RNAi-induced synthetic lethality in cancer therapy

Cancer is caused by genetic or epigenetic alterations that deregulate multiple oncogenes and tumor suppressor genes and progressively drive benign cells toward malignancy.¹ Despite the complicated deregulation of network signaling pathways, the proliferation and survival of cancer cells may be dependent upon a specific activated oncogene product. This phenomena, called “oncogene addiction,” has been targeted as a cancer cell’s “Achilles heel” to develop cancer therapies.² Using specific antibodies or small-molecule inhibitors to switch off the key pathway that the cancer cells depend upon will lead to cancer cell death while sparing normal cells. Several targeted therapeutic agents identified in this way have shown remarkable clinical efficacy. Among these agents are imatinib, which targets the oncogene BCR/ABL protein in chronic myeloid leukemia;³ gefitinib and erlotinib, which target the epidermal growth factor receptor (EGFR) in lung cancer;⁴ and trastuzimab, which targets HER2 in breast cancer.⁵ However, one of the limitations of targeted therapy is the cancer’s acquired resistance to a drug via complicated mechanisms such as a second genetic mutations and alternative pathway compensation by gene amplification. For example, Gorre et al. reported that some leukemia patients who initially responded to imatinib treatment later developed resistance to it owing to a second mutation in the kinase domain of the BCR/ABL protein that completely inhibited the function of imatinib.⁶ Similarly, in another report a second point mutation in the kinase domain of EGFR caused acquired resistance of non-small cell lung cancer to gefitinib therapy.⁷ In addition, Engleman et al. reported that amplification of the MET tyrosine kinase gene was responsible for lung cancer’s acquired resistance to treatment with EGFR tyrosine kinase inhibitor (TKI).⁸ Clearly, the adaptation of a cancer cell to selection stress from the targeted drug will be a hurdle for targeted therapy.

One idea proposed to overcome some of the problems associated with targeting oncogenes and tumor suppressor genes in cancer therapy involves the application of “synthetic lethality.” Two genes may be considered to have a synthetically lethal relationship when a mutation in either of the two genes alone has no effect on cell survival but when mutations in both genes at the same time cause cell death.⁹ Synthetic lethality was first studied in Drosophilia and yeast and later in human cells. Instead of trying to block an activated oncogenic pathway or restore a mutated tumor suppressor gene, synthetic lethality uses the tumor cell’s own genetic or metabolic changes to kill the cell. For example, an antibody activating the DR5 “death receptor” caused apoptosis in multiple human cells that overexpressed the Myc oncogene, indicating that activated Myc and DR5 are synthetically lethal.¹⁰ This concept now has been widely exploited to identify gene targets for cancer therapy.

Recently, large-scale synthetic lethality RNA interference (RNAi) library screening has identified multiple potential targets and gene interactions that can be exploited for clinical cancer treatment. Through genome wide RNAi library screening, Luo et al. identified that anaphase-promoting complex/cyclosome (APC/C) and Pololike kinase (PLK) are synthetically lethal with the RAS oncogene in colorectal cancer cell lines.¹¹ Similarly, Scholl et al. found that the STK33 gene has synthetic lethality with a RAS mutation in multiple cancer cell lines from different tumor types.¹² In targeted cancer therapy, most agents target membrane receptors, kinases, and oncogenes; however, large-scale synthetic lethality screening has identified many non-oncogenes that are synthetic lethal to the cancer cells indicating that those genes can also be targeted for cancer therapy.¹³ Agents that normally target kinases, membrane receptors, or oncogenes will not be applicable for targeting these “undruggable” synthetic lethality genes or gene-gene interactions.¹³

One of the major challenges in using RNAi in cancer treatment is the effective intracellular delivery of siRNA molecules in vivo. Remarkable progress has been made in delivering siRNA in vivo by chemically modifying the RNA duplex or conjugating the RNA with small molecules or peptides and by using specific formulations with liposomes or nanoparticles to increase the siRNA’s stability and improve targeted delivery.¹⁴ Preclinical studies have demonstrated effective responses to non-viral and viral RNAi molecules in disease models.¹⁴

In this issue of Cancer Biology & Therapy, Michiue et al. have demonstrated that synthetic lethality responses can be induced in vivo by using RNAi molecules targeting EGFRvIII and AKT.¹⁵ Amplification or truncated mutation of EGFR and hyperactivation of AKT play a major role in the development of glioblastoma, one of the deadliest malignancies.¹⁶ Agents targeting those molecules and related pathways have been actively studied and clinically tested.¹⁷ Using RNAi technology to suppress targeted gene expression is not new; however, Michiue and colleagues’ work targeting two molecules at the same time in vivo to induce synthetic lethality is the first such report in the literature. Previously, investigators from the same laboratory developed an efficient siRNA delivery approach that fused a dsRNA Binding Domain (DRBD) with a TAT Peptide transduction Domain (PTD) delivery peptide, thus enabling a much stabler and more efficient delivery of siRNAs into cells.^18,19 With this innovative siRNA delivery system, Michiue et al. have successfully suppressed targeted gene expression in vitro and in vivo. Of even greater interest, their results showed that knockdown of either the EGFRvIII gene or AKT2 alone only moderately inhibited tumor growth, whereas suppression of both EGFRvIII and AKT2 significantly inhibited tumor growth by inducing cancer cell apoptosis and prolonged the median survival duration of tumor-bearing mice from 14 to 31.5 d (Fig. 1). This study demonstrated for the first time that selectively targeting more than one molecule to induce a synthetic lethal response in vivo is feasible and effective for cancer therapy in an animal tumor model.

Figure 1.

Diaphragmatic illustration of synthetic lethality relation with EGFRvIII and AKT2.

The study of Michiue et al. has left some questions unanswered. First, the authors did not clarify whether knocking down expression of EGFRvIII will suppress AKT activity. Considering that AKT is one of the major downstream molecules of the EGFR/PI3K pathway, AKT activity should be checked when EGFRvIII is knocked down. Second, the authors did not explain why AKT2 but not AKT1 or AKT3 is specifically synthetically lethal with EGFRvIII, since these three molecules thus far are supposed to be functionally similar. It is not clear whether the synthetic lethality between EGFRvIII and AKT2 is cell type-specific or dependent on any specific cellular genetic background, for example mutations in p53 or PTEN, which both have a higher frequency of mutation in brain tumors. Testing this synthetic lethality in multiple glioblastoma cell lines with different genetic backgrounds will help answer these questions. Further studies are also needed to test whether different dosages of RNAi molecules will result in different effects in vivo, and the RNAi off-target effects and related toxicity also should be monitored.²⁰

Nevertheless, the article of Michiue et al. develops a very important concept of in vivo synthetic lethality. So far, large-scale synthetic lethality RNAi screenings are performed in vitro, but all the targets identified in vitro will eventually need to be tested in vivo. Further improvements in the approach to delivery of RNAi molecules in vivo have great potential to advance drug development and cancer treatment.