MOLECULAR FEATURES OF OXYGEN AVIDITY DISCERNED FROM STRUCTURES OF ASCARIS HEMOGLOBIN DOMAIN I AND SEVERAL MUTANTS. F. Scott Mathews^a, Louise M. Cunane^a, Jian Yang^a, Andrew P. Kloek^b and Daniel E. Goldberg^b,c. ^aDepartment of Biochemistry and Molecular Biophysics, ^bDepartment of Molecular Microbiology and ^cDepartment of Medicine, Washington University School of Medicine, St. Louis, MO 63110 USA.

The structures of the N-terminal domain of Ascaris hemoglobin and of several mutants have been determined at 2.2 Å resolution. This perienteric hemoglobin from a parasitic nematode has an exceptionally high affinity for oxygen. It is an octameric protein containing two similar heme-binding domains per subunit. The recombinant monomeric N- terminal heme-binding domain (D1) retains full oxygen avidity. Its structure reveals a characteristic globin fold. A strong hydrogen bond between tyrosine B10 and the ligand distal oxygen, combined with a weak hydrogen bond between glutamine E7 and the proximal oxygen, grip the ligand in the binding pocket. A third hydrogen bond between these two amino acids appears to stabilize the structure. Mutation of B10 YL increases the oxygen dissociation rate of D1 about 500 fold. An initial difference Fourier indicated negative density at the site of tyrosine B10, consistent with substitution by leucine; refinement of the model indicated little additional change in structure, confirming the importance of the strong hydrogen bond of the tyrosine hydroxyl to dioxygen. Mutation of E7 QL increases the oxygen off-rate 5-fold. Analysis of this mutant shows a slight reduction of the E7 side chain volume and the elimination of the hydrogen bonds both to the oxygen and the tyrosine. In addition, there appears to be a slight reorientation of the heme group toward the mutated side chain and a 1.3Å movement of the FG loop toward the exposed edge of the heme group. Mutation of E7 QN increases the oxygen off-rate 50-fold. The mutated E7 Asn is too far from the oxygen to form a hydrogen bond, but a hydrogen bond to the B10 Tyr side chain is still maintained. There appears to be no movement of the heme group, in contrast to the E7 QL mutant, but the FG loop has moved further toward the heme, 1.6 Å, and the side chain of an Asp in this loop now appears to form a hydrogen bond to a heme propionate. This latter interaction may prevent the movement of the heme group to protect the oxygen-binding pocket, seen in the E7 QL mutant, leading to a large increase in the oxygen dissociation rate.