| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Short Communication |
Virologie et Pathologie Humaine, CNRS FRE3011, Université Lyon 1, Faculté de Médecine RTH Laennec, rue G. Paradin, F-69372 Lyon, France
Correspondence
Mireille de Turenne-Tessier
tessier{at}sante.univ-lyon1.fr
 |
ABSTRACT |
|---|
 TOP ABSTRACT MAIN TEXTREFERENCES |
|---|
|
MAIN TEXT |
|---|
|
TOPABSTRACTMAIN TEXTREFERENCES |
|---|
). Although latent membrane protein 1 is essential for B-cell immortalization, the BARF1 gene is thought to play an important role in carcinoma development. Indeed, its expression was found in >90 % of EBV-positive gastric or nasopharyngeal carcinoma biopsies (Decaussin et al., 2000; zur Hausen et al., 2000; Luo et al., 2005; Seto et al., 2005) and its anti-apoptotic role was recently demonstrated in gastric cancer cells (Wang et al., 2006b). Our transfection experiments (de Turenne-Tessier et al., 2005; Ooka, 2005) demonstrated the ability of BARF1 to induce cell immortalization or malignant transformation and to activate several cellular genes, including Bcl-2. On the other hand, our observations and those of others (Tanner et al., 1997; Strockbine et al., 1998; Cohen & Lekstrom, 1999) revealed several immunomodulatory functions of the BARF1 protein.
Previous studies of the BARF1 sequence predicted a 24.5 kDa hydrophobic polypeptide with a 20 aa signal peptide, three sites for phosphorylation and one site for N-glycosylation (de Turenne-Tessier et al., 1997). However, we detected the BARF1 protein as a 31–33 kDa intracellular polypeptide in various cell types and, later, the secretion of a 29 kDa polypeptide was reported (Strockbine et al., 1998) and confirmed in our laboratory from distinct BARF1-expressing cells in the growth phase (Sall et al., 2004). BARF1 expression in human epithelial cells infected with a tetracyclin-regulatable recombinant adenovirus (rAd/B-G) allowed us to identify the signal peptide cleavage site by amino-terminal sequencing, and to discover that the BARF1 protein is secreted naturally as a macromolecule endowed with mitogenic activity in various cell cultures (Sall et al., 2004; de Turenne-Tessier et al., 2005). Here, our expression system was used further to investigate the post-translational modifications of the BARF1-encoded polypeptide and the native molecular structure of its secreted form. The data presented here on the phosphorylation and glycosylation of the BARF1 protein complement both our biophysical analyses revealing the hexameric structure of the secreted BARF1 protein (Tarbouriech et al., 2006) and the studies on BARF1 protein maturation reported by Wang et al. (2006a).
Phosphorylation of the BARF1 protein was examined in both its secreted and membrane forms, as BARF1 protein was found in both the insoluble and aqueous-phase fractions obtained by Triton X-114 extraction and partitioning of the crude cell-membrane components from rAd/B-G-infected 293-tTA cells (de Turenne-Tessier et al., 2005). BARF1 protein samples were prepared and submitted to two-dimensional (2D) electrophoresis and Western blotting as described previously (de Turenne-Tessier et al., 2005), except that blots were treated in TBS-Tw (20 mM Tris, pH 7.6; 137 mM NaCl; 0.1 % Tween 20) and saturated with 1 % BSA. By using the classical stripping/reprobing procedure (Amersham Biosciences), the BARF1 protein was detected with Ra-Barf1 polyclonal antibodies (de Turenne-Tessier et al., 2005) and its phosphorylation was examined with anti-phosphothreonine and anti-phosphoserine monoclonal antibodies (Sigma). Although no phosphorylation could be detected on BARF1-associated spots from the Triton X-114-insoluble membrane fraction, spots identified as BARF1-encoded polypeptide either in the aqueous phase of Triton X-114-soluble membrane extract or in the conditioned culture medium were recognized clearly by both anti-phosphoserine and anti-phosphothreonine antibodies (Fig. 1). As protein phosphorylation occurs in the endoplasmic reticulum (ER) and BARF1 protein follows the classical secretory pathway (see below), these results supported our previous suggestion (de Turenne-Tessier et al., 2005) that the Triton X-114-insoluble membrane fraction retained uncleaved immature BARF1 polypeptide, whereas the soluble fraction contained cleaved BARF1 protein being processed through the intracellular cisternae to the plasma membrane, without excluding the possible presence of secreted BARF1 protein acting as an autocrine growth factor. As the presence of several isoforms in BARF1 protein 2D electrophoresis patterns seemed to be only partially connected with serine/threonine phosphorylation (Fig. 1) or glycosylation (results not shown), these isoforms might also reflect unidentified post-translational modification, partial denaturation of oligomers/aggregates during isoelectric focusing and/or recombinant protein heterogeneity. Our higher detection of phosphothreonine than phosphoserine might reflect differences in phosphorylation levels and/or in antibody affinities. NetPhos (Blom et al., 1999) and Proscan (Combet et al., 2000) predictions differ, and analyses using site-specific antibodies, directed mutagenesis and/or mass spectrometry should clarify the sites and functions of the BARF1 protein multi-site phosphorylation (Salazar & Höfer, 2007).
|
N-Glycosylation was first analysed by using either peptide N-glycosidase F (NGF) or endoglycosidase H (endoH), known to remove all and only immature (non-complex) types of N-linked glycan, respectively (Maley et al., 1989). As shown in Fig. 2(a), no digestion by either glycosidase was obtained from the native form of the secreted BARF1 protein, but partial denaturation rendered the protein sensitive to either NGF or endoH; the major 29 kDa polypeptide (p29) was degraded into a 25 kDa product (p25), and the 27 kDa polypeptide (p27, detected in 293-tTA but not HeLa-rtTA cell medium) into a 23 kDa product (p23). Although some digestion by endoH was observed on the BARF1 protein treated with 15 mM dithiothreitol (DTT), heating with 0.2 % SDS was necessary and sufficient for complete N-glycan deletion. The predicted N-glycosylation of the BARF1 protein was thus confirmed and was shown to consist of mannose-rich carbohydrate side chains buried within the native molecular structure of the secreted protein. The N-glycans were examined further by testing BARF1 protein affinity for agarose-conjugated concanavalin A (conA-ag, from Canavalia ensiformis; Sigma). Overnight incubation of HeLa-rtTA cell-culture medium with conA-ag in conA buffer (0.1 M potassium phosphate buffer, pH 6, containing 1 M NaCl and 1 mM each of MgCl2, CaCl2 and MnCl2) at room temperature was followed by conA-ag washing with conA buffer and elution of the conA-bound proteins by competition with methyl-
-D-glucopyranoside (MGP, 0.5–1.0 M; Sigma). As shown in Fig. 2(a), mannose-rich N-glycosylation was confirmed by the great affinity of the secreted BARF1 protein for conA, which led us to develop a simple and efficient purification procedure. Indeed, the direct fractionation of the culture medium of 750x106 HeLa-rtTA cells with conA-ag (in the ratio of 1 µl conA-ag suspension to 1–3 µg BARF1 protein in a 20 µl final volume) resulted in the recovery of about 1 mg highly enriched protein, allowing successful biophysical analyses of the secreted BARF1 protein structure (Tarbouriech et al., 2006).
|
) and the NetOglyc 3.1 program (Julenius et al., 2005) predicted a mucin-type O-glycosylation at threonine 169 of the BARF1 sequence, we wondered whether p29 and p25 might differ from p27 and p23, respectively, due to O-GalNAc glycosylation (Van den Steen et al., 1998). When secreted BARF1 protein was treated with O-glycosidase (endo--N-acetylgalactosaminidase from Streptococcus pneumoniae, 0.02–0.10 U ml–1), which releases the Gal(
1–3)GalNAc disaccharide bound to serine or threonine as a core unit of mucin-type O-glycans, no digestion was observed for either native or partially denatured samples. On the other hand, neuraminidase from Vibrio cholerae, which releases terminal sialic acid from various glycans, reduced p29 and p25 into p27 and p23, respectively, whether the BARF1 protein had been partially denatured or not (Fig. 3a). No further downward shift was detectable when O-glycosidase was added with or after neuraminidase. Such data suggested a mucin-type O-glycosylation during which sialic acid substitution of the disaccharide Gal(1–3)GalNAc would be responsible for the resistance to O-glycosidase. With reference to recently reported data (Nagano et al., 2005), O-glycosylation was investigated further by testing the affinity of blotted BARF1 protein for peanut agglutinin (PNA, from Arachis hypogaea; Sigma), a lectin that recognizes the disaccharide Gal(1–3)GalNAc in the absence of sialic acid residues. SDS-PAGE and protein transfer were performed on native or neuraminidase-digested samples of both pure fetuin from fetal calf serum (Sigma), used as a control (Carpentier et al., 2005), and a conA-purified fraction from HeLa-rtTA-secreted BARF1 protein. After saturation with BSA, the blot was incubated overnight at 4 °C with 20 µg biotin-labelled PNA ml–1 in TBS-Tw containing 1 mM each of MgCl2, CaCl2 and MnCl2, then washed and treated with peroxidase-conjugated streptavidin for chemiluminescence analysis of lectin binding. As shown in Fig. 3(b), only the desialylated forms of BARF1 protein and fetuin were able to bind PNA efficiently, which indicated BARF1 protein modification by Gal(1–3)GalNAc, with a sialic acid (2–3) substitution preventing PNA recognition.
|
-D-galactopyranoside (benzyl--GalNAc) and monensin, which respectively prevent mucin-type O-linked glycosylation directly and indirectly in the Golgi; and brefeldin A, which disassembles the Golgi complex and thus inhibits ER
Golgiplasma membrane translocation (Mardassi et al., 1998; Tsuiji et al., 2003; Pager et al., 2004). As revealed by Western blot analysis of the crude cell extracts and conditioned culture media (results not shown), all of the agents tested suppressed or strongly restricted BARF1 protein secretion, leading to intracellular accumulation of distinct BARF1 products: matured forms in the presence of brefeldin A (5 µg ml–1), p23 only in the presence of tunicamycin (10 µg ml–1) and p27 only in the presence of benzyl--GalNAc (1.5–4.5 mM) or monensin (17 µg ml–1), with some p25 secretion being observed in the presence of tunicamycin, but not in the presence of both tunicamycin and monensin. Although Wang et al. (2006a) reported that N-glycosylation was essential for the maturation and ERGolgiplasma membrane translocation of BARF1 protein in BARF1-transfected HeLa cells, our observations demonstrated the involvement of both N- and O-glycosylation in the active secretion of BARF1 protein via the classical pathway. The weak secretion of incompletely glycosylated polypeptides was enigmatic, but has already been reported from another human recombinant adenovirus system (Mardassi et al., 1998) and from vaccinia virus-infected cells (Ng et al., 2001). As 293-tTA and HeLa-rtTA cells, both of human epithelial origin, but permissive and non-permissive, respectively, for rAd/B-G replication, did not secrete exactly the same panel of BARF1 protein isoforms (Fig. 2
), BARF1 polypeptide processing seemed to be partially dependent on the cell type and/or adenovirus expression system used.
We have demonstrated here that the detection of 23–29 kDa products from actively growing cells was related to both N- and O-glycosylation of cleaved BARF1 polypeptide. As cell incubation with both tunicamycin and monensin limited BARF1 expression to a 23 kDa polypeptide co-migrating with the product of secreted BARF1 protein digestion with both endoH and neuraminidase, the absence of N- and O-linked glycans led to a polypeptide molecular size close to that predicted from the BARF1 amino acid sequence (22.3 kDa). Our data on N-glycosylation supplement both the mutation approach of Wang et al. (2006a) and our biophysical analyses (Tarbouriech et al., 2006) to confirm the existence of a single, high-mannose-type N-glycan linked to asparagine 95 and located on the inner side of the hexameric ring structure of the BARF1 protein. Small discrepancies between our evaluation and that of Wang et al. (2006a) of N-glycan and BARF1 isoforms produced by HeLa cells might result from the use of different expression vectors and/or electrophoresis conditions. As mannose-rich N-linked glycans are not typical of cytokines (Meads & Medveczky, 2004) or secreted viral proteins (Mardassi et al., 1998; Ressing et al., 2005; Wahl-Jensen et al., 2005; Evans et al., 2006), and as N-glycosylation did not appear essential for BARF1 protein oligomeric structure (unpublished data from our sedimentation-velocity analyses; not shown), it is suggested that immature N-glycosylation results from BARF1 polypeptide oligomerization in the ER, the concealed location of asparagine 95 hindering the access of oligosaccharides to the Golgi processing machinery (Rudd & Dwek, 1997; Cals et al., 1996). Mucin-type O-glycosylation with sialic acid substitution was indicated by the sensitivity of native secreted BARF1 protein to neuraminidase, its resistance to O-glycosidase and PNA affinity for desialylated BARF1 protein. O-Glycosylation was confirmed by the effects of cell treatment with monensin or benzyl--GalNAc, which induced the production of a 27 kDa polypeptide co-migrating with secreted BARF1 protein digested by neuraminidase. Although the existence of a superficial O-glycan linked to threonine 169 was consistent with our crystallographic data (Tarbouriech et al., 2006), the exact site of BARF1 protein O-glycosylation should be confirmed by directed mutagenesis.
Given the various possible roles of protein glycans (Mitra et al., 2006) and the requirement for both N- and O-linked glycosylation for BARF1 protein secretion, the carbohydrate side chains of the BARF1 polypeptide should be characterized fully, as secreted BARF1 protein mitogenic activity (Sall et al., 2004) and functional binding to human colony-stimulating factor 1 (Strockbine et al., 1998) both suggest that secretion might be an important pathway for BARF1 functions in oncogenicity and/or immunomodulation.
|
ACKNOWLEDGEMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTMAIN TEXTREFERENCES |
|---|
Cals, M. M., Guenzi, S., Carelli, S., Simmen, T., Sparvoli, A. & Sitia, R. (1996). IgM polymerization inhibits the Golgi-mediated processing of the mu-chain carboxy-terminal glycans. Mol Immunol 33, 15–24.[CrossRef][Medline]
Carpentier, M., Morelle, W., Coddeville, B., Pons, A., Masson, M., Mazurier, J. & Legrand, D. (2005). Nucleolin undergoes partial N- and O-glycosylations in the extranuclear cell compartment. Biochemistry 44, 5804–5815.[CrossRef][Medline]
Cohen, J. I. & Lekstrom, K. (1999). Epstein–Barr virus BARF1 protein is dispensable for B-cell transformation and inhibits alpha interferon secretion from mononuclear cells. J Virol 73, 7627–7632.
Combet, C., Blanchet, C., Geourjon, C. & Deleage, G. (2000). NPS@: network protein sequence analysis. Trends Biochem Sci 25, 147–150.[CrossRef][Medline]
de Turenne-Tessier, M., Jolicoeur, P. & Ooka, T. (1997). Expression of the protein encoded by Epstein–Barr virus (EBV) BARF1 open reading frame from a recombinant adenovirus system. Virus Res 52, 73–85.[CrossRef][Medline]
de Turenne-Tessier, M., Jolicoeur, P., Middeldorp, J. M. & Ooka, T. (2005). Expression and analysis of the Epstein–Barr virus BARF1-encoded protein from a tetracycline-regulatable adenovirus system. Virus Res 109, 9–18.[CrossRef][Medline]
Decaussin, G., Sbih-Lammali, F., de Turenne-Tessier, M., Bouguermouh, A. & Ooka, T. (2000). Expression of BARF1 gene encoded by Epstein–Barr virus in nasopharyngeal carcinoma biopsies. Cancer Res 60, 5584–5588.
Evans, A. G., Moorman, N. J., Willer, D. O. & Speck, S. H. (2006). The M4 gene of 
HV68 encodes a secreted glycoprotein and is required for the efficient establishment of splenic latency. Virology 344, 520–531.[CrossRef][Medline]
Julenius, K., Molgaard, A., Gupta, R. & Brunak, S. (2005). Prediction, conservation analysis and structural characterization of mammalian mucin-type O-glycosylation sites. Glycobiology 15, 153–164.
Luo, B., Wang, Y., Wang, X. F., Liang, H., Yan, L. P., Huang, B. H. & Zhao, P. (2005). Expression of Epstein–Barr virus genes in EBV-associated gastric carcinomas. World J Gastroenterol 11, 629–633.[Medline]
Maley, F., Trimble, R. B., Tarentino, A. L. & Plummer, T. H., Jr (1989). Characterization of glycoproteins and their associated oligosaccharides through the use of endoglycosidases. Anal Biochem 180, 195–204.[CrossRef][Medline]
Mardassi, H., Gonin, P., Gagnon, C. A., Massie, B. & Dea, S. (1998). A subset of porcine reproductive and respiratory syndrome virus GP3 glycoprotein is released into the culture medium of cells as a non-virion-associated and membrane-free (soluble) form. J Virol 72, 6298–6306.
Meads, M. B. & Medveczky, P. G. (2004). Kaposi's sarcoma-associated herpesvirus encoded viral interleukin-6 is secreted and modified differently than human interleukin-6: evidence for a unique autocrine signaling mechanism. J Biol Chem 279, 51793–51803.
Mitra, N., Sinha, S., Ramya, T. N. C. & Surolia, A. (2006). N-linked oligosaccharides as outfitters for glycoprotein folding, form and function. Trends Biochem Sci 31, 156–163.[CrossRef][Medline]
Nagano, M., Stübiger, G., Marchetti, M., Gmeiner, G., Allmaier, G. & Reichel, C. (2005). Detection of isoforms of recombinant human erythropoietin by various plant lectins after isoelectric focusing. Electrophoresis 26, 1633–1645.[CrossRef][Medline]
Ng, A., Tscharke, D. C., Reading, P. C. & Smith, G. L. (2001). The vaccinia virus A41L protein is a soluble 30 kDa glycoprotein that affects virus virulence. J Gen Virol 82, 2095–2105.
Ooka, T. (2005). Biological role of the BARF1 gene encoded by Epstein–Barr virus. In Epstein–Barr Virus, pp. 613–630. Edited by E. S. Robertson. Norfolk, UK: Caister Academic Press.
Pager, C. T., Wurth, M. A. & Dutch, R. E. (2004). Subcellular localization and calcium and pH requirements for proteolytic processing of the Hendra virus fusion protein. J Virol 78, 9154–9163.
Ressing, M. E., van Leeuwen, D., Verreck, F. A., Keating, S., Gomez, R., Franken, K. L., Ottenhoff, T. H., Spriggs, M., Schumacher, T. N. & other authors (2005). Epstein–Barr virus gp42 is posttranslationally modified to produce soluble gp42 that mediates HLA class II immune evasion. J Virol 79, 841–852.
Rudd, P. M. & Dwek, R. A. (1997). Glycosylation: heterogeneity and the 3D structure of proteins. Crit Rev Biochem Mol Biol 32, 1–100.[Medline]
Salazar, C. & Höfer, T. (2007). Versatile regulation of multisite protein phosphorylation by the order of phosphate processing and protein-protein interactions. FEBS J 274, 1046–1061.[CrossRef][Medline]
Sall, A., Caserta, S., Jolicoeur, P., Franqueville, L., de Turenne-Tessier, M. & Ooka, T. (2004). Mitogenic activity of Epstein–Barr virus-encoded BARF1 protein. Oncogene 23, 4938–4944.[CrossRef][Medline]
Seto, E., Yang, L., Middeldorp, J., Sheen, T. S., Chen, J. Y., Fukayama, M., Eizuru, Y., Ooka, T. & Takada, K. (2005). Epstein–Barr virus (EBV)-encoded BARF1 gene is expressed in nasopharyngeal carcinoma and EBV-associated gastric carcinoma tissues in the absence of lytic gene expression. J Med Virol 76, 82–88.[CrossRef][Medline]
Spiro, R. G. (2002). Protein glycosylation: nature, distribution, enzymatic formation, and disease implications of glycopeptide bonds. Glycobiology 12, 43R–56R.
Strockbine, L. D., Cohen, J. I., Farrah, T., Lyman, S. D., Wagener, F., DuBose, R. F., Armitage, R. J. & Spriggs, M. K. (1998). The Epstein–Barr virus BARF1 gene encodes a novel, soluble colony-stimulating factor-1 receptor. J Virol 72, 4015–4021.
Tanner, J. E., Wei, M. X., Alfieri, C., Ahmad, A., Taylor, P., Ooka, T. & Menezes, J. (1997). Antibody and antibody-dependent cellular cytotoxicity responses against the BamH1 A rightward open-reading frame-1 protein of Epstein–Barr virus (EBV) in EBV-associated disorders. J Infect Dis 175, 38–46.[Medline]
Tarbouriech, N., Ruggiero, F., de Turenne-Tessier, M., Ooka, T. & Burmeister, W. P. (2006). Structure of the Epstein–Barr virus oncogene BARF1. J Mol Biol 359, 667–678.[CrossRef][Medline]
Tsuiji, H., Takasaki, S., Sakamoto, M., Irimura, T. & Hirohashi, S. (2003). Aberrant O-glycosylation inhibits stable expression of dysadherin, a carcinoma-associated antigen, and facilitates cell-cell adhesion. Glycobiology 13, 521–527.
Van den Steen, P., Rudd, P. M., Dwek, R. A. & Opdenakker, G. (1998). Concepts and principles of O-linked glycosylation. Crit Rev Biochem Mol Biol 33, 151–208.[Medline]
Wahl-Jensen, V., Kurz, S. K., Hazelton, P. R., Schnittler, H. J., Stroher, U., Burton, D. R. & Feldmann, H. (2005). Role of Ebola virus secreted glycoproteins and virus-like particles in activation of human macrophages. J Virol 79, 2413–2419.
Wang, L., Tam, J. P. & Liu, D. X. (2006a). Biochemical and functional characterization of Epstein–Barr virus-encoded BARF1 protein: interaction with human hTid1 protein facilitates its maturation and secretion. Oncogene 25, 4320–4331.[CrossRef][Medline]
Wang, Q., Tsao, S. W., Ooka, T., Nicholls, J. M., Cheung, H. W., Fu, S., Wong, Y. C. & Wang, X. (2006b). Anti-apoptotic role of BARF1 in gastric cancer cells. Cancer Lett 238, 90–103.[CrossRef][Medline]
Young, L. S. & Rickinson, A. B. (2004). Epstein–Barr virus: 40 years on. Nat Rev Cancer 4, 757–768.[CrossRef][Medline]
zur Hausen, A., Brink, A. A., Craanen, M. E., Middeldorp, J. M., Meijer, C. J. & van den Brule, A. J. (2000). Unique transcription pattern of Epstein–Barr virus (EBV) in EBV-carrying gastric adenocarcinomas: expression of the transforming BARF1 gene. Cancer Res 60, 2745–2748.
Received 5 April 2007;
accepted 31 May 2007.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | MICROBIOLOGY | J GEN VIROL |
| J MED MICROBIOL | ALL SGM JOURNALS | |
