Bovine type M pyruvate kinase, which normally has hyperbolic kinetics with its substrates, was inactivated by treatment with trinitrobenzenesulfonic acid. The inactivation probably occurs through trinitrophenylation of the epsilon-amino group of a lysine residue in or near the ADP binding site. Although 90 to 95% of the enzymatic activity is lost by this treatment, the molecular weight and sedimentation coefficient of the trinitrophenylated enzyme are quite similar to values obtained with the native enzyme. The inactivated, trinitrophenylated type M pyruvate kinase was hybridized in vitro with the native bovine type L enzyme, which has sigmoidal kinetics with phosphoenolpyruvate but can be activated by fructose 1,6-diphosphate to give hyperbolic kinetics. Four enzymatically active species were produced, designated L4, L3M, L2M2, and LM3, according to their subunit composition. L4 and L3M have sigmoidal kinetics with phosphoenolpyruvate and are activated by fructose diphosphate. Little or no sigmoidicity was seen for L2M2, although this species is activated to a moderate degree by fructose diphosphate. LM3 appears to have hyperbolic kinetics and is activated only slightly by fructose diphosphate. The kinetic results obtained with hybrids containing trinitrophenylated type M subunits are quite similar to the results previously reported by Dyson and Cardenas ((1973) J. Biol. Chem. 248, 8482-8488) using native type M and type L subunits, indicating that the properties of a type L subunit are profoundly affected by the nature of the other subunits present in the tetramer. In fact, type L and type M subunits in a given hybrid seem to have similar kinetic responses toward phosphoenolpyruvate and fructose diphosphate.