Hemoglobin affinity for 2,3-bisphosphoglycerate in solutions and intact erythrocytes: studies using pulsed-field gradient nuclear magnetic resonance and Monte Carlo simulations.

The diffusion coefficient (D) of 2,3-bisphosphoglycerate (DPG) was measured using pulsed-field gradient (PFG)-31P nuclear magnetic resonance spectroscopy in solutions containing 2.7-5.0 mM hemoglobin (Hb) and a range of DPG concentrations. The dependence of the measured values of D on the fraction of the total DPG in the sample that is bound to Hb enabled the estimation of the dissociation constants (Kd) of complexes of DPG with carbonmonoxygenated, oxygenated, and deoxygenated Hb; the values of Kd (mM), measured at 25 degrees C, pH 6.9 and in 100 mM bis Tris/50 mM KCl, were 1.98 +/- 0.26, 1.8 +/- 0.5 and 0.39 +/- 0.26, respectively. In intact erythrocytes the apparent diffusion coefficient, Dapp, of DPG was larger in oxygenated and carbonmonoxygenated cells (6.17 +/- 0.20 x 10(-11) m2s-1) than in deoxygenated cells (4.10 +/- 0.23 x 10(-11) m2s-1). Changes in intracellular DPG concentration (5-55 mM) in erythrocytes, brought about by incubation in a medium containing inosine and pyruvate, did not result in significant changes in the value of Dapp; this result supports the hypothesis that DPG binds to other sites in the erythrocyte. Monte Carlo simulations of diffusion in biconcave discs were used to test the adequacy of the values of Kd estimated in solution to describe the binding of DPG to Hb in oxygenated and deoxygenated erythrocytes. The results of the simulations implied that the value of Kd estimated for deoxygenated Hb-DPG was greater than expected from the experiments involving intact erythrocytes. This difference is surmised to be at least partly due to the difficulty of measuring D at low-ligand concentrations. Notwithstanding this shortcoming, the PFG method appears to be suitable for probing interactions between macromolecules and ligands when the Kd is in the millimolar range. It is one of the few techniques available in which these interactions can be studied in intact cells. In addition, the Monte Carlo simulations of the diffusion experiments highlighted important differences between theory and experiment relating to the nature of molecular motion inside the cells.