Nitrogen fixation in legume bacteroids is energized by the metabolism of dicarboxylic acids, which requires their oxidation to both oxaloacetate and pyruvate. In alfalfa bacteroids, production of pyruvate requires NAD + malic enzyme (Dme) but not NADP + malic enzyme (Tme). However, we show that
Rhizobium leguminosarumhas two pathways for pyruvate formation from dicarboxylates catalyzed by Dme and by the combined activities of phosphoenolpyruvate (PEP) carboxykinase (PckA) and pyruvate kinase (PykA). Both pathways enable N 2 fixation, but the PckA/PykA pathway supports N 2 fixation at only 60% of that for Dme. Double mutants of dmeand pckA/pykAdid not fix N 2 . Furthermore, dme pykAdouble mutants did not grow on dicarboxylates, showing that they are the only pathways for the production of pyruvate from dicarboxylates normally expressed. PckA is not expressed in alfalfa bacteroids, resulting in an obligate requirement for Dme for pyruvate formation and N 2 fixation. When PckA was expressed from a constitutive nptIIpromoter in alfalfa dmebacteroids, acetylene was reduced at 30% of the wild-type rate, although this level was insufficient to prevent nitrogen starvation. Dme has N-terminal, malic enzyme (Me), and C-terminal phosphotransacetylase (Pta) domains. Deleting the Pta domain increased the peak acetylene reduction rate in 4-week-old pea plants to 140 to 150% of the wild-type rate, and this was accompanied by increased nodule mass. Plants infected with Pta deletion mutants did not have increased dry weight, demonstrating that there is not a sustained change in nitrogen fixation throughout growth. This indicates a complex relationship between pyruvate synthesis in bacteroids, nitrogen fixation, and plant growth.