The p53 tumor suppressor gene plays a critical role in maintaining tissue homeostasis through both transcriptionally dependent and independent mechanisms. The response of p53 is dictated by the type of stress (genotoxic, oncogenic, hypoxia, metabolic, etc.) through a complex and incompletely understood number of post-translational modifications. Wild type p53 promotes apoptosis, cell cycle arrest and senescence in response to growth-restrictive conditions, including glucose deprivation.^1,2 Under such conditions, nuclear p53 has also been reported to stimulate autophagy through the transcriptional activation of genes that comprise the autophagic network, such as ULK1 and DRAM1, and modulators of autophagy, such as Sestrin2.^3–6 Unlike nuclear p53, cytoplasmic p53 appears to inhibit autophagy through mechanisms that are not well understood, but appear to involve localization of p53 to the endoplasmic reticulum (ER) and binding to FIP200, the human ortholog of yeast Atg17.⁷ In fact, some cancer-associated forms of mutant p53 found predominantly in the cytoplasm can also inhibit autophagy.⁸ This negative regulation of autophagy by p53 has been proposed to act as rheostat to prevent an excessive amount of autophagy from occurring. Thus, the role of p53 in autophagy is complex and may be dependent on the autophagic stimulus and the mutational status of p53 in the cell.

Studies describing the role p53 in the response to metabolic stress have focused primarily on the wild-type version of the protein. However, the vast majority of human tumors express high levels of mutant p53 protein that have acquired novel tumor-promoting functions distinct from those of wild type p53.⁹ To date, little is known about how mutant versions of p53 respond to growth-restrictive conditions, including the absence of glucose. In an elegant paper by Rodriguez et al.,¹⁰ the authors explored the effects of glucose restriction on the stability of a number of common cancer-associated p53 mutants and found that glucose deprivation resulted in degradation of mutant p53 protein levels. Interestingly, the negative regulation of p53 by glucose restriction was specific to mutant p53, since glucose restriction had a stabilizing effect on wild type p53. Surprisingly, the decreased levels of mutant p53 proteins were associated with rapid deacetylation and degradation through an autophagy-dependent but proteasome-independent process. Through the use of a constitutive acetylation-mimetic mutant p53, the authors demonstrated that autophagic degradation of p53 was dependent on the acetylation status of the protein. A major consequence of mutant p53 degradation in tumor cells after glucose deprivation is the loss of a critical check on the autophagic process that results in increased autophagy and leads to cell death (Fig. 1). Importantly, wild type p53 has been previously demonstrated to protect cells from glucose deprivation through induction of a reversible G₁/S phase cell cycle arrest, suggesting that normal tissues will respond to glucose shortage differently than tumors harboring mutant p53.¹