Minicircle-oriP-IFNγ: A Novel Targeted Gene Therapeutic System for EBV Positive Human Nasopharyngeal Carcinoma

Background Nonviral vectors are attractively used for gene therapy owing to their distinctive advantages. Our previous study has demonstrated that transfer of human IFNγ gene into nasopharyngeal carcinoma (NPC) by using a novel nonviral vector, minicircle (mc), under the control of cytomegalovirus (CMV) promoter was effective to inhibit tumor growth. However, therapies based on CMV promoter cannot express the targeted genes in cancer tissues. Previous studies indicated that the development of human NPC was closely associated with Epstein-Barr virus (EBV) and demonstrated the transcriptional enhancer function of oriP when bound by EBV protein. Therefore, the present study is to explore the targeted gene expression and the anti-tumor effect of a novel tumor-specific gene therapeutic system (mc-oriP-IFNγ) in which the transgene expression was under the transcriptional regulation of oriP promoter. Methodology/Principal Findings Dual-luciferase reporter assay and ELISA were used to assess the expression of luciferase and IFNγ. WST assay was used to assess the cell proliferation. RT-PCR was used to detect the mRNA level of EBNA1. RNAi was used to knockdown the expression of EBNA1. NPC xenograft models in nude mice were used to investigate the targeted antitumor efficacy of mc-oriP-IFNγ. Immunohistochemistry was used to detect the expression and the activity of the IFNγ in tumor sections. Our results demonstrated that mc-oriP vectors mediated comparable gene expression and anti-proliferative effect in the EBV-positive NPC cell line C666-1 compared to mc-CMV vectors. Furthermore, mc-oriP vectors exhibited much lower killing effects on EBV-negative cell lines compared to mc-CMV vectors. The targeted expression of mc-oriP vectors was inhibited by EBNA1-siRNA in C666-1. This selective expression was corroborated in EBV-positive and -negative tumor models. Conclusions/Significance This study demonstrates the feasibility of mc-oriP-IFNγ as a safe and highly effective targeted gene therapeutic system for the treatment of EBV positive NPC.


Introduction
The goal of cancer treatment is to selectively eliminate malignant cells while leaving normal tissues intact [1]. Therefore, targeted strategies are needed to be implemented for future therapies to ensure efficient activity at the site of patients' primary tumors or metastases without causing intolerable side effects. To this end, cellular mechanisms of gene regulation have been successfully exploited to direct therapeutic gene expression into cancer cells [2,3,4]. Transcriptional targeting is feasible because the tissue-or cancer-specific promoter can be activated in the targeted cancer cells in the presence of proper subset of activators while remaining silent in the non-targeted cells [5].
Nasopharyngeal carcinoma (NPC) is prevalent in South China, North Africa, and among Alaskan Eskimos. A unique feature of NPC is that nearly 100% of anaplastic or poorly differentiated nasopharyngeal carcinomas contain Epstein-Barr virus genomes and express EBV proteins [6], which are expressed exclusively in the malignant tissues but not in the surrounding normal tissues. This difference provides an exploitable opportunity for tumor-specific targeting. Initial genetic dissections of EBV identified one viral protein, Epstein-Barr Nuclear Antigen 1 (EBNA1), and one region of the viral genome, termed latent origin of plasmid replication (oriP), as being necessary and sufficient for replication of the viral plasmid. Previous studies have determined that EBNA1 is essential for regulating the transcription of the transforming genes of EBV [7,8,9]. Additionally, EBNA-1, the only viral protein required for the replication of EBV in latently-infected cells, is found in all EBV-associated malignancies [10]. The oriP is composed of two separable cis elements, the Family of Repeats (FR) and Dyad Symmetry element (DS) [11]. The FR element consists of 20 tandem 30-bp repeats and acts as a transcriptional enhancer for heterologous promoters when it is bound by EBNA1. Based on these features, the oriP-CMV promoter has been exploited for targeted gene therapy in EBV-positive NPC [12].
Our laboratory has investigated the potential effect of minicircle-mediated IFNc gene therapy in human nasopharyngeal carcinoma. Our data indicated that IFNc gene transfer produced an antiproliferative effect on NPC cells in vitro and a profound antitumor effect in vivo [13,14]. We also demonstrated that minicircle-CMV-IFNc was more efficient than corresponding conventional plasmids due to its capability of mediating longlasting, high level of IFNc gene expression [13]. The CMV promoter, however, is ubiquitously expressed without tumortargeting activity; thus the application of this promoter is limited due to the potential side effects caused by unwanted expression of a therapeutic gene in normal tissues [15]. Furthermore, the persistence of transgene expression from the minicircle can be achieved in cells with low turnover rates, such as hepatocytes and skeletal muscle cells [16]. These issues are especially critical when the therapeutic gene is delivered systemically, such as by intravenous injection.
To solve these problems, we have developed a novel minicircle (mc) vector in which transgene expression is under the transcriptional regulation of the oriP-CMV promoter (hereinafter to be referred as oriP promoter). The binding of EBNA1 to the FR domain in oriP region activates the transcription of downstream genes. Selective expression of the therapeutic gene is successfully achieved both in vitro and in vivo, indicating the feasibility of mc-oriP-IFNc as a safe and highly effective gene therapy system for the treatment of NPC. To our knowledge, this is the first report on a non-viral minicircle vector used in targeted gene therapy and is the first time that the oriP promoter has been combined with the minicircle system for NPC-targeted therapy. This strategy lays the foundation for targeted gene therapy of metastatic NPC by intravenous delivery of therapeutic genes.

Selective expression of luciferase in EBV-positive C666-1 cells mediated by oriP-vectors
To determine the transgene expression provided by the novel EBNA1-regulated minicircle vectors in EBV-negative cells (293, NP69, CNE-1 and CNE-2 cells) and the only available EBVpositive NPC cell line (C666-1), the minicircle-luci was compared with its parent plasmid p2WC31-luci and the intermediate plasmid from which p2WC31-luci was derived (pSP72). All these plasmids contained a luciferase expression cassette driven by an oriP promoter or cytomegalovirus promoter ( Figure 1A). Luciferase activity was assessed 48 h (72 h for C666-1 since its low growth rate) after transfection using the dual-luciferase reporter assay system. Luciferase activities were 10-60 fold lower when driven by oriP promoter than by CMV promoter in EBV-negative cells (p,0.001; Figure 2A). In contrast, luciferase activity was 2-5 fold higher when driven by oriP promoter than by CMV promoter in the EBV-positive C666-1 cells (p,0.01; Figure 2A). Occasionally, low levels of luciferase activity were detected in some EBVnegative cells when driven by the oriP promoter, which suggested the basal expression induced by the minimal CMV IE promoter included in oriP promoter. This observation is consistent with a previous study [12]. The expression levels of the minicircle groups were significantly higher than those of the corresponding parent groups, demonstrating that the minicircle is more efficient in mediating transgene expression in vitro. (p,0.05; Figure 2A). There was no consistent difference when minicircle groups were compared with pSP72 groups in different cell lines in the short term (48 hrs or 72hrs), which may be due to the size of the pSP72 vector is close to the minicircle vector.

Selective expression of IFNc in EBV-positive C666-1 cells mediated by oriP-vectors
To investigate whether mc-oriP-IFNc can mediate efficient expression of the IFNc gene in EBV-positive NPC cells, the minicircle-IFNc was compared with its parent plasmid p2WC31-IFNc and the intermediate plasmid from which p2WC31-IFNc was derived (pSP72). All of these plasmids contained an IFNc expression cassette driven by an oriP promoter or CMV promoter ( Figure 1A). 293 cells, NP69 cells, and three NPC cell lines were transfected according to the regimen shown in Table 1. The culture supernatant of each treatment was collected to investigate the cumulative production of IFNc over the indicated time course using ELISA. No IFNc was found in the culture medium from p2WC31-transfected cells (data not shown). The expression level of IFNc was obviously lower when driven by oriP promoter than by CMV promoter in EBV-negative cells (p,0.05; Figure 2B). However, there was no significant difference in IFNc expression level when driven by the oriP or CMV promoter in EBV-positive C666-1 cells (p.0.05; Figure 2B). The expression levels of the minicircle groups were significantly higher than those of the corresponding parent plasmid groups (p,0.05; Figure 2B). There was no consistent difference when the minicircle groups were compared with the pSP72 groups in different cell lines.
Mc-oriP-IFNc selectively inhibits the growth of EBVpositive C666-1 cells The killing effects of mc-oriP-IFNc on NPC cells compared with mc-CMV-IFNc were assessed using the WST assay. The WST assay was performed with Cell Counting Kit-8 (CCK-8) by using Dojindo's highly water-soluble tetrazolium salt WST-8. WST-8 is reduced by dehydrogenases in cells to give a yellow colored product (formazan), which is soluble in the tissue culture medium. The amount of the formazan dye generated by the activity of dehydrogenases in cells is directly proportional to the number of living cells. The detection sensitivity of CCK-8 is higher than other tetrazolium salts such as MTT, XTT, MTS or WST-1 (Cell Counting Kit-8 Technical Manual). Transfection of NPC cells with the mc-CMV-IFNc resulted in significantly reduced cell viability in both EBV-negative cells (CNE-1, CNE-2) and EBVpositive C666-1 cells (p,0.05; Figure 3). However, mc-oriP-IFNc caused a significant reduction of C666-1 cell viability but no impact on EBV-negative cells. Treatment of C666-1 cells with mc-oriP-IFNc or mc-CMV-IFNc achieved a similar extent of toxicity (relative growth rate of 48.8961.7% and 46.8962.07%, respectively; p.0.05), suggesting that the two promoters provided comparable levels of transgene expression. These results are consistent with previous observations reported by Li et al [12]. Furthermore, the minicircle showed more profound effects than the parent plasmid (p,0.01). In contrast to NPC cell lines, no growth inhibitory effect was observed in 293 and NP69 cells treated with IFNc gene transfer (p.0.05).

The oriP-based promoter is responsive to EBNA1
Although endpoint PCR is at best only semi-quantitative, we observed a distinct increase in EBNA1 mRNA level following overexpression (p,0.05; Figure 4A) and a clear decrease following siRNA treatment (p,0.05; Figure 4B). To further verify whether the targeted gene expression of oriP promoter was regulated by EBNA1 and transient expression of EBNA1 was sufficient to enhance the luciferase activity of the mc-oriP-luci treatment group, the following experiments were performed. EBV-negative CNE-2 cells were pre-transfected with an EBNA1 expression plasmid, followed by a second round of transfection with mc-oriP-luci 48 hrs later. EBNA1 expression increased the luciferase activity in the mc-oriP-luci treatment group (p,0.01; Figure 4C) while had no effect on the luciferase activity in the mc-CMV-luci treatment group (p.0.05; Figure 4C). The EBV-positive C666-1 cells were treated with siRNA to down-regulate EBNA1 expression. EBNA1 silencing reduced the luciferase activity in the mc-oriP-luci treatment group (p,0.0001; Figure 4D) with no effect on the luciferase activity in the mc-CMV-luci treatment group (p.0.05; Figure 4D).

Targeted efficacy of mc-oriP-IFNc on nasopharyngeal xenograft tumors
To determine whether the in vitro data of selective expression of the oriP-driven minicircle vector can be corroborated in a more complex in vivo model, CNE-2 (EBV-negative) and C666-1 (EBVpositive) tumors were treated with either mc-oriP-IFNc or mc-CMV-IFNc. Each mouse was treated with intratumoral injection of 100 ml plasmid-liposome complex containing 15 mg DNA and 60 ml lipofectamine 2000 once a week for 3 weeks (as shown in Table 1). The time-dependent evolution of tumor volume in mice inoculated with CNE-2 and C666-1 cells ( Figure 5A and B) indicated that treatment with mc-oriP-IFNc had minimal inhibitory effect on EBV-negative CNE-2 tumors compared with the control p2WC31 group ( Figure 5A). In contrast, the tumor size of EBV-positive C666-1 xenografts treated with mc-oriP-IFNc was significantly decreased compared with control groups (p,0.05; Figure 5B). Furthermore, mc-oriP-IFNc treatment had similar effect on EBV-positive C666-1 tumors compared to mc-CMV-IFNc treatment (p.0.05).

Selective expression in EBV-positive C666-1 tumors mediated by mc-oriP vectors
To assay the targeted expression of mc-oriP-IFNc in vivo, IFNc protein levels were analyzed in tumor and liver using a human IFNc ELISA kit. Intratumoral expression of mc-oriP-IFNc was detected only in the EBV-positive C666-1 tumor, compared with mc-CMV-IFNc treatment (p.0.05; Figure 6A and B). Systemic (liver) expression of mc-oriP-IFNc was not detected in mice bearing either the EBV-negative CNE-2 tumor or the EBVpositive C666-1 tumor, compared with mc-CMV-IFNc treatment (p,0.01). However, in mc-CMV-IFNc treatment groups, mice bearing the CNE-2 tumor had much higher expression level of IFNc in their livers than those bearing the C666-1 tumor ( Figure 6A and B).
To further verify the above result of different expressions in liver mediated by the mc-CMV vector, we detected the expression of luciferase gene carried by minicircle in liver and tumor tissues through immunohistochemistry. Consistent with the above result, in mc-oriP-luci treatment groups, the staining of luciferase was observed only in the EBV-positive C666-1 tumor but neither in the EBV-negative CNE-2 tumor nor in all livers of mice tested. However, in mc-CMV-luci treatment groups, the staining of luciferase was observed both in the EBV-positive C666-1 tumor and in the EBV-negative CNE-2 tumor. Liver staining of mc-CMV-luci was observed mainly in the mice bearing the EBVnegative CNE-2 tumor ( Figure S1). The result is similar to previous report [12].

Minicircle mediates more robust antitumor effect than conventional plasmid
In order to compare the antitumor effects of minicircle and its derived plasmid pSP72 in vivo, C666-1 cell-xenografted mice were treated with mc-oriP-IFNc and pSP72-oriP-IFNc. Treatment regimen was conducted with the same protocol as shown in Table 1. Our results indicated that mc-oriP-IFNc could mediate more robust antitumor effect than pSP72-oriP-IFNc at days 16 and 21 (p,0.05) ( Figure S2).

Discussion
Minicircle-mediated gene therapeutic techniques shown in previous studies were non-specific, which limited their application due to the potential systemic side effects. We report here for the first time both the non-viral minicircle vector used in targeted gene therapy and the combination of oriP promoter with minicircle system for NPC-targeted therapy. Our findings demonstrate that mc-oriP-IFNc induces antiproliferation of EBV-positive tumor cells, represses tumor growth, and prolongs the mouse life span in a manner as effective as that of mc-CMV-IFNc in vitro and in vivo. In addition, mc-oriP-IFNc had minimal or no killing effects on EBV-negative tissues. Therefore, this study indicates the feasibility of mc-oriP-IFNc as a safe and highly effective gene therapy system for NPC treatment.
Nasopharyngeal carcinoma has been shown to be the human tumor showing the most consistent association with EBV [8]. Utilization of molecules uniquely present within a tumor cell to elicit a cytotoxic response is an attractive strategy for the development of new approaches to the targeted therapy of neoplasia. The oriP promoter has been demonstrated to be very powerful in inducing targeted gene expression [12,17,18,19,20]. In previous studies, the oriP promoter was placed in adenovirus vectors, which achieved very high transduction efficiencies. Although viral-based systems have shown high transfection efficiencies in vivo, they have serious disadvantages such as immunogenicity and inflammatory responses [21].
Non-viral gene delivery strategies are usually based on plasmid DNA carrying the gene of interest. Conventional plasmid vectors include a bacterial backbone and a transcription unit. These sequences, however, may cause undesirable effects such as the production of antibodies against bacterial proteins expressed from cryptic upstream eukaryotic expression signals, changes in eukaryotic gene expression caused by antibiotic resistance markers, and immune responses to CpG sequences [22]. Compared to conventional plasmids, minicircle DNAs devoid of plasmid bacterial sequences are superior as non-viral DNA vector for multiple reasons: (a) relative safety due to the reduced numbers of inflammatory unmethylated CpG motifs; (b) more efficient transgene expression due to its reduced size; and (c) more robust and persistent transgene expression [16,23,24]. Previous studies have demonstrated that the use of minicircles may offer a promising avenue for safe and efficacious non-viral-based gene therapies [25,26,27,28]. Based on these superior attributes, we developed a recombinant minicircle vector carrying the human IFNc gene that had antiproliferative and antitumor effects against NPC in vitro and in vivo and also demonstrated that the antiproliferative effects of IFNc gene transfer on NPC cell lines could be attributed to G0-G1 arrest and apoptosis [13]. Considering the need for selective NPC-specific expression, the minicircle vector used in the current study was constructed such that the CMV enhancer promoter was replaced with the oriP-CMV promoter. The new promoter provides significantly higher levels of luciferase reporter gene expression in EBV-positive NPC cells compared with CMV promoter (p,0.05; Figure 2A). Although the results of IFNc gene expression are consistent with those of the reporter gene in EBV-negative cells, there was no significant difference between the mc-oriP-IFNc-treated and the mc-CMV-IFNc-treated groups in EBV-positive C666-1 cells, indicating approximately equivalent levels of IFNc gene expression and cytotoxicity between the respective mc-oriP and mc-CMV vectors in the C666-1 cells (p.0.05; Figures 2B and 3). One explanation for these observations is that the expression of different genes is variable and consequently, the cell number for IFNc gene expression was decreased after the cytotoxicity caused by IFNc.
In the present study, the overexpression of the EBNA1 gene in CNE-2 cells resulted in an increase in luciferase activity in the mc-oriP-luci treatment group ( Figure 4C). Inhibition of EBNA1 expression in C666-1 cells by EBNA1-specific siRNA subsequently reduced the luciferase activity in the mc-oriP-luci treatment group but had no effect on the mc-CMV-luci treatment group ( Figure 4D). We and others have consistently demonstrated that the oriP-based promoter specifically responds to EBNA1 [12,18,19,20]. Therefore, the tumor-specific promoter oriP can be used for gene therapy in EBV-associated diseases, such as Burkitt's lymphoma (BL), Hodgkin's disease (HD), and posttransplant lymphoproliferative disorders (PTLD).
In addition to investigating the selective antiproliferation effects of mc-oriP-IFNc in vitro, the targeted antitumor effects of mc-oriP-IFNc were assessed in NPC-xenografts. The fact that the in vitro data could be replicated in in vivo tumor models is exciting. Our data show that the antitumor effects and survival rates were similar between the mc-oriP-IFNc-treated and mc-CMV-IFNc-treated groups in the C666-1 tumor model (p.0.05; Figure 5). However, the oriP promoter was much less active in the EBV-negative CNE- cell-xenografted tumors treated with mc-oriP-IFNc or mc-CMV-IFNc, respectively. Mice were sacrificed after three weeks of treatment, and tumors were resected and frozen for immunohistochemistry assays. IRF-1 staining shows cytoplasmic and nuclear staining, p21 staining shows nuclear staining, and BAK staining shows cytoplasmic staining. In the mc-oriP-IFNc-treated group, intense IRF-1, p21 and BAK staining were observed only in EBV-positive C666-1 tumors. Tissue sections are shown at 6400 magnification. Scale bar represents 50 mm. doi:10.1371/journal.pone.0019407.g006 2 tumor xenografts, resulting in very limited antitumor effect and a reduced survival rate (p.0.05; Figure 5). Again, the minicircle was much more efficient than the parent plasmid p2FC31 and conventional plasmid pSP72 (p,0.05; Figure 5 and Figure S2), this may be due to the fact that minicircle DNA does not contain extraneous plasmid backbone sequences that could cause transcriptional repression in vivo [23,29], which is consistent with previous studies [13,30,31].
Transgene expression can be detected in the host liver with extensive intratumoral injection of the mc-CMV-IFNc ( Figure 6A and B), which raises a concern about potential systemic cytotoxicity. Furthermore, there is a big difference in expression of mc-CMV-IFNc in the livers of mice carrying different tumors. However, the reason for this unexpected result is unclear. A possible explanation is that the tissues of CNE-2 tumors that had higher amount of stroma are looser and softer than those of C666-1 tumors, which in turn may result in more minicircle DNAs in liver probably by the leakage. The underlying mechanism needs further to be elucidated. In addition, transgene expression in the mc-oriP-IFNc-treated groups has not been detected in liver tissues ( Figure 6A and B). These results provide support for the mc-oriP-IFNc plasmid as a promising and NPC-specific vector for gene therapy. The minicircle vector is also a versatile tool to carry other tissue-or cancer-specific promoters for the development of promising therapeutics for other diseases.
Orthotopic models of many tumor types have been developed in which metastasis occurs in a similar manner as in patients [32]. Data on orthotopic human NPC models, however, are still limited. In one report investigating this issue, a fluorescent orthotopic NPC metastatic model was established in nude mice using stable and high GFP-expressing NPC cell lines. This model will be useful for understanding the biology of metastatic NPC and for discovering effective therapies for this disease [33]. The next step in developing the oriP-based minicircle system will be to investigate its antimetastatic effect on the orthotopic human NPC model in nude mice. This study will lay the foundation for targeted gene therapy of metastatic NPC by intravenous delivery of a therapeutic gene.

Production and purification of minicircles
Minicircle-IFNc and minicircle-luciferase were produced according to the methods described by Chen et al. [24] with minor modifications. Briefly, overnight bacterial growth from a single colony of parent plasmid-transformed E. coli Top 10 in Tris-borate medium was centrifuged at 20uC and 4,000 rpm for 20 minutes. The pellet was resuspended 4:1 (v/v) in fresh Luria-Bertani broth containing 1.5% L-arabinose. The bacteria were incubated at 32uC with constant shaking at 250 rpm for 4 hours. After adding onehalf volume of fresh Luria-Bertani broth (pH 8.0) containing 1.0% L-arabinose, the incubation temperature was increased to 37uC, and the incubation was continued for an additional 4 hours. Episomal DNA circles were prepared from bacteria using plasmid purification kits from Qiagen (Chatsworth, CA).

Quantitative evaluation of mc-oriP-luciferase or mc-CMVluciferase expression
To evaluate transgene expression from mc-oriP-luciferase or mc-CMV-luciferase in EBV-negative and -positive cells, luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Promega). Cells were seeded in 24-well culture plates (5610 4 cells/well for C666-1, and 3610 4 cells/well for other cells). After one doubling, cells were cotransfected with mc-oriP-luciferase and pGL4.73 (Promega) simultaneously at a ratio of 50:1 (Table 1). Cell lysates were analyzed for luciferase activity using the Dual-Luciferase Reporter Assay System and a Luminometer (BERT-HOLD Technologies, Centro LB-960) according to the manufacturers' protocols [12].

Quantitative evaluation of mc-oriP-IFNc or mc-CMV-IFNc expression
To evaluate transgene expression from mc-oriP-IFNc or mc-CMV-IFNc in EBV-negative and -positive cells, the concentration of IFNc in the culture supernatant of transfected cell lines was measured with a human IFNc ELISA kit (R&D Systems, Minneapolis, MN) according to the manufacturer's protocol. The culture supernatant of transfected cells treated with mc-IFNc for 48 hours (C666-1 for 72 hours since its low growth rate) was collected and frozen (270uC) for activity analysis [13]. The sensitivity of the kit was up to 16 pg/ml.

Effect of mc-oriP-IFNc or mc-CMV-IFNc on cell viability
To evaluate the effect of mc-oriP-IFNc or mc-CMV-IFNc treatment on viability, the WST assay was used as described previously [13,36]. WST assay was performed using Cell Counting Kit-8 which was nonradioactive, allowed sensitive colorimetric assays for the determination of the number of viable cells in cell proliferation and cytotoxicity assays. In brief, EBV-negative and -positive cell lines were transfected with mc-oriP-IFNc and corresponding control plasmids (

RNA preparation and reverse transcription-PCR
Total RNA was prepared using the Micro-to-Midi Total RNA Purification System (Invitrogen) according to the manufacturer's instructions. RNA was submitted to DNase digestion and 1 mg aliquots were used for reverse transcription with the Reverse Transcription System (Promega). PCR reactions were performed using the following primers: human GAPDH, sense 59-AGA-AGGCTGGGGCTCATTTG-39 and antisense 59-AGGGGCC-ATCCACAGTCTTC-39; and human EBNA1, sense 59-AAG-GAGGGTGGTTTGGAAAG-39 and antisense 59-TGGAATAG-CAAGGGCAGTTC-39. The PCR reaction for EBNA1 was carried out using the following conditions: denaturation-95uC (30 s), annealing-62uC (30 s), and extension-72uC (30 s), with 40 cycles [37]. For GAPDH, the following reaction conditions were used: denaturation-95uC (30 s), annealing-58uC (30 s), and extension-72uC (30 s), with 25 cycles. The sizes of the PCR products were 258 bp for GAPDH and 206 bp for EBNA1, respectively. ImageJ software was used for quantitative analysis of EBNA1/GAPDH from three independent experiments.
Antitumor activity of mc-oriP-IFNc in a subcutaneous NPC tumor model Care, use, and treatment of all animals in this study were in strict agreement with the institutionally approved protocol according to the USPHS Guide for the care and use of laboratory animals, as well as the guidelines set forth in the Care and Use of Laboratory Animals by the Sun Yat-sen University.
Female BALB/c nude mice (4-6 weeks old) were obtained from Shanghai Slike Experimental Animals Co. Ltd. (Shanghai, China; animal experimental license no. SCXKhu2007-0005). After 1 week of adaptation, the mice were inoculated s.c. in the scapular region with 2610 6 CNE-2 cells or 1610 7 C666-1 cells to generate tumors for the following experiments. Once the tumor dimension reached approximately 5-8 mm (100 mm 3 ), the animals were randomly assigned to groups. Each mouse was treated with intratumoral injection of 100 ml plasmid-liposome complex once a week for 3 weeks, according to the regimen shown in Table 1. For the antitumor experiments, a total of 30 mice were used for either xenograft model (6 mice per group, 5 groups). Tumor volume (V) was measured and calculated according to the following formula: V = L6W 2 /2 (L, length; W, width). Tumors were resected at the end point and frozen (270uC) for analysis. For survival studies, there were eight mice in each group. Five treatment groups were included for either xenograft model. Animals were either found dead or sacrificed when tumors were observed by palpation to approach 10% of the body weight or when individual animals seemed to be stressed by weight loss, ruffled fur, and/or lethargy. All the animal experiments were conducted in accordance with the Guidelines for the Welfare of Animals in Experimental Neoplasia [13].

IFNc production by minicircle-IFNc transfected tumor tissues
Frozen samples were sonicated in 16TBS (25 mmol/L Tris, 138 mmol/L NaCl, and 3 mmol/L KCl, pH 7.4) and centrifuged at 8,0006g for 1 minute. The resulting supernatants were used for analysis. IFNc levels were determined using a human IFNc ELISA kit (R&D Systems) according to the manufacturer's recommendations.

Antibodies and immunohistochemistry
The following commercial antibodies were used: IRF-1 (Beijing Biosynthesis Biotechnology Co., Ltd, China), p21 and BAK (Boster Biological Technology Ltd, Wuhan, China), Firefly Luciferase (Abcam Inc, UK). Sections of 5 mm from formalin-fixed and paraffin-embedded tumors were deparaffinized in two 5-min washes with xylene and rehydrated through a graded alcohol series to distilled water. The sections were then treated with 0.3% H 2 O 2 in methanol for 15 min to block endogenous peroxidase activity. Before applying the primary antibody, sections were microwaved for antigen retrieval in 10 mmol/L citrate buffer (pH 6.0) for a total of 25 min, followed by equilibration in phosphate-buffered saline. Sections were treated first at room temperature for 30 minutes with goat serum blocking solution (Boster Biological Technology Ltd, Wuhan, China) in humidity chambers and then incubated with the respective primary antibody for 12-16 h at 4uC. After several washes, the sections were treated with the appropriate secondary antibody (Envison Kit; DAKO) for 30 min. The antigen-antibody complex was visualized by incubation with the DAB Kit (Envison Kit; DAKO). Finally, all sections were counterstained with hematoxylin. All immunostaining was first optimized in single tissue slides. Negative controls were obtained using the method as described above but without incubation with the appropriate primary antibody [39,40,41,42,43].