| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
-secretase
,¶
* Department of Biochemistry,
Department of Otolaryngology and Sensory Organ Surgery,
Department of Anatomy and Neuroscience, Osaka University Graduate School of Medicine, Suita, Osaka, Japan;
Department of Disease Glycomics, Research Institute for Microbial Diseases, Osaka University, Suita, Osaka, Japan; and
¶ Department of Anatomy, Hamamatsu University School of Medicine, Shizucka, Japan
1Correspondence: Department of Biochemistry, Osaka University Graduate School of Medicine, 2–2 Yamadaoka, Suita, Osaka 565-0871, Japan. E-mail: miyoshi34{at}biochem.med.osaka-u.ac.jp
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
100 kDa) from GnT-V-overexpressed cells revealed that its terminus started at His31, located at the boundary position between the transmembrane and stem regions. This secretion was not inhibited by a single amino acid mutation at the cleavage site (Leu29, Leu30 to Asp, His31 to Ala), but specifically inhibited by addition of DFK-167, a 
-secretase inhibitor, suggesting that -secretase is a plausible protease for secretion processing. In addition, transfection of the gene of familial Alzheimer’s disease (FAD)-linked presenilin-1, a component of -secretase, increased the secretion rate of endogenous GnT-V; the secretion of soluble GnT-V (100 kDa) was completely inhibited in presenilin-1/2 double-deficient cells, which have no -secretase activity. Collectively, these results demonstrate that Golgi-resident GnT-V is cleaved at the transmembrane region by -secretase, and this might control tumor angiogenesis through a novel pathway.—Nakahara, S., Saito, T., Kondo, N., Moriwaki, K., Noda, K., Ihara, S., Takahashi, M., Ide, Y., Gu, J., Inohara, H., Katayama, T., Tohyama, M., Kubo, T., Taniguchi, N., Miyoshi, E. A secreted type of ß1,6 N-acetylglucosaminyltransferase V (GnT-V), a novel angiogenesis inducer, is regulated by -secretase.
Key Words: glycosylation • presenilin-1 • tumor metastasis • proteases
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2
3)
. Recent studies in animal models and clinical studies have reinforced the importance of GnT-V in cancer metastasis. When GnT-V-null mice were mated with tumorigenic mice, tumor metastasis and tumor growth were found to be suppressed (4)
, and a high expression of GnT-V was correlated with a poor prognosis in colon cancer (5)
. An enhanced GnT-V expression was also observed in the early stages of carcinogenesis (6)
. In addition, we reported that a secreted type of GnT-V (soluble GnT-V) causes tumor angiogenesis by functioning as a releaser of fibroblast growth factor-2 (FGF-2) (7)
. This is an entirely different function from the original glycosyltransferase activity of this protein. Consecutive basic amino acid residues of 264–269 in GnT-V protein sequence stimulate the proliferation of endothelial cells (7)
. Therefore, a strategy for suppressing the release of soluble GnT-V from cancer cells could be one type of molecular therapy for cancer, especially in the early phases of carcinogenesis. For that purpose, it is important to identify protease(s) involved in the cleavage process of GnT-V.
The ß-site APP-cleaving enzyme (BACE) 1 has recently been identified as the protease responsible for the cleavage and subsequent secretion of a Golgi-resident ß-galactoside 
2,6-sialyltransferase (ST6Gal-I) (8)
. The mechanisms for this cleavage are complicated, but have now been well characterized (9)
. Soluble forms of glycosyltransferases are present in the plasma of patients with certain diseases and sometimes can be used as biomarkers for diseases (10
11
12
13)
. In the case of GnT-V, levels of GnT-V activity in the serum were correlated with the progression of hepatocellular carcinoma (14)
. However, a possible biological function of glycosyltransefrases in the plasma has not been reported except for our report concerning a secreted type of GnT-V (7)
.
In the present study we identified the N terminus of soluble GnT-V, which is secreted as an 100 kDa form and purified from conditioned medium of a human pancreatic cancer cell line transfected with GnT-V cDNA, started at His31, and we questioned what the responsible protease(s) was for its cleavage. Based on available information on the amino acid sequence around the cleavage site and the results of a specific inhibitor analysis or secretion assays, -secretase could be a plausible protease that is responsible for the secretion of GnT-V.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. These cells were cultured in a D-MEM or -MEM (Sigma, St. Louis, MO, USA) containing 10% FBS (Sigma), 100 µg/ml kanamycin (Wako, Osaka, Japan), 50 units/ml penicillin (Banyu Corp., Tokyo, Japan) under 5% CO2 at 37°C. Embryonic fibroblasts from presenilin-1 and presenilin-2 double homozygous-deficient mice (presenilin-1/2 double knockout MEFs) were kindly provided from Dr. B. De Strooper (Katholieke Universiteit, Leuven, and Flanders Institute for Biotechnology) and cultured in a D-MEM under 5% CO2 at 37°C. A human GnT-V cDNA (16)
inserted into a mammalian expression vector pCXN2, which is regulated by the beta-actin promoter (17)
, was transfected into PaCa-2 cells or transiently into MEFs by means of LIPOFECTAMINE 2000 (Invitrogen, San Diego, CA, USA). Selection was performed by the addition of 500 µg/ml G418 (Sigma) to establish GnT-V transfectants.
Western and lectin blot analysis
Cell lysates and conditioned medium were prepared as follows. Cells were washed twice with PBS, removed with a scraper, and centrifuged at 3000 rpm for 5 min. The pellet was homogenized in TNE buffer (10 mM Tris-HCl [pH 7.8], 1% Nonidet P-40, 0.15 M NaCl, 1 mM EDTA, and a protease inhibitors mixture [Wako, Osaka, Japan]) for 20 min on ice and centrifuged at 15,000 rpm for 15 min at 4°C. The supernatant was used as the cell lysate. Conditioned medium without serum for 40–48 h culture was centrifuged at 3000 rpm for 15 min at 4°C to remove any cell debris, and the supernatant was concentrated at 20-fold using a Centricon YM-50 device (Millipore, Bedford, MA, USA). Protein concentrations were determined with a bicinchoninic acid (BCA) kit (Pierce, Rockford, IL, USA) using BSA as a standard. Extracted proteins (10–20 µg) or conditioned medium (20-fold concentration) were suspended in Laemmli sample buffer with 5% 2-mercaptoethanol (18)
, subjected to 8% SDS-PAGE, then transferred to a nitrocellulose membrane (Schleicher & Schuell GmbH, Dassel, Germany). The blotted membrane was blocked with 5% skim milk in PBS for Western blot and 3% BSA for lectin blot, followed by incubation with 1:1000 diluted anti-GnT-V antibody (Ab) 24D11 (provided from Fujirebio, Hachiohji, Japan) and 10 µg/ml of biotinilated L4-PHA lectin (Seikagaku Corp., Tokyo, Japan). Details of the washing and developing procedure have been described (19)
. The reactive bands were detected by chemiluminescence using an enhanced chemiluminescence (ECL) system (Amersham Pharmacia Biotech Ltd., Buckinghamshire, UK).
Glycosidase treatment
For N-glycanase (PNGase F) treatment, cell extracts (10 µg of total proteins) and concentrated conditioned medium were denatured by boiling for 5 min in 100 mM sodium phosphate (pH 7.0), 2.5% Triton X-100, 1% SDS, and 5% 2-mercaptoethanol. After the addition of 1 U of PNGase F (Roche, Mannheim, Germany) or water to bring the final volume to 15 µl, samples were incubated for 12–18 h at 37°C, then analyzed by Western blot analysis.
Enzyme assay of GnT-V
GnT-V activity was determined as described previously (20)
. Briefly, cells were washed twice with PBS, removed with a scraper, and centrifuged at 3000 rpm for 5 min. The pellet was homogenized in TNE buffer for 20 min on ice and centrifuged at 15,000 rpm for 15 min at 4°C. The supernatant was used for the GnT-V assay. Conditioned medium without serum for 12–48 h culture was centrifuged at 3000 rpm for 15 min at 4°C to remove any cell debris, and the supernatant was concentrated 20-fold using a Centricon YM-50 device (Millipore). Protein concentrations were determined with a BCA kit (Pierce) using BSA as a standard. Extracted proteins (20–50 µg) or concentrated medium (5–10 µl) were incubated at 37°C for 8–12 h with 5 µM pyridylaminated agalacto biantennary oligosaccharide as an acceptor and 40 mM uridine diphosphate (UDP)-GlcNAc as a donor substrate. While a biantennary oligosaccharide is also a good substrate for N-acetylglucosaminyltransferase IX (GnT-IX), PaCa2 cells do not express GnT-IX, indicating that a specific activity of GnT-V can be measured in this assay system.
Purification of soluble GnT-V protein and determination of NH2-terminal amino acid sequence
GnT-V-transfected PaCa-2 (PaCa-2/V) cells were grown to confluent conditions in 15 cm dishes. After washing with PBS twice, serum-free medium was added, followed by culture for a further 48 h. The conditioned medium was collected and centrifuged at 3000 rpm for 15 min at 4°C to remove any cell debris, then frozen at –30°C until used. Conditioned medium (1500 ml) was saturated with ammonium sulfate and centrifuged at 13,500 rpm for 30 min at 4°C. The mixture was placed in a dialysis membrane for 12 h at 4°C. The medium, saturated with ammonium sulfate (400 ml), was centrifuged at 13,500 rpm for 30 min at 4°C. The pellet was dissolved in ice-cold PBS and finally adjusted to 10 ml. The solution was applied to a PD-10 column (Amersham Biosciences, Uppsala, Sweden) to remove the salts, followed by elution with 50 mM Tris-HCl (pH 7.5). The extracted fractions were applied to an immunoaffinity column of GnT-V that had been equilibrated with 50 mM Tris-HCl (pH 7.5) containing 0.5 M NaCl and 1% Triton X-100. An immunoaffinity column of GnT-V was established using the 24D11 Ab coupled to protein A-Sepharose 4B (Amersham Pharmacia Biotech Ltd.). The flow rate during loading was 6 ml/h. After washing the column with 5 column volumes of the same buffer with equilibration, GnT-V was eluted with 0.05% trifluoroacetic acid. The pH of the eluted fraction was immediately adjusted to neutrality with 2 M Tris-HCl (pH 9.0). Fractions with a high activity of GnT-V were pooled and dried using CC-181 (Tomy Seiko, Tokyo, Japan). The dried proteins were dissolved in Laemmli sample buffer with 5% 2-mercaptethanol and subjected to 6% SDS-PAGE. The proteins were then transferred to a PVDF membrane (Millipore). After staining with Coomassie brilliant blue (CBB), the soluble form of GnT-V was excised from the membrane, and the amino-terminal amino acid sequence was determined using HP G1005A Protein Sequencing System (Hewlett Packard, Palo Alto, CA, USA).
Site-directed mutagenesis
Mutations of GnT-V were generated using the Quick Change Site-Directed Mutagenesis kit (Stratagene, La Jolla, CA, USA). Primer M1F [5'-GGCTTCATTTGGGGTATGATGGATCTGCAACTTTACCATCCAGCAG-3'] and primer M1R [5'-CTGCTGGATGGTAAAGTGCAGATCCATCATACCCCAAATGAAGCC-3'] for mutating the Leu residue at position 29 to Asp (L29D), primer M2F [5'-GGCTTCATTTGGGGTATGATGCTTGATCACTTTACCATCCAGCAG-3'] and primer M2R [5'-CTGCTGGATGGTAAAGTGATCAAGCATCATACCCCAAATGAAGCC-3'] for mutating the Leu residue at position 30 to Asp (L30D), primer M3F [5'-GGCTTCATTTGGGGTATGATGCTTCTGGCATTTACCATCCAGCAG-3'] and primer M3R [5'-CTGCTGGATGGTAAATGCCAGAAGCATCATACCCCAAATGAAGCC-3'] for mutating the His residue at position 31 to Ala (H31A) were used, respectively. Original GnT-V inserted into pSVK3 vector was used as a template for generating each mutant GnT-V expression vector. Detailed procedures have been described in the manufacturer’s protocol. All mutated sequences were verified by automated DNA sequencing using a Big Dye Terminator Cycle Sequencing fibrous sheath (FS) Ready Reaction Kit and ABI Prism 3700 Genetic Analyzer (Applied Biosystems, Warrington, UK). PaCa-2 cells were transiently transfected with wild-type (WT) or mutant GnT-V in the pSVK3 expression vector (Pharmacia Biotech, Inc., Piscataway, NJ, USA).
Secretion assay
To determine the secretion rate of GnT-V in a cell, we used the activity assay of GnT-V. In the case of mutant GnT-V transfection, both cell lysate and conditioned medium derived from GnT-V-transfected PaCa-2 cells in 6-well culture plates were extracted 48 h after transfection, as described above. For SK-N-SH transfectants, 1 x 106 cells were plated on 6 cm dishes, then cultured for 10 h. The conditioned medium was exchanged with 2 ml of serum-free medium and cultured for an additional 48 h. Both cell lysate and conditioned medium were extracted using identical procedures. The GnT-V activity of these samples was assayed as described above, and the secretion rate was calculated as follows: (total activity of conditioned medium/(total activity of conditioned medium+total activity of cell lysate)) x 100.
Inhibitor treatment
After 5 x 104 of cells were cultured in a 96-well plate for 10 h, they were washed with PBS, cultured with 50 µl of serum-free medium containing 1% of DMSO with/without various concentrations of DFK-167 (a -secretase inhibitor) (21)
(Enzyme System Products, Livermore, CA, USA), then cultured for additional 12 h. The conditioned medium was collected and centrifuged at 2500 rpm at 4°C to remove any cell debris. The supernatants were subjected to the GnT-V enzymatic assay.
Northern blot analysis
Total RNAs were prepared from SK-N-SH transfectants according to the method reported (22)
. Twenty micrograms of RNA were electrophoresed on a 1% agarose gel containing 2.2 M formaldehyde and transferred onto a Zeta probe membrane (Bio-Rad, Hercules, CA, USA). The membrane filter was hybridized with a [-32P]cytidine triphosphate-labeled GnT-V cDNA fragment for 12 h at 42°C in a hybridization buffer (23)
. After washing, the filter was exposed to an X-ray film (Eastman Kodak Co.) with an intensified screen at –80°C for 20 h.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, we first checked the amount of soluble GnT-V secreted from cancer cells into the conditioned medium by Western blot of GnT-V. While the secreted GnT-V bands were barely detectable in normal conditioned medium from several cancer cell lines after a 24–48 h cell culture at 37°C, a 20-fold concentrated medium by an ultrafiltration device clearly showed an 100 kDa form of soluble GnT-V (Fig. 1
A, PaCa-2; data for the other cell lines not shown), indicating that endogenous GnT-V can freely secrete out of cancer cells like other glycosyltransferases (8
9
10
11
12
13)
. Next, to investigate this secretion mechanism we purified the soluble GnT-V and determined its NH2-terminal protein sequence using GnT-V-transfected PaCa-2 cells (PaCa-2/V) since PaCa-2 cells, a human pancreatic cancer cell line, are the most efficient secretors among the cancer cells tested (data not shown); however, the amount of its secretion was not sufficient to permit its purification. Western blot of GnT-V in the unconcentrated conditioned medium of PaCa-2/V revealed the high secretion level of soluble GnT-V as well as a high expression level of Golgi-resident GnT-V (Fig. 1A
, PaCa-2/V) and showed the molecular size of soluble GnT-V to be 100 kDa, similar to one of the Golgi-resident GnT-V for both PaCa-2 and PaCa-2/V cells (Fig. 1A
); this suggested that the cleavage site was very close to the N terminus of Golgi-resident GnT-V. Since the both bands of soluble GnT-V from PaCa-2 and PaCa-2/V cells migrated very broadly (Fig. 1A
) and it was reported that the differences in the processing of N-linked oligosaccharides could affect the molecular mass of the secreted glycosyltransferases (24
, 25)
, we next reevaluated the size of each secreted soluble GnT-V by Western blot after PNGase F treatment to remove all N-linked oligosaccharide chains. As shown in Fig. 1B
, deglycosylated, soluble forms of GnT-V expressed in PaCa-2 and PaCa-2/V cells migrated with identical molecular masses and much faster than the PNGase F-untreated soluble forms, suggesting that soluble GnT-V was highly glycosylated, and its cleavage site from both cells seemed to be the same. These results suggest that the cleavage site of GnT-V and its N-glycosylation patterns in GnT-V-overexpressed cells could be compatible with that of endogenous GnT-V in parental cells. Thereafter, we purified soluble GnT-V protein from the conditioned medium of PaCa-2/V cells to determine the cleavage site of GnT-V (see Materials and Methods). Secreted GnT-V protein purified by immunoaffinity chromatography was visualized by CBB staining on a PVDF membrane (Fig. 1C
); the band with a molecular mass of 100 kDa, corresponding to the positive band detected in the Western blot of GnT-V (data not shown), was subjected to amino acid sequencing analysis. The terminal amino acid sequence of the purified soluble GnT-V was found to be "His-Phe-Thr-Ile-Gln." This amino acid sequence was identical to the sequence from His31 to Gln35 in GnT-V. Taken together, these results indicate that the cleavage site of GnT-V is between Leu30 and His31, which is located at the boundary position between the transmembrane and the stem region (Fig. 1D
).
|
No changes in GnT-V secretion after making point mutation at the cleavage site of GnT-V
Since the NH2-terminal amino acid of the secreted type of GnT-V from PaCa-2/V cells was found to be His31 (Fig. 1D
), we constructed single amino acid mutated GnT-V expression vector in the region from Leu29 to His31 by site-directed mutagenesis (Fig. 2
A), transfected them into PaCa-2 cells, and analyzed the effects of amino acid substitutions on GnT-V secretion by Western blot and a secretion assay of GnT-V (see Materials and Methods). As shown in Fig. 2B, C
, substitution of each amino acid near the cleavage site led to no change in either molecular size (Fig. 2B
) or secretion rate of soluble GnT-V (Fig. 2C
) compared with the pattern of WT GnT-V. We carried out the same assay using COS-7 cells, which have no activity of GnT-V, and results were similar to those case of PaCa-2 cells (data not shown). Therefore, these results suggest that the protease(s) responsible for the cleavage and secretion of GnT-V at the transmembrane region is not an amino acid sequence-specific protease.
|
Cleavage and secretion of GnT-V were inhibited by -secretase inhibitor
To identify which protease(s) is/are involved in GnT-V secretion, we investigated several known proteases that cleave around the transmembrane region and require no specific amino acid sequences to function. As shown in Fig. 3
A, -secretase can cleave some membrane proteins within the transmembrane region (26
27
28
29
30)
—in particular, E-cadherin could be cleaved by -secretase at the same position of GnT-V cleavage site that we identified (30)
. Therefore, we investigated whether -secretase inhibitor could inhibit secretion of GnT-V by Western blot and a secretion assay of GnT-V after PaCa-2/V cells were treated with its inhibitor for 12 h at 37°C. DFK-167, a specific -secretase inhibitor, inhibited secretion of the 100 kDa form of soluble GnT-V in a dose-dependent manner compared with no reagent treatment (NT) or DMSO treatment only (CTRL) (Fig. 3B, C
). We also checked cell viability under this inhibitor treatment by an MTT assay, and no cell toxicity was found at the concentrations used in this experiment (data not shown). These results suggest that -secretase is involved in the cleavage of the transmembrane region of GnT-V. To exclude the possibility that DFK-167 suppresses synthesis of GnT-V, a Western blot analysis of cell lysates as well as an RT-polymerase chain reaction (RT-PCR) analysis was performed. The expression of GnT-V in cell lysates was increased by a little after DFK-167 treatment. However, the GnT-V mRNA expression was not changed with DFK 167 treatment (see supplemental data).
|
Expression of familial Alzheimer’s disease (FAD) -linked presenilin-1, a dominant active form of -secretase, promotes secretion of GnT-V. While the function and role of -secretase are not clearly known, presenilin-1 (PS-1) has been reported to play an important role -secretase activity (26
, 31
, 32
, 33)
. Moreover, the -secretase complex with FAD-linked PS-1 variants (A246E, 
E9) could increase -secretase activity because these variants’ cDNA transfection elevated the Aß 42/Aß 40 production ratio (26)
, a measure of that activity; from previous reports (15
, 26)
, the activity level of PS-1 (-secretase) is thought to be WT < A246E < E9. Therefore, to address the possibility that -secretase may be involved in GnT-V cleavage and processing, we analyzed the secretion efficiency of endogenous GnT-V by a secretion assay in specific SK-N-SH cells that were stably transfected with WT PS-1 (wild-type) or FAD-linked PS-1 (A246E, E9), as well as control vector only (mock) (15)
. GnT-V activity in the condition medium of PS-1 E9 transfectant was dramatically increased although GnT-V activity in cell lysates was not significantly changed among the four groups. Western blot analysis using cell lysates showed similar levels of GnT-V, but the molecular size of GnT-V in the E9 PS-1 transfectants (E9) was lower than the other transfectans (Fig. 4
B, upper panel), suggesting this might be due to an increased cleavage of intracellular GnT-V. In addition, we assessed the amount of ß1–6GlcNAc branching oligosaccharide, the product of GnT-V, in these transfectants by L4-PHA lectin blot analysis, which preferentially recognizes the structure of ß1–6GlcNAc branching oligosaccharides (34)
, after the same membrane of GnT-V Western blot was deprobed. As shown in Fig. 4B
, the amount of ß1–6GlcNAc branching oligosaccharide was inversely proportional to the secretion rate of GnT-V (Fig. 4B
, lower panel, asterisk), suggesting that intact intracellular GnT-V cannot work properly in the cells of higher -secretase activity. In contrast, expression of GnT-V mRNA was increased in E9 transfectants (Fig. 4C
). Increased level of mRNA might not cause the increased secretion rate of GnT-V because expression levels of intracellular GnT-V protein (Fig. 4B
, upper panel) and relative activities of GnT-V among these transfectants were similar. Taken together, these results suggest that a higher level of activity of -secretase by WT or mutated PS-1 overexpression could cause an increase in the levels of secretion and turnover of GnT-V and a decrease in the levels of Golgi-resident GnT-V, followed by a decrease in the amount of ß1–6GlcNAc branching in the cells.
|
Secretion of GnT-V from presenilin-1/2 double knockout cells
One should consider the problem of the overexpression system by presenilin-1 (PS-1) to increase -secretase activity in cells, since it is known that -secretase is a complex of comprised of several molecules including PS-1, nicastrin, PEN-2, and APH-1 (35
36
37)
. Moreover, it was reported that presenilin-2 (PS-2), which is the homologous protein of PS-1, can play a distinct role in -secretase activity without PS-1 (38
, 39)
. Therefore, to confirm that intracellular GnT-V can be cleaved and secreted out of the cells by -secretase, we investigated the secretion levels of soluble GnT-V by Western blot using presenilin-1/2 (PS-1/2) double-deficient mouse embryonic fibroblast (MEF) cells, in which there are no -secretase activities (40
, 41)
. The protein expression levels of PS-1 and GnT-V, as confirmed by Western blot using WT and PS-1/2-deficient MEFs, revealed a lack of PS-1 in its deficient MEF (Fig. 5
A, upper panel) and very low expression of GnT-V in both MEFs (data not shown); therefore, GnT-V was overexpressed in both MEFs by transient transfection to assess the GnT-V secretion levels. After transfection for 24 h at 37°C, conditioned medium and cell lysate were collected from both MEFs and subjected to SDS-PAGE, followed by Western blot of GnT-V. As expected, the 100 kDa form of soluble GnT-V was completely inhibited in PS-1/2-deficient MEF, whereas a sufficient level of expression intracellular GnT-V was observed in its deficient cells (Fig. 5B
). This result strongly suggests that -secretase is responsible for production of soluble GnT-V.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-secretase was identified as a specific protease involved in that cleavage. We previously purified GnT-V proteins from the conditioned medium of a small lung cancer cell line (42)
, and soluble GnT-V was present in the natural biological status. Many papers have reported that the levels of ß1–6GlcNAc branching, the product of Golgi-resident GnT-V, are frequently increased in a variety of malignant tumors and are closely correlated with tumor progression (3
, 43)
. Moreover, our previous findings of soluble GnT-V as an angiogenesis inducer prompted us to investigate which position in the GnT-V protein sequence represents the cleavage site and which proteases cleave GnT-V at that site. Since the expression levels of endogenously secreted GnT-V from parental cancer cells were not sufficient to permit its purification, we used PaCa-2 cells (human pancreatic cancer cell line) transfected with GnT-V. In overexpression systems, problems such as mislocalization or abnormal metabolic change of glycosyltransferases might arise, so a pulse-chase study of GnT-V was performed to check the pattern of its secretion and the life span of protein using parental and GnT-V-transfected PaCa-2 cells. The levels of secreted GnT-V were low compared with that of intracellular GnT-V, but both proteins were stable for periods of up to 6 h (data not shown). Therefore, the method to assess GnT-V secretion in this study is deemed suitable.
-Secretase belongs to the aspartic protease family and is able to cleave membrane proteins within the transmembrane region (44)
. It has been reported that -secretase is responsible for the production of Aß 42, a fragment of C99, resulting in the development of Alzheimer’s diseases (43
, 45)
. It is also known that a specific sequence of amino acids is not necessary for -cleavage to occur (46
, 47)
, and -secretase is constructed as a complex of several molecules including, at a minimum, presenilin-1, nicastrin, PEN-2, and APH-1 (35
36
37)
. A number of papers have reported that certain membrane proteins such as Notch 1 (27)
, CD44 (28)
, ErbB-4 (29)
, and E-cadherin (30)
are also cleaved by this secretase (Fig. 3A
); these substrates for -secretase are all type I membrane proteins and are found on the cell surface. However, GnT-V is a type II membrane protein present in the Golgi apparatus; therefore, if GnT-V were a candidate substrate for -secretase, this represents a new function of -secretase. In the present study, to show the involvement of -secretase in GnT-V cleavage, we transfected an active form of precenilin I into SK-N-SH cells. As shown in Fig 3
, secretion of GnT-V was dramatically increased in E9 transfectants and binding to L4-PHA lectin was decreased, which might in part include nonspecific bindings. From this result, we wondered whether the turnover of GnT-V was accelerated in cells with high -secretase activity due to its increased level of secretion; to address this, we evaluated the mRNA expression levels of GnT-V in SK-N-SH transfectants by Northern blot of GnT-V. As expected, mRNA levels of GnT-V were increased in proportion to the secretion rate of GnT-V in E9 cells (Fig. 4B
). This result suggests that increases in mRNA concomitant with increased secretion are due to differences in turnover. Furthermore, we showed that the formation of soluble GnT-V was completely inhibited by addition of -secretase inhibitor in PaCa-2/V cells in -secretase inhibitor experiments. Moreover, the lack of secreted GnT-V observed in PS-1/2 double knockout cells, in which no -secretase activity was found, supports this conclusion. While there is a difference of GnT-V level in cell lysate between WT and PS-1/2 MEF cells (Fig 5B
), it could be due to lower transfection efficiency into PS-1/2 cells. Densitometry analysis showed that the level of GnT-V in WT was 1.7-fold higher than that of precenilin 1-deficient cells. However, no bands of 100 kDa GnT-V were detected even if twice the volume of conditioned medium from precenilin 1-deficient cells were electrophoresed. Establishment of permanent GnT-V transfectants in MEF cells was difficult, although we do not know the precise reason. In addition, PNGase F treatment of soluble GnT-V indicates that Golgi-resident GnT-V could be cleaved after it was highly glycosylated (Fig. 1B
), indicating that the location where it is cleaved in cells might be around the trans-Golgi network (TGN). It was reported that -secretase activity is highest in the TGN (48)
; therefore, this supports a scenario in which GnT-V is processed by -secretase at the point of their colocalization around the TGN. As shown in Fig 4B
, a marked increase of GnT-V secretion inversely brought decreases in ß1–6 GlcNAc branching. This phenomenon might reflect certain pathological conditions. An immunohistochemical study of GnT-V and L4-PHA in melanoma tissues showed their coexpression in certain lesions, but not in other lesions (T. Handerson, E. Miyoshi, N. Taniguchi, J. Pawelak, unpublished results).
In conclusion, Golgi-resident GnT-V is cleaved mainly in the transmembrane region by -secretase, followed by its secretion. This shedding of glycosyltransferases could play an important role in the regulation of glycosylation of many glycoproteins related to tumor progression as well as the production of secreted soluble GnT-V, related to tumor angiogenesis. Therefore, a selective inhibitor for this GnT-V secretion, a -secretase inhibitor, could provide a useful strategy for the inhibition of tumor angiogenesis and progression.
|
ACKNOWLEDGMENTS |
|---|
Received for publication December 20, 2005. Accepted for publication July 17, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
5ß1 integrin in a cisplatin-resistant human squamous cell carcinoma cell line. Mol. Cancer Ther. 2,1207-1214-secretase activity. J. Med. Chem. 41,6-9[CrossRef][Medline]-secretase activity. Nature 398,513-517[CrossRef][Medline]-secretase-like protease mediates release of Notch intracellular domain. Nature 398,518-522[CrossRef][Medline]-secretase-like intramembrane cleavage of ErbB4. J. Biol. Chem. 277,6318-6323-secretase cleavage releases the E-cadherin intracellular domain and regulates disassembly of adherens junctions. EMBO J. 21,1948-1956[CrossRef][Medline]-secretase bind directly to presenilin-1. Nat. Cell Biol. 2,428-434[CrossRef][Medline]-secretase inhibitors directed to the active site covalently label presenilin 1. Nature 405,689-694[CrossRef][Medline]-secretase cleavage of betaAPP, and presenilin protein accumulation. Dev. Cell. 3,85-97[CrossRef][Medline]-Secretase complex. Neuron. 38,9-12[CrossRef][Medline]-secretase complex. Nature 422,438-441[CrossRef][Medline]-secretase activities. J. Biol. Chem. 278,22475-22481-Secretase activity requires the presenilin-dependent trafficking of nicastrin through the Golgi apparatus but not its complex glycosylation. J. Cell Sci. 116,1127-1136-secretase cleavage. J. Biol. Chem. 271,21914-21919-Secretase, evidence for multiple proteolytic activities and influence of membrane positioning of substrate on generation of amyloid beta peptides of varying length. J. Biol. Chem. 274,11914-11923-secretase activity to the trans-Golgi network. J. Neurochem. 84,1143-1153[CrossRef][Medline]This article has been cited by other articles:
|
|
 |
|
  D. Li, Y. Li, X. Wu, Q. Li, J. Yu, J. Gen, and X.-L. Zhang Knockdown of Mgat5 Inhibits Breast Cancer Cell Growth with Activation of CD4+ T Cells and Macrophages J. Immunol., March 1, 2008; 180(5): 3158 - 3165. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
