Aerobic and anoxic variants of radular retractor muscle pyruvate kinase (PK-aerobic and Pk-anoxic) from the gastropod mollusc, Busycotypus canaliculatum, were purified to homogeneity and respective specific activities of 368 and 186 mumol of product min-1 mg protein-1. Both PK variants were apparent homotetramers with native molecular masses of about 235 kDa, but differed in several other physical characteristics including pI (5.81 +/- 0.06 for PK-aerobic, 5.42 +/- 0.03 for PK-anoxic) and chromatographic behavior on several columns used during their respective purifications. The two enzymes differed greatly in several kinetic properties. Affinity for phosphoenolpyruvate was more than tenfold greater for PK-aerobic (K0.5 = 0.067 +/- 0.002 mM; h = 0.99 +/- 0.10), whereas the cooperative effect for phosphoenolpyruvate binding was greatly enhanced for PK-anoxic (K0.5 = 0.85 +/- 0.02 mM, h = 2.57 +/- 0.01). Although the affinities for the second substrate, ADP, were identical for both enzyme forms (apparent Km = 0.25 mM) pK-anoxic showed greater substrate inhibition by high concentrations of ADP. Likewise, affinities for K+ and Mg2+ were similar but PK-anoxic showed a greater degree of cooperativity with Mg2+ (h = 2.50 +/- 0.02) than did PK-aerobic (h = 1.70 +/- 0.06). Saturating concentrations of fructose 1,6-bisphosphate (50 microM) activated PK-anoxic resulting in an enzyme with properties similar to fructose-1,6-bisphosphate-activated PK-aerobic, with K0.5 values for phosphoenolpyruvate of about 0.04 mM and Hill coefficients of 1.1. PK-anoxic showed much stronger regulation by the allosteric inhibitors MgATP, phenylalanine, proline and alanine. Fructose 1,6-bisphosphate partially relieved the inhibitions by ADP, MgATP, alanine, proline and arginine phosphate of both enzyme forms. However, at 0.1 mM phosphoenolpyruvate PK-aerobic was much more sensitive to activation by fructose 1,6-bisphosphate, Ka values being 0.05 +/- 0.01 microM for PK-aerobic and 1.3 +/- 0.1 microM for PK-anoxic. In the presence of 1.0 mM alanine and 1.5 mM MgATP much higher concentrations of fructose 1,6-bisphosphate were required for activation of PK-anoxic (Ka = 5.2 +/- 0.4 microM) than for PK-aerobic (Ka = 0.02 +/- 0.01 microM). Variations in pH over the range likely occurring in vivo during anaerobiosis caused no significant additional kinetic differences between the two enzyme forms. The dissimilarity in kinetic properties of PK-aerobic and PK-anoxic indicate that red muscle PK activity is probably strongly depressed in vivo during anoxia stress.