The present study was carried out in strict accordance with the Guide for the Care and Use of Laboratory Animals of the US National Institutes of Health, and the study protocol was approved by the Committee on the Ethics of Animal Experiments of Nanjing Medical University (SYXK2009-0015). All surgeries were performed under sodium pentobarbital anesthesia, and suffering was minimized as much as possible. Female athymic nude mice at 4 weeks old were obtained from the Experimental Animal Center of University of Yangzhou, Yangzhou, China. The animals were housed in SPF-free facilities with 12-h light-dark cycles and standard pellet feed and water ad libitum.

Human embryonic kidney 293 cells (HEK293) were obtained from the American Type Culture Collection (ATTCC, Manassas, VA, USA). Human colorectal carcinoma cell lines (HCT-8, HT-29 and SW480) and human gastric epithelial cells (GES-1) were purchased by the Shanghai Cell Collection, Chinese Academy of Sciences (Shanghai, China). All the cell lines except SW480 were cultivated in Dulbecco’s modified Eagle’s medium (DMEM; Gibco-BRL, Grand Island, NY, USA) with 10% fetal bovine serum (FBS; Gibco) and SW480 cell line was maintained in RPMI-1640 (Invitrogen, Carlsbad, CA, USA) with 10% FBS. All the cells were incubated under humidified conditions of 95% air and 5% CO₂ at 37°C.

In the present study, the original adenovirus shuttle plasmid (pDC-312) was used. The mouse H19 enhancer exon 1 (258 bp) and exon 2 (360 bp) and DMR exon 1–2 (429 bp), exon 3 (207 bp) and exon 4 (156 bp) were amplified by PCR from the mouse genomic DNA, respectively, and then the two fragments were linked to a single fragment by PCR. The DMD was cloned downstream of the enhancer by the restriction endonuclease EcoRI and NheI. The mouse H19 promoter (302 bp) was amplified by PCR from mouse genomic DNA with the primers shown in Table I. We used restriction endonuclease SalI/HindIII to clone downstream of the enhancer-DMD. The human adenovirus E1A segment (1013 bp) was amplified by PCR from a TOPK plasmid, which was benevolently provided by Dr Ji-Fan Hu (Stanford University Medical School, Stanford, CA, USA) and the primers are indicated in Table I. The enhanced green fluorescent protein (EGFP) reporter gene from the pEGFP-C1 vector (Clontech Laboratories Inc., Mountain View, CA, USA) and the E1A gene were inserted downstream of H19 promoter by using restriction endonuclease BamHI and HindIII to construct pDC312-enhancer-DMD-H19-EGFP and pDC312-enhancer-DMD-H19-E1A, respectively, which were confirmed by DNA sequencing. The adenovirus Ad312-E1A was constructed by homologous recombination techniques utilizing pDC312-enhancer-DMD-H19-E1A and the adeno-virus packaging plasmid PBHGLOX1, 3CRE.

The plasmid carrying E1A gene and the adenovirus vector Ad5 were transfected into HEK293 with liposome Lipofectamine™ 2000 (Invitrogen-Life Technologies, Carlsbad, CA, USA). The Ad312-EGFP is a standard replication deficient adenovirus and constructed by cotransfection of the adenovirus shuttle vector covering EGFP with a deleted E1A/B adenoviral backbone vector. The culture solution was changed after 4–6 h and the cytopathic effect (CPE) was continuously observed through the transfection. The CPE was observed after ~10 days and the abnormal cells and supernatant were collected, frozen and thawed at −80°C/37°C three times and centrifuged at 2,500 rpm for 15 min. The supernatant was Ad-E1A and Ad-EGFP. The adenoviruses were plaque purified and propagated in HEK293, then by a CsCl gradient according to standard techniques purified again. Functional particle titers of all adenoviruses were identified using a plaque assay in HEK293. The explicit control adenovirus (H101) was benevolently supplied by Dr Sheng-Fang Ge (Shanghai Jiao Tong University School of Medicine, Shanghai, China).

The four cell lines were seeded in 96-well plates at a density of 1,000/well for the Cell Counting kit-8 (CCK-8; Dojindo Laboratories, Kumamoto, Japan) assay and in 6-well plates at a density of 10⁶/well for RT-PCR, western blot analysis and flow cytometric analysis. The cells were incubated with various concentrations of H101, AdpDC312-E1A and AdpDC312-EGFP with serum-free DMEM at 37°C for 60 min. After the incubation period, a normal growth medium replaced the serum-free DMEM with the viruses. The infected cells were expected to continue to be cultured at 37°C for further assays.

Four kinds of cells (HCT-8, HT-29, SW480 and GES-1) were infected with adenoviral vectors (Ad-EGFP, 10 PFU/cell). EGFP expression was examined at 48 h after infection using a fluorescence inversion microscope system (excitation 450–490 nm type 108; Nikon).

E1A mRNA expression was determined by real-time PCR (RT-PCR; Applied Biosystems, Waltham, MA, USA). Four types of cells (HCT-8, HT-29, SW480 and GES-1) were, respectively, infected with Ad312-E1A (10 plaques forming units/cell). Total RNA was extracted using TRIzol (Invitrogen-Life Technologies) followed by the manufacturer’s instructions. The first strand cDNA synthesis was performed in a whole volume of 25 μl: 2 μg RNA, 0.5 μg up-primer and down-primer, 200 units of M-MLV reverse transcriptase. The cDNA was then amplified in 50 μl reaction volumes containing 0.4 μmol/l of up-primer and down-primer as well as 1.25 U Taq DNA polymers (Takara, Dalian, China) with the conditions of pre-denaturation at 94°C for 5 min, subsequently, by 35 cycles of 94°C for 40 sec, 60°C for 40 sec and 72°C for 1 min and a final extension of 72°C for 7 min. The PCR products with 1013 bp were electrophoresed on a 1% agarose gel with ethidium bromide before visualizing under UV light.

Western blot analysis was performed to evaluate the E1A protein level. The HCT-8, HT-29, SW480 and GES-1 cells were washed three times with ice-cold PBS, and the cells were suspended in lysis buffer. Cell lysates were collected through 12,000 rpm, 5 min at 4°C and then separated by 12% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinylidene difluoride (PVDF). Then blocked with 5% non-fat dry milk in TBST buffer overnight at 4°C, with mouse anti-E1A antibodies (1:1,000) and rabbit anti-human β-actin antibodies (1:500) incubated at 24°C for 2 h, and washed by shaking with TBST solution. The proteins were visualized by ECL (Wuhan Boster Biological Technology, Ltd., Wuhan, China). Western blot analyses were performed at least three times.

CCK-8 assay was based on the ability of viable cells using a Cell Counting kit-8 (CCK-8; Dojindo Laboratories). The cells were seeded in 96-well plates at a density of 1,000 cells/wells and were infected with recombinant adenoviral vectors (10 PFU/cell) the next day. Following incubation for 72 h, the cells were assayed with CCK-8 reagents by measuring absorbances at 450 nm with a microplate reader (Bio-Rad Laboratories, Richmond, CA, USA) to cell growth and viability. All samples were assayed in quadruplicate and the experiments were repeated three times.

The apoptosis assay was performed by flow cytometry after double staining with Annexin V fluorescein isothiocyanate (FITC) apoptosis detection kit (Nanjing Keygen Biotech, Co., Ltd., Nanjing, China) to discriminate early apoptosis (single Annexin V-positive) and double Annexin V/propidium iodide-positive (PI) differentiated necrotic cells. The 6-well dishes were inoculated with 1×10⁶ cells/well and then infected with 10 PFU/cell of Ad312-E1A, Ad312-EGFP and H101. Cell apoptosis was analyzed at 72 h after infection.

Tumor xenografts were established by oxter injection of 5×10⁷ HT-29 cells into the 4-week-old female athymic nude mice (the Experimental Animal Center of University of Yangzhou, Yangzhou, China). The tumor volume was measured by the formula: Volume = 1/2× (length × width²). When tumors had grown to 100 mm³, the xenografted mice were randomly divided into four groups of twelve mice in each group. The Ad312-E1A, Ad312-EGFP, H101 received intratumoral injections of 10⁹ PFU every other day. The tumor was calculated by vernier calipers every two days. All the mice were euthanized at 1 month post-inoculation. The mice were euthanized by cervical dislocation at a predetermined interval of observation. The harvested tumors were fixed in 40% buffered formalin, sectioned at 5–7 mm and stained with H&E.

The tumors were stored in 10% formalin and embedded in paraffin for staining, and then incubated at 4°C overnight with the mouse anti-human E1A antibody (1:50 dilution; Abcam, Boston, MA, USA) to detect E1A protein expression. The sections were rinsed in PBS-T (0.05% Triton X-100 in PBS) with a goat anti-mouse secondary antibody (1:500 dilution) incubation for 60 min at room temperature and then incubated with streptavidin-horseradish peroxidase (BD Biosciences, San Jose, CA, USA), diaminobenzidine substrate to form the colorimetric reaction. The positive cells were calculated in six random fields at 400 magnifications with a light microscope. Only distinct staining cells were counted and the positivity rate was utilized to grade the expression levels.

Apoptosis of the tumor cells was detected by the terminal deoxynucleotidyl transferase (TdT)-mediated dUTP nick-end labeling (TUNEL) and performed with In Situ Cell Death Detection kit (Roche, Mannheim, Germany) with the manufacturer’s direction. The tumors were filled with 10% formaldehyde and paraffin-embedded sections were prepared to stain apoptotic cells. The number of TUNEL-positive cells was calculated in six random fields at 400 magnification with a light microscope, and then the apoptosis index for each field was calculated as the percent of TUNEL-positive cells relative to the total.

IBM SPSS Statistics software version 20.0 (IBM, Armonk, NY, USA) was used to analyze the data. Every assay was performed at least three times. All experimental data were expressed as the mean ± standard deviation and assessed by the Student’s t-tests and the one-way ANOVA. The results of in vivo survival experiments were assessed by GraphPad Prism 5 (GraphPad software, San Diego, CA, USA). P<0.05 was considered to be statistically significant.

We constructed a conditional replication-competent adenovirus to gene therapy for cancer cells on loss of IGF2 imprinting and p53 mutations as shown in Table II, and cells infected with H101 served as a positive control. Enhancer-DMD fragment (1376 bp) was verified through the restriction endonuclease SalI and XbaI from the recombinant plasmids pDC312-enhancer-DMD. The H19 fragment (302 bp) was confirmed through the restriction endonuclease HindIII and SalI from the recombinant plasmid pDC312-enhancer-DMD-H19. In addition, E1A (1013 bp) and EGFP (718 bp) fragment were, respectively, verified through the restriction endonuclease BamHI and HindIII from both pDC312-enhancer-DMD-H19-E1A and pDC312-enhancer-DMD-H19-EGFP, which were transfected into HEK293 cells, respectively. We observed CPE in HEK293 cells (Fig. 1) after infection with Ad312-E1A at day 10 and day 13. The successful construction of oncolytic adenovirus Ad312-E1A and Ad312-EGFP was confirmed and regulated by the IGF2 imprinting systems, and the Ad312-E1A expressed efficiently and replicated selectively in HCT-8 and HT-29 (IGF2 LOI), except for SW480 and GES-1 (IGF2 MOI). Targeted therapy for CRC was mediated-by oncolytic adenovirus vector, which was based on the IGF2 LOI genomic imprinting systems.

An EGFP reporter gene was utilized to examine the applicability of the expression system. After infection of the four cell types with Ad312-EGFP (10 PFU/cell) for 48 h, EGFP gene expression was found in LOI cells (HCT-8 and HT-29), but negative or only weakly positive EGFP was observed in MOI cells (SW480 and GES-1) for maintained normal IGF2 imprinting system (Fig. 2). Hence, the virus gene therapy system only expressed the reporter gene in the tumor cells with IGF2 LOI.

We tested E1A mRNA and protein expression in HCT-8, HT-29, SW480 and GES-1 cells which were, respectively, infected with Ad312-E1A and H101 (10 PFU/cell). The expression of E1A mRNA and protein were determined by RT-PCR and western blot analysis 48 h after infection, respectively. As shown in Fig. 3, in the Ad312-E1A group, E1A mRNA and protein were expressed in LOI cells (HCT-8 and HT-29), not in MOI cells (SW480) or normal cell (GES-1). The p53 mutant cells (HT-29) infected with H101 lead to high E1A expression in mRNA and protein level, which was hardly expressed in p53 wild-type cells (HCT-8, SW480 and GES-1), indicating that H101 expressed only in cell lines with p53 mutant.

The cytotoxic effects of adenoviral vectors expressing E1A were observed in four cell types. The groups were treated with oncolytic adenovirus (10 PFU/cell) Ad312 EGFP, Ad312-E1A, H101 and PBS at 72 h, their viability was assessed by CCK-8 assay (shown in Fig. 4A). The result of CCK-8 assay revealed that the cell viability of the LOI cells (HCT-8 and HT-29) infected with Ad312-E1A was significantly reduced when compared with the MOI cells (SW480 and GES-1) (P<0.05), which still had higher cell viability. In the same way, the viability of p53 mutant cell line (HT-29) infected with oncolytic adenovirus H101 was also significantly decreased in contrast to the p53 wild cell lines (HCT-8, SW480 and GES-1) (P<0.05), which had obviously stronger cell viability.

Apoptosis in the four cell types was calculated using flow cytometry 72 h after Ad312-E1A infection. To evaluate the cytopathic effect of adenoviral infection on the cells, apoptosis of cells infected with Ad312-EGFP was measured as the negative control (Fig. 4B). The resutls indicated that the apoptosis rate in LOI cell lines (HCT-8 and HT-29) infected with Ad312-E1A (10 PFU/cell) was significantly higher than that in the control group (P<0.05). However, there was no significant difference of apoptosis ratio between MOI cells (SW480 and GES-1) infected with Ad312-E1A (10 PFU/cell) and the negative control group (P>0.05). To evaluate the cytopathic effect of H101 and Ad312-E1A on the p53 mutant cells (HT-29), similar experimental procedure was applied, and the results showed no obvious difference between the two groups (P>0.05), as shown in Fig. 4C.

The antitumor effect of the recombinant adenoviral was tested in nude mice transplanted with HT-29 cells. We measured tumor volume once every three days for 30 days after injecting Ad312-E1A, H101, Ad312-EGFP and PBS (n=12 per group), and the average volume of the tumors were 432, 498, 2132 and 2233 mm³, as shown in Fig. 5A. After 30 days of injection, the tumor tissues were harvested, as shown in Fig. 5B. Although no significant difference existed between infected with Ad312-E1A and H101 groups, a significant antitumor efficacy was shown in these two groups compared with that in PBS and Ad312-EGFP groups (P<0.01). In addition, the average survival time of the four groups treated with Ad312-E1A, H101, Ad312-EGFP and PBS were 149, 152, 60 and 55 days, as shown in Fig. 5C. In short, the above results showed that the survival time of group with infection of Ad312-EGFP was not significantly prolonged compared with the group injected with PBS (P>0.05), and that the survival time of the group injected with the Ad312-E1A and H101 was obviously prolonged compared with PBS group (P<0.05).

To detect apoptosis in tumors from mice injected with Ad312-E1A, Ad312-EGFP, H101 and PBS, TUNEL assay was applied after the tumor tissue obtained, the number of apoptotic bodies in tumor tissue after staining with hematoxylin and eosin (H&E) was higher in the two groups with Ad312-E1A and H101 infection than in the group treated with Ad312-EGFP and PBS (shown in Fig. 6A and B). Furthermore, the expression of E1A protein was confirmed in tumor tissue from mice injected with Ad312-E1A or H101, as shown in Fig. 6C. In addition, the apoptosis index, represented by the percentage of TUNEL-positive cells, in the group of H101 (0.63±0.04) and E1A (0.59±0.05) was higher than that in group of PBS (0.19±0.06), EGFP (0.27±0.02), separately (P<0.01). Moreover, the expression rate of E1A in tumour tissue of the four groups was 0.009±0.001 in PBS group, 0.01±0.002 in Ad312-EGFP group, 0.43±0.06 in H101 group and 0.48±0.08 in Ad312-E1A group, respectively, as shown in Fig. 6D. In conclusion, cell apoptotic indexes of tumour tissue from xenograft mice in groups treated with both Ad312-E1A and H101 were obviously higher than in the groups treated with the PBS and EGFP, separately (P<0.05), and the apoptotic indexes of xenograft tumors treated with Ad312-E1A had no difference compared with the H101 group (P>0.05).