Aflatoxin B1 and Epstein-Barr virus-induced CCL22 expression stimulates B cell infection

Abstract

Epstein-Barr Virus (EBV) infects more than 90% of the adult population worldwide. EBV infection is associated with Burkitt lymphoma (BL) though alone is not sufficient to induce carcinogenesis implying the involvement of co-factors. BL is endemic in African regions faced with mycotoxins exposure. Exposure to mycotoxins and oncogenic viruses has been shown to increase cancer risks partly through the deregulation of the immune response. A recent transcriptome profiling of B cells exposed to aflatoxin B1 (AFB1) revealed an upregulation of the Chemokine ligand 22 (CCL22) expression although the underlying mechanisms were not investigated. Here, we tested whether mycotoxins and EBV exposure may together contribute to endemic BL (eBL) carcinogenesis via immunomodulatory mechanisms involving CCL22. Our results revealed that B cells exposure to AFB1 and EBV synergistically stimulated CCL22 secretion via the activation of Nuclear Factor-kappa B pathway. By expressing EBV latent genes in B cells, we revealed that elevated levels of CCL22 result not only from the expression of the latent membrane protein LMP1 as previously reported but also from the expression of other viral latent genes. Importantly, CCL22 overexpression resulting from AFB1-exposure in vitro increased EBV infection through the activation of phosphoinositide-3-kinase pathway. Moreover, inhibiting CCL22 in vitro and in humanized mice in vivo limited EBV infection and decreased viral genes expression, supporting the notion that CCL22 overexpression plays an important role in B cell infection. These findings unravel new mechanisms that may underpin eBL development and identify novel pathways that can be targeted in drug development.

Keywords: CCL22; Epstein–Barr virus; carcinogenesis; endemic Burkitt lymphoma; mycotoxins.

Fig. 1.

Mycotoxins and EBV exposure increase CCL22 expression. (A) CCL22 relative mRNA expression levels in Louckes and primary B cells exposed to aflatoxin B1 (AFB1) 50 µM and B2 (AFB2) 50 µM. (B) CCL22 protein levels detected in presence and absence of aflatoxin exposure (50 µM) by Western Blot using total cell extract (Left) and immunoprecipitated (IP) CCL22 from cell supernatant (Right). (C and D) RT-qPCR quantification of CCL22 mRNA levels in cells exposed to AFB1 (50 µM), aflatoxicol (AFL) at 25 µM, sterigmatocystin (STC) at 3.13 µM and combination of both (AFL+STC) for 48 h only or infected by EBV 24 h after treatment. (E) The RT-qPCR analyses of CCR4 mRNA levels in Louckes and primary B cells exposed to AFB1 (50 µM) and EBV. DMSO was used as solvent control in all experiments. The histograms represent data mean ± SD and significance level calculated by ANOVA test (P *≤ 0.05, or P** ≤ 0.01, or P*** ≤ 0.001 or P**** ≤ 0.0001), n = 3. (F) Representative images of immunohistochemistry staining for CCL22 performed on EBV-positive (n = 8) and EBV-negative (n = 6) Burkitt Lymphomas.

Fig. 2.

AFB1 and EBV stimulate CCL22 via the activation of NF-kB pathway. (A) RT-qPCR analysis of CCL22 mRNA expression levels in Louckes cells exposed to NF-kB inhibitor (Bay11; used at 1 µM or 10 µM) under conditions of AFB1 treatment (50 µM) for 48 h, EBV for 24 h, or a combination of both EBV and AFB1 with nonexposed cells (/) as control. (B) RT-qPCR quantification of RelA/p65 (key subunit of NF-kB) mRNA expression levels in exposed Louckes cells as (A) and transfected with siRelA/p65 or scramble control (siCtrl) (Left). RT-qPCR relative quantification of CCL22 mRNA in exposed cells transfected with siCtrl or siRelA/p65 (Right). Data are mean ± SD (n = 3) and significance level calculated by ANOVA test (P* ≤ 0.05, or P** ≤ 0.01, or P*** ≤ 0.001 or P**** ≤ 0.0001).

Fig. 3.

EBV latent viral proteins play a role in induction of CCL22 expression. (A) RT-PCR products of LMP1 and GAPDH amplification run on agarose gel confirming the efficiency of LMP1 deletion in EBV genome compared to empty vector control (/) used as negative control and EBV-infected Louckes cells as positive control (Left). mRNA quantification of EBNA2 (key EBV viral marker) and CCL22 in Louckes cells infected with EBV or recombinant EBV (Right). (B) CCL22 mRNA quantification by RT-qPCR of Louckes cells transfected with empty vector (EV), vector expressing individual EBV latent genes (LMP1, LMP2A, EBNA1, EBNA2, EBNA3A, EBNA3B, EBNA3C, and EBNALP) or infected with EBV. (C) RT-qPCR quantification of CCL22 mRNA in Louckes cells transfected with either LMP1 expressing vector or with a mixture of vectors expressing all viral latent genes except LMP1 or with all viral latent genes including LMP1 or empty vector (EV) as control. Data are mean ± SD n = 3 and significance level calculated by ANOVA test for (A) and Student’s t test for (B) and (C) (P* ≤ 0.05). ns: no significant difference between LMP1 and All viral latent genes.

Fig. 4.

CCL22 enhances EBV infection through activation of PI3K pathway. (A) qPCR quantification of viral DNA levels in EBV exposed primary B cells treated with CCL22 neutralizing antibody (CCL22 ab) or isotype antibody (isotype). (B) RT-qPCR analysis of CCL22 gene silencing efficiency in EBV infected B cells (Left) and FACS analysis data indicating the percentage of green fluorescence protein EBV-positive B cells (%GFP+ B cells) in siCCL22 (siRNA against CCL22) versus siCtrl (scramble control) treated B cells infected with EBV (Right). (C) qPCR quantification of viral DNA levels in EBV exposed primary B cells treated with human recombinant CCL22 (RhCCL22) or phosphate buffered saline (Pbs). (D) Same as in (B). Mock infected cells were used as controls in all EBV infection experiments. (E) Viral quantification results by FACS in siCCL22 treated cells exposed to AFB1 50 µM (Up) and infected by EBV together with western blot results of CCL22 protein detection in treated cells (Down). (F) Viral DNA level by qPCR in exposed B cells treated with a PI3K inhibitor (Wortmannin) or untreated, exposure of cells was with either EBV, or EBV plus RhCCL22, or EBV plus AFB1. Dimethyl sulfoxide (DMSO) was used as solvent control in aflatoxins exposures. Histograms represent mean ± SD, n = 3, and significance level calculated by Student’s t test (P* ≤ 0.05, or P** ≤ 0.01, or P*** ≤ 0.001 or P**** ≤ 0.0001).

Fig. 5.

Neutralizing CCL22 decreases EBV infection and expression of its latent and lytic genes in humanized mice tissues. (A) Schematic representation of mice intervention plan indicating the three phases composed of the humanization phase, neutralizing antibody administration phase (abCCL22 20 µg treatment per week for 4 wk) and EBV infection phase (0.5 × 10⁵ particle). The mice were divided into four groups (5/6 mice per group) with treatment or infection as: (group-A) mice treated with isotype antibody with no EBV infection (Iso/noEBV), (group-B) mice treated with isotype antibody with EBV infection (Iso+EBV), (group-C) mice treated with CCL22 blocking antibody with no EBV infection (abCCL22/noEBV) and (group-D) mice treated with CCL22 blocking antibody with EBV infection (abCCL22+EBV), (B) Cytokine CCL22 concentration measured using Luminex serology-based assay in plasma samples from the four groups of humanized mice, (C) EBV mean levels measured in mice spleen DNA using Taqman PCR, (D) Hematoxylin and eosin (H&E) staining of mice spleen sections (a–d) and EBV-encoded RNA (EBER) in- situ hybridization (ISH) staining for EBV detection (e–h). Images shown are representative images of animals per group (n = 5 or 6 animals per group). Percentage of EBER-ISH positive cells per section is noted for (e–h), (E–J) Average viral gene (EBNA1 or BZLF1 or BNLF2A or LF3 or BILF1 or BMRF1) mRNA expression measured by qPCR on transcribed RNA from the spleen. Experiments represent n = 5 or 6 mean per group ± SEM and significance level calculated by Student’s t test (P* ≤ 0.05 or P** ≤ 0.01, or P**** ≤ 0.0001). NSG: NOD.Cg-Prkdc^scid Il2rg^tm1Wjl/SzJ mice.