Anti-tumor therapy with an adeno-associated virus (AAV) suppressing the KRAS oncogene by RNA interference.

Mortality in colorectal cancer is mostly due to liver metastases. Surgical resection of liver tumors offers the only hope for cure, but only 10-20% of the patients are eligible for surgery. Therefore, alternative treatment modalities are urgently needed. Tumor formation is caused by a series of genetic changes resulting in the activation of proto-oncogenes and the inactivation of tumor suppressor genes. One of the most common genetic changes in human tumors is the mutational activation of the KRAS proto-oncogene. Activating mutations in KRAS are observed in 35% of sporadic colorectal tumors [1-3]. Therapeutics that are designed to interfere with Ras function have been developed and tested in clinical trials [4]. They include inhibitors of RAS farnesylation (FTI's) and inhibitors of the RAS downstream targets RAF, MEK and mTOR. However, none of these inhibitors selectively targets KRAS. Selective inhibition of mutant KRAS can be achieved by sequence-specific RNA interference (RNAi). Problem definition Activating mutations in KRAS are acquired at the very early pre-malignant stages of tumor development. An important prerequisite for using KRAS as a target for therapy is that also late metastatic tumors should be continuously dependent on KRAS for their maintenance and outgrowth. Although suppression of endogenous mutant KRAS can reduce the tumorigenicity of human tumor cell lines [5-8], it remains unclear to what extent metastatic human CRC tumors are dependent on the continued presence of mutant KRAS. This is especially relevant in light of the disappointing results of the above-mentioned trials [4]. Although RNA interference is now widely used to study gene function in vitro, it is less clear to what extent it is therapeutically applicable [9,10]. The successful application of RNAi in anti-tumor therapy will be critically dependent on the efficient delivery of the therapeutic nucleic acid to tumor cells in vivo, and on the efficiency and persistence of gene knockdown. Prior work Our group has recently shown that murine C26 colorectal carcinoma cells contain endogenous mutant Kras-D12 and that stable suppression of this allele by RNA interference reverts the transformed properties of C26 cells. Kras knockdown cells still form tumors but these elicit an immune repsonse which results in tumor regression and subsequent protection of the mice against wildtype tumor growth [11]. Furthermore, Kras knockdown completely abrogates the potential of C26 cells to grow out as liver metastases. Through our collaboration with Prof. Hoeben and Dr. de Vries, we have ample experience in constructing and employing AAV, adenovirus, and AAV-adenovirus hybrid vectors for gene delivery purposes [12-14]. Objective In the current project proposal we will develop RNAi vectors, designed to suppress mutant Kras/KRAS, that efficiently and selectively transduce experimental and human colorectal tumors. In addition, we aim to assess whether these vectors should be considered as an alternative/additive strategy for the treatment of Kras/KRAS-dependent tumors. Strategy We propose to realize our objective by developing AAV-based RNAi vectors. The choice of AAV as an RNAi vector was based on the following considerations: First, AAV has a stable capsid in vivo, displays low immunogenicity and has an excellent safety profile [15]. Second, AAV-mediated RNAi can produce efficient and persistent gene suppression in several in vitro and in vivo model systems (reviewed in [16]). We will select AAV-2 vectors that efficiently transduce experimental and human CRC tumors, using a library of ~4x10e6 mutant AAV-2 cap-protein variants [17]. The selectivity and efficiency of tumor cell infection by the selected mutant viruses will be assessed in a mouse model of Kras-dependent colorectal tumor growth [11]. In addition, the anti-tumor efficacy of AAV-mediated Kras knockdown will be determined. Concurrently, we will evaluate the efficiency of AAV-mediated KRAS knockdown in human CRC tumor explants, grown in vitro and in immune-deficient mice. The anti-tumor efficacy of this treatment will subsequently be assessed. 35% of CRC tumors harbor activating mutations in KRAS and are therefore potentially sensitive to AAV-mediated KRAS knockdown. Novelty Our approach to test the dependency of human CRC tumors on the continued presence of the KRAS oncogene by using fresh tumor explants in in vitro and in vivo systems has, to our knowledge, never been reported. If our approach is successful it can be used for testing the dependency of human colorectal tumor tissue on any target gene of interest. Our approach to use an existing AAV-2 library to select AAV-mutants for their ability to target fresh human tumor tissue is novel and may lead to the development of AAV vectors that selectively and efficiently transduce human CRC tissue. We will use these vectors to deliver KRAS directed shRNA expression cassettes, but they may be used in many additional applications.