Znt. J. Cancer: 67,86-94 (1996) 0 1996 Wiley-Liss, Inc. Publication of the International Union Against Cancer Publication de I’llnion Internationale Contre le Cancer EXPRESSION OF EPSTEIN-BARR VIRUS LATENT INFECTION GENES AND ONCOGENES IN LYMPHOMA CELL LINES DERIVED FROM PYOTHORAX-ASSOCIATED LYMPHOMA Hiroyuki K A ” O ’ , Yutaka YASUNAGA’, Masahiko OHSAWA’, Masafumi TANIWAKI?, Keiji IUCH13, Norifumi NAKA’, Masanori SHIMOYAMA“ and Katsuyuki AOZASA’,~ Kei TORIKAI’, ‘Department of Pathology, Osaka University Medical School, Suita 565; 2Department of Internal Medicine, Kyoto Prefectural Universiy of Medicine, Kyoto 602; 3Department of Suvery, National Kinki-Chuo Hospital, Sakai 591; and 4National Cancer Center Hospital, Tokyo 104, Japan. Malignant lymphomas frequently develop in the pleural cavity of patients with long-standingpyothorax. The term pyothoraxassociated lymphoma (PAL) has been proposed for this type of tumor. Most PALs are diffuse lymphomas of the B-cell type and contain Epstein-Barr virus (EBV) DNA. We have established 2 lymphoma cell lines from biopsy specimens of PAL cases, OPL- I and OPL-2. and examined their growth characteristics and the expression of EBV latent infection genes and oncogenes. OPL-2 exhibited a more rapid growth and higher saturation density than OPL- I, and only OPL-2 exhibited colony-forming activity in soft agar. OPL- I and -2 were positive for B-cell differentiation markers and showed clonal surface immunoglobulins. Both lines contained a single predominant form of episomal EBV DNA, indicating clonal cellular proliferation of an EBV-infected progenitor cell. OPL- I and -2 contained type B and A EBV genome, respectively. Expressionof EBV nuclear antigen (EBNA)2 mRNA and protein was detected by Northern and Western blot analysis in OPL-I, but not in OPL-2. On the other hand, the expression of latent membrane protein (LMP) I mRNA in both OPL-I and -2 was extremely weak and detectable only by reverse transcription-polymerasechain reaction. Proteinexpression of LMPl was not observed by Western blot analysis or immunocytochemistry. Both lines expressed c-myc mRNA. Only OPL-I expressed mRNA of c-fgr, an oncogene whose expression is upregulated by EBNA2. Both OPLs expressed bcl-2 mRNA without detectable expression of LMPl protein. o 1996 Wiley-Liss,Inc. Epstein-Barr virus (EBV) has been identified as an etiological agent of the African type of Burkitt’s lymphoma and of nasopharyngeal carcinoma (NPC) in southern China (Wolf et al., 1993). Latent infection genes of EBV, including EBV nuclear antigens (EBNAs) and latent membrane proteins (LMPs), are expressed in B cells during latent infection and lead to immortalization of B cells (Rickinson et al., 1992). Burkitt’s lymphoma cells express only EBNA1, which is not recognized as a target by virus-specific T cells, and thus can evade immune surveillance in vivo (Rickinson et al., 1992;Wolf et al., 1993). In the course of in vitro culture, however, most of them express all latent infection genes (Rickinson et al., 1992). Since EBNA2 and 3s, as well as LMPs, serve as target viral antigens for the elimination of infected cells by cytotoxic T cells (CTLs) (Rickinson et al., 1992), Burkitt’s lymphoma cells expressing latent infection antigens other than EBNAl are thought to be eliminated by CTLs in vivo (Rickinson et al., 1992). EBV infection in immunosuppressed patients, such as human immunodeficiency virus-infected patients or allograft recipients receiving long-term immunosuppressive therapy, leads to polyclonal B-cell proliferative disorders, frequently resulting in development of monoclonal malignant lymphomas (Rickinson et ab, 1992;Wolf et al., 1993). These lymphomas are most often diffuse lymphomas of the B-cell type and contain EBV DNA (Rickinson et al., 1992). Suppressed immune functions in these patients are thought to cause incomplete elimination of the cells expressing EBV latent infection genes (Rickinson et aL, 1992). Genetic lesions including chromosomal translocation and point mutation play an important role in EBV-related B-cell tumors (Gaidano and Dalla-Favera, 1993). The activation of c-myc in Burkitt’s lymphoma is induced by chromosomal translocation, not by EBV infection (Battey et al., 1983). Diffuse large cell lymphoma (DLCL) frequently develops in the pleural cavity of patients with long-standing pyothorax (pyothorax-associated lymphoma; PAL) (Iuchi et al., 1989). Although no systemic immunosuppressive condition has been identified, almost all these lymphomas contain EBV DNA (Fukayama et al., 1993). Among EBV latent infection genes, EBNA2 and LMPl are essential for the transformation of B cells (Wolf et a l , 1993), which is partly explained by oncogene activation, i.e., EBNA2 transactivates c - f ~ (Knutson, 1990), and LMPl transactivates bcl-2 (Henderson et al., 1991). Previous immunohistochemical studies have described the expression of EBNA2 and LMPl in PAL cells (Fukayama et al., 1993), although the relationships among EBV latent gene expression, oncogene expression and the growth of lymphoma cells as well as the contribution of EBV infection to the development of PALs remain unknown. In the current study, we established 2 lymphoma cell lines from patients with PAL and examined the growth characteristics and the expression of EBV latent infection genes along with the oncogenes c-myc, c-fgr and bcl-2. The expression of EBNA2 varied in the 2 lymphoma cell lines. However, a restricted expression of LMPl was common. MATERIAL AND METHODS Establishment of cell lines Small pieces of the biopsy specimens were minced with stainless steel mesh and washed several times with RPMI 1640 medium medium (Nissui, Tokyo, Japan) supplemented with 300 U/ml of penicillin and 300 pg/ml of streptomycin. Subsequently, culture (5 x lo6 cells/ml) was started in RPMI 1640 supplemented with 20% heat-inactivated FCS (ICN, Lisle, IL) at 37°C in 5% C 0 2 in air. No heterologous feeder layers were used for cell line establishment. Once cell growth was established, the concentration of FCS in the medium was reduced to 15%. Zmmunophenotypic analysis Three micrometer sections of formalin-fixed, paraffinembedded tissue and cytocentrifuged specimens of cultured cells were studied by immunochemistry for expression of cell surface marker proteins and EBV latent infection gene products. Appropriate tissue sections (lymphoid tissue and lymphoma) or cytospin slides (lymphoblastoid (LCL) and lymphoma cell lines) were also subjected to immunostaining as a positive control. The polyclonal antisera and the monoclonal ’To whom correspondence and reprint requests should be sent: Department of Pathology, Osaka University Medical School, 2-2 Yamada-oka, Suita 565, Japan. Fax: (81)6-879-3719. Received: October 27,1995 and in revised form March 1,1996. 87 PAL CELL LINES antibodies (MAbs) used in our study, together with their respective clusters of differentiation and EBV antigens, are shown in Table I. Cytogenetic analysis Chromosomal study was performed on samples obtained from the cell lines OPL-1 and -2 by the trypsin-Giemsa banding method. Well-spread met aphases were photographed and arranged according to the I-ecommendations of ISCN (1991) for cancer cytogenetics. Fluorescence in situ hybridization (FISH) with C, and VH segments of the IgH gene locus was performed as described previclusly (Taniwaki et al., 1994). Cells EBV-positive or -negative Burkitt’s lymphoma cell lines, Raji or Ramos, respectively, were (obtained from the Japanese Cancer Research Resources Bank. An LCL was established by infection of peripheral blood lyinphocytes from a healthy donor with EBV derived from B95-8 cells (a generous gift from Dr. K. Takada, Sapporo, Japan). Cells were grown in RPMI 1640 medium supplemented with 10% heat-inactivated FCS and collected at 3 x lo5 to S x lo5cells/ml (more than 95% viability) and used for nucleic acid and protein extraction. DNA extraction and Southern blot Total cellular DNA was extracted from cells and tissue specimens as described previously (Kanno et al., 1993). DNA was quantitated by measuring OD260and stored at -20°C until use. Ten micrograms of DNA were digested with Bam HI, Hind 111 or Pst I (GIBCO-BRL, Gaithersburg, MD), electro- phoresed in 0.7% agarose gels and Southern blotted as described previously (Kanno et al., 1993). The filters were hvbridized with human JHprobe or LMPl exon 2 oligonucleotrde probe (described below) for checking clonality of B cells or EBV genome, respectively. LMPl exon 2 is located in the Bam HI-NJhetfused terminal fragment, which includes about 500 bp terminal repeat (TR) of EBV (Fennewald et aL, 1984). Since reduplication of TR varies among EBV clones, the presence of a single predominant band of Barn HI-NJhet fragment represents clonal proliferation of EBV-infected cells (Raab-Traub and Flynn, 1986). mRNA extraction and Northern blot Total cellular RNA was extracted from cells by the acid guanidium-phenol chloroform method. mRNA was purified by using oligo-dT latex (Takara, Kyoto, Japan), quantitated by measuring OD260and ethanol-precipitated at -80°C until use. Five micrograms of mRNA were electrophoresed in 1% formaldehyde-agarose gels, and Northern blotted as previously described (Kanno et af., 1993). Preparation ofprobes and hybridization Human immunoglobulin JH (Nishida et al., 1982), human c-myc (Battey et al., 1983) and human c-fgr (Tronick et al., 1985) DNA probes were obtained from the Japanese Cancer Research Resources Bank. Human bcl-2 (Tsujimloto and Croce, 1986) DNA probe or murine actin DNA probe were kindly provided by Dr. Y. Tsujimoto or Dr. M. k3tagawa (Osaka University, Suita, Japan), respectively. These probes were labeled with C K - [ ~ ~ P ] - ~by C Trandom P priming. EBNAl TABLE I - IMIVIUNOPHENOTfPE OF PAL CELL LINES O p t - 1AND OPL-2 AND OF THE ORIGINAL TISSUE SAMPLES Antigen Antiserum (AS) or monoclonal antibody AS AS AS AS CD3 CD4 CDS CD8 CDlO CDlla CDllc CD15 CD19 CD20 CD21 CD22 CD23 CD24 CD30 CD43 CD45R CD45RO CD54 CD57 CD74 CDw75 EBNA2 LMPl AS AS AS SK7 OPD4 SK3 L17F12 SK1 SS2136 25.3.1 S-HCL-3 MMA 4G7 L26 H299 1F8 SHC1-I BU38 BA-1 Ber-H2 MTI 4KB5 MB1 UCHL-1 84H10 HNK-1 LN2 LN1 PE2 CS1-4 Source D D D D D D D BD D BD BD BD D IMT BD BD BD KM C D BD D BMB D BGL D BGL D IMT BD T T D D Case I OPL-I ++- ND - ND ND ND ND ND ND ND ND ND ND + ND ND ND ND ND ND - + +ND ND ++ + - ++- OPL-2 - - +- + - - ND - - ND ND ND ND ND ND ND ND - - + - ND + + + + ND ND ND ND ND ND ND - - - - - ND ND ND ND + ND ND - - + ND - +- + - + ND ND ND ND ND ND +- ++- Case 2 ++ - + + + + + + ++ +- ID, Dako (Glostrup, Denmark); BD, Becton Dickinson (Mountain View, CA); IMT, Immunotech (Marseille, France); KM, Kyowa Medix (Tokyo, Japan); C, Coulter (Hialeah, FL); BMB, Boehringer Mannheim (Indianapolis, IN); BGL, BioGenex (San Ramon, CA); T, Techniclone (Santa Ana, CA); PTD, not done. 88 KA”0 E T A L oligonucleotide probe (5’-TCCTAGGCCATTTCCAGGTCCTGTACCTGG-3’, in K exon) (Sample et al., 1986), EBNA2 oligonucleotide probe (5’-CCTATGTAACGCAAGATAGAATGTAGGCAT-3’ in YH exon) (Sample et ab, 1986) and LMPl oligonucleotide probe (5’-GATGAACAGCACAATTCCAAGGAACAATGC-3’, in exon 2) (Fennewald et a/., 1984) were 5-end-labeled with Y-[~~P]-ATP by T4 polynucleotide kinase (New England Biolabs, Beverly, MA). For the check of clonality of B cells using the human JH probe, Southern-blotted filters were prehybridized for 4 hr at 42°C in 6 x SSPE (0.9 M NaCI, 0.06 M Na phosphate, 6 mM EDTA), S X Denhardt’s solution, 0.1% SDS, 200 pg/ml denatured salmon testes DNA (Sigma, St. Louis, MO), 50% formamide and hybridized in a solution of the same composition containing the probe at 2 x lo6 cpm/ml. Filters were hybridized for 16 hr at 42”C, washed twice for 30 min at room temperature with 1x SSPE, 0.1% SDS and twice for 15 min at 50°C with 0.1X SSPE, 0.1% SDS and exposed at -80°C. For detection of oncogenes and actin mRNA, Northern-blotted filters were prehybridized, hybridized and washed in the same manner as described above with the exception of an additional 15 rnin washing at 60°C with 0 . 1 ~ SSPE, 0.1% SDS for actin mRNA. For checking clonality of the EBV genome and for detection of EBV latent infection gene expression using oligonucleotide probes, filters were prehybridized for 4 hr at 42°C in 6 x SSC (0.9 M NaCI, 0.09 M Na3 citrate), 5 X Denhardt’s solution, 0.5% SDS, 20 pg/ml denatured salmon testes DNA and hybridized in a solution containing the probe at 3 x lo6 cpmiml. Filters were hybridized for 16 hr at 42”C,washed with 2 x SSC, 0.1% SDS twice for 10 rnin at room temperature and once for 20 rnin at 50°C and exposed at -80°C. PCR ampliJication of EBNAZ gene and Southern blot analysis for the typing of EBVgenome For PCR amplification of the EBNA2 region of the EBV genome, 1 pg of DNA was diluted to a SO pl solution containing 0.2 mM dNTP, 3 mM MgCI2, 1.25 U of Taq polymerase (Promega, Madison, WI), 1X Taq buffer supplied by manufacturer and 1.5 JLM both primers. Primers are designed to amplify the 89 bp segment, both ends of which are conserved and the middle portion of which varies between type A and B EBV (Sixbey et al., 1989). Primer sequences were 5 ’-CCACCAGCAGCACCAGCACA-3’and 5 ‘-GGTGGCCACCATGGTGGCCC-3’. PCR was performed (1 cycle at 94°C for 3 min, 60°C for 1 min, 72°C for 1 rnin and 39 cycles at 94°C for 1 min, 60°C for 1 min, 72°C for 1 min) and amplified products were electrophoresed in 2% agarose gels and Southern blotted as previously described (Kanno et a[., 1993). EBNA2A- or -2B-specific oligonucleotide probes (EBNA2Aspecific,5 ’-TTACATCATCTACCCTCG-3’, EBNA2B-specific, S‘-GCACTTCCTCCAACTCCA-3’) were designed to hybridize to the variable region between the 2 primers (Sixbey eta/., 1989) and 3’-end-labeled with fluorescein-11-dUTP by using the ECL 3’-oligolabeling system (Amersham, Aylesbury, UK). Filters were hybridized with the labeled probe, washed and incubated with anti-fluorescein horseradish peroxidase conjugate. Signals were generated with ECL detection reagents (Amersham) and exposed at room temperature by following the procedures suggested by the manufacturer. Reverse transcription (RT)-PCR analysis For RT-PCR amplification of LMPl and actin mRNA, 1 pg of mRNA was converted to cDNA with 200 U of Moloney murine leukemia virus reverse transcriptase (GIBCO-BRL) in the presence of 5 mM MgC12, 1 mM dNTP, 20 U of RNase inhibitor (Promega), 500 nM downstream primer and 1X R T reaction buffer supplied by the manufacturer in 20 pl reaction volume. After incubation at 37°C for 60 min, the R T mixture was boiled for 5 min and subjected to PCR amplification. Ten microliters of the RT mixture (cDNA sample) was diluted to 50 p1 containing 0.2 mM dNTP, 1.5 mM MgCI2, 1.25 U of Taq polymerase, I X Taq buffer and 500 nM of both primers. Primer sequences were 5 ' -TGGAGCCCTTTGTATACTCC-3' (exon 1) and 5’-GATTCATGGCCAGAATCATC-3‘ (exon 3) for LMP1, amplifying the 405 bp segment of mRNA and the intervening 154 bp introns (Fennewald et al., 1984); and 5‘-ATCATGTTTGAGACCTTCAA-3’and 5’-CATCTCTTGCTCGAAGTCCA-3’ for human actin, amplifying the 318 bp segment of mRNA. PCR was performed (1 cycle at 94°C for 3 min, 58°C for 1 min, 72°C for 2 rnin and 34 cycles at 94°C for 1 min, 58°C for 1 min, 72°C for 2 min) and amplified products were electrophoresed, Southern blotted and hybridized with non-radiolabeled LMPl oligonucleotide probe; signals were then generated in the same manner as described in the previous section. In the case of actin, amplified products were examined in ethidium bromide-stained gels with an ultraviolet transilluminator. Western blot analysis Cells were directly lysed with Laemmli sample buffer and boiled for 5 min. Lysates were electrophoresed in SDS 7.5% polyacrylamide gels and blotted onto nitrocellulose membranes as described previously (Kanno et al., 1993). After incubation in 5% dried milk in PBS at room temperature overnight, membranes were incubated at room temperature for 3 hr with a 1/100 dilution of anti-EBNA2 (PE2) or with a 1/25 dilution of anti-LMP1 (CS1-4) mouse MAbs. After washing with PBS supplemented with 0.05% of Tween 20, membranes were incubated at room temperature for 2 hr with 1/400 dilution of horse anti-mouse immunoglobulins (Vector, Burlingame, CA), washed and subsequently incubated for 1 hr with 1/4,000 dilution of peroxidase-labeled protein A (Boehringer Mannheim, Indianapolis, IN). Signals were generated with ECL detection reagents by chemiluminescence. RESULTS Patients The cell lines OPL-1 and OPL-2 were derived from 2 patients (case 1 or 2). They had suffered from chronic pyothorax resulting from pulmonary tuberculosis and were followed at the National Kinki-Chuo Hospital, Sakai, Japan. During the course of pyothorax, a tumor developed in the thoracic wall, and open biopsy was performed. Histopathologic examination and immunophenotypic analysis of biopsy materials revealed B-cell lymphoma. Case 1 entered into complete remission with chemotherapy and has remained alive for more than 1%years since biopsy. Aggressive chemotherapy was not effective in case 2, and the patient died approx. 2 months after biopsy. The major clinical characteristics of each patient are summarized in Table 11. TABLE I1 -CLINICAL AND BIOLOGICAL CHARACTERISTICS OF THE PAL CELL LINES AND OF THE PATIENTS FROM WHICH THEY WERE DERIVED’ ~ ~~ Case 1 (OPL-I) Case 2 (OPL2) Clinico-pathologicalfeatures of original cases 76 67 Age Gears) M M Sex Duration of Dvothorax hears) 46 40 Artificial pneimothora; ’ Done Done Histologic diagnosis DIB DIB CR Died Prognosis Growth characteristics Doubling time (hours) 48 24 Saturation density (cells/rnl) 5 x 10s 1 x 106 Colony formation on soft agar (-) (+I IDIB, diffuse lymphoma of immunoblastic cell type; CR, complete remission. PAL CELL LINES Establishment of cell lines Two PAL cell lines, OPL-1 arid OPL-2, were established from biopsies of PAL cases 1 and 2, respectively. In both cases, the biopsy samples obtained prior to antineoplastic chemotherapy were diagnosed as diffuse lymphoma of the immunoblastic type ( D W After cultures were started, cells growing in suspension were visible as small clusters at day 14 in OPL-1 and day 10 in OPL-2. Once established, OPL-1 and OPL-2 cells were passed at a density of 1.5 x lo5 and 1 :< 10s cells/ml every 3 days, respectively. The growth characteristics of the 2 lines are summarized in Table 11. Immunophenotypic analysis Table I shows the results of the comparative immunophenotypic analysis on cell lines OPL-1 and OPL-2 and the original tissue samples. The B-cell origin of OPL-1 and OPL-2 was demonstrated by the expression of surface immunoglobulins together with the presence of several B-cell-restricted markers, irrespective of negative reactivity for anti-CD19 antibody in both OPLs. Analysis of immunoglobulin light chains showed the monoclonal nature of OPL-1 and OPL-2 and was in agreement with a clonal derivation from the respective tumor biopsy specimens. The expression of B-cell activation markers (CD23 and/or CD30) in OPLs suggests an LCL-like character rather than a Burkitt’s lymphoma-like character. Many tumor cells in both cases 1 and 2 were: positive for EBNA2 (Fig. la, b). On the other hand, reactivity for LMPl was weak and was found in only a small number of tumor cells in case 1, but 89 tumor cells in case 2 were LMPl negative (Fig. lc, d). As for cytocentrifuged cells from the established cell lines, more than 30% of OPL-1 cells showed a much stronger expression of EBNA2 protein, but OPL-2 cells were EBNA2 negative by immunocytochemistry (Table I and Fig. 2u, b). The expression of LMPl was not detected in OPLs by immunocytochemistry (Table I and Fig. 2c, d). Cytogenetic analysis Eighteen cells (OPL-1) and 24 cells (OPL-2) were analyzed. Both cell lines showed a highly complicated karyotype, with numerical and multiple structural rearrangements. Chromosomal numbers ranged from 46 to 207 in OPL-1, and from 53 to 129 in OPL-2. Although it is difficult to define a modal number of chromosomes in both cell lines, common structural abnormalities were identified only in OPL-2 as follows: del(l)(q21) x 1-2, add(2)(pll.l), add(7)(q32 x 2, der(ll)t(ll;l4)(pl5;qll) x 2, add (15)(p13), der(l9)t(1;19)(q21;p13.3) X 2. Band 14q32.33 was not involved in structural rearrangements. Using FISH with the IgH probe, 3 co-localized signal of C, and VH segments were observed, one on the normal chromosome 14 and the others on the telomeric region of duplicated C group-sized chromosome, possibly der(1l)t(ll;l4)(pl5;qll). The split signal of C, and VH,which indicates 14q32 translocation commonly found in B-cell malignancies, was not detected. Immunoglobulin gene rearrangement analysis DNA extracted from the biopsy specimens and OPLs were digested with Hind 111 or Pst I and subjected to Southern blot FIGURE 1 - Immunostaining for EBNA2 and LMPl of biopsy specimens of cases 1 and 2. Signals were generated with the alkaline phosphatase anti-alkaline hosphatase (APAAP) method (a, 6) or with the avidin-biotin-complex(ABC) method (c, d). (a, b) EBNA2staining of cases 1 (a) and (6). Positive staining is visible in nuclei of many tumor cells in both cases. (c, d) LMP1-stainingof cases 1 (c) and 2 (d). Scattered and weak staining is visible only in case 1 (arrow heads), but no staining is visible in case 2. Scale bar = 17 p.m. ! 90 KANNO ETAL. FIGURE 2 - Immunostaining for EBNA2 and LMPl of cytocentrifuged slides of OPL-1 and OPL-2. Signals were generated with the avidin-biotin-complex (ABC) method. (a, b) EBNA2-staining of OPL-1 (a) and 2 (b). Positive staining is visible in nuclei of approximately 30% of OPL-1 cells, but no staining is visible in OPL-2. (c, d) LMP1-stainingof OPL-1 (c) and 2 (d). No staining isvisible in OPL-1 and 2. Scale bar = 17 pm. analysis with human JH probe. OPL-2 displayed clonal rearrangements of IgH gene identical to those detectable in the biopsy specimen from which OPL-2 originated (Fig. 3). On the other hand, no clear band was observed in OPL-1 and its original biopsy specimen probably because of the heteroploid pattern of its chromosomes (results not shown). EBVgenome analysis In the Southern blot analysis of Bam HI-digested DNA samples probed with LMPl oligonucleotide probe, both cell lines and their original biopsy specimens showed a slightly broad signal. However, they contained a predominant terminal repeat in fused form, and their size was identical between each cell line and the original biopsy specimen (8 kb in case 1 and OPL-1. and 10 kb in case 2 and OPL-2) (Fig. 4a). Thus, OPLs were identical to their original lymphoma tissues at the level of episomal EBV DNA and showed a monoclonal cellular proliferation of an EBV-infected progenitor cell. The type of EBV in OPL-1 and OPL-2 was type B and type A, respectively, by PCR-Sout hern blot analysis (Fig. 46). Expression of EBNA2 and LMPl in OPLs Since EBV genes EBNA2 and LMPl are known to alter the growth phenotype of human B cells (Rickinson et al., 1992; Wolf et al., 1993), we examined the expression of these molecules in OPLs at both mRNA and protein levels. By Northern blot analysis of mRNA from OPL-1 with an EBNA2 oligonucleotide probe, we observed 2 bands (Fig. 5a). Ten FIGURE 3 - Southern blot analysis for kH micrograms of DNA were digested with Hind IrI (a) or Pst I @), electrophoresed in 0.7% agarose gel, Southern blotted and hybridized with 32P-labeled JH DNA probe (5 days of exposure at -80°C). DNA size markers are indicated o n the right. Arrows indicate rearranged bands of case 2 and OPL-2. PAL CELL LINES 91 By Northern blot analysis of mRNA from cells with an LMPl oligonucleotide probe, we observed a band of about 2.6 kb in LCL and Raji cells. No band was observed in mRNA sample from OPLs (Fig. 56), though low-level expression of LMPl mRNA in OPLs was found by RT-PCR, which is much more sensitive than Northern blot analysis (results not shown). We could not observe LMPl protein expression in OPLs by Western blot (Fig. 5e). These results taken together clearly show that EBNA2 mRNA and protein were expressed in OPL-1 cells at a higher level than in LCL and Raji cells but were not detectably expressed in OPL-2 cells. On the other hand, OPLs showed little expression of LMPl mRNA, and the expression of LMPl protein was not observed in OPLs. Oncogene expression in OPLs Since EBV infection of B cells is known to up-regulate the expression of c-&r (Knutson, 1990), 6cl-2 (Henderson et a/., 1991) and c-myc (Wennborg et al., 1987), we examined OPLs for the expression of these genes. Expression of c-myc mRNAs was noted in both OPL-1 and -2 and that of c-fgr in OPL-1 alone (Fig. 6). Expression of 6cl-2 was observed in both OPL-1 and OPL-2, with much stronger expression in OPL-2 than in OPL-1 (Fig. 6). DISCUSSION F~CURE4-Southern blot analysis for the clonality of EBV genome (a) and PCR analysis for typing of EBV genome (6). (a) Ten micrograms of DNA were digested with Bam HI, electrophoresed in 0.7% agarose gel, Southern blotted and hybridized with 32P-labeledLMPl oligonucleotide probe. DNA size markers are indicated on the left (10 days of exposure at -80°C). OPLs and their original bio sy specimens contain a predominant terminal repeat (arrowheag). (b) One microgram of DNA was amplified by PCR, electrophoresed in 2% agarose gels, Southern blotted and hybridized with non-radiolabeled EBNA2A- or EBNA2B-specific oligonucleotide probe. Signals were generated by chemiluminescence 4 min of exposure at room temperature). EBNA2 sequences amp ified b PCR in biopsy specimen of case 1 and OPL-1 hybridize only with EBNA2B-specific probe. Amplified EBNA2 products in bio sy specimen of case 2, OPL-2 and Raji cells hybridize with EB&2A-..pc ecific probe. I One was approximately 3.0 kb, the same size as that observed for EBNA2 in LCL and Raji cells. The other band, unique to OPL-1 cells, was approximately 0.5 kb smaller in size. By Western blot analysis, we observe,d a strong expression of the approximately 73 kDa EBNA2 protein in OPL-1 (Fig. 5 d ) , but only a weak expression in Raji cells (Fig. 5d, arrowhead). EBNA2 protein in OPL-1 cells was approximately 7 kDa smaller in size than that in Raji cells. Since the sizes of EBNA2 protein in type A or type B EBV-infected cells are 82-87 or 72 kDa, respectively (Dambaugh et al., 1984), the difference in size of EBNA2 protein between OPL-1 (type B EBV) and Raji (type A) reflected the type of EBV genome they contained. On the other hand, neither mRNA nor protein for EBNA2 was detected in OPL-2 by Northern or Western blot analysis. Among the several EBV genes associated with latent infection, expression of EBNAl is observed in all EBV-associated malignancies since it is essential to the replication of EBV genome (Rickinson et al., 1992). OPLs also expressed EBNA1RNA by Northern blot analysis (results not shown). On the other hand, EBNA2 and LMPl activate a variety of cellular genes including oncogenes (Wennborg et al., 1987; Knutson, 1990; Henderson et al., 1991), adhesion molecules (Rickinson et al., 1992) and activation markers (Rickinson er al., 1992; Wolf et al., 1993), suggesting that these genes have direct effects on cellular transformation. However, the level of expression of EBNA2 and LMPl varies among these malignancies (Rickinson et al., 1992; Wolf et al., 1993), as shown in Table 111. Therefore, analysis of the pattern of EBV latent infection gene expression, especially EBNA2 and LMPl, might shed light on lymphomagenesis in PAL. In our current study, we showed that the expression of EBV latent infection genes varied among PAL cell lines, although restricted expression of LMPl was a constant finding. As reported previously (Fukayama et al., 1993), however, the immunohistochemical study revealed some LMP1-positive cells in the original tumor specimen of case 1. Several explanations are possible: 1) cross reactions of anti-LMP1 antibody might have occurred; 2) LMP1-positive cells in the PAL tissue specimen might not be the most rapidly growing cells, and thus only the LMPl negative cells grew in culture; or 3) LMP1expressing cells in vivo might become LMPl negative during in vitro culture. In this regard, conversion from LMPl negative to positive has been reported during in vitro culture of Burkitt’s lymphoma cells (Rickinson et al., 1992). Although we cannot address the specific likelihood of these options, restricted expression of LMPl might be a feature characteristic of PAL. LMPl is recognized as a target molecule by EBV-specific CTLS (Rickinson et al., 1992). Restricted expression of LMPl in PALS might enable transformed cells to evade immunological surveillance in non-immunocompromisedhosts. Furthermore, induction of cellular adhesion molecules, such as LFA-1, LFA-3 and ICAM-1, which participate in CTL recognition (Rickinson et aL, 1992), might be weak due to restricted expression of LMPl. Immunocytochemical procedures revealed the expression of ICAM-1 (CD54) alone in OPL-2 but both LFA-1 (CDlla) and ICAM-1 in OPL-1. These findings suggest a role of induction mechanism other than those mediated by LMPl. 92 KA”0 ETAL. FIGURE 5 - Northern and Western blot analysis of EBV latent infection genes. Northern blot for EBNA2 (a), LMPl (b) and actin (c) mRNAs. Five micrograms of mRNAs were applied to each well, Northern blotted and hybridized with 32P-labeledEBNA2 or LMPl oligonucleotide probes or with actin DNA probes. RNA size markers are indicated on the left (7 (a), 10 (b) or 1 (c) days of exposure at -80°C). OPL-1 mRNAs hybridize with EBNA2 probe only, but not with LMPl probe. OPL-2 mRNA sample contains no detectable amount of EBNA2 and LMPl transcripts. Western blot for EBNA2 (d) or LMPl (e). Cells (3 x lo7) were lysed with 400 p,l of Laemmli sample buffer, and a 40 pl aliquot of each sample was applied to each well. After electrophoresis, proteins were transferred to a nitrocellulose membrane, incubated with anti-EBNA2 (d) or anti-LMPl (e) antibodies and visualized as described in Material and Methods. The positions of molecular size markers are indicated (in kDa) on the left (20 min of exposure at room temperature). FIGURE6 -Northern blot analysis of the expression of oncogene mRNAs. Five micrograms of mRNA samples were applied to each well, Northern blotted and probed with 32P-labeledc-myc (a), c-fgr (b) and bcl-2 (c) genomic DNA probes and with mouse actin DNA probe (d). RNA size markers are indicated on the left (3 (a, c), 4 (b) or 1 (d) days of exposure at -80°C). PAL CELL LINES TABLE III - PATTERNS OF EBV LATENT GEYE EXPRESSION IN HUMAN MALIGNANCIES Burkitt’s lymphoma Hodgkin’s disease ML/ICH OPL-1 OPL-2 EBNAl EBNAZ LMPl + + + + + - + + +- - + +/+/‘ML/ICH, malignant lymphoma developed in immunocompromised hosts; +/-,weak mRNA expression detectable by RT-PCR only. LMPl transactivates the oncogene bcl-2 (Henderson et af., 1991). OPL-2 contained bcl-2 mRNA without detectable expression of LMPl protein. ‘his finding suggested bcf-2 activation by a non-LMP1-mediated mechanism. On the other hand, the expression of [email protected], which correlated with the expression of EBNA2, was strong in the EBNA2-expressing OPL-1 but undetectable in the EBNA2-negative OPL-2. Because the growth rate of the OPL-1 was much lower than that of OPL-2, EBNA2-mediated c-f’ expression might contribute little to the growth of PAL cells. Therefore, the differences in growth rate and clncogene expression between OPL-1 and OPL-2 could not be adequately explained by the pattern of EBV latent gene expression. The latent infection of EBV might be important in the initial phase of PAL lyrnphomagenesis, and subsequent genetic lesions might be necessary for the development of overt lymphoma. EBNA2 can also be a target molecule for host CTL response (Rickinson et af., 1992). The EBNA2-positive cells were present in the original biopsy specimen of OPL-2, but not in the cells derived in culture. Did the EBNA2-positive cells turn EBNA2 negative during in vitro culture? A previous study reported the reverse phenomenon, i.e., from EBNA2 negativity to EBNA2 positivity (Rickinson et al., 1992). We assume that the EBNA2-negative cells were the most rapidly growing cells in the original tumor tissue of OPL-2 and could escape host immune surveillance. On the other hand, EBNA2positive cells must be the most rapidly growing cells in the original tumor tissue of OPL-1. Because the EBNA2 may work as a target molecule for CTL, transformed cells in case 1might require some immunosuppressive circumstances to survive against host immune surveillan~x.In fact, only OPL-1, not OPL-2, expressed mRNA of an immunosuppressive cytokine, interleukin-10 (data not shown). OPL-2 had type A EBV, and OPL-1 had type B. The difference in EBV subtype in OPLs might influence their growth characteristics since EBNA2A induces a much more enhanced growth phenotype in 13 cells than EBNAZB in vitro 93 (Rickinson et al., 1987). In fact, OPL-2 with type A EBV showed a more enhanced growth rate than that of OPL-1 with type B. However, OPL-2 did not express EBNA2. These findings also suggest that genetic lesions other than EBV infection are essential for the development of overt lymphoma. The profile of expression of EBV latent infection genes in OPL-1 also provided an interesting issue in the regulation of EBV gene expression. EBNA2 transactivates LMPl in B cells (Wang et al., 1990); however, LMPl expression in OPL-1 was restricted in spite of the higher expression of EBNA2 protein. This might not be anecdotal since the expression of LMPl in B cells is regulated by various factors (Boos et af., 1987). Since 2 different-sized EBNA2 mRNAs were observed in OPL-1, some alterations in EBNA2 transcripts might affect subsequent LMPl expression. The YH exon of EBNA2 in type B EBV is approximately 100 bp smaller than that in type A EBV (Dambaugh et al., 1984; Sample et al., 1986), which could not be distinguished in agarose gels. Thus, the larger EBNA2 mRNA in OPL-1 may represent the transcripts processed in the same manner as that in LCL and Raji cells containing type A EBV, and the smaller one might contain some alterations. Further analysis may provide interesting information not only about EBV-mediated lymphomagenesis, but also in the regulation of EBV gene expression. Before the development of PAL, patients suffer from long-standing pyothorax with a mean duration of more than 30 years (Iuchi et al., 1989). Biologically active metabolites produced in chronic inflammatory sites might induce cellular DNA aberrations, the accumulation of which might be essential for the development of overt lymphoma. In conclusion, our present study suggests that the restricted expression of LMPl is common in PAL cells. This may contribute to the escape of EBV-infected cells from host immune surveillance. However, additional events other than the latent infection of EBV might be essential for causing a growth advantage sufficient for the development of overt lymphoma. Our cell lines may provide a useful model for studying lymphomagenesis in PAL, the EBV-positive B-cell lymphoma developing in non-immunocompromised hosts with long-standing chronic inflammation. ACKNOWLEDGEMENTS We thank Dr. M.E. Bloom for critically reviewing the manuscript; Ms. Y.Kanno and Y. Matsumoto for preparing it; and Ms. I. Kuroki for technical assistance. This work was supported in part by a grant-in-aid for cancer research (4-5) from the Ministry of Health and Welfare and grants from the Ministry of Education, Science and Culture (05152111, 06770125), Japan. REFER.ENCES BAITEY,J., MOULDING,C., TAUB,R., MURPHY, W., STEWART, T., P., The human c-myc oncogene: POTTER, H., LENOIR,G. and LEDER, structural consequences of translocation into the IgH locus in Burkitt lymphoma. Cell, 34,77%787 (1983). BOOS, H., BERGER, R., KUKLIK-ROCIS, C., IFTNER, T. and MUELLERLANTZSCH, N., Enhancement of E stein-Barr virus membrane protein FMP) expression by serum, TPX or n-butyrate in latently infected aji cells. Virology, 159,161-165 (1987). T., HENNESSY, K., CHAMNANKIT, L. and KIEFF, E., U2 DAMBAUGH, region of Epstein-Barr virus DNA niay encode E stein-Barr nuclear antigen 2.froc. nat. Acad. Sci. (Wash.),81,7632-7936 (1984). FENNEWALD, S., VANSANTEN, V. and KIEFF,E., Nucleotide se uence of an mRNA transcribed in latent growth-transformingvirus injection indicates that it may encode a membrane pr0tein.J. virol., 51,411-419 (1984). FUKAYAMA, M., IBUKA,T., HAYASHI, Y., OOBA,T., KOIKE, M. and MIZUTANI, S., Epstein-Barr virus in yothorax-associated pleural lymphoma.Amer. J. Pathol., 143,1044-1849 (1993). GAIDANO, G. and DALLA-FAVERA, R., Biologic and molecular characterization of non-Hodgkin’s lymphoma. Cur. @in. Oncol., 5,776784 (1993). HENDERSON, S., ROWE,M., GREGORY, C., CROOM-CARTER, D., WANG, F., LONGNECKER, R., KIEFF,E. and RICKINSON, A., Induction of bcl-2 ex ression by Epstein-Barr virus latent membrane protein 1 rotects ingcted B cells from programmed cell death. Cell, 65, 1197-1115 (1991). I ISCN 1991), Guidelines for cancer cytogenetics. In: F. Mitelman (ed.), upplement to an international system for human cytogenetics nomenclature. S . Karger, Basel (1991). IUCHI, K., AOZASA, K., YAMAMOTO, S., MORI,T., TAJIMA, K., MINATO, K., MUKAI, K., KOMATSU, H., TAGAKI, T., KOBASHI,Y.,YAMABE,H. and SHIMOYAMA, M., Non-Hodgkin’s lymphoma of the leural cavity developing from long-standing pyothorax. Summary o f clinical and pathological findings in thirty-seven cases. Jpn. J. din. Oncol., 19, 249-257 (1989). KANNO, H., WOLFINBARGER, J.B. and BLOOM, M.E., Aleutian mink 94 PAL CELL LINES L., Detection of a second widespread strain of Epstein-Barr virus. Lancet, 761-765 (1989). M., MATSUDA, F., JAUCH,A,, NISHIDA, K., TAKASHIMA, T., KNUTSON, J.C., The level of c-fgr RNA is increased by EBNA-2, an TANIWAKI, S., SUGIYAMA, H., MISAWA,S., ABE, T. and KASHIMA, K., Epstein-Barr virus gene required for B-cell immortalization. J. ViroL, TAGAWA, Detection of 14q32 translocations in B-cell malignancies by m situ 64,2530-2536 (1990). hybridization with yeast artificial chromosome clones containing the NISHIDA, Y.,MIKI,T.. HISNIMA,H. and HONJO,T., Cloning of human human IgH gene locus. Blood, 83,2962-2969 (1994). immunoglobulin E chain genes: evidence for multiple C, genes. Proc. TRONICK,S.R., POPESCU,N.C., CHEAH, M.S.C., SWAN,D.C., nat. Acad. Sci. (Wash.),79,3833-3837 (1982). AMSBAUGH, S.C., LENGEL,C.R., DIPAOLO,J.A. and ROBBINS, K.C., Isolation and chromosomal localization of the human fp protooncoRAAB-TRAUB. N. and FLYNN, K., The structure of the termini of the Epstein-Barr virus as a marker of clonal cellular proliferation. CeN, 47, gene, a distinct member of the tyrosine kinase gene family. Proc. nat. Acad. Sci. (Wash.),82,6595-6599 (1985). 883-889 (1986). Y. and CROCE.C.M., Analysis of the structure, transcripts, RICKINSON, A.B., MURRAY, R.J., BROOKS, J., GRIFFIN, H., MOSS,D.J. TSUJIMOTO, and MASUCCI, M.G., T cell recognition of Epstein-Ban virus associ- and protein products of bcl-2, the gene involved in human follicular lymphoma. Proc. nat. Acad. Sci. (Wash.). 83,5214-5218 (1986). ated lymphomas. Cancer SZLW.,13,53-80 (1992). WANG,F., TSANG, S.-F., KURILLA, M.G., COHEN,J.I. and KIEFF,E., RICKINSON, A.B., YOUNG,L.S. and ROWE. M., Influence of the Epstein-Barr virus nuclear antigen 2 transactivates latent membrane Epstein-Barr virus nuclear antigen EBNA 2 on the growth phenotype protein LMP1. J. ViroL, 64,3407-3416 (1990). of virus-transformed B cells. J. Virol., 61, 1310-1317 (1987). WENNBORG, A,, AMAN, P., SARANATH, D., PEAR,W., SUMEGI,J. and SAMPLE, J., HUMMEL, M., BRAUN, D., BIRKENBACH, M. and KIEFF,E., KLEIN,G., Conversion of the lymphoma line “BJAB” by Epstein-Barr Nucleotide sequences of mRNAs encoding Epstein-Barr virus nuclear virus into phenotypically altered sublines is accompanied by increased proteins: a probable transcriptional initiation site. Proc. tiat Acad. Sci. c-myc mRNA levels. Int. J. Cancer, 40,202-206 (1987). (Wash.), 83,5096-5100 (1986). WOLF,H., BOGEDAIN, C. and SCHWARZMANN, F., Epstein-Barr virus SIXBEY, J.W., SHIRLEY, P., CHESNEY, P.J., BUNTIN, D.M. and RESNICK, and its interaction with the host. Intervirology, 35,26-39 (1993). disease parvovirus infection of mink peritoneal macrophages and human macrophage cell 1ines.J. ViroL, 67,2075-2082 (1993).