Silencing RNA: a novel treatment for pancreatic cancer?

The antiapoptotic gene Bcl-2 has been a target for downregulation by nucleic acid based strategies for more than a decade but the recent failure of the synthetic antisense oligonucleotide agent Genasense in phase III clinical trial caused many to think again about the worth of this approach. However, a new optimism about this and other targets is spreading through the community following several encouraging applications of the recently discovered technology of RNA interference, which is essentially a new biological version of the antisense system.¹

RNA interference is considered to have begun as an evolutionarily ancient mechanism for protecting organisms from viruses. Many viruses have RNA, rather than DNA, as their genetic material and go through at least one stage in their life cycle in which they make double stranded RNA. Perhaps not surprisingly, all multicellular organisms have evolved a well conserved protein apparatus that destroys double stranded RNA but this has also been found to play a role in maintenance of the organism’s own genome stability by suppressing the movement of mobile genetic elements, such as transposons and repetitive sequences.

The gene silencing process of RNA interference (RNAi) involves the manufacture of short double stranded RNA molecules by an enzyme called DICER, which cleaves RNA duplexes into 21–26 base pair oligomers. These small interfering RNAs (siRNA) cause sequence specific, post-transcriptional gene silencing by guiding an endonuclease, the RNAi induced silencing complex (RISC), to mRNA. This process has been seen in a wide range of organisms such as Neurospora fungus (in which it is known as quelling), plants (post-transcriptional gene silencing), and mammalian cells (RNAi). Downregulation of target gene expression has been found to involve interactions at multiple levels. Where there is complete or near complete sequence complementarity between the small RNA and the target, the Argonaute 2 component of RISC mediates cleavage of the target transcript.^2,3 In contrast, where there is sequence mismatch between the miRNA and the target transcript, the mechanism appears to involve repression of translation predominantly.⁴ More recently, it has been recognised that siRNA molecules can induce transcriptional silencing through promoter methylation.^5,6

In principle, the high sequence specificity of RNA interference might make it suitable to treat disease that is linked to selective or elevated expression of particular identified genes. This may make it particularly appropriate for combating cancers associated with mutated endogenous gene sequences. An early example of the potential power of this approach came in a study of pancreatic cancer. RAS genes are frequently mutated in human cancers, particularly in pancreatic and colon carcinomas. Mutant RAS oncogenes often contain point mutations that alter only a single amino acid, which locks the oncogenic RAS proteins in a persistently activated GTP bound state. A complication in using RAS oncogenes as targets in anticancer therapy is that at present it is not possible specifically to inhibit the biochemical function of only the oncogenic RAS alleles. This may be essential as the wild-type K-RAS gene appears to be required for viability, as evidenced by the embryonic lethal phenotype of mice nullizygous for K-ras.⁷ Retroviral delivery of siRNAs can specifically inhibit the mutant K-RASV12 allele in human pancreatic carcinoma cells, while leaving the wild-type K-RAS allele untouched.⁸ In spite of the fact that pancreatic carcinoma cells have many genetic alterations, loss of K-RASV12 expression leads to loss of tumorigenicity in experimental animal models.

In this issue of Gut, Ocker and colleagues⁹ explore the use of siRNAs against another gene that is aberrantly expressed at high frequency in pancreatic cancer, the antiapoptotic gene bcl-2 (see page 1298). Their results suggest that the target can be selectively downregulated in tumour cells in vitro and that intraperitoneal administration of the naked nucleic acid agent can produce variable antitumour effects against malignant deposits growing subcutaneously in vivo. In this study—as in most experiments using RNAi to target particular genes in mammalian cells—the results are interpreted as representing induction of sequence specific transcript cleavage. However, at this early stage in our understanding of RNAi, it is important not to rule out the possibility that interference mediated through protein translational repression or genomic modification (DNA methylation or histone modification) may also be playing a role in mediating gene specific silencing and any derived RNAi phenotype. The great attraction of therapeutic epigenetic gene specific silencing lies in its heritable nature, meaning that, unlike post-transcriptional gene silencing that requires the continued presence of an siRNA molecule targeting a coding sequence, long lasting suppression of gene expression could be achieved from a single exposure to a specific methylation inducing RNAi agent targeting a promoter sequence.

The major challenge in turning RNA interference into an effective therapeutic strategy is the delivery of RNA interference agents, whether they are synthetic, short double stranded RNAs (as in the paper by Ocker and colleagues⁹) or viral vectors directing production of double stranded RNA, to the target cells within the body. While siRNA technology has proven extremely powerful and robust for cell culture work, translating this success reliably to animals or humans is proving very difficult, due to insufficient bioavailability of the compounds. However, an important step in the right direction is that Jürgen Soutschek and colleagues have recently been able to demonstrate siRNA mediated downregulation of apolipoprotein B in the liver (and jejunum) of mice using cholesterol conjugates delivered systemically.¹⁰ The effects were preferentially seen in the liver, which is a relatively easy organ to target, and relatively high dosages (three injections each of 50 mg/kg) were required for the effect. In view of the extremely high potency of the siRNA in in vitro cell cultures, one must conclude that only a very small fraction of the injected siRNA actually reaches its molecular mRNA target in liver cells. Thus it is unfortunately not likely that simple cholesterol conjugation will solve the general delivery problem of siRNA. Other cationic cell penetrating peptides such as penetratin, Tat, and more recently transportan and oligo arginine have been proposed as general transmembrane carriers for a variety of cargoes, including oligonucleotides and PNA.^11,12 However, it appears that the main uptake route for most, if not all, of these peptides is endosomal, and thus the reagents have to escape the endosomal compartment in order to enter the cellular compartments of action: the cytoplasm and/or nucleus. The problem of cellular delivery is yet more complex for clinical application where the real challenge for an adjuvant therapy agent is delivery and maintenance of the compound in cancer cells in multiple organs in humans. However, the biotechnology industry has readily recognised the potential silencing properties of RNA mediated interference, with at least 15 companies active in the field, and two companies (Ribopharma and Benitec) have patents for RNAi based clinical applications. Significant progress in delivery technology is required before the concept of RNAi can realistically benefit cancer patients, but exploitation of the decade of clinical experience of antisense and viral gene therapy agents gives investigators a head start.