Double trouble for tumours

Article metrics

When some cancer cells delete a tumour-suppressor gene, they also delete nearby genes. It emerges that one of these latter genes has a key metabolic role, revealing a therapeutic opportunity that might be relevant for many tumours. See Letter p.119

During cancer progression, many cancer cells delete tumour-suppressor genes that block tumour development. This deletion process often also removes some neighbouring genes. The partial or complete loss of these neighbouring genes can make cancer cells vulnerable to therapeutic targeting — a concept known as collateral lethality1. On page 119, Dey et al.2 reveal an example of collateral lethality in pancreatic cancer that exploits changes in cellular metabolism.

Collateral lethality is based on the premise that cancer cells survive co-deletion of neighbouring genes that have essential functions because such a loss is functionally compensated for by a closely related gene elsewhere in the genome. Targeting these closely related genes should be lethal only to tumour cells that have lost the initial genes, offering a cancer-selective therapeutic strategy. A similar concept has been exploited clinically for tumours with a deficiency in BRCA genes, which have a role in a DNA-repair pathway3. These BRCA-deficient tumours have an impaired DNA-repair capacity and are vulnerable to the inhibition of an alternative DNA-repair pathway that is dependent on PARP proteins4. There have been promising clinical results for use of PARP inhibitors as a component of therapy for breast and ovarian cancer4.

Pancreatic ductal adenocarcinoma is an aggressive malignancy that has a dismal clinical prognosis. There is, therefore, an urgent need for improved treatments. Despite considerable advances in our understanding of pancreatic cancer biology, there is no lasting effective clinical therapy for late-stage pancreatic ductal adenocarcinoma, and current treatments provide an increase in survival of just a few months5,6.

Altered cellular metabolism is a characteristic of tumour cells, which creates an opportunity to develop new classes of anticancer agent7. Pancreatic cancer cells harbour extensively rewired metabolic networks because of the activity of the KRAS protein8. This has implications for drug development because molecules downstream of the KRAS pathway are potential cancer-specific targets for therapeutic intervention.

In addition to the high frequency of mutations of the KRAS gene that generate a tumour-promoting KRAS protein, pancreatic ductal adenocarcinoma is characterized9 by inactivation of some tumour-suppressor genes, including SMAD4. Dey et al. focused on the subset of pancreatic ductal adenocarcinoma cells that harbours a deletion of both copies of the SMAD4 gene, and they found that this subset is associated with the loss of neighbouring genes involved in essential metabolic pathways. One of these neighbouring genes is mitochondrial malic enzyme 2 (ME2). The authors noted that a decrease in ME2 expression leads to compensatory increased expression of the closely related protein mitochondrial malic enzyme 3 (ME3). The low expression of ME3 observed in normal cells makes the protein an appealing target for a cancer-specific therapeutic approach, because depletion of its function would affect only cancer cells that express ME3 at high levels and leave the normal cells unaffected.

ME2 and ME3 function in the conversion of the metabolic molecule malate to pyruvate, which is a key molecule in several metabolic pathways. These enzymes are also involved in the regeneration of the cofactor NADPH, which is needed for the metabolic processes that produce the nucleotide ATP, the cell's main energy-carrying molecule10.

The authors demonstrated that, on loss of ME2, the resulting compensatory increase in ME3 is sufficient to maintain the metabolic processes that lead to ATP production and support the high energetic and biosynthetic demands of proliferating cancer cells. Furthermore, the authors demonstrated that ME2 deletion makes cells vulnerable to ME3 loss in human pancreatic-cancer cell lines and mouse transplantation models of pancreatic cancer. They found that this sensitivity is due to several metabolic impairments of cancer cells that lack ME2, including deficiency in the function of mitochondrial organelles; low levels of NADPH; high levels of reactive oxygen species, which act in redox pathways; and impaired regeneration of the amino acid glutamate, an indispensable metabolite for pancreatic cancer cells. The authors found that deletion of both ME2 and ME3 thus results in cancer-cell death and impaired tumour maintenance (Fig. 1).

Figure 1: Pancreatic cancer cells have a metabolic vulnerability.

In normal cells, the tumour-suppressor gene SMAD4 neighbours the ME2 gene, which encodes an enzyme that has an essential role in cellular metabolism. Dey et al.2 observed that deletion of SMAD4 in human pancreatic ductal adenocarcinoma can be associated with the loss of ME2. This loss leads to an increase in expression of the closely related enzyme ME3, which can provide functional compensation for the essential metabolic functions carried out by ME2. Normal cells usually express low levels of ME3 and do not die if it is deleted. However, if ME3 is deleted in cancer cells that lack ME2, this causes cell death because of metabolic abnormalities that include low levels of the cofactor NADPH and of the amino acid glutamate, and high levels of reactive oxygen species (ROS). This approach of investigating metabolic vulnerabilities and the role of genes co-deleted with tumour-suppressor genes offers a new framework for finding therapeutic targets.

These findings reveal a previously unknown vulnerability of pancreatic ductal adenocarcinoma because of the cells' altered metabolism, and highlight the potential for drug targeting of the rewired metabolic pathways of cancer cells. Targeting of the altered redox state of pancreatic cancer has emerged11 as a promising therapeutic approach. In KRAS-driven cancer cells, the effect of excessive cellular oxidation produced by increased mitochondrial activity needs to be counteracted by high production of antioxidant molecules12. Impairing antioxidant function might drive cancer-cell oxidation to toxic levels, revealing additional possibilities for clinical approaches and providing justification for KRAS-driven cancer therapies designed to include oxidizing agents or other agents that target redox dependencies.

However, targeting metabolic and redox vulnerabilities of cancer cells could ultimately drive the development of treatment resistance. Therefore, improvements in our understanding of metabolic and redox networks will probably be needed to tackle the evasion of such selective therapies. Nevertheless, targeting the key molecules required to enable metabolic and redox rewiring in cancer cells is a promising therapeutic strategy. Although drugs targeting genes that promote cancer development (oncogenes) will doubtless continue to be a source of treatments, pursuing collateral lethality targets as an associated approach might also reveal therapeutic opportunities.

Dey et al. investigated a situation in which both copies of a metabolic gene that neighbours a tumour-suppressor gene were lost. However, the loss of one copy of other essential genes might also generate collateral dependencies. This single-copy loss might be specific to individual tumours or could be more widely relevant if linked to frequently lost tumour-suppressor genes. Such a view of cancer might expand horizons of potential therapeutic targets and enable the delivery of truly personalized medical care.Footnote 1


  1. 1.

    See all news & views


  1. 1

    Muller, F. L., Aquilanti, E. A. & DePinho, R. A. Trends Cancer 1, 161–173 (2015).

  2. 2

    Dey, P. et al. Nature 542, 119–123 (2017).

  3. 3

    Lord, C. J. & Ashworth, A. Nature 481, 287–294 (2012).

  4. 4

    Farmer, H. et al. Nature 434, 917–921 (2005).

  5. 5

    Von Hoff, D. D. et al. N. Engl. J. Med. 369, 1691–1703 (2013).

  6. 6

    Conroy, T. et al. N. Engl. J. Med. 364, 1817–1825 (2011).

  7. 7

    Vander Heiden, M. G. Nature Rev. Drug Discov. 10, 671–684 (2011).

  8. 8

    Ying, H. et al. Cell 149, 656–670 (2012).

  9. 9

    Biankin, A. V. et al. Nature 491, 399–405 (2012).

  10. 10

    Jiang, P., Du, W., Mancuso, A., Wellen, K. E. & Yang, X. Nature 493, 689–693 (2013).

  11. 11

    Chio, I. I. C. et al. Cell 166, 963–976 (2016).

  12. 12

    DeNicola, G. M. et al. Nature 475, 106–109 (2012).

Download references

Author information

Correspondence to Giulia Biffi or David A. Tuveson.

Related links

Related links

Related links in Nature Research

Cancer: The enemy of my enemy is my friend

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Biffi, G., Tuveson, D. Double trouble for tumours. Nature 542, 34–35 (2017) doi:10.1038/nature21117

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.