- Email: [email protected]
Steric effects of isoleucine 107 on heme reorientation reaction in human myoglobin
BBRC Biochemical and Biophysical Research Communications 324 (2004) 1095–1100 www.elsevier.com/locate/ybbrc
Steric effects of isoleucine 107 on heme reorientation reaction in human myoglobin Haruto Ishikawa1, Satoshi Takahashi2, Koichiro Ishimori*, Isao Morishima Department of Molecular Engineering, Graduate School of Engineering, Kyoto University, Kyoto 615-8510, Japan Received 18 September 2004
Abstract Structural factors to regulate the heme reorientation reaction in myoglobin were examined and we found that the side chain at position 107 (Ile107), which is located between the 2-vinyl and 3-methyl groups of heme, forms a kinetic barrier for the heme rotation about the a–c axis. The phenylalanine-substituted mutant showed an extremely slow heme reorientation rate, compared to that of the wild-type protein, while replacement by the decreased side chain, valine, at position 107 accelerated the reorientation reaction. Considering that the spectroscopic data show only minor structural changes in the heme environments of the Ile107 mutants, the side chain at position 107 sterically interacts with the heme peripheral groups in the activation state for the heme reorientation, which supports the intramolecular mechanism that the heme rotates about the a–c axis without leaving the ‘‘protein cage.’’ 2004 Elsevier Inc. All rights reserved. Keywords: Myoglobin; Heme orientation; NMR; Resonance Raman
Due to the pseudo-symmetric structure of the porphyrin macrocycle, the heme prosthetic group can be inserted into apomyoglobin and apohemoglobin in two structurally distinct orientations related by a 180 rotation about the a–c meso-axis [1–3] as illustrated in Figs. 1A and B. With lapse of time, the thermodynamically less stable heme orientation, ‘‘disordered’’ heme orientation (Fig. 1B), is converted into the other heme orientation, ‘‘native’’ orientation (Fig. 1A), and, in the equilibrium state, the dominant (>90%) component has the ‘‘native’’ heme orientation [4–6]. Although the heme disorder appears to have significant, but minor functional consequences in most of hemoproteins [7,8],
*
Corresponding author. Fax: +81-75-383-2541. E-mail address: [email protected] (K. Ishimori). 1 Present address: Department of Molecular Cell Biology, Graduate School of Medicine, Osaka City University, Osaka 545-8585, Japan. 2 Present address: Institute for Protein Research, Osaka University, Osaka 565-0871, Japan. 0006-291X/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2004.09.163
the heme reorientation reaction is the common process in the heme binding to apoproteins, which is inevitable for the formation of structurally matured intact hemoproteins [6]. The heme reorientation reactions are also quite interesting in that the interactions between the heme peripheral groups and the amino acid residues in the heme pocket can specifically and uniquely determine the position of the nearly symmetric heme in the protein matrix. Recently, the interactions between cofactors and the surrounding amino acid residues attracted considerable attention in design for new functional proteins [9], because it is not so simple and easy to place the artificial cofactors at the desired position in the protein matrix [10]. The molecular mechanism for the heme reorientation might shed light on new strategies for design of novel functional proteins containing cofactors. While the heme incorporation into protein is rapid and completed within 1 ms [11], the reorientation of the heme after the incorporation proceeds in the range of hours and days [12]. In the heme incorporation
1096
H. Ishikawa et al. / Biochemical and Biophysical Research Communications 324 (2004) 1095–1100
Fig. 1. The two heme orientations in myoglobin viewed from the proximal histidine side. (A) ‘‘Native’’ heme orientation; (B) ‘‘disordered’’ heme orientation.
process, hydrophobic interactions between the heme peripheral groups, and amino acid residues inside the heme pocket are primarily important to avoid the nonspecific heme binding on the protein surface in vivo [13]. On the other hand, the heme reorientation is supposed to be governed by steric interactions between the heme peripheral groups and side chains of the amino acid residues inside the heme pocket [13]. However, the detailed mechanism for the heme reorientation is not fully understood and, particularly, any amino acid residues responsible for the heme reorientation have not yet been identified. Based on the competition experiments between the heme reconstitution and heme replacements, the heme orientation reaction is proposed to occur through an intramolecular mechanism, i.e., without leaving a ‘‘protein cage’’ [14]. Although the current view of highly flexible proteins has allowed the visualization of aromatic ring reorientation such as tyrosine and phenylalanine, and the structural factors that regulate the ring reorientation reaction were suggested [15], heme is considerably larger and the experimental evidence has not yet been reported to show specific interactions between the heme and amino acid residues crucial for the heme reorientational reactions. In this paper, we focused on the buried amino acid residue located near the a-meso position that can be a steric barrier for the rotation about the a–c axis in the ‘‘protein cage.’’ The amino acid residue we mutated is isoleucine 107 (Ile107), which is located in the heme pocket, and the distances between the side chain of Ile107 and the heme peripheral groups are shown in Fig. 2. As shown in Fig. 2, the distances between the carbon atoms of the side chain and those of the 2- and 3substituents of the heme are in the range from 3.6 to ˚ . It is, therefore, likely that the volume change of 4.6 A the side chain at position 107 can perturb the heme rotation in the ‘‘protein cage.’’ We replaced isoleucine at position 107 with other hydrophobic amino acid residues having the different
Fig. 2. The view of heme and Ile107 in oxygenated myoglobin from the distal histidine side. The distances between the side chain and heme peripheral groups were calculated by the X-ray structure of oxygenated ˚ ) between the myoglobin. The red dotted lines indicate the distances (A ˚) c-carbon of Ile107 and heme peripheral groups. The distances (A between the d-carbon of Ile107 and heme peripheral groups are depicted by the yellow dotted lines.
side chain volumes [16], phenylalanine, leucine, valine, and alanine, to examine the steric effects of amino acid residue at position 107 on the heme orientation reaction. As it has previously been shown, the extent and dynamics of the heme disorder in reconstituted myoglobins can be measured by the NMR spectra of the paramagnetically shifted heme peripheral methyl protons [2,17,18]. The cyanide adducts of myoglobins show the slow heme reorientation, enough to measure the NMR spectra, and give the best resolved heme methyl resonances [1].
Materials and methods The original expression vector of human Mb, pMb3(pLcIIFXMb), is a gift from Varadarajan and Boxer (Stanford University, USA). The procedures for site-directed mutagenesis are described in a previous paper [16]. DNA sequencing for all mutants was performed by the DyeDeoxy Terminator method using ABI 373S DNA sequencer and was analyzed by 373S DNA sequencing system. The procedure to prepare recombinant Mb has been described elsewhere in our previous paper [16]. Apo Mb was prepared by the acid-butanone method [19]. Typically, 1 ml of purified protein solution (about 1 mM) was dialyzed against the chilled 0.1 M phosphate buffer at pH 8.0 and precipitate was removed by centrifugation. The final protein concentration was determined from the absorbance at 280 nm (e = 15.8 mM1 cm1). The hemin was dissolved in minimal volume of 0.1 M NaOH and added, with stirring (about 10 mM), to the cold apoMb solution. A 2fold excess of cyanide was added to the mixture within 5 min after addition of the hemin. The excess hemin was removed using a PD-10 column (Amersham–Pharmacia) by the gel filtration. 1 H NMR spectra of the cyanomet derivatives were recorded on a BRUKER AVANCE DRX 500 MHz system at 25 C. We used a WET pulse sequence for the measurements of the hyperfine-shifted proton resonances to minimize the water signal in the sample. The time course of the reaction was followed by the previously assigned heme methyl peaks.
H. Ishikawa et al. / Biochemical and Biophysical Research Communications 324 (2004) 1095–1100 kobsd was calculated with use of the Eq. (1) lnððAt Ae Þ=ðA0 Ae ÞÞ ¼ k obsd ;
ð1Þ
where A0, At, and Ae are the mole fractions of the native orientation at time zero, time = t, and at equilibrium (time = 1), respectively. With use of the methyl peak areas for the ‘‘major’’ and ‘‘minor’’ components that are defined as M and m, respectively, we calculated the mole fraction of the native orientation at time = i, Ai, as follows: Ai ¼ ½Mi =½M þ mi :
ð2Þ
The observed rate constant kobs is related to the forward, kf, and reverse, kb, rate constants. k obsd ¼ k f þ k b :
ð3Þ
The equilibrium constant for the heme disorder, KD, can be defined as K D ¼ k f =k b ¼ ½Me =½me ;
ð4Þ
where [M]e and [m]e are the intensities of the major and minor heme orientations at equilibrium [14]. We estimated the KD value by extrapolating the intensity of the major and minor heme orientations to infinite time and confirmed the KD value (11) was almost independent of the mutation. kf and kb can be calculated from Eqs. (3) and (4).
1097
Resonance Raman scattering for the deoxy derivatives was excited at 441.6 nm with a He/Cd laser (Kinmon Electronics, CDR80SG) and detected at room temperature by a single polychromator (Ritu, DG1000) equipped with a cooled CCD camera (Astromed, CCD3200). The sample solutions were sealed in quartz cells which were rotated at 1000 rpm. The frequencies of the Raman spectra were calibrated with indene. Sample concentration was 40 lM.
Results and discussion Fig. 3 illustrates the time-dependent changes for the NMR signals from 5-methyl and 8-methyl groups in the ‘‘native’’ and ‘‘disordered’’ heme orientations, respectively, for reconstituted wild-type and mutant myoglobins [1]. In reconstituted wild-type myoglobin (Fig. 3A), a pair of NMR signals was observed at 27.7 and 28.1 ppm immediately after the reconstitution and the signal at 28.1 ppm, arising from the 8-methyl group in the disordered heme, lost the signal intensity with
Fig. 3. 1H NMR spectra (500 MHz) of the human cyanomet wild-type (A), I107F (B), I107L (C), I107V (D), and I107A (E) myoglobins at pH 8.0 immediately after reconstituting apoMb with hemin (upper trace), after 150 h (2nd trace), after 250 h (3rd trace), and after 1 month (bottom trace). These spectra were recorded on a BRUKER AVANCE DRX 500 MHz system at 25 C. 5 M and 8 m represent the NMR signals from the 5-methyl group of the ‘‘native’’ heme and the 8-methyl group of the ‘‘disordered’’ heme, respectively. In the I107V mutant, the NMR spectrum for the after 1 month was not measured due to denaturation of the mutant protein.
1098
H. Ishikawa et al. / Biochemical and Biophysical Research Communications 324 (2004) 1095–1100
time. After 10 days (third trace), nearly 90% of the ‘‘disordered’’ heme orientation was converted into the ‘‘native’’ heme orientation. The first order rate constant for the heme reorientation from the ‘‘disordered’’ form to the ‘‘native’’ form (kf) was estimated as 43 · 104 h1, and that for the back reaction (kb) was 3.9 · 104 h1 (Table 1). These heme reorientation rates are virtually the same as those previously reported [14]. The rate constants are almost insensitive to the isovolume mutation to Leu in the I107L mutant as shown in Fig. 3C (kf = 44 · 104 h1, kb = 4.0 · 104 h1). In the heme reorientation reactions for the Ile107 mutants, the most prominent change in the time course of the NMR signals was observed for the phenylalaninesubstituted (I107F) mutant (Fig. 3B). The heme reorientation reaction in the I107F mutant, which can also be fitted by a single exponential decay and give similar KD value to that of the wild-type protein, was extremely slow and, even after 1 month, about 40% of the heme orientation was still ‘‘disordered.’’ The heme reorientation rate was only 7% of that of wild-type myoglobin. Such drastically reduced heme reorientation rates in the I107F would be an unambiguous demonstration that the bulky side chain at position 107 can be a steric barrier for the heme reorientation reaction. To further examine the structural factors that drastically decelerate the heme reorientation reaction in the I107F mutant, we measured the resonance Raman spectra of the mutant (Fig. 4). While the spectral perturbations by the replacement of Ile with Phe are rather small and the deviation of the stretching mode for the iron–histidyl bond, one of the markers for the heme environmental structure, in the mutant from that in the wild-type protein was also less than 1 cm1, some of the vibration modes of the porphyrin ring were shifted from those of the wild-type protein (Fig. 4). One of the distinct changes observed for the I107F mutant is the position of the m8 line. The m8 line, which has been assigned to the stretching mode between the heme iron and pyrrole nitrogen [20], appeared at 342 cm1 for the wild-type protein, while the I107F mutant gave the m8 line at 337 cm1. One of the out-of-plane modes from the methine wagging, c7 [20], also shifted from 303 to 300 cm1. In addition, the significant signal broadening
Table 1 Rate constants for heme reorientation reactions in wild-type and mutant myoglobins Myoglobin
Side chain volume19 (cm3 mol1)
kf (·104 h1)
kb (·104 h1)
Wild-type I107F I107L I107V I107A
168.8 203.4 167.9 141.7 91.5
43 2.7 44 75 31
3.9 0.25 4.0 6.8 2.8
Fig. 4. Resonance Raman spectra for the human deoxy myoglobin exited by He/Cd laser. The sample concentration is about 40 lM in 100 mM Tris–Cl, pH 7.5, at room temperature.
was detected for the m9 line, a symmetric bending mode for the Cb–C1 bond [20]. Although the correlation between the heme environmental changes and the positional shift of the m8 and c7 lines or broadening of the m9 line is not clear, it can be concluded that the side chain at position of 107 can interact with the peripheral groups and meso-positions of the porphyrin ring and to affect the vibration modes. Such small structural changes in the heme environment are also evident in the NMR spectrum for the cyanide bound I107F mutant. As displayed in Fig. 5, the spectral pattern for the I107F mutant is quite similar to that of the wild-type protein. The prominent retardation in the heme reorientation reaction for the I107F mutant is, therefore, not due to the structural changes of the protein moiety near the heme peripheral groups. Instead, the bulky side chain at position 107 would increase the activation energy [21] for the rotation of the heme about the a–c axis, corresponding to the heme rotation for the reorientation in the ‘‘protein cage’’ [14]. In sharp contrast to the replacement with the bulky side chain, introduction of an amino acid residue with the small side chain volume, valine, into position 107 (I107V) accelerated the heme reorientation reaction and 150 h are enough to complete the heme reorientation reaction (Fig. 2D; kf = 75 · 104 h1, kb = 6.8 · 104 h1). The resonance Raman (Fig. 4) and NMR (Fig. 5) spectra of the I107V mutant also exhibit only minor perturbations on the heme environmental structure. Contrary to the bulky side chain in the I107F mutant, the reduced steric hindrance at position 107 in the I107V mutant would lower the activation energy for the
H. Ishikawa et al. / Biochemical and Biophysical Research Communications 324 (2004) 1095–1100
Fig. 5. 1H NMR spectra (500 MHz) of the human cyanomet wildtype, I107F, I107L, I107V, and I107A myoglobins at pH 8.0 immediately after reconstituting apoMb with hemin. These spectra were recorded on a BRUKER AVANCE DRX 500 MHz system at 25 C. Five and one molar represent the NMR signals from the 5- and 1-methyl groups of the ‘‘native’’ heme, respectively. Three and 8 m are the NMR signals from the 3- and 8-methyl groups of the ‘‘disordered’’ heme, respectively. His93NdH denotes the NMR signal from the proton at the Nd position of the ‘‘distal’’ histidine (His93).
heme reorientation reaction to facilitate the heme rotation in the ‘‘protein cage.’’ On the other hand, drastic reduction of the side chain volume at position 107 significantly retarded the heme reorientation reaction. The reorientation rate for the alanine-substituted (I107A) mutant (Fig. 3E; kf = 31 · 104 h1, kb = 2.8 · 104 h1), whose side chain volume (91.5 cm3 mol1) at position 107 is much smaller than that of the I107V mutant (168.8 cm3 mol1), was slower than that of the wild-type protein (kf = 43 · 104 h1, kb = 3.9 · 104 h1). We have not yet found out the structural factors responsible for the retardation of the heme reorientation reaction in the I107A mutant, but it should be noted here that the resonance Raman spectrum for the I107A mutant is quite similar to that of the I107F mutant. The m8 line of the I107A mutant appeared at 337 cm1 as found for the I107F mutant, while the wild-type protein and the other mutants gave the m8 line at 342 cm1. In addition to the position of the m8 line, the broad m9 lines were observed for the I107A and I107F mutants. The structural perturbations on the metal–pyrrole nitrogen and
1099
pyrrole Cb–C1 bonds in the I107A mutant are similar to those in the I107F mutation. One of the possible reasons for the close similarity between the two mutants would be that, due to the formation of the large space by the drastically reduced side chain at position 107, the side chain of some other amino acid residue is located near the a-meso position, resulting in a new steric barrier for the heme reorientation. In the I107A mutant, the reduced steric hindrance at position 107 would be no longer a kinetic barrier for the heme reorientation and, instead of Ile107, side chains of other amino acid residues would play a primary role in forming the kinetic barrier. In summary, we found that the side chain at position of 107 can be a steric barrier for the rotation of the heme about the a–c axis in myoglobin. Based on the structural characterization of the I107 mutants, the mutation at position 107 primarily affects the kinetic barrier between the ‘‘native’’ and ‘‘disordered’’ heme orientation. Involvement of the amino acid residues inside the heme pocket in the formation of the energetic barrier for the heme reorientation reaction is clear evidence for the crucial roles of the steric interactions between the side chain of the amino acid residues and heme peripheral groups in the correct heme binding of hemoproteins, which supports the intramolecular reaction mechanism for the heme reorientation in the ‘‘protein cage’’ of myoglobin.
Acknowledgments This work was supported by Grants-in-Aid (12002008, I.M.; 14658217, K.I.) from Ministry of Education, Culture, Sports, Science, and Technology in Japan. H.I. was supported by JSPS Research Fellowships for Young Scientists. We are indebted to Prof. T. Kitagawa (National Institutes of Natural Science) for the measurements of the resonance Raman spectra.
References [1] G.N. La Mar, N.L. Davis, D.W. Parish, K.M. Smith, Heme orientational disorder in reconstituted and native sperm whale myoglobin. Proton nuclear magnetic resonance characterizations by heme methyl deuterium labeling in the Met-cyano protein, J. Mol. Biol 168 (1983) 887–896. [2] G.N. La Mar, Y. Yamamoto, T. Jue, K.M. Smith, R.K. Pandey, 1 H NMR characterization of metastable and equilibrium heme orientational heterogeneity in reconstituted and native human hemoglobin, Biochemistry 24 (1985) 3826–3831. [3] C. Kiefl, N. Sreerama, R. Haddad, L. Sun, W. Jentzen, Y. Lu, Y. Qiu, J.A. Shelnutt, R.W. Woody, Heme distortions in spermwhale carbonmonoxy myoglobin: correlations between rotational strengths and heme distortions in MD-generated structures, J. Am. Chem. Soc. 124 (2002) 3385–3394. [4] M.F. Perutz, Stereochemistry of cooperative effects in haemoglobin, Nature 228 (1970) 726–739.
1100
H. Ishikawa et al. / Biochemical and Biophysical Research Communications 324 (2004) 1095–1100
[5] J.C. Kendrew, Bodo, G., Dintzis, H.M., Parrish, R.G., Wyckoff, H., Phillips, D.C., A Three-dimensional model of the myoglobin molecule obtained by X-ray analysis, Nature 181 (1958) 662–666. [6] A.J. Mathews, T. Brittan, Haem disorder in recombinant- and reticulocyte-derived haemoglobins: evidence for stereoselective haem insertion in eukaryotes, Biochem. J. 357 (2001) 305–311. [7] W. Light, R. Rohlfs, G. Palmer, J. Olson, Functional effects of heme orientational disorder in sperm whale myoglobin, J. Biol. Chem. 262 (1987) 46–52. [8] H.S. Aojula, M.T. Wilson, G.R. Moore, D.J. Williamson, 1H NMR and CD studies of haem orientational disorder in spermwhale myoglobin and human haemoglobin, Biochem. J. 250 (1988) 853–858. [9] T. Ueno, M. Ohashi, M. Kono, K. Kondo, A. Suzuki, T. Yamane, Y. Watanabe, Crystal structures of artificial metalloproteins: tight binding of FeIII(Schiff-base) by mutation of Ala71 to Gly in apo-myoglobin, Inorg. Chem. 43 (2004) 2852–2858. [10] H.J. Yang, T. Matsui, S. Ozaki, S. Kato, T. Ueno, G.N. Phillips Jr., S. Fukuzumi, Y. Watanabe, Molecular engineering of myoglobin: influence of residue 68 on the rate and the enantioselectivity of oxidation reactions catalyzed by H64D/V68X myoglobin, Biochemistry 42 (2003) 10174–10181. [11] Q.H. Gibson, E. Antonini, Kinetic studies on the reaction between native globin and haem derivatives, Biochem. J. 77 (1960) 328–341. [12] T. Jue, R. Krishnamoorthi, G.N. La Mar, Proton NMR Study of the heme-apoprotein reaction for myoglobin, J. Am. Chem. Soc. 105 (1983) 5701–5703.
[13] K. Ishimori, I. Morishima, Study of the specific heme orientation in reconstituted hemoglobins, Biochemistry 27 (1988) 4747–4753. [14] G.N. La Mar, H. Toi, R. Krishnamoorthi, Proton NMR investigation of the rate and mechanism of heme rotation in sperm whale myoglobin: evidence for intramolecular reorientation about a heme twofold axis, J. Am. Chem. Soc. 106 (1984) 6395– 6401. [15] I.D. Campbell, C.M. Dobson, R.J.P. Williams, Proton magnetic resonance studies of the tyrosine residues of hen lysozymeassignment and detection of conformational mobility, Proc. R. Soc. Lond. B. Biol. Sci. 189 (1975) 503. [16] H. Ishikawa, T. Uchida, S. Takahashi, K. Ishimori, I. Morishima, Ligand migration in human myoglobin: Steric effects of isoleucine 107(G8) on O2 and CO binding, Biophys. J. 80 (2001) 1507–1517. [17] Y. Yamamoto, G.N. La Mar, 1H NMR study of dynamics and thermodynamics of heme rotational disorder in native and reconstituted hemoglobin A, Biochemistry 25 (1986) 5288–5297. [18] G.N. La Mar, D.L. Budd, D.B. Viscio, K.M. Smith, K.C. Langry, Proton nuclear magnetic resonance characterization of heme disorder in hemoproteins, Proc. Natl. Acad. Sci. USA 75 (1978) 5755–5759. [19] F.W.J. Teale, Cleavage of the haem–protein link by acid methylethylketone, Biochim. Biophys. Acta 35 (1959) 543. [20] S. Hu, K.M. Smith, T.G. Spiro, Assignment of protoheme resonance Raman spectrum by heme labeling in myoglobin, J. Am. Chem. Soc. 118 (1996) 12638–12646. [21] C. Chothia, Structural invariants in protein folding, Nature 254 (1975) 304–308.
Steric effects of isoleucine 107 on heme reorientation reaction in human myoglobin Haruto Ishikawa1, Satoshi Takahashi2, Koichiro Ishimori*, Isao Morishima Department of Molecular Engineering, Graduate School of Engineering, Kyoto University, Kyoto 615-8510, Japan Received 18 September 2004
Abstract Structural factors to regulate the heme reorientation reaction in myoglobin were examined and we found that the side chain at position 107 (Ile107), which is located between the 2-vinyl and 3-methyl groups of heme, forms a kinetic barrier for the heme rotation about the a–c axis. The phenylalanine-substituted mutant showed an extremely slow heme reorientation rate, compared to that of the wild-type protein, while replacement by the decreased side chain, valine, at position 107 accelerated the reorientation reaction. Considering that the spectroscopic data show only minor structural changes in the heme environments of the Ile107 mutants, the side chain at position 107 sterically interacts with the heme peripheral groups in the activation state for the heme reorientation, which supports the intramolecular mechanism that the heme rotates about the a–c axis without leaving the ‘‘protein cage.’’ 2004 Elsevier Inc. All rights reserved. Keywords: Myoglobin; Heme orientation; NMR; Resonance Raman
Due to the pseudo-symmetric structure of the porphyrin macrocycle, the heme prosthetic group can be inserted into apomyoglobin and apohemoglobin in two structurally distinct orientations related by a 180 rotation about the a–c meso-axis [1–3] as illustrated in Figs. 1A and B. With lapse of time, the thermodynamically less stable heme orientation, ‘‘disordered’’ heme orientation (Fig. 1B), is converted into the other heme orientation, ‘‘native’’ orientation (Fig. 1A), and, in the equilibrium state, the dominant (>90%) component has the ‘‘native’’ heme orientation [4–6]. Although the heme disorder appears to have significant, but minor functional consequences in most of hemoproteins [7,8],
*
Corresponding author. Fax: +81-75-383-2541. E-mail address: [email protected] (K. Ishimori). 1 Present address: Department of Molecular Cell Biology, Graduate School of Medicine, Osaka City University, Osaka 545-8585, Japan. 2 Present address: Institute for Protein Research, Osaka University, Osaka 565-0871, Japan. 0006-291X/$ - see front matter 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2004.09.163
the heme reorientation reaction is the common process in the heme binding to apoproteins, which is inevitable for the formation of structurally matured intact hemoproteins [6]. The heme reorientation reactions are also quite interesting in that the interactions between the heme peripheral groups and the amino acid residues in the heme pocket can specifically and uniquely determine the position of the nearly symmetric heme in the protein matrix. Recently, the interactions between cofactors and the surrounding amino acid residues attracted considerable attention in design for new functional proteins [9], because it is not so simple and easy to place the artificial cofactors at the desired position in the protein matrix [10]. The molecular mechanism for the heme reorientation might shed light on new strategies for design of novel functional proteins containing cofactors. While the heme incorporation into protein is rapid and completed within 1 ms [11], the reorientation of the heme after the incorporation proceeds in the range of hours and days [12]. In the heme incorporation
1096
H. Ishikawa et al. / Biochemical and Biophysical Research Communications 324 (2004) 1095–1100
Fig. 1. The two heme orientations in myoglobin viewed from the proximal histidine side. (A) ‘‘Native’’ heme orientation; (B) ‘‘disordered’’ heme orientation.
process, hydrophobic interactions between the heme peripheral groups, and amino acid residues inside the heme pocket are primarily important to avoid the nonspecific heme binding on the protein surface in vivo [13]. On the other hand, the heme reorientation is supposed to be governed by steric interactions between the heme peripheral groups and side chains of the amino acid residues inside the heme pocket [13]. However, the detailed mechanism for the heme reorientation is not fully understood and, particularly, any amino acid residues responsible for the heme reorientation have not yet been identified. Based on the competition experiments between the heme reconstitution and heme replacements, the heme orientation reaction is proposed to occur through an intramolecular mechanism, i.e., without leaving a ‘‘protein cage’’ [14]. Although the current view of highly flexible proteins has allowed the visualization of aromatic ring reorientation such as tyrosine and phenylalanine, and the structural factors that regulate the ring reorientation reaction were suggested [15], heme is considerably larger and the experimental evidence has not yet been reported to show specific interactions between the heme and amino acid residues crucial for the heme reorientational reactions. In this paper, we focused on the buried amino acid residue located near the a-meso position that can be a steric barrier for the rotation about the a–c axis in the ‘‘protein cage.’’ The amino acid residue we mutated is isoleucine 107 (Ile107), which is located in the heme pocket, and the distances between the side chain of Ile107 and the heme peripheral groups are shown in Fig. 2. As shown in Fig. 2, the distances between the carbon atoms of the side chain and those of the 2- and 3substituents of the heme are in the range from 3.6 to ˚ . It is, therefore, likely that the volume change of 4.6 A the side chain at position 107 can perturb the heme rotation in the ‘‘protein cage.’’ We replaced isoleucine at position 107 with other hydrophobic amino acid residues having the different
Fig. 2. The view of heme and Ile107 in oxygenated myoglobin from the distal histidine side. The distances between the side chain and heme peripheral groups were calculated by the X-ray structure of oxygenated ˚ ) between the myoglobin. The red dotted lines indicate the distances (A ˚) c-carbon of Ile107 and heme peripheral groups. The distances (A between the d-carbon of Ile107 and heme peripheral groups are depicted by the yellow dotted lines.
side chain volumes [16], phenylalanine, leucine, valine, and alanine, to examine the steric effects of amino acid residue at position 107 on the heme orientation reaction. As it has previously been shown, the extent and dynamics of the heme disorder in reconstituted myoglobins can be measured by the NMR spectra of the paramagnetically shifted heme peripheral methyl protons [2,17,18]. The cyanide adducts of myoglobins show the slow heme reorientation, enough to measure the NMR spectra, and give the best resolved heme methyl resonances [1].
Materials and methods The original expression vector of human Mb, pMb3(pLcIIFXMb), is a gift from Varadarajan and Boxer (Stanford University, USA). The procedures for site-directed mutagenesis are described in a previous paper [16]. DNA sequencing for all mutants was performed by the DyeDeoxy Terminator method using ABI 373S DNA sequencer and was analyzed by 373S DNA sequencing system. The procedure to prepare recombinant Mb has been described elsewhere in our previous paper [16]. Apo Mb was prepared by the acid-butanone method [19]. Typically, 1 ml of purified protein solution (about 1 mM) was dialyzed against the chilled 0.1 M phosphate buffer at pH 8.0 and precipitate was removed by centrifugation. The final protein concentration was determined from the absorbance at 280 nm (e = 15.8 mM1 cm1). The hemin was dissolved in minimal volume of 0.1 M NaOH and added, with stirring (about 10 mM), to the cold apoMb solution. A 2fold excess of cyanide was added to the mixture within 5 min after addition of the hemin. The excess hemin was removed using a PD-10 column (Amersham–Pharmacia) by the gel filtration. 1 H NMR spectra of the cyanomet derivatives were recorded on a BRUKER AVANCE DRX 500 MHz system at 25 C. We used a WET pulse sequence for the measurements of the hyperfine-shifted proton resonances to minimize the water signal in the sample. The time course of the reaction was followed by the previously assigned heme methyl peaks.
H. Ishikawa et al. / Biochemical and Biophysical Research Communications 324 (2004) 1095–1100 kobsd was calculated with use of the Eq. (1) lnððAt Ae Þ=ðA0 Ae ÞÞ ¼ k obsd ;
ð1Þ
where A0, At, and Ae are the mole fractions of the native orientation at time zero, time = t, and at equilibrium (time = 1), respectively. With use of the methyl peak areas for the ‘‘major’’ and ‘‘minor’’ components that are defined as M and m, respectively, we calculated the mole fraction of the native orientation at time = i, Ai, as follows: Ai ¼ ½Mi =½M þ mi :
ð2Þ
The observed rate constant kobs is related to the forward, kf, and reverse, kb, rate constants. k obsd ¼ k f þ k b :
ð3Þ
The equilibrium constant for the heme disorder, KD, can be defined as K D ¼ k f =k b ¼ ½Me =½me ;
ð4Þ
where [M]e and [m]e are the intensities of the major and minor heme orientations at equilibrium [14]. We estimated the KD value by extrapolating the intensity of the major and minor heme orientations to infinite time and confirmed the KD value (11) was almost independent of the mutation. kf and kb can be calculated from Eqs. (3) and (4).
1097
Resonance Raman scattering for the deoxy derivatives was excited at 441.6 nm with a He/Cd laser (Kinmon Electronics, CDR80SG) and detected at room temperature by a single polychromator (Ritu, DG1000) equipped with a cooled CCD camera (Astromed, CCD3200). The sample solutions were sealed in quartz cells which were rotated at 1000 rpm. The frequencies of the Raman spectra were calibrated with indene. Sample concentration was 40 lM.
Results and discussion Fig. 3 illustrates the time-dependent changes for the NMR signals from 5-methyl and 8-methyl groups in the ‘‘native’’ and ‘‘disordered’’ heme orientations, respectively, for reconstituted wild-type and mutant myoglobins [1]. In reconstituted wild-type myoglobin (Fig. 3A), a pair of NMR signals was observed at 27.7 and 28.1 ppm immediately after the reconstitution and the signal at 28.1 ppm, arising from the 8-methyl group in the disordered heme, lost the signal intensity with
Fig. 3. 1H NMR spectra (500 MHz) of the human cyanomet wild-type (A), I107F (B), I107L (C), I107V (D), and I107A (E) myoglobins at pH 8.0 immediately after reconstituting apoMb with hemin (upper trace), after 150 h (2nd trace), after 250 h (3rd trace), and after 1 month (bottom trace). These spectra were recorded on a BRUKER AVANCE DRX 500 MHz system at 25 C. 5 M and 8 m represent the NMR signals from the 5-methyl group of the ‘‘native’’ heme and the 8-methyl group of the ‘‘disordered’’ heme, respectively. In the I107V mutant, the NMR spectrum for the after 1 month was not measured due to denaturation of the mutant protein.
1098
H. Ishikawa et al. / Biochemical and Biophysical Research Communications 324 (2004) 1095–1100
time. After 10 days (third trace), nearly 90% of the ‘‘disordered’’ heme orientation was converted into the ‘‘native’’ heme orientation. The first order rate constant for the heme reorientation from the ‘‘disordered’’ form to the ‘‘native’’ form (kf) was estimated as 43 · 104 h1, and that for the back reaction (kb) was 3.9 · 104 h1 (Table 1). These heme reorientation rates are virtually the same as those previously reported [14]. The rate constants are almost insensitive to the isovolume mutation to Leu in the I107L mutant as shown in Fig. 3C (kf = 44 · 104 h1, kb = 4.0 · 104 h1). In the heme reorientation reactions for the Ile107 mutants, the most prominent change in the time course of the NMR signals was observed for the phenylalaninesubstituted (I107F) mutant (Fig. 3B). The heme reorientation reaction in the I107F mutant, which can also be fitted by a single exponential decay and give similar KD value to that of the wild-type protein, was extremely slow and, even after 1 month, about 40% of the heme orientation was still ‘‘disordered.’’ The heme reorientation rate was only 7% of that of wild-type myoglobin. Such drastically reduced heme reorientation rates in the I107F would be an unambiguous demonstration that the bulky side chain at position 107 can be a steric barrier for the heme reorientation reaction. To further examine the structural factors that drastically decelerate the heme reorientation reaction in the I107F mutant, we measured the resonance Raman spectra of the mutant (Fig. 4). While the spectral perturbations by the replacement of Ile with Phe are rather small and the deviation of the stretching mode for the iron–histidyl bond, one of the markers for the heme environmental structure, in the mutant from that in the wild-type protein was also less than 1 cm1, some of the vibration modes of the porphyrin ring were shifted from those of the wild-type protein (Fig. 4). One of the distinct changes observed for the I107F mutant is the position of the m8 line. The m8 line, which has been assigned to the stretching mode between the heme iron and pyrrole nitrogen [20], appeared at 342 cm1 for the wild-type protein, while the I107F mutant gave the m8 line at 337 cm1. One of the out-of-plane modes from the methine wagging, c7 [20], also shifted from 303 to 300 cm1. In addition, the significant signal broadening
Table 1 Rate constants for heme reorientation reactions in wild-type and mutant myoglobins Myoglobin
Side chain volume19 (cm3 mol1)
kf (·104 h1)
kb (·104 h1)
Wild-type I107F I107L I107V I107A
168.8 203.4 167.9 141.7 91.5
43 2.7 44 75 31
3.9 0.25 4.0 6.8 2.8
Fig. 4. Resonance Raman spectra for the human deoxy myoglobin exited by He/Cd laser. The sample concentration is about 40 lM in 100 mM Tris–Cl, pH 7.5, at room temperature.
was detected for the m9 line, a symmetric bending mode for the Cb–C1 bond [20]. Although the correlation between the heme environmental changes and the positional shift of the m8 and c7 lines or broadening of the m9 line is not clear, it can be concluded that the side chain at position of 107 can interact with the peripheral groups and meso-positions of the porphyrin ring and to affect the vibration modes. Such small structural changes in the heme environment are also evident in the NMR spectrum for the cyanide bound I107F mutant. As displayed in Fig. 5, the spectral pattern for the I107F mutant is quite similar to that of the wild-type protein. The prominent retardation in the heme reorientation reaction for the I107F mutant is, therefore, not due to the structural changes of the protein moiety near the heme peripheral groups. Instead, the bulky side chain at position 107 would increase the activation energy [21] for the rotation of the heme about the a–c axis, corresponding to the heme rotation for the reorientation in the ‘‘protein cage’’ [14]. In sharp contrast to the replacement with the bulky side chain, introduction of an amino acid residue with the small side chain volume, valine, into position 107 (I107V) accelerated the heme reorientation reaction and 150 h are enough to complete the heme reorientation reaction (Fig. 2D; kf = 75 · 104 h1, kb = 6.8 · 104 h1). The resonance Raman (Fig. 4) and NMR (Fig. 5) spectra of the I107V mutant also exhibit only minor perturbations on the heme environmental structure. Contrary to the bulky side chain in the I107F mutant, the reduced steric hindrance at position 107 in the I107V mutant would lower the activation energy for the
H. Ishikawa et al. / Biochemical and Biophysical Research Communications 324 (2004) 1095–1100
Fig. 5. 1H NMR spectra (500 MHz) of the human cyanomet wildtype, I107F, I107L, I107V, and I107A myoglobins at pH 8.0 immediately after reconstituting apoMb with hemin. These spectra were recorded on a BRUKER AVANCE DRX 500 MHz system at 25 C. Five and one molar represent the NMR signals from the 5- and 1-methyl groups of the ‘‘native’’ heme, respectively. Three and 8 m are the NMR signals from the 3- and 8-methyl groups of the ‘‘disordered’’ heme, respectively. His93NdH denotes the NMR signal from the proton at the Nd position of the ‘‘distal’’ histidine (His93).
heme reorientation reaction to facilitate the heme rotation in the ‘‘protein cage.’’ On the other hand, drastic reduction of the side chain volume at position 107 significantly retarded the heme reorientation reaction. The reorientation rate for the alanine-substituted (I107A) mutant (Fig. 3E; kf = 31 · 104 h1, kb = 2.8 · 104 h1), whose side chain volume (91.5 cm3 mol1) at position 107 is much smaller than that of the I107V mutant (168.8 cm3 mol1), was slower than that of the wild-type protein (kf = 43 · 104 h1, kb = 3.9 · 104 h1). We have not yet found out the structural factors responsible for the retardation of the heme reorientation reaction in the I107A mutant, but it should be noted here that the resonance Raman spectrum for the I107A mutant is quite similar to that of the I107F mutant. The m8 line of the I107A mutant appeared at 337 cm1 as found for the I107F mutant, while the wild-type protein and the other mutants gave the m8 line at 342 cm1. In addition to the position of the m8 line, the broad m9 lines were observed for the I107A and I107F mutants. The structural perturbations on the metal–pyrrole nitrogen and
1099
pyrrole Cb–C1 bonds in the I107A mutant are similar to those in the I107F mutation. One of the possible reasons for the close similarity between the two mutants would be that, due to the formation of the large space by the drastically reduced side chain at position 107, the side chain of some other amino acid residue is located near the a-meso position, resulting in a new steric barrier for the heme reorientation. In the I107A mutant, the reduced steric hindrance at position 107 would be no longer a kinetic barrier for the heme reorientation and, instead of Ile107, side chains of other amino acid residues would play a primary role in forming the kinetic barrier. In summary, we found that the side chain at position of 107 can be a steric barrier for the rotation of the heme about the a–c axis in myoglobin. Based on the structural characterization of the I107 mutants, the mutation at position 107 primarily affects the kinetic barrier between the ‘‘native’’ and ‘‘disordered’’ heme orientation. Involvement of the amino acid residues inside the heme pocket in the formation of the energetic barrier for the heme reorientation reaction is clear evidence for the crucial roles of the steric interactions between the side chain of the amino acid residues and heme peripheral groups in the correct heme binding of hemoproteins, which supports the intramolecular reaction mechanism for the heme reorientation in the ‘‘protein cage’’ of myoglobin.
Acknowledgments This work was supported by Grants-in-Aid (12002008, I.M.; 14658217, K.I.) from Ministry of Education, Culture, Sports, Science, and Technology in Japan. H.I. was supported by JSPS Research Fellowships for Young Scientists. We are indebted to Prof. T. Kitagawa (National Institutes of Natural Science) for the measurements of the resonance Raman spectra.
References [1] G.N. La Mar, N.L. Davis, D.W. Parish, K.M. Smith, Heme orientational disorder in reconstituted and native sperm whale myoglobin. Proton nuclear magnetic resonance characterizations by heme methyl deuterium labeling in the Met-cyano protein, J. Mol. Biol 168 (1983) 887–896. [2] G.N. La Mar, Y. Yamamoto, T. Jue, K.M. Smith, R.K. Pandey, 1 H NMR characterization of metastable and equilibrium heme orientational heterogeneity in reconstituted and native human hemoglobin, Biochemistry 24 (1985) 3826–3831. [3] C. Kiefl, N. Sreerama, R. Haddad, L. Sun, W. Jentzen, Y. Lu, Y. Qiu, J.A. Shelnutt, R.W. Woody, Heme distortions in spermwhale carbonmonoxy myoglobin: correlations between rotational strengths and heme distortions in MD-generated structures, J. Am. Chem. Soc. 124 (2002) 3385–3394. [4] M.F. Perutz, Stereochemistry of cooperative effects in haemoglobin, Nature 228 (1970) 726–739.
1100
H. Ishikawa et al. / Biochemical and Biophysical Research Communications 324 (2004) 1095–1100
[5] J.C. Kendrew, Bodo, G., Dintzis, H.M., Parrish, R.G., Wyckoff, H., Phillips, D.C., A Three-dimensional model of the myoglobin molecule obtained by X-ray analysis, Nature 181 (1958) 662–666. [6] A.J. Mathews, T. Brittan, Haem disorder in recombinant- and reticulocyte-derived haemoglobins: evidence for stereoselective haem insertion in eukaryotes, Biochem. J. 357 (2001) 305–311. [7] W. Light, R. Rohlfs, G. Palmer, J. Olson, Functional effects of heme orientational disorder in sperm whale myoglobin, J. Biol. Chem. 262 (1987) 46–52. [8] H.S. Aojula, M.T. Wilson, G.R. Moore, D.J. Williamson, 1H NMR and CD studies of haem orientational disorder in spermwhale myoglobin and human haemoglobin, Biochem. J. 250 (1988) 853–858. [9] T. Ueno, M. Ohashi, M. Kono, K. Kondo, A. Suzuki, T. Yamane, Y. Watanabe, Crystal structures of artificial metalloproteins: tight binding of FeIII(Schiff-base) by mutation of Ala71 to Gly in apo-myoglobin, Inorg. Chem. 43 (2004) 2852–2858. [10] H.J. Yang, T. Matsui, S. Ozaki, S. Kato, T. Ueno, G.N. Phillips Jr., S. Fukuzumi, Y. Watanabe, Molecular engineering of myoglobin: influence of residue 68 on the rate and the enantioselectivity of oxidation reactions catalyzed by H64D/V68X myoglobin, Biochemistry 42 (2003) 10174–10181. [11] Q.H. Gibson, E. Antonini, Kinetic studies on the reaction between native globin and haem derivatives, Biochem. J. 77 (1960) 328–341. [12] T. Jue, R. Krishnamoorthi, G.N. La Mar, Proton NMR Study of the heme-apoprotein reaction for myoglobin, J. Am. Chem. Soc. 105 (1983) 5701–5703.
[13] K. Ishimori, I. Morishima, Study of the specific heme orientation in reconstituted hemoglobins, Biochemistry 27 (1988) 4747–4753. [14] G.N. La Mar, H. Toi, R. Krishnamoorthi, Proton NMR investigation of the rate and mechanism of heme rotation in sperm whale myoglobin: evidence for intramolecular reorientation about a heme twofold axis, J. Am. Chem. Soc. 106 (1984) 6395– 6401. [15] I.D. Campbell, C.M. Dobson, R.J.P. Williams, Proton magnetic resonance studies of the tyrosine residues of hen lysozymeassignment and detection of conformational mobility, Proc. R. Soc. Lond. B. Biol. Sci. 189 (1975) 503. [16] H. Ishikawa, T. Uchida, S. Takahashi, K. Ishimori, I. Morishima, Ligand migration in human myoglobin: Steric effects of isoleucine 107(G8) on O2 and CO binding, Biophys. J. 80 (2001) 1507–1517. [17] Y. Yamamoto, G.N. La Mar, 1H NMR study of dynamics and thermodynamics of heme rotational disorder in native and reconstituted hemoglobin A, Biochemistry 25 (1986) 5288–5297. [18] G.N. La Mar, D.L. Budd, D.B. Viscio, K.M. Smith, K.C. Langry, Proton nuclear magnetic resonance characterization of heme disorder in hemoproteins, Proc. Natl. Acad. Sci. USA 75 (1978) 5755–5759. [19] F.W.J. Teale, Cleavage of the haem–protein link by acid methylethylketone, Biochim. Biophys. Acta 35 (1959) 543. [20] S. Hu, K.M. Smith, T.G. Spiro, Assignment of protoheme resonance Raman spectrum by heme labeling in myoglobin, J. Am. Chem. Soc. 118 (1996) 12638–12646. [21] C. Chothia, Structural invariants in protein folding, Nature 254 (1975) 304–308.