All organisms have an internal circadian clock, which ensures that physiological functions occur at the right time of day — for instance, that the intestine, pancreas and liver are ready to metabolize food when you eat rather than when you are sleeping. Writing in Cancer Discovery, Dong et al.1 report that the cells responsible for initiating a specific type of aggressive brain tumour, glioblastoma, rely on an altered circadian clock to grow. What’s more, drug-based inhibition of the cells’ molecular clock can kill them.
Glioblastomas are the most prevalent and aggressive tumour of the central nervous system. Fewer than 6% of patients survive for five years after diagnosis2. Cells in a glioblastoma often have varied gene-expression profiles. This, coupled with the fact that glioblastoma-initiating stem cells (GSCs) act to maintain the tumour, means that glioblastomas can rapidly develop resistance to conventional therapies3. New treatments are therefore urgently needed.
Disruption of the circadian clock, either because of lifestyle choices or because of mutations in core clock genes, is associated with a higher incidence of tumours4. In some tissues, clock genes can be co-opted to promote cancer (they are said to act as oncogenes), whereas in others they act as tumour suppressors5–7. The origin of such differences is an open question that, when answered, will help researchers to identify the mechanisms by which tumour cells hijack the molecular clock machinery to increase their chances of survival.
Dong and colleagues show that two key clock genes, BMAL1 and CLOCK, are co-opted to act as oncogenes in glioblastoma. The authors first observed that the genes are essential for the survival and proliferation of GSCs in vitro. By contrast, neither differentiated glioblastoma cells nor normal neural stem cells (from which GSCs arise8) seem to depend on the genes in this way. The authors validated these findings by showing a strong correlation between the expression of some of the core clock components and patient outcomes.
The researchers went on to show that a process called circadian reprogramming might explain why GSCs depend on the circadian clock. Circadian reprogramming involves changes in the circadian-clock output — that is, in the collection of genes in a given cell or tissue that are under the control of the clock, and so are expressed in oscillating rhythms across the day. Dong et al. demonstrated that the circadian-clock output of GSCs includes genes involved in glucose metabolism and lipid synthesis, whereas the circadian-clock output of normal neural stem cells does not. Changes in glucose metabolism and lipid synthesis have been previously shown to aid cancer progression9.
In addition, Dong and colleagues observed that the metabolic capacity of GSCs changed in the absence of BMAL1 and CLOCK. The group showed that circadian reprogramming in GSCs is mediated by changes in chromatin — the DNA–protein complex in which DNA is packaged. More regions of chromatin are open in GSCs than in normal neural stem cells, allowing the BMAL1 and CLOCK proteins to bind to and activate different genes. The authors then linked these data by showing that BMAL1 and CLOCK regulate the expression of genes involved in lipid metabolism in GSCs, indicating that the oncogenic activity of the clock genes might involve metabolic pathways (Fig. 1).
Previous reports have described circadian reprogramming in response to various stimuli, such as changes in diet, physiological ageing or exercise10–13. In all these cases, circadian reprogramming is a fast and effective way to respond to changing external demands. Circadian reprogramming has also been observed between organs — for instance, reprogramming in the livers of mice that have developed lung cancer probably ensures that the liver provides sufficient energy for the tumour cells to grow efficiently14. The picture that is emerging is of circadian reprogramming as a common mechanism to help cells, tissues and whole organisms adapt to change, whether they are healthy or cancerous.
In a final set of experiments, Dong et al. showed that small molecules that repress BMAL1, either directly or indirectly, strongly inhibit the self-renewing potential of GSCs. Mice that carried GSCs from patients survived longer if they were treated with one of these molecules than they did without treatment.
Caution is needed when considering translating these findings to humans, because the small molecules used by Dong and colleagues also affect the activity of the clock machinery systemically, potentially perturbing normal physiological processes in healthy tissues — this might induce damage accumulation and signs of premature ageing15. A better alternative might be to target the factors that induce circadian reprogramming in GSCs. Such an approach should block circadian-related changes in gene expression in cancer cells without perturbing the clock in the rest of the organism.
What might these factors be? There is likely to be a mixture, some intrinsic to the cells, others extrinsic, probably acting synergistically. For example, as in the current study, a change in energy requirements when a cell becomes cancerous can lead to changes in the metabolic products generated in that cell; this, in turn, can affect chromatin remodelling, changing the catalogue of genes available to be activated and thereby altering the rhythmic transcription of genes16. Outside the cell, signalling pathways involving the hormone insulin and the neurotransmitter molecule adrenaline are both altered in tumours, and can re-entrain the cancer-cell clock, thus integrating whole-body information into the cell’s circadian output15.
These systemic pathways might represent therapeutic targets to treat cancer. However, the complex effect of these pathways on circadian reprogramming in cancer cells is still poorly understood. Nonetheless, Dong et al. have opened a new chapter in the search for therapeutic targets for aggressive and incurable glioblastomas.
Nature 574, 337-338 (2019)