Dimeric phosphoglucose isomerase from rabbit muscle was immobilized by reaction with cyanogen-bromide-activated Sepharose 4B. The catalytic parameters and stability properties of the free and matrix-bound isomerases were essentially identical. Total monomerisation of the matrix-bound enzyme was achieved with 8 M urea as determined by a study in which one subunit was labelled with i~do['~C]acetate and the other with the 3H-labelled reagent. Although matrixbound monomers were devoid of isomerase activity, they were still capable of binding the substrate. Matrix-bound monomers exhibited the ability to redimerize with soluble isomerase subunits from either rabbit or human yielding catalytically active dimers. Yeast isomerase monomers, in contrast, did not yield active hybrids with the rabbit monomers. Furthermore, soluble subunits, which had been inactivated with pyridoxal 5'-phosphate were also capable of hybridizing with and inducing catalytic activity in the matrix-bound monomers. These studies indicate the prerequisite of dimer formation for catalytic activity but the independent action of the catalytic centers of the dimer.Dissociation of phosphoglucose isomerase into subunits has been achieved in 3 mM sodium dodecylsulfate, 6 M guanidine hydrochloride and 8 M urea [l -31. The demonstration of the dimeric structure of the enzyme, identical polypeptide chains [4] and of two substrate binding sites [5] does not exclude the functional non-equivalence of the two active sites.

Binding studies utilizing the techniques of gel filtration, rate dialysis and equilibrium dialysis yielded a value of 2.0 binding sites per molecule for either substrate or inhibitors like 6‐phospho‐gluconate and pyridoxal 5′‐phosphate to rabbit muscle phosphoglucose isomerase, thereby establishing that the enzyme has one catalytic site per subunit. For the substrates and 6‐phospho‐gluconate studies were performed at various pH values. For 6‐phosphogluconate, the dissociation constant was found to change dramatically with increasing pH, following a pattern analogous to that deduced from catalytic rate measurements. For the catalytic equilibrium mixture of glucose 6‐phosphate and fructose 6‐phosphate, the Kd showed excellent agreement with the Km values reported earlier. The binding studies are discussed in relation to the current concept of the catalytic mechanism.