Effect of Saposins A and C on the Enzymatic Hydrolysis of Liposomal Glucosylceramide*

The degradation of glucosylceramide in lysosomes is accomplished by glucosylceramidase with the assistance of, at least, another protein, saposin C (Sap C), which is generated from a large precursor together with three other similar proteins, saposins A, B, and D. In the present study, we have examined the effects of saposins on the enzymatic hydrolysis of glucosylceramide inserted in large and small phospholipid liposomes. The glucosylceramide contained in large unilamellar vesicles (LUV) was degraded by glucosylceramidase at a rate 7–8-fold lower than glucosylceramide inserted in small unilamellar vesicles (SUV). The separate addition of either Sap A or Sap C to the LUV system partially stimulated the sphingolipid degradation while saposins B and D had no effect. In the presence of both Sap A and Sap C, the rate of sphingolipid degradation was higher than the sum of the rates with the two saposins individually, indicating synergism in their actions. The stimulatory effect of the two saposins depended on the incorporation of an acidic phospholipid such as phosphatidylserine (PS) into LUV. The characteristics of glucosylceramidase activation by Sap C were different from those of Sap A. Sap C increased the rate of hydrolysis of both the artificial water soluble substrate, 4-methylumbelliferyl-β-d-glucopyranoside, and the lipid substrate, glucosylceramide, while Sap A only stimulated degradation of the sphingolipid. Also the binding properties of Saps A and C were markedly different. At acidic pH values, Sap C bound to PS-containing LUV and promoted the association of glucosylceramidase with the membrane. In contrast, Sap A had poor affinity for the membrane even in the presence of glucosylceramide; moreover, Sap A did not potentiate the capacity of Sap C to mediate glucosylceramidase binding. In conclusion, our results show that both Sap A and Sap C are required for maximal hydrolysis of glucosylceramide inserted in PS-containing LUV, that their effects are synergistic, and that their mode of action is different. Sap C is responsible for the membrane binding of glucosylceramidase, while Sap A stimulation is possibly related to its effect on the conformation of the enzyme. It can be envisaged that Sap A in conjunction with Sap C might have a physiological role in glucosylceramide degradation.

Sap C is released together with three other similar proteins, Saps A, B, and D, from a common precursor called prosaposin (1-4, 8 -10). All saposins appear to be involved in the catabolism of sphingolipids (1)(2)(3)(4). A patient lacking the four saposins in consequence of a mutation in the prosaposin initiation codon showed a combined sphingolipid storage disorder (11,12). A selective deficiency of Sap C caused glucosylceramide accumulation in a variant form of Gaucher disease (13). Mutations affecting the coding region of Sap B caused a variant form of metachromatic leukodystrophy with storage of sulfatides (14,15). An isolated deficiency of either Sap A or D has not been reported so far, and thus the physiological consequences of the absence of one of these two saposins are unknown.
To explore the role of Sap A, the effect of this saposin on the degradation of sphingolipids has been tested in vitro. It was reported that Sap A is able to increase substrate hydrolytic rates of glucosylceramidase and galactosylceramidase (16), an observation that prompted the authors to propose an involvement of Sap A in the metabolism of gluco-and galactosylceramides (16). Another group claimed that Sap A was able to stimulate the glucosylceramidase activity at high concentrations (17) while at low physiological concentrations, it bound to glucosylceramidase without activating effects (18). It was thus considered unlikely that Sap A is an important activator of the enzyme.
While the function of Sap A is still debated, that of Sap C as glucosylceramidase activator is well assessed (1)(2)(3)(4)(5)13). In vitro, the Sap C-induced stimulation of glucosylceramidase requires the presence of a negatively charged phospholipid such as phosphatidylserine (PS) (5,19). In earlier studies, we provided evidence that Sap C was able to favor the association of glucosylceramidase with PS membranes and that the physical characteristics of PS vesicles markedly influenced their interaction with glucosylceramidase (20,21).
Most reports dealing with the reconstitution of glucosylceramidase activity by saposins have involved enzyme assays in which the substrate was an artificial water-soluble compound, 4-methylumbelliferyl-␤-D-glucopyranoside (MU-Glc) (17,18,20). When the characteristics of glucosylceramidase activity were investigated with the lipid substrate glucosylceramide inserted into sonicated small unilamellar vesicles (SUV), high activity was achieved upon addition of anionic phospholipids to the lipid bilayer (22,23). The rate of glucosylceramide degradation depended on the binding of glucosylceramidase to SUV, and the presence of acidic phospholipids in the vesicles promoted enzyme binding even in the absence of saposins (23). Since at least Sap C is essential in vivo for the enzymatic degradation of glucosylceramide (13), the question arises whether small liposomes are a reliable membrane model for investigations on the involvement of saposins in sphingolipid hydrolysis. It is well known that the lipid packing in small vesicles is not optimal (24,25); the lipid surface disorder might facilitate the insertion of glucosylceramidase into the bilayer and obliterate the requirement for saposins.
To have a better insight into the molecular mechanism of glucosylceramide degradation and the interactions involved, we have compared the characteristics of the hydrolysis of the sphingolipid inserted either in large or small vesicles. The effects of saposins on the catalytic process and their mode of action have been investigated.

EXPERIMENTAL PROCEDURES
Materials-Glucosylceramide, purified from spleen of patients with Gaucher's disease, was labeled with tritium in the glucose moiety according to McMaster and Radin (26). Phosphatidylcholine (PC) from egg yolk and PS from bovine brain were from Avanti Polar Lipids, Inc. (Alabaster, AL). L-␣-dipalmitoyl [dipalmitoyl-1-14 C]-PC (110 mCi/mmol) was from NEN Research Products, DuPont de Nemours (Germany). Cholesterol (chol) and MU-Glc were from Sigma. All other chemicals were of the purest available grade.
Saps A, B, C, and D Preparation-Saps B, C, and D were purified from spleens of patients with Type 1 Gaucher's disease following a procedure previously reported (27); it consisted of heat treatment of a water homogenate, ion exchange chromatography on DEAE-Sephacel, gel filtration on Sephadex G-75, and reverse-phase high pressure liquid chromatography on a protein C4 column (Vydac). Sap A, also from spleens of Gaucher's disease patients, was purified according to the procedure of Morimoto et al. (16). The purity of the final saposin preparations was verified by N-terminal sequence analysis, SDS-polyacrylamide gel electrophoresis, and Western blotting.
Saposin Antibodies-Antibodies against Saps A and C were raised in rabbits by injecting 200 g of the individual saposins three times over a period of 2 months.
Glucosylceramidase Preparation and Assay-Glucosylceramidase was purified from human placenta following the procedure described by Murray et al. (28).
To measure the glucosylceramidase activity, either the water-soluble substrate, MU-Glc, or the lipid substrate, glucosylceramide, were utilized. In the first case, the assay mixture contained in a final volume of 0.1 ml: buffer A, pH 5.0 (10 mM acetate, 150 mM NaCl, 1 mM EDTA), 2.5 mM MU-Glc, 3 ng of purified placental glucosylceramidase (about 9 units), 5 g of albumin, and 100 g of LUV without glucosylceramide (see below). The assay mixtures were incubated for 30 min at 37°C. The extent of reaction was estimated fluorometrically as described previously (29).
In the second case, the assay mixture contained in a final volume of 0.1 ml: buffer A, pH 4.5 or pH 5.0, 3 ng of glucosylceramidase, 5 g of albumin, and 100 g of glucosylceramide-containing vesicles (see below). Glucosylceramide included in the liposomes was supplemented with the labeled compound to a specific activity of 3000 dpm/nmol. The assay mixtures were incubated for 30 min at 37°C. The incubation was terminated by the addition of 0.4 ml of chloroform/methanol (2:1) and 50 l of a 0.1% glucose solution. After shaking and centrifugation at 4000 rpm, the enzymatically released [ 3 H]glucose present in the aqueous phase was estimated as described (23).
In some cases also, a detergent assay has been utilized. It contained 0.1/0.2 M citrate/phosphate buffer, pH 5.6, 2.5 mM MU-Glc, 0.1% (v/v) Triton X-100, and 0.25% (w/v) sodium taurocholate (21). 1 unit of glucosylceramidase is defined as the amount of enzyme that hydrolyzes 1 nmol of MU-Glc/h in the detergent assay.
Vesicle Preparation-LUV were prepared by filter exclusion using a high pressure extrusion apparatus (Lipex Biomembranes, Vancouver, Canada) as described previously (21,30). In short, the dry lipids were dispersed by Vortex mixing in 2 mM L-histidine, 2 mM TES, 150 mM NaCl, 1 mM EDTA, pH 7.4. The suspension was submitted to 10 cycles of freezing and thawing and then extruded 15 times through two stacked 0.1-m pore size polycarbonate filters (Nucleopore Corp., Pleasanton, CA). Four kinds of LUV were prepared: 1) LUV without glucosylceramide (chol:PC:PS, molar ratio 25:55:20); 2) LUV without glucosylceramide and without PS (chol:PC, molar ratio 25:75); 3) LUV with glucosylceramide but without PS (chol:PC:glucosylceramide, molar ratio 25:70:5); 4) LUV with glucosylceramide and PS (chol:PC:PS:glucosylceramide, molar ratio 25:50:20:5). The last lipid mixture was also utilized to prepare SUV by sonication (23). Briefly, the lipid suspension was sonicated under nitrogen in a Branson B15 Sonifier (3 min with a cup horn at a power setting of 100 watts, followed by 6 min with a microtip at a power setting of 100 watts). The preparation was centrifuged at 100,000 ϫ g for 30 min, and the supernatant was used for the experiments.
All vesicles were supplemented with trace amounts of labeled PC, and their concentration was determined by radioactivity measurements.
Binding of Saposins and of Glucosylceramidase to Vesicles-For saposins binding studies, Sap A (10 g) and Sap C (10 g), separately or in combination, were incubated with LUV (100 g total lipid) in 0.2 ml of buffer A adjusted to pH 4.5 at 37°C for 15 min. The mixture was then centrifuged with a 42.2 Ti rotor (Beckman) in polycarbonate centrifuge tubes (7 ϫ 20 mm) at 80,000 ϫ g for 1 h. More than 95% of vesicles were found in the pellet as determined by radioactivity measurements. Conversely, in control experiments without liposomes, the saposins were found not to sediment during the ultracentrifugation. After separation of the supernatant, the pelletted vesicles were rinsed once with 0.2 ml of buffer A and finally resuspended in 0.2 ml of the same buffer. Saposins in the initial supernatant (free saposins) and in the resuspended vesicles (liposome-bound saposins) were precipitated by addition of 20 g of albumin and 2 ml of cold acetone. After 2 h at Ϫ20°C, the precipitated saposins were collected by centrifugation, solubilized with 30 l of electrophoresis sample buffer and identified by SDSpolyacrylamide gel electrophoresis and immunoblotting (see below).
For glucosylceramidase binding studies, the enzyme (1000 units) was incubated at 37°C for 15 min with LUV (100 g total lipid), and the indicated amounts of saposins in 0.2 ml of buffer A, pH 4.5, were added with 10 mM dithioerythritol and 20 g of albumin. The mixture was centrifuged as above. The amount of glucosylceramidase in the supernatant was determined by measuring the enzyme activity according to the detergent assay (see above). The amount of liposome-bound glucosylceramidase was expressed relative to the amount of enzyme in the supernatant of a sample centrifuged in the absence of liposomes.
SDS-Polyacrylamide Gel Electrophoresis and Immunoblotting-A discontinuous Tricine/SDS-polyacrylamide gel electrophoresis system was performed with 16.5% acrylamide (31). After electrophoresis, the saposins were electroblotted to polyvinylidene difluoride membranes (Bio-Rad) and detected with antibodies against each saposin using a Vectastain ABC kit (Vector Labs, Inc., Burlingame, CA) according to the manufacturer instructions. Antibodies raised against Sap A did not cross-react with Sap C and vice versa.

Influence of Liposome Size on the Enzymatic Hydrolysis of
Liposomal Glucosylceramide-We have previously shown that PS LUV, as opposed to PS SUV, are unable to fully stimulate glucosylceramidase activity toward the artificial substrate MU-Glc (21). Here, we have investigated whether the liposome size has a similar effect on the enzyme activity toward glucosylceramide, the physiological enzyme substrate, inserted in vesicles containing, besides glucosylceramide (5%), cholesterol, PC, and an acidic phospholipid such as PS. Large and small liposomes of identical composition were utilized. As shown in Fig. 1, glucosylceramidase at acidic pH values was able to efficiently degrade the sphingolipid contained in SUV but not that in LUV; glucosylceramidase activity measured with the LUV system was only 10 -15% of the activity measured with the SUV system. This same pattern was observed when the percentage of the sphingolipid in the bilayer was increased from 5 to 10% (data not shown). To reduce changes in the surface of the phospholipid vesicles, the lower glucosylceramide concentration was utilized for further investigations.
Influence of Saposins on the Enzymatic Hydrolysis of Liposomal Glucosylceramide-The ability of the four saposins to promote the hydrolysis of glucosylceramide in LUV was next investigated ( Fig. 2A). Sap C and Sap A increased the degradation rate of the sphingolipid while Saps B and D had no effect. Despite the stimulation induced by either Sap A or Sap C, the hydrolysis rate of glucosylceramide inserted in LUV was still considerably lower than that achieved with the SUV system. To explore the possibility of an additive or synergistic effect, Sap A and Sap C were simultaneously added to the enzymatic assay. As shown in Fig. 2B, Sap A caused an additional increase of the hydrolysis rate over that obtained with 5 M Sap C and vice versa, the total stimulation being higher than the sum of the single effects of the two saposins.
We then investigated whether Sap A and Sap C could further increase the high degradation rate of glucosylceramide inserted in PS-containing SUV. As shown in Fig. 3A, in this system, the two saposins also stimulated enzyme activity (about 30% activation). In the LUV system, the stimulatory effects, and especially the synergism of Sap A and Sap C in the admixture, were much more evident (700 -800% activation).
The optimum pH for the sphingolipid degradation was the same both in the presence and in the absence of saposins, namely pH 5.0 for the SUV system and pH 4.5 for the LUV system.
To check the importance of PS for glucosylceramidase stim-ulation by Sap A and Sap C, the enzyme assays were repeated utilizing glucosylceramide-containing LUV devoid of PS. Under these conditions, the enzyme activity decreased dramatically, even in the presence of both saposins (Fig. 3A).
The effect of Sap A on the glucosylceramide hydrolysis observed in this paper was unexpected; in fact, we had previously found that the hydrolysis of the water-soluble enzyme substrate, MU-Glc, was unaffected by Sap A (32). Fig. 3B confirms that, in the presence of PS-containing LUV, Sap A poorly activates the MU-Glc hydrolysis. The enzyme activity toward this substrate is fully stimulated by Sap C, independently of Sap A. It is thus evident that the Sap A stimulation is specific for glucosylceramide hydrolysis.
Binding Properties of Sap A, Sap C, and Glucosylceramidase-We recently found that Sap A, as opposed to Sap C, fails to bind tightly to phospholipid membranes (32). Since Sap A, especially in combination with Sap C, is able to promote the degradation of glucosylceramide inserted in LUV, we considered the possibility that the sphingolipid and/or Sap C could increase the affinity of Sap A for the lipid surface. Sap A, either alone or in admixture with Sap C, was thus incubated with PS-containing LUV with or without glucosylceramide; the LUV-associated saposins were then separated from nonbound saposins by ultracentrifugation. The identity of the saposins present in the supernatant and in the pellet was evaluated by immunoblotting with specific anti-Sap A or anti-Sap C antibodies. Fig. 4 shows that Sap A poorly associates with the vesicles (compare lanes 2 and 4 with lanes 3 and 5), even in the presence of glucosylceramide (Fig. 4A) and Sap C. Centrifugation separated the two combined proteins; most of Sap A was recovered in the supernatant while most of Sap C was bound to the vesicles either in the presence or in the absence of the sphingolipid (Fig. 4, A and B, lane 5). These results clearly show that the binding properties of Sap A and Sap C are quite different, irrespective of whether the phospholipid membranes contain

glucosylceramide.
We had previously observed that Sap A is unable to promote the binding of glucosylceramidase to LUV composed of cholesterol, PC, and PS (32). We have now investigated whether Sap A could at least potentiate the Sap C capacity to mediate the glucosylceramidase association with PS-containing membranes. As shown in Table I, at pH 4.5 (the optimum for enzyme activity in the LUV system), only a minor percentage (10%) of glucosylceramidase is bound to the bilayer when saposins are omitted. Consistent with our earlier observations (32), Sap C is able to promote the extensive enzyme association with LUV. The further addition of Sap A does not increase the adsorption of glucosylceramidase, even at suboptimal Sap C concentrations. In conclusion, Sap C is the only saposin responsible for the enzyme interaction with membranes.
It was confirmed that the presence of an acidic phospholipid in the bilayer is essential for enzyme binding; Sap C, either alone or in combination with Sap A, is unable to promote a tight association of the enzyme with liposomes devoid of PS (Table I). DISCUSSION The purpose of this study was to evaluate the involvement of saposins in the enzymatic degradation of glucosylceramide and to consider the mechanism whereby saposins and lipids may modulate glucosylceramidase activity toward its natural substrate. Glucosylceramide was inserted in LUV having the lipid composition estimated to be characteristic of most membranes, namely containing cholesterol and phospholipids (1:3). In many biological studies, the average size of the liposomes used as the model of biological membranes is a critical parameter for the interpretation of the results. In the present research, we have found that glucosylceramide can be degraded more efficiently when inserted in PS-containing SUV rather than in PS-con-taining LUV. This result stresses the importance of the assay system chosen for the study of glucosylceramidase activity and properties. The different performance of the LUV and SUV systems can only be explained on the basis of lipid organization. In fact, there were no compositional differences between vesicles. Probably, the loose packing of the lipids in the outer surface of small vesicles favors the enzyme interaction with glucosylceramide. The influence of the bilayer curvature on glucosylceramidase activity might have a physiological relevance also on consideration that cerebrosides, such as the glucosylceramide substrate, are possibly involved in the formation and maintenance of highly curved membranes (33).
When glucosylceramide is contained in LUV, where the lipid packing arrangement approaches that of a planar bilayer, glucosylceramidase poorly interacts with its natural substrate unless two saposins, Sap C and A, are added to the liposomal system. Each of the two saposins by itself has a partial effect; only in admixture do they increase the rate of glucosylceramide hydrolysis up to a level comparable with that observed in the SUV system. The combined effects of the two saposins were always greater than the sum of the stimulation induced by the two saposins separately. Such synergism suggests a different mode of action of Saps A and C. The idea that the two saposins act through different mechanisms is strongly supported by their distinct effects on the glucosylceramidase activity toward two substrates; Sap A stimulates the hydrolysis of glucosylceramide, but not that of the artificial substrate MU-Glc, while Sap C stimulates the hydrolysis of MU-Glc more than that of glucosylceramide. We previously showed that Sap C has the greatest membrane affinity of the four saposins (32). An acid-induced increased hydrophobicity has been demonstrated as the first step in the Sap C association with membranes (32). After interaction with Sap C, PS-containing LUV acquire the capacity to bind glucosylceramidase (20). The present observation that the Sap C-mediated adhesion of glucosylceramidase to lipid bilayers results in only a partial stimulation of glucosylceramide hydrolysis indicates that additional factors are required for an optimal interaction between the enzyme and its lipid substrate. The presence of Sap A appears to fulfill this requirement.
Differently from Sap C, Sap A binds poorly to bilayers (present work, and Ref. 32). In parallel, Sap A neither has a significant effect by itself on the interaction of the enzyme with membranes nor potentiates the Sap C capacity to promote this interaction. To explain the stimulation of glucosylceramide hydrolysis by Sap A, a mechanism different from that of Sap C can be envisaged. As proposed by other groups (16,17), Sap A might form a complex with glucosylceramidase, causing a conformational change in the structure of the enzyme that results in increased catalysis. In this context, the role of Sap A would be mainly related to the enzyme-glucosylceramide interaction while that of Sap C to the membrane localization of the enzyme.
In general, the activity of membrane-bound enzymes, such as glucosylceramidase, is modulated by their lipid environment. Glucosylceramidase has a strong affinity for acidic phospholipids such as PS (5); these lipids are essential for the binding of glucosylceramidase to liposomes (23,29) and for the reconstitution of the enzyme activity either in the presence or absence of saposins (present paper, and Refs. 5 and 20). The presence of PS markedly increases also the extent of Sap C binding and the consequent perturbation of the bilayer structure (32). Thus, the Sap C stimulation of glucosylceramidase is possibly related to the Sap C-induced destabilization of membranes, which might favor the interaction of the enzyme with PS. Our past (29,32) and present results suggest that a change in PS content might regulate the amount of Sap C and glucosylceramidase in lysosomal membranes, allowing acidic phospholipids to have a dynamic function in the glucosylceramide degradation.
In contrast with our findings, some authors have claimed that Sap C does not interact with acidic phospholipids and that this saposin stimulates glucosylceramidase by binding directly to the enzyme (34). The two groups who previously tested the effect of the simultaneous addition of Saps A and C on the glucosylceramidase activity utilized MU-Glc but not glucosylceramide as substrate; under these conditions, they found that Sap A, although being able by itself to activate glucosylceramidase, had no additive effect over the optimal catalytic rate achieved by saturating amounts of Sap C (18,34). The authors concluded that Sap A and Sap C have the same mode of action and that the two saposins compete for a binding site on glucosylceramidase (18,34). In our experience, it is critically important to use appropriate conditions for testing the effects of saposins. We could evidence the additivity of the Sap A and Sap C effects only when the degradation of glucosylceramide was measured (see Fig. 3). We did not observe any Sap A stimulation of the MU-Glc hydrolysis, a finding in agreement with Kondoh et al. (35).
When Sap C is absent in consequence of a genetic defect, the accumulation of intralysosomal glucosylceramide and a Gaucher-like disease have been observed (13). In this variant of Gaucher's disease, glucosylceramidase and Sap A are present at almost normal levels. For authors who believe that Sap A and Sap C act through the same mechanism, this observation argues against a possible physiological function of Sap A as a glucosylceramidase activator (18). Theoretically, if the two saposins have similar properties, a normal Sap A might accomplish the function of a deficient Sap C. By contrast, if the two saposins have different mechanisms of action, as our results indicate, the glucosylceramide accumulation caused by the absence of Sap C might not be prevented by the presence of Sap A.
In conclusion, our work provides new information regarding the influence of saposins on glucosylceramide degradation. According to our findings, glucosylceramidase activity toward its lipid substrate can be modulated by several factors, the more important being the presence of two saposins, Sap C and Sap A, and the membrane structure and composition. The fact that Sap A and Sap C can synergistically affect glucosylceramide hydrolysis suggests the possibility that saposins might accomplish their physiological function not only as single proteins but also in conjunction with each other.