p53 is one of the most widely known tumour suppressors, which integrates a host of stress signals and orchestrate specific cellular responses, including transient cell cycle arrest, cellular senescence and apoptosis — all processes that prevent tumorigenesis.
Recent studies have challenged the relative importance of these actions, as brought to light in a review this week in Nature Reviews Cancer. The Comment piece also discusses more obscure roles of p53 in modulating metabolism, maintaining ‘stemness’, invasion, metastasis and communication within the tumour microenvironment.
p53 is inactivated in more than half of all sporadic human cancers
During tumour development, a TP53 mutation, either sporadic or inherited, is typically followed by loss of heterozygosity, which results in complete p53 deficiency. p53 deficiency can enhance the initiation or progression of cancer, depending on the tumour type, and tumours that lack p53 are commonly characterized by more malignant characteristics, such as a lack of cellular differentiation, genetic instability, and increased invasiveness and metastatic potential. These effects are probably conferred both by loss of wild-type p53 function and by oncogenic gain-of-function properties that characterize some p53 mutants.
Since it’s such a ubiquitous molecular actor, I was surprised to read that “the mechanisms that underlie p53-mediated tumour suppression remain incompletely understood… the downstream genes and pathways that are crucial for tumour suppression remain unresolved.”
What we know is that when a stressor is detected, p53 is displaced from the MDM2 and MDM4 which normally keep it on lockdown, thereby allowing it to stabilise and become activated.
I won’t go into the details of p53’s general tumour suppression effects as they’re not new, and the authors actually segregate the functions in cell cell cycle arrest, DNA repair, apoptosis and senescence as ‘the classical view’.
Far more compelling is the discussion of roles in regulation of -metabolism – a concept I’d not encountered until earlier today in a Nature Cell Biology paper, Metabolic control of YAP and TAZ by the mevalonate pathway.
This is a particularly nice one, tying together a central pathway in metabolism with p53 and Rho GTPases. Summary in brief:
In tumour cells, YAP/TAZ activation is promoted by increased levels of mevalonic acid produced by SREBP transcriptional activity, which is induced by its oncogenic cofactor mutant p53.
Statins [inhibit] the enzyme HMG-CoA reductase (HMGCR). This enzyme catalyses the production of mevalonic acid (MVA), which represents the rate-limiting step of cholesterol biosynthesis (the mevalonate pathway)… results suggested an unsuspected relevance of the mevalonate pathway in sustaining YAP/TAZ activity, as well as a role for statins as YAP/TAZ inhibitors.
statins inhibited YAP/TAZ activity in all of the cell lines tested, in a MVA-dependent manner. YAP/TAZ inhibition was obtained also on HMGCR depletion by RNAi, supporting the notion that an active mevalonate pathway is required for YAP/TAZ function… results demonstrated that the mevalonate pathway is required to sustain the YAP/TAZ gene expression program.
The mevalonate pathway is crucial for the biosynthesis of cholesterol but also of other crucial metabolites. Interestingly, only the farnesyl diphosphate synthase inhibitor (bisphosphonate) zoledronic acid and the geranylgeranyl transferase inhibitor GGTI-298 were able to reproduce the effect of statins on YAP/TAZ localization,
Adding back MVA could not rescue the inhibitory effect of zoledronic acid and GGTI-298, consistent with these inhibitors acting downstream to the HMGCR enzyme (Fig. 2a). In contrast, re-exposure to geranylgeranyl pyrophosphate (GGPP), but not to farnesyl phyrophosphate or squalene, rescued statin and bisphosphonate-dependent inhibition of YAP/TAZ, both in terms of localization and transcriptional activity. These findings demonstrate that protein geranylgeranylation is responsible for the positive effect of the mevalonate pathway on YAP/TAZ activity.
Investigating what geranylgeranylated factors may foster YAP/TAZ activity, we focused on the Rho family of GTPases, recently identified as one of the key upstream inputs that, by modulating actin cytoskeleton, positively control YAP/TAZ activity. The enzymatic activity of Rho GTPases relies on their ability to properly localize at the plasma membrane, a process promoted by the transfer of a geranylgeranyl moiety to a carboxy-terminal cysteine residue.
…statins inhibit YAP/TAZ by depleting cells of GGPP and, as such, of active Rho GTPases.
The link to statins is intriguing, since so many are using them as prophylaxis as well as treatment after diagnosis, and it’s always nice for an unintended consequence to be positive. The authors urge caution in interpreting their results, but still, sounds good to me.
Recently the provocative idea has been proposed that dysregulation of the mevalonate pathway, and expression of HMGCR itself, may have sufficient oncogenic potential to drive malignant progression and anchorage-independent growth, in line with the correlation of high HMGCR mRNA levels with poor patient prognosis and reduced survival. Our results suggest that aberrant YAP/TAZ may be the most likely candidate mediators of these responses, as we show that statins, bisphosphonates and GGT inhibitors work through YAP/TAZ regulation, and that these drugs have the potential to target YAP/TAZ malignant effects in cancer cells. However, it should be noted that the statin concentrations used here largely exceed the plasma concentrations of statins achieved in patients for cardiovascular disease prevention.
In sum, the discovery that YAP/TAZ is controlled by mevalonate and Rho GTPases reveals unexpected connections between metabolism, proliferation and stemness
Returning to Bieging et al, they outline a ‘revised view’ (main image, above) to include nutrient deprivation, hypoxia, oxidative stress and hyperproliferative signals (capable of promoting chronic DNA damage triggered by replicative/oxidative stress or telomere attrition) and ribonucleotide depletion in a broader idea of what tumour suppression really means.
Emerging functions of p53 came into the spotlight toward the end of the essay. ‘Policing metabolism’ through dampening the Warburg effect, p53 stimulates oxidative phosphorylation through activation of cytochrome oxidase 2 synthesis, inhibiting glycolysis through Tigar's transcriptional activation, and transcriptional repression of GLUT1&4.
Autophagy comes under this heading of metabolic regulation, in an clever little manoeuvre dependent on intracellular localisation:
Depending on the location of the p53 tumor suppressor protein, it plays a different role in regulating autophagy as well. When in the nuclear region, p53 acts as a transcription factor in order to activate DRAM1 and Sestrin2 which activates autophagy. In the cytoplasm, p53 inhibits autophagy. Thus, to induce autophagy, p53 is degraded through proteasomes.
The situation’s far more complex than this though, with autophagy promoting or inhibiting tumorigenesis dependent on the context. Its transcriptional activity as outlined above induces autophagy, but p53 also activates a host of target genes encoding proteins in various steps of the process (upstream regulators, core machinery components and lysosomal constituents).
Furthermore, the inhibition of autophagy through ablation of autophagy related 5 (Atg5) — a central component of the autophagy machinery — results in defective p53‑dependent apoptosis and promotes transformation of oncogene-expressing MEFs, which is similar to p53 loss.
p53 impedes tumorigenesis at least in part through the induction of autophagy. Activation of autophagy by p53 ensures efficient apoptosis, and it potentially has other tumour-suppressive effects, such as limiting ROS accumulation. These observations additionally highlight the crosstalk between the regulation of metabolism by p53 and canonical p53 functions such as apoptosis.
Other roles emerging through research include
- slowing down stem cells, suppressing the reprogramming of differentiated somatic cells which takes place via induction of Cdkn1a (p21); also through transactivating mir-145 and again inhibiting essential pluripotency genes via mir34ac
- imposing a barrier to invasion/metastasis (including to the infamous epithelial-to-mesenchymal transition) by modulating activities of the SNAIL, TWIST and ZEB transcription factor families and transactivation of mir-200c
- non-cell autonomous functions arising from crosstalk between tumour and its microenvironment through the likes of thrombospondin 1 and stimulating innate immune response via inflammatory cytokines.
One hypothesis is that p53’s mechanism of action may vary with tissue context (given to explain differing results between studies), which would have implications for the many cancers it’s implicated in.
If the components that are crucial for p53‑mediated tumour suppression are known, more reliable expression signatures that reflect functional p53 status can be used for diagnosis or prognostication. Moreover, identifying the key targets and pathways that are involved in the function of p53 in tumour suppression provides more flexibility for therapeutic intervention. As restoration of wild-type p53 function is not a trivial proposition, identifying a more targetable component or pathway downstream of p53 could be a key to attacking p53‑deficient tumours.
» Bieging (2014) Unravelling mechanisms of p53-mediated tumour suppression. Nature Reviews Cancer, 14, 359–370