Previous Article | Next Article ![]()
Journal of Virology, March 2003, p. 3859-3865, Vol. 77, No. 6
0022-538X/03/$08.00+0 DOI: 10.1128/JVI.77.6.3859-3865.2003
Copyright © 2003, American Society for Microbiology. All Rights Reserved.
Laboratoire de Virologie Moléculaire, UMR5537, CNRS, Faculté de Médecine R.T.H. Laennec, Université Claude Bernard Lyon-1, 69372 Lyon, France,1 Institute for Genetic Medicine, Hokkaido University, Sapporo, Japan2
Received 23 July 2002/ Accepted 5 December 2002
|
|
|---|
|
|
|---|
EBV-positive Akata BL cells spontaneously lost the EBV genome during long-term culture, becoming nontumorigenic, while parental EBV-positive Akata cells induced tumors in SCID mice, suggesting an important oncogenic role of EBV in these cells (8). In fresh BL biopsy tissues and early-passage BL-derived cell lines, only EBNA1, EBERs, and BARF0 are expressed and these are classified as type I latency genes. In the EBV-positive Akata BL cell line, three of the genes mentioned above plus LMP2A were expressed (7, 12). Transfection of the EBNA1 gene into EBV-negative Akata cells did not restore tumorigenicity to Akata cells. Recent data showed that transfection of EBERs, the second viral gene expressed in EBV-positive Akata cells and in EBV-negative Akata cells, induced tumor formation in SCID mice (7, 11). In these EBERs-expressing cells, activation of the antiapoptotic Bcl2 gene was observed (7) but this activation was not confirmed by another laboratory (12). In our previous study, we reported that about 1 to 2% of EBV-positive Akata cells expressed a lytic antigen (22). In these cells, we observed expression of the BARF1 gene localized in the BamHI A fragment of the viral genome (23).
The BARF1 gene, encoding a 31-kDa early protein, has been shown to be able to induce malignant transformation in BALB/c3T3 cells and in the human B-cell line Louckes (15, 19, 20). Injection of BARF1-expressing BALB/c3T3 cells into newborn rats resulted in the induction of aggressive tumors, while injection of BARF1-expressing Louckes human EBV-negative B cells into the same mouse induced the formation of a small tumor that regressed after 3 weeks. In addition, this viral gene was capable of immortalizing primary epithelial cells (20) and was expressed in NPC biopsy tissues (4). Thus, the BARF1 gene has oncogenic activity (9). Expression of the BARF1 gene in the EBV-negative Louckes cell line induces expression of the proto-oncogene c-myc and B-cell activation antigens CD23 and CD21 (10). However, the known activities of BARF1 are not limited to its oncogenic ability; BARF1 has also been shown to serve as a target for antibody-dependent cytotoxicity (18). Recent reports have shown that this viral protein could play a role as a soluble receptor for human colony-stimulating factor 1 (16) and could regulate the immune response by inhibiting alpha interferon secretion from mononuclear cells (3).
We wondered if BARF1 is capable of inducing a tumoral transformation when it is expressed in EBV-negative Akata cells.
Activation of BARF1 gene expression in tumors. Since 1 to 2% of EBV-positive Akata cells expressed early proteins (22), we asked whether BARF1 is expressed in this cell line. To detect BARF1 expression, we carried out reverse transcription RT-PCR and an immunoblot assay. RT-PCR analysis was carried out with RNA extracted from EBV-positive Akata, M81, B95-8, and P3HR-1 cells. The RT-negative control came from a PCR without reverse transcriptase (Fig. 1 right part). The presence of BARF1 transcription was clear not only in EBV-positive Akata cells but also in other EBV-positive B-cells (Fig. 1, left part). We were then interested in determining whether the BARF1 gene is expressed in tumors induced by injection of EBV-positive Akata cells in SCID mice. After 3 months of Akata cell injection, we harvested tumor biopsy tissues and examined the expression of early proteins like BARF1, ZEBRA, EA-D, and BALF2 by both the Western blot and immunofluorescence assay methods. Figure 2 shows representative immunoblots of protein extracted from tumors induced by EBV-positive Akata cells. Translational expression of BARF1 was very weak in EBV-positive Akata cells (Fig. 2A, lane 2) and absent in EBV-negative Akata cells (Fig. 2A, lane 1), as well as in another EBV-negative B-cell line, Louckes (Fig. 2A, lane 5), while its expression became more important in two EBV-positive Akata cell-derived tumors (Fig. 2A, lanes 3 and 4). Expression of EA-D, ZEBRA, and BALF2 was weakly positive in cell cultures but became significant in tumors, while expression of the EBV gp220-320 late protein was never detected in tumors or in Akata cell cultures (Fig. 2B). Expression of the early proteins was also analyzed by using an immunofluorescence assay (6) with anti-BARF1, anti-ZEBRA, anti-EA-D, and anti-BALF2 antibodies (Fig. 3). Surprisingly, a large proportion of cells from tumors effectively expressed the above-mentioned early proteins. Before injection, about 1% of EBV-positive Akata cells expressed lytic proteins (BARF1, EA-D, BALF2, and ZEBRA) while a large proportion of tumor-derived cells became positive for early antigens (more than 50% for BARF1, ZEBRA, and EA-D and 30% for BALF2). EBV-positive Akata cells do not expressed LMP1 as a type I latency protein. We also observed no reactivation of LMP1 in tumors derived from EBV-positive Akata cells. Neither RT-PCR nor immunoblot analysis showed any positive response in cell cultures or in tumors (Fig. 2B).
![]() View larger version (15K): [in a new window] |
FIG. 1. Transcriptional expression of BARF1 in EBV-positive Akata cells. BARF1 expression was examined by RT-PCR in EBV-positive Akata cells, as well as in other EBV-positive B-cell lines. Total RNA was extracted from EBV-positive Akata, M81 (B cells immortalized with NPC-derived EBV), B95-8, and P3HR1 cells. Following treatment with DNase I, RNA was reverse transcribed and amplified with PCR primers specific for the BARF1 gene (5'-GGGGATCCCAGAGCAATGGCCAGGTTC-3' and 5'-GGGGATCCAAGGTGAAATAGGCAAGTGCG-3'). To determine whether the RNA preparation was contaminated with DNA, a PCR was carried out without reverse transcriptase (RT-). PCR products were hybridized by Southern blot analysis with a BARF1-specific probe as previously described (14). Amplified fragments were electrophoresed, transferred onto a nylon filter, and then hybridized with the 32P-labeled BARF1 sequence as the probe.
|
![]() View larger version (42K): [in a new window] |
FIG. 2. Expression of BARF1 and latent, early, and late antigens encoded by EBV in tumors induced after injection of EBV-positive Akata cells. (A) Western blot analysis of BARF1 expression in tumors A and B produced by distinct SCID mice (lanes 3 and 4) and in EBV-positive and EBV-negative Akata cells (lanes 1 and 2). A xenograft of NPC C18 expressing a high level of BARF1 protein (as previously described [4]) was used as a positive control (lane 6). Another EBV-negative B-cell line, Louckes, was used as a negative control (lane 5). (B) Expression of BZLF1, BALF2, EA-D, gp350, and LMP1 in tumor cells was analyzed by Western blotting with antibodies against the BZLF1 (OT20), EA-D (OT13J), gp350, LMP1, and BALF2 (OT13B) proteins. B95-8 treated with TPA-SB was used as a positive control (lane 1).
|
![]() View larger version (76K): [in a new window] |
FIG. 3. Immunofluorescence analysis of EBV proteins in tumors. Akata cells were extracted from tumors, washed twice with PBS, deposited onto a slide, and fixed with acetone for 15 min (right side). EBV-positive Akata cells before injection are on the left side. The slides were treated with antibodies against the BARF1 (anti-Pep 2A serum, a polyclonal rabbit antiserum prepared against a synthetic peptide corresponding to a presumed epitope, amino acids 172 to 180 [NGGVMKEKD], of the BARF1 protein) (A), BZLF1(OT20) (B), EA-D (OT13J) (C), and BALF2 (OT13B) (D) proteins.
|
![]() View larger version (46K): [in a new window] |
FIG. 4. Expression of BARF1 and Bcl2 in EBV-positive Akata cells and BARF1 transfectants. (A) BARF1 expression was analyzed by the RT-PCR method. P3HR-1 induced by TPA-SB was used as a positive control. RT-PCR was carried out as described in the legend to Fig. 1. The size of the amplified fragment was 668 bp for the BARF1 gene. EBV-negative Akata cells transfected with the pZip vector were used as a negative control. (B) Expression of BARF1 protein in EBV-positive Akata cells and in BARF1 transfectants was analyzed by Western blotting with a Pep-1 antibody, a polyclonal rabbit antiserum prepared against a synthetic peptide corresponding to a presumed epitope, amino acids 203 to 209 (GKNDKEE), of the BARF1 protein (4). Two xenografts of NPC C17 and C18 were used as positive controls for BARF1 detection (lanes 1 and 8). A 31-kDa protein corresponding to the BARF1 protein was found in C18 and C17 at a high intensity, but BARF1 transfectants showed a weak response. (C) Activation of Bcl2 expression by the BARF1 gene. Bcl2 expression was analyzed in vector DNA-transfected Akata cells (lane 1), EBV-negative Akata cells (lane 2), EBV-positive Akata cells (lane 3), and BARF1 transfectants pZC55A, pZC55B, pZC55C, and pZCC55D (lanes 4 to 7). Jurkat cells (lane 8) were used as a positive control, giving a band of 26 kDa corresponding to the Bcl2 protein. ß-Actin was used as the loading control.
|
Growth in soft agar. We analyze first the growth properties of four BARF1 transfectants by examining their ability to form colonies in soft agar (SeaPlaque; FMC Bioproducts). Colony formation efficiency was determined by three independent experiments for each BARF1 transfectant. Cells (0.2 x 106) were cultured in complete medium containing 0.33% agarose as already described (15). High cloning efficiency was observed in three BARF1-expressing clones. The averages of these experiments were 52% for pZC55A, 71% for pZC55C, and 64% for pZC55D(Fig. 5, right side). The average obtained for pZC55B was 14%. No colony formation was observed in cells transfected by vector DNA under the same condition (Fig. 5, left side).
![]() View larger version (75K): [in a new window] |
FIG. 5. Soft-agar assay. Anchorage dependence of cell growth was tested in 0.33% agarose in RPMI 1640 medium with 10% FCS as previously described (15). Seven milliliters of 0.66% agarose (SeaPlaque FMC) in RPMI 1640 medium was layered into a 10-cm-diameter dish. Cells (0.2 x 106) suspended in 4 ml of complete medium containing 0.33% agarose were layered over the 0.66% agarose underlayer. Cultures were fed with complete medium twice a week. The colonies were photographed after 4 weeks. (Right) Vector DNA-transfected, EBV-negative Akata cells. (Left) BARF1-transfected, EBV-negative Akata cell clone pZC55D.
|
|
View this table: [in a new window] |
TABLE 1. Tumorigenicity of EBV-negative Akata cells transfected with BARF1 in SCID micea
|
![]() View larger version (41K): [in a new window] |
FIG. 6. Sensitivity to apoptosis of EBV-negative Akata cells expressing the BARF1 gene. Cells were washed twice in serum-free medium and resuspended at a concentration of 5 x 103/ml in medium containing 10 or 0.1% FCS. Cells were then harvested from the cultures after 4 days. The viability of EBV-negative Akata cells, EBV-positive Akata cells, Akata cells transfected with the vector alone (pZip-Akata), and four BARF1 transfectants (pZC55A, pZC55B, pZ55C, and pZC55D) was assayed by trypan blue exclusion (left side). All of the cell clones showed good viability. The kinetics of cell viability (right side) were determined in the presence of 0.1% FCS for 4 days, and the percentage of viable cells was determined every day. The values shown are averages of three similar experiments. All BARF1 transfectants and EBV-positive Akata cells were more resistant with 0.1% of serum than were EBV-negative Akata cells (Akata EBV-) or vector DNA-transfected, EBV-negative Akata cells (pZip-Akata).
|
![]() View larger version (85K): [in a new window] |
FIG. 7. Gel electrophoretic analysis of DNAs extracted from cells. EBV-negative (lane 6) and -positive (lane 5) Akata cells and BARF1 transfectants (pZC55A, pZC55B, pZ55C, and pZC55D) were cultured for 3 days in either 0.1% FCS (lanes 1 to 6) or 10% FCS (lane 7). For analysis of apoptotic cell death, 106 cells were washed twice with PBS and harvested by centrifugation. The pellet was lysed in a solution containing 1% (wt/vol) sodium dodecyl sulfate, 0.5 mg of proteinase K (Boehringer) per ml, and 10 mM Tris-HCl (pH 8.0); incubated for 12 h at 37°C; and then incubated with RNase for a further 1 h. Protein was subsequently precipitated by addition of 5 volumes of a saturated NaCl solution. The DNA was precipitated with 3 volumes of ethanol at -20°C. A 1.0-µg DNA sample was loaded onto a 1.5% agarose gel containing ethidium bromide, electrophoresed, and photographed under UV light.
|
From our results obtained with BARF1-transfected, EBV-negative Akata cells, the subclones expressing a high level of p31 BARF1 protein become tumorigenic and resistant to apoptosis induced by serum starvation. A subclone expressing very low p31 levels was less tumorigenic and entered into apoptosis, as did EBV-positive Akata cells. EBV-negative Akata cells and EBV-negative Akata cells transfected with the pZip vector alone entered into apoptosis more rapidly than the others. Thus, BARF1 put EBV-negative Akata cells into an antiapoptotic state under serum starvation conditions. Bcl2 activation, apoptotic resistance, and tumorigenicity in SCID mice are therefore likely due to BARF1 expression. In particular, the activation of Bcl2 expression could come from the BARF1 gene because a similar activation of Bcl2 expression was observed in BARF1-transfected BALB/c3T3 cells (15). In contrast to BARF1-transfected subclones expressing a high level of p31, one of the BARF1 subclones (pZC55B) expressing a very low level of p31 protein activates Bcl2 less efficiently. These cells were less resistant to apoptosis in 0.1% serum, were poorly tumorigenic in SCID mice, and had more fragmentation of DNA in 0.1% serum. These results reinforce the antiapoptotic and oncogenic roles of BARF1 in Akata cells. In our experiments, it was difficult to show a significant difference in Bcl2 expression between EBV-positive and -negative Akata cells while we did observe significant activation of Bcl2 protein in BARF1-transfected cells. The level of Bcl2 activation with the BARF1 gene was about three- to fourfold (values estimated after scanning with a densitometer) in three BARF1-transfected, EBV-negative Akata cells compared with those obtained from EBV-positive control Akata cells. These results suggest that BARF1 is able to activate Bcl2 not only in B cells but also in rodent fibroblasts (15). It is worth noting that it was very difficult to obtain tumors when we injected EBV-positive and BARF1-transfected Akata cells into nude mice. Even with SCID mice, the development of tumors with these Akata cells took at least 12 to 16 months. A similar phenomenon was also observed with EBERs-transfected Akata cells (7). Since EBERs and BARF1 are both tumorigenic in SCID mice, it will be important to examine their oncogenic role by using a recombinant virus defective for one or both of these genes.
This work was supported by the Association pour la Recherche sur le Cancer (ARC) and the Programme de Recherche Fondamentale en Microbiologie, Maladies Infectieuses et Parasitaires (Ministère de l'Education National de la Recherche et de la Technologie, France).
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»