Specificities of the two genetically distinct beta-galactosidases in human sphingolipidoses

Validation of high-sensitivity assays to quantitate cerebrospinal fluid and serum β-galactosidase activity in patients with GM1-gangliosidosis

GM1-gangliosidosis (GM1) is a lysosomal storage disorder caused by mutations in the galactosidase beta 1 gene (GLB1) that leads to reduced β-galactosidase (β-gal) activity. This enzyme deficiency results in neuronal degeneration, developmental delay, and early death. A sensitive assay for the measurement of β-gal enzyme activity is required for the development of disease-modifying therapies. We have optimized fluorometric assays for quantitative analysis of β-gal activity in human cerebrospinal fluid (CSF) and serum for the development of a GLB1 gene replacement therapy. Assay analytical performance was characterized by assessing sensitivity, precision, accuracy, parallelism, specificity, and sample stability. Sensitivity of the CSF and serum β-gal activity assays were 0.05 and 0.20 nmol/mL/3 h, respectively. Assay precision represented by inter-assay percent coefficient of variation of the human CSF and serum was <15% and <20%, respectively. The effect of pre-analytical factors on β-gal activity was examined, and rapid processing and freezing of samples post-collection was critical to preserve enzyme activity. These assays enabled measurement of CSF and serum β-gal activities in both healthy individuals and patients with GM1-gangliosidosis. This CSF β-gal activity assay is the first of its kind with sufficient sensitivity to quantitatively measure β-gal enzyme activity in CSF samples from GM1 patients.

Genetic background markedly influences vulnerability of the hippocampal neuronal organization in the "twitcher" mouse model of globoid cell leukodystrophy

The twitcher mouse is well known as a naturally occurring authentic mouse model of human globoid cell leukodystrophy (GLD; Krabbe disease) due to genetic deficiency of lysosomal galactosylceramidase. The twitcher mice used most commonly are on the C57BL/6J background. We generated twitcher mice that were on the mixed background of C57BL/6J and 129SvEv, the standard strain for production of targeted mutations. Twitcher mice on the mixed background were smaller and had a shorter lifespan than were those on the C57BL/6J background. Many twitcher mice on the mixed background developed generalized seizures around 30 days that were never seen in twitcher mice on the C57BL/6J background. Neuropathologically, although the degree of the typical demyelination with infiltration of macrophages was similar in the central and peripheral nervous systems, in both strains, marked neuronal cell death was observed only in twitcher mice on the mixed background. In the hippocampus, the neuronal cell death occurred prominently in the CA3 region in contrast to the relatively well-preserved CA1 and CA2 areas. This neuropathology has never been seen in twitcher mice on the C57BL/6J background. Biochemically, the brain of twitcher mice on the mixed background showed much greater accumulation of lactosylceramide. Genetic background must be carefully taken into consideration when phenotype of mutant mice is evaluated, particularly because most targeted mutants are initially on a mixed genetic background and gradually moved to a pure background. These findings also suggest an intriguing possibility of important function of some sphingolipids in the hippocampal neuronal organization and maintenance.

Globoid cell leukodystrophy (Krabbe's disease): update

The classic globoid cell leukodystrophy (Krabbe's disease) is caused by genetic defects in a lysosomal enzyme, galactosylceramidase. It is one of the two classic genetic leukodystrophies, together with metachromatic leukodystrophy. The mode of inheritance is autosomal recessive. Typically, the disease occurs among infants and takes a rapidly fatal course, but rarer late-onset forms also exist. Clinical manifestations are exclusively neurologic with prominent white-matter signs. The pathology is unique, consisting of a rapid and nearly complete disappearance of myelin and myelin-forming cells--the oligodendrocytes in the central nervous system and the Schwann cells in the peripheral nervous system, reactive astroytic gliosis, and infiltration of the unique and often multinucleated macrophages ("globoid cells") that contain strongly periodic acid-Schiff (PAS)-positive materials. A normally insignificant but highly cytotoxic metabolite, galactosylsphingosine (psychosine), is also a substrate of galactosylceramidase and is considered to play a critical role in the pathogenesis. The galactosylceramidase gene has been cloned, and a large number of disease-causing mutations have been identified. Equivalent genetic galactosylceramidase deficiency occurs in several mammalian species, such as mouse, dog, and monkey. Recently, deficiency of one of the sphingolipid activator proteins, saposin A, was demonstrated to cause a late-onset, slowly progressive globoid cell leukodystrophy at least in the mouse, with all of the phenotypic consequences of impaired degradation of galactosylceramidase substrates. Human globoid cell leukodystrophy owing to saposin A deficiency might be anticipated and should be suspected in human patients with a late-onset leukodystrophy with normal galactosylceramidase activity when other possibilities are also excluded. The only serious attempt at treating human patients is bone marrow transplantation, which can provide significant alleviation of symptoms, particularly in those patients with later-onset, more slowly progressive globoid cell leukodystrophy.

Optimal assay conditions for enzymatic characterization of homozygous and heterozygous twitcher mouse

The neurological mouse mutant twitcher is characterized by a genetic deficiency of galactosylceramide beta-galactosidase (galcerase) (EC 3.2.1.46) which also represents lactosylceramide beta-galactosidase I (lactosidase I) activity. The assay conditions for both these activities in several mouse tissues have been optimized to facilitate the enzymatic characterization of homozygous and heterozygous twitcher mice. Galcerase in mouse tissues is optimally activated by 7.0 mg/ml of sodium taurocholate (pure) and 1.5-2.0 mg/ml of oleic acid in this system. When lactosylceramide is used as the substrate, no more than 1 mg/ml of taurocholate is appropriate in the assay, since higher concentrations of this pure bile salt stimulate another enzyme, lactosylceramide beta-galactosidase II (lactosidase II), which is unaffected in twitcher mice. At the optimized condition, lactosidase I in the twitcher mouse amounts to 3-4% of control activity in agreement with the residual galcerase (2%) in this mouse mutant. These assay conditions provide better sensitivity to discriminate heterozygotes from controls until 40 days of age from measurement of this activity in clipped tail samples.

Aggregation-dissociation and stability of acid beta-galactosidase purified from porcine spleen

Sucrose gradient centrifugation of the monomeric form (A1) of porcine spleen beta-galactosidase showed a pH-dependent equilibrium between monomer at neutral pH (pH 7.0) and dimer at acidic pH (pH 5.4-3.0), independent of ionic strength. While the oligomeric form (Ao), which was hardly dissociated under physiological conditions, was dissociated only with some protein denaturing agents into similar catalytic subunit to the A1. Both the A1 and Ao were equally active and stable at acidic pH, in the physiological condition inside lysosome (around pH 4.6).

A beta-galactosidase isoenzyme from Turbo cornutus with substrate specificity toward GM1-ganglioside and glycoproteins

beta-Galactosidase from T. cornutus was resolved into two activity peaks by gel filtration column chromatography. The pH optima of the two peaks designated P1 and P2, were 5.5 and 3.0, respectively, when p-nitrophenyl-beta-D-galactopyranoside was used as the substrate. The molecular weights of P1 and P2 were 700,000 +/- 70,000 and 78,000 +/- 7800, respectively, when estimated by gel filtration chromatography. The activities of both forms of the enzymes are stimulated by anions such as Cl-, Br- and NO-3. While the activity of P1 was stimulated by low anion concentrations, P2 requires 700 times higher anion concentration for similar enhancement of activity. P1, the high molecular weight form hydrolyzes mainly galactose from small molecular weight beta-galactosides, such as p-nitrophenyl-beta-D-galactopyranoside, 4-methylumbelliferyl-beta-D-galactopyranoside, lactose, lactosylceramide and 3-O-beta-D-galactopyranosyl-D-arabinose, whereas P2, the low molecular weight form cleaves, in addition, all the beta-galactosides tested, including 2-hexadecanoylamino-4-nitrophenyl-beta-D-galactopyranoside, GM1-ganglioside, asialo-GM1-ganglioside, asialo fetuin, alpha 1-acid glycoproteins and the tryptic peptides of the glycoproteins. The optimal conditions for the hydrolysis of the terminal galactose from GM1-ganglioside which does not occur in gastropods, such as T. cornutus, was found to require 40 mM NaCl and 1 mM sodium taurodeoxycholate at pH 3.0 in 50 mM sodium citrate buffer, conditions similar to those by mammalian beta-galactosidase.

Aggregation properties of beta-galactosidase of human urine and degradation of its natural substrates by a purified preparation of the enzyme

Acid beta-D-galactosidase (beta-D-galactoside galactohydrolase, EC 3.2.1.23) was purified to near homogeneity from normal human urine by two affinity chromatography steps. On polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate the major protein band had an apparent molecular weight of 59000, thus being 5000 daltons smaller than the protein purified from human liver. Upon gel filtration on Sephadex G-150 column the purified enzyme had an apparent molecular weight of 70000 of pH 7.0. At pH 4.0 partial aggregation to a dimer of an apparent molecular weight of 150000 was found. Addition of 0.1 M galactose caused at pH 3.5, but not at pH 4.0 and 7.0, an increased formation of multimeric beta-galactosidase which eluted with the void volume of the column. Crude beta-galactosidase from human urine showed a higher aggregation tendency than the purified enzyme. None of the conditions produced an enzyme species of an apparent molecular weight of less than 40000. pH-activity profiles were measured against p-nitrophenyl-beta-D-galactoside, 3H-labelled GM1-ganglioside, [3H]keratan sulfate and the pentasaccharide O-beta-(1 leads to 4)-[6-3H]galactopyranosyl-O-beta-(1 leads to 2)-2-deoxy-2-acetamidoglycopyranosyl-O-alpha-(1 leads to 6)-mannopyranosyl-O-beta-(1 leads to 4)-mannopyranosyl-2-deoxy-2-acetamidoglucopyranoside. While p-nitrophenyl-beta-D-galactopyranoside and GM1-ganglioside were optimally hydrolyzed at pH 4.0, keratan sulfate and the pentasaccharide were optimally degraded at pH 4.3 and pH 5.0, respectively. With the chromogenic substrate and with GM1-ganglioside Km values of 0.33 mM were calculated. At pH 3.5 the hydrolysis of the synthetic substrate did not follow Michaelis-Menten kinetics. Two enzyme species appeared with Km values of 0.006 mM and 3.2 mM, respectively. The affinity of beta-galactosidase for [3H]keratan sulfate and the 3H-labelled pentasaccharide was at least one order of magnitude lower than for the amphiphilic substrates. Keratan sulfate and GM1-ganglioside did not act as competitive inhibitors of p-nitrophenyl-beta-galactosidase at the concentration tested. These findings could be explained by the existence of different binding sites for the substrates used.

Studies of a synthetic substrate in canine globoid cell leukodystrophy

Gal et al. ((1977) Clin. Chim. Acta 77, 53-59) reported the use of a new synthetic substrate, 2-hexadecanoylamino-4-nitrophenyl-beta-D-galactopyranoside for the diagnosis of human globoid cell leukodystrophy. Assay of beta-galactosidase in brain homogenates from normal, carrier, and globoid cell leukodystrophy-affected dogs utilizing this new substrate demonstrated overlapping activities. Instead of reflecting specific D-galactosyl-N-acylsphingosine galactohydrolase (EC 3.2.1.46), the 2-hexadecanoylamino-4-nitrophenyl-beta-D-galactopyranoside beta-galactosidase activity in canine brain is highly correlated with nonspecific 4-methylumbelliferyl beta-galactosidase. Optimization of the 2-hexadecanoyl-amino-4-nitrophenyl-beta-D-galactopyranoside assay system for canine brain and the use of varying concentrations of taurocholate or taurodeoxycholate in the assay mixture did not alter the lack of specificity. These results indicate a significant difference in the nature of the underlying defect in galactosylceramide beta-galactosidase in canine globoid cell leukodystrophy compared to human globoid cell leukodystrophy.

Hydrolysis of galactose from glycolipids and glycoprotein by young rat brain beta-galactosidases immobilized to Sepharose 4B

An extract of glycosidic enzymes from young rat brain was immobilized to cyanogen bromide-activated Sepharose 4B. Most glycosidases retained approximately 10--25% of their activities after immobilization. Immobilized beta-galactosidases were used repeatedly without detectable loss of enzyme activity in the hydrolysis of p-nitrophenyl-beta-D-galactopyranoside. In addition to the synthetic substrate, the immobilized rat brain beta-galactosidases could also hydrolyze galactose from lactose, galactosylcerebroside, asialofetuin, and GM1-ganglioside. The hydrolysis of GM1- to GM2-ganglioside was confirmed on TLC.

Human brain cerebroside beta-galactosidase: deficiency of transgalactosidic activity in Krabbe's disease

Under experimental conditions optimal for the assay of D-galactosyl-N-acylsphingosine galactohydrolase (EC 3.2.1.46) activity, homogenates of neurologically normal human brain tissue could transfer galactose from galactosyl ceramide (gal-cer), lactosyl ceramide (lac-cer), 4-methylumbelliferyl-beta-galactoside (4-MU-gal), or p-nitrophenyl-beta-galactoside (PNP-gal) to [1-14C]oleoyl sphingosine, but homogenates of brain tissue from patients with Krabbe's disease lacked this ability. The rate of hydrolysis of ganglioside GM1 and, to a lesser extent, of PNP-gal by homogenates of Krabbe's brain tissue was also decreased. Activity of PNP-beta-galactosidase in normal brain tissue, like that of cerebroside beta-galactosidase from the same source, was considerably more heat-stable than the activity of either 4-MU-beta-galactosidase or the predominant GM1 beta-D-galactosidase (EC 3.2.1.23). Lac-cer and GM1, as well as 4-MU-gal and PNP-gal, were competitive inhibitors of human-brain cerebroside beta-galactosidase. These findings confirm the ability of mammalian cerebroside beta-galactosidase to catalyze a transgalactosylation reaction and provide additional information on the substrate specificity of human brain cerebroside beta-galactosidase.

Globoid cell leukodystrophy (Krabbe's disease). Metabolic studies with cultured fibroblasts

Metabolism of tritium-labelled galactosylceramide and lactosylceramide added to the culture medium was examined in cultured skin fibroblasts from 4 patients with globoid cell leukodystrophy (GLD) and 4 control individuals. The uptake of [3H]galactosylceramide and [3H]lactosylceramide by the fibroblasts continued actively at least up to 3 days. Approximately 30--40% of the galactosylceramide, which had been taken up, was released subsequently from the cells in a 4-day period, whereas only 10% of lactosylceramide was released during the same period. The GLD fibroblasts showed no abnormality in the kinetics of the uptake and in the release of these glycosphingolipids which are natural substrates of the beta-galactosidase genetically deficient in the disorder. This finding differs from that reported for fibroblasts from patients with metachromatic leukodystrophy, which showed abnormal accumulation and retention of sulfatide added to the culture media. However, degradation of added galactosylceramide to [3H]galactose by the GLD fibroblasts was only 25% of the control cells, while lactosylceramide was degraded at 70% of the normal rate. These findings are consistent with the known substrate specificities of the two acidic beta-galactosidases in human tissues; galactosylceramide is hydrolyzed almost exclusively by galactosylceramidase, while lactosylceramide can be hydrolyzed by both galactosylceramidase and GM1-ganglioside beta-galactosidase.

The two human lactosylceramidases and their respective enzyme activity deficiency diseases: inhibition studies using p-nitrophenyl-beta-D-galactoside

Total lactosyl ceramide beta-galactosidase (LC) activity from normal and pathologic human leukocytes and tissues was subdivided into LC I (EC 3.2.1.46) and LC II (EC 3.2.1.23) activity by means of specific inhibition of LC II with 5 mM p-nitrophenyl-beta-D-galactoside (Ki = 1.5 mM). In globoid-cell leudodystrophy, inhibition of total LC was nearly complete (only LC II is active), whereas in GM1-gangliosidosis Type 1, very little inhibition was found (only LC I is actict). Total LC activity was not significantly low in either of the diseases, which have different genetic origins. The ratio of LC I to LC II activity may display remarkable genetic variation in normal probands.

Genetic linkage studies of the human glycosphingolipid beta-galactosidases

The genetic linkage relationships of the human glycosphingolipid beta-galactosidases were determined using human--mouse somatic cell hybrids. A new method was devised for the estimation of human galactosylceramide, lactosylceramide, and GMI-ganglioside beta-galactosidase activities in the presence of their mouse counterparts, which takes advantage of the reproducible specific activity of lysosomal hydrolases under a given set of culture conditions and is based on differences in both pH optima and sensitivity to chloride ion. Human and mouse chromosomes were identified by their characteristic banding patterns obtained after quinacrine staining, and the optimum glycolipid beta-galactosidase activity was determined for three different substrates. A ratio was defined for each activity which was the specific activity at the human pH optimum divided by the specific activity at the mouse pH optimum. Linear regression analysis was used to test for concordant segregation between pH ratios for each enzyme and the frequency of occurrence of different human chromosomes in the man--mouse somatic hybrid clones. The results obtained from two independent series of hybrid clones indicated that human beta-galactosidase activities consistently segregated with human chromosome 12 in these somatic cell hybrids.

GM1-ganglioside beta-galactosidase in leukocytes and cultured fibroblasts

GM1-ganglioside hydrolysis by leukocytes and fibroblasts, tissues easily obtainable from patients, was investigated using 3H-labeled GM1 and was found to be at least as active as that reported for any other tissue. Sodium taurocholate was required for the reaction, the crude bile salt at an optimum concentration of 0.4% producing twice as much activity as pure taurocholate at its optimum concentration of 0.8%. Leukocyte GM1-ganglioside beta-galactosidase and 4-MU-beta-gal cleaving activities were similar, 134.5 +/- 23.3 and 179.8 +/- 25.4 nmol/h/mg protein, respectively. In cultured skin fibroblasts and amniotic fluid cells these enzyme activities were 4 to 5 times higher. Homozygotes for GM1-gangliosidosis showed negligible activity while in heterozygotes the leukocyte GM1-cleaving activity was reduced to one-third of control values. In leukocytes from patients with four other sphingolipid storage diseases the activity was either normal (Krabbe's, Tay-Sachs, Metachromatic leukodystrophy) or increased (adult Gaucher's).

Glycosphingolipid hydrolases: properties and molecular genetics

This is a review of the properties and molecular genetics of six lysosomal hydrolases: beta-galactosidase, hexosaminidases A and B, alpha-galactosidase, beta-glucosidase and alpha-fucosidase. Each enzyme is discussed with regards to isoenzymes and substrate specificity, subunit structure, genetic relationship of isoenzymes and genetic variants. The molecular genetics of human diseases caused by deficiencies of each enzyme are discussed.

Lactosylceramidase assays for diagnosis of globoid cell leukodystrophy and GM1-gangliosidosis

An assay procedure was developed for accurate estimation of lactosylceramidase II in the presence of relatively high activity of lactosylceramidase I. The procedure involves determination of lactosylceramide-cleaving activities under two different assay conditions, and lactosylceramidase II activity is calculated by the difference. Applicability of the procedure was evaluated with separated soluble fractions of the two beta-galactosidases from normal human brains, and with whole homogenates of gray and white matter, liver and cultured fibroblasts from control individuals and from patients with globoid cell leukodystrophy or GM1-gangliosidosis. The use of the lactosylceramidase I assay procedure developed by Wenger, D.A., Sattler, M., Clark, C. and McKelvey, H. ((1974) Clin. Chim. Acta 56, 199-206) and of the present procedure permits accurate diagnosis of both globoid cell leukodystrophy and GM1-gangliosidosis with one natural substrate, lactosylceramide, irrespective of the relative proportion of the two beta-galactosidases in the tissue.

Substrate specificities of the two genetically distinct human brain beta-galactosidases

The two human brain beta-galactosidases were solubilized and fractionated by Sephadex G-200 gel filtration, free from each other. Substrate specificities of the two enzymes were examined for galactosylceramide, lactosyl-[N-stearoyl]ceramide, lactosyl-[N-lignoceroyl]ceramide, galactosyl-N-acetylgalactosaminyl-[N-stearoyl]ceramide, lactosyl-[N-lignoceroyl]ceramide, galactosyl-N-acetylgalactosaminyl-[N-acetylneuraminyl]galactosyl-glucosylceramide (GMI-ganglioside), galactosyl-N-acetylgalactosaminyl-galactosyl-glucosylceramide (asialo GM1-ganglioside), and 4-methylumbelliferyl beta-galactoside. Under appropriately optimized conditions, either of the two beta-galactosidases could hydrolyze all of the substrates, although with widely varying rates. Relative specific activities of galactosylceramide beta-galactosidase toward galactosylceramide, lactosyl-[N-steroyl]ceramide, lactosyl-[N-lignoceroyl]ceramide. GM1-ganglioside, asialo GM1-ganglioside, and 4-methylumbelliferyl beta-galactoside were 100, 510, 250, 39, 41 and 120, respectively. Relative specific activities of GM1-ganglioside beta-galactosidase toward the same series of the substrates were 0.3, 78, 19, 100, 150 and 240; However, the optimal assay conditions for any given natural substrate were sufficiently different for each beta-galactosidase so that diagnostic assays for the two genetic diseases due to beta-galactosidase deficiencies could be carried out in whole tissues. Since the relative distribution of the two enzymes vary greatly in different tissues, contributions by the two enzymes to degradation of the natural glycosphingolipids in vivo may well vary in different organs. These findings may have an important bearing on the biochemical pathogenesis of these genetic disorders.