Cytolytic virus activation therapy and treatment monitoring for Epstein-Barr virus associated nasopharyngeal carcinoma in a mouse tumor model

Abstract

Undifferentiated nasopharyngeal carcinoma (NPC) is 100% associated with Epstein-Barr virus (EBV). Expression of viral proteins in the tumor cells is highly restricted. EBV reactivation by CytoLytic Virus Activation (CLVA) therapy triggers de novo expression of early viral kinases (PK and TK) and uses antiviral treatment to kill activated cells. The mechanism of tumor elimination by CLVA was analyzed in NPC mouse model using C666.1 cells. Valproic acid (VPA) was combined with gemcitabine (GCb) to stimulate EBV reactivation, followed by antiviral treatment with ganciclovir (GCV). A single cycle of CLVA treatment resulted in specific tumor cell killing as indicated by reduced tumor volume, loss of EBV-positive cells in situ, and paralleled by decreased EBV DNA levels in circulation, which was more pronounced than treatment with GCb alone. In vivo reactivation was confirmed by presence of lytic gene transcripts and proteins in tumors 6 days after GCb/VPA treatment. Virus reactivation was visualized by [¹²⁴ I]-FIAU accumulation in tumors using PET-scan. This studied showed that CLVA therapy is a potent EBV-specific targeting approach for killing tumor cells. The [¹²⁴ I]-FIAU appears valuable as PET tracer for studies on CLVA drug dosage and kinetics in vivo, and may find clinical application in treatment monitoring.

Keywords: EBV DNA load; Epstein-Barr virus; PET-scan; cytolytic virus activation therapy; nasopharyngeal carcinoma; targeted cancer therapy; treatment monitoring.

Figure 1

(A) Time schedule of drug(s) administration: Gemcitabine (GCb) treatment was started on day 15 (D15) after tumor inoculation (D0) and 4 days later (D19) animals received a 2nd dose of GCb (n = 15). Valproic acid (VPA) was given to GCb‐treated animals (n = 10) from D15 and one group was treated with VPA‐only (n = 5). Ganciclovir (GCV) was administered daily starting on D19 from tumor inoculation in GCb/VPA treated mice (n = 5) and one group was treated only with GCV (n = 5). (B) Scheme of [¹²⁴I]‐FIAU administration before PET‐scan and ex vivo tissue distribution in GCb/VPA treated mice (n = 8); only one dose of GCb was administered (D19)

Figure 2

(A) Tumor volumes represented as average per group (n = 5) of left (full circle) and right (open circle) tumors. In mice treated with GCb, GCb/VPA, and GCb/VPA/GCV tumor volumes decreased in time. Arrows represent administration of GCb, that is, 15 and 19 days after tumor inoculation. Numeric details of mean tumor volumes and standard deviations (SD) are presented in Supplemental Table S1. (B) EBV DNA load in circulation calculated as mean value plus SD (n = 5). Note the continued increase in untreated mice, while in VPA and GCV‐only treated mice a plateau was reached and high levels were measured. Combined/complete treatment reduced the EBV DNA load (starting from day 20) to virtually negative values at end of treatment, and the level in GCb‐only treated mice later raised again in time. Arrows represent administration of GCb, that is, 15 and 19 days after tumor inoculation. Upper bold line (purple) denotes timeframe of VPA administration and the line beneath (green) duration of GCV administration

Figure 3

The kinetics of lytic induction of EBV in C666.1 cells treated with combination of GCb and VPA for 1‐8 days. (A) EBV RNA profiling revealed strong upregulation of all lytic transcripts after 3 days of treatment. Transcripts were calculated as target molecules/cell and normalized to human cellular housekeeping U1A. The fold change was determined by normalization to the level in the untreated cells. (B) Western blot analysis shows a typical example of lytic induction profiling in cultured C666.1 cells detecting Zebra and PK proteins, using cellular β‐actin as a loading control

Figure 4

Lytic transcripts of EBV detected in tumors of 6 (D6) and 9 (D9) days GCb/VPA treated mice (n = 4); the immediate early Zebra, Rta, early PK, TK, and late VCA‐p18 transcripts were calculated as target molecules/cell and normalized to human cellular housekeeping U1A. Fold change was determined by normalization of the average level of each specific mRNA per treated group to the level of that mRNA in the untreated group

Figure 5

Hematoxylin and eosin (HE) staining (A‐C) and EBER in situ hybridization (D‐F) were performed on tumors from control, GCb‐only, and complete CLVA treated mice for 15 days. In untreated tumors, a high density fields of EBER‐positive tumor cells were detected (D), which correspond to the architecture of HE stained tumor (A). GCb‐only treatment reduced the number of tumor cells (E). In the complete CLVA treated tumors only few residual tumor cells remained (F) and phagocytosis by activated macrophages was observed (illustrated by arrowheads). Detection of Zebra‐positive cells in tumor tissues of control and 6, 9, and 15 days GCb/VPA treated mice (G‐J; respectively). CD68 staining was performed to confirm the presence of infiltrated macrophages in 6 (L) and 9 days (M) treated tumors, while the untreated tumors were entirely negative (K). Macrophages in mouse lung alveoli (N) were stained as a positive control. Note the apoptotic cells (indicated by arrow) and tumor cells digested by macrophages (arrowhead). 40× magnification

Figure 6

(A) In vitro assay showing accumulation of [¹²⁴I]‐FIAU in C666.1 cells after 5 days induction by GCb/VPA. (B) PET image 6 days after GCb/VPA treatment. (C) Ex vivo tissue distribution demonstrating tumor specific uptake of [¹²⁴I]‐FIAU after 9 days of GCb/VPA treatment presented as average %ID/g in four animals. Note the high unspecific [¹²⁴I]‐FIAU accumulation in thyroid