Deletion of the TP53 gene, an event seen in colorectal cancers, often occurs with co-deletion of a gene that encodes an enzyme subunit governing gene transcription. This creates a vulnerability ripe for therapeutic development. See Letter p.697
There are increasing numbers of cancer drugs that successfully target cell-growth proteins that are activated by mutation. But when proteins that restrict cell growth, called tumour suppressors, are inactivated by mutation or deletion, few therapeutic strategies exist. Often, tumour-suppressor loss creates a pathway-specific cellular dependency, such as the reliance of BRCA-mutant breast-cancer cells on the enzyme poly(ADP-ribose) polymerase1,2. Regrettably, we have no general strategy for targeting tumour-suppressor loss, and nowhere is a creative solution to this challenge more urgently needed than in the context of the tumour suppressor p53. Alteration of the p53-encoding gene TP53 is the most common genetic event observed among 12 of the most common cancer types (42% of tumours)3. Yet despite definitive biological characterization over decades of study, there is no drug that acts directly on the p53 pathway. On page 697 of this issue, Liu et al.4 report that deletion of TP53 in many colorectal cancers is often accompanied by deletion of one copy of a subunit of the enzyme RNA polymerase II. This co-deletion renders such hemizygous cells sensitive to polymerase inhibition by the natural toxin α-amanitin.
p53 is a transcription factor that activates the expression of cell-cycle checkpoint genes such as CDKN1A in response to DNA damage, thereby arresting cell proliferation. In cancer, alteration of TP53, either by mutation or deletion, promotes cell proliferation and survival. Mutant p53 lacks a capacity for transcriptional activation, and so direct-acting small molecules would need to have a compensatory, corrective function. Such molecules have proved elusive. Hemizygous deletion of TP53 results in reduced abundance of p53, establishing a rationale for stabilizing remaining p53 by targeting the protein MDM2 — a ubiquitin-ligase enzyme that promotes p53 degradation5. Derivatives of the small molecule nutlin-3a, which disrupts the MDM2–p53 interaction, have recently transitioned to human clinical investigation. But there remains a pressing need for therapeutics that target p53 loss.
Deletions in cancer genomes are not limited to tumour-suppressor genes, and often involve bystander or 'passenger' genes. There is an increasing appreciation that collateral genetic damage to passenger genes may create unique vulnerabilities in cancer cells. For example, a study6 of the aggressive brain tumour glioblastoma multiforme found that the common causative mutation — deletion of both copies of the gene that encodes the metabolic enzyme enolase 1 (ENO1) — creates a heavy reliance on the remaining copies of the gene encoding the back-up enzyme enolase 2 (ENO2). Drugs that target ENO2 selectively killed ENO1-deleted cancer cells.
Another study7 built on the observation that most human tumour cells harbour partial loss of at least 10% of their genome, typically as a result of the loss of whole chromosomes or chromosome arms. Among partially lost genes, the authors of that study found that hemizygous deletion of the gene PSMC2, which encodes a regulatory subunit of the essential proteasome protein-degradation complex, rendered ovarian-cancer cell lines highly sensitive to reduced PSMC2 expression (generated by the technique of RNA interference), both in vitro and in vivo.
Liu et al. extend this logic to study hemizygous loss of TP53 in colorectal cancer. The authors find that regional deletions on chromosome 17 commonly involve both TP53 and a neighbouring gene, POLR2A. This gene encodes the largest subunit (subunit A) of RNA polymerase II, which transcribes messenger RNA from protein-coding genes, an essential function of all human cells. The authors report that hemizygous loss of POLR2A results in decreased abundance of RNA polymerase II in colorectal-cancer cells, leading them to suggest that these altered cancer cells could be more vulnerable to transcriptional inhibition than unaltered normal cells.
Using chemical and genetic approaches in defined colorectal-cancer cell lines, Liu et al. observed a crucial dependence on the remaining copy of POLR2A (Fig. 1). They found that hemizygosity for POLR2A renders colorectal cancer cells ten times more sensitive to α-amanitin, and highly sensitive to silencing of POLR2A-gene expression by RNA interference. Careful controls validated these findings, including desensitization to α-amanitin by restoration of normal POLR2A expression, and comparisons of engineered paired cell lines with or without hemizygous alteration of POLR2A.
To assess the potential clinical relevance of these findings, the authors conducted RNA-interference and α-amanitin-efficacy studies in immunocompromised mice bearing human tumours. To overcome the inadequacies of α-amanitin as a therapeutic agent (such as its systemic toxicity), an antibody–drug conjugate was used to deliver α-amanitin to tumours through binding to the cell-surface protein EpCAM. In all studies, sensitivity of POLR2A-hemizygous cells to transcriptional inhibition was confirmed.
This discovery opens a therapeutic window for transcriptional inhibition in a common, genetically defined subtype of human cancer. Limitations of the approach include the likelihood of drug resistance arising through duplication (or further amplification) of the remaining intact POLR2A gene, and the reliance on the natural poison α-amanitin. Nevertheless, these studies establish powerful reagents and a strong rationale for studying inhibitors of transcription and transcriptional signalling in cancer. Immediate next steps for this research are likely to include assessment of inhibitors of transcriptional kinase enzymes (for example, CDK7 inhibition by THZ-1) and modulators of chromatin-associated transcription factors (such as BET bromodomain inhibition by JQ1). Altogether, this study provides further validation that investigation of genomic loss can reveal previously unrecognized and actionable dependencies, establishing hope of a therapeutic inroad to managing TP53 alteration in cancer.Footnote 1
Bryant, H. E. et al. Nature 434, 913–917 (2005).
Farmer, H. et al. Nature 434, 917–921 (2005).
Kandoth, C. et al. Nature 502, 333–339 (2013).
Liu, C. et al. Nature 520, 697–701 (2015).
Wade, M., Li, Y.-C. & Wahl, G. M. Nature Rev. Cancer 13, 83–96 (2013).
Muller, F. L. et al. Nature 488, 337–342 (2012).
Nijhawan, D. et al. Cell 150, 842–854 (2012).
About this article
Journal of Agricultural and Food Chemistry (2019)
Highly Selective 5-Formyluracil Labeling and Genome-wide Mapping Using (2-Benzimidazolyl)Acetonitrile Probe
HMGB1 represses the anti-cancer activity of sunitinib by governing TP53 autophagic degradation via its nucleus-to-cytoplasm transport
Synergistic cytotoxic effects of a combined treatment of a Pinellia pedatisecta lipid-soluble extract and cisplatin on human cervical carcinoma in vivo
Oncology Letters (2017)
FXR1 regulates transcription and is required for growth of human cancer cells with TP53/FXR2 homozygous deletion