Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M701200200 on May 31, 2007

J. Biol. Chem., Vol. 282, Issue 30, 21934-21944, July 27, 2007
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
282/30/21934    most recent
M701200200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mani, K.
Right arrow Articles by Fransson, L.-A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mani, K.
Right arrow Articles by Fransson, L.-A.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Heparan Sulfate Degradation Products Can Associate with Oxidized Proteins and Proteasomes*

Katrin Mani1, Fang Cheng, and Lars-Åke Fransson2

From the Department of Experimental Medical Science, Section of Neuroscience, Lund University, Biomedical Centre A13, SE-221 84, Lund, Sweden

Received for publication, February 8, 2007 , and in revised form, May 10, 2007.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
The S-nitrosylated proteoglycan glypican-1 recycles via endosomes where its heparan sulfate chains are degraded into anhydromannose-containing saccharides by NO-catalyzed deaminative cleavage. Because heparan sulfate chains can be associated with intracellular protein aggregates, glypican-1 autoprocessing may be involved in the clearance of misfolded recycling proteins. Here we have arrested and then reactivated NO-catalyzed cleavage in the absence or presence of proteasome inhibitors and analyzed the products present in endosomes or co-precipitating with proteasomes using metabolic radiolabeling and immunomagnet isolation as well as by confocal immunofluorescence microscopy. Upon reactivation of deaminative cleavage in T24 carcinoma cells, [35S]sulfate-labeled degradation products appeared in Rab7-positive vesicles and co-precipitated with a 20 S proteasome subunit. Simultaneous inhibition of proteasome activity resulted in a sustained accumulation of degradation products. We also demonstrated that the anhydromannose-containing heparan sulfate degradation products are detected by a hydrazide-based method that also identifies oxidized, i.e. carbonylated, proteins that are normally degraded in proteasomes. Upon inhibition of proteasome activity, pronounced colocalization between carbonyl-staining, anhydro-mannose-containing degradation products, and proteasomes was observed in both T24 carcinoma and N2a neuroblastoma cells. The deaminatively generated products that co-precipitated with the proteasomal subunit contained heparan sulfate but were larger than heparan sulfate oligosaccharides and resistant to both acid and alkali. However, proteolytic degradation released heparan sulfate oligosaccharides. In Niemann-Pick C-1 fibroblasts, where deaminative degradation of heparan sulfate is defective, carbonylated proteins were abundant. Moreover, when glypican-1 expression was silenced in normal fibroblasts, the level of carbonylated proteins increased raising the possibility that deaminative heparan sulfate degradation is involved in the clearance of misfolded proteins.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Cell-associated heparan sulfate (HS)3-containing proteoglycans (PG) regulate a great diversity of biological processes involved in development, tissue repair, and tumorigenesis (1, 2). HS chains are also found both in intracellular tangles and in extracellular amyloid deposits characteristic of Alzheimer disease (3, 4). The two major cell-associated HS proteoglycan families are the syndecans and the glypicans. The syndecan core proteins are all transmembrane proteins, whereas the glypicans are attached to membrane lipids via a C-terminal glycosylphosphatidylinositol anchor (Fig. 1). The ectodomains, which carry the HS chains, vary greatly in size within the syndecan family but display considerable homology among the glypicans. The latter also share a characteristic pattern of 14 conserved Cys residues, and their HS attachment sites are concentrated to a short region near the C terminus.

Cell surface HS proteoglycans are constitutively endocytosed and degraded. Their HS chains can be cleaved into oligosaccharides by heparanase and terminally degraded by a set of exoglycosidases and sulfatases in lysosomes (5). Glypican-1 (Gpc-1), the most widely expressed member of the glypican family, is lipid raft-associated (6) and probably endocytosed via caveolae (see Fig. 1). In addition to heparanase-catalyzed degradation, Gpc-1 can undergo nitric oxide (NO)- or nitroxyl (HNO)-catalyzed deaminative cleavage of its HS chains at N-unsubstituted glucosamine residues (GlcNH+3) (7–9). The deaminative cleavage generates anhydromannose (anMan) which remains as the reducing terminal sugar of the released HS chains or oligosaccharides (10, 11). Endogenously generated anMan-positive degradation products have been detected both in cell cultures and in vivo by confocal immunofluorescence microscopy using a specific mAb (8, 9, 12, 13).

The NO/HNO required for deaminative cleavage of HS seems to be derived from preformed S-nitroso (SNO) groups in the Gpc-1 core protein. In the presence of copper ions and an NO donor, purified Gpc-1 can be S-nitrosylated in vitro (8). In vivo, copper ions can be provided by cuproproteins, such as the prion protein in fibroblasts or N2a neuroblastoma cells (14, 15), by the brain-specific glycosylphosphatidylinositol-linked splice variant of ceruloplasmin in glial cells (16), or by the Alzheimer amyloid precursor protein in neural cells (13). Accordingly, Gpc-1 is not S-nitrosylated in prion null fibroblasts, but S-nitrosylation is restored upon ectopic expression of the prion protein (14). No anMan-positive products are generated when prion protein lacking the copper binding domain is expressed in these cells (15). In T24 carcinoma cells, Gpc-1-SNO colocalizes with caveolin-1 (9), indicating that S-nitrosylation takes place at an early stage of recycling (see Fig. 1).

Gpc-1-SNO autodegrades its own HS chains to anMan-containing oligosaccharides when exposed to ascorbate in vitro (8). Studies on cultured cells have shown that endogenously generated anMan-positive products colocalize primarily with Rab7, a marker for late endosomes (17, 18). In Niemann-Pick C1 (NPC-1) fibroblasts, which display an endosomal transport block, staining by the anMan-specific mAb was very weak (17). However, formation of anMan-positive products that colocalized with Rab7 could be enhanced by exogenously supplied ascorbate. When N2a neuroblastoma cells were treated with 3-beta[2(diethylamino)ethoxy]androst-5-en-17-one (U18666A), a compound that mimics the NPC phenotype, formation of anMan-positive products was depressed. Also in these cells, ascorbate restored formation of anMan-positive products that colocalized with Rab7. Inhibition of endosomal acidification in T24 cells, which blocks transfer from early (Rab5) to late (Rab7) endosomes, abrogated generation of anMan-positive products (18). This could also be overcome by simultaneous addition of ascorbate, which induced formation of anMan-positive products that colocalized with Rab7.

[35S]Sulfate-labeled anMan-positive products generated constitutively in T24 cells comprise free HS chains and relatively large HS oligosaccharides (18). The latter are especially prominent in Rab7-positive vesicles. Moreover, anMan-positive HS degradation products have been detected both in the cytosol (18) and in the nucleus (19), suggesting that deaminatively generated HS degradation products may exit from the endosomes (see Fig. 1).

Because HS can be associated with intracellular aggregates derived from misfolded proteins (3), we speculated that HS and its oligosaccharide degradation products might be involved in the degradation and/or clearance of misfolded recycling proteins. It has been shown that a proteasome {alpha}-subunit interacts specifically with Rab7 and thereby recruits 20 S proteasomes to multivesicular late endosomes (20). We, therefore, decided to examine if anMan-containing HS oligosaccharides can associate with misfolded proteins and proteasomes. By inhibiting and then activating deaminative cleavage in the presence of proteasome inhibitors and analyzing HS degradation products co-isolating with Rab7-positive endosomes or colocalizing/co-precipitating with oxidized proteins or proteasomes, we have obtained results indicating that the anMan-containing HS degradation products can associate with proteins that are normally degraded in proteasomes.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Cells, Antibodies, and Reagents—Human T24 carcinoma cells, HFL-1 fibroblasts, and mouse N2a neuroblastoma cells were obtained from ATCC and human NPC-1 fibroblasts (GM03123) from the Corielle Institute. Cells were maintained in minimal essential medium supplemented with 10% fetal calf serum. Polyclonal antisera against human Gpc-1, Rab7, and Rab9 and mAbs against Rab9 and anMan-terminating HS oligosaccharides suitably tagged secondary antibodies as well as suramin, difluoromethylornithine (DFMO), U18666A, HS, N-desulfated HS, sodium L-ascorbate, enzymes, prepacked columns, and other chemicals were generated or obtained as described previously (7–9, 17, 18). A polyclonal antibody against Rab5 and a mAb against ubiquitin (Ub) were from Santa Cruz and a polyclonal antibody against the human 20 S proteasome {alpha}1 subunit was from Calbiochem. Immune-Star horse-radish peroxidase chemiluminescent kit was purchased from Bio-Rad, and the OxyBlot protein oxidation detection kit was from Chemicon. The commercially obtained antibodies were tested by Western blotting according to the manufacturers. Proteinase K was from Sigma-Aldrich, and a proteasome inhibitor set I, including two aldehyde-terminating hydrophobic peptides, and lactacystin were obtained from Calbiochem. Polyacrylamide gels were from Invitrogen, Sweden, and protein G-Sepharose 4B was from Sigma-Aldrich. Dynabeads M-280 sheep anti-rabbit IgG was obtained from Dynal ASA, Oslo, Norway. Dehydroascorbic acid was generated by leaving a 1 M stock solution of ascorbate in contact with air for 24 h.

Confocal Microscopy—The various procedures including seeding of cells, fixations, the use of primary and secondary antibodies, generation of images by sequential scans, and data processing were the same as those used previously (8, 9, 17, 18). The second antibody used was either goat anti-mouse total Ig when the primary antibody was a mouse monoclonal or goat anti-rabbit IgG when the primary antibody was a rabbit polyclonal. The second antibodies were tagged with either fluorescein isothiocyanate or Texas Red and appropriately combined for colocalization studies. For detection of carbonyls, the method of Hernebring et al. (21) was used. In the controls, the primary antibody was omitted. For confocal microscopy we used a Nikon Eclipse E800 microscope equipped with a 100x objective and a Bio-Rad MRC 1024 confocal laser scanning system. Images shown were obtained at a focal plane that was at the center of the cell and of 0.3–0.5-µm thickness, depending on the intensity of the staining. Identical focal plane thickness and exposure settings were used for image capture in the colocalization experiments. Images were digitized and transferred to Adobe PhotoShop for merging, annotation, and printing.

Immunoisolation and Separation of Radiolabeled Degradation Products—Confluent T24 cells were preincubated with 5 mM DFMO for 24 h and then with 0.2 mM suramin and 50 µCi/ml [35S]sulfate for another 24 h as described previously (7). Cells were homogenized in phosphate-buffered saline containing 0.25 M sucrose, 0.1 M EDTA, 0.5 mM phenylmethylsulfonyl fluoride. Rab5-, Rab-7, or Rab-9 positive vesicles as well as 20 S proteasome {alpha}1 subunit-containing material were isolated from cell homogenates by using polyclonal antibodies against the various proteins as primary antibodies and Dynabeads M-280 sheep anti-rabbit IgG as magnetic secondary antibody. The magnetic particles were recovered using a magnetic particle concentrator and extensively washed with phosphate-buffered saline (at least 10 times). The radiolabeled, magnetically isolated vesicles were lysed in 0.15 M NaCl, 10 mM EDTA, 2% Triton X-100, 10 mM KH2PO4, pH 7.5, and mixed with an equal volume of 8 M guanidinium chloride. Lysates were then subjected to gel filtration chromatography on Superose 6 or Superdex peptide in 4 M guanidinium chloride/0.2% (v/v) Triton X-100 as described previously (7–9). 20 S proteasome immuno-isolates were directly dissolved in elution buffer and chromatographed.

For immunoisolation of radiolabeled, carbonyl-containing proteins, an adaptation of the method of Hernebring et al. (21) was used. In short, confluent NPC-1 fibroblasts were incubated with 50 µCi/ml [35S]sulfate or [35S]Met/Cys for 24 h. A sucrose homogenate (1.1 ml) was treated with 150 µl of 2,4-dinitrophenyl (DNP)-hydrazine (10x) for 15 min. After the addition of 115 µl of neutralization solution, the sample was treated with 150 µl of anti-DNP antibody (1:62) at 4 °C overnight. The immune complex was then recovered using a magnetic secondary antibody, dissolved in elution buffer, and chromatographed on Superose 6.

Co-immunoprecipitation, SDS-PAGE, and Western Blotting—Confluent T24 cells (0.5 x 106 cells) were preincubated with 5 mM DFMO for 24 h and then with 0.2 mM suramin and 10 µg/ml U18666A for another 24 and 16 h, respectively. Deaminative HS cleavage was reactivated by 1 mM ascorbate in the presence of proteasome inhibitors (25 µM MG-132, 25 µM proteasome inhibitor-I, and 1.5 µM lactacystin) for 4 h. In the case of N2a neuroblastoma cells, only inhibition of proteasome activity was performed.

Sucrose homogenates of the cells containing ~2 mg of protein were treated with DNP hydrazine as described above and then first treated with protein G-Sepharose 4B and then with the mAb directed against anMan-containing HS degradation products or the anti-20 S proteasome {alpha}1 subunit antibody or no antibody in the presence of added Triton X-100 (1%, v/v). Immune complexes were recovered on protein G-Sepharose 4B. The gels were washed 6 times with 0.15 M NaCl, 10 mM Tris, pH 7.4, containing 0.2% (v/v) Tween 20. Bound material was released by boiling in SDS buffer and subjected to SDS-PAGE on 4–12% gels under reducing conditions followed by transfer to blotting membrane as described (7). Membranes were blotted first with anti-DNP (1:150) then stripped in 100 mM 2-mecaptoethanol, 2% (w/v) SDS, 62.5 mM Tris-HCl, pH 6.7, at 50 °C for 30 min and re-blotted with anti-Ub (1:200) or in reverse order. After the washings, membranes were probed with goat anti-mouse horseradish peroxidase-conjugated antibody (1:20,000) for 1 h at 20 °C and developed using a Fujifilm ECL detector. Band intensity was estimated by densitometric tracing.

Silencing of Gpc-1 Expression and Detection of Carbonylated Proteins—Confluent HFL-1 fibroblasts (0.5 x 106 cells) were transfected with a vector containing a sequence corresponding to nucleotides 977–995 in human Gpc-1 or a scrambled sequence as described previously (22). The extent of silencing was assessed by SDS-PAGE and Western blotting using anti-Gpc-1 (1:1,000) and anti-rabbit horseradish peroxidase-conjugated antibody.

A sucrose homogenate of the cells containing ~2 mg of protein was treated with DNP hydrazine as described above and then first treated with protein G-Sepharose 4B and then with anti-DNP antibody in presence of added Triton X-100 (1%, v/v). Immune complexes were recovered by magnetic secondary antibodies as described above. Bound material was released by boiling in SDS buffer and subjected to SDS-PAGE on 4–12% gels under reducing conditions followed by staining for protein by Coomassie colloidal blue (Invitrogen). Band intensity was estimated by densitometric tracing.

Degradations—Samples were digested with 100 µg/ml proteinase K in 50 mM Tris-HCl, 5 mM CaCl2, pH 7.5, at 37 °C for 8 h or with HS lyase (alias heparitinase or heparinase III) in the presence of proteinase inhibitors or treated with alkali (0.5 M NaOH, 0.05 M NaBH4 at room temperature overnight), acid (0.1 M HCl at room temperature for 4 h), or HNO2 at pH 1.5 or 3.9, all as described previously (7–9, 17, 18).

Dot Blot Assay—N-Desulfated HS (3 µg), either untreated or treated with HNO2 at pH 3.9, was blotted onto immunoblot membranes (polyvinylidene difluoride), and carbonyl staining was performed as described (21) except that goat anti-rabbit horseradish peroxidase-conjugated antibody was used as secondary antibody and visualized by photoimaging.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Deaminatively Generated HS Degradation Products Accumulate When Proteasome Activity Is Inhibited—Previous studies have shown that treatment of T24 carcinoma cells with DFMO increases the number of GlcNH+3 residues, i.e. NO-sensitive sites, in HS of Gpc-1 (23). Hence, pretreatment with DFMO is a way to increase the yield of anMan-containing HS chains and oligosaccharides upon deaminative cleavage. Furthermore, when heparanase is inhibited by suramin, more HS is available for NO-dependent cleavage (Fig. 1).

Hence, to maximize the yield and to investigate the nature of the deaminative HS degradation products present in endosomes of T24 cells, DFMO- and suramin-treated cells were incubated with [35S]sulfate, and Rab5-, Rab7-, and Rab9-positive vesicles were immunomagnet-isolated from cell homogenates as described previously (18). The vesicles were lysed, and their contents were chromatographed on Superose 6 under dissociative conditions (4 M guanidinium chloride, 0.2% Triton X-100 at pH 5.8). As shown previously (18), a relatively homogenous pool of large [35S]sulfate-labeled oligosaccharides was present in the Rab5- and Rab7-positive vesicles but absent from the Rab9-positive vesicles. The major part of these oligosaccharides were in the Rab7-positive vesicles (Fig. 2A).

Previous studies have shown that staining for anMan-positive HS degradation products is depressed in NO-depleted T24 cells (9). This is an indirect effect caused by depletion of Gpc-1-SNO and diminished capacity for deaminative autocleavage of HS. A more direct effect is obtained with U18666A, which inhibits NO release from Gpc-1-SNO and thereby reduces formation of anMan-containing HS oligosaccharides (18). Accordingly, when DFMO- and suramin-pretreated and [35S]sulfate-labeled T24 cells were exposed to U18666A, the total yield of intracellular [35S]sulfate-labeled oligosaccharides was reduced, especially in the Rab7-positive vesicles (Fig. 2B, solid line). As shown previously (17, 18), when U18666A-treated T24 cells are subsequently exposed to ascorbate, anMan-positive staining increases. Simultaneously, a heterogeneous population of [35S]sulfate-labeled degradation products was recovered from the Rab7-positive vesicles (Fig. 2, B, open symbols, and C, solid line). However, the amounts obtained gradually declined with time (Fig. 2C, dashed and dotted lines).


Figure 1
View larger version (41K):
[in this window]
[in a new window]

  		FIGURE 1.
Proposed traffic route for recycling Gpc-1. Lipid raft-associated Gpc-1 is depicted (top left) as a globular protein with Cys thiols (SH), three heparan sulfate chains (long black bars) in which the glucosamine residues are sometimes N-unsubstituted (GlcNH+3, star), and a glycosylphosphatidylinositol-anchor (black oval with two small bars) plugged into a lipid raft (dark rectangle). Endocytosis (top right), probably via caveolae, is proposed to transport Gpc-1 via Rab5-positive early endosomes to Rab7-positive late endosomes and finally via Rab9-positive transporting endosomes back to the Golgi (18). At the beginning of this traffic route, Cys residues in the core protein are S-nitrosylated (SNO) in a redox reaction that requires nitric oxide (NO) and a cuproprotein in which Cu(II) is reduced to Cu(I). The HS chains of Gpc-1-SNO can be degraded either by heparanase (an endoglucuronidase) or by an NO-catalyzed deaminative cleavage. Degradation by heparanase before deaminative cleavage will release HS oligosaccharides containing the GlcNH+3 residues (bottom right). This precludes NO-catalyzed degradation. Deaminative cleavage is initiated by release of NO/HNO from Gpc-1-SNO in endosomes. This results in cleavage of the HS chains at the GlcNH+3 residues generating anMan-containing HS oligosaccharides. These oligosaccharides accumulate in late endosomes, and some of them may even exit into the cytosol (bottom left). The Rab9-positive endosomes transport the truncated Gpc-1 back to the Golgi, where new HS chains can be extended on the remaining stubs (black arrowheads).

 
Proteasomes can bind to Rab7-positive late endosomes, especially in the presence of proteasome inhibitors (20). To examine if contact between endosomes and proteasomes affected deaminative HS cleavage, we reactivated this process in the presence of proteasome inhibitors. The content of Rab7-positive vesicles obtained from [35S]sulfate-labeled T24 cells treated with DFMO-, suramin-, U18666A-, and ascorbate in the presence of proteasome inhibitors was analyzed by gel chromatography on Superose 6. When proteasome activity was inhibited during a 4-h ascorbate treatment, there was a 2–3-fold increase in radiolabeled products (cf. Fig. 2, C, dashed line with 2D, solid line; note also the different scales). The products eluted from Superose 6 like free HS chains (position 40) and oligosaccharides (position 50).

We also analyzed [35S]sulfate-labeled material that co-precipitated with the 20 S proteasome {alpha}1 subunit. When cells were treated with only DFMO, suramin, and U18666A, the amount of radioactive material co-precipitating with the proteasomal marker was negligible. In contrast, when cells were subsequently treated with ascorbate for 1 h, a heterogeneous population of radiolabeled products was present in the 20 S proteasome immuno-isolates (Fig. 2E, solid line). After 4 h of ascorbate treatment, the products in the 20 S proteasome immuno-isolates had almost disappeared (Fig. 2E, dashed line). However, when proteasome activity was inhibited during the 4-h ascorbate treatment, there was a more than 10-fold increase in [35S]sulfate-labeled products that co-precipitated with the 20 S proteasome subunit (cf. Fig. 2, E, dashed line, with F, solid line). Overall, the yield of radiolabeled products was higher in the Rab7-positive vesicles than in the 20 S proteasome subunit immuno-isolate after a 1-h ascorbate treatment but much higher in the 20 S proteasome immuno-isolate than in the Rab7 vesicles after a 4-h treatment with both ascorbate and proteasome inhibitors.

We then performed a wash-out chase of the radiolabeled DFMO-, suramin-, U18666A-, ascorbate-, and proteasome inhibitor-treated cells with fresh medium without ascorbate and proteasome inhibitors and analyzed radioactive products obtained from the Rab7-positive vesicles and from the 20 S proteasome subunit immuno-isolate. Surprisingly, in the Rab7-positive vesicles, the amount of [35S]sulfate-labeled products increased markedly. Especially material eluting like HS chains increased 4–5-fold during the first hour of chase (Fig. 2D, solid and dashed lines). Also, higher molecular size material accumulated, and after 3 h of chase lower molecular size products also began to appear (Fig. 2D, dashed and dotted lines). Simultaneously, the amount of [35S]sulfate-labeled products co-precipitating with the 20 S proteasome subunit decreased during the first hour of chase (Fig. 2F, solid and dashed lines).

In summary, the above results showed that, upon reactivation of deaminative HS cleavage in U18666A-treated T24 cells, [35S]sulfate-labeled products appeared in Rab7-positive vesicles and co-precipitated with a 20 S proteasome subunit. Simultaneous inhibition of proteasome activity resulted in a sustained accumulation of degradation products in the Rab7-positive vesicles and even more so in the 20 S proteasome subunit immuno-isolate. Upon removal of the proteasome inhibitors, the degradation products accumulated to even greater levels in the Rab7-positive vesicles, whereas they diminished in the 20 S proteasome subunit immuno-isolate.

Associations between Carbonyls, anMan-containing HS Oligosaccharides, and the 20 S Proteasome {alpha}1 Subunit in T24 Carcinoma Cells—The [35S]sulfate-labeled products present in the Rab7-positive vesicles and co-precipitating with the 20 S proteasome {alpha}1 subunit (Fig. 2) may comprise both free and protein-bound glycosaminoglycan degradation products. Deaminatively generated HS oligosaccharides contain a reducing terminal anMan residue with a reactive aldehyde, i.e. a carbonyl (Fig. 3A). Because these oligosaccharides are formed in a reductive environment (18), it cannot be excluded that the terminal anMan is reduced to anhydromannitol (anManOH), which lacks a carbonyl (Fig. 3A), although it can still be recognized by the specific mAb used in this study (12). Alternatively, the aldehyde of unreduced HS oligosaccharides reacts with amino groups in proteins forming an unstable Schiff base (Fig. 3B). By subsequent reduction or Amadori rearrangement, the linkage between oligosaccharide and protein can be stabilized (Fig. 3C).

Oxidation and carbonylation of proteins (Fig. 3) are associated with misfolding, and furthermore, oxidized protein aggregates are poor substrates for proteolysis in proteasomes (24). Hernebring et al. (21) have shown that undifferentiated embryonic stem cells contain carbonylated proteins that are eliminated by proteasomal degradation upon differentiation. To detect carbonyls they used a method that is based on the reaction of DNP hydrazine with carbonyl groups forming DNP-hydrazone that can be immunochemically detected by confocal microscopy. This method may be expected to also detect the carbonyl (aldehyde) in anMan-containing HS oligosaccharides. We tested this by subjecting N-desulfated HS to DNP hydrazine after deaminative cleavage by HNO2 at pH 3.9, which should generate anMan-containing degradation products. As shown in Fig. 3D, the anMan-containing products (II) reacted with DNP hydrazine and were detected using the method of Hernebring et al. (21). We, therefore, examined T24 cells for colocalization between carbonyl- and anMan-staining.


Figure 2
View larger version (18K):
[in this window]
[in a new window]

  		FIGURE 2.
Analysis of 35S-labeled products deaminatively generated in T24 carcinoma cells and present in Rab7-positive endosomes or co-precipitated with the 20 S proteasome {alpha}1 subunit. Cells were grown to confluence in 25-cm2 wells and radiolabeled with 50 µCi/ml [35S]sulfate for 24 h. The panels show results of gel chromatography on Superose 6 of products present in lysates of Rab7-positive vesicles (A–D) and in 20 S proteasome subunit {alpha}1 immunoisolates obtained from DFMO- and suramin-treated cells (E–F) without further additions (A), with the addition of U18666A (B) and with ({circ}) or without (——) subsequent treatment with 1 mM ascorbate for 1 h and from DFMO-, suramin-, and U18666A-treated cells after further treatment with 1 mM ascorbate in the absence (C–E) or presence (D–F) of proteasome inhibitors (PI). In C and E cells were treated with ascorbate for 1 h (——), 4 h (------), or 8 h (......). In D and F cells were treated with ascorbate and proteasome inhibitors for 4 h (——) and then chased in fresh medium without ascorbate or proteasome inhibitors for 1 h (------) or 3 h (......). Homogenates of the cells were subjected to immunomagnet isolation of Rab7- or 20 S{alpha}1 subunit-positive immuno-isolates, respectively, as described under "Experimental Procedures" and chromatographed. The entire content of each tube was subjected to radioactivity determination. Vo, void volume; Vt, total volume; PG, elution position of Gpc-1 proteoglycan; HS, elution position of HS chains. Cells were pretreated with 5 mM DFMO and 0.2 mM suramin for 24 h. Treatment with 10 µg/ml U18666A was for 16 h. Inhibition of proteasome activity was performed with 25 µM MG-132, 25 µM proteasome inhibitor-I, and 1.5 µM lactacystin.

 
In DFMO-, suramin-, and U18666A-treated T24 cells, where both heparanase- and NO-dependent cleavages of HS were thus suppressed, carbonyl groups were still detected (Fig. 4A, red). Because anMan staining was undetectable (Fig. 4A, no green or yellow), the carbonyl staining may be due to oxidations in the side groups of amino acids in proteins (21). However, it is also possible that the DNP-staining is more sensitive than the anMan-staining in detecting HS oligosaccharides.

When deaminative HS degradation was reactivated by subsequent exposure to ascorbate and anMan-positive products were generated, there was perinuclear colocalization between carbonyl- and anMan-containing products (Fig. 4B, yellow arrow). There was also some anMan reactivity that did not colocalize with the carbonyl staining (Fig. 4B, green arrow), suggesting that in some HS oligosaccharides anMan was converted to anManOH.

When reactivation of deaminative HS degradation was performed in the presence of proteasome inhibitors, carbonyl reactivity increased, and some cells showed massive accumulation of products that were both carbonyl- and anMan-positive (Fig. 4C, yellow arrow in inset). The observed colocalization between anMan- and carbonyl-staining may thus partly be due to accumulation of anMan-containing HS oligosaccharides that reacted both with DNP hydrazine and with the mAb (Fig. 3A). Alternatively or in addition, there could be a colocalization between anManOH-containing HS oligosaccharides and oxidized proteins (Fig. 3A), or the colocalization could be due to formation of unstable (Fig. 3B) or stable HS-protein conjugates (Fig. 3C).

In parallel, we examined possible colocalizations between anMan reactivity and proteasomes using the polyclonal antibody to the 20 S proteasome {alpha}1 subunit (Fig. 4, D–F). When deaminative degradation was reactivated by ascorbate, there was colocalization between the proteasomal marker and anMan-staining in the same area (Fig. 4E, yellow arrow) where the carbonyl- and anMan-staining colocalized (Fig. 4B, yellow arrow). Upon simultaneous inhibition of proteasome activity, the proteasome-anMan colocalization was enhanced (Fig. 4F, yellow arrows), but some of the HS degradation products still appeared to be carbonyl-negative (Fig. 4F, green arrow). In summary, upon inhibition of proteasome activity a pronounced colocalization between carbonyl-staining, anMan-containing HS degradation products, and proteasomes was observed, suggesting enhanced association between HS, oxidized proteins. and proteasomes.


Figure 3
View larger version (36K):
[in this window]
[in a new window]

  		FIGURE 3.
Possible reactions between an anMan-terminating HS oligosaccharide and an amino group in a protein (A–C) and dot-blot carbonyl staining of anMan (D). A, the terminal anMan in a deaminatively generated HS oligosaccharide contains a free aldehyde group (CHO, circled), which is a functional carbonyl. This aldehyde can either be reduced to CH2OH, forming anManOH, or react with amino groups in proteins forming a Schiff base (B). This is a reversible reaction, sensitive to neighboring acidic or basic groups. The unstable bond between HS and protein could be irreversibly stabilized by reduction or via Amadori rearrangement to a HS-protein conjugate containing another carbonyl group (circled; C). This is a form of neo-glycation, such as that formed between open-chain forms of sugars and proteins. This process is usually slow due to low abundance of open-ring sugar structures. However, anMan has a free aldehyde and could, thus, react more rapidly. Oxidized proteins may also contain carbonyls as indicated. Hence, DNP hydrazide, which reacts with carbonyls, may detect anMan-terminating HS oligosaccharides (A), oxidized proteins unconjugated or conjugated with HS (A–C), and the carbonyl of a stable HS-protein conjugate (C). anManOH-containing oligosaccharides should not react with DNP hydrazide. The mAb that recognizes deaminatively generated HS chains and oligosaccharides reacts both with anMan- and anManOH-terminating sequences, indicating that the C-1 carbonyl is not a critical subsite in the epitope (12). Therefore, the mAb is expected to detect anMan in B. Apparently, the mAb also recognizes anMan-containing oligosaccharides that have been derivatized with DNP. D, dot-blot carbonyl staining of N-desulfated HS (I) and N-desulfated HS after deaminative cleavage by HNO2 at pH 3.9 (II).

 
Identification of HS Oligosaccharide-Protein Conjugates—The [35S]sulfate-labeled degradation products isolated from Rab7-positive vesicles of DFMO- and suramin-treated T24 cells comprised a relatively homogenous oligosaccharide population that was well retarded on Superose 6 (Fig. 2A). However, when deaminative cleavage was reactivated after inhibition by U18666A (Fig. 2B), most of the [35S]sulfate-labeled degradation products isolated from the Rab7-positive vesicles (Fig. 2C) as well as those co-precipitating with the 20 S proteasome {alpha}1 subunit (Fig. 2, E and F) were heterogeneous and appeared to be larger than an oligosaccharide (Fig. 2A). Because the Superose 6 chromatography was performed under dissociative conditions, the products could consist of HS oligosaccharide-protein conjugates (Fig. 3).


Figure 4
View larger version (107K):
[in this window]
[in a new window]

  		FIGURE 4.
Colocalization of carbonyls, anMan-positive HS degradation products, and the 20 S proteasome {alpha}1 subunit in T24 carcinoma cells. Cells were treated with DFMO, suramin (Sur), U18666A, ascorbate (Asc), and proteasome inhibitors (PI) as indicated beside the panels. The merged confocal laser immunofluorescence microscopy images show cells stained for carbonyls (>CO, red), anMan-containing HS oligosaccharides (anMan, green), and the 20 S proteasome {alpha}1 subunit (20S{alpha}, red). Staining for carbonyls preceded exposure to the anMan-specific mAb. Cells were treated with 5 mM DFMO and 0.2 mM suramin for 24 h, with 10 µg/ml U18666A for 16 h, 1 mM ascorbate for 4 h, and 25 µM MG-132, 25 µM proteasome inhibitor-I, and 1.5 µM lactacystin for 4 h. Inset in C, blow-up of indicated area. Bar, 20 µm. Reducing monosaccharides and certain of their phosphate esters should react with DNP hydrazine but should be soluble in the fixatives.

 
To isolate and analyze these radiolabeled products, we incubated T24 cells with [35S]sulfate and treated them first with DFMO to increase the number of NO-sensitive sites and with suramin and U18666A to arrest HS degradation and then with ascorbate to reactivate deaminative cleavage in the presence of proteasome inhibitors. We recovered a 20 S proteasome {alpha}1 subunit immuno-isolate, and the products associated with the proteasomes were separated on Superose 6 into two major size-pools (I and II in Fig. 5A). To test if the [35S]sulfate-labeled products in pools I and II were linked to protein, they were digested with proteinase K. After digestion, both the 35S-labeled products eluted as oligosaccharides from Superose 6 (Fig. 5, B and C) but were excluded from Superdex peptide (results not shown).

Some of the 35S-labeled products in pools I and II could be intact PG or PG with truncated side chains. If so, their glycan chains should be released by alkaline scission of the xylose-to-serine bond, which connects the glycan to the protein. However, most of the products in pool I were alkali-resistant (Fig. 5D), and the same result was obtained with the 35S-labeled material from pool II (data not shown). Because aldimines (Fig. 3B) should be acid-labile, we also treated the 35S-labeled material in pools I and II with acid, but they were both resistant (data not shown), suggesting that the 35S-labeled products were not linked to protein via aldimine bonds.


Figure 5
View larger version (27K):
[in this window]
[in a new window]

  		FIGURE 5.
Characterization of [35S]sulfate-labeled products co-precipitating with the 20 S proteasome {alpha}1 subunit. T24 carcinoma cells were grown to confluence in 75-cm2 dishes and radiolabeled as described under "Experimental Procedures." The panels show gel chromatography of intact or variously degraded [35S]sulfate-labeled products present in the 20 S proteasome {alpha}1 subunit immuno-isolate obtained from DFMO-, suramin-, and U18666A-treated and [35S]sulfate-labeled T24 cells that were further treated with 1 mM ascorbate and proteasome inhibitors for 4 h. In A–D, samples were chromatographed on Superose 6, and in E–F samples were chromatographed on Superdex peptide. In A the 35S-labeled products were separated into one high molecular size (I) and one low molecular size (II) pool. In B a portion of the 35S-labeled products present in pool I was digested with proteinase K (I-PK) and re-chromatographed on Superose 6. In C a portion of the 35S-labeled products present in pool II was digested with proteinase K (II-PK) and re-chromatographed on Superose 6. In D a portion of the 35S-labeled products present in pool I was exposed to alkali (I-OH-) and re-chromatographed on Superose 6. In E a portion of the 35S-labeled products present in pool I was digested with HS lyase (I-HS'ase) and chromatographed on Superdex peptide. In F a portion of the 35S-labeled products present in pool II was chromatographed on Superdex after deaminative degradation at pH 1.5 (II-HNO2-1.5). Cells were treated with 5 mM DFMO and 0.2 mM suramin for 24 h, with 10 µg/ml U18666A for 16 h, with 1 mM ascorbate for 4 h, and with 25 µM MG-132, 25 µM proteasome inhibitor-I, and 1.5 µM lactacystin for 4 h. Vo, void volume; Vt, total volume.

 
To show that the large molecular size 35S-labeled products (pools I and II in Fig. 5A) contained HS, they were digested with HS lyase in the presence of proteinase inhibitors or treated with HNO2 at pH 1.5. HS lyase digestion of 35S-labeled products from pool I generated 35S-labeled digestion products that were included on Superdex peptide (Fig. 5E), whereas undigested material eluted in the void volume (data not shown). 35S-Labeled material in pool II, which also eluted in the void volume of Superdex peptide (data not shown), was degraded by HNO2 at pH 1.5 (Fig. 5F). Hence, both pools I and II contained HS, most likely covalently linked to protein. It cannot be excluded that the radiolabeled material also included degradation products derived from other sulfated glycosaminoglycans.

Co-immunoprecipitation of Carbonylated Proteins, 20 S Proteasome {alpha}1 Subunit, and anMan-containing HS Oligosaccharides in T24 Carcinoma Cells—The radiolabeled HS-protein conjugates recovered from the 20 S proteasome {alpha}1 subunit immuno-isolate in the presence of proteasome inhibitors (Fig. 5A) may have been formed by a reaction between anMan-terminating HS chains or oligosaccharides and proteins, possibly damaged by oxidation. We, therefore, attempted to co-immunoprecipitate carbonylated proteins and anMan-containing HS oligosaccharides.

To capture HS-protein conjugates we pretreated T24 cells with DFMO, suramin, and U18666A. Then we activated deaminative cleavage with ascorbate in the presence of proteasome inhibitors. Cells were homogenized in sucrose-containing buffer, carbonyls were derivatized with DNP hydrazine, and products reacting either with the mAb against anMan-positive HS oligosaccharides or with the antibody against the 20 S proteasome {alpha}1 subunit in the presence of added Triton X-100 were recovered from the homogenates. These immunoisolates were subjected to protein separation by SDS-PAGE, and DNP-conjugated proteins were detected by Western blotting using anti-DNP.

DNP hydrazine is expected to penetrate membranes and tag all carbonylated compounds, including oxidized proteins and anMan-containing HS oligosaccharides. In the presence of Triton X-100, DNP-tagged compounds present inside vesicles should also be accessible to immunoprecipitation.

The results showed that the same DNP-tagged proteins were present both in the proteasome and in the anMan immunoisolates obtained from cells that were not subject to inhibition of proteasome activity (Fig. 6A). However, a 55-kDa carbonyl-containing protein band was especially prominent in the anMan immuno-isolate. In this putative HS-protein conjugate, the carbonyl may be in the protein or in the oligosaccharide or both (see Fig. 3).

Moreover, the amount of the 55-kDa carbonylated HS-protein conjugate increased more than 2-fold, as determined by densitometry, upon reactivation of deaminative cleavage in the presence of proteasome inhibitors (Asc + PI), whereas the other protein bands almost disappeared. After stripping and re-blotting with anti-Ub (Fig. 6B), the 55-kDa carbonyl-containing protein recovered in the proteasome immuno-isolate stained for Ub, whereas the corresponding protein recovered in the anMan immuno-isolate showed weak Ub staining.

When immunoprecipitation was performed in the absence of added Triton X-100, the same pattern of DNP-tagged proteins was obtained (result not shown). However, the 55-kDa component obtained in the anMan immuno-isolate from proteasome-inhibited cells was less pronounced, suggesting that HS-tagged proteins were present both inside as well as outside the endosomal compartments.


Figure 6
View larger version (54K):
[in this window]
[in a new window]

  		FIGURE 6.
Co-immunoprecipitation of carbonylated proteins, 20 S proteasome {alpha}1 subunit, and anMan-containing HS oligosaccharides in T24 carcinoma cells. Cells were pretreated with DFMO, suramin, and U18666A. Some cells were then exposed to ascorbate in the presence of proteasome inhibitors (Asc + PI). Cells were homogenized in sucrose-containing buffer, and carbonylated compounds were tagged with DNP hydrazine and immunoprecipitated with antibodies either against the 20 S proteasome {alpha}1 subunit or the anMan-containing HS oligosaccharides in the presence of Triton X-100. The immuno-isolates were subjected to SDS-PAGE (4–12% gels), and DNP- or Ub-tagged proteins, respectively, were visualized by Western blotting. Cells were treated with 5 mM DFMO and 0.2 mM suramin for 24 h, 10 µg/ml U18666A for 16 h, 1 mM ascorbate for 4 h, and 25 µM MG-132, 25 µM proteasome inhibitor-I, and 1.5 µM lactacystin for 4 h. Membranes were first blotted with anti-DNP (A), then stripped and re-blotted with anti-Ub (B). Treatments and the migration of molecular size standards are indicated.

 
Accumulation of Carbonylated Proteins in NPC-1 Fibroblasts—As shown previously (17), staining for deaminative HS degradation products by the anMan-specific mAb is greatly depressed in NPC-1 fibroblasts, suggesting that impaired transport through late endosomes also affects N-denitrosylation and subsequent deaminative HS cleavage. If carbonyl staining detects anMan-containing HS oligosaccharides, such staining should also be depressed in NPC-1 cells. On the other hand, if HS oligosaccharides are normally involved in clearance of oxidized proteins from endosomes, NPC-1 fibroblasts may accumulate carbonylated proteins.

We visualized carbonylated compounds in normal (HFL-1) and NPC-1 fibroblasts by confocal microscopy. Subconfluent HFL-1 fibroblasts, which are anMan-positive (13, 17), were also carbonyl-positive (Fig. 7A), and most of this staining colocalized with the anMan-staining both in the perinuclear area and near the cell surface (Fig. 7B, yellow/brown). There was also separate green staining suggesting that some anManOH-containing HS degradation products were formed. Treatment with U18666A to suppress deaminative cleavage did not completely abolish the carbonyl reactivity despite the disappearance of anMan-staining (result not shown). As mentioned above, the two staining procedures could have different sensitivity, but a decrease in anMan-containing oligosaccharides and a corresponding increase in oxidized proteins could also give this result. Indeed, inhibition of proteasome activity in HFL-1 fibroblasts increased carbonyl staining, some of which colocalized with the anMan-staining, whereas some of it was separate (result not shown; see silencing experiments below).


Figure 7
View larger version (50K):
[in this window]
[in a new window]

  		FIGURE 7.
Visualization and isolation of carbonyl-positive products generated in normal (HFL-1) and NPC-1 fibroblasts. A–C, the confocal laser immunofluorescence microscopy images show cells stained for carbonyls (>CO, red) and anMan-containing HS oligosaccharides (anMan, green). Bar, 20 µm. In D–E NPC-1 cells were grown to confluence in 25-cm2 wells and labeled either with [35S]sulfate (D) or [35S]Met/Cys (E) as described under "Experimental Procedures." Carbonyl-reactive material was recovered using the DNP hydrazide procedure and chromatographed on Superose 6. Total radioactivity in each tube is recorded. Vo, void volume; Vt, total volume.

 
NPC-1 fibroblasts displayed intense carbonyl-staining (Fig. 7C) despite undetectable staining with the anMan-specific mAb (17). The carbonyl-staining was not visibly affected by further treatments with suramin, ascorbate, or proteasome inhibitors over a 4-h period (result not shown) despite restored formation of anMan-positive HS degradation products (17).

To identify sulfated oligosaccharides and oxidized proteins, NPC-1 fibroblasts were incubated either with [35S]sulfate or [35S]Met/Cys for 24 h. Carbonylated or proteasome-associated products were immuno-magnet-isolated by using the DNP hydrazine procedure (21) and the antibody to the 20 S proteasome {alpha}1 subunit, respectively. No [35S]sulfate-labeled oligosaccharides were recovered using the DNP hydrazine procedure (Fig. 7D, positions 40–50). However, some high molecular size material eluting in or near the void volume was obtained. This could correspond to PGs with carbonylated protein cores. In contrast, after labeling with [35S]Met/Cys, ~50-fold greater amounts of radiolabeled carbonylated products were recovered (Fig. 7E). A similar yield and Superose 6 profile were obtained when [35S]Met/Cys-labeled products were recovered using the antibody to the 20 S proteasome {alpha}1 subunit (result not shown).

Effect of Silencing of Gpc-1 Expression on the Level of Carbonylated Proteins in HFL-1 Fibroblasts—If the anMan-containing HS oligosaccharides that became conjugated to proteins were derived from Gpc-1, silencing of Gpc-1 expression may affect the level of carbonylated proteins. To test this, HFL-1 fibroblasts were transfected with vectors expressing either a scrambled or a Gpc-1-specific siRNA to reduce Gpc-1 expression. After 48 h, cells were homogenized in sucrose, and carbonylated compounds were tagged with DNP, recovered by immunoisolation using anti-DNP in the presence of added Triton X-100, and subjected to protein separation by SDS-PAGE. The gels were finally stained for protein (Fig. 8). A 48-h silencing of Gpc-1 expression to ~60% that of the level obtained with a scrambled siRNA resulted in an increase in carbonylated proteins. The major protein eluting at 72 kDa increased 2-fold, as determined by densitometry (Fig. 8A).

Colocalization of anMan-Containing HS Oligosaccharides, Rab7, Rab9, 20 S Proteasome {alpha}1 Subunit, and Carbonyls in N2a Neuroblastoma Cells—Constitutive generation of anMan-positive HS degradation products in unperturbed N2a neuroblastoma cells was described in previous studies (17, 18). To investigate possible association between HS degradation products, endosomes, and proteasomes in undifferentiated N2a neuroblastoma cells, which grow in conglomerates (Fig. 9), we employed confocal immunofluorescence microscopy. When proteasomes and anMan-containing HS oligosaccharides were visualized using the polyclonal antibody to the 20 S proteasome {alpha}1 subunit, which also reacts with mouse proteasomes, and the mAb specific for anMan/anManOH-containing oligosaccharides, respectively, there was limited colocalization (Fig. 9A). However, inhibition of proteasome activity resulted in increased staining for anMan/anManOH-positive products (Fig. 9, A and B, insets) that often formed caps or clusters where they colocalized with the proteasomal marker (Fig. 9B, yellow).


Figure 8
View larger version (40K):
[in this window]
[in a new window]

  		FIGURE 8.
SDS-PAGE of carbonylated proteins in HFL-1 fibroblasts transfected with vectors expressing a scrambled or a Gpc-1 specific siRNA (A) and SDS-PAGE and Western blotting of Gpc-1 from cells transfected with vectors expressing a scrambled or a Gpc-1 specific siRNA (B). A, cells were transfected with the respective vectors and 48 h later homogenized in sucrose-containing buffer; carbonylated compounds were tagged with DNP, immunoisolated by using anti-DNP, and subjected to SDS-PAGE, and the gels were stained for protein. B, in parallel, cells were lysed in Triton X-100, and proteins were separated by SDS-PAGE and Western-blotted with anti-Gpc-1.

 


Figure 9
View larger version (71K):
[in this window]
[in a new window]

  		FIGURE 9.
Visualization of the proteasome 20 S{alpha} subunit, anMan-positive HS degradation products Rab7 and -9, and carbonyl-positive products in N2a neuroblastoma cells. Cells were either left untreated or treated with proteasome inhibitors (PI) or lipopolysaccharide (LPS) and DHA as indicated in the panels. The confocal laser immunofluorescence microscopy images show cells stained for proteasome 20 S{alpha} subunit (red), anMan-containing HS oligosaccharides (anMan, green), Rab7 (red), Rab9 (red), and carbonyls (>CO, red). The treatments were: PI, 25 µM MG-132, 25 µM proteasome inhibitor-I and 1.5 µM lactacystin for 4 h; lipopolysaccharide, 1 µg/ml overnight; DHA, 1 mM for 4 h. Bar, 20 µM. N2a cells grow in clusters, sometimes overlapping each other. Therefore, all cells may not be in focus.

 
Lipopolysaccharide stimulates NO formation via inducible NO synthase (25). This raises the level of protein S-nitrosylation (26), also of Gpc-1 (17, 18), which thereby increases the capacity for deaminative degradation of Gpc-1 HS. As shown above, NO release and subsequent deaminative cleavage can be induced by exogenously supplied ascorbate. However, dehydroascorbic acid (DHA) can also be used because it is taken up by many cells and converted to ascorbate in the cytosol (27). When N2a cells were exposed to both lipopolysaccharide and DHA, there was increased formation of anMan-positive products that colocalized intensely with the proteasome marker in places where the cells made contact (Fig. 9C).

Previous studies showed that anMan-positive HS degradation products colocalize with Rab7, a late endosomal marker, in N2a cells (18). When proteasome activity was inhibited in N2a cells and staining for anMan-positive products increased (Fig. 9, D and E, insets), colocalization between Rab7 and anMan-positive products intensified, forming caps or clusters at the cell periphery (Fig. 9E). Even Rab9, a marker for endosomes transporting cargo to the Golgi, colocalized with anMan-positive products upon inhibition of proteasome activity (Fig. 9F). In the absence of proteasome inhibitors, such colocalization was limited (Fig. 9F, inset).

Treatment with DNP hydrazide, which reacts with carbonyls, yielded relatively modest staining with anti-DNP, and it colocalized almost entirely with the anMan-staining, sometimes forming clusters in the peripheral cytoplasm of N2a cells (Fig. 9, G and H, yellow and brown). This is in keeping with the presence of free unreduced HS oligosaccharides and/or stabilized HS-protein conjugates (see Fig. 3C). After treatment with proteasome inhibitors, the colocalization appeared to increase slightly (Fig. 9H). In summary, N2a cells also exhibit ascorbate- or proteasome inhibitor-induced association between anMan-containing HS degradation products, late endosomes, and proteasomes that is detectable without pretreatments with DFMO and suramin.


Figure 10
View larger version (122K):
[in this window]
[in a new window]

  		FIGURE 10.
Co-immunoprecipitation of ubiquitinated proteins, 20 S proteasome {alpha}1 subunit, and anMan-containing HS oligosaccharides in N2a neuroblastoma cells. Untreated cells or cells exposed to proteasome inhibitors (Asc + PI) were homogenized in sucrose-containing buffer, treated with DNP hydrazine, and immunoprecipitated with antibodies either against the 20 S proteasome {alpha}1 subunit or the anMan-containing HS oligosaccharides, all in the presence of Triton X-100. The immuno-isolates were subjected to SDS-PAGE (4–12% gels), and Ub- or DNP-tagged proteins, respectively, were visualized by Western blotting. Cells were treated with 25 µM MG-132, 25 µM proteasome inhibitor-I, and 1.5 µM lactacystin for 4 h. Membranes were first blotted with anti-Ub, then stripped and re-blotted with anti-DNP. Treatments and the migration of molecular size standards are indicated.

 
Co-immunoprecipitation of Ubiquitinated Proteins, 20 S Proteasome {alpha}1 Subunit and anMan-containing HS Oligosaccharides in N2a Neuroblastoma Cells—Sucrose homogenates of untreated or proteasome inhibitor-treated N2a cells were exposed to DNP, and proteasome- or HS-associated proteins were isolated in the presence of added Triton X-100 as described for T24 cells and subjected to SDS-PAGE. Western blotting was performed with both anti-Ub and anti-DNP. Untreated cells yielded two major Ub-reactive proteins of ~28–30 and 60–65 kDa that co-precipitated mostly with the anti-proteasome 20 S {alpha}1 subunit but also with the mAb, detecting anMan-containing HS oligosaccharides (Fig. 10). The sharpness of the bands suggests that the proteins were not conjugated with poly-Ub.

Proteasome-inhibited cells afforded similar components, although the smaller component was not detected in the anMan immuno-isolate. However, the larger protein appeared in equal amounts in the proteasome and anMan immuno-isolates.

Stripping and re-probing with anti-DNP yielded no signal (result not shown), suggesting that there was no or little protein carbonyl formation in N2a cells. It is possible that the anMan-positive HS-protein conjugate was derived from a HS-protein conjugate stabilized by reduction (Fig. 3B). Such conjugates should be carbonyl-negative but still recognized by the anMan mAb.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
Because HS proteoglycans are usually considered to be associated with cell surfaces and extracellular matrices, the origin of HS chains associated with intracellular, cytosolic protein aggregates (3) has been obscure. Because Gpc-1 recycles via endosomes, the HS chains could be derived from this PG. Intracellular deaminative degradation of HS in Gpc-1 appears to begin in Rab5-positive endosomes and is completed in Rab7-positive endosomes, where the bulk of the anMan/anManOH-positive HS degradation products appears (17, 18). Deaminative cleavage requires release of NO/HNO from Gpc-1-SNO, which may be S-nitrosylated during secretion at the cell surface or during endocytosis.

Traffic through the endosomal compartments is inhibited by the cationic steroid U18666A. Simultaneously, this compound suppresses generation of anMan/anManOH-containing HS degradation products (17). Upon treatment with ascorbate, anMan/anManOH-positive products reappear in Rab7-positive endosomes. The ascorbate form of vitamin C is preferentially taken up by a specific sodium-ascorbate symport, as in fibroblasts (18), whereas DHA can be taken up via the less specific hexose transporters GLUT1, -3, and -4 in other cells. Inside cells, DHA is reduced to ascorbate (27). Accordingly, exogenously supplied DHA can also activate NO-dependent degradation of HS, as in N2a cells (18).

The present results showed that when deaminative cleavage of HS is reactivated in T24 carcinoma cells, [35S]sulfate-labeled degradation products appear both in Rab7-positive endosomes and in a proteasomal subunit immuno-isolate. Sustained accumulation at these sites is obtained when reactivation of HS degradation is performed in the presence of proteasome inhibitors. Upon removal of the inhibition, there appeared to be a transfer of degradation products from proteasomes back to endosomes. To maximize the yield of anMan-containing HS degradation products, T24 cells had to be pre-exposed to both DFMO and suramin. However, in N2a cells colocalization between anMan-positive HS degradation products, Rab7-positive endosomes and proteasomes were induced when only proteasome inhibitors were added to the cell cultures.

We also demonstrate that anMan-containing HS degradation products can be detected both by a DNP hydrazide-based method (21) and by the anMan-specific mAb (12). The mAb appears to recognize anMan even when it is conjugated to DNP. Because the DNP hydrazide method also detects carbonylated proteins, colocalizations between carbonyl and anMan staining can also be due to associations between HS degradation products and oxidized, i.e. carbonylated, proteins. Indeed, when deaminative HS degradation was reactivated in T24 cells by ascorbate in the presence of proteasome inhibitors, very few free HS oligosaccharides were observed in the Rab7-positive vesicles or co-precipitating with the 20 S proteasome {alpha}1 subunit. Instead, the deaminatively generated HS degradation products were most likely covalently bound to proteins. These HS-protein conjugates were resistant to both acid and alkali. The precise nature of the covalent bonds remains to be elucidated. Both unstable and stable HS-protein conjugates may be formed (see Fig. 3).

A carbonyl-containing, putative HS-protein conjugate of 55 kDa was recovered from T24 cells by immunoprecipitation with the mAb against anMan-containing HS oligosaccharides. The level of this component increased upon activation of deaminative cleavage in the presence of inhibitors of proteasome activity. Under these conditions the putative conjugate appeared to be present both inside as well as outside the endosomal compartment.

In N2a cells, a Ub-tagged, carbonyl-negative, 60–65-kDa protein was recovered by immunoprecipitation with the mAb against anMan-containing HS oligosaccharides, indicating that non-carbonylated proteins may also be conjugated to HS. The Rab7-proteasome-HS colocalizations were often concentrated to clusters near the cell surface, as if they were involved in exocytosis.

In NPC-1 fibroblasts traffic through the endosomal compartments is genetically defective, which results in diminished formation of anMan-containing HS degradation products (17). We show here that NPC-1 fibroblasts accumulate carbonylated proteins that co-precipitate with the proteasomal subunit. Moreover, suppression of Gpc-1 expression in normal fibroblasts increases the level of carbonylated proteins.

The present results are, thus, consistent with an involvement of Gpc-1-derived anMan-containing HS chains or oligosaccharides in the clearance of certain oxidized or non-oxidized proteins, perhaps preferentially aggregation-prone, recycling metalloproteins that are not easily degraded in lysosomes. Oxidized proteins often colocalize with proteasomes (28) in places where the anMan-containing HS degradation products also appear. The anMan-containing degradation products may form both reversible and irreversible conjugates. How HS-protein complexes are transported through the endosomal membrane to reach the proteasomes remains to be understood.


   FOOTNOTES
 
* The work was supported by grants from the Swedish Science Council (VR-M), the Bergvall, Crafoord, Hedborg, Jeansson, Kock, Segerfalk, Zoega, and Österlund Foundations, and the Medical Faculty of Lund University. 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. Back

1 To whom correspondence may be addressed. Tel.: 46-46-222-8573; Fax: 46-46-222-0615; E-mail: katrin.mani{at}med.lu.se. 2 To whom correspondence may be addressed. Tel.: 46-46-222-8573; Fax: 46-46-222-0615; E-mail: lars-ake.fransson{at}med.lu.se.

3 The abbreviations used are: HS, heparan sulfate; anMan, anhydromannose; anManOH, anhydromannitol; DFMO, difluoromethylornithine; DHA, dehydroascorbic acid; DNP, 2,4-dinitrophenyl; GlcNH+3, N-unsubstituted glucosamine; Gpc-1, glypican-1; Gpc-1-SNO, S-nitrosylated glypican-1; HNO, nitroxyl; NPC-1, Niemann-Pick type C-1 disease; PG, proteoglycan; SNO, S-nitroso group; U18666A, 3-beta-[2-(diethylamino)ethoxy] androst-5-en-17-one; Ub, ubiquitin; mAb, monoclonal antibody. Back


   ACKNOWLEDGMENTS
 
We thank Prof. Catharina Svanborg at the Department of Microbiology, Immunology, and Glycobiology, Lund University, for use of microscope facilities and Drs. Gunnar Peijler, Uppsala University, and Peter Påhlsson, Linköping University, Sweden for a generous gift of monoclonal antibodies.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Couchman, J. R. (2003) Nat. Rev. Mol. Cell Biol. 4, 926-937[CrossRef][Medline] [Order article via Infotrieve]
  2. Fransson, L.-Å., Belting, M., Cheng, F., Jönsson, M., Mani, K., and Sandgren, S. (2004) Cell. Mol. Life Sci. 61, 1016-1024[CrossRef][Medline] [Order article via Infotrieve]
  3. Goedert, M., Jakes, R., Spillantini, M. G., Hasegawa, M., Smith, M. J., and Crowther, R. A. (1996) Nature 383, 550-553[CrossRef][Medline] [Order article via Infotrieve]
  4. Castillo, G. M., Lukito, W., Wight, T. N., and Snow, A. D. (1999) J. Neurochem. 72, 1681-1687[CrossRef][Medline] [Order article via Infotrieve]
  5. Yogalingam, G., and Hopwood, J. J. (2001) Hum. Mutat. 18, 264-281[CrossRef][Medline] [Order article via Infotrieve]
  6. Watanabe, N., Araki, W., Chui, D.-H., Makifuchi, T., Ihara, Y., and Tabira, T. (2004) FASEB J. 18, 1013-1015[Abstract/Free Full Text]
  7. Mani, K., Jönsson, M., Edgren, G., Belting, M., and Fransson, L.-Å. (2000) Glycobiology 10, 577-586[Abstract/Free Full Text]
  8. Ding, K., Mani, K., Cheng, F., Belting, M., and Fransson, L.-Å. (2002) J. Biol. Chem. 277, 33353-33360[Abstract/Free Full Text]
  9. Cheng, F., Mani, K., van den Born, J., Ding, K., Belting, M., and Fransson, L.-Å. (2002) J. Biol. Chem. 277, 44431-44439[Abstract/Free Full Text]
  10. Horton, D., and Philips, K. D. (1973) Carbohydr. Res. 30, 367-374[CrossRef]
  11. Vilar, R. E., Ghael, D., Li, M., Bhagat, D. D., Arrigo, L. M., Cowman, M. K., Dweck, H. S., and Rosenfeld, L. (1997) Biochem. J. 324, 473-479[Medline] [Order article via Infotrieve]
  12. Pejler, G., Lindahl, U., Larm, O., Scholander, E., Sandgren, E., and Lundblad, A. (1988) J. Biol. Chem. 263, 5197-5201[Abstract/Free Full Text]
  13. Cappai, R., Cheng, F., Ciccotosto, G. D., Needham, E. B., Masters, C. L., Multhaup, G., Fransson, L.-Å., and Mani, K. (2005) J. Biol. Chem. 280, 13913-13920[Abstract/Free Full Text]
  14. Mani, K., Cheng, F., Havsmark, B., Jönsson, M., Belting, M., and Fransson, L.-Å. (2003) J. Biol. Chem. 278, 38956-38965[Abstract/Free Full Text]
  15. Cheng, F., Lindqvist, J., Haigh, C. L., Brown, D. R., and Mani, K. (2006) J. Neurochem. 98, 1445-1457[CrossRef][Medline] [Order article via Infotrieve]
  16. Mani, K., Cheng, F., Havsmark, B., David, S., and Fransson, L.-Å. (2004) J. Biol. Chem. 279, 12918-12923[Abstract/Free Full Text]
  17. Mani, K., Cheng, F., and Fransson, L.-Å. (2006) Glycobiology 16, 711-718[Abstract/Free Full Text]
  18. Mani, K., Cheng, F., and Fransson, L.-Å. (2006) Glycobiology 16, 1251-1261[Abstract/Free Full Text]
  19. Mani, K., Belting, M., Ellervik, U., Falk, N., Svensson, G., Sandgren, S., Cheng, F., and Fransson, L.-Å. (2004) Glycobiology 14, 387-397[Abstract/Free Full Text]
  20. Dong, J., Chen, W., Welford, A., and Wandinger-Ness, A. (2004) J. Biol. Chem. 279, 21334-21342[Abstract/Free Full Text]
  21. Hernebring, M., Brolén, G., Aguilaniu, H., Semb, H., and Nyström, T. (2006) Proc. Natl. Acad. Sci. U. S. A. 103, 7700-7705[Abstract/Free Full Text]
  22. Belting, M., Mani, K., Jönsson, M., Cheng, F., Sandgren, S., Jonsson, S., Ding, K., Delcros, J.-G., and Fransson, L.-Å. (2003) J. Biol. Chem. 278, 47181-47189[Abstract/Free Full Text]
  23. Ding, K., Sandgren, S., Mani, K., Belting, M., and Fransson, L.-Å. (2001) J. Biol. Chem. 276, 46779-46791[Abstract/Free Full Text]
  24. Davies, K. J. A. (2001) Biochimie (Paris) 83, 301-310
  25. Lowenstein, C. J., Glatt, C. S., Bredt, D. S., and Snyder, S. H. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 6711-6715[Abstract/Free Full Text]
  26. Gow, A. J., Chen, Q., Hess, D. T., Day, B. J., Ischiropoulos, H., and Stamler, J. S. (2002) J. Biol. Chem. 277, 9637-9640[Abstract/Free Full Text]
  27. Wilson, J. X. (2005) Annu. Rev. Nutr. 25, 105-125[CrossRef][Medline] [Order article via Infotrieve]
  28. Jung, T., Bader, N., and Grune, T. (2007) Arch. Biochem. Biophys., in press

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 Proc. Natl. Acad. Sci. USAHome page
J. Duan, F. Y. Avci, and D. L. Kasper
Microbial carbohydrate depolymerization by antigen-presenting cells: Deamination prior to presentation by the MHCII pathway
PNAS, April 1, 2008; 105(13): 5183 - 5188.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
282/30/21934    most recent
M701200200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mani, K.
Right arrow Articles by Fransson, L.-A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mani, K.
Right arrow Articles by Fransson, L.-A.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2007 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement