Cancer becomes lethal when tumour cells spread from their primary site in the body to invade distant organs — a process termed metastasis. For this complex event to occur, the cells must invade their surrounding tissue, enter the bloodstream and colonize another location, where secondary tumours called metastases form. Several of the early steps in metastasis, including cell migration, can be induced by the abnormal activation of a normal developmental program called epithelial–mesenchymal transition (EMT), in which the epithelial cells that line body surfaces take on the characteristics of mesenchymal cells, which have migratory properties1. Writing in Nature, Wang et al.2 identify a previously unknown mechanism by which a molecule generated in cellular metabolism inhibits the induction of EMT and thereby restrains the metastasis of lung cancer in mice.
The molecules formed during metabolism can have a key role in supporting the survival, proliferation and metastasis of tumour cells. Cancer cells have a higher than normal level of nutrient uptake and altered metabolic pathways, and these properties ensure that tumours make the metabolites they need to grow3. When tumour cells migrate into the bloodstream, they experience cellular stress. This is characterized by an increase in molecules called reactive oxygen species, and metabolic alterations that counter such stress can promote metastasis3,4. But whether metabolic pathways affect other aspects of metastasis has been poorly understood.
To investigate this further, Wang and colleagues individually blocked the expression of 111 metabolic enzymes in human lung cancer cells that had arisen from epithelial cells. Using these cells grown in vitro, the authors found that inhibiting production of the enzyme UGDH impaired the migratory capacity of the cells. UGDH converts UDP-glucose (UDP-Glc) to UDP-glucuronic acid (UDP-GlcUA), which is needed to make polysaccharide molecules such as hyaluronic acid, a component of the extracellular matrix material in the tissues in which epithelial cells reside. Hyaluronic acid can activate receptors on the surface of cells to initiate EMT, and its accumulation in tumours is often associated with poor clinical outcome5.
Surprisingly, when the authors inhibited UGDH expression, cell migration was not impaired as a result of reduced levels of UDP-GlcUA or hyaluronic acid, but because of an accumulation of UDP-Glc. Because EMT in cancer cells is associated with an increase in their migration1, the authors investigated whether UDP-Glc has an effect on the induction of EMT. They found that depletion of UGDH, and hence UDP-Glc accumulation, was accompanied by a decrease in the stability of messenger RNA that encodes a transcription-factor protein called SNAIL. This transcription factor regulates the expression of genes associated with EMT1. When the authors engineered cancer cells so that SNAIL was produced, the cells migrated even when UGDH was depleted. These results indicate that UGDH acts in a pathway that regulates cell migration by affecting SNAIL production (Fig. 1).
How might a metabolic enzyme such as UGDH affect mRNA stability? The authors focused on HuR, a protein that binds to and stabilizes mRNA targets6, including the transcript that encodes SNAIL. They found that UDP-Glc binds directly to HuR, thereby preventing the protein from interacting with the mRNA that encodes SNAIL. The authors engineered a form of HuR that had mutations in the amino-acid residues predicted to coordinate its binding to UDP-Glc. Compared with cells that had wild-type HuR, those with the mutant form were found to be more likely both to form metastases in mice and to migrate in vitro across a membrane in a culture dish.
This suggests that an interaction between UDP-Glc and HuR prevents HuR from acting in a pathway to induce cellular programs that promote metastasis. When the authors injected tumour cells into mice and gave UDP-Glc to some of them, those that received UDP-Glc had fewer metastases than the animals that did not receive it.
There are intriguing hints that the authors’ findings might have relevance for human cancer. In lung cancer, the receptor EGFR is commonly activated by mutations7, and the authors found that an increase in signalling through this receptor is associated with increased stability of SNAIL-encoding mRNA in human lung cancer cells grown in vitro. They observed that EGFR activation triggers the phosphorylation (the addition of a phosphate group) of the amino-acid residue tyrosine 473 (Y473) in UGDH, inducing a physical interaction between HuR and UGDH.
Wang and colleagues speculate that phosphorylated UGDH bound to HuR causes the local conversion of UDP-Glc to UDP-GlcUA, thereby alleviating UDP-Glc’s inhibition of the interaction between HuR and SNAIL-encoding mRNA and promoting SNAIL accumulation (Fig. 1). The authors engineered human lung cancer cells to express UGDH lacking a tyrosine residue at position 473, and found that such cells formed fewer metastases in mice than did cells that had wild-type UGDH. Wang et al. also noted that, in people with lung cancer, phosphorylation of Y473 in UGDH was more common in metastases than in primary tumours, and that this phosphorylation was associated with a poor clinical prognosis.
The authors’ findings add to growing evidence that metabolites can affect gene-expression programs8. The best-known examples of this are cases in which metabolites provide substrates for enzymes that regulate gene expression by modifying chemical groups attached to DNA or to the histone proteins that bind to DNA. However, UDP-Glc instead affects gene expression by physically preventing the interaction between a protein and mRNA. How UDP-Glc specifically affects HuR’s interaction with SNAIL-encoding mRNA, without impairing its interaction with other mRNAs, is an open question. Given the links between SNAIL expression, EMT and the extracellular matrix, it is tempting to speculate that coupling the production of SNAIL to the metabolites that generate hyaluronic acid might be an efficient way to coordinate the changes in both metabolism and protein production that are needed to promote metastasis.
Thus, in contrast to the metabolites that accumulate through cancer-associated mutations in metabolic enzymes and promote tumour progression8, UDP-Glc limits progression. This discovery widens our horizons regarding the ways in which metabolites can influence cancer. Although it has long been recognized that the metabolic profiles of cancer cells differ from those of normal cells, we are only beginning to appreciate the complexity of the metabolic alterations involved in tumour growth.
Nature 571, 39-40 (2019)