The heme propionate groups of both myoglobin (Mb) and cytochrome b5 form hydrogen bonds with nearby surface amino acids residues that are believed to stabilize the heme-protein complex. To evaluate the magnitude of this stabilization, the kinetics of heme dissociation from variants of horse heart Mb and cytochrome b5 in which these hydrogen bonding interactions have been systematically eliminated were studied by the method of Hargrove and colleagues (1994), and their thermal stability was assessed. Elimination of each hydrogen bond was found to decrease the thermal stability of the proteins and increase the rate constant for heme dissociation in a progressive fashion. For the Mb derivatives, 1H-NMR studies indicate that the elimination of individual hydrogen bonds also affects the rate at which the heme orientational equilibrium is achieved. In both types of kinetics experiment, the effects of decreasing the number of potential hydrogen bonding interactions are found to be cumulative. Despite their kinetic effects, elimination of these hydrogen bonding interactions had no influence on the initial distribution of heme orientational isomers immediately following reconstitution or on the equilibrium constant of heme orientational disorder. The interactions between the heme propionates and nearby protein residues play a partial role in the stabilization of the heme-protein complex and are a major factor in the kinetic "trapping" of the minor heme orientation. Comparisons of the various rate constants determined for the mechanism of heme binding and reorientation suggests that the intramolecular reorientation mechanism is slightly favored over the intermolecular mechanism.

The interaction of azide with variants of horse heart myoglobin (Mb) has been characterized by Fourier transform infrared (FTIR), electron paramagnetic resonance (EPR), and UV-VIS absorption spectroscopy and by molecular modeling calculations. Distal histidine variants (His64Thr, His64Ile, His64Lys) and charged surface variants (Val67Arg, Lys45Glu, Lys45Glu/Lys63Glu) were included in this study. All variants, with the exception of Val67Arg, have a lower azide affinity than the wild-type protein. Analysis of the temperature dependence of the FTIR spectra (277-313 K) revealed that the wild-type protein and all variants exhibit a high-spin/low-spin equilibrium. Introduction of positively charged amino acid residues shifts nu max for the low-spin form to higher energy while negatively charged residues shifted this maximum to lower energy. The low azide binding affinity exhibited by the His64Thr and His64Ile variants is accompanied by a shift of the nu max for the low-spin infrared band to lower energy and by a significant increase in the corresponding half-bandwidths. This observation indicates greater mobility of the bound azide ligand in these variants. The His64Lys variant exhibits two infrared bands attributable to low-spin forms that are assigned to two different conformations of the lysyl residue. In one conformation, the lysine is proposed to form a hydrogen bond with the bound azide similar to that proposed to occur between the distal histidine and bound azide, and in the other conformation no interaction occurs.

The structural and functional consequences of the introduction of a negatively charged amino acid into the active site of horse heart myoglobin have been investigated by replacement of the proximal Ser92 residue (F7) with an aspartyl residue (Ser92Asp). UV-visible absorption maxima of various ferrous and ferric derivatives and low-temperature EPR spectra of the metaquo (metMb) derivative indicate that the active site coordination geometry has not been perturbed significantly in the variant. 1H-NMR spectroscopy provides direct evidence for the existence of a distal water molecule as the sixth ligand in the oxidized form of the variant at pD 5.7. Spectrophotometric pH titration of the Ser92Asp variant is consistent with this finding and with a pKa = 8.90 +/- 0.02 [25.0 degrees C, mu = 0.10 M (NaCl)] for titration of the distal water molecule, identical to the value reported for the wild-type protein. X-ray crystallography of the metMb derivative indicates that the heme substituents conserve their orientations in the variant protein, except for a slight reorientation of the pyrrole A propionate group to which Ser92 normally hydrogen bonds and reorientation of the carboxyl end of the pyrrole D propionate group. No change is observed in conformation of the proximal (His93) or distal (Wat156) heme ligands. 1H-NMR spectroscopy of the metMbCN form of the protein indicates that a slight rotation of the proximal His93 ligand has occurred in this derivative. Resonance Raman experiments indicate increased conformational heterogeneity in the proximal pocket of the variant. Failure to detect electron density for the Asp residue in the X-ray diffraction map of the variant protein and high average thermal factors for the pyrrole A propionate substituent are consistent with this observation. The variant exhibits novel pH-dependent behavior in the metMb form, as shown by 1H-NMR spectroscopy, and provides evidence for a heme-linked titratable group with a pKa of 5.4 in this derivative. The metMbCN and deoxyMb derivatives also exhibit pH-dependent behavior, with pKas of 5.60 +/- 0.07 and 6.60 +/- 0.07, respectively, compared to the wild-type values of 5.4 +/- 0.04 and 5.8 +/- 0.1. The heme-linked ionizable group is proposed to be His97 in all three derivatives. The reduction potential of the variant is 72 +/- 2 mV vs SHE [25.0 degrees C, mu = 0.10 M (phosphate), pH 6.0], an increase of 8 mV over the wild-type value. The possible influence of a number of variables on the magnitude of the reduction potential in myoglobin and other heme proteins is discussed.