HIF1α delays premature senescence through the activation of MIF

Abstract

Cellular senescence has emerged as a programmed cellular stress response that is induced due to the accumulation of damage to a cell. Whether through the shortening of telomeres associated with a high number of cell divisions, activation of oncogenes, or DNA damage due to oxidative stress, induction of senescence in primary cells leads to an irreversible arrest phenotype that is characterized by the activation of the p53 and Rb proteins, as well as extensive chromatin modifications associated with the silencing of S-phase-promoting genes (Narita et al. 2003). In this way, senescence can be seen as a tumor suppressor mechanism that prevents excessive cellular divisions, or division of damaged cells (Ben-Porath and Weinberg 2005).

There is increasing evidence that senescence plays a critical role as a tumor suppressor in vivo. Senescent cells have been found recently in early-stage human prostate cancer specimens and premalignant melanocytic nevi, as well as in experimental models of lung adenocarcinoma and Ras-driven lymphoma (Braig et al. 2005; Chen et al. 2005; Collado et al. 2005; Michaloglou et al. 2005). These findings not only substantiate the significance of in vitro models of senescence, but also suggest novel therapeutic avenues aimed at reinitiating senescent programs in malignant cells (Lowe et al. 2004; Sharpless and DePinho 2005).

In normal tissue culture conditions, murine embryonic fibroblasts (MEFs) survive eight to 10 population doublings before they undergo premature cellular senescence. This senescence has been attributed to “culture shock” due to the nonphysiological conditions in which cells are grown (Sherr and DePinho 2000). Accordingly, it has been recently observed that senescence of MEFs in vitro can be abrogated by maintaining cells in a more physiological oxygen environment (3% O₂) (Parrinello et al. 2003). The mechanism leading to this reduction in senescence appears to be tied to the decrease in DNA damage that the cells endure in hyperoxic tissue culture conditions that are termed “normoxia” (i.e., atmospheric O₂ levels, 21%). A similar result has also been observed for oncogene-induced senescence, in which oncogenes such as Ras can induce premature senescence of MEFs (Serrano et al. 1997; Lee et al. 1999). When grown in the presence of antioxidants or in lower oxygen tensions, however, Ras-expressing MEFs continue to proliferate (Lee et al. 1999). These data suggest that reduction in oxidative damage due to reactive oxygen species (ROS) is sufficient to inhibit premature senescence.

Normal tissues are not typically exposed to the high levels of oxygen found in the environment. In vivo oxygen levels range from roughly 2%–3% in the brain, liver, and myocardium; 9%–10% in the spleen; and up to 13%–14% in the alveoli of the lung (Vaupel et al. 1989). Lower oxygen levels characterize normal and pathologic states including wound healing, ischemic disease, and cancer. The hypoxia-inducible factors (HIFs) are a family of transcriptional regulators that are important in the cellular response to hypoxia. They transcriptionally control a diverse number of genes including those involved in glycolytic metabolism, vascular remodeling, and erythropoeisis. HIFs are regulated primarily at the level of protein stability by the von Hippel Lindau protein (VHL), which directs the HIFα subunits to the proteosome for rapid degradation in oxic conditions (Kim and Kaelin 2004; Schofield and Ratcliffe 2004). In the mildly hypoxic conditions common to many tissues, HIFα subunits are stabilized and active (Stewart et al. 1982; Bedogni et al. 2005). Active HIF therefore correlates with resistance to premature senescence, both of which occur in physiological oxygen levels.

Whether HIFs play a direct role in preventing senescence under hypoxic conditions has not been determined. Recently, it has been observed in some endometrial cancer cell lines that modulation of the HIF pathway can affect senescence. Overexpression of a key negative regulator of HIFs (EGLN1), or use of a HIF inhibitor (YC-1), brought about proliferative arrest and the onset of a senescence-like phenotype (Kato et al. 2006). Collectively, these observations prompted us to ask whether HIFs were able to prevent senescence in a genetically defined system. In MEFs, the HIF1 complex is the primary functional member of the HIF family, as HIF2α is sequestered in the cytoplasm and held transcriptionally inactive (Park et al. 2003). Therefore, we used conditional loxP HIF1α MEFs to test the hypothesis that HIF1 may act to delay premature senescence. We found that HIF1α does play a role in the enhanced proliferation of cells under low oxygen conditions, and modulates senescence in response to hyperoxic conditions (21% O₂). This effect was also seen when oxidative stress was induced by γ-irradiation to MEFs maintained in physiological oxygen tensions. We further found that the anti-senescent effect of HIF1α is mediated in part by the transcriptional regulation of the macrophage migration inhibitory factor (MIF). Together, these findings offer new insights into the modulation of the tumor suppressor mechanism of senescence and its regulation under physiological conditions.

Acknowledgments

Footnotes

References