Fig 1.
EBV lytic gene expression is associated with advanced NPC and immunosuppression.
(A) Heatmap illustrating relative EBV gene expression profiles in NPC samples from different TNM stages. Unsupervised clustering of genes (x-axis) and nasopharyngeal carcinoma samples (y-axis) was performed by complete-linkage clustering. (B) Heatmap showing genes differentially expressed between the early-stage NPCs samples and the late-stage NPCs samples. RNA-seq reads were aligned to hg19/GRCh37 human reference genome and visualized on standard scaling. (C)The down-regulated genes in advanced NPC were analyzed by DAVID-KEGG. Top 10 enriched signaling pathways were displayed (the abscissa is -log10 (P-value); the ordinate is pathway name). (D) Cluster network diagram of the genes that are down-regulated in advanced NPC via Metascape. The darkness of colors represents the enrichment into the pathway or biological process.
Fig 2.
CNE2 cells in EBV abortive lytic phase exhibit enhanced capacity of monocytes chemotaxis.
(A) The switch of EBV phase in EBV+CNE2-Day7 and -Day14 was presented by GFP flow cytometry. (B) ZTA expression levels in EBV+CNE2-Day7 and -Day14 were analyzed by Western blotting. (C) EBER1 expression levels were quantified by Reverse-transcriptional PCR. (D) The number of immune cells recruited by CNE2 in different EBV phase (n = 4). (E) The trend of cytokine mRNA level change was detected by real-time PCR (n = 3). Data are presented as the mean±SEM. * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001, NS, not significant.
Fig 3.
CNE2 cells in EBV abortive lytic phase promote monocyte differentiation toward TAM-like phenotype and away from DCs.
(A) Changes in cell morphology and cell surface markers during DC differentiation. Scale bar, 50 μm. (B) and (C) Supernatants of CNE2 with different EBV life cycle were used to treat DCs. The difference in the yield of DCs was examined by CD14&CD1a expression profile. the proportion of DC-d-Ms was examined by CD14&CD68 expression profile (n = 4). (D) Changes in cell morphology and cell surface markers during Mφs differentiation. Scale bar, 50 μm. (E) Supernatants of CNE2 in different EBV life cycle were used to treat Mφs, the proportion of M1 and M2 subtypes were determined by detecting CD86 and CD163 (n = 4). Data are presented as the mean±SEM. * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001, NS, not significant.
Fig 4.
ZTA exhibits the ability to recruit and induce directed differentiation of monocytes.
(A) The mRNA expression levels of GM-CSF, IL-8 and GRO-α were determined by real-time PCR in CNE2 cells that express the EBV lytic genes as indicated (n = 3). (B) Stable expression of ZTA in CNE2 cells confirmed by Western blot. (C) Cytokine antibody array identified the differences of cytokine secretion between the supernatants of CNE2-ZTA and CNE2-CTR. The spot signal densities were quantified by Image J. (D) The concentrations of GM-CSF, IL-8 and GRO-α in the supernatants of each CNE2-derived cell line were measured by ELISA (n = 3). (E) The number of immune cells recruited by CNE2-ZTA and CNE2-CTR (n = 4). (F) and (G) After treatment with the supernatants of CNE2-ZTA and CNE2-CTR, the difference in the yield of DCs was examined by CD14 and CD1a expression profile, the proportion of DC-d-Ms was examined by CD14 and CD68 (n = 4). (H) The proportion of M1 and M2 subtypes were determined by detecting CD86&CD163 (n = 4). (I) Neutralizing antibodies were used to block the recruitment of monocytes to CNE2-ZTA. The proportion of CD14+ cells was determined by flow cytometry (n = 4). (J) Neutralizing antibodies were used to block the ability of CNE2-ZTA to induce TAMs. The proportion of CD163+ cells was determined by flow cytometry (n = 4). Data are presented as the mean±SEM; * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001, NS, not significant.
Fig 5.
ZTA Knockout EBV lost the capacity of recruiting and polarizing monocyte.
(A) and (B) The CRISPR/Cas9-mediated knockout efficiency of ZTA was estimated by measuring ZTA mRNA expression (n = 3), as well as ZTA protein levels. (C) EBER1 expression levels in the knockout and control cells were quantified by RT-PCR. (D) The GFP profiles of EBV+CNE2 after ZTA deletion were detected by flow cytometry. (E) and (F) The mRNA expression levels of GM-CSF, IL-8, GRO-α in ZTA knockout and control EBV+CNE2 cells were detected by real-time PCR (n = 3), the concentrations of GM-CSF, IL-8 and GRO-α in the supernatants were measured by ELISA (n = 3). (G) The number of immune cells recruited by CNE2-EBV+SgCTR and CNE2-EBV+SgZTA (n = 4). (H) and (I) After treatment with the supernatants of CNE2-EBV+SgCTR and CNE2-EBV+SgZTA, the yield of DCs was examined by CD14 and CD1a expression profile. The proportion of DC-d-Ms was examined by CD14 and CD68 (n = 4). (J) M1 and M2 proportions were identified by CD86 and CD163 expression (n = 4). Data are presented as the mean±SEM. * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001, NS, not significant.
Fig 6.
EBV activation, TAMs infiltration and angiogenesis in advanced NPC samples.
(A) Clinical NPC samples were analyzed by hematoxylin and eosin (H&E) staining (100X 400X). (B) and (C) Representative images of tumor showing CD1a+ DCs and CD163+ TAMs. Cells were immunostained with the corresponding antibodies (400X). (D) Representative images of tumor showing PAS (Purple) expression and CD34+ cells (Brown) (400X). Black arrows indicate classical angiogenesis (PAS+/CD34+). (E) EBV-DNA titer in plasma from circulating blood, obtained from the hospital information system. (F) The positive area of each stain was quantified using ImageJ in randomly chosen 400X fields (n = 3 per sample).Data are presented as the mean±SEM; * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001, NS, not significant.
Fig 7.
TAMs induced by EBV abortive lytic cycle promote NPC angiogenesis and invasion.
(A) Effect of macrophage supernatants from different groups on angiogenesis of CNE2 analyzed by Matrigel tube assay in vitro (n = 3). (B) Effect of macrophage supernatants from different groups on migration and invasion abilities of CNE2 cells assayed by Transwell migration assay and Transwell-Matrigel invasion assay (n = 3 per sample). (C) to (E) Tumors xenografts in the abdominal cavity of mice were analyzed by hematoxylin and eosin (H&E) staining (100X) and vascular immunohistochemical staining. Representative images of tumor showing PAS (Purple) expression and CD34+cells (Brown) (400X) were presented. Red arrows indicate VM channels (PAS+/CD34-), black arrows indicate classical angiogenesis (PAS+/CD34+). Quantification of classical angiogenesis determined by microscopy with 400X magnification in randomly chosen fields (n = 3 per sample). Data are presented as the mean±SEM; * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001, NS, not significant.
Fig 8.
Schematic model for the role of EBV-abortive lytic cycle in modifying monocytes to establish the tumor-promoting microenvironment that promotes NPC late-stage progression.