To assess the effect of EBV DNA on the transcriptional profile of mouse PBMCs, RNA-seq was used. Study protocols were approved by the Animal Care and Use Committee (IACUC) at the American University of Beirut (AUB). BALB/c mouse PBMCs were isolated from 4 to 6 week old female mouse blood using Histopaque^® (Sigma-Aldrich, Saint Louis, Missouri) and then were cultured in a 96-well plate whereby each well contained 25 × 10⁴ PBMCs in 250 μL of the RPMI 1640 culture medium (Lonza, Basel, Switzerland) supplemented with 10% FBS (Sigma-Aldrich, Saint Louis, Missouri) and 1% penicillin-streptomycin (Lonza, Basel, Switzerland). Cells were cultured for 12 h at 37°C with 5% CO₂ in the presence or absence of 9 × 10³ copies of EBV DNA. Triplicates of each treatment were performed. Cells were then collected and total RNA was extracted using Qiazol (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. The isolated RNA was subsequently sequenced by Macrogen Inc. (Seoul, Korea) using the NovaSeq 6000 S4 Reagent Kit and an Illumina NovaSeq 6000 System (Illumina Inc., San Diego, CA, United States). Reads were trimmed with Trimmomatic and the trimmed reads were mapped to the Mus musculus mm10 reference genome using HISAT2; featureCounts was then used to assign sequence reads to genomic features and Limma-Voom was subsequently employed for differential expression (DE) analysis.

To visualize and analyse functional association network interactions involving the DEGs, GeneMANIA (Warde-Farley et al., 2010) using Cytoscape v3.8.2 was employed setting the max resultant genes to 20 and max resultant attributes to 10. Gene set enrichment (GSE) was conducted using EGSEA (Alhamdoosh et al., 2017) which combines data from 12 enrichment algorithms allowing analyses across multiple gene set collections. EGSEA v1.10.0 and EGSEAdata v1.10.0 were used with the Wilkinson p-value combining method employed for this analysis.

Four to 6 week-old female BALB/c mice were used to isolate immune cell populations. For isolation of DCs, B cells and T cells, mouse spleens were each digested in 10 mL of a digestion mix consisting of 10 mg/mL collagenase IV (Roche) and 10 mg/mL DNAseI (Sigma-Aldrich, Saint Louis, Missouri) for 1 h at 37°C with gentle shaking. Splenocytes were then passed through Corning^® 70 μm cell strainers and washed with 1X PBS (Sigma-Aldrich, Saint Louis, Missouri). Cells obtained per spleen were then incubated with 5 mL of RBC lysis buffer (Qiagen, Hilden, Germany) for 10 min at room temperature. After two additional washes with 1X PBS, cells were incubated in T25 flasks in complete RPMI for 24 h at 37°C in 5% CO₂. Cells were subsequently washed with 1X PBS and enumerated using a TC20™ Automated Cell Counter (Bio-Rad). To isolate elicited macrophages from peritoneal compartments, mice were each injected with 2.5 mL of 3% Brewer’s Thioglycolate medium (Scharlau, Barcelona, Spain) in the peritoneal cavity. After 4 days, mice were sacrificed and a small incision along the mouse midline was introduced with sterile scissors. The intact peritoneal wall was exposed and about 10 mL of cold sterile 1X PBS per mouse was injected in the cavity. The compartment was massaged for 5 min and the peritoneal fluid was then aspirated and centrifuged at 400 g for 10 min at 4°C. After two washes with 1X PBS, cells were resuspended in DMEM/F12-10 medium (Lonza, Basel, Switzerland) supplemented with 10% FBS (Sigma-Aldrich, Saint Louis, Missouri) and 1% penicillin-streptomycin (Lonza, Basel, Switzerland) and enumerated. A total of 10⁷ cells per well were cultured in 6-well plates in 2 mL of DMEM/F12-10 medium (Lonza, Basel, Switzerland) supplemented with 10% FBS and 1% penicillin-streptomycin (Lonza, Basel, Switzerland) then incubated overnight at 37°C in 5% CO₂.

To isolate T cells, B cells, DCs, and macrophages, Fluorescence Activated Cell Sorting (FACS) was performed using a BD FACSAria™ II-SORP (BD Biosciences, San Jose, CA). For total leukocytes, an anti-mouse CD45 labelled with APC/Cy7 (BioLegend Inc. San Diego, CA) was employed. Then, for total T cells, an anti-mouse CD3 labelled with Alexa Fluor^® 647 (BioLegend Inc. San Diego, CA) was used. For B cells, PE-labelled anti-CD19 (BioLegend Inc. San Diego, CA) was employed and an anti-CD11c labelled with Brilliant Violet 421™ (BioLegend Inc. San Diego, CA) was used to collect pan DCs including plasmacytoid and myeloid DCs. The antibodies used for sorting of macrophages from the peritoneal exudate included anti-mouse CD11b and F4/80 labelled with PE and Brilliant Violet 421™ (BioLegend Inc. San Diego, CA) respectively.

Isolated mouse B cells, DCs and macrophages were cultured each alone or with T cells in 96-well plates. Each APC population was co-cultured with T cells at a ratio of 1:10 which was found to be an optimal ratio in preliminary experiments for DCs. A similar ratio was then used for the other APC types to allow potency comparisons across cell types assessed. Thus, 15,000 of each APC type was cultured with 150,000 T cells per well in a total volume of 250 µL of complete RPMI. Mouse PBMCs (25 × 10⁴ per well) were also examined. Cells were either left untreated or cultured in the presence of 9 × 10³ copies of EBV DNA or 9 × 10³ EBV virions. The viral DNA was isolated from a P3HR-1 strain of EBV, which was also the strain used for testing the effects of the EBV viral particles. S. epidermidis DNA (1.7pg, equivalent to the weight of 9 × 10³ copies of EBV DNA) was used as a non-viral control DNA. Transcriptional analyses were conducted after 12 h of culture while supernatant levels of mediators were assayed 24 h after incubation.

For flow cytometry analysis of IL-17A positive cells, 25 × 10⁴ PBMCs were treated with 9 × 10³ copies of EBV DNA, 1.7 pg of S. epidermidis DNA, or mock treated (with culture medium). After 24 h of incubation cells were stained with anti-mouse CD3 labelled with Alexa Fluor^® 647 (BioLegend Inc. San Diego, CA) and anti-mouse IL-17A labelled with Brilliant Violet 605™ (BioLegend Inc. San Diego, CA) then analyzed by flow cytometry.

Mouse IL-17A, IFNγ and TNF-α ELISA Kits (Abcam, Cambridge, United Kingdom) were used to determine culture supernatant levels of these mediators. Kits were used as per the manufacturer’s instructions.

To determine the relative expression of the IL-17A, IL-21, IL-23a, IFNγ, TNF-α, IL-12b, TLR9, TLR3, and CTLA4 genes, total RNA was isolated using Qiazol (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. cDNA was synthesized employing the QuantiTect Reverse Transcription Kit (Qiagen, Hilden, Germany) and then real time RT-PCR was carried out using iTaq Universal SYBR Green Supermix (Bio-Rad) and employing the Bio-Rad CFX96 Real Time PCR Detection System (Bio-Rad). β-actin expression was used for expression normalization per sample. Previously published primers for IL-21 (Pesce et al., 2006), IL-23a (Teng et al., 2014), IL-12b (Sun et al., 2017), IL-17A, IFNγ, TNF-α, TLR9, TLR3, CTLA4, and β-actin (Andari et al., 2021) were used; primers were obtained from Macrogen Inc.

Statistical analyses for ELISA and real-time PCR were performed using GraphPad Prism. Means were compared using the two-tailed paired Student’s t-test. p-values less than 0.05 were considered statistically significant.

To examine the effect of EBV DNA on the transcriptional profile of mouse peripheral blood mononuclear cells (PBMCs), cells were cultured in the presence or absence of the viral DNA then subjected to RNA extraction followed by RNA-seq. Analysis of differentially-expressed genes (DEGs) revealed a total of nine genes whose expression was significantly altered by the EBV DNA treatment. The expression of eight genes, including IL-17A, IFNγ, IL-21 and TNF-α, was significantly increased in the EBV DNA-treated group and one gene, CTLA4, showed a significant decrease in expression. Figure 1A is a volcano plot representing gene expression comparisons and Figure 1B is a heat map of the DEGs.

Constructing an interaction gene network with GeneMANIA using Cytoscape to predict interactions between DEGs and other genes resulted in a network of 29 genes and 264 interaction edges (Figure 2). Node degrees were calculated to identify hub genes in the network; this analysis indicated that MyD88 followed closely by IFNγ were hub genes with the highest degree of connectivity in this module (Table 1).

To assess the enrichment of gene sets detected to be differentially expressed in mouse PBMCs across human gene sets, Ensemble Gene Set Enrichment Analysis (EGSEA) was conducted (Figure 3) first using the Human MSigDB c5 gene ontology collection. We observed enrichment of gene sets involved mainly in inflammatory pathways such as those involved in positive regulation of IL-17 production, regulation of granulocyte macrophage colony stimulating factor production, positive regulation of myeloid leukocyte differentiation, and positive regulation of tissue remodelling. On the other hand, analysis using the Kyoto Encyclopedia of Genes and Genomes (KEGG) collections revealed enrichment of gene sets involved in proinflammatory diseases such as inflammatory bowel disease, rheumatoid arthritis, and type 1 diabetes mellitus, all of which are diseases associated with EBV infections.

To validate the RNA-seq differential expression signature observed, the expression of all 9 DEGs was assessed in PBMCs treated with EBV DNA (Figure 4). Results paralleled the RNA-seq observations with enhanced expression of IL-17A, IL-21, IL-23a, IFNγ, TNF-α, IL-12b, TLR9 and TLR3 but a decreased expression of CTLA4. Culturing PBMCs with the virus itself also resulted in a similar expression pattern. Treatment with control bacterial DNA from S. epidermidis was able to enhance the expression of IFNγ and IL-12b, albeit not to the extent induced by EBV DNA; the bacterial DNA was not able to increase the expression of the other genes assessed. To examine the effects of another mammalian DNA virus, we assessed those of human adenovirus type 1 subgroup C; this viral DNA was also not able to replicate the expressional signature of EBV DNA. It was able to enhance the expression of IFNγ, TNF-α and IL-12b but to a much lesser extent compared to EBV DNA and it was not able to increase the expression of the other genes. This indicates that this IL-17A-centric signature is not induced by just any type of DNA and is likely more unique to EBV, and possibly to other herpesviruses as well.

Analysis of the proinflammatory network of genes triggered by EBV DNA indicated IL-17A to be one of the most highly upregulated genes with MyD88 seemingly a hub gene in this interaction network. With MyD88 being an adapter protein integral to TLR signalling, the role we observed for endosomal TLRs in the response to EBV DNA and the types of cytokines observed to be upregulated, T cell stimulation by innate immune responses to the viral DNA was a likely scenario. Hence, we sought to determine whether professional antigen presenting cells (APCs) mediate such a stimulation. Therefore, we examined the three types of professional APCs, B cells, dendritic cells (DCs) and macrophages; these cells were cultured alone or with T cells in the presence or absence of EBV DNA. This was followed by IL-17A level assessment in culture supernatants.

IL-17A levels were found to be significantly increased by 4.28 folds (p = 0.0001) in supernatants from mouse DCs and T cells co-cultured in the presence of EBV DNA compared to co-cultures of DCs and T cells without EBV DNA (Figure 5A). This co-culture of DCs and T cells with EBV DNA had a level similar to that of PBMCs cultured with EBV DNA; this PBMC control culture showed a 2.46-fold increase (p = 0.0006) in IL-17A levels compared to PBMCs cultured in the absence of the viral DNA. Mouse peritoneal elicited macrophages co-cultured with T cells showed a 3.12-fold increase in IL-17A levels upon treatment with EBV DNA compared to elicited macrophages and T cells co-cultured without EBV DNA (p = 0.0095) (Figure 5B). However, this increased level was 47.5% less than that observed in supernatants from PBMCs cultured in the presence of EBV DNA and hence not significant when compared to it (p = 0.0715). On the other hand, B cells co-cultured with T cells in the presence of EBV DNA resulted in some increase in IL-17A levels compared to a co-culture of these cell populations in the absence of EBV DNA; however, the level was 73% less than that of PBMCs treated with EBV DNA (Figure 5C). This indicates that DCs are the most potent in inducing an IL-17A response from T cells in the presence of EBV DNA.

To examine whether the increased production of IL-17A upon EBV treatment of mouse immune cells was due to enhanced differentiation of T cells into Th17 cells, PBMCs were treated with EBV DNA, S. epidermidis DNA or mock treated with culture medium. Cells were subsequently analyzed by flow cytometry (Figure 6). We observed no marked differences between the different treatments which indicates that there is increased production of IL-17A from the EBV DNA-stimulated cells rather than enhanced differentiation into Th17 cells.

To examine whether EBV particles are capable of triggering IL-17A production from DC and T cell co-cultures similar to the viral DNA, mouse cells were cultured with the viral particles. Subsequently, IL-17A levels were determined in culture supernatants (Figure 7A). We observed a 4.43-fold (p = 0.0017) increase in IL-17A levels from supernatants of DCs and T cells co-cultured in the presence of EBV compared to a co-culture of these cells in the absence of the virus. An increase of 2.42 folds (p = 0.0026) in IL-17A levels was also observed when PBMCs were cultured with the virus. Hence, the IL-17A response to the virus itself largely paralleled that to the viral DNA. On the other hand, control bacterial DNA from S. epidermidis was incapable of enhancing the production of IL-17A from PBMCs or DC-T cell co-cultures.

We also assessed whether other proinflammatory mediators observed to be triggered by EBV DNA were also triggered by the virus itself. Hence, IFNγ (Figure 7B) and TNF-α (Figure 7C) levels were examined in culture supernatants. A 41.43% increase in the level of IFNγ (p = 0.0383) was detected in supernatants from DC and T cell co-cultures in the presence of the virus compared to cells grown in the absence of the virus; PBMCs, on the other hand, showed a 32.92% (p = 0.0024) increase in IL-17A levels when cultured with EBV. EBV DNA was similarly capable of significantly increasing IFNγ from these cell types. Worth noting, is that unlike its inability to enhance production of IL-17A from PBMCs or DC-T cell co-cultures, S. epidermidis DNA was able to significantly increase IFNγ levels from these cells. Levels of TNF-α were also significantly elevated upon culturing PBMCs or DCs and T cells with the virus; similar observations were made in cultures of these cells with EBV DNA but not with S. epidermidis DNA.