'- 'vvv I'lh'OVIV'vvv 00.0. 1.15. 1-..»- 0‘... . . ..OO. ...—"'1. :pv‘ -- ~00..-«‘-'.. 01", ‘P'o‘u‘lVO .OI~.-O- — cor. cooa"‘1"""".: vVU II . I W . ... . . . .... ..h .b. m .s .. - _. . .. . . y. g .. mm M R N . . _ . E . _ r I 5 .. m A. r ...... . R ..n _ N . E R ...... N N _ 0 .. N x... K . _ N . 2w ......C . . i M _ 0 . . S . I . E \\ . . , . . . ,. . . . .,....—.. .. . . . . N ..,. . 4.. ..... ....c. .. .. . .. .... . .. ..- .V . .. . .....T... . . . ... . . .. ..‘ . . ... .. A ... . .u .... .. o . . .. J... . ....~.(.r.........¢....:......... #24: .HVV'II... ....4115; '9 .u 0.. t 0:... . .., . . . . .- .. ., . . . _. . -. ,_ .. . .. .. , a...i.H.,.u.._.._._m.,..._.,..,v,.£..nw. £9. wafirvm..%fikq. 31115518 / vi ABSTRACT THE ISOLATION OF CERAMIDE TRIHEXOSIDASES FROM NORMAL PLASMA BY Kenneth J. Dean Cermide trihexosidases were isolated from ficin (fig latex) and human plasma (Cohn fraction IV-l). The ceramide trihexosidases in Cohn fraction IV—l were recovered by butanol extraction. It was found that the ceramide trihexosidases in butanol were only soluble in water to the extent that butanol was soluble in water, suggesting that butanolceramide trihexosidase miceles were present in the aqueous solutions. The enzyme activity exhibited a bimodal pH Optimum, with maximum hydrolysis of trihexosyl ceramide at pH 4.6 and 6.5. The enzyme was purified further by 60% acetone preci- pitation and affinity chromatography on melibiose phenylhydrazone bound to dextran-coated glass beads. Five enzymatically active proteins were isolated by affinity chromatography, two of which were active at pH 4.5, and three that were active at pH 6.5. No activity toward the artificial substrate p-nitrophenyl-a-Q-galactopyranoside Kenneth J. Dean was detected in eluate from the affinity column, or at any point in the isolation prior to affinity chromatography, suggesting that these a-galactosidases.were not soluble in butanol. The retention of the ceramide trihexosidases by the melibiose phenylhydrazone affinity adsorbent was found to be pH dependent, and the enzymes were eluted comparably with borate buffer, pH 8.5, or Tris-HCl, pH 8.5. THE ISOLATION OF CERAMIDE TRIHEXOSIDASES FROM NORMAL PLASMA BY Kenneth J. Dean A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Master of Science Department of Biochemistry 1975 to Rebecca Trithart Dean ii ACKNOWLEDGMENTS I would like to express my sincere appreciation to Dr. Charles C. Sweeley, my teacher and advisor, for his assistance and guidance throughout the course of my studies. His concern for the professional development of this graduate student was particularly encouraging. I would also like to express my appreciation to Drs. John F. Holland, Bader Siddiqui, Subroto Chatterjee, and to Clifford Wong, Stephen Gates and Sun-sang Sung for their helpful suggestions and valuable discussions. This research was supported by the National Institutes of Health. iii TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . . . . . . . . . LITERATURE REVIEW . . . . . . . . . . . . . . . . . Discovery and Characterization of the Neutral Glycosphingolipids The Isolation and Characterization of Glycosphingolipid Hydrolases . . . . . . . . Galactocerebrosidase . . . . . . . . . . . . Glucocerebrosidase . . . . . . . . . . . . . Lactosyl Ceramide B-Galactosidase . . . . . Ceramide Trihexosidase . . . . . . . . . . . B-EfAcetylhexosaminidase . . . . . . . . . . The Sphingolipidoses . . . . . . . . . . . . . . Sandhoff's Disease . . . . . . . . . . . . . Fabry's Disease . . . . . . . . . . . . . . Gaucher's Disease . . . . . . . . . . . . . Krabbe's Disease . . . . . . . . . . . . . . Enzyme Replacement Therapy . . . . . . . . . . . MATERIALS AND METHODS . . . . . . . . . . . . . . . MATERIALS . . . . . . . . . . . . . . . . . . . METHODS . . . . . . . . . . . . . . . . . . . . Purification of a-Galactosidases from Ficin Preparation of Affinity Column Adsorbents . Affinity Chromatography . . . . . . . . . . Assays with Ficin . . . . . . . . . . . . . Purification of a-Galactosidases from Cohn Fraction IV-l . . . . . . . . . . . . . Assays with Cohn Fraction IV—l . . . . . . . RESULTS . . . . . . . . . . . . . . . . . . . . . . Preparation of Affinity Adsorbents . . . . . . . Purification of Ficin a-Galactosidases . . . . . iv Page vi vii (‘0 57 57 60 60 60 66 68 71 72 76 76 85 Page Purification of Ceramide Trihexosidases from Cohn Fraction IV—l . . . . . . . . . . . . 105 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . 121 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . 129 LIST OF TABLES Table Page 1. Recovery of Sugars from Acid Methanolysis of Melibiose Phenylhydrazone Glass Beads . . 83 2. Affinity Chromatography of Ficin . . . . . . . 98 3. Purification of Ceramide Trihexosidases from Cohn Fraction IV-l . . . . . . . . . . 120 vi Figure l. 10. ll. 12. 13. 14. 15. LIST OF FIGURES The Current Concept of Glycosphingolipid Metabolism and Enzyme Defects Resulting in Sphingolipidoses . . . . . . . . . . . . . ’- Glycosphingolipids which Accumulate in Fabry's Disease . . . . . . . . . . . . . . . . . . . Steps in the Preparation of Melibiose Phenyl- hydrazone Glass Beads . . . . . . . . . . . . Standard Plot of Copper Sulphate Concentra- tion 0 O O O O C O O O O O O O O O O O O O O Gas-Liquid Chromatography of Trimethylsilyl Methyl Glycosides from Glass Beads . . . . . Standard Plot of Protease Assay . . . . . . . . Standard Plot of a-Galactosidase Assay Using p-Nitrophenyl-a-g-Galactopyranoside . . . . . Standard Plot of B-Galactosidase Assay Using p-Nitrophenyl-B-g-Galactopyranoside . . . . . Standard Plot of Ceramide Trihexosidase Assay Using Trihexosyl Ceramide . . . . . . . . . . Thin Layer Chromatography Plate of Ficin Ceramide Trihexosidase Assay . . . . . . . . Affinity Chromatography of Ficin . . . . . . . Sodium Dodecyl Sulphate Polyacrylamide Gel Electrophoresis of Ficin . . . . . . . . . . Chromatography of Ficin on Acetaldehyde Phenyl- hydrazone Glass Beads . . . . . . . . . . . . Standard Plot of Ceramide Trihexosidase Assay O O O O O O O O O O O O O O O O O O O O The pH Optima for Ceramide Trihexosidase . . . vii Page 29 35 61 79 81 86 88 90 92 99 96 101 103 106 109 Figure Page 16. Thin Layer Chromatography Plate of Cohn Fraction IV-l Ceramide Trihexosidase ACtiVity o o o o o o o o o o o o o o o o o o 111 17. Thin Layer Chromatography and Strip Scan of 3H-Labelled Trihexosyl Ceramide 113 18. Gas-Liquid Chromatography of Silylated Methyl Glycosides from 3H-Labelled Trihexosyl Ceramide . . . . . . . . . . . . . . . . . . 115 19. Affinity Chromatography of Cohn Fraction IV‘]. 0 o o o o o o o o o o o o o o o o o o 118 viii INTRODUCTION AND LITERATURE REVIEW The study of sphingolipids has become an important area of research in the hundred years since Thudichum's first report of their isolation from the supposed single constituent of the brain, protagon. Sphingolipids have been found in all organs of the body. Members of this class of lipids have been characterized as blood group substances, cell surface antigens and mediators of intercellular responses and have been found to accumulate in several genetic diseases classified as hereditary errors of lipid metabolism. The investigation of sphingolipids and sphingolipid metabolism can be subdivided into the discovery and characterization of sphingolipids, the isolation and characterization of sphingolipid hydrolases, the character- ization of the several heritable Sphingolipidoses as sphingolipid hydrolase or transferase deficiencies, and attempts to treat the Sphingolipidoses by enzyme replacement therapy. This is by no means a complete review of the chemistry and metabolism of sphingolipids. Primary emphasis will be placed on the characterization and metabolism of the major plasma glycosphingolipids in man. DISCOVERY AND CHARACTERIZATION OF THE NEUTRAL GLYCOSPHINGOLIPIDS Thudichum reported the discovery of sphingolipids in brain in 1874 (1). Using alcohol-ether preparations of brain, Thudichum discovered and named sphingomyelin, sphingosine, aesthesine (a ceramide), phrenosine (a cerebroside), cerasine (a cerebroside), and psychosine (galactosyl sphingosine). A summary of his work on the isolation and characterization of these lipids appeared in 1901 (2). Thudichum's formulation of sphingosine was not correct, but it was not until 1947 that Carter and his co- workers proposed the correct structure for sphingosine (3). The isolation and characterization of the individual lipids were continued by Thierfelder, Klenk, Levine and Rosenheim (4-6), and the discovery of a new class of acidic glycosphingolipids, gangliosides, was reported by Klenk in 1935 (6). Structural determinations of the neutral glycosphingolipids followed soon after Carter's elucidation of the structure of sphingosine. In 1950, Nakayama concluded from methylation studies that the glycosidic bond of galactocerebroside was between the primary hydroxyl group of Sphingosine and the anomeric carbon of galactose (7). These findings were verified by Carter and Greenwood by acetylation and platinum-catalyzed reduction of galactocerebroside (8). Further studies by Fujino and Negishi in 1956 using B-galactosidase, revealed that this linkage was of the B-configuration (9). In 1958, Rosenberg and Chargaff, using infrared spectroscopy, determined that the glycosidic linkage of gluccerebroside isolated from Gaucher spleen was also of the B-configuration (10). Carter et_al. confirmed these findings by comparing the products of chemical degradation of cerasine (galactocerebroside) with those of glucocerebroside from Gaucher spleen. It was concluded that the lipid moieties of cerasine and Gaucher cerebroside were identical, and that these compounds differ only in ther sugar constituents (11). In 1942, Klenk and Rennkamp (12) isolated a cerebroside with two hexose residues, glucose and galactose, from bovine spleen. Rapport gt_§l. isolated a glycolipid with haptenic properties from human carcinoma tissue, cytolipin H, which was also found to contain glucose, galactose, fatty acid and sphingosine in equimolar amounts (13-15). In 1961, Rapport and coworkers, using hapten inhibition of cytolipin H compliment fixation by B-lactose, demonstrated that the disaccharide moiety of cytolipin H was B-g-galactosyl- glucosyl (16). Further hapten inhibition studies and chemical analyses suggested that cytolipin H and ox spleen cytoside were identical except for the fatty acid residues of the ceramide group (17). Makita and Yamakawa verified these findings in 1962 using silicic acid and Florisil chromato- graphy to isolate glycosphingolipids from human, equine and bovine spleens. Ceramide dihexoside from the various sources was found to be identical except for variations in the fatty acid constituent (18). Further studies on the structure of ceramide dihexoside from human erythrocytes by Yamakawa et al., using permethylation and gas-liquid chromatography revealed the linkage of galactose and glucose to be 1+4. Infrared spectroscopy confirmed the previous report that these sugars were linked in the B-configuration (19). In the course of examining mammalian spleen glycos- phingolipids, Makita and Yamakawa discovered a trihexosyl ceramide in bovine spleen consisting of galactose and glucose. Hydrolysis revealed the galactose-reducing spot to be more intense than glucose, and examination by thin-layer chromatography showed that trihexosyl ceramide migrated more slowly than lactosyl ceramide (18). The following year Svennerholm and Svennerholm isolated and quantitated the distribution of cerebrosides, lactosyl ceramide, ceramide trihexoside, and ceramide trihexoside-Efacetyl-hexosamines in human serum, liver, and spleen. They found the ratio of galactose to glucose in ceramide trihexoside to be 2:1; and using mild acid hydrolysis were able to isolate galactose; ceramide monohexoside and hihexoside had RF values on thin- layer chromatography that were identical with those of cerebroside and lactosyl ceramide, respectively (20). Sweeley and Klionsky confirmed these findings using ceramide trihexoside isolated from Fabry kidney (21). A previously unknown ceramide dihexoside (designated GL—2 by these workers), having the same RF value on thin layer chromatography as lactosyl ceramide, was also found in Fabry kidney. However, the carbohydrate moiety of GL-2 was found to consist entirely of galactose. Further studies by these investigators revealed the structure of ceramide trihexoside to be galactosyl-(l+4)- galactosyl-(l+4)-glucosyl-(1+1)—ceramide (22). The structure of GL-2 was found to be galactosyl-(l+4)-galac- tosyl-(l+1)-ceramide (23). However, the anomeric configur- ation of the glycosidic bonds was not unequivocally deter- mined until 1971. Hakomori gt_al., using nuclear magnetic resonance spectroscopy, infrared spectroscopy, and a- galactosidase from ficin (fig latex) and B-galactosidase from jack bean determined that the terminal galactosyl linkage was in the a—configuration and the internal galactosyl-glucosyl linkage was in the B-configuration (24). Digalactosyl ceramide was also found to contain a terminal a—linkage and an internal galactosyl-ceramide B-linkage using these enzymes (25). The next higher homolog in this series, ceramide tetra- hexoside, was discovered in 1951 by Klenk and Lauenstein (26). 'These workers reported the presence of a glycolipid in human blood containing galactosamine residues. Further investiga- tions were conducted by Yamakawa and Suzuki, who named this glycolipid globoside and determined its composition to be ceramide y-acetylgalactosaminyl trihexoside (27). Suzuki and Iida found that globoside exhibited haptenic properties with human erythrocytes of blood groups A and B (28). Structural studies conducted by Yamakawa and coworkers using erythrocyte globoside led them to propose the structure: E—acetylgalacto— saminyl—(B l+6)-galactosy1-(l+4)—galactosyl-(l+4)-g1ucosy1- (1+l)-ceramide (29). In 1963 a glycolipid hapten with composition similar to erythrocyte globoside was isolated from human kidney by Rapport gt_§1., who named this compound cytolipin K (30). Further investigations of human kidney glycolipids by Makita lead to the isolation of a hexosamine- containing glycolipid with the same composition as human erythrocyte globoside (31). The structure of this glycos- phingolipid was found to correspond to the structure of human erythrocyte globoside, except for the terminal glycosidic linkage, which was found to be geacetylgalacto- saminyl (81+3) galactosyl rather than (81+6) as originally reported (32). The immunological identity of cytolipin K with human kidney and erythrocyte globoside was determined by Rapport and Graf using double diffusion immunoprecipi- tation (Ouchterlony method) (33). The anomeric linkages of the internal glycosidic bonds were determined by Hakomori gt_al. and were found to be Efacetylgalactosaminyl-(Bl+3)- galactosyl-(al+4)-galactosyl-(Bl+4)-glucosyl-(Bl+l)-ceramide (24). Higher homologs in this series of neutral glycosphingo- lipids are not known in man at the present time. However, a ceramide pentahexoside called the Forsman hapten (34) has been found in other mammalian species, as it has not been detected in man, it will not be further described. THE ISOLATION AND CHARACTERIZATION OF THE GLYCOSPHINGOLIPID HY DROLASES Studies on the discovery and characterization of the glycolipid hydrolases can be grouped into three categories: the discovery and isolation of the enzymes; the characteri- zation of the enzymes with respect to activators, pH optimum, thermostability and as glycoproteins; and the discovery of enzymes other than the neutral glycosphingolipid hydrolases will be limited to their roles in the hereditary sphingoli- pidoses and will be reserved for the discussion of those disorders. GALACTOCEREBROSIDASE The first report of glycophingolipid hydrolase activity appeared in 1936. Thannhauser and Reichel reported the ability of crude brain and spleen preparations to cleave galactose from cerebron (galactosylceramide) when stimulated with sulfhydryl reagents (cysteine, hydrogen sulfide or glutathione) (35). In 1965, Kopaczyk and Radin examined the metabolism in vivo of 14C-lignoceroyl psychosine (galactocerebroside) in rat brain (36). Two days following injection of 14C-galactocerebroside, ceramide was found to contain the highest specific activity of all lipids except for galactocerebroside. Free ceramide underwent rapid hydrolysis in the brain, with label appearing in ester linkages and gangliosides. These findings led to the con- clusion that the initial step in the catabolism of galactocerebroside is cleavage of the glycosidic bond by B-galactosidase to yeild ceramide, which undergoes further hydrolysis. Further studies by Hajra et a1. led to the 8 isolation of galactocerebrosidase from rat and porcine brain (37). It was found that hydrolysis of cerebroside could be stimulated by cholate or taurocholate, and optimal activity was exhibited at pH 4.5. The enzyme preparation hydrolyzed both galactocerebroside and the water soluble artificial substrate, o-nitrophenyl-B-Q-galactopyranoside. However, during the purification, the activity toward the artificial substrate decreased, while the activity toward cerebroside increased. These investigators were able to separate these B-galactosidase activities by column chromatography. In 1970, Suzuki and Suzuki (38) examined brain, liver, and spleen from patients with globoid cell leukodystrophy (Krabbe's disease) to determine the nature of the enzymatic defect. These workers found normal B-galactosidase activity toward p- nitrophenyl-B-g-galactopyranoside in these tissues while galactocerebroside B-galactosidase activity was 5-10% of normal. These findings were confirmed by Austin et_al. (39) who found normal B-galactosidase activity toward o-nitrophenyl- B-g—galactopyranoside in brain, liver and kidney from patients with Krabbe's disease. Galactocerebrosidase activity was markedly deficient in tissues from patients with Krabbe's disease. These findings indicate that there are two specific B-galactosidases, an enzyme specific for galactosylceramide, and another B-galactosidase active toward the water-soluble substrate o-nitrophenyl-B-g-galactopyranoside. Further characterization of rat brain galactocerebro- sidase by Bowen and Radin revealed that galactocerebrosidase 9 is a lysosomal enzyme. The addition of cholate was found to be necessary during purification to keep the enzyme de- aggregated. The addition of Triton X-100 to enzyme solutions was not completely effective in preventing the enzyme from aggregating (40, 41). Recently, Suzuki and Suzuki (42) characterized B- galactosidases active toward 4-methylumbelliferyl-B-Q— galactopyranoside were obtained by each of these fraction- ation procedures. Two bands of galactocerebrosidase activity were obtained by electrofocusing, while the enzyme activity was eluted in a single band from gel filtration on Sephadex G-200. These findings suggest that multiple forms of each of these enzymes are present in normal human liver. Norden et_al. found that GMl B-galactosidase from human liver were retained on Conconavalin A-Sepharose chromatography and could be eluted with a-methylmannoside (43). These findings imply that these enzymes are glycoproteins. Further characterization of galactocerebroside B-galactosidase with respect to carbohydrate and amino acid composition, subunit, and active site await the purification of sufficient quantities of the enzyme. GLUCOCEREBROSIDASE In 1965, Brady et a1. isolated an enzyme from rat and human spleen that cleaved glucocerebroside specifically. The most purified enzyme preparation obtained did not catalyze the hydrolysis of galactocerebroside or the water-soluble artificial substrate o-nitrophenyl-B-gr galactopyranoside. The hydrolysis of p-nitrophenyl-B-Q- 10 glucopyranoside decreased during the purification of the glucocerebrosidase activity, suggesting that these reactions are catalyzed by different enzymes. Optimal hydrolysis of glucocerebroside was found to occur at approximately pH 6.0. The non-ionic detergent Cutscum was used to solubilize the substrate (44). The following year Gatt and Rapport isolated glucocerebrosidase from ox brain (45). It was found that pre- cipitated glucocerebrosidase could be resuspended with 1% sodium cholate. Hydrolysis of glucocerebroside was stimulated by the addition of the non-ionic detergent Triton X-100 or by anionic detergents such as sodium taurocholate or sodium cholate. The pH Optimum for hydrolysis of glucocerebroside was found to be buffer-dependent, with optimal activity at pH 5.0 using acetate buffer, and pH 5.6 using pyridine buffer (46). In 1971, Ho and O'Brien isolated two protein factors from human spleen designated Factor P and Factor C; when combined they exhibited high glucocerebrosidase activity (47). Factor P was characterized as a thermostable, water-soluble acid glycoprotein that was devoid of B-glucosidase activity using either artificial or natural substrates. Factor C was characterized as a particle (or membrane)-bound, thermolabile protein that could be solubilized only in the presence of Triton X-100. Factor C exhibited low levels of B-glucosidase activity toward both 4-methylumbe11iferyl-B-g-glucopyranoside and glucocerebroside. When Factors P and C were combined, 4-methylumbelliferyl-B-Q-glucopyranosidase activity was 11 stimulated 7.4-fold and glucocerebrosidase activity was stimulated 39-fold. Both activities were optimal at pH 4.5 in the presence of Triton X-100, which stimulated activity. Optimal activity was attained when a 1:10 (w:w) ratio of Factor P to Factor C was used (48). The glucocerebrosidase activity observed with the combined factors was 70- to 80- fold higher than levels reported by Brady et_§1, in human spleen (49). Pentchev gt_al. have purified human placental gluco- cerebroside B-glucosidase 4000-fold, yielding a homogeneous preparation as determined by sodium dodecyl sulphate gel electrophoresis (50). The crude glucocerebrosidase was found to be bound to subcellular particulate material, and sodium taurocholate was required to solubilize the enzyme and extract it from this material. Hydrolysis of gluco- cerebroside by placental glucocerebrosidase alone was maximal at pH 6.0. The addition of a heat-stable factor from spleen shifted the pH optimum for hydrolysis of glucocerebroside to approximately pH 5.8. The pH optimum for hydrolysis of 4-methylumbelliferyl-B-g-glucopyranoside was also shifted downward, from pH 6.5 to pH 5.0. The hydrolysis of the artificial substrate was also stimulated 4-fold by addition of the heat-stable factor from spleen. The homogeneous placental glucocerebrosidase had an apparent molecular weight of 240,000 as determined by gel filtration. Glycerol and dithiothreitol stabilized the purified glucocerebrosi- dase. 12 The discrepancy in pH optima reported by Ho and O'Brien for human spleen glucocerebrosidase and Pentchev et_al. for placental glucocerebrosidase may actually indicate the existence of multiple forms of glucocerebrosidase activity rather than differences in methods or errors by these workers. The existence of multiple forms of glycosphingolipid hydro- lases with differing pH optima has been described by Mapes gE_gl. (73) for ceramide trihexosidase (to be discussed). These isozymes were found to differ in sialic acid content and could be interconverted using neuraminidase and sialyl transferase. The interconversion of splenic and placental glucocerebrosidase catalyzed by neuraminidase or sialyl transferase has not been reported. The clarification of this discrepancy awaits further study of these glucocerebrosidase preparations. LACTOSYL CERAMIDE B-GALACTOSIDASE In 1965, Gatt and Rapport reported the hydrolysis of lactosyl ceramide by rat brain (51). The enzymatic activity was recovered from particulate matter sedimenting between 800 and 15,000xg by extraction with 0.5% sodium cholate. The hydrolysis of lactosyl ceramide by B-galactosidase from rat brain required Triton X-100 and sodium cholate or taurocholate as activators. The enzyme preparation did not catalyze the hydrolysis of galactocera- broside or psychosine. Maximal hydrolysis of lactosyl ceramide by this partially purified enzyme preparation occurred at pH 5.0. Further investigations by these workers resulted in the partial purification of B-galactosidase from 13 rat and calf brains (52). The pH optima of these B-galacto- sidase preparations as determined with p-nitrophenyl-B-Q— galactopyranoside, was pH 4.5 for calf brain B-galactosidase and pH 3.1 for rat brain B-galactosidase. The pH optimum with the artificial substrate was approximately 2 pH units lower than the optimum observed using lactosyl ceramide as substrate. This finding was confirmed by Gatt and Rapport in 1966 (53). Maximal hydrolysis of lactosyl ceramide by rat brain B-galactosidase was again found to occur at pH 5.0. Enzymatic activity toward lactosyl ceramide required Triton x-1oo and sodium cholate or taurocholate as activators, whereas activity toward the artificial substrate did not require the addition of detergents. The Michaelis constants for the two enzymatic activities were also different. The Km of lactosyl ceramide B-galacto- sidase was 2.2 x 10'.5 M. The Km of p-nitrophenyl-B-g- galactopyranoside was about 20-fold higher, 4 x 10"4 M. In 1968, Bowen and Radin obtained a B-galactosidase preparation from rat brain that had approximately equal activities toward both galactocerebroside and lactosyl ceramide. It was also found that lactosyl ceramide inhibited hydrolysis of cere- broside. These workers concluded that their enzyme prepar- ation was not pure, and that inhibition of galactocerebro- sidase activity by lactosyl ceramide is the result of competitive attachment of lactosyl ceramide to the active site of the galactocerebrosidase without catalytic activity toward this lipid (54). In further studies, Radin et a1. examined 14 the changes in lactosyl ceramide B-galactosidase activity in rat brain accompanying development (55). These studies indicated that hydrolysis of lactosyl ceramide and p—nitro- phenyl—B-g—galactopyranoside are probably carried out by the same enzyme since changes in the activity toward both of these substrates were nearly identical throughout maturation (monitored from age 5-320 days) in rats. In 1974, Wenger gt_al. found that lactosyl ceramide B-galactosidase activity is deficient in patients with globoid cell leukodystrophy (Krabbe's disease) in addition to the well-documented galactocerebroside B-galactosidase deficiency (56). Activity toward the artificial substrate 4-methylumbelliferyl-B-g—galactopyranoside in brain tissue from patients with Krabbe's disease was normal. This finding does not agree with the previous conclusion by Radin gt_al. (55) that hydrolysis of lactosyl ceramide and the artificial substrate is catalyzed by the same enzyme. Wenger recently characterized lactosyl ceramide B- galactosidase from human brain (57). He found that the thermal denaturation patterns for galactocerebroside B- galactosidase and lactosyl ceramide B-galactosidase and lactosyl ceramide B-galactosidase were identical under all conditions, but significantly different from the heat de- naturation pattern for 4-methy1umbelliferyl-B-Q-galactopyra- noside. The effects of stimulators such as sodium cholate and oleic acid were the same. The substrates were found to be mutually competitive inhibitors, and the activities 15 co-chromatographed on Sephedex G-200 and DEAE cellulose chromatography. Wenger concluded that lactosyl ceramide B-galactosidase and galactocerebroside B-galactosidase may be identical. Suzuki and Suzuki have reported conflicting results for lactosyl ceramide B-galactosidase activity in liver tissue from patients with Krabbe's disease (58, 59). These workers reported normal activity for both lactosyl ceramide and 4- methylumbelliferyl-B-g-galactopyranoside B-galactosidases, and deficiencies of galactocerebroside and psychosine B-galactosidases in liver tissue from patients with Krabbe's disease. Recently, Miyatake and Suzuki obtained a partially purified rat brain B-galactosidase preparation that exhibited B~galactosidase activity toward galactocerebroside, lactosyl ceramide, asialo-GMl-ganglioside, GMl-ganglioside, and 4- methylumbelliferyl-B-g-galactopyranoside (60). Contrary to the report by Wenger (57), B-galactosidase activities toward 4-methylumbelliferyl-B-g-galactopyranoside, lactosyl ceramide, asialo-GMl-ganglioside and GMl-ganglioside all co-chromato- graphed and were purified approximately 50-fold, whereas galactocerebroside B-galactosidase was purified only 3-fold by this method. The behavior of lactosyl ceramide B- galactosidase with inhibitors was different from the behavior of asialo-GMl-ganglioside or GMl—ganglioside B-galactosidases, and lactosyl ceramide was a poor inhibitor of these two gang— 1ioside B-galactosidases. These workers concluded that lactosyl ceramide B-galactosidase is a distinct enzyme from l6 galactocerebroside B-galactosidase, and asialo-GMl-ganglio- side and GMl-ganglioside B-galactosidases. However, they suggested that hydrolysis of the ganglioside substrates may be carried out by the same enzyme. The clarification of this controversy awaits purifica— tion of these enzymes to homogeneity and their further char- acterization. CERAMIDE TRIHEXOSIDASE In 1967, Brady et a1. isolated an enzyme from rat intestinal tissue that catalyzed the hydro- lysis of the terminal galactose residue from galactosyl- galactosylglucosylceramide (61). The crude ceramide tri- hexosidase was particle-bound and could be released from the particulate material by treatment with sodium cholate. Hydrolysis of trihexosyl ceramide by the partially purified enzyme preparation was optimal at pH 5.0. It was found that 2 mg/ml sodium cholate stimulated ceramide trihexosidase activity, but addition of sodium cholate beyond this amount resulted in a decrease in enzymatic activity. The Michaelis constant was found to be 3.7 x 10'4M. In 1971, Beutler and Kuhl attempted to characterize ceramide trihexosidase activity from normal and Fabry leukocytes and fibroblasts using the artificial substrate 4-methylumbelliferyl-a-g- galactopyranoside to monitor activity (62, 63). These workers isolated two isozymes of a-galactosidase from normal human leukocytes and fibroblasts designated Forms A and B. Form A was thermolabile and Form B was thermostable when heated to 45°C for 2 hours. Form A had a lower Km with the 17 fluorogenic substrate than did Form B, and was found to be electrophoretically more mobile than Form B at pH 7.0. These findings were confirmed by Wood and Nadler using 4-methyl- umbelliferyl-a-gegalactopyranoside to monitor a-galactosidase activity in fibroblasts (64). In 1971, Ho gt_al. conducted further investigations using 4-methylumbelliferyl-a-g- galactopyranoside to characterize*the a-galactosidase activity in human fibroblasts and liver (65). Hydrolysis of the artificial substrate was optimal near pH 4.5. The Michaelis constant for the enzyme was found to be 4 x 10"3 M. Electro- phoresis of normal human liver preparations yielded 7 peaks of a-galactosidase activity ranging from pI 3.6-6.8. Ho §E_gl. suggested that the two least acidic protein peaks represented molecular aggregates of the give other protein peaks. The three most acidic peaks co-migrated as Form A on starch gel electrophoresis and the remaining two peaks co- migrated as Form B on starch gel electrophoresis. The a- galactosidase Form A isozyme and one of the Form B isozymes were thermolabile when heated to 50°C for 15 minutes. Treat- ment of the a-galactosidases with neuraminidase resulted in an enzyme preparation that gave a single peak of a-galacto- sidase activity with pI 4.6 and pH optimum at pH 4.35. Beutler and Kuhl attempted to purify d-galactosidases Form A and B from human placenta in 1972 (66). Forms A and B were separated by DEAE cellulose chromatography and the partially purified enzymes were examined by isoelectric focusing. These proteins had pI 4.7 for a-galactosidase A 18 and p1 4.42 for a-galactosidase B. Hydrolysis of the artificial substrate was optimal at pH 4.5 for both isozymes. The Km of a-galactosidase A was 3.4 x 10'3M and the Km of a-galactosidase B was 40.6 x 10'3M, as compared with the Km of 3.7 x 10'3M obtained by Ho et_al. (65) for a-galactosidases from human liver. These findings suggest that liver and placenta may possess different forms of a- galactosidase A and B. In 1973, Beutler gt_al. substantiated this hypothesis by isolating a-galactosidase form human intestinal mucosal scrapings, liver, leukocytes, placenta, fibroblasts, heart, skeletal muscle, kidney and spleen (67). Each tissue sample was examined by starch gel electrophoresis for a-galactosidase isozymes. a-Galactosidases from each tissue exhibited a unique electrophoretic pattern. It was found that a-galactosidase A from all tissues except intes- tinal mucosal scrapings was susceptible to attack by neuraminidase. Neuraminidase-treated a-galactosdiase A differed electrophoretically from Form B. These investiga- tors concluded that a-galactosidase A and B differ by more than neuraminyl substituients. The artificial substrate 4-methylumbelliferyl-a-Q- galactopyranoside was used exclusively in the discovery and characterization of a-galactosidase A and B. No attempts were made to correlate activity toward this synthetic sub- strate analog with the ability of enzyme preparations to hydrolyze the natural substrate, trihexosyl ceramide. In 1972, Mapes and Sweeley isolated a-galactosidases, from human Cohn fraction IV-l (plasma protein), which were 19 specific for either synthetic substrate or trihexosyl ceramide (68). Affinity chromatography of partially purified Cohn fraction a-galactosidases yielded two peaks of ceramide tri- hexosidase activity that were well separated from protein activity toward the synthetic substrates, also eluted from the column. This finding may indicate that the a-galactosi- dases A and B characterized using artificial substrates, but untested for their ability to hydrolyze the natural lipid substrate, trihexosyl ceramide, may not be glycolipid hydro- lases. Johnson and Brady published the partial purification of placental ceramide trihexosidase in 1972 (69). These investi- gators obtained three protein peaks of a-galactosidase activity by carboxymethyl-Sephadex chromatography of the placental preparation. All three a-galactosidase isozymes were active toward both 4-methylumbelliferyl-a-g-galactopy- ranoside and trihexosyl ceramide, although the third peak eluted form the column contained 95% of the ceramide tri- hexosidase activity. The first two peaks contained predomi- nantly 4-methylumbelliferyl-a-g-galactopyranoside a-galacto- sidase activity, and were not considered to be true ceramide trihexosidases. The thermostability of these isozymes, their mobility on starch gel electrophoresis and their isoelectric focusing profiles were not determined. Unfortunately they cannot be compared by these methods to the a-galactosidase A and B isozymes characterized by Beutler and Kuhl from human placenta (66). The Michaelis constants toward trihexosyl 20 ceramide and 4-methylumbelliferyl-a-gfgalactopyranoside for the third a-galactosidase isozyme eluted from the column were 840 x lO-3M and 1900 x 10'3M respectively. These values are much higher than any reported previously. Hydrolysis of tri- hexosyl ceramide was optimal at pH 4.4. In 1973, Ho obtained a partially purified a-galactosi- dase preparation from human liver (70). DEAE-cellulose chromatography of the partially purified liver preparation yielded two peaks of a-galactosidase active toward 4-methyl- umbelliferyl-a-g-galactopyranoside. The first peak contained all of the ceramide trihexosidase activity and 98% of the artificial substrate activity. This fraction was thermola- bile at 50°C, losing 90% of its activity toward both sub- strates in 36 minutes. These findings, the elution profile from DEAE-cellulose and the thermolability of the first peak, are consistent with the results obtained with a-galactosidase A by Beutler and Kuhl (66). Unfortunately, the mobility on starch gel electrophoresis and the electrofocusing profile of the liver preparation were not determined by Ho. A more pre- cise comparison of this liver a-galactosidase preparation with a-galactosidases isolated by Beutler and Kuhl (66, 67) and Mapes and Sweeley (68) will be of considerable importance. Hydrolysis of trihexosyl ceramide by the liver enzyme was optimal at pH 3.5 and activity toward the artificial sub- strate was optimal at pH 4.5. The pH optimum for hydrolysis of the natural lipid substrate is thus much more acidic than the values reported by Johnson and Brady (69) for placental 21 ceramide trihexosidase (pH 4.4, or Mapes g£_al. (71) for human plasma (pH 5.4 and pH 7.2) or Brady et_al, (61) for rat intes- tine (pH 5.0). It is not clear whether the differences in pH optimum indicate differences between isozymes of ceramide tri- hexosidase (each worker uses a different source for ceramide trihexosidase) or differences in assay methods. Each worker employs different detergents to solubilize and stimulate hydrolysis of trihexosyl ceramide. The effects of these detergents in the micro-environment of the micelle may be reflected by different pH optima for hydrolysis of trihexosyl ceramide. The Michaelis constant for the liver ceramide trihexo— sidase was 5 x 1075M and that for 4-methylumbelliferyl-a-g- galactopyranoside was 2 x 10'3M. These values are much lower than the Km values observed for either substrate by other workers. In 1973, Mapes §t_al. reported the purification and characterization of a-galactosidase isozymes from human Cohn fraction IV-l (72-74). Two forms of ceramide trihexosidase seperated by affinity chromatography of Cohn fraction IV-l. Both of these ceramide trihexosidase isozymes were devoid of artificial substrate activity. Both forms were relatively 'thermostable at 50°C as compared to the thermostable loehavior of ceramide trihexosidase from liver. The Edichaelis constant for the first form eluted from the (affinity column was 4.5 x 10’4M and that for the second ceramide trihexosidase isozyme was 5 x 10'4M. Both isozymes 22 were examined by electrofocusing and had pI 3.0. The first ceramide trihexosidase isozyme eluted from the affinity column, active at pH 5.4, was treated with neuraminidase. Electrofocusing of the partially desialylated isozymes yielded a complex profile of protein peaks active at pH 5.4 and 7.2. Several of these forms corresponded to isozymes active at pH 7.2 obtained by affinity chromatography of whole plasma. When the least acidic ceramide trihexosidase isozyme, active at pH 7.2, was treated with a crude porcine sialyl transferase fraction, a complicated profile of several protein bands active at pH 5.4 and 7.2 was obtained on electro- focusing, including a form active at pH 5.4 that corresponded with the most active sialylated form. These workers postula- ted that attack by neuraminidase or addition of sialic acid residues by sialyl transferase could be a mechanism of regulating ceramide trihexosidase activity in vivo. It now appears that there are two general classes of a-galactosidase activity: activity toward the artificial sub- strates 4-methylumbelliferyl-a-g-galactopyranoside and p- nitrophenyl-a-g-galactopyranoside, originally characterized by thermostability and starch gel electrophoresis studies; and activity toward glycosphingolipids. Clearly, some :relationship must exist between these a-galactosidases as «evidenced by deficiencies of both classes of a-galactosidase activity in Fabry's disease (63, 71, 75, 76). Whether this is a precursor-degraded product relationship, or the result of a more complicated subunit-activator mechanism of 23 regulating activity towards a particular substrate is not known. Mapes gt_gl, have demonstrated that isozymes of each class of a-galactosidase exist. The physiological signifi- cance of these isozymes is not known. The different pH optima obtained for ceramide trihexosidase from placenta, plasma, liver and recently kidney (77, 78), and spleen (79) may indicate the possibility of organ-specific production of a particular ceramide trihexosidase isozyme. Such a theory would require that Fabry's disease results from several genetic mutations, rather than a single genetic error. This theory, seems unlikely, but organ—specific activators for ceramide trihexosidase could confer organ-specific charac- teristics, as observed for ceramide trihexosidase, and one gene mutation could cause alteration of all of the forms. B-N-ACETYLHEXOSAMINIDASE In 1936, Watanabe reported the ability of crude rabbit and beef liver preparations to hydro- lyze phenyl-Bffl-acetylglucosaminide. Optimum hydrolysis of the artificial substrate was at pH 3.7 (80 - 83). Other workers reported the partial purification of y-acetyl-B- glucosaminidase from ox testis (84, 85) and rat kidney (86) following Watanabe's discovery. The pH optimum for hydroly- sis of phenyl-B-§facetylglucosaminide by the partially purified rat kidney enzyme was found to be buffer dependent. ‘Optimal hydrolysis of the substrate was observed at pH 4.6 using acetate buffer, and pH 4.0 using citrate buffer. In 1961, Woollen gt_al. examined partially purified ram testis preparations for Efacetyer-glucosaminidase and Efacetylfig- 24 galactosaminidase activities using p-nitrophenyl-Eracetyl—B- g-glucosaminide and -ga1actosaminide as substrates (87). They were unable to separate these enzymes by ammonium sulfate and calcium phosphahagel fraction, or by heat inactivation. The pH optimum for hydrolysis of p-nitrophenyl-Efacetyl-B-g- glucosaminide was at pH 4.3. Optimal E—acetyl-B-galactosa- minidase activity was bimodal, with peaks at pH 4.45 and pH 4.7. The two artificial substrates were found to be mutually competitive inhibitors. Furthermore, when other inhibitors such as 2-acetamido-2-deoxy-gluconolactone or -galactono— lactone were used, the Ki values were almost identical. Woollen §£_gl. suggested that each substrate (or inhibitor) was bound at the same site on the enzyme surface. They postulated that the same enzyme was responsible for both HI acetyl-B-g-glucosaminidase and -galactosaminidase activities. Further studies by these investigators using partially purified rat kidney preparations (88), human pregnancy serum, pig epididymis, ram testis, various rat tissues and viper venom (89) supported the previous finding that these enzyme activities copurified, but did not yield kinetic data supporting a single-site for binding of both substrates. In 1967, Frohwein and Gatt isolated three enzymes from calf lbrain active toward Efacetylhexosamines (90). One enzyme hydrolyzed both B-E-acetylglucosaminides and ~galactosami- nides, and the other two enzymes were specific for either B-Ntacetylglucosaminide or -galactosaminide. The ability of these enzymes to hydrolyze glycosphingolipids was 25 examined (91). It was found that only the nonspecific hexo- saminidase was capable of hydrolyzing glycosphingolipids, with globoside and asialo-GMz-ganglioside hydrolyzed more readily than GMz-ganglioside. The hydrolysis of asialo-GMz-ganglio- side was strongly inhibited by GMz-ganglioside (Tay-Sachs ganglioside), suggesting that these reaction are catalyzed by the same enzyme. In 1968, Robinson and Stirling found that E-acetyl-B- hexosaminidase from human spleen could be separated into two forms on starch gel electrophoresis (92). The negatively charged form was designated hexosaminidase A and the posi- tively charged form was designated hexosaminidase B. It was found that the two forms could also be separated by DEAE cellulose chromatography. Both components A and B showed activity toward both p-nitrophenyl glycosides. Treatment of hexosaminidase A with neuraminidase produced several inter- mediate forms with decreasing anodic mobility on starch gel electrophoresis. If neuraminidase was added in excess the complete conversion of hexosaminidase A to a form resembling hexosaminidase B was observed without the production of intermediate forms. The conversion of hexosaminidase A to a form resembling hexosaminidase B by neuraminidase has been confirmed by Murphy and Craig, using bacterial (93) and human (94) neuraminidases. In 1969, Okada and O'Brien discovered a deficiency of hexosaminidase A in tissues from patients ‘with Tay-Sachs disease (GMZ-gangliosidosis, type 1) (95). Tay-Sachs disease is characterized by the accumulation of 26 GMZ and asialo-GMz-ganglioside in brain. Kolodny et_§l, demonstrated the absence of GMz-ganglioside B-gfacetyl- galactosaminidase in muscle tissue of patients with Tay- Sachs disease (96). These findings suggest that hexo- saminidase A is specific for GMz-ganglioside. This was confirmed by Sandhoff, in 1970 using hexosaminidase A and B from human liver (97). Hexosaminidase A catalyzed the hydrolysis of GMz-ganglioside, but hexosaminidase B did not. Wenger gt_al. examined the substrate specificity of hexo- saminidase A and B from human liver in 1972 (98). They found that asialo-GMz-ganglioside was hydrolyzed by both hexosaminidase A and B and suggested that elevated levels of asialo-GMz-ganglioside in Tay-Sachs disease may be due to competitive inhibition of hexosaminidase B by accumulating GMz-ganglioside. The Km of globoside for hexosaminidase B was approximately one-third the Km for hexosaminidase A. Considering the Michaelis constant to be an indication of substrate affinity, it appears likely that hexosaminidase B is responsible for hydrolysis of globoside in vivo. This conclusion is supported by the failure of patients with Tay— Sachs disease to accumulate globoside, whereas patients with Sandhoff's disease (hexosaminidase A and B deficiency) accumulate globoside as well as GMZ and asialo-GMZ ganglio- side. In 1974, Srivastava gt_§l. purified hexosaminidase A (and B from human placenta, performed structural studies and jprepared antiserum to each isozyme (99-101). These workers found cross-reactivity of antibodies against hexosaminidase 27 A and B, indicating a close similarity between hexosaminidase A and B, or the existence of a common subunit. Structural studies revealed that hexosaminidase B consists of only one subunit of molecular weight approximately 17,600. Hexo- saminidase A was found to consist of one major subunit of molecular weight approximately 17,600, and two minor subunits of molecular weight 17,000-18,000. The molecular weight of the holoenzymes were about 100,000. These findings indicate that hexosaminidase A and B differ by more than sialic acid residues, as suggested earlier. Support for this concept was obtained with the discovery by Li gt_§l. of a heat stable glycoprotein from human liver, of molecular weight approxi- mately 22,000, which was necessary for hydrolysis of GMZ- ganglioside by hexosaminidase A from liver and urine (102, 103). The presence of this activator greatly stimulated the hydrolysis of GMz-ganglioside by hexosaminidase A. _Hexo— saminidase B was unable to cleave GMz-ganglioside even in the presence of the activator. These findings suggest that classical Tay-Sachs disease may be due to the absence of a glycoprotein subunit that is an obligatory component of hexo- saminidase A. However, the role of sialic acid residues on the surface of hexosaminidase A, and the relationship between asialo-hexosaminidase A and hexosaminidase B, have not been determined. Srivastava et_al. suggest that carbo— hydrate moieties on the surface of hexosaminidase may play a role in the binding of subunits to generate the holoenzyme (101).' The sialic acid residues may denote organ specificity to a particular hexosaminidase isozyme. 28 THE SPHINGOLIPIDOSES The Sphingolipidoses are hereditary disorders of lipid metabolism characterized by the absence of specific sphingo— lipid hydrolase(s) or transferase(s), and the concomitant accumulation of the lipid substrate(s) of the missing enzyme(s). Historically, the Sphingolipidoses were first described symptomatically and characterized as sphingolipid storage disorders. Only in the last ten years were these disorders found to result from the absence of catabolic enzymes, and only in the last year has a sphingolipidosis (GM3—Gangliosidosis) been found to result from a deficiency of an anabolic enzyme (145). The pathways of glycosphingo- lipid metabolism, consisting of the sequential hydrolysis of terminal hexose units (or addition as in the case of GM3- Gangliosidosis) and the points of interruption by the Sphingolipidoses are shown in Figure l. The gangliosides and the gangliosidoses are included for completeness, although this discussion will be limited to neutral glyco- sphingolipid metabolism. SANDHOFF'S DISEASE (GMleANGLIOSIDOSIS, VARIENT 2) The simplistic view of sphingolipidosis previously presented is complicated by the multiplicity of forms of glycophingolipid hydrolases described in the previous section. Two forms of hexosaminidase, forms A and B, are known and variations in the levels of these forms of hexosaminidase have resulted in the characterization of three clinical abnormalities with hexosaminidase deficiencies. These disorders are classical 29 .m3ouum cmxoun ma meMOHpcfl mum monocflmwaomcflsmm 0:» NO one :H mswuHSmmH mmflocmfloflmop mewusm .mmmmaoupmn pHmHHomsflnmmoomHm oamaummm ma >Hm>flmmmoosm UmUMHme mum mpHmHHomcwnmmoomam one manomflmflaomcanmm CH msAuHSmmm mofioamfloamma msmusm cam Emflaonmumz owmfiaomCHnmmoomao mo umooqoo ucmuuso one .H musmflm 30 05000023 _ 00.50.00 _ F 000005 0.320000 H 0002000....0-‘ ._. 0 3028828.». F - - J . 00.100 00.020.200.00 -mv. 000005 03.50“. 0002020200- B 50-20 A I 0000.50.00.30 -m. . L59.8.3 N. 000005 0.0.30.2 30.20100 1 I 0002020200- U _ 000005 fling 3-201818 m 0:0 4 0002320090: H L m ....> .280_§_S8-~20 8830 ....tzusm 30-20-818-323 mewm mgmomnjo 05000028100 > 30963930... 0:909:83: 000.02.30.30. .8.“ come 60 31 0250.00 000200020- 0. 2.0.0.0 000200—0290-2 H a 20882080020 00200000.; 3558202093004-2-00:-0.20 <2«z 30-20-30 w 30-20-60 > 082.20 H m 9.5 4 000.252.609.01 0 00> 008.02.800-05 a $805 0.22080— 05 .2 800.018.02.00 000005 02000-202 .- < 0322:6098... \83220 0002020200- 0.x 424.2 v L.00 0.01.00.02.00 30. 20-60-042.818 00020202007 200202.080 082.20 DOE—0.59.00 422.02 ..00-0.0-_00-0H0 m msoum pooHa 000 0000000 0.mun0m £003 00:0Hu0m 0H 0u0H5Edoo0 0©HE0H00 H>00000H0mflp 0:0 0©HE0A00 Hamox0£ane 0000000 0.>Hn0m 0H 0u0HSEsoom coaez mcwmflaomswnmmoomaw .N musmflm HOCH2 HOCH2 HOCHz H0 0 0 9H 0 O O'CHz‘CH'CH'CH=CH' (CH2 1'2 "CH: H OH OH NH I OH H 0-0 (sz )- n=22,24 CH3 L 1 R Trihexosyl Ceramide HOCHz HOCHz 0 H0 0 O-R OH OH H Digalactosyl Ceramide HOCHz HOCHz HOCHz Blood Group B Glycolipid 37 protein fractions analagous to the normal plasma ceramide tri- hexosidase, but devoid of all activity (72). Beutler and Kuhl (63) and Wood and Nadler (64) have examined normal and Fabry leukocytes and skin fibroblasts for ceramide trihexosidase activity. These workers isolated two forms of ceramide trihexosidase activity from normal cells, designated Forms A and B, which were distinguished by their thermostability. Examination of Fabry leukocytes and skin fibroblasts revealed the absence of Form A (thermolabile) and the presence of residual, but reduced amounts of the B form (thermostable). The latter reports were carried out using 4-methylumbelliferyl-a-gégalactopyranoside rather than trihexosyl ceramide to measure enzyme activity. Mapes and Sweeley have demonstrated the substrate specificity of ceramide trihexosidase to be absolute for the glycolipid substrate. These investigators have separated ceramflmatri- hexosidase activity Specific for trihexosyl ceramide from a-galactosidase activity specific for p-nitrophenyl- and 4- methylumbelliferyl-a-g-galactoside by affinity chromato- graphy (72). Affinity chromatography of Fabry plasma did produce several protein fractions active toward artificial substrates. However, these enzymes were not characterized. Fabry's disease is an X-linked trait that is not usually manifest in heterozygous females, who possess reduced ceramide trihexosidase activity. Hemizygous males usually exhibit the characteristic skin lesions and other manifes- tations of the disease and accumulate trihexosyl ceramide 38 and digalactosyl ceramide in all tissues. Hemizygotes with- out skin lesions have also been reported (123). No cases of homozygotes have been reported. Fabry's disease, like Sandhoff's disease does not appear to have any ethnic predi- lection. GAUCHER'S DISEASE (GLUCOSYL CERAMIDE LIPIDOSIS) In 1882, Phillip Gaucher described a disorder characterized by pro- gressive hepatosplenomegaly, that he believed to be an epithelial carcinoma of the spleen (124). Further investi- gations by other workers led to the conclusion that a substance was being deposited in the reticuloendothelial cells (125). In 1924, Lieb isolated large amounts of a cerebroside from Gaucher spleen, which he believed to be galactocerebroside (126-128). A discrepency between the optical rotation of Gaucher cerebroside and galactocerebro- side (129) led to the proper identification of Gaucher cerebroside as glucosyl ceramide (10, 130). In 1965, Brady §E_al. demonstrated the presence of glucocerebrosidase in normal spleen, using 14 C-glucosyl ceramide; the activity was reduced in the spleen of Gaucher patients (49), demonstrating that Gaucher's disease results from a deficiency of a cata- bolic enzyme. In 1973, Ho and O'Brien isolated two factors from normal human spleen, factors P and C, which combined to yield high glucocerebrosidase activity as previously described (47, 131). Factor P was isolated from Gaucher spleen and was found to have 10-15 times normal activity, whereas factor C activity was nearly absent. These workers postulated that 39 Gaucher's disease is a deficiency of factor C, the membrane associated, thermolabile component of glucocerebrosidase. The elevation of factor P activity in Gaucher spleen (factor P is the soluble, thermolabile component) may denote the role of this factor as a regulator of glucocerebrosidase activity. Gaucher's disease, like Tay-Sachs disease, is trans- mitted as an autosomal recessive trait (132). Three variants of Gaucher's disease have been characterized symptomatically. These are: type 1, chronic nonneuronopathic (adult) Gaucher's disease; type 2, acute neuronopathic Gaucher's disease; and type 3, subacute neuronopathic (juvenile) Gaucher's disease. There is a predelection for type 1 Gaucher's disease among Jews and a less striking prevalence of type 2 Gaucher's disease among this ethnic group. Type 3 Gaucher's disease is prevalent among the Norbotten families in Sweden but, unlike types 1 and 2, has not been manifested among Jews (133). Although all of these forms of Gaucher's disease are transmitted as autosomal recessive traits, they have not been characterized as distinct enzymatic defects. It is possible that the three forms of Gaucher's disease could represent deficiencies of factor P, factor C, or a deficiency of both factors. KRABBE'S DISEASE (GLOBOID CELL LEUKODYSTROPHY) In 1916, Krabbe described a disorder in two infant siblings charac- terized by a diffuse sclerosis of the brain and the presence of globoid cells in the white matter of the brain. He also noted familial occurrence of the disease (134). The 40 suggestion that globoid cells contained kerasin (galactocere- broside) was first made by Hallervorden in 1948 (135). Further investigations by other workers verified the accumu- lation of galactosyl ceramide in globoid cells (136-138). The ability of high levels of galactosyl ceramide to induce the formation of globoid cells was demonstrated by Austin g£;al. (139, 140) in 1961. These workers administered intracranial injections of galactosyl ceramide to rats and observed the production of globoid cells that were indistinguishable from those characteristic of Krabbe's disease. Galactosyl cera- mide was the only lipid effective in eliciting this, the globoid cell response. Glucosyl ceramide, sulfatide, psycho- sine, lactosyl ceramide, ceramide and other glycolipids all failed to stimulate the production of globoid cells. In 1970, Suzuki et_al. (38) and Austin §t_al. (39) demonstrated the deficiency of galactosyl ceramide B-galacto- sidase in brain, liver, spleen, and kidney of patients with Krabbe's disease using 3 H-galactosyl ceramide. In 1972, Miyatake and Suzuki demonstrated a deficiency of galactosyl sphingosine (psychosine) B-galactosidase in brain, liver and kidney of patients with Krabbe's disease (141). Using galactosyl sphingosine B-galactosidase from rat brain (which may be different from the human brain B-galactosidase) these workers demonstrated that galactosyl ceramide and galactosyl sphingosine were mutually competitive inhibitors (142). More recently, Wenger et al. have found lactosyl ceramide B-galac- tosidase is also deficient in tissues from patients with 41 Krabbe's disease (56). In further investigations, Wenger isolated a B-galactosidase from normal human brain which catalyzed the hydrolysis of both galactosyl ceramide galac- tosylglucosylceramide (57). Both activities had identical heat denaturation patterns. The effects of stimulators such as sodium cholate and oleic acid were the same. The sub- strates were mutually competitive inhibitors, and the activi- ties co-chromatographed on Sephadex G-200 and DEAE cellulose. These studies are not complete, but the findings support the hypothesis that Krabbe's disease is a deficiency of a B-galac- tosidase active toward psychosine, galactosyl ceramide and lactosyl ceramide. Genetically, Krabbe's disease is an autosomal recessive trait with no ethnic preponderance (143). Globoid cell leukodystrophy has also been reported to occur in the West Highland and Cairns Terriers (144). This canine disorder has been studied extensively and it also appears to be transmitted as an autosomal recessive trait (143). ENZYME REPLACEMENT THERAPY The infusion of exogenous enzymes into patients with sphingolipidoses has been attempted as a therapy for these genetic diseases. Clearly enzyme replacement therapy would not provide a cure for sphingolipidosis. A cure would re- quire genetic engineering, which is beyond the expertise of present-day science. Indeed, the gene mutations at the root of these diseases have not yet been identified. Enzyme 42 replacement therapy could prove to be a method of regulating the accumulation of sphingolipids in patients with an enzyme deficiency. However, replacement therapy may not be as straight-forward as injecting purified enzymes into patients suffering from sphingolipidosis. Normal glycosphingolipid hydrolases have been found in multiple forms, and the organ specificity of a particular isozyme may determine the uptake of that isozyme by the target organ (146-150). Chemical or enzymatic modification of the enzyme (151-154) or encapsu- lation in membranes (155-157) or liposomes (158-162) may pro- vide mechanisms of directing uptake of the enzyme by the target organ and increasing the lifetime of the enzyme in circulation. With these considerations in mind, the attempts that have been made to treat Metachromatic Leukodystrophy (163, 164), Sandhoff's disease (165, 166), Tay-Sachs disease (170), Gaucher's disease (167), and Fabry's disease (168, 169, 172, 173) will be examined. The first attempt to treat a sphingolipidosis by enzyme replacement therapy was in Metachromatic Leukodys- trophy, which is characterized by the accumulation of cere- broside sulphates in various organs of the body, but pre- dominantly affects the white matter of the brain (174). Therefore, the primary target organ for enzyme replacement therapy is the brain, and successful treatment requires that the injected enzyme must pass the blood-brain barrier, enter the brain, and degrade the accumulating sulfatides. 43 In 1967, Austin attempted to treat Metachromatic Leukodystrophy by intrathecal administration of a partially purified arylsulphatase A from human urine (164). No changes were observed in the patient's clinical status, and prein- fusion and postinfusion levels of arylsulphatase A activity in the organs were not changed. In 1969, Greene et_al. attempted to treat Metachromatic Leukodystrophy by intravenous and intrathecal infusion of a partially purified arylsulpha- tase from beef brain (163). Following intravenous infusion of arylsulphatase A no enzyme activity was detected in the brain or the cerebrospinal fluid, but some arylsulphatase activity was detected in the liver. Intrathecal administra- tion of arylsulphatase A resulted in a peak of enzyme activity in the cerebrospinal fluid one hour following in- jection. Activity decreased rapidly until only very low arylsulphatase activity could be detected twenty hours following injection. Arylsulphatase activity could not be detected in the brain or liver five hours following the intrathecal injection. The failure of the administered enzyme to enter the target organ, the brain, and the presence of intravenously- infused arylsulfatase A in the liver, the site of degrada- tion of desialylated glycoproteins, indicate that the beef brain arylsulphatase A may be inappropriate for enzyme replacement therapy via direct infusion. The enzyme apparently lacks the appropriate marker to gain passage across the blood-brain barrier, or uptake by neuronal tissue. 44 This could be a reflection of specificity for the source from which the enzyme was isolated. Beef brain arylsulfatase and human brain arylsulfatase could be different proteins, with different uptake markers. Many proteins and drugs do not pass the blood-brain barrier, and it is more likely that glycosphingolipid hydrolases cannot cross the blood-brain barrier at all, but must be synthesized in situ. The failure of enzyme replacement therapy to alter the clinical status of patients with diseases involving brain damage, as in Meta- chromatic Leukodystrophy, Krabbe's disease and Tay-Sachs disease may not indicate the inefficacy of the therapy as much as the pathology of the disease. Enzyme replacement therapy may provide no treatment at all for those disorders involving brain damage. Indeed, irreversible damage may even be suffered in utero. In 1973, Johnson §t_al. attempted to treat a patient with Sandhoff's disease with hexosaminidase A purified from human urine (165). Hexosaminidase levels in serum, urine, cerebrospinal fluid, liver and brain were monitored before and after injection of purified hexosaminidase A. Following administration of the enzyme, serum hexosaminidase A rose to normal levels, and then decreased rapidly to preinjection levels 75 minutes followinginfusion. The hexosaminidase activity in the liver was nine times preinfusion levels when sampled 45 minutes following infusion. No changes in hexosa- minidase levels in urine, cerebrospinal fluid or brain were observed following administration of the enzyme, and no 45 change in the clinical status of the patient was observed. The rapid decrease in serum hexosaminidase is similar to the exponential disappearance of desialylated glycoproteins re- moved by the liver. This suggests that urinary hexosamini- dase A may be a somewhat degraded form of hexosaminidase, lacking the appropraite markers for uptake by the target organ, the brain. The increase in hepatic hexosaminidase levels following administration of the enzyme supports this hypothesis. Indeed, Hickman gt_al. observed that highly purified hexosaminidase A from human urine was taken up poorly, if at all by fibroblasts from a patient with Sandhoff's disease (148). These findings suggest that urine may not be a good source for enzymes to be used for enzyme replacement therapy because specific recognition sites are required for uptake by tissues other than the liver. In 1972, Desnick et_al. attempted to treat a patient with Sandhoff's disease by infusion of fresh whole plasma or fractionated plasma concentrates containing hexosaminidase activity (166). Maximal enzyme activity for each infusion occurred immediately following infusion, and decreased rapidly until preinfusion levels were reached after thirty hours. The concentration of glycosphingolipids in plasma were monitored after infusion. Levels of GM3-ganglioside and trihexosyl ceramide, the products of hexosaminidase activity on GMZ-ganglioside and globoside respectively, were increased over preinfusion levels. Similarly, levels of globoside in urinary sediment decreased following infusions. The hexosaminidase activity in the organs was not monitored, 46 and the clinical status of the patient was not altered. The increase in levels of GM3-ganglioside and trihexosyl ceramide in serum, and the decrease of globoside levels in urine sedi- ment indicate some breakdown of the accumulating GMZ-ganglio- side and globoside by infused hexosaminidase. The decrease of globoside levels in urine sediment may also indicate the entry of infused hexosaminidase into kidney, but it is difficult to evaluate the degree of success of this treatment because hexosaminidase levels in the organs were not monitored. The rate of decrease in hexosaminidase activity in serum was somewhat slower than the exponential decrease in hexosamini- dase activity observed by Johnson §E_al. (165) using hexo- saminidase A from urine. The increased lifetime of the plasma hexosaminidase in circulation may indicate that it is more highly sialylated than hexosaminidase from urine or that it contains cell-specific receptor sites. Indeed, in 1973, Ikonne and Ellis demonstrated that hexosaminidase A from serum and liver are different enzymes, that could be separated by DEAE-cellulose chromatography (171). These in- vestigators found that serum hexosaminidase A was susceptible to attack by neuraminidase while hexosaminidase from liver was not. These findings suggest that enzymes isolated from human plasma may have greater theraputic value than those from liver or urine, which may be desialylated and thereby destined for more rapid removal from circulation by the liver. 47 In 1973, O'Brien infused two patients with Tay-Sachs disease with normal plasma and monitored hexosaminidase A activity in serum. Hexosaminidase A activity reached maxi- mallevels in the patients' serum 2-4 hours following in- fusion, and rapidly decreased until activity could not be detected 24 hours later. Levels of hexosaminidase A activity in the patients' tissues and urine and levels of GMZ- and GM3-ganglioside in plasma and urine were not monitored. The rate of decrease of hexosaminidase A levels fluctuated greatly between patients, and between infusions of plasma in the same patient and an evaluation of the survival of hexosaminidase A in circulation is therefore difficult. However, the hexosaminidase A in a second infusion into one of the patients corresponded to the observed survival of in- fused plasma hexosaminidase in a patient with Sandhoff's disease (166). In 1974, Brady et_al. attempted to treat a patient with subacute neuronopathic Gaucher's disease and a patient with chronic, nonneuronopathic Gaucher's disease using glucocere- bresidase purified from human placenta (167). The glucocere- brosidase activity in plasma reached maximal activity immediately following infusion of both patients, then de- creased rapidly to preinfusion levels four hours following administration of the enzyme, with a half-life of about 18 minutes. Hepatic glucocerebroside levels were monitored in both patients. A 26% decrease in hepatic glucocerebroside was observed in both patients following intravenous injection 48 of the enzyme. The levels of glucocerebroside in plasma of both patients were also monitored. In the patient with sub- acute neuronopathic Gaucher's disease, plasma glucocerebro- side had decreased 20% in three days following injection. In the patient with acutermmneuronopathic Gaucher's disease, plasma glucocerebroside had dropped 11% by 14 days following administration of the enzyme. Glucocerebroside levels in erythrocytes in both patients dropped from levels two and a half times normal to essentially normal values following ad- ministration of the enzyme. The sizes of the patients' livers were not determined. The decrease in glucocerebrosidase in plasma following administration of the enzyme corresponds to the exponential rate of removal of desialylated glyc0proteins from circulation by the liver. Brady gt_al. performed experiments with primates using intravenous injections of purified glucocerebrosidase from human placenta and found the enzyme to be present predomi- nantly in the liver (167). This finding suggests that placental glucocerebrosidase is rapidly removed from circu- lation by liver and may not be taken up by other organs. These workers were unable to determine the efficacy of placental glucocerebrosidase in reducing glucocerebroside in spleen, the other target organ in Gaucher's disease. Both patients had been splenectomized several years prior to enzyme replacement therapy. The "success" of enzyme replace- ment therapy in these patients with Gaucher's disease is largely due to the uptake of glucocerebrosidase by the liver. 49 The function of the liver as the binding site for desialy- lated glycoproteins appears to have contributed to the success of enzyme replacement therapy in Gaucher's disease. However, the use of placental enzymes for treatment of other sphingolipidoses, where the liver is not a target organ, is discouraged by the results of treatment of Gaucher's disease using placental glucocerebrosidase. In 1970, Mapes gt_al. attempted to treat two hemizygotes with Fabry's disease by infusion of normal plasma (168). Ceramide trihexosidase activity and levels of trihexosyl cer- amflmain plasma were monitored, but tissues were not monitored for enzyme activity. Maximal ceramide trihexosidase activity in plasma was observed six hours following infusion. The ceramide trihexosidase maxima were 22- and 35-fold higher than had been expected on the basis of the activity in the in- fused plasma. The enzyme activity decreased rapidly from 6 to 12 hours following injection, to levels somewhat lower than those predicted on the basis of infused ceramide tri- hexosidase activity. These levels of ceramide trihexosidase activity decreased gradually until activity could not be de- tected 7 days following injection. The concentration of tri- hexosyl ceramide in the patients' plasma decreased to a minimum 10 to 20 hours following administration of the enzyme, and then increased rapidly beyond initial levels. From this maximum, the concentration of trihexosyl ceramide in plasma decreased to about 50% of the initial concentra- tions ten days following infusion. Lactosyl ceramide levels 50 and examinations of tissues for ceramide trihexosidase activity were not reported. In 1972, Sweeley gt_§l. reported a second attempt to treat hemizygotes with Fabry's disease by plasma infusion (172). Plasma a-galactosidase activities were monitored using an artificial substrate, 4-methylumbelliferyl-a-g- galactopyranoside, in three hemizygous patients with Fabry's disease. A fourth patient was also infused with normal plasma. However, in addition to monitoring levels of a-galactosidasein plasma using the artificial substrate, the ceramide trihexosid- ase: activity was also followed using the natural substrate. The a-galactosidase activity rapidly increased to a maximum immediately following administration of the enzyme, and no enhancement of a-galactosidase activity beyond levels pre- dicted on the basis of infused enzyme activity were observed in any of the patients. The ceramide trihexosidase activity monitored in the fourth patient, however reached an initial maximum at pH 5.4 approximately three hours following in- fusion. The activity observed was approximately six times greater than had been anticipated. The enzyme activity de- creased rapidly and no ceramide trihexosidase activity was detected between 4-8 hours following infusion. Between 8-12 hours following administration of the enzyme, a second maxi- mum in ceramide trihexosidase activity was observed. This activity was approximately twelve times greater than had been anticipated. The enzyme activity decreased rapidly from this maximum to levels that corresponded to the anticipated cera- mide trihexosidase activity, based on the amount of ceramide 51 trihexosidase administered. Levels of ceramide trihexosidase in tissues were not determined. These findings by Mapes §E_§l. and Sweeley gt_al. indicate that complicated changes of the administered plasma ceramide trihexosidase may occur in the patient's tissues. When a sample of normal plasma was diluted with an aliqout of Fabry plasma, no activation was observed (172). The longevity of infused plasma ceramide trihexosidase in the patients' plasma may indicate that cera- mide trihexosidase from plasma is resistant to uptake by the parenchymal cells of the liver. In 1973, Brady e£_al. attempted to treat two hemizygotes with Fabry's disease using plasma infusions, leukocytes and platelets suspended in plasma, and purified ceramide trihexo- sidase from placenta (169). The activity of a-galactosidase in plasma was monitored using the artificial substrate, 4- methylumbelliferyl-a-Q—galactopyranoside. Levels of globo- side, trihexosyl ceramide and lactosyl ceramide in plasma were also monitored. No increase in a-galactosidase activity was observed in the patient's plasma following infusion of normal plasma. Except for a slight decrease in globoside levels, no change in levels of circulating glycolipids was found. When platelets and leukocytes suspended in plasma were infused, the circulating a-galactosidase activity rose 68% over the preinfusion levels. This high level of a-galac- tC>Sidase activity persisted for four hours following infusion, and gradually decreased to preinfusion levels 24 hours following administration of the enzyme. Levels of trihexosyl ceramide in plasma had decreased to_64% of the 52 preinfusion level 40 minutes followig injection. Trihexosyl ceramide gradually returned to preinfusion levels in plasma 24 hours following administration of the enzyme. The ad- ministration of ceramide trihexosidase purified from human placenta resulted in maximal a-galactosidase activity imme- diately following infusion of the enzyme. The a-galactosid- asea activity decreased at an exponential rate until pre- infusion levels were reached one hour following injection. The trihexosyl ceramide in the patient's plasma decreased to a minimum 40 minutes following administration of the enzyme, which was 58% of the preinfusion levels and then gradually increased to the preinfusion levels 48 hours following in- jection of the placental enzyme. No change was found in a-galactosidase levels in intestine following administration of the enzyme, but a-galactosidase activity was detected in white blood cells and predominantly in the liver. The a-galactosidase in the liver was 2 to 4 times greater than the total enzymatic activity injected. No a-galactosidase activity was detected in urine following infusion of placen- tal ceramide trihexosidase. The failure of these workers to detect a-galactosidase activity in the patient's plasma following infusion of normal plasma may be due to the long infusion period of four hours (173). The infusion of normal leukocytes and platelets into patients with Fabry's disease represents the first attempt to treat a sphingolipidosis using enzyme-loaded cells. The survival in plasma, of a-galactosidase administered in leuko- cytes and platelets was 24 hours compared to 1 hour for 53 purified placental enzyme administered directly. The decrease in trihexosyl ceramide levels following administration of the leukocyte and platelet suspension indicates the effectiveness of this method of therapy in treating sphingolipidosis. The exponential disappearance of purified placental a-galactosi- dase corresponds to the exponential removal of desialylated glycoproteins by liver parenchymal cells (151, 152). This is supported by the presence of a-galactosidase activity in the liver of patients infused with placental enzyme. The use of the artificial substrate, 4-methylumbelliferyl-a-g-galacto- side, to moniter a-galactosidase may not be a true indication of ceramide trihexosidase activity in the patients' plasma. Sweeley §E_al. found that a-galactosidase activity and cera- mide trihexosidase activity in plasma of a Fabry hemizygote infused with normal plasma did not correspond (172). Indeed, the ceramide trihexosidase had a much greater lifetime than the a-galactosidase activity monitored with 4-methylumbelli- feryl-a-g-galactopyranoside. The rapid decrease in trihex- osyl ceramide in plasma followed by the gradual increase to approximately preinfusion levels by 48 hours following infusion of placental enzyme may correspond to the fluctuation of trihexosyl ceramide observed by Mapes gt_al. (168) Unfor- tunately, the time intervals and the extent of the study by Brady §£_al. do not provide an accurate evaluation of this facet of the therapy. In 1974, Christensen attempted to treat a hemizygote patient with Fabry's disease by infusion of normal plasma 54 (173). The a-galactosidase activity in the patient's plasma was monitored for 8 hours following injection using 4-methyl- umbelliferyl-a-g-galactopyranoside to monitor activity. The a-galactosidase activity rose to a maximal level immediately following infusion of normal plasma. The maximal activity roughly corresponded to the anticipated activity based on the a-galactosidase activity administered. The a-galactosidase activity decreased in the first hour from this maximum to approximately half the maximal level initially observed. The levels of a-galactosidase in the patient's plasma decreased gradually from this point and did not reach preinfusion levels in the 8 hour period over which activity was monitored. These findings correspond to the results of infusion of plasma in Fabry hemizygotes reported by Sweeley et_al. (172). The survival of infused plasma a-galactosidase and ceramide trihexosidase in circulation appears to be longer than the survivial of placental a-galactosidase. This suggests that plasma may be a better source of ceramide trihexosidase for treatment of Fabry's disease. However, the failure of workers using plasma to monitor levels of ceramide trihexosi- dase in organs, and the failure of workers using placental enzyme to monitor ceramide trihexosidase activity using the natural substrate, leave many questions unanswered concern- ing the efficacy of these treatments. THE ISOLATION OF a-GALACTOSIDASES BY AFFINITY CHROMATOGRAPHY Affinity chromatography has been employed in the isola- tion of a-galactosidases from ficin (fig latex) (72, 186), 55 coffee beans (187), human plasma, urine and kidney (72). In 1973, Kanfer §t_al. isolated a-galactosidases from ficin by affinity chromatography on g-galactono-a-lactone bound to Sepharose (186). The ficin a-galactosidases were eluted in one peak from the affinity column, but were not examined for contaminating enzyme activities, such as pro- tease, B-galactosidase, or ceramide trihexosidases, which are also present in ficin. The "galactonate" affinity adsorbent was not specific for a-galactosidases; these workers also isolated B-galactosidases from jack bean meal in a single peak using this adsorbent. None of the enzymes were examined by electrophoresis for purity. In 1974, Harpaz et_al. isolated a-galactosidases from coffee beans by affinity chromatography on N-e-aminocaproyl- a-g-galactopyranosylamino—Sepharose (187). The enzymes were isolated in a single peak from the affinity column and poly- acrylamide gel electrophoresis revealed three bands that were active toward p-nitrophenyl-a-g-galactopyranoside. In 1973, Mapes and Sweeley examined ficin, human plasma, kidney and urine by affinity chromatography on p- aminophenyl melibioside bound to affinose (72). The ficin a-galactosidases were isolated in three protein peaks from the affinity column. The first peak was active toward both p-nitrophenyl—a-g-galactopyranoside and trihexosyl ceramide, the second peak was active only toward trihexosyl ceramide and the third peak was active only toward the artificial substrate. The eluate from the affinity column was not 56 examined for protease or B-galactosidase activities, nor were the a-galactosidases examined by electrophoresis. Affinity chromatography of whole plasma resulted in the isolation of six protein peaks active toward the artificial substrate p- nitrophenyl—a-g-galactopyranoside and five protein peaks active toward trihexosyl ceramide, two of which were active at pH 5.4, and three that were active at pH 7.2. Affinity chromatography of urine resulted in a similar protein profile, however the urine appeared to be enriched in ceramide tri- hexosidases active at pH 7.2. Affinity chromatography of a human kidney preparation resulted in the isolation of one protein peak active toward the artificial substrate, and one protein peak active toward trihexosyl ceramide at pH 5.4. These findings suggest that different isozymes of ceramide trihexosidase, perhaps organ-specific isozymes, exist. The present research establishes a method of isolating ceramide trihexosidase from plasma (Cohn fraction IV-l), suitable for pilot scale isolation of ceramide trihexosidase and eventual evaluation of enzyme replacement therapy in Fabry's disease. These methods were first examined using ficin a-galactosidases to establish the efficacy of the pro- cedures, particularly affinity chromatography. MATERIALS AND METHODS MATERIALS SOURCES OF ENZYMES Ficin Cohn Fraction IV-l Galactose Dehydrogenase Galactose Oxidase SOLVENTS General Solvents Dry Methanol Dry Pyridine Dry Acetonitrile CHEMICALS p-Nitrobenzoyl Chloride 57 Sigma Chemical Co. St. Louis, Mo. Michigan Department of Health Lansing, Mi. Boehringer Mannheim Corp. New York, N.Y. Worthington Biochemical Corp. Freehold, N.J. Solvents were redistilled by constant-flow rotary evaporation. Methanol was dried by re- fluxing over magnesium turnings containing iodine as catalyst. Methanol was distilled and stored over molecular sieves. Pyridine was dried by re- fluxing over barium oxide. It was then distilled and stored over potassium hy- droxide. Acetonitrile was dried by refluxing over barium hydroxide. It was distilled and stored over molecular sieves. Aldrich Chemical Co. Milwaukee, Wis. 58 Putrescine Triethylamine Cyanogen Bromide Stannous Chloride Sodium Nitrite a-g-(+)-Melibiose Fluram Sodium Borohydride ENZYME SUBSTRATES p-Nitrophenyl-a-g- GalactOpyranoside p-NitrOphenyl-B-g- Galactopyranoside Casein (Practical (Grade) SILYLATING REAGENTS Trimethylchlorosilane Hexamethyldisilazane Dimethyldichlorosilane CHROMATOGRAPHIC SUPPLIES Silicic Acid (Unisil) Silica Gel G Thin Layer Chromatography Plates Sigma Chemical Co. St. Louis, Mo. Aldrich Chemical Co. Milwaukee, Wis. Aldrich Chemical Co. Milwaukee, Wis. J. T. Baker Chemical Co. Phillipsburg, N.J. J. T. Baker Chemical Co. Phillipsburg, N.J. Sigma Chemical Co. St. Louis, Mo. Roche Diagnostics Nuttley, N.J. Sigma Chemical Co. St. Louis, Mo. Sigma Chemical Co. St. Louis, Mo. Sigma Chemical Co. St. Louis, Mo. Sigma Chemical Co. St. Louis, Mo. Pierce Chemical Co. Rockford, Ill. Pierce Chemical Co. Rockford, Ill. Applied Science Labora- tories, Inc. State College, Pa. Clarkson Chemical Co. Williamsport, Pa. Analtech, Inc. Newark, Del. 59 3% SE-30 on Chromosorb W.H.P. 5% SE-30 on Chromosorb W.H.P. Dextran-Coated Glasso Beads (Pore Size 550A Particle Size 177-2403) MISCELLANEOUS REAGENTS grade. Sodium Taurocholate Sodium Cholate Methyl Orange B-Napthol Sodium Borohydride-H3 Bio-Solv BBS-3 Scintillation Toluene PPO (2,5-diphenylox- azole) Dimethyl-POPOP (1,4-bis- 2-[4-methyl-5-phenylox- azolyl] -benzene) Hewlett-Packard Avondale, Pa. Supelco, Inc. Bellefonte, Pa. Pierce Chemical Co. Rockford, Ill. Calbiochem San Diego, Calif. Sigma Chemical Co. St. Louis, Mo. Fisher Scientific Co. Fair Lawn, N.J. Sigma Chemical Co. St. Louis, Mo. New England Nuclear Boston, Mass. Beckman Instruments, Inc. Fullerton, Calif. Beckman Instruments, Inc. Fullerton, Calif. Research Products Inter- national Corp. Elk Grove Village, Ill. Research Products Inter- national Corp. Elk Grove Village, 111. All other chemicals and materials were of reagent 60 METHODS PURIFICATION OF a-GALACTOSIDASES FROM FICIN The a-galacto- sidases from ficin were purified in a two step procedure which included the aqueous extraction of a-galactosidases and affinity chromatography. Crude ficin (500 mg) was dissolved in 10 ml of 0.05 M sodium acetate buffer, pH 4.5, at 4°C. The solution was stirred slowly on a magnetic stirrer at 4°C for one hour and then centrifuged at 12,200xg at 4°C in a Sorvall RC2-B centrifuge for 10 minutes to remove undissolved material. Then supernatant solution was applied to the affinity column (5mm x 80 mm) with no further preparation. Ficin solutions were prepared just prior to affinity chroma- tography and were not stored frozen, or at 4°C, in advance. PREPARATION OF AFFINITY COLUMN ADSORBENTS Melibiose phenyl- hydrazone glass beads for the purification of a-galactosidases and acetaldehyde phenylhydrazone glass beads to serve as a control adsorbent were prepared (see Figure 3). Dextran coated glass beads were activated and coupled with putrescine as described by Cuatrecasas (175, 176), as follows: Dextran- coated glass beads (59) were placed in a 150 ml beaker and 20 m1 of 2 M sodium carbonate at room temperature was added. Cyanogen bromide (lOg) dissolved in acetonitrile (5 ml) was added slowly with stirring. The mixture was stirred vigor- ously With a glass rod for 15 minutes. As the mixture became hot it was cooled momentarily in an ice bath. Manual stirring with a glass rod was much preferred to the use of magnetic stirring bars, which could break the dextran-coated glass 61 00000 00000 0000000»: 1000000 0000n0002 mo 00000000000 0:0 :0 00000 .m 005000 62 Preparation of Glass Bead Affinity Adsorbent \ 51’0” 310 Glass >1‘0’gl: (CH2 )3- NH— 51‘ / o" ‘ Dextran CNBr i (9 \SI’OH Et “”2 Glass Si-O-§i-(CH2)3 WNH- o‘CNH / Et 0’ 51‘0" / Dextran HZN-(CH2)E-NH J, (+3 \S .2 0H ““2 (9 Glass S1\51—o-§1—(CH2)3 NH- -6 0” ,“2 Et O-C-NH-(CH )-NH SLOH 2 4 2 /’ i Dextran 8 N0 0’ O 2 \, 1 51’0" Et "“2 GD \ 0H HH2 Glass [Si-O- i- (CH 2)3 NH-C-O 8 $1 Et -NH- (CH 2)4 NH- / 0” Dextran HCI SnCl2 63 Preparation of Glass Bead Affinity Adsorbent (continued) \ Glass \Ei-O-g -(CH2) 3-NH-gl0 0" ”“2 i 30 -CI NH-(CH )- -NH- QZNH 51‘0" 2 4 / Dextran HC1 NaNOz \ 0H Q 5i Et 1:“ 0H0 Glass i-0-§i-(CH2)3-NH- - “Hz 89 5’ Et —E-NH- -(CH2)4- NH H=H+ c1 / k0H Dextran HCl SnCl2 \ 51/0" 0 \ 51: HQ mass 51-0- i-(CH2)3- -HH—(52 ED 54‘ Et -NH-( CH2)4-NH-H OZNH-NH 0H Dextran / H OH H H a—D-Mel ibiose 0---CH2 5 NH‘ Et G1ass \Si-O-gi- Mcnfl-uu—E— " M 80' / ‘OH Dextran "$3” HOH f5}; H -o 64 beads, exposing uncoated porous glass surfaces that could serve as sites for ionic bonding of proteins, resulting in non-specific adsorption. After 15 minutes the mixture was filtered on a Buchner funnel and the glass beads were washed with 0.2 M sodium bicarbonate (30 ml) at 4°C. A solution of putrescine (Sg) dissolved in 0.2 M sodium bicarbonate (20 ml) was added to the glass beads and the mixture was decanted in- to a 150 ml beaker and stirred slowly for 30 minutes at 4°C with a glass rod. The Buchner funnel was rinsed with 0.2 M sodium bicarbonate (10 ml). The filtration and addition of the putrescine solution were carried out in less than 90 seconds. The putrescine-treated glass beads were then fil- tered on a Buchner funnel and washed with distilled water and allowed to dry on the funnel. The putrescine-treated glass beads were coupled with p-nitrobenzoyl chloride by combining the dry glass beads (59) with p-nitrobenzoyl chloride (lg), triethylamine (550 mg) and chloroform (70 ml). The glass beads must be dry because water will react rapidly with p-nitrobenzoyl chloride to yield hydrogen chloride and p-nitrobenzoic acid. The mixture was refluxed for one hour with swirling on a New Brunswick Shaker at slow speed to prevent bumping. When glass beads were refluxed without swirling, violent bumping occurred. Attempts to eliminate bumping by bubbling nitrogen through the mixture or placing porcelain chips in the mixture failed, and the use of a magnetic stirring bar was not desirable. The New Brunswick Shaker proved to be the most effective 65 method of agitating the mixture to prevent bumping. The p- nitrobenzamide glass beads were decanted onto a coarse sintered glass funnel and washed with chloroform (100 ml) and water (100 ml). The p-nitrobenzamide glass beads were reduced using a 20-fo1d excess of stannous chloride in concentrated HCl. A solution of stannous chloride (39) in concentrated HCl (30 m1) at 4°C was added to the p-nitrobenzamide glass beads (59). The mixture was allowed to stand for 2.5 hours at 4°C. After 2.5 hours the glass beads were decanted onto a coarse sin- tered glass funnel and washed with water (1 liter). The hydrolysis of dextran in concentrated HCl at 4°C was not observed. The multiple linkages between dextran and the glass beads may contribute to the stability of the dextran in con- centrated HCl, and the cold temperature at which this re- action is carried out reduces the likelihood of significant hydrolysis of dextran from the surface of the glass beads. The arylamine glass beads (59) were diazotized in a 300 ml round-bottom flask in 2N HCl (70 ml). The mixture was cooled to 4°C in an ice bath and sodium nitrite (49) was added slowly. The round-bottom flask in the ice bath was evacuated on a water aspirator for 30 minutes to remove air bubbles from the porous glass beads. After 30 minutes the diazotized glass beads were decanted onto a funnel with a coarse glass frit and washed with 1% aqueous sulfamic acid (50 ml) (to destroy unreacted sodium nitrite) and distilled water (100 m1). 66 The diazotized glass beads were reduced in a solution of stannous chloride as described for the nitrobenzamide derivative. The phenylhydrazine glass beads were coupled with a-g-melibiose to generate an a-galactosyl affinity adsorbent, and with acetaldehyde to prepare a control adsorbent. The melibiose phenylhydrazone glass beads were prepared by re- fluxing phenylhydrazine glass beads (59) with melibiose (3g) in distilled water (30 ml) and a few drops of acetic acid. The mixture was refluxed for 60 minutes with swirling on a New Brunswick Shaker to prevent bumping. After 60 minutes the beads were decanted onto a coarse sintered glass funnel and washed with distilled water. The acetaldehyde phenylhydrazone glass beads were pre- pared by refluxing phenylhydrazone glass beads (59) with acetaldehyde and a few drops of acetic acid. Acetic acid is necessary to maintain an acid pH during the reaction, since acetaldehyde polymerizes under alkaline conditions. The mixture was refluxed for 60 minutes on a New Brunswick Shaker to prevent bumping. After 60 minutes the acetaldehyde phenyl— hydrazone glass beads were decanted onto a funnel with a coarse glass frit and washed with distilled water. AFFINITY CHROMATOGRAPHY The glass column to be used for affinity chromatography was silylated prior to use to elimi- nate the possibility of non-specific, ionic bonding of proteins to the polar glass surface of the column. This precaution may be unnecessary considering that the surface 67 area of the column is quite small compared to the surface area of the dextran-coated glass beads (=70 m2/ 9 glass beads). The column (5 x 80 mm) was rinsed with concentrated HCl, water and hexane and allowed to dry in the air. The column was sealed with a cork at the bottom and filled with an 8% solution of dimethyldichlorosilane in hexane. The column stood at room temperature for 30 minutes and then was drained and rinsed with hexane. The column was next filled with dry methanol and allowed to stand at room tem- perature for 15 minutes. The silylated column was drained, rinsed with acetone and allowed to dry in the air. The column was packed by first placing the glass beads in a round-bottom glask with 0.05 M sodium acetate buffer pH 4.5 (approximately 40 ml). The round-bottom flask was evacuated for 15 minutes with gentle tapping to eliminate air bubbles from the porous glass beads. The column was fitted with a silylated glass wool plug, filled with acetate buffer and air bubbles were dislodged. The glass beads were poured into the column with care taken so that the column did not run dry. The packed column was equilibrated at 4°C and washed with several volumes of 0.05 M sodium acetate buffer, pH 4.5, prior to use. The crude ficin solution (0.75 ml) was applied to the column and drained to the top of the adsorbent. The column was incubated at 4°C for 30 minutes and then washed with 0.05 M sodium acetate buffer, pH 4.5 at 4°C (40 ml) to elute unbound material from the column. The enzymes were eluted using 0.025 M sodium borate buffer, pH 8.5, 68 at 4°C. Acetate and borate eluates from the column were collected in 1 ml fractions. A constant flow rate of approx- imately l ml/minute was maintained using an LKB 12000 Vario- perpex peristaltic pump. ASSAYS Protein was determined by flourescence using Fluram (fluorescamine) (177). Fluram was prepared by dissolving 30 mg of Fluram in 100 ml of dry acetonitrile. Care must be taken to exclude water from the Fluram reagent and the Fluram-acetonitrile solution since the reagent reacts rapidly with water to produce an inert, nonfluorescent product. The protein solution to be assayed (0.005 ml) was added to 0.75 ml of 0.2M sodium phosphate buffer pH 9.0. The solution was placed on a vortex and 0.25 ml of Fluram solution was jetted in while vortexing. Fluram reacts more rapidly with primary amines than with water; however, it is necessary to disperse the Fluram throughout the reaction mixture immediately to en- sure the derivatization of all primary amine groups before decomposition with water can occur. Fluorescence was measured on an Aminco J4-7439 Fluoro-Colorimeter using a Corning 7-51 (Aminco J4-7l99) primary filter (390 nm excita- tion) and a Turner 110-828 secondary filter (475 nm emission). Bovine serum albumin solutions from 0.01-0.10 mg/ ml were employed as standards. Assays for a-galactosidase activity were carried out using the synthetic substrate p-nitrophenyl-a-g—galactopyra- noside and trihexosyl ceramide. Enzyme assays using p-nitro- pheny1-a-g-galactopyranoside were conducted by adding 0.1 ml 69 of 0.01M p-nitrophenyl—a-g-galactopyranoside solution prepared in 0.05M sodium acetate buffer pH 4.5 to 0.1 ml of enzyme sol- ution and 0.3 ml of acetate buffer. The solution was incuba- ted at 23°C for 2 hours and the assay was terminated by the addition of 0.5 ml of saturated sodium borate, pH 9.7. The optical density of the solution was measured at 410 nm on a Gilford 2400 Spectrophotometer. 'Ceramide trihexosidase activity in ficin was assayed by adding 0.1 ml of enzyme sol- ution to 0.9 m1 of reaction mixture containing 500 nmoles of trihexosyl ceramide solubilized in 4 mg sodium cholate, 15 mg sodium taurocholate, 0.08 ml n-butanol and 0.05M sodium ace- tate buffer, pH 4.5, in a final volume of 0.9 ml. The sol- ution was incubated at 25°C for 4 hours and the assay was terminated by the addition of 2 m1 of chloroform-methanol 2:1 (v/v), vortexed and centrifuged at l300xg for 5 minutes in a Sorvall GLC-l clinical centrifuge. The upper phase was re- move and the lower phase was washed two more times with equal volumes of theoretical upper phase (chloroform-methanol- water 3:48:47 (v/v/v) (183). The combined upper phases were dried under nitrogen. Galactose hydrolyzed from trihexosyl ceramide was estimated by incubating the dried upper phase in 0.9 m1 of galactose dehydrogenase at 37°C for 1 hour. Galactose dehydrogenase catalyzes the oxidation of galactose to galactonate, accompanied by the reduction of NAD to NADH. The NADH generated by this reaction is quantited by fluores- cence on a computer-assisted fluorimeter at 340 nm excitation and 460 nm emission wavelengths. Galactose was used as a standard. 70 Enzyme assays for B-galactosidase were conducted using p—nitrophenyl-B-g—galactopyranoside. A 0.01 M solution of p- nitrophenyl—B-g—galactopyranoside was prepared in citrate- phosphate buffer, pH 3.5, according to Gomori (178). This substrate solution (0.1 ml) was added to 0.1 ml of enzyme solution and 0.3 m1 of citrate-phosphate buffer, pH 3.5. The solution was incubated for 2 hours at 25°C and the assay was terminated by the addition of 0.5 m1 of saturated sodium borate, pH 9.7. The optical density was measured at 410 nm as previously described. Protease was assayed, using technical grade casein purified according to Dunn (179) as substrate, as described by Liener and Friedenson (180). A 1% casein solution in 0.1 M sodium phosphate buffer containing 0.007 M mercaptoethanol and 0.001 M EDTA was prepared by heating the 1% casein sus- pension in a boiling water bath for 10-15 minutes to bring about complete solution of the casein. Aliquots (0.1 m1) of the enzyme solution to be assayed were placed in conical centrifuge tubes and diluted to 1 ml with phosphate- mercaptoethanol-EDTA buffer, pH 7.0. Casein solution (1 ml) was added and the tubes were incubated at 37°C for exactly 20 minutes. The assays were terminated by the addition of 3 ml of 5% trichloroacetic acid and allowed to stand for 1 hour at room temperature before centrifugation at 1300xg for 20 minutes in a Sorvall GLC-l clinical centrifuge. The optical density of the supernatant was read at 280 nm in a Gilford 2400 spectrophotometer. The readings were corrected 71 for blank values. The blanks were prepared by mixing 1 ml of casein solution with 3 ml of 5% trichloroacetic acid and then adding 1 m1 of phosphate-mercaptoethanol-EDTA buffer or an aliquot (0.1 ml) of enzyme solution diluted to 1 m1. PURIFICATION OF a-GALACTOSIDASES FROM COHN FRACTION IV-l The a-galactosidases from Cohn fraction IV—l, prepared by Method 6 of the low temperature ethanol fraction of plasma described by Cohn gt_al. (181), were purified in a three step process which included n—butanol extraction, acetone precipitation and affinity chromatography. Cohn fraction IV-l (2009) was placed in a stainless steel Waring blender with n-butanol (350 m1) at 4°C adjusted to pH 4.5 using 0.1N HCl. The mixture was dispersed for 1 hour at 4°C at slow speed using a variable power supply to regulate speed. After 1 hour, undissolved material was re- moved by suction filtration on a Buchner funnel with Whatman No. 3 filter paper moistened with n-butanol. The filtrate (350 ml) was removed and cooled to -20°C. Acetone (525 m1) at -20°C was added slowly through a separatory funnel with slow stirring over a period of 1 hour, to a final concentra- tion of 60% acetone. The precipitate was removed by centri- fugation at -20°C at 5,000xg for 10 minutes. The precipi- tate was redissolved in citrate phosphate buffer, pH 6.5, prepared according to Gomori (178), containing 5mg/ml sodium cholate and made up to 8% n-butanol. The solution was sonicated in an ultrasonic cleaner filled with water and ice at 4°C for several 5-10 second intervals. The enzyme 72 solution (50 ml) was filtered through an Amicon 0.45 micron filter and the clear solution was applied to a 10 x 100 mm affinity column packed with melibiose phenylhydrazone glass beads. The enzyme solution was passed through the column at a rate of about 1 ml/minute. The sclution was drained to the top of the column and the column was incubated at 4°C for 30 minutes. The column was waShed with 40 ml of citrate- phosphate-butanol buffer, pH 6.5 to remove unbound protein. The enzymes were then eluted with 80 m1 of 0.025 M sodium borate buffer with 8% butanol, pH 8.5. Eluent from the column was collected in 1 m1 fractions. ASSAYS Protein was assayed using Fluram as previously des- cribed. Ceramide trihexosidase assay solutions were made up to 8% butanol. When butanol solutions of ceramide trihexosi- dase were assayed, 0.08 ml of enzyme solution was used to limit the reaction mixture to 8% butanol. When enzyme solutions in 8% butanol buffers were assayed, 0.5 ml of enzyme solution was used and 0.04 ml of butanol was added to bring the reaction mixture up to 8% butanol. The reaction mixture contained 500 nmoles trihexosyl ceramide solubilized in 4 mg sodium cholate, 15 mg sodium tauroCholate, 8,8 mg sodium chloride and the appropriate volume of enzyme and butanol. Assays at pH 4.5 were brought to 1 m1 total volume using 0.05 sodium acetate buffer, pH 4.5. Assays at pH 6.5 were brought up to 1 m1 total volume with citrate-phosphate buffer, pH 6.5 prepared according to Gomori (178). The pH of the assays was adjusted using 0.1 N HCl or 0.1 N NaOH. 73 The reaction mixtures were incubated at 25°C for 4 hours and the assays were terminated by the addition of 2 m1 of chloro- form:methanol (2:1). The solutions were extracted and the galactose generated by the enzyme action was quantitated as previously described for ficin ceramide trihexosidase. Ceramide trihexosidase activity was also assayed using [3Hl-labelled trihexosyl ceramide prepared according to the method of Suzuki and Suzuki (182), as follows. An aliquot of the glycolipid was placed in a large screw-cap test tube with 4 ml of freshly distilled tetrahydrofuran and 4 ml of 0.1 M potassium phosphate buffer, pH 7.0. Galactose oxidase (427 units) dissolved in 0.5 ml of 0.1M potassium phosphate buffer, pH 7.0, was added and the solution was incubated at room tem- perature for 4 hours with gentle shaking. After 4 hours an additional 427 units of galactose oxidase dissolved in 0.5 m1 of 0.1M potassium phosphate buffer, pH 7.0, and the incuba- tion was continued overnight. Tetrahydrofuran was then re- moved from the solution under a stream of nitrogen, and 5 volumes of chloroform-methanol 2:1 (v/v) were added. The upper phase was removed and the lower phase was washed once with pure theoretical upper phase (chloroform-methanol-water 3:48:47 (v/v/v) (183)). The sample was dried and 5 m1 of tetrahydrofuran, [3H]-sodium borohydride (lOmCi/ml in 0.1N NaOH) (0.4 ml) was added. The sample was incubated at room temperature with shakin overnight. Excess borohydride was destroyed by the addition of 0.7 ml of 10N acetic acid in the hood. Tetrahydrofuran was then removed under a stream 74 of nitrogen and 5 volumes of chloroformrmethanol 2:1 (v/v) were added. The upper phase was removed and the lower phase was washed once with theoretical upper phase. The lower phase was dried under nitrogen and 5 ml of tetrahydrofuran and 10 mg of unlabelled sodium borohydride was added. The solution was incubated over night at room temperature with shaking. Excess borohydride was.destroyed by the addition of 0.7 ml of 10N acetic acid in the hood. Tetrahydrofuran was then removed under a stream of nitrogen and 5 volumes of chloroform-methanol 2:1 (v/v) were added. The upper phase was removed and the lower phase was washed 10-15 times with theoretical upper phase. The [3H]-1abe11ed glycolipid was purified by preparative thin layer chromatography in chloroform—methanol-water 65:25:4 (v/v/v). The labelled gly- colipid was located by scanning the thin layer plate on a Varian Aerograph Berthold Radio Scanner. The glycolipid was eluted from the silica gel scrapings with chloroform- methanol-water 100:50:10 (v/v/v). Radioactivity was determined with a Beckman LS-150 liquid scintillation counter. Scintillation solvent was pre- pared according to Suzuki and Suzuki (182) and contained 79 PPO (2, S-diphenyloxazole), 0.69 of dimethyl-POPOP (1,4-bis- 2-[4-methy1-5-phenyloxazoly1]-benzene), and 100 ml of Bio- Solv BBS-3 in 1000 m1 of toluene. Prior to the addition of scintillation solvent the sample was dried under a stream of nitrogen and redissolved in 0.5 m1 of water. Ten milliliters of scintillation solvent were added and the sample was counted. 75 Ceramide trihexosidase assays with the [3HJ-labelled glycolipid were conducted as follows. Trihexosyl ceramide (100 nmoles) diluted to 30,000 cpm with unlabelled glyco- lipid, 3.75 mg sodium taurocholate, 1.0 mg sodium cholate, 2.2 mg sodium chloride, 20 ul butanol extract or 50 ul of aqueous exzyme solution, 50 ul of 0.5M sodium acetate, pH 4.5, or citrate-phosphate buffer prepared according to Gomori (178) and water up to a final volume of 250 p1 was incubated at 37°C for 2 hours. Carrier galactose (250 nmoles in 250 pl) was added and the assay was terminated by the addition of 4 volumes of chloroform-methanol 2:1 (v/v). The upper phase was removed and the lower phase was washed once with pure theoretical upper phase. The upper phases were pooled and washed once with chloroform and then trans- ferred to scintillation vials. The upper phases were dried under a stream of nitrogen and counted as previously des- cribed. RESULTS PREPARATION OF AFFINITY ADSORBENTS The preparation of the affinity adsorbents was monitored qualitatively throughout the synthesis of the ligand. The coupling of putrescine to cyanogen bromide- activated dextran-coated glass beads was confirmed by com- plex formation of the amino group with methyl orange (p- [[(p-dimethy1amine) phenyl]azolbenzenesulphonic acid sodium salt). An aliquot of glass beads was placed in a screw-cap test tube and 1-2 ml of 20% aqueous methyl orange solution was added. The tube was sealed and heated at 80°C for 15 minutes. The beads were removed and washed exhaustively with distilled water. The presence of putrescine bound to the glass beads was confirmed by the orange color retained by the beads. Uncoupled beads did not retain the orange color. The preparation of p-nitrobenzamide glass beads was confirmed using anhydrous aluminum chloride reagent. An aliquot of the p-nitrobenzamide glass beads and 1-2 ml of chloroform were placed on a watch glass. Anhydrous aluminum chloride was sprinkled lightly over the glass beads. The aromatic p-nitrobenzamide glass beads became yellow with the addition of anhydrous aluminum chloride. A yellow color was also obtained by dissolving p-nitrobenzoyl chloride in chloroform and adding anhydrous aluminum chloride. 76 77 Putrescine-glass beads tested with aluminum chloride re— mained white. The reduction of p-nitrobenzamide glass beads to p— aminobenzamide glass beads was monitored with methyl orange as previously described. The diazotization of the arylamine glass beads was confirmed using a solution of B—napthol (550 mg) in 10 m1 of 10% sodium hydroxide. Diazotized glass beads became bright red under these conditions due to the formation of a diazo- dye analogous to para-red. The reduction of the diazotized glass beads to phenyl- hydrazine glass beads was quantitated using a 0.05 M aqueous copper sulphate solution. Approximately lg of phenylhydra- zine glass beads, weighed exactly, was placed in a screw- capped test tube and 3 m1 of 0.05 M copper sulphate was added. The test tube was sealed with a Teflon-lined cap, sonicated to dislodge air bubbles from the porous glass, and heated at 80°C for 30 minutes. Nitrogen gas is generated by this reaction, and’the precipitation of copper metal is also observed. The tube was removed from the oven, cooled, and centrifuged to obtain a clear copper sulphate supernatant. Care must be taken that the tube does not leak when heated to 80°C, or the concentration of copper sulphate will in- crease, resulting in erroneous computations. The optical density of the supernatant solution at 830 nm was determined using a Gilford 2400 spectrophotometer. Copper sulphate standards from 0.01 M-0.l M were used to quantitate the 78 change in concentration of the copper sulphate solution as a result of oxidation of the phenylhydrazine glass beads. The standard plot of optical density at 830 nm at increasing concentration of copper sulphate is shown in Figure 4. The phenylhydrazine bound to glass beads was found to be 4.2 nmoles/g glass beads. Melibiose bound to glass beads in the final product was quantitated by placing approximately 250 mg of melibiose glass beads, weighed exactly, in a screw-capped test tube with mannitol as an internal standard. The melibiose glass beads were hydrolyzed using 3 m1 of 0.75 N methanolic HCl and the liberated methyl glycosides were silylated and quantitated by gas-liquid chromatography on a 6 foot 5% SE-30 column at 170°C according to the method of Vance and Sweeley (184). The methyl glycosides obtained from 24 hours of acid methano- lysis and silylation with 0.01 ml of pyridine-hexamethyl- disalizane-trimethychlorosilane 4:4:2 (v/v/v) are shown in Figure 5. The recovery of sugars from acid methanolysis in 0.75 N methanolic HCl at 80°C for 6 and 24 hours are shown in Table 1. After 6 hours of incubation, approximately 91% of the galactose was cleaved from the glass beads, but only 45% of the dextran surface was hydrolyzed, compared with the results of 24 hours of incubation. The resistance of the dextran surface to 6 hours of acid-catalyzed methanolysis is an indication of the stability of this support. 79 .mpme mmmHm mcflumupmnawcmnm ou momma mmmam @mNHUONMfiU mo coauosomu on» mumuwusmsv on own: was coflumuucmocoo mumnmasm Hmmmoo mcwmmmHOCw nuHB Es 0mm um anamsmo Havaumo mo uoam onmosmum use COHumnucmocoo mumsmasm Hmmmoo mo uon pumocmum .v wusmflm 80 op.o mo.o z ”cowsu. 8.0 «0.0 No 0 . a . ~.o ¢.o o.o m.c c.— 08800 81 .Oooha um HmEHmQHOmH afisaoo omsmm mm uOOM w m auflz oomufisqm camnmoumeouno mmm cow 2 can m cm nuH3 omnaamcm can A>\>\>V Nueuw mcmHHmouoHsoahsumEHuu ImcmNmemfivahsumemxmnimcwpwumm mo HE H.o nuflz omumamawm wumz Amy momma mmmHm mcwnmuowsHmcmam can Adv momma mmmam mconuomnamcmam mmoHnHHoE Eonm mopfimooham Hmsumfi was momwm unwao Eoum muofimoomaw Hanumz HhHHmHmQHGEHHB mo mammamoumfioucu UHSUHA mow .m wusmfim Response Detector 82 M (1) RS M inuies be Minutes 83 Table 1. Recovery of Sugars From Acid Methanolysis of Melibiose Phenylhydrazone Glass Beads The recovery of sugars from acid catalyzed methanolysis of melibiose phenylhydrazone glass beads was examined after 6 and 24 hours of incubation in 0.75 N methanolic HCl at 80°C. The products were silylated and analyzed with an F and M 400 gas chromatograph equipped with a 6 foot 5% SE-30 column isothermal at 170°C. — . ~. 84 Hydrolysis of Melibiose Phenylhydrazone Glass Beads nmoles c1eavedqper_g_g]ass beads 6 hours 24 hours Galactose 3.60 3.95 Glucose 4.90 10.95 85 PURIFICATION OF FICIN a-GALACTOSIDASES The standard curve used in the protease assay, performed with ficin is shown in Figure 6. The assay was linear over the range of ficin concentration examined. Assays were typically performed with 0.1 m1 of ficin solution. The a-galactosidase assay performed with p—nitrophenyl- ajg-galactopyranoside is shown in Figure 7. It was necessary to dilute crude ficin solutions to 50% of their original con- centration with 0.05 M sodium acetate buffer, pH 4.5 to avoid deviations from linearity at higher concentrations of ficin. The B-galactosidase assay performed with p-nitrophenyl-a-g- galactopyranoside is shown in Figure 8. It was not necessary to dilute crude ficin solutions to obtain linearity with B-galactosidase: ficin contains approximately 80 times more u-galactosidase than B-galactosidase. The ceramide trihexosidase activity in ficin is shown in Figure 9. The lipoidal product of ceramide trihexosidase activity, in the lower phase from the Folch partition of the ficin ceramide trihexosidase assay, was examined by thin- layer chromatography on a silica gel G plate using chloro- form-methanol-water 65:25:4 (v/v/v). As shown in Figure 10, exposure of the plate to iodine vapor revealed that lactosyl ceramide was present in the lower phase of the enzyme assay but not in the assay blank. The affinity chromatographic purification of a-galacto- sidases and ceramide trihexosidases from crude ficin is shown in Figure 11 and is summarized in Table 2. Eluent from the 86 .Uocflfimxm mason soflumuucmocoo on» Ho>o nausea on on UQSOM mm3 paw .aoflumuucoocoo sfiofim mCHmmmuosH £ua3 omcHmem mmz CHUHM mouse «0 muw>fluom mummuoum one ammm< mmmmpoum mo uOHm Unmocmum .m wusmwm 2.5 52... 25.5 m.o ~.o m.o m.o ¢.o m.o N.o p.o _ . m . q _ a . 08300 88 .mowmosmu>m0pomHmmIMIdlamcmnmouuflcum mumuum Insm HMAUMMfluum may mean: coflumuucoocoo seven msflmmouocw nuflz omcfiemxo mm3 afloam mosuo mo mufl>fluom mmmoflmouumflmmua one modmocmummouomHmulmlcnamconmouuHZIm mcflma mmmm< ommowmouomamwla uon oumocmum .h musmflm 89 0~.0 0P.0 0~.0 ep.0 apes crust «uses 2... 25 00.0 00.0 00.0 «0.0 q q - om 00F 00F 00m 0mm Ju/pazfiloupfiu leg-o-dud salomu 90 .oofimocmnhmouomammlmlm chmsmouuflcum muouum loam Hmflowmflunm mnu mcflms cofiumuusmocoo GHOHM mcflmmmuocfl saws oocflsmxm mma cfiofim oosuo mo >ue>wuom mmmoflmouomammlm one woflmocmummouomamwlmlmIahcmnmouuMZIm mchD >mwm¢ wmmoflmOHUMHmwlm mo uoam oucosmum .0 musmflm 9 1 J J I l l L O In C N '- '- Ju/pazfilwpm 109+de salon" 3.0 2.5 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 0.02 Crude Ficin (m1) 92 .mowemumo Hamoxosfluu wumuumnsm Hmnsum: map mcflms coflumnucmocoo afloflm mcwmmmuoca spas omcwemxm mm3 cwoflm mosuo mo huw>wuom mmmowmoxmnwuu mowfimumo use woflEmumO ammomeHuB moan: humus ommoflmoxmnflue moHEmumU mo scam oumccmum .0 musmflm 80)- l O IN 93 l I l I I O O O O O to m d’ m N Jq/asealau asoioeleg salomu 0.25 0.20 0.15 0.10 0.05 Crude Ficin (m1) 94 Figure 10. Thin Layer Chromatography Plate of Ficin Ceramide Trihexosidase Assay The lower phase from the Folch partition of ficin ceramide trihexosidase assay was examined by thin layer chromatography on a silica gel G plate using chloroform- methanol-water 65:25:4 (v/v/v). The lipids were visualized by staining with iodine vapors. 9-5 Lactosyl Cor-Me Tri hexosyl Cor-i de Origin Ficin Standards Ass Ceramide Monty Tri hexosidase Activity 96 .mcofluomnm HE a a“ omuomaaoo mm3 mumsam one .m.0 mm .uommsn mumuon EsfloOm 2 mm0.0 cues soap .HMfiHoqu mason Is: o>oEmu op Hummus mumuoom nuw3 nousaw mos aEsHoo one .Hmmmsn wsmm was Buss ompmunsanswm some can pass Ass cm x mo cesaoo muwcflmmm oconuomsHmsocm mm0flnwame on» 0» vowammm ocm .m.v mm .Hommsn mumuwom ESHQOm z m0.0 a“ pm>aommflp mm3 Ame n.5m0 :Hon seven mo unmoumoum50uno muwcflmma .HH musmflm (Kosso asoemld) 092 00 Ju/pazKIOJpKu agousqns salomu Q m E g ' <2 8 - 1 3 31 8 g ":33 £8 .3; °_"=: Ase-{£8 . :33: (...? .8 §§88§ : frets '2 ' ¢ ¢ ¢ ¢ ¢ ‘1 . 48 - ., ’8 0‘: a: (r ‘ ‘0,‘_ — .——. ire . " g s 8 ° (NJ/5w) ugasmd ; 1 L J .3 a s 0 Fraction 98- .H£\ommmmHmH mmouomHmm moHoEc mum wufl>fluum mahucm mo muss: one .omNfiHmEESm ma mammamouofionno mufisflmmm an seven Eouw ommowmoxmnfluu mpflEmumo ocm mucoflmouomammla mo sofluwHOmH use CHUHm mo mammamoum50nso muflcflmmm .m manna 0.00P 0.0000 0.00 00.0 00 cowuuasm 0.00 0.000 0.000 00.0 00-00 meowuumgm 0.00 F.¢~ 0.000 00.0F mango mmmuwmoxmgvth vaEMLmQ 0.0P 0.Fu0 ~.eop PF.0 00 newscasm 0.¢_ 0.000 0.000 00.0 00-00 meowuumgm I. e.po 0.000 00.0F mango ommuwmopumpmuua ARV ‘ A0sxmuwc=0 covamuPewtsa 0u0>wpu< mupcsv A050 spat outsamam us>wsu< gusseta 0.00 0.000 0.000 mmmuwmoxmzwgh mvwsmsmu 0.00 0.00_ 0.00_ mmomuoga 0.00 0.000 0.000 ammvwmouum—muum m.FOP 0.000 0.000 Amupeav mmmuwmopumpm0-a ~.oop mo.os co.ms A020 crasora 0Lm>ouom & umtm>oomm uwwpqn< sgaarmosneorgu surestt< =_upi 100 column was monitored for protein, protease, a-galactosidase, B-galactosidase and ceramide trihexosidase activities. The protease and B-galactosidase activities were not retained by the column and were therefore recovered in the acetate buffer. The a-galactosidase and ceramide trihexosidase activities were retained by the column. Attempts to isolate the individual enzymes by elution with Triton X-100 were not successful since only broad, unresolved bands were obtained. When 0.025 M borate buffer, pH 8.5, was applied to the column three protein peaks were eluted. The first peak con- tained both a-galactosidase activity and ceramide trihexosi- dase activity. Subsequent protein peaks contained enzyme activity specific for either p-nitrophenyl-a-g-galactoprano- side or trihexosyl ceramide. Electrophoresis of proteins retained by the affinity column on 9% polyacrylamide gels with 1% sodium dodecyl sulphate is shown in Figure 12. The specificity of the melibiose phenylhydrazone liquid was examined using acetaldehyde phenylhydrazone glass beads as a control. When crude ficin solutions were applied to a column containing acetaldehyde glass beads (5 x 80 mm) all of the a-galactosidase and ceramide trihexosidase activities were eluted in the 0.05 M sodium acetate, pH 4.5 eluate, as shown in Figure 13. Low levels of protein were detected in the borate eluate from this column, but this protein fraction was not active toward a-galactosidase, 8- galactosidase, ceramide trihexosidase, or protease. 101 Figure 12. Sodium Dodecyl Sulphate Polyacrylamide Gel Electrophoresis of Ficin Samples obtained by affinity chromatography of ficin were examined by electrophoresis on 9% polyacrylamide gels with 1% sodium dodecyl sulphate. (A) crude ficin: (B) the first protein peak eluted from the affinity column with borate buffer; (C) the second peak from the affinity column: (D) the third peak from the affinity column. Gels were stained for protein with Coomassie G (192)- 102 | I l l ' I II I ._ . i II. I In! III III I‘ll! 103 .m.0 mm .memsn oumuon ESHGOm 2 000.0 £003 con» 0cm .Hommsn wumumom no as 00 £003 wwusHm mm3 :Edaou one .Hmmmsn 08mm may nuw3 @mumuoflafisvm soon own van» mummn mmmam mconuomnamcmnm ocmnwpamuoom mo AEE 00 x 00 :EsHoo m on Umflammm can .m.v mm .nmmmsn oumumom Esflcom 2 00.0 CH pm>aommflo was A05 m.000 :Howm momom mmmaw oconHomsawcmcm momnooamumoa co cfiowm mo mammamoumfiounu .ma musmflm 104 00. m C? (Aosso esoemd) 082 go o, ‘52- f mmoEmoxmfszEEoo Iol 08288200 - 0». LT $8.20 An. 23 0820200. 00 .5 £220 _ _ p IQII IOI + 00:02... 0.010 39.3 62000 EmNO N O V. O (Iw/fiw) mama 00 L 0 ID L E} 0 Q m/ pezKIOJqu ewusqns se|ou1u 105 PURIFICATION OF CERAMIDE TRIHEXOSIDASES FROM COHN FRACTION IV-l The ceramide trihexosidase assay performed with Cohn fraction IV-l is shown in Figure 14. The galactose cleaved from trihexosyl ceramide with increasing concentration of butanol extracted Cohn fraction IV-l was determined using a computer-assisted fluorimeter. Assays conducted with 0.01- 0.08 ml of butanol extract were adjusted to 8% butanol in the assay solution; assays with 0.01 ml, 0.15 ml and 0.20 ml of butanol extract were 10%, 15% and 20% butanol, respec- tively. Butanol is not soluble in water above 9.1% (185) and the assays conducted at, or above, 10% butanol were therefore biphasic. An experiment was performed to deter- mine whether the decreased ceramide trihexosidase activity in the biphasic assays with 0.15 and 0.20 ml of butanol ex- tract was due to solubility of substrate in the butanol phase, or solubility of ceramide trihexosidase in the aqueous phase. An assay was conducted with 0.20 ml of butanol extract (20% butanol) but just prior to incubation, the butanol phase was removed and incubated separately. Ceramide trihexosidase activity corresponding to the pre- viously observed activity was found in the aqueous phase but none was observed in the butanol phase. It was ob- served that the decreased ceramide trihexosidase activities obtained with 0.15 and 0.20 ml of butanol extract were 60% and 44%, respectively, of the anticipated ceramide trihexo- sidase activity (indicated by the broken line in Figure 14). 106 .0Hm>euommmmu .Hocmusn 000 can 00H .moa mum3 uomuuxo Hocwuso 00 as 00.0 0cm .ma.0 .0H.0 cues mammwm wee .Hocmusn 00 on woumsncm mnmz uomuuxm Hocmusn no He 00.0IH0.0 Eoum pmuosccoo mwmmmd .uomuuxm Hosanna mo mcoflumuucmocoo meemmmuocfl nufi3 nmcwmew mm3 HI>H cofluomsm c300 monomsuxouaocmuso mo wufl>fluom mmmvemoxmnauu moflemuwo one 0.4 00 pm 0mmm4 ommwfimoxonflue mUHEmumo 00 uon Unmocmum .0H musmflm 1(17 00.0 0—.0 0—.0 vp.0 Ape. uuotuxm poeuusm Np.0 0—.0 00.0 00.0 00.0 00.0 - q a q - 1— q « 0.0 0.0 0.0 0.0p 0.0— Ju/paneala 350132199 salomu 108 These values correspond to the theoretical proportion of butanol in the aqueous phases of these assays. The pH optima for ceramide trihexosidase activity from Cohn fraction IV—l is shown in Figure 15. Assays from pH 3.5—pH 5.0 were conducted in 0.05 M sodium acetate buffer: assays above pH 5.0 were conducted in citrate-phosphate buffer prepared according to Gomori (178). All assays con- tained 8% butanol. It was found that Cohn fraction IV—l ceramide trihexosidase exhibited two optimal activities at pH 4.5 and pH 6.6. The product of ceramide trihexosidase activity at pH 4.5 and pH 6.5, recovered in the lower phase from the Folch partition of the Cohn fraction IV-l ceramide trihexisodase assay, is shown in Figure 16. The preparation of 3H-labelled trihexosyl ceramide was monitored by thin—layer chromatography on silica gel G plates in chloroform-methanol-water 65:25:4 (v/v/v) and stained by exposure to iodine vapor. The 3H-labelled trihexosyl cera- mide was scanned on a Varian Aerograph Berthold strip scanner as shown in Figure 17. The only radioactive material detected by strip scanning was trihexosyl ceramide. The amount of 3H-labelled trihexosyl ceramide recovered from this procedure was quantitated by gas-liquid chromatography according to the method of Vance and Sweeley (184), as shown 3 in Figure 18. The specific activity of the H-labelled lipid was 26,768 cpm/nmole trihexosyl ceramide. 109 Figure 15. The pH Optima for Ceramide Trihexosidase The pH optima for ceramide trihexosidase from Cohn fraction IV-l was determined using a computer-assisted fluorimeter. Assays from pH 3.5-pH 5.0 were conducted in 0.05 M sodium acetate buffer: assays above pH 5.0 were conducted in citrate phosphate buffer prepared according to Gomori (178). 111) 10.0 P 9.0 " — P o o o o 7 6 8.0 ' nextsxmumaopaa «mono-poo mo—osc 5.0 " 4.0 T 8.0 7.0 6.0 5.0 4.0 3.0 111 Figure 16. Thin Layer Chromatography Plate of Cohn Fraction IV—l Ceramide Trihexosidade Activity The lower phase from the Folch partition of ceramide trihexosidase assays at pH 4.5 and 6.5 were examined by thin layer chromatography on a silica gel G plate using chloroform-methano1-water 65:25:4 (v/v/v). The lipids were visualized by staining with iodine vapors. 112 Cerebroside Lactosyl Ceramide Trihexosyl Ceramide Globoside Standards Ceramide Blank Ceramide Blank Trihexosidase pH 4.5 Trihexosidase pH 6.5 Activity Activity pH 4.5 DH 6.5 113 Figure 17. Thi -Layer Chromatography and Strip Scan of H-Labelled Trihexosyl Ceramide 3H—Labelled trihexosyl ceramide prepared according to the method of Suzuki and Suzuki (182) was examined by thin- layer chromatography on a silica gel G plate using chloroform-methanol-water 65:25:4 (v/v/v). The plate was 3 scanned for H-labelled material on a Varian Aerograph Berthold strip scanner. 1Q14 Cerebroside Lactosyl Ceramide _, c '9: mm ' Trihexosyl Ceramide idifl? )Y‘L“ Globoside ,8 1133 ph Origin Standards Trihexosyl Ceramide 115 Figure 18. Gas—Liquid Chromatography of Silylated Methyl Glycosides from 3H—Labelled Trihexosyl Ceramide The recovery of trihexosyl ceramide from the 3H- 1abelling-procedure of Suzuki and Suzuki (182) was determined by gas-liquid chromatography. The methyl glycosides were silylated with 0.1 ml of pyridine- hexamethyldisilzaane-trimethy1chlorosi1ane 4:4:2 (v/v/V) and analyzed with an F and M gas chromatograph equipped with a 6 foot 5% SE-30 column maintained at 170°C. 116 (D U) C U) (D Y 0f Silylated 3H-Labelled E L— nide from the 3H- 2 O -. lki (182) was a) a 1Y- The methyl 0) Of pyridine— lane 4:4:2 (v/v/v) O iatograph equipped ed at 170°C. U L l 1 I l I . O 4 8 I2 M inU‘I’eS 117 Affinity chromatography of partially purified ceramide trihexosidase from Cohn fraction IV-l is shown in Figure 19, and the purification is summarized in Table 3. a—Galactosid- asses active toward the artificial substrate, p—nitrophenyl- atg-galactopyranoside, were not present in butanol extracts of Cohn fraction IV-l, and were not detected in eluent from the affinity column. The specificity of the melibiose phenylhydrazone ligand was examined using acetaldehyde phenylhydrazone glass beads as a control. When the redissolved acetone precipitate was passed through a column of acetaldehyde phenylhydrazone glass beads (5 x 80 mm) no protein was retained by the column, and all of the ceramide trihexosidase activity was recovered in the citrate-phosphate eluate from the column. An experiment was performed to determine whether the ceramide trihexosidases retained by the melibiose phenyl- hydrazone affinity adsorbent were eluted due to the pH change from the citrate-phosphate buffer, pH 6.5, to the borate buffer, pH 8.5, or due to the borate itself. The re- dissolved acetone precipitate was applied to the melibiose phenylhydrazone glass beads and washed with citrate- phosphate buffer, pH 6.5, containing 8% butanol as before. However, the column was then washed with 0.1 M Tris-HCl, pH 8.5, containing 8% butanol instead of with borate buffer. The ceramide trihexosidases were eluted from the affinity adsorbent by the Tris-HCl buffer in a pattern identical to that obtained by elution with the borate buffer. 118 Figure 19. Affinity Chromatography of Cohn Fraction IV-l Affinity chromatography of the acetone precipitate from 200 g of Cohn fraction IV-l redissolved in 50 ml of citrate-phosphate buffer, pH 6.5 prepared according to Gomori (178), with 0.5% sodium cholate and 8% butanol. The column was washed with 40 ml of the citrate- phosphate buffer to remove unbound material, and then with 70 ml of 0.025 M sodium borate buffer, pH 8.5, with 8% butanol. 119 8 Jq/DSSDGIGJ 950430106 selowu § c_®_.o._n_ .181 0.010 082865182828 Iol 0.310 082862302828 Iol II- c288“. 0.0 Ia .823 $0 2200 8308 2000.0 J ON. (501) manna 120 0.0 :0 a. 000 0 000.0 000 00p.0 xup>puu< 000 peace 0.0 :0 an 000 0 000.0 000 00,.0 00w>Puu< 00» p000» 02000000000000 0000000< .000 0 P0 0__ 000.0 0.0 :0 00000000000 0 00 00 000.0 0.0 :0 0.00 0000000 000 .00 00 000.0P 0.0 :0 p 00_ 00 000.0 0.0 :0 0.000 co_uumgux0 —0cmu=0 .H atgx0e\mmpoecv 000000000030 zuw>puo< Ag:\mm_0e:0 A000 0—00 000>0000 a upmmuw0m 000>Puu< 0000000 0000 pu>0 00000000 0:00 500» ommnpmcxocwge 0 0050000 00 000000000000 .0 0.000 I illllllllllllll l I! .l i . l DISCUSSION In all previous work on the affinity chromatography of d-galactosidases, agarose was employed as the support for the affinity ligands. The use of this support for pilot-scale enzyme isolation posed several problems. Agarose is a poly- saccharide and the possibility of bacterial contamination could result in contamination of the isolated enzymes with pyrogens, making them unsuitable for intravenous injections. The agarose beads cannot withstand autoclaving or ethanol treatment to prevent bacterial growth. In addition, the pilot-scale isolation of ceramide trihexosidase will involve passing large volumes of enzyme solutions through the affi- nity column, and agarose beads could compress under this pressure, limiting the flow rate of the column. For these reasons glass beads rather than agarose beads were employed as the support for the melibiose phynylhydrazone ligand. Glass beads are stable in organic solvents such as ethanol and can therefore be maintained sterile. The glass beads are also rigid, and will not compress under pressure, threeby ensuring consistent flow rates through the column. A disadvantage of glass beads is the possibility of nonspecific, ionic bonding of proteins to the polar surface 121 122 of the glass. When ficin was applied to affinity column prepared from y-aminopropyltriethoxysilane—substituted glass beads (not dextran—coated), nonspecific adsorption of pro- tein to the exposed surface of the support interfered with the isolation of a-galactosidases from the column. The a- galactosidases were isolated in sharp peaks against a high background of protein that gradually eluted from the column (188). The use of dextran-coated glass beads alleviated this problem by effectively shielding the polar glass Sur- face of the affinity adsorbent from the medium. The amount of melibiose immobilized on the dextran- coated glass beads (3.95 “moles/g glass beads) was determined by gas-liquid chromatography as previously described. However, it was convenient to assay qualitatively for the presence of various functional groups during the preparation of the affinity adsorbent. Methyl orange, anhydrous alumi- num chloride, and sodium B-naptholate were used to determine primary amines, aromatic rings and phenyldiazonium salts, respectively. Previously, sodium 2,4,6-trinitrobenzene- sulphonate (picryl sulphonate) was employed as an indicator for primary amines, unsubstituted hydrazides, carboxylic acids and bromoacetyl derivatives of agarose (175). This reagent, however, requires a 2 hour incubation at room tem- perature for the reaction to be completed, whereas reactions with the presently employed reagents are complete in 10-15 minutes. Furthermore, the colored derivatives produced with picryl sulphonate are not specific for particular functional 123 groups as yellow color is produced with unsubstituted agarose, carboxylic acids and bromoacetyl derivatives. The quantitation of phenylhydrazine immobilized to glass beads was developed as a monitor for the amount of acetalde- hyde that could be substituted onto glass beads. Acetaldehyde phenylhydrazone glass beads (4.2 nmoles/g glass beads) were prepared as a control to determine the specificity of the melibiose phenylhydrazone affinity adsor- bent. This product was used rather than unsubstituted dextran-coated glass beads so that the possibility could be tested of binding to the extender arm, rather than binding specific for the melibiose phenylhydrazone. When ficin was applied to a column containing acetaldehyde phenylhydrazone glass beads, low levels of protein were retained by the column, but a-galactosidase and ceramide trihexosidase activities were not retained by the column, as shown in Figure 13. The low level of protein retained by the column may be due to nonspecific binding to the extender arm, or to the glass surface of the column. Ficin contains high levels of protease, and consequently crude ficin contains many low molecular weight proteins as shown by electrophoresis (see Figure 12). It may be that these small proteins pass be- tween the dextran and the glass surface of the affinity support and are retained by the adsorbent. Affinity chromatography of ficin on melibiose phenyl— hydrazone glass beads resulted in the retention of three protein peaks (see Figure 11), similar to the elution 124 profile for ficin obtained by Mapes and Sweeley (72). Sodium dodecyl sulphate polyacrylamide gel electrophoresis revealed that none of the protein peaks consisted of a single homo- genous protein. The conclusion that multiple bands on electrophoresis indicate impurities may not be entirely accurate, however, in View of several recent reports indica- ting that subunits in glycosidases (99,103,189) often occur and can be separated electrophoretically (99). For example, in 1974, Srivastava gt_§l, examined hexosaminidase A and B from human placenta by sodium dodecyl sulphate-guanidine hydrochloride polyacrylamide gel electrophoresis and urea- starch gel electrophoresis (99). When the purified placental hexosdaminidases were examined by electrOphoresis on native polyacrylamide gels each isozyme migrated as a single, homo- geneous protein band (100). However, when the hexosamini- dases were examined by sodium dodecyl sulphate-guanidine hydrochloride electrophoresis and urea-starch gel electro- phoresis hexosaminidase A dissociated into three protein bands, and hexosaminidase B migrated as two protein bands. Harpaz §£_al. (187) found three a-galactosidase isozymes in coffee beans when eluate from their affinity column was examined on native polyacrylamide gels. Had these isozymes been examined by sodium dodecyl sulphate poly- acrylamide gel electrophoresis, perhaps several protein bands would have been detected. In 1970, Mapes §E_gl. reported that ceramide trihexosi- dase activity in human plasma was optimal at two pH's, pH 5.4 12S and 7.2 (71) and that q—galactosidase activity in plasma was optimal at pH 3.0 and 6.0 (72). These values are different from single pH optimum reported by other workers (69, 70, 76, 78) which was more acidic than those reported by Mapes et_al. (ranging in human sources of enzyme from pH 3.5-4.6). None of the latter values were obtained with ceramide tri- hexosidase from human plasma. I have obtained a bimodal pH optimum for ceramide trihexosidase activity in Cohn frac- tion IV-l (see Figure 15); however, the more acidic value (pH 4.6) is in agreement with the pH optima obtained by other workers for human placenta ceramide trihexosidase (69) and human kidney ceramide trihexosidase (78). The ability of n-butanol to solubilize and stabilize ceramide trihexosidase was first reported by Mapes gt_al. in 1973 (73). Advantage has been taken of this unusual property in the isolation of ceramide trihexosidase from Cohn fraction IV-l as well as the observation that ceramide tri- hexosidase is relatively stable in this organic solvent. The initial velocity of ceramide trihexosidase was examined with increasing concentrations of crude enzyme in the butanol extracts (Figure 14). Decreased rates were observed in biphasic assays (> 0.10 ml butanol extract) and it was concluded that the observed rate was due to enzyme activity in the aqueous phase and not the total activity in the added butanol, since the decreased ceramide trihexosi- dase activity in these biphasic assays was proportional to the butanol concentration in the aqueous phase. The assay 126 with 0.15 ml of butanol extract should theoretically consist of 0.09 ml of butanol in the aqueous phase (60% of the ali- quot of butanol extract added) and 0.06 ml of butanol in the upper phase. The observed rate of hydrolysis of trihexosyl ceramide was 60% of the anticipated value (indicated by the broken line in Figure 14). The assay with 0.20 ml of butanol extract should theoretically contain 0.09 ml of butanol in the aqueous phase (45% of the aliquot of butanol extract added) and 0.11 ml of butanol in the upper phase. The observed rate of hydrolysis of trihexosyl ceramide was 44% of the anticipated value. This indicates that the cera- mide trihexosidase partitions exactly as the butanol does, suggesting that butanol-ceramide trihexosidase micelles may be present in the aqueous phase and are responsible for the observed ceramide trihexosidase activity. The use of butanol to extract ceramide trihexosidase proved to be a much more specific method of extraction com- pared with the aqueous extraction employed by Mapes §E_§1. (73). The specific activity of the butanol extract was approximately BOO-fold greater than that of the aqueous extract. Difficulties were encountered in redissolving the acetone-precipated ceramide trihexosidase. Sodium cholate has been used successfully to solubilize glycosphingolipid hydrolases (40, 41, 45, 51, 61), and was employed in this research to redissolve the acetone precipitate. Recently, it was found that 0.1 M urea is also effective in redis- solving this fraction (190). 127 Affinity chromatography of the redissolved acetone pre- cipitate resulted in an elution profile very similar to that obtained by Mapes and Sweeley (72) for whole human plasma; however, no a-galactosidase activity toward the artificial substrate, p-nitrophenyl-a-g-galactopyranoside, was detected. The specific activities of the ceramide trihexosidases obtained by affinity chromatography were approximately 2- fold greater than those observed by Mapes and Sweeley (72). This finding may reflect the difference in pH, or other assay conditions, employed in the present research. The recovery of the ceramide trihexosidases from the affinity column were relatively low, perhaps because of relative instability of the enzymes at relatively high pH (human spleen ceramide trihexosidase is most stable at 2pH 6.5 (191)) or the absence of stabilizing protein, such as albumin, removed by affinity chromatography. The conditions under which the acetone-precipitated ceramide trihexosidase is redissolved may not be conducive to the survival of the enzyme prior to affinity chromatography and this may be a step where the overall yield can be improved. It was found that 0.1 M Tris HCl, pH 8.5, was as effective as 0.025 M sodium borate, pH 8.5, for elution of the ceramide trihexosidase activity from the affinity column. This finding indicates that the change in pH from 6.5 to pH 8.5 is responsible for elution of the enzymes from the affinity adsorbent, rather than complexation of the borate with carbohydrates on the surface of the enzyme, or with the affinity adsorbent. 128 The application of this method of isolating ceramide trihesosidases from Cohn fraction IV-l to a pilot-scale operation can provide 2,500 units of ceramide trihexosidase activity per kilogram of Cohn fraction IV-l at pH 4.5 and 6.5. If methods of improving the overall yield of the pro- cedure are found, one kilogram of Cohn fraction IV-l could provide up to 50,000 units of ceramide trihexosidase activity at pH 4.5 and 6.5. The theraputic value of ceramide trihexosidase isolated by this method for enzyme replacement therapy could be considered. As previously discussed, plasma may be the best source of ceramide trihexosidase for treatment of Fabry's disease due to the extended lifetime of ceramide trihexosidase from plasma in circulation (approximately 160 hours (168)), and decreases in levels of trihexosyl ceramide in patients treated with ceramide trihexosidase from plasma (168). However, the theraputic value of ceramide trihexosid- aseas isolated by the present method may not coincide with those of whole plasma, previously used to treat patients with Fabry's disease. BIBLIOGRAPHY 10. 11. 12. 13. 14. 15. BIBLIOGRAPHY Thudichum, J. L. W., Report Med. Off. Priv. Council, New Series, No. 3, Appendix 5, 113 (1874) Thudichum, L. L. W., Die Chemische Konstitution des Gehirns des Munshen und der Tiere, F. Pietzcher, Tubingen (1901) Carter, H. E., Glick, F. J., Norris, W. P. and Phillips, G. E., J. Biol. Chem. 170, 285 (1947 Thierfelder, H., and Klenk, E., Die Chemie der Cerebroside und Phosphatide, J. Springer, Berlin (1930) Levene, P. N., and West, C. J., J. Biol. Chem. 16, (1914) Klenk, E., Ztscher. Pysiol. Chem.235, 24 (1935) Nakayama, T., J. Biochem. 31, 309 (1950) Carter, H. E., and Greenwood, F. L., J. Biol. Chem. 199, 283 (1952) Fujino, Y., and Negishi, T., Bull. Agr. Chem. Soc. Jap. 29, 183 (1956) Rosenberg, A., and Chargoff, E., J. Biol. Chem. 233, 1323 (1958) Carter, H. E., Rothfus, J. A., and Gigg, R., J. Lipid Res. 2, 228 (1961) Klenk, E., and Rennkamp, R., Z. Physiol. Chem.273, 253 (1942) Rapport, M. M., Graf, L., and Alonzo, N. F., Cancer 11, 1136 (1958) Rapport, M.M., Graf, L., Skipski, V. P., and Alonzo, N. F., Cancer 12, 438 (1959) Graf, L., and Rapport, M. M., Cancer Res. 20, 546 (1960) 129 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 3O 31. 32. 33. 34. 35. 130 Rapport, M. M., Graf, L., and Yarif, J., Arch. Biochem. Biophys. 92, 438 (1961) Rapport, M. M., Graf, L., and Alonzo, N. F., J. Lipid Res. 1, 301 (1960) Makita, A., and Yamakawa, T., J. Biochem. 55, 124 (1962) Yamakawa, T., Kiso, N., Handa, 8., Makita, A., and Yokoyam, S., J. Biochem. 55, 226 (1962) Svennerholm, E., and Svennerholm L., Nature 198, 688 (1963) Sweeley, C. C., and Klionsky, B., J. Biol. Chem. 238, 3148 (1963) Sweeley, C. C., and Klionsky, B., Abstracts Sixth Int. Congr. Biochem., New York (1964) Makita, A., and Yamakawa, T., J. Biochem. 55, 365 (1964) Hakomori, S. I., Siddiqui, B., Li, Y. T., Li, S. C., and Hellerqvist, C. G., J. Biol. Chem. 246, 2271 (1971) Li, T. T., Li, S. C., and Dawson, G., Biochim. Bithys. Acta 260 88 (1972) Klenk, E., and Launstein, K., A. Physiol. Chem. 288, 220 (1951) Yamakawa, T., and Suzuki, 8., J. Biochem. 59, 175 (1952) Yamakawa, T., and Iida, T., Jap. J. Exptl. Med. 23, 327 (1953) Yamakawa, T., Yokoyama, S., and Handa, N., J. Biochem. 55, 28 (1963) Rapport, M. M., Graf, L., and Schneider, L., Fed. Proc. 22, 498 (1963) Makita, A., J. Biochem. 55, 269 (1964) Makita, A., Iwanaga, M., and Yamakawa, T., J. Biochem. 55, 202 (1964) Rapport, m. m., and Graf, L., Fed. Proc. 25, 176 (1965) Forsmann, J., Biochem. Z. 51, 78 (1911) Thannhauser, S. J., and Reichel, M., J. Biol. Chem. 113, 311 (1936) 36. 37 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 131 KOpaczyk, K. C., and Radin, N. S., J. Lipid Res. 5, 140 (1965) Hajra, A. K., Bowen, D. M., Kishimoto, Y., and Radin, N. S., J. Lipit Res. 1, 379 (1966) Suzuki, K., and Suzuki, Y., Proc. Nat. Acad. Sci. U. S. A. 55, 302 (1970) Austin, J., Suzuki, K., Armstrong, D., Brady, R., Bachhawait, B. K., Schlenker, J., and Stumpf, D., Arch. Neurol. 25, 502 (1970) Bowen, D. M., and Radin, N. S., Biochim. BiOphys. Acta 152, 587 (1968) Bowen, D. M., and Radin, N. S., Biochim. Biophys. Acta 152, 599 (1968) Suzuki, Y., and Suzuki, K., J. Biol. Chem. 249, 2098 (1974) Norden, A. G., Tennant, L. L., and O'Brien, J. S., 5; Biol. Chem. 249, 7969 (1974) Brady, R. 0., Kanfer, J., and Shapiro, D., J. Biol. Chem. 240, 39 (1965) Gatt, S., and Rapport, M. M., Biochim. Biophys. Acta 113, 567 (1966) Gatt, S., Biochem. J. 101, 687 (1966) Ho, M. W., and O'Brien, J. S., Proc. Nat. Acad. Sci. U. S. A. 55, 2810 (1971) Ho, M. W., O'Brien, J. S., Radin, N. S., and Erickson, J. S., Biochem. J. 131, 173 (1973) Brady, R. 0., Kanfer, J. N., and Shapiro, D., Biochem. Biophys. Res. Comm. $5, 221 (1965) Pentchev, P. G., Brady, R. 0., Hibbert, S. R., Gal, A. E., and Shapiro, D., J. Biol. Chem. 248, 5256 (1973) Gatt, S., and Rapport, M. M., Israel J. Med. Sci. 5, 624 (1965) Gatt, S., and Rapport, M. M., Biochim. BioPhys. Acta 113, 567 (1966) Gatt, S., and Rapport, M. M., Biochim. J. 101, 680 (1966) Bowen, D. M., and Radin, N.S., Biochim. Biophys. Acta 152, 599 (1968) 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 132 Radin, N. S., Bradley, R. M., and Brady, R. 0., Brain Res. 24, 497 (1969) Wenger, D. A., Sattler, M., and Hiatt, W., Proc. Nat. Acad. Sci. U. S. A. 71, 854 (1974) Wenger, D. A., Chem. Phys. Lipids 25, 327 (1975) Suzuki, Y., and Suzuki, K., J. Biol. Chem. 249, 2105 (1974) Suzuki, Y., and Suzuki, K., J. Biol. Chem. 249, 2113 (1974) Miyatake, T., and Suzuki, K., J. Biol. Chem. 250, 585 (1975) Brady, R. 0., Gal, A. E., Bradley, R. M., and Martinsson, E., J. Biol. Chem. 242, 1021 (1967) Beutler, E., and Kuhl, W., J. Lab. Clin. Med. 78, 987 (1971) Beutler, E., and Kuhl, W., Amer. J. Hum. Genet. 24, 237 (1972) Wood, S., and Nadler, H. L., Amer. J. Hum. Genet. 24, 250 (1972) Ho, M. W., Buetler, E., Tennant, L., and O'Brien, J. S., Amer. J. Hum. Genet. 24 256 (1972) Beutler, E., and Kuhl, W., J. Biol. Chem. 247, 7195 (1972) Beutler, E., Guinto, E., and Kuhl, W., Amer. J. Hum. Genet. 25, 42 (1973) Mapes, C. A., and Sweeley, C. C., FEBS Letters 25, 279 (1972) Johnson, W. G., and Brady, R. 0., in Methods in Enzymology (Vol. XXXVIII), V. Ginsberg, Ed., Academic Press, New York 849, (1955) Ho, M. W., Biochem. J., 253 l (1973) Mapes, C. A., Anderson, R. L., and Sweeley, C. C., FEBS Letters 2, 180 (1970) Mapes, C. A., and Sweeley, C. C., J. Biol. Chem. 248, 2461 (1973) 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 86. 87. 88. 89. 90. 91. 92. 133 Mapes, C. A., Suelter, C. J., and Sweeley, C. C., J. Biol. Chem. 248, 2471 (1973) Mapes, C. A., and Sweeley, C. C., Arch. Biochem. Biophys. 158, 297 (1973) Brady, R. 0., Gal, A. E., Bradley, R. M., Martinsson, E., Warshaw, A. L., and Laster, L., New Eng. J. Med. 276, 1163 (1967) Desnick, R. J., Allen, K. Y., Desnick, S. J., Raman, M. K., Bernlohr, R. W., and Krivit, W., J. Lab. Clin. Med. fill 157 (1973) Kano, I., and Yamakawa, T., J. Biochem. 75, 347 (1974) Kano, I., and Yamakawa, T., Chem. Phys. Lipids 25, 283 (1974) Snyder, P. D., Jr., Wold, F., Bernlohr, R. W., Dullam, C., Desnick, R. J., Krivit, W., and Condie, R. M., Biochim. Biophys. Acta 350, 432 (1974) Watanabe, K. J. Biochem. 25, 287 (1936) Watanabe, K. J. Biochem. 25, 297 (1936) Watanabe, K. J. Biochem. 24, 305 (1936) Watanabe, K. J. Biochem. 24, 315 (1936 ——- East, M. E., Madinaveita, J., and Todd, A. R., Biochem. J. 55, 872 (1941) Linker, A., Meyer, K., and Weissman, B., J. Biol. Chem. 213, 237 (1955) Pugh, D., Leaback. D. H., and Walker, P. G., Biochem. J. 55, 464, (1957) Woollen, J. W., Heyworth, R., and Walker, P. G., Biochem. J. 25, 111 (1961) Walker, P. G., Woollen, J. W., and Heyworth, R., Biochem. J. 25, 288 (1961) Woollen, J. W., Walker, P. G., and Heyworth, R., Biochem. J. 25, 294 (1961) Frohwein, Y. Z., and Gatt, S., Biochem. 5, 2775 (1967) Frohwein, Y. Z., and Gatt, S., Biochem. 5, 2783 (1967) Robinson, D., and Stirling, J. L., Biochem. J. 107, 321 (1968) 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 134 Murphy, J. V., and Craig, L., Clin. Chem. Acta 42, 267 (1972) Murphy, J. V., and Craig, L., Clin. Chem. Acta 51, 67 (1974) Okada, S., and O'Brien, J. S., Science 165, 698 (1969) Kolodny, E. H., Brady, R. O., and Volk, B. W., Biochem. BiOphys. Res. Comm. 21, 526 (1969) Sandhoff, K., FEBS Letters, 22, 342 (1970) Wenger, D. A., Okada, S., and O'Brien, J. S., Arch. Biochem. Biophys. 153, 116 (1972) Srivastava, S., Yoshida, A., Awasthi, Y. C., and Beutler, E., J. Biol. Chem. 249, 2049 (1974) Srivastava, S., Awasthi, Y. C., Yoshida, A., and Beutler, E., J. Biol. Chem. 249, 2043 Srvastava, S., and Beutler, E., J. Biol. Chem. 249, 2054 (1974) Li, Y. T., Mazzotta, M. Y., Wan, C. C., and Li, S. C., J. Biol. Chem. 248, 7512 (1973) Li, S. C., Wan, C. C., Mazzotta, M. Y., and Li, Y. T., Carb. Res. 25, 189 (1974 Sandhoff, K., FEBS Letters 5, 351 (1969) Sandhoff, K., Andreae, U., and Jatzkewitz, H., Life Sciences 2, 283 (1968) Pilz, H., Muller, D., Sandhoff, D., and ter Meulen, V., Deutsch. Med.Wochschr. 22, 1833 (1968) O'Brien, J. S., Lancet 2, 805 (1969) Suzuki, K., Suzuki, J., Rapin, I., Suzuki, Y., and Ishu, N., Neurology 25, 190 (1970) O'Brien, J. S., and Okada, S., Amer. J. Hum. Genet. 22, 524 (1969) O'Brien, J. S., Okada, S., Ho, H. M., Fillerup, D. L., Veath, M. L., and Adams, K., in Lipid Storage Diseases, Enzymatic Defects and Clinical Implications, J. Berhnson and H. J. Grossman, (Eds.) Academic Press, New York, 224 (1971) 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 135 Anderson, W., Brit. J. Derm 25, 113 (1898) Fabry, J., Arch. Derm. Syph. 52, 187 (1898 Ruiter, M., Pompen, A. W. M., and Wyers, H. J. G., Dermatologia (Basel) 25, 1 (1947) Pompen, A. W. M., Ruiter, M., and Wyers, H. J. G., Acta Med. Scand. 128, 234 (1947) Scriba, K., Verh. Deutsch. Ges. Path. 25, 221 (1950) Wherrett, J. R., and Hakomori, S. I., J. Biol. Chem. 248, 3046 (1973) Kint, J. A., Science 167, 1268 (1970 Romeo, G., and Migeon, B. R., Science 170, 180 (1970) Kint, J. A., and Dacremont, G., Arch. Int. Physiol. Biochem. 15, 169 (1970) Brady, R. O., Ulendorf, B. W., and Jacobson, C. B., Science 172, 174 (1971) Dubach, U. C., Enderlin, F., and Mannhar, M., Germ. Med. Mths. 55, 34 (1969) Mapes, C. A., and Sweeley, C. C., Biochem. BiOphys. Res. Comm. 52 1317 (1973) Clarke, J. T. R., Knaack, J., Crawhill, J. C., Wolfe, L. S., New Eng. J. Med. 284, (1971) Gauche, P. C. E., Thesis, University of Paris (1882) as described by R. 0. Brady and F. M. King in Lysosomes and Storage Diseases, H. G. Hers and F. Van Hoff, (Eds.), Academic Press, New York, 381 (1973) Marchand, F., Mfinchen. Med. Wschr. 55, 1102 (1907) Leib. H., Z. Physiol. Chem. 140, 305 (1924) Lieb, H., Z. Physiol. Chem. 170, 60 (1927) Lieb, H., Z. Physiol. Chem. 181, 208 (1929) Klenk, E., and Harle, R., Z. Physiol. Chem. 189, 243 (1930) Agion, A., Thesis, University of Paris (1934) as described by D. S. Fredrickson and H. R. Sloan in The Metabolic Basis of Inherited Diseases (3rd Ed.) J. B. Stanbury, J. B. Wyngaarden, and D. S. Fredrickson (Eds.), McGraw-Hill Book Co., New York, 730 (1972) 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 136 Ho, M. W., Biochem. J. 136, 721 (1973) Brady, R. 0., and King, F. M., in Lysosomes and Storage Diseases, H. G. Hers, and F. Van Hoff, (Eds.) Academic Press, New York, 381 (1973) Fredrickson, D. S., and Sloan, H. R., in The Metabolic Basis of Inherited Diseases (3rd Ed.), J. B. Stanbury, J. B. Wyngaarden, and D. S. Fredrickson, (Eds.), McGraw- Hill Book Co., New York, 730 (1972) Krabbe, K., Brain 25, 74 (1916) Hallervorden, J., Vehr. Deutsch. Ges. Pathol. 22, 96, (1948) as described by K. Suzuki and Y. Suzuki in 255 Metabolic Basis of Inherited Diseases, J. B. Stanbury, J. B. Wyngaarden, and D. S. Fredrickson, (Eds.), McGraw-Hill Book Co., New York, 760 (1972) Austin, J. H., J. Neurochem. 25, 921 (1963) Blackwood, W., and Cumings, J. N., J. Neurochem. Neuro- surg. Psyciat. 21, 33 (1954) Diezel, P. B., Arch. Path. Anat. 227, 206 (1955) Austin, J., Lehfeldt, D., and Maxwell, W., J. NeurOpath. Exp. Neurol. 25, 284 (1961) Austin, J., and Lehfeldt, D., J. Neuropath. Exp. Neurol 25, 265 (1965) Miyatake, T., and Suzuki, K., Biochem Bi0phys. Res. Comm. 55, 538 (1972) Miyatake, T., and Suzuki, K., J. Biol. Chem 247, 5398 (1972) Suzuki, K., and Suzuki, Y., in The Metabolic Basis of Inherited Diseases (3rd Ed.), J. B. Stanbury, J. B. Wygaarden, and D. S. FredriEkson, (Eds.), McGraw-Hill Book Co., New York, 760 (1972) Fankhauser, R., Lubinbfihl, H., and Hartley, W. J., Schweiz. Arch. Tierheilk. 105, 198 (1963) as described by K. Suzuki and Y. Suzuki, Ibig Max, 8. R., Maclaren, N. K., Brady, R. 0., Bradley, M. S., Rennels, M. B., Yanaka, J., Garcia, J. H., and Cornblath, M., New Eng; J. Med. 291 929 (1974) Brot, F. S., Glaser, J. H., Roozen, K. J., and Sly, W. S., Biochem. Biophys. Res. Comm. 51, 1 (1972) 147. 148. 149. 150. 151. 156. 157. 158. 159. 160. 161. 162. 137 Hickman, S., and Neufeld, E. F., Biochem. Biophys. Res. Comm. 55, 992 (1972) Hickman, S., Shapiro, L. J., and Neufeld, E. F., 5iochem. Biophys. Res. Comm. 51, 55 (1974) Schneck, L., Amsterdam, D., Brooks, S. E., Rosenthan, A. L., and Volk, B. W., Pediatrics 52, 221 (1973) Kolodny, E. H., Milunsky, A., and Sheny, G. S., in Enzyme Therapy in Genetic Diseases, R. J. Desnick, R. W. Bernlohr, and W. Krivit, (Eds.) Birth Defects: Origninal Article Series, IX, 130 (1930) Morrell, A. G., Gregoiadis, G., Scheinberg, I. H., Hickman, J., and Ashwell, G., J. Biol. Chem. 246, 1461 (1971) Morrell, A. G., Irvine, R. A., Sternlieb, I., Scheinberg, I. H., and Ashwell, G., J. Biol. Chem. 243, 155 (1968) Pricer, W. E., and Ashwell, G., J. Biol. Chem. 246, 4825 (1971) Rogers, J. C., and Kornfield, S., Biochem. Biophys. Res. Comm. 55, 622 (1971) Chang, T. M. S., Artificial Cells, Charles C. Thomas, Springfield, 7 (1972) Ihler, G. M., Glew, R. H., and Schnure, F. W., Proc. Nat. Acad. Sci. U. S. A. 15, 2663 (1973) Rifkind, R. A., Amer. J. Med. 41, 711 (1966) Gregoriadis, G., and Buckland, R. A., Nature 244, 170 (1973) Gregoriadis, G., and Ryman, B. E., Eur. J. Biochem. 140 485 (1972) Gregoriadis, G., and Ryman, B. E., Biochem. J. 129, 123 (1972) Weissmann, G., Bloomgarden, D., Kaplan, R., Cohen, C., Hoffstein, S., Collins, T., Gotlieb, A., Nagle, D., Proc. Nat. Acad. Sci. U. S. A. 12, 88 (1975) Gregoriadis, G., in Enzyme Therapy in 2ysosoma1 Storagg Diseases, J. M. Tager, G. J. M. Hoogehwenkel, W. T. Daems, (Eds.), North Holland Publishing Co., Amsterdam, 131, (1974) 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 138 Greene, H. L., Hug, G., and Schubert, W.K., §£c55 Neurol. 20, 147 (1969) Austin, J. H., in Inborn Errors of Sphingolipid Metabolism, S. M. Aroson and B. W. Volk, (Eds.), Pergamon Press, New York, 359 (1967) Johnson, W. G., Desnick, R. J., Long, D. M., Sharp, H. L., Krivit, W., Brady, B., and Brady, R. O., in Enzymen Therapy in Genetic Diseases, R. J. Desnick, R. W. Bernlohr, and W. Krivit, (Eds.), Birth Defects: Original Article Series, IX, 120 (1973) Desnick, R. J., Snyder, P. D., Desnick, S. J., Krivit, W., and Sharp, H. L., in Sphingolipids, Sphingolipidoses and Allied Disorders, B. W. Volk and S. M. Aronson, (Eds.), Plenum Press, New York, 351 (1972) Brady, R. 0., Pentchev, P. G., Gal, A. E., Hibbert, S. R., and Dekaban, A. S., New Eng. J. Med. 291, 989 (1974) Mapes C. A. Anderson, R. L., Sweeley, C. C., Desnick, R. J., and Krivit, W., Science 169, 987 (1970) Brady, R. 0., Tallman, R. F., Johnson, W. G., Gal, A. E., Leahy, W. R., Quirk, J. M., and Dekaban, A. S., New Eng. J. Med. 289, 9 (1973) O'Brien, J. S., Fed. Proc. 32, 191 (1973) Ikonne, J. U., and Ellis, R. B., Biochem. J. 135, 457 (1973) Sweeley, C. C., Mapes, C. A., Krivit, W., and Desnick, R. J., in Sphingolipids, Spingolipidoses and Allied Disorders, B. W. Volk and S. M. Aronson, (Eds.), Plenum Press, New York, 287 (1972) Christensen, E., New Eng5 J. Med. 290, 629 (1974) Austin, J. H., Neurology 25, 470 (1960) Cuatrecases, P., J. Biol. Chem. 245, 3059 (1970) March, 8. C., Parikh, I., and Cuatrecases, P., Anal. Biochem. 55, 149 (1974) Udenfriend, S., Stein, S., Boehlen, P., Dairman, W., Leimgruber, W., and Weigele, M., Science 178, 871 (1972) Gomori, G., in Methods and Enzymology (Vol. I), S.P. Colowick and N. 0. Kaplan, (Eds.), Academic Press, New York, 138 (1955) 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. 191. 192. 139 Dunn, M. 8., Biochem. Prep 2, 22 (1949) Liener, I. E., and Friedenson, B., in Methods in Enzymology (Vol. XIX), S. P. Colowick and N. 0. Kaplan, (Ed.), Academic Press, New York, 261 (1955) Cohn, E. J., Strong, L. E., Hughes, W. L., Mulford, D. J., Ashworth, J. M., Melin, M., and Taylor, H. L., J. Amer. Chem. Soc. 55, (1946) Suzuki, Y., and Suzuki, K., J. Lipid Res., 13, 687 (1972) Folch, J., Lees, M., and Stanley, G. H. S., J. Biol. Chem. 226, 497 (1957) Vance, D. E., and Sweeley, C. C., J. Lipid Res.5, 621 (1967) The Merk Index (Eighth Edit.), P. G. Stecher (Ed.), Merk and Company, Inc., Rahway, 178 (1968) Kanfer, J. N., Petrovich, G., and Mumford, R. A., Anal. Biochem. 55, 301 (1973) Harpaz, N., Flowers, H. M., and Sharon, N., Biochem. Biophys. Acta 341, 213 (1974) Dean, K. J., Brozowki, T., and Sweeley, C. C., Unpublished Experiments Brady, R. O., Pentcev, P. G., and Gal, A. E., Fed. Proc. 25, 1310 (1975) Sgouris, J., Personal Communication Desnick, R. J., Personal Communication Molick, N., and Berzie, A., Anal. Biochem. 49, 173 (1973) "'TIT!17IL)ITIL 11111111131711)? Ilflfllfllfll?)fl'“ 533