The human cancer cell lines DU145 (prostate adenocarcinoma), MDA-MB-231 (breast adenocarcinoma), A549 (lung adenocarcinoma), HCT116 (colon carcinoma) and A375 (skin melanoma), were cultured in Dulbecco’s modified Eagle’s medium (DMEM) with 10% fetal bovine serum (FBS) (HyClone, Logan, UT, USA). U373-MG, U87-MG, U251N and U343 (human malignant glioma), Jurkat (human T cell leukemia), SK-MEL-2 and SK-MEL-3 (human skin melanoma) cells were cultured in RPMI-1640 with 10% FBS. HeLa (human cervical adenocarcinoma) cells were cultured in minimal essential medium (MEM) with 10% FBS. HEM (human epidermal melanocyte) cells were cultured in melanocyte growth medium containing low levels of serum and fully supplemented with growth factors and antibiotics. 293A cells (a subclone of the 293 human embryonic kidney cell line) were cultured in DMEM with 10% FBS (HyClone). Cells were maintained in a 37°C humidified atmosphere containing 5% CO₂.

Anti-mTOR, anti-phospho-Akt, anti-Bcl-2, anti-Bcl-xL, anti-LC3B antibodies and wortmannin were purchased from Cell Signaling Technology (Beverly, MA, USA). Fluorescein isothiocyanate (FITC)-conjugated anti-CD46, allophycocyanin (APC)-conjugated anti-mouse IgG1 and mouse IgG fluorescence control isotype antibodies were purchased from BD Biosciences (San Jose, CA, USA). Anti-CAR and anti-actin antibodies, 3-methyladenine (3-MA) and rapamycin were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). All other chemicals, including pepstatin A, E-64d and curcumin were purchased from Sigma-Aldrich (St. Louis, MO, USA).

For the construction of chimeric sequences of Ad5 and Ad35 fiber regions, a synthetic gene (2374 bp) encoding Ad5 and Ad35 fibers was cloned into XbaI/KpnI-digested pBluescript II SK(-) and named pBSK[2374]. DNA synthesis was carried out by Epoch Biolabs (Sugar Land, TX, USA). The synthesized DNA sequences spanned a region from the 30470th (XbaI) to the 33599th (KpnI) nucleotide of the human Ad5 genome sequence, with substitution of the Ad35 fiber shaft and knob domain. To remove the KpnI site present within the Ad35 shaft, the nucleotide sequence ggtacc was replaced with the sequence ggcacc, without altering the amino acid sequence. The resultant sequences were a mixture of the Ad5 fiber tail domain and the Ad35 fiber shaft and knob domain. Following the digestion of pBSK[2374] with XbaI/KpnI, the insert was subcloned into XbaI/KpnI-digested pSK[5543], a structure that has been previously described in detail (34). The resultant product was used as a shuttle vector. This newly constructed adenovirus shuttle vector with chimeric fibers, pSK[5543 with 5/35 fiber], was digested with XmnI and the adenoviral vector vmdl324Bst containing the Ad5 genome lacking the E1 region (340–4640 in the nucleotide sequence of Ad5) and the E3 region (28592–30470 in the nucleotide sequence of Ad5) and the IX gene was linearized with SpeI for homologous DNA recombination in the E. coli strain, BJ5183. To generate Ad5/F35 chimeric fiber-incorporated green fluorescent protein (GFP)-expressing adenovirus, further homologous recombination was performed after subcloning the GFP gene from pEGFP-N1 (AflII-blunt-BamHI) into the pSP72-ΔE3-adenovirus death protein (ADP) Ad shuttle vector (35), after digestion with KpnI-blunt-BamHI for substitution of ADP with GFP. Subsequently, the second homologous recombination was performed in BJ5183 following digestion of the shuttle vector with XmnI and of the adenoviral vector with Srf I. Finally, pCA14 containing the IX gene, which plays a role in efficient viral proliferation, was linearized with XmnI for homologous DNA recombination with the previous recombinant adenoviral vectors following digestion with BstBI. To verify the respective homologous recombinants, plasmid DNA was purified from an E. coli culture grown overnight, digested with HindIII and the digestion pattern was analyzed. The appropriate homologous recombinant adenoviral plasmid DNA was digested with PacI and transfected into 293A cells to generate GFP-expressing chimeric adenoviruses.

These chimeric adenoviruses, with or without the GFP gene, were propagated in 293A cells and amplified for purification, according to standard methods. Titration was performed by estimating the infectious viral particles with a standard plaque assay kit developed by Qbiogene (Carlsbad, CA, USA) in 293A cells. The Adeno-X Rapid Titer kit (Clontech; Mountain View, CA, USA) was used for the titration of GFP-expressing adenovirus.

For the construction of adenoviral vectors, a synthetic gene (3484 bp) was used as a backbone construct for the shuttle vector and the DNA was named pBSK[3484]. It consisted of inverted terminal repeats (ITR) - packaging signal - mouse survivin promoter-E1A-BGH polyA-E1B and 55-kDa gene cassette-Ad E1 right region (in the order listed). The mouse survivin promoter may be replaced with other promoter types (CMV promoter, human survivin promoter), through digestion with KpnI and XhoI, respectively, for the examination of promoter activity. pcDNA3.1/Hygro was used as a template for PCR amplification of the CMV promoter, with the sense primer, 5′-CGGGGTACCGATGTACGGGCCAGAT-3′, and the anti-sense primer, 5′-CCGCTCGAGAATTTCGATAAGCCAG-3′. Following digestion of the PCR product of the CMV promoter with KpnI/XhoI, it was inserted into KpnI/XhoI-digested pBSK[3484]. Human survivin promoter was synthesized from +1 to −268 nucleotides of the survivin promoter and then inserted into KpnI/XhoI-digested pBSK[3484]. For the replacement of the CMV promoter or the human survivin promoter, the E1B 55-kDa gene cassette was deleted by EcoRI and SalI digestion, followed by blunting, so as to provide oncolytic activity. These various types of pSK[3484] constructs were again digested with FspI and BamHI and ligated into SspI and BglII-digested pCA14 for the formation of the final E1 shuttle vector. The pCA14-[3484] shuttle vector was linearized by XmnI digestion. The linearized pCA14-[3484] was co-transformed into E. coli BJ5183, together with BstBI-digested vmdl324Bst, for homologous recombination. To verify the respective homologous recombinants, the plasmid DNA, purified from an overnight-grown E. coli culture, was digested with HindIII and the digestion pattern was analyzed. The homologous recombinant adenoviral plasmid DNA was digested with PacI and transfected into 293A cells to generate tumor-selective, replication-competent Ad-3484-CMV-ΔE1B-derived ΔE1B19/55 or Ad-3484-human survivin promoter-derived ΔE1B19.

Cancer cell surface receptors (CAR and CD46) were quantified by flow cytometric analysis. After various cancer cells were trypsinized and washed twice with ice-cold PBS, they were separately incubated with mouse anti-CAR antibody (Santa Cruz Biotechnology) for 1 h at 4°C. After washing twice with ice-cold PBS, cells were incubated with APC-conjugated anti-mouse IgG antibody for 45 min at 4°C in the dark. Subsequently, the cells were washed twice with ice-cold PBS, incubated again with FITC-conjugated anti-CD46 for 1 h at 4°C in the dark, then washed twice with ice-cold PBS. Mouse IgG fluorescence control antibody was used as the negative control. Finally, the cells were suspended again in PBS and analyzed using a FACS Calibur flow cytometer (BD Biosciences, Lincoln Park, NJ, USA).

To evaluate the CPE of replication-competent adenoviruses, cells were first plated to ∼80% confluency into 24- or 48-well plates. They were pre-treated with 3-MA (Santa Cruz Biotechnology), wortmannin (Cell Signaling Technology), or rapamycin (Santa Cruz Biotechnology) for 1 h and were then infected with replication-competent adenoviruses of various multiplicities of infection (MOI). After 8 h of infection, cells were monitored daily under a microscope. When the infected cells exhibited cell lysis at the lowest MOI, the remaining cells on the plate were fixed with 4% paraformaldehyde and stained with 0.5% crystal violet.

Cells were lysed with 1X Laemmli lysis buffer [62.5 mM Tris (pH 6.8), 2% sodium dodecyl sulfate (SDS), 10% glycerol, 0.002% bromophenol blue] and boiled for 10 min. Protein content was measured using BCA Protein Assay reagent (Pierce, Rockford, IL, USA). The samples were diluted with 1X lysis buffer and β-mercaptoethanol was added to a concentration of 350 mM. Equal amounts of protein were loaded onto 10% SDS polyacrylamide gels. SDS-polyacrylamide gel electrophoresis (PAGE) analysis was performed according to the procedure of Laemmli using a Hoefer gel apparatus.

Proteins were separated by SDS-PAGE and electrophoretically transferred onto polyvinylidene fluoride (PVDF) membranes. Each PVDF membrane was blocked with 5% non-fat dry milk in PBS-Tween-20 (0.1%, v/v) at room temperature for 1 h. The membrane was then incubated with the primary antibody (diluted according to the manufacturer’s instructions) for 2 h. Horseradish peroxidase-conjugated anti-rabbit or anti-mouse IgG was used as a secondary antibody. Immunoreactive proteins were visualized by chemiluminescence (ECL; Amersham, Arlington Heights, IL, USA).

The oncolytic effect of adenoviruses was assessed using the CytoTox 96^® Non-Radioactive Cytotoxicity Assay kit (Promega, Madison, WI, USA). This assay can quantitatively measure lactate dehydrogenase (LDH) levels, a stable cytosolic enzyme that is released upon cell lysis. After the various cancer cells were treated with 3-MA or rapamycin for 1 h, they were infected with two types of oncolytic adenovirus of various MOIs and incubated for 3–7 days in 48-wells. A total of 50 μl of supernatant from each well was transferred onto a new 96-well flat-bottom (enzymatic assay) plate and 50 μl of Reconstituted Substrate Mix were added to each well, followed by incubation for 30 min at room temperature in the dark. After 30 min, 50 μl of stop solution were added to each well and the absorbance was examined at 490 nm within 1 h, using an ELISA plate reader. For LDH positive control, lysis solution (10X) was added to the control cells and the lysed cells were incubated for 45 min, after which time the plate was centrifuged at 250 x g for 4 min. Following centrifugation, 50 μl of supernatant were transferred to an enzymatic assay plate for LDH assay, as described above.

Data are presented as the means ± SEM of replicate samples, as described in the figure legends. The Student’s t-test was used to compare two different groups. A value of p<0.05 was considered to indicate a statistically significant difference.

The infectivity of Ad5/F35 was expected to be higher than that of Ad5 in the cancer cell lines expressing CAR or CD46. We first examined the transduction efficiency of GFP-expressing Ad5/F35 in a variety of human cancer cell lines to identify the types of cancer cells that may be suitable targets for chimeric Ad5/F35. Two types of adenoviruses that were replication-defective or tumor-selective and replication-competent were produced and each type exhibited an alteration in the fiber region. Their detailed structures are shown in Fig. 1. In general, melanoma cells are known to require high doses of adenovirus for sufficient gene transfer, due to the the paucity of CAR expression (19). However, it is unlikely that transduction efficiency correlates with a specific type of cancer cell. Compared to Ad5, chimeric Ad5/F35 did not exhibit higher transduction efficiency in SK-MEL-2, SK-MEL-3 and other cancer cell lines. To further investigate the effect of chimeric Ad5/F35 on gene transfer efficiency in melanoma cells (SK-MEL-2), chimeric Ad5/F35 adenovirus was infected at various MOIs. As shown in Fig. 2B, the transduction efficiency of Ad5/F35 at different MOIs was very similar to that of Ad5, which was consistent with the findings of Yu et al(21,22).

Despite our expectations, the transduction efficiency of chimeric Ad5/F35 in most cancer cells was not enhanced compared to that of Ad5. Therefore, we investigated whether a correlation exists between the transduction efficiency of Ad5 or Ad5/F35 and CAR or CD46 expression levels in the plasma membrane of the cancer cells. We did not discover any direct correlation between the transduction efficiency and CD46 expression following the transfection of cancer cells with Ad5/F35. CAR and CD46 were highly expressed in the majority of the cancer cell lines, apart from the SK-MEL-3 cells (Fig. 3). This indicated that the infectivity of Ad5/F35 is not modified in relation to cell surface receptors. To confirm that the transduction efficiency of Ad5/F35 and the levels of CD46 were not the main factors responsible for optimal gene transfer, we investigated the possibility that autophagy facilitates a faster induction of Ad5/F35 viral particle production compared to Ad5. According to recent studies, adenovirus has the ability to induce autophagy (36–39); therefore, autophagy may be used to facilitate the release of viral progeny, promote adenovirus replication and improve oncolysis (23,39).

To estimate the oncolytic activity induced by adenoviruses, various tumor cells were infected with two types of replication-competent adenovirus (Ad-3484-CMV-ΔE1B and Ad-3484-hsurvivin-ΔE1B19KD). For Ad-3484-CMV-ΔE1B, E1A gene expression was regulated by the CMV promoter and the entire E1B gene was deleted, whereas in the case of Ad-3484-hsurvivin-ΔE1B19KD, E1A expression was regulated by the human survivin promoter and only E1B55KD was expressed. The oncolytic activity induced by these viruses was very similar in all the cancer cells examined (Fig. 4), whereas no oncolytic activity was observed in normal cells (data not shown). In this study, we also compared the oncolytic activity induced by Ad-3484-CMV-ΔE1B with Ad5 fiber (control) with that of Ad-3484-CMV-ΔE1B-2374 with chimeric Ad5/F35 in various cancer cell lines (Fig. 5A). As previously mentioned, we also examined whether the autophagic condition is essential for oncolytic activity. For this experiment, rapamycin was used as an autophagy inducer and 3-MA, as well as wortmannin [phosphoinositide 3-kinase (PI3K) inhibitors], were used as autophagy inhibitors, based on their inhibitory effect on class III PI3K activity, which is essential for the induction of autophagy (40). As shown in Fig. 5A, the oncolytic activity induced by chimeric Ad5/F35 was higher in DU145 and A549 cells following rapamycin treatment, but not in SK-MEL-2, SK-MEL-3, A375 and HeLa cells. To verify this comparative potency of the treatments, LDH cytotoxicity assay was performed. We discovered that the oncolytic activity induced by chimeric Ad5/F35 was higher in the DU145 and A549 cells following pre-treatment with rapamycin, by comparing the survival rates of the cells treated with rapamycin to those of the cells transfected with chimeric Ad5/F35 without rapamycin pre-treatment (DU145, 96→61% and A549, 76→59%), at an MOI of 10 (Fig. 5B). We then aimed to determine the factor(s) that mainly control(s) the autophagy-enhanced oncolytic effect. We first examined the endogenous levels of several key molecules related to survival or anti-apoptosis, such as mTOR, phosphorylated Akt and Bcl-2 or Bcl-xL (Fig. 5C). As shown in Fig. 5C, DU145, A375 and A549 cells have higher survival potentials, whereas SK-MEL-2, SK-MEL-3 and HeLa cells have lower survival potentials. We also calculated the doubling time of each cell line in search of a correlation between oncolytic activity and proliferation rate (Table I). Based on our overall results, we discovered a correlation between oncolytic activity following rapamycin pre-treatment and survival potential. Accordingly, rapamycin-pre-treated DU145 cells exhibited the highest oncolytic activity. To confirm this finding, several other cancer cells with a higher survival potential were also examined using rapamycin, 3-MA and wortmannin, after narrowing down the range of MOIs. Our data indicated that Ad-3484-CMV-Δ E1B-2374 with chimeric Ad5/F35 increased its oncolytic activity in the presence of rapamycin in DU145, A549, SK-MEL-28 and A375 cells, whereas no increase in oncolytic activity in the presence of rapamycin was observed in SK-MEL-2, U251N and HCT116 cells. SK-BR3 and SK-OV3 cells exhibited some oncolytic activity induced by these adenoviruses, irrespective of chemical treatment; however, the oncolytic effect of the Ad-3484-CMV-ΔE1B-2374 virus was less significant than that of the control virus on HeLa cells (Fig. 6A). As shown in Fig. 6A, the selective improvement in the oncolytic effect induced by Ad-3484-CMV-ΔE1B-2374 + rapamycin was also observed in DU145, A549, SK-MEL-28, SK-BR3 and SK-OV3 cells, as indicated by the survival rates (DU145, 69→32%; A549, 56→36%; SK-MEL-28, 52→37%; SK-BR3, 67→45% and SK-OV3, 83→50%), at 50 MOI. As regards the latter two cell lines (SK-BR3 and SK-OV3), the selectivity was only observed with Ad-3484-CMV-ΔE1B-2374, irrespective of the presence of inhibitors (Fig. 6A). Similar patterns of oncolytic activity were also observed by LDH cytotoxicity assay (Fig. 6B). The survival potential of each cell line was investigated (Fig. 6C) and the doubling time was estimated (Table II). However, as shown in Fig. 6D, the reduced oncolytic activity induced by Ad5/F35 in HeLa cells may be explained by the possibility that the cellular autophagic condition in HeLa cells is already established without an autophagy inducer, which results in less oncolytic activity induced by Ad5/F35 than by Ad5. Based on these results (Figs. 5 and 6, Tables I and II), we created a frequency diagram to demonstrate the correlation between survival or cell proliferation rate and oncolytic activity induced by Ad5/F35 + autophagy inducer. As shown in Fig. 7, apart from the HCT116 cells, the majority of cancer cells exhibited higher survival potentials and increased oncolytic activities induced by Ad5/F35 + autophagy inducer. However, we discovered that the cell proliferation rate was not related to oncolytic activity induced by Ad5/F35 + autophagy inducer. Taken together, our results suggest that rapamycin, being an autophagy inducer, exerts a pronounced effect on cancer cells with a higher survival potential.