Spectrophotometric Analysis Of Hemoproteins

The unique chromophore of hemopro-teins allows the use of a number of spec-trophotometric techniques to yield detailed information on the nature of the heme lig-and. Although individual spectroscopy techniques can yield valuable information, it is the combined use of techniques such as UV/visible, magnetic circular dichroism (MCD), resonance Raman, and NMR that can provide powerful insight as to the nature of the coordination, ligation, and oxidation state of the heme.

It will not be the focus of this review to give a detailed account of each technique, but present a brief synopsis of the information that can be gained as it relates to hemoprotein structure and function. On studies of the human heme oxygenase (HO-1), we have previously used a combination of spectrophotometric techniques such as UV/visible, resonance Raman, MCD, and NMR spectrophotometry to determine the ligation, coordination, and electronic nature of the heme. In addition to providing valuable information on the ligation and coordination state of the heme, these techniques have aided greatly in elucidating and confirming the mechanism of action of the enzyme.

HO-1 catalyzes the rate limiting step in the conversion of heme to biliverdin (Figure 2), and although not a hemoprotein on reconstitution with heme, the enzyme forms a 1:1 ferric heme:HO-1 (heme:HO-1) complex with electronic spectra similar to that of myoglobin (128). At pH 6.0, the heme:HO-1 complex has a Soret at 404 nm and a charge transfer band at 632 nm, characteristic of a high spin species (Figure 3b). On increasing the pH from 6.0 to 10.0, the heme Soret shifts from 404 to 409 nm, and the charge transfer band at 630 nm decreases with the appearance of a/p bands at 539 and 574 nm, a process attributed to a pH-dependent high-spin to low-spin conversion of a water ligand to a hydroxide (Figure 3b) (40,111). The UV/visible spectra of the reduced complex in the presence of CO has a Soret at 420 nm and a/p bands at 567 and 537 nm, consistent with a histidine proximal ligand.

Although the UV/visible spectrum can provide some limited information on the proximal ligand of a hemoprotein of unknown structure, it is not conclusive. MCD is a difference technique that will have both positive and negative components yielding a more detailed spectrum than the corresponding UV/visible spectrum. The increased sensitivity of MCD has been shown to be an effective "fingerprinting tool" for the determination of the ligand coordination, oxidation state, and spin state of hemoproteins (15,123). Assignment of the proximal ligand in a hemoprotein of unknown structure by MCD, like UV/visible spectroscopy and resonance Raman (see below), is dependent on comparison to a protein whose proximal ligand is known, such as myoglobin

(histidine), catalase (tyrosine), and cytochrome P-450 (cysteine). The sample to be analyzed is prepared with a known sixth ligand such as cyanide, which together with the 4 pyrroles can then be compared to the ferric cyano complex of hemopro-teins, whose proximal ligands are known. The electronic absorption and MCD spectra of the cyanoferric heme:HO-1 complex is very similar to that of cyanoferric myo-globin, thus confirming the proximal histi-

dine ligation in the heme:HO-1 complex (Figure 3 a) (40). The derivative bands of the heme:HO-1 complex are blue-shifted when compared to myoglobin, but are still very similar in structure (Figure 3b) (40).

Raman spectroscopy is a vibrational spectroscopy technique that can detect transitions between different ligation states of porphyrins as the spin state is changed (107,109). The change in spin state alters the size of the iron and its displacement

Figure 1. (a) Labeling of active-site residues of hemo-proteins with photoac-tivable probes. (b) Structure of the photoactivatable probes of trifluoromethyl-diazirinylphenyldiazene as synthesized by Tschirret-Guth et al. (118).

from the heme plane which directly effects the vibrational frequencies of the porphyrin ring bond stretches. This technique can be used to characterize the spin state and coordination state in both the ferric and ferrous oxidation states of a given hemoprotein (108). As in the MCD experiments, the resonance Raman spectra of the heme:HO-1 complex show a marked change in the heme axial coordination and spin state on changing pH (111). At pH 6.0, the spectrum shows the heme to be primarily 6-coordinate high-spin (6CHS) similar to myoglobin, in which the heme is coordinated through a proximal histidine with water bound as the sixth ligand. Increasing the pH to 8.0 significantly alters the Raman spectrum, with the bands at 1483 (U3) and 1565 cm-1 (u2) of the ferric 6CHS species decreasing, and the bands corresponding to the ferric 6-coordinate low-spin species (6CLS) at 1503 (U3), 1582 (u2), and 1638 (U10) increasing (Figure 4). This again supports the ionization of a coordinating water molecule to a hydroxide above pH 8.0.

Resonance Raman studies of the heme: HO-1 complex provided direct confirmation of histidine as the proximal ligand (111). The iron—imidazole or iron—thiolate ligand is not generally detectable due to the weak coupling of the n electronic system. However, in the case of the iron—imidazole (Fe-His) ligation, the reduced ferrous 5-coordinate species shows significant enhancement due to the iron being out of the plane of the porphyrin. This transition provides a characteristic marker for the Fe-His stretching mode in the low frequency ferrous spectra. Interestingly, the position

Figure 2. Conversion of heme to biliverdin by HO-1. The abbreviations are as follows; Me, methyl; V, vinyl; Pr, propionic acid.

of the band is also a strong indication of the ionization state of the proximal histi-dine. The band at 216 cm-1 in the heme:HO-1 complex (111) is close to that of myoglobin (221 cm-1) (53,55), where the proximal histidine is only weakly hydrogen-bonded (Figure 5). The presence of strong hydrogen bonding to the proximal histidine in cytochrome c peroxidase shifts the iron—imidazole stretch to 233 cm-1, while complete ionization gives a value closer to 246 cm-1 (105,115). In cytochrome c peroxidase, it has been shown that removal of the hydrogen bonding network by mutation of Asp-235 to an Ala shifts the Fe-His stretching mode from 246 to 206 cm-1. The relatively low value for the heme:HO-1 complex can therefore be interpreted as a proximal histidine that is weakly or not hydrogen-bonded (106).

It is interesting that the structure of the heme:HO-1 complex resembles the oxygen-binding proteins more than the oxygen-activating peroxidases. The presence in the heme:HO-1 complex of a histidine proximal ligand that is neither ionized nor hydrogen-bonded has strong implications for the mechanism of action of HO-1. The ability of the proximal ligand to destabilize the dioxygen bond and assist in cleavage to form the activated ferryl species is largely dependent on the electron density present on the proximal ligand. As the electron density increases by partial or complete deprotona-tion of the proximal histidine, the ability to carry out dioxygen bond cleavage increases.

In previous studies, we have generated a ferryl species analogous to that of com

Heme Absorption Spectrum

Figure 3. (a) MCD and (b) electronic absorption spectra of the cyanoferric heme:HO-1 complex and of the cyanoferric sperm whale myoglobin. The

MCD spectra of the human HO-1 and sperm whale myoglobin have previously been reported in References 40 and 16, respectively.

Figure 3. (a) MCD and (b) electronic absorption spectra of the cyanoferric heme:HO-1 complex and of the cyanoferric sperm whale myoglobin. The

MCD spectra of the human HO-1 and sperm whale myoglobin have previously been reported in References 40 and 16, respectively.

pound II in peroxidase and shown that this species is not an intermediate in the normal physiological reaction of heme oxygenase (128). Subsequent studies with ethyl-hydroperoxide showed that the reaction proceeded by the normal pathway with the formation of a-meso-ethoxyheme, analogous to the formation of a-meso-hydroxy-heme in the enzymatic conversion of heme to biliverdin (130). The ethylhydroperox-ide reaction, therefore, favors an electro-philic mechanism with the addition of the terminal ethoxy or hydroxy (Fe-O-OH) moiety to the heme (Figure 6). The addition of the terminal oxygen of the ferric peroxo anion (Fe-O-O-) to the a-meso-edge of the heme is ruled out as the terminal oxygen of the ethylhydroperoxide species is blocked. The proposed elec-

trophilic mechanism is consistent with the presence of a proximal ligand that is neither ionized or hydrogen-bonded.

Isotopic labeling and 2-dimensional NMR studies on the rat HO-1 have revealed an unusual heme electronic structure (43). The proton contact shift patterns of the heme reveal large differences in the delocalized spin density in which the spin density is highest at positions 3 (methyl) and 2 (vinyl) and much smaller at positions 1 (methyl) and 4 (vinyl) (Figure 7). The asymmetric spin density within a pyrrole is unprecedented in any low-spin ferric hemoprotein, where it is more common to find asymmetry between different pyrroles. Based on studies with model hemes, the distribution of the spin density suggests a direct electronic perturbation of the heme

Figure 4. Resonance Raman spectra of the heme:HO-1 complex in the spin and coordination state frequency region as a function of pH (413.1 nm excitation). The resonance Raman data for the heme:HO-1 complex has previously been reported (111).

by the protein environment such that there is an increase in electron density at the a-meso-edge of the heme. The increased electron density at the a-meso-edge would facilitate the electrophilic addition of the protonated terminal oxygen to the a-meso-position (Figure 6).

The extensive characterization of the heme:HO-1 complex by spectroscopic techniques such as MCD, resonance Raman, and NMR have greatly aided the biochemical and enzymatic studies in elucidating the mechanism of action of HO-1. It is also gratifying that the recent 3-dimen-sional structure of the heme:HO-1 complex has in large part confirmed many aspects of the previous spectroscopic data (97).

0 0

Post a comment