Loss-of-function genetic screens as a tool to improve the diagnosis and treatment of cancer

Abstract

A major impediment to the effective treatment of cancer is the molecular heterogeneity of the disease, which is also reflected in an equally diverse pattern of clinical responses to therapy. Currently, only few drugs are available that can be used safely and effectively to treat cancer. To improve this situation, the development of novel and highly specific targets for therapy is of utmost importance. Possibly even more importantly, we need better tools to predict which patients will respond to specific therapies. Such drug response biomarkers will be instrumental to individualize the therapy of patients having seemingly similar cancers. In this study, we discuss how RNA interference-based genetic screens can be used to address these two pressing needs in the care for cancer patients.

The need for new drug targets and better biomarkers of drug response

Traditionally, new cancer drugs have been identified mainly through empirical approaches (Chabner and Roberts, 2005). This has led to a generation of chemotherapeutic agents that is relatively nonspecific in targeting cancer cells, yielding considerable side effects. Such conventional therapies often benefit only a minority of patients, because of both intrinsic and acquired drug resistance. Another major factor limiting the effective treatment of cancer is the molecular heterogeneity of the disease. Because of this heterogeneity only few cancer types have a common ‘driver’ mutation: a genetic alteration that is instrumental for the malignant phenotype of the cancer. One such recurrent driver mutation in chronic myelogenous leukemia is the BCR-ABL fusion gene, which is targeted by the drug imatinib mesylate (de Klein et al., 1982; Druker et al., 1996). Imatinib mesylate is a member of a new class of anticancer drugs, the ‘targeted-therapeutics’ (Sawyers, 2004; Luo et al., 2009b). These drugs are developed based on knowledge of which genes (often kinases) are specifically altered in cancer. Therefore, such targeted therapies are highly cancer cell-selective and have fewer side effects. Unfortunately, most cancer types do not have such frequently occurring oncogenic driver event, making it difficult to expand on the success of this highly effective drug. Consequently, the arsenal of targeted cancer therapeutics is small, making the more conventional chemotherapy still the mainstay of treatment in today's cancer clinic. In this review, we will discuss how loss-of-function genetic screens in mammalian cells using RNA interference can contribute to the identification of completely new classes of highly selective cancer drug targets.

A second major obstacle in the treatment of cancer is the unpredictable response to therapy. For instance, in breast cancer, only 1 in 30 post-menopausal women benefit from chemotherapy (EBCTCG, 2005). Availability of biomarkers of drug responses could greatly help in the development of a more personalized approach to cancer treatment, in which patients having seemingly similar cancers can be pre-selected for specific therapies based on their predicted responses to therapy. In this study too, RNA interference-based genetic screens can be very efficient in identifying genes that control how cells respond to cancer drugs. Consequently, such genes are prime candidate biomarkers to foretell drug responsiveness in the clinic. Strategies to discover such biomarkers through RNAi-based genetic screens are discussed here also.

Loss-of-function genetic screening tools

In recent years, an array of new high-throughput technologies (such as DNA sequencing, single-nucleotide polymorphism genotyping, comparative genomic hybridization, proteomics and gene expression micro-arrays) have all been implemented in drug development efforts. A drawback of most of these approaches is that the data generated are mostly correlative and therefore do not directly identify the driver event among the many genetic alterations present in each cancer. Loss-of-function genetic screens on the other hand can also be carried out in a high throughput manner and, as a consequence of the functional nature of the approach, leads to the identification of causal factors only.

Two complementary types of functional genetic screens can be carried out (Brummelkamp and Bernards, 2003; Grimm, 2004). Gain-of-function genetic screens involve the ectopic expression of genes. This is often brought about by expression of collections of complementary DNAs (cDNAs). Such gain-of-function screens are relatively easy to perform and can be carried out using heterologous cDNA collections (that is, cDNAs from another species) in simple model organisms or by ectopic expression of homologous cDNAs in cells from higher organisms. For discovery of drug targets, however, loss-of-function screens are more suitable as the genetic event mimics the intended effect of the cancer drug: reduced gene product activity. In mammalian cells high throughput loss-of-function screens were not feasible for a long time. This situation changed in 2001, when it was discovered that RNA interference (RNAi) can also be used in mammalian cells to suppress gene expression (Elbashir et al., 2001). As many different collections of RNAi resources have been created that cover the entire human and mouse genomes (Brummelkamp et al., 2002a; Brummelkamp and Bernards, 2003; Grimm, 2004; Moffat and Sabatini, 2006; Iorns et al., 2007). Different reagents can be used to trigger RNA interference: the original synthetic short duplex RNAs and vector-encoded short hairpin RNAs (shRNAs (Brummelkamp et al., 2002a)). These two RNAi reagents each come in several flavors. For example, apart from the chemically synthesized small interference RNAs (siRNAs), siRNAs have been made by nuclease cleavage of long in vitro transcribed double-stranded RNAs: esiRNAs (Buchholz et al., 2006). For plasmid-derived shRNAs different viral delivery systems are available (for example, Moloney virus-based vectors, lentiviral vectors and adenoviral vectors (Abbas-Terki et al., 2002; Brummelkamp et al., 2002b; Michiels et al., 2002)). Some libraries consist of shRNA vectors that contain a shRNA embedded in a microRNA precursor for more efficient knockdown, but the beneficial effect of this has been disputed (Li et al., 2007) (Figure 1).

Figure 1
figure1

RNA interference (RNAi) screening modalities: RNAi can be brought about by different molecules, the vector-based short hairpin RNA (shRNA) and the chemically or enzymatically generated small interference RNAs (siRNA). The choice of screening model is dependent on the species of RNAi used, shRNAs can be used in both polyclonal and single-well screens, siRNAs can only be used in single-well assays. This also limits the phenotypes that can be screened. Polyclonal assays are more applicable to screen growth phenotypes although single-well assays can be used to screen complex phenotypes.

There is no simple answer to the question which RNAi reagent is better, as each reagent has its own pros and cons. A major advantage of the use of plasmid-encoded shRNA is the possibility to perform long-term cell culture experiments, such as clonogenic assays and drug resistance assays. shRNA-based genetic screens are particularly useful when one searches for genes whose suppression allows cells to proliferate in the presence of an anti-proliferative signal, such as a drug-induced growth arrest or a physiological growth arrest such as senescence. In this scenario, shRNA screens can be carried out in a polyclonal format, in which a single dish of cells is infected in culture with a large number of shRNA vectors, after which the cells are exposed to the anti-proliferative signal. Cells that become resistant to the growth arrest through knockdown of a specific gene will become positively selected. As each shRNA vector contains a unique shRNA sequence to knockdown a specific gene, this sequence can be used as a molecular ‘bar code’ to identify the knockdown vector that conferred the growth advantage to the cell. The bar-coded shRNA that confers resistance to the growth arrest signal will become more prominent in the population over time, which is detected as an increased abundance of the bar code sequence. This approach, known as shRNA bar code screening, has been applied successfully to find genes whose inactivation confers resistance to a p53 induced growth arrest (Berns et al., 2004; Brummelkamp et al., 2006), the breast cancer drug Herceptin (Berns et al., 2007) and genes that allow growth under non-adherent conditions (Westbrook et al., 2005) (Figure 2). The screening of shRNA collections by this bar coding approach to identify shRNAs that are negatively selected in culture represents a more significant technical challenge. Detecting the loss of a shRNA from a large pool is more difficult because not every cell infected with a given shRNA vector has sufficient gene knockdown to provoke the desired phenotype. Obviously, only cells with sufficient knockdown of a lethal gene are negatively selected in a population. In addition, there may be cells in the population that harbor a knockdown vector without having sufficient gene knockdown to provoke the lethal phenotype. The presence of such cells limits the signal one can obtain in this type of bar code screening approach. One solution to this problem is the use of relative small pools of 1000 shRNAs in negative selection assays (Ngo et al., 2006; Shaffer et al., 2008). More recently, this approach was optimized as well as used successfully to screen larger shRNA libraries (Rines et al., 2006; Luo et al., 2008; Schlabach et al., 2008; Silva et al., 2008).

Figure 2
figure2

The short hairpin RNA (shRNA) bar code screening: shRNA bar code screening is a very efficient technique to identify shRNAs from a large collection of vectors that give rise to a specific phenotype. Plasmids encoding shRNAs are available in various viral vectors for high-efficiency infection of target cell lines. After sufficient numbers of cells are infected (in order to have the entire library of vectors represented) two replicate cell populations are created. One population is used as reference and left untreated, whereas the other population is treated with the stimulus of interest. After phenotypic selection has taken place the bar code identifier can be retrieved by PCR. The bar code identifier can be a separate DNA sequence in the plasmid but also an integral part of the shRNA cassette. Hybridization of the bar code identifier to a specific microarray reveals the abundance of shRNA in both the reference and experimental population.

Small interference RNAs on the other hand are very suitable for high throughput single-well assays, in which every well contains a siRNA reagent that targets a single transcript. An advantage of the single-well screening format is that more complex biological phenotypes can be screened. Such complex phenotypes can be detected using for instance cell sorting or high throughput microscopy. Microscopic images can subsequently be analysed by software that can be programmed to extract features such as cell shape, DNA content, subcellular localization of proteins and other parameters (Wheeler et al., 2005; Pepperkok and Ellenberg, 2006; Rines et al., 2006). This combination of single-well screening and reading of complex phenotypes is referred to as ‘high content screening’ (Neumann et al., 2006). The cell cycle is one of the best-studied biological processes using RNAi combined with high content screening (Kittler et al., 2004; Bjorklund et al., 2006; Mukherji et al., 2006; Kittler et al., 2007). The parameters examined in such screens include cell number, cell cycle phase, apoptosis (sub G1) and the total number of chromosomes (ploidy). Among the hits in these screens are many genes that have been reported earlier to regulate mammalian cell cycle, showing the validity of the approach. High content screening also enables the identification of novel morphological phenotypes (Kiger et al., 2003; Jones et al., 2009). Single-well screens can also be performed using shRNA vectors (Hopkins and Groom, 2002; Moffat et al., 2006), but it requires significant automation to generate and handle individual shRNA vectors in single-well screening format.

Loss-of-function screening strategies to identify drug targets

Suitable targets for cancer therapy are hard to find. Most small molecule drugs today inactivate proteins by binding to the catalytic site of an enzyme. This mostly restricts the development of cancer drugs to the products of oncogenes, not tumor suppressor genes. Moreover, many of the well-known oncogenes (for example, RAS and MYC) are not considered to be ‘druggable’, that is, proteins whose activity can be readily inhibited by a low-molecular-weight compound (Hopkins and Groom, 2002). Thus far only a small number of proteins have been identified that can be efficiently inhibited by small molecules for cancer therapy. Many of these targets belong to the family of kinases and are either specifically (over-) expressed in cancer cells (that is, HER1 and HER2) or activated by mutation/translocation (for example, HER1 and BRAF and BCR-ABL). Inhibitors of these targets can be very efficacious in blocking cancer growth with relatively mild side effects (Carter et al., 1992; Druker et al., 1996). The identification of new drug targets with similar cancer selectivity is urgently needed. There are four approaches to finding novel drug targets through RNAi-based genetic screens:

Pathway screens

An obvious starting point to identify novel drug targets is by searching for novel components of cancer-relevant signaling pathways. Although many signaling pathways have been studied for considerable time, genetic screens may identify hitherto unknown components of these pathways, which may serve as potential targets for therapy. Initial RNAi genetic screens in Drosophila using reporter gene assays established the utility of signaling pathway screens (Baeg et al., 2005; DasGupta et al., 2005). More direct measurement of pathway activity was used to screen for modulators of ERK signaling, which was evaluated by immuno-fluorescence (Friedman and Perrimon, 2006).

A particularly powerful combination is to screen subsets consisting of families of druggable genes for their ability to modulate a cancer-relevant signaling pathway. Any validated hit from such a screen would directly place a druggable gene in a cancer-relevant pathway (Figure 3). As a proof of concept, we made a library of shRNA vectors targeting some 60 ubiquitin-specific proteases (DUBs) and searched for DUB enzymes that control the activity of NF-κB (Brummelkamp et al., 2003; Oosterkamp et al., 2006). This led to the identification of the familial cylindromatosis tumor suppressor gene CYLD in the NF-κB signaling pathway. This work suggested that tumors in cylindromatosis patients are caused, at least in part, by activation of NF-kB. Indeed, in a pilot clinical study, inhibitors of NF-kB were found to be useful for the treatment of patients suffering from cylindromatosis (Oosterkamp et al., 2006).

Figure 3
figure3

Pathway screening using RNA interference (RNAi) libraries targeting druggable genes: choosing druggable genes as a starting point for RNAi screening can lead to rapid identification of a novel drug target in a cancer-relevant pathway. The first step is to select the gene family of interest. Subsequently, a targeted RNAi library of either shRNAs of siRNAs must be constructed. This library can be screened in various screening models including reporters and high-content-based assays. Small molecules targeting the hits identified from the screen can subsequently be used to perform in vivo experiments and finally clinical trials.

Phenotypic screens

Another screening format is to search for modulators of a specific biological phenotype, such as apoptosis, migration, invasion, senescence and responses to cytokines. One of the first genome-wide screens in mammalian cells aimed at identifying genes whose suppression allowed bypass of a p53-dependent senescence-like growth arrest in human fibroblasts, which led to the identification of five novel components of the p53 pathway (Berns et al., 2004). A screen for bypass of growth arrest caused by activation of p53 with specific small molecules allowed the identification of further components of the p53 pathway (Brummelkamp et al., 2006).

Oncogene expression can also induce a senescence-like proliferation arrest. Wajapeyee et al. (2008) identified genes whose suppression confers resistance to growth arrest induced by a BRAF oncogene. Surprisingly, the authors found that suppression of IGFBP7, which encodes a secreted protein, was required for cells to proliferate in the presence of mutant BRAF. Moreover, injection of recombinant IGFBP7 in mice carrying a BRAF mutant tumor led to inhibition of tumor growth, providing a potential new therapeutic strategy for tumors having a mutant BRAF oncogene.

Tumor cells often have an abrogated cell death or apoptosis pathway. Several screens have been performed to identify genes that, when knocked down, either suppress or stimulate cell death (Aza-Blanc et al., 2003; MacKeigan et al., 2005; Hitomi et al., 2008). By screening siRNA libraries targeting kinases and phosphatases MacKeigan et al. (2005) identified a surprisingly large number of genes that are involved in regulating apoptosis under physiological conditions. In addition, a number of phosphases have been identified to be essential for the cyotoxic effects of certain chemotherapeutic drugs.

Synthetic lethal screens

Most targeted therapies take advantage of the fact that certain oncoproteins are hyperactive in cancer cells, either as a consequence of activating mutations or over-expression. As was pointed out above, such targets are often not ‘druggable’ or only expressed in a small fraction of the tumors, two factors that limit their clinical utility. In 1997, it was proposed to exploit the phenomenon of ‘synthetic lethality’ to find completely novel classes of highly cancer-selective drug targets (Hartwell et al., 1997). This strategy takes advantage of the notion, first observed in lower organisms, that some genes are only lethal to cells if a second non-lethal mutation is also present (Tong et al., 2001). These types of genetic interactions are likely to be present in mammals also and therefore could be exploited for novel cancer treatment strategies (Wang et al., 2004; Kaelin, 2005). In one of the first examples, it was shown that cells with an amplified MYC oncogene are more sensitive to apoptosis induced by death receptor ligands (Wang et al., 2004; Rottmann et al., 2005). It is important to point out that the genes that are synthetic lethal with cancer-specific lesions are not necessarily mutated or over-expressed themselves in cancer. Consequently, large-scale cancer genome re-sequencing efforts and gene expression profiling might not identify synthetic lethality genes. As such, synthetic lethal interactions represent a potential untapped reservoir of highly cancer-selective drug targets.

Finding synthetic lethal interactions in mammalian cells is not a trivial exercise, but is greatly facilitated by large-scale RNAi genetic screens. Large-scale efforts to identify synthetic lethal interactions were initially examined in yeast (Pan et al., 2004, 2006). The advantage of performing screens in yeast is that, rather than knockdown by RNAi, knockout strains can be used. On the other hand, the increased complexity of mammalian signaling pathways may limit the value of the lessons learnt from simple model organisms. To date, synthetic lethal screens have been performed for two specific oncogenic lesions in mammalian cells: targeting the activated KRAS oncogene (Sarthy et al., 2007; Bommi-Reddy et al., 2008; Luo et al., 2009a; Scholl et al., 2009) and loss of the VHL tumor suppressor (Bommi-Reddy et al., 2008). Most of these screens were performed using single-well assays with only relatively small collections of RNAi reagents. Only one study used a genome-wide shRNA library that was screened using micro-array bar code technology (Luo et al., 2009a). Genes identified in the screens above include Polo-like kinase 1 (PLK1), SURVIVIN and the STK33 kinase. Interestingly, clinical trials using small molecule inhibitors of PLK1 (BI-2536) and SURVIVIN (YM155) are already ongoing (Mross et al., 2008; Satoh et al., 2009). On the basis of the results from the activated RAS RNAi synthetic lethal screens, it will be interesting to monitor the specific response of RAS mutant versus wild-type tumors in these ongoing clinical trials.

Synthetic lethal screens can also be used to find genes that are specifically toxic in two closely related cancers. For instance, Ngo et al. (2006) screened for genes that are essential for survival of activated B-cell-like diffuse large B-cell lymphoma, but not for the related germinal centre B-cell-like diffuse large B-cell lymphoma cells. They identified the NF-kB pathway as essential only for survival of activated B-cell-like diffuse large B-cell lymphoma, suggesting that this cancer might benefit from NF-κB inhibition.

In vivo screens

The identification of genes that regulate the process of metastasis is a very difficult and laborious task. Therefore, screening in a surrogate assay for in vivo metastasis can be used to examine large collections of shRNAs or siRNAs. Hits from this screen can subsequently be tested in vivo, as the number of genes to be tested is significantly smaller. The latter approach was used to identify the GAS1 metastasis suppressor from a complex library of shRNAs. Furthermore, when the expression level of GAS1 in melanoma was analysed, a clear correlation between low GAS1 messenger RNA (mRNA) and metastatic potential was identified (Gobeil et al., 2008).

Another way to reduce the complexity of a genome-wide RNAi set for in vivo genetic screen is to use data obtained from clinical specimen. This was elegantly shown in a study by Zender et al. (2008) (Tyner et al., 2009). Using comparative genomic hybridization profiles from human liver cancer samples they identified some 300 genes that were frequently lost in these tumors, making these genes candidate tumor suppressor genes for liver cancer. On the basis of this information, they generated a mini shRNA targeting these 300 genes. These shRNAs were introduced into immortalized, but non-oncogenic, hepatocytes that were transplanted into recipient mice. In the tumors that formed in these mice shRNAs for 16 genes were selectively enriched, showing a role for these genes in hepatocyte oncogenicity. This shows that the use of selected shRNA libraries allows for the very efficient identification of genes that are functionally involved in cancer progression.

A variation on this theme is to perform these types of screens in tumor cells isolated from patients. This was shown in a recent report that described the siRNA screening of tyrosine kinases in cells derived from 30 patients suffering from leukemia (Tyner et al., 2009). In 10 out of these 30 patients, at least one kinase was identified that was essential for tumor cell survival. Furthermore, the inhibition of the JAK2 kinase by a small molecule inhibitor was shown to be selectively toxic only in the tumor lines that depend on JAK2.

Development of predictive biomarkers of therapy efficacy

Identical treatment regimen can produce remarkably different responses in patients with seemingly indistinguishable tumors at the macromolecular level. This phenomenon reflects the heterogeneity that is seen when individual tumors are studied in more detail. Understanding the factors that predict whether patients will respond to certain treatment is very useful to create efficient therapies. Although different techniques can be used to identify biomarkers of drug response, the use of RNAi to find these factors has the advantage that it will identify biomarkers based on their causal role in the response to the drug.

Two different types of drug response biomarkers can be distinguished. First, biomarkers may foretell whether patients are resistant to a certain treatment and second, biomarkers may predict drug sensitivity. The former biomarkers have two clinical applications: they may identify patients that fail to respond to a given therapy upfront, avoiding unnecessary toxicity. Second, understanding resistance may uncover strategies to overcome resistance. The biomarkers of drug sensitivity on the other hand may help in identifying synergistic drug combinations. RNAi can be used for the identification of both types of biomarkers through screening for modulators of drug resistance.

Biomarkers of drug resistance

Multiple mechanisms have been suggested for the resistance of cancer cells to conventional chemotherapy, for example, the upregulation of drug pumps and (in)-activation of pathways on which the drug works (Gottesman et al., 2002; Rodriguez-Antona and Ingelman-Sundberg, 2006; Savage et al., 2009). However, in most cases, the factors that contribute to drug resistance are not known and RNAi screens can be useful to identify genes that modulate drug responses.

When performing a functional genetic screen to find drug resistance genes it is important to use concentrations of drug that are close to levels obtained in the clinic. In long-term shRNA-based screens, drug concentration can be relatively low. In contrast, siRNA single-well screens usually require much higher concentration of drug, as the phenotype must be measured in a much shorter time frame. For drug resistance RNAi screens, the use of shRNA vectors is therefore preferred. Drug resistance screens have been performed successfully with various anticancer drugs. One approach was designed to identify genes that are essential for the action of the HER2 antibody trastuzumab (Herceptin) (Berns et al., 2007). A single gene was identified that confers resistance to trastuzumab when it was knocked down by shRNAs. This gene-encoded PTEN, the inhibiting phosphatase of the PI3-kinase catalytic subunit PIK3CA. When patient samples were analysed on their activation of the PI3-kinase pathway, a clear correlation with trastuzumab response was found. Patients with an activated PI3-kinase pathway, loss of PTEN or mutation in PIK3CA, showed worse response to trastuzumab than patients that did not have these mutations. This example shows that results from in vitro RNAi screens can yield biomarkers having clinical utility to predict drug responses.

In related studies, loss of CDK10 expression was identified as a major factor in mediating resistance to tamoxifen in breast cancer and CDK10 levels were found to be predictive of tamoxifen response in clinical samples (Iorns et al., 2008). Similarly, ZNF423 was identified as a critical mediator of the response to retinoic acid in neuroblastoma and RAD23B was found to control responses to histone deacetylase inhibitors. In all cases, the genetic screen did not only provide potentially useful biomarkers of drug responses, but also yielded novel insights into the signaling pathways that mediate drug resistance (Fotheringham et al., 2009; Huang et al., 2009). Such studies may therefore point at specific combination therapies that will be more powerful or suggest ways to overcome therapy resistance.

Enhancers of drug efficacy

Another way of finding modulators of drug responses is to search for genes whose suppression enhances the response to a given cancer drug. This is in fact a variation on the theme of synthetic lethality and is also referred to as ‘chemical synthetic lethality’. Like most genetic screens, this principle was also pioneered in lower organisms. For instance, testing of a set of compounds that are currently used in the clinic in a heterozygous yeast strain identified strains that show a decrease in fitness in presence of the drug (Lum et al., 2004). Performing RNAi-based drug enhancer screens in mammalian cells requires some technical adaptations to the conventional approach. To find factors that can augment the effects of a drug, the screen should be performed in presence of a low concentration of the drug, which causes only a small decrease of cell viability. In this way genes that significantly enhance loss of cell viability can be identified. Most drug screens are performed in a single-well format, which allow for effective gene knockdown because of highly efficient transfection with siRNAs.

Thus, far screens have been performed aiming at the identification of sensitizers of well the well-known anticancer drugs gemcitabine, cisplatin and paclitaxel (MacKeigan et al., 2005; Bartz et al., 2006; Giroux et al., 2006; Ji et al., 2007; Swanton et al., 2007; Whitehurst et al., 2007). Although cisplatin has been approved for the treatment of human malignancies since 1978, it is still unclear why some tumors respond better to treatment than others. By screening a genome-wide siRNA library to identify genes that modulate cisplatin efficacy, it was uncovered that cells that have an inactivation of both TP53 and a member of the BRCA network are more sensitive to cisplatin treatment. Although this had been reported before, this example underscores the power of large-scale loss-of-function genetic screens to find enhancers of drug responses.

Another example of a large RNAi-based approach with the objective to find genes that, when suppressed, sensitize cells to paclitaxel treatment identified several members of the proteasome (Whitehurst et al., 2007). This interesting observation raises the possibility of combining treatment of taxanes with proteasome inhibitors. Intriguingly, several clinical trials combing paclitaxel and the proteasome inhibitor bortezomib were already performed before the results from this study became available (Ma et al., 2007; Cresta et al., 2008; Jatoi et al., 2008). From these phase I/II trials it was concluded that the combination therapy is well tolerated but unfortunately not very efficacious.

Selecting hits from RNA interference screens as candidate drug targets

Not every gene identified in an RNAi screen is a suitable target for drug discovery. First, one needs to verify that the effect observed in the RNAi screen is caused by knockdown of the intended target gene. Although RNAi was initially thought to be exquisitely specific, some concerns have been raised after reports of so-called ‘off-target’ effects (Jackson et al., 2003). The term ‘off-target effect’ is used to describe the nonspecific suppression of mRNAs by siRNA or shRNAs. This suppression can either be through cleavage of mRNA or mRNA translation inhibition through miRNA-like processes. Especially this last phenomenon is difficult to predict, as it only requires a partial complementarity between the siRNA/shRNA and the mRNA. Several approaches can be used to validate that phenotypes produced by RNAi are caused by ‘on-target’ rather than ‘off-target’ mechanisms. The first approach is the use of several siRNAs/shRNAs that target different regions in a given mRNA. Having several independent siRNAs/shRNAs that are able to produce the same phenotype makes chances for an ‘off-target’ effects very slim. Another way to validate the specificity of an RNAi generated phenotype is the use of RNAi-resistant cDNAs, this allows for the re-expression of the targeted mRNA, which should lead to reversal of the phenotype (Echeverri et al., 2006).

If available, one can also validate RNAi screening hits using conventional gene knockout cells. However, it is possible that gene knockout triggers compensatory adaptation mechanisms that mask the phenotype of a knockout over time. Therefore, suppression of a gene by RNAi may reflect more accurately the pharmacological inhibition of a protein than a full gene knockout. It should also be kept in mind that complete gene knockout (null allele) does not necessarily yield the same phenotype as incomplete gene suppression by RNAi.

Finally, it must be appreciated that substantial phenotypic differences can occur between pharmacological inhibition of an existing protein, and the removal of the protein from the cell by RNAi. For example, if a protein functions in a multi-protein complex, pharmacological inhibition of the target may preserve the integrity of the complex, whereas removal of the protein can cause such a complex to fall apart, removing also potential other functions of the complex.

Future directions for RNA interference in cancer research

In less than a decade, RNA interference has evolved into a major tool for drug discovery and even a potential therapeutic agent in it's own right (Goff, 2008; Castanotto and Rossi, 2009). Large RNAi collections have been made available to the scientific community enabling hundreds of RNAi screens, ranging in size from only dozens of genes to entire mammalian genomes (see Table 1 for a summary of some of the most noteworthy screens). Together, these screens have assigned (additional) functions to many genes in processes ranging from cancer, development, stem cell maintenance to viral replication. It should be kept in mind that many genes reported in these screens have not been validated in additional biological assays, and therefore such gene lists should be viewed with caution (Goff, 2008). Of particular concern is the possible context dependency of genetic interactions identified in RNAi screens. For instance, gene A may be synthetic lethal with the loss of gene B in lung cancer cells in which the screen was performed, but not in breast cancer. Similarly, a gene may be involved in resistance to a given drug in colon cancer, but not have a role in resistance to the same drug in prostate cancer. Understanding context dependency of genetic interactions is the major challenge for the next decade and this will be a prerequisite for a more profound understanding of how patients respond to cancer drugs. Such context dependency will ultimately be governed by cross talk between signaling pathways. Thus, if we want to fully understand how cells respond to perturbations of specific signals with targeted cancer therapies, we will need to map these interactions between signaling pathways. RNAi is exquisitely well suited to identify functional interactions between signaling pathways. We wholeheartedly support the massive efforts that are undertaken worldwide to re-sequence thousands of cancer genomes. It will no doubt provide a deep insight into the genes that are deregulated in cancer. However, we will only be able to truly appreciate the meaning of the multitude of mutations in the major signaling pathways, if we also understand how these pathways are interconnected. We therefore need in addition to the re-sequencing project a large-scale effort to map functional interactions between signaling pathways using the approaches described here. The technology is readily available and proof of concept has been delivered. All we need is a concerted funding effort to make it happen. Anyone interested?

Table 1 Some noteworthy RNAi screens performed in mammalian cells in recent years

Conflict of interest

The authors declare no conflict of interest.

References

  1. Abbas-Terki T, Blanco-Bose W, Deglon N, Pralong W, Aebischer P . (2002). Lentiviral-mediated RNA interference. Hum Gene Ther 13: 2197–2201.

    CAS  PubMed  Google Scholar 

  2. Aza-Blanc P, Cooper CL, Wagner K, Batalov S, Deveraux QL, Cooke MP . (2003). Identification of modulators of TRAIL-induced apoptosis via RNAi-based phenotypic screening. Mol Cell 12: 627–637.

    CAS  PubMed  Google Scholar 

  3. Baeg GH, Zhou R, Perrimon N . (2005). Genome-wide RNAi analysis of JAK/STAT signaling components in Drosophila. Genes Dev 19: 1861–1870.

    CAS  PubMed  PubMed Central  Google Scholar 

  4. Bartz SR, Zhang Z, Burchard J, Imakura M, Martin M, Palmieri A et al. (2006). Small interfering RNA screens reveal enhanced cisplatin cytotoxicity in tumor cells having both BRCA network and TP53 disruptions. Mol Cell Biol 26: 9377–9386.

    CAS  PubMed  PubMed Central  Google Scholar 

  5. Berns K, Hijmans EM, Mullenders J, Brummelkamp TR, Velds A, Heimerikx M et al. (2004). A large-scale RNAi screen in human cells identifies new components of the p53 pathway. Nature 428: 431–437.

    CAS  Article  Google Scholar 

  6. Berns K, Horlings HM, Hennessy BT, Madiredjo M, Hijmans EM, Beelen K et al. (2007). A functional genetic approach identifies the PI3 K pathway as a major determinant of trastuzumab resistance in breast cancer. Cancer Cell 12: 395–402.

    CAS  Google Scholar 

  7. Bjorklund M, Taipale M, Varjosalo M, Saharinen J, Lahdenpera J, Taipale J . (2006). Identification of pathways regulating cell size and cell-cycle progression by RNAi. Nature 439: 1009–1013.

    PubMed  Google Scholar 

  8. Bommi-Reddy A, Almeciga I, Sawyer J, Geisen C, Li W, Harlow E et al. (2008). Kinase requirements in human cells: III. Altered kinase requirements in VHL−/− cancer cells detected in a pilot synthetic lethal screen. Proc Natl Acad Sci USA 105: 16484–16489.

    CAS  PubMed  Google Scholar 

  9. Brummelkamp TR, Bernards R . (2003). New tools for functional mammalian cancer genetics. Nat Rev Cancer 3: 781–789.

    CAS  PubMed  Google Scholar 

  10. Brummelkamp TR, Bernards R, Agami R . (2002a). A system for stable expression of short interfering RNAs in mammalian cells. Science 296: 550–553.

    CAS  PubMed  PubMed Central  Google Scholar 

  11. Brummelkamp TR, Bernards R, Agami R . (2002b). Stable suppression of tumorigenicity by virus-mediated RNA interference. Cancer Cell 2: 243–247.

    CAS  PubMed  PubMed Central  Google Scholar 

  12. Brummelkamp TR, Fabius AW, Mullenders J, Madiredjo M, Velds A, Kerkhoven RM et al. (2006). An shRNA barcode screen provides insight into cancer cell vulnerability to MDM2 inhibitors. Nat Chem Biol 2: 202–206.

    CAS  PubMed  Google Scholar 

  13. Brummelkamp TR, Nijman SM, Dirac AM, Bernards R . (2003). Loss of the cylindromatosis tumour suppressor inhibits apoptosis by activating NF-kappaB. Nature 424: 797–801.

    CAS  PubMed  PubMed Central  Google Scholar 

  14. Buchholz F, Kittler R, Slabicki M, Theis M . (2006). Enzymatically prepared RNAi libraries. Nat Methods 3: 696–700.

    CAS  PubMed  Google Scholar 

  15. Carter P, Presta L, Gorman CM, Ridgway JB, Henner D, Wong WL et al. (1992). Humanization of an anti-p185HER2 antibody for human cancer therapy. Proc Natl Acad Sci USA 89: 4285–4289.

    CAS  Google Scholar 

  16. Castanotto D, Rossi JJ . (2009). The promises and pitfalls of RNA-interference-based therapeutics. Nature 457: 426–433.

    CAS  PubMed  PubMed Central  Google Scholar 

  17. Chabner BA, Roberts Jr TG . (2005). Timeline: chemotherapy and the war on cancer. Nat Rev Cancer 5: 65–72.

    CAS  PubMed  Google Scholar 

  18. Cresta S, Sessa C, Catapano CV, Gallerani E, Passalacqua D, Rinaldi A et al. (2008). Phase I study of bortezomib with weekly paclitaxel in patients with advanced solid tumours. Eur J Cancer 44: 1829–1834.

    CAS  PubMed  Google Scholar 

  19. DasGupta R, Kaykas A, Moon RT, Perrimon N . (2005). Functional genomic analysis of the Wnt-wingless signaling pathway. Science 308: 826–833.

    CAS  Google Scholar 

  20. de Klein A, van Kessel AG, Grosveld G, Bartram CR, Hagemeijer A, Bootsma D et al. (1982). A cellular oncogene is translocated to the Philadelphia chromosome in chronic myelocytic leukaemia. Nature 300: 765–767.

    CAS  Google Scholar 

  21. Druker BJ, Tamura S, Buchdunger E, Ohno S, Segal GM, Fanning S et al. (1996). Effects of a selective inhibitor of the Abl tyrosine kinase on the growth of Bcr-Abl positive cells. Nat Med 2: 561–566.

    CAS  PubMed  PubMed Central  Google Scholar 

  22. EBCTCG (2005). Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomised trials. Lancet 365: 1687–1717.

    Google Scholar 

  23. Echeverri CJ, Beachy PA, Baum B, Boutros M, Buchholz F, Chanda SK et al. (2006). Minimizing the risk of reporting false positives in large-scale RNAi screens. Nat Methods 3: 777–779.

    CAS  PubMed  Google Scholar 

  24. Elbashir SM, Harborth J, Lendeckel W, Yalcin A, Weber K, Tuschl T . (2001). Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411: 494–498.

    CAS  Google Scholar 

  25. Firestein R, Bass AJ, Kim SY, Dunn IF, Silver SJ, Guney I et al. (2008). CDK8 is a colorectal cancer oncogene that regulates beta-catenin activity. Nature 455: 547–551.

    CAS  PubMed  PubMed Central  Google Scholar 

  26. Fotheringham S, Epping MT, Stimson L, Khan O, Wood V, Pezzella F et al. (2009). Genome-wide loss-of-function screen reveals an important role for the proteasome in HDAC inhibitor-induced apoptosis. Cancer Cell 15: 57–66.

    CAS  PubMed  Google Scholar 

  27. Friedman A, Perrimon N . (2006). A functional RNAi screen for regulators of receptor tyrosine kinase and ERK signalling. Nature 444: 230–234.

    CAS  PubMed  Google Scholar 

  28. Giroux V, Iovanna J, Dagorn JC . (2006). Probing the human kinome for kinases involved in pancreatic cancer cell survival and gemcitabine resistance. Faseb J 20: 1982–1991.

    CAS  PubMed  Google Scholar 

  29. Gobeil S, Zhu X, Doillon CJ, Green MR . (2008). A genome-wide shRNA screen identifies GAS1 as a novel melanoma metastasis suppressor gene. Genes Dev 22: 2932–2940.

    CAS  PubMed  PubMed Central  Google Scholar 

  30. Goff SP . (2008). Knockdown screens to knockout HIV-1. Cell 135: 417–420.

    CAS  PubMed  PubMed Central  Google Scholar 

  31. Gottesman MM, Fojo T, Bates SE . (2002). Multidrug resistance in cancer: role of ATP-dependent transporters. Nat Rev Cancer 2: 48–58.

    CAS  Google Scholar 

  32. Grimm S . (2004). The art and design of genetic screens: mammalian culture cells. Nat Rev Genet 5: 179–189.

    CAS  PubMed  Google Scholar 

  33. Hartwell LH, Szankasi P, Roberts CJ, Murray AW, Friend SH . (1997). Integrating genetic approaches into the discovery of anticancer drugs. Science 278: 1064–1068.

    CAS  Google Scholar 

  34. Hitomi J, Christofferson DE, Ng A, Yao J, Degterev A, Xavier RJ et al. (2008). Identification of a molecular signaling network that regulates a cellular necrotic cell death pathway. Cell 135: 1311–1323.

    CAS  PubMed  PubMed Central  Google Scholar 

  35. Hopkins AL, Groom CR . (2002). The druggable genome. Nat Rev Drug Discov 1: 727–730.

    CAS  PubMed  Google Scholar 

  36. Huang S, Laoukili J, Epping MT, Koster J, Holzel M, Westerman BA et al. (2009). ZNF423 is critically required for retinoic acid-induced differentiation and is a marker of neuroblastoma outcome. Cancer Cell 15: 328–340.

    CAS  PubMed  PubMed Central  Google Scholar 

  37. Iorns E, Lord CJ, Turner N, Ashworth A . (2007). Utilizing RNA interference to enhance cancer drug discovery. Nat Rev Drug Discov 6: 556–568.

    CAS  Google Scholar 

  38. Iorns E, Turner NC, Elliott R, Syed N, Garrone O, Gasco M et al. (2008). Identification of CDK10 as an important determinant of resistance to endocrine therapy for breast cancer. Cancer Cell 13: 91–104.

    CAS  PubMed  Google Scholar 

  39. Jackson AL, Bartz SR, Schelter J, Kobayashi SV, Burchard J, Mao M et al. (2003). Expression profiling reveals off-target gene regulation by RNAi. Nat Biotechnol 21: 635–637.

    CAS  PubMed  PubMed Central  Google Scholar 

  40. Jatoi A, Dakhil SR, Foster NR, Ma C, Rowland Jr KM, Moore Jr DF et al. (2008). Bortezomib, paclitaxel, and carboplatin as a first-line regimen for patients with metastatic esophageal, gastric, and gastroesophageal cancer: phase II results from the North Central Cancer Treatment Group (N044B). J Thorac Oncol 3: 516–520.

    PubMed  PubMed Central  Google Scholar 

  41. Ji D, Deeds SL, Weinstein EJ . (2007). A screen of shRNAs targeting tumor suppressor genes to identify factors involved in A549 paclitaxel sensitivity. Oncol Rep 18: 1499–1505.

    CAS  PubMed  Google Scholar 

  42. Jones TR, Carpenter AE, Lamprecht MR, Moffat J, Silver SJ, Grenier JK et al. (2009). Scoring diverse cellular morphologies in image-based screens with iterative feedback and machine learning. Proc Natl Acad Sci USA 106: 1826–1831.

    CAS  PubMed  Google Scholar 

  43. Kaelin Jr WG . (2005). The concept of synthetic lethality in the context of anticancer therapy. Nat Rev Cancer 5: 689–698.

    CAS  PubMed  PubMed Central  Google Scholar 

  44. Kiger AA, Baum B, Jones S, Jones MR, Coulson A, Echeverri C et al. (2003). A functional genomic analysis of cell morphology using RNA interference. J Biol 2: 27.

    CAS  PubMed  PubMed Central  Google Scholar 

  45. Kittler R, Pelletier L, Heninger AK, Slabicki M, Theis M, Miroslaw L et al. (2007). Genome-scale RNAi profiling of cell division in human tissue culture cells. Nat Cell Biol 9: 1401–1412.

    CAS  PubMed  Google Scholar 

  46. Kittler R, Putz G, Pelletier L, Poser I, Heninger AK, Drechsel D et al. (2004). An endoribonuclease-prepared siRNA screen in human cells identifies genes essential for cell division. Nature 432: 1036–1040.

    CAS  Google Scholar 

  47. Kolfschoten IG, van Leeuwen B, Berns K, Mullenders J, Beijersbergen RL, Bernards R et al. (2005). A genetic screen identifies PITX1 as a suppressor of RAS activity and tumorigenicity. Cell 121: 849–858.

    CAS  Google Scholar 

  48. Li L, Lin X, Khvorova A, Fesik SW, Shen Y . (2007). Defining the optimal parameters for hairpin-based knockdown constructs. RNA 13: 1765–1774.

    CAS  PubMed  PubMed Central  Google Scholar 

  49. Lum PY, Armour CD, Stepaniants SB, Cavet G, Wolf MK, Butler JS et al. (2004). Discovering modes of action for therapeutic compounds using a genome-wide screen of yeast heterozygotes. Cell 116: 121–137.

    CAS  PubMed  Google Scholar 

  50. Luo B, Cheung HW, Subramanian A, Sharifnia T, Okamoto M, Yang X et al. (2008). Highly parallel identification of essential genes in cancer cells. Proc Natl Acad Sci USA 105: 20380–20385.

    CAS  PubMed  Google Scholar 

  51. Luo J, Emanuele MJ, Li D, Creighton CJ, Schlabach MR, Westbrook TF et al. (2009a). A genome-wide RNAi screen identifies multiple synthetic lethal interactions with the Ras oncogene. Cell 137: 835–848.

    CAS  PubMed  PubMed Central  Google Scholar 

  52. Luo J, Solimini NL, Elledge SJ . (2009b). Principles of cancer therapy: oncogene and non-oncogene addiction. Cell 136: 823–837.

    CAS  PubMed  PubMed Central  Google Scholar 

  53. Ma C, Mandrekar SJ, Alberts SR, Croghan GA, Jatoi A, Reid JM et al. (2007). A phase I and pharmacologic study of sequences of the proteasome inhibitor, bortezomib (PS-341, Velcade), in combination with paclitaxel and carboplatin in patients with advanced malignancies. Cancer Chemother Pharmacol 59: 207–215.

    CAS  PubMed  Google Scholar 

  54. MacKeigan JP, Murphy LO, Blenis J . (2005). Sensitized RNAi screen of human kinases and phosphatases identifies new regulators of apoptosis and chemoresistance. Nat Cell Biol 7: 591–600.

    CAS  PubMed  Google Scholar 

  55. Michiels F, van Es H, van Rompaey L, Merchiers P, Francken B, Pittois K et al. (2002). Arrayed adenoviral expression libraries for functional screening. Nat Biotechnol 20: 1154–1157.

    CAS  PubMed  Google Scholar 

  56. Moffat J, Grueneberg DA, Yang X, Kim SY, Kloepfer AM, Hinkle G et al. (2006). A lentiviral RNAi library for human and mouse genes applied to an arrayed viral high-content screen. Cell 124: 1283–1298.

    CAS  Google Scholar 

  57. Moffat J, Sabatini DM . (2006). Building mammalian signalling pathways with RNAi screens. Nat Rev Mol Cell Biol 7: 177–187.

    CAS  PubMed  Google Scholar 

  58. Mross K, Frost A, Steinbild S, Hedbom S, Rentschler J, Kaiser R et al. (2008). Phase I dose escalation and pharmacokinetic study of BI 2536, a novel Polo-like kinase 1 inhibitor, in patients with advanced solid tumors. J Clin Oncol 26: 5511–5517.

    CAS  PubMed  Google Scholar 

  59. Mukherji M, Bell R, Supekova L, Wang Y, Orth AP, Batalov S et al. (2006). Genome-wide functional analysis of human cell-cycle regulators. Proc Natl Acad Sci USA 103: 14819–14824.

    CAS  PubMed  Google Scholar 

  60. Neumann B, Held M, Liebel U, Erfle H, Rogers P, Pepperkok R et al. (2006). High-throughput RNAi screening by time-lapse imaging of live human cells. Nat Methods 3: 385–390.

    CAS  PubMed  Google Scholar 

  61. Ngo VN, Davis RE, Lamy L, Yu X, Zhao H, Lenz G et al. (2006). A loss-of-function RNA interference screen for molecular targets in cancer. Nature 441: 106–110.

    CAS  Google Scholar 

  62. Oosterkamp HM, Neering H, Nijman SM, Dirac AM, Mooi WJ, Bernards R et al. (2006). An evaluation of the efficacy of topical application of salicylic acid for the treatment of familial cylindromatosis. Br J Dermatol 155: 182–185.

    CAS  PubMed  Google Scholar 

  63. Pan X, Ye P, Yuan DS, Wang X, Bader JS, Boeke JD . (2006). A DNA integrity network in the yeast Saccharomyces cerevisiae. Cell 124: 1069–1081.

    CAS  PubMed  Google Scholar 

  64. Pan X, Yuan DS, Xiang D, Wang X, Sookhai-Mahadeo S, Bader JS et al. (2004). A robust toolkit for functional profiling of the yeast genome. Mol Cell 16: 487–496.

    CAS  PubMed  Google Scholar 

  65. Pepperkok R, Ellenberg J . (2006). High-throughput fluorescence microscopy for systems biology. Nat Rev Mol Cell Biol 7: 690–696.

    CAS  PubMed  Google Scholar 

  66. Rines DR, Gomez-Ferreria MA, Zhou Y, DeJesus P, Grob S, Batalov S et al. (2008). Whole genome functional analysis identifies novel components required for mitotic spindle integrity in human cells. Genome Biol 9: R44.

    PubMed  PubMed Central  Google Scholar 

  67. Rines DR, Tu B, Miraglia L, Welch GL, Zhang J, Hull MV et al. (2006). High-content screening of functional genomic libraries. Methods Enzymol 414: 530–565.

    CAS  PubMed  Google Scholar 

  68. Rodriguez-Antona C, Ingelman-Sundberg M . (2006). Cytochrome P450 pharmacogenetics and cancer. Oncogene 25: 1679–1691.

    CAS  PubMed  Google Scholar 

  69. Rottmann S, Wang Y, Nasoff M, Deveraux QL, Quon KC . (2005). A TRAIL receptor-dependent synthetic lethal relationship between MYC activation and GSK3beta/FBW7 loss of function. Proc Natl Acad Sci USA 102: 15195–15200.

    CAS  PubMed  Google Scholar 

  70. Sarthy AV, Morgan-Lappe SE, Zakula D, Vernetti L, Schurdak M, Packer JC et al. (2007). Survivin depletion preferentially reduces the survival of activated K-Ras-transformed cells. Mol Cancer Ther 6: 269–276.

    CAS  PubMed  Google Scholar 

  71. Satoh T, Okamoto I, Miyazaki M, Morinaga R, Tsuya A, Hasegawa Y et al. (2009). Phase I study of YM155, a novel survivin suppressant, in patients with advanced solid tumors. Clin Cancer Res 15: 3872–3880.

    CAS  PubMed  Google Scholar 

  72. Savage P, Stebbing J, Bower M, Crook T . (2009). Why does cytotoxic chemotherapy cure only some cancers? Nat Clin Pract Oncol 6: 43–52.

    CAS  PubMed  Google Scholar 

  73. Sawyers C . (2004). Targeted cancer therapy. Nature 432: 294–297.

    CAS  PubMed  PubMed Central  Google Scholar 

  74. Schlabach MR, Luo J, Solimini NL, Hu G, Xu Q, Li MZ et al. (2008). Cancer proliferation gene discovery through functional genomics. Science 319: 620–624.

    CAS  PubMed  PubMed Central  Google Scholar 

  75. Scholl C, Frohling S, Dunn IF, Schinzel AC, Barbie DA, Kim SY et al. (2009). Synthetic lethal interaction between oncogenic KRAS dependency and STK33 suppression in human cancer cells. Cell 137: 821–834.

    CAS  PubMed  Google Scholar 

  76. Shaffer AL, Emre NC, Lamy L, Ngo VN, Wright G, Xiao W et al. (2008). IRF4 addiction in multiple myeloma. Nature 454: 226–231.

    CAS  PubMed  PubMed Central  Google Scholar 

  77. Silva JM, Marran K, Parker JS, Silva J, Golding M, Schlabach MR et al. (2008). Profiling essential genes in human mammary cells by multiplex RNAi screening. Science 319: 617–620.

    CAS  PubMed  PubMed Central  Google Scholar 

  78. Stegmeier F, Rape M, Draviam VM, Nalepa G, Sowa ME, Ang XL et al. (2007). Anaphase initiation is regulated by antagonistic ubiquitination and deubiquitination activities. Nature 446: 876–881.

    CAS  Google Scholar 

  79. Swanton C, Marani M, Pardo O, Warne PH, Kelly G, Sahai E et al. (2007). Regulators of mitotic arrest and ceramide metabolism are determinants of sensitivity to paclitaxel and other chemotherapeutic drugs. Cancer Cell 11: 498–512.

    CAS  Google Scholar 

  80. Tang W, Dodge M, Gundapaneni D, Michnoff C, Roth M, Lum L . (2008). A genome-wide RNAi screen for Wnt/beta-catenin pathway components identifies unexpected roles for TCF transcription factors in cancer. Proc Natl Acad Sci USA 105: 9697–9702.

    CAS  PubMed  Google Scholar 

  81. Tong AH, Evangelista M, Parsons AB, Xu H, Bader GD, Page N et al. (2001). Systematic genetic analysis with ordered arrays of yeast deletion mutants. Science 294: 2364–2368.

    CAS  PubMed  PubMed Central  Google Scholar 

  82. Turner NC, Lord CJ, Iorns E, Brough R, Swift S, Elliott R et al. (2008). A synthetic lethal siRNA screen identifying genes mediating sensitivity to a PARP inhibitor. Embo J 27: 1368–1377.

    CAS  PubMed  PubMed Central  Google Scholar 

  83. Tyner JW, Deininger MW, Loriaux MM, Chang BH, Gotlib JR, Willis SG et al. (2009). RNAi screen for rapid therapeutic target identification in leukemia patients. Proc Natl Acad Sci USA 106: 8695–8700.

    CAS  Google Scholar 

  84. Wajapeyee N, Serra RW, Zhu X, Mahalingam M, Green MR . (2008). Oncogenic BRAF induces senescence and apoptosis through pathways mediated by the secreted protein IGFBP7. Cell 132: 363–374.

    CAS  PubMed  PubMed Central  Google Scholar 

  85. Wang Y, Engels IH, Knee DA, Nasoff M, Deveraux QL, Quon KC . (2004). Synthetic lethal targeting of MYC by activation of the DR5 death receptor pathway. Cancer Cell 5: 501–512.

    CAS  PubMed  Google Scholar 

  86. Westbrook TF, Martin ES, Schlabach MR, Leng Y, Liang AC, Feng B et al. (2005). A genetic screen for candidate tumor suppressors identifies REST. Cell 121: 837–848.

    CAS  PubMed  PubMed Central  Google Scholar 

  87. Wheeler DB, Carpenter AE, Sabatini DM . (2005). Cell microarrays and RNA interference chip away at gene function. Nat Genet 37 (Suppl): S25–S30.

    CAS  PubMed  Google Scholar 

  88. Whitehurst AW, Bodemann BO, Cardenas J, Ferguson D, Girard L, Peyton M et al. (2007). Synthetic lethal screen identification of chemosensitizer loci in cancer cells. Nature 446: 815–819.

    CAS  PubMed  Google Scholar 

  89. Zender L, Xue W, Zuber J, Semighini CP, Krasnitz A, Ma B et al. (2008). An oncogenomics-based in vivo RNAi screen identifies tumor suppressors in liver cancer. Cell 135: 852–864.

    CAS  PubMed  PubMed Central  Google Scholar 

Download references

Author information

Affiliations

Authors

Corresponding author

Correspondence to R Bernards.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Mullenders, J., Bernards, R. Loss-of-function genetic screens as a tool to improve the diagnosis and treatment of cancer. Oncogene 28, 4409–4420 (2009). https://doi.org/10.1038/onc.2009.295

Download citation

Keywords

  • RNAi screen
  • biomarkers of drug response
  • cancer therapy

Further reading

Search

Quick links