“…In fact, no shifts were observed in the 2,3-DPG 31P NMR lines when Pi was added to model solutions [ 191, and the Pi line itself did not shift in intact erythrocytes when the state of the hemoglobin changed from oxy to deoxy. The small increase of the adenosine triphosphate (ATP) level originated by the incubation (-0.2 mM) [39], together with the small percentages of binding of oxyhemoglobin to either ATP or ATP-Mg*+ complex [41], exclude any significant contribution of this compound to the observed shifts. Even assuming that other minor metabolites, such as, adenosine diphosphate (ADP), inosine monophosphate (IMP), 2-phosphoglycerate (2-PG), nicotinamide dinucleotide (NADH), bind to oxyhemoglobin with binding constants of the same order of magnitude as the 2,3-DPG-oxy constant, only a small contribution to the shifts is expected, because of their low concentration (<l mM).…”