Search

Among the genetic leukodystrophies known to occur in man, the fundamental genetic defects have been clarified in two disorders, metachromatic leukodystrophy and globoid cell leukodystrophy (Krabbe's disease). Nevertheless, the biochemical sequences with which the underlying genetic defects lead to the clinical and pathological features of the diseases and to disruption of the normal brain function are not well understood. Comparison of the two classical leukodystrophies provides us an opportunity to consider the biochemical pathogenesis of these disorders, some common between them and others apparently unique. That both metachromatic leukodystrophy and globoid cell leukodystrophy manifest themselves almost exclusively as diseases of the myelin sheath is relatively easy to understand because both are caused by genetic abnormalities in the metabolism of the characteristic constituents of myelin, sulfatide and galactosylceramide. The presence of the abnormal and characteristic globoid cells in the white matter of Krabbe's disease patients appears to be due to a unique property of galactosylceramide in that, when present in free form in the brain, it elicits infiltration of macrophages which transform themselves to globoid-like cells. No other lipids, including sulfatide, are known to induce similar tissue reactions. The most conspicuous difference between the two diseases is the presence or absence of abnormal accumulation of the substrates, the degradation of which is genetically blocked in the respective diseases. In metachromatic leukodystrophy, sulfatide accumulates to abnormally high levels, as logically expected for a "lysosomal storage disease". In Krabbe's disease, on the other hand, galactosylceramide is always much less than normal despite the genetic block in its catabolic pathway. This paradoxical finding can be explained by the "psychosine hypothesis".(ABSTRACT TRUNCATED AT 250 WORDS)

This paper addresses the similarities and differences in the topology of the catalytic centres of human liver cytosolic beta-glucosidase and placental lysosomal glucocerebrosidase, and utilizes well-documented reversible active-site-directed inhibitors. This comparative kinetic study was performed mainly to decipher the chemical and structural nature of the active site of the cytosolic beta-glucosidase, whose physiological function is unknown. Specifically, analysis of the effects of a family of alkyl beta-glucosides consistently displayed 100-250-fold lower inhibition constants with the cytosolic broad-specificity beta-glucosidase compared with the placental glucocerebrosidase; for example, with octyl beta-D-glucoside the Ki values were 10 microM and 1490 microM for the cytosolic and lysosomal beta-glucosidases respectively. Furthermore the higher affinity of the cytosolic beta-glucosidase than glucocerebrosidase for the amphipathic alkyl beta-D-glucosides was validated by the greater increase in the free energy of binding with increasing alkyl chain length [delta delta G0 (K,)/CH2: lysosomal enzyme, 2.01 kJ/mol (480 cal/mol); cytosolic enzyme, 3.05 kJ/mol (730 cal/mol)]. The implications of the presence of highly non-polar domains in the active site of the cytosolic beta-glucosidase are discussed with regard to its potential physiological substrates.

We developed a sensitive and simple procedure for determination of galactosylsphingosine (psychosine), using HPLC. The method involved extraction of lipids, separation by cation-exchange and C18 reverse-phase columns, and derivatization with o-phthalaldehyde. The fluorescent galactosylsphingosine was detected by HPLC. The amount of galactosylsphingosine was accurately assayed by simultaneous determination of glucosylsphingosine, as the internal standard. The detection limit was 0.5 ng/assay tube, and the quantitative range of the method was up to 750 ng. This procedure was applied to tissue from the twitcher mouse, an animal model of human globoid cell leukodystrophy, as well as tissue from normal and carrier mice. In the latter mice, a small amount of galactosylsphingosine was detected in the spinal cord (21.6-37.2 ng/100 mg wet weight) but not in the cerebrum and sciatic nerve. Marked accumulation of galactosylsphingosine was noted in the nervous tissues of the twitcher strain, even on postnatal day 4. The concentration of galactosylsphingosine was greater in the peripheral than in central nervous tissues. The spinal cord and brainstem contained more galactosylsphingosine than did the cerebrum and cerebellum. The concentration increased with age from 764 ng/100 mg in the sciatic nerve at 4 days to 5,910 ng/100 mg at 37 days. These data correlate well with the pathological changes; tissues containing higher concentrations of galactosylsphingosine show earlier and more severe pathological changes than those containing lower concentrations, thereby indicating the close link of galactosylsphingosine to the pathogenesis of the twitcher mouse.

Psychosine (galactosylsphingosine) is enzymatically synthesized from UDP-galactose and sphingosine and degraded by galactosylceramidase. Galactosylceramidase is genetically deficient in Krabbe disease (globoid cell leukodystrophy) and psychosine accumulates in the brain of humans, dogs or mice affected by the disease. Psychosine has a very potent inhibitory effect on cytochrome c oxidase (COX) in mitochondria. When COX is purified, psychosine does not suppress the enzyme activity. However, a clear inhibitory effect is seen when the enzyme is "reconstituted" with sonicated phosphatidylcholine, suggesting that the inhibition of the enzymatic activity is caused by perturbation of the environment of COX in the mitochondrial membrane. Studies using analogues suggest that the free amino group in the sphingosine moiety plays an important role in exerting the effects of psychosine. The effects of psychosine are not only potent but fast and reversible. It is noteworthy that a powerful inhibitor of cellular respiration is synthesized in mammals' brain, an organ vulnerable to hypoxia.

Oligodendrocytes were isolated from the brain of the twitcher (twi/twi), an authentic murine model of globoid cell leukodystrophy (GLD), carrier (+/twi) and their littermate controls (+/+) and were maintained in vitro for 24 days. By 4 days in vitro (4 DIV) oligodendrocytes developed thorny processes and their morphology were closely similar to each other regardless of their genetic status. After 17 DIV, however, oligodendrocytes in twi/twi progressively degenerated and only 12% of oligodendrocytes counted at 10 DIV survived at 24 DIV in twi/twi, while in +/twi and +/+, mean survival rates were 85 and 83% respectively. Characteristic inclusions of GLD were detected in the perikarya of degenerating twi/twi oligodendrocytes indicating that metabolic defect was expressed even in isolated oligodendrocytes. These results further support the hypothesis that the primary pathogenetic event in murine GLD, twitcher, is degeneration of oligodendrocytes due to progressive accumulation of the toxic metabolite, galactosylsphingosine (psychosine).

Enzymatic properties of beta-galactosidases with galactosylsphingosine (psychosine) and lactosylsphingosine as the substrates were examined. Although bile salts were stimulatory on the hydrolysis of the glycolipids in normal brain and cultured fibroblasts, the hydrolytic activities could be readily assayed, without detergents. The in vitro hydrolysis of lactosylsphingosine in cultured fibroblast homogenates was catalyzed by two enzymes, as is the case with the hydrolysis of galactosylceramide and lactosylceramide. Lactosylsphingosine beta-galactosidase activities assayed in the absence and the presence of taurocholate (probably lactosylceramidase I) were deficient in fibroblasts from patients with globoid cell leukodystrophy, while the activity assayed with sodium cholate (probably lactosylceramidase II) was deficient in GM1 gangliosidosis fibroblasts. In contrast, galactosylsphingosine beta-galactosidase was not activated by cholate and the enzyme activities assayed with the no-additive and taurocholate systems were deficient in brain and fibroblasts from patients with globoid cell leukodystrophy, thereby indicating that the hydrolysis of galactosylsphingosine is catalyzed by one enzyme, galactosylceramidase I. Exogenous lipids and an activator protein purified from normal spleen activated galactosylsphingosine beta-galactosidase but they were inhibitory to lactosylsphingosine beta-galactosidase. Because the Km values of lactosylsphingosine beta-galactosidase assayed with cholate were several magnitude higher than those obtained with the no-additive system and because lactosylsphingosine is readily hydrolyzed with the no-additive system in vitro, it is likely that the in vivo hydrolysis of the lipid is catalyzed by only one enzyme, lactosylceramidase I.

Arylsulphatase A and 2',3'-cyclic nucleotide 3'-phosphohydrolase activities of myelin deficient mutant mice brains were studied. The results indicated that there were no changes in arylsulphatase A activity of the developing mutant brain, whereas the activity of 2',3'-cyclic nucleotide 3'-phosphohydrolase decreased considerably. The data obtained in this study suggest that in brain arylsulphatase A activity is localized in cells other than oligodendroglia.

The synthesis of cerebronic acid from lignoceric acid by rat brain postnuclear fraction is stimulated by sphingosine and psychosine (galactosyl sphingosine), which are the long chain bases of cerebrosides. Maximum stimulation by sphingosine occurred at 80 μM and by psychosine at 160 μM. A further increase in concentration of either of these bases decreased the degree of the stimulation. All newly synthesized cerebronic acid is found as a component of ceramide and cerebroside with or without the addition of these bases. When radioactive psychosine ([6- 3 H]galactosyl sphingosine) was added to the incubation mixture, the incorporation of 3 H into cerebroside was much less than would be expected if psychosine acts as the acceptor of the newly synthesized cerebronic acid. threo -Psychosine (a diastereomer of the naturally occurring compound) stimulated the α-hydroxylation of lignoceric acid as much as erythro -psychosine (the natural form), while glucosyl sphingosine showed only a slight stimulation. d -Cerebronic acid, the naturally occurring isomer and the product of the α-hydroxylation, strongly inhibited the α-hydroxylation. On the other hand, l -cerebronic acid (the enantiomorph of naturally occurring cerebronic acid) and 2-hydroxypalmitic acid (which is not a detectable brain cerebroside component) did not inhibit the reaction. Free cerebronic acid did not accumulate in the system even in the presence of a large excess of d -cerebronic acid. d -[1- 14 C]Cerebronic acid which was added to the incubation system, was not incorporated into ceramide or cerebroside under these conditions. These findings suggest that the newly synthesized cerebronic acid is incorporated into sphingolipids without being released as free acid.

Galactosylsphingosine, glucosylsphingosine and sphingosine all inhibited cytochrome c oxidase activity in mitochondria from rat liver; more than 50% inhibition was caused by 5 microM lipid (0.1 mumol/mg mitochondrial protein). However, these lysosphingolipids did not suppress the activity of purified cytochrome c oxidase. When the enzyme was "reconstituted" with phosphatidylcholine, the lysosphingolipids clearly inhibited the activity. On the other hand, galactosylsphingosine, glucosylsphingosine and sphingosine all hemolyzed erythrocytes, indicating that lysosphingolipids can disrupt the membrane. Thus, it appears that the inhibition of cytochrome c oxidase, a membrane-bound enzyme in mitochondria, is due to perturbation of the environment of the enzyme and that the primary attacking site of the lysosphingolipids is the membrane. Because the potency to inhibit cytochrome c oxidase and to hemolyze erythrocytes did not differ among these lysosphingolipids and because galactosylceramide caused neither inhibition of cytochrome c oxidase nor hemolysis, the free amino group in the lysosphingolipids seems to be essential to give the effects. In addition, both inhibition of cytochrome c oxidase and hemolysis caused by lysosphingolipids were completely abolished by albumin, suggesting that toxic effects of lysosphingolipids may not be apparent in blood.

A new method was developed for the simultaneous determination of psychosine and cerebrosides in tissues. Total lipids extracted from the tissues were treated with [3H]acetic anhydride in toluene-methanol. Known amounts of nonradioactive N-acetylpsychosine were added to the reaction product and then subjected to mild alkaline methanolysis. After the product was washed, it was fractionated by silica gel column chromatography and the fraction containing glycolipids was benzoylated. The benzoylated product was finally fractionated on TLC. The amounts of benzoylated derivatives of nonhydroxy- and hydroxycerebrosides and N-acetylpsychosine were determined using a scanning densitometer. The amounts of psychosine in tissues were calculated from the radioactivity in the spot of N-acetylpsychosine and the recovery of added carrier N-acetylpsychosine. This method allowed us to determine 5 to 1000 pmol of psychosine and 1 to 20 nmol of cerebrosides in peripheral nerves and other tissues of the twitcher mouse as well as transfected Schwann cells derived from the sciatic nerves.

Three major clinical variants of Gaucher disease have been defined: Type I, chronic nonneuronopathic; Type II, acute neuronopathic; and Type III, subacute neuronopathic. In a search for the underlying molecular basis of the neurological manifestations, the concentration and composition of cholesterol, phospholipids, neutral glycosphingolipids, and gangliosides were examined in cerebral and cerebellar cortices of five cases of Type II, eight cases of Type III, and one case of presumed Type I/III. In Type II the concentration of glucosylceramide was 140-530 mumol/kg in cerebral cortex and 51-450 mumol/kg in cerebellar cortex, the highest values found in the most fulminant cases. These concentrations were 20-80 times greater than normal in cerebral cortex and 5-40 times normal in cerebellar cortex. In type III the concentration of glucosylceramide was 37-65 and 59-1750 mumol/kg in cerebral and cerebellar cortex, respectively. The highest concentrations were found in the cerebellum of patients who had survived splenectomy for several years. The ceramide composition of the accumulated glucosylceramide suggested that brain gangliosides were the major precursors of the glucosylceramide in brains of Type II but in cerebellar cortex in Type III was partly of extracerebral origin. The levels of lactosylceramide and oligohexaosylceramides were slightly raised in all brain specimens from the Gaucher cases. The ganglioside concentration was normal, whereas there was a certain increase in the proportion of GM2 and GM3 gangliosides. The brain glycosphingolipid changes in the Type I/III case were similar but slightly less than those in Type III cases of corresponding age. Glucosylsphingosine (psychosine), never detected in normal human brain, was demonstrated in brains from all the Gaucher cases. The psychosine concentration was highest in Type II cases, 3.8-8.8 and 3.9-12.3 mumol/kg in cerebral and cerebellar cortex, respectively, with the highest values found in the most fulminant cases. In type III the psychosine concentration varied more widely, 0.8-4.6 and 1.4-6.3 mumol/kg in cerebral and cerebellar cortex, respectively. The lowest value, 0.7 mumol/kg, was found in the Type I/III case. Our method detected psychosine down to 0.01 mumol/kg, which means that the concentration of psychosine was increased at least 100- to 1000-fold in Gaucher grey matter. We suggest that the accumulation of the cell-toxic substance psychosine is the basis for the extensive neuronal cell loss in Gaucher disease, which is most striking in Type II disease.

Thirteen years ago I found that galactosyl ceramide undergoes metabolic conversion in the brain of young ratsa). Month-old rats were injected with 14C-galactose, rats were sacrificed at various intervals, and the brain lipids were fractionated according to the following scheme (fig. 1). The radioactivity in the galactose of each lipid group is shown in fig. 2, which plots total activity against time. The top curve, from cerebrosides, shows a slow but distinct conversion of brain cerebroside. Judging by the somewhat slower rate of rise in sulfatide activity (bottom curve), this conversion could - in part, at least - be due to conversion of cerebroside to sulfatide. Confirmation of this guess came from the work of Bachhawat and McKhann some years later z, 3). (The middle curve, for gangliosides, shows this group of lipids has a higher turnover rate than cerebrosides.) There was some reason to believe that another portion of the metabolic conversion of cerebrosides involved hydrolysis of the galactosidic linkage. Fujino 4) and Thannhauser and Reichel ~) had reported rather incomplete evidence for the presence in brain of a cerebroside galactosidase, and it was also known that cerebrosides can disappear in demyelinative processes. Accordingly, we made some radioactive cerebroside and looked for the hydrolysis. The preparation of the labeled lipid was based on a procedure of Carter and Fujino 6) for the formation of psychosine sulfate (fig. 3). The first reaction is formulated in an unusual way in order to explain something we found a few years later: when a mixture of cerebrosides is used, containing both hydroxy and nonhydroxy fatty acids, one obtains a great deal of unhydrolyzed cerebroside of the latter type. Carter and Koob simultaneously made a similar observation. We have not made a serious study of the reaction, but it looks as though only the hydroxy fatty acid amide bond is attacked by the barium ions, and I suggest the effect involves formation of a barium chelate ofhydroxy fatty acids, shown in the top equation. The last step in the synthesis is an adaptation of the procedure of Shapiro

1. A method for the synthesis of 3-O-alkylated cerebrosides has been developed. The synthesis of 1-O-β-D-galactosyl-N-palmitoyl-3-O-hexadecyldihydrosphingosine has been performed. 2. The introduction of voluminous substituents into the molecule of a dihydroceramide and a dihydrocerebroside adjacent to the amide grouping makes the amide bond in these compounds resistant to alkaline hydrolysis. 3. Conditions for the alkaline hydrolysis of cerebrosides and 3-O-alkylcerebrosides leading to psychosine and 3-O-alkylpsychosines with yields of 30–35% have been proposed.

We have studied the possible involvement of the calcium- and phospholipid/diacylglycerol-dependent enzyme, protein kinase C (PKC) in mediating insulin action in the central nervous system (CNS) by testing the effect of direct activation or blockade of the CNS PKC system on the plasma glucose responses to central insulin injection in mice. Insulin (0.1–1 μg), injected into the CNS, produced rapid transient hypoglycemia. This effect appeared to involve interaction of insulin with specific receptors, since insulin analogs exhibiting diminished receptor binding affinity and peripheral bioactivity compared to the native hormone were much less active (i.e., insulin ⋙ acetyl 3 insulin > proinsulin > IGF-I) or not active at all (i.e., insulin chain A and B). Central injection of the specific PKC activator, 12-O-tetradecanoylphorbol-13-acetate (TPA) (0.01–0.5 μg), but not the inactive TPA analog, 4-α☎orbol or the unstable synthetic diacylglycerol analog, 1-oleoyl-2-acetyl-sn-glycerol (OAG), significantly enhanced the hypoglycemic response to co-administered insulin (0.5 μg) or the insulin derivative, acetyl 3 insulin (2.5 μg). Central TPA had no effect on basal glucose levels. Furthermore, central administration of the selective PKC blockers, polymyxin B (PMB, 1–25 μg) or 1-β-galactosylsphingosine (psychosine, 0.5–10 μg) but not their respective inactive analogs, polymyxin E and sphingomyelin, strongly inhibited the hypoglycemic response to insulin (1 μg) or acetyl 3 insulin (5 μg). PMB and psychosine, injected alone had no effect on basal glucose levels. These findings, of a significant enhancement or blockade of the hypoglycemic response to central insulin following direct, selective activation or inhibition, respectively, of the CNS PKC system are consistent with the view that PKC might play a role in the mediation of insulin action in the CNS.

The primary genetic defect underlying Krabbe disease or globoid cell leukodystrophy is considered to be a deficiency of galactosylceramide-beta-galactosidase. In the present study of the brains from 18 patients who had died from Krabbe disease at 7-37 months of age, the concentration of galactosylceramide of cerebral and cerebellar white matter was severely reduced to 10-20% of that in age-matched controls. The lowest values were found in the most long-standing cases. Lactosylceramide was reduced to about 50% of normal, while globotriaosylceramide, blobotetraosylceramide and III3-alpha-fucosylneolactotetraosylceramide were increased 10 to 100-fold. Two glycosphingolipids, which have never before been isolated from normal human brains were now isolated and characterized: galactosylsphingosine (psychosine) and galactosyl beta 1 leads to 4 galactosylceramide. We were unable to identify galactosylsphingosine in normal human brains with certainty. We estimate its concentration in the cerebral white matter in Krabbe disease to be increased at least 100-fold (higher than normal). Psychosine was isolated also from the cerebral cortex in Psychosine was isolated also from the cerebral cortex in Krabbe disease after derivatization to the N-acetyl form. Its concentration there was 1 nmol/g tissue compared with 6-10 nmol/g in the white matter. All the neutral glycosphingolipids were isolated and their structure proved by the quantitative determination of their components, degradation by acid and specific glycohydrolases and permethylation and gas-liquid chromatographic-mass spectrometric assay of the methylated sugars. The paradoxical findings of a severely reduced concentration of galactosylceramide and a primary deficiency of cerebroside-beta-galactosidase can be explained by the present finding of the accumulation of galactosylsphingosine in the brains from patients who had died from Krabbe disease. The enzyme has a broad specificity and it normally also degrades galactosylsphingosine. Because of competitive inhibition by the accumulated galactosylceramide its lysosomal hydrolysis will be blocked. The concentration of psychosine will steadily increase and reach toxic levels and kill the oligodendroglial cells. This results in an arrest of the galactosylceramide biosynthesis. Therefore, we feel that galactosylsphingosine and not galactosylceramide is the primary storage substance in the brain in Krabbe disease that the disease is a psychosine lipidosis.

beta-Galactosidase (beta-D-galactoside galactohydrolase, EC 3.2.1.23) purified from Aspergillus oryzae was modified with 2,4,6-trichloro-s-triazine derivatives of polyethylene glycol (activated BPEG) having molecular weights of 600, 1500, 2000, and 4000. Polyethylene glycol derivatives were attached to 6 of the 12 amino groups exposed on the surface of the enzyme. Upon modification, the enzymatic activity for a water-soluble substrate, o-nitrophenyl beta-D-galactopyranoside, was reduced with increasing molecular weight of the activated BPEG. On the contrary, the enzymatic activity for another substrate, 4-methylumbelliferyl beta-D-galactopyranoside, was increased upon modification. The Michaelis constants of native and modified enzymes for these two substrates were virtually the same. The effect of the modification was more marked in the enzymatic hydrolysis of the beta-galactosidic bond of amphipathic substrates. A fluorescent analog of naturally occurring galactocerebroside, 1-O-galactosyl-2-N-(1-dimethylaminonaphthalene-5-sulfonyl)-sphingosine, was hydrolyzed more rapidly by the modified enzyme than by the native one. The enzyme modified with activated BPEG of 1500 Da had the highest activity for this substrate. The beta-galactosidic bond of the terminal galactose of GM1-ganglioside (II3NeuAcGgOse4Cer, galactosyl-N-acetylgalactosaminyl-(N-acetylneuraminosyl)-galactosyl -glucosylceramide) was cleaved by the modified but not by the native enzyme.

Lysosomal beta-glucosidase ('glucocerebrosidase') in peripheral blood lymphocyte and spleen extracts from normal individuals and Ashkenazi-Jewish Gaucher disease type-1 patients were investigated using several modifiers of glucosyl ceramide hydrolysis. The negatively charged lipids, phosphatidylserine and taurocholate, had differential effects on the hydrolytic rates of the normal and Gaucher disease enzymes from either source. With the normal enzyme, either negatively charged lipid (up to 1 mmol/l) increased the reaction rates, while decreasing hydrolytic rates were obtained at greater concentrations. In comparison, the peak activities of the Gaucher enzymes were observed at about 2-3 mmol/l or 5-8 mmol/l of phosphatidylserine or taurocholate, respectively. These negatively charged lipids altered only the velocity of the reactions; the apparent Km values were not affected. Taurocholate or phosphatidylserine also facilitated the interaction of the normal enzyme with conduritol B epoxide, a covalent inhibitor of the catalytic site. Compared to the normal enzyme, the Ashkenazi-Jewish Gaucher type-1 enzyme required about 5-fold greater concentrations of conduritol B epoxide for 50% inhibition. Neutral or cationic acyl-beta-glucosides were found to be competitive or noncompetitive inhibitors of the enzymes, respectively. Alkyl beta-glucosides were competitive (or linear-mixed type) inhibitors of the normal splenic or lymphocyte enzyme with competitive inhibition constants (Ki) inversely related to the chain length. With octyl and dodecyl beta-glucoside nearly normal competitive Ki values were obtained with the splenic enzymes from Gaucher patients. These Ki values were not influenced by increasing phosphatidylserine or taurocholate concentrations. In contrast, the cationic lipids, sphingosyl-1-O-beta-D-glucoside (glucosyl sphingosine) and its N-hexyl derivative, were noncompetitive inhibitors whose apparent Ki values for the normal enzyme were 30 and 0.25 mumol/l, respectively. The Ki values for these sphingosyl glucosides were about increased 5 times for the Gaucher type-1 enzymes from Ashkenazi-Jewish Gaucher disease type-1 patients. The Ki values of glucosyl sphingosine for the normal or mutant enzymes were directly related to increasing concentrations of phosphatidylserine or taurocholate. This latter site appears to be specifically altered by a mutation in the structural gene for lysosomal beta-glucosidase in the Ashkenazi-Jewish form of type-1 Gaucher disease.

Two genetically distinct acid beta-galactosidases are apparently involved in the hydrolysis of galactosylceramide in fibroblasts. These beta-galactosidases were activated by different bile salts. The classical galactosylceramidase (galactosylceramidase I, EC 3.2.1.46) was activated by sodium taurocholate, while the other galactosylceramidase (galactosylceramidase II) was activated by sodium cholate. The former was genetically lacking in globoid cell leukodystrophy (GLD) and the latter in GM1 gangliosidosis. Galactosylceramidase II cross-reacted with antibody raised against purified GM1 ganglioside beta-galactosidase (EC 3.2.1.23) from the human placenta. The purified beta-galactosidase had galactosylceramidase II activity, which was competitively inhibited by GM1 ganglioside. Thus, galactosylceramidase II seems to be identical to GM1 ganglioside beta-galactosidase and lactosylceramidase II. Galactosylceramidase II had a very low affinity for galactosylsphingosine. In the galactosylceramide-loading tests using fibroblasts from patients with GLD and GM1 gangliosidosis, both cell lines hydrolyzed the incorporated galactosylceramide, with lower rates than control fibroblasts but higher than the fibroblasts from patients with I-cell disease, in which both galactosylceramidase I and II were deficient. These results indicate that galactosylceramide is hydrolyzed by two genetically distinct beta-galactosidases and explain well that galactosylsphingosine but not galactosylceramide accumulates in the brain of patients with GLD.

Metachromatic leucodystrophies (MLD) comprise a small group of heredodegenerative disorders of the nervous system. Deficiency of sulfatide-sulfatase or arylsulfatase A is the common defect in all forms of MLD leading to lysosomal sulfatide storage in the nervous tissue and in the kidney. On the basis of animal experiments, experiments with cultured fibroblasts of the patients as well as ultrastructural studies in a case of prenatal MLD, the following pathomechanism is proposed: 1. Lysosomal degradation of a large portion newly synthetised sulfatide normally regulating the net synthesis and incorporation of sulfatide into myelin, causes early accumulation of sulfatide in lysosomes of myelinating cells and in neurons in the genetic deficiency of arylsulfatase A. 2. Early accumulation of sulfatide does not lead to disturbance in myelination. Demyelination occurs possibly by storage of a cytotoxic compound, psychosin sulfate, also a substrate for the missing enzyme. Prevention of MLD is possible by prenatal diagnosis of arylsulfatase A deficiency in cultured amniotic cells. Enzyme substitution of the missing arylsulfatase A is possible by exogenous uptake of the enzyme in cultured fibroblasts. Thereby the defect of sulfatide degradation can be corrected. Although principles of enzyme substitution have been demonstrated, the problems of treating patients with MLD with arylsulfatase A infusions have yet to be overcome.

We examined chronological changes of myelin proteins of the brainstem and spinal cord of the twitcher mouse (15, 20, and 30 days old), a murine model of human globoid cell leukodystrophy caused by a genetic deficiency of galactosylceramidase I activity. The yield of myelin was normal until postnatal day 20, whereas galactosylsphingosine (psychosine) accumulated with age in myelin. The protein profiles of myelin and the activity of 2',3'-cyclic nucleotide 3'-phosphodiesterase in the myelin remained normal throughout the experimental period. Fatty acylation of proteolipid protein (PLP) was examined in a cell-free system by incubation of myelin with [3H]palmitic acid, CoA, and ATP, and was normal at postnatal day 15, but decreased after postnatal day 20. Decreased fatty acylation of PLP was also observed in the twitcher mouse at postnatal day 20 when the isolated myelin was incubated with [14C]palmitoyl-CoA in the absence of ATP and CoA, or the slices of brainstem and spinal cord were incubated with [3H]palmitic acid. The activity of fatty acid:CoA ligase was reduced in myelin. These data suggest that decreased acylation of PLP in twitcher mouse myelin is probably due to reduced activities for both activation and transfer of fatty acid into PLP and that metabolic disturbance is present in myelin because acylation of PLP has been shown to occur in myelin membrane. Although psychosine (200 microM) inhibited only 17% of the acylation in vitro, it may be responsible for the reduced acylation of PLP in vivo.

Progressive accumulation of a cytotoxic metabolite, galactosylsphingosine (psychosine), was found in the brain of the twitcher mouse, a mutant caused by genetic deficiency of galactosylceramidase. Similar abnormal accumulation was also found in the brain of the genetic galactosylceramidase deficiency disease in the dog and in human patients (globoid cell leukodystrophy or Krabbe disease). Galactosylphingosine was absent in the brains of normal and heterozygous mice. The finding provides support for the psychosine hypothesis as the biochemical pathogenetic mechanism of globoid cell leukodystrophy. Analogous mechanisms may be important in the pathogenesis of other genetic lysosomal diseases.

Galactosylceramides and their sulphates are the main constituents of myelin sheath of the nerve cell. Two genetically determined disorders are the results of an inability to enzymatically hydrolyse these glycolipids. Thus the deficiency of galactosylceramide beta-galactosidase results in globoid cell leucodystrophy and the reduced activity of enzyme, arylsulphatase A is responsible for the disease Metachromatic leucodystrophy. Both these disorders are fatal and are characterized by marked demyelination and severe mental retardation. Since homognenous enzyme preparations of galactosylceramide beta-galactosidase and arylsulphatase A are now available, a possibility of enzyme replacement therapy in globoid and metachromatic leucodystrophies has been discussed.

Tetracosanoic-(15,16-3H) acid was used to form lignoceroyl psychosine (kerasin). Emulsion of the labelled sphingolipid was injected directly into the brain of adult rabbits and the resultant metabolic products were examined for radioactivity over a period of 21 days. In six animals the precursor was administered after application of cold on the exposed dura. This procedure was found to be an easy practical method for inducing local brain necrosis which resembled histologically and biochemically sudanophilic myelin breakdown occurring in some demyelinating diseases. Small amounts of the labelled material were used in an autolysis series and in an in-vitro study with brain homogenate.

Abnormal accumulation of lysosulfatide (sulfogalactosylsphingosine) was evident in autopsied tissues from a boy with late-infantile metachromatic leukodystrophy. The concentration was high in the cerebral white matter, spinal cord and sciatic nerve (116-787 pmol/mg protein) and low in the cerebral gray matter, kidney and liver (4-40 pmol/mg protein). As is the case with galactosylsphingosine, lysosulfatide inhibited cytochrome c oxidase activity, in a dose-dependent manner. Judging from the tissue distribution of the accumulated lysosulfatide and because of the cytotoxicity, the lysosulfatide presumably explains the demyelination seen in the nervous tissues of patients with metachromatic leukodystrophy.

Jungalwala et al. (1975) described a very sensitive method for the separation and quantitation of phospholipids containing primary amino groups such as phosphatidylethanolamine (PE), phosphatidylserine (PS) and lysophosphatidylethanolamine by high performance liquid chromatography. Compounds which contain primary amino groups can be converted into their U.V. absorbing N-biphenylcarbonyl derivatives. Based on this property, we developed a high performance liquid Chromatographic method for the analysis of brain psychosine, with the use of U.V. detection at 280 nm.

Galactosylsphingosine (psychosine) in somatic organs from a patient with globoid cell leukodystrophy and from the twitcher mouse, an animal model of human globoid cell leukodystrophy was assayed. There was an abnormal accumulation of galactosylsphingosine as in nervous tissues, albeit the concentrations being lower than those in nervous tissues. Galactosylsphingosine accumulation in the kidney of the twitcher mouse increased with age. These findings indicate that globoid cell leukodystrophy is a generalized galactosylsphingosine storage disease.

Glucosylsphingosine has been isolated for the first time as a natural constituent from Gaucher's spleen. On thin-layer chromatography, it migrates with authentic glucosylsphingosine, yielding a positive color reaction with ninhydrin for the amino group and with alpha-naphthol-sulfuric acid for the carbohydrate residue. N-Acylation with palmitic acid gave rise to glucosylceramide, which was cleaved by purified glucosylceramide: beta-glucosidase to ceramide. Gas-liquid chromatography of the trimethylsilyl derivative showed a retention time similar to authentic glucosylsphingosine. Gas-liquid chromatographic analysis of the trimethylsilyl derivatives after methanolysis revealed the presence of only glucose and C(18)-sphingosine. Mass spectral data further supported the structural identity with glucosylsphingosine.

A radioimmunoassay was evolved for detection of antibodies reacting with lactosylsphingosine in the serum. Positive results were obtained in cases of digestive tract neoplasms, cervical carcinoma and burns.

Saponification of cerebroside sulfate (sulfatide) by refluxing with 1 N KOH in 90% n-butanol for 1 h yielded ceramide, sphingosine, lysosulfatide (psychosine-3'-sulfate ester) and a hitherto unknown compound. The latter compound was identified as 3,6-anhydrogalactosyl sphingosine (3',6'-anhydropsychosine) from its mass spectrum. The structure of lysosulfatide was confirmed by reacylating it to sulfatide by condensing it with lignoceroyl chloride. The resulting sulfatide, which was chromatographically identical to control sulfatides, was not oxidized by periodate. The sulfatide was also permethylated and methanolyzed. The sugar moiety obtained was identified as methyl 2,4,6-tri-O-methylgalactoside by gas-liquid chromatography and thin-layer chromatography. The presence of the sulfate group in lysosulfatide was further confirmed by IR spectroscopy and the presence of radioactivity when it was prepared from [35S]sulfatide. The effect of the sulfate group on cleavage of the galactoside linkage and on the formation of the 3,6-anhydro derivative is discussed.

When entrapped into liposomes composed of phosphatidylcholine and other lipids, β-galactosidase (β- d -galactoside galactohydrolase, EC 3.2.1.23) purified from Aspergillus oryzae could cleave the β-galactosidic bond of the terminal galactose of galactocerebroside and GM1-ganglioside (II3NeuAc-GgOse4Cer, galactosyl-N-acetylgalactosaminyl-(N-acetylneuraminosyl)-galactosylglucosylceramide), while the free enzyme could not. The products of the hydrolysis of galactocerebroside were found to be β-galactose and ceramide, which was confirmed by using a fluorescent analog of galactocerebroside, 1-O-galactosyl-2-N-(1-dimethylaminonaphthalene-5-sulfonyl)-sphingosine, as substrate. The formation of GM2-ganglioside (II3NeuAc-GgOse3Cer, N-acetylgalactosaminyl-(N-acetylneuraminosyl)-galactosylglucosylceramide) by the hydrolysis of GM1-ganglioside was also demonstrated. The lipid composition of the liposomes influenced the amount of the enzyme entrapped and the activity of the trapped enzyme. A large amount of the enzyme was entrapped into the liposomes composed of phosphatidylcholine-cholesterol-stearoylamine (molar ratio, 7:2:1). The enzyme trapped in the liposomes and that in those of phosphatidylcholine-cholesterol-sulfatide (molar ratio, 7:2:1) had higher activity on galactocerebroside and GM1-ganglioside than that in other liposomes. The activity of β-galactosidase trapped in liposomes was increased in the presence of detergent, while that of the free enzyme was not changed. By a similar procedure to introduce enzymes into hydrophobic environments, enzymes other than β-galactosidase might come to possess different substrate specificities.

In 1916, the Danish neurologist, Knud Krabbe (9), delineated a new type of infantile familial diffuse brain sclerosis and gave the first description of the characteristic giant cells, which are the histological hallmark of the disease. During the 60s, cumulative analytical and experimental findings resulted in the concept of Krabbe disease as galactosylceramidosis, a concept corroborated by the demonstration of a generalized deficiency of galactosylceramide β-galactosidase (21). But further investigations disclosed also additional enzymic deficiencies, involving the degradation of galactosylsphingosine (13), monogalactosyl diglyceride (37) and lactosylceramide (38), all substrates with a terminal β-galactosidic linkage. Though this does not imply the assumption of a multiple enzyme deficiency, since a single enzyme seems to catalyze the hydrolysis of galactosylceramide and of the three other substrates (12, 14, 32, 37), it does open up new approaches in our search for a better comprehension of the pathophysiology of Krabbe disease. The impaired degradation of galactosylceramide appears to be directly involved in the formation of the globoid cells (1, 15, 20), but it cannot by itself explain other characteristics of the disease, such as the lack of demonstrable accumulation of galactosylceramide in extra-neural organs or in brain outside the globoid cells.

Psychosine, sphingosylphosphorylcholine (52–104 μM), and other glycosphingolipids stimulate mitochondrial respiration (up to 500%) and inhibit oxidative phosphorylation to varying degrees. Above 104 μM these functions as well as uptake of Ca2+ are prevented. At 104 μM sphingosylphosphorylcholine inhibits the mitochondrial ATPase reaction in submitochondrial particles by 48%. Both sphingosylphosphorylcholine and psychosine enhance the active phosphate-dependent swelling of mitochondria. Passive swelling occurs in the presence of rotenone (when swelling does not normally occur) and under hypotonic conditions. A direct interaction of sphingosylphosphorylcholine with membranes is demonstrated by a discharge of the proton gradient across mitochondrial membranes, hemolysis of red blood cells, and binding to inner and outer mitochondrial membranes. Thus lysosphingolipids bind strongly to mitochondrial membranes and markedly alter mitochondrial function. This alteration would affect the ATP levels, thereby altering a wide range of ATP-dependent cellular functions. These results offer a partial explanation for the pathogenesis of representative lysosomal storage diseases.

Glucosylsphingosine is a potent inhibitor of lysosomal glucocerebrosidase and the broad-specificity, cytosolic beta-glucosidase of human liver. In the present study, it was demonstrated that the broad-specificity beta-glucosidase was also inhibited by galactosylsphingosine. The inhibition was observed when the enzyme was assayed for beta-glucosidase, beta-galactosidase, beta-xylosidase, and alpha-arabinosidase activities. Inhibition was of the mixed-type and the degree of inhibition depended on the substrate. For example, galactosylsphingosine was a more potent inhibitor of beta-glucosidase activity (I0.5 = 0.3 mM) than beta-xylosidase activity (I0.5 = 1.2 mM). In addition, the observation that the broad-specificity, cytosolic beta-glucosidase was inhibited by hydrophobic glycosphingolipids prompted the definition of a revised purification procedure which took advantage of hydrophobic affinity chromatography. This revised purification scheme employed Octyl-Sepharose and yielded the largest (68,000 Da) and most active (470,000 nmol h-1 mg protein-1) beta-glucosidase preparation yet described. The beta-glucosidase preparation contained 19% serine and 17% glycine, while 24% of the total amino acids were hydrophobic. The results of this study document the presence of a sphingolipid binding site on the broad-specificity beta-glucosidase. The implications of galactosylsphingosine inhibition of cytosolic beta-glucosidase and the possible role of the enzyme in glycosphingolipid metabolism are discussed.

1. Arylsulphatase A (arylsulphate sulphohydrolase E.N. 3.1.6.1) has been purified 7200-fold from human placenta using concanavalin A Sepharose chromatography. 2. Ultracentrifugation studies indicated that the purified enzyme was homogenous with respect to sedimentation coefficient and molecular weight and has a molecular weight of 102000. 3. The purified enzyme could hydrolyze cerebroside 3-sulphate, seminolipid and sulphogalactosylsphingosine under identical conditions. 4. The kinetic parameters for the hydrolysis of all sulphate esters used in the present study were the same. 5. Both seminolipid and sulphogalactosylsphingosine were competitive inhibitors for the hydrolysis of cerebroside-3-sulphate with an inhibition constant of 0.2 mM. 6. Kinetic parameters, metal ion effect and heat inactivation profile of enzyme suggest that the same active site of enzyme is responsible for the hydrolysis of cerebroside 3-sulphate, seminolipid and sulphogalactosylsphingosine.

Possible use of the fluorescently labelled cerebroside, 1-O-(beta-D-galactosyl)-2-N-[6-(2-antroyl)hexanoyl]-4-sphingeni n (Gal-A-sphingenin) as a substrate for galactosylceramidase (GC) from human skin fibroblasts was investigated. Studies involving TLC and fluorimetric methods revealed enzymatic splitting of Gal-A-sphingenin whose degree correlated with the amount of the enzyme and incubation time. Some kinetic parameters of GC were determined, using Gal-A-sphingenin as substrate. It was shown that the values of specific activity of GC (1.6 nmol/mg protein/hr) and Km (0.025 mM) for Gal-A-sphingenin agree well with the corresponding values obtained with the use of the galactocerebroside as a natural substrate. In experiments with mixtures of Gal-A-sphingening and galactocerebroside used as substrates in different molar ratios, it was shown that the enzyme splits each of the substrates with equal velocity. The results of experiments with the enzyme samples from skin fibroblasts of healthy individuals and of patients with GM1-gangliosidosis (GM1-beta-D-galactosidase deficiency) suggest that Gal-A-sphingenin is a specific substrate for GC.

Lysosphingolipids potently and reversibly inhibited protein kinase C activity and binding of phorbol dibutyrate in vitro and in human platelets. As with activation of protein kinase C by phosphatidylserine and sn-1,2-diacylglycerol, inhibition was subject to surface dilution. Accordingly, inhibition in mixed micelle assays was dependent on the molar percentage of lysosphingolipids rather than the bulk concentration. Lysosphingolipids inhibited protein kinase C activity at molar percentages similar to those required for activation by phosphatidylserine and sn-1,2-diacylglycerol. Since lysosphingolipids accumulate in Krabbe's disease, Gaucher's disease, and other sphingolipidoses, the hypothesis that lysosphingolipid inhibition of protein kinase C represents the missing functional link between the accumulation of sphingolipids and the pathogenesis of these disorders appears to unify existing data. The accumulation of lysosphingolipids would cause progressive dysfunction of signal transduction mechanisms vital for neural transmission, differentiation, development, and proliferation and would eventually lead to cell death.

3-Keto derivatives were prepared in good yield by the oxidative procedure with 2,3-dichloro-5,6-dicyanobenzoquinone from N -acetyl sphingosine, N -palmitoyl sphingosine, N -lignoceroyl sphingosine, and N -lignoceroyl psychosine. None of these 3-keto derivatives, except the one from N -acetyl sphingosine, have been previously reported. Ceramides were isolated from a calf brain and reacted with 2,3-dichloro-5, 6-dicyanobenzoquinone. Ceramides containing sphingosine (4-sphingenine) were converted to 3-keto derivative, while those containing dihydrosphingosine (sphinganine) remained intact under these conditions. The 3-keto ceramides were then separated from the ceramides containing dihydrosphingosine by preparative thin layer chromatography. Similarly cerebrosides from the same calf brain were oxidized and fractionated to 3-ketocerebrosides (from cerebrosides containing sphingosine) and unreacted cerebrosides (cerebrosides containing dihydrosphingosine). The fatty acid composition of these four sphingolipids were determined. Both the ceramides and the cerebrosides containing sphingosine had more unsaturated fatty acids than the corresponding dihydrosphingosine-containing compounds. The ratio of C 16 -C 20 fatty acids to C 22 -C 26 acids was higher in the ceramides containing sphingosine than in ceramides containing dihydrosphingosine, while the ratio was reversed in cerebrosides. The possible precursor-product relationship among these lipids is discussed.