Interconversion of Fast and Slow Forms of Cytochrome bo from Escherichia coli

The fully oxidized fast form of cytochrome bo from Escherichia coli is shown to convert spontaneously to a slow form when stored at -20 °C in 50 mM potassium borate, pH 8.5, containing 0.5 mM potassium EDTA. Evidence for the conversion, and that the form produced is analogous to the slow form of bovine heart cytochrome c oxidase, comes from decreases in the extents of fast (k = 1-2 x 10³ M^-1 s^-1) H₂O₂ binding and fast (k = 20-30 M^-1 s^-1) cyanide binding; (b) changes in the optical spectrum that are like those induced by formate, i.e., a blue shift in the Soret absorption band, loss of absorbance in the α and β bands, and a red shift in the “630 nm” charge-transfer band; (c) changes in the EPR spectrum that are like those induced by formate, i.e., disappearance of signals at g = 8.6 and g = 3.71, and appearance of signals at g ≈ 13, g = 3.14, and g = 2.58; and (d) appearance of a slow phase of reduction of heme o by dithionite. The mutant enyzme E286Q also converts to a slow form under the same conditions, as shown by a decrease in the extent of fast H₂O₂ binding; (b) changes in the optical spectrum like those seen with wild-type enzyme; and (c) changes in the EPR spectrum that are like those induced by formate, i.e., disappearance of signals at g = 7.3 and g = 3.6 and appearance of signals at g ≈13, g = 3.18, and g = 2.59. Hence, models for the fast to slow conversion that specifically involve ligation of the high-spin heme or Cub by E286 (or by the equivalent residue, E242, in bovine heart cytochrome c oxidase) are unlikely to be correct.