The room temperature magnetic susceptibilities of human and carp oxy-and carbonmonoxyhemoglobin solutions were measured in a 30-gauss (1 G = 10-4 T) field with a superconducting magnetometer. To within experimental uncertainty, the susceptibility was the same for both the oxy and carbonmonoxy forms, and salt concentration did not affect it. A variety of sample preparations were used; the iron chemical state was verified by Mossbauer spectroscopy. A value of -0.580 ± 0.010 x 10-6 centimeter-gram-second (cgs) system was obtained for the mass susceptibility of the protein.We attribute the paramagnetism sometimes observed in oxyhemoglobin solutions to the presence of a small amount of the deoxy form.A controversy has recently grown over the magnetic susceptibility of oxy-and carbonmonoxyhemoglobin under various solution conditions. The data of Pauling and Coryell (1), Taylor and Coryell (2), and Havemann et al. (3) indicate that both forms of the protein are totally diamagnetic at room temperature. Recent work (4, 5) shows spin unpairing in oxyhemoglobin with a singlet-triplet separation of 144 cm'1. Further evidence for an excited state was the observation of temperature-dependent quadrupole splittings and linewidths in the Mossbauer spectra of oxyhemoglobins (6). The diamagnetism of carbonmonoxyhemoglobin came under question with reports of salt-dependent effects (7). Most recently, one researcher has again found oxyhemoglobin to be diamagnetic (8) but in solution conditions vastly different from those in which paramagnetism was seen.The resolution of this conflict is imperative for the development of valid electronic models of reversible oxygen binding to hemes. Therefore, we have constructed a superconducting magnetometer to measure the magnetic susceptibilities of oxy-and carbonmonoxyhemoglobin at room temperature and have attempted to reproduce the conditions of the work which showed paramagnetism (4, 5, 7) as closely as possible.
MATERIALS AND METHODSA superconducting magnetometer built in our laboratory was used to make the measurements. Fig. 1). These samples proved to be more paramagnetic than water. A Mossbauer spectrum of one of these, frozen to 77 K immediately after the susceptibility measurement, indicated about 9% deoxyhemoglobin, a spin 2 compound. After this discovery, Mossbauer spectra were taken immediately after all susceptibility measurements. Mossbauer measurements, although less sensitive than optical spectroscopy, avoid dilution and the consequent risk of changing the oxy/deoxy equilibrium.We found that it was difficult to prepare samples free of deoxyhemoglobin by exposing concentrated solutions (15 mM in heme) to air. If a concentrated, mostly oxygenated sample were exposed to CO, all deoxyhemoglobin would be replaced by carbonmonoxyhemoglobin, but a significant fraction of the protein remained in the oxygenated form. By diluting a sample to about 5 mM in heme, we could reliably prepare pure oxy or carbonmonoxy proteins. More concentrated solutions were then obtained through the u...