Clinical Genetics 1985: 21: 443450

Glycosphingolipid studies of visceral tissues and brain from type 1 Gaucher

disease variants OLLE NILSSON,’ GREGORY A. GRAB OW SKI,^ MARK D. LUDMAN,~ ROBERT J. DESNICK~ AND LARS

SVENNERHOLM~ IDepartment of Psychiatry and Neurochemistry, University of Goteborg, St. Jorgen’s Hospital,

Sweden, and 2The Center for Jewish Genetic Diseases, Division of Medical Genetics of the Mount Sinai School of Medicine, New York, NY, U.S.A.

Glucosylceramide and glucosylsphingosine isolated from spleen, liver and brain were quantita- ted and characterized in two unrelated patients with Gaucher disease, neither of whom had clinical or neuropathologic evidence of neuronal involvement. Visceral glucosylceramide ac- cumulation did not differ in the two patients. Hepatic glucosylsphingosine content was 2-fold greater in a young severely affected 3-year-old American Black patient compared to that in a 56-year-old Ashkenazi Jewish patient. In contrast, significant differences in glycosphingolipid content and composition were observed in the brains of these two cases. Cerebral and cerebellar cortical glucosylceramide accumulated to a greater extent (3-fold) in the severely affected 3- year-old patient compared to that in the older case. The compositions of the acyl and sphingosyl base residues of glucosylceramide in the cerebral and cerebellar cortices from the Ashkenazi Jewish patient were similar to those in normal individuals. In comparison, the gray matter glucosylceramide in the severely affected patient had increased percentages of stearic acid (18:O) and eicosasphingenine (d20: I), suggesting that the accumulated substrate was derived from the brain ganglioside pool. Glucosylsphingosine was found in large amounts only in cerebral and cerebellar cortices from the severely affected patient. The glycolipid content and composition in this patient was similar to that found in the Norrbottnian (Type 3) form of Gaucher disease. The differences in glucosylceramide acyl and sphingosyl base composition in gray matter from the severely affected patient and that in the Ashkenazi Jewish patient suggested that the accumulated substrates were metabolized differently by the residual enzymes in each case. Since both glucosylceramide and glucosylsphingosine are substrates of the normal enzyme, the data suggested that the allelic mutations of acid P-glucosidase can differentially alter the catabolism of these substrates and result in the different phenotypes of Gaucher disease.

Received 3 Septemher 1984, revised, accepted for publication I3 January I985

Key words: Gaucher disease; /I-glucosidase; glucosylceramide; glucosylsphingosine; psychosine.

Gaucher disease, a glycosphingolipid stora- et al. 1982, Svennerholm et al. 1982). The ge disease, is characterized by the deficient clinical manifestations result from the ac- activity of the lysosomal enzyme, acid j3- cumulation of this enzyme’s natural sub- glucosidase (N-acyl-sphingosyl- 1 -0-j3-D- strate, glucosylceramide, primarily within glucoside: glucohydrolase, EC 3.2.1.45) cells of monocyte/macrophage origin (Des- (Brady et al. 1965, Patrick 1965, Desnick nick et al. 1982, Parkin & Brunning 1982).

444 N I L S S O N E T A L

Hypertrophy and hyperplasia of these Gau- cher cells lead to hepatosplenomegaly, skel- etal deterioration and other visceral mani- festations. Three phenotypes of Gaucher disease have been classified by the absence or presence and severity of central nervous system involvement: Type 1, non-neuron- opathic, Type 2, acute neuronopathic, and Type 3, sub-acute neuronopathic (Fredrick- son & Sloan 1972, Dreborg et al. 1980, Des- nick et al. 1982, Svennerholm et al. 1982).

Despite these distinct clinical courses, comparative characterizations of the stor- age products isolated from the visceral or- gans of each of these subtypes have not revealed differential abnormalities of glycol- ipid content or composition (Nilsson et al. 1982). Since the major distinguishing fea- tures are referable to central nervous in- volvement, comparative studies of the stor- age materials in brain may provide greater insight into the functional enzyme abnor- malities which distinguish the non-neuron- opathic and neuronopathic forms of this disease.

In this communication, the glycolipid ab- normalities in spleen, liver and brain gray matter from two unrelated patients with Type 1 Gaucher disease phenotype, but who had remarkably different clinical courses, are compared and contrasted for the first time with previous findings from patients with Gaucher Type 2 and Norrbottnian Type 3 disease (Nilsson et al. 1982, Nils- son & Svennerholm 1982a).

Case Reports Patient 1, a 3-year-old American Black male, was diagnosed as homozygous for the Gaucher disease gene at age 6 months. The diagnosis was established by the presence of rapidly progressive hepatosplenomegaly and lymphadenopathy, as well as typical Gaucher cells in bone marrow. Biochemical demonstration of deficient acid p-gluco- sidase activity in lymphocytes and cultured

fibroblasts confirmed the diagnosis. Exten- sive physicokinetic and immunologic stu- dies of the residual enzyme in this patient will be published separately (Grabowski et al. 1985). No evidence of central nervous system involvement was found by compre- hensive neurologic examination, CT scans and electroencephalographic studies per- formed at 2.7 years. Death from cardiac failure occurred at 3 years. Postmortem light and electron microscopic examination of the brain revealed only perivascular ac- cumulation of Gaucher cells in a small frontal subcortical area without signs of neuronal involvement.

Patient 2 was a 56-year-old Ashkenazi Jew- ish male who was diagnosed as having Gau- cher disease at age 20 years when he was evaluated for splenomegaly. The diagnosis was confirmed by the demonstration of defi- cient acid p-glucosidase in various tissues. Death occurred at age 56 years from Pseu- domonas aueriginosa septicemia secondary to osteomyelitis of the humerus. Postmor- tem light and electron microscopic examina- tion of the brain revealed evidence of an old subarachnoid bleed without any other abnormalities. Detailed physicokinetic studies of the residual splenic enzyme of patient 2 will be published separately (Gatt et al. 1985).

Material and Methods

Tissue Preparation Tissues were obtained at autopsy within 1 h of death from each patient. Lipids were analyzed from approximately 2 g of gray matter isolated from cerebrum and cerebel- lum, and about 3 g of splenic and liver tissue from each case.

Lipid Extraction and Isolation The tissues were homogenized in water (1 : 3; w/v) with a scissor homogenizer and ex-

B R A I N G L Y C O L I P I D S I N T Y P E 1 G A U C H E R D I S E A S E 445

HOMOGENIZATION

EXTRACTION

SEPHADEX G-25 I

ION EXCHANGE

I

I

SILICA'COLUMN CHROMATOGRAPHY

D i s c a r d e d I 10 vol c

10 vol C-HAc, 19/1

10 vol C-M, 9/1 Cereb ros ides

10 v o l C-M, 4/1 Loc tosy l ce ramide

10 v o l C-M 1/1

Psychosine, o l i g o h e x o s y l c e r a m i d e s

Fig. 1. Schematic illustration of the isolation pro- cedures.

tracted twice with twenty volumes of chloroform: methanol :water (4: 8 : 3; v/v/v) (Nilsson & Svennerholm 1982b). The total lipid extract was evaporated to dryness, dissolved in chloroform: methanol: water (60: 30: 3.5; v/v/v) and purified by chroma- tography on Sephadex G-25 (Wells & Dittmer 1963). The purified total lipid ex- tract was used for quantitative determi- nation of cholesterol, phospholipids and to- tal ganglioside sialic acid content. Other gly- colipids were determined after separation of the total lipid extract as shown in Figure 1, as previously described (Nilsson & Svenner- holm 1982a). Briefly, the total lipid extract was separated into neutral and acidic lipid fractions by anion exchange chromatogra- phy. The neutral lipid fraction was subjected to silica gel column chromatography and the monohexosylceramide, dihexosylcer- amide and oligohexosylceramide fractions were isolated. The latter fraction also con- tained glucosylsphingosine. The mono- hexosyl fraction isolated from the brain tis- sue was separated into galactosyl- and glu- cosyl-ceramide by preparative TLC on

borate impregnated plates using chloro- form:methanol:water (40: 10: 1; v/v/v) as the developing solvent. The oligohexosyl- ceramides were purified from phospholipids by peracetylation and chromatography ac- cording to Saito & Hakomori (1971).

Individual glycolipids were resolved by TLC, visualized with cupric acetate and quantified by densitometric scanning at 450 nm (Nilsson & Svennerholm 1982b). Gluco- sylsphingosine was determined as its N-ace- tyl derivative, which was formed during the peracetylation and subsequent removal of 0-acetyl groups with mild alkali (Nilsson & Svennerholm 1982a). The N-acetyl gluco- sylsphingosine was resolved on HPTLC- plates using a ch1oroform:methanol: water (40: 10: 1; v/v/v) solvent system. The sphin- gosyl and acyl compositions of the glucosyl- ceramides were determined by gas liquid chromatography after acid hydrolysis with 1.0 M HCl in methano1:water (82: 18; v/v) (Mansson et al. 1978). The sphingosine base compositions were determined as their cor- responding aldehydes obtained after per- iodate oxidation (Mansson et al. 1978).

Results The results of the quantitative lipid analyses of the visceral and neural tissues from each patient and normal controls are summa- rized in Tables 1 and 2. In the spleen and liver, the amount of glucosylceramide in pa- tients l and 2 were increased about 300- fold, to values similar to those observed in Type 2 and Norrbottnian Type 3 homozygo- tes (Nilsson et al. 1982). Glucosylsphingo- sine also was accumulated in the spleen and liver of both patients. In spleen, the concen- trations of glucosylsphingosine in patients 1 and 2 were similar to those observed in spleens from Type 2 and Norrbottnian Type 3 homozygotes (Nilsson et al. 1982). In con- trast, the glucosylsphingosine concentration in the liver of patient 1 was about 2-fold

446 N I L S S O N E T A L .

Table 1

L ip id composi t ion in spleen a n d l iver

Phospho- Ganglioside Glucosyl Glucosyl Cholesterol lipids NeuAc ceramide sphingosine

rnrnollkg wet weight

9.34Z0.2 16.4*1.8 0.28*0.07 8.04Z 0.4 25.2 i 2.5 0.22 4 0.02

yrnollkg wet weight

Normal Controls' (n = 5): Spleen Liver

Patient I : Spleen Liver

Patient 2. Spleen Liver

0.1oio.01 ND2 0.040 * 0.005 ND2

14.6 20.4 1.65 8.2 23.1 0.78

30.0 0.16 13.7 0.10

9.2 19.2 0.95 7.2 22.4 0.61

23.8 0.13 13.8 0.04

' Nilsson et al. (1982) ND not detected.

higher than that in patient 2, and was simi- lar to those found in Type 2 homozygotes or splenectomized Norrbottnian Type 3 pa- tients (Nilsson et al. 1982). The glucosyl- sphingosine concentration in the liver of pa- tient 2 was the same as that in non-splenec-

tomized Norrbottnian Type 3 homozygotes. In addition, an increased concentration of gangliosides was found in liver and spleen of both patients 1 and 2 .

Quantitative and qualitative differences in brain gray matter glycolipids were ob-

Table 2

Lipid composition in the brain

Phospho- Ganglioside Galactosyl- Glucosyl- Glucosyl- Cholesterol lipids NeuAc ceramide ceramide sphingosine

hmol/kg wet weight

ND2 ND2 ND2

mrnol/kg wet weight

41.1f3.09 3.19i0.32 0.98' 38.9 2.61 1.61 45.0 3.24 6.03

Normal controls: C.C. 6 2 4 rno ( n = l l ) Cb.C. !54 rno

5-20 yrs

Type /I: C.C (n=5)3 Cb.C. (n=4) l

C.C. (n=8)3 Cb.C. (n=6)3

Patient 1: C.C. Cb.C.

Patient 2: C.C. Cb.C.

Norrbottnian Type ///I

23.6k2.54 19.6 28.9

5' 10 16

26.9 k 1 .O 25.64 2.6

43.6 i 2.4 3.2 f 0.45 0.91 f 0.47 42.942.2 2.9f0.12 3.25f0.97

140-5304 51-4504

3 . 8 8 . 8 4

3.9-12.34

25.3f 2.3 28.4f3.3

46.0+3.9 3.2k0.39 1.28f0.70 46.71t3.7 2.9f0.22 4.55& 1.32

37-654 59-1 7504

0.@-t.64 1.4-6.34

31 .O 26.7

45.8 3.51 1.45 45.1 3.13 2.79

92 94

2.4 1.7

32.7 40.2

53.6 3.44 3.26 58.4 2.96 7.88

36 28

0.2 0.06

'Svennerholm et al. (1982); 2Not detected; 3From Nilsson 8. Svennerholm (1982a), C.C.=cerebral cortex, Cb.C. = cerebellar cortex; 4The highest values were found in splenectornized cases.

B R A I N G L Y C O L I P I D S I N T Y P E 1 G A U C H E R D I S E A S E 447

Table 3

Ceramide composition in glucosylceramide isolated from brain

Sphingosine base Fatty acid composition composition

16:O 18:O 20:O 22:O 22: l 23:O 24:O 24: l d18:O d18:l d20: l t20:O

Normal controls’ C.C. (2-27 rnos) 2 42 4 9 - 2 17 14 Cb.C. (3-8 mos) 3 36 5 11 2 4 17 15 9 8 1 7 1

(5-20yrs) 6 34 7 12 1 4 13 16 6 8 1 9 2

C.C. 4 6 3 5 1 1 - 4 1 3 - 6 74 18 1 Cb.C. 7 58 10 10 - 2 11 2 5 70 24 1

Patient 1

Patient 2 C.C. 5 4 7 6 1 8 2 5 1 7 - 5 79 14 1 Cb.C. 4 3 7 7 1 5 1 5 2 6 6 5 81 12 -

Nilsson & Svennerholrn (1982a).

served between these two patients (Table 2 and 3). The concentration of glucosylcer- amide was increased in both cases but was approximately three times higher in the tis- sues of patient 1 than in patient 2. The concentrations were of the same magnitude (although slightly lower in patient 2) as those observed in comparable tissues from Norrbottnian Type 3 non-splenectomized cases (Nilsson & Svennerholm 1982a).

Analyses of the acyl and sphingosine moi- eties of glucosylceramide revealed differ- ences in the composition of brain ceramides between these two cases (Table 3) . Glucosyl- ceramide in cerebral and cerebellar cortices from patient 1 contained a higher pro- portion (60%) of stearic acid (C- 18 : 0) and eicosasphingenine (d20: 1). This ceramide composition was similar to that of glucosyl- ceramide isolated from cerebral and cerebel- lar cortices of Gaucher Type 2 cases and from cerebral cortex of Norrbottnian Type 3 cases (Nilsson & Svennerholm 1982a). In contrast, brain glucosylceramide from pa- tient 2 contained a lower percentage of ste- aric acid and higher proportions of behenic acid (C-22:O) and (2-24 fatty acids. In ad- dition, the ceramide composition of gluco-

sylceramide in brain from patient 2 re- sembled that of control subjects.

The amounts of glucosylsphingosine in the brain from patients 1 and 2 were differ- ent (Table 2). The levels of glucosylsphingo- sine in cerebral (2.4 ,umol/kg) and cerebellar (1.7 pmol/kg) cortices from patient 1 were 10- and 30-fold increased over the levels in patient 2 , respectively. Moreover, the con- centrations of glucosylsphingosine in cere- bral and cerebellar tissues from patient 1 were similar to those found in Norrbottnian Type 3 homozygotes (Nilsson & Svenner- holm 1982a). In contrast, the concen- trations of glucosylsphingosine in cerebral and cerebellar cortices from patient 2 were very low (0.2 and 0.06 ,umol/kg, respective- ly).

Discussion

The clinical delineation of the three pheno- types of Gaucher disease is based primarily on the absence or presence and severity of neuronopathic manifestations. Although re- cent physicokinetic (Grabowski et al. 1985) and immunologic (Ginns et al. 1983, Gra- bowski et al. 1985) studies have suggested

448 N I L S S O N E T A L .

different properties of the residual acid j?- glucosidase in the neuronopathic (Types 2 and 3) and the non-neuronopathic (Type 1) diseases, the specific nature of the abnormal function of the residual enzyme and the re- sulting chemical basis for the differential pathophysiology underlying the three clini- cally distinct phenotypes remain unknown. In this communication, detailed analyses of the composition of accumulated brain gly- colipids are reported for the first time in two cases of Gaucher disease without clinical evidence of neuronopathic disease. Patient 1 was an American Black male who had severe visceral disease manifestations, no evidence of neurologic involvement and ex- pired at 3 years of age. In contrast, patient 2 was an Ashkenazi Jewish male who had the typical features of Type 1 disease, no neurologic manifestations and expired at 56 years of age. These analyses demonstrated quantitative and qualitative differences in brain substrate accumulation in these pa- tients with Type l disease. These findings were instructive as they provide information on the levels of glucosylceramide and gluco- sylsphingosine in presumed Type 1 homo- zygotes as a function of disease severity, which can be compared to those in Type 2 and 3 homozygotes.

Although glucosylceramide accumulation was demonstrated in the gray matter from both cases, the amount in the 3-year-old patient 1 was almost three times that ob- served in the 56-year-old patient 2. Further- more, glucosylsphingosine, a putative neur- otoxin (Nilsson et al. 1982, Nilsson & Sven- nerholm 1982a), was accumulated in cerebral and cerebellar cortices from patient 1 at levels 10- to 30-fold higher, respectively, than those in patient 2.

The amounts of accumulated glucosylcer- amide and glucosylsphingosine in the brain tissues from the more severely affected pa- tient l were similar to those previously ob- served in the brain of non-splenectomized

Norrbottnian Type 3 homozygotes (Nils- son & Svennerholm 1982a), but were sub- stantially less than that found in compar- able Type 2 disease tissues. In contrast, the concentrations of glucosylsphingosine in patient 2, who had a classical Type 1 disease course, was markedly less in both cerebral and cerebellar cortices than the respective levels in patient 1 (or the Type 2 and Norrbottnian Type 3 homozygotes). It is notable that the low levels of glucosylsphin- gosine in the brain of patient 2 represented a period of accumulation over five decades, reaching a level only slightly above the limit of detection (0.01 pmollkg) (Nilsson & Svennerholm 1982a). In contrast, in the Type 2 and Norrbottnian Type 3 cases, as well as in patient 1, the accumulation of glucosylceramides and glucosylsphingosine in the cortices reached much higher concen- trations in less than a decade. These findings suggest that the chemical pathology in pa- tient l , although clinically and neuropathol- ogically classified as a Type 1 homozygote, possibly had an acid j?-glucosidase mutation which was similar to those in Norrbottnian Type 3 patients.

The differences in the composition of the acyl and sphingosine moieties of brain glu- cosylceramides from patients 1 and 2 sug- gested that the accumulated substrates in these homozygotes were derived from dis- tinct metabolic sources. The greater pro- portion of stearic acid (18:O) and eicosa- sphingenine (d20: 1) in the accumulated brain glucosylceramide from patient 1 sug- gested that part of the stored substrate was derived from gangliosides, since the cerami- des derived from GMI, GDI,, GDlband GTlb gangliosides are enriched for stearic acid and eicosasphingenine (Mansson et al. 1978). In contrast, the predominance of C22 : 0 and C24 fatty acids, as well as d 18 : 1 sphingosine in glucosylceramide from the cerebral and cerebellar cortices of patient 2 suggested that a significant percentage of

B R A I N G L Y C O L I P I D S I N T Y P E 1 G A U C H E R D I S E A S E 449

this accumulated substrate was from non- ganglioside derived metabolic sources.

The results of the chemical analyses of the glycolipid abnormalities in the brains from the two cases clinically diagnosed as Type 1 disease suggested that they did not belong to the same disease subtype. Indeed, recent enzymologic studies using cultured fibroblasts (Grabowski et al. 1985) or delipi- dated splenic extracts (Gatt et al. 1985) indi- cated that the residual acid P-glucosidase activities from patients 1 and 2 were kinetic- ally different, suggesting the expression of different mutant alleles. The enzyme from patient 1 had kinetic properties identical to those observed in non-Ashkenazi Type 1, Type 2 and Norrbottnian Type 3 homozygo- tes as well as the enzyme in normal subjects (Grabowski et al. 1985). The residual acid P-glucosidase from patient 2 had the typical properties found in other Ashkenazi Type 1 homozygotes. These enzymologic studies, together with the chemical analyses pre- sented here, suggest that the degradation of glucosylsphingosine is dependent on the specific mutations which differentially alter the residual /?-glucosidase activity in each of the Gaucher disease subtypes and their variants.

The data reported here suggest differ- ences in the metabolism of glucosylceramide and glucosylsphingosine in the brain be- tween the three types of Gaucher disease (Table 2, 3). To further evaluate the impor- tance of the glucosylsphingosine accumu- lation in the brain as an explanation for the different clinical expressions of Gaucher disease, additional analyses of several “classical” and non-classical cases of Gau- cher disease obviously are needed.

Acknowledgments

The authors thank Dr. S. Song for expert neuropathology and Mrs. Linda Lug0 for her expert clerical assistance in the prep- aration of this manuscript.

This work was supported in part by grants from the Swedish Medical Research Council (03X3-62), the Medical Faculty (University of Gotenborg), the National In- stitutes of Health (AM 26824), the New York Heart Association, the March of Di- mes Birth Defects Foundation (1-857), and the Florence and Theodore Baumritter Foundation to the Mount Sinai Center for Jewish Genetic Diseases. MDL is the recipi- ent of an NIH postdoctoral fellowship in medical genetics (5 T32 HD07105). GAG is the recipient of an NIH Career Develop- ment Award (1 KO4 AM 01351) and an Irma T. Hirschl Career Scientist Award.

References

Brady, R. O., J. N. Kanfer & D. Shapiro (1965). Metabolism of glucocerebrosides. 11. Evidence of an enzymatic deficiency in Gaucher’s dis- ease. Biochem. Biophys. Rex Comrnun. 18,

Desnick, R. J., S. Gatt & G. A. Grabowski (eds.) (1982). Gaucher Disease: A Century of Delin- eation and Research, New York, Alan R. Liss, Inc., p. 736.

Dreborg, S., A. Erikson & B. Hagberg (1980). Gaucher’s disease - Norrbottnian type. Eur. J . Pediatr. 133, 107-1 18.

Fredrickson, D. S. & H. R. Sloan (1972). Gluco- sylceramide lipidoses: Gaucher’s disease. In The Metabolic Basis of Inherited Disease. 3rd Edit. J. B. Stanbury, J. B. Wyngaarden & D. S. Fredrickson, (eds.). New York, McGraw- Hill Book Co. pp. 730-759.

Gatt, S., T. Dinur, K. Osiecki, R. J. Desnick & G. A. Grabowski (1985). Use of activators and inhibitors to define the properties of the active site of normal and Gaucher disease lysosomal B-glucosidase. Enzyme (in press).

Ginns, E. I., F. P. W. Tegelaers, R. Barnveld, H. Galjaard, A. J. J. Reuser, R. 0. Brady, J. M. Tager & J. A. Barranger (1983). Determination of Gaucher’s disease phenotypes with mon- oclonal antibody. Clin. Chim. Acta 131,

Grabowski, G. A., J. Goldblatt, T. Dinur, J. Kru- se, L. Svennerholm, S. Gatt & R. J. Desnick (1 985). Genetic heterogeneity in Gaucher dis- ease: Physiokinetic and immunologic studies of

221-225.

283-288.

450 N I L S S O N E T A L

the residual enzyme in cultured fibroblasts from non-neuronopathic and neuronopathic patients. Am. J . Med. Genet. (in press).

Mansson, J. E., T. Varnier & L. Svennerholm (1978). Changes in fatty acid and sphingosine composition of the major gangliosides of hu- man brain with age. J . Neurochem. 30,

Nilsson, O., J. E. Mansson, G. Hakansson & L. Svennerholm (1982). The occurrence of psy- chosine and other glycolipids in spleen and liver from the three major types of Gaucher disease. Biochim. Biophys. Acfa 712, 453463.

Nilsson, 0. & Svennerholm (1 982a). Accumu- lation of glucosylceramide and glucosyl sphin- gosine (psychosine) in cerebrum and cerebel- lum in infantile and juvenile Gaucher disease. J . Neurochem. 39, 709-718.

Nilsson, 0. & L. Svennerholm (1982b). Charac- terization and quantitative determination of gangliosides and neutral glycosphingolipids in human liver. J . Lipid Res. 23, 327-334.

Parkin, J. L. & R. D. Brunning (1982). Pathology of the Gaucher cell. In Gaucher Disease: A Century of Delineation and Research. R. J. Des- nick, S. Gatt & G. A. Grabowski, (eds). New York, Alan R. Liss, Inc., pp. 151-176.

273-275.

Patrick, D. A. (1965). A deficiency of glucocereb- rosidase in Gaucher’s disease. Biochem. J . 97,

Saito, T. & S. 1. Hakomori (1971). Quantitative isolation of total glycosphingolipids from ani- mal cells. J . Lipid Res. 12, 257-259.

Svennerholm, L., A. Dreborg, 0. Eriksson, C. G. Groth, P. 0. Hillborg, G. Hakansson, 0. Nilsson & E. Tibblin (1982). Gaucher disease of the Norrbottnian type (type 111). Phenotypic manifestations. In Gaucher Disease: A Century of Delineation and Research. R. J. Desnick, S. Gatt & G. A. Grabowski, (eds). New York, Alan R. Liss, Ine., pp. 67-94.

Wells, M. A. & J. C. Dittmer (1963). The use of Sephadex for the removal of non-lipid con- taminants from lipid extracts. Biochemistry 2,

17C-18C.

1259-1263.

Address: Dr. Lars Svennerholm University of Giitehorg St. Jorgen’s Hospital S-422 03 Hisings Backa Sweden