Isolation of two forms of an activator protein for the enzymic sphingomyelin degradation from human Gaucher spleen

Stimulation of lysosomal sphingomyelin degradation by sphingolipid activator proteins

Lysosomal breakdown of glycosphingolipids with short hydrophilic carbohydrate headgroups is achieved by the simultaneous action of specific hydrolases and sphingolipid activator proteins (SAPs). Activator proteins are considered to facilitate the enzyme/substrate interaction between water-soluble enzymes and membrane-bound substrates. Sphingomyelin, containing the small hydrophilic phosphorylcholine moiety, is hydrolysed by acid sphingomyelinase (acid SMase). Recent experimental data on the in vivo and in vitro role of activator proteins in sphingomyelin breakdown by acid SMase are reviewed. These data combined with the results using homogenous protein preparations as well as a liposomal assay system mimicking the physiological conditions suggest that lysosomal sphingomyelin degradation is not critically dependent on any of the known activator proteins. Moreover, evidence is provided that the assumed intramolecular activator domain of acid SMase and especially the presence of negatively charged lipids in the lysosomes are sufficient for sphingomyelin turnover.

Saposins (sap) A and C activate the degradation of galactosylceramide in living cells

In loading tests using galactosylceramide which had been labelled with tritium in the ceramide moiety, living skin fibroblast lines derived from the original prosaposin-deficient patients had a markedly reduced capacity to degrade galactosylceramide. The hydrolysis of galactosylceramide could be partially restored in these cells, up to about half the normal rate, by adding pure saposin A, pure saposin C, or a mixture of these saposins to the culture medium. By contrast, saposins B and D had little effect on galactosylceramide hydrolysis in the prosaposin-deficient cells. Cells from beta-galactocerebrosidase-deficient (Krabbe) patients had a relatively high residual galactosylceramide degradation, which was similar to the rate observed for prosaposin-deficient cells in the presence of saposin A or C. An SV40-transformed fibroblast line from the original saposin C-deficient patient, where saposin A is not affected, showed normal degradation of galactosylceramide. The findings support the hypothesis, which was deduced originally from in vitro experiments, that saposins A and C are the in vivo activators of galactosylceramide degradation. Although the results with saposin C-deficient fibroblasts suggest that the presence of only saposin A allows galactosylceramide breakdown to proceed at a normal rate in fibroblasts, it remains to be determined whether saposins A and C can substitute for each other with respect to their effects on galactosylceramide metabolism in the whole organism.

Model SV40-transformed fibroblast lines for metabolic studies of human prosaposin and acid ceramidase deficiencies

Skin fibroblasts from patients with Farber disease (acid ceramidase deficiency) and from two siblings of the only known family affected with prosaposin deficiency were transformed by transfection with a plasmid carrying the SV40 large T antigen. The prosaposin-deficient transformed cell lines conserved their original metabolic defects, and in particular they were free of detectable immunoreactivity when using anti-saposin B and anti-saposin C antisera. Ultrastructurally, the cells contained heterogeneous lysosomal storage products. As found for their parental cell lines, the SV40-transformed fibroblasts exhibited deficient in vitro activities of lysosomal ceramidase and beta-galactosylceramidase, but a normal activity of acid sphingomyelinase. As observed for SV40-transformed fibroblasts from Farber disease, degradation of radioactive glucosylceramide or low density lipoprotein-associated radiolabelled sphingomyelin by the prosaposin-deficient cells in situ showed a clear impairment in the turnover of lysosomal ceramide. Ceramide storage in prosaposin-deficient cells was also demonstrated by ceramide mass determination. In contrast to acid ceramidase deficient cells, both the accumulation of ceramide and the reduced in vitro activity of acid ceramidase in cells from prosaposin deficiency could be corrected by addition of purified saposin D. The data confirm that prosaposin is required for lysosomal ceramide degradation, but not for sphingomyelin turnover. The SV40-transformed fibroblasts will be useful for pathophysiological studies on human prosaposin deficiency.

Specific recognition of N-acetylneuraminic acid in the GM2 epitope by human GM2 activator protein

GM2 Activator is a low molecular weight protein cofactor that stimulates the enzymatic conversion of GM2 into GM3 by human beta-hexosaminidase A and also the conversion of GM2 into GA2 by clostridial sialidase (Wu, Y.-Y., Lockyer, J.M., Sugiyama, E., Pavlova, N.V., Li, Y.-T., and Li, S.-C. (1994) J. Biol. Chem. 269, 16276-16283). Among the five known activator proteins for the enzymatic hydrolysis of glycosphingolipids, only GM2 activator is effective in stimulating the hydrolysis of GM2. However, the mechanism of action of GM2 activator is still not well understood. Using a unique disialosylganglioside, GalNAc-GD1a, as the substrate, we were able to show that in the presence of GM2 activator, GalNAc-GD1a was specifically converted into GalNAc-GM1a by clostridial sialidase, while in the presence of saposin B, a nonspecific activator protein, GalNAc-GD1a was converted into both GalNAc-GM1a and GalNAc-GM1b. Individual products generated from GalNAc-GD1a by clostridial sialidase were identified by thin layer chromatography, negative secondary ion mass spectrometry, and immunostaining with a monoclonal IgM that recognizes the GM2 epitope. Our results clearly show that GM2 activator recognizes the GM2 epitope in GalNAc-GD1a. Thus, GM2 activator may interact with the trisaccharide structure of the GM2 epitope and render the GalNAc and NeuAc residues accessible to beta-hexosaminidase A and sialidase, respectively.

Prosaposin deficiency: further characterization of the sphingolipid activator protein-deficient sibs. Multiple glycolipid elevations (including lactosylceramidosis), partial enzyme deficiencies and ultrastructure of the skin in this generalized sphingolipid storage disease

Sphingolipid activator protein (SAP) deficiency, previously described in two sibs and shown to be caused by the absence of the common saposin precursor (prosaposin), was further characterized by biochemical lipid and enzyme studies and by ultrastructural analysis. The 20-week-old fetal sib had increased concentrations of neutral glycolipids, including mono-, di-, tri- and tetrahexosylceramide, in liver, kidney and cultured skin fibroblasts compared with the controls. Glucosylceramide and lactosylceramide were particularly elevated. The kidney of the affected fetus showed additional increases in the concentration of sulphatide, galactosylceramide and digalactosylceramide. Free ceramide was stored in the liver and kidney, and GM3 and GM2 gangliosides were elevated in the liver, but not the brain, of the fetus. Phospholipids, however, were normal in the affected fetus. In the liver biopsy of the propositus, who later died at 16 weeks of age, only a few lipids could be studied. Glucosylceramide, dihexosylceramide and ceramide were elevated in agreement with our previous study. Enzyme studies were undertaken using detergent-free liposomal substrate preparations and fibroblast extracts. The sibs' beta-glucocerebrosidase and beta-galactocerebrosidase activities were clearly reduced, but their sphingomyelinase activities were normal. The normal activity of the latter enzyme and the almost normal tissue concentration of sphingomyelin in prosaposin deficiency suggest that the prosaposin-derived SAPs are not required for sphingomyelinase activity in vivo. In keeping with the biochemical findings, skin biopsies from the sibs showed massive lysosomal storage with a vesicular and membranous ultrastructure. The function of SAPs in sphingolipid degradation and the role of SAPs for enzyme activity in vitro are discussed. In addition, the similarity in neutral glycolipid accumulations in Niemann-Pick disease type C and in prosaposin deficiency are noted. The phenotype of the prosaposin deficient sibs resembled acute neuronopathic (type 2) Gaucher disease more than Farber disease in several aspects, but their genotype was unique.

Additional biochemical findings in a patient and fetal sibling with a genetic defect in the sphingolipid activator protein (SAP) precursor, prosaposin. Evidence for a deficiency in SAP-1 and for a normal lysosomal neuraminidase

It has been shown that sphingolipid activator proteins (SAPs) 1 and 2 are encoded on the same gene along with two other putative activator proteins [Fürst, Machleidt & Sandhoff (1988) Biol. Chem. Hoppe-Seyler 369, 317-328 and O'Brien, Kretz, Dewji, Wenger, Esch & Fluharty (1988) Science 241, 1098-1101]. We have undertaken further biochemical investigations on a patient and fetal sibling, who were previously shown to have a unique sphingolipid storage disorder associated with an SAP-2 deficiency [Harzer, Paton, Poulos, Kustermann-Kuhn, Roggendorf, Grisar & Popp (1989) Eur. J. Pediatr. 149, 31-39]. The severity of their disorder suggested that other products of the SAP precursor or prosaposin gene may also be deficient. The turnover of cerebroside sulphate and globotriaosylceramide were investigated and were both impaired in fibroblasts from the patient and fetus. However, the activities of cerebroside sulphate sulphatase and globotriaosylceramide alpha-galactosidase in vitro were normal in cells from the fetus and patient respectively. In addition, there was an increase in cerebroside sulphate concentration in the kidney of the affected fetus. These results indicate that, in addition to the SAP-2 deficiency, there was a defect in SAP-1 function in this disorder. Additional increases in the concentration of monohexosyl- and dihexosyl-ceramide in the fetal kidney probably reflect the deficiency of SAP-2 in the case of monohexosylceramides, and the combined activator deficiency in the case of dihexosylceramides. Lactosylceramide-loading studies confirmed that there was a defect in the turnover of this lipid in fibroblasts from the affected patient and fetus but not from a patient with an isolated SAP-1 deficiency, or from patients with Krabbe disease, GM1 gangliosidosis or galactosialidosis. It has been suggested [Potier, Lamontagne, Michaud & Tranchemontagne (1990) Biochem. Biophys. Res. Commun. 173, 449-456] that the prosaposin gene also codes for lysosomal neuroaminidase. However, we found normal neuraminidase activity in fibroblasts from our patient, using assay conditions which are diagnostic for sialidosis patients. The role of prosaposin gene products in sphingolipid metabolism is discussed in view of our biochemical findings in this genetic disorder.

Activator proteins and topology of lysosomal sphingolipid catabolism

The lysosomal degradation of several sphingolipids by acid hydrolases is dependent on small non-enzymic cofactors, called sphingolipid activator proteins some of which have been identified as sphingolipid binding proteins. This review summarizes the information available on the structure, function, biosynthesis, gene organization and pathobiochemistry of the known sphingolipid activator proteins. It also offers models for their mode of action and for the topology of lysosomal digestion of glycolipids.

Gaucher's disease in the presence of normal glucocerebrosidase activity

We encountered an infant with clinical and histopathologic features of Gaucher's disease (infantile, type 2) with normal glucocerebrosidase (D-glucosyl-N-acylsphingosine glucohydrolase, E.C.3.2.1.45) activity. Biochemical analysis was performed on leukocytes, cultured skin fibroblasts, and liver. Normal activity of glucocerebrosidase previously has been reported in an older child with juvenile onset (type 3) Gaucher's disease and attributed to a deficiency of a sphingolipid activator protein. These rare cases illustrate and expand our concept of Gaucher's disease and may have both diagnostic and therapeutic implications.

Evidence for two cDNA clones encoding human GM2-activator protein

Two cDNAs encoding GM2 activator, pGM2A (648 bp) and GAP (1093 bp), were isolated from human placenta lambda gt11 libraries. The DNA sequence of pGM2A from 1 to 302 was almost identical with GAP, but diverged from 303-648. PCR was used to demonstrate the presence of both species of GM2 activator in placental RNA. Both cDNAs hybridized to mRNAs of approximately 2.3 kb and to identical single bands on genomic Southern blots.

Sphingolipid activator protein 1 deficiency in metachromatic leucodystrophy with normal arylsulphatase A activity. A clinical, morphological, biochemical, and immunological study

A 7-year-old boy had clinical features of metachromatic leucodystrophy (MLD), however, an increased urinary sulphatide excretion was found in the presence of normal arylsulphatase A (and alpha-galactosidase A) activity. A rectal biopsy showed metachromatically staining storage macrophages as well as nonmetachromatic, but PAS-positive, submucosal neurons filled with membranous cytoplasmic bodies. These two types of storage material led to testing for a sphingolipid activator protein (SAP) deficiency. Loading tests with sulphatide and globotriaosylceramide showed deficient turnover of both sphingolipids in cultured fibroblasts. Using the Ouchterlony method, there was no reactivity between a described anti-SAP 1 antiserum and the patient's fibroblast extracts. This new case of SAP-1 deficient MLD was compared with the four cases of this variant known from the literature. Our results indicate that rectal biopsy morphology and lipid loading biochemistry should prove useful for the screening of SAP defects.

Intracellular transport and metabolism of sphingomyelin

SM is unique among the phospholipids because it is restricted to the lumenal aspect of organelles involved in the secretory and endocytic pathways. Given the intracellular sites of SM biosynthesis and hydrolysis, and the interconnections between these sites by vesicle-mediated transport pathways, the basic mechanism for maintaining the intracellular distribution of SM seems clear. It remains to be determined how SM metabolism and transport are coordinated to maintain the SM content of each organelle. For example, the size of the SM pool at the cell surface is maintained by regulation of at least five processes: transport of newly synthesized SM from the Golgi apparatus, plasma membrane lipid recycling, local SM synthesis, local SM hydrolysis, and SM transport from the cell surface to lysosomes. Although SM cannot undergo spontaneous transbilayer movement, SM metabolism generates both DAG, Cer and (indirectly) SPhB which can rapidly 'flip-flop', and thus gain access to the cytoplasmic leaflet of a membrane. It is of particular interest that these lipid species may be involved in the regulation of PK-C, suggesting that SM metabolism could play a role in signal transduction. However, physiological effects of endogenous Cer and SPhB remain elusive, even though the pharmacological effect of SPhB on PK-C is well established. Aside from the direct generation of second messengers, stimulation of SM hydrolysis has also been shown to induce cholesterol movement from the cell surface to intracellular membranes. It is not known whether this reflects the possibility that cholesterol may act as a second messenger. Alternatively, this phenomenon suggests that SM metabolism may cause rapid changes in the physical properties of the cell surface. For example, erythrocytes extensively treated with exogenously-added SMase will undergo endovesiculation It is tempting to speculate that any involvement of SM in the regulation of intracellular processes requires a combination of both the generation of biochemical second messengers and the alteration of membrane biophysical properties that can result from SM metabolism.

Sphingomyelin and derivatives as cellular signals

This comprehensive review was necessitated by recent observations suggesting that sphingomyelin and derivatives may serve second messenger functions. It has attempted to remain true to the theme of cellular signalling. Hence, it has focussed on the lipids involved primarily with respect to their metabolism and properties in mammalian systems. The enzymology involved has been emphasized. An attempt was made to define directions in which signals may be flowing. However, the evidence presented to date is insufficient to conclusively designate the mechanisms of stimulated lipid metabolism. Hence, the proposed pathways must be viewed as preliminary. Further, the biologic functions of these lipids is for the most part uncertain. Thus, it is difficult to presently integrate this sphingomyelin pathway into the greater realm of cell biology. Nevertheless, the present evidence appears to suggest that a sphingomyelin pathway is likely to possess important bioregulatory functions. Hopefully, interest in this novel pathway will grow and allow a more complete understanding of the roles of these sphingolipids in physiology and pathology.

Activator protein deficient Gaucher's disease. A second patient with the newly identified lipid storage disorder

A report is presented based on the biochemical and immunochemical studies of various tissues from a 15-year-old boy with a neuronopathic form of Gaucher's disease. Qualitative and quantitative lipid analyses revealed a storage of glucosylceramide. The striking feature was that, employing the usual assay methods, a normal activity of the lysosomal enzyme glucosylceramidase was revealed, despite massive lipid accumulation. Immunochemical assays of hepatic and splenic tissue extracts from this atypical Gaucher's patient disclosed the absence of A1 activator protein, which is necessary for the enzyme degradation of glucosylceramide in vivo. This is the second documented case of a patient presenting with glucosylceramide activator protein deficiency.

Sphingolipid activator protein deficiency in a 16-week-old atypical Gaucher disease patient and his fetal sibling: biochemical signs of combined sphingolipidoses

We describe a patient who presented shortly after birth with hyperkinetic behaviour, myoclonia, respiratory insufficiency and hepatosplenomegaly. Gaucher-like storage cells were found in bone marrow. A liver biopsy showed massive lysosomal storage morphologically different to that in known lipid storage disorders. Biochemically, the patient had partial deficiencies of beta-galactocerebrosidase, beta-glucocerebrosidase and ceramidase in skin fibroblast extracts, but the sphingomyelinase activity was normal. Glucosyl ceramide and ceramide were elevated in liver tissue. Loading of cultured fibroblasts with radioactive sphingolipid precursors indicated a profound defect in ceramide catabolism. Immunological studies in fibroblasts showed a total absence of cross-reacting material to sphingolipid activator protein 2 (SAP-2). The patient died at 16 weeks of age. The fetus from his mother's next pregnancy was similarly affected. The possibility that the disorder results from a primary defect at the level of SAP-2 is discussed. We have named this unique disorder SAP deficiency.

Saposin A: second cerebrosidase activator protein

Saposin A, a heat-stable 16-kDa glycoprotein, was isolated from Gaucher disease spleen and purified to homogeneity. Chemical sequencing from its amino terminus and of peptides obtained by digestion with protease from Staphylococcus aureus strain V-8 demonstrated that saposin A is derived from proteolytic processing of domain 1 of its precursor protein, prosaposin. Processing of prosaposin (70 kDa) also generates three other previously reported saposin proteins, B, C, and D, from its second, third, and fourth domains. Similar to saposin C, saposin A stimulates the hydrolysis of 4-methylumbelliferyl beta-glucoside and glucocerebroside by beta-glucosylceramidase and of galactocerebroside by beta-galactosylceramidase, mainly by increasing the maximal velocity of both reactions. Saposin A is as active as saposin C in these reactions. Saposin A has no significant effect on other sphingolipid and 4-methylumbelliferyl glycoside hydrolases tested. Saposin A has two potential glycosylation sites that appear to be glycosylated. After deglycosylation, saposin A had a subunit molecular mass of 10 kDa and was as active as native saposin A. However, reduction and alkylation abolished the activation. A three-dimensional model comparing saposins A and C reveals significant sequence homology between them, especially preservation of conserved acidic and basic residues in their middle regions. Each appears to possess a conformationally rigid hydrophobic pocket stabilized by three internal disulfide bridges, with amphipathic helical regions interrupted by helix breakers.

Complete amino-acid sequence of the naturally occurring A2 activator protein for enzymic sphingomyelin degradation: identity to the sulfatide activator protein (SAP-1)

The naturally occurring A2 activator protein for enzymic sphingolipid degradation is characterized by complete amino-acid sequence and carbohydrate content. It consists of 79 amino-acid residues and has a molecular mass of 8.875 kDa. The polypeptide chain contains 2 mol of N-acetylglucosamine, bound to asparagine in position 21, as well as 2 mol of galactose and mannose per mol protein. The primary structure of the A2 activator protein is identical to that of the sulfatide activator protein (SAP-1). Possible differences in the carbohydrate content are discussed.

Saposin D: a sphingomyelinase activator

Saposin D, a newly discovered heat-stable, 10 kDa glycoprotein, was isolated from Gaucher spleen and purified to homogeneity. Chemical sequencing from its amino terminus demonstrated colinearity between its amino acid sequence and the deduced amino acid sequence of the fourth domain of prosaposin, the precursor of saposin proteins. Saposin D specifically stimulates acid sphingomyelinase but has no significant effect on the other hydrolases tested.

Analysis of the multiple forms of Gaucher spleen sphingolipid activator protein 2

Gaucher spleen sphingolipid activator protein 2 was fractionated into concanavalin A binding- and non-binding fractions. These fractions each contained several bands on non-denaturing polyacrylamide gel electrophoresis (PAGE). The two fractions were further fractionated by electroblotting the proteins from preparative gels onto nitrocellulose, staining with Ponceau S to locate the bands of protein and then eluting the protein components from the nitrocellulose. A total of ten fractions, each containing only one or two major components, was collected. All of these subfractions activated beta-glucocerebrosidase and sphingomyelinase and most subfractions also activated beta-galactocerebrosidase. The structural relationship of the bands was investigated using endoglycosidase digestions. The results indicated that the two bands with the fastest mobility on non-denaturing PAGE did not contain any carbohydrate. The remaining bands showed only limited or partial digestion with endoglycosidase H and endoglycosidase D, but were readily hydrolysed with endoglycosidase F. The products of these digestions included bands with similar mobilities to the non-carbohydrate containing bands.

The precursor of sulfatide activator protein is processed to three different proteins

The enzymic degradation of a number of sphingolipids in the lysosomes is stimulated by small acid glycoproteins named activator proteins. We purified and sequenced a new protein, called component C, which seems to be related to sulfatide activator and to a recently described activator of glucosylceramidase (A1 activator) (Kleinschmidt, T., Christomanou, H. & Braunitzer, G. (1987) Biol. Chem. Hoppe-Seyler 368, 1571-1578). It consists of 78 amino acids and carries one carbohydrate chain at aparagine 20. Component C shows 21.5% sequence homology to sulfatide activator and 34.2% homology to A1 activator. Structural similarities between these three proteins have also been detected. Recently the cDNA sequence of the sulfatide activator precursor has been published (Dewji, N.N., Wenger, D.A. & O'Brien, J.S. (1987) Proc. Natl. Acad. Sci. U.S.A. 84, 8652-8656). We could align the protein sequences of sulfatide activator, A1 activator and component C with that of this large precursor protein. After minor corrections of the DNA sequence we obtained total fit. Thus it seems that three different proteins are derived from the sulfatide activator precursor by proteolytic processing. Possible processing sites were found on the precursor at sites adjacent to the N-termini and C-termini of the mature proteins. The processing of sulfatide activator was studied by Fujibayashi and Wenger (Fujibayashi, S. & Wenger, D.A. (1986) Biochim. Biophys. Acta 875, 554-562). Their data support our assumption that processing occurs by simultaneous cleavage at all possible sites.

N-terminal amino-acid sequence of a sphingolipid activator protein missing in a new human Gaucher disease variant

A naturally occurring non-enzymic sphingolipid activator protein (A1a activator) shown previously to be immunochemically not detectable in a new variant of human Gaucher disease (glucosylceramide-lipidosis) without glucosylceramidase deficiency was characterized by partial sequence analysis. The N-terminal amino-acid sequence of the A1a activator--a glycoprotein with high carbohydrate content--could be determined up to position 38. About 20% of the polypeptide chain are shorter by two amino-acid residues at the N-terminal end. Position 22 seems to be occupied by a carbohydrate-binding asparagine. The N-terminus of the A1a activator does not show any homology with the activator for the enzymic sulfatide degradation.

Immunochemical characterization of two activator proteins stimulating enzymic sphingomyelin degradation in vitro. Absence of one of them in a human Gaucher disease variant

Two nonenzymic activator proteins shown previously to strongly stimulate enzymic sphingomyelin degradation in vitro were purified from human Gaucher type 1 and control spleen. Activator A1 (molecular mass 6,500 Da) had affinity for ConA-Sepharose, while activator A2 (molecular mass 3,500 Da) did not. Monospecific antibodies to each activator protein were prepared in rabbits by immunization with protein purified from type 1 Gaucher spleen. A1 and A2 activators from Gaucher type 1 spleen were shown to be immunochemically identical to A1 and A2 activators from control spleen. However, A1 and A2 activators, whether isolated from Gaucher type 1 or control spleen, were shown to be distinct proteins. Immunochemical examination of all collected fractions during the purification revealed the existence of a third activator (molecular mass 6,000 Da), which was antigenically identical to A1 activator but had no affinity for ConA-Sepharose. The two forms of A1 activator showed similar mobility on immunoelectrophoresis differing from that of A2 activator. Fibroblast extracts from controls and patients with different variants of Gaucher disease were investigated using immunodiffusion against antisera to A1 or A2 activator. In contrast to normal and Gaucher (types 1, 2 and 3) cell extracts, those of a Gaucher patient with normal glucosylceramidase activity had no visible precipitin line towards the antiserum against the two forms of A1 activator. The lack of crossreacting material to antibodies against A1 activator was confirmed by radial immunodiffusion and rocket immunoelectrophoresis. A1 activator stimulated the basal glucosylceramidase activity 5-6 fold in fibroblasts from this patient, whereas the normal effect was only a 1.2-1.5-fold stimulation. The immunological results together with the biochemical data provide evidence for the lack of an activator protein in a variant form of human Gaucher disease for the first time.

Preliminary evidence for a processing error in the biosynthesis of Gaucher activator in mucolipidosis disease types II and III

Activator protein (AP), which stimulated fibroblast sphingomyelinase activity, was isolated from the spleen of a patient with Gaucher's disease type I by the combined techniques of heat and alcohol denaturation, DEAE-cellulose column chromatography, gel filtration, preparative polyacrylamide-gel electrophoresis and decyl-agarose chromatography. Urea/sodium dodecyl sulphate (SDS)/polyacrylamide-gel electrophoresis showed two bands, one with an Mr of approx. 3,000 and the other with an Mr of 5,000-6,500. Similarly, SDS/polyacrylamide-gel electrophoresis performed in the absence of urea revealed the presence of two components, one of which adsorbed to a concanavalin A (Con A) column. Both components stimulated sphingomyelinase activity. On a non-denaturing polyacrylamide gel containing Triton X-100, four major components, two of which bound to Con A, were detected with the dye Stains-All. Cross-reacting material (CRM) to polyclonal Gaucher spleen AP antibodies was detected in normal fibroblasts and in fibroblasts from patients with sphingomyelinase and beta-glucocerebrosidase deficiency states (Niemann-Pick and Gaucher's diseases respectively). CRM in normal fibroblasts adsorbed to Con A columns and had the same mobility on SDS/polyacrylamide-gel electrophoresis as Con A-adsorbing Gaucher spleen AP. Normal AP was not observed in mucolipidosis type II (I-cell disease) fibroblasts; instead, extracts from these cells revealed the presence of two closely migrating bands with higher Mr values than normal fibroblast CRM. Furthermore, extracts of media from I-cell fibroblast cultures, but not from control or Gaucher fibroblast cultures, contained AP activity towards sphingomyelinase and beta-glucocerebrosidase. Fibroblasts from a patient with mucolipidosis type III (pseudo-Hurler polydystrophy) showed an intermediate pattern consisting of normal as well as the higher-Mr CRM. Our data provide evidence for the existence of AP in cultured skin fibroblasts and suggest that these proteins may be targetted to the lysosome by post-translational modification in a similar manner to that reported for lysosomal enzymes.