int. J. Cancer: 50,589-592 (1992) 0 1992 Wiley-Liss, Inc.
I
Publicationof the International Union Against Cancer Publicationde I'Union Internationale Contre le Cancer
CELL PHENOTYPE (CD23)-DEPENDENT VARIATION IN EBV GENOME COPY NUMBERS WITHIN LYMPHOBLASTOID CELL LINES (LCL) Vidar WENDEL-HANSEN1'4,Wen TA02,Mats ERICSON3,George KLEIN' and Anders ROSkN' 'Hybridoma Group, Department of Medical Cell Genetics, 'Department of Tumour Biology and 'Department of Physiological Chemistiy, Karolinska institutet, S-104 01 Stockholm, Sweden. Three Epstein-Barr-virus-transformed lymphoblastoid cell show that EBV-carrying cells which enter the lytic cycle, as lines (LCL) were analysed on the basis of their CD23 expression. indicated by the appearance of the EAiVCA antigen complex, Levels of EBV-DNA were compared in the positive and negative do not express CD23. subpopulations. Two lines were further analysed with regard t o EBNA, cytoplasmic immunoglobulin (clg) and lytic (EA/VCA) MATERIAL AND METHODS protein expression. Both subpopulations had a similar MHC class4 transcription, but the CD23- subpopulation had a lower Antibodies plating efficiency and a lower rate of DNA synthesis. In the B6, The mouse monoclonal antibody (MAB) MHM6 (Rowe et NADSO and 0467.3 cell lines, CD23- cells contained 2 f 0.2 6.4 f 3.0 times less EBV DNA than the corresponding CD23+ al., 1982) was used for staining the CD23 antigen. FITCpopulation. EBNA was expressed in 8 I 2 4.2% - 93 t 3.8% of conjugated, affinity-purified anti-mouse immunoglobulin was the CD23' cells and in 0 - 46 2 8.0% of the CD23- cells. No obtained from Dakopatts (Glostrup, Denmark). Phycoerythrin CD23+cells in 86 or NADSO contained any EAIVCA, while 19 f (PE)-conjugated anti-mouse immunoglobulin was obtained 2.8% - 24 2 4.2% of the CD23- cells were positive for the from Vector (Burlingame, CA). lytic-cycle-associatedantigens. Ofthe CD23- cells, 70 2 8.6% 86 t 6.0% were positive for cytoplasmic immunoglobulin Cell lines compared t o 14.7 t 2.7% - 14.9 t 1.8% in the corresponding Cell lines are listed in Table I. All cells were cultured at 37°C CD23' population. We have previously shown that only 18% of in a 5% CO, atmosphere, with RPMI 1640 containing 5% fetal the clg-positive cells were EBNA-positive in the B6 line com- calf serum, 1 mM L-glutamine, 100 U/ml penicillin and 100 pared to 94% in the clg- population. This was open to 2 pg/ml streptomycin. alternative interpretations: loss of EBV genomes from a fraction of the cells with subsequent differentiation t o secretory immu- Two-colorimmunofluorescence noglobulin production, or down-regulation of EBNA expression Cells were stained for CD23 in suspension by indirect IF in differentiating, EBV-genome-positive cells. Our present findings speak for the first alternative, indicating that a certain using PE-conjugated anti-mouse Ig. Cell smears were preproportion of the cells may lose their EBV genomes in both pared, fixed and stained for EBNA, EA or human immunogloblong-establishedand freshly transformed LCLS. This is accompa- ulin using FITC-conjugated antibodies. EBNA was stained by nied by a reduced percentage of EBNA-positive cells, the anti-complement IF (ACIF) using the serum of an EBVdisappearance of at least one activation marker (CD23) associ- seropositive donor, KF (Reedman and Klein, 1973). ACIF ated with the virally induced blast transformation, and an stains EBNA 1, 2, 3 and 6 and has a titer of 1:320 (EBNA l), increased synthesis of clg. 1:160 (EBNA 2), and 1:160 (EBNA 6). The lytic-cycle-
EBV-immortalized lymphoblastoid cell lines carry multiple episomal copies of the viral genome (Adams, 1979) which replicate with the cell cycle. The average number of viral genome copies remains relatively constant in each line over long periods of time (Adams et al., 1973). This has led to the tacit assumption that all cells within a given EBV-transformed B-cell clone carry approximately the same number of genome copies. The question has not been tested experimentally, however. We have previously shown that 1@33% of the cells within 4 different EBV-transformed LCLs matured continuously to plasmacytoid cells that contained clg and expressed little or no detectable EBNA (Wendel-Hansen et al., 1987). Of these cells, 75% failed to incorporate bromodeoxyuridine, suggesting that they have exited from the cell cycle. This raised the question whether EBV-DNA-carrying cells down-regulate EBNA when they differentiate into secretory cells or, alternatively, whether the accidental loss of the viral genomes may have permitted the cells to proceed with their secretory differentiation. We have now approached this problem from another direction. Conceivably, variations in the EBV-DNA content of lymphoblastoid cells may be reflected in differences in their expression of EBV-induced activation markers. We have chosen CD23, regularly expressed on EBV-transformed B-cells (Gordon et al., 1989). EBV-negative Burkitt lymphoma (BL) lines do not express CD23, as a rule, but transfection with either EBNA-2 (Wang et al., 1987) or with the virally encoded latent membrane protein LMP (Wang et al., 1988) may trigger CD23 expression. We report that CD23-positive and -negative cells differ in the number of EBV genomes per cell. We also
associated EAiVCA complex was stained by direct IF, using FITC-conjugated human immunoglobulins (Esther). It was used at a working dilution of 1:60. FACS separation Cells were stained for CD23 in suspension by indirect IF using FITC-conjugated antimouse Ig. The 10% most intensely stained and the 10% least intensely stained cells were separated on a FACS IV (Becton Dickinson, Mountain View, CA). After separation, the cells were either used directly or pelleted and stored at -70°C until further use. Plating eflcienq Cells were stained with the MHM6 MAb and sorted on the FACS under sterile conditions, then seeded into 96-well microplates in 2 concentrations, 1 and 10 cells/well, using irradiated (5,000 rad) human peripheral blood lymphocytes as feeder cells. The microplates were screened for outgrowth after 6 weeks. DNA synthesis Cells were sorted under sterile conditions and seeded into 96-well microplates at densities of 5 x lo' and 5 x lo4 cells/well. After 48 hr, 1 pCi/well of 3H-thymidine was added to the cultures and 24 hr later the cells were harvested to filters and the incorporation of 'H-thymidine was analyzed with a
'To whom correspondence and reprint requests should be sent. Received: June 17, 1991 and in revised form September 30,1991.
590
WENDEL-HANSEN ET AL.
TABLE I - LIST OF EBV-TRANSFORMED LYMPHOBLASTOID CELL LINES STUDIED Cell line
le. class
Suecific features
Reference
NADSO IgM, K, A Freshly established line. Studied from 3 to 6 months after infection B6 IgM, K Tetanus toxoid antiKozbor and body producer Roder, 1981 0467.3 IgM, A Chiorazzi et al., 1982 Fluorescence intensity
Betaplate scintillation counter (LKB Wallac, Turkku, Finland).
DNA slot-blotting Cells were suspended in 0.4 M NaOH and heated at 80°C for 10 min. The samples were cooled briefly, titrated in 2 steps and applied to a Nylon filter (Hybond-N, Amersham, Ailesbury, UK) pre-wetted with 0.4 M NaOH using a slot-blot manifold (BioRad, Richmond, CA). The filters were rinsed twice for 5 min each with 2 x SSC (standard sodium citrate; 0.15 M sodium chloride, 0.016 M sodium citrate) to neutralize the NaOH, then dried in room temperature for at least 1 hr and baked at 80°C for 30 min. DNA hybridization The EBV fragment used for hybridization was the BamHIW fragment inserted in pBR 322. Approximately 1 pg of plasmid was labelled with P3'-dCTP using random priming and separated on a Sephadex G-50 spin column. Hybridization was performed as previously described (Siimegi et al., 1983). To assess the amount of DNA applied in each slot, the filter was stripped and rehybridized with 2 pg of radiolabelled human placental DNA, digested with EcoRI for 3 hr. The films were scanned with a Grundig Pasecon TV camera and transferred to a Zeiss/Kontron I3AS image analysis computer. The integrated optical densities (IOD) of the signals were measured and the size for integration was fixed for individual bands. Northern-blotanalysis Total RNA isolated from 2 x 10' FACS-sorted CD23' and CD23- cells from the B6 cell line by the method of Auffray and Rougeon (1980) was subjected to electrophoresis in a 1% agarose gel containing 1% formaldehyde. Ten micrograms were loaded into each sample well. Electrophoretically separated RNA molecules were blotted to a nylon membrane (Schleicher and Schuell, Dassel, Germany) by capillary transfer, followed by baking for 1 hr at 80°C and UV cross-linking for 3 min. The probe used for hybridization was the HLA-DRa fragment isolated from plasmid pDR-a-1 (Larhammar et al., 1982) by cleavage with PstI. The resulting 2 insert fragments were not separated from each other. 3ZP-dCTP-labelledprobes were used at final concentrations of 0.5 x lo6 cpm/ml in each hybridization experiment. Conditions for hybridization were as described (Andersson et al., 1987). The membranes were washed once in 2 x SSC containing 0.1% SDS at room temperature for 30 min, and twice in 0.5 x SSC, 0.1% SDS at 65°C for 1 hr each. Washed membranes were exposed to Kodak XAR-5 films for 16 hr at -70"C, with an intensifying screen. Densitometry was performed as described above. RESULTS
FACS separation FACS separation was evaluated by re-running the separated subpopulations and comparing them with the unseparated population (Fig. 1). The expected difference in antigen expression was also verified by visual examination in fluorescence microscopy.
Fluorescence intensity
FIGURE 1 - FACS analysis of FACS-sorted cells. The 10% most and least CD23-positive cells were separated. The 2 populations were then re-run and compared to the total population (n = 10). ( a ) NADSO (polyclonal); (b)B6 (monoclonal).
Two-color irnrnunofluorescence Each individual cell was examined for surface (CD23) and intracellular (cIg, EBNA, EA/VCA) staining. The CD23positive and -negative fractions were compared with regard to EBNA and cIg (Fig. 2) and EA/VCA (Table 11). In the B6 cell line, all EBNA-positive cells were CD23', whereas in NAD50, 17% of the EBNA+ cells were CD23-. In both cell lines investigated, the proportion of cIg-positive cells was significantly higher in the CD23- population. All EA/VCA-positive cells were found in the CD23- fraction (Table 11). Plating eficiency The plating efficiency of the CD23' and CD23- subpopulations was compared by estimating the frequency of micro-wells containing growing cells after 6 weeks of incubation. In each of 2 experiments, 192 wells were screened for each cell line and cell density. At 1 cell/well, the outgrowth frequency (mean values) of the CD23+ cells was 6% in both B6 and NADSO, as compared to 0% (NAD50) and 2% (B6) in the CD23population. At 10 cells/well, the outgrowth frequencies were 26% (NAD50) and 20% (B6) in the CD23' population versus 13% (NADSO) and 16% (B6) in the CD23- population. DNA synthesis Eight wells were counted for each subpopulation and density. The CD23' population had a higher rate of DNA synthesis than the CD23- (Table 111). DNA hybridization on FACS-sorted cells The CD23' cells contained 6.4 ? 3.0, 2 +- 0.2 and 2.6 ? 1.0 (for B6, NAD50 and 0467.3, respectively; ?SD, n = 4) times more EBV DNA than the CD23- cells, as estimated by optical densitometry on the autoradiograms of titrated samples (Fig. 3). Hybridization with randomly radiolabelled human DNA was performed to confirm that similar amounts of DNA were present in each slot. In order to evaluate the influence of virus-producing cells, the FACS-sorted subpopulations were stained for VCA, using a human hyperimmune serum and indirect IF. The frequency of positive cells was below 0.5% in both fractions. Northern blotting The yield of total RNA was 21 pg from the CD23' population, and 17 pg from the CD23- population. The integrated optical density values for the 2 subpopulations of the B6 cell line after hybridization with the MHC class-I1 probe was 32.8 for the CD23+ and 24.3 for the CD23population. DISCUSSION
EBV-transformed LCLs proliferate by multiplication of virally transformed immunoblasts that carry multiple E3Vgenome copies and express the virally encoded EBNA 1-6 plus
CD23-DEPENDENT VARIATION
9*4.2%
591
2.0+0.7%
NADSO
B6
FIGURE 2 - Analysis of two-color fluorescence. CD23-positive and -negative cells were examined for EBNA and cytoplasmic immunoglobulin expression. A total of 500 cells were examined in each experiment (mean value and SD of 3 experiments). TABLE I1 - EAIVCA EXPRESSION IN RELATION TO CD23’ AND CD23SUBPOPULATIONS (PER CENT POSITIVE CELLS, MEAN VALUE AND SD OF 3 EXPERIMENTS)
CD23 in total population EAIVCA in total population EAIVCA in CD23- subpopulation EAIVCA in CD23’ subpopulation
B6
NADSO
8324 4+1 24 4 0
73+6 5f1 18 f 3 0
*
TABLE 111 - ‘11 T H Y W l D l N t IUCORPOKATION (cpm x 10 ’) OF FACS SOKT‘FDC t L L S (MEAZ V A L U t AUD SD OF 3 FXPtKIMEh’TS)
B6
NADSO Cell
I
Publicationof the International Union Against Cancer Publicationde I'Union Internationale Contre le Cancer
CELL PHENOTYPE (CD23)-DEPENDENT VARIATION IN EBV GENOME COPY NUMBERS WITHIN LYMPHOBLASTOID CELL LINES (LCL) Vidar WENDEL-HANSEN1'4,Wen TA02,Mats ERICSON3,George KLEIN' and Anders ROSkN' 'Hybridoma Group, Department of Medical Cell Genetics, 'Department of Tumour Biology and 'Department of Physiological Chemistiy, Karolinska institutet, S-104 01 Stockholm, Sweden. Three Epstein-Barr-virus-transformed lymphoblastoid cell show that EBV-carrying cells which enter the lytic cycle, as lines (LCL) were analysed on the basis of their CD23 expression. indicated by the appearance of the EAiVCA antigen complex, Levels of EBV-DNA were compared in the positive and negative do not express CD23. subpopulations. Two lines were further analysed with regard t o EBNA, cytoplasmic immunoglobulin (clg) and lytic (EA/VCA) MATERIAL AND METHODS protein expression. Both subpopulations had a similar MHC class4 transcription, but the CD23- subpopulation had a lower Antibodies plating efficiency and a lower rate of DNA synthesis. In the B6, The mouse monoclonal antibody (MAB) MHM6 (Rowe et NADSO and 0467.3 cell lines, CD23- cells contained 2 f 0.2 6.4 f 3.0 times less EBV DNA than the corresponding CD23+ al., 1982) was used for staining the CD23 antigen. FITCpopulation. EBNA was expressed in 8 I 2 4.2% - 93 t 3.8% of conjugated, affinity-purified anti-mouse immunoglobulin was the CD23' cells and in 0 - 46 2 8.0% of the CD23- cells. No obtained from Dakopatts (Glostrup, Denmark). Phycoerythrin CD23+cells in 86 or NADSO contained any EAIVCA, while 19 f (PE)-conjugated anti-mouse immunoglobulin was obtained 2.8% - 24 2 4.2% of the CD23- cells were positive for the from Vector (Burlingame, CA). lytic-cycle-associatedantigens. Ofthe CD23- cells, 70 2 8.6% 86 t 6.0% were positive for cytoplasmic immunoglobulin Cell lines compared t o 14.7 t 2.7% - 14.9 t 1.8% in the corresponding Cell lines are listed in Table I. All cells were cultured at 37°C CD23' population. We have previously shown that only 18% of in a 5% CO, atmosphere, with RPMI 1640 containing 5% fetal the clg-positive cells were EBNA-positive in the B6 line com- calf serum, 1 mM L-glutamine, 100 U/ml penicillin and 100 pared to 94% in the clg- population. This was open to 2 pg/ml streptomycin. alternative interpretations: loss of EBV genomes from a fraction of the cells with subsequent differentiation t o secretory immu- Two-colorimmunofluorescence noglobulin production, or down-regulation of EBNA expression Cells were stained for CD23 in suspension by indirect IF in differentiating, EBV-genome-positive cells. Our present findings speak for the first alternative, indicating that a certain using PE-conjugated anti-mouse Ig. Cell smears were preproportion of the cells may lose their EBV genomes in both pared, fixed and stained for EBNA, EA or human immunogloblong-establishedand freshly transformed LCLS. This is accompa- ulin using FITC-conjugated antibodies. EBNA was stained by nied by a reduced percentage of EBNA-positive cells, the anti-complement IF (ACIF) using the serum of an EBVdisappearance of at least one activation marker (CD23) associ- seropositive donor, KF (Reedman and Klein, 1973). ACIF ated with the virally induced blast transformation, and an stains EBNA 1, 2, 3 and 6 and has a titer of 1:320 (EBNA l), increased synthesis of clg. 1:160 (EBNA 2), and 1:160 (EBNA 6). The lytic-cycle-
EBV-immortalized lymphoblastoid cell lines carry multiple episomal copies of the viral genome (Adams, 1979) which replicate with the cell cycle. The average number of viral genome copies remains relatively constant in each line over long periods of time (Adams et al., 1973). This has led to the tacit assumption that all cells within a given EBV-transformed B-cell clone carry approximately the same number of genome copies. The question has not been tested experimentally, however. We have previously shown that 1@33% of the cells within 4 different EBV-transformed LCLs matured continuously to plasmacytoid cells that contained clg and expressed little or no detectable EBNA (Wendel-Hansen et al., 1987). Of these cells, 75% failed to incorporate bromodeoxyuridine, suggesting that they have exited from the cell cycle. This raised the question whether EBV-DNA-carrying cells down-regulate EBNA when they differentiate into secretory cells or, alternatively, whether the accidental loss of the viral genomes may have permitted the cells to proceed with their secretory differentiation. We have now approached this problem from another direction. Conceivably, variations in the EBV-DNA content of lymphoblastoid cells may be reflected in differences in their expression of EBV-induced activation markers. We have chosen CD23, regularly expressed on EBV-transformed B-cells (Gordon et al., 1989). EBV-negative Burkitt lymphoma (BL) lines do not express CD23, as a rule, but transfection with either EBNA-2 (Wang et al., 1987) or with the virally encoded latent membrane protein LMP (Wang et al., 1988) may trigger CD23 expression. We report that CD23-positive and -negative cells differ in the number of EBV genomes per cell. We also
associated EAiVCA complex was stained by direct IF, using FITC-conjugated human immunoglobulins (Esther). It was used at a working dilution of 1:60. FACS separation Cells were stained for CD23 in suspension by indirect IF using FITC-conjugated antimouse Ig. The 10% most intensely stained and the 10% least intensely stained cells were separated on a FACS IV (Becton Dickinson, Mountain View, CA). After separation, the cells were either used directly or pelleted and stored at -70°C until further use. Plating eflcienq Cells were stained with the MHM6 MAb and sorted on the FACS under sterile conditions, then seeded into 96-well microplates in 2 concentrations, 1 and 10 cells/well, using irradiated (5,000 rad) human peripheral blood lymphocytes as feeder cells. The microplates were screened for outgrowth after 6 weeks. DNA synthesis Cells were sorted under sterile conditions and seeded into 96-well microplates at densities of 5 x lo' and 5 x lo4 cells/well. After 48 hr, 1 pCi/well of 3H-thymidine was added to the cultures and 24 hr later the cells were harvested to filters and the incorporation of 'H-thymidine was analyzed with a
'To whom correspondence and reprint requests should be sent. Received: June 17, 1991 and in revised form September 30,1991.
590
WENDEL-HANSEN ET AL.
TABLE I - LIST OF EBV-TRANSFORMED LYMPHOBLASTOID CELL LINES STUDIED Cell line
le. class
Suecific features
Reference
NADSO IgM, K, A Freshly established line. Studied from 3 to 6 months after infection B6 IgM, K Tetanus toxoid antiKozbor and body producer Roder, 1981 0467.3 IgM, A Chiorazzi et al., 1982 Fluorescence intensity
Betaplate scintillation counter (LKB Wallac, Turkku, Finland).
DNA slot-blotting Cells were suspended in 0.4 M NaOH and heated at 80°C for 10 min. The samples were cooled briefly, titrated in 2 steps and applied to a Nylon filter (Hybond-N, Amersham, Ailesbury, UK) pre-wetted with 0.4 M NaOH using a slot-blot manifold (BioRad, Richmond, CA). The filters were rinsed twice for 5 min each with 2 x SSC (standard sodium citrate; 0.15 M sodium chloride, 0.016 M sodium citrate) to neutralize the NaOH, then dried in room temperature for at least 1 hr and baked at 80°C for 30 min. DNA hybridization The EBV fragment used for hybridization was the BamHIW fragment inserted in pBR 322. Approximately 1 pg of plasmid was labelled with P3'-dCTP using random priming and separated on a Sephadex G-50 spin column. Hybridization was performed as previously described (Siimegi et al., 1983). To assess the amount of DNA applied in each slot, the filter was stripped and rehybridized with 2 pg of radiolabelled human placental DNA, digested with EcoRI for 3 hr. The films were scanned with a Grundig Pasecon TV camera and transferred to a Zeiss/Kontron I3AS image analysis computer. The integrated optical densities (IOD) of the signals were measured and the size for integration was fixed for individual bands. Northern-blotanalysis Total RNA isolated from 2 x 10' FACS-sorted CD23' and CD23- cells from the B6 cell line by the method of Auffray and Rougeon (1980) was subjected to electrophoresis in a 1% agarose gel containing 1% formaldehyde. Ten micrograms were loaded into each sample well. Electrophoretically separated RNA molecules were blotted to a nylon membrane (Schleicher and Schuell, Dassel, Germany) by capillary transfer, followed by baking for 1 hr at 80°C and UV cross-linking for 3 min. The probe used for hybridization was the HLA-DRa fragment isolated from plasmid pDR-a-1 (Larhammar et al., 1982) by cleavage with PstI. The resulting 2 insert fragments were not separated from each other. 3ZP-dCTP-labelledprobes were used at final concentrations of 0.5 x lo6 cpm/ml in each hybridization experiment. Conditions for hybridization were as described (Andersson et al., 1987). The membranes were washed once in 2 x SSC containing 0.1% SDS at room temperature for 30 min, and twice in 0.5 x SSC, 0.1% SDS at 65°C for 1 hr each. Washed membranes were exposed to Kodak XAR-5 films for 16 hr at -70"C, with an intensifying screen. Densitometry was performed as described above. RESULTS
FACS separation FACS separation was evaluated by re-running the separated subpopulations and comparing them with the unseparated population (Fig. 1). The expected difference in antigen expression was also verified by visual examination in fluorescence microscopy.
Fluorescence intensity
FIGURE 1 - FACS analysis of FACS-sorted cells. The 10% most and least CD23-positive cells were separated. The 2 populations were then re-run and compared to the total population (n = 10). ( a ) NADSO (polyclonal); (b)B6 (monoclonal).
Two-color irnrnunofluorescence Each individual cell was examined for surface (CD23) and intracellular (cIg, EBNA, EA/VCA) staining. The CD23positive and -negative fractions were compared with regard to EBNA and cIg (Fig. 2) and EA/VCA (Table 11). In the B6 cell line, all EBNA-positive cells were CD23', whereas in NAD50, 17% of the EBNA+ cells were CD23-. In both cell lines investigated, the proportion of cIg-positive cells was significantly higher in the CD23- population. All EA/VCA-positive cells were found in the CD23- fraction (Table 11). Plating eficiency The plating efficiency of the CD23' and CD23- subpopulations was compared by estimating the frequency of micro-wells containing growing cells after 6 weeks of incubation. In each of 2 experiments, 192 wells were screened for each cell line and cell density. At 1 cell/well, the outgrowth frequency (mean values) of the CD23+ cells was 6% in both B6 and NADSO, as compared to 0% (NAD50) and 2% (B6) in the CD23population. At 10 cells/well, the outgrowth frequencies were 26% (NAD50) and 20% (B6) in the CD23' population versus 13% (NADSO) and 16% (B6) in the CD23- population. DNA synthesis Eight wells were counted for each subpopulation and density. The CD23' population had a higher rate of DNA synthesis than the CD23- (Table 111). DNA hybridization on FACS-sorted cells The CD23' cells contained 6.4 ? 3.0, 2 +- 0.2 and 2.6 ? 1.0 (for B6, NAD50 and 0467.3, respectively; ?SD, n = 4) times more EBV DNA than the CD23- cells, as estimated by optical densitometry on the autoradiograms of titrated samples (Fig. 3). Hybridization with randomly radiolabelled human DNA was performed to confirm that similar amounts of DNA were present in each slot. In order to evaluate the influence of virus-producing cells, the FACS-sorted subpopulations were stained for VCA, using a human hyperimmune serum and indirect IF. The frequency of positive cells was below 0.5% in both fractions. Northern blotting The yield of total RNA was 21 pg from the CD23' population, and 17 pg from the CD23- population. The integrated optical density values for the 2 subpopulations of the B6 cell line after hybridization with the MHC class-I1 probe was 32.8 for the CD23+ and 24.3 for the CD23population. DISCUSSION
EBV-transformed LCLs proliferate by multiplication of virally transformed immunoblasts that carry multiple E3Vgenome copies and express the virally encoded EBNA 1-6 plus
CD23-DEPENDENT VARIATION
9*4.2%
591
2.0+0.7%
NADSO
B6
FIGURE 2 - Analysis of two-color fluorescence. CD23-positive and -negative cells were examined for EBNA and cytoplasmic immunoglobulin expression. A total of 500 cells were examined in each experiment (mean value and SD of 3 experiments). TABLE I1 - EAIVCA EXPRESSION IN RELATION TO CD23’ AND CD23SUBPOPULATIONS (PER CENT POSITIVE CELLS, MEAN VALUE AND SD OF 3 EXPERIMENTS)
CD23 in total population EAIVCA in total population EAIVCA in CD23- subpopulation EAIVCA in CD23’ subpopulation
B6
NADSO
8324 4+1 24 4 0
73+6 5f1 18 f 3 0
*
TABLE 111 - ‘11 T H Y W l D l N t IUCORPOKATION (cpm x 10 ’) OF FACS SOKT‘FDC t L L S (MEAZ V A L U t AUD SD OF 3 FXPtKIMEh’TS)
B6
NADSO Cell
