An NMR characterization of the 98Arg
Cys variant of iron (III)-containing cytochrome
b562 from
Escherichia coli has been performed and the solution structure obtained. This variant has acovalent bond between the heme and Cys 98, thus mimicking the heme binding in cytochrome
c. TheR98C cytochrome is shown to have a significantly increased stability, compared to that of wild type,toward thermal and chemical denaturation. In water at 20
C it is 5.60 kJ mol
-1 more stable than the WTprotein, measured by equilibrium guanidine hydrochloride denaturation. The structure has been obtainedthrough two-dimensional total correlation spectroscopy (TOCSY) and nuclear Overhauser effectspectroscopy (NOESY) experiments and through three-dimensional NOESY-
15N heteronuclear multiplequantum coherence (HMQC). By these methods, 85% of protons and 100% of backbone nitrogens wereassigned. 2145 meaningful nuclear Overhauser effects (NOEs) (20 NOEs per residue), 45 backbone
3Jvalues, and 397 pseudocontact shifts were used to obtain a family of 35 members, which were then energy-minimized. The root-mean-square deviation (RMSD) with respect to the average structure is 0.50 ± 0.07for the backbone and 1.01 ± 0.08 for the heavy atoms. The magnetic anisotropy resulting from analysisof the pseudocontact shifts indicates an anisotropy that is an intermediate between that of the wild-type,which is the smallest, and cytochrome
c. The
g values confirm a higher anisotropy of the variant withrespect to the wild-type protein. The chirality of the heme 2
carbon is the same as that in all naturallyoccurring cytochromes
c. The overall secondary structure and tertiary structure are very similar to thewild type. The removal of Arg 98 causes a change in the pH-dependent properties. The p
Ka, proposed tobe due to deprotonation of the coordinated histidine, is 1.5 units higher than in the wild type, consistentwith the lack of the positive charge of Arg 98 close to the ionizable group. This is further support for thecoordinated histidine being the titratable group with an alkaline p
Ka in the wild-type protein. The patternof the shifts of the heme methyl groups is different than in the wild-type protein, presumably due toalteration of the electronic structure by the presence of the covalent bond between the protein and theheme. The difference in stability between the variant and wild-type protein is discussed in terms of thestructural information.