A new oncolytic adenoviral vector carrying dual tumour suppressor genes shows potent anti-tumour effect

Abstract Cancer Targeting Gene-Viro-Therapy (CTGVT) is a promising cancer therapeutical strategy that strengthens the anti-tumour effect of oncolytic virus by expressing inserted foreign anti-tumour genes. In this work, we constructed a novel adenoviral vector controlled by the tumour-specific survivin promoter on the basis of the ZD55 vector, which is an E1B55KD gene deleted vector we previously constructed. Compared with the original ZD55 vector, this new adenoviral vector (ZD55SP/E1A) showed much better ability of replication and reporter gene expression. We then combined anti-tumour gene interleukine-24 (IL-24) with an RNA polymerase III-dependent U6 promoter driving short hairpin RNA (shRNA) that targets M-phase phosphoprotein 1 (MPHOSPH1, a newly identified oncogene) by inserting the IL-24 and the shRNA of MPHOSPH1 (shMPP1) expression cassettes into the new ZD55SP/E1A vector. Our results demonstrated excellent anti-tumour effect of ZD55SP/E1A-IL-24-shMPP1 in vitro on multiple cancer cell lines such as lung cancer, liver cancer and ovarian caner. At high multiplicity-of-infection (MOI), ZD55SP/E1A-IL-24-shMPP1 triggered post-mitotic apoptosis in cancer cells by inducing prolonged mitotic arrest; while at low MOI, senescence was induced. More importantly, ZD55SP/E1A-IL-24-shMPP1 also showed excellent anti-tumour effects in vivo on SW620 xenograft nude mice. In conclusion, our strategy of constructing an IL-24 and shMPP1 dual gene expressing oncolytic adenoviral vector, which is regulated by the survivin promoter and E1B55KD deletion, could be a promising method of cancer gene therapy.


Introduction
Current anti-mitotic drugs for cancer, including vinca alkaloids and taxanes, function by disturbing spindle assembly, which activates the spindle assembly checkpoint (SAC) and causes mitotic arrest and triggers apoptosis [1]. However, all these drugs also exhibit side effects, including neurotoxicity, presumably due to the perturbation of microtubules in neurons [2]. As an effort to develop new anti-mitotic targets without this defect, in this study, we focused on a novel combination of IL-24 and MPHOSPH1, a newly discovered oncogene. MPHOSPH1 is a member of the kinesin superfam-ily. It is a plus-end-directed molecular motor playing important roles in cytokinesis [3]. Kanehira et al. recently found that MPHOSPH1 was up-regulated in bladder cancer, and MPHOSPH1 knockdown resulted in cytokinesis defect that significantly inhibited the growth of bladder cancer cells [4], which made it a potential target for cancer gene therapy. In this study, we found for the first time that in addition to previously reported bladder cancer cells, MPHOSPH1 was also up-regulated in multiple cancer cells. MPHOSPH1 knockdown caused mitotic arrest without activation # These authors contributed equally to this work. of SAC, which may be attributed to a lack of effect on microtubules (spindle), followed by post-mitotic apoptosis.
In 2001, we designed a cancer treating strategy called CTGVT, which combines gene therapy with oncolytic viral therapy by inserting an anti-tumour gene into the oncolytic viral vector [16]. Our previous CTGVT work used the ZD55 vector, which was based on the deletion of E1B55KD gene, as the oncolytic adenoviral vector (OV) [17]. Deletion of adenoviral E1B55KD, an adenovirus early gene, resulted in the loss of replication capacity of adenovirus (Ad) in normal cells but not in tumour cells, and thus confer new tumour-specific replication property to adenovirus [18]. Similar E1B55KD gene deletion strategy has also been used in ONYX-015, the first generation of commercial oncolytical virus product [19]. We have also demonstrated that the ZD55-gene system could be a potent anti-tumour strategy [5,17,20,21]. Despite its successful applications, disadvantages of this adenovirus have also been reported. For example, we observed impaired replication ability and even certain liver toxicity of ZD55 vector [17,22]. To improve the replication ability of the ZD55 vector in tumour tissue and its safety in normal tissue, in this study, we replaced its E1A promoter with a 269 bp tumour-specific survivin promoter, and named this new vector ZD55SP/E1A. The replication of this new vector is controlled by two mechanisms (both genetic deletion of E1B55KD and survivin promoter-controlled E1A). We then used this dual regulated vector ZD55SP/E1A to express tumour suppressor gene IL-24 and shRNAs targeting MPHOSPH1 (shMPP1) in cancer cells. Our results showed that ZD55SP/E1A-IL-24-shMPP1 had excellent anti-tumour effects both in vitro and in vivo. We also found that cIAP down-regulation may contribute to the enhanced antitumour effects of ZD55SP/E1A-IL-24-shMPP1. Taken together, our results suggested that ZD55SP/E1A-IL-24-shMPP1 could be a highly promising agent for cancer gene therapy.

Cell lines and culture conditions
Detailed information for cell lines used in this study and their culturing conditions can be found in the Supporting Material.

Construction of the adenoviral plasmids
The adenoviral packaging plasmid pBHGE3 was obtained from Microbix Biosystems (Toronto, Ontario, Canada). The adenoviral shuttle plasmids pZXC2, pZD⌬55 and pCA13-IL-24 were constructed as we previously described [5,17,23]. Detailed information on the construction of adenoviral plasmids can be found in the Supporting Material.

Luciferase assay
pGL3-Basic (promoter-free) and pGL3-Control (SV40 promoter) were used as the luciferase reporter plasmids. The survivin promoter on pTG19-SP was cloned into pGL3-Basic to generate the luciferase reporter plasmid pGL3-SP. One day before transfection, 0.5-2ϫ10 5 cells were seeded in 24well plates. The cells were transfected with pGL3-SP by Lipofectamine2000 (Invitrogen, Carlsbad, CA, USA) and also with pGL3-Basic and pGL3-Control as controls. The cells were harvested 48 hrs later, and the luciferase activity was assayed by the dual-luciferase reporter assay system (Promega, Madison, WI, USA).

RT-PCR analysis
Total RNA was isolated from cells using Trizol (Invitrogen). A total of 2 g of RNA was used to synthesize the first single-strand cDNA for PCR amplification using RevertAid™ First Strand cDNA Synthesis Kit (Fermentas, Vilnius, Lithuania). The relative quantification of MPHOSPH1 cDNA and adenovirus gene E3 cDNA was measured by using the ABI 7500 Fast System and the Bio-Rad CFX96™ Real-Time PCR Detection System with Power SYBR ® Green PCR Master Mix (Applied Biosystems, Carlsbad, CA, USA), with GAPDH cDNA as an internal standard, with primers as follows: Forward

Cytotoxicity assay
Tumour cells and normal cells were seeded in 24-well plates, and 24 hrs later, they were infected with Ads at various MOIs. Four days after infection, the medium was removed, and the cells were exposed to 2% crystal violet in 20% methanol for 20 min. The plates were then washed with water and documented as photographs.

Cell viability assay
SW620 and BEL-7404 were seeded in 96-well plates, and 24 hrs later, they were infected with Ads at MOI of 1. At 24, 48, 72 and 96 hrs after the infection, 20 ml of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT; 5 mg/ml) was added to each well, and the plates were incubated at 37ЊC for four hrs. The medium was then removed and 100 ml of 0.04M HCl-isopropanol solution was added to each well. Absorbance at 595 and 655 nm was measured using a Microplate Reader (Thermo, Asheville, NC, USA). Ten replicate wells were counted for each assay.

Cell nuclei and skeleton staining assay
Cells were seeded on glass cover slips in six-well plates, and 24 hrs later they were infected with Ads at MOI of 1. Cells were then fixed with 0.

Western blot assay
Cells were harvested and washed with PBS, and then lysed in SDS-PAGE sample buffer. Protein concentrations were determined using the Bio-Rad protein assay buffer. A total of 50 g of protein was separated on 8-15% polyacrylamide gels and transferred to Polyvinylidene Fluoride (PVDF) membranes (Millipore, Billerica, MA, USA). Detailed information for the antibodies used in this study can be found in the Supporting Material.

Flow cytometric analysis
For propidium iodide (PI) staining, cells were harvested and washed with PBS, fixed with 70% ethanol at 4ЊC for 30 min. The cells were then washed with PBS and incubated at 37ЊC for 30 min. with RNase A, washed with PBS and stained with PI (50 g/ml) for 30 min. to be ready for fluorescence-activated cell sorting (FACS) analysis. For JC-1 staining, cells were harvested and resuspended in 0.5 ml of medium, and 0.5 ml of 5 g/ml JC-1 fluorescent probe (Sigma-Aldrich) was added into the medium; the cells were incubated at 37ЊC for 20 min. and washed with PBS and were ready for FACS analysis. For fluorescein isothiocyanate (FITC) and PI double-staining FACS, an Annexin V-FITC Kit (Bender MedSystems, Vienna, Austria) was used, and the cells were treated according to the manufacturer's instructions. The FACS assay was performed on a FACSCalibur flow cytometer (BD Biosciences, Franklin Lakes, NJ, USA) immediately after staining.

SA-␤-gal staining assay
␤-Galactosidase staining was performed with a senescence-associated ␤-Galactosidase Staining Kit (Beyotime, Shanghai, China). Cells were washed three times with PBS and fixed with 4% paraformaldehyde for 15 min. at room temperature. Next, the cells were incubated overnight at 37ЊC in darkness with the working solution containing 0.05 mg/ml X-gal.

Animal experiments
Animal experiments were performed according to the SIBS Guideline for the Care and Use of Laboratory Animals. Male BALB/c nude mice (4-week old) were purchased from the Shanghai Experimental Animal Center (Shanghai, China). To establish xenograft tumours, 5 ϫ 10 6 SW620 cells in 150 l DMEM were subcutaneously injected into the right flank of each mouse. When tumours reached 100-150 mm 3 in volume, mice were divided randomly into five groups (seven mice per group). A total of 2 ϫ 10 9 PFUs of Ads per mouse or PBS were injected every other day into the tumours for four times. Tumour volume (mm 3 ) was measured by a Vernier calliper every week and calculated by length (mm) ϫ width (mm) ϫ width (mm)/2.

Immunohistochemistry and terminal deoxynucleotidyl transferase-mediated dUTP biotin nick end-labelling assay
Tumour tissues were fixed in 4% formaldehyde for 2 days, dehydrated in 30% sucrose solution and embedded in Jung tissue freezing medium (Leica Microsystems, Nussloch Gmbh, Nussloch, Germany) at Ϫ80ЊC. Detailed procedures of immunohistochemistry and terminal deoxynucleotidyl transferase-mediated dUTP biotin nick end labelling (TUNEL) assay can be found in the Supporting Material.

Statistical analysis
All data were expressed as the mean Ϯ S.D. and were analysed using independent sample t-tests and one-way ANOVA using SPSS Base 10.0. Results were considered statistically significant when P Ͻ 0.05.

Clone of survivin promoter and its tumour-specific transcriptional activity
Survivin is a member of the inhibitor of apoptosis (IAP) family. We had previously took advantage of survivin differential expression in tumour tissues versus normal tissues [24][25][26], to construct oncolytic adenovirus with a 1.1 kb survivin promoter [27]. In this study, a further shortened 269 bp survivin promoter was used to construct a luciferase reporter plasmid containing this short promoter. Promoter activity was determined by the luciferase activity after transient transfection. In SW620 and HeLa cells, the luciferase expression level driven by the short survivin promoter was similar to that driven by the strong SV40 viral promoter (pGL3-Control), and was much higher than that observed with the promoterless control (Fig. 1A). In contrast, in transient transfected normal MRC-5 and HLF-1 cells, only baseline levels of luciferase expression were observed (Fig. 1B). All these results suggested that the short 269 bp survivin promoter has a strict tumour-selective transcriptional activity.

MPHOSPH1 is up-regulated in various tumour cells
It has been reported that MPHOSPH1 is up-regulated in bladder cancer cells but not in normal human tissues [4]. To evaluate MPHOSPH1 levels in different cancer cells, we employed quantitative real-time PCR to analyse the levels of MPHOSPH1 mRNA. We found that MPHOSPH1 was up-regulated not only in the bladder cancer cells (SCaBER and T24 cells), but also in many other types of cancer (HeLa, A549, BEL-7404 and SW620 cells); whereas it remained at low levels in normal cells (MRC-5 and L-02 cells; Fig. 2), which suggested MPHOSPH1 may be an ideal target in cancer gene therapy.

The new OV ZD55SP/E1A showed better replication ability and higher foreign gene delivery efficiency than ZD55
The OV ZD55SP/E1A is constructed based on the ZD55 vector [17]. In this study, the E1A promoter in the original wild-type serotype 5 Ad was replaced by the short 269 bp survivin promoter, and the E1B55KD was also deleted. Expression cassettes of both IL-24 and shMPP1 were then inserted into this new vector (Fig. 3A). We measured the relative replicating multiples of both oncolytic Ads ZD55SP/E1A and ZD55 by RT-PCR using the adenovirus gene E3 cDNA. Our result suggested that ZD55SP/E1A had a better replication ability than that of the ZD55 in tumour cells SW620 and HeLa, while in normal MRC-5 cells, both Ads showed low replication abilities (Fig. 3B). We also assessed the geneexpression ability of this new OV, the result is consistent with our replication assay results (Fig. 3B-E). ZD55SP/E1A-EGFP, an oncolytic adenovirus carrying EGFP reporter gene, showed much stronger fluorescence than ZD55-EGFP after infecting A549 and SW620 cells for 96 hrs, while very weak fluorescence was observed in infected normal MRC-5 cells (Fig. 3C-E). All these results suggested that compared with the ZD55 vector, the OV ZD55SP/E1A has comparable safety in normal cells, and much better replication ability and higher extraneous gene delivery efficiency in tumour cells.
We subsequently inserted both the IL-24 and shMPP1 expression cassettes into the ZD55SP/E1A OV, and the results on the identification of rec-Ads are provided in the Supporting Material. A total of 0.7 g of reporter plasmid together with 0.07 g of pCMV-␤-gal were co-transfected. Forty-eight hours later, the cells were harvested, and the luciferase activity was assayed by the dual-luciferase reporter assay system. For two Ads expressing the individual genes and the dual expressing construct, similar levels of IL-24 expression and inhibition of MPP1 were identified ( Fig. S2C and D).

ZD55SP/E1A-IL-24-shMPP1 specifically inhibits tumour cell proliferation
MPHOSPH1 has been reported as a potential oncogene, and knockdown of MPHOSPH1 suppressed the growth of bladder cancer cells [4]. Here, to evaluate the anti-tumour ability of the OV expressing shMPP1 in vitro, we measured the cytopathic effect and cell viability by crystal violet staining and MTT assay. We found that besides the bladder cancer cells (SCaBER, data not shown), a number of other cancer cells (SW620, A549 and BEL-7404) were also sensitive to ZD55SP/E1A-IL-24-shMPP1 infection, whereas the normal MRC-5 and BEAS-2B cells showed much lower sensitivity (Fig. 4A), demonstrating the safety of ZD55SP/E1A-IL-24-shMPP1 in normal cells.
Further MTT results revealed the oncolytic effect of different adenoviruses in details. As shown in Figure 4B, from day 3, a visible decrease of cell viability was observed in all of the oncolytic Ads treating groups compared with the control group; on day 4, the decrease was 17.1% in BEL-7404 cells and 38.8% in SW620 cells treated with ZD55SP/E1A, while the decrease was 64.9% in BEL-7404 cells and 82.4% in SW620 cells treated with ZD55SP/E1A-IL-24-shMPP1 (P Ͻ 0.001), suggesting ZD55SP/E1A-IL-24-shMPP1 has the highest anti-tumour activity among the four rec-Ads we constructed.
To validate that the cell death induced by the Ads was apoptosis mediated, we infected SW620 cells with various rec-Ads, and 48 hrs later, nuclear fragmentation and chromatin condensation was found in the rec-Ads-treated cells (Fig. 5D). FACS analysis was also performed 48 hrs after infection; SW620 cells were stained with annexin V-FITC and PI and evaluated for apoptosis. As shown in Figure 5E, the proportion of annexin V-positive SW620 cells infected with ZD55SP/E1A-IL-24-shMPP1 was 24.53%, much higher than the control (3.22%). We then applied JC-1 staining to examine the mitochondrial transmembrane electrical potential (⌬⌿m) of SW620 cells by FACS. We found that 72 hrs after ZD55SP/E1A-IL-24-shMPP1 infection, 35.49% of the cells were detected with altered ⌬⌿m (Fig. 5E). Therefore, we believed that ZD55SP/E1A-IL-24-shMPP1 may induce apoptosis via a mitochondria-dependent signalling pathway.
We further studied the levels of apoptosis-associated proteins after ZD55SP/E1A-IL-24-shMPP1 infection using Western blotting. As shown in Figure 5F, the cleavage products of Caspase-8, Caspase-9, Caspase-3 and PARP could be detected 48 hrs after ZD55SP/E1A-IL-24-shMPP1 infection, and by the time of 72 hrs, the cleavage products were enhanced, suggesting ZD55SP/E1A-IL-24-shMPP1 induces Caspase-8-and Caspase-9-mediated apoptosis. Another interesting result was that although all the rec-Ads can induce expression of cIAP, which can efficiently block TNF-mediated apoptosis [28,29], only ZD55SP/E1A-IL-24-shMPP1 infection significantly inhibited cIAP expression 72 hrs after infection (Fig. 5F). Thus, the suppression of the cIAP may be the underlying mechanism by which the combination of IL-24 and shMPP1 shows strong synergistic anti-tumour effects.

ZD55SP/E1A-IL-24-shMPP1 induces mitotic arrest and senescence
To further clarify the relationship between polyploid status and apoptosis induced by the rec-Ads, A549 cells were infected with ZD55SP/E1A-IL-24-shMPP1 at MOI of 5, and 48 hrs later, we observed a remarkable accumulation of cells with Ͼ4N DNA content by FACS. However, the sub-G1 peak was not distinct at 48 hrs, whereas 24 hrs later, the sub-G1 peak markedly increased from 1.53% to 26.04% with the peak of Ͼ4N visually decreased (Fig. 6A). This observation suggested that most of the polyploid cells induced by ZD55SP/E1A-IL-24-shMPP1 were undergoing apoptosis. We then used phase-contrast videomicroscopy to examine the behaviour of the A549 cells infected with ZD55SP/E1A-IL-24-shMPP1 at MOI of 5 over a 72-hr duration. Approximately 36 hrs after infection, we observed a high proportion of multinuclear cells, and during the subsequent 24 hrs, many multinuclear cells were observed to be in a prolonged M phase. And with the time passing, a large number of cells in mitotic arrest underwent apoptosis (Fig. 6B).
Since p53 is often associated with post-mitotic apoptosis in cancer cells [30], we performed Western blot to investigate p53 and its downstream target genes of infected A549 cells, in which p53 was reported as wild type without any mutation [31]. As shown in Figure 6C, p53 was detected 24 hrs after infection, and enhanced expression of BTG2, Puma and Bax were observed as well. This result suggested that ZD55SP/E1A-IL-24-shMPP1 induced apoptosis after mitotic arrest through a p53 signalling pathway. In addition, no up-regulation of Mad2, a critical SAC triggering protein [32,33], was found during the mitotic arrest (Fig. 6C), suggesting SAC was not activated by the infection of ZD55SP/E1A-IL-24-shMPP1, and similar results were also observed in SW620 cells (data not shown).
We also found that compared with other rec-Ads, only the ZD55SP/E1A-IL-24-shMPP1 infection at low MOI induced senes-cence, a morphological change different from apoptosis or cytopathic effect. We used the SA-␤-gal staining assay [34] to detect senescence in A549 cells induced by ZD55SP/E1A-IL-24-shMPP1. Senescence was observed at 0.5 MOI, but when the MOI was raised to 2 or higher, only apoptosis or cytopathic effect was detected (Fig. 6D), while no senescence was observed in other rec-Ads-treated cells (data not shown) suggest that even at low  MOI, ZD55SP/E1A-IL-24-shMPP1 could still inhibit tumour cells growth through triggering senescence. Western blot analysis revealed that p53 was also induced by the infection of ZD55SP/E1A-IL-24-shMPP1 at low MOI in A549 cells (Fig. 6E). The expression of cell cycle inhibition genes, p21 and p27, were up-regulated by infection as well, implying the cell cycle arrest caused by the infection of ZD55SP/E1A-IL-24-shMPP1 at low MOI, which thus further triggered senescence.

ZD55SP/E1A-IL-24-shMPP1 inhibits tumour growth in vivo
The in vivo oncolytic potential of ZD55SP/E1A-IL-24-shMPP1 was evaluated in a SW620 tumour xenografts model. As shown in Figure 7A, in PBS-treated group, the tumours grew rapidly and reached approximately 2100 mm 3 within 3 weeks, suggesting the high malignancy of the SW620 cells used in this study. However, the growth of ZD55SP/E1A-IL-24-shMPP1-treated tumours was obviously suppressed. Two months after treatment, the average volume of the ZD55SP/E1A-IL-24-shMPP1-treated tumours was 576 mm 3 , significantly smaller than those with other rec-Ads treatments (P Ͻ 0.001; Fig. 7A) [21].
Immunohistochemistry was also used to evaluate the effect of the Ads on IL-24 and MPHOSPH1 expression in vivo. Seven days after treatment, the tumour specimens were separated from tumour-bearing mice. As shown in Fig. 7B, the MPHOSPH1 expression levels in the tumours treated with ZD55SP/E1A-IL-24-shMPP1 and ZD55SP/E1A-shMPP1 were markedly lower than those treated with PBS or ZD55SP/E1A; and IL-24 was also detected in tumours treated with ZD55SP/E1A-IL-24 and ZD55SP/E1A-IL-24-shMPP1. In all rec-Ads-treated tumours, the TUNEL analysis showed the cancer cells were undergoing apoptosis; in contrast, the PBS-treated tumour cells showed no signs of apoptosis. We thus deduced that ZD55SP/E1A-IL-24-shMPP1 may exert anti-tumour effect in vivo by combining the effects of up-regulation of IL-24, down-regulation of MPHOSPH1, and the oncolytic activity of the oncolytic vector, which subsequently induced apoptosis in tumour cells.

Discussion
Survivin has been reported as a cancer-specific up-regulated gene, and we had previously used its 1.1-kb promoter to construct a cancer targeted OV [26,27]. To shorten its length and increase the capacity of OV in this study, we cloned a 269-bp survivin promoter inspired by the work of Xu et al. that identified this 269-bp protomer fragment has the strongest transcriptional activity within the full-length survivin promotor [25], and investigated its promoter properties. Our results revealed that the 269-bp survivin promoter had a remarkable tumour-selective transcriptional activity, and compared with the 1.1-kb promoter, the 269-bp promoter mediated much stronger reporter expression in tumour cells (data not shown). Because of its multiple advantages, for example, the small size, strict and strong tumour-specific transcriptional activity, the 269-bp survivin promoter is an ideal tumour selective target for cancer gene therapy.
We have previously reported that the CTGVT enhanced the antitumour effect of an oncolytic adenovirus by amplifying the tumour suppressor gene it carries during replication [17,23,35]. An appropriate tumour suppressor gene is critical for the success of the CTGVT, in this study, we investigated the anti-tumour effects of co-expression of two tumour suppressor genes, IL-24 and shMPP1, with our newly constructed ZD55SP/E1A vector. Our previous results that IL-24 an excellent candidate of anti-oncogene carried by oncolytic Ads [5,6]. Another potential benefit of using IL-24 in this study is that the interferon effects, which can be induced by Ad infection and IL-24 treatment [36,37], can up-regulate IL-24 and might further enhance anti-tumour effects of the oncolytic Ads [38,39], and it will be of great interest to investigate this interferon effects in detail in the future. Compared to IL-24, MPHOSPH1 was only recently identified in bladder cancer as a novel oncogene [4]. Here, we demonstrated that MPHOSPH1 was up-regulated in many types of cancer cells, such as cells of liver cancer, lung cancer, colorectal cancer and ovarian cancer; moreover, we also targeted MPHOSPH1 for the first time in cancer gene therapy and achieved satisfying tumour suppression effects by inducing mitotic arrest and apoptosis of tumour cells.
Another interesting finding is that mitotic arrest induced by MPHOSPH1 knockdown did not up-regulate Mad2 that triggers SAC activation [32,33], which is normally associated with the vinca alkaloids or taxanes treatment [1]. A possible explanation is that, rather than the neuron-microtubulal perturbation side effect of alkaloids or taxanes treatment [2], MPHOSPH1 knockdown may exert effect on cytokinesis and trigger mitosis arrest without disturbing the cell spindle assembly. Thus, comparing with the traditional vinca alkaloids or taxanes treatment, MPHOSPH1 suppression may represent a better anti-mitotic cancer therapy strategy. However, more evidence is needed to test this hypothesis.
We also noted that although the MPHOSPH1 knockdown (by ZD55SP/E1A-shMPP1 infection) induced significant polyploidy in tumour cells, the further combination of IL-24 (ZD55SP/E1A-IL-24-shMPP1) significantly raised the anti-tumour activity (Fig. 6A). According to our Western blotting results, the ZD55SP/E1A-IL-24-shMPP1 was the only Ad resulted in the down-regulation of cIAP (Fig. 5F); therefore, IL-24 combined with shMPP1 may enhance anti-tumour effect by suppressing cIAP. It has been reported that cIAP can be induced upon virus infection [40,41], we also found here that adenovirus infection induced cIAP expression; because cIAP is believed to protect the infected cells via blocking apoptosis [40,41], thus it may explain the weak anti-tumour effects of OVs respectively harbouring IL-24 or shMPP1 alone. It will be of our future interests to study the relationship between IL-24 and MPHOSPH1. Our animal studies provide evidence that the combination of anti-mitotic MPHOSPH1 and IL-24 genes with CTGVT is a promising strategy for cancer gene therapeutics. Immunohistochemistry study showed that 30-50% of the tumour cells stained brown by anti-hexon antibody (data not shown), which may be an index of infected cell number in vivo. Immunohistochemistry assay also indicated that three functions including viral oncolysis, IL-24 tumour suppression and MPHOSPH1 inhibition are involved in the apoptosis of tumour cells in vivo (Fig. 7B). Although the viral oncolysis and the single tumour suppressor gene expression both inhibit the xenograft tumours growth, the combined dual-construct nevertheless showed the highest anti-tumour ability in vivo (Fig. 7A).
Much work has been done to study the relationship between the cellular genome and cancer [42][43][44][45], however, it remains in debate whether the genome change of cell promotes or suppresses tumourigenesis. In this study, we found that the polyploidy induced by ZD55SP/E1A-IL-24-shMPP1 actually inhibited tumourigenesis through apoptosis. Because observed tumourigenesis suppression were obtained by oncolytic adenovirus infection, which can cause oncolytic effects by itself, it is thus possible that under other conditions, the cell genome change (aneuploidy or polyploidy) may be found to induce tumourigenesis. Our results may provide a new approach on this research.
Another interesting question raised regards how the infected cells with Ͼ4N DNA content came into being (Fig. 6A). A possible interpretation is that at anaphase or telophase, the cells with 4N DNA content may escape from the mitotic arrest induced by ZD55SP/E1A-IL-24-shMPP1 infection and exit mitosis without division (mitotic slippage), and subsequently initiate the next round of cell cycle and enter the S phase, at which their DNA is replicated (DNA content increased). Thus, the observed apoptosis of cells with Ͼ4N DNA may represent a type of post-mitotic apoptosis. We also found that p53, which promotes post-mitotic apoptosis due to its accumulation in the prolonged mitotic arrest [30], was involved in the apoptosis and senescence of the mitosisarrested tumour cells induced by infection of ZD55SP/E1A-IL-24-shMPP1. Furthermore, it is interesting to note that infections of different MOI led to the different cellular changes (apoptosis or senescence) in this study. We believe a possible explanation is that the infection results depend on the initially infected cell number. When the initial percentage of infected cells is low, the infection might result in chronic effect such as senescence, whereas for high percentage of infected cells, acute effects such as CPE or apoptosis might be caused, although our results suggest that both low and high MOI infections will lead to more than 90% of the cells infected eventually (Fig. 3). However, further investigations are still in need.  Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.