The binding of azide to human methemoglobin A0. Error analysis for the interpolative and noninterpolative methods.

The binding azide to human methemoglobin A0 has been studied at 6 degrees, pH 7, and I = 0.2 by three spectroscopic methods: (1) the conventional interpolative method, (2) an interpolative dialysis technique, and (3) a noninterpolative method. The interpolative methods assume that the fractional spectral change equals the fraction of heme sites bound by ligand, while the noninterpolative method measures the extent of binding directly, i.e., without the interpolative assumption. Both experiment and error analysis show that method 1 has low precision, and consequently, gives an inherently unreliable binding isotherm. Method 2 achieves high experimental and intrinsic precision. However, method 3, which also has high precision, clearly proves that the interpolative assumption of method 2 is incorrect. That is, the true fractional extent of binding becomes equal to the fractional spectral change only after about 97% of heme sites have been bound with ligands.