Investigations have been carried out on both of the isozymic forms of phosphofructokinase (PFK) in some enterobacteria, together with a detailed characterization of the non allosteric form from Escherichia coli.
Using both nucleotide linked-Agarose and triazine dye linked-Sepharose, homogeneous preparations of high specific activity of the major allosteric isozymic form were obtained from Escherichia coli, Enterobacter aeroqenes, Hafnia alvei and Serratia marcescens.
Molecular weight estimations have shown that all are oligomeric tetramers ranging in size from 120 000 to 160 000 for the En.aeroqenes and Se.mercescens PFK respectively. The Ha.alvei enzyme has a molecular weight of 140 000 equivalent to that from E.coli. Enzyme subunits possessed molecular weights between 32 000 to 39 000. Additional physical properties show the three non coliforms to produce enzymes less heat stable at 50°C than that from E.coli.
Differences also occur in the primary amino acid composition. Isoelectric points from 5.2 and 5.0 were recorded for the E. coli and En.aeroqenes PFK respectively, and 4.5 for the Se.marcescens enzymes. All of the pure enzymes possessed sulphydryl and arginyl groups essential for catalysis. The Se.marcescens PFK recorded the highest inhibition from the arginine-specific reagent, 2,3 butanedione. The acid isoelectric point, indicates that a greater amount of negativity charged amino acid residues are present in this protein. Each PFK cross-reacted with the antisera against the E.coli antigen, although it was evident that En.aeroqenes, Ha.alvei and Se.marcescnes proteins possess antigenic determinants not in common with those from the E.coli protein.
Kinetic properties show a common allosteric response towards the substrate fructose 6-phosphate, activation by both ADP and NH4Cl, and strong inhibition by phosphoenolpyruvate. The Ha.alvei PFK differed in some respect, showing a lack of change in cooperativity in the presence of NH4Cl. Collectively, the results confirm the present taxonomic boundaries allocated for these genera.
The use of phosphoenolpyruvate in an assay procedure to estimate the presence of the non allosteric isozyme, showed that all of the enterobacteria tested produced variable but low levels of this enzyme. When grown on glucose, cell-free extracts of Citrobacter freundii and Proteus vulqarus recorded the highest amounts of non allosteric PFK production at 0.099 and 0.076 Units/mg respectively. Growth of E.coli on a variety of carbon sources indicated the production of this enzyme was highest on glucose with synthesis being constitutive and less than 10% of the total PFK produced. This was in contrast to inducible enzyme synthesis of the allosteric protein which was derepressed under anaerobiosis.
The allosteric enzyme has a strong affinity for Blue Dextran linked-Sepharose 4B. The non allosteric enzyme showed a tendency to bind in a weak manner to the same resin when in the crude form.
A successful attempt was made to isolate the non allosteric PFK from E.coli. A combination of techniques involving salt and acid precipitation, ion-exchange and gel filtration chromatography,-and preparative isoelectric focusing produced a homogeneous preparation. The enzyme was a dimeric protein of molecular weight 67 000 with an isoelectric point of 5.1, and possessed two pH optima at 8.4 and 6.8 respectively. Fructose 6-phosphate saturation curves were hyperbolic and no catalytic response occurred in the presence of the metabolites known to effect the allosteric PFK. Whilst these results indicate a nonfunctional metabolic role for this protein, additional information contradicts this feature. The affinity for nucleotide triphosphates was equivalent, although a decreased maximum velocity became evident with UTP and CTP. Manganese can replace magnesium as a cofactor with a marked 2.5 fold increase in maximum velocity. The enzyme was activated by both MnATP2- and Mn2+ at the active site. Together with the acidic pH optimum and the kinetic response shown by the non allosteric enzyme, it is suggested that this enzyme can be an important catalyst in E.coli during fermentative glycolysis.
The enzyme was found to be extremely labile in the pure form, requiring the presence of bovine serum albumin and low temperatures to minimize activity losses. This feature may have contributed to the failure to raise antisera in rabbits when inoculated as the pure antigen.
A reported tetrameric configuration and a molecular weight of 140 000 of the non allosteric PFK isolated from a mutant of E.coli, prompted an examination of the contrasting molecular weight result from the present study. In crude extracts, the enzyme exists as a tetramer of molecular weight 140 000 which dissociated to a dimeric protein of 67 000 molecular weight under mild denaturing conditions. Incubation at 50°C for 20 minutes resulted in the formation of the dimer which in turn had an increased affinity for Cibracon Blue linked-Sepharose 6B. Binding to this chromophore was not specific and assays for proteolytic activity indicated that a non-specific protease present in crude cell-free extracts of E.coli was responsible for the increase in affinity of the crude enzyme, but not for the dissociation. The latter was shwon to be related to the presence of weak electrostatic forces between protomers.
ATP eluted less than 20% of the non allosteric enzyme bound to a Blue Dextran column. This fraction was completely inactivated following dialysis against Tris buffer at pH 8.4, and required the presence of 2-mercaptoethanol for the restoration of activity. Incubation in a mixture of fructose 6-phosphate (5.0 mM), ADP (0.5 mM) restored this activity which was enhanced when 10.0 mM ATP was added to the mixture.
A study of this mechanism showed that interconversion was directly related to protein concentration occurring only at levels greater than 0.0011 mg/ml and less than 0.088 mg/ml. The presence of ATP did not stabilize the active protein, with dissociation and activity losses becoming apparent after 2 hours at 4°C . It is suggested that this feature, while unlikely to occur in the cell, does indicate that a critical protein and metabolite concentration is necessary for stabilisation of this enzyme in both invitro and insitu environments. The metabolic role of both isozymic forms and their relation to the mechanism of the 'Pasteur effect' in E.coli are discussed.