Loss of p53 protein strikes a nerve to aid tumour growth

Malignant tumours are a complex, yet organized, diverse ensemble of cells. Tumour cells are surrounded by other types of cell, which collectively form the tumour microenvironment. Components of this microenvironment include fibroblast cells, which can promote the growth and spread of tumours to distant sites, and immune cells. The latter have antitumour functions that are often suppressed by cancer cells; indeed, therapies that boost such immune cells are revolutionizing the treatment of certain cancers. By contrast, the interactions between cancer cells and neurons in the tumour microenvironment are less-well understood. Writing in Nature, Amit et al.1 reveal how tumours influence neurons to promote tumour growth, and show how this discovery might lead to new anticancer therapies.

The interplay between cancer and neurons has negative clinical consequences for people with prostate tumours2. Individuals who have a higher number of new neurons (in structures called nerve fibres) in the tumour microenvironment tend to have more-aggressive tumour features, such as further tumour growth and migration to distant sites, and a decrease in survival time2.

Studies last year found that cancer cells and neurons can interact directly with each other through connections called synapses, that these connections aid the growth of brain tumours35 called gliomas, and that this interaction is associated with lethal cancer spread5. These and other findings36 contribute to a growing body of evidence that neurons are crucial components of the tumour microenvironment. However, what prompts the formation of neurons in the microenvironment had not been understood until now.

Amit and colleagues took on this challenge by focusing on tumours known as head and neck cancers, which can arise in the oral cavity. In humans, these tumours are often characterized by mutations that inactivate the gene TP53. This gene encodes a protein (p53) that functions as a tumour suppressor and that can modulate the tumour microenvironment7. By analysing four different mouse models of this disease and data obtained from biopsies of people with head and neck cancer, the authors found that tumours with mutant versions of p53 have a higher number of associated newly formed neurons than do those with wild-type p53. Moreover, an increased number of such neurons correlated with a shorter survival time.

To try to determine whether cancer cells with mutant p53 might stimulate neurons to form, Amit et al. analysed the factors released by human cancer cells that have mutant or wild-type p53. Both types of cell secreted vesicles that contained small RNA molecules called microRNAs (miRNAs). The vesicles in the two cell types were of a comparable number and size, but their contents differed (Fig. 1). Only the vesicles secreted from tumours with mutant p53 were devoid of an miRNA termed miR-34a, which is a tumour suppressor. When vesicles from tumours lacking p53 were injected into mice with tumours that had wild-type p53, the tumours with wild-type p53 grew larger and had more surrounding neurons than normal, indicating that the contents of these vesicles drive the formation of new neurons. This is the first report showing that miR-34a, the main function of which is to keep in check the proliferation of normal and cancer cells8, is important in counteracting the formation of neurons in the tumour microenvironment.

Figure 1

Figure 1 | Tumours manipulate neighbouring neurons to boost cancer growth. Amit et al.1 analysed head and neck cancers using clinical data and mouse models. a, Tumour cells that expressed wild-type p53 protein released vesicles containing small RNA molecules called microRNAs (miRNAs) that were taken up by neighbouring neurons. An miRNA known as miR-34a blocks neuronal proliferation, and the neurons were maintained in their current state. By contrast, tumours that had a mutant version of p53 released vesicles that lacked miR-34a. In this case, neurons increased in number in the vicinity of the tumour, and these cells were reprogrammed as adrenergic neurons that express the molecule noradrenaline. These neurons had more axonal branches than did those near tumours that expressed wild-type p53. Interactions between adrenergic neurons and the tumour aided cancer growth. b, When mice received a transplant of p53-deficient tumour cells, treatment with a drug (carvedilol) that blocks adrenergic signalling pathways slowed tumour growth. This might provide a new therapeutic tool for targeting tumours that need neighbouring adrenergic neurons for their growth.

Amit and colleagues analysed how these newly formed neurons promote tumour growth. The authors examined the neurons present in tumours with mutant and wild-type p53. Intriguingly, in the former set, the neurons (presumably including those already present in the area where the tumour formed) had undergone a functional change to become a type of neuron known as an adrenergic neuron — which uses the adrenergic signalling pathway and is activated in the ‘fight-or-flight’ response. This adrenergic feature (which has hallmarks including expression of the molecule noradrenaline) was crucial for sustaining cancer growth.

Interestingly, previous epidemiological analysis9 revealed that the use of the drug carvedilol, which blocks adrenergic signalling and is prescribed for conditions such as high blood pressure, is associated with a reduced risk of cancer onset. Now, Amit et al. raise the question of whether carvedilol’s anticancer properties might be due to its ability to target adrenergic neurons, given the effectiveness of the drug in treating mice with p53-deficient tumours (Fig. 1). The authors’ findings are of particular interest because these insights might offer a way to combat the tumour-driven formation of adrenergic neurons and to counteract their tumour-promoting effects. It will be important to establish whether adrenergic neurons’ contribution to tumour growth is limited to just head and neck cancers that have mutant p53, or whether this phenomenon could also be a feature of other types of tumour, as suggested by the epidemiological evidence for carvedilol use9.

Mutant versions of the gene encoding p53 are among the most common alterations in certain human cancers, occurring in approximately 60% of colon cancers, 50–80% of lung cancers and 95% of ovarian tumours10. Given the high prevalence of p53 abnormalities in cancer, numerous efforts have been made to design compounds that target mutant p53 to force it to act like wild-type p53, and promising results have been obtained in early-phase clinical trials of such drugs11. It would be worth testing whether using both carvedilol and a drug that targets mutant p53 is more effective than either compound alone in treating these lethal forms of cancer.

Amit and colleagues’ discovery that the absence of functional p53 influences the formation of neighbouring neurons might have relevance for interpreting reports showing that fluctuations in the levels of wild-type p53 are observed in nerve regeneration12. Thus, the authors’ findings might have repercussions that reach beyond the field of cancer research to regenerative medicine. Perhaps therapies that modulate the activity of p53 will have a future role in aiding the repair or regeneration of neurons, an outcome that would make a profound difference to the lives of people who have neurodegenerative diseases or other types of nerve injury.

Nature 578, 367-369 (2020)

doi: https://doi.org/10.1038/d41586-020-00328-6


  1. 1.

    Amit, M. et al. Nature 578, 449–454 (2020).

    Article  Google Scholar 

  2. 2.

    Magnon, C. et al. Science 341, 1236361 (2013).

    PubMed  Article  Google Scholar 

  3. 3.

    Venkataramani, V. et al. Nature 573, 532–538 (2019).

    PubMed  Article  Google Scholar 

  4. 4.

    Venkatesh, H. S. et al. Nature 573, 539–545 (2019).

    PubMed  Article  Google Scholar 

  5. 5.

    Zeng, Q. et al. Nature 573, 526–531 (2019).

    PubMed  Article  Google Scholar 

  6. 6.

    Yu, K. et al. Nature 578, 166–171 (2020).

    PubMed  Article  Google Scholar 

  7. 7.

    Bieging, K. T. et al. Nature Rev. Cancer 14, 359–370 (2014).

    PubMed  Article  Google Scholar 

  8. 8.

    Zhang, L. et al. J. Exp. Clin. Cancer Res. 38, 53 (2019).

    PubMed  Article  Google Scholar 

  9. 9.

    Lin, C. S. et al. Int. J. Cardiol. 184, 9–13 (2015).

    PubMed  Article  Google Scholar 

  10. 10.

    Kandoth, C. et al. Nature 502, 333–339 (2013).

    PubMed  Article  Google Scholar 

  11. 11.

    Bykov, V. J. N. et al. Nature Rev. Cancer 18, 89–102 (2018).

    PubMed  Article  Google Scholar 

  12. 12.

    Krishnan, A. et al. Eur. J. Neurosci. 43, 297–308 (2016).

    PubMed  Article  Google Scholar 

Download references


Nature Briefing

An essential round-up of science news, opinion and analysis, delivered to your inbox every weekday.

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing