Structural studies on pyruvate kinase

Pyruvate kinase from brewers' yeast (Saccharomyces cerevisiae or Saccharomyces carlsbergensis) is an allosteric enzyme that is activated by fructose diphosphate. The reaction velocity is sigmoidally dependent on the concentration of the substrate phosphoenolpyruvate in the absence of the activator. It was considered to be a particularly suitable choice for study by physical methods in an attempt to understand the mechanism of allosteric effects because there were already available, for comparative purposes, the results of several lines of investigation of rabbit muscle pyruvate kinase, which is not activated by fructose diphosphate. The work described in this thesis follows on from that of Ashton (1971), who devised a method for the preparation of pyruvate kinase from S. cerevisiae, and who found that the molecular weight of the native enzyme was 161,000, apparently contributed by eight monodisperse subunits of 20,000 each. This value for the native molecular weight is in reasonable agreement with that of 168,000 reported by Kuczenski and Suelter (1970) for pyruvate kinase prepared from S. cerevisiae by the method of Hunsley and Suelter (1969). The enzyme prepared by both these methods was cold-labile. A higher molecular weight, 190,000, was reported by Bischofberger et al. (1971) for the enzyme prepared from S. carlsbergensis by the method of Haeckel et al. (1968), and this preparation was not cold-labile. However, both Kuczenski and Suelter (1970) and Bischofberger et al. (1971) reported that the enzyme had four subunits. In this present study, pyruvate kinase was first prepared by the method of Ashton (1971). The binding of Mn2+ to the enzyme was measured by the magnetic resonance technique of proton relaxation rate enhancement. Both strong and weak binding sites for the metal were detected, although exact analysis of the binding curve was not possible. These measurements were also sensitive to the binding of the substrate phosphoenolpyruvate and of the effector fructose diphosphate, and an increase in the affinity of the enzyme for the substrate was demonstrated in the presence of the latter. In the absence of metal ions, the effector fructose diphosphate was found to catalyse the inactivation of the enzyme. The molecular weight of the subunits was re-investigated, but Ashton's results (1971) were not confirmed; sedimentation equilibrium measurements were consistent with four subunits per mole, but electrophoresis in polyacrylamide gels containing sodium dodecylsulphate revealed hetero geneity of the subunits, presumably caused by proteolytic attack during the preparation of the enzyme. Various alterations were made to Ashton's method of preparation in an attempt to prevent proteolysis, but eventually a new method was devised in which: (i) the initial autolytic degradation of the yeast cells, which had been used in all three published methods of preparation, was replaced by the use of freeze-thawing to rupture the cells; (ii) the temperature was kept below 5°C at all stages of the preparation, and (iii) the protease inhibitor phenylmethylsulphonylfluoride was added in the early stages of the preparation. The specific activity of the enzyme thus obtained, up to 400 U/mg, was significantly higher than had previously been observed in any laboratory. The molecular weight of yeast pyruvate kinase prepared by this method was determined as 213,000 by sedimentation equilibrium, and as 217,000 from a combination of the sedimentation and diffusion coefficients. The diffusion coefficient was determined by intensity fluctuation spectroscopy of scattered laser light. Sedimentation equilibrium measurements in 6M guanidinium chloride - 1% 2-mercaptoethanol and electrophoresis in sodium dodecyl-sulphate gels showed that there were 4 subunits, of molecular weight about 54,000, per mole of native enzyme. Like pyruvate kinase prepared from S. cerevisiae by the methods of Hunsley and Suelter (1969) and Ashton (1971), the enzyme was cold-labile, and the inactivation was catalysed by the allosteric effector fructose diphosphate. A free N-terminus was not detected by dansylation, although N-terminal alanine had been observed with the enzyme prepared by the method of Ashton. When the proton relaxation enhancement experiments were repeated, it was found that there were no longer any strong binding sites for Mn2+, and the metal binding was too weak to obtain meaningful results by this technique. The reactivity of the thiol groups of the protein was investigated: two groups per mole were particularly reactive with both 5,5'-dithiobis-(2-nitrobenzoate) and iodoacetamide; 14 groups per mole reacted with the former reagent in the absence of denaturant, and a total of 24 reacted when the protein was denatured. Two 'spin-labels' were covalently attached to the enzyme, but the spectrum of neither was sensitive to the binding of ligands. The results are compared with those reported by other investigators for pyruvate kinase from yeast, and also with those for pyruvate kinase from other sources. Some of the work presented in this thesis has also been published (Fell et al. 1972 and 1974).