Hex (β-hexosaminidase) is a soluble glycohydrolase involved in glycoconjugate degradation in lysosomes, however its localization has also been described in the cytosol and PM (plasma membrane). We previously demonstrated that Hex associated with human fibroblast PM as the mature form, which is functionally active towards GM2 ganglioside. In the present study, Hex was analysed in a lysosomal membrane-enriched fraction obtained by purification from highly purified human placenta lysosomes. These results demonstrate the presence of mature Hex associated with the lysosomal membrane and displaying, as observed for the PM-associated form, an acidic optimum pH. When subjected to sodium carbonate extraction, the enzyme behaved as a peripheral membrane protein, whereas Triton X-114 phase separation confirmed its partially hydrophilic nature, characteristics which are shared with the PM-associated form of Hex. Moreover, two-dimensional electrophoresis indicated a slight difference in the pI of β-subunits in the membrane and the soluble forms of the lysosomal Hex. These results reveal a new aspect of Hex biology and suggest that a fully processed membrane-associated form of Hex is translocated from the lysosomal membrane to the PM by an as yet unknown mechanism. We present a testable hypothesis that, at the cell surface, Hex changes the composition of glycoconjugates that are known to be involved in intercellular communication and signalling.
Glycohydrolases were believed to be mainly concentrated inside endosomal/lysosomal compartments, where they were involved in the degradation of macromolecules into monomeric subunits [1,2]. Nevertheless, previous studies showed the presence of glycohydrolases at the PM (plasma membrane) and point to their potential role in modifying glycosphingolipids during signalling and cellular communication [3,4]. For instance, co-culture experiments have provided evidence for the activity of the PM sialidase Neu3 towards gangliosides exposed to the surface of adjacent cells has been demonstrated previously [5,6]. Moreover, the presence of the lysosomal sialidase Neu1 anchored to both lysosomal membrane and the PM has been established, and an important immunoregulatory role for this enzyme has been suggested [7,8]. In particular, the translocation of the sialidase Neu1 from lysosomes to the PM has been observed during T-cell activation, where its ability to desialylate GM3 is required for IL-4 (interleukin-4) production in activated T-cells [9–11].
Hex (β-hexosaminidase, EC 126.96.36.199) is an acidic glycohydrolase that catalyses the cleavage of terminal β-N-acetylglucosamine or β-N-acetylgalactosamine from a broad range of glycoconjugated substrates in lysosomes [12,13]. However, the presence of Hex has also been described in the cytosol  and associated with the PMs of different cell populations, such as ascidia and Drosophila melanogaster sperm [15–17], human erythrocytes , leukaemic cells , lymphocytes and monocytes purified from peripheral blood of multiple sclerosis patients . In normal human tissue, there are two major Hex isoenzymes: the heterodimer Hex A (pI=4.8) and the homodimer Hex B (pI=6.9) . Minor forms of Hex, characterized by their intermediate pI values, have also been described in both normal and pathological cells . Hex isoenzymes originate from the association of two different subunits, α and β, which are encoded by two evolutionarily related genes which have been mapped to chromosomes 15 and 5 respectively [23,24]. Subunit dimerization occurs in the endoplasmic reticulum, where neo-synthesized precursor forms of Hex A (αβ) and Hex B (ββ) undergo extensive post-translational modifications, such as glycosylation, and culminates in the final proteolytic processing in lysosomes. The pro-α subunit is cleaved into αm and αp polypeptides, whereas the mature β-subunit is composed of three polypeptides, βa, βb and βp [21,25].
We previously demonstrated the existence of a mature form of Hex A associated with the PM of human fibroblasts and proved its ability to hydrolyse the natural substrate GM2 . This observation suggests that PM-associated Hex is of lysosomal origin, as the final maturative process of this enzyme only occurs in lysosomes . It is evident that the presence of Hex, as well as other acidic glycohydrolases of lysosomal origin, presents new perspectives on the role of these enzymes.
In the present study, we demonstrate the association of mature Hex with the lysosomal membrane. To perform this, highly purified lysosomes were further processed in order to remove the soluble content and to recover the purified lysosomal membrane-enriched fraction. Moreover, the nature of the association of Hex with the lysosomal membrane was further elucidated. Our results show that a fraction of the fully processed forms of Hex A and Hex B isoenzymes are non-covalently attached to the inner aspect of the lysosomal membrane.
MUGS (4-methylumbelliferyl-N-acetyl-β-D-glucosaminide-6-sulfate) was purchased from Toronto Research Chemicals. MUG (4-methylumbelliferyl-N-acetyl-β-D-glucosaminide), 4-methylumbelliferyl-α-mannopyranoside, 4-methylumbelliferyl-β-glucoside, 4-methylumbelliferone, protease inhibitor cocktail for mammalian cell extracts, Triton X-114, NP-40 (Nonidet P40), ASB-14, HRP (horseradish peroxidase)-conjugated monoclonal anti-rabbit IgG antibody, HRP-conjugated polyclonal goat anti-mouse IgG antibody and HRP-conjugated monoclonal anti-goat/sheep IgG antibody were purchased from Sigma. Monoclonal mouse anti-LAMP-2 (lysosome-associated membrane protein-2) antibody was from Santa Cruz Biotechnology. DEAE-cellulose was from Whatman Biochemicals. ECL® (enhanced chemiluminescence), IPG (immobilized pH gradient) strips, ExcelGel 2D Homogeneous 12.5, 2D Clean-Up Kit and Carbamylate Calibration Kit for 2D were from Amersham Bioscencies. Trans-Blot nitrocellulose membrane, bovine serum albumin and Bio-Rad protein assay reagent were from Bio-Rad. Centricon YM-10 centrifugal filters were from Millipore. All other reagents were of analytical grade.
Lysosomal membrane-enriched fraction preparation
PL (purified human placenta lysosome) (2 ml) obtained from fresh human placenta using a Percoll gradient as described previously  was thawed and mixed with 40 μl of protease inhibitor cocktail. The mixture was sonicated (three times for 15 s each) and diluted with 10 mM Na/P buffer (sodium phosphate buffer) (pH 6.0) to a final volume of 10 ml. This lysosomal homogenate was subjected to ultracentrifugation at 48000 rev./min for 2 h at 4°C h using a Beckman Optima Max MLS-50 rotor, and the supernatant fraction (S1) and the pellet fraction that contained the enriched lysosomal membranes (P1) were separated. The P1 fraction was resuspended in 0.5 ml of 10 mM Na/P buffer (pH 6.0) in the presence of protease inhibitors, sonicated as above and diluted with the same buffer to a final volume of 10 ml, before centrifugation at 40000 rev./min for 2 h at 4°C using a Beckman Optima Max MLS-50 rotor. This yielded the supernatant S2 and pellet P2 fractions. The latter fraction was extracted once more as detailed above and the supernatant S3 and pellet P3 fractions were obtained.
Determination of enzyme activity and protein concentration
Total Hex and Hex A activity was measured using 3 mM MUG and 3 mM MUGS respectively in 0.1 M citric acid/0.2 M disodium phosphate buffer (pH 4.5) . The optimum pH of Hex in the lysosomal membrane-enriched fraction and Hex isoenzymes, recovered after DEAE-chromatography, was determined using 3 mM MUG in 0.1 M citric acid/0.2 M disodium phosphate buffer at different pH values (pH 3.5–7.5). The thermal stability of the Hex isoenzymes, recovered after DEAE-chromatography, was determined by incubating the samples at 52°C for 1 h. Samples were cooled on ice for 1 h and then assayed for Hex activity by using the MUG substrate as detailed above.
The activities of acid α-Man (α-mannosidase, EC 188.8.131.52) and β-Gluc (β-glucosidase, EC 184.108.40.206) were determined using solutions containing the corresponding 4-methylumbelliferyl-glycoside substrates (3 mM) as described previously [29,30]. The fluorescence of the liberated 4-methylumbelliferone was measured using a PerkinElmer LS B50 fluorimeter (excitation at 360 nm and emission at 446 nm). One unit is the amount of enzyme that hydrolyses 1 μmol of substrate/min at 37°C.
The protein concentration was determined using the method of Bradford , with bovine serum albumin used as a standard. Specific activity is expressed as m-units per mg of protein.
The soluble lysosomal content fraction S1 and the lysosomal membrane-enriched fraction P3 were resuspended in 10 mM Na/P buffer (pH 6.0) containing 0.1% NP-40, sonicated and centrifuged at 14000 g for 15 min. The resultant supernatants were analysed by ion-exchange chromatography on DE-52 DEAE-cellulose as described previously . Briefly, a 1 ml column was equilibrated with 10 mM Na/P buffer (pH 6.0) and the protein retained by the column was eluted using a linear 0.0–0.5 M NaCl gradient in 50 ml of 10 mM Na/P buffer (pH 6.0). Finally, the proteins were eluted with 1.0 M NaCl in 10 mM Na/P buffer (pH 6.0). The flow rate was 1 ml/min. Fractions (1 ml each) were collected and assayed for Hex activity using the MUG and MUGS substrates as detailed above.
SDS/PAGE and Western blotting analysis
PL and the fractions resulting from the lysosomal membrane-enrichment procedure (P1 and S1; P2 and S2; and P3 and S3) were subjected to SDS/PAGE (12% gels) under reducing conditions as described by Laemmli . Proteins were transferred on to nitrocellulose and the α and β peptides of Hex were detected using an anti-α-subunit antibody (1:2500 dilution) (raised against a mixture of synthetic peptides of the sequence of the mature human enzyme) and a goat anti-β-subunit antibody (1:2500 dilution) raised against Hex B purified from human placenta . The blot was also developed with a monoclonal antibody raised against LAMP-2 (1:200 dilution). Blots were analysed using the ECL® detection system and densitometric analysis was performed using the ImageMaster2D software (Amersham Biosciences).
Analysis of the amino acid sequences of Hex α-subunit (GenBank® Entrez protein accession number P06865) and β-subunit (GenBank® Entrez protein accession number P07686) was performed using free internet software. Prediction of the presence of transmembrane domains was performed using TMpred Prediction of Transmembrane Regions and Orientation software (http://www.ch.embnet.org/software/TMPRED_form.html). Prediction of GPI (glycosylphosphatidylinositol)-anchor motifs was performed using big-PI Predictor GPI Modification Site Prediction software (http://mendel.imp.ac.at/gpi/gpi_server.html). Prediction of fatty-acyl- or prenyl-anchor motifs was performed using the following software: Myristoylator (predicts N-terminal myristoylation of proteins by neural networks, http://www.expasy.ch/tools/myristoylator/) and PrePS Prenylation Prediction Suite (http://mendel.imp.ac.at/sat/PrePS/index.html).
NaCl and sodium carbonate extractions
Lysosomal membrane-enriched fraction P3 was resuspended in 0.1 M sodium carbonate (pH 11.5)  or in 10 mM Na/P buffer containing 1 M NaCl (pH 6.0). As a control, the P3 fraction was also resuspended in 10 mM Na/P buffer (pH 6.0). After 30 min incubation on ice, samples were ultracentrifuged at 40000 rev./min for 2 h at 4°C using a Beckman Optima Max MLS-50 rotor. Both the supernatant and pellet fractions were adjusted to the same final volume and analysed by immunoblotting.
Triton X-114 phase separation
Triton X-114 phase separation was performed as described previously . Briefly, the P3 fraction was resuspended in 0.2 ml of 10 mM Tris/HCl (pH 7.4) containing 150 mM NaCl and 0.6% Triton X-114 and incubated at 0°C for 1 h. The detergent-treated sample was then layered on to a 6% (w/v) sucrose cushion (0.3 ml), incubated at 30°C for 3 min and finally centrifuged at 300 g for 3 min. After centrifugation, the upper aqueous phase was recovered, re-extracted with 0.5% Triton X-114 and subjected to a second separation step through the same sucrose cushion. The detergent and aqueous phases were adjusted to the same final volume and analysed by immunoblotting.
Two-dimensional electrophoresis and Western blotting analysis of lysosomal membrane-associated Hex
The soluble lysosomal content fraction S1 and the lysosomal membrane-enriched fraction P3 were resuspended in 40 mM Trizma base containing 2% (w/v) SDS and 60 mM DTT (dithiothreitol). Samples were heated at 95°C for 5 min and any insoluble debris was removed by centrifugation at 16000 g for 1 min at 4°C. The 2D Clean-Up kit was used to prepare concentrated protein samples that were then resuspended in rehydration solution [7 M urea, 2 M thiourea, 2% (w/v) ASB-14, 0.5% IPG buffer (pH 3–10) and 40 mM DTT] with the addition of carbamylated protein standards for two-dimensional electrophoresis. Pre-cast linear IPG strips (11 cm, pH 3–10) were swollen at 30 V for 14 h at 20°C. Isoelectric focusing was conducted at 20°C, with a current limit of 50 μA/IPG strip in an IPGphor IEF System (Amersham Biosciences) using the following steps: 1 h at 200 V, 1 h at 500 V, 1 h at 1000 V, 0.5 h at 4000 V (gradient) and 6 h at 4000 V. After isoelectric focusing, the IPG strips were transferred into equilibration buffer [50 mM Tris/HCl (pH 8.8), 6 M urea, 30% (v/v) glycerol and 2% (w/v) SDS] in the presence of 1% DTT for 15 min, and then incubated in equilibration buffer with 4% (w/v) iodoacetamide for a further 15 min. The second dimension was performed on ExcelGel 2D Homogeneous 12.5 (Amersham Biosciences). The resolved proteins were transferred on to nitrocellulose using a Multiphor II NovaBlot Unit (Amersham Biosciences). After blotting, proteins were visualized by Ponceau Red staining. The blot was then incubated with an antiserum raised against Hex B  and developed by ECL®.
Hex activity and isoenzymes in lysosomal membrane-enriched fractions
Fractions enriched in lysosomal membranes were obtained from PL through three successive steps of sonication and centrifugation as described in the Experimental section. For each of the three steps, we recovered a pellet fraction further enriched in lysosomal membrane (P1, P2 and P3) and a supernatant fraction containing the soluble lysosomal content (S1, S2 and S3). All pellet and supernatant fractions were assayed for Hex activity using two substrates, MUG and MUGS, the former being hydrolysed by both α- and β-subunits, and the latter specifically by the α-subunit. We also assayed the activity of another soluble lysosomal glycohydrolase, α-Man , and the activity of the lysosomal membrane marker β-Gluc [28,30]. All fractions displayed Hex activity toward both MUG and MUGS substrates, with an acidic optimum pH of pH 4.5. Results reported in Figure 1 show that the increasing purification level of the lysosomal membrane achieved by three steps of sonication and ultracentrifugation gives rise to a gradual increase in the specific activity of β-Gluc. However, the specific activity of α-Man decreased progressively. In the P3 fraction, it was only barely detectable. The specific activity of Hex, assayed using the MUG substrate, displayed a peculiar pattern, as a decrease was observed from PL to the P2 fraction, which can be attributed, as for α-Man, to the removal of soluble Hex. Subsequently in the P3 fraction we detected a reproducible enrichment of total Hex-specific activity with respect to the P2 fraction. Further sonication and ultracentrifugation of the P3 fraction did not further increase the specific activity of Hex (results not shown). A similar trend was also detected when Hex activity was assayed with the MUGS substrate, but the increase in the specific activity present in the P3 fraction with respect to the P2 fraction was less pronounced than for MUG activity (Figure 1). The Hex isoenzyme pattern in the lysosomal membrane preparation P3 fraction and in the soluble lysosomal content fraction (S1) was investigated by DEAE-cellulose chromatography. The subunit composition in the peaks of activity observed that were eluted from the column was calculated from the relative rates of hydrolysis of the MUG and MUGS substrates. With both fractions, similar elution profiles were obtained. Hex B (ββ dimer, as demonstrated by its inability to hydrolyse MUGS) was not retained by the column, and Hex A (αβ dimer) was eluted with a maximum activity at approx. 0.12 mM NaCl. Nevertheless, the elution profiles disclosed different proportions of Hex B and Hex A between the membrane-associated and soluble fractions (Table 1). The same amount of Hex activity (detected using MUG) was loaded on to the columns, and the loss of activity was attributed to an inactivation of the α-subunit during the purification procedure. To better characterize the Hex isoenzymes, we determined the optimum pH and the thermal stability of Hex B and Hex A eluted by ion-exchange chromatography of the S1 and P3 fractions. Regarding the thermal stability and, thus, the relative isoenzyme content, there were no differences observed between the soluble and the membrane-associated Hex forms (Table 1).
Specific activity of lysosomal glycohydrolases during the enrichment procedure
|.||Percentage of Hex activity recovered from DEAE* .||NaCl† .||Optimum pH‡ .||Thermal stability§ .|
|.||S1 .||P3 .||S1 .||P3 .||S1 .||P3 .||S1 .||P3 .|
|.||Percentage of Hex activity recovered from DEAE* .||NaCl† .||Optimum pH‡ .||Thermal stability§ .|
|.||S1 .||P3 .||S1 .||P3 .||S1 .||P3 .||S1 .||P3 .|
Values represent the percentage of Hex activity towards MUG substrate recovered as Hex B and Hex A, with respect to the total MUG activity loaded. Equal amounts of MUG activity were loaded on the column for each experiment.
The salt concentration (mM) required for Hex A elution within the gradient.
The optimum pH for Hex isoenzyme activity, which was determined after dissolving MUG substrate in buffers at different pH values (pH 3.5–7.5).
The percentage of residual MUG activity of Hex B and Hex A after 1 h incubation at 52°C with respect to non-heated controls.
SDS/PAGE and Western blotting analysis of Hex in lysosomal membrane-enriched fractions
The pellet fractions containing the enriched lysosomal membranes (P1, P2 and P3) and the supernatant fractions containing the lysosomal matrix proteins (S1, S2 and S3) were analysed by Western blotting and immunodetection of Hex using a specific antiserum raised against placental Hex B, which is able to specifically detect the β-subunit (Figure 2A). A constant amount of protein was loaded in each well. As shown in the left-hand panel, in the PL, P1, P2, P3 and S1 fractions, different amounts of mature Hex β-subunit were detected as indicated by a single band of approx. 28 kDa [21,25]. Densitometric analysis (shown in the right-hand panel) indicated a trend in the Hex β-subunit polypeptide abundance in the above fractions that was observed for the specific activity towards the MUG substrate. A strong enrichment was observed in the P3 fraction. Similar results were obtained when the Hex α-subunit was immunodetected (Figure 2B): mature α-subunit (54 kDa) was found in all samples except for the S3 fraction. The clear enrichment of this subunit in the P3 fraction compared with the P2 fraction highlighted the presence of both mature Hex α- and β-subunits in the membrane-associated enzyme and suggested that the low enrichment of MUGS activity in the P3 fraction compared with the P2 fraction may have been caused by an inactivation of Hex A during the fractionation. In order to verify the effectiveness of our lysosomal membrane-enrichment procedure, we also monitored LAMP-2, a typical lysosomal integral-membrane protein , using a LAMP-2-specific antibody (Figure 2C). The results highlighted a gradual increase in the intensity of the LAMP-2 band up to 24-fold in the P3 fraction compared with PL. In a separate experiment, we repeated the sonication and ultracentrifugation steps for a fourth time, but this did not produce any further enrichment of LAMP-2 in the pellet fraction (results not shown).
Western blots and densitometric analysis of the fractions obtained from the lysosomal membrane-enriched procedure
Peripheral membrane protein extraction and Triton X-114 phase separation
In order to understand the nature of the anchoring of the predominantly soluble matrix enzyme to the lysosomal membrane, we first analysed the amino-acid sequence of the Hex subunits. Bioinformatic analysis (see the Experimental section) of the Hex α- and β-subunit amino-acid sequences did not reveal the presence of any motifs compatible with the addition of GPI, fatty-acyl or prenyl-lipid-bilayer anchors. In fact, no consensus sequences for these post-translational modifications were found. Instead TMpred Prediction of Transmembrane Regions and Orientation software highlighted the presence of two putative membrane-spanning domains in the Hex β-subunit (Met13–Thr31 and Val436–Gln458) and one hydrophobic amino-acid stretch in the Hex α-subunit (Leu6–Trp24), but it has been demonstrated that the amino-acid sequences between Met13–Thr31 and Leu6–Trp24 in the α- and β-subunits respectively are removed in the endoplasmic reticulum by signal peptidase during N-terminal signal-peptide cleavage . Crystallographic analysis of the Hex B structure showed that the Val436–Gln458 stretch comprises amino-acid residues involved in subunit dimerization, a crucial event in acquiring the enzymatic activity .
On the basis of these considerations, we examined the possibility that Hex may be associated with the lipid bilayer as a peripheral membrane protein. Using a procedure as described by Fujiki et al. , aliquots of the lysosomal membrane-enriched P3 fraction were incubated for 30 min on ice with 0.1 M sodium carbonate (pH 11.5), with 1 M NaCl in 10 mM Na/P buffer (pH 6.0), and with 10 mM Na/P buffer (pH 6.0). After the incubation, the mixtures were subjected to ultracentrifugation and the supernatant and pellet fractions obtained were analysed by immunoblotting for Hex α- and β-subunits and LAMP-2 polypeptides as before. The results (Figure 3) demonstrated an almost complete solubilization of both Hex α- and β-subunits in the presence of sodium carbonate. Both the α- and β-subunit bands were present in the supernatant fraction, whereas LAMP-2 was completely recovered in the insoluble pellet fraction. In contrast, the treatment of the P3 fraction with 1 M NaCl in 10 mM Na/P buffer (pH 6.0) and with 10 mM Na/P buffer (pH 6.0) alone caused no appreciable solubilization of the Hex α- and β-subunits or of LAMP-2.
Extraction of Hex polypeptides from purified lysosomal membranes in the presence of NaCl and sodium carbonate solutions
To obtain more information about the nature of the interaction between Hex and the lipid bilayer, we also performed Triton X-114 extraction of the P3 fraction, followed by aqueous/detergent phase separation. The detergent and aqueous phases obtained after centrifugation were analysed by immunoblotting. As shown in Figure 4, Hex α- and β-subunits partitioned in the aqueous phase, whereas LAMP-2 was present exclusively in the detergent phase, as expected.
Partitioning of Hex isoenzymes and LAMP-2 using Triton X-114 phase separation
Comparative analysis of soluble and membrane-associated Hex by two-dimensional electrophoresis
To investigate the possible differences between soluble and membrane-associated Hex, we performed two-dimensional electrophoresis and Western blotting analysis of both the S1 and P3 fractions. CPK (creatine phosphokinase) and GAPDH (glyceraldehyde-3-phosphate dehydrogenase) carbamylated standards provided a series of spots across a pI range of 4.9–7.1 and 4.7–8.3 respectively, which allowed the identification of the pI of the protein by ImageMaster2D analysis. Spots detected on the blot by the specific antibody raised against Hex B are indicated in Figure 5. The results obtained after Western blotting analysis of the lysosomal matrix protein (Figure 5A) demonstrated the presence of two spots of the same apparent molecular mass with pIs of 6.00 and 6.66, corresponding to the mature Hex βa- and βb-subunits respectively .
Characterization of Hex polypeptides in the soluble lysosomal content fraction S1 and the purified lysosomal membrane fraction P3 using two-dimensional electrophoresis and Western blotting
These polypeptides were generated by the final proteolytic processing of the Hex B precursor which takes place in lysosomes . The two-dimensional pattern of the P3 fraction (Figure 5B) showed two spots of slightly different molecular mass with pIs of 5.80 and 6.63. Analogous experiments performed using a specific antibody raised against Hex A did not reveal any differences between the soluble and membrane-associated Hex α-subunit (results not shown).
Glycosphingolipids are ubiquitous membrane components of eukaryotic cells that are abundant in the PM, where they function as modulators of signal transduction regulating the proliferation, survival and differentiation of cells . These lipids play a key role as signalling molecules in both physiological and pathological processes regulating cell-to-cell and/or cell–environment interactions [3,38]. Several studies have previously revealed the presence of glycohydrolases not only in lysosomes, but also at the cell surface [5,18,39,40]. They stress the potential role of the remodelling of glycosphingolipids during cellular processes [6,7]. In a previous study, we demonstrated that a mature form of the Hex A isoenzyme is associated to the human fibroblast PM and is catalytically active toward its natural substrate, GM2 .
Here we demonstrate that lysosomal membranes that were purified from human placenta lysosomes  by repeated cycles of sonication and ultracentrifugation sedimentation also contain mature Hex isoenzymes in non-covalently-attached forms. It is conceivable that the procedure used here for the purification of the lysosomal membrane may result in contamination of the final preparation with lipofuscin-like lysosomal inclusions, which tend to sediment with the membranes. However, Hex enzymes are not present in the lipofuscin-like lysosomal inclusion body , and therefore this kind of contamination is rather unlikely.
We assayed Hex activity towards MUG and MUGS substrates and examined the presence of the two polypeptides that constitute the Hex isoenzymes, α and β, using specific antibodies in all fractions obtained during the purification. The results revealed that, in the first two steps, a major portion of Hex behaved as a soluble lysosomal enzyme similar to α-Man, which was detected as a control. In the third step, Hex, unlike α-Man, behaved as a lysosomal membrane-associated protein and displayed an enrichment comparable with that of β-Gluc. The latter is known to bind to the lysosomal integral membrane protein LIMP-2 (lysosomal integral membrane protein-2) . The conclusion is that a significant amount of Hex is associated with the lysosomal membrane. A further step of sonication and centrifugation was not useful to detach this fraction of the enzyme from the membrane, since it resulted in a selective loss of the enzymatic activity of the α-subunit. This was demonstrated by a rapid decrease in MUGS activity, but not in the amount of α-polypeptide that was detected by Western blotting.
Lysosomal membrane-associated Hex activity displayed an acidic optimum pH and thermal stability, as did the soluble counterparts. Western blotting analysis clearly demonstrated that both α- and β-subunits associated with the lysosomal membrane fraction in their fully processed forms. In an attempt to elucidate the isoenzyme composition of lysosomal membrane-associated Hex, we performed DEAE-cellulose chromatography of the P3 fraction. The subunit composition of the peaks eluted from the column was determined by assaying the eluates with MUG and MUGS. This analysis clearly revealed the presence of both Hex A and Hex B isoenzymes, which eluted at the same position as the soluble lysosomal counterpart. The relative amount of Hex A and B is different between the soluble fraction and the lysosomal membrane-enriched P3 fraction, with Hex B being the predominant form, but this can be attributed to the instability of the α-subunit activity.
Bioinformatic analysis of the Hex α- and β-subunit amino-acid sequences excluded the presence of sequences compatible with a GPI-anchored, fatty acylated or prenylated protein because no post-translational modification consensus sequences were found. In order to verify whether Hex behaves as a peripheral membrane protein associated with the lipid bilayer by ionic and/or hydrogen interactions involving transmembrane proteins or phospholipids resident in lysosomal membrane, the lysosomal membrane-enriched P3 fraction was treated with 1 M NaCl or with 0.1 M sodium carbonate (pH 11.5) . Under the former condition, Hex remained associated with the lysosomal membrane, as did the lysosomal transmembrane protein LAMP-2, whereas the extreme pH condition produced the complete solubilization of both Hex subunits, but not of the integral lysosomal membrane LAMP-2 (Figure 3). Triton X-114 phase separation of the lysosomal membrane-enriched P3 fraction allowed the complete segregation of Hex α- and β-subunits in the aqueous phase (Figure 4), thus demonstrating the hydrophilic nature of Hex subunits . In contrast, LAMP-2 partitioned completely into the detergent phase as expected. In summary, these findings support the hypothesis that a portion of the lysosomal Hex behaves as a peripheral membrane-associated protein.
Comparative analysis by two-dimensional electrophoresis and Western blotting revealed that the mature βa and βb polypeptides that were recovered in the lysosomal membrane-associated form displayed a slightly different apparent molecular mass and pI compared with their soluble lysosomal counterparts (Figure 5B). In contrast, no differences were observed between the soluble and the membrane-associated forms of the α-subunit (results not shown). We speculate that the observed differences between the apparent molecular masses and isoelectric points of the matrix and the membrane-associated forms of Hex β polypeptides could be due to an as yet unidentified post-translational modification. This may result in binding to a component of the membrane.
The present study demonstrated the existence of membrane-associated forms of both Hex A and Hex B isoenzymes in lysosomal membranes isolated from human placenta. The lysosomal membrane-associated forms displayed the same catalytic properties as the soluble forms and the enzyme that has been found at the PM in a previous study . Therefore we consider the existence of two pools of mature Hex: a membrane-associated pool and a soluble pool. The membrane-associated Hex is present in both lysosomal membranes and the PM. It should be of interest to examine the pathway delivering the membrane-associated lysosomal Hex to the cell surface and the possible functions of the cell surface-associated enzyme in modulating the glycolipid pattern at the cell boundary.
immobilized pH gradient
lysosome-associated membrane protein-2
- Na/P buffer
sodium phosphate buffer
purified human placenta lysosome
This work was supported by COFIN (Cofinanziamento Ministero dell'Istruzione, dell'Università e della Ricerca Scientifica)-PRIN (Progetti Diricerca di Interesse Nazionale) 2004050497_002 and FIRB (Fondo Investimenti Ricerca di Base) RBAUO1MSFR grants to C.E.
These authors contributed equally to this work.