Role of B-Cell Proliferation in the Establishment
http://www.100md.com
病菌学杂志 2005年第15期
Center for Emerging Infectious Diseases, Yerkes National Primate Research Center, Emory University School of Medicine, Atlanta, Georgia 30329
Graduate Program in Molecular Cell Biology, Washington University School of Medicine, St. Louis, Missouri 63110
Departments of Immunology and Pediatrics, University of Washington, Seattle, Washington 98195
ABSTRACT
Murine gammaherpesvirus 68 (HV68) provides a tractable small animal model with which to study the mechanisms involved in the establishment and maintenance of latency by gammaherpesviruses. Similar to the human gammaherpesvirus Epstein-Barr virus (EBV), HV68 establishes and maintains latency in the memory B-cell compartment following intranasal infection. Here we have sought to determine whether, like EBV infection, HV68 infection in vivo is associated with B-cell proliferation during the establishment of chronic infection. We show that HV68 infection leads to significant splenic B-cell proliferation as late as day 42 postinfection. Notably, HV68 latency was found predominantly in the proliferating B-cell population in the spleen on both days 16 and 42 postinfection. Furthermore, virus reactivation upon ex vivo culture was heavily biased toward the proliferating B-cell population. DNA methyltransferase 1 (Dnmt1) is a critical maintenance methyltransferase which, during DNA replication, maintains the DNA methylation patterns of the cellular genome, a process that is essential for the survival of proliferating cells. To assess whether the establishment of HV68 latency requires B-cell proliferation, we characterized infections of conditional Dnmt1 knockout mice by utilizing a recombinant HV68 that expresses Cre-recombinase (HV68-Cre). In C57BL/6 mice, the HV68-Cre virus exhibited normal acute virus replication in the lungs as well as normal establishment and reactivation from latency. Furthermore, the HV68-Cre virus also replicated normally during the acute phase of infection in the lungs of Dnmt1 conditional mice. However, deletion of the Dnmt1 alleles from HV68-infected cells in vivo led to a severe ablation of viral latency, as assessed on both days 16 and 42 postinfection. Thus, the studies provide direct evidence that the proliferation of latently infected B cells is critical for the establishment of chronic HV68 infection.
INTRODUCTION
Epstein-Barr virus (EBV) is a human lymphotropic virus that establishes a life-long infection, predominantly in memory B cells (17, 35, 36). EBV has been associated with numerous human malignancies, including Burkitt's lymphoma, Hodgkin's lymphoma, and posttransplantation lymphoproliferative disease, and is capable of transforming B cells in vitro (34). Since EBV has a very limited host range, the study of this virus in vivo is difficult. Murine gammaherpesvirus 68 (HV68) is a closely related gammaherpesvirus that infects small rodents and has recently been used to characterize the pathogenesis of gammaherpesviruses in vivo (31). Similar to EBV, HV68 establishes latency in B cells and persists for the lifetime of the host in the memory B-cell compartment (10, 42).
In the peripheral blood, EBV latency is restricted to resting memory B cells, while at sites of active virus replication, such as the tonsils, EBV infection of na?ve, germinal-center, and memory B cells can be detected (3, 4). The current model for the establishment of EBV latency suggests that virus infection initiates through the infection of any resting B cell, including the na?ve B-cell population (reviewed in references 17, 35, and 36). As a consequence of the expression of a subset of viral genes, referred to as the growth program, EBV is able to activate infected B cells and drive these cells to proliferate (15). These proliferating lymphoblasts are then thought to migrate and form germinal centers. Germinal centers are the sites of B-cell differentiation during the immune response that lead to the development of antibody-producing plasma cells and memory B cells. During the germinal-center reaction, the EBV latency-associated membrane antigens LMP1 and LMP2A are thought to mimic signals normally provided through CD40 and the B-cell receptor, respectively, helping to drive differentiation to memory B cells and promote cell survival.
This model is supported by data that illustrate that there are distinct expression patterns of EBV latency genes found in each differentiation stage of the infected B cell (4). EBV was observed to be present in all B-cell subsets in the tonsils of individuals who had ongoing viral replication. However, in the rare tonsils that had no evidence of ongoing viral replication, no infected na?ve B cells or germinal center B cells were detected, and all the virus-infected cells were in the memory compartment (4). These data further support the hypothesis that the longevity of the memory B-cell compartment provides a stable reservoir for EBV latency while allowing for reactivation and reseeding latency to the memory B-cell reservoir through the infection of na?ve B cells present at the site of virus shedding.
Although there are several studies that support this model, there are several points that have yet to be proven experimentally. Some of the ideas discussed above have been developed from parallels between viral gene expression and normal B-cell biology. The role of the immune response and the presence of antigen have yet to be fully deciphered during the germinal center reaction. Alternative models have also been proposed for EBV infection and establishment of latency. Immunohistochemical and in situ hybridization studies have shown that EBV-infected B cells are rarely seen in germinal centers in the tonsils, but instead are found in the extrafollicular regions (18). These data possibly suggest that EBV-infected B cells acquire the phenotype of germinal-center B cell but may not actually participate in the germinal-center reaction. An alternative model is that EBV does not participate in the germinal center reaction, but rather directly infects memory B cells. EBV-positive and -negative germinal-center B cells isolated from sections of tonsils taken from infectious mononucleosis patients by micromanipulation were analyzed for somatic hypermutation of their V genes (19). EBV-positive cells located in the germinal centers proliferated without ongoing somatic hypermutation compared to the EBV-negative germinal-center B cells from the same sections. The authors of that study concluded that EBV-infected cells do not participate in the germinal-center reaction during infectious mononucleosis.
The difficulty with experimentally proving the proposed models is directly related to the limitations in manipulating EBV infection in vivo. In an attempt to begin addressing the questions raised above, HV68 infection in mice is being utilized to track gammaherpesvirus infection in vivo. Previously, it was shown that at early stages of HV68 latency following intranasal infection (day 16 postinfection), virus infection of na?ve, germinal-center, and memory B cells is observed (10, 42). However, by later latency times (3 and 6 months postinfection), viral genomes are only detectable in germinal-center and memory B-cell populations (42). Again, this leaves unresolved the question of whether there is direct infection of na?ve, germinal-center, and memory B cells, with the memory B-cell population being the long-lived population. To address this question, we investigated whether B-cell proliferation is a critical intermediate step in the establishment of latency in the memory B-cell reservoir. Here we show that HV68 is predominantly found in the proliferating B-cell population on days 16 and 42 postinfection. These data correlate with previous data from our lab showing a greater frequency of virus present in the germinal-center B-cell population at these times postinfection (42).
To determine whether B-cell proliferation is a critical intermediate step in the establishment of latency in the memory B-cell compartment, we utilized a conditional knockout mouse to delete the gene encoding DNA methyltransferase 1 (Dnmt1), an essential enzyme needed during cellular proliferation (9, 14, 21). Dnmt1 is the maintenance methyltransferase, which preferentially methylates hemimethylated DNA (5). The Cre-recombinase-mediated deletion of Dnmt1 causes the demethylation and uniform cell death of several different cell types, including cultured fibroblasts and B-cell progenitors, in vitro (7, 14). In vivo, the conditional deletion of Dnmt1 in postmitotic neurons did not affect the levels of global methylation or cell survival during postnatal life, but a Dnmt1 deletion in proliferating mitotic central nervous system precursors led to DNA hypomethylation and functionally impaired cells that were selected against at postnatal stages (9). Finally, the deletion of Dnmt1 impaired the proliferation and survival of T-lineage cells (21). Here we show that the deletion of Dnmt1 in HV68-infected cells severely impairs the establishment of latency at day 16 postinfection and that by day 42 postinfection, latency in the spleen is almost completely ablated. These studies argue persuasively that HV68-infected B cells undergo several rounds of proliferation prior to establishing latency in the resting memory B-cell compartment.
MATERIALS AND METHODS
Mice, viruses, and cells. C57BL/6J mice were purchased from Jackson Laboratory (Bar Harbor, Maine). Dnmt2lox/2lox mice were constructed as previously described (21). Briefly, these mice contain loxP sites flanking exons 4 and 5 of the Dnmt1 gene. Cre-recombinase-mediated deletion leads to an out-of-frame splicing from exons 3 to 6, resulting in a null Dnmt1 allele (9, 21). All protocols for animal studies were approved by the Institutional Animal Care and Use Committee of Emory University. Mice were used at 8 to 12 weeks of age. HV68 strain WUMS (ATCC VR-1465) was used for all wild-type virus infections. Virus passaging, maintenance, and titration were all performed as previously described (8). Mice were anesthetized with isofluorane and inoculated intranasally with 1,000 PFU of virus in 20 μl of Dulbecco modified Eagle medium (DMEM). NIH 3T12 cells and mouse embryonic fibroblast (MEF) cells were maintained at 37°C and 5% CO2 in DMEM supplemented with 10% fetal calf serum (FCS), 100 U of penicillin per ml, 100 mg of streptomycin per ml, and 2 mM L-glutamine (cMEM). MEF cells from C57BL/6 embryos were obtained as previously described (26).
Construction of HV68-Cre virus. To generate the targeting construct for the introduction of Cre-recombinase into HV68, the genomic region between nucleotide positions 45237 and 48347 was excised as a HindIII-SacI fragment from a pUC13 plasmid vector carrying the entire BamHI C HV68 genomic fragment (22). The fragment was then reinserted between the HindIII and SacI sites of pUC13 to create the plasmid JE109. JE109 was then digested with HindIII, and a HindIII/EcoRI adaptor (Ezclonesystems, New Orleans, LA) was inserted to generate plasmid JE110. A kanamycin resistance cassette (Kanr) flanked by FLP recombinase target sites was excised as a KpnI-HindIII fragment from the plasmid pCP15 and inserted into KpnI-HindIII-digested pMECA to generate the plasmid pMECAKan. The Cre-recombinase expression cassette from pCREin was excised with PacI and inserted into the PacI site of pMECAKan to generate pMECAKanCre (30). The Cre-recombinase expression cassette from pCREin includes the human cytomegalovirus (HCMV) immediate-early promoter upstream of Cre. In addition, the Cre gene contains an engineered intron to prevent the expression of this protein in bacterial cells (30). The Cre expression cassette and the kanamycin resistance cassette were excised from pMECAKanCre as an EcoRI-NsiI fragment. Following treatment with Klenow, the fragment was inserted into the PmlI site of JE110 (corresponding to HV68 genomic position 46347) to generate JE111. To isolate the targeting construct for use in recombination, JE111 was digested with EcoRI. The insert DNA was gel purified twice and subsequently electroporated into Escherichia coli strain DH10B containing the HV68 bacterial artificial chromosome (BAC) and a temperature-sensitive RecA expression plasmid, p2650. The p2650 plasmid contains a temperature-sensitive origin of replication which is unable to function at 42°C, leading to the loss of p2650 function in cells grown at 42°C. Following transformation, the cells were allowed to recover at 37°C for 2 hours and then grown on chloramphenicol (Cam) and kanamycin (Kan) selection media, with incubation at 42°C overnight followed by an additional 12 h at 37°C. HV68-BAC provides Camr in this selection process, and therefore only genomes which had recombined the Cre-Kanr cassette into the locus between ORF27 and ORF29b were selected. Camr Kanr colonies were replica plated onto carbenicillin selection medium to confirm the loss of p2650. Colonies were isolated, and the insertion of the Cre-Kanr cassette was confirmed by Southern blotting. To remove the Kanr cassette, electrocompetent cells were prepared from one such confirmed Camr Kanr clone. The competent cells were transformed with a plasmid expressing the FLP recombinase, i.e., pCP20, which has a temperature-sensitive origin of rep-lication that will not function at 42°C. Transformants were grown in medium containing Cam, Kan, and carbenicillin at 30°C overnight. Colonies were streaked onto Cam plates and grown at 42°C overnight to remove pCP20. Colonies were replica plated onto either chloramphenicol or kanamycin selection medium and grown overnight at 37°C. DNAs from Camr Kans transformants were analyzed by Southern blotting to confirm the loss of the Kanr cassette. HV68-Cre BAC DNA was transfected into NIH 3T12 fibroblast cells, and the virus was isolated as described previously (23). Self-excision of the BAC sequence was confirmed by Southern blotting and flow cytometry. The HV68 genome cloned as a BAC was a kind gift of Ulrich Kozinowski (1, 2). The pCREin plasmid was a generous gift from Greg Smith (30, 43).
Limiting-dilution PCR analysis. The frequency of cells harboring the HV68 genome was determined using a single-copy-sensitivity nested PCR assay as previously described (41). Briefly, cells were resuspended in isotonic buffer and subjected to a series of six threefold serial dilutions. Starting with 104 cells per well, cells were plated in a background of 104 uninfected NIH 3T12 cells in 96-well PCR plates. Twelve replicate PCRs were performed for each cell dilution per experiment. The cells were subjected to lysis by proteinase K at 56°C for 6 h, followed by enzyme inactivation for 95°C for 20 min. Round 1 and round 2 PCRs were performed as previously described on a PrimusHT thermal cycler (MWG Biotech, Highpoint, NC). Reaction products were separated in 2% agarose gels. All assays demonstrated near-single-copy sensitivities, with no false positives. All data analysis was performed with GraphPad Prism software (San Diego, CA). Data were subjected to nonlinear regression analysis with a sigmoidal dose-response algorithm for best fit. Frequencies of viral genome-positive cells were obtained from the nonlinear regression fit of the data where the regression line intersected 63.2%. Based on the Poisson distribution, this is the frequency at which there is at least one event present in a population.
Limiting-dilution ex vivo reactivation analyses. Limiting-dilution analysis to detect reactivation from latency was performed as previously described, with the following modifications and additions (40). Briefly, 24 hours prior to the assay, MEF cells were plated in 96-well or 24-well plates at 5 x 103 cells per well or 5 x 104 cells per well, respectively. Spleen cells obtained from infected mice on days 16, 42, 90, and 180 postinfection were resuspended in cMEM and plated in twofold serial dilutions starting with either 105 or 106 cells per well onto a MEF indicator monolayer. Twenty-four replicate wells were plated per dilution. The wells were scored microscopically for cytopathic effect (CPE) on the MEF monolayer at 21 days postplating. To detect preformed infectious virus, cells were hypotonically lysed while simultaneously being mechanically disrupted using a Mini-Beadbeater 8 and mechanical disruption beads (Biospec Products, Bartlesville, OK). Latent viruses cannot reactivate from dead cells, and this method kills >99% of cells. No measurable preformed infectious virus was detected at the specified time points postinfection.
Proliferation sorting. Single-cell suspensions of spleen cells were prepared from infected mice, and red blood cells were lysed. Cells were labeled and stained as previously described (13). Briefly, cells were washed once in phenol-free Hanks' balanced salt solution (HBSS) supplemented with 2% FCS (HBSS/2) and then resuspended with HBSS/2 containing 10 μmol Hoechst dye (Hst) (Molecular Probes, Eugene, OR). Hoechst 33342 was used for sorting with a UV laser (13), and Hoechst 34580 was used for sorting with a violet laser (28). The spleen cells were incubated for 45 min at 37°C. Pyronin Y (Sigma Chemicals, St. Louis, MO) was added to give a final concentration of 2.5 μg/ml, and the cells were incubated for an additional 45 min at 37°C. The cells were then stained with anti-mouse CD19 conjugated to allophycocyanin (APC) (BD Pharmingen, San Diego, CA) for a final 20 min on ice in the dark. The cells were then washed once in HBSS/2 with Hst and once with just HBSS/2. Cells were sorted on either a MoFlo (Cytomation, Fort Collins, CO) or FACSAria (BD Biosciences Immunocytometry Systems, San Jose, CA) flow cytometer. The purity of sorted cells always exceeded 90%.
BrdU staining. Mice were given bromodeoxyuridine (BrdU) in their drinking water for 8 days at a concentration of 0.8 mg/ml. Fresh BrdU water was given to the mice daily. Spleen cells were harvested from treated mice and stained for intracellular BrdU using a BrdU flow kit (BD Pharmingen) and following the manufacturer's recommended protocol. In brief, spleen cells were surface stained with anti-CD19-APC and anti-immunoglobulin D (anti-IgD)-fluorescein isothiocyanate (FITC) (BD Pharmingen) for 15 min on ice. Cells were washed, fixed, permeabilized in BD Cytofix/Cytoperm buffer for 15 min on ice, washed with BD Perm/Wash buffer, and incubated with BD Cytofix/Cytoperm Plus buffer for 10 min on ice. Spleen cells were washed with BD Perm/Wash buffer and refixed with BD Cytofix/Cytoperm buffer on ice for 5 min. Spleen cells were washed with BD Perm/Wash buffer and treated with DNase at a concentration of 300 μg/ml for 1 h at 37°C. Spleen cells were washed with BD Perm/Wash buffer and stained with anti-BrdU-FITC or anti-BrdU-APC diluted in BD Perm/Wash buffer, with incubation for 20 min at room temperature. Spleen cells were washed with BD Perm/Wash buffer and resuspended in phosphate-buffered saline with 2% FCS for analysis on a FACSCalibur instrument (BD Biosciences).
Plaque assays. Plaque assays were performed using NIH 3T12 monolayers under noble agar overlays as previously described (40), with the following changes. NIH 3T12 cells were plated in six-well plates at 2 x 105 cells per well the day prior to infection. Lungs were thawed and homogenized by mechanical disruption with beads (Immsilica beads; Biospec Products) in a Mini-Beadbeater 8 (Biospec Products, Bartlesville, OK), and then 10-fold serial dilutions were made in complete DMEM and plated on NIH 3T12 monolayers. Infections were performed in a 200-μl volume, and plates were rocked every 15 min for 1 hour at 37°C. Samples were overlaid with 3 ml of a 1:1 mixture of 3% noble agar and 2x minimum essential medium (MEM) supplemented with 20% FCS, 2x penicillin-streptomycin, and 2x L-glutamine. An additional 2 ml of noble agar with supplemented 2x MEM was added on day 3 postplating. Monolayers were stained on day 7 by the addition of 2 ml of neutral red overlay (0.01% neutral red in serum-free DMEM). After 18 to 24 h, plaques were counted. The limit of detection for this assay is 10 PFU per organ.
Immunoblots. NIH 3T12 fibroblast cells were infected with either HV68-WT or HV68-Cre at a multiplicity of infection (MOI) of 5 for 1 h at 37°C. The cells were then harvested at appropriate times and lysed in ELB buffer (50 mM HEPES, pH 7.2, 250 mM NaCl, 2 mM EDTA, 0.1% NP-40 substitute, and 0.01% protease inhibitor cocktail [Sigma-Aldrich]) for 20 min on ice. Sample buffer (2x) was added, and samples were boiled for 10 min before being subjected to standard sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes. Membranes were blotted with an anti-Cre antibody (1:10,000 dilution) (Novagen, San Diego, CA), a donkey anti-rabbit secondary antibody conjugated to Alexa Fluor 594 (Molecular Probes), and anti-?-actin directly conjugated to FITC (Sigma-Aldrich) and then visualized on a Typhoon 9400 instrument (Amersham Biosciences, Piscataway, NJ).
RESULTS
HV68 infection drives B-cell proliferation in vivo. HV68 undergoes a lytic infection in the lungs of mice after intranasal infection, which is followed by the establishment of a latent infection that persists for the lifetime of the host. Acute virus replication is cleared by days 9 to 15 postinfection, and the early and intermediate stages of latency are typically analyzed on days 16 and 42 postinfection, respectively (32, 37, 40, 42). On day 16 postinfection, the predominant latency reservoirs for HV68 are na?ve, germinal-center, and memory B cells, but by day 42 postinfection, HV68 genomes are found predominantly in the more mature surface IgD– B-cell population (10, 42). By 3 and 6 months postinfection, surface IgD– B cells (the majority of which have undergone heavy-chain isotype switching) are the only population harboring detectable levels of HV68 genomes (10, 42). It is currently unclear how HV68 accesses the memory B-cell reservoir. This could occur either through direct infection of existing memory B cells or by infecting na?ve B cells which subsequently go through the germinal-center reaction to become memory B cells. To further characterize the initial events in the establishment of B-cell latency, we assessed the levels of B-cell proliferation at early and late times during the establishment of latency.
To determine the level of B-cell proliferation on days 16 and 42 postinfection, bromodeoxyuridine (BrdU) was administered in the drinking water to mice for 8 days prior to the harvest of spleen cells. BrdU is a thymidine analog that is incorporated into the DNA of cells undergoing proliferation. On day 16 postinfection, approximately 35% of CD19+ B cells were BrdU positive, while on day 42 postinfection approximately 26% of CD19+ B cells were BrdU positive (Fig. 1A and Table 1), compared to approximately 10 to 15% BrdU-labeled CD19+ cells in na?ve mice (Table 1). Interestingly, ca. 60% of the IgD– B cells had proliferated within the 8 days before the harvest at day 42 postinfection, in contrast to only 37% of IgD– B cells harvested from na?ve mice (Fig. 1B). In addition, an increase in na?ve B-cell (surface IgD+) proliferation was also observed in mice at 42 days postinfection (16.3%) compared to na?ve mice (10.6%).
Proliferating B cells are the predominant latently infected cell population in the spleen at early and intermediate times postinfection. Since we observed a large number of proliferating B cells in the spleens of HV68-infected mice on days 16 and 42 postinfection, we wanted to determine whether HV68 infection was biased toward the proliferating B-cell population. Because the methods used to identify proliferating cells using BrdU labeling are incompatible with the analysis of HV68 latency, we employed an alternative approach. To fractionate live B cells into proliferating and nonproliferating populations, spleen cells were labeled with a Hoechst dye and pyronin Y in conjunction with staining of the cells with anti-CD19, followed by flow cytometry (11-13, 20). The Hoechst dye labels DNA content, allowing cells with greater than 2N DNA to be distinguished from cells containing 2N DNA, while pyronin Y stains RNA, discriminating between resting cells (G0 phase of the cell cycle) and cells in cycle (G1, S, G2, and M phases of the cell cycle) based on their RNA contents (11-13, 20). Based on this staining, approximately 30% of CD19+ cells were proliferating at day 16 postinfection (Fig. 2A), which was consistent with the percentage of proliferating splenic B cells determined by BrdU staining (Fig. 1A). To assess the accuracy of cell sorting, we determine the expression of D-type cyclins in each population. Cyclins D1, D2, and D3 are not appreciably expressed in quiescent cells (G0 phase of cell cycle) but are upregulated upon entry into the G1 phase of the cell cycle (29). An immunoblot for cyclins D1, D2, and D3 demonstrated an absence of detectable levels of D-type cyclins in the nonproliferating cell fraction and the presence of D-type cyclins in the proliferating population (data not shown).
To compare the levels of HV68 latency between the proliferating and nonproliferating B-cell populations, two standard measures of latency were used to determine the frequency of cells harboring latent virus ex vivo, namely, a limiting-dilution reactivation assay and a limiting-dilution nested PCR analysis (LDPCR) (8). The reactivation assay consists of limiting-dilution plating of spleen cells on a monolayer of mouse embryonic fibroblast cells (MEFs), which are permissive for virus growth, and monitoring the wells for cytopathic effect (CPE) on the MEF monolayer. Also plated in parallel are samples of cells that have been hypotonically lysed and mechanically disrupted in order to distinguish between preformed infectious viruses and viruses reactivating from latency. In the analyses presented here, there were never any significant levels of preformed infectious virus detected at any of the time points postinfection (data not shown). The LDPCR assay is performed following serial dilutions of virally infected spleen cells starting at 104 cells per well. The cells are lysed with proteinase K, followed by two rounds of nested PCR to detect the presence of the HV68 genome. For both assays, the data are subjected to nonlinear regression analysis with a sigmoidal dose-response algorithm for best fit.
On day 16 postinfection, there was a fivefold higher frequency of HV68 latently infected cells, as assessed by the LDPCR assay, in the proliferating B-cell population than in the nonproliferating B-cell population (Fig. 2B). Furthermore, when these populations were assayed for spontaneous reactivation ex vivo, there was a ca. 50-fold higher frequency of cells reactivating virus from the proliferating B-cell population than from the nonproliferating B-cell population (Fig. 2C), indicating that the virus reactivates more efficiently from latently infected proliferating B cells than from latently infected resting B cells. These data correlate with data from our laboratory demonstrating an increase in reactivation of HV68 from spleen cells ex vivo following stimulation with anti-Ig and anti-CD40 antibodies (24).
By day 42 postinfection, there was an even greater bias of HV68 latency toward the proliferating B-cell population, with a ca. 100-fold higher frequency of HV68 genome-positive cells present in the proliferating B-cell population than in the nonproliferating B-cell population (Fig. 2D and Table 1). These data correlated well with previous data from our lab which demonstrated that HV68 genomes are found predominantly in germinal-center (PNAhi) B cells on both days 16 and 42 postinfection (42). By 3 months postinfection, the percentage of splenic B cells that were proliferating had dropped to nearly background levels (Table 1) and the bias of viral latency toward proliferating B cells was also reduced compared to that observed on day 42 (Fig. 2E and Table 1).
Insertion of Cre-recombinase into HV68 between open reading frames 27 and 29b does not affect acute virus replication or latency in normal C57BL/6 mice. To investigate the role of B-cell proliferation in the establishment of HV68 latency, we utilized a mouse conditional for an enzyme that is required for the survival of proliferating, but not quiescent, cells. The conditional mouse we chose to employ targets the cellular enzyme DNA methyltransferase 1 (Dnmt1). The deletion of Dnmt1 in proliferating cells leads to hypomethylation of the cellular genome, ultimately leading to decreased cell replication and finally cell death (14). In conditional Dnmt1 knockout mice (Dnmt12lox/2lox), exons 4 and 5 of the Dnmt1 gene are flanked by loxP sites (21), such that the expression of Cre-recombinase leads to the deletion of these exons, resulting in a null Dnmt1 allele. To utilize these mice for our infection studies, we constructed a recombinant HV68 virus that expresses Cre-recombinase, as described in Materials and Methods. The Cre-recombinase expression cassette, driven by the HCMV immediate-early promoter, was targeted to the region of the viral genome between ORF27 and ORF29b of HV68 in E. coli (Fig. 3) (1, 30). The resulting recombinant virus, HV68-Cre, was shown to express Cre-recombinase by immunoblotting following the infection of NIH 3T12 murine fibroblasts (Fig. 4A).
A common strategy employed for the generation of herpesvirus BACs has been the insertion of loxP sites flanking the BAC sequences to allow removal of the BAC during virus production. Thus, recombinant HCMV and pseudorabies virus BACs have been generated that express Cre-recombinase, creating viruses capable of self-excising their BAC sequences (30, 43). Since the BAC vector in HV68-BAC, which includes a green fluorescent protein (GFP) expression cassette, is also flanked by loxP sites, we could assess whether the inserted Cre-recombinase expression cassette was functional in the context of virus infection. To do this, we monitored the loss of GFP expression as an indicator of excision of the BAC sequences following infection of NIH 3T12 fibroblasts. At a high multiplicity of infection (MOI = 1), there was a complete absence of GFP expression in the HV68-Cre-infected culture, in contrast to the case for the wt-BAC virus, as assessed by flow cytometry (Fig. 4B). The removal of the BAC vector from the recombinant virus was confirmed by Southern blot analysis (data not shown). Thus, the inserted Cre-recombinase expression cassette is functional in the context of the viral genome.
To determine whether the insertion of Cre-recombinase into the locus between ORF27 and ORF29b affects the pathogenesis of HV68 in vivo, C57BL/6 mice were infected intranasally with 1,000 PFU of either HV68-Cre or wild-type HV68 (HV68-WT). On days 4 and 9 postinfection, lungs were analyzed for levels of acute lytic replication by a plaque assay. HV68-Cre and the wild-type virus exhibited very similar levels of acute viral replication load at both time points assayed postinfection (Fig. 5A). To determine whether the insertion of Cre-recombinase into HV68 affects the establishment and maintenance of latency, the frequencies of latently infected spleen cells from mice inoculated with either HV68-Cre or HV68-WT were compared. On day 16 postinfection, there was no significant difference in the frequencies of latently infected spleen cells from mice infected with either HV68-Cre or HV68-WT, as determined by the limiting-dilution PCR assay or the limiting-dilution reactivation assay (Fig. 6A and B). Furthermore, by day 42 postinfection, there were nearly identical frequencies of HV68-Cre and HV68-WT latently infected cells (Fig. 6C). Thus, we concluded that the insertion of the Cre-recombinase expression cassette into the region between open reading frames 27 and 29b has no appreciable effect on the replication of HV68 in vivo or on the establishment and maintenance of latency in wild-type C57BL/6 mice.
As an additional control, we infected mice harboring a conditional enhanced green fluorescent protein (EGFP) expression cassette. The mouse line examined, ROSA26-EGFPf/+, has EGFP knocked into the ROSA26 locus, such that the Cre-recombinase-mediated excision of "Stop" sequences results in the ubiquitous expression of EGFP (22a). Notably, we examined infections of this mouse strain with WT HV68 and the Cre-expressing recombinant virus and did not detect any difference in the establishment or maintenance of latency on days 16 and 42 postinfection (data not shown). As such, we can rule out any influence of Cre-mediated recombination of an irrelevant cellular gene on virus infection.
Infection of conditional Dnmt1 knockout mice with HV68-Cre leads to ablation of HV68 latency. To investigate the role of Dnmt1 on the establishment and maintenance of latency of HV68, Dnmt12lox/2lox mice were infected with either HV68-WT or HV68-Cre. Importantly, as expected, deletion of the Dnmt1 allele had no effect on acute virus replication in the lungs following intranasal infection of mice (Fig. 5B). On days 4 and 9 postinfection, there were nearly identical levels of virus in the lungs, as determined by a plaque assay (Fig. 5B). However, the deletion of Dnmt1 severely impaired the establishment of latency in the spleen, as measured on day 16 postinfection (Fig. 7). In the Dnmt1 conditional mice on day 16 postinfection, 1 in 500 spleen cells were latently infected with HV68-WT, while only 1 in 20,000 spleen cells were latently infected in HV68-Cre-infected mice (Fig. 7A). The defect in the establishment of latency was also observed by a decreased frequency of cells reactivating virus ex vivo (Fig. 7B).
Notably, no global defects were seen in the activation of the immune response or in the levels of B-cell proliferation between Dnmt1 conditional mice infected with HV68-WT and those infected with HV68-Cre (data not shown). The absence of a global defect was expected since (i) there was no defect in acute virus replication and (ii) only a very small percentage of spleen cells are latently infected with HV68, and hence only a small percentage of B cells would undergo Cre-recombinase-mediated deletion of the Dnmt1 allele.
By day 42 postinfection, latency in the Dnmt12lox/2lox mice infected with the HV68-Cre virus was almost completely ablated, whereas Dnmt12lox/2lox mice infected with HV68-WT had similar levels of viral genome-positive cells in the spleen to those in wild-type C57BL/6 mice (compare Fig. 6C and 7C). These data strongly indicate that the establishment of HV68 latency requires infection of a B-cell population(s) that subsequently undergoes several rounds of proliferation prior to establishing life-long latency in the memory B-cell compartment.
To extend the above analyses, we also examined the establishment and maintenance of latency in specific cell populations purified from infected Dnmt1 conditional mice (Fig. 8). On day 16 postinfection of the Dnmt1 conditional mice, as previously shown for infections of C57BL/6 mice, there was a high frequency of WT virus in splenic B cells (ca. 1 in 200) and only a very low frequency of WT virus in the splenic non-B-cell fraction (<1 in 10,000) (Fig. 8A). In contrast, the majority of latency observed on day 16 upon infection with the Cre-recombinase-expressing virus was present in the non-B-cell fraction (ca. 1 in 10,000), with only low levels of virus (<1 in 10,000 infected cells) detectable in the B-cell fraction (Fig. 8A). To assess whether latency recovers in Dnmt1 conditional mice infected with the Cre-expressing virus at late times postinfection, we examined the surface IgD+ (na?ve B cells) and IgD– splenic (germinal-center and memory B cells) B-cell populations at 3 months postinfection. As previously shown, with the WT virus nearly all of the detectable latency was present in the surface IgD– B-cell fraction (Fig. 8B). However, infection of the Dnmt1 conditional mice with the Cre-recombinase-expressing virus resulted in nearly undetectable levels of virus in the surface IgD-negative B-cell fraction, with slightly higher levels of virus infection present in the na?ve B-cell fraction (equivalent to the levels of virus infection in the na?ve B-cell population observed with WT virus infection) (Fig. 8B). Thus, these data directly demonstrate an inability of the Cre-recombinase-expressing virus to establish latency in the memory B-cell reservoir following the infection of Dnmt1 conditional mice.
DISCUSSION
The long-term latency reservoir for HV68 is memory B cells, although at early times postinfection latency can be detected in na?ve, germinal-center, and memory B cells (10, 42). We have hypothesized that, like the case for EBV infection in humans, the establishment of HV68 latency in memory B cells is the result of virus infection of na?ve B cells, followed by the differentiation of latently infected na?ve B cells through the highly proliferative germinal-center reaction and ultimately into the memory B-cell reservoir. The observed bias of virus infection toward the proliferating B-cell population during the establishment of latency shown here is consistent with this model. The major alternative model for the establishment of virus latency in memory B cells is based on HV68 directly infecting na?ve, germinal-center, and memory B cells, with the observed differential persistence of virus infection in memory B cells simply reflecting the longevity of this reservoir. With this model, we might also observe a bias of virus infection towards proliferating B cells, but this would not be an obligate step toward establishing latency in the memory B-cell reservoir. However, we cannot rule out the possibility that HV68 infection of memory B cells drives them to proliferate, which would also lead to their attrition upon the loss of Dnmt1 activity. Thus, the results shown here demonstrate that the proliferation of latently infected B cells precedes the establishment of long-term virus latency in the memory B-cell compartment. However, these studies cannot conclusively distinguish between (i) the direct infection of memory B cells followed by virus-driven proliferation and (ii) virus infection of na?ve and/or germinal-center B cells with subsequent differentiation to latently infected memory B cells. In addition, we do not know whether virus infection drives the observed proliferation of latently infected B cells or whether the virus preferentially infects proliferating B cells.
A further consideration for ablating Dnmt1 function comes from previous studies with EBV, which have shown that there is a tight correlation between the methylation of discrete regions of the viral genome and the establishment of long-term latency in the memory B-cell reservoir (33). It has been hypothesized that EBV genome methylation is critical for silencing the expression of several latency genes that are expressed during the growth-transforming latency program (EBNAs 2, 3a, 3b, and 3c and LMPs 1 and 2A), some of which are targets for CD8+ T-cell control of EBV infection (25). Although no link between methylation of the HV68 genome and the establishment of latency has been established to date, it is possible that the deletion of Dnmt1 leads to decreased levels of viral genome methylation in proliferating latently infected B cells. This may lead to HV68-infected cells being more vulnerable to host immune system recognition and elimination as a consequence of altered viral gene expression. Unfortunately, it is currently not possible to distinguish between effects on the methylation state of the viral genome and those on the cellular genome. However, it is important that the loss of latently infected cells by either mechanism due to the deletion of Dnmt1 is dependent on the proliferation of latently infected B cells. Since data in the literature have shown that cells that have a deletion of Dnmt1 survive only two to three rounds of cell division prior to dying and that B cells that undergo isotype switching go through at least three to six rounds of cell division (7, 14, 33), we currently favor the hypothesis that the loss of latently infected B cells is directly due to the loss of cellular genome methylation.
In the case of EBV infection, there is a tight anatomical compartmentalization of latency programs. In peripheral blood, EBV latency is exclusively found in circulating resting memory B cells (3). Similarly, by 3 months postinfection of mice, HV68 latency in the blood is almost exclusive found within the circulating isotype-switched memory B-cell population (39). In contrast, in the tonsils of seropositive individuals that are shedding virus, EBV infection of na?ve, germinal-center, and memory B cells is observed (10). The latter is very reminiscent of what is observed during the establishment of HV68 latency in the spleen (e.g., on day 16). Whether there is ongoing reseeding of B-cell latency at sites of HV68 reactivation and shedding is currently unknown, but this seems likely. Thus, for both EBV and HV68 infections, there appear to be distinct anatomical restrictions with respect to the forms of latency observed.
As stated above, the proposed model for the establishment of HV68 latency closely mirrors the proposed model for the establishment of EBV latency in humans. Models of EBV infection and the establishment of viral latency propose that the virus infects na?ve B cells and then drives these B cells to enter the germinal center reaction (35, 36). EBV has also been proposed to increase the survival rate of germinal center B cells, allowing the infected cells to successfully differentiate into memory B cells (27). This model has been derived from several experiments where peripheral blood mononuclear cells were sorted into different stages of B-cell differentiation and analyzed for the expression of EBV proteins (4). Distinct EBV latency-associated gene expression patterns which are consistent with this widely accepted model have been identified for specific B-cell populations. As with the model for HV68, there are several alternative scenarios for the establishment of EBV latency in the memory B-cell compartment, including the direct infection of memory B cells. Due to the limited host range of EBV, further experiments to track the virus in vivo will be difficult.
There is a large amount of evidence suggesting that EBV is capable of driving the germinal-center reaction in infected B cells. LMP1 and LMP2A are EBV-encoded proteins that are capable of providing prosurvival signals. LMP1 has been shown to activate the NF-B pathway similar to the engagement of CD40, while LMP2A has immunoreceptor tyrosine activation motifs in its cytoplasmic domain, perhaps preventing activation-induced cell death during the germinal-center reaction as well as providing survival signals (6, 16). Several studies have suggested that EBV might stimulate cells to acquire a germinal-center B-cell phenotype but that these infected cells do not actually participate in the germinal center reaction (18).
To date, no homologs of EBV LMP1 or LMP2A have been identified in HV68, and there is currently little evidence to suggest that HV68 drives B cells to undergo the germinal-center reaction. We have observed in the course of studies characterizing a HV68 mutant which cannot express the latency-associated M2 gene product that there is an accumulation of latently infected na?ve B cells in the spleen compared to wild-type virus-infected cells during the establishment of latency. One interpretation of this analysis is that the M2 gene product is involved in the virus-driven differentiation of latently infected na?ve B cells (40). Alternatively, it is possible that HV68 takes advantage of the B-cell proliferation that is triggered upon virus infection by infecting antigen-stimulated B cells during the acute phase of infection. Through this mechanism, the virus may passively end up in memory B cells. A prediction from the latter model is that latently infected memory B cells would be specific for HV68. Another possibility is that even if the virus does not actively drive the germinal-center reaction, it may provide survival signals for the infected cell to escape activation-induced cell death. HV68 does encode several proteins that could play critical roles in infected germinal-center B cells, including a bcl2 homolog (v-bcl2), a D-type cyclin homolog (v-cyclin), and a G-coupled receptor (v-GPCR) (38, 39). In addition, there are still many genes encoded within the viral genome with unknown functions.
In summary, the results presented here demonstrate that the establishment of chronic HV68 infection requires B-cell proliferation prior to the establishment of long-term latency in the memory B-cell compartment. These studies point to a role for gammaherpesviruses utilizing normal B-cell biology during the course of a viral infection. Further analysis is needed to fully understand the role of both virally encoded genes and cellular genes in this process.
ACKNOWLEDGMENTS
Samuel H. Speck is supported by grants CA43143, CA52004, CA58524, and CA095318. Janice M. Moser is supported by a postdoctoral fellowship from the American Cancer Society.
We thank Robert Karaffa and Michael Hulsey for their assistance and expertise with fluorescence-activated cell sorting and Joshy Jacob, Craig Chappell, and the Speck lab for helpful comments on this research.
REFERENCES
Adler, H., M. Messerle, and U. H. Koszinowski. 2001. Virus reconstituted from infectious bacterial artificial chromosome (BAC)-cloned murine gammaherpesvirus 68 acquires wild-type properties in vivo only after excision of BAC vector sequences. J. Virol. 75:5692-5696.
Adler, H., M. Messerle, M. Wagner, and U. H. Koszinowski. 2000. Cloning and mutagenesis of the murine gammaherpesvirus 68 genome as an infectious bacterial artificial chromosome. J. Virol. 74:6964-6974.
Babcock, G. J., L. L. Decker, R. B. Freeman, and D. A. Thorley-Lawson. 1999. Epstein-Barr virus-infected resting memory B cells, not proliferating lymphoblasts, accumulate in the peripheral blood of immunosuppressed patients. J. Exp. Med. 190:567-576.
Babcock, G. J., D. Hochberg, and D. A. Thorley-Lawson. 2000. The expression pattern of Epstein-Barr virus latent genes in vivo is dependent upon the differentiation stage of the infected B cell. Immunity 13:497-506.
Bird, A. P., and A. P. Wolffe. 1999. Methylation-induced repression—belts, braces, and chromatin. Cell 99:451-454.
Caldwell, R. G., J. B. Wilson, S. J. Anderson, and R. Longnecker. 1998. Epstein-Barr virus LMP2A drives B cell development and survival in the absence of normal B cell receptor signals. Immunity 9:405-411.
Cherry, S. R., C. Beard, R. Jaenisch, and D. Baltimore. 2000. V(D)J recombination is not activated by demethylation of the kappa locus. Proc. Natl. Acad. Sci. USA 97:8467-8472.
Clambey, E. T., H. W. Virgin, and S. H. Speck. 2000. Disruption of the murine gammaherpesvirus 68 M1 open reading frame leads to enhanced reactivation from latency. J. Virol. 74:1973-1984.
Fan, G., C. Beard, R. Z. Chen, G. Csankovszki, Y. Sun, M. Siniaia, D. Biniszkiewicz, B. Bates, P. P. Lee, R. Kuhn, A. Trumpp, C. Poon, C. B. Wilson, and R. Jaenisch. 2001. DNA hypomethylation perturbs the function and survival of CNS neurons in postnatal animals. J. Neurosci. 21:788-797.
Flano, E., I. Kim, D. L. Woodland, and M. A. Blackman. 2002. g-Herpesvirus latency is preferentially maintained in splenic germinal center and memory B cells. J. Exp. Med. 196:1363-1372.
Glimm, H., I. H. Oh, and C. J. Eaves. 2000. Human hematopoietic stem cells stimulated to proliferate in vitro lose engraftment potential during their S/G(2)/M transit and do not reenter G(0). Blood 96:4185-4193.
Gothot, A., R. Pyatt, J. McMahel, S. Rice, and E. F. Srour. 1997. Functional heterogeneity of human CD34(+) cells isolated in subcompartments of the G0/G1 phase of the cell cycle. Blood 90:4384-4393.
Holyoake, T., X. Jiang, C. Eaves, and A. Eaves. 1999. Isolation of a highly quiescent subpopulation of primitive leukemic cells in chronic myeloid leukemia. Blood 94:2056-2064.
Jackson-Grusby, L., C. Beard, R. Possemato, M. Tudor, D. Fambrough, G. Csankovszki, J. Dausman, P. Lee, C. Wilson, E. Lander, and R. Jaenisch. 2001. Loss of genomic methylation causes p53-dependent apoptosis and epigenetic deregulation. Nat. Genet. 27:31-39.
Joseph, A. M., G. J. Babcock, and D. A. Thorley-Lawson. 2000. Cells expressing the Epstein-Barr virus growth program are present in and restricted to the naive B-cell subset of healthy tonsils. J. Virol. 74:9964-9971.
Kilger, E., A. Kieser, M. Baumann, and W. Hammerschmidt. 1998. Epstein-Barr virus-mediated B-cell proliferation is dependent upon latent membrane protein 1, which simulates an activated CD40 receptor. EMBO J. 17:1700-1709.
Kuppers, R. 2003. B cells under influence: transformation of B cells by Epstein-Barr virus. Nat. Rev. Immunol. 3:801-812.
Kurth, J., M. L. Hansmann, K. Rajewsky, and R. Kuppers. 2003. Epstein-Barr virus-infected B cells expanding in germinal centers of infectious mononucleosis patients do not participate in the germinal center reaction. Proc. Natl. Acad. Sci. USA 100:4730-4735.
Kurth, J., T. Spieker, J. Wustrow, G. J. Strickler, L. M. Hansmann, K. Rajewsky, and R. Kuppers. 2000. EBV-infected B cells in infectious mononucleosis: viral strategies for spreading in the B cell compartment and establishing latency. Immunity 13:485-495.
Ladd, A. C., R. Pyatt, A. Gothot, S. Rice, J. McMahel, C. M. Traycoff, and E. F. Srour. 1997. Orderly process of sequential cytokine stimulation is required for activation and maximal proliferation of primitive human bone marrow CD34+ hematopoietic progenitor cells residing in G0. Blood 90:658-668.
Lee, P. P., D. R. Fitzpatrick, C. Beard, H. K. Jessup, S. Lehar, K. W. Makar, M. Perez-Melgosa, M. T. Sweetser, M. S. Schlissel, S. Nguyen, S. R. Cherry, J. H. Tsai, S. M. Tucker, W. M. Weaver, A. Kelso, R. Jaenisch, and C. B. Wilson. 2001. A critical role for Dnmt1 and DNA methylation in T cell development, function, and survival. Immunity 15:763-774.
Mackett, M., J. P. Stewart, S. de V Pepper, M. Chee, S. Efstathiou, A. A. Nash, and J. R. Arrand. 1997. Genetic content and preliminary transcriptional analysis of a representative region of murine gammaherpesvirus 68. J. Gen. Virol. 78:1425-1433.
Mao, X., Y. Fujiwara, A. Chapdelaine, H. Yang, and S. H. Orkin. 2001. Activation of EGFP expression by Cre-mediated excision in a new ROSA26 reporter mouse strain. Blood 97:324-326.
Moorman, N. J., D. O. Willer, and S. H. Speck. 2003. The gammaherpesvirus 68 latency-associated nuclear antigen homolog is critical for the establishment of splenic latency. J. Virol. 77:10295-10303.
Moser, J. M., J. W. Upton, K. S. Gray, and S. H. Speck. 2005. Ex vivo stimulation of B cells infected with gammaherpesvirus 68 triggers reactivation from latency. J. Virol. 79:5227-5231.
Paulson, E. J., and S. H. Speck. 1999. Differential methylation of Epstein-Barr virus latency promoters facilitates viral persistence in healthy seropositive individuals. J. Virol. 73:9959-9968.
Pollock, J. L., and H. W. Virgin. 1995. Latency, without persistence, of murine cytomegalovirus in the spleen and kidney. J. Virol. 69:1762-1768.
Schuster, V. H., and M. Muschen. 2003. Epstein-Barr virus and the B cell: a secret romance. Trends Microbiol. 11:243-245.
Shapiro, H. M., and N. G. Perlmutter. 2001. Violet laser diodes as light sources for cytometry. Cytometry 44:133-136.
Sherr, C. J. 2000. The Pezcoller lecture: cancer cell cycles revisited. Cancer Res. 60:3689-3695.
Smith, G. A., and L. W. Enquist. 2000. A self-recombining bacterial artificial chromosome and its application for analysis of herpesvirus pathogenesis. Proc. Natl. Acad. Sci. USA 97:4873-4878.
Speck, S. H., and H. W. Virgin. 1999. Host and viral genetics of chronic infection: a mouse model of gamma-herpesvirus pathogenesis. Curr. Opin. Microbiol. 2:403-409.
Sunil-Chandra, N. P., S. Efstathiou, J. Arno, and A. A. Nash. 1992. Virological and pathological features of mice infected with murine gammaherpesvirus 68. J. Gen. Virol. 73:2347-2356.
Tangye, S. G., A. Ferguson, D. T. Avery, C. S. Ma, and P. D. Hodgkin. 2002. Isotype switching by human B cells is division-associated and regulated by cytokines. J. Immunol. 169:4298-4306.
Thompson, M. P., and R. Kurzrock. 2004. Epstein-Barr virus and cancer. Clin. Cancer Res. 10:803-821.
Thorley-Lawson, D. A. 2001. Epstein-Barr virus: exploiting the immune system. Nat. Rev. Immunol. 1:75-82.
Thorley-Lawson, D. A., and A. Gross. 2004. Persistence of the Epstein-Barr virus and the origins of associated lymphomas. N. Engl. J. Med. 350:1328-1337.
Tibbetts, S. A., J. Loh, V. Van Berkel, J. S. McClellan, M. A. Jacoby, S. B. Kapadia, S. H. Speck, and H. W. T. Virgin. 2003. Establishment and maintenance of gammaherpesvirus latency are independent of infective dose and route of infection. J. Virol. 77:7696-7701.
Virgin, H. W., and S. H. Speck. 1999. Unraveling immunity to gamma-herpesviruses: a new model for understanding the role of immunity in chronic virus infection. Curr. Opin. Immunol. 11:371-379.
Virgin, H. W., P. Latreille, P. Wamsley, K. Hallsworth, K. E. Weck, A. J. Dal Canto, and S. H. Speck. 1997. Complete sequence and genomic analysis of murine gammaherpesvirus 68. J. Virol. 71:5894-5904.
Weck, K. E., M. L. Barkon, L. I. Yoo, S. H. Speck, and H. I. Virgin. 1996. Mature B cells are required for acute splenic infection, but not for establishment of latency, by murine gammaherpesvirus 68. J. Virol. 70:6775-6780.
Weck, K. E., S. S. Kim, H. I. Virgin, and S. H. Speck. 1999. B cells regulate murine gammaherpesvirus 68 latency. J. Virol. 73:4651-4661.
Willer, D. O., and S. H. Speck. 2003. Long-term latent murine gammaherpesvirus 68 infection is preferentially found within the surface immunoglobulin D-negative subset of splenic B cells in vivo. J. Virol. 77:8310-8321.
Yu, D., G. A. Smith, L. W. Enquist, and T. Shenk. 2002. Construction of a self-excisable bacterial artificial chromosome containing the human cytomegalovirus genome and mutagenesis of the diploid TRL/IRL13 gene. J. Virol. 76:2316-2328.(Janice M. Moser, Jason W.)
Graduate Program in Molecular Cell Biology, Washington University School of Medicine, St. Louis, Missouri 63110
Departments of Immunology and Pediatrics, University of Washington, Seattle, Washington 98195
ABSTRACT
Murine gammaherpesvirus 68 (HV68) provides a tractable small animal model with which to study the mechanisms involved in the establishment and maintenance of latency by gammaherpesviruses. Similar to the human gammaherpesvirus Epstein-Barr virus (EBV), HV68 establishes and maintains latency in the memory B-cell compartment following intranasal infection. Here we have sought to determine whether, like EBV infection, HV68 infection in vivo is associated with B-cell proliferation during the establishment of chronic infection. We show that HV68 infection leads to significant splenic B-cell proliferation as late as day 42 postinfection. Notably, HV68 latency was found predominantly in the proliferating B-cell population in the spleen on both days 16 and 42 postinfection. Furthermore, virus reactivation upon ex vivo culture was heavily biased toward the proliferating B-cell population. DNA methyltransferase 1 (Dnmt1) is a critical maintenance methyltransferase which, during DNA replication, maintains the DNA methylation patterns of the cellular genome, a process that is essential for the survival of proliferating cells. To assess whether the establishment of HV68 latency requires B-cell proliferation, we characterized infections of conditional Dnmt1 knockout mice by utilizing a recombinant HV68 that expresses Cre-recombinase (HV68-Cre). In C57BL/6 mice, the HV68-Cre virus exhibited normal acute virus replication in the lungs as well as normal establishment and reactivation from latency. Furthermore, the HV68-Cre virus also replicated normally during the acute phase of infection in the lungs of Dnmt1 conditional mice. However, deletion of the Dnmt1 alleles from HV68-infected cells in vivo led to a severe ablation of viral latency, as assessed on both days 16 and 42 postinfection. Thus, the studies provide direct evidence that the proliferation of latently infected B cells is critical for the establishment of chronic HV68 infection.
INTRODUCTION
Epstein-Barr virus (EBV) is a human lymphotropic virus that establishes a life-long infection, predominantly in memory B cells (17, 35, 36). EBV has been associated with numerous human malignancies, including Burkitt's lymphoma, Hodgkin's lymphoma, and posttransplantation lymphoproliferative disease, and is capable of transforming B cells in vitro (34). Since EBV has a very limited host range, the study of this virus in vivo is difficult. Murine gammaherpesvirus 68 (HV68) is a closely related gammaherpesvirus that infects small rodents and has recently been used to characterize the pathogenesis of gammaherpesviruses in vivo (31). Similar to EBV, HV68 establishes latency in B cells and persists for the lifetime of the host in the memory B-cell compartment (10, 42).
In the peripheral blood, EBV latency is restricted to resting memory B cells, while at sites of active virus replication, such as the tonsils, EBV infection of na?ve, germinal-center, and memory B cells can be detected (3, 4). The current model for the establishment of EBV latency suggests that virus infection initiates through the infection of any resting B cell, including the na?ve B-cell population (reviewed in references 17, 35, and 36). As a consequence of the expression of a subset of viral genes, referred to as the growth program, EBV is able to activate infected B cells and drive these cells to proliferate (15). These proliferating lymphoblasts are then thought to migrate and form germinal centers. Germinal centers are the sites of B-cell differentiation during the immune response that lead to the development of antibody-producing plasma cells and memory B cells. During the germinal-center reaction, the EBV latency-associated membrane antigens LMP1 and LMP2A are thought to mimic signals normally provided through CD40 and the B-cell receptor, respectively, helping to drive differentiation to memory B cells and promote cell survival.
This model is supported by data that illustrate that there are distinct expression patterns of EBV latency genes found in each differentiation stage of the infected B cell (4). EBV was observed to be present in all B-cell subsets in the tonsils of individuals who had ongoing viral replication. However, in the rare tonsils that had no evidence of ongoing viral replication, no infected na?ve B cells or germinal center B cells were detected, and all the virus-infected cells were in the memory compartment (4). These data further support the hypothesis that the longevity of the memory B-cell compartment provides a stable reservoir for EBV latency while allowing for reactivation and reseeding latency to the memory B-cell reservoir through the infection of na?ve B cells present at the site of virus shedding.
Although there are several studies that support this model, there are several points that have yet to be proven experimentally. Some of the ideas discussed above have been developed from parallels between viral gene expression and normal B-cell biology. The role of the immune response and the presence of antigen have yet to be fully deciphered during the germinal center reaction. Alternative models have also been proposed for EBV infection and establishment of latency. Immunohistochemical and in situ hybridization studies have shown that EBV-infected B cells are rarely seen in germinal centers in the tonsils, but instead are found in the extrafollicular regions (18). These data possibly suggest that EBV-infected B cells acquire the phenotype of germinal-center B cell but may not actually participate in the germinal-center reaction. An alternative model is that EBV does not participate in the germinal center reaction, but rather directly infects memory B cells. EBV-positive and -negative germinal-center B cells isolated from sections of tonsils taken from infectious mononucleosis patients by micromanipulation were analyzed for somatic hypermutation of their V genes (19). EBV-positive cells located in the germinal centers proliferated without ongoing somatic hypermutation compared to the EBV-negative germinal-center B cells from the same sections. The authors of that study concluded that EBV-infected cells do not participate in the germinal-center reaction during infectious mononucleosis.
The difficulty with experimentally proving the proposed models is directly related to the limitations in manipulating EBV infection in vivo. In an attempt to begin addressing the questions raised above, HV68 infection in mice is being utilized to track gammaherpesvirus infection in vivo. Previously, it was shown that at early stages of HV68 latency following intranasal infection (day 16 postinfection), virus infection of na?ve, germinal-center, and memory B cells is observed (10, 42). However, by later latency times (3 and 6 months postinfection), viral genomes are only detectable in germinal-center and memory B-cell populations (42). Again, this leaves unresolved the question of whether there is direct infection of na?ve, germinal-center, and memory B cells, with the memory B-cell population being the long-lived population. To address this question, we investigated whether B-cell proliferation is a critical intermediate step in the establishment of latency in the memory B-cell reservoir. Here we show that HV68 is predominantly found in the proliferating B-cell population on days 16 and 42 postinfection. These data correlate with previous data from our lab showing a greater frequency of virus present in the germinal-center B-cell population at these times postinfection (42).
To determine whether B-cell proliferation is a critical intermediate step in the establishment of latency in the memory B-cell compartment, we utilized a conditional knockout mouse to delete the gene encoding DNA methyltransferase 1 (Dnmt1), an essential enzyme needed during cellular proliferation (9, 14, 21). Dnmt1 is the maintenance methyltransferase, which preferentially methylates hemimethylated DNA (5). The Cre-recombinase-mediated deletion of Dnmt1 causes the demethylation and uniform cell death of several different cell types, including cultured fibroblasts and B-cell progenitors, in vitro (7, 14). In vivo, the conditional deletion of Dnmt1 in postmitotic neurons did not affect the levels of global methylation or cell survival during postnatal life, but a Dnmt1 deletion in proliferating mitotic central nervous system precursors led to DNA hypomethylation and functionally impaired cells that were selected against at postnatal stages (9). Finally, the deletion of Dnmt1 impaired the proliferation and survival of T-lineage cells (21). Here we show that the deletion of Dnmt1 in HV68-infected cells severely impairs the establishment of latency at day 16 postinfection and that by day 42 postinfection, latency in the spleen is almost completely ablated. These studies argue persuasively that HV68-infected B cells undergo several rounds of proliferation prior to establishing latency in the resting memory B-cell compartment.
MATERIALS AND METHODS
Mice, viruses, and cells. C57BL/6J mice were purchased from Jackson Laboratory (Bar Harbor, Maine). Dnmt2lox/2lox mice were constructed as previously described (21). Briefly, these mice contain loxP sites flanking exons 4 and 5 of the Dnmt1 gene. Cre-recombinase-mediated deletion leads to an out-of-frame splicing from exons 3 to 6, resulting in a null Dnmt1 allele (9, 21). All protocols for animal studies were approved by the Institutional Animal Care and Use Committee of Emory University. Mice were used at 8 to 12 weeks of age. HV68 strain WUMS (ATCC VR-1465) was used for all wild-type virus infections. Virus passaging, maintenance, and titration were all performed as previously described (8). Mice were anesthetized with isofluorane and inoculated intranasally with 1,000 PFU of virus in 20 μl of Dulbecco modified Eagle medium (DMEM). NIH 3T12 cells and mouse embryonic fibroblast (MEF) cells were maintained at 37°C and 5% CO2 in DMEM supplemented with 10% fetal calf serum (FCS), 100 U of penicillin per ml, 100 mg of streptomycin per ml, and 2 mM L-glutamine (cMEM). MEF cells from C57BL/6 embryos were obtained as previously described (26).
Construction of HV68-Cre virus. To generate the targeting construct for the introduction of Cre-recombinase into HV68, the genomic region between nucleotide positions 45237 and 48347 was excised as a HindIII-SacI fragment from a pUC13 plasmid vector carrying the entire BamHI C HV68 genomic fragment (22). The fragment was then reinserted between the HindIII and SacI sites of pUC13 to create the plasmid JE109. JE109 was then digested with HindIII, and a HindIII/EcoRI adaptor (Ezclonesystems, New Orleans, LA) was inserted to generate plasmid JE110. A kanamycin resistance cassette (Kanr) flanked by FLP recombinase target sites was excised as a KpnI-HindIII fragment from the plasmid pCP15 and inserted into KpnI-HindIII-digested pMECA to generate the plasmid pMECAKan. The Cre-recombinase expression cassette from pCREin was excised with PacI and inserted into the PacI site of pMECAKan to generate pMECAKanCre (30). The Cre-recombinase expression cassette from pCREin includes the human cytomegalovirus (HCMV) immediate-early promoter upstream of Cre. In addition, the Cre gene contains an engineered intron to prevent the expression of this protein in bacterial cells (30). The Cre expression cassette and the kanamycin resistance cassette were excised from pMECAKanCre as an EcoRI-NsiI fragment. Following treatment with Klenow, the fragment was inserted into the PmlI site of JE110 (corresponding to HV68 genomic position 46347) to generate JE111. To isolate the targeting construct for use in recombination, JE111 was digested with EcoRI. The insert DNA was gel purified twice and subsequently electroporated into Escherichia coli strain DH10B containing the HV68 bacterial artificial chromosome (BAC) and a temperature-sensitive RecA expression plasmid, p2650. The p2650 plasmid contains a temperature-sensitive origin of replication which is unable to function at 42°C, leading to the loss of p2650 function in cells grown at 42°C. Following transformation, the cells were allowed to recover at 37°C for 2 hours and then grown on chloramphenicol (Cam) and kanamycin (Kan) selection media, with incubation at 42°C overnight followed by an additional 12 h at 37°C. HV68-BAC provides Camr in this selection process, and therefore only genomes which had recombined the Cre-Kanr cassette into the locus between ORF27 and ORF29b were selected. Camr Kanr colonies were replica plated onto carbenicillin selection medium to confirm the loss of p2650. Colonies were isolated, and the insertion of the Cre-Kanr cassette was confirmed by Southern blotting. To remove the Kanr cassette, electrocompetent cells were prepared from one such confirmed Camr Kanr clone. The competent cells were transformed with a plasmid expressing the FLP recombinase, i.e., pCP20, which has a temperature-sensitive origin of rep-lication that will not function at 42°C. Transformants were grown in medium containing Cam, Kan, and carbenicillin at 30°C overnight. Colonies were streaked onto Cam plates and grown at 42°C overnight to remove pCP20. Colonies were replica plated onto either chloramphenicol or kanamycin selection medium and grown overnight at 37°C. DNAs from Camr Kans transformants were analyzed by Southern blotting to confirm the loss of the Kanr cassette. HV68-Cre BAC DNA was transfected into NIH 3T12 fibroblast cells, and the virus was isolated as described previously (23). Self-excision of the BAC sequence was confirmed by Southern blotting and flow cytometry. The HV68 genome cloned as a BAC was a kind gift of Ulrich Kozinowski (1, 2). The pCREin plasmid was a generous gift from Greg Smith (30, 43).
Limiting-dilution PCR analysis. The frequency of cells harboring the HV68 genome was determined using a single-copy-sensitivity nested PCR assay as previously described (41). Briefly, cells were resuspended in isotonic buffer and subjected to a series of six threefold serial dilutions. Starting with 104 cells per well, cells were plated in a background of 104 uninfected NIH 3T12 cells in 96-well PCR plates. Twelve replicate PCRs were performed for each cell dilution per experiment. The cells were subjected to lysis by proteinase K at 56°C for 6 h, followed by enzyme inactivation for 95°C for 20 min. Round 1 and round 2 PCRs were performed as previously described on a PrimusHT thermal cycler (MWG Biotech, Highpoint, NC). Reaction products were separated in 2% agarose gels. All assays demonstrated near-single-copy sensitivities, with no false positives. All data analysis was performed with GraphPad Prism software (San Diego, CA). Data were subjected to nonlinear regression analysis with a sigmoidal dose-response algorithm for best fit. Frequencies of viral genome-positive cells were obtained from the nonlinear regression fit of the data where the regression line intersected 63.2%. Based on the Poisson distribution, this is the frequency at which there is at least one event present in a population.
Limiting-dilution ex vivo reactivation analyses. Limiting-dilution analysis to detect reactivation from latency was performed as previously described, with the following modifications and additions (40). Briefly, 24 hours prior to the assay, MEF cells were plated in 96-well or 24-well plates at 5 x 103 cells per well or 5 x 104 cells per well, respectively. Spleen cells obtained from infected mice on days 16, 42, 90, and 180 postinfection were resuspended in cMEM and plated in twofold serial dilutions starting with either 105 or 106 cells per well onto a MEF indicator monolayer. Twenty-four replicate wells were plated per dilution. The wells were scored microscopically for cytopathic effect (CPE) on the MEF monolayer at 21 days postplating. To detect preformed infectious virus, cells were hypotonically lysed while simultaneously being mechanically disrupted using a Mini-Beadbeater 8 and mechanical disruption beads (Biospec Products, Bartlesville, OK). Latent viruses cannot reactivate from dead cells, and this method kills >99% of cells. No measurable preformed infectious virus was detected at the specified time points postinfection.
Proliferation sorting. Single-cell suspensions of spleen cells were prepared from infected mice, and red blood cells were lysed. Cells were labeled and stained as previously described (13). Briefly, cells were washed once in phenol-free Hanks' balanced salt solution (HBSS) supplemented with 2% FCS (HBSS/2) and then resuspended with HBSS/2 containing 10 μmol Hoechst dye (Hst) (Molecular Probes, Eugene, OR). Hoechst 33342 was used for sorting with a UV laser (13), and Hoechst 34580 was used for sorting with a violet laser (28). The spleen cells were incubated for 45 min at 37°C. Pyronin Y (Sigma Chemicals, St. Louis, MO) was added to give a final concentration of 2.5 μg/ml, and the cells were incubated for an additional 45 min at 37°C. The cells were then stained with anti-mouse CD19 conjugated to allophycocyanin (APC) (BD Pharmingen, San Diego, CA) for a final 20 min on ice in the dark. The cells were then washed once in HBSS/2 with Hst and once with just HBSS/2. Cells were sorted on either a MoFlo (Cytomation, Fort Collins, CO) or FACSAria (BD Biosciences Immunocytometry Systems, San Jose, CA) flow cytometer. The purity of sorted cells always exceeded 90%.
BrdU staining. Mice were given bromodeoxyuridine (BrdU) in their drinking water for 8 days at a concentration of 0.8 mg/ml. Fresh BrdU water was given to the mice daily. Spleen cells were harvested from treated mice and stained for intracellular BrdU using a BrdU flow kit (BD Pharmingen) and following the manufacturer's recommended protocol. In brief, spleen cells were surface stained with anti-CD19-APC and anti-immunoglobulin D (anti-IgD)-fluorescein isothiocyanate (FITC) (BD Pharmingen) for 15 min on ice. Cells were washed, fixed, permeabilized in BD Cytofix/Cytoperm buffer for 15 min on ice, washed with BD Perm/Wash buffer, and incubated with BD Cytofix/Cytoperm Plus buffer for 10 min on ice. Spleen cells were washed with BD Perm/Wash buffer and refixed with BD Cytofix/Cytoperm buffer on ice for 5 min. Spleen cells were washed with BD Perm/Wash buffer and treated with DNase at a concentration of 300 μg/ml for 1 h at 37°C. Spleen cells were washed with BD Perm/Wash buffer and stained with anti-BrdU-FITC or anti-BrdU-APC diluted in BD Perm/Wash buffer, with incubation for 20 min at room temperature. Spleen cells were washed with BD Perm/Wash buffer and resuspended in phosphate-buffered saline with 2% FCS for analysis on a FACSCalibur instrument (BD Biosciences).
Plaque assays. Plaque assays were performed using NIH 3T12 monolayers under noble agar overlays as previously described (40), with the following changes. NIH 3T12 cells were plated in six-well plates at 2 x 105 cells per well the day prior to infection. Lungs were thawed and homogenized by mechanical disruption with beads (Immsilica beads; Biospec Products) in a Mini-Beadbeater 8 (Biospec Products, Bartlesville, OK), and then 10-fold serial dilutions were made in complete DMEM and plated on NIH 3T12 monolayers. Infections were performed in a 200-μl volume, and plates were rocked every 15 min for 1 hour at 37°C. Samples were overlaid with 3 ml of a 1:1 mixture of 3% noble agar and 2x minimum essential medium (MEM) supplemented with 20% FCS, 2x penicillin-streptomycin, and 2x L-glutamine. An additional 2 ml of noble agar with supplemented 2x MEM was added on day 3 postplating. Monolayers were stained on day 7 by the addition of 2 ml of neutral red overlay (0.01% neutral red in serum-free DMEM). After 18 to 24 h, plaques were counted. The limit of detection for this assay is 10 PFU per organ.
Immunoblots. NIH 3T12 fibroblast cells were infected with either HV68-WT or HV68-Cre at a multiplicity of infection (MOI) of 5 for 1 h at 37°C. The cells were then harvested at appropriate times and lysed in ELB buffer (50 mM HEPES, pH 7.2, 250 mM NaCl, 2 mM EDTA, 0.1% NP-40 substitute, and 0.01% protease inhibitor cocktail [Sigma-Aldrich]) for 20 min on ice. Sample buffer (2x) was added, and samples were boiled for 10 min before being subjected to standard sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes. Membranes were blotted with an anti-Cre antibody (1:10,000 dilution) (Novagen, San Diego, CA), a donkey anti-rabbit secondary antibody conjugated to Alexa Fluor 594 (Molecular Probes), and anti-?-actin directly conjugated to FITC (Sigma-Aldrich) and then visualized on a Typhoon 9400 instrument (Amersham Biosciences, Piscataway, NJ).
RESULTS
HV68 infection drives B-cell proliferation in vivo. HV68 undergoes a lytic infection in the lungs of mice after intranasal infection, which is followed by the establishment of a latent infection that persists for the lifetime of the host. Acute virus replication is cleared by days 9 to 15 postinfection, and the early and intermediate stages of latency are typically analyzed on days 16 and 42 postinfection, respectively (32, 37, 40, 42). On day 16 postinfection, the predominant latency reservoirs for HV68 are na?ve, germinal-center, and memory B cells, but by day 42 postinfection, HV68 genomes are found predominantly in the more mature surface IgD– B-cell population (10, 42). By 3 and 6 months postinfection, surface IgD– B cells (the majority of which have undergone heavy-chain isotype switching) are the only population harboring detectable levels of HV68 genomes (10, 42). It is currently unclear how HV68 accesses the memory B-cell reservoir. This could occur either through direct infection of existing memory B cells or by infecting na?ve B cells which subsequently go through the germinal-center reaction to become memory B cells. To further characterize the initial events in the establishment of B-cell latency, we assessed the levels of B-cell proliferation at early and late times during the establishment of latency.
To determine the level of B-cell proliferation on days 16 and 42 postinfection, bromodeoxyuridine (BrdU) was administered in the drinking water to mice for 8 days prior to the harvest of spleen cells. BrdU is a thymidine analog that is incorporated into the DNA of cells undergoing proliferation. On day 16 postinfection, approximately 35% of CD19+ B cells were BrdU positive, while on day 42 postinfection approximately 26% of CD19+ B cells were BrdU positive (Fig. 1A and Table 1), compared to approximately 10 to 15% BrdU-labeled CD19+ cells in na?ve mice (Table 1). Interestingly, ca. 60% of the IgD– B cells had proliferated within the 8 days before the harvest at day 42 postinfection, in contrast to only 37% of IgD– B cells harvested from na?ve mice (Fig. 1B). In addition, an increase in na?ve B-cell (surface IgD+) proliferation was also observed in mice at 42 days postinfection (16.3%) compared to na?ve mice (10.6%).
Proliferating B cells are the predominant latently infected cell population in the spleen at early and intermediate times postinfection. Since we observed a large number of proliferating B cells in the spleens of HV68-infected mice on days 16 and 42 postinfection, we wanted to determine whether HV68 infection was biased toward the proliferating B-cell population. Because the methods used to identify proliferating cells using BrdU labeling are incompatible with the analysis of HV68 latency, we employed an alternative approach. To fractionate live B cells into proliferating and nonproliferating populations, spleen cells were labeled with a Hoechst dye and pyronin Y in conjunction with staining of the cells with anti-CD19, followed by flow cytometry (11-13, 20). The Hoechst dye labels DNA content, allowing cells with greater than 2N DNA to be distinguished from cells containing 2N DNA, while pyronin Y stains RNA, discriminating between resting cells (G0 phase of the cell cycle) and cells in cycle (G1, S, G2, and M phases of the cell cycle) based on their RNA contents (11-13, 20). Based on this staining, approximately 30% of CD19+ cells were proliferating at day 16 postinfection (Fig. 2A), which was consistent with the percentage of proliferating splenic B cells determined by BrdU staining (Fig. 1A). To assess the accuracy of cell sorting, we determine the expression of D-type cyclins in each population. Cyclins D1, D2, and D3 are not appreciably expressed in quiescent cells (G0 phase of cell cycle) but are upregulated upon entry into the G1 phase of the cell cycle (29). An immunoblot for cyclins D1, D2, and D3 demonstrated an absence of detectable levels of D-type cyclins in the nonproliferating cell fraction and the presence of D-type cyclins in the proliferating population (data not shown).
To compare the levels of HV68 latency between the proliferating and nonproliferating B-cell populations, two standard measures of latency were used to determine the frequency of cells harboring latent virus ex vivo, namely, a limiting-dilution reactivation assay and a limiting-dilution nested PCR analysis (LDPCR) (8). The reactivation assay consists of limiting-dilution plating of spleen cells on a monolayer of mouse embryonic fibroblast cells (MEFs), which are permissive for virus growth, and monitoring the wells for cytopathic effect (CPE) on the MEF monolayer. Also plated in parallel are samples of cells that have been hypotonically lysed and mechanically disrupted in order to distinguish between preformed infectious viruses and viruses reactivating from latency. In the analyses presented here, there were never any significant levels of preformed infectious virus detected at any of the time points postinfection (data not shown). The LDPCR assay is performed following serial dilutions of virally infected spleen cells starting at 104 cells per well. The cells are lysed with proteinase K, followed by two rounds of nested PCR to detect the presence of the HV68 genome. For both assays, the data are subjected to nonlinear regression analysis with a sigmoidal dose-response algorithm for best fit.
On day 16 postinfection, there was a fivefold higher frequency of HV68 latently infected cells, as assessed by the LDPCR assay, in the proliferating B-cell population than in the nonproliferating B-cell population (Fig. 2B). Furthermore, when these populations were assayed for spontaneous reactivation ex vivo, there was a ca. 50-fold higher frequency of cells reactivating virus from the proliferating B-cell population than from the nonproliferating B-cell population (Fig. 2C), indicating that the virus reactivates more efficiently from latently infected proliferating B cells than from latently infected resting B cells. These data correlate with data from our laboratory demonstrating an increase in reactivation of HV68 from spleen cells ex vivo following stimulation with anti-Ig and anti-CD40 antibodies (24).
By day 42 postinfection, there was an even greater bias of HV68 latency toward the proliferating B-cell population, with a ca. 100-fold higher frequency of HV68 genome-positive cells present in the proliferating B-cell population than in the nonproliferating B-cell population (Fig. 2D and Table 1). These data correlated well with previous data from our lab which demonstrated that HV68 genomes are found predominantly in germinal-center (PNAhi) B cells on both days 16 and 42 postinfection (42). By 3 months postinfection, the percentage of splenic B cells that were proliferating had dropped to nearly background levels (Table 1) and the bias of viral latency toward proliferating B cells was also reduced compared to that observed on day 42 (Fig. 2E and Table 1).
Insertion of Cre-recombinase into HV68 between open reading frames 27 and 29b does not affect acute virus replication or latency in normal C57BL/6 mice. To investigate the role of B-cell proliferation in the establishment of HV68 latency, we utilized a mouse conditional for an enzyme that is required for the survival of proliferating, but not quiescent, cells. The conditional mouse we chose to employ targets the cellular enzyme DNA methyltransferase 1 (Dnmt1). The deletion of Dnmt1 in proliferating cells leads to hypomethylation of the cellular genome, ultimately leading to decreased cell replication and finally cell death (14). In conditional Dnmt1 knockout mice (Dnmt12lox/2lox), exons 4 and 5 of the Dnmt1 gene are flanked by loxP sites (21), such that the expression of Cre-recombinase leads to the deletion of these exons, resulting in a null Dnmt1 allele. To utilize these mice for our infection studies, we constructed a recombinant HV68 virus that expresses Cre-recombinase, as described in Materials and Methods. The Cre-recombinase expression cassette, driven by the HCMV immediate-early promoter, was targeted to the region of the viral genome between ORF27 and ORF29b of HV68 in E. coli (Fig. 3) (1, 30). The resulting recombinant virus, HV68-Cre, was shown to express Cre-recombinase by immunoblotting following the infection of NIH 3T12 murine fibroblasts (Fig. 4A).
A common strategy employed for the generation of herpesvirus BACs has been the insertion of loxP sites flanking the BAC sequences to allow removal of the BAC during virus production. Thus, recombinant HCMV and pseudorabies virus BACs have been generated that express Cre-recombinase, creating viruses capable of self-excising their BAC sequences (30, 43). Since the BAC vector in HV68-BAC, which includes a green fluorescent protein (GFP) expression cassette, is also flanked by loxP sites, we could assess whether the inserted Cre-recombinase expression cassette was functional in the context of virus infection. To do this, we monitored the loss of GFP expression as an indicator of excision of the BAC sequences following infection of NIH 3T12 fibroblasts. At a high multiplicity of infection (MOI = 1), there was a complete absence of GFP expression in the HV68-Cre-infected culture, in contrast to the case for the wt-BAC virus, as assessed by flow cytometry (Fig. 4B). The removal of the BAC vector from the recombinant virus was confirmed by Southern blot analysis (data not shown). Thus, the inserted Cre-recombinase expression cassette is functional in the context of the viral genome.
To determine whether the insertion of Cre-recombinase into the locus between ORF27 and ORF29b affects the pathogenesis of HV68 in vivo, C57BL/6 mice were infected intranasally with 1,000 PFU of either HV68-Cre or wild-type HV68 (HV68-WT). On days 4 and 9 postinfection, lungs were analyzed for levels of acute lytic replication by a plaque assay. HV68-Cre and the wild-type virus exhibited very similar levels of acute viral replication load at both time points assayed postinfection (Fig. 5A). To determine whether the insertion of Cre-recombinase into HV68 affects the establishment and maintenance of latency, the frequencies of latently infected spleen cells from mice inoculated with either HV68-Cre or HV68-WT were compared. On day 16 postinfection, there was no significant difference in the frequencies of latently infected spleen cells from mice infected with either HV68-Cre or HV68-WT, as determined by the limiting-dilution PCR assay or the limiting-dilution reactivation assay (Fig. 6A and B). Furthermore, by day 42 postinfection, there were nearly identical frequencies of HV68-Cre and HV68-WT latently infected cells (Fig. 6C). Thus, we concluded that the insertion of the Cre-recombinase expression cassette into the region between open reading frames 27 and 29b has no appreciable effect on the replication of HV68 in vivo or on the establishment and maintenance of latency in wild-type C57BL/6 mice.
As an additional control, we infected mice harboring a conditional enhanced green fluorescent protein (EGFP) expression cassette. The mouse line examined, ROSA26-EGFPf/+, has EGFP knocked into the ROSA26 locus, such that the Cre-recombinase-mediated excision of "Stop" sequences results in the ubiquitous expression of EGFP (22a). Notably, we examined infections of this mouse strain with WT HV68 and the Cre-expressing recombinant virus and did not detect any difference in the establishment or maintenance of latency on days 16 and 42 postinfection (data not shown). As such, we can rule out any influence of Cre-mediated recombination of an irrelevant cellular gene on virus infection.
Infection of conditional Dnmt1 knockout mice with HV68-Cre leads to ablation of HV68 latency. To investigate the role of Dnmt1 on the establishment and maintenance of latency of HV68, Dnmt12lox/2lox mice were infected with either HV68-WT or HV68-Cre. Importantly, as expected, deletion of the Dnmt1 allele had no effect on acute virus replication in the lungs following intranasal infection of mice (Fig. 5B). On days 4 and 9 postinfection, there were nearly identical levels of virus in the lungs, as determined by a plaque assay (Fig. 5B). However, the deletion of Dnmt1 severely impaired the establishment of latency in the spleen, as measured on day 16 postinfection (Fig. 7). In the Dnmt1 conditional mice on day 16 postinfection, 1 in 500 spleen cells were latently infected with HV68-WT, while only 1 in 20,000 spleen cells were latently infected in HV68-Cre-infected mice (Fig. 7A). The defect in the establishment of latency was also observed by a decreased frequency of cells reactivating virus ex vivo (Fig. 7B).
Notably, no global defects were seen in the activation of the immune response or in the levels of B-cell proliferation between Dnmt1 conditional mice infected with HV68-WT and those infected with HV68-Cre (data not shown). The absence of a global defect was expected since (i) there was no defect in acute virus replication and (ii) only a very small percentage of spleen cells are latently infected with HV68, and hence only a small percentage of B cells would undergo Cre-recombinase-mediated deletion of the Dnmt1 allele.
By day 42 postinfection, latency in the Dnmt12lox/2lox mice infected with the HV68-Cre virus was almost completely ablated, whereas Dnmt12lox/2lox mice infected with HV68-WT had similar levels of viral genome-positive cells in the spleen to those in wild-type C57BL/6 mice (compare Fig. 6C and 7C). These data strongly indicate that the establishment of HV68 latency requires infection of a B-cell population(s) that subsequently undergoes several rounds of proliferation prior to establishing life-long latency in the memory B-cell compartment.
To extend the above analyses, we also examined the establishment and maintenance of latency in specific cell populations purified from infected Dnmt1 conditional mice (Fig. 8). On day 16 postinfection of the Dnmt1 conditional mice, as previously shown for infections of C57BL/6 mice, there was a high frequency of WT virus in splenic B cells (ca. 1 in 200) and only a very low frequency of WT virus in the splenic non-B-cell fraction (<1 in 10,000) (Fig. 8A). In contrast, the majority of latency observed on day 16 upon infection with the Cre-recombinase-expressing virus was present in the non-B-cell fraction (ca. 1 in 10,000), with only low levels of virus (<1 in 10,000 infected cells) detectable in the B-cell fraction (Fig. 8A). To assess whether latency recovers in Dnmt1 conditional mice infected with the Cre-expressing virus at late times postinfection, we examined the surface IgD+ (na?ve B cells) and IgD– splenic (germinal-center and memory B cells) B-cell populations at 3 months postinfection. As previously shown, with the WT virus nearly all of the detectable latency was present in the surface IgD– B-cell fraction (Fig. 8B). However, infection of the Dnmt1 conditional mice with the Cre-recombinase-expressing virus resulted in nearly undetectable levels of virus in the surface IgD-negative B-cell fraction, with slightly higher levels of virus infection present in the na?ve B-cell fraction (equivalent to the levels of virus infection in the na?ve B-cell population observed with WT virus infection) (Fig. 8B). Thus, these data directly demonstrate an inability of the Cre-recombinase-expressing virus to establish latency in the memory B-cell reservoir following the infection of Dnmt1 conditional mice.
DISCUSSION
The long-term latency reservoir for HV68 is memory B cells, although at early times postinfection latency can be detected in na?ve, germinal-center, and memory B cells (10, 42). We have hypothesized that, like the case for EBV infection in humans, the establishment of HV68 latency in memory B cells is the result of virus infection of na?ve B cells, followed by the differentiation of latently infected na?ve B cells through the highly proliferative germinal-center reaction and ultimately into the memory B-cell reservoir. The observed bias of virus infection toward the proliferating B-cell population during the establishment of latency shown here is consistent with this model. The major alternative model for the establishment of virus latency in memory B cells is based on HV68 directly infecting na?ve, germinal-center, and memory B cells, with the observed differential persistence of virus infection in memory B cells simply reflecting the longevity of this reservoir. With this model, we might also observe a bias of virus infection towards proliferating B cells, but this would not be an obligate step toward establishing latency in the memory B-cell reservoir. However, we cannot rule out the possibility that HV68 infection of memory B cells drives them to proliferate, which would also lead to their attrition upon the loss of Dnmt1 activity. Thus, the results shown here demonstrate that the proliferation of latently infected B cells precedes the establishment of long-term virus latency in the memory B-cell compartment. However, these studies cannot conclusively distinguish between (i) the direct infection of memory B cells followed by virus-driven proliferation and (ii) virus infection of na?ve and/or germinal-center B cells with subsequent differentiation to latently infected memory B cells. In addition, we do not know whether virus infection drives the observed proliferation of latently infected B cells or whether the virus preferentially infects proliferating B cells.
A further consideration for ablating Dnmt1 function comes from previous studies with EBV, which have shown that there is a tight correlation between the methylation of discrete regions of the viral genome and the establishment of long-term latency in the memory B-cell reservoir (33). It has been hypothesized that EBV genome methylation is critical for silencing the expression of several latency genes that are expressed during the growth-transforming latency program (EBNAs 2, 3a, 3b, and 3c and LMPs 1 and 2A), some of which are targets for CD8+ T-cell control of EBV infection (25). Although no link between methylation of the HV68 genome and the establishment of latency has been established to date, it is possible that the deletion of Dnmt1 leads to decreased levels of viral genome methylation in proliferating latently infected B cells. This may lead to HV68-infected cells being more vulnerable to host immune system recognition and elimination as a consequence of altered viral gene expression. Unfortunately, it is currently not possible to distinguish between effects on the methylation state of the viral genome and those on the cellular genome. However, it is important that the loss of latently infected cells by either mechanism due to the deletion of Dnmt1 is dependent on the proliferation of latently infected B cells. Since data in the literature have shown that cells that have a deletion of Dnmt1 survive only two to three rounds of cell division prior to dying and that B cells that undergo isotype switching go through at least three to six rounds of cell division (7, 14, 33), we currently favor the hypothesis that the loss of latently infected B cells is directly due to the loss of cellular genome methylation.
In the case of EBV infection, there is a tight anatomical compartmentalization of latency programs. In peripheral blood, EBV latency is exclusively found in circulating resting memory B cells (3). Similarly, by 3 months postinfection of mice, HV68 latency in the blood is almost exclusive found within the circulating isotype-switched memory B-cell population (39). In contrast, in the tonsils of seropositive individuals that are shedding virus, EBV infection of na?ve, germinal-center, and memory B cells is observed (10). The latter is very reminiscent of what is observed during the establishment of HV68 latency in the spleen (e.g., on day 16). Whether there is ongoing reseeding of B-cell latency at sites of HV68 reactivation and shedding is currently unknown, but this seems likely. Thus, for both EBV and HV68 infections, there appear to be distinct anatomical restrictions with respect to the forms of latency observed.
As stated above, the proposed model for the establishment of HV68 latency closely mirrors the proposed model for the establishment of EBV latency in humans. Models of EBV infection and the establishment of viral latency propose that the virus infects na?ve B cells and then drives these B cells to enter the germinal center reaction (35, 36). EBV has also been proposed to increase the survival rate of germinal center B cells, allowing the infected cells to successfully differentiate into memory B cells (27). This model has been derived from several experiments where peripheral blood mononuclear cells were sorted into different stages of B-cell differentiation and analyzed for the expression of EBV proteins (4). Distinct EBV latency-associated gene expression patterns which are consistent with this widely accepted model have been identified for specific B-cell populations. As with the model for HV68, there are several alternative scenarios for the establishment of EBV latency in the memory B-cell compartment, including the direct infection of memory B cells. Due to the limited host range of EBV, further experiments to track the virus in vivo will be difficult.
There is a large amount of evidence suggesting that EBV is capable of driving the germinal-center reaction in infected B cells. LMP1 and LMP2A are EBV-encoded proteins that are capable of providing prosurvival signals. LMP1 has been shown to activate the NF-B pathway similar to the engagement of CD40, while LMP2A has immunoreceptor tyrosine activation motifs in its cytoplasmic domain, perhaps preventing activation-induced cell death during the germinal-center reaction as well as providing survival signals (6, 16). Several studies have suggested that EBV might stimulate cells to acquire a germinal-center B-cell phenotype but that these infected cells do not actually participate in the germinal center reaction (18).
To date, no homologs of EBV LMP1 or LMP2A have been identified in HV68, and there is currently little evidence to suggest that HV68 drives B cells to undergo the germinal-center reaction. We have observed in the course of studies characterizing a HV68 mutant which cannot express the latency-associated M2 gene product that there is an accumulation of latently infected na?ve B cells in the spleen compared to wild-type virus-infected cells during the establishment of latency. One interpretation of this analysis is that the M2 gene product is involved in the virus-driven differentiation of latently infected na?ve B cells (40). Alternatively, it is possible that HV68 takes advantage of the B-cell proliferation that is triggered upon virus infection by infecting antigen-stimulated B cells during the acute phase of infection. Through this mechanism, the virus may passively end up in memory B cells. A prediction from the latter model is that latently infected memory B cells would be specific for HV68. Another possibility is that even if the virus does not actively drive the germinal-center reaction, it may provide survival signals for the infected cell to escape activation-induced cell death. HV68 does encode several proteins that could play critical roles in infected germinal-center B cells, including a bcl2 homolog (v-bcl2), a D-type cyclin homolog (v-cyclin), and a G-coupled receptor (v-GPCR) (38, 39). In addition, there are still many genes encoded within the viral genome with unknown functions.
In summary, the results presented here demonstrate that the establishment of chronic HV68 infection requires B-cell proliferation prior to the establishment of long-term latency in the memory B-cell compartment. These studies point to a role for gammaherpesviruses utilizing normal B-cell biology during the course of a viral infection. Further analysis is needed to fully understand the role of both virally encoded genes and cellular genes in this process.
ACKNOWLEDGMENTS
Samuel H. Speck is supported by grants CA43143, CA52004, CA58524, and CA095318. Janice M. Moser is supported by a postdoctoral fellowship from the American Cancer Society.
We thank Robert Karaffa and Michael Hulsey for their assistance and expertise with fluorescence-activated cell sorting and Joshy Jacob, Craig Chappell, and the Speck lab for helpful comments on this research.
REFERENCES
Adler, H., M. Messerle, and U. H. Koszinowski. 2001. Virus reconstituted from infectious bacterial artificial chromosome (BAC)-cloned murine gammaherpesvirus 68 acquires wild-type properties in vivo only after excision of BAC vector sequences. J. Virol. 75:5692-5696.
Adler, H., M. Messerle, M. Wagner, and U. H. Koszinowski. 2000. Cloning and mutagenesis of the murine gammaherpesvirus 68 genome as an infectious bacterial artificial chromosome. J. Virol. 74:6964-6974.
Babcock, G. J., L. L. Decker, R. B. Freeman, and D. A. Thorley-Lawson. 1999. Epstein-Barr virus-infected resting memory B cells, not proliferating lymphoblasts, accumulate in the peripheral blood of immunosuppressed patients. J. Exp. Med. 190:567-576.
Babcock, G. J., D. Hochberg, and D. A. Thorley-Lawson. 2000. The expression pattern of Epstein-Barr virus latent genes in vivo is dependent upon the differentiation stage of the infected B cell. Immunity 13:497-506.
Bird, A. P., and A. P. Wolffe. 1999. Methylation-induced repression—belts, braces, and chromatin. Cell 99:451-454.
Caldwell, R. G., J. B. Wilson, S. J. Anderson, and R. Longnecker. 1998. Epstein-Barr virus LMP2A drives B cell development and survival in the absence of normal B cell receptor signals. Immunity 9:405-411.
Cherry, S. R., C. Beard, R. Jaenisch, and D. Baltimore. 2000. V(D)J recombination is not activated by demethylation of the kappa locus. Proc. Natl. Acad. Sci. USA 97:8467-8472.
Clambey, E. T., H. W. Virgin, and S. H. Speck. 2000. Disruption of the murine gammaherpesvirus 68 M1 open reading frame leads to enhanced reactivation from latency. J. Virol. 74:1973-1984.
Fan, G., C. Beard, R. Z. Chen, G. Csankovszki, Y. Sun, M. Siniaia, D. Biniszkiewicz, B. Bates, P. P. Lee, R. Kuhn, A. Trumpp, C. Poon, C. B. Wilson, and R. Jaenisch. 2001. DNA hypomethylation perturbs the function and survival of CNS neurons in postnatal animals. J. Neurosci. 21:788-797.
Flano, E., I. Kim, D. L. Woodland, and M. A. Blackman. 2002. g-Herpesvirus latency is preferentially maintained in splenic germinal center and memory B cells. J. Exp. Med. 196:1363-1372.
Glimm, H., I. H. Oh, and C. J. Eaves. 2000. Human hematopoietic stem cells stimulated to proliferate in vitro lose engraftment potential during their S/G(2)/M transit and do not reenter G(0). Blood 96:4185-4193.
Gothot, A., R. Pyatt, J. McMahel, S. Rice, and E. F. Srour. 1997. Functional heterogeneity of human CD34(+) cells isolated in subcompartments of the G0/G1 phase of the cell cycle. Blood 90:4384-4393.
Holyoake, T., X. Jiang, C. Eaves, and A. Eaves. 1999. Isolation of a highly quiescent subpopulation of primitive leukemic cells in chronic myeloid leukemia. Blood 94:2056-2064.
Jackson-Grusby, L., C. Beard, R. Possemato, M. Tudor, D. Fambrough, G. Csankovszki, J. Dausman, P. Lee, C. Wilson, E. Lander, and R. Jaenisch. 2001. Loss of genomic methylation causes p53-dependent apoptosis and epigenetic deregulation. Nat. Genet. 27:31-39.
Joseph, A. M., G. J. Babcock, and D. A. Thorley-Lawson. 2000. Cells expressing the Epstein-Barr virus growth program are present in and restricted to the naive B-cell subset of healthy tonsils. J. Virol. 74:9964-9971.
Kilger, E., A. Kieser, M. Baumann, and W. Hammerschmidt. 1998. Epstein-Barr virus-mediated B-cell proliferation is dependent upon latent membrane protein 1, which simulates an activated CD40 receptor. EMBO J. 17:1700-1709.
Kuppers, R. 2003. B cells under influence: transformation of B cells by Epstein-Barr virus. Nat. Rev. Immunol. 3:801-812.
Kurth, J., M. L. Hansmann, K. Rajewsky, and R. Kuppers. 2003. Epstein-Barr virus-infected B cells expanding in germinal centers of infectious mononucleosis patients do not participate in the germinal center reaction. Proc. Natl. Acad. Sci. USA 100:4730-4735.
Kurth, J., T. Spieker, J. Wustrow, G. J. Strickler, L. M. Hansmann, K. Rajewsky, and R. Kuppers. 2000. EBV-infected B cells in infectious mononucleosis: viral strategies for spreading in the B cell compartment and establishing latency. Immunity 13:485-495.
Ladd, A. C., R. Pyatt, A. Gothot, S. Rice, J. McMahel, C. M. Traycoff, and E. F. Srour. 1997. Orderly process of sequential cytokine stimulation is required for activation and maximal proliferation of primitive human bone marrow CD34+ hematopoietic progenitor cells residing in G0. Blood 90:658-668.
Lee, P. P., D. R. Fitzpatrick, C. Beard, H. K. Jessup, S. Lehar, K. W. Makar, M. Perez-Melgosa, M. T. Sweetser, M. S. Schlissel, S. Nguyen, S. R. Cherry, J. H. Tsai, S. M. Tucker, W. M. Weaver, A. Kelso, R. Jaenisch, and C. B. Wilson. 2001. A critical role for Dnmt1 and DNA methylation in T cell development, function, and survival. Immunity 15:763-774.
Mackett, M., J. P. Stewart, S. de V Pepper, M. Chee, S. Efstathiou, A. A. Nash, and J. R. Arrand. 1997. Genetic content and preliminary transcriptional analysis of a representative region of murine gammaherpesvirus 68. J. Gen. Virol. 78:1425-1433.
Mao, X., Y. Fujiwara, A. Chapdelaine, H. Yang, and S. H. Orkin. 2001. Activation of EGFP expression by Cre-mediated excision in a new ROSA26 reporter mouse strain. Blood 97:324-326.
Moorman, N. J., D. O. Willer, and S. H. Speck. 2003. The gammaherpesvirus 68 latency-associated nuclear antigen homolog is critical for the establishment of splenic latency. J. Virol. 77:10295-10303.
Moser, J. M., J. W. Upton, K. S. Gray, and S. H. Speck. 2005. Ex vivo stimulation of B cells infected with gammaherpesvirus 68 triggers reactivation from latency. J. Virol. 79:5227-5231.
Paulson, E. J., and S. H. Speck. 1999. Differential methylation of Epstein-Barr virus latency promoters facilitates viral persistence in healthy seropositive individuals. J. Virol. 73:9959-9968.
Pollock, J. L., and H. W. Virgin. 1995. Latency, without persistence, of murine cytomegalovirus in the spleen and kidney. J. Virol. 69:1762-1768.
Schuster, V. H., and M. Muschen. 2003. Epstein-Barr virus and the B cell: a secret romance. Trends Microbiol. 11:243-245.
Shapiro, H. M., and N. G. Perlmutter. 2001. Violet laser diodes as light sources for cytometry. Cytometry 44:133-136.
Sherr, C. J. 2000. The Pezcoller lecture: cancer cell cycles revisited. Cancer Res. 60:3689-3695.
Smith, G. A., and L. W. Enquist. 2000. A self-recombining bacterial artificial chromosome and its application for analysis of herpesvirus pathogenesis. Proc. Natl. Acad. Sci. USA 97:4873-4878.
Speck, S. H., and H. W. Virgin. 1999. Host and viral genetics of chronic infection: a mouse model of gamma-herpesvirus pathogenesis. Curr. Opin. Microbiol. 2:403-409.
Sunil-Chandra, N. P., S. Efstathiou, J. Arno, and A. A. Nash. 1992. Virological and pathological features of mice infected with murine gammaherpesvirus 68. J. Gen. Virol. 73:2347-2356.
Tangye, S. G., A. Ferguson, D. T. Avery, C. S. Ma, and P. D. Hodgkin. 2002. Isotype switching by human B cells is division-associated and regulated by cytokines. J. Immunol. 169:4298-4306.
Thompson, M. P., and R. Kurzrock. 2004. Epstein-Barr virus and cancer. Clin. Cancer Res. 10:803-821.
Thorley-Lawson, D. A. 2001. Epstein-Barr virus: exploiting the immune system. Nat. Rev. Immunol. 1:75-82.
Thorley-Lawson, D. A., and A. Gross. 2004. Persistence of the Epstein-Barr virus and the origins of associated lymphomas. N. Engl. J. Med. 350:1328-1337.
Tibbetts, S. A., J. Loh, V. Van Berkel, J. S. McClellan, M. A. Jacoby, S. B. Kapadia, S. H. Speck, and H. W. T. Virgin. 2003. Establishment and maintenance of gammaherpesvirus latency are independent of infective dose and route of infection. J. Virol. 77:7696-7701.
Virgin, H. W., and S. H. Speck. 1999. Unraveling immunity to gamma-herpesviruses: a new model for understanding the role of immunity in chronic virus infection. Curr. Opin. Immunol. 11:371-379.
Virgin, H. W., P. Latreille, P. Wamsley, K. Hallsworth, K. E. Weck, A. J. Dal Canto, and S. H. Speck. 1997. Complete sequence and genomic analysis of murine gammaherpesvirus 68. J. Virol. 71:5894-5904.
Weck, K. E., M. L. Barkon, L. I. Yoo, S. H. Speck, and H. I. Virgin. 1996. Mature B cells are required for acute splenic infection, but not for establishment of latency, by murine gammaherpesvirus 68. J. Virol. 70:6775-6780.
Weck, K. E., S. S. Kim, H. I. Virgin, and S. H. Speck. 1999. B cells regulate murine gammaherpesvirus 68 latency. J. Virol. 73:4651-4661.
Willer, D. O., and S. H. Speck. 2003. Long-term latent murine gammaherpesvirus 68 infection is preferentially found within the surface immunoglobulin D-negative subset of splenic B cells in vivo. J. Virol. 77:8310-8321.
Yu, D., G. A. Smith, L. W. Enquist, and T. Shenk. 2002. Construction of a self-excisable bacterial artificial chromosome containing the human cytomegalovirus genome and mutagenesis of the diploid TRL/IRL13 gene. J. Virol. 76:2316-2328.(Janice M. Moser, Jason W.)