Ultrafast Ligand Rebinding to Protoheme and Heme Octapeptide at Low Temperature

Ultrafast flash photolysis is used to investigate the rebinding of carbon monoxide to protoheme (PH) and heme octapeptide (HO) in a glycerol-water glass at 100 K. Kinetic decay data are obtained at several probing wavelengths in the Soret bands. Previous work [Hill, J. R., et al. Springer Ser. Chem. Phys. 1986, 46, 433] had shown the existence of two distinct ligand rebinding processes. The first, denoted process I*, is exponential in time, with a rate of ≈3 × 10¹⁰ s^-1. Process I* involves non-Arrhenius rebinding, as evidenced by a roughly linear dependence of the rate constant on temperature. The second, denoted process I, involves nonexponential rebinding which does obey the Arrhenius expression for a system with distributed activation barriers. Kinetic measurements performed at the isosbestic point between carboxy heme (denoted Fe-L) and deoxy heme (denoted Fe + L) demonstrate the existence of two short-lived intermediates (denoted Fe^I-L and Fe*-L). Multiple-flash experiments at 15 K show that these species cannot be easily interconverted, implying they exist as two distinct, inhomogeneous sets of reacting molecules. One reacting subensemble consists of Fe-L species which, upon photolysis, relax at a rate of ≈10¹² s^-1 into Fe + L and then rebind ligands via process I. The other consists of molecules which are unable to relax into Fe + L states, and which rebind ligands via process I*. The difference between the two subensembles is attributed to solvent interactions which constrain the geometry of some molecules to inhibit the relaxation process. The rebinding parameters of PH and HO are compared at 100 K. The HO differs from PH by the addition of a basket handle peptide chain which donates a proximal histidine. HO relaxes faster (1 ps) after deligation than does PH (1.5 ps). Process I* rebinding to unrelaxed hemes has nearly the same rate in PH and HO. This observation is consistent with a picture of ligand rebinding to planar, unrelaxed hemes which differ only in proximal substitution. Process I is much slower in HO, probably because proximal tension exerted by the peptide chain on the relaxed heme increases the activation barrier.