EVENTS | VIEW CALENDAR
Guest Commentary: New frontiers for target ID in oncology--Where will the next generation of targets come from?
Cancer therapy is fast approaching a clinical crossroads: to treat a cancer patient with an approach based on stimulating an immune response or to target the cancer cells directly with cytotoxic or a targeted therapy?
The arrival of this decision point is being hastened by the recent FDA and EMEA approval of the immune checkpoint inhibitors ipilimumab, nivolumab and pembrolizumab. Furthermore, there are hopes that the use of T cells expressing modified or chimeric antigen receptors (CARs) might extend beyond the hematological malignancies in which they have shown initial success.
However, this move towards immunotherapy is unlikely to be all-encompassing—clinical trials have identified numerous cancer types where immune checkpoint inhibitors have had little efficacy. Current thinking is that there will be cancer types for which an immunotherapeutic approach will be part of successful therapeutic modalities, but for others, an approach focused on targeting the cancer cells directly will remain the gold standard. However, for the promise of both strategies to be realized, new therapeutics need to be found.
Inciting tumor regression
Inciting an immune response in a patient with cancer is not straightforward. A productive immune response requires many different arms of the immune system, and multiple checks and balances exist to regulate the response and avoid autoimmunity. Several molecules crucial for regulating this coordinated response have been identified, in particular those operating in the immune checkpoint. Much like a neuronal synapse, immune cells form a synapse when interacting with target cells or antigen-presenting cells. The expression levels of immune synapse proteins vary as the immune response progresses, allowing effector immune cells to be maximally stimulated at the start of the response and then stood down once the infected cells have been cleared. Thus, the immune checkpoint has both activating and inhibitory signals.
For example, expression of CTLA4, the target of ipilimumab, occurs late in an immune response to provide a competing signal to CD28 (an activation signal), resulting in the inhibition of effector T cells. T cells in cancer patients can still recognize tumor antigens, but no longer show a productive immune response, leading to the hypothesis that T cells could be re-educated by inhibiting negative regulators such as CTLA4 and PD-1 (the target of nivolumab and pembrolizumab). This approach in essence releases the brakes on the immune response.
The increased overall survival rates observed in melanoma patients treated with immune checkpoint inhibitors are an aggregate of very different fates. Only a minority of patients respond, but the key result driving the excitement around immune checkpoint inhibitors is that some patients obtain a long-lasting benefit from this therapy. However, the majority of patients with melanoma or lung cancer do not fully respond to immune checkpoint inhibitors, and in other types of tumors, immune checkpoint inhibitors have had little impact to date. Understanding why some patients do not respond is one route to identifying the next wave of immuno-oncology targets.
The immune cells that reside in or near to the immune microenvironment can contribute to the growth of the tumor rather than impede its expansion. In particular, cytokines and chemokines produced by type 2 macrophages and myeloid-derived suppressor cells are known to inhibit T cell function and promote tumor cell proliferation. Strategies to inhibit and/or re-educate macrophages and myeloid-derived suppressor cells such that they once again contribute to a productive antitumor immune response is one route through which more patients might benefit from treatment with immune checkpoint inhibitors. There is also scope to identify new targets in T cells and natural killer cells, away from the immune checkpoint, that enable a more productive antitumor response.
However, productive targeting of all tumor types with immune-based therapies is unlikely to be possible. Thus, more direct approaches, such as targeting the signaling alterations in epithelial cells, remain crucial. Drug discovery opportunities in the most obvious targets of enzymes and receptors bearing frequent activating mutations have now been mined out. Remaining opportunities center on exploiting mutations in hard-to-drug oncogenes, such as KRAS or loss-of-function mutations in tumor suppressors such as RB1. Although one cannot develop a drug to a target that is lost, mutation of tumor suppressors might compromise the biology of the cancer cell, making it dependent on processes that are dispensable in healthy cells. This opens the way to synthetic lethal approaches.
ⱥ + ⱦ = death
Work in model organisms, notably yeast and Drosophila, has underlined the prevalence of synthetic lethality, a situation where loss-of-function mutations are tolerated until a gene responsible for fulfilling a redundant function is also mutated. Does this phenomenon have applications beyond basic research? In 1997, Leland Hartwell and colleagues proposed that insights from yeast genetics be applied to the search for anticancer drugs, anticipating a target discovery pipeline where synthetic lethal screening in yeast was at the forefront. However, the past 20 years have seen advances in mammalian functional genomics that have made it feasible to carry out such screens in cancer cells.
There are at least three contexts where synthetic lethality might be exploited for cancer therapy. First, many cancer driver mutations targeting tumor suppressors represent loss-of-function mutations in the complex network of proteins responsible for DNA damage repair or restraining inappropriate cell cycle progression. If backup pathways required for survival in these genetic contexts could be targeted, cancer cells might succumb, while the normal cells of the body that retain the pathway mutated in the cancer survive. Second, oncogenic mutations might rewire cancer cell signaling to create an increased dependency on factors downstream of the lesion. This could create a situation where signal transduction pathway components are essential in the cancer cell but dispensable in healthy cells. Third, successful tumor evolution requires substantial changes in the microenvironment and a tolerance of conditions known to induce cell stress, such as hypoxia or alterations in pH. This might cause a substantial dependence on certain stress pathways, which could be targeted to selectively kill cancer cells.
One therapeutic strategy exploiting tumor suppressor loss with a drug has now proved successful in the clinic. Mutations in the breast cancer susceptibility genes BRCA1 and BRCA2 compromise error-free repair of DNA via homologous recombination, creating a dependence on single-stranded DNA repair mechanisms, which require PARP1. The PARP inhibitor olaparib is now approved for the treatment of ovarian cancer patients with germline mutations in BRCA1 and BRCA2 who have been previously treated with chemotherapy.
Considerable effort has also been applied to targeting cancer cells with mutant TP53 with CHEK1 inhibitors, because in the absence of functional p53, cells rely in part on CHEK1 kinase signaling to restrain cell cycle progression. Thus, CHEK1 inhibitors should be selectively lethal to cells lacking wild-type p53. However, despite supportive preclinical data, this hypothesis has not been confirmed in the clinic. This might be connected with CHEK1 being essential in human cells rather than dispensable, as it is in fission yeast where it was first discovered.
Finding new targets
Finding new targets for immunotherapy and identifying synthetic lethal pathways present a similar challenge. This is one of emergent properties—the targets that we are trying to identify are not necessarily predictable based on the knowledge that we currently have at our disposal. We need to take a holistic approach to find these targets rather than to try and second-guess them. Again, this is not a new approach in biology, but it is one for which we have a new tool, that of CRISPR-Cas9 whole-genome screening.
The advent of gene editing using CRISPR-Cas9—essentially repurposing an atavistic adaptive immune response in bacteria—provides a powerful new tool to interrogate gene function on a genome-wide level. CRISPR-Cas9 screens make use of two aspects of the CRISPR machinery: short guide RNAs (sgRNAs) and the Cas9 nuclease. Cas9 binds to the sgRNA, which in turn can bind with the matching genomic DNA sequence where Cas9 introduces DNA double-strand breaks. These are repaired by non-homologous end joining, usually resulting in small nucleotide insertions and/or deletions interrupting the normal reading frame, thereby disrupting gene function.
Where Cas9 is expressed with sgRNAs libraries targeting early exons, a panel of knockout lines can be generated with high on-target specificity. When these libraries are delivered via lentiviral transduction in pooled-based screens, it is possible to identify from many thousands of genes in a single experiment, genes that are required for the maintenance of a specific phenotype.
Results from several CRISPR-Cas9 screens have been reported, and indications are that the technique is more sensitive and more selective at identifying relevant hits than the RNA interference-based technologies used for the past 10 years. We can now expect this approach to be applied in earnest to identify synthetic lethal targets in cancer cells and new targets in immune cells. With all the tools at our disposal for validation (including RNA interference, gene editing and novel mechanisms of RNA-mediated gene regulation such as CRISPRi and CRISPRa), we can expect a boost in the productivity of the drug discovery enterprise. Resources should become better focused on targets where inhibitors/modulators will prove effective in the clinic. A relevant and robust cancer therapy pipeline now seems much more plausible that it did just five years ago.
Nicola J. McCarthy is the oncology program manager for Horizon Discovery in the United Kingdom. She joined Horizon in 2014 after five years as the chief editor of Nature Reviews Cancer. Her role at Horizon includes project management and research strategy. She has a Ph.D. in cell biology from the University of Birmingham.
Jonathan D. Moore is the chief scientific officer of Horizon Discovery, and joined Horizon to manage its target ID and validation alliances and became CSO in early 2014. He currently manages a $15-million portfolio of internal and partnered research investments. He has a Ph.D. in biochemistry/molecular biology from the University of Newcastle upon Tyne.