INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Epstein-Barr virus (EBV), a human herpesvirus, is the etiological agent of infectious mononucleosis and is associated with various lymphoid and epithelioid malignancies, such as Burkitt's lymphoma and nasopharyngeal carcinoma. In vivo, it has a narrow host range and is known to experimentally infect some New World monkeys, like the cotton-top tamarin and owl monkey (4, 20, 22, 28), but not other animal species. In vitro, EBV preferentially infects human and some nonhuman primate B lymphocytes and transforms them into indefinitely growing lymphoblastoid cells. EBV utilizes CD21 molecules for attachment to cells, as these are abundantly expressed on B lymphocytes (5). This probably explains why EBV infects B lymphocytes very efficiently but not other types of cells. Recently, we have demonstrated that epithelioid cells are also infectable with EBV in vitro, though the infective efficiency is much lower than that for B lymphocytes (9, 17, 32). Our studies have also indicated that the infection of epithelial cells is mediated via a receptor different from CD21, since infection is not blocked by pretreatment of cells with anti-CD21 antibodies (32). Besides binding to the cell surface, there are several steps for establishing infection. T-lymphocyte-derived Molt-4 cells are positive for CD21 expression and allow EBV binding, but the virus cannot penetrate the cells, because virus-cell fusion does not occur (15, 18). In EBV-infected cells, the viral genome is maintained as a plasmid and replicates once per cell cycle; otherwise, the virus is not transmitted to each daughter cell. Replication of the EBV plasmid is initiated at a unique site termed the latent origin of plasmid replication, oriP, in a sequence-specific manner (30) and from a broad region upstream of oriP (12) in a manner that resembles the delocalized initiation pattern in mammalian chromosomes (26). oriP contains EBV-determined nuclear antigen 1 (EBNA1)-binding sites, and EBNA1 binding to oriP is essential for stable replication and transmission of the oriP plasmid into daughter cells (30). Plasmids bearing oriP and the EBNA1 gene are maintained extrachromosomally in human cells, but murine cells cannot support plasmid replication (31), indicating that the barrier to establishing stable EBV infection exists at the step of plasmid maintenance.

Several attempts have been made to infect nonprimate mammalian cells with EBV (1, 2, 27), but all failed to establish stable infection. The present study aimed to establish stable EBV infection in nonprimate mammalian cells. This study was also based on the expectation that there might be a lytic host for EBV replication in nonhuman cells and that animal models might be developed based on the knowledge of what animal species are susceptible to EBV infection in vitro. To overcome the initial barrier of EBV infection, we generated an adenovirus vector carrying human CD21 cDNA. In addition, the use of recombinant EBV (rEBV) carrying a selectable marker (21) allowed us to select EBV-infected cells with certainty. The results indicated that one canine cell line and two rat cell lines were susceptible to EBV infection, and stably EBV-infected cell clones were isolated. The EBV genome was maintained as a plasmid in these canine and rat cells.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Cells and EBV. Seven nonprimate mammalian cell lines were used in this study, including one canine cell line, MDCK; one hamster cell line, BHK (13); four rat cell lines, Rat-1, 9L (29), cKDH (14), and c-SST-2 (8); and one mouse cell line, NIH 3T3 (10). They were maintained in Dulbecco's modified Eagle's medium (Sigma, St. Louis, Mo.) or minimum essential medium Eagle (Sigma) containing 10% fetal bovine serum (FBS; GIBCO BRL, Rockville, Md.) and antibiotics. The cultures were reseeded by treating the cells with 0.1 or 0.25% trypsin-1 mM EDTA-phosphate-buffered saline (PBS) solution every 3 days. Recombinant Akata EBV with the neomycin resistance gene was obtained by treating rEBV-infected Akata cells with anti-immunoglobulin G (IgG) antibodies (Dako, Glostrup, Denmark).

Construction and preparation of an adenovirus vector carrying human CD21 cDNA (Adv-CD21). Human CD21 cDNA (16) with the simian virus 40 promoter was cloned into the multicloning site of the shuttle plasmid pE1sp1A, which can be used to construct adenovirus type 5 vectors with inserts in early region 1 (E1) (Microbix Biosystems Inc., Toronto, Canada). pE1sp1A contains adenovirus sequences from bp 22 to 5790 with a deletion of E1 sequences from bp 342 to 3523. A polycloning site is present at the position of the deletion of the E1 gene. The plasmid pJM17 contains the full-length adenovirus genome with an insertion of a pBR322 derivative within the coding region of the E1 gene, which makes the viral genome too large to package. Cotransfection of both plasmids into 293 cells allows generation of infectious vectors of a packageable size by in vivo recombination. The plasmid pJM17 (3 µg) was cotransfected with the plasmid pE1sp1A (5 µg) into 293 cells at subconfluence in a 10-cm-diameter dish by the lipofection method. After 30 h of transfection, the cells were transferred to a 96-well plate at 100 cells/well/100 µl of fresh medium containing 1% FBS. Fifty microliters of medium was added every 3 days. Cytopathic effects appeared after 12 to 18 days of transfection, and the virus was harvested. A part of the virus preparation was used to infect CD21-negative HeLa cells, and CD21 expression was checked by flow cytometry. The titer of the virus was about 10⁹ PFU/ml.

Adv-CD21 and EBV infection. Nonprimate mammalian cells to be used as virus recipients were detached by treating them with 0.1 or 0.25% Trypsin-1 mM EDTA-PBS and were seeded into six-well culture plates at 8 × 10⁴ to 12 × 10⁴ per well with 3 ml of the medium. The next day, the culture medium was removed and 6 × 10⁸ PFU of Adv-CD21 in 1 ml of medium was added to the cultures. After a 90-min incubation, the cultures were washed and incubated for 2 days. For BHK cells, the FBS concentration of the medium was reduced to 3% to prevent cell overgrowth. The cells were then removed from the culture plates with 2 mM EDTA-PBS. The cells in one well were used to determine the CD21 expression by flow cytometric analysis. The cells in another well were centrifuged and suspended in 1 ml of rEBV preparation. After 2 h of incubation, the cells were washed, resuspended in fresh medium, and seeded in a well of a six-well plate. The next day, the cells were transferred to a 10-cm-diameter dish to prevent overgrowth. After 3 days of EBV infection, the cells were harvested and examined by immunofluorescence assay for the expression of EBNA. The remaining cells were reseeded into 24-well plates at 2 × 10⁴ per well in 1 ml of selecting medium containing G418 (200 µg/ml for rodent cells and 500 µg/ml for MDCK cells; GIBCO BRL). The medium was changed every 3 days until G418-resistant clones emerged (3 to 5 weeks).

Flow cytometric analysis. To examine the expression of CD21, cells were detached from the culture plates by treatment with 2 mM EDTA-PBS at 37°C for 10 to 30 min, washed with cold PBS containing 1% bovine serum albumin, and reacted with mouse monoclonal antibody (MAb) HB-5a (Becton Dickinson, Mountain View, Calif.). The second and third reactions were done with biotinylated goat anti-mouse Ig (Dako) and R-phycoerythrin-conjugated streptavidin (Dako), respectively, followed by flow cytometric analysis on a FACScan (Becton Dickinson). The cells were washed with PBS between reactions.

Immunofluorescence. Expression of EBNA was examined on acetone-methanol (1:1)-fixed cells by anticomplement immunofluorescence with reference human serum (titer, ×640). Expression of EBV lytic antigens was tested on acetone-fixed cells by indirect immunofluorescence with MAb R3 (19) (a gift of G. Pearson, Georgetown University, Washington, D.C.) specific to the viral early protein encoded by BMRF1, Cl-51 (23) specific to the viral capsid antigen gp110 encoded by BALF4, and C1 (24) (a gift of D. A. Thorley-Lawson, Tufts University, Boston, Mass.) specific to the viral envelope protein gp350/220 encoded by BLLF1. The second antibody was a fluorescein isothiocyanate-labeled F(ab')₂ fragment of a rabbit antibody to mouse IgG (Dako).

Immunoblotting. Cells were lysed in sodium dodecyl sulfate-polyacrylamide gel electrophoresis loading buffer, sonicated, and boiled for 5 min. A volume of lysate equal to 2 × 10⁵ to 4 × 10⁵ cells was separated in 10% polyacrylamide gels and transferred to a nitrocellulose membrane (Schleicher & Schuells, Dassel, Germany). After being blocked with 5% nonfat dry milk in Tris-buffered saline (TBS-M [pH 7.6]), the membrane was incubated for 2 h at room temperature with human serum diluted 1:50 in TBS-M to detect EBNAs, washed three times with TBS-M containing 0.1% Tween 20, and then serially reacted for 30 min with biotinylated rabbit anti-human IgG (diluted 1:500 in TBS-M [Dako]) and alkaline phosphatase-conjugated streptavidin (Amersham International plc, Little Chalfont, United Kingdom) (diluted 1:3,000 in TBS-M), with washing between reactions. Expression of EBNA2 and LMP1 was examined by using MAbs PE2 (a gift of E. Kieff, Harvard Medical School, Boston, Mass.) and CSl-4 (Dako), and antibody reaction and washing were done in TBS and TBS-0.1% Tween 20 solutions, respectively. The second antibody reaction was done with horseradish peroxidase-conjugated sheep antibodies to human IgG (Amersham) (diluted 1:3,000 in TBS-M). After the second antibody reaction, the filters were washed five times with TBS-0.1% Tween 20, immersed in the enhanced chemiluminescence solutions (Amersham) as specified by the manufacturer, and subjected to autoradiography.

RT-PCR. Reverse transcription (RT)-PCR analysis was carried out to investigate the expression of EBV latent and lytic genes and the utilization of EBNA gene promoters (Qp, Cp, and Wp) as described previously (9). Total cellular RNA was isolated by guanidium isothiocyanate-phenol extraction using TRIzol reagent (GIBCO BRL) according to the manufacturer's protocol. The extracted RNA was heated for 10 min at 70°C and rapidly cooled on ice. cDNA synthesis was performed for 60 min at 37°C with Moloney murine leukemia virus reverse transcriptase (GIBCO BRL) using 100 pmol of random hexamer (Takara, Otsu, Japan) followed by 10 min of heating at 94°C to inactivate the reverse transcriptase. The cDNA samples were then subjected to 30 cycles of PCR in a thermal cycler. Each cycle consisted of denaturation for 30 s at 94°C, annealing for 30 s at 45 to 55°C, and extension for 1 min at 72°C. The reaction mixture contained buffers and reagents as described previously (9), with 20 pmol of each primer and cDNA (equivalent to 5 × 10⁴ cells/tube) in a volume of 50 µl. Five microliters of the PCR products was electrophoresed on a 2% agarose gel and blotted onto nylon membranes (Hybond N⁺; Amersham), and specific amplified DNA was detected by ECL 3'-oligolabeling and detection systems (Amersham). The quality of RNA was checked by parallel amplification of -actin mRNA.

Gardella gel analysis. Analysis of linear and circular viral DNAs in rEBV-infected mammalian cells was carried out by the method of Gardella et al. (7) with modifications. The cells were lysed in a well of agarose gel. Cellular DNA and integrated viral DNA are too large to enter the gel, but circular and linear EBV DNA will enter it. Different mobilities of circular and linear EBV DNA can be recognized. A horizontal gel of 0.75% agarose (20 by 20 cm) was prepared, and the area above the wells (2.5 by 12 cm) was removed and replaced with 0.8% agarose containing 2% sodium dodecyl sulfate and 1 mg of pronase (Sigma)/ml. Cells (1.5 × 10⁶ to 2 × 10⁶) were placed in each well at 4°C. After electrophoresis at 0.7 V/cm at 4°C for 3 h, the voltage was increased to 3.5 V/cm for an additional 14 h. The viral DNAs were transferred to a nylon membrane (Amersham), hybridized with a ³²P-labeled BamHI-W fragment of Akata EBV DNA, and detected by autoradiography.

Southern blot analysis. Purified cellular DNA (5 µg) was digested with EcoRI, size fractionated by electrophoresis in a 0.8% agarose gel, and transferred to a nylon membrane (Amersham). To detect the EBV genome, the EcoRI-K fragment of Akata EBV was used as a probe. Probe labeling and detection of viral DNA signals were carried out using the ALKphos Direct kit (Amersham) following the manufacturer's instructions.

Real-time quantitative RT-PCR. Quantitative PCR was performed in 20-µl glass capillary tubes using a Lightcycler (Roche Molecular Biochemicals), which was equipped with a thermal cycler and real-time detector of fluorescence. Fifty nanograms of first-strand cDNA was amplified specifically by PCR at a final concentration of 1× PCR reaction buffer (GIBCO BRL), 0.01% bovine serum albumin (Takara), 3 mM MgCl₂, 200 µM deoxynucleoside triphosphate (Takara), 0.05 U of recombinant Taq polymerase (GIBCO BRL)/µl, 0.5 µM (each) sense and antisense gene-specific primers, and a final concentration of 1:50,000 of SYBR green I (FMC Bioproducts). The sequences of the primers for each gene are listed in Table 1.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

CD21 expression on Adv-CD21-infected nonprimate mammalian cells. No CD21 expression was found on cells of the seven nonprimate mammalian cell lines by flow cytometric analysis. To force CD21 expression on these cells, we generated an adenovirus vector carrying human CD21 cDNA under control of a simian virus 40 promoter (Adv-CD21). After 2 days of Adv-CD21 infection, expression of CD21 was examined by flow cytometric analysis. As shown in Fig. 1, all cells of the seven cell lines became positive for CD21. An adenovirus vector carrying the Escherichia coli lacZ gene was used as a negative control and yielded no CD21 expression upon infection of these cells. The levels of CD21 expressed on Adv-CD21-infected nonprimate mammalian cells were similar to that on Akata cells but were lower than on Raji cells.

View larger version (29K): [in this window] [in a new window]   FIG. 1.   CD21 expression on nonprimate mammalian cells after infection with an adenovirus vector (Adv-CD21) carrying cDNA of the human CD21 gene. After 2 days of Adv-CD21 infection, the cells were examined by flow cytometry for the expression of CD21. Cells of each cell line infected with an adenovirus vector carrying the E. coli lacZ gene (Adv-LacZ) were used as negative controls. CD21-positive Raji and Akata cells (both derived from Burkitt's lymphoma) were used as positive controls. The solid and dotted lines indicate staining with an anti-CD21 MAb (HB5a) and isotype control, respectively. The vertical axis denotes the number of cells counted, and the horizontal axis denotes fluorescence intensity (log scale).

EBV infection of CD21-expressing nonprimate mammalian cells. CD21-expressing nonprimate mammalian cells were infected with rEBV carrying a neomycin resistance gene. Three days postinfection, EBNA expression was determined by anticomplement immunofluorescence staining. EBNA was detected in 0.3% of MDCK cells and 0.1% of 9L cells. Expression of EBNA in c-SST-2 cells could not be determined because of a high background in nuclear staining. Cells of the other four cell lines were negative for EBNA expression. rEBV-infected cells were maintained in medium containing G418. G418-resistant cell clones emerged in three of the seven cell lines, MDCK, 9L, and c-SST-2. All of these cell clones were nearly 100% EBNA positive (Table 2 and Fig. 2).

View larger version (20K): [in this window] [in a new window]   FIG. 2.   Immunofluorescence staining of EBNA in recombinant EBV-infected, G418-resistant clones of nonprimate mammalian cells. Non-EBV-infected cells are shown as negative controls.

EBV gene expression in rEBV-infected nonprimate mammalian cells. rEBV-infected MDCK, 9L, and c-SST-2 cells were examined for the expression of EBV latent genes. EBNAs and LMP1 were examined by immunoblotting, and other EBV latent genes and EBNA1 promoter usage were examined by RT-PCR. MDCK cell clones were positive for EBV-encoded small RNA (EBER) and EBNA-1, LMP2A, and BARF0 genes but negative for the other EBV genes. EBNA1 promoter analysis indicated that Qp was active, while Cp and Wp were silent (Fig. 3). LMP2A expression was lower than that in lymphoblastoid cell lines immortalized by EBV infection (LCL). These patterns of viral gene expression are those typically observed in Burkitt's lymphoma (termed latency I). In 9L and c-SSTT-2 cells, all EBV latent genes, including those for six EBNAs, three LMPs, and BARF0 and EBER, were detectable, similar to the pattern observed in LCL (termed latency III), though the expression levels of EBNA2, EBNA3s, LMP1, and LMP2B were lower than in LCL. EBNA1 promoter usage studies revealed that all promoters, Cp, Wp, and Qp, were active in EBV-infected rat cells (Fig. 4A). To make clear whether EBV-infected rat cells were a mixture of latency I and latency III cells, cell clones were isolated from each EBV-infected rat cell line by the limiting-dilution method. As a result, all cell clones examined showed a pattern of EBV expression similar to that of parental cells and utilized all three EBNA promoters (Fig. 4B), suggesting that EBV-infected rat cells are a mixture of latency I and latency III cells at the clonal level.

View larger version (28K): [in this window] [in a new window]   FIG. 3.   Immunoblot analysis of EBV latent gene expression in EBV-infected nonprimate mammalian cells. The top and middle blots were treated with EBNA2 and LMP1 MAbs, respectively. The bottom blot was treated with a standard EBNA-positive human serum. LCL, a B-lymphoblastoid cell line immortalized with Akata EBV as a positive control.

View larger version (47K): [in this window] [in a new window]   FIG. 4.   (A) RT-PCR analysis of EBV latent gene expression and EBNA promoter usage in EBV-infected nonprimate mammalian cells. Akata cells were used as a positive control for detection of Qp-initiated EBNA mRNA, and LCL was used as a positive control for detection of other latent gene products and Cp- or Wp-initiated EBNA mRNAs. 9L cells served as a negative control. (B) Analysis of the clonal difference in EBNA gene promoter usage. Cell subclones from EBV-infected 9L and c-SST-2 cell clones were examined.

Form of EBV DNA in EBV-infected nonprimate mammalian cells. To examine whether EBV DNA was maintained as a plasmid or integrated into cellular DNA in EBV-infected cells, we performed Gardella gel analysis. This method allows resolution of circular and replicating linear viral DNAs from EBV-infected cells. As shown in Fig. 5, neither linear nor circular EBV DNA was detected in non-EBV-infected Akata cells, whereas both linear and circular viral DNAs were present in EBV-positive Akata cells. Linear viral DNA in Akata cells reflected spontaneous EBV replication and increased greatly when the cells were treated with anti-Ig antibodies. In all EBV-infected nonprimate mammalian cells, a band corresponding to circular viral DNA was detected, indicating that EBV DNA was maintained as a plasmid in the cells. Furthermore, MDCK cell clones and one of two 9L clones contained linear viral DNA, reflecting spontaneous induction of viral replication.

View larger version (36K): [in this window] [in a new window]   FIG. 5.   Gardella gel analysis of EBV-infected nonprimate mammalian cells. Akata, EBV-negative Akata cells; Akata+, EBV-positive Akata cells; Akata+(Ind), EBV-positive Akata cells treated with anti-Ig antibodies. EBV-positive Akata cells contain ~20 copies of the EBV genome per cell in a closed circular form. Linear viral DNA in Akata cells reflects spontaneous EBV replication, in which amplification of linear viral DNA was induced after anti-Ig treatment.

Induction of lytic infection in EBV-infected nonprimate mammalian cells. For each of the EBV-infected cell lines, MDCK, 9L, and c-SST-2, two clones were examined by immunofluorescence assay for the expression of the early antigens BMRF1, capsid antigen gp110, and envelope protein gp350/220 (Table 3). Spontaneous activation of BMRF1 protein was observed in the two MDCK clones and one 9L clone, but gp110 and gp350/220 were not detected in any cell clones examined. In MDCK cells, BMRF1 protein was induced in 68 to 70% of cells by simultaneous treatment with 3 mM n-butyric acid (n-BA) and 20 ng of 12-O-tetradecanoylphorbol-13-acetate (TPA)/ml. However, the induction level of gp110 was lower (16 to 20%), and gp350/220 could not be detected in MDCK cells. The same was seen in 9L and c-SST-2 cells. In contrast, in Akata cells, induction of BMRF1 protein accompanied induction of gp110 and gp350/220, followed by the production of progeny viruses. RT-PCR analysis also indicated that gp350/220 was not induced in MDCK, 9L, and c-SST-2 cells (Fig. 6). On the other hand, Southern blot analysis indicated that EBV DNA replication was efficiently induced in the two MDCK clones, one 9L clone, and one c-SST-2 clone (Fig. 7). We further examined whether infectious viruses were produced in these cells. Culture supernatants obtained from n-BA- and TPA-treated cells were inoculated into cord blood B lymphocytes, but EBV-immortalized lymphoblastoid cell lines could not be obtained. These results indicated that MDCK, 9L, and c-SST-2 cells were susceptible for inducing signals for lytic infection but could not support the full virus program.

View larger version (25K): [in this window] [in a new window]   FIG. 6.   Real-time quantitative RT-PCR analysis of EBV lytic gene expression in EBV-infected nonprimate mammalian cells. To induce the lytic cycle, MDCK, 9L, and c-SST-2 cells were treated with 3 mM n-BA plus 20 ng of TPA/ml for 3 days, and Akata cells were treated with anti-Ig for 2 days. All analyses were performed by real-time quantitative RT-PCR assay using a Lightcycler (6). The results are expressed as the ratio to the value of GAPDH (K × number of cytokine copies/5 × 103 copies of GAPDH; K, constant).

View larger version (26K): [in this window] [in a new window]   FIG. 7.   Induction of EBV replication in EBV-infected nonprimate mammalian cells. To induce the lytic cycle, MDCK, 9L, and c-SST-2 cells were treated with 3 mM n-BA plus 20 ng of TPA/ml for 3 days, and Akata cells were treated with anti-Ig for 2 days. Five micrograms of cellular DNA was digested with the EcoRI restriction enzyme, blotted, and hybridized with an EBV EcoRI-K probe of Akata EBV. Akata+, EBV-positive Akata cells.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

EBV readily infects human B cells by binding to CD21, which is the receptor for EBV, on the cell surface. The absence of the human CD21 molecule is the first barrier to the establishment of EBV infection of nonprimate mammalian cells. To overcome this barrier, an adenovirus vector carrying human CD21 cDNA was introduced, and CD21 was artificially expressed on nonprimate mammalian cell lines. We have presented evidence that EBV infected three of seven mammalian cell lines in a CD21-dependent manner when CD21 was expressed exogenously. There have been several reports of successful EBV infection of nonprimate mammalian cells by introducing the human CD21 gene, but all were transient infections (1, 2, 27). This is the first report of stable EBV infection in nonprimate mammalian cells. In our previous studies, we demonstrated that CD21-negative human epithelial cells were susceptible to EBV infection (9, 17, 32) and that infection was mediated via a new receptor different from CD21 (9, 32). However, without CD21 transfer, none of the nonprimate mammalian cells examined here could be infected with EBV, suggesting that nonprimate mammalian cells had this new receptor deleted.

Unlike Akata cells, nonprimate cells displayed relative resistance to EBV infection, though the barrier of EBV binding was overcome. The expression levels of CD21 in all seven nonprimate cell lines after CD21 transfer were similar to that in Akata cells, though the frequencies of EBV-infected cells, assayed by EBNA expression 3 days postinfection, were much lower in nonprimate cells than in Akata cells, even though the same virus preparation was used for infection. Moreover, the frequencies of EBNA-positive cells that became stably EBV infected were much lower in nonprimate mammalian cells than in Akata cells (Table 2). These results indicate that barriers against EBV infection exist at the steps of virus adsorption, entry, expression of viral genes, and maintenance of the viral genome in nonprimate mammalian cells. The EBNA1/oriP-based vector has been successfully used in numerous studies in human cells. In contrast, there have been only two reports describing the successful use of an EBNA1/oriP-based vector in rat glioblastoma and pheochromocytoma cells (11, 25), and the first report of the EBNA1/oriP vector described a failure of EBV episome maintenance in several murine cell lines (31). Our results also indicate that, among rodent cells, at least rat cells can support stable replication of the EBV plasmid.

Another interesting aspect of this work is the efficient induction of the lytic cycle but failure to synthesize the envelope protein gp350/220 in EBV-infected nonprimate cells. Similar results have recently been reported in transiently EBV-infected MDCK cells (2). The discordance between early and late viral proteins is often seen in human B-lymphocyte cultures (3). It remains to be clarified whether the abortive replicative infection in MDCK, 9L, and c-SST-2 cells reflects intrinsic disabilities of canine and rat species.

Some New World monkeys can be infected by EBV. Both the cotton-top tamarin and the owl monkey are susceptible to EBV-induced B-cell lymphomas (4, 22). An infectious mononucleosis-like syndrome is induced in the EBV-infected common marmoset (28). However, there are no animals that can be persistently EBV infected without immediate lymphoma development, and monkeys are not easy to keep. The present results raise the possibility that transgenic animals with human CD21 could become models for EBV infection. Transgenic rats with human CD21 would be susceptible to EBV infection and would make it possible to study many aspects of EBV pathogenesis, including acute and persistent infections and their control by host immune responses. They would also be useful for evaluating drug and vaccine candidates against EBV infection.