Limits of cryofixation as seen by Fourier transform infrared spectra of metmyoglobin azide and carbonyl hemoglobin in vitrified and freeze-concentrated aqueous solution.

The limits of cryofixation were probed by investigating metmyoglobin azide and carbonyl hemoglobin in approximately 5 wt% aqueous solution by Fourier transform infrared spectroscopy. Spectra of solutions cooled slowly and recorded in steps between 295 K and 190 K are compared with those obtained by "hyperquenching" either into their glassy states at 80 K, or into freeze-concentrated solution at 170 K. For metmyoglobin azide we conclude from an analysis of its covalently and ionically bound azide that it is impossible to freeze-in its high-spin/low-spin equilibrium even by hyperquenching, and that its vitrified state must correspond to a temperature T < 226 K for the Fe(II) site of the protein. In the amide I spectral region of carbonyl hemoglobin (HbCO), a band at approximately 1654 cm-1 due to alpha-helical structures is the dominant band in spectra recorded at ambient temperature and in the vitrified state, but in the spectrum of HbCO quenched at similar rates into a freeze-concentrated state, a band at approximately 1650 cm-1, tentatively assigned to unordered structures, becomes the dominant feature. This band is absent in the spectra of freeze-concentrated samples obtained by heating a vitrified sample to 170 K. We surmise that HbCO is dehydrated by freeze-concentration to a larger extent in solution quenched rapidly at 170 K than in a vitrified solution heated to 170 K, and that this dehydration is the primary cause for HbCO's perturbation. We conclude that freeze-concentration induced by heating a vitrified solution can cause less perturbations of a protein than does quenching into a freeze-concentrated state. Therefore it can be advantageous for the practice of freeze-etching to vitrify first a solution by "hyperquenching" and thereafter freeze-etch at e.g. approximately 170 K.