The 10.8-A structure of Saccharomyces cerevisiae phosphofructokinase determined by cryoelectron microscopy: localization of the putative fructose 6-phosphate binding sites

Phosphofructokinase plays a key role in the regulation of the glycolytic pathway and is responsible for the phosphorylation of fructose 6-phosphate to fructose 1,6-bisphosphate. Although the structure of the bacterial enzyme is well understood, the knowledge is still quite limited for higher organisms given the larger size and complexity of the eukaryotic enzymes. We have studied phosphofructokinase from Saccharomyces cerevisiae in the presence of fructose 6-phosphate by cryoelectron microscopy and image analysis of single particles and obtained the structure at 10.8A resolution. This was achieved by optimizing the illumination conditions to obtain routinely 8-A data from hydrated samples in an electron microscope equipped with an LaB(6) and by improving the image alignment techniques. The analysis of the structure has evidenced that the homology of the subunits at the sequence level has transcended to the structural level. By fitting the X-ray structure of the bacterial tetramer into each dimer of the yeast octamer the putative binding sites for fructose 6-phosphate were revealed. The data presented here in combination with molecular replacement techniques have served to provide the initial phases to solve the X-ray structure of the yeast phosphofructokinase.     

Fructose 2,6-bisphosphate as a contaminant of commercially obtained fructose 6-phosphate: Effect on PPi:fructose 6-phosphate phosphotransferase

Fructose 6-phosphate from several commercial sources was shown to be contaminated with fructose 2,6-bisphosphate. This contaminant was identified by its activation of PP_i:fructose 6-phosphate phosphotransferase, extreme acid lability and behaviour on ion-exchange chromatography. The apparent kinetic properties of PP_i:fructose 6-phosphate phosphotransferase from castor bean endosperm were considerably altered when contaminated fructose 6-phosphate was used as a substrate. Varying levels of fructose 2,6-bisphosphate in the substrate may account for differences that have been observed in the properties of the above enzyme from several plant sources.     

Effects of AMP and Cibacron blue F3G-A on the fructose 6-phosphate binding of yeast phosphofructokinase.

The positive effector 5′-AMP of yeast phosphofructokinase does not influence the binding of fructose 6-phosphate to the enzyme. Cibacron blue F3G-A considered an ATP analogue decreases the affinity of the enzyme to fructose 6-phosphate without exerting an effect on the cooperativity of fructose 6-phosphate binding. The peculiarities of the interactions of AMP and Cibacron blue with fructose 6-phosphate binding demonstrate compatibility of the allosteric kinetics with the binding behavior of the enzyme.     

Endocytosis of mannose-6-phosphate binding sites by mouse T-lymphoma cells

The endocytosis and intracellular transport of mannose-6-phosphate conjugated to bovine serum albumin (Man-6-P:BSA) by mouse T-lymphoma cells were investigated in detail using several methods of analysis, both morphological and biochemical. Man-6-P:BSA was labeled with fluorescein or 125I and used to locate both surface and intracellular Man-6-P binding sites by light or electron microscopy, respectively. Incubation of cells with either fluorescent- or 125I-labeled Man-6-P:BSA at 0 degree C revealed a uniform distribution of the Man-6-P binding sites over the cell surface. Competition experiments indicate that the Man-6-P:BSA binding sites on the cell surface are the same receptors that can recognize lysosomal hydrolases. After as little as 1 min incubation at 37 degrees C, endocytosis of Man-6-P binding sites was clearly observed to occur through regions of the plasma membrane and via vesicles that also bound anticlathrin antibody. After a 5-15-min incubation of cells at 37 degrees C, the internalized ligand was detected first in the cis region of the Golgi apparatus and then in the Golgi stacks using both autoradiography and immunocytochemistry to visualize the ligand. The appearance of Man-6-P:BSA in the Golgi region after 15-30 min was confirmed by subcellular fractionation, which demonstrated an accumulation of Man-6-P:BSA in light membrane fractions that corresponded with the Golgi fractions. After a 30-min incubation at 37 degrees C, the internalized Man-6-P binding sites were localized primarily in lysosomal structures whose membrane but not lumen co-stained for acid phosphatase. These results demonstrate a temporal participation of clathrin-containing coated vesicles during the initial endocytosis of Man-6-P binding sites and that one step in the Man-6-P:BSA transport pathway between plasma membrane and the lysosomal structure can involve a transit through the Golgi stacks.     

An equilibrium binding study of the interaction of fructose 6-phosphate and fructose 1,6-bisphosphate with rabbit muscle phosphofructokinase.

Equilibrium binding studies of the interaction of rabbit muscle phosphofructokinase with fructose 6-phosphate and fructose 1,6-bisphosphate have been carried out at 5 degrees in the presence of 1-10 mM potassium phosphate (pH 7.0 and 8.0), 5 mM citrate (pH 7.0), or 0.22 mm adenylyl imidodiphosphate (pH 7.0 and 8.0). The binding isotherms for both fructose 6-phosphate and fructose 1,6-bisphosphate exhibit negative cooperativity at pH 7.0 and 8.0 in the presence of 1-10 mM potassium phosphate at protein concentrations where the enzyme exists as a mixture of dimers and tetramers (pH 7.0) or as tetramers (pH 8.0) and at pH 7.0 in the presence of 5 mM citrate where the enzyme exists primarily as dimers. The enzyme binds 1 mol of either fructose phosphate/mol of enzyme monomer (molecular weight 80,000). When enzyme aggregation states smaller than the tetramer are present, the saturation of the enzyme with either ligand is paralleled by polymerization of the enzyme to tetramer, by an increase in enzymatic activity and by a quenching of the protein fluorescence. At protein concentrations where aggregates higher than the tetramer predominate, the fructose 1,6-bisphosphate binding isotherms are hyperbolic. These results can be quantitatively analyzed in terms of a model in which the dimer is associated with extreme negative cooperativity in binding the ligands, the tetramer is associated with less negative cooperativity, and aggregates larger than the tetramer are associated with little or no cooperativity in the binding process. Phosphate is a competitive inhibitor of the fructose phosphate sites at both pH 7.0 and 8.0, while citrate inhibits binding in a complex, noncompetitive manner. In the presence of the ATP analog adenylyl imidodiphosphate, the enzyme-fructose 6-phosphate binding isotherm is sigmoidal at pH 7.0, but hyperbolic at pH 8.0. The characteristic sigmoidal initial velocity-fructose 6-phosphate isotherms for phosphofructokinase at pH 7.0, therefore, are due to an heterotropic interaction between ATP and fructose 6-phosphate binding sites which alters the homotropic interactions between fructose 6-phosphate binding sites. Thus the homotropic interactions between fructose 6-phosphate binding sites can give rise to positive, negative, or no cooperativity depending upon the pH, the aggregation state of the protein, and the metabolic effectors present. The available data suggest the regulation of phosphofructokinase involves a complex interplay between protein polymerization and homotropic and heterotropic interactions between ligand binding sites.     

Structural similarities among enzyme pterin binding sites as demonstrated by a monoclonal anti-idiotypic antibody

BALB/c mice were immunized with a synthetic co-factor of the aromatic amino acid hydroxylases, 6,7-dimethyl-5,6,7,8-tetrahydropterin, conjugated to albumin. Hybridoma cell lines isolated from the immunized mice secreted monoclonal antibodies reacting specifically with the pterin molecule and monoclonal antibodies which were found to bind phenylalanine hydroxylase. Several lines of evidence were consistent with the anti-phenylalanine hydroxylase antibodies being anti-idiotype antibodies mimicking the pterin molecule and binding to the pterin binding site of phenylalanine hydroxylase. (a) An anti-idiotype monoclonal antibody, NS7, when reimmunized into mice produced anti-pterin antibodies consistent with NS7 being an internal image anti-idiotypic antibody. (b) NS7 antibody was prevented from binding to phenylalanine hydroxylase when a competitive inhibitor of phenylalanine hydroxylase enzyme activity, 6,7-dimethyl-7,8-dihydropterin, was bound to phenylalanine hydroxylase. (c) NS7 antibody was shown to bind to a wide range of pterin-requiring enzymes: phenylalanine, tyrosine and tryptophan hydroxylases, dihydropteridine reductase, dihydrofolate reductase, and sepiapterin reductase. Thus the NS7 antibody has successfully mimicked a common portion of the pterin cofactors utilized by these enzymes and demonstrated structure homology in their pterin binding sites despite their diverse function and little amino acid sequence homology except among the three aromatic amino acid hydroxylases.     

The tryptophan synthase alpha 2 beta 2 complex: a comparison of the reactivity of amino groups in the alpha and beta 2 subunits and in the complex by differential labeling studies

The interaction of the alpha and beta 2 subunits of tryptophan synthase of Escherichia coli to form an alpha 2 beta 2 complex has been probed by differential labeling studies. In the first step the separate alpha or beta 2 subunit or the alpha 2 beta 2 complex was labeled by reductive methylation with trace amounts of [3H]HCHO in the presence of NaCNBH3. In the second step the 3H-labeled preparation was fully labeled under denaturing conditions with [14C]HCHO and NaCNBH3. Peptides containing labeled monomethyl or dimethyl amino groups were isolated after thermolytic digestion or after cyanogen bromide treatment. The 3H/14C ratio of each peptide is a measure of the relative reactivity of the amino group or groups in each peptide. The most reactive amino group in the alpha subunit, lysine-109, is strongly shielded from modification in the alpha 2 beta 2 complex. The most reactive amino group in the beta 2 subunit, the amino-terminal threonine, is not shielded from modification in the alpha 2 beta 2 complex.     

Identification of a region of UDP-galactose:N-acetylglucosamine beta 4-galactosyltransferase involved in UDP-galactose binding by differential labeling

The location of regions in the primary structure of UDP-galactose:N-acetylglucosamine beta 4-galactosyl-transferase (GT) that are involved in binding UDP-galactose has been investigated by differential chemical modification with two different reagents in the presence and absence of UDP-galactose. Treatment with periodate-cleaved UDP and NaCNBH3 resulted in a loss of 80% of GT activity, which was largely prevented by UDP-galactose. Stoichiometry of labeling and peptide maps of the modified enzyme samples indicated partial labeling at many sites. A major site of reaction in the absence of UDP-galactose that was essentially unmodified in its presence was found to correspond to Lys341 in the cDNA sequence of GT. As a second approach, the reactivities of the amino groups of GT were compared in the presence and absence of saturating levels of UDP-galactose by trace acetylation with [3H]acetic anhydride. UDP-galactose binding was found to perturb the reactivities of a number of lysines in the C-terminal region of GT, the most pronounced effect being a reduction in the reactivity of Lys351. The two procedures thus identified a region between residues 341 and 351 as being associated with UDP-galactose binding. This region overlaps a small section in the sequence of GT that was previously noted to be similar to part of bovine alpha-1,3-galactosyltransferase (Joziasse, D. H., Shaper, J. H., Van den Eijnden, D. H., Van Tunen, A. J., and Shaper, N. L. (1989) J. Biol. Chem. 264, 14290-14297). Sequence comparisons indicate that extended regions at the C terminus of each enzyme encompassing this area may represent homologous UDP-galactose-binding domains.     

Quantum chemical study of silanediols as metal binding groups for metalloprotease inhibitors

DFT (B3LYP and M06L) as well as ab initio (MP2) methods with Dunning cc-pVnZ (n?=?2,3) basis sets are employed for the study of the binding ability of the new class of protease inhibitors, i.e., silanediols, in comparison to the well-known and well-studied class of inhibitors with hydroxamic functionality (HAM). Active sites of metalloproteases are modeled by [R₃M-OH₂]²⁺ complexes, where R stands for ammonia or imidazole molecules and M is a divalent cation, namely zinc, iron or nickel (in their different spin states). The inhibiting activity is estimated by calculating Gibbs free energies of the water displacement by metal binding groups (MBGs) according to: [R₃M-OH₂]²⁺ + MBG → [R₃M-MBG]²⁺ + H₂O. The binding energy of silanediol is only a few kcal mol^?1 inferior to that of HAM for zinc and iron complexes and is even slightly higher for the triplet state of the (NH₃)₃Ni²⁺ complex. For both MBGs studied in the ammonia model the binding ability is nearly the same, i.e., Fe²⁺(t) > Ni²⁺(t) > Fe²⁺(q) > Ni²⁺(s) > Zn²⁺. However, for the imidazole model the order is slightly different, i.e., Ni²⁺(t) > Fe²⁺(t) > Fe²⁺(q) > Ni²⁺(s) ≥ Zn²⁺. Equilibrium structures of the R₃Zn ²⁺ complexes with both HAM and silanediol are characterized by the monodentate binding, but the bidentate character of binding increases on going to iron and nickel complexes. Two types of intermediates of the water displacement reactions for [(NH₃)₃M-OH₂]²⁺ complexes were found which differ by the direction of the attack of the MBG. Hexacoordinated complexes exhibit bidentate bonding of MBGs and are lower in energy for M=Ni and Fe. For Zn penta- and hexacoordinated complexes have nearly the same energy. Intermediate complexes with imidazole ligands have only octahedral structures with bidentate bonding of both HAM and dimethylsilanediol molecules.     

NMR studies on mechanism of isomerisation of fructose 6-phosphate to glucose 6-phosphate catalysed by phosphoglucose isomerase from Thermococcus kodakarensis

The fate of hydrogen atoms at C-2 of glucose 6-phosphate (G6P) and C-1 of fructose 6-phosphate (F6P) was studied in the reaction catalysed by phosphoglucose isomerase from Thermococcus kodakarensis (TkPGI) through 1D and 2D NMR methods. When the reaction was performed in ²H₂O the hydrogen atoms in the aforementioned positions were exchanged with deuterons indicating that the isomerization occurred by a cis-enediol intermediate involving C-1 pro-R hydrogen of F6P. These features are similar to those described for phosphoglucose isomerases from rabbit muscle and Pyrococcus furiosus.     

Rabbit liver fructose 1,6-bisphosphatase: labeling of the active and allosteric sites with pyridoxal 5-phosphate and sequence of a nonapeptide from the active site

The active and allosteric sites of fructose 1,6-bisphosphatase (Fru-P₂ase, EC 3.1.3.11) were labeled by reaction with pyridoxal phosphate and sodium borohydride in the presence of the allosteric inhibitor AMP or the substrate, Fru-P₂, respectively. Modification of the active site results in loss of activity. Modification of the allosteric site decreases the sensitivity of the enzyme to inhibition by AMP and alters its ability to bind to blue dextran-Sepharose. The allosteric and active sites have been located on different cyanogen bromide peptides; the sequence of a nonapeptide from the active site is (H)GlyLysLeuArgLeuLeu TyrGluCys(OH). The lysyl residue is modified by pyridoxal phosphate.     

Measurement of the rate of substrate cycling between fructose 6-phosphate and fructose 1,6-bisphosphate in skeletal muscle by using a single-isotope technique.

The effects of several agents on the rates of the fructose 6-phosphate/fructose 1,6-bisphosphate substrate cycle were measured in incubated epitrochlearis muscles of the rat by monitoring the transfer of radiolabel from [6-14C]glucose to the 1-position of glucose residues in glycogen. The cycling rates observed were almost identical with those previously obtained by using the well-established dual-isotope technique. In particular, it was found that the beta-adrenoceptor agonist isoprenaline increased the cycling rate about 12-fold.     

Dissecting differential binding of fructose and phosphate as leaving group/nucleophile of glucosyl transfer catalyzed by sucrose phosphorylase

Site-directed mutagenesis was used to examine the specificity of Leuconostoc mesenteroides sucrose phosphorylase for utilization of fructose and phosphate as leaving group/nucleophile of the reaction. The largest catalytic defect in Arg(137)-->Ala (approximately 60-fold) and Tyr(340)-->Ala (approximately 2500-fold) concerned phosphate dependent half-reactions whereas that in Asp(338)-->Asn (approximately 7000-fold) derived from disruption of steps where fructose departs or attacks. The relative efficiencies for enzyme glucosylation by sucrose compared with alpha-d-glucose-1-phosphate and enzyme deglucosylation by phosphate compared with fructose were 5.5 and 6.2 for wild-type, 19 and 2.0 for Arg(137)-->Ala, 950 and 0.17 for Tyr(340)-->Ala, and 0.05 and 180 for Asp(338)-->Asn, respectively. Asp(338) and Tyr(340) have a key role in differential binding of fructose and phosphate, respectively.     

Involvement of a divalent cation in the binding of fructose 6-phosphate to Trypanosoma cruzi phosphofructokinase: kinetic and magnetic resonance studies

When Mg2+ ions were replaced by Mn2+ in the assay of Trypanosoma (Schizotrypanum) cruzi phosphofructokinase (ATP:D-fructose-6-phosphate 1-phosphotransferase, EC 2.7.1.11) the Km for D-fructose 6-phosphate (F6P) was reduced threefold while the corresponding constant for ATP was essentially unaffected. A detailed kinetic investigation showed that the apparent Km for F6P decreased monotonically with increasing free Mn2+ concentrations, from a limiting value of 5.7 mM in its absence to a limiting value of 1.1 mM in the presence of saturating concentrations of the ion; the Vmax of the enzyme was, on the other hand, not affected by the concentration of Mn2+. Conversely, it was shown that the apparent Km for Mn2+ at fixed MnATP concentrations decreased with increasing F6P concentrations, from a limiting value of 30 microM in the absence of the sugar phosphate to 9 microM at saturating concentrations of the substrate, while the apparent Vmax increased monotonically from zero to its limiting value. Both electron paramagnetic resonance and water proton longitudinal relaxation studies showed binding of one Mn2+ ion per 18,000 Da catalytic subunit of enzyme in the absence of F6P, with a dissociation constant of 57 +/- 4 microM, comparable to the apparent Km for the ion in the absence of F6P. The presence of saturating level of F6P decreases the value of the dissociation constant of Mn2+ to a limiting value of 7.9 microM in agreement with the results of the kinetic analysis. The substrate F6P decreases the enhancement of the water proton longitudinal relaxation rate in a saturable fashion, suggesting displacement of water molecules coordinated to the enzyme-bound Mn2+ ion by the sugar phosphate. Computer fitting of the several dissociation constants and relaxation enhancements for binary and ternary complexes gives a value of 7.9 mM for the dissociation constant of the enzyme-F6P complex in the absence of Mn2+ and 1.1 mM in the presence of saturating concentrations of the ion, in excellent agreement with the respective Km values of F6P extrapolated to zero and saturating Mn2+, respectively. Studies of the frequency dependence of the water proton longitudinal relaxation rate enhancements in the presence of both binary (enzyme-Mn2+) and ternary (enzyme-Mn2(+)-F6P) complexes, are most simply explained by assuming two exchangeable water molecules in the coordination sphere of the enzyme-bound Mn2+ in the binary complex, while in the ternary complex the data are consistent with the displacement of one of the water molecule from the coordination sphere with no significant alteration of the correlation time. Overall, the kinetic and binding data are consistent with the formation of an enzyme-metal-F6P bridge complex at the active site of T. cruzi phosphofructokinase, a coordination scheme which is unique among the phosphofructokinases.     

Photoaffinity labeling of the acetylcholine binding sites on the nicotinic receptor by an aryldiazonium derivative

p-(Dimethylamino)benzenediazonium fluoroborate (DDF) behaves, in the dark, as a reversible competitive antagonist of the electrical response of Electrophorus electricus electroplaque to acetylcholine and of the acetylcholine-gated single-channel currents recorded in the C2 mouse cell line. This chemically stable but highly photoreactive compound binds irreversibly to the acetylcholine receptor when irradiated by visible light. In vivo, it irreversibly blocks the postsynaptic response of E. electricus electroplaque to agonists. In vitro, it reduces the alpha-bungarotoxin-binding capacity of acetylcholine receptor rich membrane fragments prepared from Torpedo marmorata electric organ. Once reversibly bound to the T. marmorata acetylcholine receptor, this ligand can be selectively photodecomposed by an energy-transfer reaction involving a tryptophan residue(s) of the protein. By use of reagent concentrations that are below the dissociation constant at equilibrium, up to 60% of the agonist-binding sites are covalently labeled. Under these conditions the alpha subunit of the acetylcholine receptor is preferentially labeled, and this labeling is partially prevented by agonists or competitive antagonists. This protective effect is substantially increased by prior incubation with phencyclidine, a compound known to prevent the binding of DDF at the level of the high-affinity site for noncompetitive blockers [Kotzyba-Hibert, F., Langenbuch-Cachat, J., Jaganathen, J., Goeldner, M. P., & Hirth, C. G. (1985) FEBS Lett. 182, 297-301]. The incorporation of about one molecule of label in an agonist/competitive antagonist protectable manner per alpha-bungarotoxin-binding site suffices to fully block alpha-bungarotoxin binding to the membrane-bound receptor. Thus, DDF behaves as a monovalent photoaffinity label of the acetylcholine-binding site.     

Rat brain hexokinase: further studies on the specificity of the hexose and hexose 6-phosphate binding sites

Based on the lack of correlation between the ability of various hexoses to serve as substrate and the ability of the corresponding hexose 6-phosphates to inhibit brain hexokinase (ATP:D-hexose 6-phosphotransferase, EC 2.7.1.1), R. K. Crane and A. Sols (1954, J. Biol. Chem. 210, 597-606) proposed that this enzyme possesses two discrete sites capable of binding hexose moieties, one serving as the substrate binding site and a second, regulatory in function, to which inhibitory 6-phosphates bind. Subsequent work has provided further experimental support for this proposal. The pioneering work by Crane and Sols focused primarily on the specificity of these sites with respect to requirements for orientation of hydroxyl substituents at the various positions of the pyranose ring. The present study explores additional aspects of the specificity of these sites, namely, the effect of substitution of a sulfur atom in place of the oxygen in the pyranose ring on ability to serve as substrate or inhibitor, and the effect of modification in charge of the substituent at the 6-position on inhibitory effectiveness. 5-Thioglucose is a linear competitive (versus glucose) inhibitor of rat brain hexokinase, with a Ki of about 0.2 mM, and is a linear mixed inhibitor (versus ATP), with Ki values in this same range. 5-Thioglucose is not, however, readily phosphorylated by brain hexokinase. Thus, although 5-thioglucose binds with moderate affinity to the glucose binding site, it is not effectively used as a substrate of the enzyme. Inhibition of brain hexokinase by glucose 6-phosphate or its analogs has been found to require a dianionic substituent at the 6-position. The 6-fluorophosphate derivative and glucose 6-sulfate are poor inhibitors of the enzyme, and the Ki for inhibition by 1,5-anhydroglucitol 6-phosphate increases markedly at pH values below the pK of the 6-phosphate group, indicating that the monoanionic form is ineffective as an inhibitor. In contrast to the detrimental effect that substitution of the oxygen atom in the pyranose ring with a sulfur has on ability to serve as substrate, 5-thio analogs are considerably more effective as inhibitors, the Ki for inhibition by 5-thioglucose 6-phosphate being 10-fold lower than that seen with glucose 6-phosphate. This effect of the heteroatom substitution can partially offset the decreased inhibition resulting from monoanionic character at the 6-position, but the 6-fluorophosphate derivative of 5-thioglucose 6-phosphate still inhibits with a Ki about 1000-fold greater than that seen with 5-thioglucose 6-phosphate.(ABSTRACT TRUNCATED AT 400 WORDS)     

Identification of proteins functionally altered by chemical modification of the transfer RNA and polyuridylic acid binding sites of 30 S ribosomal subunits

Previous studies have shown that Rose Bengal-sensitized photo-oxidation of 30 S ribosomal subunits causes inactivation of tRNA binding and partial loss of poly(U) binding activities (Noller et al., 1971). The present studies, reconstitution of 30 S subunits from 16 S RNA, total protein from modified subunits, and purified proteins from untreated subunits, show that proteins S2 and S3 together completely restore these activities to the reconstituted subunits. The modified proteins are capable of in vitro assembly, and give rise to particles with normal sedimentation constants, showing that restoration of activity is not simply due to correction of an assembly defect.Protein S3 restores poly(U) binding and tRNA binding to the same extent, accounting for the lowered mRNA binding activity of the modified particles as well as a corresponding fraction of the tRNA binding activity. Protein S2 restores the remaining fraction of the tRNA binding activity, but has no effect on poly (U) binding. In 50 S-stimulated tRNA binding, proteins S1 and S5 are required in addition to S2 and S3 for full activity.     

Control of the fructose 6-phosphate/fructose 1,6-bisphosphate cycle in isolated hepatocytes by glucose and glucagon. Role of a low-molecular-weight stimulator of phosphofructokinase

1. Recycling of metabolites between fructose 6-phosphate and triose phosphates has been investigated in isolated hepatocytes by the randomization of carbon between C₍₁₎ and C₍₆₎ of glucose formed from [1-¹⁴C]galactose. 2. Randomization of carbon atoms was regularly observed with hepatocytes isolated from fed rats and was then little influenced by the concentration of glucose in the incubation medium. It was decreased by about 50% in the presence of glucagon. 3. Randomization of carbon atoms by hepatocytes isolated from starved rats was barely detectable at physiological concentrations of glucose in the incubation medium, but was greatly increased with increasing glucose concentrations. It was nearly completely suppressed by glucagon. These large changes can be attributed to parallel variations in the activity of phosphofructokinase. 4. The main factors that appear to control the activity of phosphofructokinase under these experimental conditions are the concentration of fructose 6-phosphate, the concentration of fructose 1,6-bisphosphate and also the affinity of the enzyme for fructose 6-phosphate. 5. The affinity of phosphofructokinase for fructose 6-phosphate was diminished by incubation of the cells in the presence of glucagon and also by filtration of an extract of hepatocytes through Sephadex G-25 and by purification of the enzyme. When assayed at 0.25 or 0.5mm-fructose 6-phosphate, the activity of phosphofructokinase present in a liver Sephadex filtrate was increased by a low-molecular-weight effector, which could be isolated from a liver extract by ultrafiltration, gel filtration or heat treatment, but was rapidly destroyed in trichloroacetic acid, even in the cold. This effector appears to be a highly acid-labile phosphoric ester. Its concentration was greatly increased in hepatocytes incubated in the presence of glucose and was decreased in the presence of glucagon.