This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Guerreiro-Cacais, A. O.
Right arrow Articles by Levitsky, V.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Guerreiro-Cacais, A. O.
Right arrow Articles by Levitsky, V.

 Previous Article  |  Next Article 

Journal of Virology, February 2007, p. 1390-1400, Vol. 81, No. 3
0022-538X/07/$08.00+0     doi:10.1128/JVI.01999-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Inhibition of Heavy Chain and ß2-Microglobulin Synthesis as a Mechanism of Major Histocompatibility Complex Class I Downregulation during Epstein-Barr Virus Replication{triangledown}

Andre Ortlieb Guerreiro-Cacais,1,4,{dagger} Mehmet Uzunel,3 Jelena Levitskaya,2,4,{dagger} and Victor Levitsky1,4*

IRIS Center for Strategic Research, Department of Microbiology, Tumor and Cell Biology,1 Oncology and Pathology Department, Cancer Centrum Karolinska, Karolinska Institute, Stockholm, Sweden,2 Department of Clinical Immunology, Karolinska University Hospital, Huddinge 141 86, Sweden,3 Division of Biomedical Sciences, Johns Hopkins Singapore, Singapore4

Received 14 September 2006/ Accepted 5 November 2006


arrow
ABSTRACT
 
The mechanisms of major histocompatibility complex (MHC) class I downregulation during Epstein-Barr virus (EBV) replication are not well characterized. Here we show that in several cell lines infected with a recombinant EBV strain encoding green fluorescent protein (GFP), the virus lytic cycle coincides with GFP expression, which thus can be used as a marker of virus replication. EBV replication resulted in downregulation of MHC class II and all classical MHC class I alleles independently of viral DNA synthesis or late gene expression. Although assembled MHC class I complexes, the total pool of heavy chains, and ß2-microglobulin (ß2m) were significantly downregulated, free class I heavy chains were stabilized at the surface of cells replicating EBV. Calnexin expression was increased in GFP+ cells, and calnexin and calreticulin accumulated at the cell surface that could contribute to the stabilization of class I heavy chains. Decreased expression levels of another chaperone, ERp57, and TAP2, a transporter associated with antigen processing and presentation, correlated with delayed kinetics of MHC class I maturation. Levels of both class I heavy chain and ß2m mRNA were reduced, and metabolic labeling experiments demonstrated a very low rate of class I heavy chain synthesis in lytically infected cells. MHC class I and MHC class II downregulation was mimicked by pharmacological inhibition of protein synthesis in latently infected cells. Our data suggest that although several mechanisms may contribute to MHC class I downregulation in the course of EBV replication, inhibition of MHC class I synthesis plays the primary role in the process.


arrow
INTRODUCTION
 
Epstein-Barr virus (EBV) is a human gammaherpesvirus, which causes infectious mononucleosis and is associated with a number of malignancies arising from B cells, epithelial cells, and other cell lineages. EBV establishes latent infection predominantly, if not exclusively, in B lymphocytes, but virus replication can occur in both B lymphocytes and epithelial cells of the oropharynx. In the course of virus replication, approximately 70 lytic cycle proteins are expressed in EBV-infected cells in three temporal stages: immediate early (IE), early, and late. Entry into the lytic cycle is triggered by expression of either of the two IE genes BZLF1 and BRLF1. Both the BZLF1 and BRLF1 proteins function as transcriptional activators and initiate the ordered cascade of viral lytic gene expression culminating in the release of infectious virus (reviewed in reference 18).

The majority of EBV carriers mount strong cytotoxic T-lymphocyte (CTL) responses specific to EBV lytic cycle antigens (5, 6, 15). EBV-specific CTLs appear to play the most important role in the immunological control of EBV replication and EBV-induced malignant transformation. CTLs recognize their target cells through major histocompatibility complex (MHC) class I molecules loaded with antigen-derived peptides. Peptide ligands are generated in the cytosol by the proteolytic action of the proteasome and trimming peptidases and then translocated into the endoplasmic reticulum (ER) by heterodimeric transporters associated with antigen presentation 1 and 2 (TAP1 and TAP2). In the ER, MHC class I heavy chains interact with the chaperone calnexin, followed by interaction with the chaperones calreticulin, ERp57, and ß2-microglobulin (ß2m) and a preformed complex of TAPs and tapasin. This peptide loading complex is essential for efficient assembly of MHC class I complexes and their egress from the ER through the Golgi to the cell surface (reviewed in reference 8).

To escape from CD8+ T-cell recognition and destruction, viruses have developed strategies to inhibit the expression of MHC class I. These mechanisms include transcriptional downregulation of the heavy chain expression, interference with antigen processing by the proteasome, blocking of peptide transport into the ER, dislocation of MHC class I heavy chains from the ER to the cytosol for subsequent degradation, retention of MHC class I in the ER, and targeted degradation of class I from post-ER compartments. Several of these strategies for herpes simplex virus, varicella-zoster virus, Kaposi's sarcoma-associated herpesvirus, and most notably cytomegalovirus (CMV) have been dissected in detail previously (reviewed in references 19 and 36).

EBV replication also results in downregulation of MHC class I molecules at the surface of infected cells. However, the mechanisms of this process are not well understood. Previous studies have shown that in lymphoblastoid cell lines, the EBV lytic cycle and MHC class I downregulation are paralleled by downregulation of MHC class II, CD40, and CD54 while CD19, CD80, and CD86 are not affected, pointing to a possible specificity of the process (16). Reduction of MHC class I expression was found to be an early lytic cycle event that was not dependent on viral DNA replication. The immediate-early BZLF1 protein was implicated in MHC class I downregulation in lymphoblastoid cell lines through inhibition of latent membrane protein 1 (LMP1), which upregulates MHC class I expression in cells latently infected by EBV and is also expressed throughout the lytic cycle. A more efficient system has been developed recently to study EBV replication based on transfection of EBV-positive Akata cells with a reporter which is expressed during the lytic cycle, allowing identification and isolation of cells supporting virus replication from those in latency (29). Experiments performed with this model revealed that the capacity of the TAP1/TAP2 heterodimer to translocate peptides into the ER is inhibited by up to 70% during EBV replication, and because the expression of MHC class I heavy chains was found to be unaltered, it was suggested that this mechanism plays the primary role in downregulation of MHC class I molecules in the course of EBV replication (29).

In this study, we made use of a similar model based on a recombinant Akata virus carrying a cassette expressing green fluorescent protein (GFP) and the neomycin resistance gene inserted into the open reading frame of the nonessential thymidine kinase gene (34). We show that in these cells, the GFP gene is expressed during the lytic cycle with kinetics similar to those of the early EBV lytic genes. Thus, cells expressing GFP can be identified, separated, and used for immunological or biochemical analysis of changes caused by virus replication in EBV-infected cells. Our results show that even though multiple mechanisms might operate during EBV lytic replication, inhibition of MHC class I heavy chains and ß2m synthesis is one of the key mechanisms of EBV-induced MHC class I downregulation.


arrow
MATERIALS AND METHODS
 
Cell lines and induction of EBV replication. The Bx1g variants of the Burkitt's lymphoma cell line Akata (Akata Bx1g), the gastric carcinoma cell line AGS (AGS Bx1g), and the colorectal epithelial carcinoma cell line HCT116 (HCT116 Bx1g) were obtained by the infection of EBV-negative Akata, AGS, and HCT116 cells with a recombinant virus originated from the EBV strain derived from the wild-type Akata cells. The recombinant virus was generated by introducing a cassette containing the neomycin resistance gene under the control of a thymidine kinase promoter and a modified GFP gene under the control of the immediate-early cytomegalovirus promoter into the open reading frame of the nonessential thymidine kinase gene (refer to reference 24 for more details on the generation of the virus). The Bx1g cell variants were grown in medium supplemented with 500 µg/ml of Geneticin sulfate (G418; Sigma Aldrich, Saint Louis, MO). In Akata Bx1g cells, the EBV lytic cycle was induced by culturing in the presence of 50 µg/ml goat F(ab')2 fragments specific to human immunoglobulin G (IgG; MP Biomedicals, Aurora, OH) for the indicated periods of time. The lytic cycle in AGS Bx1g and HCT116 Bx1g cells was induced by culturing for 24 h with 30 ng/ml of 12-O-tetradecanoylphorbol-13-acetate (TPA; Sigma Aldrich) and 0.5 mM sodium butyrate (Calbiochem, San Diego, CA). The chemicals were then removed, and the culturing was continued for the indicated periods of time. All cells were grown in RPMI 1640 supplemented with L-glutamine and 10% fetal calf serum.

To inhibit viral DNA replication and late lytic cycle gene expression or to block protein synthesis, cells were cultured in the presence of 0.2 mM acyclovir (Alpharma, Oslo, Norway) or 50 µg/ml cycloheximide (Sigma Aldrich), respectively.

Antibodies. BZ.1 (Dako, Glostrup, Denmark) is a monoclonal antibody (MAb) specific to the immediate-early BZLF1 protein of EBV. 85K and 2L10 are MAbs specific to the p85 early antigen of EBV and to the late envelope glycoprotein gp350/220, respectively. W6/32 (HB95; ATCC) is a pan-MHC class I-specific MAb which recognizes heavy chains associated with ß2m. r{alpha}HC (rabbit anti-heavy chain) is a rabbit antiserum that recognizes free heavy chains of MHC class I molecules encoded by the human leukocyte antigen A (HLA-A), B, or C locus (32) and HC10 is a MAb that binds to a linear epitope exposed on free heavy chains of the HLA-B and -C loci (32). Both Abs were a kind gift of Hidde Ploegh (Harvard Medical School, Boston, MA). The anti-HLA-A (clone 108-2C5) and -HLA-B (clone Joan-1) MAbs were both from Lab Vision (Fremont, CA). The L31 Ab recognizes a linear {alpha}1 domain epitope exposed on free heavy chains of HLA-C (a kind gift of Louise Berg, Karolinska Institute, Stockholm, Sweden). Abs specific to ERp57 or calreticulin (both from StressGen, Victoria, Canada), calnexin (Abcam, Cambridge, United Kingdom), or tapasin (BD Biosciences) were used for both flow cytometry and immunoblotting. Abs specific to ß2m, clone TÜ99 (BD Biosciences, San Jose, CA), and TAP1 or TAP2 (BD Biosciences), as well as fluorescein isothiocyanate-conjugated Ab against MHC class I, MHC class II, CD19, and CD86 (all from BD Biosciences) were used for flow cytometry. Rabbit antiserum specific to ß2m (Dako), Abs against the proteasomal subunits LMP2, LMP7, MECL1 (monoclonal antibodies), and {alpha}3 (rabbit polyclonal) (Affinity Research, Plymouth Meeting, PA) as well as GFP-specific Ab (Roche, Indianapolis, IN) or actin-specific Ab (Sigma Aldrich) were used only for immunoblotting. The tripeptidyl peptidase II (TPPII)-specific MAb was from Immunesystem (Uppsala, Sweden). The TAP1- and TAP2-specific Abs used in immunoblotting were a kind gift from John Trowsdale (Cambridge Institute for Medical Research, Cambridge, United Kingdom). All antibody preparations used as isotype controls for flow cytometry were purchased from either Dako or BD Biosciences. The goat anti-mouse allophycocyanin (APC)-conjugated antibody for indirect immunostaining was purchased from BD Biosciences. The relevant secondary antibodies for immunoblotting were purchased form Amersham Biosciences (Uppsala, Sweden).

Flow cytometry. At the indicated time points after induction, Akata Bx1g, AGS Bx1g, or HCT116 Bx1g cells were washed twice in phosphate-buffered saline (PBS), fixed for 25 min in Cytofix/Cytoperm (BD Biosciences), washed three times in Perm/Wash solution (BD Biosciences), and blocked for 30 min on ice with serum from an EBV-negative individual (5% in PBS). Samples were then incubated for 30 min on ice with primary unconjugated Abs, washed, and incubated with goat anti-mouse APC-conjugated Ab for an additional 30 min on ice. All acquisitions were performed on a FACScalibur flow cytometer and analyzed using CellQuest (BD Biosciences). Double stainings were performed by labeling cells first with unconjugated BZLF1-specific Ab, whose binding was revealed by staining with a relevant anti-mouse conjugate, as described above, followed by staining with MHC class I-, class II-, CD19-, or CD86-specific directly conjugated Ab. Staining for HLA-C with L31 Ab, which recognizes free HLA-C heavy chains, was performed after the cells were treated for 2 min with a low-pH buffer (0.131 M citric acid, 0.066 M Na2HPO4, pH 3.3) to elute the peptides and dissociate the heavy chain from ß2m.

Analysis of protein expression by immunoblotting. Akata Bx1g cells were induced with IgG-specific Abs for either 24 or 48 h and then sorted using a MoFlo sorter (Dako Cytomation) into GFP+ and GFP– cells. 7AAD (BD Biosciences) was used to exclude dead cells during sorting. Cells were counted and pelleted, and the same numbers of GFP-positive cells and GFP-negative cells were lysed in reducing electrophoresis sample buffer (1 x 106 cells/100 µl). Electrophoresis was carried out on 10% polyacrylamide gels or 8 to 16% polyduramide gradient gels (Cambrex, East Rutherford, NJ) and proteins were transferred onto nitrocellulose membranes (Bio-Rad, Hercules, CA). Membranes were blocked in PBS containing 5% skimmed milk and probed with different primary Abs followed by secondary antibodies conjugated to horseradish peroxidase. The signal was detected using Super Signal chemiluminescence detection solution (Pierce, Rockford, IL), and images were acquired on a LAS-3000 luminescent image analyzer (Fujifilm Medical Systems, Stamford, CT).

Pulse-chase experiments, immunoprecipitation and MHC class I maturation. Lytic cycle was induced in Akata Bx1g cells for 24 or 48 h, which were then sorted into GFP+ (lytic) and GFP– (latent), as described above. For metabolic labeling, 2 x 106 cells/sample were incubated in methionine- and cysteine-free DMEM (Invitrogen, Carlsbad, CA) at 37°C for 1 h, spun down, and resuspended in fresh prewarmed methionine- and cysteine-free Dulbecco's modified Eagle's medium (DMEM) with 150 µCi of [35S]methionine-[35S]cysteine (Redivue Pro-mix L-[35S] in vitro cell labeling mix; Amersham) and cultured for 2 h at 37°C.

For short pulse-chase experiments, cells were labeled for 2 min with 400 µCi Redivue Pro-mix. Labeling was terminated by the addition of nonradioactive L-methionine and L-cysteine (1 mM) (Sigma), and samples were pelleted and frozen at time zero or at 5 and 30 min of chase.

To assess the role of protein degradation, cells were incubated in methionine- and cysteine-free DMEM at 37°C for 1 h in the presence of the proteasomal inhibitor lactacystine (10 µM) and the lysosomal inhibitor chloroquine (4 µM) (Sigma). Cells were labeled for 20 min with 400 µCi Redivue Pro-mix, washed, and then pelleted.

For the assessment of MHC class I maturation, cells were pulsed for 30 min, labeling was terminated as described above, cells were washed once and reincubated in prewarmed DMEM containing excess unlabeled L-methionine and L-cysteine and 10% fetal calf serum. Samples were collected immediately after the pulse (time zero) and at 1 h and 2 h of chase.

Samples were lysed in lysis buffer (10 mM Tris, pH 7.4, 1% Triton X-100, 1 mM EDTA, 150 mM NaCl, 2 mM phenylmethylsulfonyl fluoride) on ice. After 30 min, nuclei were removed by centrifugation (relative centrifugal force, 3,000; 10 min). MHC class I molecules were immunoprecipitated with either W6/32 or r{alpha}HC Abs after preclearing with mouse or rabbit serum, respectively.

For endo-ß-N-acetylglucosaminidase H (endoH) digestion, immune complexes bound to protein A Sepharose beads were eluted by incubation at 95°C for 5 min in elution buffer (0.1 M sodium acetate, 3 mM EDTA, 0.25% sodium dodecyl sulfate [SDS], pH 6.0). One half of each sample was digested overnight at 37°C with 2 mU endoH (Roche, Mannheim, Germany), and the other half was mock digested. Samples were mixed with reducing electrophoresis loading buffer and analyzed by SDS-polyacrylamide gel electrophoresis (PAGE) on a 10% polyacrylamide gel. Gels were fixed (40% methanol, 10% acetic acid, 5% glycerol), vacuum dried, and incubated in a Hypercassette (Amersham), and the signal was revealed in a Storm 860 PhosphorImager (Molecular Dynamics, GE Healthcare).

Real-time RT-PCR (RNA preparation and reverse transcription). After induction of the lytic cycle, Akata Bx1g cells were sorted into GFP– and GFP+ populations, carefully counted, and RNA was extracted from 1 x 106 cells of each population according to the instructions of the QIAGEN RNA blood minikit (Hilden, Germany). RNA was eluted with 30 µl of RNase-free water. The cDNA synthesis was performed in a total volume of 50 µl containing 30 µl RNA, 1 x first-strand buffer (50 mM Tris-HCl, pH 8.3, 75 mM KCl, 3 mM MgCl2), 105 µg/ml pdN6 (Amersham Biosciences, Piscataway, NJ), 1 mM of each deoxynucleoside triphosphate (Amersham Biosciences), 1 mM dithiothreitol, 0.48 units/µl RNasin (Promega, Madison, WI), and 4.8 units/µl Moloney murine leukemia virus reverse transcriptase (RT; Invitrogen, Paisley, Scotland), at 37°C for 1.5 h. The reaction was stopped by heating at 68°C for 15 min.

Real-time PCR. A total of 2.5 µl of cDNA was used in a 25-µl reaction mixture containing 1x SYBR green (Applied Biosystems, Foster City, CA) and 300 nM of each primer. Primers and primer sequences were as follows: HLA class I-F, 5'-ACC TGG AGA ACG GGA AGG A-3'; HLA class I-R, 5'-TGT GAT CTC CGC AGG GTA GAA-3'; ß2m-F, 5'-GAT GAG TAT GCC TGC CGT GTG-3'; ß2m-R, 5'-CAA TCC AAA TGC GGC ATC T-3'; G6PD-F, 5'-TGC CCC CGA CCG TCT AC-3'; G6PD-R, 5'-ATG CGG TTC CAG CCT ATC TG-3'; ABL-F, 5'-CGA AGG GAG GGT GTA CCA TTA C-3'; ABL-R, 5'-CGT TGA ATG ATG ATG AAC CAA CTC-3'. The PCR was performed and analyzed on the ABI 7000 sequence detection system with the following PCR conditions: 50°C for 2 min, 95°C for 10 min, followed by 40 PCR amplification cycles with 95°C for 15 s and 60°C for 1 min. Relative quantification of gene expression was calculated according to the Delta-Delta CT method. The formula used was Formula, where {Delta}CT equals the cycle threshold (CT) of a given gene in GFP+ cells minus the CT of the same gene in GFP– cells.


arrow
RESULTS
 
An alternative model for EBV lytic replication. Analysis of EBV lytic replication has been hampered by the lack of suitable permissive systems of virus infection. Recently, Ressing et al. described a new experimental model of EBV replication based on transfection of Burkitt's lymphoma Akata cells with a plasmid encoding the rat CD2 extracellular and transmembrane domains, with the cytosolic domain replaced with GFP (29). Expression of the fusion protein in this construct is driven by the BMRF1 early gene promoter. Upon induction of the lytic cycle by cross-linking of surface immunoglobulins, transfected cells supporting virus replication could be either studied based on their GFP expression or isolated with anti-rat CD2 beads (29).

We made use of an alternative system in which a cassette encoding the GFP gene under control of the CMV immediate-early promoter was inserted into the open reading frame of the nonessential thymidine kinase gene of the wild-type Akata virus (34). This recombinant virus was used to infect an EBV-negative variant of the Akata cell line, the gastric carcinoma cell line AGS, and the colorectal epithelial carcinoma cell line HCT116, generating EBV-positive cell lines referred to as Akata Bx1g, AGS Bx1g, and HCT116 Bx1g, respectively.

Cultures of cells infected with the recombinant virus typically contained a small population of cells in lytic replication (Fig. 1A, upper panels; data not shown for HCT116 Bx1g cells). In Akata Bx1g cells, the expression of the immediate-early gene BZLF1 quite strictly coincided with expression of GFP, whereas both GFP single-positive and GFP/BZLF1 double-positive populations could be clearly distinguished in AGS Bx1g cells and HCT116 Bx1g cells before and after induction of the lytic cycle. This probably reflects the higher steady-state activity of the CMV promoter in epithelial cells. Upon induction of the lytic cycle by surface IgG cross-linking or TPA and sodium butyrate, the proportions of Akata Bx1g and AGS Bx1g cells replicating EBV increased from approximately 2% to 40% and from 3% to 10%, respectively.


Figure 1
View larger version (48K):
[in this window]
[in a new window]

  		FIG. 1. EBV lytic replication results in GFP expression and MHC class I downregulation in Akata Bx1g and AGS Bx1g cells. Akata cells carrying a recombinant EBV strain with the GFP gene inserted under the immediate-early CMV promoter into the virus genome (Akata Bx1g cells) or AGS gastric epithelial cells infected with the same virus (AGS Bx1g cells) were treated as described in Materials and Methods to induce the EBV lytic cycle. Staining with the indicated specific antibodies and fluorescence-activated cell sorter (FACS) analysis were performed as described in Materials and Methods. A. EBV replication, GFP expression, and MHC class I downregulation coincide in Akata Bx1g and AGS Bx1g cells. To induce EBV replication, cells were cultured either with F(ab')2 fragments specific to human IgG for the indicated periods of time (Akata Bx1g) or with TPA and sodium butyrate (AGS Bx1g) for 24 h, washed, and incubated in standard culture medium for the remaining time of induction. The four upper panels show intracellular staining for BZLF1. The lower panels show surface staining for MHC class I using W6/32 antibody. In both cases, anti-mouse APC-conjugated antibody was used to detect binding of primary antibodies. The R1 and R2 regions in the lower panels define two different cell populations, GFP– and GFP+ cells, respectively, that were used for FACS analysis and cell sorting in this and all subsequent experiments. B. Signals used to activate the EBV lytic cycle in Akata or AGS cells induce overall upregulation of MHC class I, which is overridden by virus replication. Surface staining for MHC class I was done with W6/32 antibody and APC-conjugated anti-mouse secondary antibody. The upper panel shows the expression of MHC class I in the total population of either control (noninduced) or anti-IgG-treated (Akata-induced) or TPA-butyrate-treated (AGS-induced) cells. The four lower panels show either control or induced cells, each divided into GFP– and GFP+ populations. The histograms are presented to illustrate the distribution of cells with different levels of MHC class I expression in cultures replicating the virus. C. Wild-type Akata cells downregulate MHC class I and MHC class II molecules upon induction of the EBV lytic cycle. Wild-type EBV-positive Akata cells were incubated with F(ab')2 fragments specific to human IgG for 24 h. Indirect intracellular staining for BZLF1 was followed by staining with directly conjugated MHC class I-, class II-, CD19-, or CD86-specific antibody. Cells were gated as BZLF1 positive (pos) or BZLF1 negative (neg). The geometric means of fluorescence intensities (Geo MFI) of the gated populations are shown. D. HLA-A, -B, and -C alleles are downregulated to similar extents during EBV replication. At 24 h after induction, the levels of surface expression of classical MHC class I molecules were assessed in Akata Bx1g cells with either a pan-MHC class I-specific antibody (W6/32) or antibodies specifically recognizing HLA-A, -B, or -C alleles. The geometric means of fluorescence intensity obtained with each of the indicated antibodies for GFP-positive cells are shown as percentages relative to those of anti-IgG-treated GFP-negative (latent) cells (the means and standard deviations of three experiments).

In accordance with previous reports, cells supporting lytic replication expressed decreased surface levels of MHC class I (Fig. 1A, lower panels, and B) and MHC class II (Fig. 1C). Induction of the virus lytic cycle increased the number of cells with low levels of surface MHC class I in Akata Bx1g, AGS Bx1g (Fig. 1A and B, lower panels), and HCT116 Bx1g (data not shown) cells, although the kinetics of this process was slower in the epithelial cells, with visible increase of the GFP/BZLF1 double-positive population only 48 h posttreatment and onwards. In our experimental model, Ig cross-linking or TPA-sodium butyrate treatment led to an increase in MHC class I expression on the surface of cells that remained in latency compared to that on control untreated cells (Fig. 1B, upper panels). This observation contrasts with the findings of Ressing et al., who observed no difference in MHC class I expression between anti-Ig-treated and untreated cells latently infected with the virus (29). Therefore, in all experiments, we compared GFP+ cells replicating the virus with the GFP– population within the same cultures exposed to lytic cycle-inducing stimuli.

To exclude the possibility that the observed effect of virus replication on MHC class I expression was caused by the expression of GFP or other unknown effects mediated by the recombinant virus, wild-type Akata cells were induced to replicate EBV and then costained with the BZLF1-specific antibody and one of the antibodies specific to MHC class I, MHC class II, CD19, or CD86 (Fig. 1C). As previously reported, surface expression of CD19 and CD86 remained unchanged, whereas levels of MHC class I and II were decreased in cells replicating the virus. All classical MHC class I alleles were downregulated by approximately 60% on average in lytically infected Akata Bx1g cells as revealed by staining with antibodies that recognize pan-HLA-A, pan-HLA-B, or pan-HLA-C (Fig. 1D).

To get an insight into the role of different EBV proteins in the downregulation of MHC class I expression, we first assessed the kinetics of this process in relation to the kinetics of expression of immediate-early, early or late EBV proteins following induction of the lytic cycle. The immediate-early gene BZLF1 was already expressed after 6 h of induction of the lytic cycle in Akata Bx1g cells (Fig. 2). Early gene expression, assessed by staining for the EA-R protein, was undetectable at 6 h but picked up at 12 h, when virtually all cells expressing immediate-early genes also expressed early genes (Fig. 2A). At this time point, a small proportion of cells already expressed late genes, as assessed by staining for the envelope glycoprotein gp350/220, but late gene expression by the majority of GFP+ cells was seen only at 24 h. The percentage of cells that express lytic cycle genes peaked at 24 h, followed by a decrease at 48 h, which could be due to cell death as a result of lytic replication. Surface MHC class I was slightly upregulated at 6 h postinduction in cells that expressed BZLF1 in comparison to cells that were subjected to anti-Ig treatment but remained in latency. This upregulation was quickly reverted, and by 12 h, MHC class I was reduced by 20%; the extent of downregulation continued to increase, reaching its peak at 48 h. The degree of downregulation varied between individual experiments but averaged 60% by 48 h.


Figure 2
View larger version (50K):
[in this window]
[in a new window]

  		FIG. 2. MHC class I downregulation is independent of virus DNA replication. A. MHC class I downregulation is an early event in the virus lytic cycle. The lytic cycle was induced in Akata Bx1g cells by IgG cross-linking, and the levels of expression of immediate-early (BZLF1), early (EA-R-p85), and late (gp350/220) lytic cycle antigens were assessed by intracellular staining. In parallel, surface MHC class I was detected at each time point after induction. Percentages of cells positive for each of the antigens at the indicated time points are shown on the left y axis, whereas the relative MHC class I downregulation in GFP-positive cells compared with the GFP-negative cells is shown on the right y axis. B. MHC class I downregulation is independent of virus DNA replication and late gene expression. Akata Bx1g cells were induced by IgG cross-linking in the presence or absence of 0.2 mM acyclovir. At the indicated time points, cells were collected and stained for intracellular BZLF1 or gp350/220 expression. The numbers inside the dot plots indicate percentages of cells in the relevant quadrants. Mean fluorescence intensity of BZLF1-specific staining is shown for cells positive for both GFP and BZLF1. In parallel, samples were also stained for surface MHC class I, and the percentage of MHC class I downregulation was assessed as described above. The results of this analysis are shown on the graph.

The fact that the surface levels of MHC class I were already decreased by 12 h postinduction, as well as the fact that assembled molecules should have an average half-life of 6 to 8 h, suggested that MHC class I downregulation is induced by an immediate-early or early EBV proteins. In agreement with this assumption and previously published data (16), acyclovir greatly decreased late gene expression in induced Akata cells and reduced percentages of cells expressing BZLF1 but did not significantly affect the levels of BZLF1 expression and MHC class I downregulation in BZLF1-positive cells (Fig. 2B).

The total pool of MHC class I heavy chains and ß2m is downregulated, whereas free heavy chains accumulate in cells replicating EBV. To understand the fate of MHC class I in cells replicating EBV, we first asked whether the surface downregulation of assembled molecules reflected the overall decrease of heavy chain expression in these cells. Akata Bx1g cells were sorted into GFP– (latent) and GFP+ (lytic) populations at 48 h after induction. Western blot analysis of total cell lysates showed that lytic replication reduces the expression of MHC class I heavy chains as detected by two different heavy chain-specific antibodies (Fig. 3A). Levels of ß2m were also strongly reduced. In accordance with downregulation of assembled MHC class I at the cell surface (Fig. 1), cells subjected to surface IgG cross-linking but remaining in latency (GFP–) expressed slightly increased amounts of heavy chains in total cell lysates compared to noninduced controls (Fig. 4A). Cells in the lytic cycle had reduced levels of total heavy chains in comparison to either noninduced or induced but latently infected cells.


Figure 3
View larger version (30K):
[in this window]
[in a new window]

  		FIG. 3. Free heavy chains accumulate during lytic replication. A. Total MHC class I heavy chains and ß2m are downregulated in GFP-positive Akata Bx1g cells. Immunoblotting of total cell lysates of Akata Bx1g cells induced for 48 h with anti-IgG antibodies and sorted into GFP+ (lytic) and GFP– (latent) cells. The heavy chain expression was assessed by two different antibodies: r{alpha}HC, a rabbit antiserum that recognizes heavy chains of HLA-A, -B, and -C, and HC10, a mouse monoclonal antibody specific for heavy chains of the HLA-B and -C loci. B. Both surface and total pools of assembled MHC class I molecules as well as ß2m are downregulated during EBV replication while free MHC class I heavy chains accumulate in lytically infected cells. The upper panel shows density plots of Akata Bx1g cells 48 h after induction, surface stained with W6/32 (for assembled MHC class I), HC10 (for free MHC class I heavy chains), or TÜ99 (for ß2m) antibodies obtained in one representative experiment. The lower graphs show the percentage of change in the expression of indicated molecules at the cell surface or in permeabilized GFP+ cells relative to expression levels detected in GFP (latent) cells. The means and standard deviations of values obtained in 5 to 10 independent experiments are shown. FC, flow cytometry.


Figure 4
View larger version (48K):
[in this window]
[in a new window]

  		FIG. 4. Levels of expression of several proteins that participate in MHC class I processing and presentation are modulated during EBV lytic replication. A. Immunoblotting of total cell lysates of Akata Bx1g sorted 24 h after induction as described in the legend to Fig. 3. The control lane (Ctr) refers to cells kept under the same conditions as induced cells but without anti-IgG antibodies. The results of two separate representative experiments are shown. B. Relative change in the expression of the indicated chaperones and peptide transporters as assessed by flow cytometry (FC) of intact (surface) or permeabilized (total) cells or by immunoblotting (IB) of total cell lysates. The relative change in percentage refers to the change in the level of these proteins in cells supporting replication (GFP+) in relation to cells that remained in latency (GFP–). Means and standard deviations of at least three independent experiments are shown. N.D., not detected.

Next, we studied the expression of assembled MHC class I, free heavy chains, and ß2m both at the cell surface and in permeabilized cells, the latter reflecting the total pool of the analyzed molecules. The density plots in Fig. 3B show a typical surface staining for these molecules at 48 h postinduction. Both assembled MHC class I and ß2m were decreased in permeabilized cells and on the surface of GFP+ cells compared to GFP– cells at 24 h postinduction, and this downregulation became more pronounced at 48 h (Fig. 3B). In contrast, the levels of free heavy chain, detected by the MAb HC10, were unchanged at 24 h and tended to increase at 48 h in lytic cells compared to those in cells in latency. Also, downregulation of ß2m at the cell surface was statistically more significant (P value in nonpaired t test, 0.01) than that of assembled MHC class I after 48 h of lytic cycle induction. A stronger downregulation of ß2m was also revealed upon staining of permeabilized cells after 24 h of induction (P value, 0.002). Taken together, these results reveal that during EBV replication, expression levels of both heavy chains and ß2m are downregulated, and ß2m seems to be reduced to a greater extent. At the same time, both the total pool and surface fraction of free heavy chains are increased, suggesting that free heavy chains selectively accumulate at the surface of cells supporting the lytic cycle.

Proteins that aid the generation of folded and peptide-loaded MHC class I are modulated during EBV lytic replication. To investigate whether other mechanisms, in addition to decreased expression of heavy chains and ß2m, contribute to MHC class I downregulation in lytically infected cells, expression of several components of the MHC class I processing pathway in GFP+ and GFP– Akata Bx1g cells was analyzed by immunoblotting. GFP– cells were always negative for BZLF1 expression regardless of the time period between induction of the lytic cycle and cell sorting (24 or 48 h). No significant changes could be observed in the expression of LMP2, LMP7, and MECL-1 subunits of the immunoproteasome (Fig. 4A), nor was there a change in the expression of the nonproteolytic constitutive {alpha}3 proteasomal subunit. TPPII, another cytosolic peptidase implicated in the generation of MHC class I peptide epitopes, was slightly downregulated.

To compare the expression levels of peptide transporters and ER chaperones that play a role in the assembly and peptide loading of MHC class I, both immunoblotting and flow cytometry analysis of protein expression at the cell surface and in permeabilized cells were used as described for Fig. 3 and in Materials and Methods. Significant accumulation of calnexin was observed mostly in the intracellular compartments of cells in the lytic cycle after 24 h of induction. In contrast, calreticulin accumulated on the surface, while its overall expression was reduced by up to 50% (Fig. 4). Expression of the chaperone ERp57 showed a general reduction. Tapasin, TAP1, and TAP2 could not be detected at the cell surface by flow cytometry, and their overall levels were not modulated significantly, as detected by immunoblotting, with the exception of TAP2, whose expression was decreased by approximately 30% (Fig. 4).

MHC class I synthesis is reduced during lytic replication. Herpesviruses target MHC class I molecules for degradation through a number of mechanisms (31). Therefore, we investigated whether the decrease of MHC class I heavy chain expression in the course of EBV replication is due to inhibition of heavy chain synthesis or targeted protein degradation. Induced and sorted Akata Bx1g cells were labeled with [35S]methionine and [35S]cysteine, and their lysates were immunoprecipitated with W6/32 or r{alpha}HC antibodies specific to assembled MHC class I or free heavy chains, respectively. After 2 h of labeling, most labeled heavy chains were associated with ß2m and recovered with W6/32 antibody from control cells (Fig. 5A). Comparison of GFP– and GFP+ cells revealed a drastic reduction in 35S-labeled W6/32- and r{alpha}HC-reactive molecules in cells replicating the virus. The amount of de novo-synthesized MHC class I molecules was strongly reduced, by approximately 85% in GFP+ cells. Trichloroacetic acid precipitation of [35S]methionine-labeled cellular proteins revealed that cells in the lytic cycle incorporated the amino acid at 25% of the level of GFP– cells in latency (data not shown). This is consistent with overall inhibition of protein synthesis in cells replicating the virus. However, heavy chain downregulation observed by immunoblotting of total lysates of the same cells was relatively moderate, reaching only about 40% reduction in band intensity (Fig. 5B).


Figure 5
View larger version (61K):
[in this window]
[in a new window]

  		FIG. 5. Reduced synthesis of MHC class I is associated with inhibition of transcription during lytic replication. Akata Bx1g cells were induced with anti-IgG for either 48 h (A and B) or 24 h (C, D, E, F, and G) and then sorted into GFP– and GFP+ populations. Both populations were incubated for 1 h at 37°C in methionine- and cysteine-free DMEM prior to metabolic labeling. The bands shown in all experiments correspond to the heavy chain of MHC class I ({approx}45 kDa). A. Synthesis of MHC class I heavy chains is inhibited during EBV replication. Cells were metabolically labeled for 2 h as described in Materials and Methods and immunoprecipitated with W6/32 or r{alpha}HC antibody. Samples were separated by SDS-PAGE, and autoradiography of dried gels was performed using a PhosphorImager. B. The levels of newly synthesized MHC class I heavy chains are reduced during lytic replication. Equal numbers of GFP– and GFP+ cells were either lysed directly in SDS-containing lysis buffer or labeled with [35S]methionine-[35S]cysteine mix for 2 h and immunoprecipitated with W6/32 as described above. Total cell lysates (TCL) were separated by SDS-PAGE, transferred onto nitrocellulose filters, and probed for the heavy chain and actin specific antibody as a loading control. C. Rapid degradation does not appear to account for the decreased levels of newly synthesized heavy chains. Cells were labeled with [35S]methionine-[35S]cysteine mix for 2 min and incorporation of radioactivity was terminated by the addition of 1 mM of nonradioactive L-methionine and L-cysteine. Samples were pelleted and frozen either immediately (time zero) or after 5 and 30 min of chase, followed by sequential immunoprecipitation with W6/32 and r{alpha}HC antibody. Electrophoresis and autoradiography were performed as described above. D. Inhibitors of proteasomal or lysosomal degradation do not recover expression of newly synthesized heavy chains. Cells were incubated in methionine- and cysteine-free DMEM for 1 h in the presence of both proteasomal inhibitor lactacysin (10 µM) and lysosomal inhibitor chloroquine (4 µM), labeled for 20 min with 400 µCi [35S]methionine-[35S]cysteine followed by sequential immunoprecipitation with W6/32 and r{alpha}HC, electrophoresis, and autoradiography. E. MHC class I heavy chain and ß2m mRNA are reduced in cells in lytic replication. After sorting, GFP– and GFP+ cells were carefully counted and pellets of equal numbers of cells were used for reverse transcription and RT-PCR analysis of MHC class I and ß2m mRNA. Values of relative expression of these mRNAs in GFP+ relative to GFP– cells are shown. Means and standard deviations of three independent experiments. F. Inhibition of protein synthesis mimics the effect of lytic replication on downregulation of assembled surface MHC class I and class II molecules. Akata Bx1g cells were either cultured in control medium, treated with 50 µg/ml cycloheximide or induced into lytic cycle for 24 h and then stained for surface MHC class I and II. The expression levels of MHC class I and II on both GFP+ and cycloheximide-treated cells are shown as percentages relative to those in GFP– or untreated cells, respectively. G. Assembled MHC class I molecules synthesized in lytically infected cells are delayed in their maturation. Cells metabolically labeled for 30 min were incubated in prewarmed DMEM containing excesses of unlabeled L-methionine and L-cysteine and 10% fetal calf serum. Samples were collected immediately after the pulse (time zero) and at 1 h or 2 h of incubation. After immunoprecipitation with W6/32, assembled MHC class I was eluted from the Sepharose beads, digested by endoH (2 mM) and separated by SDS-PAGE together with nondigested controls. The line graph shows ratios between the intensities of endoH-resistant and -sensitive bands in GFP+ and GFP– cells.

These results could reflect two non-mutually exclusive scenarios, that heavy chains are synthesized at a low rate or newly made molecules are rapidly degraded. To address this question, we performed pulse-chase experiments on sorted cell populations, in which cells were metabolically labeled for 2 min and chased immediately at time zero or 5 or 30 min postlabeling. The cell lysates were sequentially immunoprecipitated with W6/32 and r{alpha}HC antibodies (Fig. 5C). In cells infected with EBV latently, r{alpha}HC-reactive molecules were visible at time zero and their intensity decreased at the subsequent time points, whereas the amount of assembled, W6/32-reactive MHC class I molecules increased over time. However, both W6/32- and r{alpha}HC-reactive molecules were practically undetectable in cells replicating the virus, supporting the notion that the rate of MHC class I synthesis is low during the EBV lytic cycle. This experiment still did not rule out completely the possibility that heavy chains are synthesized but degraded immediately. Therefore, we pretreated cells with a combination of lactacystin, a specific inhibitor of the proteasome, and chloroquine, an inhibitor of lysosomal degradation, and then performed metabolic labeling and sequential immunoprecipitation as described above (Fig. 5D). Immunoprecipitation of lysates of GFP– cells with the W6/32 and r{alpha}HC antibodies showed an accumulation of relevant protein moieties in cells pretreated with the inhibitors. In contrast, neither W6/32- nor r{alpha}HC-reactive species were revealed in cells replicating EBV even upon inhibition of protein degradation. Immunoprecipitation of ß2m also performed in this series of experiments demonstrated that its synthesis was inhibited in GFP+ cells to an extent comparable to that observed for MHC class I heavy chains (data not shown).

To investigate whether decreased transcription/increased turnover of mRNA contributes to the decrease of MHC class I heavy chain synthesis, the abundance of mRNA message for the classical MHC class I molecules and ß2m was determined using real-time RT-PCR analysis. Since CT values for mRNA of housekeeping genes (G6PD [glucose-6-phosphate dehydrogenase] and Abl protein kinase genes) were increased in GFP+ cells (data not shown), normalization to housekeeping gene expression levels could not be used to assess the abundance of MHC class I and ß2m mRNAs. Instead, normalization was achieved by extracting mRNA from equal numbers of sorted GFP– and GFP+ cells. As shown in Fig. 5E, the mRNA message for MHC class I was reduced by an average of 70%, whereas the ß2m message was reduced by approximately 60%.

To test directly whether global inhibition of protein synthesis can cause MHC class I downregulation in Akata Bx1g cells within the timeframe of experiments performed in this study, cells were either treated to induce the lytic cycle or incubated with 50 µg/ml of cycloheximide for 24 h (Fig. 5F). GFP+ cells were compared to anti-Ig-treated GFP– cells, while untreated cells served as a control for cycloheximide-treated cells. EBV replication and inhibition of protein synthesis by cycloheximide resulted in comparable reduction of MHC class I and class II expression.

Inhibition of TAP-mediated peptide translocation into the ER during EBV replication has been previously suggested (29). Our results revealed a slight inhibition of TAP2 expression as well as downregulation of TPPII and ERp57 in cells replicating the virus. All these changes could decrease the efficiency of peptide supply and loading in lytically infected cells, thereby affecting the kinetics of MHC class I progression through the ER and Golgi apparatus. To assess the efficiency of this process, Akata Bx1g cells were induced, sorted, labeled for 30 min with [35S]methionine-[35S]cysteine, and chased at the indicated time points after the pulse. After immunoprecipitation with W6/32 antibodies, samples were divided and either mock treated or treated with endoH, an enzyme which removes N-linked oligosaccharides attached to proteins in the ER but does not trim modified versions of these sugars acquired in the Golgi compartment. It has been previously demonstrated that the kinetics of MHC class I maturation is relatively slow in Akata cells due to their intrinsically low TAP activity (10, 17). Accordingly, a proportion of MHC class I molecules remained endoH sensitive even after 2 h of chase in the control GFP-negative cells (Fig. 5G). In agreement with the previous experiments, bands of very low intensity were detected in MHC class I immunoprecipitates from the lysates of GFP+ Akata Bx1g cells. Although it was clear that MHC class I molecules in cells replicating EBV progress to the Golgi, densitometry of endoH-sensitive and endoH-resistant bands indicated that there was a delay in MHC class I maturation during the EBV lytic cycle. However, this delay was revealed only upon a comparison of samples obtained at 2 h of chase and its significance for the phenomenon of MHC class I downregulation remains unclear.


arrow
DISCUSSION
 
Study of EBV replication was hampered for a long time by the lack of a permissive in vitro system. We hereby present a model for studying EBV replication that is based on a recombinant virus which drives GFP expression in infected cells during the lytic cycle, enabling us to distinguish and sort cells supporting virus replication from cells that remain in latency. We found that, at least in Akata Bx1g cells, GFP expression, driven by the immediate-early CMV promoter from a cassette inserted into the thymidine kinase gene of the EBV genome, largely coincides with expression of early and late lytic cycle genes. The fact that high levels of GFP expression in Akata Bx1g cells are observed only upon induction of the lytic cycle is most likely explained by upregulation of the CMV promoter activity or regulatory effects of the two thymidine kinase promoters which lie upstream of the CMV promoter in the recombinant EBV strain and become activated in cells supporting EBV replication. An increase in EBV copy numbers during the lytic cycle is unlikely to contribute to this phenomenon, because inhibition of EBV DNA synthesis by acyclovir does not significantly affect GFP expression in anti-Ig-stimulated Akata cells. In the present study, we utilized this model to analyze the influence of the EBV lytic cycle on the MHC class I processing and presentation pathway. We observed that EBV replication results in MHC class I downregulation at the cell surface of Burkitt's lymphoma Akata cells, as well as the epithelial cell lines AGS and HCT116, infected with the recombinant EBV strain (Fig. 1). Kinetic analysis of viral gene expression in Akata cells indicated that the MHC class I downregulation precedes the expression of late genes of the EBV lytic cycle. Accordingly, treatment with acyclovir, which blocks late EBV gene expression and virus DNA replication, did not prevent MHC class I downregulation in cells expressing GFP and BZLF1 following anti-Ig cross-linking (Fig. 2). It is not surprising that the recombinant virus retains sensitivity to acyclovir even though its thymidine kinase gene is disrupted, because EBV thymidine kinase does not phosphorylate acyclovir and the effect of this drug is determined by the exquisite sensitivity of the viral DNA polymerase to phosphorylated acyclovir presumably generated by cellular or other viral kinases (12, 26).

MHC class I downregulation in the course of EBV replication observed in our model recapitulates recent findings by Ressing et al., who characterized this phenomenon by using EBV-positive Akata cells transfected with a reporter plasmid in which expression of the extracellular and transmembrane domains of the rat CD2 molecule, with GFP replacing the functional cytosolic domain, is driven by the early BZLF1-responsive promoter (29). Here we extend this finding towards more extensive characterization of the MHC class I-processing pathway during the course of EBV replication. We demonstrate that the expression levels of MHC class I molecules encoded by each of the three loci, A, B, or C, are decreased to similar extents on the surface of cells in the lytic cycle (Fig. 1D). This correlates with significantly decreased levels of MHC class I heavy chains and ß2m in total cell lysates (Fig. 3 and 4). Interestingly, Ressing et al. have not observed any modulation of heavy chain levels in their model of EBV replication (29). This discrepancy can be partly explained by the fact that in the previous study, cells replicating the virus were compared to untreated cells in culture, while we observed a significant MHC class I upregulation on the surface of cells which were subjected to lytic cycle-inducing stimuli, e.g., anti-Ig antibody, but remained in latency (Fig. 1B). Therefore, in our analysis, we always compared cells in the lytic cycle to similarly treated cells in latency obtained by sorting the same cell culture into GFP+ and GFP– populations. In contrast to the changes observed in the total pool of heavy chains, free heavy chains accumulated in cells supporting EBV replication (Fig. 3B). In activated human T cells, MHC class I molecules coexist in two conformation states, folded and misfolded, and can be identified by the monoclonal antibodies W6/32 and HC10, respectively. In these cells, the pool of misfolded heavy chains is found in association with calreticulin and ERp57 (3, 30). Our results point to a possible role of both calreticulin and calnexin in stabilizing different pools of free heavy chains. Although calreticulin and, to a lesser extent, calnexin accumulate on the surface of cells supporting the virus lytic cycle, the levels of calreticulin in total cell lysates are reduced, whereas calnexin is increased (Fig. 4). This suggests that a proportion of free heavy chains could be stabilized by calnexin intracellularly, whereas calreticulin aids their stabilization at the surface. It is also possible that some free heavy chains associated with chaperones and especially with calreticulin acquire a folded conformation, which is seen by the W6/32 antibody. That could explain why ß2m seems to be downregulated to a greater extent than assembled MHC class I molecules in cells replicating the virus (Fig. 3B).

Although it has been suggested that free heavy chains and MHC class I molecules devoid of peptides can interact with NK cell receptors and modulate the functional outcome of target cell recognition by NK cells (2, 7, 22), it remains to be established whether the accumulation of free MHC class I heavy chains at the cell surface serves any immunological function in cells replicating EBV.

Calreticulin is found in the ER membrane, the nucleus, and on the cell surface of most mammalian cells. Cellular stress, including apoptosis, induces expression of calreticulin and leads to an increase of its cell surface expression and redistribution, whereby calreticulin becomes associated with phosphatidylserine (11). At least in part, the surface accumulation of calreticulin could be a reflection of apoptotic changes associated with EBV replication but could also have different functional implications. Calreticulin has been shown to interact with perforin and inhibit perforin-mediated killing. This protection seems to occur independently of direct interaction with perforin, and it has been proposed that calreticulin stabilizes the membrane and prevents pore formation induced by perforin (9). Although MHC class I downregulation may lead to increased sensitivity of cells to NK lysis (20), functional consequences of EBV replication in terms of its effect on recognition by CTLs or NK cells remain to be characterized, and calreticulin could affect perforin-mediated lysis of EBV-infected cells during the lytic cycle.

Pulse-chase analysis of MHC class I molecules in cells replicating EBV showed a strong reduction in the levels of newly made heavy chains and ß2m (Fig. 5). Targeting proteins for degradation is a common strategy utilized by herpesviruses to ensure their persistence and/or escape from immune surveillance (see reference 27 for a review). In particular, cytomegalovirus encodes at least two proteins, US2 and US11, which target MHC class I heavy chains for retrograde transport from the ER to cytosol (21, 35). A decrease in the amount of newly made heavy chains was clearly detected after brief metabolic labeling and could not be rescued by a combination of proteasomal and lysosomal inhibitors, strongly suggesting that targeting of heavy chains for protein degradation does not play a role in the observed MHC class I downregulation (Fig. 5C and D). Ressing et al. have shown that the functional activity of the TAP heterodimer is inhibited during EBV replication, although the expression of TAP1 and TAP2 appeared to be unaffected (29). In our system, expression of TAP2, but not TAP1, was decreased by about 30% in cells supporting virus replication (Fig. 4). We also observed downregulation of ERp57, another functionally important member of the peptide loading complex, and TPPII, an endopeptidase implicated in the generation of peptide ligands presented by MHC class I molecules. These changes could, indeed, result in a lower efficiency of MHC class I assembly, which is usually reflected in delayed class I maturation, as assessed by changes in the endoH sensitivity of sugars attached to these molecules and modified during their progression from the ER to the Golgi compartment. We detected a delay in MHC class I maturation during the EBV lytic cycle in Akata cells; however, this delay was revealed only upon comparison of samples obtained at 2 h of chase (Fig. 5G), and its significance for the phenomenon of MHC class I downregulation remains unclear. It is conceivable that strong inhibition of MHC class I heavy chain synthesis observed during EBV replication minimizes competition of newly produced heavy chains for available peptide ligands, even if the latter are supplied at a decreased rate due to inhibition of the TAP heterodimer. Whether inhibition of TAP function plays a more significant role at a particular stage of the virus lytic cycle remains to be determined.

Nonspecific inhibition of protein synthesis by cycloheximide causes downregulation of MHC class I comparable to that observed in Akata cells during the EBV lytic cycle (Fig. 5F). The notion that the inhibition of MHC class I heavy chain and ß2m synthesis plays the primary role in MHC class I downregulation in the course of EBV replication is consistent with a recent demonstration of skewing of EBV lytic cycle-specific CTL responses towards the immediate-early and early proteins (28). In this scenario, the immediate-early and early proteins could be still processed and presented before the virus-induced inhibition of protein synthesis comes into force and the existing pool of nonassembled heavy chains and ß2m gets depleted. A detailed discussion of mechanisms which could account for the inhibition of MHC class I synthesis by the immediate-early or early proteins of the EBV lytic cycle is outside the scope of this article. It has been proposed that the activity of LMP1, which induces the IRF7 transcription factor and upregulates MHC class I expression (37, 38), can be counteracted by BZLF1 through its interference with IRF7 (14). MHC class I upregulation induced by B-cell-receptor triggering involves NF-{kappa}B and can also be counteracted by BZLF1 through its interference with the NF-{kappa}B p65 subunit (13, 25). However, these mechanisms do not seem to be sufficient, because the level of class I synthesis in cells replicating EBV is much lower than that in noninduced control Akata cells, which have not received an activating signal and do not express LMP1. Other mechanisms may be at play in inhibiting basic transcription of MHC class I molecules and potentially include BZLF1-dependent inhibition of transcription factors such as CREB and TBP (1, 23), as well as inhibition of nuclear export of spliced cellular mRNAs by BMLF1 (4, 33). Regardless of the specific molecular event(s) involved in this phenomenon, our data strongly suggest that virus-mediated inhibition of protein synthesis through reduction of mRNA expression is the major mechanism of MHC class I downregulation during the EBV lytic cycle.


arrow
ACKNOWLEDGMENTS
 
This study was supported by grants awarded to V.L. by the Swedish Cancer Society, Swedish Children Cancer Foundation, and IRIS Center of the Swedish Foundation for Strategic Research.

We thank Lindsey Hutt-Fletcher (Louisiana State University, Shreveport, LA) for providing us with the recombinant EBV strain and EBV-infected cell lines.


arrow
FOOTNOTES
 
* Corresponding author. Present address: Nobels väg 16, 17177 Stockholm, Sweden. Phone: 46-8-52486981. Fax: 46-8-331339. E-mail: Victor.Levitsky{at}ki.se. Back

{triangledown} Published ahead of print on 15 November 2006. Back

{dagger} Present address: Division of Biomedical Sciences, Johns Hopkins in Singapore, 31 Biopolis Way #02-01, Nanos Building, Singapore 138669. Back


arrow
REFERENCES
 
    1
  1. Adamson, A. L., and S. Kenney. 1999. The Epstein-Barr virus BZLF1 protein interacts physically and functionally with the histone acetylase CREB-binding protein. J. Virol. 73:6551-6558.[Abstract/Free Full Text]
    2. 2
  2. Allen, R. L., T. Raine, A. Haude, J. Trowsdale, and M. J. Wilson. 2001. Leukocyte receptor complex-encoded immunomodulatory receptors show differing specificity for alternative HLA-B27 structures. J. Immunol. 167:5543-5547.[Abstract/Free Full Text]
    3. 3
  3. Arosa, F. A., O. de Jesus, G. Porto, A. M. Carmo, and M. de Sousa. 1999. Calreticulin is expressed on the cell surface of activated human peripheral blood T lymphocytes in association with major histocompatibility complex class I molecules. J. Biol. Chem. 274:16917-16922.[Abstract/Free Full Text]
    4. 4
  4. Buisson, M., F. Hans, I. Kusters, N. Duran, and A. Sergeant. 1999. The C-terminal region but not the Arg-X-Pro repeat of Epstein-Barr virus protein EB2 is required for its effect on RNA splicing and transport. J. Virol. 73:4090-4100.[Abstract/Free Full Text]
    5. 5
  5. Callan, M. F. 2003. The evolution of antigen-specific CD8+ T cell responses after natural primary infection of humans with Epstein-Barr virus. Viral Immunol. 16:3-16.[CrossRef][Medline]
    6. 6
  6. Callan, M. F. C., N. Annels, N. Steven, L. Tan, J. Wilson, A. J. McMichael, and A. B. Rickinson. 1998. T cell selection during the evolution of CD8+ T cell memory in vivo. Eur. J. Immunol. 28:4382-4390.[CrossRef][Medline]
    7. 7
  7. Carbone, E., G. Stuber, S. Andree, L. Franksson, E. Klein, A. Beretta, A. G. Siccardi, and K. Karre. 1993. Reduced expression of major histocompatibility complex class I free heavy chains and enhanced sensitivity to natural killer cells after incubation of human lymphoid lines with beta 2-microglobulin. Eur. J. Immunol. 23:1752-1756.[Medline]
    8. 8
  8. Cresswell, P., A. L. Ackerman, A. Giodini, D. R. Peaper, and P. A. Wearsch. 2005. Mechanisms of MHC class I-restricted antigen processing and cross-presentation. Immunol. Rev. 207:145-157.[CrossRef][Medline]
    9. 9
  9. Fraser, S. A., R. Karimi, M. Michalak, and D. Hudig. 2000. Perforin lytic activity is controlled by calreticulin. J. Immunol. 164:4150-4155.[Abstract/Free Full Text]
    10. 10
  10. Frisan, T., Q. J. Zhang, J. Levitskaya, M. Coram, M. G. Kurilla, and M. G. Masucci. 1996. Defective presentation of MHC class I-restricted cytotoxic T-cell epitopes in Burkitt's lymphoma cells. Int. J. Cancer 68:251-258.[CrossRef][Medline]
    11. 11
  11. Gardai, S. J., K. A. McPhillips, S. C. Frasch, W. J. Janssen, A. Starefeldt, J. E. Murphy-Ullrich, D. L. Bratton, P. A. Oldenborg, M. Michalak, and P. M. Henson. 2005. Cell-surface calreticulin initiates clearance of viable or apoptotic cells through trans-activation of LRP on the phagocyte. Cell 123:321-334.[CrossRef][Medline]
    12. 12
  12. Gustafson, E. A., A. C. Chillemi, D. R. Sage, and J. D. Fingeroth. 1998. The Epstein-Barr virus thymidine kinase does not phosphorylate ganciclovir or acyclovir and demonstrates a narrow substrate specificity compared to the herpes simplex virus type 1 thymidine kinase. Antimicrob. Agents Chemother. 42:2923-2931.[Abstract/Free Full Text]
    13. 13
  13. Gutsch, D. E., E. A. Holley-Guthrie, Q. Zhang, B. Stein, M. A. Blanar, A. S. Baldwin, and S. C. Kenney. 1994. The bZIP transactivator of Epstein-Barr virus, BZLF1, functionally and physically interacts with the p65 subunit of NF-kappa B. Mol. Cell. Biol. 14:1939-1948.[Abstract/Free Full Text]
    14. 14
  14. Hahn, A. M., L. E. Huye, S. Ning, J. Webster-Cyriaque, and J. S. Pagano. 2005. Interferon regulatory factor 7 is negatively regulated by the Epstein-Barr virus immediate-early gene, BZLF-1. J. Virol. 79:10040-10052.[Abstract/Free Full Text]
    15. 15
  15. Hislop, A. D., N. E. Annels, N. H. Gudgeon, A. M. Leese, and A. B. Rickinson. 2002. Epitope-specific evolution of human CD8(+) T cell responses from primary to persistent phases of Epstein-Barr virus infection. J. Exp. Med. 195:893-905.[Abstract/Free Full Text]
    16. 16
  16. Keating, S., S. Prince, M. Jones, and M. Rowe. 2002. The lytic cycle of Epstein-Barr virus is associated with decreased expression of cell surface major histocompatibility complex class I and class II molecules. J. Virol. 76:8179-8188.[Abstract/Free Full Text]
    17. 17
  17. Khanna, R., C. Leanne, N. Kienzle, D. J. Moss, S. R. Burrows, and K. K. Khanna. 1997. Engagement of CD40 antigen with soluble CD40 ligand up-regulates peptide transporter expression and restores endogenous processing function in Burkitt's lymphoma cells. J. Immunol. 159:5782-5785.[Abstract]
    18. 18
  18. Kieff, E. 1996. Epstein-Barr virus and its replication, p. 2343-2396. In B. N. Fields, D. M. Knipe, and P. M. Howley (ed.), Fields virology, vol. 2. Lippincott-Raven Publishers, Philadelphia, PA.
    19. 19
  19. Lilley, B. N., and H. L. Ploegh. 2005. Viral modulation of antigen presentation: manipulation of cellular targets in the ER and beyond. Immunol. Rev. 207:126-144.[CrossRef][Medline]
    20. 20
  20. Ljunggren, H. G., and K. Karre. 1990. In search of the ‘missing self’: MHC molecules and NK cell recognition. Immunol. Today 11:237-244.[CrossRef][Medline]
    21. 21
  21. Machold, R. P., E. J. Wiertz, T. R. Jones, and H. L. Ploegh. 1997. The HCMV gene products US11 and US2 differ in their ability to attack allelic forms of murine major histocompatibility complex (MHC) class I heavy chains. J. Exp. Med. 185:363-366.[Abstract/Free Full Text]
    22. 22
  22. Mandelboim, O., H. T. Reyburn, M. Vales-Gomez, L. Pazmany, M. Colonna, G. Borsellino, and J. L. Strominger. 1996. Protection from lysis by natural killer cells of group 1 and 2 specificity is mediated by residue 80 in human histocompatibility leukocyte antigen C alleles and also occurs with empty major histocompatibility complex molecules. J. Exp. Med. 184:913-922.[Abstract/Free Full Text]
    23. 23
  23. Mauser, A., S. Saito, E. Appella, C. W. Anderson, W. T. Seaman, and S. Kenney. 2002. The Epstein-Barr virus immediate-early protein BZLF1 regulates p53 function through multiple mechanisms. J. Virol. 76:12503-12512.[Abstract/Free Full Text]
    24. 24
  24. Molesworth, S. J., C. M. Lake, C. M. Borza, S. M. Turk, and L. M. Hutt-Fletcher. 2000. Epstein-Barr virus gH is essential for penetration of B cells but also plays a role in attachment of virus to epithelial cells. J. Virol. 74:6324-6332.[Abstract/Free Full Text]
    25. 25
  25. Morrison, T. E., and S. C. Kenney. 2004. BZLF1, an Epstein-Barr virus immediate-early protein, induces p65 nuclear translocation while inhibiting p65 transcriptional function. Virology 328:219-232.[CrossRef][Medline]
    26. 26
  26. Pagano, J. S., and A. K. Datta. 1982. Perspectives on interactions of acyclovir with Epstein-Barr and other herpes viruses. Am. J. Med. 73:18-26.[CrossRef][Medline]
    27. 27
  27. Petersen, J. L., C. R. Morris, and J. C. Solheim. 2003. Virus evasion of MHC class I molecule presentation. J. Immunol. Methods 171:4473-4478.
    28. 28
  28. Pudney, V. A., A. M. Leese, A. B. Rickinson, and A. D. Hislop. 2005. CD8+ immunodominance among Epstein-Barr virus lytic cycle antigens directly reflects the efficiency of antigen presentation in lytically infected cells. J. Exp. Med. 201:349-360.[Abstract/Free Full Text]
    29. 29
  29. Ressing, M. E., S. E. Keating, D. van Leeuwen, D. Koppers-Lalic, I. Y. Pappworth, E. J. Wiertz, and M. Rowe. 2005. Impaired transporter associated with antigen processing-dependent peptide transport during productive EBV infection. J. Immunol. 174:6829-6838.[Abstract/Free Full Text]
    30. 30
  30. Santos, S. G., S. J. Powis, and F. A. Arosa. 2004. Misfolding of major histocompatibility complex class I molecules in activated T cells allows cis-interactions with receptors and signaling molecules and is associated with tyrosine phosphorylation. J. Biol. Chem. 279:53062-53070.[Abstract/Free Full Text]
    31. 31
  31. Seliger, B., U. Ritz, and S. Ferrone. 2006. Molecular mechanisms of HLA class I antigen abnormalities following viral infection and transformation. Int. J. Cancer 118:129-138.[CrossRef][Medline]
    32. 32
  32. Stam, N. J., H. Spits, and H. L. Ploegh. 1986. Monoclonal antibodies raised against denatured HLA-B locus heavy chains permit biochemical characterization of certain HLA-C locus products. J. Immunol. 137:2299-2306.[Abstract]
    33. 33
  33. Swaminathan, S. 2005. Post-transcriptional gene regulation by gamma herpesviruses. J. Cell Biochem. 95:698-711.[CrossRef][Medline]
    34. 34
  34. Wang, X., and L. M. Hutt-Fletcher. 1998. Epstein-Barr virus lacking glycoprotein gp42 can bind to B cells but is not able to infect. J. Virol. 72:158-163.[Abstract/Free Full Text]
    35. 35
  35. Wiertz, E. J. H., T. R. Jones, L. Sun, M. Bogyo, H. J. Geuze, and H. L. Ploegh. 1996. The human cytomegalovirus US11 gene product dislocates MHC class I heavy chains from the endoplasmic reticulum to the cytosol. Cell 84:769-779.[CrossRef][Medline]
    36. 36
  36. Yewdell, J. W., and A. B. Hill. 2002. Viral interference with antigen presentation. Nat. Immunol. 3:1019-1025.[CrossRef][Medline]
    37. 37
  37. Zhang, L., and J. S. Pagano. 2000. Interferon regulatory factor 7 is induced by Epstein-Barr virus latent membrane protein 1. J. Virol. 74:1061-1068.[Abstract/Free Full Text]
    38. 38
  38. Zhang, L., and J. S. Pagano. 2001. Interferon regulatory factor 7 mediates activation of Tap-2 by Epstein-Barr virus latent membrane protein 1. J. Virol. 75:341-350.[Abstract/Free Full Text]

Journal of Virology, February 2007, p. 1390-1400, Vol. 81, No. 3
0022-538X/07/$08.00+0     doi:10.1128/JVI.01999-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:

  • Roder, G., Geironson, L., Bressendorff, I., Paulsson, K. (2008). Viral Proteins Interfering with Antigen Presentation Target the Major Histocompatibility Complex Class I Peptide-Loading Complex. J. Virol. 82: 8246-8252 [Full Text]  
  • Zuo, J., Thomas, W., van Leeuwen, D., Middeldorp, J. M., Wiertz, E. J. H. J., Ressing, M. E., Rowe, M. (2008). The DNase of Gammaherpesviruses Impairs Recognition by Virus-Specific CD8+ T Cells through an Additional Host Shutoff Function. J. Virol. 82: 2385-2393 [Abstract] [Full Text]  

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Guerreiro-Cacais, A. O.
Right arrow Articles by Levitsky, V.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Guerreiro-Cacais, A. O.
Right arrow Articles by Levitsky, V.

Copyright © 2009 by the American Society for Microbiology.   For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»