StatinAE gene transfer using a Sleeping Beauty transposon vector resulted in antitumoral effects in an animal model of colorectal cancer metastatic to the liver. Significant tumor growth inhibition was seen in mice injected with a high (25 μg) dose of antiangiogenic transposon plasmid, in which co-injection of the SB transposase did not achieve any further tumor regression. In contrast, antitumor activity of the Statin AE transposon administered at a lower dose (10 μg) was dependent on co-infusion of SB transposase-encoding plasmid. These results suggest that antiangiogenic gene therapy using the SB non-viral transposon system has the potential to be an effective treatment for colorectal cancer metastatic to the liver, and that this process is dependant on the transposition process under transposon dose-limiting conditions.
Although antiangiogenic proteins show great promise in preclinical cancer models, their effectiveness in the clinic, especially when administered alone, has been limited [34
]. The effect of VEGF on tumor and vascularization has been studied by many groups, demonstrating it's pivotal role in increasing vascular permeability, tumor growth and metastasis [37
]. In pre-clinical studies, VEGF targeting strategies have demonstrated significant antivascular effects and tumor growth inhibition. Although several drugs and small molecule inhibitors of VEGF are being clinically tested, their efficacy as monotherapeutic agents in advanced stage disease has been discouraging, due to short half-lives and dose limiting toxicities [39
]. However, there is a significant survival benefit when VEGF-targeted therapy is combined with standard chemotherapy for metastatic colorectal cancer [36
]. Also, most inhibitors of angiogenesis exert their effect against newly formed blood vessels rather than existing vasculature. The inability of these agents to completely eradicate disease presents a serious challenge for patients with advanced malignant tumors. For most angiogenesis inhibitors, tumor growth resumes after cessation of therapy [39
]. This necessitates administration of high doses and prolonged treatment with antiangiogenic proteins to obtain a sustained therapeutic effect.
Delivery of gene sequences encoding recombinant inhibitors of angiogenesis potentially allows for localized and sustained expression, which could reduce the risk of systemic toxicity and perhaps increase efficacy. Other advantages include the possibility of expressing multiple antiangiogenic gene products that act via different mechanisms, thereby increasing the effectiveness of the therapy. The antitumor effects of introducing gene sequences encoding endogenous anti-angiogenic proteins endostatin and angiostatin have been characterized in several studies, and their synergistic effects when used in combination (either in individual or separate vectors) is well documented [29
]. We therefore investigated an angiostatin-endostatin fusion protein previously shown to confer potent antitumor effects in a subcutaneous melanoma model in vivo [32
The integrating Sleeping Beauty (SB) transposon system circumvents the primary limitation of non-viral plasmid based gene transfer, i.e. the transient duration of gene expression that fails to give way to long-term expression. SB-mediated transposition has been shown to occur in a variety of cultured cell types, in zebrafish [41
] and mouse embryos and germ cells [42
], in human primary blood lymphocytes [26
], and in mouse somatic tissues, including the lung and the liver [25
]. SB-mediated transposition in mouse liver has been verified in several laboratories by recovery and sequencing of transposon-chromosome junction sequences [22
SB has also been used successfully to deliver antitumor genes to neoplastic tissue. Ohlfest et al successfully used the SB transposon system to deliver a cocktail of antiangiogenic genes to human glioblastoma xenografts in mice, subsequently observing increased survival and sustained regression of tumor [27
]. Wu et al compared the efficacy of interferon-gamma immunogene therapy using non-integrating plasmid vectors vs SB plasmid vectors in a syngeneic glioma model. Only animals co-injected with SB transposase plasmid exhibited prolonged expression of interferon-gamma and a significant increase in survival (3 weeks), while expression in animals treated with transposon plasmid alone was undetectable after 1 week [53
While hydrodynamic tail vein injection has been used very effectively for delivering DNA to liver tissue in rodents, extending this technique to humans is still implausible. However, several laboratories have reported the use of balloon occlusion catheters for successful gene delivery to the liver in large animals. Delivery of DNA into occluded rabbit liver under X-ray guidance [54
], into the left lateral lobe of pig liver by catheterization and occlusion of the portal vein [55
], and to the whole liver of pig and dog via the inferior vena cava (IVC) with double balloon occlusion above and below the IVC-hepatic vein conjunctions [56
] have been reported. Liu et al have developed a device that uses high pressure from a gas cylinder and a computer-controlled switch to drive and regulate DNA injection in pigs [58
]. Use of this device, combined with vessel occlusion and image guided catheterization to achieve site specificity, was found to provide effective gene delivery [59
]. Overall, these results from large animal studies demonstrate that modifications of the hydrodynamic technique can potentially be applied to humans.
Our results demonstrate that animals injected with both transposon and transposase-encoding plasmids survive significantly longer than untreated control animals, or animals treated with transposon plasmid alone. However, transgene expression declines over time, leading to eventual emergence of tumor metastases. Transgene-specific immune responses, both humoral and cell-mediated, have been described. Aronovich et al have shown that prolonged expression of ß-glucuronidase after SB transposon-mediated delivery elicited an immune response against transgene-expressing cells, which were subsequently eliminated (22). Lutzko et al have characterized humoral immune responses against transgene products by ELISA, and have demonstrated cellular immune responses using lymphocyte proliferation assays [60
]. Relative persistence of transgene expression is also mouse-strain dependant. Injection of recombinant adenovirus expressing human alpha-1-antitrypsin (hAAT), resulted in persistent, circulating levels of hAAT in C57Bl/6 mice, while Balb/c mice rapidly neutralized the transgene product [61
]. Hodges et al have shown that hydrodynamic injection of CpG replete, supercoiled human Factor IX encoding plasmid in Balb/c mice resulted in loss of transgene expression after 3-4 weeks, while therapeutic levels of Factor IX expression were maintained in mice that received CpG depleted plasmid [62
]. Any or all of these factors could account for the decline in gene expression observed in our study. Modulation or counteraction of the immune response is therefore essential in maintaining sustained transgene expression and continued suppression of tumor metastases.
Here, we have shown that the SB transposon system can be used to successfully deliver antiangiogenic genes and inhibit colorectal tumors metastatic to the liver. We observed that treatment with a high dose of antiangiogenic transposon plasmid appeared to be effective with no requirement for co-delivery of SB transposase. However, antitumor effectiveness of Statin-AE transposon when administered at low dose (10 μg) was dependent on SB transposase. Since transposon delivery in large animals and in humans is likely to be infrequent, the low dose treatment is considered to be more representative of the low level of gene transfer that is to be expected in a clinical setting. In this case, co-delivery of the SB transposase encoding plasmid is required, thereby enabling sustained expression of the therapeutic antiangiogenic transgene through transposition, and thus continued inhibition of tumor growth.