Interaction of immobilized phosphofructokinase with soluble muscle proteins

Selected glycolytic enzymes (including phosphoglucose isomerase, aldolase, glyceraldehyde phosphate dehydrogenase, enolase, pyruvate kinase and lactate dehydrogenase), as well as glycogen phosphorylase, creatine kinase, and adenylate kinase, bound to phosphofructokinase immobilized on an agarose gel. The affinity of phosphofructokinase to these various proteins differed, with phosphorylase exhibiting the strongest binding. Binding was reversed either by: (1) elution with high-ionic-strength buffer (0.4 M KCl); (2) the addition of a 5-10 mM concentration of ATP; or (3) high concentrations of fructose 6-phosphate (5 mM).     

Interaction of phosphofructokinase with antibodies. Kinetic properties of phosphofructokinase in complexes with antibodies.

The allosteric properties of phosphofructokinase (EC 2.7.1.11) from rabbit muscle are influenced by enzyme concentration, most probably due to changes in the association state of the enzyme. In this study, the behaviour of dispersed pre-cipitates of phosphofructolinase as produced by treatment with antibodies has been investigated. The enzyme is not capable of rapid dissociation in the precipitated state as is confirmed by the lack of inactivation upon dilution and by the absence of shifts in substrate saturation curves as measured in the presence of different concentrations of the enzyme. The Hill coefficient of phosphofructokinase is decreased from 1.96 to 1.04 by antibody treatment. The V at neutral pH is increased 3-fold while the K0.5 for fructose 6-phosphate is reduced significantly. On the other hand, antibody-treated phosphofructokinase retains its sensitivity to allosteric activation by glucose 1,6-bisphosphate in the rpesence of high ATP concentrations.     

Interaction of calmodulin with muscle phosphofructokinase. Interplay with metabolic effectors of the enzyme under physiological conditions

The hysteretic calmodulin-induced inactivation of muscle phosphofructokinase and the calmodulin-mediated reactivation are essentially dependent on environmental conditions. The interplay of calmodulin during these reactions and at allosteric conditions with Mg . ATP, fructose 6-phosphate, adenosine 5'-[beta, gamma-imido]triphosphate and with the allosteric effectors AMP, ADP, fructose 1,6-bisphosphate, fructose 2,6-bisphosphate and glucose 1,6-bisphosphate was studied by two techniques. (a) A two-step technique with a preincubation of enzyme, calmodulin and effectors in close to physiological concentrations before dilution into an optimal activity assay. It reveals aggregation and slowly reversible conformation changes. (b) A direct assay of dilute enzyme at allosteric conditions. Dominating in the interplay of calmodulin with metabolic effectors is the competitive-like action of calmodulin on Mg . ATP binding to the regulatory sites of the enzyme. At high enzyme concentrations in the absence of hexose phosphates, i.e. at noncatalytic conditions calmodulin counteracts the stabilization of the highly active tetrameric form caused by Mg . ATP. In the allosteric assay it counteracts the ATP-induced allosteric inhibition. In both cases calmodulin acts synergistic with AMP and ADP. To a minor degree calmodulin also counteracts the stabilization of the tetrameric form caused by fructose 6-phosphate and hexose bisphosphate, now however antagonistically to AMP and ADP. By the demonstrated interactions the enzyme can be slowly and hysteretically shifted between an active tetrameric and an inactive dimeric state under control metabolic conditions and of Ca2+ and calmodulin. Resting conditions will inactivate and high contractile activity reactivate available enzyme.     

Interaction of calmodulin with the phosphofructokinase target sequence

Ca4.calmodulin (Ca4.CaM) inhibits the glycolytic enzyme phosphofructokinase, by preventing formation of its active tetramer. Fluorescence titrations show that the affinity of complex formation of Ca4.CaM with the key 21-residue target peptide increases 1000-fold from pH 9.0 to 4.8, suggesting the involvement of histidine and carboxylic acid residues. 1H NMR pH titration indicates a marked increase in pKa of the peptide histidine on complex formation and HSQC spectra show related pH-dependent changes in the conformation of the complex. This unusually strong sensitivity of a CaM-target complex to pH suggests a potential functional role for Ca4.CaM in regulation of the glycolytic pathway.     

Mouse muscle phosphofructokinase is partially phosphorylated.

Phosphofructokinase isolated from mouse skeletal muscle 18 hours after intraperitoneal injection of [${}^{32}\text{P]-}\text{PO}{}_4{}^{\mathrm{3?}}$ contained 0.12 to 0.15 moles of covalently bound phosphate per protomer on the basis of the specific activity of radiolabel in the γ-position of ATP. Under identical conditions, muscle pyruvate kinase and aldolase had no covalently bound phosphate.     

Interaction of calmodulin with muscle phosphofructokinase. Changes of the aggregation state, conformation and catalytic activity of the enzyme

Phosphofructokinase from muscle has been shown to be a calmodulin-binding protein [Mayr, G.W. and Heilmeyer, L.M.G., Jr (1983) FEBS Lett. 159, 51-57]. Details of the influence of calmodulin on the aggregation state, the conformation and the catalytic properties of phosphofructokinase have been studied by enzymatic and light-scattering analyses. Calmodulin acts as a Ca2+-dependent hysteretic inhibitor of the highly active enzyme. At least one mole of calmodulin binds to each protomer of the enzyme, induces a shift from the highly active tetrameric towards an inactive dimeric state and slowly changes the conformation of the dimers. Dissociation of calmodulin from conformationally changed dimers by removal of Ca2+ stops the inactivation. Without a significant regain of catalytic activity large polymers are rapidly formed. For a reactivation of the inactivated enzyme, calmodulin has to remain associated and the incubation conditions must be changed in a way to allow for a back isomerization and reassociation of dimers. The isomerization reaction is promoted by Mg . ATP, the reassociation reaction most effectively by fructose bisphosphate. A model for the calmodulin-phosphofructokinase interaction is proposed.     

Protein-induced inactivation and phosphorylation of rabbit muscle phosphofructokinase.

Several previously untested proteins promote the reversible inactivation of rabbit skeletal muscle phosphofructokinase. Grouped in decreasing order of effectiveness, they include the following: skeletal muscle troponin C greater than troponin, the two smooth muscle myosin light chains, alpha-actinin, and S-100 much greater than parvalbumin and soybean trypsin inhibitor. The efficiency of troponin C in this process may even exceed that previously reported for calmodulin. Sequences near calcium binding site III are apparently involved in the troponin C-phosphofructokinase interaction. Troponin C and calmodulin exert calcium-dependent effects on the physical and chemical properties of muscle phosphofructokinase. When calcium is present, comigration with either protein allows the enzyme to enter the stacking gel during urea-polyacrylamide gel electrophoresis. Both enhance the phosphorylation of phosphofructokinase catalyzed by the cAMP-dependent protein kinase, with phosphate incorporations approaching 2 mol of P/mol of protomer. Reaction occurs at Ser774 and at Ser376--a novel site whose phosphorylation is highly sensitive to troponin C and less so to calmodulin. Maximum phosphorylation has slight effect on the catalytic activity of the enzyme under standard assay conditions. The troponin C induced or calmodulin-induced phosphorylation of phosphofructokinase requires calcium and is strongly inhibited by either fructose 2,6-bisphosphate or fructose 1,6-bisphosphate. Inactivation occurs in the presence or absence of calcium, with generally higher concentrations of effectors required for protection in the latter case. Liver and yeast phosphofructokinases shows little activity loss in the presence of either calmodulin or troponin C. We have developed and tested a general mathematical model for the protein-induced inactivation of phosphofructokinase which may find application to other systems.     

Structure of cross-linked rabbit muscle phosphofructokinase in solution.

Cross-linked rabbit muscle phosphofructokinase in the active tetrameric and octameric state was studied in solution by hydrodynamic methods and small angle x-ray scattering techniques. The translational diffusion coefficients were determined by means of inelastic light scattering and were found to be 3.60 (+/- 0.02) x 10(-7) cm2 . s-1 for the tetramer and 2.54 (+/- 0.15) x 10(-7) cm2 . s-1 for the octamer. From small angle x-ray scattering measurements the radius of gyration, the specific inner surface area, and the volume were determined for both enzyme forms, revealing that the octameric cross-linked form is approximately spherical, with a diameter of 120.0 A, whereas the tetrameric form is asymmetric having an axial ratio of 2. By comparison of the scattering curves with triaxial geometric bodies which are equivalent in scattering, the tetrameric enzyme is described as a rectangular prism, with overall dimensions of A = 131.0 A, B = 131.0 A, and C = 65.0 A, and the octameric form as that of a cube with A = B = C = 120.0 A. The shape of the protomer, having a radius of gyration of 24.8 A, in the tetramer and octamer is similar to that for the native tetramer at pH 10 in the presence of 5 mM fructose 6-phosphate or 15 mM fructose 1,6-bis-phosphate. From the different shapes of the scattering curves of the native phosphofructokinase at pH 7.5 in the presence of 15 mM ATP and of the cross-linked tetramer or octamer, it can be inferred that the shapes of the protomers are different: in the presence of ATP the protomers are elongated, having an axial ratio of 1.8 to 2.0; the cross-linked state reveals a spherical protomer of radius 33.0 A, similar to that of the native enzyme at pH 7.5 in the presence of fructose 6-phosphate or fructose 1,6-bisphosphate.     

A calorimetric study of the interaction of ATP with rabbit muscle phosphofructokinase.

The heat of interaction of ATP with phosphofructokinase from rabbit muscle was determined at 25 degrees C in 0.1 M potassium phosphate, pH 7.0 and 8.0. The limiting value of the enthalpy change at high ATP concentrations was found to be -11.5 kcal mol of enzyme polypeptide chains. Since phosphate and imidazole have very different heats of ionization (+0.8 and +7.5 kcal/mol, respectively), this suggests that the binding of at least two protons to the enzyme occurs concomitantly with the binding of ATP at the regulatory site.     

Cloning of human muscle phosphofructokinase cDNA

Three overlapping cDNA clones for human muscle phosphofructokinase (HMPFK) covering the complete coding sequence were isolated. The sequence included a poly(A) tail, a 399 bp 3'-untranslated region, a 2337 bp coding region for 779 amino acid residues and a part of the 5'-untranslated region. Homologies between HMPFK and rabbit muscle phosphofructokinase (RMPFK) were 96% of the amino acids and 89% of the nucleotides in the coding region. Like RMPFK, HMPFK also possessed the internal homology between C- and N-halves in its primary structure. Cloning of HMPFK cDNA will help to identify the molecular defect in patients with glycogenosis type VII (HMPFK deficiency).     

Allosteric regulatory properties of muscle phosphofructokinase

We have reviewed the allosteric regulatory properties of skeletal muscle phosphofructokinase and recent results on the phosphorylation of this enzyme. The number and affinities of various ligand binding sites are described, and a simple three state model is presented to explain the kinetic and ligand-binding properties of the enzyme. Data describing a lack of fit to a concerted transition model are presented. The widespread occurrence of partial phosphorylation of phosphofructokinase at a specific site near the carboxyl terminus is documented, as well as the lack of significant kinetic consequences of such phosphorylation.     

Computer modeling of muscle phosphofructokinase kinetics

The kinetics of the phosphofructokinase reaction were studied by computer modeling. A general random order, two-state allosteric model, of which the Monod--Wyman--Changeux model is a limiting case, was found to most accurately reproduce the experimental observations of Pettigrew & Frieden (1979 a,b). A simplified model with Hill coefficients was found to fit almost as well. In these models substrates bind preferentially to and stabilize the enzyme in the R state, and ATPH3-, the inhibitory species, binds preferentially to and stabilizes the enzyme in the T state. Enzymatic activity is regulated by conversion from the R to the T state, which is effected by protonation, especially of the uncomplexed enzyme, but the experimental data are inadequate for accurate estimation of the pKa of the enzyme. Random order binding of substrates is an important cause of sigmoidal kinetics. Additional experiments that would aid in the discrimination among rival models are described.