Heme-hemopexin

E. Reduction of Heme-Hemopexin and Binding of Exogenous Ligands... 

The dissociation rate of heme from methemoglobin and consequent formation of the heme-hemopexin complex is facile at 37°C, and the presence of small amounts of H2O2 (even below levels obtained from the respiratory burst of neutrophils) dramatically increases heme dissociation from oxyhemoglobin (55). The binding of heme by hemopexin prevents the oxidation of lipoprotein (50,55,56) and lipid and membrane damage (57-59). 

The hemopexin heme transport system thus acts as an early warning system for cells by activating signaling pathways (including the N-terminal c-Jun kinase, kinases to release RbIA/NFkB family members for nuclear translocation) and transcription of the HO-1 and MT-1 genes. The details of this aspect of hemopexin function with redox-active copper as an initial event in the coordinated induction of HO-1 and MT-1 by heme-hemopexin have recently been reviewed (89) and are not presented in detail here. 

Fig. 3. Absorbance spectra of ferri- and ferro-heme-hemopexin in the Soret and visible regions. The spectra of rabbit ferri- and ferro-mesoheme-hemopexin (solid hne and dashed line, respectively) are shown and are typical of low-spin bis-histidyl heme proteins. Other species display similar spectra, and a variety of other 2,4-substituted heme analogs are also bound by hemopexin.

Pig. 4. CD spectra in the near and far UV of apo- and heme-hemopexin. The CD spectra of rabbit apo- and heme-hemopexin (solid line and dashed line, respectively) at pH 7.4 in 0.05 M sodium phosphate buffer are shown. The increases in ellipticity in the near UV are attributable to changes in tertiary conformation leading to altered environments of aromatic residues, particularly tryptophan. The unusual positive ellipticity in the far UV is attributable to tryptophan-tryptophan interactions that are perturbed by heme binding 124, 130). This positive signal precludes analysis of the secondary structure of hemopexin using current CD-based algorithms. 

Fig. 6. Deconvolved amide F region FTIR spectra of apo- and heme-hemopexin. The amide F FTIR spectra of apo- and heme-hemopexin in D2 O were recorded and curve-fitted to resolve the individual bands. The differences between the original and fitted curves are shown in the upper traces in the panels. The estimated helix (15%), beta (54%), turn (19%), and coil (12%) content of the apo-protein are not significantly changed upon heme binding 104). This analysis was required because of the positive 231-nm elhpticity band in hemopexin and is consistent with the derived crystal structure results.

Determination of the structure of the entire rabbit heme-hemopexin complex (11) (Fig. 8) clearly revealed the expected complementary structures of the N- and C-domains. More importantly, several fundamental... 

Fig. 8. Crystal structure of heme-hemopexin. The crystal structure of the rabbit mesoheme-hemopexin complex (PDB accession number IQHU) (11) showed heme to be bound in a relatively exposed site between the N- and C-domains with one axial His ligand being contributed by the hinge or linking region between the domains and the other by the C-domain. Also noteworthy is the disposition of the heme with its propionate residues pointing inward and neutralized by positive charges in the binding site.

Fig. 9. EPR spectra of heme-hemopexin and heme-N-domain. X-band EPR spectra at 4 K of ferri-mesoheme-hemopexin (a) and ferri-mesoheme-N-domain (b) are shown. The concentration of both heme complexes was 0.15 mM in 50 50 (v/v) 10 mM sodium phosphate/150 mM NaCl (pH 7.2) glycerol. The g-value scale is noted at the top and the -values observed are noted in each spectrum. Although both complexes are low-spin (some adventitious high-spin iron is present), the differences in g-values indicate nonidentical heme environments in the two complexes (.114).

Exposure of ferri-heme-hemopexin to imidazole or KCN can displace one or both of the heme coordinating His residues, but millimolar concentrations are required (138). Other potential ligands such as azide or fluoride are inactive. This coordination stability of the ferri-heme-hemopexin bis-histidyl complex, despite the exposed heme site, is home out by thermal imfolding studies (Section IV,F). Reduction of the heme-hemopexin complex, however, has dramatic effects on its stability. 

The biphasic reaction with CO points to the existence of multiple heme-hemopexin conformers, and this is borne out by spectral analyses. The absorbance spectra of rabbit ferri-, ferro-, and CO-ferro-mesoheme-hemopexin are entirely analogous to those of other bis-histidyl heme proteins such as cytochrome 65 142), but the CD spectra exhibit unusual features (Fig. 11). Of particular interest are the weak signal of the ferro complex and the bisignate signal of the CO-ferro complex (also seen in the NO-ferro-mesoheme-hemopexin complex (140) and in human ferri-protoheme—hemopexin (139)). 

The ferro-complex CD spectrum shows that reduction of the heme iron alters the heme environment. Redox-induced protein conformation changes could alter the S5unmetry in the heme pocket or produce two binding modes for the reduced complex whose asymmetries nearly cancel each other. Redox-linked conformational changes are especially interesting in view of recent findings of oxido-reductase activity associated with the heme-hemopexin-receptor interaction (89). 

Fig. 12. Schematic views of bis-histidyl ferri-, ferro-, and CO-ferro-heme-hemopexin. Unlike myoglobin with one open distal site, heme bound to hemopexin is coordinated to two strong field ligands, either of which a priori may be displaced by CO. This may well produce coupled changes in protein conformation like the Perutz mechanism for 02-binding by hemoglobin (143). The environment of heme bound to hemopexin and to the N-domain may be influenced by changes in the interactions of porphyrin-ring orbitals with those of aromatic residues in the heme binding site upon reduction and subsequent CO binding.

However, this explanation is not sufficient to accoimt for the bipha-sic CD spectrum of human ferri-protoheme—hemopexin (with 2,4-vinyl substituents), as well as the much weaker human CO-ferro-heme-hemopexin bisignate signal compared to the rabbit congener (139), and hence other factors must be involved. Several potential effectors exist (a) exciton coupling (b) the conformers produced by a 180° rotation about the a- and y-meso-carbon axis and consequent nonisometric interactions of the as5unmetric 2,4- and 9,10-substituents (c) the aromatic tryptophan residues near the heme binding site (s) and (d) two independent binding modes or sites. 

A bisignate spectrum is characteristic of exciton coupling between identical chromophores on one molecule (144), an impossibility with the 1 1 heme-hemopexin complex. Bisignate CD spectra have been observed with bilirnbin-albnmin however, bilirubin, nnlike the planar molecnle heme, can adopt an extended, helical stacked conformation... 

Fig. 13. DSC of apo-hemopexin and heme-hemopexin. DSC recordings of apo- and heme-hemopexin in 0.05 M sodium phosphate at pH 7.4 and of a mixture of the two are presented. The stabihzation of hemopexin upon heme binding (706) is evident. Other results established an even greater stabihzation of N-domain by heme.

Importantly, heme coordination in ferri-heme-hemopexin, as monitored by Soret absorbance, is strikingly more stable than in the reduced form, both with and without NaCl (Tm 70°C and 51°C 55.5°C and... 

Fig. 15. Effects of pH on apo- and heme-hemopexin. The Soret region absorbance (filled squares) of rabbit heme-hemopexin was monitored in two separate titrations, from pH 7.4 to 11.8 in one and from pH 7.4 to 3.8 in the other. Similarly, theellipticity at 231 nm of apo-hemopexin (open circles) and of heme-hemopexin (filled circles) was assessed from pH 7.4 to 11.8 and from pH 7.4 to 1.7 111). The heme complex and the tertiary structure are unaffected by pH in the region from pH 6 to 9, and other values are normalized to these.

In patients clearance of intravenous heme is rapid until hemopexin levels are depleted (148), and lack of interaction with hemopexin may explain the higher clinical efficacy of heme-arginate compared with hemin itself (149, 150). In intact animals, i.v. heme causes rapid association of hemopexin but not albumin with the liver (47, 63, 68), and heme uptake from heme-albumin complexes into isolated rat hepato-cytes occurs via diffusion of heme released from its loose complex with BSA (137). Moreover, unlike uptake from heme-hemopexin, free heme uptake by cells occurred even at 4°C, as expected for nonspecific membrane association and in total disagreement with a membrane-receptor-mediated or active transport uptake process. 

The heme—hemopexin—receptor complex is internalized via receptor-mediated endocytosis (160), presumably with the heme in the reduced state, and experiences a lower pH in the endosome. In the ferro-form, the acidic pH in the endosome may further weaken the association between heme and hemopexin, and heme is likely to be released at this... 

Ren Y, Smith A. 1995. Mechansim of metallothionein gene regulation by heme-hemopexin. J. Biol. Chem. 270 23988-95... 

---