Upregulation of LMP1 Expression by Histone Deacetylase Inhibitors in ...

3 downloads 54 Views 237KB Size Report
Upregulation of LMP1 Expression by Histone Deacetylase Inhibitors in an. EBV Carrying NPC Cell Line. JUN NISHIKAWA,1,2 LORAND L. KIS,1 ANQUAN LIU,1 ...
Virus Genes 28:1, 121–128, 2004 Ó 2004 Kluwer Academic Publishers. Manufactured in The Netherlands.

Upregulation of LMP1 Expression by Histone Deacetylase Inhibitors in an EBV Carrying NPC Cell Line JUN NISHIKAWA,1,2 LORAND L. KIS,1 ANQUAN LIU,1 XIANGNING ZHANG,1 MIKI TAKAHARA,1 KENTARO BANDOBASHI,1 CSABA KISS,1 NOEMI NAGY,1 KIWAMU OKITA,2 GEORGE KLEIN1 & EVA KLEIN1,* 2

1 Microbiology and Tumor Biology Center, Karolinska Institute, S-171 77 Stockholm, Sweden Department of Gastroenterology & Hepatology, Yamaguchi University, School of Medicine, Minami-Kogushi 1-1-1, Ube, Yamaguchi 755-8505, Japan

Received August 27, 2003; Accepted September 14, 2003

Abstract. Objectives: In about 60% of Epstein–Barr virus (EBV) carrying nasopharyngeal carcinomas (NPC) LMP1 expressing cells can be detected. The frequency of LMP1 positive cells and the expression level varies from cell to cell in the different tumors. Cell lines derived from EBV positive NPCs loose the virus during in vitro culture. The in vitro infected NPC cell line TWO3-EBV used in our study carries the neomycin-resistance gene containing EBV and expresses low level of LMP1. With this cell line it was thus possible to study the regulation of LMP1 expression by modification of chromatin acetylation state. Study design: The TWO-EBV cell line was treated with n-butyrate (NB) or trichostatin A (TSA). Results: Shown by immunoblotting, the LMP1 level was elevated in the treated samples. Already 2 h after TSA exposure LMP1 expression was higher and it increased up to 24 h. Immunofluorescence staining showed that nearly all cells were LMP1 positive. Neither EBNA2 nor BZLF1 were induced. Tested first 2 h after the treatment, acetylated histone H3 and H4 were already detectable, and their level increased up to 8 h. Chromatin immunoprecipitation (ChIP) verified that the LMP1-promoter (LMP1p) (ED-L1) was acetylated after TSA treatment. Conclusion: EBV carrying epithelial cells do not express EBNA-2. We showed that LMP1 expression was upregulated by histone deacetylase inhibitors in an in vitro infected, EBV carrier NPC cell line. Key words: Epstein–Barr virus (EBV), histone acetylation, histone deacetylase inhibitor, latent membrane protein 1(LMP1), nasopharyngeal carcinoma (NPC)

Introduction Epstein–Barr virus (EBV) genome can be detected in several malignancies, both of lymphoid and epithelial cell origin [1]. The expression of the virally encoded proteins differs in the various tumor cell types. In the in vitro transformed B cells 9 EBV encoded proteins are expressed (six nuclear; *Author for all correspondence: E-mail: [email protected]

EBNAs and three membrane associated; LMPs). B cell malignancies exhibit 3 patterns of EBV latency, such as type I – expressing only EBNA1 protein – in Burkitt lymphomas (BL); type II – expressing EBNA1 and LMP1, LMP2s – in Hodgkin lymphoma (HL), and type III – expressing EBNA1-6 and LMP1 and LMP2s – in immunoblastic lymphomas which arise in immunocompromised patients. In EBV carrying T/NK cell lymphomas and in nasopharyngeal carcinomas (NPC), the cell virus interaction is type II [2].

122

Nishikawa et al.

Thus they express LMP1. The role of EBV in the malignant transformation in HL, in T/NK lymphoma and in NPC is unknown. LMP1 is essential for the EBV-induced immortalization of B lymphocytes in vitro [3]. In epithelial cells transfected LMP1 can block differentiation [4] and in keratinocytes it can induce morphologic transformation [5]. About 60% of NPCs that carry EBV express LMP1 with highly variable frequencies of positive cells in the tumor tissues. In addition, the LMP1 expression level in the individual cells varies also [6]. It was suggested that the methylation status of the LMP1 promoter (LMP1p) regulates the control of LMP1 expression in the tumor. The protein was detected in NPCs with unmethylated LMP1p and tumors with highly methylated LMP1p were found to be LMP1 negative [7]. When propagated in vitro or in nude mice the EBV positive NPCs were found to loose the virus. Histone acetylation regulates gene expression through opening the chromatin structure for transcription. The dynamics of histone acetylation– deacetylation and consequently the gene transcription is regulated by histone acetyltransferases and histone deacetylases (HDAC) [8]. The TWO3-EBV subline of a NPC cell line, was established by in vitro infection with a neomycin resistance gene carrying EBV strain [9]. This cell line expresses only a limited number of EBV genes, such as EBNA1, EBV-encoded small RNAs (EBERs), transcripts from the BamHI-A region (BARFO), latent membrane protein 2A (LMP2A) and a very low level of LMP1, in that the line conforms with the NPC tumors, expressing a type II EBV latency pattern. In B cells LMP1 expression is modulated by EBVencoded proteins, of which EBNA-2 plays a pivotal role, and by cellular factors that are part of the general signal transduction pathways. Epithelial cells do not express EBNA2. Therefore, the TWO3-EBV cell line provided the opportunity to study the modulation of EBNA2 independent LMP1 expression. Our results showed that acetylation of the chromatin by treatment with histone deacetylase inhibitors (HDACI), n-butyrate (NB) and trichostatin A (TSA) markedly raised the LMP1 level. NB, but not TSA treated cells entered the lytic cycle. Chromatin immunoprecipitation (ChIP) [10]

verified that the ED-L1 LMP1 promoter (LMP1p) was acetylated after TSA treatment.

Materials and Methods Cell Culture and HDACI Treatment An EBV-positive subline of an EBV-negative nasopharyngeal carcinoma cell line TWO3 was established by infection with a neomycin resistance gene carrying EBV recombinant. It was designated as TWO3-EBV [9]. All cells are EBNA1 positive when the culture is kept under selection by G418 (500 lg/ml). TWO3-EBV cells were incubated with various concentrations of NB and TSA. For time course experiments and the assay of NFkB activity and MHC class I/II expression, the cells were treated with 400 ng/ml TSA and 5 mM NB. For protein synthesis inhibitor studies cells were grown in medium containing 10 lg/ml CHX alone or CHX and TSA and were harvested at the indicated timepoints. Western Blotting Cells were lysed in SDS-polyacrylamide gel electrophoresis loading buffer, sonicated, and boiled for 5 min. A volume of lysate equal to 105 cells was separated in 10% polyacrylamide gels and transferred to PVDF membranes (BioRad). After overnight blocking with 5% nonfat dry milk in PBS (PBS-M), the filter was incubated for 2 h at room temperature with primary antibody in PBSM, washed three times with PBS containing 0.1% Tween 20 (PBS-T), and then reacted for 30 min with horseradish peroxidase-conjugated second antibody (diluted 1:2000 in PBS-M). After the second antibody reaction, the filters were washed five times with PBS-T, immersed in the enhanced chemiluminescence solutions (Amersham Biosciences) as specified by the manufacturer, and subjected to autoradiography. The antibodies used were anti-EBNA2 mAb (PE2, Dako), anti-LMP1 antibody (S12), anti-BZLF1 antibody (BZ1, Dako) and anti-acetylated histone H3 (06-599) and histone H4 (06-866) (Upstate Biotechnology, Lake Placid, NY).

LMP1 Induction by Histone Acetylation

Immunofluorescence Staining The cells were centrifuged on glass slides in cytospin cytocentrifuge at 1,200 rpm for 5 min and fixed in methanol:acetone (1:1). After rehydratation in PBS the slides were incubated with the antiLMP1 monoclonal Ab mixture CS1-4 (DAKO) (dilution 1:200) for 1 h at room-temperature. After washing, the slides were incubated with FITC conjugated rabbit-anti-mouse Ab (DAKO, Copenhagen, Denmark) (dilution 1:200) for 30 min at room temperature. The nuclei were visualized with 0.5 g/ml Hoechst 33258 that was added to the conjugated antibodies. The slides were mounted with 70% glycerol, 2.5% DABCO (Sigma), pH 8.5, in PBS. Images were generated with a Leitz DM RB microscope using a PL APO Ph X 63 oil immersion objective. The SNK-6 NK-cell lymphoma line [11] and the uninfected TWO3 cells were used as LMP1 positive and negative staining controls, respectively. Northern Blotting Thirty lg of total RNA was fractionated on 1% agarose formaldehyde gel and transferred to Hybond-N membranes (Amersham Biosciences). The probes were generated by PCR and labeled by a32P-dCTP using High Prime labeling kit (Roche). The membranes were hybridized overnight in Rapid-hyb buffer (Amersham Biosciences). After stringent washing, the membranes were exposed to Phosphorimager (Amersham Biosciences) for 2 days. ChIP Assay The ChIP assay was performed according to manufacture’s recommendation (Upstate Biotechnology). Briefly, TSA-treated and -untreated cells (107 cells) were fixed with formaldehyde to cross link histone to DNA. Fixed cells were pelleted by centrifugation and washed twice with ice-cold PBS containing protease inhibitor (1 mM phenylmethylsulfonyl fluoride (PMSF)). The cells were then resuspended in SDS/lysis buffer containing 1 mM PMSF and sonicated (three times for 10 s each) to make soluble chromatin. Samples of total chromatin were taken at this point to use as a positive control in the PCRs (input chromatin). After the

123

pre-clear of the cell lysates by incubation with Salmon Sperm DNA/protein A-Agarose, samples were incubated with antibodies against acetylated histone H3 (06-599) and histone H4 (06-866) (Upstate Biotechnology, Lake Placid, NY) for immunoprecipitations overnight at 4°C. DNA–protein complexes were collected with Salmon Sperm DNA/protein A-Agarose followed by several rounds of washing. Bound DNA-protein complexes were eluted from the antibodies with two incubations in elution buffer (1% SDS, 0.1 M NaHCO3). Cross-links were reversed by the sequential addition of proteinase K, followed by incubation at 65°C for 5 h. Samples were then extracted twice with phenol/chloroform and precipitated with ethanol. DNA fragments were recovered by centrifugation, resuspended in water, and used for PCR amplifications. To examine the degree of histone acetylation at the ED-L1 LMP1p, PCR carried out by primer pair covering the LMP1 promoter from )206 to +1 relative to the initiation site at three different cycles (27, 31 and 35 cycles). The PCR products were fractionated on 2.5% agarose gels, stained with ethidium bromide. The primers were designated on the basis of our LMP1 northern-blotting results that showed that the LMP1 mRNA in the TSA treated cells was of 2.8 kb, indicative of a transcript originating from the ED-L1 LMP1 promoter (not shown). NFkB Activity Luciferase reporter plasmid that contains three NFkB binding sites were transfected into TWO3 cells by using Lipofectamine reagent (Invitrogen) according to manufacture’s protocol. Twenty-four hours after transfection, the cells were exposed to TSA and incubated for 14 h. The luciferase activity was assayed according to the manufacturer’s instruction (Promega). Flow Cytometry For expression of MHC class I/II, we followed the procedure described in our previous publications [12]. Antibodies used in flow-cytometry were antiMHC class I antibody (W6/32) and anti-MHC class II antibody (L243). RPE-conjugated rabbit anti-mouse Ig F(ab)2 (DAKO) reagent was used as secondary antibody.

124

Nishikawa et al.

Results Increased LMP1 Expression Induced by NB and TSA In the NB and TSA treated TWO3-EBV cultures cell proliferation was arrested and cell viability

decreased. After 48 h about 60% of cells survived in the NB and 80% in the TSA treated cultures. In immunoblotting a very week LMP1 expression was seen in TWO3-EBV. After 48 h exposure to NB and to TSA, LMP1 expression was markedly elevated in a dose dependent manner (Fig. 1A). The LMP1 proteins were full-length,

Fig. 1. (A) LMP1 expression detected by Western blotting. TWO-EBV cells were treated with TSA (25, 50, 100, 200 and 400 ng/ml) and NB (0.3, 0.6, 1.2, 2.5 and 5 mM) for 48 h. LCL and uninfected TWO3 cells were used as positive respectively negative controls. (B) BZLF1 expression detected by Western blotting. TWO3-EBV cells were treated with TSA (200 and 400 ng/ml) and NB (2.5 and 5 mM) for 48 h. (C) Expression of LMP1 in the TSA treated TWO3-EBV cells, assessed by indirect immunofluorescence with the CS1-4 monoclonal Abs mixture. SNK-6 and uninfected TWO3 cells were used as positive and negative staining controls, respectively.

LMP1 Induction by Histone Acetylation

125

Fig. 2. Expression of EBV genes and acetylated histones in TSA treated cells by Western blotting. TWO3-EBV cells were treated with 400 ng/ml TSA and assayed after 2, 4, 8 and 24 h. LCL and uninfected TWO3 cells were used as positive respectively negative controls for LMP1 and EBNA-2. EBNA-2 was not detected in the TSA treated TWO3 cells.

their size being about 60 kDa. In addition, treatment with high dose of NB induced a very week band corresponding to the truncated protein, often referred to as lytic LMP1 [13] (Fig. 1A). In accordance with the known effect of NB as an inducer of EBV lytic cycle [14], expression of the EBV immediate early gene, BZLFl was detected in the NB treated culture (2.5 and 5 mM) (Fig. 1B). TSA did not induce BZLFl even if it was used in high doses (200 and 400 ng/ml); 48 h after NB treatment, 20% of cells were BZLFl positive when visualized by immunofluorescence, while less than 2% positive cells were seen in the TSA treated culture (not shown). Expression of LMP1 at single cell level was visualized by immunofluorescence. While only occasional cells were positive in the untreated samples, almost all cells were positive in the TSA treated cells. The intensity of staining varied considerably from cell to cell. The pattern of LMP1 staining differed from that of the SNK-6, NK lymphoma cells, used as positive staining control. While in the SNK-6 cells the staining was localized in caps, in the epithelial cells the distribution was even on the circumference, with the appearance of small brilliant dots (Fig. 1C).

EBV Gene Expression and Histone Acetylation in the TSA Treated Culture Time course experiments showed that already 2 h after TSA exposure, the LMP1 level was elevated and it increased till 24 h (Fig. 2). EBNA2 was not induced (not shown). Tested first 2 h after the treatment, acetylated histone H3 and H4 were already detectable, their level increased up to 8 h; at 24 h their expression level was lower (Fig. 2).

Induction of LMP1 mRNA LMP1 message was clearly detectable 4 h after TSA treatment and increased later. LMP1 message was induced by the protein synthesis inhibitor, cycloheximide (Fig. 3). In the CHX treated cells the level of LMP1 mRNA was detectable earlier and its level was considerably higher than in the TSA treated cells. Comparison of the samples treated for 8 h with CHX alone or in combination with TSA showed a weaker band in the latter. In view of the complexity of LMP1 regulation interpretation of this result needs further analysis.

126

Nishikawa et al.

Fig. 3. Expression of LMP1 mRNA in the TSA treated cells by Northern blotting. Total RNAs were extracted from cells treated with 10 lg/ml CHX alone or CHX and 400 ng/ml TSA. 18S ribosomal RNA showed that the RNAs were loaded equally.

Histone Acetylation of the LMP1 Promoter in the TSA Treated Culture In chromatin immunoprecipitation (ChIP) assay we detected acetylation of the LMP1p in the TSA treated samples (Fig. 4). The primers were designated on the basis of our LMP1 northern-blotting results that showed that the LMP1 mRNA in the TSA treated cells was of 2.8 kb, indicative of a transcript originating from the ED-L1 LMP1 promoter. The bands of 207 bp PCR product from the ED-L1 LMP1 promoter were detected at 31 cycles for ChIP with the antibody against acety-

lated histone H3 and at 27 cycles for H4. In the untreated cells only traces were seen even after 35 cycles. Increase of NFkB Activity and HLA Class I Expression in the TSA Treated Cells LMP1 is known to activate the NFkB signaling pathway and to upregulate the expression of MHC class I molecules. Therefore we attempted to assess the influence of EBV and of the TSA induced LMP1 in these two cellular properties. NFkB activity in the EBV carrying cells was higher than in

Fig. 4. ChIP assay for LMP 1p in TSA treated cells. PCR for the LMP1p carried out at three different cycles (27, 31 and 35 cycles) in TSA treated and untreated TWO3-EBV cells. Samples of total chromatin were used as positive control in the PCRs (input). Immunoprecipitated samples without antibodies were used as a negative control ()).

LMP1 Induction by Histone Acetylation

Fig. 5. NFkB activity and HLA class I expression in the TSA treated cells. (A) NFkB activity was assayed in the samples transfected with luciferase reporter construct. The mean of the activity in two independent experiments was calculated and shown as relative activity to the level found in the untreated TWO3 cells. (B) The expression of MHC class I molecule was assayed by flow-cytometry. The expression in TSA treated TWO3-EBV is shown as a bold line, untreated TWO3-EBV, dotted line. In the cells stained only with secondary antibody the plain lines overlapped.

the uninfected TWO3 cells and it was elevated 5.7fold after TSA treatment (Fig. 5A). In the TWO3EBV cells TSA induced a 4-fold increase of the NFkB activity in the untreated cells. The EBV carrier state and TSA treatment led to 85-fold increase of NFkB activity. TSA enhanced the expression of surface MHC class I molecule in the TWO3-EBV cells (Fig. 5B). The cells did not express MHC class II molecules even after treatment with TSA (not shown). Discussion In about half of EBV positive NPCs tissues LMP1 can be detected. Immunohistochemical staining

127

showed that the expression of LMP1 varies in the different NPC tumors and the intensity of staining from cell to cell is heterogeneous. Often only a small fraction of the cells were LMP1 positive [6]. Studies in our laboratory showed a correlation between the methylation of LMP1p and the absence of LMP1 protein in the tumor [7]. Survey of the clinical course indicated that the LMP1 positive cases had relatively favorable prognosis [15]. This correlation was assumed to be due to the LMP1 specific immune response. Therefore it was attempted to administrate CTLs raised against LMP1 as a therapeutic measure to patients [16]. Several studies in BL lines showed a correlation between the acetylation state of histones and the LMP1 expression [17–19]. Similarly, we found that in the EBV carrying NPC cells that shows very low level of LMP, its expression was considerably elevated by histone acetylation at the LMP1p. In B cell lines the expression of LMP1 is regulated by EBNA-2 involving the transcription factors that have histone acetyltransferase activity, such as CBP, P300 and PCAF [17]. This effect counteracts the Mad, Max and Sim3a complexed HDAC that repress the transcription of LMP1 [18]. In the EBNA-2 negative P3HR1 cells inhibition of HDAC led to the up-regulation of LMP 1 as well [19]. Our results on LMP1 expression of epithelial cells are similar, thus both in B and in epithelial cells LMP1 can be induced by histone-acetylation. The mechanism for the lack of EBNA-2 expression is different in the two cells. While in P3HR1 cells the EBNA-2 gene is deleted, the epithelial cells do not express EBNA2 because they lack the required B cell specific transcription factors [20]. Regulation of LMP1 expression is complex, involving collaboration of both cellular and EBV encoded proteins. These exert counteracting positive and negative effects. In B cells EBNA-2 plays a crucial role that involves elimination of the silencer Mad, Max, Sim3a complex. The regulation seems to take place at the level of chromatin acetylation [18]. Notch signaling was shown to substitute the role of EBNA-2 in the modulation of EBV gene expression [21]. Our finding on induction of LMP1 message in absence of protein synthesis may be due to elimination of short-lived proteins that constitute the silencer complex, and/or due to stabilization of LMP1 mRNA in these cells. Our

128

Nishikawa et al.

experiment indicated that TSA lowered the effect of CHX. This may be due to the involvement of several proteins in the activation of the LMP-1 promoter, the synthesis of which is influenced differently by these treatments. Interestingly, expression of LMP-1 induced by Notch receptor activation was shown to require CHX in a BL that lacks EBNA-2. NB is known to induce the lytic cycle in EBV carrying cells [13]. In P3HR1 cells BZLF1 message was induced by TSA at least 8 h after treatment [22]. NB but not TSA induced the EBV lytic cycle in the TWO-EBV cells. The demethylating agent 5azacytidine did not induce BZLF1 and induced only very weak LMP1 expression (data not shown). In addition to the HDAC inhibitory activity, NB affects also phosphorylation and methylation of histones and other nuclear proteins [23]. Thus induction of the lytic cycle by NB cannot be attributed solely to its effect on histone acetylation. The EBV carrier state of the TWO3 cells was shown to increase cell motility in vitro and tumorigenicity in nude mice. Our experiment shows an additional trait of the virus carrying cells, they had considerably elevated NFkB activity. As mentioned above it has been shown that methylation of the LMPp was associated with lack of LMP1 expression in NPC tissues. In our experiments with the epithelial cells that carry EBV after in vitro infection, the demethylating agent had only very week effect on LMP1 expression, while acetylation of the LMP1p induced it clearly. Interestingly the protein expression in the TSA treated cells resembled that in the tumor tissue showing considerable cell to cell variation. Whether chromatin acetylation mechanism regulates LMP1 expression in the tumor tissue remains to be established.

Acknowledgements This work was supported by The Swedish Cancer Society, Sweden, and the Cancer Research Institute/Concern Foundation.

References 1. Zur Hausen H., Schulte-Holthausen H.S., Klein G., Henle W., Henle G., Clifford P., and Santesson L., Nature 228, 1056–1058, 1970. 2. Rickinson, A., and Kieff E., in Knipe D., Howley P., Griffin D., Martin M., Lamb R., Roizman B., and Straus S. (ed.), Fields Virology, Fourth Edition, Vol. 2, Lippincott Williams and Wilkins, Philadelphia, USA, 2001, pp. 2575– 2628. 3. Kaye K.M., Izumi K.M., and Kieff E., Proc Natl Acad Sci USA 90, 9150–9154, 1993. 4. Dawson, C.W., Rickinson A.B., and Young L.S., Nature 344, 777–780, 1990. 5. Fahraeus, R., Rymo L., Rhim J.S., and Klein G., Nature 345, 447–449, 1990. 6. Niedobitek G., Meru N., and Delecluse H.J., Int J Exp Pathol 82, 149–170, 2001. 7. Hu L.F., Minarovits J., Cao S.L., Contreras-Salazar B., Rymo L., Falk K., Klein G., and Ernberg I., J Virol 65, 1558–1567, 1991. 8. Luo R.X. and Dean DC., J Natl Cancer Inst 91, 1288– 1294, 1999. 9. Teramoto N., Maeda A., Kobayashi K., Hayashi K., Oka T., Takahashi K., Takada K., Klein G., and Akagi T., Lab Invest 80, 303–312, 2000. 10. Braunstein M., Sobel R.E., Allis C.D., Turner B.M., and Broach J.R., Mol Cell Biol 16, 4349–4356, 1996. 11. Nagata H., Konno A., Kimura N., Zhang Y., Kimura M., Demachi A., Sekine T., Yamamoto K., and Shimizu N. Blood 97, 708–713, 2001. 12. Nagy N., Maeda A., Bandobashi K., Kis L.L., Nishikawa J., Trivedi P., Faggioni A., Klein G., and Klein E., Int J Cancer 100, 433–440, 2002. 13. Erickson K.D. and Martin J.M., Virology 234, 1–13, 1997. 14. Kallin B., Luka J., and Klein G., J Virol 32, 710–716, 1979. 15. Hu L.F., Chen F., Zhen Q.F., Zhang Y.W., Luo Y, Zheng X., Winberg G., Ernberg I., and Klein G., Eur J Cancer 31, 658–660, 1995. 16. Gottschalk S., Edwards O.L., Sili U., Huls M.H., Goltsova T., Davis A.R., Heslop H.E., and Rooney C.M., Blood 101, 1905–1912, 2003. 17. Wang L., Grossman S.R., and Kieff E., Proc Natl Acad Sci USA 97, 430–435, 2000. 18. Sjoblom-Hallen A., Yang W., Jansson A., and Rymo L., J Virol 73, 2983–2993, 1999. 19. Park J.H., and Faller D.V., Virology 303, 345–363, 2002. 20. Tierney R., Kirby H., Nagra J., Rickinson A., and Bell A., J Virol 74, 10458–10467, 2000. 21. Strobl L.J., Hofelmayr H., Marschall G., Brielmeier M., Bornkamm G.W., and Zimber-Strobl U., J Virol 74, 1727– 1735, 2000. 22. Chang L.K., and Liu S.T., Nucleic Acids Res, 28, 3918– 3925, 2000. 23. Vigushin D.M. and Coombes R.C., Anticancer Drugs 13, 1–13, 2002.