Age-associated decline of MondoA drives cellular senescence through impaired autophagy and mitochondrial homeostasis

Cellular senescence describes coordinated growth arrest frequently resulting in stalling of the cell cycle between the G1 and S phases. One of the hallmarks of senescence is the induction of the senescence-associated secretory phenotype, which is characterized primarily by inflammatory cytokines. Senescence is triggered by a variety of factors like stress, DNA damage, and mitochondrial dysfunction. Mitochondrial dysfunction can result in increased cellular damage from reactive oxygen species further necessitating the induction of senescence. Various forms of senescence can be alleviated by intracellular antioxidant systems like thioredoxin and peroxiredoxins. Senescence can be induced to maintain tissue homeostasis, as a protective stress response, or as an anti-cancer mechanism. Conversely, senescent cells drive aging and the pathology of many age-related diseases.¹

There is growing evidence that autophagy plays a role in aging and cellular senescence, but the precise mechanism is unknown. This hypothesis is supported by the fact that constitutive activation of autophagy, through activation of Beclin-1 or inhibition of Rubicon, leads to longevity in C. elegans, Drosophila, and mice. To investigate the role of autophagy in senescence in a mammalian system, two different cell lines were examined: human retinal pigment epithelial cells were treated with DNA damaging agent doxorubicin to induce DNA damage-induced senescence, and human lung fibroblast like cells at late passage (>70 passages) to induce replicative senescence. Under both mechanisms of senescence induction, autophagy (measured by LC3 flux) was decreased and levels of the autophagy inhibitor, Rubicon, were increased. Additionally, siRNA knockdown of autophagy inducing factors ATG7 and ATG13 increased the expression of senescent markers, secretion of senescent associated factors, and numbers of senescent cells independent of mitochondrial dysfunction. Furthermore, depletion of Rubicon reduced senescent cell numbers but did not rescue cell cycle arrest after DNA damage-induced senescence. These data indicate an inverse relationship between autophagy induction and cellular senescence.²

MondoA is a key transcription factor in autophagy and senescence

MondoA is a transcription factor that can dimerize with Max-like protein X (Mlx). These transcription factors shuttle between the cytosol and nucleus depending on intra and extra-cellular signals such as glucose levels and mitochondrial energy levels.³ Additionally, a longevity study in C. elegans identified MondoA and TFEB, an autophagy and lysosome master regulator, as mutual regulators. TFEB was found to play an important role in MondoA expression and both modulated each other’s nuclear translocation. This finding was also observed in HeLa cells providing support for MondoA playing a role in both longevity and autophagy in mammals.⁴

To further examine the role of MondoA in senescent cells, the DNA damage-induced senescence model was used. While MondoA mRNA expression levels were not affected by knockdowns of ATG7 or ATG13, MondoA siRNA knockdown decreased LC3 autophagic flux and decreased expression of some autophagy genes (e.g., ATG7 but not ATG5). Additionally, MondoA, but not Mlx, knockdown increased the number of senescent cells and enhanced cellular senescence expression signatures. These data implicate a pathway in which MondoA induces transcription of autophagy genes to counter cellular senescence.²

Transcriptome analysis of MondoA-deficient cells during DNA damage-induced senescence showed increased expression of pathways related to the senescence-associated secretory phenotype (IFN-γ responses, cytokine production, IκBK-NF-κB signaling) and decreased expression of pathways related to cell division, kinase binding, and stress responses. These pathways align with MondoA playing a role in cellular senescence. To further elucidate the pathway by which MondoA regulates senescence, the authors examined fifteen genes that were downregulated when MondoA-deficient cells treated with doxorubicin. They found that Prdx3 was drastically decreased and that knockdown of Prdx3 lead to alterations in protein levels of senescent markers.

Prdx3 

Prdx3 is a member of the peroxiredoxin family, which catalyzes the reduction of peroxides thereby protecting cells from oxidative stress. This protein family has also been shown to be involved in cell cycle regulation. Interestingly, Prdx3 - like MondoA - can localize to the mitochondria. This led researchers to investigate whether Prdx3 and MondoA contribute to mitochondrial function and whether mitochondrial dysfunction could drive senescence.

MondoA knockdown cells, Prdx3 knockdown cells, and senescence-induced cells had increased mitochondrial tubular network formation indicative of cell cycle stalling at the G1-S phase. MondoA- and Prdx3-deficient cells had decreased mitochondrial fission. MondoA and Prdx3, therefore, play a role in mitochondrial homeostasis and function, and disruptions in these processes can lead to cellular senescence. Knockdown of other mitophagy and mitochondrial homeostasis genes also enhanced cellular senescence, further linking mitochondrial dysfunction to senescence.

Linking mitochondrial dysfunction with senescence and autophagy

MondoA interaction with Prdx3 affects cellular senescence through mitochondrial function and regulates Rubicon to impact autophagy, but the question remains whether these pathways are linked or are independent. Knockdown of Rubicon did not affect Prdx3 expression levels and knockdown of Prdx3 did not affect Rubicon expression levels. Additionally, knockdown of Rubicon in MondoA-deficient cells rescued autophagy signaling, but it did not inhibit mitochondrial dysfunction. This work supports the concept that MondoA influences cellular senescence through two independent pathways: Prdx3-regulated mitochondrial function and Rubicon-regulated autophagy.

The kidney is a site of senescent cell accumulation, and cellular senescence accelerates kidney fibrosis after acute kidney injury. MondoA, Prdx3, and Rubicon were investigated in the context of this pathology. In aged mouse and human kidneys, Prdx3 expression was decreased, and Rubicon expression was increased. MondoA knockout mice had increased cellular senescence after acute kidney injury. Human tissue biopsies showed decreased MondoA protein levels in samples from aged and acute kidney injury tissue compared to samples from healthy, young tissue. The aged and acute kidney injury samples also showed elevated levels of cellular senescence markers. Whether the cells with decreased MondoA expression are the cells with a senescent phenotype is an ongoing area of investigation.

This paper provides a mechanistic link between autophagy and cellular senescence. It also provides a potential pathway whereby acute kidney injury progresses to fibrosis. These data provide additional therapeutic potential for age-related diseases caused by cellular senescent cell phenotypes.