disulfide bonds present in the protein. Trp, Tyr, and Phe absorb at around 290 nm, 280 nm, and 260-270 nm, respectively. In addition, the disulfide bond band extends from about 260 nm to its maximum at around 330 nm and typically produces a weak signal. However, its intensity depends on the C-S-S bond angle, dihedral angle of the disulfide bond, and the effects of the surrounding groups.4445

Chloroperoxidase has 18 Phe, 5 Trp, and 11 Tyr in its primary structure. Thus, the broad positive signal in the 275-300 nm region observed for WT CPO is likely produced by the Trp and Tyr residues. The negative band at around 265 nm could be a combination of absorptions arising from Phe and a disulfide bond between Cys79 and Cys87. All three mutants produce negative signals in the aromatic region, which indicate changes in the rigidity or environment of the aromatic residues and possibly of the disulfide bond region.

The region between 400-600 nm provides information about the heme environment. Heme absorbs at around 350 nm (delta band), 410 nm (Soret), and between 500-600 nm (Q bands). In the CD spectrum of WT CPO, the heme group has strong absorptions at 428 nm (Soret) and 545 nm (Q band). Additional weak negative signals are located at 372 nm and 480 nm. The CD spectra of the mutants are markedly different from that of the WT (Figure 3B and Table S1). WT CPO has a positive signal in the aromatic region and two strong heme group signals at 428 nm (Soret) and 545 nm (Q band). These absorptions disappear in the spectrum of the R26A mutant that has a broad negative peak in the 500-600 nm region with two maxima at 561 nm and 572 nm. The CD spectrum of R26A suggests that the heme environment in R26A is quite different from that in the WT. In addition, the absence of the Soret signal may reflect poor incorporation of the heme into the R26A mutant, which is also supported by the lower RZ value. Furthermore, the Soret band in the R26A UV-Vis spectrum decreases significantly within 1-2 months of protein storage, indicating low mutant stability. A very weak signal at around 344-350 nm (delta band) is present in the spectra of all three mutants, which indicates that the electronegative environment of the heme has changed in the H-bond deletion mutants. In the spectra of N33A and R26A/N33A, the Soret bands are observed at 424 nm and 420 nm, respectively, and are broader than that of the WT with a shoulder at around 400 nm. In the spectra of both N33A and R26A/N33A mutants, there are no signals at 370 nm and 480 nm; however, three and two Q bands are present, respectively. These observations suggest that the heme active centers in those two mutants are similar to each other but different from that in the WT.

To further investigate the heme center of the mutants, UV-Vis measurements were taken for both WT and mutant CPO (Figure 3C). At pH 6.0, the WT spectrum shows, as expected, the Soret peak at 398 nm as well as the β, α, and CT bands (Figure 3C and Table 1). The high-spin markers at 515 nm and 650 nm indicate that the heme iron is in the high-spin pentacoordinated state. The spectroscopic signature of high-spin pentacoordinated WT CPO remains stable over the acidic and neutral pH range, shifting to a low-spin state at pH 8.0 and above (Figure 4A), which is consistent with previously reported data showing that transition to the alkaline low-spin, hexacoordinate form occurs between pH 7 and 8.46

The Soret, β, and α bands of the mutants are red-shifted (Table 1). These spectral features are consistent with the presence of a low-spin heme with a sixth ligand that can be either water or the side chain of an amino acid residue in the distal pocket. The presence of a CT band in the R26A and R26A/N33A spectra suggests that a small fraction of the protein may be in a high-spin state. Furthermore, the α/β band intensity ratio in the WT spectrum is similar to that in R26A and lower than that in the N33A and R26A/N33A mutant spectra. These ratios suggest that the a2u and a1u porphyrin orbitals in WT and R26A may be closer in energy than those in N33A and R26A/N33A. In addition, the spectra of all mutants possess a clear δ-band, which is also an indication of a low-spin heme.

The stability of the heme environment at different pH in the mutant CPO proteins is different from that in WT CPO with the double mutant being the most stable of the three mutants over a wide pH range (Figure 4B-D). To further examine the effect of the mutations on the heme environment of CPO, UV-Vis spectra of the mutant CPO proteins and their complexes with CO (Figure 5 and Table S2) and CN^- (Figure 6 and Table 1) were compared with those of the corresponding WT CPO. The absorption spectra of the ferrous mutant CPO proteins are fundamentally indistinguishable from that of the ferrous WT CPO with the Soret at 407-408 nm and Q-bands at 550-551 nm. Furthermore, the Soret absorption of the ferrous mutant CPO-CO complexes is also essentially identical to that of the WT CPO-CO adduct with the Soret bands at 444±1 nm. These spectral properties suggest that disrupting the proximal H-bonding network does not affect the binding of a strong neutral ligand, such as CO, to CPO. Interestingly, the N33A-CO and R26A/N33A-CO complexes show the delta band at 358 nm and 357 nm, respectively, with the first mutant having a more defined peak (Figure 5C, D). The presence of the delta band indicates differences in the electronegative environment of the heme in these two mutants, which can influence the catalytic properties of CPO.

The UV-Vis spectra of the ferric WT and mutant cyanide complexes were obtained at pH 6.0 (Figure 6) in the presence of CN^- excess. Interestingly, the spectral features of mutant CPO-CN gradually change with a slight Soret band blue-shift (Figure 6B-D and Table 1). No further spectral changes were observed beyond 1 hour of CN^- addition, indicating an equilibrium was reached. However, no spectral changes were observed for WT CPO-CN whose Soret band remained at 439 nm (Figure 6A), in agreement with the observations reported previously.47

The slow binding equilibrium between free and cyanide-bound mutant CPO suggests that the sixth coordination site in the mutant CPO is occupied by either an H₂O molecule or an exogenous ligand. Therefore, cyanide has to compete with the existing axial ligand which hinders the speed of CN^- binding and may also result in a lower binding affinity for the mutant than WT CPO. To determine the cyanide dissociation constant, both the WT and mutant CPO proteins were titrated with potassium cyanide at pH 6.0 (Figure 7A-D). The presence of the isosbestic point in all CPO spectra is consistent with the equilibrium between the uncomplexed CPO and CPO-CN species. The Kd values were calculated using equation 1 (Figure 7E) and are listed in Table 2 indicating lower affinity of cyanide for the mutant CPO. The Kd result for the R26A mutant should be treated with caution since only one trial was performed because of the low yield and stability of this mutant. Decreased strength of CN^- binding to the mutants having a low-spin heme is consistent with a decreased CN^- affinity to the WT (Kd of 4 mM) having a low-spin heme at alkaline pH.47However, the WT and mutant low-spin forms must be different since the low-spin WT is enzymatically inactive, while the mutants are active.The Hill plot (Figure 7F) shows that CN^- binding to the R26A, N33A, and R26A/N33A mutants follows the same stoichiometry as in binding to the WT: one molecule of cyanide binds to one molecule of CPO.

To examine the effect of hydrogen bond deletion between Arg26-Asn37 and Ala27-Asn33 on the catalytic properties of CPO, 4 enzymatic assays at various pH were performed: chlorination, dismutation, epoxidation, and peroxidation. WT CPO catalyzed chlorination reaction most efficiently at a very acidic pH with the optimal pH at 2.75 (Figure 8A) while the CPO mutants were essentially devoid of the chlorination activity in the tested pH range of 2.75-5.0, suggesting that Cpd X is likely to be responsible for the differences in the enzyme activities observed at various pH.

Hydrogen peroxide decomposition, a two-electron disproportionation reaction, is catalyzed by WT CPO most efficiently at acidic pH with the highest rate at pH 4.5-5.0 (Figure 8B). R26A CPO displayed catalatic activity of <6% at pH 4.5 and 5.0 and ≤2% at other tested pH compared to WT CPO. No catalatic activity was observed at any tested pH (3.0-5.5) in the N33A and R26A/N33A CPO mutants. Essentially absent catalatic activity of the mutant CPO suggested a reduction in the heme redox potential and/or an obstruction of H₂O₂ binding to Cpd I of the mutant CPO. The impaired disproportionation pathway eliminates the inevitable competition epoxidation has to deal with in WT CPO, overriding the adverse effect of a reduction in the heme redox potential and consequently conferring significantly enhanced epoxidation activity to the mutant CPO as observed in the epoxidation assay discussed next.

The rate of mutant CPO-catalyzed epoxidation of styrene was dramatically higher than that of WT CPO at all tested pH (3.0-10.0) (Figure 8C). Moreover, in the reactions catalyzed by the mutant CPO enzymes, styrene epoxidation does not seem to depend heavily on pH since mutant CPO oxidizes styrene at acidic, neutral, and basic pH with similar efficiency, except at pH 10.0, where the enzyme inactivation starts occurring (Figure 4C, D). A significant decrease in WT CPO activity is observed above pH 8.0 because of the formation of the inactive alkaline species of WT CPO (Figure 4A).46 However, epoxidation catalyzed by the mutant CPO remained highly active at both pH8.0 and pH 9.0, especially the R26A/N33A mutant (Figure 8C), suggesting that the mutant CPO did not form the inactive alkaline form in this pH range as WT CPO did (Figure 4A). This observation implies that Cpd I can be successfully generated in the mutant CPO over a broader pH range than in WT CPO.

The rate of peroxidation reaction catalyzed by WT CPO peaked at pH 2.75 and reduced sharply as pH was increased to 3.5 (Figure 8D). However, all mutant CPO proteins displayed the highest peroxidation activity at pH 3.5 and remained active even at pH 4.5 when WT CPO is devoid of the peroxidation activity, suggesting a possible increase in the pKa of Cpd II or the distal acid-base catalyst, E183.