Glycobiology Advance Access originally published online on October 27, 2007
Glycobiology 2008 18(1):53-65; doi:10.1093/glycob/cwm121
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Overexpression of the 3'-Phosphoadenosine 5'-Phosphosulfate (PAPS) Transporter 1 Increases Sulfation of Chondroitin Sulfate in the Apical Pathway of MDCK II Cells
2 Department of Molecular Biosciences, University of Oslo, Box 1041 Blindern, 0316 Oslo, Norway
1 To whom correspondence should be addressed: Tel: +47-22856753; Fax: +47-22854443; e-mail: kristian.prydz{at}imbv.uio.no
Received on July 13, 2007; revised on October 3, 2007; accepted on October 24, 2007
 |
Abstract |
|---|
 Top Abstract IntroductionResultsDiscussionMaterials and methodsSupplementary DataConflict of interest statementReferences |
|---|
The canine 3'-phosphoadenosine 5'-phosphosulfate (PAPS) transporter 1 fused to GFP was stably expressed with a typical Golgi localization in MDCK II cells (MDCK II-PAPST1). The capacity for PAPS uptake into Golgi vesicles was enhanced to almost three times that of Golgi vesicles isolated from untransfected cells. We have previously shown that chondroitin sulfate proteoglycans (CSPGs) are several times more intensely sulfated in the basolateral than the apical secretory pathway in MDCK II cells (Tveit H, Dick G, Skibeli V, Prydz K. 2005. A proteoglycan undergoes different modifications en route to the apical and basolateral surfaces of Madin-Darby canine kidney cells. J Biol Chem. 280:29596–29603). Here we demonstrate that increased availability of PAPS in the Golgi lumen enhances the sulfation of CSPG in the apical pathway several times, while sulfation of CSPGs in the basolateral pathway shows minor changes. Sulfation of heparan sulfate proteoglycans is essentially unchanged. Our data indicate that CSPG sulfation in the apical pathway of MDCK II cells occurs at suboptimal conditions, either because the sulfotransferases involved have high Km values, or there is a lower PAPS concentration in the lumen of the apical secretory route than in the basolateral counterpart.
Key words: Golgi apparatus / MDCK epithelial cells / PAPS / proteoglycan / sulfation
|
Introduction |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary DataConflict of interest statementReferences |
|---|
Proteoglycans (PGs) are proteins carrying long glycosaminoglycan (GAG) chains, consisting of repeating disaccharide units modified by epimerization and sulfation. Although most GAGs are linear structures, great structural diversity has been observed, due to variability in the modification patterns. PGs carrying GAG chains with characteristic modification patterns are important in cell–cell recognition events, cell migration, organ morphogenesis, and other developmental processes (Deepa et al. 2004
; Kreuger et al. 2006). Although the building blocks and enzymatic events in GAG synthesis and modification are well characterized, less is known about how these processes are regulated. The aim of this work is to study whether increased availability of the sulfate donor 3'-phosphoadenosine 5'-phosphosulfate (PAPS) in the lumen of the secretory pathway influences PG synthesis.
Nucleotide sugars, such as UDP-N-acetyl-glucosamine (UDP-GlcNAc), UDP-N-acetyl-galactosamine (UDP-GalNAc), and UDP-glucuronic acid (UDP-GlcUA), are substrates for the GAG polymerizing enzymes, while the sulfate group of PAPS is transferred onto the polymerizing chain in various positions by specific sulfotransferases (Kjellen and Lindahl 1991). Polymerization and modification take place in the secretory pathway, initiated in the ER/ERGIC and commencing through the Golgi apparatus. Activation of sugars and sulfate by nucleotides occurs in the cytoplasm from where the activated forms are translocated into the Golgi lumen by specific nucleotide sugar transporters (NSTs) (Prydz and Dalen 2000). Molecular cloning of NSTs of the solute carrier family 35 (SLC35) (Ishida and Kawakita 2004) has identified proteins which together possess all the necessary activities to translocate substrates for heparan sulfate (HS) and chondroitin sulfate (CS) synthesis (Ashikov et al. 2005). Specific sulfotranferases transfer sulfate from PAPS to designated sites on the sugar residues (Kusche-Gullberg and Kjellen 2003). In higher eukaryotes, PAPS is synthesized in the cytoplasm or nucleus from sulfate and ATP by PAPS synthetase (Besset et al. 2000; Li et al. 1995). PAPS may then be translocated into the Golgi lumen by specific PAPS transporter molecules in the Golgi membrane (Schwarz et al. 1984). In cell culture, synthesis of PAPS can be inhibited by addition of chlorate, a competitive inhibitor of sulfate binding to PAPS synthetase (Baeuerle and Huttner 1986). Reduced PAPS levels lead to decreased sulfation of proteins and PGs (Fjeldstad et al. 2002; Keller et al. 1989). Upregulation of PAPS synthesis by overexpression of PAPS synthetase does not seem to increase sulfation of PGs (Girard et al. 1998). This points to the PAPS transporter as a possible regulator of PAPS availability. Earlier biochemical studies have identified Golgi membrane proteins with PAPS binding capability or PAPS transport activity (Lee et al. 1984; Mandon et al. 1994; Ozeran et al. 1996a). Recently, two human PAPS transporters were cloned, PAPST1 and PAPST2 (Kamiyama et al. 2003, 2006). The PAPST1 homolog in drosophila, Slalom (Sll), has been mutated, resulting in reduced sulfation of HSPG, with concomitant reduction in Wg and Hh signaling during development (Luders et al. 2003). The activity of the PAPS transporters mentioned above has been characterized, but their respective involvement in PG synthesis remains to be clarified.
The MDCK epithelial cell line is a well-established model for studies of polarized protein and lipid sorting. The cell line exists as two variants, MDCK I and MDCK II, with some differences in PG synthesis (Svennevig et al. 1995). MDCK II cells produce mainly PGs of CS and HS type, while only small amounts of keratan sulfate (KS) and hyaluronic acid (HA) have been observed (Schepers et al. 2003; Svennevig et al. 1995; Toma et al. 1996). Grown on filters, MDCK cells establish tight epithelial monolayers. From these, HSPGs are secreted mainly basolaterally, contributing to the matrix of the basal membrane, while chondroitin sulfate proteoglycans (CSPGs) are mainly secreted apically (Kolset et al. 1999). The structures of CS chains added to protein cores in the apical and basolateral secretory pathways differ dramatically, even for the same protein core, as observed in MDCK II cells expressing the serglycin-green fluorescent protein (GFP) fusion protein (Tveit et al. 2005; Vuong et al. 2006) and as observed for the endogenous PG versican in this work. These observations indicate that separate platforms for apical and basolateral CS synthesis are established before the terminal sorting station, the trans-Golgi network (TGN) (Tveit et al. 2005). One explanation for the pronounced difference in sulfation intensity of apical and basolateral CSPGs could be differential availability of the substrate, PAPS, to the two pathways. We therefore wanted to study factors involved in PAPS transport and address their relationship with polarized sulfation and secretion of PGs in MDCK II cells.
In the present study we have investigated how PG synthesis is affected by enhanced expression of canine PAPST1 in MDCK II cells. To elucidate the effect of PAPS transport on GAG synthesis, we have cloned the canine PAPST1 and introduced it as a GFP-fusion protein in MDCK II cells in addition to the endogenous activity. The PAPST1-GFP localized to the Golgi apparatus and contributed to an increased PAPS uptake into the Golgi lumen. The resulting changes in PG synthesis and modification in the apical and basolateral secretory pathways are reported.
|
Results |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary DataConflict of interest statementReferences |
|---|
Identification of expression and cloning of canine PAPST1
A cDNA library derived from MDCK II cells, as described in the Materials and methods section, was subjected to PCR with designated primers (see supplementary data). After the expression of PAPST1 was confirmed, the full length coding sequence of canine SLC35 B2 was cloned as described in detail in supplementary data. The canine SLC35 B2 from MDCK II cells, acc.nr. EF568109, has a coding sequence of 1299 bp, corresponding to 433 amino acids. The deduced protein sequence is very well conserved compared to the human and other mammalian PAPS transporter 1 proteins, with only 15 amino acid substitutions separating human and canine PAPST1.
PAPST1-GFP localizes to the Golgi apparatus and increases PAPS uptake into Golgi vesicles
To study the involvement of PAPST1 in transport of PAPS from the cytoplasm to the Golgi lumen, we constructed a fusion protein with GFP C-terminally. MDCK II cells were transfected, and stable transfectants expressing PAPST1-GFP (MDCK II-PAPST1) were generated. Fourteen clones were expanded after selection, and the expression of PAPST1-GFP assessed by live confocal imaging. Three of the fourteen clones expressed PAPST1-GFP. All the three clones expressed the fluorescent fusion protein with a typical Golgi staining. The Golgi localization was confirmed by co-localization with the Golgi marker BODIPY-TR-C5-ceramide (Figure 1). Golgi vesicles from MDCK II and MDCK II-PAPST1 cells were enriched by subcellular fractionation to determine [35S]PAPS uptake (Fjeldstad et al. 2002). Golgi enriched fractions from MDCK II-PAPST1 cells showed an increased uptake of [35S]PAPS (cpm/µg Golgi protein), compared to Golgi fractions from nontransfected MDCK II cells (Figure 2). Thus, the expression of GFP-PAPST1 and localization to the Golgi apparatus resulted in an increased PAPS uptake into the Golgi lumen.
|
|
Overexpression of PAPST1-GFP alters sulfation of secreted macromolecules
MDCK II cells grown on filter supports form tight polarized monolayers and exhibit polarized sorting of several classes of molecules, including sulfated macromolecules (Svennevig et al. 1995
). Although some sulfate is carried by glycoproteins, such as the major secreted glycoprotein gp80, most of the sulfate is incorporated into PGs. During transport to the cell surface, PGs acquire sulfate groups to different extents. The amount of sulfate delivered with macromolecules to the apical and basolateral domains of epithelial cells depends on the types of macromolecules transported and their degree of sulfation. Incorporation of [35S]sulfate into macromolecules was increased to different extents compared to untransfected MDCK II cells in the three clones expressing PAPST1-GFP (not all three clones shown). For molecules secreted apically, the incorporation of sulfate was increased two to three times. For the basolateral pathway, only a slight increase was observed (Figure 3). The increased expression of the PAPST1 thus levels out much of the difference in sulfation intensity observed for macromolecules in the apical and basolateral secretory pathways of MDCK II cells with a basal PAPST1 level. To study the biochemical basis for these alterations, one of the three transfected cell lines expressing PAPST1-GFP was chosen for further characterization of PG synthesis and sulfation.
|
Altered sulfation of proteoglycans secreted apically
PGs carrying heparan sulfate (HS) and/or chondroitin sulfate (CS) are the major sulfated macromolecules secreted by MDCK II cells. After SDS–PAGE analysis, PGs appear as a smear in the upper region of the gel (above 220 kD), due to the large mass and heterogeneous nature of the associated GAG chains (Figure 4A). Different radioactive precursors added to the culture medium are incorporated into the different moieties of the PGs. Metabolic labeling with [35S]sulfate reveals the amount of sulfate carried by GAG chains. For MDCK II cells, a strong signal is observed in the PG-region basolaterally (lane 4) compared to the much weaker apical signal (lane 3, Figure 4A). Indeed, most of the sulfate is bound to molecules secreted to the basolateral medium. MDCK II-PAPST1 cells, however, exhibit [35S]sulfate signals of similar strength for macromolecules secreted apically and basolaterally (lane 1 and 2, Figure 4A). The increase in sulfation observed for the apical pathway could be due to an overall increase in apical PG secretion, which would be reflected in an increased secretion of [35S]cystein/methionine (Cys/Met) labeled protein cores to the apical medium. The [35S]Cys/Met-labeled protein cores above 220 kD are mainly PGs. Degradation of HS- and chondroitin sulfate-glycosaminoglycans (CS-GAGs) in [35S]Cys/Met-labeled samples resulted in reduced molecular weight of the most prominent protein bands, confirming PG protein cores (data not shown). The slight increased secretion of protein cores both to apical and basolateral media in PAPST1-MDCK II (comparing lanes 5 and 7, and lanes 6 and 8, Figure 4A) cannot account for the increased apical sulfation. Especially since the pattern of PG protein core secretion to the two opposite medium compartments is similar for MDCK II and MDCK II-PAPST1 cells (comparing lanes 5 and 6, and lanes 7 and 8, Figure 4A). In both cases, most of the PG protein cores are secreted basolaterally. Alternatively, the cause could be an increased incorporation of sulfate into GAGs of the PGs secreted apically. The load of GAG chains carried by a PG can be determined by metabolic incorporation of [3H]GlcN during polymerization. Incorporation of [3H]GlcN in macromolecules above 220 kD is mainly as GAGs in PGs. After degradation of HS- and CS-GAGs in [3H]GlcN-labeled samples, the signal from [3H] almost disappeared for basolateral sample. For the apical sample, the signal from [3H] was reduced, and the most prominent protein band was reduced in molecular mass (data not shown). The secretion pattern of GAGs is quantitatively similar for MDCK II and MDCK II-PAPST1 cells (lanes 9–12, Figure 4A), but there is a difference in the average molecular mass for the PGs in the respective apical medium (lanes 9 and 11), implying a shift to shorter GAG chains for MDCK II-PAPST1 (as seen in Figure 9). Quantification of the incorporated sulfate relative to the incorporated [35S]Cys/Met and [3H]GlcN, respectively, reveals an increase in sulfation of the PGs secreted to the apical medium from the epithelial cell layer in MDCK II-PAPST1 (Figure 4B), implying a higher sulfate density of GAGs secreted by MDCK II-PAPST1 cells compared to the corresponding GAGs secreted by MDCK II cells. For PGs secreted basolaterally, the sulfate density is similar for MDCK II and MDCK II-PAPST1 cells.
|
|
Increased sulfate incorporation into apically secreted chondroitin sulfate proteoglycans
The observed increase in sulfation of macromolecules secreted apically could be a general enhancement of sulfation of both HS- and CS-GAGs released to the apical side of the epithelium, or it could be due to increased sulfation of one GAG-type only. In previous investigations of PGs synthesized by MDCK II cells, HSPG was predominantly secreted to the basolateral medium, but a significant amount of HSPG was also secreted apically (approximately 30% of secreted HSPG, based on distribution of [35S]sulfate). CSPG is synthesized in lesser amounts, but is mainly secreted to the apical side (Kolset et al. 1999
). To examine the contribution of HS and CS, respectively, to the total amount of sulfated PGs, we carried out specific degradation of each GAG type and quantified the remaining undigested material. For MDCK II cells we confirmed the pattern of secretion observed previously (Svennevig et al. 1995). For basolateral samples, most of the sulfate-labeled macromolecules in the PG-region were degraded by HNO2, an HS-GAG degrading agent (Figure 5A, lanes 10–12). There was a minor contribution from CSPG basolaterally, seen as a HNO2-resistent band in the upper part of the gel (Figure 5A, lane 11), which could be degraded by cABC (Figure 5A, lane 10). The distribution of CS/HS was quantified by phosphorimaging (Figure 5B). In the apical medium from MDCK II cells, there was a more even contribution from CS- and HS-GAGs (Figure 5A, lanes 7–8). MDCK II-PAPST1 exhibits a basolateral secretion pattern similar to that of MDCK II cells (Figure 5A, lanes 4–6 and 10–12), while a marked difference in apical secretion was observed for the two cell lines. Degradation of CS with cABC (Figure 5A, lane 1) removes most of the [35S]sulfate signal in the PG-region for MDCK II-PAPST1 cells. In this cell line, most of the sulfate delivered with macromolecules to the apical medium is carried by CSPG. There is a more condensed band left after cABC degradation, which also barely can be seen for MDCK II (Figure 5A, lanes 1 and 7). This is probably a CSPG protein core carrying sulfated stubs after GAG degradation and/or sulfated N-glycans. Quantification of sulfate incorporation by phosphorimaging demonstrates that the CS/HS [35S]sulfate ratio has changed (Figure 5B) due to an increased incorporation of sulfate into apically secreted CS-GAGs in MDCK II-PAPST1 cells.
|
Increased sulfation of the CSPG versican, but not of the HSPG perlecan
Previous studies have demonstrated that MDCK II cells secrete the CSPG versican and the HSPG perlecan in a polarized manner (Svennevig et al. 1995
). By immune precipitation (IP) of versican from apical and basolateral media, we could compare the sulfate incorporation into CS chains attached to this core protein in the two cell lines. If the Golgi apparatus of MDCK II-PAPST1 cells has an increased capacity for sulfation in the apical pathway, then versican secreted apically from these cells could carry more sulfate compared to versican secreted from untransfected MDCK II cells. To determine the incorporation of sulfate compared to the total amount of GAG chains and protein cores, we labeled filter-grown cells metabolically with [35S]sulfate, [3H]GlcN, or [35S]Cys/Met, as described for previous experiments. The same approach was taken for the basolaterally secreted HSPG perlecan. Medium samples containing versican and perlecan were treated with cABC or HNO2 after IP to confirm that the expected type of GAG chains was attached to each of the two PGs (data not shown). IP of versican after [35S]sulfate labeling gave a broad band in the upper PG-region of the SDS–PAGE gels, above 220 kD. The amount of [35S]sulfate incorporated into versican secreted from MDCK II-PAPST1 cells was markedly increased in both the apical and the basolateral medium compared to that observed for MDCK II wild-type cells (Figure 6A, lanes 1–4). Also the secretion pattern was changed. Phosphorimaging revealed that there was about three times more sulfate attached to versican secreted to the apical medium compared to the basolateral counterpart (Figure 6A, lanes 1 and 2). For MDCK II cells the amount of sulfate associated with versican was almost equal in the two media (Figure 6A, lanes 3 and 4). A difference in the molecular weight of versican secreted by the two cell lines was also observed. Versican secreted basolaterally from MDCK II cells was seen both as a high molecular weight species remaining in the application well, and as a broader band further into the gel (Figure 6, lane 4). The large molecular weight material remaining in the well was not seen for MDCK II-PAPST1 basolateral medium samples (Figure 6A, lane 2). For these cells, the basolateral versican had a lower molecular weight than the apical counterpart. Metabolic labeling with [35S]Cys/Met demonstrated that most of the versican was secreted apically from both cell lines (Figure 6A, lanes 5–8), with an apical signal six to seven times stronger than the basolateral counterpart. Incorporation of [3H]GlcN reveals the amount of GAGs carried by versican in the two cell lines (Figure 6A, lanes 9–12). For the MDCK II-PAPST1 cell line, there is a strong apical signal. The secretion pattern is slightly more polarized than for MDCK II cells, but the difference is less pronounced than what was observed after [35S]sulfate labeling. The signals were quantified and the distribution of apically and basolaterally secreted versican, labeled with [35S]sulfate, [35S]Cys/Met, or [3H]GlcN, was calculated as percent of total versican secretion (Figure 6B). The MDCK II-PAPST1 cells displayed a shift in [35S]sulfate-labeled versican distribution compared to wild-type cells. The distribution of protein cores and GAG chains was similar in the two cell lines. This demonstrates an increased sulfation of versican secreted apically in MDCK II-PAPST1 cells.
|
[35S]sulfate labeling followed by immune precipitation of perlecan gave a band above 220 kD on SDS–PAGE gels. [35S]sulfate-labeled perlecan was mainly secreted to the basolateral medium, but a small fraction was also detected in the apical medium. Slight differences in the molecular size of perlecan were observed on opposite sides of the cell monolayer. Basolaterally, the band was more intense and stretched further into the gel, (Figure 7A, comparing lanes 2 and 4 to lanes 1 and 3). Only slight differences were observed when comparing perlecan expression in MDCK II-PAPST1 cells to that of MDCK II cells, regardless the metabolic label used (Figure 7A). The distribution of apically and basolaterally secreted perlecan labeled with [35S]sulfate was similar in the two cell lines (Figure 7B). The distribution of [35S]Cys/Met or [3H]GlcN incorporated into perlecan was also the same. Thus, perlecan sulfation and secretion appeared similar in the two cell lines. Our results demonstrate that the increased sulfation observed when overexpressing PAPST1 was essentially caused by increased sulfation of CS. This was exemplified by the immune precipitation of the endogenous CSPG versican, and was particularly evident for the major fraction of versican which was secreted to the apical medium. The sulfation level of the HSPG perlecan was essentially unchanged.
|
Disaccharide analysis of secreted proteoglycan GAG chains confirms alteration in sulfate density in CS but not in HS
To confirm alterations in sulfation of the GAG chains, the disaccharide composition of HS- and CS-GAGs isolated from PGs secreted to the apical and basolateral media of the two cell lines was analyzed. Filter-grown MDCK II and MDCK II-PAPST1 cells were allowed to reach confluency before incubation in the serum-free medium for 24 h at 37°C and 5% CO2. Macromolecules secreted during the incubation period were collected and processed for disaccharide analysis, as described in the Materials and methods section. Data from at least three replicas were used for determination of sulfate density and sulfation pattern.
Analysis of CS (Figure 8C and D) revealed four different types of disaccharides. The dominating type of CS disaccharides of GAGs secreted from MDCK II cells was the unsulfated 
Di-0S. Even after adjustment by subtracting the amount of Di-HA (degraded hyaluronic acid), a side-product after cABC degradation, Di-0S counts for more than 60% of the CS disaccharides. The HA contribution was estimated by digestion with Hyaluronidase (St. Hyal.) of [3H]GlcN-labeled material followed by separation by gel filtration or digestion by Hyaluronidase SD (S. Diag.) and subsequent disaccharide analysis (data not shown). CS-GAGs secreted to the two sides of the cell layer differed in amount and composition. Most of the CS-GAGs secreted (75%) was recovered apically, confirming earlier findings (Kolset et al. 1999). The average sulfate density of apically and basolaterally secreted CS-GAGs differed (Table I), the basolateral CS-GAGs being more intensely sulfated. This is in line with the results from IP of radioactively labeled versican, which exhibited different sulfation densities at the two sides of the cell layer. The most prominent of the sulfated disaccharides was Di-4S. Small amounts of Di-6S and the disulfated-type Di-4,6diS were also detected. The signal from Di-6S seemed to be more prominent for basolateral than for apical samples (Figure 8D). This is in agreement with what we previously have observed for the CS-GAGs of serglycin, when this PG was expressed in MDCK II cells (Tveit et al. 2005). The transfected cell line MDCK II-PAPST1 displayed the same types of disaccharides as wild-type cells, but with a dramatically altered distribution. The relative amounts of all sulfated CS disaccharides were increased, with a corresponding decrease in the level of the unsulfated disaccharide. Increased sulfation of CS disaccharides was observed for PGs secreted both to the apical and the basolateral medium of MDCK II-PAPST1 cells, but was most prominent for apical PGs (Figure 8C and D). The total amount of apically secreted CS-GAGs (the sum of the disaccharide components) was reduced by 30% for MDCK II-PAPST1 cells compared to MDCK II cells. Thus, the increased sulfation was accompanied by a reduced amount of GAG disaccharides. This indicates that increased sulfation either leads to shorter, or to a reduced number of GAG chains. When apical and basolateral populations of [35S]sulfate-labeled GAG chains were subjected to gel filtration chromatography, a shift toward shorter GAG chains was observed for PGs secreted from MDCK II-PAPST1 cell monolayers into both the apical and the basolateral medium, compared to the corresponding for MDCK II cells (Figure 9). The difference in chain length is probably not related to increased degradation. When metabolically labeled PGs (from a previous experiment) were added to the cells and incubated under the same conditions as for metabolic labeling, we could not detect any decrease in chain length (data not shown).
|
|
HS disaccharide analysis (Figure 8A and B) revealed two variants of unsulfated disaccharides (
UA-GlcNH and UA-GlcNAc), three types of sulfated disaccharides (UA-GlcNS, UA-GlcNAc-6S, and UA-2S-GlcNAc), two disulfated disaccharides (UA-GlcNS-6S and UA-2S-GlcNS), and finally one trisulfated disaccharide (UA-2S-GlcNS-6S). The unsulfated UA-GlcNAc was the dominant disaccharide in all samples, comprising around 40–50% of the total. In CS-GAGs, disaccharides with one sulfate group dominated among the sulfated ones. In HS, the main sulfated disaccharide was UA-GlcNS and the trisulfated UA-2S-GlcNS-6S. HS-GAGs from MDCK II cells consisted of several disaccharide types carrying two or three sulfate groups, and were thus on average more densely sulfated than the CS-GAGs. HS-GAGs secreted to the two sides of the cell layer differed in amount, but not in composition. Five to six times more of HS-GAGs were secreted basolaterally than apically for both MDCK II and MDCK II-PAPST1 cells (Figure 8A and B). Average sulfation density per disaccharide was around 1.0 (Table I). The average sulfation density of basolaterally secreted PGs in MDCK II-PAPST1 seemed somewhat decreased compared to MDCK II, but not significantly. The distribution between the two disulfated disaccharides, UA-GlcNS-6S and UA-2S-GlcNS, differs in MDCK II and MDCK II-PAPST1 cells. The amount of UA-2S-GlcNS is increased in MDCK II-PAPST1 cells in both apical and basolateral HS. Still HS-GAG sulfation does not contribute to the increased sulfation of apically secreted PGs in MDCK II-PAPST1 cells. |
Discussion |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary DataConflict of interest statementReferences |
|---|
PAPST1 is involved in Golgi uptake of PAPS in MDCK II cells
Synthesis and modification of proteoglycan GAG chains take place in the lumen of the Golgi apparatus. These processes are regulated by Golgi enzymes carrying out polymerization of GAGs and subsequent epimerization and sulfation. In addition, the availability of substrates, nucleotide sugars and activated sulfate (PAPS), is also crucial for the PG output. These substrates are imported from the cytoplasm, across the Golgi membrane by specific transporters (as reviewed in Berninsone and Hirschberg 2000). The primary sequences of several of these multimembrane-spanning transport proteins have become available after their molecular cloning. The transporters have been divided into subgroups of the SLC35, based on their sequences (as reviewed in Ishida and Kawakita 2004).
PAPS transport activity of Golgi vesicles was established two decades ago (Schwarz et al. 1984), but attempts to biochemically purify the protein harboring this activity gave contradictory results. Mandon et al. described a 75 kDa protein with an apparent Km of 1.7 µM (Mandon et al. 1994), while Ozeran et al. reported a 230 kDa protein with a Km of 0.83 µM (Ozeran et al. 1996a, 1996b). The recent molecular cloning of the human SLC35B2 and B3 demonstrated PAPS transport activity for the protein products designated PAPST1 and PAPST2 (Kamiyama et al. 2003, 2006). PAPST1 consists of 432 amino acids (corresponding to a mass of 
48 kDa) with an apparent Km of 0.8 µM (Kamiyama et al. 2003). The drosophila homolog of PAPST1, Slalom, is identified by molecular cloning and harbors PAPS transport activity (Luders et al. 2003). We have cloned the canine PAPST1 and overexpressed it in MDCK II cells. This increased [35S]PAPS uptake into Golgi vesicles and the sulfation intensity of apically secreted CS-GAGs. We have previously manipulated MDCK II GAG synthesis by reducing the availability of PAPS in the Golgi lumen by treatment with chlorate or Brefeldin A (Fjeldstad et al. 2002). Low concentrations of chlorate reduced the sulfation of CS-GAGs and proteins, while higher concentrations were needed to reduce HS-GAG sulfation. Brefeldin A treatment discriminated between CS and HS in a similar way, but in addition, the reduction was specific for the apical pathway. Identifying PAPST1 in MDCK II cells and overexpressing this transporter allowed us to study the opposite situation, increased availability of PAPS in the Golgi lumen.
Increased Golgi uptake of PAPS influences CS-GAG synthesis
Nucleotide sugar transporters (NSTs) of the Golgi apparatus have been proven vital for PG synthesis. In drosophila, both the UDP-sugar transporter Fringe-connection (Selva et al. 2001) and the PAPS transporter Slalom (Luders et al. 2003) participate in PG synthesis. The connection between NSTs and PG synthesis has also been demonstrated for mammalian cells. Overexpression of a human UDP-GlcNAc/UDP-Glc transporter in HCT116 cells leads to increased levels of cell surface HS (Suda et al. 2004). In MDCK II cells, a UDP-Gal transporter has been shown to regulate keratan sulfate synthesis and to some degree influence CS sulfation (Toma et al. 1996). In this study, we have undertaken a detailed study of HS and CS GAGs synthesized by wild-type MDCK II cells and MDCK II cells stably overexpressing the canine PAPST1. There is a clear difference in the extent of GAG sulfation on PGs secreted by the two cell lines. Both by metabolic labeling and disaccharide analysis, we could detect increased CS sulfation, while HS sulfation was essentially unchanged. To further verify this GAG specific alteration in sulfation, we studied the sulfation level of two different endogenously synthesized PGs, the large CSPG versican and the HSPG perlecan. Versican was clearly more intensely sulfated in MDCK II-PAPST1 than in MDCK II cells, while the HSPG perlecan exhibited similar sulfation levels in both cell types.
Increased CS-GAG sulfation in the apical pathway
The MDCK II cell line is an established model for studies of polarized sorting and secretion of lipids and proteins, including PGs. CSPGs are mainly secreted to the apical medium from polarized MDCK II cell monolayers, while most of the HSPGs are secreted basolaterally (Svennevig et al. 1995). The MDCK II-PAPST1 cell line exhibits the same PG secretion pattern as the parental cell line, as seen for the [35S]Cys/Met-labeled protein cores and [3H]GlcN-labeled GAGs. The incorporation and distribution of [35S]sulfate, however, are changed. PGs secreted apically by MDCK II-PAPST1 cells carry two to three times more sulfate than their counterparts in MDCK II cells. The increase in CSPG sulfation mainly affects the apical PG population due to the secretion pattern of CS- and HSPGs. CSPGs constitutes a much larger fraction of apically secreted PGs than the basolateral counterpart, which mainly consists of HSPGs. Thus, alterations in CS sulfation have a greater impact on the apically secreted PG population than the basolateral. In addition, there are differences between apically and basolaterally secreted CS. Our best interpretation of the experimental data is that CS-GAG synthesis in the apical and basolateral pathway of MDCK II cells takes place in different Golgi environments. We have previously reported that a nonendogenous PG, serglycin, when expressed in MDCK II cells, acquires different CS-GAG sulfation patterns, sulfation intensities, and GAG chain lengths in the apical and the basolateral secretory pathways (Tveit et al. 2005). Now, we demonstrate a similar difference in sulfation intensity for the endogenously expressed CSPG versican. In MDCK II cells, the basolaterally secreted versican carries more sulfate per protein core and GAG chain than the versican secreted to the apical medium. Disaccharide analysis of CS secreted from MDCK II cells revealed that basolaterally secreted CS GAGs have a higher average sulfate density (Table I). The Di-6S is more prominent in CS-GAGs secreted basolaterally than those secreted apically. Overexpression of PAPST1 enhances the sulfation of CS. In MDCK II-PAPST1 cells, the sulfate density of CS is increased for both apical and basolateral GAGs, but much more dramatically for CS secreted apically, where the sulfate density changes from 0.17 to 0.90 (sulfate groups per disaccharide). Also IP of versican demonstrated more dramatic changes to the apically secreted variant of this CSPG. The ratio of GAG-incorporated sulfate to protein core synthesis increased two to three times, for apical media from MDCK II-PAPST1 cells over the corresponding from MDCK II cells. Only a minor increase was observed for the basolaterally secreted versican.
Differences in HS-GAG and CS-GAG synthesis
The synthesis of HS- and CS-GAGs is initiated by the addition of xylose to a serine residue in the protein core followed by completion of a linker tetrasaccharide. Addition of the fifth sugar is a decision point, where either GlcNAc, destined for HS, or GalNAc, destined for CS is added (as reviewed in Prydz and Dalen 2000). The enzymes involved in polymerization and modification of CS have been postulated to reside in the medial-Golgi and TGN, while HS synthesis is thought to take place prior to the TGN in most cell types (Spiro et al. 1991). The cis-, medial-, or trans-Golgi localization of endogenous PAPST1 in MDCK II cells, or any other cell type, has not been firmly determined, but transfection experiments in other cell lines have revealed that human PAPST1 partly co-localizes with trans-Golgi markers (Kamiyama et al. 2003). A restricted localization to early regions of the Golgi apparatus, harboring HS synthesizing enzymes, could favor synthesis of HS above CS in MDCK II wild-type cells. The PAPST1-GFP does not display a restricted Golgi localization in MDCK II-PAPST1 cells, when compared to the fluorescent lipid Golgi marker BODIPY-TR-C5-ceramide, but the resolution of the method is not sufficient to exclude a differential localization of PAPST1-GFP among Golgi subcompartments. We propose that the presence of PAPST1-GFP in MDCK II-PAPST1 cells increases the availability of PAPS throughout the Golgi apparatus, resulting in more similar sulfation intensities for CS and HS. MDCK II cells, both CS and HS sulfation, are sensitive to the PAPS concentration. There is clearly a potential for increased CS sulfation in MDCK II cells. The HS sulfation density is essentially unchanged, but the alteration in HS disulfated disaccharide composition between the two cell lines could reflect a response to altered PAPS concentration. Relatively small amounts (2–5 mM) of the PAPS synthase inhibitor, chlorate, led to a dramatic reduction in CS sulfation, while HS sulfation was essentially unaffected (Fjeldstad et al. 2002; Safaiyan et al. 1999). The higher sensitivity of CS sulfation to a reduction in the PAPS concentration could be explained by the observation that CS sulfotransferases in general exhibit higher Km values for PAPS than sulfotransferases involved in HS synthesis (Kolset et al. 2004). Increasing the PAPST1 level by transfection increases the uptake of PAPS into the Golgi lumen, which again leads to more efficient sulfate transfer by the CS sulfotransferases, and, hence, the sulfate density of CS-GAGs is increased. In line with existing data, CS synthesis seems more flexible and prone to alterations than HS synthesis. A similar relationship between CS and HS synthesis has been observed for two syndecans expressed in epithelial cells. Syndecan-1 and -4 carry CS-GAGs that differ significantly in structure and sulfate density, while the HS-GAGs of the same syndecans are structurally indistinguishable (Deepa et al. 2004).
The most prominent difference in PG synthesis between the MDCK II-PAPST1 and the MDCK II cell lines is the increased sulfation of CS-GAGs in the apical pathway. However, we could also detect differences in GAG chain lengths between the two cell lines. Both SDS–PAGE analysis of total PGs and of immune precipitated versican indicated PGs with higher molecular weight in the parental cell line, MDCK II. When [35S]sulfate-labeled GAGs from the two cell lines were analyzed by gel filtration, the profiles differed markedly. Both for apical and basolateral GAGs the sulfate label was relatively speaking shifted toward a pool of shorter GAGs in MDCK II-PAPST1. At present, six enzymes that might take part in CS synthesis have been cloned and characterized; chondroitin sulfate synthase 1–3 (Kitagawa et al. 2001; Yada, Gotoh et al. 2003; Yada, Sato et al. 2003), CS-GalNAc-transferase 1–2 (Gotoh et al. 2002; Uyama et al. 2003), and CS glucuronyl-transferase (Gotoh et al. 2002). The different enzymes displayed variable activities toward nonsulfated and sulfated chondroitin substrates. The polymerizing enzymes also differed in activity toward substrates of different sulfate constitution. 6-O-sulfation has been shown to reduce the activity of GAG polymerases (Sato et al. 2003). The observed shortening of GAG chains with increased sulfate density supports the view that sulfation has an impact on chain termination. This is in agreement with the observed increase in CS- and HS-GAG chain lengths observed in MDCK II cells upon treatment with the inhibitor of PAPS synthesis, chlorate (Vuong et al. 2006). Further insight into the activities and Golgi localization of enzymes involved in CS biosynthesis in MDCK II cells would give valuable information concerning the coordinated synthesis and sorting of PGs and the consorted function of the enzymes involved.
Previous studies have indicated that HS and CS synthesis occurs in sequence, where HS is completed before the trans-Golgi network (TGN), while CS synthesis is completed in the TGN (reviewed in Prydz and Dalen 2000). Our data indicate that PGs that are secreted at the apical and basolateral surfaces of MDCK II epithelial cells are segregated before their synthesis is completed in the Golgi apparatus. This view is supported by the observed differences in GAG chain length and sulfation intensity, observed for CS-PGs in general, for the endogenous CSPG versican (in the present work), and for the serglycin-GFP expressed after transfection (Vuong et al. 2006). The increased CS sulfation observed in the apical pathway of MDCK II-PAPST1 cells, where the PAPS concentration in the Golgi lumen was increased, did not affect the sorting polarity of versican protein cores. Still the additional sulfate groups added to CS-GAGs did not distribute with the same overall pattern as for CS-GAGs in MDCK II (wild-type cells), but were rather favoring shorter CS chains. These results point to GAG synthesis as a more regulated process than simply a sequential synthesis of HS and CS on the way through the Golgi apparatus.
|
Materials and methods |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary DataConflict of interest statementReferences |
|---|
Molecular cloning of canine PAPST1 and construction of GFP-fusion protein
The human PAPST1 sequence by Kamiyama et al. (2003)
was used as a query in a BLAST (Altschul et al. 1997) search via Entrez in GenBank. Homologous sequences were compared by sequence alignments in ClustalW (Chenna et al. 2003) at EBI-EMBL. Parts of the sequence exhibiting high degree of conservation were used for primer design by the Primer3 program (Rozen and Skaletsky 2000). Selected primer sequences are listed in supplementary data. Primers were ordered from Medprobe/Eurogentec, Liege, Belgium. A cDNA library was made from MDCK II cells by the use of the Direct mRNA Kit (Qiagen, Hilden, Germany) and reverse transcriptase by Superscript II (Invitrogen, Carlsbad, CA). The resulting cDNA was used directly as a template in PCR with selected gene specific primers. Additionally, isolated mRNA from MDCK II cells was used in the RNA-ligase-mediated and oligo-capping rapid amplification of cDNA ends (RLM-RACE) method performed by the GeneRacer Kit (Invitrogen). When performing PCR with RACE-cDNA as a template, a combination of GeneRacer 5' or 3' primer and a selected gene-specific primer was used (see supplementary data). PCR was performed using Taq polymerase (New England Biolabs, Ipswich, MA). PCR products were sequenced directly or after standard molecular cloning using pGEM-T-easy Vector System 1 (Promega, Madison, WI). All PCR products or clones were sequenced by GATC, Constance, Germany or by the inhouse sequence facility. Primer pair PAPST-GFP (see supplementary data) harboring an XhoI restriction site and a kozak sequence in the 5'-end primer and a deletion of stop-codon and a BamHI restriction site in the 3'-end primer were used to amplify the full-length cDNA. The product was cloned into a pEGFP-N3 vector (Clontech, Mountain View, CA) giving a C-terminal GFP-tagged fusion protein, PAPST1-GFP. Restriction endonucleases and Taq polymerase were all purchased from New England Biolabs.
Cell culture, transfection, and selection of PAPST1-GFP expressing tranfectants
MDCK II cells were grown in DMEM (Cambrex, East Rutherford, NJ), 5% Fetal bovine serum (PAA, Pasching, Austria), 1% L-glutamine, and 1% penicillin/streptomycin (Cambrex) at 37°C and 5% CO2. The cells were passed at confluency, every third or fourth day. The cells were sustained to a maximum of 25 passages. Transfection of MDCK II cells was done at 50–70% confluence by adding 4 µg of pEGFP-N3-PAPST1 vector and 12 µL of FuGENE 6 (Roche Applied Sciences, Basel, Switzerland) under normal growth conditions. After 72 h, the cells were passed and the medium was changed to the selection medium with 1 mg/mL G-418 (Duchefa Biochemie, Haarlem, The Netherlands). Resistant colonies were selected and screened by two criteria. First, GFP fluorescence in the Golgi region of live cells was assessed using confocal imaging (Confocal IX81 Olympus Fluorview FV1000). Golgi membranes were visualized by the Golgi marker BODIPY-TR-C5-ceramide (Invitrogen). Secondly, expression was verified by Western blotting onto PVDF membranes (GE Healthcare, Fairfield, CT) after SDS–PAGE on Criterion-XT gels (Bio-Rad, Hercules, CA) using rabbit anti-GFP (Ab-6556, Abcam, Cambridge, UK) as primary antibody and donkey antirabbit HRP-conjugated IgG (NA943, GE Healthcare) as secondary antibody. Western blots were developed by ECL + (GE Healthcare). Three out of 20 selected transfectants expressed detectable levels of PAPST1-GFP.
Preparation of Golgi vesicles and uptake of PAPS
Subcellular fractionation procedures for MDCK II cells and MDCK II-PAPST1 cells were modified from the method described previously (Fjeldstad et al. 2002). Cells were grown to confluency on 500 cm2 plates (Corning, Corning, NY). The cells were scraped off the plates, washed, and homogenized at 4°C in homogenization buffer (250 mM sucrose, 3 mM imidazole, pH 7.4). After centrifugation at 10 min at 800 x g, 8.5 mL of the post nuclear supernatant (PNS) was mixed with 6.5 mL of 2 M sucrose, 10 mM CsCl, and 1 mM Hepes. A step-gradient of five steps was made: 5 mL of 1.3 M sucrose and 1 mM Hepes, 5 mL of 1.15 M sucrose and 1 mM Hepes, 15 mL of PNS mix, 6 mL 0.9 M sucrose and 1 mM Hepes, and finally 6 mL of homogenization buffer. The gradient was centrifuged in an SW 32 rotor (Beckman Coulter, Fullerton, CA) for 4 h and 20 min at 28,000 rpm using an LK-80 Ultracentrifuge (Beckman Coulter). The enriched Golgi fraction (2–3 mL) at the upper interface was transferred to a new centrifuge tube and diluted to 5 mL with homogenization buffer. Golgi vesicles were pelleted by centrifugation in an LMA-80 rotor (Beckman Coulter) for 1 h at 50,000 rpm using an Optima Ultracentifuge (Beckman Coluter). Uptake of PAPS by Golgi vesicles using 2 µM [35S]PAPS (PerkinElmer, Waltham, MA) and measurement of the protein content of Golgi fractions were performed as described in Fjeldstad et al. (2002).
Radioactive metabolic labeling
MDCK II wild-type and MDCK II-PAPST1 cells were seeded onto 4.7 cm2 polycarbonate filters (Costar 3412, Corning), approximately 1 x 106 cells per filter. After 3–4 days under normal growth conditions the cells were transferred to appropriate labeling media with 2% FCS and radioactive metabolic label added to the basolateral media; 0.3 mCi/mL of [35S]sulfate (PerkinElmer) in RPMI 1640 without sulfate (Invitrogen), 0.3 mCi/mL of [3H]GlcN (PerkinElmer) in RPMI 1640 without glucose (Invitrogen), or 0.3 mCi/mL of [35S]Cys/[35S]Met (PerkinElmer) in DMEM without cysteine or methionine (Sigma, St. Louis, MO). All media were collected after 24 h of labeling and diluted to equal volumes before downstream analysis.
Analysis of labeled macromolecules and proteoglycans
Analysis of macromolecules in apical and basolateral media were done as previously described (Svennevig et al. 1995). Excess radioactive label was removed by gel filtration with Sephadex G-50 Fine (GE Healthcare), and incorporated radioactivity was determined by scintillation counting (TR 1900, Packard) using Ultima Gold XR (PerkinElmer). In some experiments, the composition of GAGs was analyzed. This was done by specific degradation as described earlier (Svennevig et al. 1995); CS by chondroitinase ABC (Seikagaku, Tokyo, Japan) or HS by HNO2. SDS–PAGE of labeled macromolecules was performed with Criterion XT 4–12% gradient gels (Bio-Rad). Thirty microliter of eluted macromolecules was diluted in 4 x XT-sample buffer (Bio-Rad) and 20 x XT reducing agent (Bio-Rad) added. SDS–PAGE was run for 1 h and 10 min at 140 V in XT-MOPS buffer (Bio-Rad). [14C]-labeled protein molecular weight standard was used (GE Healthcare). After fixation, the gels were treated with Amplify (GE Healthcare) and dried. Gels containing [35S] were subjected to phosphorimaging using screens appropriate for the Typhoon 9400 (GE Healthcare). Quantification was done by ImageQuant TL v2003.02 software (GE Healthcare). Gels containing [3H] were subjected to autoradiography using MP hyperfilm (GE Healthcare). The developed films were scanned by a regular scanner, converted to tiff image files, and quantified with ImageQuant.
Immune precipitation of versican and perlecan
Immune precipitation of versican was done using a mouse monoclonal antibody against large human chondroitin sulfate (2-B-1, Seikagaku). This antibody recognizes an epitope in the EGF domain of versican and cross-reacts with canine versican. The antibody was bound to Dynabeads-protein-G (Invitrogen) according to the manufacturers description. One microgram of antibody and 30 µL of bead slurry were used to immune precipitate versican from 500 µL of the medium. After IP overnight at 4°C, the beads were washed three times with IP wash containing 1% BSA (50 mM Tris, pH 7.4, 150 mM NaCl, 0.05% Triton X-100) and three times with IP wash without BSA. Immune precipitation of perlecan was carried out with a rabbit polyclonal serum against human perlecan, kindly provided by R. Iozzo. The IP was performed in the same manner, but with the use of Dynabeads-protein-A (Invitrogen). After IP, the beads were subjected to SDS–PAGE, detection and quantification as described above.
Isolation of GAGs from media and preparation for disaccharide analysis
MDCK II and MDCK II-PAPST1 cells were plated onto Costar filters as described above. After 3–4 days, the filters were rinsed in DMEM, and transferred to fresh DMEM with 1% L-Gln. After 24 h, the apical (1 mL) and basolateral (2 mL) media were collected and subjected to Sephadex G50-Fine (GE Healthcare) gel filtration. Eluates of media from six filters were pooled and dried under vacuum, constituting one sample. Isolation of GAGs was based on the method used by Ledin et al. (2004) and performed as described in Tveit et al. (2005). To prepare disaccharides, a scheme using enzymatic degradation and subsequent filtration was used. For CS the GAG pools were digested with 50 mU of chondroitinase ABC (Seikagaku) in a final volume of 50 µL of 40 mM Tris-acetate buffer, pH 8.0. CS digestion was allowed to proceed for 3 h at 37°C, followed by heat inactivation. Samples were then dried and dissolved in 100 µL H2O. From the CS digests, 10 µL was removed for analysis of the CS disaccharides by reverse phase ion pair high performance liquid chromatography (RPIP-HPLC) (see Analysis of disaccharides). To check for contribution from hyaluronic acid, some samples were divided, and equal amounts were treated either by chondroitinase ABC as above, or by 5 mU hyaluronidase SD (Seikagaku). For HS disaccharide preparation, the remaining 90 µL of the CS digest was subjected to filtration by a 3 kDa cutoff Microcon filter (Millipore, Billerica, MA). The concentrate containing intact HS-GAGs was freeze-dried and dissolved in 20 µL of H2O. Two equal aliquots were dried and prepared for heparitinase digestion. One of the aliquots was treated with 0.4 mU of each of heparitinases I, II, and III (Grampian Enzymes, Orkney, Scotland, UK) in 15 µL of heparitinase buffer (5 mM Hepes, pH 7.0, 50 mM NaCl, 1 mM CaCl2, 0.7 mg/mL BSA) and incubated for 16 h at 37°C. The other aliquot was incubated under the same conditions without enzymes. The reaction was stopped by heat inactivation. All the samples were dried and dissolved in 45 µL of H2O before analysis by the RPIP-HPLC.
Analysis of disaccharides
Quantitative analysis of disaccharides derived from GAG chains and their sulfation patterns was performed by RPIP-HPLC (Staatz et al. 2001) on a Luna 5 µ C18 reversed phase column (4.6 x 150 mm) from Phenomenex, Torrance, CA in acetonitrile (8.5%) and tetra-n-butylammonium hydrogen sulfate (1.2 mM; Fluka, St. Louis, MO) by applying a stepwise gradient of 0.2 M NaCl from 1 to 53%. The flow rate was 1.1 mL/min, and the fluorescent labeling reaction was performed by the addition of 2-cyanoacetamide (0.25%; Sigma) in NaOH (0.5%) at a flow rate of 0.35 mL/min. Signals were quantified against known amounts of standard disaccharides analyzed in parallel runs. HA, CS, and HS disaccharide standards were purchased from Sigma and Grampian Enzymes (Orkney, Scotland, UK). In addition, a hyaluronidase digest of hyaluronic acid (Calbiochem, Merck, Darmstadt, Germany) was run as a standard. The HPLC equipment (pump, autosampler, and fluorescence detector) was purchased from Dionex, Sunnyvale, CA, including the apparatus for online postcolumn delivery of solutions (PC10 postcolumn pneumatic delivery package). The chromatography software used was Chromeleon from Dionex.
Glycosaminoglycan isolation and gel fitration
[35S]sulfate-labeled GAGs from apical and basolateral media were isolated as described for the disaccharide analysis. The resulting GAGs were subjected to Sepharose CL-6B (GE Healthcare) gel filtration to compare GAG chain lengths as described (Vuong et al. 2006).
|
Supplementary Data |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary DataConflict of interest statementReferences |
|---|
Supplementary data for this article is available online at http://glycob.oxfordjournals.org.
|
Conflict of interest statement |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary DataConflict of interest statementReferences |
|---|
None declared.
|
Acknowledgements |
|---|
The presented work was supported by The Research Council of Norway (FUGE), The Norwegian Cancer Society, and The Blix Foundation. Gunnar Dick held a PhD fellowship from the University of Oslo.
|
Abbreviations |
|---|
cABC, chondroitinase ABC; Cys/Met, cystein/methionine; CS, chondroitin sulfate; ER, endoplasmic reticulum; ERGIC, ER to Golgi intermediate compartment; GAG, glycosaminoglycan; GalNAc, N-acetyl-galactosamin; GFP, green fluorescent protein; GlcNAc, N-acetyl-glucosamine; GlcUA, glucuronic acid; HA, hyaluronic acid; HS, heparan sulfate; IP, immune precipitation; KS, keratan sulfate; MDCK, Madin-Darby Canine Kidney; NST, nucleotide sugar transporter; PAPS, 3'-phosphoadenosine 5'-phosphosulfate; PG, proteoglycan; PNS, post nuclear supernatant; RPIP-HPLC, reverse phase ion pair high performance liquid chromatography; SLC35, solute carrier family 35; TGN, trans-Golgi network; UDP-GlcNAc, UDP-N-acetyl-glucosamine; UDP-GalNAc, UDP-N-acetyl-galactosamine; UDP-GlcUA, UDP-glucuronic acid
|
References |
|---|
|
TopAbstractIntroductionResultsDiscussionMaterials and methodsSupplementary DataConflict of interest statementReferences |
|---|
Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ. Gapped BLAST and PSI-BLAST: A new generation of protein database search programs. Nucleic Acids Res (1997) 25:3389–3402.
Ashikov A, Routier F, Fuhlrott J, Helmus Y, Wild M, Gerardy-Schahn R, Bakker A. The human solute carrier gene SLC35B4 encodes a bifunctional nucleotide sugar transporter with specificity for UDP-xylose and UDP-N-acetylglucosamine. J Biol Chem (2005) 280:27230–27235.
Baeuerle PA, Huttner WB. Chlorate—a potent inhibitor of protein sulfation in intact cells. Biochem Biophys Res Commun (1986) 141:870–877.[CrossRef][Web of Science][Medline]
Berninsone PM, Hirschberg CB. Nucleotide sugar transporters of the Golgi apparatus. Curr Opin Struct Biol (2000) 10:542–547.[CrossRef][Web of Science][Medline]
Besset S, Vincourt JB, Amalric F, Girard JP. Nuclear localization of PAPS synthetase 1: A sulfate activation pathway in the nucleus of eukaryotic cells. Faseb J (2000) 14:345–354.
Chenna R, Sugawara H, Koike T, Lopez R, Gibson TJ, Higgins DG, Thompson JD. Multiple sequence alignment with the Clustal series of programs. Nucleic Acids Res (2003) 31:3497–3500.
Deepa SS, Yamada S, Zako M, Goldberger O, Sugahara K. Chondroitin sulfate chains on syndecan-1 and syndecan-4 from normal murine mammary gland epithelial cells are structurally and functionally distinct and cooperate with heparan sulfate chains to bind growth factors. A novel function to control binding of midkine, pleiotrophin, and basic fibroblast growth factor. J Biol Chem (2004) 279:37368–37376.
Fjeldstad K, Pedersen ME, Vuong TT, Kolset SO, Nordstrand LM, Prydz K. Sulfation in the Golgi lumen of madin-darby canine kidney cells is inhibited by brefeldin a and depends on a factor present in the cytoplasm and on Golgi membranes. J Biol Chem (2002) 277:36272–36279.
Girard JP, Baekkevold ES, Amalric F. Sulfation in high endothelial venules: Cloning and expression of the human PAPS synthetase. Faseb J (1998) 12:603–612.
Gotoh M, Sato T, Akashima T, Iwasaki H, Kameyama A, Mochizuki H, Yada T, Inaba N, Zhang Y, Kikuchi N, et al. Enzymatic synthesis of chondroitin with a novel chondroitin sulfate N-acetylgalactosaminyl- transferase that transfers N-acetylgalactosamine to glucuronic acid in initiation and elongation of chondroitin sulfate synthesis. J Biol Chem (2002) 277:38189–38196.
Gotoh M, Yada T, Sato T, Akashima T, Iwasaki H, Mochizuki H, Inaba N, Togayachi A, Kudo T, Watanabe H, et al. Molecular cloning and characterization of a novel chondroitin sulfate glucuronyltransferase that transfers glucuronic acid to N-acetylgalactosamine. J Biol Chem (2002) 277:38179–38188.
Ishida N, Kawakita M. Molecular physiology and pathology of the nucleotide sugar transporter family (SLC35). Pflugers Arch (2004) 447:768–775.[CrossRef][Web of Science][Medline]
Kamiyama S, Sasaki N, Goda E, Ui-Tei K, Saigo K, Narimatsu H, Jigami Y, Kannagi R, Irimura T, Nishihara S. Molecular cloning and characterization of a novel 3'-phosphoadenosine 5'-phosphosulfate transporter, PAPST2. J Biol Chem (2006) 281:10945–10953.
Kamiyama S, Suda T, Ueda R, Suzuki M, Okubo R, Kikuchi N, Chiba Y, Goto S, Toyoda H, Saigo K, et al. Molecular cloning and identification of 3'-phosphoadenosine 5'-phosphosulfate transporter. J Biol Chem (2003) 278:25958–25963.
Keller KM, Brauer PR, Keller JM. Modulation of cell surface heparan sulfate structure by growth of cells in the presence of chlorate. Biochemistry (1989) 28:8100–8107.[CrossRef][Web of Science][Medline]
Kitagawa H, Uyama T, Sugahara K. Molecular cloning and expression of a human chondroitin synthase. J Biol Chem (2001) 276:38721–38726.
Kjellen L, Lindahl U. Proteoglycans: Structures and interactions. Ann Rev Biochem (1991) 60:443–475.[CrossRef][Web of Science][Medline]
Kolset SO, Prydz K, Pejler G. Intracellular proteoglycans. Biochem J (2004) 379:217–227.[CrossRef][Web of Science][Medline]
Kolset SO, Vuong TT, Prydz K. Apical secretion of chondroitin sulphate in polarized Madin-Darby canine kidney (MDCK) cells. J Cell Sci (1999) 112:1797–1801.[Abstract]
Kreuger J, Spillmann D, Li JP, Lindahl U. Interactions between heparan sulfate and proteins: The concept of specificity. J Cell Biol (2006) 174:323–327.
Kusche-Gullberg M, Kjellen L. Sulfotransferases in glycosaminoglycan biosynthesis. Curr Opin Struct Biol (2003) 13:605–611.[CrossRef][Web of Science][Medline]
Lee RW, Suchanek C, Huttner WB. Direct photoaffinity labeling of proteins with adenosine 3'-[32P]phosphate 5'-phosphosulfate. Atractyloside inhibits labeling of a Mr = 34,000 protein in an adrenal medullary Golgi fraction. J Biol Chem (1984) 259:11153–11156.
Ledin J, Staatz W, Li JP, Gotte M, Selleck S, Kjellen L, Spillmann D. Heparan sulfate structure in mice with genetically modified heparan sulfate production. J Biol Chem (2004) 279:42732–42741.
Li H, Deyrup A, Mensch JR Jr, Domowicz M, Konstantinidis AK, Schwartz NB. The isolation and characterization of cDNA encoding themouse bifunctional ATP sulfurylase-adenosine 5'-phosphosulfate kinase. J Biol Chem (1995) 270:29453–29459.
Luders F, Segawa H, Stein D, Selva EM, Perrimon N, Turco SJ, Hacker U. Slalom encodes an adenosine 3'-phosphate 5'-phosphosulfate transporter essential for development in Drosophila. Embo J (2003) 22:3635–3644.[CrossRef][Web of Science][Medline]
Mandon EC, Milla ME, Kempner E, Hirschberg CB. Purification of the Golgi adenosine 3'-phosphate 5'-phosphosulfate transporter, a homodimer within the membrane. Proc Natl Acad Sci (1994) 91:10707–10711.
Ozeran JD, Westley J, Schwartz NB. Identification and partial purification of PAPS translocase. Biochemistry (1996) 35:3695–3703.[CrossRef][Web of Science][Medline]
Ozeran JD, Westley J, Schwartz NB. Kinetics of PAPS translocase: Evidence for an antiport mechanism. Biochemistry (1996) 35:3685–3694.[CrossRef][Web of Science][Medline]
Prydz K, Dalen KT. Synthesis and sorting of proteoglycans. J Cell Sci (2000) 113:193–205.[Abstract]
Rozen S, Skaletsky H. Primer3 on the WWW for general users and for biologist programmers. Methods Mol Biol (2000) 132:365–386.[Medline]
Safaiyan F, Kolset SO, Prydz K, Gottfridsson E, Lindahl U, Salmivirta M. Selective effects of sodium chlorate treatment on the sulfation of heparan sulfate. J Biol Chem (1999) 274:36267–36273.
Sato T, Gotoh M, Kiyohara K, Akashima T, Iwasaki H, Kameyama A, Mochizuki H, Yada T, Inaba N, Togayachi A, et al. Differential roles of two N-acetylgalactosaminyltransferases, CSGalNAcT-1, and a novel enzyme, CSGalNAcT-2. Initiation and elongation in synthesis of chondroitin sulfate. J Biol Chem (2003) 278:3063–3071.
Schepers MS, Asselman M, Duim RA, Romijn JC, Schroder FH, Verkoelen CF. Pericellular matrix formation by renal tubule epithelial cells in relation to crystal binding. Nephron Exp Nephrol (2003) 94:e103–e112.[CrossRef][Medline]
Schwarz JK, Capasso JM, Hirschberg CB. Translocation of adenosine 3'-phosphate 5'-phosphosulfate into rat liver Golgi vesicles. J Biol Chem (1984) 259:3554–3559.
Selva EM, Hong K, Baeg GH, Beverly SM, Turco SJ, Perrimon N, Hacker U. Dual role of the fringe connection gene in both heparan sulphate and fringe-dependent signalling events. Nat Cell Biol (2001) 3:809–815.[CrossRef][Web of Science][Medline]
Spiro RC, Freeze HH, Sampath D, Garcia JA. Uncoupling of chondroitin sulfate glycosaminoglycan synthesis by brefeldin A. J Cell Biol (1991) 115:1463–1473.
Staatz WD, Toyoda H, Kinoshita-Toyoda A, Chhor K, Selleck SB. Analysis of proteoglycans and glycosaminoglycans from Drosophila. Methods Mol Biol (2001) 171:41–52.[Medline]
Suda T, Kamiyama S, Suzuki M, Kikuchi N, Nakayama K, Narimatsu H, Jigami Y, Aoki T, Nishihara S. Molecular cloning and characterization of a human multisubstrate specific nucleotide-sugar transporter homologous to Drosophila fringe connection. J Biol Chem (2004) 279:26469–26474.
Svennevig K, Prydz K, Kolset SO. Proteoglycans in polarized epithelial Madin-Darby canine kidney cells. Biochem J (1995) 311:881–888.[Web of Science][Medline]
Toma L, Pinhal MA, Dietrich CP, Nader HB, Hirschberg CB. Transport of UDP-galactose into the Golgi lumen regulates the biosynthesis of proteoglycans. J Biol Chem (1996) 271:3897–3901.
Tveit H, Dick G, Skibeli V, Prydz K. A proteoglycan undergoes different modifications en route to the apical and basolateral surfaces of Madin-Darby canine kidney cells. J Biol Chem (2005) 280:29596–29603.
Uyama T, Kitagawa H, Tanaka J, Tamura J, Ogawa T, Sugahara K. Molecular cloning and expression of a second chondroitin N-acetylgalactosaminyltransferase involved in the initiation and elongation of chondroitin/dermatan sulfate. J Biol Chem (2003) 278:3072–3078.
Vuong TT, Prydz K, Tveit H. Differences in the apical and basolateral pathways for glycosaminoglycan biosynthesis in Madin-Darby canine kidney cells. Glycobiology (2006) 16:326–332.
Yada T, Gotoh M, Sato M, Shionyu M, Go M, Kaseyama H, Iwasaki H, Kikuchi N, Kwon YD, Togayachi A, et al. Chondroitin sulfate synthase-2 Molecular cloning and characterization of a novel human glycosyltransferase homologous to chondroitin sulfate glucuronyltransferase, which has dual enzymatic activities. J Biol Chem (2003) 278:30235–30247.
Yada T, Sato M, Kaseyama H, Gotoh M, Iwasaki H, Kikuchi N, Kwon YD, Togayachi A, Kudo T, Watanabe H, et al. Chondroitin sulfate synthase-3. Molecular cloning and characterization. J Biol Chem (2003) 278:39711–39725.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||