Efficacy of ionizing radiation combined with adenoviral p53 therapy in EBV‐positive nasopharyngeal carcinoma

We have previously shown that adenoviral-mediated wild-type p53 (Ad5CMV-p53) gene therapy, either when used alone (Li et al., 1997) or in combination with ionizing radiation (XRT) (Li et al., 1999), were cytotoxic to two nasopharyngeal carcinoma (NPC) cell lines in vitro. These cell lines, however, have lost their Epstein-Barr virus (EBV). Given the clinical scenario where 80% of human patients with NPC harbour the latent gene products of EBV in the malignant epithelial cells (Liebowitz, 1994; Raab-Traub, 1992), it is necessary to evaluate this therapeutic strategy in a more relevant cancer model. The C666-1 is one of the few human NPC cell lines available that consistently express EBV (Cheung et al., 1999; Hui et al., 1998). Another important reason to evaluate this strategy, is that some latent EBV products, such as EBNA-5 or BZLF1, can physically interact with and perhaps abrogate p53 function (Szekely et al., 1993; Zhang et al., 1994), although the latter is a controversial point (Allday et al., 1995). Therefore, before this gene therapy approach can be explored further, its efficacy in this model has to be evaluated. The C666-1 cells (Cheung et al., 1999; Hui et al., 1998) were maintained in RPMI 1640 supplemented with 10% fetal bovine serum (Wisent Inc). The Akata (Takada et al., 1991) and B95.8 cells (Miller et al., 1972) served as positive control cells for EBV gene products, and MCF-7 cells (Soule et al., 1973) served as negative controls for EBV. The human fetal renal 293 and HeLa cell lines used for the adenoviral studies have been described previously (Li et al., 1997, 1999). The experiments were conducted using the C666-1 cells in an exponential growth phase. After 3 days of growth, the cultures were infected with either Ad5CMV-p53 or Ad5CMV-B-gal for 24 hr. Cells were then irradiated (Gamma-cell 40 Exactor, Nordion International Inc., Canada) using single exposures ranging from 0 to 12 Gy. They were then harvested at different time points for protein or RNA extraction, morphologic analysis for apoptosis, or the MTT assay. Results from the RT-PCR analysis for LMP1, EBNA1 and EBNA2 in the C666-1 cells (Fig. 1) demonstrate that under all experimental conditions, these latent EBV genes were consistently present, regardless of exposure to Ad5CMV-B-gal or Ad5CMV-p53, with or without ionizing radiation (XRT). The p53 gene was directly sequenced from exons 5–9 since >85% of p53 mutations in human malignancies are located within these regions (Gangopadhyay et al., 1997), and we identified a deletion in codon 249 (exon 7). Figure 2 demonstrates that instead of the wild-type AGG sequence, the last "G" is absent, resulting in a premature stop codon (TAA) at 263. RT-PCR analysis of LMP1, EBNA1 and EBNA2 mRNAs in C666-1 cells. The cells were Ad5CMV-p53 or Ad5CMV-B-gal-infected at 10 pfu/cell for 1 hour; followed by XRT (0 or 6 Gy) administered 24 hours post-infection. Total RNA was extracted at 72 hours after XRT, and analyzed by reverse transcription and PCR amplification using the specific respective combinations of oligonucleotide primers. PCR products were electrophoresed on 2% agarose gel, and visualized by ethidium bromide staining. Direct sequencing of the p53 gene from exons 5–9 was conducted after extraction of genomic DNA from the C666-1 cells, and then PCR-amplified using pairs of nested primers. After purification of the PCR products, sequencing was performed using the chain-termination sequencing method; reaction samples were then loaded onto polyacrylamide gels, and visualized using autoradiography. Western blot analysis of the C666-1 cells was conducted after Ad5CMV-B-gal (10 pfu/cell), Ad5CMV-p53 (10 pfu/cell), or XRT (0 or 6 Gy), for p53, p21WAF1/CIP1 and Bcl-2 protein expression. Figure 3 demonstrates that under control conditions, C666-1 cells have low levels of p53 protein expression, with modest levels of Bcl-2 and p21WAF1/CIP1 expression. XRT alone appeared to induce a minimal increase in p53 expression, but p21WAF1/CIP1 expression was induced to a greater extent, particularly at 72 hr post-XRT. Ad5CMV-p53 treatment predictably resulted in progressively higher p53 protein expression, in a time-dependent fashion, along with increasing p21WAF1/CIP1 expression. XRT combined with Ad5CMV-p53 did not cause a detectable increase in p53 protein expression. Bcl-2 expression did not change appreciably under any of these experimental conditions. Western blot analysis of Ad5CMV-p53 mediated gene transfer and expression in C666-1 cells. The cells were Ad5CMV-p53 or AdCMV-B-gal-infected at 10 pfu/cell for 1 hour; followed by XRT (0 or 6 Gy) administered 24 hours post-infection. The cells were harvested at 72 hours after XRT. For each sample, 20 μg of cell lysate was separated on an 8–16% SDS-PAGE, electroblotted onto nitrocellulose membrane, and probed with the respective mAb's for p53, Bcl-2, p21WAF1/CIP1, or EAd. The signals were detected using the ECL method. Interestingly, one of the early lytic EBV proteins EAd, which is strongly expressed constitutively in the B95-8 cell line (positive control), appeared to be induced after either XRT or Ad5CMV-p53 exposure. Its expression did not appear until 72 hr after either treatment, and was particularly enhanced with the combination of XRT plus Ad5CMV-p53 therapy. Figure 4 clearly demonstrates decreasing cell viability associated with increasing dose of Ad5CMV-p53. XRT alone (6 Gy) reduced viability to 70%, which was further decreased to 0.4% when combined with Ad5CMV-p53 gene therapy. As we have observed previously (Li et al., 1999), the two curves are not parallel, suggesting a more than additive cytotoxic interaction between the two modalities. Apoptosis was evaluated using morphologic criteria under fluorescent microscopy after the cells were stained with acridine orange-ethidium bromide (Li et al., 1997, 1999). Under control, XRT alone, or Ad5CMV-B-gal infection conditions, ≤ 3% of C666-1 cells displayed apoptosis. However, the combination of Ad5CMV-B-gal infection plus 6 Gy doubled the proportion of apoptotic cells to 6%. Ad5CMV-p53 infection alone further increased the proportion of cells undergoing apoptosis to 20%, which was amplified by almost 3-fold, up to 58%, when combined with XRT. The effect of AD5CMV-p53 ± XRT on cell viability. C666-1 cells were seeded in triplicate in 96-well plates, and then infected with Ad5CMV-p53 at moi's of 0, 2, 10, 25 or 50 pfu/cell for C666-1 cells for 24 hours, and then irradiated at 0 or 6 Gy. Culture medium alone was used as mock infections. Cell viability was analyzed by using MTT assay on Day 6 after irradiation. Each data point represents the mean plus the standard error from 3 separate experiments. This current study demonstrates that despite the expression of EBNA1, EBNA2, and LMP1, Ad5CMV-p53 gene therapy was still effectively cytotoxic, and induced apoptosis in the EBV-positive C666-1 cells. The specific role of EBV in causality or progression of NPC remains controversial. The EBV DNA has been shown to be clonal, suggesting a causative role for EBV in NPC development (Pathmanathan et al., 1995). LMP-1 has also been demonstrated to be transforming (Fahraeus et al., 1990), and oncogenic (Moorthy and Thorley-Lawson, 1993; Wang et al., 1985), which might be mediated through inhibition of apoptosis (Gregory et al., 1991; Okan et al., 1995). However, despite the presumed anti-apoptotic signals that can be mediated by these latent EBV products (Fries et al., 1996; Gregory et al., 1991; Kawanishi, 1997; Okan et al., 1995), we clearly demonstrate that Ad5CMV-p53 gene therapy, by inducing significant expression of exogenous p53, can still effectively cause cytotoxicity, mediated through apoptosis. The up-regulation of p21WAF1/CIP1 expression after either XRT or Ad5CMV-B-gal without induction of p53 differs from our previous observations where the kinetics of p21WAF1/CIP1 followed p53 protein levels relatively faithfully (Li et al., 1997, 1999). This suggests that p21WAF1/CIP1 induction by these "stresses" in the C666-1 cells is occurring independently of p53, which has been reported in other systems (MacLeod et al., 1995). The precise mechanism of this up-regulation however, remains to be elucidated, and whether EBV is playing a role in this process is currently unknown. One of the most novel observations from this study relates to the induction of EAd, an early lytic EBV protein, after XRT, Ad5CMV-B-gal, or Ad5CMV-p53 gene transfer. Sera IgG antibodies to EAd in NPC patients can indicate tumour progression, but its presence has never been directly detected in tumour samples (Rickinson and Kieff, 1996). Our data clearly demonstrate that at one doubling-time after stress exposure (72 hr), this early lytic protein is expressed, with a marked induction after exposure to XRT plus Ad5CMV-p53 (Fig. 3). The mechanism of this induction is unknown, but it is likely mediated through activation of the immediate early EBV genes such as BZLF1 or BRLF1. Teleologically, one can hypothesize that induction of EAd is an EBV-protective mechanism whereby once the host cell is "stressed," the EBV switches to a lytic cycle, preceding its own replication, in order to retain self-preservation (Zhang et al., 1998). In summary, our current work demonstrates that despite the presence of latent gene products of EBV in an NPC cell line, gene transfer therapy using Ad5CMV-p53 is effectively cytotoxic, mediated through apoptosis. Insofar as NPC appears to be a radio-sensitive tumour in human patients, at least 25% of such patients still experience failure at the primary site (Al-Sarraf et al., 1998; Payne et al., 1996), suggesting that such a combination of therapeutic strategies might translate into an improved cure rate for these patients. However, before this therapy can be applied to human patients with NPC, we need to develop effective strategies to address both the delivery and targeting challenges of cancer gene therapy. Recent data using systemic delivery of a transferrin-liposome complex (Xu et al., 1999) appears to be very promising in addressing the delivery issue. We are also in the process of evaluating targeted expression of therapeutic genes by exploiting the presence of EBNA1 in NPC, which hopefully will ultimately allow us to translate these strategies into a clinical benefit for patients with NPC. Jian-Hua Li* , Dolly Huang , Bao-Fu Sun* , Xianghe Zhang* , Jaap Middeldorp§, Henry Klamut* , Fei-Fei Liu Fei-Fei.Liu@rmp.uhn.on.ca* ¶ **

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.