Abstract

Strains of Escherichia coli bearing different forms of phosphofructokinase were used to assess the occurrence of futile cycling in cell resuspensions supplied with glycerol as gluconeogenic carbon source. A model was used to simulate results of different kinds of experiments for different levels of futile cycle. The main predictions of the model were experimentally confirmed in a strain with a mutant phosphofructokinase-2 (phosphofructokinase-2*) which is not inhibited by MgATP. The intracellular fructose 1,6-bisphosphate concentration reaches significantly higher levels in the mutant-bearing strain than in strains with either phosphofructokinase-1 or -2. Also, this strain showed a higher rate and level of in vivo radioactive labelling of fructose 1,6-bisphosphate, from a trace of [U-14C]glucose supplied during gluconeogenesis, indicating higher kinase activity in these conditions. Cell resuspensions of the mutant-bearing strain produced higher levels of radioactively labelled CO2 when supplied with [U-14C]glycerol as the only carbon source. Simultaneously, fewer glycerol carbons were incorporated into HClO4-insoluble macromolecules. Finally, radioactive CO2 output was measured in resuspensions supplied with glycerol as the major carbon source with traces of either [1-14C]glucose or [6-14C]glucose. It was found that, whereas in the strains with either of the wild-type phosphofructokinase isoenzymes, radioactive CO2 output from [1-14C]glucose was higher than with [6-14C]glucose, the reverse is found for the strain with phosphofructokinase-2*. This result also agrees with the corresponding prediction of the model. Using the radioactivity flux rates predicted by the model, an explanation linking the futile cycle to the differential labelling of CO2 is advanced. Finally, on the basis of these results it is proposed that strains bearing phosphofructokinase-2* sustain higher rates of futile cycling during gluconeogenesis than strains bearing either of the wild-type isoforms of phosphofructokinase.