Structures of deoxy and carbonmonoxy haemoglobin Kansas in the deoxy quaternary conformation

Abstract

Haemoglobin Kansas (Asn102(G4)β → Thr) is the only widely studied mutant or modified haemoglobin having four functional haems and displaying lower than normal oxygen affinity. Two forms of this mutant have been investigated by X-ray diffraction. The deoxy form, whose crystals are isomorphous with the native form, has been examined directly by the difference Fourier technique (3.4 Å). In addition to the replaced residue itself, the difference electron density map shows signs of slight movements throughout the region between α and β haem pockets. However, neither type of chain shows stereochemical evidence of a very abnormal oxygen affinity in the tetramer. The nature of the perturbation is different from that proposed in the earlier, low-resolution study of Greer (1971 a). Exposure of deoxy crystals to carbon monoxide produces two new crystal forms similar but not identical to the native deoxy form. One of these structures has been solved by rotation and translation function methods and a difference map between carbonmonoxy haemoglobin Kansas in the deoxy quaternary structure and native deoxy haemoglobin has been calculated at 3.5 Å resolution. Carbon monoxide molecules are observed at three of the four haems, and two unidentified large peaks ‡ ‡ Note added in proof: The large positive peak near the Cys93β SH group of THbCO Kansas (peak X in Plate IX) appears to be an artefact due to the presence at this site of a mercury atom in p-mercuribenzoate Hb (the most important heavy atom derivative used in determining the native structure). Since the native Fourier has a large negative peak here (due to small phasing errors), the reconstituted Kansas cell did also; when this structure was “subtracted” from the THbCO Kansas crystal structure (having near zero density at this site) a large positive difference peak resulted. The effect was not noticed before in simple difference Fouriers (i.e. deoxyHb Kansas) because in that case neither set of ¦F¦'s contains any information regarding heavy atoms or errors in phasing. The measured crystal sulphydryl reactivity is now seen to be correct (see Materials and Methods, section vii). appear next to the sulphydryl groups of Cys93β. None of the interchain salt bridges which stabilize the deoxy quaternary state appears to be broken, but large tertiary structural changes are seen in the liganded chains. It seems that when the molecule is subjected to the lattice constraints of the deoxy crystal, the salt bridges do not break upon ligand binding, even though the pH dependence of the first Adair constant and the linearity of proton release with ligand uptake both imply that this does happen to stripped HbA in solution.