Introduction Epstein–Barr virus (EBV) is a human γ-herpes virus that establishes a persistent infection in >90% of the world's population. Like other herpes viruses, EBV has two alternative lifestyles: latent (non-productive) infection, and lytic (productive) replication. Following primary infection, EBV persists within memory B lymphocytes in a latent state for the life of the host. A low level of reactivation into lytic replication allows viral shedding into the saliva and transmission of the virus in vivo [1]. The lifelong persistent infection established by EBV is harmless in almost every host and rarely causes disease, unless the host–virus equilibrium is upset. Thus, viral persistence represents a balance between viral latency, viral replication, and host immune responses. The lytic phase of viral replication can be triggered in vitro by a variety of reagents and stimuli, including halogenated pyrimidine [2], phorbol esters [3], calcium ionophores [4], transforming growth factor β [5], butyrate [6], and triggering of the B cell receptor (BCR) with anti-immunoglobulin (anti-Ig) antibody (Ab) [7]. Less is known about the physiological stimuli that control activation of the virus productive cycle in vivo, although replication seems to occur following plasma cell differentiation [8]. It has been well documented that EBV is causally associated with various malignancies, including endemic Burkitt lymphoma (BL), nasopharyngeal carcinoma, and B cell lymphoma, in immunocompromised hosts [9]. Both EBV infection and intense exposure to Plasmodium falciparum malaria (holoendemic malaria) are recognized cofactors in the pathogenesis of BL, which is the most common paediatric cancer in equatorial Africa, accounting for up to 74% of childhood malignant disorders [10]. Development of BL, a B cell malignancy, is heralded by high Ab titers to replicative antigens indicative of EBV reactivation [11]. Recent reports indicate that the impact of malaria infection on EBV persistence is reflected by an increased viral replication. Children living in malaria-endemic areas have an elevated EBV load [12,13], and acute malaria infection leads to increased levels of circulating EBV that are cleared following anti-malaria treatment [14]. The mechanisms that may lead to viral reactivation during P. falciparum malaria are not well understood. The identification of a polyclonal B cell activator and Ig-binding protein in P. falciparum is of particular relevance in this context. We demonstrated that the cystein-rich inter-domain region 1α (CIDR1α) of the P. falciparum erythrocyte membrane protein 1 (PfEMP1) induces proliferation and activation of B cells, preferentially of the memory subset, where EBV is known to reside [15,16]. To understand the relative contribution of malarial antigens on EBV reactivation, we used the prototype EBV-positive BL cell line Akata as a model to determine whether CIDR1α could induce reactivation of the EBV lytic cycle in latently infected B cells. Furthermore, we analyzed the effect of the CIDR1α on freshly isolated peripheral blood mononuclear cells (PBMCs) from EBV-positive healthy donors and from children with BL living in malaria-endemic areas. The results support the hypothesis that CIDR1α is one of the molecules involved in EBV reactivation during the course of malaria infection. Our data provide new insights into how malaria infection may contribute to BL development. Results P. falciparum–Infected Red Blood Cells and Purified CIDR1α Bind to EBV-Carrying B Cells During the blood stage of P. falciparum malaria, infected red blood cells (iRBCs) express high levels of PfEMP1, reaching their maximum at the trophozoite stage (28–32 h post-invasion). The CIDR1α domain of PfEMP1 (clone FCR3S1.2-var1) has a multi-adhesive phenotype and binds to different cell surface receptors, such as CD36, PECAM-1 (CD31), and immunoglobulins (Igs) [17]. CIDR1α also binds to isolated B cells via an interaction that involves surface Igs [15]. To establish whether iRBCs and the soluble form of CIDR1α interact with EBV-carrying B cells, we used the EBV-positive BL cell line Akata as a model. Akata cells stained with PKH67 (green) were co-incubated with PKH26 (red)-stained uninfected red blood cells (RBCs), or with enriched iRBCs at the trophozoite stage (28 h post-invasion, 75%–80% final parasitemia), at a ratio of 1:2, respectively. RBCs did not bind to Akata (Figure 1A), but co-incubation with iRBCs led to the formation of conjugates that varied in size but frequently involved two to five iRBCs/Akata cell (Figure 1B). A higher magnification of the conjugates showed a polarization of the iRBC, where the parasites were found at the proximity of the membrane's tight junction between the two cell types (Figure 1C). Figure 1 P. falciparum iRBCs and Soluble CIDR1α Bind to the EBV-Positive B Cell Line Akata (A–C) Akata cells stained with PKH67 (green) were co-incubated with RBCs stained with PK26 (red) (A) or with purified iRBCs (B, C) at a ratio of 1:2 (Akata:iRBC). Binding was evaluated by light microscopy (left panel) and immunofluorescence microscopy (right panel). (C) Magnification of the binding between iRBCs and Akatas. Localization of the parasite in the iRBC (indicated by white arrows) was revealed using phase contrast microscopy (left panel). (D–F) Akata cells were incubated with soluble GST (1 μM) (D) or CIDR1α (1 μM) (E) for 1 h at RT and stained with anti-GST Ab and anti-mouse Alexa-488-conjugated Ab. The binding was evaluated by light microscopy (left panel, [D, E]), immunofluorescence microscopy (right panel, [D, E]), and fluorescent-activated cell sorting (FACS) analysis (F). Being that CIDR1α is a domain expressed on iRBCs with multi-adhesive phenotypes [17], we investigated its ability to bind Akata cells. A soluble form of CIDR1α was produced as a glutathione-S-transferase (GST) fusion protein, and the GST protein alone was used as control. Immunofluorescence studies with anti-GST fluorescent Abs demonstrated that CIDR1α, but not the GST control, binds to the membrane of Akata cells (Figure 1D and 1E). Flow cytometry analysis showed a peak shift representing an increased mean fluorescence intensity (MFI) as compared to the MFI values obtained when Akata cells were incubated with GST control protein or with the isotype control Ab (Figure 1F). Thus, both iRBCs and the recombinant CIDR1α domain of PfEMP1 bind to the EBV-carrying B cell line Akata. CIDR1α and iRBC Stimulation Lead to Increased EBV Viral DNA Load in Akata Cells In contrast to the variety of reagents and signals able to induce EBV lytic production in vitro [18], the physiological signals involved in the activation of the virus productive cycle have not been well characterized yet, although plasma cell differentiation seems to represent one such trigger [8]. Anti-Ig treatment, which leads to BCR signalling [19], has served as a relevant in vitro model of reactivation for inducing virus replication in some EBV-carrying BL cell lines, including Akata [7,20]. Because CIDR1α is an Ig-binding protein [17], we analyzed its functional impact on the reactivation of lytic virus production and used the well-characterized Akata cell line model as a read-out system. First, we analyzed whether stimulation of Akata cells with CIDR1α would affect the number of viral DNA copies produced. Cells were cultured with increasing concentrations of CIDR1α, GST (range 0,5–2 μM, corresponding to 25–100 μg/mL), or in medium alone. After 48 h, we quantified the EBV viral DNA copy number in the cultures (cells + supernatant) by monitoring the EBV LMP1 gene, which is present as a single copy in the virus genome. As shown in Figure 2A and 2B, stimulation with CIDR1α increased the viral DNA load in a dose-dependent manner. Cells incubated with GST contained numbers of EBV genomes comparable to that of cells cultured in medium alone. In two out of four independent experiments, there was a statistical significance in relative increase of EBV load between the concentrations of 0,5 μM and 2 μM (p = 0,03). Figure 2 CIDR1α and iRBC Stimulation Lead to Increased Viral Load in Akata Cells (A) Akata cells were cultured in medium alone (referred to as Ag concentration 0) with GST (white circles) or with CIDR1α (black circles) at protein concentrations ranging from 0,5 to 2,5 μM in the presence or absence of zVAD. After 48 h of incubation, the viral genome copy number was determined by real-time PCR. Results from a representative experiment (out of four independent experiments) are expressed as numbers of EBV genome copies per μg of total DNA. (B) Akata cells were cultured with GST or CIDR1α (0–2,5 μM), and with anti-Ig (10 μg/mL). After 48 h of incubation, the viral genome copy number was determined by real-time PCR. Results are expressed as percentage of relative increase in EBV copy number in cultures stimulated with CIDR1α versus cultures stimulated with GST, and represent the mean of four independent experiments ± standard deviation. *, p 2vs0,5 μM = 0,03. (C) Akata cells were cultured with increasing concentrations of crude extracts from iRBCs and RBCs. After 48 h of incubation, the viral genome copy number was determined by real-time PCR. Results from one representative experiment (out of three performed) are expressed as percentage of relative increase in EBV copy numbers in cultures stimulated with iRBCs versus cultures stimulated with RBC extracts. We have recently demonstrated that CIDR1α induces B cell proliferation and protects B cells from apoptosis [16]. It could be then argued that the augmented viral load might simply result from a net increase in the number of cells in the culture. Cell cycle analysis performed by propidium iodide staining after 24 and 48 h of incubation did not reveal any significant change in the proportion of dead ( 95% as revealed by cytofluorimetric (FACS) analysis after staining with the pan–T cell Ab CD3 and the monocyte marker CD14. Parasites. The highly rosetting and auto-agglutinating P. falciparum parasite clone FCR3S1.2 was obtained through cloning by micromanipulation of the mother clone FCR3 [33] and cultured according to standard methods. Parasites were maintained and expanded in RBCs (O Rh+) at 5% haematocrit. iRBCs were cultured in RPMI 1640 medium supplemented with 10% B+ human serum, 0,6% HEPES (GIBCO-BRL), 25 μg/mL gentamycin (GIBCO-BRL), 0,25% sodium bicarbonate (GIBCO-BRL), and 2 mM L-glutamine (GIBCO-BRL). Trophozoites were enriched by magnetic-assisted cell sorting, using a Vario-MACS (Miltenyi Biotec, http://www.miltenyibiotec.com). Synchronous P. falciparum parasite cultures, grown 24–28 h post-invasion, were washed twice in RPMI 1640 and resuspended in 5–10 mL of phosphate-buffered saline (PBS) with 2% bovine serum albumin (BSA). Rosettes were disrupted mechanically by repeated passage through a 0,6-mm-thick injection needle. The material was slowly added to a MACS separation column mounted in the magnet and rinsed with 50 mL of 2% BSA in PBS. The iRBCs were eluted in 50 mL of 2% BSA in PBS after removing the column from the magnet, spun down at 500g, and resuspended in 1 mL of RPMI 1640. Following enrichment, the parasitemia was evaluated by fluorescence microscopy after addition of one drop of acridine orange (10 μg/mL). Parasite extracts. Parasite extracts were prepared as previously described [34]. Briefly, parasites at trophozoite stage (iRBCs, 28 h post-invasion) or RBCs were washed, resuspended in PBS, and sonicated (25w) on ice at short intervals for 2 min. The extracts were then centrifuged at 500g for 10 min at 4 °C and filter-sterilized. After determination of protein concentration by Bradford Assay (Bio-Rad, http://www.bio-rad.com), the extracts were diluted with PBS. Recombinant CIDR1α. The sequence of the CIDR1α domain of the PfEMP1 from clone FCR3S1.2-var1 was optimized for codon adaptation in Escherichia coli, re-synthesized chemically (GeneArt, http://www.geneart.com), cloned into the pGEX4T-1 vector (Amersham Pharmacia, http://www.gelifesciences.com), and expressed in E. coli BL21 CodonPlus-RIL (Stratagene, http://www.stratagene.com) as fusion protein with the GST. Non-recombinant pGEX4T-1 was used to produce GST as control protein. Protein purification was carried out according to an optimized protocol [35], and the recombinant protein was purified on glutathione-sepharose column GSTrap FF (Amersham Pharmacia) according to the manufacturer's instructions. Throughout the paper, we refer to the recombinant CIDR1α-GST fusion protein as CIDR1α, while GST is referred to as control protein. Binding assays. Akata iRBCs: Akata cells and enriched iRBCs (trophozoite stage) were stained with PKH67 (green) and PKH26 (red) (Sigma-Aldrich), respectively, according to manufacturer's instructions. The binding was evaluated by fluorescence microscopy after co-incubation in PBS at room temperature (RT) for 1 h at a 1:2 ratio (Akata:iRBC). Akata CIDR1α: Cells incubated for 1 h at RT in PBS containing GST or CIDR1α (1 μM) were washed twice in PBS and incubated for 30 min at RT with anti-GST Abs (Sigma-Aldrich) diluted 1:500 in PBS. After two washes with PBS, anti-mouse IgG Alexa-488-conjugated Ab (Molecular Probes, http://probes.invitrogen.com) diluted 1:100 in PBS was added for 30 min at 4 °C. Cell binding was analyzed by fluorescence microscopy, and the fluorescence intensity was measured with a FACSCalibur flow cytometer and analyzed with Cell Quest Pro software (Becton Dickinson, http://www.bd.com). Stimulation assays. Twenty-four hours before experiments, Akata and Akata-GFP cells were suspended at a concentration of 106/mL in complete medium and complete medium containing G418 (500 μg/mL), respectively. Fresh PBMCs were washed with PBS and resuspended in complete medium at a concentration of 105 cells per ml. The cells were then seeded in 96-well plates and cultured in medium alone or with increasing concentrations of CIDR1α, GST (range 0–4 μM, corresponding to 0–200 μg/mL), or with anti-Ig (10 μg/mL) (Jackson ImmunoResearch, http://www.jacksonimmuno.com). After 48 h, cells were harvested for analysis. All tests were set up in multiple replicates. EBV DNA detection by real-time PCR. DNA was extracted from cells and supernatants using the QIAamp Blood kit (Qiagen, http://www.qiagen.com) according to manufacturer's instructions and eluted in 50 μL of DEPC-treated water (Ambion, http://www.ambion.com). Purity and DNA concentration were evaluated using a NanoDrop ND-1000 spectrophotometer (http://www.nanodrop.com). The PCR primers and probe used for the quantification of EBV genomes were selected from the LMP-1 gene as previously described [13,14]. The primers used were the EBV-LMP1 forward primer 5′-AAGGTCAAAGAACAAGGCCAAG-3′ and the EBV-LMP1 reverse primer 5′-GCATCGGAGTCGG-3′. The fluorogenic probe (PE Applied Biosystems, http://www.appliedbiosystems.com) was synthesized using a FAM reporter molecule attached to the 5′ end and a TAMRA quench-er linked to the 3′ end (5′-AGGAGCGTGTCCCCGTGGAGG-3′). A standard curve was prepared using serial dilutions of DNA derived from the EBV-positive BL line Namalwa that contains two copies of EBV genome per cell. Detection was performed using an ABI Prism 7700 Sequence detection System (PE Applied Biosystems). Briefly, cycling parameters were 50 °C for 2 min, 95 °C for 10 min, 45 cycles 95 °C for 15 s, and 60 °C for 1 min. The EBV DNA copy number was calculated as mean of triplicates. Cell cycle analysis. Cells were washed twice in ice-cold PBS and resuspended in 100 μL of PBS containing 50 μg propidium iodide/mL and 0.1% (v/v) triton X-100. After incubation at 4 °C for 8 h, cell DNA analysis was performed by flow cytometry (FACSCalibur, Becton Dickinson) using Cell Quest Pro software (Becton Dickinson). p-Values ≤ 0,05 were regarded as statistically significant. Western blot analysis. Expression of BZLF1, the early lytic antigen, was assessed using a specific mouse Ab (DAKO, http://www.dako.com). Total extracts from 106 cells were separated in 10% SDS polyacrylamide gels, blotted onto nitrocellulose filters probed with 1:500 dilution of the DAKO monoclonal Ab for 1 h at RT, and then reacted with horseradish-peroxidase–labelled anti-mouse Abs (Amersham). Immunocomplexes were visualized by enhanced chemiluminescence according to the manufacturer's instructions (Amersham). To control for equal loading, the total protein concentration of each sample was checked using Bradford assay. Statistical analysis. Computations were performed with the Prism 4 package (GraphPad Software, http://www.graphpad.com). For parametric data, differences between groups were analyzed with Student t-test; for nonparametric data, differences between groups were analyzed with the Wilcoxon rank sum test. Results are expressed as mean ± SEM, and unless otherwise stated, considered statistically significant at p < 0,05.