Persistent Infection of Epstein-Barr Virus-Positive B Lymphocytes by Human Herpesvirus 8

Blood donors.

Peripheral blood mononuclear cells (PBMC) were isolated from EDTA-treated blood of four EBV-negative and EBV-positive healthy individuals by Ficoll-Hypaque discontinuous gradient centrifugation (Lymphoflot; Biotest AG, Dreieich, Germany) as specified by the manufacturer. Blood donors were not from HHV-8 risk groups. EBV serologic testing was performed with the following commercial kits detecting virus capsid antigen (VCA) immunoglobulin G (IgG) (Fresenius, Bad Homburg, Germany) and IgM (Viramed, Martinsried, Germany) and Epstein-Barr virus nuclear antigen (EBNA) IgG antibodies (Biotest, Dreieich, Germany). EBV-positive individuals were positive for VCA IgG and EBNA IgG antibodies but negative for VCA IgM antibodies. EBV-negative individuals were negative for VCA IgG, VCA IgM, and EBNA IgG.

Cell culture and EBV and/or HHV-8 infection.

BCBL-1 (38), BC-1 (35), B95-8 (CRL1612), Raji (CCL 86), and LCLs were cultured in RPMI (Life Technologies, Paisley, Scotland) supplemented with 20% heat-inactivated fetal calf serum (FCS; Life Technologies), 100 IU of penicillin (Life Technologies) per ml, 100 mg of streptomycin (Life Technologies) per ml, 2 mM l-glutamine (Life Technologies), and 0.05 mM 2-mercaptoethanol (Sigma, St. Louis, Mo.). For cells at low density, the 2-mercaptoethanol was replaced by 20 mM bathocuproine disulfonic acid (Sigma), 50 μm α-thioglycerol (Sigma), and 1 mM sodium pyruvate (Life Technologies). To induce the expression of lytic viral proteins, cells were treated with 20 ng of phorbol 12-myristate 13-acetate (TPA; Sigma) per ml and 3 mM sodium n-butyrate (Sigma) for 24 h.

Supernatants of either BCBL-1, BC-1, B95-8, or Raji cells grown at very high densities (5 × 10⁵ per ml) were used to infect PBMC. Supernatants were filtered through 0.4-μm filters and serially diluted with cell culture medium in 96-well plates, and PBMC were added to 10⁴ cells per well. For serial virus passage, supernatants of ABEH-1, ABE-1, GMEH-1, and GME-1 cells were generated similarly and added to 10⁴ PBMC from an EBV⁺ donor.

Isolation of cellular DNA and detection of viral DNA by PCR.

Total cellular DNA was prepared from 10⁶ cells by the method of Miller et al. (26). The primer used to amplify HHV-8 DNA flanked the K12 open reading frame (K12for, 5′ cggaattcatggatagaggcttaacg 3′; K12rev, 5′ cgctcgagtcagtgcgcgcccgttgc 3′). The length of the expected amplified product is 196 bp. The primers used to amplify EBV DNA from the BALF5 open reading frame were Epolfor (5′ aggttggcggggctcagggc 3′) and Epolrev (5′ agcacaggctagccggcctg 3′), and the amplified product was 343 bp in size. Each reaction mixture contained 500 ng of cellular DNA, 100 ng of each primer, and 5 U of Taq polymerase (Perkin-Elmer Cetus, Branchburg, N.J.). The mixture was cycled in a DNA thermal cycler 2400 (Perkin-Elmer) for 30 cycles of amplification (96°C for 1 min; 60°C for 1 min; 72°C for 1 min). One-tenth of the PCR products were analyzed on 2% agarose–40 mM Tris acetate–1 mM EDTA and visualized after ethidium bromide staining. As negative and positive PCR controls, water and plasmid DNAs of EBV polymerase and K12 cloned into pBluescript KS II⁺ were used, respectively.

Preparation of RNA and RT-PCR.

RNA was extracted from 10⁶ cells with Tri Reagent (Molecular Research Center, Cincinnati, Ohio) as specified by the manufacturer. A 5-μg portion of total RNA in 10 μl of water and 1 μg of oligo(dT)₁₈ primer were heated to 70°C for 10 min and then cooled to 4°C on ice. A 10-μl volume of reaction mixture (100 mM Tris-HCl [pH 8.3], 6 mM MgCl₂, 150 mM KCl, 1 mM each deoxynucleoside triphosphate (Boehringer, Mannheim, Germany), 30 U of RNase inhibitor [MBI Fermentas, Vilnius, Lithuania], 200 U of Superscript [Life Technologies]) was added, and the mixture was incubated at 37°C for 1.5 h and heated to 67°C for 15 min. A 2-μl volume of the reverse transcription reaction mixture was amplified with gene-specific primers in a 100-μl PCR mixture containing 10 mM Tris-HCl (pH 8.3), 1.5 mM MgCl₂, 0.2 mM each deoxynucleoside triphosphate, 100 ng of each primer, and 5 U of Taq polymerase. The mixture was cycled in a DNA thermal cycler 2400 for 30 cycles of amplification (96°C for 1 min; 60°C for 1 min; 72°C for 1 min). One-tenth of the PCR mixtures were analyzed on 1 or 2% agarose–40 mM Tris-acetate–1 mM EDTA and visualized after ethidium bromide staining.

The oligonucleotide primers used in this study were K12for, K12rev, VP23for (5′ cgcgggtctagaatcgcactcgacaagagtata 3′), VP23rev (5′ cgcggggaattctttagcgtggggaataccaacagga 3′), Actfor (5′ cgcgaatccccccagtgtgacatgg 3′), and Actrev (5′ cgcggatcccagccaggtccagacg 3′). As negative controls for these experiments, reverse transcription-PCR (RT-PCR) was performed with 1 μg of total RNA. As negative and positive PCR controls, water and plasmid DNAs of full-length cDNA of VP23, actin, and K12 cloned into pBluescript KS II+ were used, respectively.

Flow cytometry.

To analyze the surface phenotype of LCLs, BC-1 and BCBL-1 cells (5 × 10⁵) were washed in phosphate-balanced saline (PBS) and FACS buffer (PBS containing 2.5% FCS and 0.02% sodium azide) and stained directly with anti-CD3-phycoerythrin (PE), anti-CD14-PE, anti-CD16-PE, anti-CD19-PE, anti-CD56-PE (Pharmingen, Hamburg, Germany), anti-IgG-fluorescein isothiocynate (FITC), and anti-IgM-FITC (Sigma). Isotype controls were included in each assay. For indirect staining, cells were incubated first with mouse anti-CD19, anti-CD20, anti-CD23, or anti-CD30 (Dako Diagnostika, Hamburg, Germany) and then with goat anti-mouse IgG-FITC (Dianova, Hamburg, Germany). As the negative control, only goat anti-mouse IgG-FITC was used in each assay. The cells were washed three times and analyzed on a FACScan with Cellquest software (Becton Dickinson, San Jose, Calif.).

Immunofluorescence and confocal microscopy.

After stimulation for 24 h with TPA and n-butryate as described above, the cells were dried on poly-l-lysine-coated coverslips (Marienfeld, Bad Mergentheim, Germany), fixed with acetone, and blocked for 1 h with 5% FCS in PBS. After incubation for 1 h with monoclonal antibody (MAb) vp4G2 (anti-HHV-8 VP23), kap5C4 (anti-HHV-8 K12), or 817 (anti-EBV VCA) (Chemicon, Temecula, Calif.) diluted in 5% FCS in PBS, the cells were incubated for 2 h with Cy3-conjugated goat anti-rat IgG or goat anti-mouse IgG (Sigma) diluted 1:200 in 5% FCS in PBS. MAbs vp4G2 (23a) and kap5C4 (21a) were generated as described elsewhere. The cells were washed and examined under a TNT confocal microscope DMIRB (Leica, Bensheim, Germany).

Immortalization of PBMC and persistent viral infection by EBV and HHV-8.

Supernatants of BCBL-1 (HHV-8⁺ PEL cell line), BC-1 (HHV-8⁺ EBV⁺ PEL cell line), B95-8 (EBV⁺ marmoset cell line), and Raji (human EBV-transformed cell line with a deletion in the EBV genome resulting in an inability to release virus) cells were used to infect PBMC from four EBV⁻ and four EBV⁺ individuals. Incubation of PBMC with control supernatant or supernatant of Raji cells did not result in the outgrowth of LCLs, independent of EBV status, indicating that the EBV present in PBMC from EBV⁺ individuals was not sufficient to transform B cells under the conditions used (Table 1). Incubation with supernatants of B95-8 resulted in the outgrowth of LCLs in PBMC of each individual, again independent of EBV status. Supernatants of HHV-8⁺ BCBL-1 cells resulted in the outgrowth of LCLs if PBMC were derived from EBV⁺ but not from EBV⁻ individuals. Interestingly, supernatants of HHV-8⁺ and EBV⁺ BC-1 cells resulted in the immortalization of PBMC from both EBV⁻ and EBV⁺ individuals. Since supernatants from BCBL-1 cells (HHV-8⁺ EBV⁻) immortalized PBMC from EBV⁺ but not EBV⁻ donors, we hypothesized that HHV-8 either provides a cofactor promoting the immortalization of EBV-transformed B cells or itself has transforming activity if EBV or EBV-derived cofactors are present. Since supernatants of BCBL-1 cells did not promote the immortalization of B lymphocytes derived from EBV⁻ donors in the assay system we used, we thus far have no hints for the latter.

Initially, six cell lines were established from PBMC of each donor and supernatant. PBMC from one donor were used twice, with a similar outcome. Thus, a total of 54 cell lines were established by using either BCBL-1 or BC-1 supernatant. HHV-8⁺ EBV⁺ LCLs were tested after various passages for the presence of EBV and HHV-8 DNA by PCR (Fig. 1). Both EBV and HHV-8 were present in all cell lines derived from PBMC of EBV⁺ donors infected with supernatant of the HHV-8⁺ cell lines BC-1 and BCBL-1, as well as in cell lines derived from EBV⁻ donors infected with supernatant of BC-1 cells. In the experiment in Fig. 1, two LCLs, ABEH-1 and GMEH-1, were tested after 3, 6, 9, 12, 15, 18, and 21 passages and found to be positive for both EBV and HHV-8 with no apparent decline in viral load. Four immortalized HHV-8⁺ EBV⁺ cell lines are continuously kept in culture for currently more than 25 passages and longer than 9 months. The other HHV-8⁺ EBV⁺ cell lines were frozen in liquid nitrogen since they appeared to be similar. Once the HHV-8⁺ EBV⁺ cell lines were established, there was no difference in terms of the efficiency of long-term culture in comparison to EBV⁺ LCLs. All four cell lines kept in continuous cell culture for more than 9 months were found to be positive for both EBV and HHV-8. Furthermore, none of the HHV-8⁺ EBV⁺ LCLs have become negative for either HHV-8 or EBV thus far.

Phenotypic analysis of EBV⁺ HHV-8⁺ LCLs.

Next, we determined the phenotype of the immortalized cell lines. All HHV-8⁺ EBV⁺ LCLs tested were positive for CD19 (B-lymphocyte antigen) but negative for CD3 (T-lymphocyte marker), CD14 (monocytes/macrophages), and CD16 (NK cells) (Fig. 2), indicating that these cells are of B-lymphocyte origin. Moreover, HHV-8⁺ EBV⁺ LCLs expressed surface IgM, as well as B-cell activation markers CD23, CD30, and CD80 similarly to EBV-transformed LCLs (Table 2). Surface IgG, CD20, and CD56 were not found on these cells. The phenotype differed from BCBL-1 and BC-1 cell lines by some markers; e.g., BCBL-1 cells were negative for CD80, and BC-1 cells were negative for CD30. Intriguingly, the surface phenotype as well as growth characteristics varied between LCLs from different donors. Whereas most of the HHV-8⁺ EBV⁺ LCLs grow in large conglomerates in suspension similar to EBV-transformed LCLs, cells of the HHV-8⁺ EBV⁺ LCL COEH-1 show semiadherent growth and are spindle shaped (data not shown).

Latent infection of transformed B-cell lines by HHV-8.

We next tested HHV-8⁺ EBV⁺ LCLs for expression of the latent viral transcript T0.7, which codes for kaposin. T0.7 was detected in the noninduced HHV-8⁺ EBV⁺ LCL GMEH-1 after 5, 10, and 15 passages by RT-PCR, and expression could be increased by induction with phorbol ester and n-butyrate (Fig. 3). The signal resulting from the amplification of actin mRNA was similar for all samples tested, indicating that the RNA content was comparable. The RT-PCR result was confirmed by Northern blot analysis, which revealed the presence of T0.7 in unstimulated ABEH-1 and GMEH-1 cells after more than 25 passages. In accordance with the RT-PCR results, kaposin protein was detected in the noninduced HHV-8⁺ EBV⁺ LCLs GMEH-1 and ABEH-1 by immunofluorescence with the MAb kap5C4 (Fig. 4). Kaposin was also detected on BCBL-1 cells but not on B95-8 cells.

Induction of lytic viral proteins in HHV-8⁺ EBV⁺ LCLs.

We next investigated whether latent HHV-8 can be reactivated in HHV-8⁺ EBV⁺ LCLs. GMEH-1 and ABEH-1 LCLs after 5, 10, and 15 culture passages were stimulated with phorbol ester and n-butyrate and subsequently tested for lytic viral transcripts coding for VP23 by RT-PCR (Fig. 3). In all immortalized HHV-8⁺ EBV⁺ LCLs tested, lytic HHV-8 mRNA was detected after induction by phorbol ester and n-butyrate, similar to BC-1 and BCBL-1 cells. In accordance with the RT-PCR results, the viral capsid protein VP23 was detected by immunofluorescence with the MAb vp4G2 in cells of the HHV-8⁺ EBV⁺ LCL ABEH-1 and in BCBL-1 cells. No reactivity was seen with B95-8 (Fig. 4) and EBV⁺ LCLs of the same donors (data not shown), indicating that there was no cross-reactivity with EBV proteins. Reactivity against EBV VCA was detected in B95-1 and in ABEH-1 cells but not in BCBL-1 cells.

Serial passage of HHV-8 from HHV-8⁺ EBV⁺ LCLs.

Since lytic HHV-8 genes could be induced in HHV-8⁺ EBV⁺ LCLs, we hypothesized that HHV-8⁺ EBV⁺ LCLs are able to produce infectious viral particles and infect new cells. PBMC of an EBV⁺ donor were incubated with supernatants of HHV-8⁺ EBV⁺ and EBV⁺ LCLs (Table 3). As anticipated, supernatants of the EBV⁺ LCLs ABE-1 and GME-1 immortalized B cells. Intriguingly, incubation with supernatants of HHV-8⁺ EBV⁺ LCLs also promoted the outgrowth of cell lines, whereas control supernatant of Raji cells or culture medium had no such effect. Six second-generation cell lines were established for each supernatant, and one second-generation cell line of each supernatant was tested by immunofluorescence analysis after two passages. Cell lines that were derived from PBMC incubated with supernatant of ABE-1 and GME-1 cells were found to be positive for EBV VCA and negative for HHV-8 kaposin. Cell lines that were derived from PBMC incubated with supernatant of cells of the HHV-8⁺ EBV⁺ LCLs ABEH-1 and GMEH-1 were positive for both HHV-8 kaposin and EBV VCA, indicating that HHV-8 has been transferred to these cells.