Effects of Heme Electronic Structure and Distal ... - ACS Publications

Jan 27, 2016 - Saburo Neya,. ∥. Akihiro Suzuki,. ⊥ and Yasuhiko Yamamoto*,†,#. †. Department of Chemistry, University of Tsukuba, Tsukuba 305-...
12 downloads 0 Views 1MB Size
Download PDF
Article pubs.acs.org/IC

Effects of Heme Electronic Structure and Distal Polar Interaction on Functional and Vibrational Properties of Myoglobin Yuki Kanai,† Ryu Nishimura,† Kotaro Nishiyama,†,∇ Tomokazu Shibata,† Sachiko Yanagisawa,‡ Takashi Ogura,‡ Takashi Matsuo,§ Shun Hirota,§ Saburo Neya,∥ Akihiro Suzuki,⊥ and Yasuhiko Yamamoto*,†,# †

Department of Chemistry, University of Tsukuba, Tsukuba 305-8571, Japan Department of Life Science, Graduate School of Life Science, University of Hyogo, RSC-UH Leading Program Center, Sayo-cho, Sayo-gun, Hyogo 679-5148, Japan § Graduate School of Materials Science, Nara Institute of Science and Technology, Ikoma, Nara 630-0192, Japan ∥ Department of Physical Chemistry, Graduate School of Pharmaceutical Sciences, Chiba University, Chuoh-Inohana, Chiba 260-8675, Japan ⊥ Department of Materials Engineering, Nagaoka National College of Technology, Nagaoka 940-8532, Japan # Life Science Center of Tsukuba Advanced Research Alliance, University of Tsukuba, Tsukuba 305-8577, Japan ‡

S Supporting Information *

ABSTRACT: We analyzed the oxygen (O2) and carbon monoxide (CO) binding properties, autoxidation reaction rate, and FeO2 and FeCO vibrational frequencies of the H64Q mutant of sperm whale myoglobin (Mb) reconstituted with chemically modified heme cofactors possessing a variety of heme Fe electron densities (ρFe), and the results were compared with those for the previously studied native [Shibata, T. et al. J. Am. Chem. Soc. 2010, 132, 6091−6098], and H64L [Nishimura, R. et al. Inorg. Chem. 2014, 53, 1091− 1099], and L29F [Nishimura, R. et al. Inorg. Chem. 2014, 53, 9156−9165] mutants in order to elucidate the effect of changes in the heme electronic structure and distal polar interaction contributing to stabilization of the Fe-bound ligand on the functional and vibrational properties of the protein. The study revealed that, as in the cases of the previously studied native protein [Shibata, T. et al. Inorg. Chem. 2012, 51, 11955−11960], the O2 affinity and autoxidation reaction rate of the H64Q mutant decreased with a decrease in ρFe, as expected from the effect of a change in ρFe on the resonance between the Fe2+−O2 bond and Fe3+−O2−-like species in the O2 form, while the CO affinity of the protein is independent of a change in ρFe. We also found that the well-known inverse correlation between the frequencies of Fe-bound CO (νCO) and Fe−C (νFeC) stretching [Li, X.-Y.; Spiro, T. G. J. Am. Chem. Soc. 1988, 110, 6024−6033] is affected differently by changes in ρFe and the distal polar interaction, indicating that the effects of the two electronic perturbations due to the chemical modification of a heme cofactor and the replacement of nearby amino acid residues on the resonance between the two alternative canonical forms of the FeCO fragment in the protein are slightly different from each other. These findings provide a new insight for deeper understanding of the functional regulation of the protein.

■

INTRODUCTION

possessing a heme Fe atom with a variety of electron densities have been investigated in order to reveal the principle of the electronic tuning of the Fe reactivity through analysis of the relationship between the protein function and the electron density of the heme Fe atom (ρ Fe). Although heme modification studies have been extensively performed to elucidate the effect of a change in the heme electronic structure on the protein function, such studies have often been hindered by the fact that the functional regulation of the protein is

Over the last half century, myoglobin (Mb), an oxygen (O2) storage hemoprotein, has served as a paradigm for the structure−function relationships of metalloproteins.1−9 Determination of the three-dimensional structures of native and mutant proteins has provided a wealth of information about how the functional properties of a heme cofactor are regulated through interaction with nearby amino acid residues in the proteins.2−6,10 We have been trying to elucidate the regulation of a protein function through electronic tuning of the intrinsic heme Fe reactivity.11−15 In our study, the native and mutant proteins reconstituted with chemically modified heme cofactors © 2016 American Chemical Society

Received: November 3, 2015 Published: January 27, 2016 1613

DOI: 10.1021/acs.inorgchem.5b02520 Inorg. Chem. 2016, 55, 1613−1622

Article

Inorganic Chemistry

Figure 1. Schematic representation of the structures of the heme cofactors used in this study, i.e., protoheme (Proto), mesoheme (Meso), 3,8dimethyldeuteroporphyrinatoiron(III) 17,18 (3,8-DMD), 13,17-bis(2-carboxylatoethyl)-3,7,12,18-tetramethyl-2,8-bis(trifluoromethyl)porphyrinatoiron(III)16 (7-PF), and 13,17-bis(2-carboxylatoethyl)-3,7-diethyl-12,18-trimethyl- 2,8-ditrifluoromethylporphyrinatoiron(III)11 (2,8DPF). Abbreviation: PH represents CH2CH2COOH.

of the intrinsic heme Fe reactivity through ρFe and the heme environment furnished by nearby amino acid residues.14,15 It is well-documented that His64 significantly increases the O2 affinity of a protein by stabilizing the binding of O2 to the heme Fe atom through the distal H-bond, and hence plays a central role in the control of O2 versus CO discrimination by the protein.1,7−9,19,20 We revealed through studies of the reconstituted protein system of the H64L mutant, the His64 residue being replaced by Leu, that the O2 versus CO discrimination by the protein can be controlled solely through electronic tuning of the intrinsic heme Fe reactivity through ρFe.14 In addition, we have demonstrated that the O2 and CO ligand binding preference of the L29F mutant, the Leu29 residue being replaced by Phe, can be controlled through changes in ρFe.15 In this study, we analyzed the O2 and CO binding properties, and autoxidation of the reconstituted protein system of the H64Q mutant, the His64 residue being replaced by Gln, to investigate the relationship between structural/electronic perturbation at the distal H-bond due to the H64Q mutation and the control of the intrinsic heme Fe reactivity through ρFe. Besides His, Gln most frequently occupies the 64th residue in the protein,21,22 and these residues are thought to play similar roles as a distal residue to each other, because of their similarity in molecular size and ability to participate in hydrogen bonding as a proton donor as well as a proton acceptor between them (Scheme S1).6,23−30 Hence studies of the reconstituted H64Q mutant system are expected to provide valuable information about the relationship between the nature of the distal H-bond and the control of the intrinsic heme Fe reactivity through ρFe. In addition, the vibrational frequencies of the Fe2+-O2 and Fe2+CO fragments of the H64Q mutant were also characterized in order to elucidate the effects of changes in ρFe on the vibrational properties.

achieved through both the heme electronic structure and the heme environment furnished by nearby amino acid residues. Consequently, elucidation of the electronic mechanism responsible for control of the Mb function requires studies on a series of proteins with heme cofactors possessing various ρFes. We have performed substitution of strongly electronwithdrawing trifluoromethyl (CF3) group(s), as heme side chain(s), for large and stepwise alterations of the heme electronic structure using the reported procedures,11,16 and then have constructed a unique system composed of Mbs reconstituted with artificial heme cofactors such as mesoheme (Meso), 3,8-dimethyldeuteroporphyrinatoiron(III)17,18 (3,8DMD), 13,17-bis(2-carboxylatoethyl)-3,8-diethyl-2,12,18-trimethyl-7-trifluoromethylporphyrinatoiron(III)16 (7-PF), and 13,17-bis(2-carboxylatoethyl)-3,7,12,18-tetramethyl-2,8-bis(trifluoromethyl)porphyrinatoiron(III)11 (2,8-DPF), i.e., Mb(Meso), Mb(3,8-DMD), Mb(7-PF), and Mb(2,8-DPF), respectively. These heme cofactors differ in the numbers of CF3, CH3, and C2H5 side chains (Figure 1), and the substitution of CF3 group(s) causes large and stepwise alterations of the heme electronic structure. We found through studies of these unique reconstituted proteins that the O2 affinity of the proteins is greatly affected by ρFe, whereas the carbon monoxide (CO) affinity is almost independent of ρFe.11,14,15 This finding could be reasonably interpreted in terms of the difference in electronic nature between O2 and CO, and hence the Fe2+-O2 and Fe2+-CO fragments in the protein (see Scheme S1 in the Supporting Information). The Fe2+-O2 fragment is bent due to end-on(η1) coordination of O2, while the Fe2+-CO one prefers a perpendicular geometry as to the heme plane, in order to maximize Fe dπ → CO π* backbonding. In addition, electron transfer from Fe2+ to the bound ligand is greater for O2, which has lower lying π* orbitals than CO, and then the developing negative charge is more strongly stabilized by a distal polar interaction, such as the well-known hydrogen bonding interaction between the distal His (His64) NεH proton and the bound O2 (distal H-bond) in the native protein,3 and nearby partial positive charges. Consequently, a decrease in ρFe hinders the formation of the Fe3+-O2−-like species through obstruction of Fe−O bond polarization, resulting in a shift of the resonance toward the Fe2+-O2 species. Since O2 dissociation from the heme Fe atom occurs only at the Fe2+−O2 bond, stabilization of the Fe2+−O2 bond over the Fe3+−O2−-like one with decreasing ρFe results in an increase in the O2 dissociation rate,11 which lowers the O2 affinity. In contrast, the effect of a change in ρFe on the CO affinity is small. Our strategy could be combined with site-directed mutagenesis to elucidate the relationship between electronic tuning

■

MATERIALS AND METHODS

Materials and Protein Samples. All reagents and chemicals were obtained from commercial sources and used as received. The expression and purification of the H64Q (H64Q(Proto)) and H64L (H64L(Proto)) mutants of sperm whale Mb were carried out according to the methods described by Rohlfs et al.31 Mesoheme (Meso) was purchased from Frontier Scientific Co. 3,8-DMD,17,18 7PF,16 and 2,8-DPF11 were synthesized as previously described. The apoproteins of the H64Q and H64L mutants were prepared,32 and then reconstituted with various heme cofactors, as described previously.11 The reconstituted proteins were kept at 4 °C for about 48 h for equilibration of the well-known heme orientational disorder (see Figure S1 in the Supporting Information).33 The CO and O2 forms of the H64Q mutants were prepared through reduction of the proteins using Na2S2O4 (Nacalai Chemicals Ltd.) in the presence of CO gas (Japan Air Gases) and in air, respectively, and then excess 1614

DOI: 10.1021/acs.inorgchem.5b02520 Inorg. Chem. 2016, 55, 1613−1622

Article

Inorganic Chemistry

Figure 2. 600 MHz 1H NMR spectra of H64Q(Proto)CN, H64Q(Meso)CN, H64Q(3,8-DMD)CN, H64Q(7-PF)CN, and H64Q(2,8-DPF)CN at pH 7.40 in 90% H2O/10% 2H2O at 25 °C. The assignments of Ile99 proton signals and the tentative ones of heme methyl proton signals are given with the spectra, and the M and m forms represent the two different orientations of 7-PF, relative to the protein (see Figure S1 in the Supporting Information).33 The Ile99 CγH proton signals, together with heme methyl proton ones, of the mutant proteins are connected by a broken line. pseudo first-order rate constant for O2 association (kobs(on)(O2)), which can be expressed in terms of kon(O2) and koff(O2) as kobs(on)(O2) = kon(O2) × [O2] + koff(O2), since koff(O2) ≪ kon(O2) × [O2], kon(O2) can be determined from the pseudo first-order rate constant for O2 association (kobs(O2)) through the equation kobs(on)(O2) = kon(O2) × [O2], where [O2] = 2.52 × 10−4 M.37 Pseudo first-order rate constants for O2 dissociation (kobs(off)(O2)) for H64Q(Proto), H64Q(Meso), H64Q(3,8-DMD), and H64Q(7-PF) were measured through analysis of the time evolution of the absorbance at 414, 399, 399, and 403 nm, respectively, after rapid mixing of their O2 forms with various [Na2S2O4] using a UNISOKU RSP-601 stopped-flow apparatus. koff(O2)’s were determined as plateau values as [Na2S2O4] → ∞.37 Since H64Q(2,8-DPF) exhibited rather fast O2 dissociation, the time evolution of the absorbance at 416 nm after rapid mixing of the O2 form of the protein with 100 mM potassium phosphate buffer, pH 7.40, equilibrated with 1 atm of CO gas, was measured and analyzed to determine koff(O2).34 The CO associations for H64Q(Proto), H64Q(Meso), H64Q(3,8DMD), H64Q(7-PF), and H64Q(2,8-DPF) were similarly measured through analysis of the time evolution of the absorbance at 419, 420, 408, 408, and 414 nm, respectively, after photolysis of their CO forms under 1 atm of CO, i.e., [CO] = 9.85 × 10−4 M,38 since koff(CO) ≪ kon(CO) × [CO], kon(CO) can be determined from the pseudo firstorder rate constant for CO association (kobs(on)(CO)) through the equation, kobs(on)(CO) = kon(CO) × [CO]. Pseudo-first-order rate constants for CO dissociation (koff(CO)) for the mutant proteins were determined through analysis of the displacement of Fe-bound CO followed by the oxidation of heme iron by K3Fe(CN)6.38 The CO dissociation for the mutant proteins

reducing agent was removed by passing the samples through a Sephadex G-15 (Sigma-Aldrich Co.) column. The met-cyano forms of the H64Q mutants were prepared by adding KCN (Nacalai Chemicals Ltd.) to the met-forms of the proteins obtained through oxidation of the proteins using K3Fe(CN)6 (Nacalai Chemicals Ltd.). For NMR samples, the proteins were concentrated to ∼0.2 to ∼1 mM in 100 mM phosphate buffer, pH 7.40, and then 10% 2H2O was added to the protein solutions. The pH of each sample was measured with a Horiba F-22 pH meter equipped with a Horiba type 6069-10c electrode. NMR Spectroscopy. 1H NMR spectra of the met-cyano forms of the H64Q mutants were recorded on a Bruker AVANCE-600 spectrometer operating at the 1H frequency of 600 MHz. Typical 1 H NMR spectra consisted of about 2−9k transients with a 60 kHz spectral width and 16k data points. The signal-to-noise ratio of the spectra was improved by apodization, which introduced 10 Hz line broadening. The chemical shifts of 1H NMR spectra are given in ppm downfield from the residual 1H2HO peak at 4.75 ppm, as a secondary reference. Kinetic Measurements of O2 and CO Binding. Kinetic measurements of O2 and CO of the proteins were carried out using the stopped-flow rapid mixing and conventional flash photolysis techniques as described previously (see Figures S2−S6 in the Supporting Information).1,34−36 The O2 associations for H64Q(Proto), H64Q(Meso), H64Q(3,8-DMD), H64Q(7-PF), and H64Q(2,8-DPF) were characterized through analysis of the time evolution of the absorbance at 432, 425, 423, 430, and 420 nm, respectively, after photolysis of their O2 forms, under air, using a 5 ns-pulse Nd:YAG laser (532 nm, Continuum Surelite II). The fitting of the time evolution of the absorbance to the first-order rate equation yielded a 1615

DOI: 10.1021/acs.inorgchem.5b02520 Inorg. Chem. 2016, 55, 1613−1622

Article

Inorganic Chemistry

Table 1. O2 and CO Binding Parameters for the H64Q, H64L, and L29F Mutant, and Native Proteins at pH 7.40 and 20 °C, and the Autoxidation Reaction Rates (kox) of the Proteins O2 binding protein

heme

H64Qd

Proto Proto Meso 3,8-DMD 7-PF 2,8-DPF Proto Meso 3,8-DMD 7-PF 2,8-DPF Protok Mesol 3,8-DMDm 7-PFl 2,8-DPFl Proto Meso 3,8-DMD 7-PF 2,8-DPF

H64Li

Mb

L29Fn

koff(O2) (s−1)

kon(O2) (μM−1 s−1) 24 38 23 28 36 57 228 222 292 212 305 14 8.2 12 8.3 16 26 21 25 22 31

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

5e 8 5 6 7 11 68 67 88 64 92 3 1.6 2 1.6 3 5 4 5 4 6

CO binding a

130 130 54 75 170 1800 3200 2300 1900 5700 31000 12 5.7 9.4 17 110 1.5 0.78 0.80 3.8 18

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

26e 30 11 15 34 600h 1300 970 800 2400 13000 2 1.1 1.9 3 22 0.3 0.2 0.2 0.8 4

K(O2) (μM−1) 0.18 0.29 0.42 0.37 0.21 0.032 0.071 0.097 0.15 0.037 0.0098 1.2 1.4 1.3 0.5 0.15 17 27 31 5.8 1.7

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

kon(CO) (μM−1 s−1)

0.05e 0.09 0.12 0.10 0.06 0.012 0.05 0.069 0.11 0.026 0.007 0.3 0.4 0.3 0.1 0.04 5 8 9 2 0.5

0.98 0.76 0.55 0.39 0.55 0.88 30 29 32 21 29 0.51 0.38 0.16 0.32 0.69 0.23 0.19 0.21 0.38 0.33

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

0.20e 0.15 0.11 0.08 0.11g 0.18 9.0 8.7 9.6 6.3j 8.7 0.06 0.07 0.03 0.06 0.13 0.05 0.04 0.04 0.08g 0.07

koff(CO) (s−1) 0.012 0.012 0.018 0.021 0.058 0.036 0.061 0.25 0.23 0.052 0.053 0.019 0.048 0.024 0.032 0.036 0.016 0.013 0.012 0.014 0.017

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

0.002e 0.002 0.004 0.004 0.012 0.007 0.018 0.075 0.069 0.016 0.016 0.005 0.009 0.005 0.006 0.007 0.003 0.003 0.002 0.003 0.003

K(CO)b (μM−1) 82 63 31 19 9.5 24 492 116 139 406 547 27 7.9 6.7 10 19 14 15 18 27 19

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

23e 18 9 5 2.7 7 207 48.7 58.4 170 230 8 2.4 2.0 3 6 4 4 5 8 5

K(CO)/K(O2)

koxc (h−1)

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

0.21f 1.3 4.5 3.0 0.76 0.52 2.1 6.5 5.4 1.3 1.0 0.28 0.97 0.83 0.17 0.052 0.005o n.d.p n.d.p n.d.p n.d.p

460 220 75 52 46 790 6900 1200 930 11000 56000 23 5.5 5.1 20 130 0.82 0.56 0.58 4.7 11

180e 90 30 21 18 380 5700 990 770 9100 46000 9 2.3 2.0 8 50 0.3 0.2 0.2 2 4

a

Calculated from the ratio of the rate constants, K(O2) = kon(O2)/koff(O2). bCalculated from the ratio of the rate constants, K(CO) = kon(CO)/ koff(CO). ckox’s of the H64Q mutants and native proteins were measured at pH 6.10 and 35 °C, and those of the H64L mutants at pH 7.40 and 25 °C. The values for the L29F mutants were not determined due to slow autoxidation. dThe experimental errors for the kon(O2), koff(O2), kon(CO), and koff(CO) and values are ±20%. eObtained from ref 31. fMeasured at pH 7.0 and 37 °C.50 gCalculated from the weight-average of fitted parameters because the time course was biphasic, and the sum of two exponentials provided a better fit to the data. hMeasured by the CO replacement method.34−36 iObtained from refs 14 and 15. jBiphasic time evolution of CO association was fitted by the sum of two exponentials, and the value closer to those for the other H64L mutant proteins is indicated; the other one is 5.69 ± 1.7 μM−1 s−1.14 kMeasured at pH 7.0 and 20 °C.31 l Obtained from ref 11. mObtained from refs 13 and14. nObtained from ref 15. oMeasured at pH 7.0, 37 °C.49 pn.d. represents not determined. Resonance Raman Spectroscopy. Resonance Raman scattering was performed with excitation at 413.1 nm and a power of ∼40 μW with a Kr+ laser (Spectra Physics, BeamLok 2060), dispersed with a polychromator (Ritsu Oyo Kogaku MC-100DG, equipped with 1200 grooves/mm grating), and detected with a liquid nitrogen-cooled charge coupled device (CCD) detector (LN/CCD-1100-PB/VISAR/ 1, Roper Scientific).39 The protein concentrations were approximately 20 μM in 100 mM potassium phosphate buffer, pH 7.40. Raman shifts were calibrated with indene as a frequency standard. The positions of the bands were determined through fitting with Voigt profiles, which comprised convolutions of Gaussian and Lorentzian functions,40 and the accuracy of the peak positions of well-defined Raman bands was ±1 cm−1.

was characterized utilizing the following reactions, i.e., the displacement of Fe-bound CO and the oxidation of heme iron by K3Fe(CN)6, koff (CO)

kox

MbFe(II)(CO) XooooooooY MbFe(II) + CO ⎯→ ⎯ MbFe(III)(H 2O) kon(CO)

where MbFe(II) and MbFe(III)(H2O) represent the deoxy and met-aquo forms of the protein, respectively, and kox is the rate constant for the oxidation of the heme iron. Under the experimental conditions of high [CO] and [K3Fe(CN)6], where a steady-state assumption can be made for [MbFe(II)], the pseudo first-order rate constant for the oxidation of the proteins (kobs(ox)) can be expressed in terms of a pseudo firstorder rate constant for CO dissociation (koff(CO)), [K3Fe(CN)6], and a constant, c, as kobs(ox) = koff(CO) × [K3Fe(CN)6]/(c + [K3Fe(CN)6]). The saturated value in plots of kobs(ox) against [K3Fe(CN)6] affords koff(CO). kobs(ox)’s for H64Q(Proto), H64Q(Meso), H64Q(3,8-DMD), H64Q(7-PF), and H64Q(2,8-DPF) with various [K3Fe(CN)6] were determined through analysis of the time evolution of the absorbance at 578, 557, 557, 415, and 417 nm, respectively. Measurement of Autoxidation Reaction Rates. UV−vis absorption spectra were recorded for ∼2 to ∼ 8 μM H64Q mutants at pH 6.10 and 35 °C, and H64L ones at pH 7.40 and 25 °C using a Beckman DU 640 spectrophotometer. Autoxidation of a protein was followed by recording UV−vis spectra over 250−750 nm using a Beckman DU 640 spectrophotometer equipped with a six-cell changer and a Peltier temperature control module. The autoxidation reactions were characterized by only two spectral species with clear isosbestic points, and the observed time evolution could be represented well by a simple first order reaction mechanism, i.e., d[OxyMb]/dt = −kox[OxyMb], where kox is the apparent autoxidation reaction rate constant.

■

RESULTS H NMR Spectra of Met-cyano Forms of the H64Q Mutants. 600 MHz 1H NMR spectra of the met-cyano forms of the H64Q mutants, i.e., H64Q(Proto)CN, H64Q(Meso)CN, H64Q(3,8-DMD)CN, H64Q(7-PF)CN, and H64Q(2,8DPF)CN, are shown in Figure 2. The NMR spectral pattern of H64Q(Proto)CN was essentially identical to that previously reported by Zhao et al.28 The NMR signals due to Ile99 side chain protons are highly sensitive to the heme active site structure,11,14,15 and hence the similarity in the shifts of the Ile99 CδH3 and CγH proton signals, resolved at ∼ −3 and ∼ −7 ppm, respectively, among the mutant proteins indicated that the orientations of the heme cofactors with respect to the polypeptide chains in these mutant proteins are similar to each other. In addition, the observation of two sets of heme methyl 1616

1

DOI: 10.1021/acs.inorgchem.5b02520 Inorg. Chem. 2016, 55, 1613−1622

Article

Inorganic Chemistry

Vibrational Frequencies of Fe-Bound O2 and CO of the H64Q Mutants. Resonance Raman spectra of the O2 and CO forms of the H64Q mutants possessing various heme cofactors were obtained in order to determine the vibrational frequencies of the Fe−O stretching (νFeO), C−O stretching (νCO), Fe−C stretching (νFeC), and Fe−C−O bending (δFeCO) frequencies in the proteins (Figure 3, Table 2, and see Figures

proton signals in the spectrum of H64Q(7-PF)CN, the major form (M):minor form (m) ratio (M:m) being 2:1, is due to the presence of well-known heme orientational isomers33 (see Figure S1 in the Supporting Information), and this result was different from those for Mb(7-PF)CN,11 H64L(7-PF)CN,14 and L29F(7-PF)CN,15 i.e., M:m = 1:2. Analysis on the NMR signal intensities of heme methyl protons indicated that H64Q(Proto)CN and H64Q(Meso)CN exhibit M:m = 28:1 and M:m = 20:1, respectively. The spectra in Figure 2 indicated that the orientations of the heme cofactors with respect to the polypeptide chains in these mutant proteins are similar to each other. These structural data provide a basis for analysis of the functional properties. Effects of Heme Cofactor Modifications on Functional Properties of the H64Q Mutants. The kinetic parameters for H64Q(Proto) at pH 7.40 and 20 °C, i.e., kon(O2) = 38 ± 8 μM−1 s−1, koff(O2) = 130 ± 30 s−1, kon(CO) = 0.76 ± 0.15 μM−1 s−1, and koff(CO) = 0.012 ± 0.002 s−1 (Table 1 and see Figures S2−S6 in the Supporting Information), agreed with those reported previously for the protein at pH 7.0 and 20 °C.31 The presence of the heme orientational isomers in H64Q(Proto), H64Q(Meso), and H64Q(7-PF) did not disturb the measurements of the kinetic parameters, except for that of kon(CO) of H64Q(7-PF), which exhibited biphasic time evolution of CO association interpretable in terms of the sum of two exponentials with the abundance of each component comparable to the ratio of the M and m forms (see Figure S5 in the Supporting Information). koff(O2) was greatly affected by the heme modification, although the other kinetic values were almost unaffected, as reported for the native protein11 and the H64L14 and L29F15 mutants. koff(O2)’s of H64Q(Meso) and H64Q(3,8-DMD) were smaller by a factor of ∼1/2 relative to that of H64Q(Proto), and those of H64Q(7-PF) and H64Q(2,8-DPF) were almost comparable to and larger by a factor of ∼14 relative to that of H64Q(Proto), respectively. Furthermore, since 7-PF and 2,8-DPF can be considered as counterparts of Meso and 3,8-DMD, respectively, the functional consequences of the substitution of one and two CF3 groups can be elucidated from the results of pairwise comparative studies on H64Q(Meso) and H64Q(7-PF), and H64Q(3,8-DMD) and H64Q(2,8-DPF), respectively. The O2 affinity of the mutant proteins was lowered by a factor of ∼1/2 on the substitution of one CF3 group, as demonstrated for the H64Q(Meso)/H64Q(7-PF) system; i.e., K(O2)’s of H64Q(Meso) and H64Q(7-PF) were 0.42 ± 0.12 μM−1 and 0.21 ± 0.06 μM−1, respectively, and then by a factor of ∼1/12 on the substitution of two CF3 ones, as revealed on analysis of the H64Q(3,8-DMD)/H64Q(2,8-DPF) system; i.e., K(O2)’s of H64Q(3,8-DMD) and H64Q(2,8-DPF) were 0.37 ± 0.10 μM−1 and 0.032 ± 0.012 μM−1, respectively (Table 1). In contrast to the O2 binding, the CO affinity of the mutant proteins was essentially independent of the CF3 substitutions (Table 1). The autoxidation reaction rate constants (kox) of the H64Q mutants at pH 6.10 and 35 °C, and the H64L ones at pH 7.40 and 25 °C were determined (see Figures S7 and S8 in the Supporting Information) and are compared with those of the native proteins in Table 1. As in the cases of the native protein and H64L mutants, the ranking of the heme cofactors, 2,8-DPF < 7-PF < Proto < 3,8-DMD < Meso, in order of increasing kox, also holds for the H64Q mutant, and paralleled it, in order of increasing ρFe, confirming that the autoxidation is regulated by ρFe.12

Figure 3. High frequency regions of visible resonance Raman spectra of the CO forms of the H64Q mutants possessing the indicated heme cofactors, at pH 7.40 and 25 °C. The positions of the individual component νCO bands of the proteins were determined through fitting with Voigt profiles40 (see Figure S12 in the Supporting Information).

S9−S13 in the Supporting Information). νFeO’s of 568−570 cm−1 determined for the H64Q mutants were similar to those of the native protein and H64L and L29F mutants reported previously. νCO, νFeC, and δFeCO of H64Q(Proto) were determined to be 1943, 509, and 578 cm−1, respectively, these values being essentially identical to the corresponding ones previously reported by Zhao et al.28 The νCO’s of the H64Q mutants were greater by 1−7 cm−1 relative to those of the native proteins possessing identical heme cofactors. Comparison of νCO’s of the mutant proteins yielded a difference of 8 cm−1 for the H64Q(Meso)/H64Q(7-PF) system, i.e., 1940 and 1948 cm−1 for H64Q(Meso) and H64Q(7-PF), respectively, which is double the value, 17 cm−1, for the H64Q(3,8-DMD)/H64Q(2,8-DPF) one, i.e., 1941 and 1958 cm−1 for H64Q(3,8-DMD) and H64Q(2,8-DPF), respectively (Table 2). These results demonstrated the additive effect of the heme π-system perturbations on νCO for the H64Q mutants, as in the cases of the native proteins, and H64L and L29F mutants.13−15,41 As in the cases of various hemoproteins,42,43 an inverse correlation between νCO and νFeC, as expected from an admixture of the two alternative canonical forms of the Fe1617

DOI: 10.1021/acs.inorgchem.5b02520 Inorg. Chem. 2016, 55, 1613−1622

Article

Inorganic Chemistry

Table 2. Vibrational Frequencies of the Fe-Bound CO and Fe-Bound O2 of the H64Q, H64L, and L29F Mutant, and Native Proteins at pH 7.40 and 25 °C vCOa (cm−1) Heme Proto Meso 3,8-DMD 7-PF 2,8-DPF

H64Q 1943 1940 1941 1948 1958

H64L

e

1969 1965 1964 1972 1978

vFeCb (cm−1) f

Mb

h

1943 1939h 1940h 1945h 1951h

g

L29F

H64Q

H64L

1933 1928 1928 1936 1944

509 510 509 507 496

490 489 491 489 483

e

δFeCOc (cm−1)

f

Mb

L29F

512 515 514 514 512

524 526 528 523 518

g

H64Q

H64L

578 579 579 575 574

574 575 575 573 570

e

vFeOd (cm−1)

f

L29F

H64Q

H64Le

Mbf

L29Fg

576 576 576 575 574

581 581 583 580 578

569 570 568 570 569

569 573 570 568 568

571 573 571 571 569

569 570 569 568 568

Mb

g

a

The C−O stretching frequency of Fe-bound CO. bThe Fe−C stretching frequency of Fe-bound CO. cThe Fe−C−O bending frequency of Febound CO. dThe Fe−O stretching frequency of Fe-bound O2. eTaken from ref 14. fTaken from ref 13. gTaken from ref 15. hThe weighted-average νCO value taken from ref 13.

ρFe in a similar manner to the cases of the native protein, and H64L and L29F mutants (Figure 4). The log(koff(O2))-νCO

CO fragment illustrated in Scheme 1,43−45 has also been observed for the H64Q mutants (Figure 7A). In addition, Scheme 1. Resonance between the Two Canonical Forms of the Fe2+-CO Fragment of the H64Q Mutants of Mb, Represented by Valence Bond Formalism42−45a

a

The distal Gln64 provides a positive electrostatic potential near the O atom of Fe2+-bound CO, and hence contributes to stabilization of Fe2+(δ+)-COδ−.

comparison of νFeC’s of the H64Q mutants revealed a decrease of 3 cm−1 on the substitution of one CF3 group, as demonstrated for the H64Q(Meso)/H64Q(7-PF) system, i.e., 510 and 507 cm−1 for H64Q(Meso) and H64Q(7-PF), respectively, and one of 13 cm−1 on the substitution of two CF3 groups, as observed for the H64Q(3,8-DMD)/H64Q(2,8DPF) system, i.e., 509 and 496 cm−1 for H64Q(3,8-DMD) and H64Q(2,8-DPF), respectively (Table 2). Hence, in contrast to the case of νCO, an additive effect of the heme π-system perturbations was not observed for νFeC. These results could be interpreted in terms of the effect of ρFe on the nature of the Fe2+-CO bond (see below). Furthermore, the νFeC span of the H64Q mutant system, i.e., 14 cm−1, was larger than that of the native protein one, i.e., 3 cm−1. Finally, the effect of the H64Q mutation on δFeCO of the protein varied with the heme cofactor (Table 2).

Figure 4. Plots of the quantities log(kon(O2)), log(koff(O2)), and log(K(O2)) against νCO of the H64Q (red) mutants, together with those of the native protein11 (black), and H64L14 (blue) and L29F15 (purple) mutants, possessing Proto(□), Meso (Δ), 3,8-DMD (○), 7PF (▲), and 2,8-DPF (●).

■

DISCUSSION Electronic Control of Functional Properties of the H64Q Mutants. We showed previously that the O2 affinity of Mb decreases, due to an increase in koff(O2), with a decrease in ρFe.11,14,15 In order to determine whether the H64Q mutants also exhibit this tendency, we first examined plots of the quantities log(kon(O2)), log(koff(O2)), and log(K(O2)) against νCO (νCO-log(kon(O2)), νCO-log(koff(O2)), and νCO-log(K(O2)) plots, respectively) for the H64Q mutant system, because νCO is a sensitive measure of ρFe.13 The plots clearly demonstrated that the O2 affinity of the H64Q mutant is regulated through

plots for the H64Q mutants could be represented by a straight line with a slope of +0.08 (1/cm−1), and the obtained slope was similar to those of the plots for the native protein, and H64L and L29F mutants previously reported, i.e., +0.08 ∼ +0.10 (1/cm−1) for the three reconstituted protein systems,14,15 demonstrating that the regulation of koff(O2) through ρFe is independent of the nature of the distal H-bond. Comparison of koff(O2)’s of the four reconstituted protein systems, i.e., the native protein, and H64Q, H64L, and L29F mutants, revealed that koff(O2) of the H64L mutant decreased 1618

DOI: 10.1021/acs.inorgchem.5b02520 Inorg. Chem. 2016, 55, 1613−1622

Article

Inorganic Chemistry

by Shikama,46−48 the stabilization of the Fe2+−O2 bond over the Fe3+−O2−-like one with decreasing ρFe is also expected to decrease kox. In fact, kox’s of the proteins correlated well with νCO’s in such a manner that an increase in νCO indicative of a decrease in ρFe results in a decrease in kox (Figure 6). The relationship between the O2 affinity, i.e., K(O2), and kox (K(O2)-kox relationship) of the H64Q mutant was essentially similar to that of the native protein (Figure 5), demonstrating that the K(O2)-kox relationship is independent of any amino acid replacements. In contrast to the cases of the reconstituted native protein and mutant systems, site-directed mutants exhibited the opposite K(O2)-kox relationship; i.e., a mutant possessing lower (higher) O2 affinity exhibits a larger (smaller) kox (Figure 5),7,20,28,49,50 because amino acid replacement(s) lead to enhancement (deterioration) of the stability of the O2 form of the protein, which in turn results in decreasing (increasing) koff(O2) and kox. In fact, the K(O2)-kox relationship of the native protein and mutants reconstituted with identical heme cofactors exhibited a similar tendency to that for sitedirected mutant proteins (Figure 5). These results confirmed that the combined use of perturbation of the heme electronic structure, through chemical modification of the peripheral side chain(s), and that of the protein structure, through amino acid replacement(s), enables independent control of the O2 affinity and autoxidation rate of the protein, which will be useful for designing hemoprotein-based blood substitutes. Effects of Heme Cofactor Modifications on the Vibrational Frequencies of Fe-bound O2 and CO of the H64Q Mutants. As shown in Table 2, νFeO’s of the H64Q mutants were lower by 1−3 cm−1 relative to those of the proteins possessing identical heme cofactors, except for H64Q(2,8-DPF), of which νFeO was identical to that of Mb(2,8-DPF). The low sensitivity of νFeO to replacement of amino acid residues in the heme pocket could be explained on the basis of the Fe−O2 fragment of the O2 form of a protein exists as an admixture of two alternative canonical forms,51,52 i.e., the Fe2+−O2 and Fe3+−O2−-like species (Scheme 2). The Fe2+−O2 σ bond is formed through overlapping of the Fe dz2 and O2 π* orbitals, and transfer of an electron from the Fe t2g orbital to the O2 nonbonding π* one results in the formation of the Fe3+−O2−-like species.48 Consequently, the bond order of the Fe−O2 bond is independent of the resonance in Scheme 2, leading to the low sensitivity of νFeO to the heme environment as well as the heme electronic structure. Furthermore, the weakening of the Fe−O bond in the proteins, as manifested in νFeO, however, cannot account for the increase in koff(O2) of the H64Q mutants with decreasing ρFe, as was also reported previously for the native protein, and H64L and L29F mutants.14,15 On the other hand, νCO’s of the proteins could be reasonably interpreted in terms of the effects of the heme electronic structure and the distal polar interaction on the resonance in Scheme 1.14,15,45 Comparative studies on νCO’s of the native protein, and H64Q, H64L, and L29F mutants revealed systematic effects of changes in ρFe and the distal polar interaction on the resonance in Scheme 1. The ranking of the heme cofactors, Meso ≈ 3,8-DMD < Proto < 7-PF < 2,8-DPF, in order of increasing νCO, consistently holds for each of the reconstituted protein systems. In addition, the ranking of the proteins, L29F < native < H64Q < H64L, in order of increasing νCO, holds for a given heme cofactor. νCO and νFeC are related to ρFe and the distal polar interaction through the resonance between the two canonical forms of the Fe2+−CO fragment

by a factor of 1/43−1/17 on the introduction of Gln64, through the H64Q mutation, and then by factors of 1/400− 1/200 and 1/2900−1/1500 due to His64 and Phe29 in the native protein and L29F mutant, respectively. These results semiquantitatively reflected the extents of the contributions of the individual residues to the stabilization of the Fe2+−O2 bond. Assuming synergetic stabilization of the Fe2+−O2 bond by the residues, the reductions of koff(O2) of the H64L mutant due to Gln64 and Phe29 were estimated to be ∼6 to ∼ 13% and ∼1 to ∼ 6%, respectively, relative to that by His64 (see Figure S14 in the Supporting Information). Despite the similarity in molecular properties between Gln and His residues, Zhao et al.28 also demonstrated that stabilization of the Fe2+−O2 bond by Gln64 in the protein is not as great as that by His64. In contrast to O2 binding, the similar plots for CO binding, i.e., νCO-log(kon(O2)), νCO-log(koff(O2)), and νCO-log(K(O2)) plots, for the H64Q mutant system indicated that the CO binding properties of the H64Q mutants were not greatly affected by ρFe (see Figure S15 in the Supporting Information). Relationship between O2 Affinity and Autoxidation. We have shown through studies of the reconstituted native proteins that a protein possessing a smaller (larger) ρFe exhibits lower (higher) O2 affinity and a smaller (larger) kox (Figure 5).12 This could also be explained in terms of the effects of changes in ρFe on the resonance in Scheme 2. As described above, the decrease in the O2 affinity with decreasing ρFe is due to an increase in koff(O2) as a result of shifting of the resonance toward the Fe2+-O2 species. In addition, considering the reaction mechanism for autoxidation of the protein proposed

Figure 5. Plots of the quantity log(kox) against the quantity log(K(O2)) for the H64Q (red) and H64L (blue) mutants, and native (black) proteins possessing Proto (□), Meso (Δ), 3,8-DMD (○), 7-PF (▲), and 2,8-DPF (●), and the site-directed mutants (×) investigated by Olson and co-workers.50 The lines were drawn using least-squares fitting: solid and broken lines for the reconstituted and site-directed mutant proteins. K(O2) of the reconstituted proteins were calculated from kon(O2) and koff(O2) measured at pH 7.40 and 20 °C. The kox of the H64Q mutants and native proteins were measured at pH 6.10 and 35 °C, and those of the H64L mutants at pH 7.40 and 25 °C. The kox and K(O2) of the site-directed mutant proteins at pH 7.0 and 37 °C were taken from ref 50 (see Figure S16 of the Supporting Information). Although, due to the disagreement in the experimental conditions used for the measurements, the plots for the reconstituted proteins cannot be directly compared with those for the mutants, it is apparent that the functional consequences of the heme side-chain modifications and amino acid replacements are different from each other. 1619

DOI: 10.1021/acs.inorgchem.5b02520 Inorg. Chem. 2016, 55, 1613−1622

Article

Inorganic Chemistry Scheme 2. Oxygenation and Autoxidation of the H64Q Mutants of Mba

a

The binding of O2 to the heme Fe atom is stabilized by the hydrogen bonding between the Fe2+-bound O2 and Gln64 (distal H-bond).6,23−30 Structure (C) of the O2 form51,52 and structure (D) of the met-form25 are only proposed ones.

CO bond, the greater will be the bond order of the Fe2+−CO bond, and the smaller will be the C−O bond order. Consequently, an increase in ρFe results in a decrease and an increase in νCO and νFeC, respectively, and vice versa. In fact, inverse correlations between νCO and νFeC have been reported for various hemoproteins,42,43 and similar νCO-νFeC inverse correlations were also observed for the H64Q mutants, together with the native protein, and H64L and L29F mutants (see below). In addition, we have shown through comparative studies on the native protein, and H64L and L29F mutants that, for a given heme cofactor, a positive electrostatic potential due to His64 decreases νCO of the H64L mutant by 24−27 cm−1,14 and those due to the multipole of the phenyl ring of the introduced Phe29 further decrease νCO by 7−12 cm−1 (Table 2).15 Hence, the finding that the νCO’s of the H64Q mutants were smaller by 20−26 cm−1 relative to those of the H64L mutants possessing identical heme cofactors (Table 2) indicated that the positive electrostatic potential of Gln64 is slightly weaker than that of His64, this being consistent with the weaker distal H-bond of Gln64 than that of His64 in the O2 forms of the proteins.28

Figure 6. Plots of the quantity log(kox) against νCO of the H64Q (red) and H64L (blue) mutants, and native proteins (black) possessing Proto (□), Meso (Δ), 3,8-DMD (○), 7-PF (▲), and 2,8-DPF (●). kox of the H64Q mutants and native proteins were measured at pH 6.10 and 35 °C, and those of the H64L mutants at pH 7.40 and 25 °C. The νCO’s were measured at pH 7.40 and 25 °C.

(Scheme 1).13−15,41−43,45 The larger ρFe is, the better the heme Fe atom can serve as a π donor to CO. The stronger the Fe2+−

Figure 7. Plots of νFeC against νCO (νCO-νFeC plots) for the individual protein systems, i.e., the native (black), and H64Q (red), H64L (blue), and L29F (purple) mutant systems (A), and the native and mutant proteins possessing identical heme cofactors (B): Proto(□), Meso (Δ), 3,8-DMD (○), 7-PF (▲), and 2,8-DPF (●). The νFeC/νCO slopes of the plots in panels A and B are ∼ −0.2 to ∼ −0.8 and ∼ −0.9 to ∼ −1.0, respectively. 1620

DOI: 10.1021/acs.inorgchem.5b02520 Inorg. Chem. 2016, 55, 1613−1622

Inorganic Chemistry

■

As described above, the νCO-νFeC inverse correlation originates from the resonance in Scheme 1.42−45 This resonance is affected by ρFe13−15 and the distal polar interaction,45 and the influences of these factors could be separately evaluated through comparative studies on the native protein, and H64Q, H64L, and L29F mutants, because the effects of changes in ρFe and the distal polar interaction are thought to be manifested in the plots of νFeC against νCO (νCOνFeC plots) of a given reconstituted protein system (Figure 7A), and the native and mutant proteins possessing identical heme cofactors (Figure 7B), respectively. As shown in Figure 7, the slope of the νCO-νFeC plots (νFeC/νCO slope) for the individual reconstituted protein systems (Figure 7A) is distinctly smaller compared with that for the native and mutant proteins possessing identical heme cofactors (Figure 7B); i.e., the νFeC/νCO slopes of the plots in Figure 7A,B were ∼ −0.2 to ∼ −0.8 and ∼ −0.9 to ∼ −1.0, respectively. The smaller νFeC/νCO slopes of the plots for the individual reconstituted protein systems could be explained on the basis of the effect of ρFe on the electronic nature of the Fe2+−CO bond. The CO ligand is a σ donor as well as a π acceptor, and dπ electrons of the Fe atom are donated back to the empty π* orbitals of CO, i.e., backbonding.43 This back-bonding increases the Fe2+−CO bond order, while decreasing the Fe2+C−O one, and hence results in increasing νFeC and conversely decreasing νCO. An increase in ρFe is expected to enhance the back-bonding from Fe2+ to CO, but at the same time weaken the Fe2+−C σ bond by competing for the Fe dz2 orbital. Consequently, for a given change of ρFe, the variation of νFeC is expected to be less than that of νCO, resulting in a reduced νFeC/νCO slope. On the other hand, the distal polar interaction contributes to stabilization of the developing negative charge produced by the back-bonding, and hence Structure II over Structure I in Scheme 1. Thus, this study demonstrated that Fe-bound CO can be used as a highly sensitive molecular probe of the heme electronic structure and the heme environment furnished by nearby amino acid residues.

AUTHOR INFORMATION

Corresponding Author

*Phone/Fax: +81 29 853 6521. E-mail: yamamoto@chem. tsukuba.ac.jp. Present Address ∇

Department of Chemistry, Graduate School of Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 1130033, Japan.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Professor John S. Olson of Rice University for supplying the plasmids containing the genes coding for the recombinant myoglobin, and his valuable guidance and stimulating discussion. We are also grateful to Prof. Masao Ikeda-Saito of Tohoku University for his valuable advice. We also thank Ms. Chihiro Maki for the measurements of kox’s of the H64Q mutants, respectively. This work was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology, Japan.

■

REFERENCES

(1) Antonini, E.; Brunori, M. Hemoglobin and Myoglobin in their Reactions with Ligands; North Holland Publishing: Amsterdam, 1971. (2) Phillips, S. E. V. J. Mol. Biol. 1980, 142, 531−554. (3) Phillips, S. E. V.; Schoenborn, B. P. Nature 1981, 292, 81−82. (4) Hanson, J. C.; Schoenborn, B. P. J. Mol. Biol. 1981, 153, 117− 146. (5) Shaanan, B. J. Mol. Biol. 1983, 171, 31−59. (6) Nagai, K.; Luisi, B.; Shih, D.; Miyazaki, G.; Imai, K.; Poyart, C.; De Young, A.; Kwiatkowsky, L.; Noble, R. W.; Lin, S.-H; Yu, N.-T. Nature 1987, 329, 858−860. (7) Springer, B. A.; Egeberg, K. D.; Sligar, S. G.; Rohlfs, R. J.; Mathews, A. J.; Olson, J. S. J. Biol. Chem. 1989, 264, 3057−3060. (8) Olson, J. S.; Phillips, G. N., Jr. JBIC, J. Biol. Inorg. Chem. 1997, 2, 544−552. (9) Kachalova, G. S.; Popov, A. N.; Bartunik, H. D. Science 1999, 284, 473−476. (10) Quillin, M. L.; Arduini, R. M.; Olson, J. S.; Phillips, G. N., Jr. J. Mol. Biol. 1993, 234, 140−155. (11) Shibata, T.; Nagao, S.; Fukaya, M.; Tai, H.; Nagatomo, S.; Morihashi, K.; Matsuo, T.; Hirota, S.; Suzuki, A.; Imai, K.; Yamamoto, Y. J. Am. Chem. Soc. 2010, 132, 6091−6098. (12) Shibata, T.; Matsumoto, D.; Nishimura, R.; Tai, H.; Matsuoka, A.; Nagao, S.; Matsuo, T.; Hirota, S.; Imai, K.; Neya, S.; Suzuki, A.; Yamamoto, Y. Inorg. Chem. 2012, 51, 11955−11960. (13) Nishimura, R.; Shibata, T.; Tai, H.; Ishigami, I.; Ogura, T.; Nagao, S.; Matsuo, T.; Hirota, S.; Imai, K.; Neya, S.; Suzuki, A.; Yamamoto, Y. Inorg. Chem. 2013, 52, 3349−3355. (14) Nishimura, R.; Shibata, T.; Ishigami, I.; Ogura, T.; Tai, H.; Nagao, S.; Matsuo, T.; Hirota, S.; Shoji, O.; Watanabe, Y.; Imai, K.; Neya, S.; Suzuki, A.; Yamamoto, Y. Inorg. Chem. 2014, 53, 1091−1099. (15) Nishimura, R.; Matsumoto, D.; Shibata, T.; Yanagisawa, S.; Ogura, T.; Tai, H.; Matsuo, T.; Hirota, S.; Neya, S.; Suzuki, A.; Yamamoto, Y. Inorg. Chem. 2014, 53, 9156−9165. (16) Toi, H.; Homma, M.; Suzuki, A.; Ogoshi, H. J. Chem. Soc., Chem. Commun. 1985, 1791−1792. (17) Chang, C. K.; Ward, B.; Ebina, S. Arch. Biochem. Biophys. 1984, 231, 366−371. (18) Neya, S.; Suzuki, M.; Hoshino, T.; Ode, H.; Imai, K.; Komatsu, T.; Ikezaki, A.; Nakamura, M.; Furutani, Y.; Kandori, H. Biochemistry 2010, 49, 5642−5650. (19) Olson, J. S.; McKinnie, R. E.; Mims, M. P.; White, D. K. J. Am. Chem. Soc. 1983, 105, 1522−1527.

■

CONCLUSION We have demonstrated that, as in the cases of the previously studied native and mutant proteins, the O2 affinity and autoxidation reaction rate of the H64Q mutant decrease with a decrease in ρFe, and that the O2 affinities as well as the autoxidation reaction rates of the proteins are affected comparably by a given change in ρFe. These findings indicate that the effects of alterations in the heme electronic structure and heme environment on the protein function are independent of each other. Since changes in the O2 affinity and autoxidation reaction rate due to the chemical modification of a heme cofactor and amino acid replacement of the protein are related to positive and negative correlation, respectively, these two manipulations can be combined to independently optimize the functional properties. This finding will provide a new strategy for designing hemoprotein-based blood substitutes.

■

Article

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.5b02520. Scheme S1 and Figures S1−S16 (PDF) 1621

DOI: 10.1021/acs.inorgchem.5b02520 Inorg. Chem. 2016, 55, 1613−1622

Article

Inorganic Chemistry (20) Springer, B. A.; Sligar, S. G.; Olson, J. S.; Phillips, G. N., Jr. Chem. Rev. 1994, 94, 699−714. (21) Bashford, D.; Chothia, C.; Lesk, A. M. J. Mol. Biol. 1987, 196, 199−216. (22) Suzuki, T.; Imai, K. Cell. Mol. Life Sci. 1998, 54, 979−1004. (23) Yamamoto, Y.; Osawa, A.; Inoue, Y.; Chujo, R.; Suzuki, T. Eur. J. Biochem. 1990, 192, 225−229. (24) Bartnicki, D. E.; Mizukami, H.; Romero-Herrera, A. E. J. Biol. Chem. 1983, 258, 1599−1602. (25) Krishnamoorthi, R.; La Mar, G. N.; Mizukami, H.; Romero, A. J. Biol. Chem. 1984, 259, 265−270. (26) Krishnamoorthi, R.; La Mar, G. N.; Mizukami, H.; Romero, A. J. Biol. Chem. 1984, 259, 8826−8831. (27) Vyas, K.; Rajarathnam, K.; Yu, L. P.; Emerson, S. D.; La Mar, G. N.; Krishnamoorthi, R.; Mizukami, H. J. Biol. Chem. 1993, 268, 14826−14835. (28) Zhao, X.; Vyas, K.; Nguyen, B. D.; Rajarathnam, K.; La Mar, G. N.; Li, T.; Phillips, G. N., Jr.; Eich, R. F.; Olson, J. S.; Ling, J.; Bocian, D. F. J. Biol. Chem. 1995, 270, 20763−20774. (29) Yamamoto, Y.; Suzuki, T. Biochim. Biophys. Acta, Protein Struct. Mol. Enzymol. 1996, 1293, 129−139. (30) Yamamoto, Y.; Kurihara, N.; Egawa, T.; Shimada, H.; Ishimura, Y. Biochim. Biophys. Acta, Protein Struct. Mol. Enzymol. 1999, 1433, 27− 44. (31) Rohlfs, R. J.; Mathews, A. J.; Carver, T. E.; Olson, J. S.; Springer, B. A.; Egeberg, K. D.; Sligar, S. G. J. Biol. Chem. 1990, 265, 3168− 3176. (32) Teale, F. W. J. Biochim. Biophys. Acta 1959, 35, 543. (33) La Mar, G. N.; Budd, D. L.; Viscio, D. B.; Smith, K. M.; Langry, K. C. Proc. Natl. Acad. Sci. U. S. A. 1978, 75, 5755−5759. (34) Matsuo, T.; Dejima, H.; Hirota, S.; Murata, D.; Sato, H.; Ikegami, T.; Hori, H.; Hisaeda, Y.; Hayashi, T. J. Am. Chem. Soc. 2004, 126, 16007−16017. (35) Olson, J. S. Methods Enzymol. 1981, 76, 631−651. (36) Matsuo, T.; Ikegami, T.; Sato, H.; Hisaeda, Y.; Hayashi, T. J. Inorg. Biochem. 2006, 100, 1265−1271. (37) Matsuo, T.; Tsuruta, T.; Maehara, K.; Sato, H.; Hisaeda, Y.; Hayashi, T. Inorg. Chem. 2005, 44, 9391−9396. (38) Matsuo, T.; Ito, K.; Nakashima, Y.; Hisaeda, Y.; Hayashi, T. J. Inorg. Biochem. 2008, 102, 166−173. (39) Kitanishi, K.; Kobayashi, K.; Kawamura, Y.; Ishigami, I.; Ogura, T.; Nakajima, K.; Igarashi, J.; Tanaka, A.; Shimizu, T. Biochemistry 2010, 49, 10381−10393. (40) Reilly, J. T.; Walsh, J. M.; Greenfield, M. L.; Donohue, M. D. Spectrochim. Acta, Part A: Molecular Spectroscopy 1992, 48, 1459−1479. (41) Tsubaki, M.; Nagai, K.; Kitagawa, T. Biochemistry 1980, 19, 379−385. (42) Li, X.-Y.; Spiro, T. G. J. Am. Chem. Soc. 1988, 110, 6024−6033. (43) Ray, G. B.; Li, X.-Y.; Ibers, J. A.; Sessler, J. L.; Spiro, T. G. J. Am. Chem. Soc. 1994, 116, 162−176. (44) Alben, J. O.; Caughey, W. S. Biochemistry 1968, 7, 175−183. (45) Li, T.; Quillin, M. L.; Phillips, G. N., Jr.; Olson, J. S. Biochemistry 1994, 33, 1433−1446. (46) Sugawara, Y.; Shikama, K. Eur. J. Biochem. 1980, 110, 241−246. (47) Shikama, K. Coord. Chem. Rev. 1988, 83, 73−91. (48) Shikama, K. Prog. Biophys. Mol. Biol. 2006, 91, 83−162. (49) Carver, T. E.; Brantley, R. E., Jr.; Singleton, E. W.; Arduini, R. M.; Quillin, M. L.; Phillips, G. N., Jr.; Olson, J. S. J. Biol. Chem. 1992, 267, 14443−14450. (50) Brantley, R. E., Jr.; Smerdon, S. J.; Wilkinson, A. J.; Singleton, E. W.; Olson, J. S. J. Biol. Chem. 1993, 268, 6995−7010. (51) Pauling, L. Nature 1964, 203, 182−183. (52) Maxwell, J. C.; Volpe, J. A.; Barlow, C. H.; Caughey, W. S. Biochem. Biophys. Res. Commun. 1974, 58, 166−171.

1622

DOI: 10.1021/acs.inorgchem.5b02520 Inorg. Chem. 2016, 55, 1613−1622