32P-labelling anomalies in human erythrocytes. Is there more than one pool of cellular Pi?

1. Human erythrocytes were incubated in autologous plasma containing [32P]Pi, and sampled by a method which avoids washing the cells. 2. In experiments of up to 3 h duration, the specific radioactivity of cellular Pi stabilized at a value below that of extracellular Pi. This can be explained on the basis of a single cellular Pi pool exchanging with a large unlabelled pool of cellular organic phosphates. 3. However, a rapid initial phase of labelling, occurring within 30 s, was inconsistent with the situation described in point 2. A possible explanation is that about 1/4 of cellular Pi occurs in a separate, fast-labelling pool. 4. When the extracellular Pi concentration was doubled, most of the corresponding increase in the steady-state cellular Pi concentration was accounted for by the apparent fast-labelling Pi pool, which also doubled. 5. The observed initial rate of labelling of cellular organic phosphates [which probably occurs through the reaction catalysed by glyceraldehyde-3-phosphate dehydrogenase (E.C. 1.2.1.12)] was considerably lower than that predicted from the flux through the Embden-Meyerhof pathway. This implies that the enzyme is exposed to Pi whose specific radioactivity is lower than the mean specific radioactivity of cellular Pi, and fails to support earlier suggestions that this enzyme uses extracellular Pi. 6. In 3 h incubations, the rate of organic phosphate labelling was roughly constant throughout, even though the specific radioactivity of cellular Pi had risen slowly to a plateau. Viewed in conjunction with point 5, this again suggests some inhomogeneity in cellular Pi. 7. Cellular Pi and extracellular Pi only reached isotopic steady state after 2 days. At this stage some organic phosphates were probably still incompletely labelled. 8. We conclude that, whatever their physical or technical reasons, such labelling inhomogeneities and slow attainment of isotopic steady state may cause serious misinterpretation of results if ignored during 32P-labelling of intact cells.