Characterization of tumor-associated fucogangliosides from PC 12 pheochromocytoma cells

PC 12h pheochromocytoma cells were subcutaneously transplanted into rat. We found the transplanted tumors accumulated some fucogangliosides associated with PC 12 cells. These gangliosides were isolated and purified by DEAE-Sephadex A-25 and Iatrobeads column chromatographies. Their structures were determined by fast atom bombardment mass spectrometry, proton nuclear magnetic resonance spectrometry, permethylation study, and sequential degradation using various exoglycosidases and mild acid hydrolysis. Two tumor-associated fucogangliosides were found to possess the blood group B determinant as follows: G6: IV2Fuc alpha, IV3Gal alpha, II3NeuAc, GgOse4Cer; G11: IV2Fuc alpha, IV3Gal alpha, II3 (NeuAc)2, GgOse4Cer. A ganglioside with the similar structure as ganglioside G6 was isolated from rat hepatoma cells (Holmes, E.H., and Hakomori, S-I. (1982) J. Biol. Chem. 257, 7698-7703). However, ganglioside G11 has not previously been reported in the literature. These fucogangliosides reacted with the monoclonal antibody prepared by immunizing mice with PC 12h cells. Other fucogangliosides were also found to accumulate in the transplanted tumor tissues. They were identified as fucosyl-GM1 and fucosyl-GDlb. These fucogangliosides did not react with the monoclonal antibody against PC 12h cells.     

Diplococcal beta-galactosidase with a specificity reacting to beta 1-4 linkage but not to beta 1-3 linkage as a useful exoglycosidase for the structural elucidation of glycolipids

Diplococcal beta-galactosidase, which is known to be useful for the structural studies of glycoprotein-linked oligosaccharides, was found to show the same substrate specificity in cleaving Gal beta 1-4 linkages of glycolipids as that of the oligosaccharides. The optimum conditions of beta-galactosidase in the 80% ammonium sulfate precipitates of the culture medium of Streptococcus (Diplococcus) pneumoniae were determined with nLcOse4Cer radiolabeled by the galactose oxidase-NaB3H4 procedure. Detergent was required for the highest activity, and different combinations of several buffers and detergents showed different properties in stimulating beta-galactosidase, and in enhancing or suppressing N-acetyl-beta-hexosaminidase which was contaminated in the enzyme preparation. The optimum pH was found to be at 6.5, and specific activity and Km were 8.1 nmol/mg protein/h and 1 nmol, respectively. While more than 70% of beta-galactose was liberated from LacCer and nLcOse4Cer within 1 h under the optimum conditions to form GlcCer and nLcOse3Cer, respectively, none was liberated from LcOse4Cer, GalCer, GgOse4Cer, GbOse3Cer, IV3 alpha GalnLcOse4Cer, and Il3NeuAcGgOse4Cer, showing the substrate specificity solely to Gal beta 1-4 linkage.     

Purification and properties of rat liver globotriaosylceramide synthase, UDP-galactose:lactosylceramide alpha 1-4-galactosyltransferase

The enzyme which catalyzes the transfer of galactose from UDP-galactose to lactosylceramide (LacCer) was obtained in a 32,000-fold purified and apparently homogeneous form from rat liver by a procedure involving affinity chromatography on UDP-hexanolamine-Sepharose and LacCer-Sepharose. The enzyme is composed of two nonidentical subunits whose apparent molecular weights are 65,000 and 22,000. Methylation and hydrolysis of the product formed by incubation of the enzyme with UDP-galactose and [3H]LacCer yielded 2,3,6-tri-O-methyl-[3H]galactose, indicating that a galactose residue was introduced to position C-4 of the terminal galactose of the LacCer. The product also specifically reacted with monoclonal antibody directed to globotriaosylceramide (Gal alpha 1-4Gal beta 1-4Glc beta 1-1Cer). This indicates that the purified enzyme is exclusively alpha 1-4-galactosyltransferase. Studies on substrate specificity indicate that the purified enzyme is highly specific for the synthesis of GbOse3Cer and is clearly distinct from the enzymes responsible for the formation of iGbOse3Cer (Gal alpha 1-3Gal beta 1-4Glc-Cer) and blood group-B substance, which possess alpha 1-3 galactosidic linkages at the nonreducing termini. The enzyme is also distinct from the alpha 1-4-galactosyltransferase which catalyzes the formation of galabiaosylceramide (Gal alpha 1-4Gal beta 1-1Cer) and IV4Gal-nLacOse4 (P1 antigen). These studies represent the first report of the properties of a highly purified alpha-galactosyltransferase catalyzing the transfer of sugar residues to glycolipids.     

An amino acid region at the N-terminus of rat hepatoma alpha1-->2 fucosyltransferase modulates enzyme activity and interaction with lipids: strong preference for glycosphingolipids containing terminal Galbeta1-->3GalNAc-structures

A GDP-fucose:GM1 alpha1-->2 fucosyltransferase (FucT) is induced during early stages of chemical hepatocarcinogenesis in parenchymal cells of Fischer 344 rats fed a diet supplemented with 0.03% N-2-acetylaminofluorene (AAF). This enzyme is undetectable in normal rat liver tissues but is highly expressed in many rat hepatoma cell lines, including rat hepatoma H35 cells. Enzymatic properties and acceptor specificity of native rat hepatoma H35 cell alpha1-->2FucT, expressed recombinant full-length H35 cell alpha1-->2FucT, and a truncated form missing the first 27 amino acid residues from the N-terminus, comprising the cytoplasmic and transmembrane domains of the enzyme, were studied. The results indicate that the recombinant full-length enzyme has a specific activity over 80-fold higher than the truncated enzyme. Both the native and recombinant full-length enzymes display significant activity in the absence of detergent or phospholipid and optimal activity in the presence of Triton CF-54 detergent. The truncated enzyme is optimally activated by CHAPSO, showing little activity in its absence. These findings are in agreement with previous studies demonstrating a requirement of a lipidic environment for optimal activity with this enzyme and suggest that the N-terminal transmembrane domain is important either in the maintenance of an active conformation or in allowing efficient interaction with acceptor glycolipids. Both the full-length and truncated enzymes transfer fucose not only to GM1 and asialo-GM1 (Gg4) but also to galactosyl globoside (Gb5) as well. Weak or undetectable transfer to lacto- and neolacto-series acceptors was observed, demonstrating a strong preference for terminal Galbeta1-->3GalNAc- structures. The structures of two reaction products generated by expressed recombinant full-length alpha1-->2FucT, which are known to be important tumor-associated antigens (fucosyl-GM1 and fucosyl-Gb5), were unambiguously confirmed by 1H-NMR spectral analysis.     

Changes in protein kinase activities during 2-acetylaminofluorene-induced hepatocarcinogenesis in the rat

Livers of rats fed the carcinogen 2-acetylaminofluorene (AAF) at a concentration of 0.025% were analyzed for protein kinase activities with [gamma 32P]ATP as substrate and either endogenous or exogenous (casein or histone) protein acceptors both in the presence or absence of cyclic nucleotides. Total protein kinase activity of the nuclear fraction, with exogenous histone or casein as substrate, was elevated during the first week of carcinogen administration. Total cytoplasmic kinase activities exhibited a pattern of activity change with maxima at about 25 and 42-49 days after the onset of carcinogen administration. Cyclic AMP levels rose steadily to approximately a 4-fold elevation by day 49 in livers of animals receiving carcinogen with the increase beginning prior to the development of externally visible nodular hyperplastic lesions. The findings demonstrate consistent and reproducible patterns of change in protein kinase activities that accompany AAF-induced hepatocarcinogenesis in the rat and provides the basis for a more detailed investigation of specific kinases.     

Photometric and fluorometric continuous kinetic assay of acid phosphatases with new substrates possessing longwave absorption and emission maxima

For the measurement of the enzymatic activity of GM1-ganglioside (II3 NeuAcGgOse4Cer, galactosyl-N-acetylgalactosaminyl-(N-acetylneuraminosyl) galactosyl-glucosylceramide) beta-galactosidase in crude enzyme samples, a microassay using nonradioisotopic GM1-ganglioside was devised. To reduce the volume of the reaction mixture and eliminate the interferences due to the fluorescent contaminants in the reaction mixture, NADH, a product after the oxidation of the released galactose with NAD and beta-galactose dehydrogenase, was fluorometrically estimated by use of high-performance liquid chromatography. By this method, as little as 10 pmol of galactose can be detected. Using rat brain homogenates as an enzyme sample, the several parameters were reexamined to define the optimal conditions for the assay. This assay method was also applied to human cultured skin fibroblast homogenates, and it was found that this method can be used for the diagnosis of GM1-gangliosidosis, instead of the usual method using the radioisotope-labeled natural substrate.     

Characterization of solubilized GlcAT-1 (UDP-GlcA: nLcOse4Cer beta 1-3 glucuronyltransferase) activity from embryonic chicken brain and its inhibition by D-erythro-sphingosine

Glycolipid glucuronyltransferase activity (GlcAT-1) has been solubilized and characterized from 19-day-old embryonic chicken brain Golgi-rich membranes. The enzyme catalyzes the biosynthesis in vitro of GlcA beta 1-3nLcOse4Cer glycolipid using neolactetraosylceramide (nLcOse4Cer, Gal beta 1-4GlcNAc beta 1-3Gal beta-1-4Glc-Cer) as the substrate. The membrane-bound enzyme shows optimum activity in the presence of neutral detergents such as Triton CF-54, Triton DF-12, and Nonidet P-40. Approximately 60% of the enzyme activity can be solubilized from the Golgi membrane by Nonidet P-40. The solubilized GlcAT-1 activity is inhibited by different salts such as NaCl, NaBr, NaI, and NaOAc, but not by sodium fluoride (up to 0.4 M concentration). Desialyzed alpha 1 acid glycoprotein (SA alpha 1AGP) can be used as a substrate for glucuronyltransferase. Competition studies between glycolipid (nLcOse4Cer) and glycoprotein SA alpha 1AGP) substrates show a mixed type of inhibition. Phospholipids, in particular phosphatidylglycerol, stimulate solubilized GlcAT-1 activity, while D-erythro-sphingosine, a metabolite of glycosphingolipids, is inhibitory (50% inhibition at 0.8 mM D-erythro-sph). These results demonstrate that both phospholipid as well as sphingosine might be involved in modulating glucuronyltransferase activity.     

Latent form of glutathione reductase in the rat liver

The existence of membrane-bound forms of glutathione reductase in rat liver and transplantable hepatoma G-27 was demonstrated, using differential centrifugation techniques. The activity of the sedimentable form of the liver enzyme was detected only in the presence of detergents. Conditions for the manifestation of the latent glutathione reductase activity in whole liver homogenates and in the 105000 g pellet were determined. Solubilization of the latent form of the enzyme in the presence of sodium deoxycholate increases 2-fold the glutathione reductase activity in liver homogenates (but not in hepatoma). Simultaneous determination of the disulfidereductase, nonspecific NADPH-oxidase and gamma-glutamyltransferase (membrane-bound enzyme of glutathione metabolism) activities was performed.     

Incorporation of urease into liposomes.

1. Plasma membranes were isolated from ascites hepatoma AH-130 and rat livers with or without partial hepatectomy or bile duct ligation. Reciprocal manifestations of two marker enzymes for plasma membranes were observed in these membrane preparations; alkaline phosphatase activity was found much higher in the hepatoma membrane than in any preparations of the liver membranes, whereas 5'-nucleotidase activity was much lower in the former than in the latter. 2. Effects of lectins and anti-plasma membrane antiserum on these two marker enzymes were examined. The results showed that about 50% of apparent activity of 5'-nucleotidase found in the hepatoma membrane was exhibited by alkaline phosphatase. 3. Localizations of alkaline phosphatase and 5'-nucleotidase in polyacrylamide gels after electrophoresis were demonstrated using 5'-AMP and 5-Br, 4-Cl-indoxylphosphate as substrate. There was a difference in electrophoretic mobility between the alkaline phosphatase of the hepatoma and that of the liver.     

Cytoplasmic changes in level and distribution of glucose-6-phosphatase activities from rat liver during diethylnitrosamine-induced carcinogenesis.

Glucose-6-phosphatase (EC 3.1.3.9) activities were determined in isolated microsomes, cytoplasmic smooth and rough membranes, ribosomes and free cytosol from rat liver undergoing carcinogenesis by diethylnitrosamine (DENA) and compared with cytoplasmic fractions isolated in parallel from healthy animals from the same age.With continuous administration of a low dose of DENA (2.6 mg/kg rat per day for 20 weeks in the drinking water) livers of carcinogen treated rats became heavier than the control livers but the body weight decreased. About 70% of total glucose-6-phosphatase activity could be detected in the microsomal fraction. While there was no significant difference in this activity in both animal groups up to the 4th week, glucose-6-phosphatase of cancerous liver showed a distinct decrease of activity compared with normal liver.During cancer induction this enzyme became more soluble, confirmed by the observation that it was detached from firmer structures of cytoplasm as rough membranes and polysomes and translocated to smooth membranes and the soluble cytoplasmic fraction successively. The corresponding increase in glucose-6-phosphatase activity in the 105 000 g supernatant appears to be due to the loss of enzyme activity in a distinct cytoplasmic membrane fraction. These data strongly suggest that in parallel with alteration of cytoplasmic membrane structures during carcinogen feeding glucose-6-phosphatase is detached from heavier components of the cytoplasm while total activity decreased. Possible mechanisms of these findings are discussed.     

The serine protease from rat liver and hepatoma 8999. Location and role in mitochondrial protein degradation.

1. Hepatoma 8999 showed extremely high activity of serine protease, but similar activities of other lysosomal proteases to those of normal rat liver. 2. Serine protease from rat liver formed a single immunoprecipitation band against antiserum to purified protease from hepatoma 8999. 3. The serine proteases in rat liver and hepatoma 8999 were restricted to the inner membranes of the mitochondrial fraction. 4. Polyacrylamide gel electrophoresis with sodium dodecylsulfate showed that hepatoma 8999 mitochondria contained less of the slowest moving protein component than rat liver mitochondrial protein. This component was found to be the best substrate for mitochondrial serine protease in both liver and hepatoma 8999. 5. The role of serine protease in mitochondrial protein degradation is discussed on the basis of these results.     

Identification of mammalian aspartate-4-decarboxylase

Several animal tissues were examined for aspartate-4-decarboxylase (EC 4.1.1.12) activity. Highest activity was seen in murine livers, in rodent livers, and in rodent kidneys. The rat liver enzyme was membrane associated and could be solubilized and partially purified with the aid of detergents. The purification studies, and studies on the stoichiometry and kinetics of the reaction, showed that aspartate is directly converted to alanine. Such a metabolic reaction had not been reported before in animals. The rat liver enzyme differed significantly from the microbial aspartate-4-decarboxylases. Among other things, the rat liver beta-decarboxylase could be purified away from a cysteine sulfinate desulfinase activity. Also, unlike the bacterial enzymes, the mammalian beta-decarboxylase could not be inactivated by preincubation with aspartate or cysteine sulfinate. These later observations strongly suggest that the mammalian aspartate-4-decarboxylase does not have an inherent transaminase activity. Like many decarboxylases, rat liver aspartate-4-decarboxylase could be inhibited by reagents which react with carbonyl groups; however, the enzyme showed no dependence on pyridoxal 5'-phosphate.     

Developmental changes of gangliosides of the rat stomach. Appearance of a blood group B-active ganglioside

Rat stomach gangliosides were purified and their distribution in the different tissue compartments was established. Three major monosialogangliosides were found: GM3, GM1, and a ganglioheptaosylceramide carrying a blood group B determinant. This latter structure was characterized by exoglycosidase degradation, immunostaining with a monoclonal anti-blood group B antibody on thin layer chromatogram, permethylation analysis, electron-impact mass spectrometry of the permethylated-reduced and trimethylsilylated molecule, and 1H NMR spectroscopy of the native ganglioside. It was found to be (Formula: see text) i.e. a GM1 structure substituted with the blood group B determinant and was called B-GM1. A similar structure has been previously identified in precancerous rat liver and chemically induced rat hepatoma (Holmes, E. H., and Hakomori, S. (1982) J. Biol. Chem. 257, 7698-7703). Fucosyl-GM1 was also detected as a minor ganglioside in rat gastric mucosa. The ganglioside profile was modified during the postnatal development. The contribution of GM3 and GD3, which accounted for 95% of the ganglioside sialic acid at birth, decreased during the first 3 weeks of life. GM1, fucosyl-GM1, and B-GM1 were not detected at birth. The concentration of the fucogangliosides increased during the 2nd and 3rd weeks after birth, was stable during the 4th week and then decreased, whereas that of GM1 increased steadily between 6 days and 2 months of age. B-GM1, which has been defined as a tumor-associated ganglioside in the rat liver, was found to be a developmentally regulated antigen of the normal rat stomach.     

Age-related induction and disappearance of carcinogen-DNA-adducts in livers of rats exposed to low levels of 2-acetylaminofluorene

It was investigated whether in vivo aging of rat liver is associated with changes in the induction and rate of disappearance of DNA damage. For this purpose 6- and 36-month-old rats were intraperitoneally injected with a single, low dose (5 mg/kg body wt.) of the model liver carcinogen 2-acetylaminofluorene (AAF). Using the 32P-postlabeling assay we found that N-(deoxyguanosin-8-yl)-2-aminofluorene (dG-C8-AF) was the major DNA-adduct formed. The minor adduct N-(deoxyguanosin-8-yl)-2-acetylaminofluorene (dG-C8-AAF) could only be detected after doses of 20 mg/kg or more. Quantitation of adduct levels at various time points after treatment indicated a rapid induction of AF-adducts, which were already present at 6 h after treatment. The subsequent loss of AF-adducts was relatively slow, as was indicated by the presence of a substantial amount of AF-adducts as late as 21 days after treatment. Slight age-related differences in the pattern of induction and disappearance of AF-adducts and a somewhat higher level of persisting lesions in old than in young rats were observed.     

Species specificity affects the choice of S9 preparations for use in the hamster embryo cell transformation system

Before their use as a source of carcinogen-activating enzymes in the hamster embryo cell transformation assay, liver, kidney, lung, and small intestine S9 fractions from Syrian golden hamsters and Sprague-Dawley rats were evaluated for toxicity to hamster embryo target cells. Sprague-Dawley rat liver and kidney S9 were highly toxic to the hamster embryo cells (90 to 100%). When retested at lower concentrations these tissue fractions were still quite toxic (up to 75%). In contrast, hamster liver and kidney S9 were considerably less toxic (14 to 25%). The S9 preparations were also evaluated for their ability to metabolize N-2-acetylaminofluorene to 2-aminofluorene and N-hydroxy-acetylaminofluorene, products that transform hamster embryo cells. Large amounts of N-hydroxy-acetylaminofluorene were formed in the presence of preparations from hamster liver and small intestine, whereas kidney and lung S9 fractions were considerably less active. No detectable levels of N-hydroxy-acetylaminofluorene were formed after incubation of N-2-acetylaminofluorene with any of the rat S9 preparations. High levels of deacetylase activity were found in hamster liver and small intestine S9 fractions, at least eightfold higher than those obtained from equivalent rat preparations. Hamster kidney and lung S9 fractions showed low levels of deacetylase activity. There was no detectable activity in equivalent preparations from rats. When tested with N-2-acetylaminofluorene in the hamster embryo cell clonal transformation system, transformed colonies were obtained with hamster liver S9, with and without an external NADPH-generating system.     

The chemical carcinogen-induced enzyme, GDP-fucose: GM1 alpha 1----2 fucosyltransferase in rat liver and hepatoma: modulation by and association with phospholipids

The enzyme GDPFuc:GM1 alpha 1----2 fucosyltransferase, induced by chemical carcinogens in precancerous rat liver as well as rat hepatoma cells, was found previously to be membrane bound, and was inactivated by various detergents, while the activities of many other transferases are generally enhanced by detergents (Holmes, E.H. & Hakomori, S. (1983) J. Biol. Chem. 258, 3706-3717). The effects of phospholipids and detergents on rat hepatoma H35 cells, the conditions of solubilization and subsequent affinity chromatography of the enzyme, and a possible association of phospholipids with the enzyme have been studied with the following major results: The alpha 1----2 fucosyltransferase activity in Golgi membrane was diminished on treatment of membranes with phospholipase A1 or phospholipase C. The enzyme activity was stimulated 7-fold in the presence of cardiolipin or phosphatidylglycerol (and 3-fold by phosphatidylethanolamine) but not other phospholipids. The stimulatory effect of phosphatidylglycerol was eliminated when a variety of ionic or non-ionic detergents were added to the reaction mixture, with the exception of the cationic detergent G-3634-A, which provided a 10-fold total stimulation in the presence of phosphatidylglycerol. The kinetic analysis indicated that addition of phosphatidylglycerol has a negligible effect on apparent Km values but increases the Vmax of the enzyme 5- to 6-fold. The enzyme activity was solubilized by the dialyzable detergent CHAPSO without inhibition of the enzyme activity, and the solubilized enzyme in the presence of 0.4% CHAPSO is partially purified by chromatography on GDP-hexanolamine-Sepharose. Removal of CHAPSO from the affinity purified enzyme by dialysis resulted in a 66% loss of the original activity, which was restored by addition of phosphatidylglycerol. Chromatography of the affinity-purified enzyme with 3H-labeled phosphatidylglycerol on a Biogel A0.5 column indicated an association of the enzyme with the phospholipid that occurred only in the absence of detergent. These results suggest that phospholipid has a direct effect on the enzyme and that the inhibitory effect of detergents can be ascribable to disturbing interaction between phospholipids and the enzyme. A possible role of specific phospholipids on in vivo transferase activity for glycolipids is discussed.     

Protein degradation in lysosomes from chemically induced malignant rat hepatoma

Hepatocellular carcinomas were induced in rat liver by exposing the animals to diethylnitrosamine and 2-acetylaminofluorene in combination with partial hepatectomy. Light and electron microscopy demonstrated that the general appearance of the tumour tissue was that of highly differentiated malignant hepatocytic cells. Morphometrically there was a difference between normal and malignant cells in that the entire lysosomal apparatus was twice as large in malignant cells as in normal cells. This was mainly due to an increase in the fractional volume of autophagic vacuoles. A total lysosomal fraction (dense bodies and autophagic vacuoles) was isolated and characterized from both control and tumour livers. Marker enzyme analysis showed that the lysosomal enzyme activities were significantly lower in malignant liver tissue. Injection of leupeptin, an inhibitor of cathepsins B, H, and L, into rats did not increase the fractional volume of autophagic vacuoles in tumour tissue as much as in normal liver tissue. The proteolytic rate was lower in the lysosomal fraction from hepatoma cell tissue compared with the lysosomal fraction from normal cell tissue. This could conceivably be due to the lower activities of lysosomal enzymes. However, if the recovery of lysosomes is taken into account no clear-cut difference in lysosomal proteolysis between control and malignant liver tissue was noted. Accordingly, in malignant liver tissue a proteolytic balance is obtained characterized by an increased fractional volume of AVs and lower rate of protein degradation in individual lysosomes.     

Effect of gangliosides on the cytotoxic activity of natural killers from Syrian hamsters

The ability of various gangliosides to inhibit the cytotoxic activity of natural killers (NK-cells) from Syrian hamsters towards human lymphoma MOLT-4 cells was studied. The inhibitory effect was found to depend on the structure and concentration of the gangliosides. At concentrations corresponding to those in the blood of tumour-bearing hosts, SiaLacCer and Sia2LacCer inhibited the NK-activity. A significant inhibition was also found for (NeuAc)2GgOse3Cer, whereas NeuGcGgOse3Cer and NeuAcGgOse4Cer were practically inactive. Previously it was shown that Sia2LacCer which is either absent or very low in normal blood, is produced by a number of tumours and that tumour cells "shed" considerable amounts of gangliosides. On this ground, it was proposed that elevated concentrations of SiaLacCer and Sia2LacCer in the blood of tumour-bearing animals may inhibit the NK-activity and thus contribute to the "escape" of tumour cells from host immune surveillance.     

Inactivation of liver aldolase in fasting rabbits: Evidence for the accumulation of two different immunoreactive forms

During prolonged starvation the activity of aldolase in crude rabbit liver extracts decreases to less than one-half the value observed in extracts of livers from fed animals. The specific activity of the enzyme purified by adsorption on phosphocellulose and elution with substrate is also approximately one-half that of the purified native enzyme. Both the level of enzyme activity and the specific activity are restored to normal within 36 h of refeeding. After removal of active aldolase from the liver extracts by adsorption on phosphocellulose an additional immunoreactive protein can be isolated by adsorption on antialdolase-Sepharose and elution with 4 m MgCl₂. This protein is devoid of catalytic activity and in livers of fasted rabbits accounts for nearly 40% of the total immunoreactive material. It has also been detected in extracts prepared from livers of fed rabbits, where it accounts for 10–20% of the total protein adsorbed by antialdolase-Sepharose. The low-activity enzyme isolated from livers of fasted rabbits cannot be reactivated by sulfhydryl compounds; it shows similar sensitivity to heat and denaturing agents as the enzyme isolated from livers of fed rabbits. The activity ratios with fructose 1,6-bisphosphate, fructose 1-phosphate, and triose phosphate are similar to those observed for the native liver enzyme.