Arylsulfatase G, a Novel Lysosomal Sulfatase*

The sulfatases constitute a conserved family of enzymes that specifically hydrolyze sulfate esters in a wide variety of substrates such as glycosaminoglycans, steroid sulfates, or sulfolipids. By modifying the sulfation state of their substrates, sulfatases play a key role in the control of physiological processes, including cellular degradation, cell signaling, and hormone regulation. The loss of sulfatase activity has been linked with various severe pathophysiological conditions such as lysosomal storage disorders, developmental abnormalities, or cancer. A novel member of this family, arylsulfatase G (ASG), was initially described as an enzyme lacking in vitro arylsulfatase activity and localizing to the endoplasmic reticulum. Contrary to these results, we demonstrate here that ASG does indeed have arylsulfatase activity toward different pseudosubstrates like p-nitrocatechol sulfate and 4-methylumbelliferyl sulfate. The activity of ASG depends on the Cys-84 residue that is predicted to be post-translationally converted to the critical active site Cα-formylglycine. Phosphate acts as a strong, competitive ASG inhibitor. ASG is active as an unprocessed 63-kDa monomer and shows an acidic pH optimum as typically seen for lysosomal sulfatases. In transfected cells, ASG accumulates within lysosomes as indicated by indirect immunofluorescence microscopy. Furthermore, ASG is a glycoprotein that binds specifically to mannose 6-phosphate receptors, corroborating its lysosomal localization. ARSG mRNA expression was found to be tissue-specific with highest expression in liver, kidney, and pancreas, suggesting a metabolic role of ASG that might be associated with a so far non-classified lysosomal storage disorder.

Twelve of the 17 sulfatases encoded in the human genome have been characterized biochemically (16). Based on their subcellular localization they can be divided into lysosomal and non-lysosomal enzymes. The latter are found either at the cell surface (Sulf1, Sulf2), in the endoplasmic reticulum (arylsulfatases C, D, and F), or in the Golgi apparatus (arylsulfatase E) and act at neutral pH. In contrast, all lysosomal sulfatases (arylsulfatases A and B, iduronate-2-sulfatase, heparan-N-sulfatase, glucosamine-6-sulfatase, and galactosamine-6-sulfatase) share an acidic pH optimum (17). The genetic deficiency of each of these six lysosomal sulfatases causes specific and severe lysosomal storage disorders, namely metachromatic leukodystrophy and mucopolysaccharidoses type VI, II, IIIA, IIID, and IVA, respectively, which highlights the essential and non-redundant function of these enzymes (2). In affected patients, the degradation of a specific sulfated compound is blocked, leading to its accumulation in the lysosomes and in the extracellular fluids. Lysosomal storage impairs autophagic delivery of bulk cytosolic contents to lysosomes, finally resulting in accumulation of toxic proteins, cellular damage, and apoptosis (18). The total number of lysosomal hydrolases, according to proteomic analyses, is estimated to be in the range of 50 -60 (19). In principle, a genetic mutation of any of these proteins can cause a lysosomal storage disorder. For many sulfated substrates like sulfo-proteins (tyrosine-, threonine-or serine-O-sulfate (20), the selectin ligand 6-sulfo sialyl Lewis X , as well as the heparan sulfate constituents N-acetylglucosamine 3-sulfate and glucuronate 2-sulfate), the corresponding sulfatases and possible associated storage disorders have not been identified yet. Thus, the discovery and characterization of novel lysosomal enzymes will likely be important in the identification and molecular understanding of so far non-classified inherited genetic diseases.
The mammalian arylsulfatase G (ASG) was identified in 2002 through bioinformatic searches of expressed sequence tag databases (21). The ARSG gene is located on chromosome 17q24.2, * This work was supported by the Deutsche Forschungsgemeinschaft and the Fonds der Chemischen Industrie. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental Figs. S1 and S2 and a supplemental reference. 1  consists of 11 exons, and encodes a 525-amino acid protein that shares a high degree of similarity with all sulfatases, in particular with arylsulfatase A (50% sequence similarity and 37% identity). The enzyme was tentatively classified as an arylsulfatase, although no activity toward the commonly used arylsulfate pseudosubstrates p-nitrocatechol sulfate (pNCS) and 4-methylumbelliferyl sulfate (4-MUS) could be detected. In overexpressing COS-7 cells, ASG was found to be localized in the endoplasmic reticulum. In this study, however, we demonstrate for the first time that ASG is an active arylsulfatase enzyme of the lysosome.

EXPERIMENTAL PROCEDURES
Construction of Expression Plasmids-The full-length cDNA sequence encoding human ASG, designated KIAA1001 (accession number NM_014960. 1,22), was obtained from the Kazusa Institute (Kisarazu, Chiba, Japan). Two sequence isoforms of ASG are known. Whereas genomic and expressed sequence tag sequences, both for human and murine ASG, contain a GCA (Ala) codon in position 501, this codon is replaced by CCA (Pro) in the KIAA1001 clone (21). The ASG A501P cDNA was amplified by PCR, thereby adding a 3Ј-RGS-His 6 -encoding sequence followed by a stop codon and a HindIII site (reverse primer 5Ј-CCCAAGCTTAGTGATGGTGATGGTGATGCG-ATCCTCTTGCGGCTTGACAGCGGC-3Ј). The PCR product was digested with HindIII and NcoI, which cuts in the endogenous Kozak sequence of ASG, and cloned into pMPSV-EH-ASA (23), thus exchanging the entire coding region of arylsulfatase A in-frame to that of ASG. The insert was then subcloned as a 5Ј-EcoRI-3Ј-HindIII/blunt fragment into the EcoRI/SmaI-opened multiple cloning site of pCI-neo (Promega). The QuikChange mutagenesis protocol (Stratagene) was used to generate the wild-type ASG-encoding sequence with a GCA codon in position 501 (accession number NM_014960.2). For this purpose, complementary primers were used (forward 5Ј-CGACAACATCTCCAGCGCAGA-TTACACTCAGG-3Ј, reverse 5Ј-CCTGAGTGTAATCTG-CGCTGGAGATGTTGTCG-3Ј). The ASG C84A mutant was prepared accordingly (forward 5Ј-GCTGCCTCCACC-GCCTCACCCTCCCG-3Ј, reverse 5Ј-CGGGAGGGTGAG-GCGGTGGAGGCAGC-3Ј). All constructs were full-length sequenced in the coding region to preclude any PCR-derived errors. Unless otherwise stated, all experiments were performed with wild-type ASG.
Cell Culture and Transfections-HT1080 human fibrosarcoma cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum (PAN Biotech GmbH, Aidenbach, Germany) and 1% penicillin/streptomycin (Invitrogen) under 5% CO 2 atmosphere at 37°C. Transfections with BamHI-linearized pCI-neo-ASG plasmids were performed using MATra-A transfection reagent (IBA, Göttingen, Germany) following the protocol recommended by the manufacturer. After growth for 12 days in medium containing 800 g/ml G-418 sulfate (PAA Laboratories, Pasching, Austria), drug-resistant cells were cloned and expanded. Clones were screened by Western blot analysis for ASG expression.
Purification from Cell Culture Supernatants-HT1080 cells stably overexpressing ASG-His were grown to near confluence in growth medium containing 10% fetal calf serum on 15-cm cell culture dishes (Nunc) or Cellmaster PS roller bottles (Greiner Bio-One). During ASG production, the amount of serum was reduced to 1%. The conditioned medium was collected every 48 -72 h, cleared by spinning, and subjected to ammonium sulfate precipitation (50% w/v). The precipitate was dialyzed overnight at 4°C against binding buffer (20 mM Tris-HCl, 500 mM NaCl, 40 mM imidazole, pH 7.4). The dialyzed material was cleared by centrifugation (18,000 ϫ g, 3 ϫ 30 min, 4°C), passed through a 0.22-m nitrocellulose filter, and bound onto a HisTrap HP 1-ml column connected to an Ä KTA Explorer 10 system (GE Healthcare). The column was washed with binding buffer, and proteins were gradually eluted with elution buffer (20 mM Tris-HCl, 500 mM NaCl, 500 mM imidazole, pH 7.4). Peak fractions, as analyzed by Western blotting and pNCS assays, were pooled, dialyzed against buffer A (20 mM MES, pH 6.0), and applied onto a RESOURCE S 1-ml cation exchange column (GE Healthcare). Elution was carried out using a gradient with buffer B (20 mM MES, 1 M NaCl, pH 6.0). The ASG-His protein eluted at 140 mM NaCl. Peak fractions were pooled and concentrated by speed vac, if necessary. Proteins were analyzed by SDS-PAGE on 15% polyacrylamide gels and stained with Roti-Blue colloidal Coomassie (Carl Roth, Karlsruhe, Germany). Protein concentrations were determined using Coomassie Plus Bradford reagent (Pierce). Average yields were ϳ40 g of purified ASG/liter of medium.
Peptide N-Glycosidase F and Endoglucosaminidase H Treatment-HT1080 cells stably overexpressing ASG-His were grown to confluency. The cells were harvested by rubber policeman and extracted by sonication in lysis buffer (10 mM HEPES, 0.5 M NaCl, pH 7.4). The lysate was cleared by centrifugation (72,000 ϫ g, 20 min, 4°C). Cell lysates and purified ASG samples were denatured and subjected to treatment with peptide N-glycosidase F and endoglucosaminidase H (both from Roche Applied Science) as described (24). In case of peptide N-glycosidase F deglycosylation, samples were denatured with 0.5% SDS and 0.2% ␤-mercaptoethanol at pH 7.2 for 5 min at 95°C. Samples were mixed with peptide N-glycosidase F in incubation buffer containing 1.2% Triton X-100 at pH 7.2. For endoglucosaminidase H treatment, samples were denatured at pH 5.0 with 0.01% SDS and 0.7% ␤-mercaptoethanol for 5 min at 95°C and subjected to deglycosylation. In both cases, samples were incubated at 37°C for 2 or 24 h and analyzed by Western blotting.
Western Blot-For Western blot analysis, a mouse monoclonal antibody directed against the RGS-His 6 epitope was used as a primary antibody (Qiagen). Signals were detected using a horseradish peroxidase-conjugated goat anti-mouse IgG secondary antibody (Molecular Probes) and ECL detection reagent (Pierce). Signals were quantified using the AIDA 4.06 software package (Raytest, Straubenhardt, Germany).
Enzymatic Assays-Activities of ASG toward pNCS or p-nitrophenyl sulfate (pNPS) were assayed using 10 mM pNCS or pNPS in either 0.5 M sodium acetate, (pH 4.5-6.0), MES (pH 6.5), or Tris-HCl (pH 7.0 -8.0). Samples containing ASG as indicated were incubated at 37°C for 1 h in 150 l of reaction volume. Reactions were terminated by addition of 150 l of 1 M NaOH. Michaelis-Menten kinetic analysis was performed at pH 5.6 using 0.5-30 mM pNCS in 0.5 M sodium acetate. Absorbances were measured at 515 nm (e 515 ϭ 12400 M Ϫ1 cm Ϫ1 ) in the case of pNCS or at 405 nm (e 405 ϭ 18000 M Ϫ1 cm Ϫ1 ) for pNPS. Inhibition kinetics included Na 2 HPO 4 , NaHSO 4 , or warfarin (Sigma-Aldrich) at the indicated concentrations. Activity measurements with 4-MUS were performed accordingly, using 10 mM 4-MUS in the respective buffers. Reactions were stopped by addition of 150 l of 1 M Na 2 CO 3 /NaHCO 3 , pH 10.7. The fluorescence of 4-methylumbelliferone, compared with a calibration curve, was measured with an excitation wavelength of 360 nm and an emission wavelength of 465 nm. All absorbance and fluorescence measurements were performed using an infinite M200 microplate reader (TECAN, Crailsheim, Germany). The temperature optimum was determined using a gradient thermocycler.
Immunofluorescence-Stably transfected HT1080 cells expressing ASG-His were grown on poly-L-lysine-coated coverslips for 24 h and labeled with 50 nM LysoTracker Red DND-99 (Molecular Probes) for 2 h in serum-free Dulbecco's modified Eagle's medium. The cells were briefly washed with phosphatebuffered saline, fixed with 4% paraformaldehyde for 20 min at room temperature, and permeabilized with 0.1% Triton X-100 for 10 min. After blocking with 2% fetal calf serum, the cells were incubated with a mouse monoclonal anti-RGS-His 6 antibody (Qiagen) for 1 h. The primary antibody was detected with an Alexa-488-conjugated goat anti-mouse IgG secondary antibody (Molecular Probes). Immunofluorescence images were obtained on a Leica DM5000 B microscope equipped with an HCX PL APO ϫ100 oil immersion objective.
Gel Filtration Analysis-Purified ASG was subjected to gel filtration on a Superdex 75 3.2/30 PC column (GE Healthcare), equilibrated with either 50 mM NaAc, 150 mM NaCl, pH 5.6, or 50 mM Tris-HCl, 150 mM NaCl, pH 7.4. Proteins were eluted at a flow rate of 40 l/min using an Ä KTA Ettan LC system (GE Healthcare). Fractions were examined for ASG activity by pNCS assays.
Reverse Transcription PCR Expression Analysis-Reverse transcription PCR experiments were performed using a panel of normalized cDNAs prepared from eight normal human tissues (MTC panel human I; Clontech). An internal 880-bp fragment from human ARSG was amplified by PCR (forward primer 5Ј-TTCATCCAGCGTGCAAGCACCAGC-3Ј and reverse primer 5Ј-CCTACCAAATTGCCTGCCGCTGTC-3Ј). PCR was carried out for 35 cycles with 62°C annealing temper-ature. Normalization was confirmed by primers specific for glycerol aldehyde-3-phosphate dehydrogenase.

RESULTS
Purification and Arylsulfatase Activity of ASG-To investigate the biochemistry and cell biology of ASG, we generated HT1080 cell lines stably expressing human ASG with a C-terminal RGS-His 6 tag. The secreted protein was purified from the conditioned medium using a combination of Ni 2ϩ affinity and cation exchange chromatography (Fig. 1, A and B). The purified protein had a size of 63 kDa and was identified as ASG by matrix-assisted laser desorption ionization time-of-flight mass spectrometry of tryptic peptides.
Purified ASG was obtained in catalytically active form. ASG cleaves the pseudosubstrate pNCS and also, but with lower activity, pNPS and 4-MUS. To find optimal conditions, we analyzed the pH dependence of the reaction with pNCS and 4-MUS and found pH optima at 5.4 and 5.6, respectively (Fig.  1C), thus giving a first indication for a possible lysosomal function of ASG. Furthermore, we analyzed the temperature optimum of ASG (Fig. 1D). During a 60-min incubation, ASG cleaved pNCS most efficiently at 45-50°C. The corresponding Arrhenius plot was linear over the temperature range from 0 to 30°C (see supplemental Fig. S1). From this plot, an activation energy of E a ϭ 43.1 kJ/mol was calculated. ASG activity was retained even at relatively high temperatures above 55°C, thus characterizing ASG as a thermostable enzyme. The extraordinary stability of ASG is underlined by a half-life of more than 100 days during storage at 4°C and pH 6.0 (data not shown). Characterization of ASG Mutants/Isoforms-The active site C ␣ -formylglycine residue is essential for catalysis in all human sulfatases that have been characterized so far (2,13). The socalled sulfatase signature 84-CSPSR-88 of ASG predicts Cys-84 to be post-translationally modified to FGly. To confirm the importance of FGly in ASG, we mutagenized Cys-84 to Ala and compared the activity of the mutant to the wild-type enzyme by performing pNCS activity assays with cell lysates from stably transfected cells. We also analyzed the activity of the ASG A501P isoform encoded by the sequence of the human expressed sequence tag clone KIAA1001 (see "Experimental Procedures" and Refs. 21,22). The ASG C84A mutant showed no activity above background levels resulting from endogenous sulfatases (Fig. 2). Compared with wild-type ASG, the A501P isoform displayed only ϳ35% activity. Wild-type ASG was used for all further studies.
Arylsulfatase Substrates, Kinetics, and Inhibitors of ASG-To further investigate the enzymology of ASG, we analyzed the influence of substrate concentration on the activity toward pNCS (Fig. 3A). ASG follows Michaelis-Menten kinetics and shows hyperbolic substrate saturation. At substrate concentrations above 15 mM substrate inhibition is observed. The doublereciprocal Lineweaver-Burk transformation (Fig. 3B) yields the following kinetic constants: K m ϭ 4.2 mM, V max ϭ 63.5 mol/ (min⅐mg) ϭ 63.5 units/mg. Compared with other arylsulfatases, these values fall within the normal range. Typical activities toward pNCS are 40 -100 units/mg (27). Because of its limited solubility at acidic pH, no substrate saturation could be obtained with 4-MUS as a substrate (data not shown). With an estimated V max of only 0.2-0.4 units/mg and a K m of 5-15 mM, 4-MUS represents a rather poor substrate for ASG. Compared with pNCS, the cleavage of the closely related pNPS, merely lacking one hydroxyl group, is even slower (ϳ0.05 units/mg at pH 5.6), indicating a stringent substrate specificity that might also apply for physiological substrates.
Many sulfatases are inhibited by their product, sulfate, or by its analog, phosphate (28). Interestingly, inhibition of ASG by phosphate is much stronger than by sulfate (Fig. 3C). Whereas sulfate has an IC 50 value of ϳ1 mM, inhibition by phosphate shows an IC 50 of ϳ50 M at 10 mM pNCS (Fig. 3C, inset). To determine the type of inhibition, substrate saturation curves were recorded at different phosphate concentrations, which are presented in Fig. 3D as Lineweaver-Burk plots. The regression lines show a common intercept on the ordinate axis as expected for a competitive inhibitor. From the slopes of the regression lines plotted against inhibitor concentrations, a K i value of 17 M was extrapolated. In contrast to arylsulfatase E (29), ASG is not inhibited by warfarin even at concentrations close to the solubility limit (data not shown).
ASG Is Active in the Monomeric State-To analyze whether or not ASG forms dimeric or oligomeric complexes, purified ASG was subjected to gel filtration on a Superdex 75 column at either pH 5.6 or 7.4. Fractions were tested for sulfatase activity (Fig. 4). At both pH values, neither protein UV peaks nor sulfatase activities provided any indication for dimerization or oligomerization of ASG as described for its closest relative, arylsulfatase A (30). ASG eluted as a 63-kDa protein, which corresponds to the monomeric molecular mass also observed in SDS-PAGE. Thus, unlike, for example, arylsulfatase B, ASG does not undergo lysosomal processing (31).
Characterization of the ASG Glycosylation State-The ASG sequence contains four potential N-glycosylation sites (asparagine residues 117, 215, 356, and 497). Treatment with peptide N-glycosidase F reduces the apparent sizes of both the intracellular and the secreted protein from a broad, glycosylated band at 63 kDa (with micro-heterogeneity in its N-glycans) to a sharp band of ϳ53 kDa (Fig. 5A). This difference in size suggests that all four N-glycosylation sites are utilized, assuming that the average mass is Յ1.9 kDa for a high mannose and Յ2.9 kDa for a complex type oligosaccharide (24). Furthermore, the intracellular ASG is largely sensitive toward treatment with endoglucosaminidase H, whereas the secreted ASG is resistant (Fig. 5A,  lane 8). Thus, high mannose type oligosaccharides added to ASG in the ER are processed to complex type N-glycan structures during maturation in the Golgi. The intracellular ASG represents a mixture containing mainly high mannose N-glycans, one of which, however, was endoglucosaminidase H-resistant in the majority of ASG molecules (Fig. 5A, lane 4). Endoglucosaminidase H treatment did not result in a complete deglycosylation even at extended incubation times and increased endoglucosaminidase H concentrations.
ASG Binds to Mannose 6-Phosphate Receptors-Most lysosomal proteins are transported to the lysosomes via the mannose 6-phosphate receptor (M6PR) pathway (32,33). In a comprehensive proteome analysis of lysosomal proteins, an affinity column with immobilized M6PRs has successfully been used to identify novel lysosomal candidate proteins (26). Here, we used this affinity column to assay whether or not ASG binds to M6PRs in vitro (Fig. 5B). The M6PR matrix was incubated with purified ASG and washed with glucose 6-phosphate to remove unspecifically bound protein. As expected for a lysosomal protein, ASG eluted specifically with mannose 6-phosphate (Fig. 5B, lanes E1-4) but not with glucose 6-phosphate ( lanes  W5-7), thus providing further evidence for a lysosomal localization. About 28% of ASG loaded onto the affinity matrix was recovered in the elution fractions. The observed binding efficiency corresponds to those of reference lysosomal proteins and is probably due to partial modification with mannose 6-phopshate residues in the overexpressing cells (26). Intracellular Localization of ASG-To corroborate the lysosomal localization of ASG as suggested by its acidic pH optimum and binding to M6PRs, we analyzed the localization of overexpressed ASG in HT1080 cells by immunofluorescence. For lysosomal staining, Lyso-Tracker Red was used, an aldehydefixable, fluorescent amine that accumulates within acidic cell compartments, predominantly lysosomes, trans-Golgi, and perinuclear late endosomes (34,35). Cells were incubated with LysoTracker Red for 2 h prior to fixation, permeabilization, and incubation with anti-RGS-His 6 antibodies to detect ASG. The ASG staining (Fig. 6) revealed vesicular spots throughout the cytoplasm, similar to the distribution observed for the lysosomal arylsulfatase A (36). Although some of the LysoTracker signal was lost during the fixation and permeabilization steps, most of the ASG-positive vesicular structures co-localized with the Lyso-Tracker staining, thus substantiating ASG as a lysosomal sulfatase. The detection of lysosomal structures was obscured in many smaller cells due to the intense perinuclear ER/Golgi staining. ER localization of ASG is most likely due to folding/ maturation problems as commonly observed for many overexpressed proteins entering the secretory route. This might explain why ASG has been characterized previously as an ERlocalized protein in transiently transfected COS-7 cells using Myc-tagged ASG (21). Co-immunofluorescence analysis using lysosomal membrane glycoprotein 1 (LAMP1) as an alternative marker of lysosomes/late endosomes confirmed the predominantly lysosomal localization of ASG also in low expressing HT1080 cells (see supplemental Fig. S2).
Tissue-specific Expression of ASG-Tissue expression analysis can add important information about the biological role of sulfatases in vivo. Previous reverse transcription PCR experiments with five different mouse tissues revealed a rather ubiq-   FIGURE 4. Gel filtration analysis of ASG. ASG was subjected to gel filtration at pH 5.6 and 7.4 on a Superdex 75 column. Fractions were tested for arylsulfatase activity using pNCS. The molecular mass of ASG was determined from the UV peaks compared with a standard curve obtained from separating bovine serum albumin (66 kDa), ovalbumin (45 kDa), cytochrome c (12.5 kDa), and aprotinin (6.5 kDa) (not shown). The exclusion limit of the column is 100 kDa, as indicated. Activity and UV peaks (not shown) were found to coincide at both pH values tested.
uitous expression of murine ASG (21). To analyze the expression of human ASG, we used a normalized cDNA panel from eight different human tissues. Interestingly, tissue-specific differences were observed with highest ASG expression levels in liver, kidney, and pancreas (Fig. 7). No or only very low expression was detected in brain, lung, heart, and skeletal muscle. Differential expression could explain why ASG has not earlier been identified as a lysosomal protein, e.g. in mannose 6-phosphate glycoproteome analysis of human brain extracts (37).

DISCUSSION
Of the 17 known human sulfatases (16), six have been reported to be localized in lysosomes. As typically seen for lysosomal enzymes, all of these sulfatases have acidic pH optima ranging from 3.8 to 5.7 (17). Moreover, all six sulfatases have been linked with six individual lysosomal storage disorders (2). Here we show that the human arylsulfatase G also belongs to this important group of lysosomal enzymes.
ASG is a stable 63-kDa glycoprotein that hydrolyzes different pseudosubstrates at acidic pH but has only ϳ10% activity at pH 7.5. Mutation of the active site Cys residue that is predicted to be post-translationally converted to FGly leads to a complete loss of enzymatic activity, emphasizing the importance of FGly for catalysis. The ASG A501P isoform, corresponding to the KIAA1001 expressed sequence tag clone (22), displays significantly reduced activity. The observed differences in comparison with wild-type ASG could be caused by compromised protein folding due to the presence of a proline residue. It should be noted that the Ala codon is conserved in the genomes of mouse, rat, cow, and even chicken and zebrafish.
In transfected HT1080 cells, ASG co-localizes with the lysosomal markers LysoTracker Red and LAMP1. Furthermore, ASG binds to immobilized mannose 6-phosphate receptors, which are required for transport of most lysosomal proteins into the lysosomes. Combined with the acidic pH optimum of enzymatic activity, these findings demonstrate that ASG is an active, lysosomal arylsulfatase. Although the immunofluorescence data do not fully exclude the possibility that ASG has a dual lysosomal/ER localization, this seems unlikely with respect to enzyme function.
Arylsulfatase A, the sulfatase with highest homology to ASG, is a lysosomal sulfatase that acts as an octamer at acidic pH (30). Deficiency of arylsulfatase A causes metachromatic leukodystrophy, a lysosomal storage disorder associated with progressive demyelination of the central and peripheral nervous system. During biogenesis, arylsulfatase A forms dimeric complexes in the ER that are transported via the mannose 6-phosphate receptor pathway to the lysosomes. Because of the pH shift, the arylsulfatase A dimers oligomerize to octamers, thereby stabilizing the enzyme against lysosomal proteases.

RGS-His 6
LysoTracker overlay FIGURE 6. Subcellular localization of ASG in HT1080 cells. HT1080 cells stably expressing RGS-His 6 -tagged ASG were incubated for 2 h with Lyso-Tracker Red, fixed with 4% paraformaldehyde, and permeabilized with 0.1% Triton X-100. ASG was detected by indirect immunofluorescence using a mouse anti-RGS-His 6 antibody. The overlay reveals a colocalization of ASG and LysoTracker Red. Boxed areas are shown below at higher magnification. Bars represent 10 m. It should be noted that the intense perinuclear Lyso-Tracker staining is due to the non-confocal image acquisition. . Reverse transcription PCR analysis of ARSG mRNAs in human tissues. Normalized cDNAs derived from different human tissues were subjected to PCR to amplify a fragment specific for ASG (880 bp). Normalization was confirmed using primers specific for glycerol aldehyde-3-phosphate dehydrogenase (GAPDH). NC, negative control (without cDNA). See "Experimental Procedures" for further details.
Inability to octamerize, as observed for the common P426L allele of arylsulfatase A, contributes to metachromatic leukodystrophy pathology (30). In contrast to arylsulfatase A, gel filtration analysis of ASG showed no indication for a possible dimerization or oligomerization. The active state of ASG is the unprocessed 63-kDa monomer. Kinetic analysis revealed a strong, competitive inhibition of ASG by phosphate (K i ϭ 17 M). This inhibition might represent an additional regulatory mechanism to control ASG activity. For instance, secreted ASG that has escaped from lysosomes would be entirely inactivated in the extracellular environment due to the neutral pH as well as the high phosphate concentrations (e.g. 0.6 -1.3 mM in blood serum). To the best of our knowledge, absolute intralysosomal phosphate concentrations have not been directly determined; however, the uptake of phosphate into human fibroblast lysosomes has been characterized as a saturable transport system with K m ϭ 5 M (38). This suggests that ASG activity could be modulated and restricted to specific biological conditions by changes in phosphate concentrations. Inhibition of ASG by sulfate is probably of less relevance, because physiological sulfate concentrations (0.3-0.5 mM in blood serum) (39) do not result in a significant reduction of ASG activity.
The physiological substrates of ASG remain to be characterized. For many sulfated biomolecules, the corresponding sulfatases have not been identified so far. The lysosomal localization and differential expression with highest levels in liver and kidney suggest the loss of ASG activity to be associated with a previously unidentified lysosomal storage disorder. A comprehensive analysis of physiological substrates will be required in order to elucidate the metabolic role of ASG. The generation of ASG knock-out mice is in progress in our laboratory and will hopefully help in identifying the physiological substrates and in understanding the biological significance of this sulfatase, which might ultimately lead to the identification of ASG-deficient patients. Such patients could then be treated by either enzyme replacement, substrate reduction, enzyme enhancement, or gene therapy as currently being tested and practiced for other lysosomal enzyme deficiencies. The finding that ASG contains mannose 6-phosphate particularly qualifies this sulfatase for enzyme replacement therapy.