Human B Cells Immortalized with Epstein-Barr Virus Upregulate CCR6 and CCR10 and Downregulate CXCR4 and CXCR5

Abstract

Compared to peripheral blood resting B cells, Epstein-Barr virus (EBV)-immortalized B cells consistently express CCR6 and CCR10 at high levels and CXCR4 and CXCR5 at low levels. Accordingly, these cells vigorously responded to the ligands of CCR6 and CCR10 but not to those of CXCR4 and CXCR5. In a human EBV-negative B-cell line, BJAB, stable expression of EBNA2 upregulated CCR6, while stable expression of EBNA2 as well as LMP1 downregulated CXCR4. On the other hand, upregulation of CCR10 or downregulation of CXCR5 was not induced in BJAB by stable expression of EBNA2 or LMP1. Thus, these changes may be due to a plasmablast-like stage of B-cell differentiation fixed by EBV immortalization. EBV-infected B cells in infectious mononucleosis are known to avoid germinal centers and accumulate under the mucosal surfaces. EBV-associated opportunistic lymphomas also tend to occur in extranodal sites. These preferred sites of in vivo localization are consistent with the unique profile of chemokine receptor expression exhibited by EBV-immortalized B cells.

Epstein-Barr virus (EBV) is a ubiquitous human B-lymphotropic herpesvirus capable of efficiently immortalizing primary B cells into continuously growing lymphoblastoid cells in vitro. Primary infection of adolescents and young adults by this virus often leads to infectious mononucleosis (IM). Following primary infection, EBV establishes a lifelong latent infection in B cells. EBV is also associated with a variety of human malignancies such as Burkitt's lymphoma (BL) in central Africa, nasopharyngeal carcinoma in southern China and other foci around the world, opportunistic B-cell lymphomas in AIDS and immunosuppressed patients, and a substantial fraction of Hodgkin's disease (for a review, see reference 14). Recent studies have shown that migration and tissue microenvironmental localization of various lymphocyte classes and subsets are finely regulated through expression of a specific set of chemokine receptors in accordance with their differentiation pathways and maturation stages (19). It is thus conceivable that chemokines and their receptors play important roles in migration and tissue localization of EBV-infected B cells in conditions such as IM and opportunistic B-cell lymphomas. However, expression of chemokine receptors on EBV-immortalized B cells and EBV-associated tumor cells has not been studied in detail yet. Here, we systematically examined expression of chemokine receptors in EBV-immortalized B cells in comparison with peripheral blood resting B cells and BL-derived B-cell lines.

By using semiquantitative reverse transcription-PCR (RT-PCR), we examined expression of all known 18 chemokine receptors (CCR1 to CCR10, CXCR1 to CXCR6, XCR1, and CX3CR1) in fresh peripheral blood CD19⁺ B cells, EBV-immortalized B cells, and BL-derived B-cell lines. BCL-NU, BCL-SH, BCL-SM, and BCL-TOS were EBV-immortalized polyclonal human B-cell lines established from peripheral blood mononuclear cells from healthy adult donors by infection with the B95-8 strain of EBV as described previously (18). AKATA, Daudi, Jijoye, Raji, and AG876 were EBV-positive BL lines, while Ramos was an EBV-negative BL line. BJAB was an EBV-negative non-BL-type human B-cell line. As shown in Fig. 1, peripheral blood resting B cells expressed CXCR4, CXCR5, and CCR7 at high levels and CCR6 at low levels. Primary B cells also expressed CXCR2, CXCR3, and CCR3 at low levels (data not shown). Most BL-derived cell lines expressed only CXCR4 and CXCR5 at high levels, consistent with their germinal center cell origin (3). In contrast, EBV-immortalized B cells expressed CXCR4 and CXCR5 at low levels. Furthermore, EBV-immortalized B-cell lines consistently expressed CCR6, CCR7, and CCR10 at high levels. We also examined surface expression of various chemokine receptors on EBV-immortalized B-cell lines and BL-derived cell lines by flow cytometry using available specific monoclonal antibodies (MAbs). Figure 2 shows the representative results. EBV-immortalized B-cell lines indeed expressed CCR6 at high levels and CXCR4 and CXCR5 at low levels. On the other hand, most BL-derived cell lines expressed CXCR4 and CXCR5 at high levels.

FIG. 1.

RT-PCR analysis for expression of various chemokine receptors. CD19⁺ B cells were purified from peripheral blood mononuclear cells by positive selection with magnetically activated cell sorting (Miltenyi Biotech, Bergisch Gladbach, Germany) after labeling with anti-CD19 microbeads. The purity of CD19⁺ B cells as examined by flow cytometry after staining with fluorescein isothiocyanate-labeled anti-CD19⁺ was >95%. By using Trizol reagent (GIBCO-BRL, Gaithersburg, Md.), total RNA was prepared from peripheral blood mononuclear cells treated with phytohemagglutinin (PHA) for 3 days, fresh peripheral blood CD19⁺ B cells, EBV-immortalized polyclonal B-cell lines (BCL-NU, BCL-SH, BCL-SM, and BCL-TOS), EBV-positive BL cell lines (AKATA, Daudi, Jijoye, Raji, and AG876), an EBV-negative BL cell line (Ramos), and a human non-BL B-cell line (BJAB). RNA was further purified using RNeasy (Qiagen, Hilden, Germany). Total RNA (1 μg) was reverse transcribed using oligo(dT)₁₈ primer and SuperScript II reverse transcriptase (GIBCO-BRL). Resulting first-strand DNA (20 ng of total RNA equivalent) and original total RNA (20 ng) were amplified in a final volume of 20 μl containing 10 pmol of each primer and 1 U of Ex-Taq polymerase (Takara Shuzo, Kyoto, Japan). Amplification conditions were denaturation at 94°C for 30 s (5 min for the first cycle), annealing at 60°C for 30 s, and extension at 72°C for 30 s (5 min for the last cycle) for 33 cycles for all chemokine receptors and 27 cycles for glyceraldehyde-3-phosphate dehydrogenase (G3PDH). Amplification products (10 μl each) were separated by electrophoresis on 2% agarose and stained with ethidium bromide. The primers used were as follows: +5′-GGCTGCTGGGGACTGTCTATGAAT-3′ and −5′-GCCCGGCCGATGTTGTTG-3′ for CXCR1, +5′-CCGCCCCATGTGAACCAGAA-3′ and −5′-AGGGCCAGGAGCAAGGACAGAC-3′ for CXCR2, +5′-CAACGCCACCCACTGCCAATACAA-3′ and −5′-CAGGCGCAAGAGCAGCATCCACA-3′ for CXCR3, +5′-ATCTTCCTGCCCACCATCTACTCCATCATC-3′ and −5′-ATCCAGACGCCAACATAGACCACCTTTTCA-3′ for CXCR4, +5′-AACTACCCGCTAACGCTGGAAATGGAC-3′ and −5′-CACGGCAAAGGGCAAGATGAAGACC-3′ for CXCR5, +5′-ATGGCAATGTCTTTAATCTCGACAA-3′ and −5′-TGAAAGCTGGTCATGGCATAGTATT-3′ for CXCR6, +5′-CAACTCCGTGCCAGAAGGTGAA-3′ and −5′-GCCAGGGCCCAAATGATGAT-3′ for CCR1, +5′-CCAACGAGAGCGGTGAAGAAGTC-3′ and −5′-TCCGCCAAAATAACCGATGTGAT-3′ for CCR2, +5′-GAGCCCGGACTGTCACTTTTG-3′ and −5′-CAGATGCTTGCTCCGCTCACAG-3′ for CCR3, +5′-AAGAAGAACAAGGCGGTGAAGATG-3′ and −5′-AGGCCCCTGCAGGTTTTGAAG-3′ for CCR4, +5′-CTGGCCATCTCTGACCTGTTTTTC-3′ and −5′-CAGCCCTGTGCCTCTTCTTCTCAT-3′ for CCR5, +5′-CCTGGGGAATATTCTGGTGGTGA-3′ and −5′-CATCGCTGCCTTGGGTGTTGTAT-3′ for CCR6, +5′-GTGCCCGCGTCCTTCTCATCAG-3′ and −5′-GGCCAGGACCACCCCATTGTAG-3′ for CCR7, +5′-GGCCCTGTCTGACCTGCTTTTT-3′ and −5′-ATGGCCTTGGTCTTGTTGTGGTT-3′ for CCR8, +5′-CACTGTCCTGACCGTCTTTGTCT-3′ and −5′-CTTCAAGCTTCCCTCTCTCCTTG-3′ for CCR9, +5′-TGCTGGATACTGCCGATCTACTG-3′ and −5′-TCTAGATTCGCAGCCCTAGTTGTC-3′ for CCR10, +5′-TGACCATCCACCGCTACC-3′ and −5′-ATCTGGGTCCGAAACAGC-3′ for XCR1, +5′-TGGCCTTGTCTGATCTGCTGTTTG-3′ and −5′-ATGGCTTTGGCTTTCTTGTGGTTC-3′ for CX3CR1, +5′-GCCAAGGTCATCCATGACAACTTTGG-3′ and −5′-GCCTGCTTCACCACCTTCTTGATGTC-3′ for G3PDH. Representative results from three independent experiments are shown. The lower panel shows the signal intensity ratio between each chemokine and G3PDH.

FIG. 2.

Flow cytometric analysis for surface expression of various chemokine receptors. Cells were washed with phosphate-buffered saline containing 2% fetal calf serum and incubated for 30 min with the following murine MAbs: anti-human CCR1 (clone 141), anti-human CCR2 (clone 48607.121; R&D Systems, Minneapolis, Minn.), anti-human CCR3 (clone 444.11), anti-human CCR4 (KM2160), anti-human CCR5 (clone 2D7; Pharmingen), anti-CCR6 (clone 53103.111; R&D Systems), anti-human CXCR4 (clone 12G5; DAKO, Kyoto, Japan), anti-human CXCR5 (clone 51505.111; DAKO), and anti-CX3CR1 (clone 2A9-1). The mouse primary antibodies were detected by using fluorescein isothiocyanate-conjugated sheep (Fab′)₂ anti-mouse immunoglobulin G (Sigma). Dead cells were stained with propidium iodide to be gated out. Cells were analyzed on a FACSCalibur (Becton Dickinson), and the data were collected in the log mode. Ten thousand cells were analyzed per sample. Results from BCL-NU (an EBV-immortalized B-cell line) and Raji (an EBV-positive BL cell line) are shown as representative.

Collectively, EBV-immortalized B cells consistently expressed CCR6, CCR7, and CCR10 at high levels and CXCR4 and CXCR5 at low levels. To evaluate the functional significance of these observations, we next carried out cell migration and calcium mobilization assays. As shown in Fig. 3A, EBV-immortalized B cells vigorously migrated to LARC/CCL20 (CCR6 ligand), SLC/CCL21 (CCR7 ligand), and MEC/CCL28 (CCR10 ligand). On the other hand, these cells showed only weak migration to SDF-1α/CXCL12 (CXCR4 ligand) and no migration to BLC/CXCL13 (CXCR5 ligand). As shown in Fig. 3B, LARC, SLC, and MEC also induced vigorous calcium mobilization in EBV-immortalized B cells. In contrast, SDF-1α and BLC hardly induced such responses. These results demonstrated the functional significance of the observed expression pattern of chemokine receptors in EBV-immortalized B cells.

FIG. 3.

Cell migration and calcium mobilization assays. (A) The chemotaxis assay was carried out using a CHEMOTX chemotaxis chamber with a 5-μm pore size (Neuro Probe, Gaithersburg, Md.). Recombinant SDF-1α/CXCL12, BLC/CXCL13, eotaxin/CCL11, LARC/CCL20, I-309/CCL1, and MEC/CCL28 were purchased from R&D Systems. Cells were suspended at 10⁷/ml in phenol red-free RPMI 1640 containing 1 mg of bovine serum albumin (Sigma)/ml and 20 mM HEPES, pH 7.4 (chemotaxis assay medium). Cells were applied to the upper wells of the CHEMOTX chemotaxis chamber (25 μl/well). Chemokines in chemotaxis assay medium were applied to lower wells (30 μl/well). After 1.5 h at 37°C, cells that had migrated into the lower wells were lysed with 0.1% Triton X-100 and measured using PicoGreen double-stranded DNA quantitation reagent (Molecular Probes, Eugene, Oreg.). All assays were done in triplicate. Representative results from three separate experiments are shown. Each point represents the mean ± the standard error of the mean. (B) Calcium mobilization assay. Cells were suspended at 10⁶ cells/ml in Hanks' balanced salt solution containing 1 mg of bovine serum albumin (Sigma)/ml and 10 mM HEPES, pH 7.4, and loaded with 3 μM fura2-AM fluorescence dye (Molecular Probes) at room temperature for 1 h in the dark. After being washed twice, cells were resuspended at 5 × 10⁶ cells/ml. Cells in 0.1 ml were centrifuged, resuspended in fresh assay medium, and then applied to an F2000 fluorescence spectrophotometer (Hitachi, Tokyo, Japan). Cells were stimulated with each chemokine at 100 nM. Emission fluorescence at 510 nm was measured upon excitation at 340 and 380 nm with a time resolution of 5 points/s to obtain the fluorescence intensity ratio (R340/380). Representative results from three separate experiments are shown.

Upregulation of CCR6 and CCR10 and downregulation of CXCR4 and CXCR5 in EBV-immortalized B cells could be due to the effects of viral latent genes. To address this possibility, we first carried out semiquantitative RT-PCR to analyze the expression of all 10 EBV latent genes in the human B-cell lines examined for Fig. 1. The results are summarized in Table 1. As expected, four EBV-immortalized B-cell lines (BCL-NU, BCL-SH, BCL-SM, and BCL-TOS) consistently expressed all the known EBV latent genes, consistent with their being at stage 3 of EBV latency (14). On the other hand, the expression profiles of EBV latent genes in EBV-positive BL cell lines were much more restricted and less consistent than in EBV-immortalized B-cell lines, most probably due to most BL cell lines being at stages 1 to 2 of EBV latency (14). EBV-negative B-cell lines were totally negative for expression of EBV latent genes.

TABLE 1.

Summary of expression of EBV latent genes in human B-cell linesa^a

Gene	Expression in cell linesb:
	BCL-NU	BCL-SH	BCL-SM	BCL-TOS	AKATA	Daudi	Jijoye	Raji	AG876	Ramos	BJAB
EBNA1	+	±	±	±	−	±	±	+	±	−	−
EBNA2	++	++	++	++	±	−	−	++	++	−	−
EBNA3A	±	±	±	±	−	−	−	−	−	−	−
EBNA3B	±	±	±	±	−	−	−	−	±	−	−
EBNA3C	±	±	±	±	−	−	−	−	−	−	−
EBNA-LP	±	±	±	±	−	−	−	±	−	−	−
LMP1	++	++	++	++	±	−	+	++	+	−	−
LMP2A	++	++	++	++	+	±	++	−	++	−	−
LMP2B	+	+	±	+	−	−	−	−	−	−	−
BARF0	+	++	++	+	±	±	±	±	++	−	−

^aaPCR primers used were +5′-AGCTTCCCTGGGATGAGCGT-3′ and −5′-TCTTCCCCGTCCTCGTCCAT-3′ for EBNA1, +5′-ATTTTGTCACCTCTGTCACAACC-3′ and −5′-TGGTTCTGGACTATCTGGATCAT-3′ for EBNA2, +5′-TCTTCCATGTTGTCATCCAGGG-3′ and −5′-CTTAGGAAGCGTTTCTTGAGCTT-3′ for EBNA3A, +5′-TTCCATGTTGCAATCGGACC-3′ and −5′-AAAGTGACCTAGCACGACGT-3′ for EBNA3B, +5′-GGGCTGTCAAGCAATCGCAC-3′ and −5′-GTGGTGCATTCCACGGGTAA-3′ for EBNA3C, +5′-CAGCAAGAAGAGGAGGTGGTAAG-3′ and −5′-CTCAAAGTGGTCTCTAATGCGTAG-3′ for EBNA-LP, +5′-TTGGTGTACTCCTACTGATGATCACC-3′ and −5′-AGTAGAATCCAGATACCTAAGACAAGT-3′ for LMP1, +5′-ATGACTCATCTCAACACATA-3′ and −5′-CATGTTAGGCAAATTGCAAA-3′ for LMP2A, +5′-CAGTGTAATCTGCACAAAGA-3′ and −5′-CATGTTAGGCAAATTGCAAA-3′ for LMP2B, and +5′-TGTCCAGCGCTCTGGTCG-3′ and −5′-CCACGGCAACCCTTCCAC-3′ for BARF0. Amplification conditions for the EBV latent genes were 36 cycles of denaturation at 94°C for 30 s (5 min for the first cycle), annealing at 58°C for 40 s, and extension at 72°C for 40 s (5 min for the last cycle). Amplification conditions for glyceraldehyde-3-phosphate dehydrogenase were 27 cycles of denaturation at 94°C for 30 s (5 min for the first cycle), annealing at 60°C for 30 s, and extension at 72°C for 30 s (5 min for the last cycle). Representative results from two independent experiments are shown.

^bThe ratios of signal intensity of the EBV genes to those of glyceraldehyde-3-phosphate dehydrogenase are indicated as follows: −, <0.1; ±, 0.1 to 0.5; +, 0.5 to 1.0; and ++, >1.0.

To directly address the roles of EBV latent genes, we next employed BJAB (an EBV-negative human B-cell line) infected with the B95-8 strain of EBV (BJAB-EBV) and clones of BJAB stably transfected with an expression vector without or with EBNA2 or LMP1 (indicated below as BJAB-vector, BJAB-EBNA2, and BJAB-LMP1; three clones each). We first examined the expression of EBNA2 and LMP1 in these BJAB sublines by RT-PCR. As shown in Fig. 4A, BJAB-EBV expressed EBNA2 and LMP1 transcripts at levels lower than those in EBV-immortalized B cells, while clones of BJAB-EBNA2 and BJAB-LMP1 consistently expressed transcripts of EBNA2 and LMP1, respectively, at levels comparable to those in EBV-immortalized B cells. On the other hand, BJAB clones transfected with the vector alone were totally negative for EBNA2 or LMP1 mRNA. We further carried out immunoblot analysis for EBNA2 and LMP1. As shown in Fig. 4B, BJAB-EBV produced EBNA2 and LMP1 at levels much lower than those in EBV-immortalized B cells, while BJAB-EBNA2 and BJAB-LMP1 clones consistently produced EBNA2 and LMP1, respectively, at levels comparable to those in EBV-immortalized B cells.

FIG. 4.

Induction of CCR6 by EBNA2 and downregulation of CXCR4 by EBNA2 and LMP1 in BJAB transfectants. BJAB, a human B-cell line without EBV, and BJAB infected with the B95-8 strain of EBV (BJAB-EBV) were provided by E. Kieff. Clones were isolated from BJAB cells stably transfected with an expression vector, pZioNeoSV(x)I, carrying either EBNA2 (BJAB-EBNA2) or LMP1 (BJAB-LMP1). For controls, clones of BJAB stably transfected with the vector alone (BJAB-vector) were used. (A and C) Total RNA was prepared from BJAB, BJAB-EBV, BJBA-EBNA2 (three clones), BJAB-LMP1 (three clones), and BJAB-vector (three clones). RT-PCR analysis was carried out for EBNA2 and LMP1 (A) and for the indicated chemokine receptors (C) (for primers, see Fig. 1 legend and Table 1). The lower panel in panel C shows the signal intensity ratio between each chemokine receptor and glyceraldehyde-3-phosphate dehydrogenase (G3PDH). Representative results from two independent experiments are shown. (B) Immunoblot analysis for EBNA2 and LMP1 was carried out with cellular extracts from the indicated cell lines (25 μg of protein each) and MAb anti-EBNA2 (PE2) or anti-LMP1 (CS1-4) (both purchased from DAKO). Bound antibodies were visualized with horseradish peroxidase-conjugated anti-mouse immunoglobulin G (Bio-Rad) and the ECL system (Amersham). Representative results from two independent experiments are shown.

Upon these confirmations, we examined these cell lines for expression of various chemokine receptors by RT-PCR. As shown in Fig. 4C, CCR6 was clearly upregulated in BJAB-EBV and three clones of BJAB-EBNA2. Furthermore, CXCR4 was consistently downregulated in three clones of BJAB-EBNA2 as well as those of BJAB-LMP1. However, CXCR4 expression was scarcely affected in BJAB-EBV. The discrepancy might be due to the low levels of EBNA2 and LMP1 in BJAB-EBV (Fig. 4A and B). On the other hand, CCR10 was not induced in BJAB-EBV, BJAB-EBNA2, or BJAB-LMP1. Similarly, CXCR5 was not downregulated in BJAB-EBV, BJAB-EBNA2, or BJAB-LMP1. We further examined surface expression of CXCR4, CXCR5, and CCR6 on these BJAB sublines by flow cytometry. Representative results are shown in Fig. 5. Surface expression of CCR6 was moderately but clearly upregulated in BJAB-EBV and BJAB-EBNA2. Surface expression of CXCR4 was downregulated in BJAB-EBV (slightly), BJAB-EBNA2 (moderately), and BJAB-LMP1 (clearly). On the other hand, surface expression of CXCR5 on BJAB-EBNA2 or BJAB-LMP1 was not significantly altered from that for the parental BJAB. These results were consistent with those from RT-PCR analysis (Fig. 4C). Collectively, EBNA2 clearly upregulates CCR6, while EBNA2 and LMP1 are both capable of downregulating CXCR4.

FIG. 5.

Flow cytometric analysis for expression of various chemokine receptors on BJAB and its sublines. Flow cytometric analysis was carried out for surface expression of indicated chemokine receptors on BJAB, BJAB-EBV, BJAB-LMP1, and BJAB-EBNA2 (for details, see the legend to Fig. 2). Representative results from two independent experiments are shown.

We have shown for the first time that EBV-immortalized B cells consistently express CCR6 and CCR10 at high levels and CXCR4 and CXCR5 at low levels (Fig. 1 and 2). Furthermore, we have shown that, at least in the cellular background of BJAB, CCR6 is upregulated by EBNA2 whereas CXCR4 is downregulated by both EBNA2 and LMP1 (Fig. 4 and 5). EBNA2 is known to induce cellular genes such as CD21 and CD23 (4). Thus, this list now includes CCR6. At present, we do not know the mechanism of downregulation of CXCR4 by EBNA2 or LMP1. Notably, however, infection of human T cells with human herpesvirus 6 or 7 has also been shown elsewhere to downregulate CXCR4 (17). CCR10 was originally shown to be selectively expressed by skin-seeking memory T cells expressing cutaneous lymphocyte antigen (6). Its expression in normal B cells has not been described so far (16). Thus, the consistent expression of CCR10 in EBV-immortalized B cells is rather striking. Furthermore, CXCR5 is considered to be expressed at high levels in all stages of mature B cells (2). Recently, however, Hargreaves et al. have demonstrated low levels of CXCR5 expression in plasma cells obtained from immunized mice (5). Given no induction of CCR10 or downregulation of CXCR5 in BJAB cells upon EBV infection or by stable expression of EBNA2 or LMP1 (Fig. 4), their levels of expression in EBV-immortalized B cells may be related to their plasmablast-like stage of B-cell differentiation fixed through EBV immortalization (14, 18).

Upon antigenic stimulation and help from T cells and dendritic cells, B cells become activated and may differentiate into short-lived plasma cells producing low-affinity antibodies. Alternatively, activated B cells may participate in the germinal center reaction, in which rapid proliferation is coupled to a process of antigenic receptor mutation and selection called affinity maturation, eventually leading to differentiation into either memory B cells or plasma cells producing high-affinity antibodies (8). Thus, there are two types of plasma cells, one without and the other with the germinal center reaction. EBV-immortalized B cells are known to spontaneously secrete monoclonal immunoglobulins and thus can be regarded to be at the stage of plasmablasts (18). However, B cells immortalized in vitro with EBV naturally have not participated in the germinal center reaction. Thus, they probably represent the particular phenotype of plasmablasts directly differentiated from antigen-stimulated B cells. Therefore, upregulation of CCR10 and downregulation of CXCR5 may reflect the phenotype of plasma cells directly differentiated from stimulated B cells.

Notably, AG876, which showed downregulation of CXCR4 and expression of CCR6 and CCR7 at relatively elevated levels (Fig. 1), expressed EBNA2 and LMP1 at relatively strong levels (Table 1). However, Raji, which also expressed EBNA2 and LMP1 at high levels (Table 1), did not show expression of CCR6 or downregulation of CXCR4 (Fig. 1). Similarly, Daudi, which did not show expression of EBNA2 or LMP1 (Table 1), weakly expressed CCR6 (Fig. 1). We do not know the cause of these discrepancies, but the cellular background and the virus types (type 2 EBV in AG876) (14) may also have roles in the expression of these chemokine receptors.

In IM patients, EBV-infected B cells are known to infiltrate into organs such as the salivary glands, tonsils, and lymph nodes (14). In the lymphoid organs, EBV-infected B cells selectively locate in the interfollicular region but not in the germinal center (11). They also accumulate around the crypts and infiltrate the epithelial layer in tonsils (1). The present results provide a potential explanation for these previous histological observations. SLC/CCL21 and ELC/CCL19, the ligands for CCR7, are mainly produced in the interfollicular region (19), while BLC/CXCL13, the ligand for CXCR5, is produced in the germinal center (19). Thus, the strong expression of CCR7 on EBV-infected B cells may promote their efficient homing into the interfollicular region of the secondary lymphoid organs (19). On the other hand, the continuous expression of CCR7 at high levels together with expression of CXCR5 at low levels on EBV-infected B cells may prevent their efficient migration into the germinal center (13). SDF-1α/CXCL12, the ligand for CXCR4, is expressed within the splenic red pulp and lymph node medullary cords as well as in the bone marrow (5). Thus, downregulation of CXCR4 may also prevent EBV-infected B cells from migrating into these anatomical sites. CCR6 is the receptor for LARC/CCL20, which is preferentially expressed by surface-lining mucosal epithelial cells and epidermal keratinocytes (10, 15, 19). CCR10 is the shared receptor for ILC/CTACK/CCL27 and MEC/CCL28 (6, 12, 16). ILC is expressed by epidermal keratinocytes (7, 9), while MEC is expressed in the selective mucosal epithelial cells such as those in the salivary glands and colon (12). Thus, the strong upregulation of CCR6 and CCR10 in EBV-infected B cells may coordinately promote the migration of these cells toward the mucosal surface of tissues such as tonsils and salivary glands as well as toward the skin in IM patients as well as in healthy carriers (1). EBV-associated opportunistic B-cell lymphomas in AIDS and immunosuppressed patients are also essentially EBV-immortalized B cells. These tumors are typically extranodal and frequently involve organs such as the digestive tissues and skin (14). Thus, their preferred sites of occurrence may also be related to the unique expression profiles of chemokine receptors as seen for EBV-immortalized B cells.