Identification of cellular senescence-specific genes by comparative transcriptomics

Abstract

Cellular senescence is defined as permanent cell cycle arrest induced by various stresses. Although the p53 transcriptional activity is essential for senescence induction, the downstream genes that are crucial for senescence remain unsolved. Here, by using a developed experimental system in which cellular senescence or apoptosis is induced preferentially by altering concentration of etoposide, a DNA-damaging drug, we compared gene expression profiles of senescent and apoptotic cells by microarray analysis. Subtraction of the expression profile of apoptotic cells identified 20 genes upregulated specifically in senescent cells. Furthermore, 6 out of 20 genes showed p53-dependent upregulation by comparing gene expression between p53-proficient and -deficient cells. These 6 genes were also upregulated during replicative senescence of normal human diploid fibroblasts, suggesting that upregulation of these genes is a general phenomenon in senescence. Among these genes, 2 genes (PRODH and DAO) were found to be directly regulated by p53, and ectopic expression of 4 genes (PRODH, DAO, EPN3, and GPR172B) affected senescence phenotypes induced by etoposide treatment. Collectively, our results identified several proteins as novel downstream effectors of p53-mediated senescence and provided new clues for further research on the complex signalling networks underlying the induction and maintenance of senescence.

Proline dehydrogenase promotes senescence through the generation of reactive oxygen species

“…Since PRODH is known to produce ROS as a byproduct of proline oxidation (Donald et al, 2001), and ROS induce senescence through oxidative damage (Campisi, 2013;Salama et al, 2014;Lu and Finkel, 2008), we first examined whether overexpression of PRODH results in ROS accumulation. In accordance with our previous results (Nagano et al, 2016), ectopic expression of PRODH modestly induced senescence in U2OS cells, as judged using SA-β-Gal and EdU assays ( Fig. 3A-C).…”

“…Consistent with this, we showed that PRODH overexpression induces DNA damage. Since we have previously revealed that PRODH is upregulated by p53 during senescence (Nagano et al, 2016), these results indicate that PRODH upregulated by p53 in response to senescence-inducing stresses amplifies DNA damage through ROS generation, thereby inducing senescence.…”

“…PRODH is a p53-inducible gene as a result of its intronic p53 responsive elements (Polyak et al, 1997;Raimondi et al, 2013;Nagano et al, 2016), and upregulation of PRODH results in the formation of proline-dependent ROS (Donald et al, 2001). Although PRODH has been suggested to play a role in p53-induced apoptosis (Polyak et al, 1997;Hu et al, 2007), our results clearly show that PRODH is a crucial molecule in DNA damageinduced senescence.…”

“…However, ROS and the resulting oxidative damage to DNA can induce senescence as well as apoptosis (Chen and Ames, 1994;te Poele et al, 2002;Walker et al, 1991), and our previous results obtained from comparative transcriptomic analyses suggest that PRODH can preferentially function in senescence rather than apoptosis. Although we have observed that overexpression of PRODH in normal and tumor cells induces senescent phenotypes, such as loss of proliferative capacity and the expression of senescence-associated β-galactosidase (SA-β-Gal; encoded by GLB1) (Nagano et al, 2016), whether and how PRODH functionally contributes to the senescence program remains uncertain.…”

“…We have recently revealed, by comparing the transcriptome between senescent and apoptotic cells, that the proline dehydrogenase gene (PRODH) is upregulated specifically in senescent cells and that p53 binds to and activates PRODH in response to senescence-inducing DNA damage (Nagano et al, 2016). PRODH is a mitochondrial inner membrane-associated enzyme catalyzing the first and rate-limiting step in proline catabolism, forming Δ 1 -pyrroline-5-carboxylic acid (P5C), and notably ROS are generated as byproducts of the reaction (Phang et al, 2008).…”