Deleted in Breast Cancer 1 regulates cellular senescence during obesity

Chronic obesity leads to inflammation, tissue dysfunction, and cellular senescence. It was proposed that cellular senescence during obesity and aging drives inflammation and dysfunction. Consistent with this, clearance of senescent cells increases healthspan in progeroid mice. Here, we show that the protein Deleted in Breast Cancer-1 (DBC1) regulates cellular senescence during obesity. Deletion of DBC1 protects preadipocytes against cellular senescence and senescence-driven inflammation. Furthermore, we show protection against cellular senescence in DBC1 KO mice during obesity. Finally, we found that DBC1 participates in the onset of cellular senescence in response to cell damage by mechanism that involves binding and inhibition of HDAC3. We propose that by regulating HDAC3 activity during cellular damage, DBC1 participates in the fate decision that leads to the establishment of cellular senescence and consequently to inflammation and tissue dysfunction during obesity.


Introduction
Obesity, a major health problem in the USA and many developed countries (Flegal et al., 2012), is associated with an increase in cellular senescence and inflammation (Tchkonia et al., 2010). Cellular senescence has been proposed to promote chronic, "sterile" inflammation through the senescence-associated secretory phenotype (SASP) (Tchkonia et al., 2010). Supporting this notion, some of us found that eliminating senescent cells from progeroid mice improves healthspan (Baker et al., 2011). The physiological and molecular events that lead to cellular senescence, however, are still poorly understood. We have been studying the role of the protein Deleted in Breast Cancer-1 (DBC1) in energy metabolism (Chini et al., 2013). DBC1 regulates several nuclear proteins, including SIRT1 and HDAC3 Chini et al., 2013). Both SIRT1 and HDAC3 regulate cellular senescence (Ghosh, 2008;Feng et al., 2009). We investigated whether DBC1 plays a role in cellular senescence and the SASP during obesity. We found that preadipocytes isolated from WT and DBC1 KO mice after 12 weeks of high-fat diet feeding exhibit less senescence, indicated by lower levels of p16 Ink4a and p21, as well as the SASP markers, MCP-1, TNF-a, and IL-6 ( Fig. 1A). Also, we found fewer c-H2.AX (a marker of activated DNA damage responses)-positive preadipocytes isolated from DBC1 KO mice (Fig. 1B). Several markers of antioxidant defense mechanisms were upregulated in preadipocytes from DBC1 KO mice (Fig. 1C). Consistent with our in vitro results, DBC1 KO mice have less cellular senescence in adipose tissue during high-fat diet feeding measured by cytoplasmic ( Fig.  S1A) SA-bGal activity and p16 Ink4a expression ( Fig. 1D-F). The effect of DBC1 on cellular senescence may not be linked to chronological aging, as there was no difference between WT and DBC1 KO mice fed with normal chow during 16 months ( Fig. 1G-H). Nevertheless, DBC1 KO mice had less inflammation in fat tissue (Fig. 1H). We are currently investigating whether there is a difference on cellular senescence that may appear later in life.
Next, we investigated whether deletion of DBC1 protects against DNA damage-induced cellular senescence. We induced DNA damage by H 2 O 2 treatment in 3T3-L1 preadipocytes stable expressing scrambled shRNA (Control shRNA) or DBC1 shRNA. We found increased cellular SA-bGal activity in the control shRNA cells exposed to H 2 O 2 , but not in cells expressing DBC1 shRNA ( Fig. 2A). Control cells showed a dose-dependent increase in expression of p53 and p21 after H 2 O 2 treatment. However, there were no changes in p53 and p21 in cells expressing DBC1 shRNA ( Fig. 2A). The effect of DBC1 on the response to H 2 O 2induced DNA damage was only related to cellular senescence, as apoptosis was not affected by DBC1 knockdown (Fig. S1B). DBC1 binds and inhibits HDAC3 . Indeed, HDAC3 regulates DNA damage response (Bhaskara et al., 2010) and inhibits expression of the senescence mediator p16 Ink4a (Zheng et al., 2012). We found that the effect of DBC1 knockdown on senescence was completely abrogated by cotransfection with HDAC3 siRNA, but not by SIRT1 siRNA (Fig. 2C and Fig. S1C). Indeed, DBC1 knockdown increased HDAC3 activity in 3T3-L1 cells (Fig. 2D). Furthermore, HDAC3 siRNA, restored p21 expression driven by H 2 O 2 treatment in DBC1 shRNA-expressing cells ( Fig. 2E and Fig. S1D). Also, knockdown of DBC1 resulted in less c-H2.AX-positive cells ( Fig. 2F and Fig. S1E), an effect that was lost when HDAC3 was knocked down together with DBC1 (Fig. 2F). Interestingly, treatment with H 2 O 2 led to a rapid increase in DBC1 binding to HDAC3 (Fig. 2G), which correlated with an increase in histone H3 acetylation (Ac-H3K9, Fig. 2H), a target site for HDAC3 (Bhaskara et al., 2010). Finally, we found that DBC1 is present in both p16 and p21 promoter regions in 3T3-L1 cells (Fig. 2I), with a binding profile similar to the one of HDAC3 (Fig. S1F), which suggests that DBC1 binding to the chromatin is bridged by HDAC3. DBC1 is regulated by the checkpoint kinase ATM (Yuan et al., 2012), and HDAC3 is required for the DNA damage response Deleted in Breast Cancer-1 and Senescence, C. Escande et al. (Bhaskara et al., 2010). We propose that during the cellular stress driven by obesity, DBC1 has an active role in the onset of cellular senescence and inflammation. It is plausible that in the event of chronic damage or stress, DBC1 plays a role in checkpoint control, contributing to a switch in cell fate and promoting cellular senescence.

Supporting Information
Additional Supporting Information may be found in the online version of this article at the publisher's web-site.
Fig. S1 (A) DAPI counterstaining of fat tissue SA-bGal staining described in Figure 1D, showing cytoplasmic localization of the bGal signal.  Figure 2E. (E) Quantitation of the effect of DBC1, SIRT1, and HDAC3 siRNA on c-H2.AX foci in 3T3-L1 preadipocytes after incubation with H 2 O 2 (200 lM) shown in Figure 2F. Connecting lines show significant differences between conditions (P < 0.05, ANOVA, n = 3). (F) Chromatin immunoprecipitation (ChIP) for the p21 and p16 promoter regions in 3T3-L1 preadipocytes using an antibody against HDAC3. Nonspecific IgG was used as control. The results shown are the average AE SEM of 4 independent ChIP. (*P < 0.01; t-test).
Deleted in Breast Cancer-1 and Senescence, C. Escande et al. 953