Elsevier

Cellular Signalling

Volume 24, Issue 11, November 2012, Pages 2007-2023
Cellular Signalling

The canonical NF-κB pathway differentially protects normal and human tumor cells from ROS-induced DNA damage

https://doi.org/10.1016/j.cellsig.2012.06.010Get rights and content

Abstract

DNA damage responses (DDR) invoke senescence or apoptosis depending on stimulus intensity and the degree of activation of the p53–p21Cip1/Waf1 axis; but the functional impact of NF-κB signaling on these different outcomes in normal vs. human cancer cells remains poorly understood. We investigated the NF-κB-dependent effects and mechanism underlying reactive oxygen species (ROS)-mediated DDR outcomes of normal human lung fibroblasts (HDFs) and A549 human lung cancer epithelial cells. To activate DDR, ROS accumulation was induced by different doses of H2O2. The effect of ROS induction caused a G2 or G2-M phase cell cycle arrest of both human cell types. However, ROS-mediated DDR eventually culminated in different end points with HDFs undergoing premature senescence and A549 cancer cells succumbing to apoptosis. NF-κB p65/RelA nuclear translocation and Ser536 phosphorylation were induced in response to H2O2-mediated ROS accumulation. Importantly, blocking the activities of canonical NF-κB subunits with an IκBα super-repressor or suppressing canonical NF-κB signaling by IKKβ knock-down accelerated HDF premature senescence by up-regulating the p53–p21Cip1/Waf1 axis; but inhibiting the canonical NF-κB pathway exacerbated H2O2-induced A549 cell apoptosis. HDF premature aging occurred in conjunction with γ-H2AX chromatin deposition, senescence-associated heterochromatic foci and beta-galactosidase staining. p53 knock-down abrogated H2O2-induced premature senescence of vector control- and IκBαSR-expressing HDFs functionally linking canonical NF-κB-dependent control of p53 levels to ROS-induced HDF senescence. We conclude that IKKβ-driven canonical NF-κB signaling has different functional roles for the outcome of ROS responses in the contexts of normal vs. human tumor cells by respectively protecting them against DDR-dependent premature senescence and apoptosis.

Graphical abstract

Highlights

► H2O2 driven ROS activates NF-κB and arrests normal HDF and A549 tumor cell cycling. ► HDFs prematurely senesce while A549 cells become apoptotic in response to ROS. ► Suppressing NF-κB activation exacerbates the ROS-mediated HDF senescence response. ► Inhibiting canonical NF-κB enhances ROS-induced A549 tumor cell apoptosis. ► The p53–p21Cip1/Waf1 axis drives the ROS-mediated senescence response of HDFs.

Introduction

Oxidative DNA damage results from the accumulation of reactive oxygen species (ROS) due to an imbalance between ROS production and removal by antioxidant systems. ROS cause a variety of genetic damage, mostly base lesions and DNA single strand breaks (SSBs) but also double stranded DNA breaks (DSBs). In effect, replication can stall due to DNA damage resulting from ROS in the cell, and failure to stabilize stalled replication forks can result in their collapse and ultimately genetic instability [1], [2]. Oxidative damage can provoke cell cycle arrest and senescence, but it can also induce apoptosis with normal cells being more resistant to apoptosis than tumor cells [1], [2], [3].

Normal human diploid fibroblasts (HDFs) undergo a finite number of cell divisions in culture, a phenomenon termed cellular or replicative senescence. In vitro cell propagation causes telomere shortening, recognized as DSBs that activates a DNA damage checkpoint response (DDR), culminating in replicative senescence. In addition, human cells undergoing senescence and aging mice accumulate DNA lesions with irreparable DSBs outside telomeres, suggesting that their accumulation may have a causal role in mammalian aging [4], [5], [6]. DDR involves activation of the kinases ATM and Chk2 and their downstream effector p53 and its target p21Cip1/Waf1 [7], [8]. Normal HDFs can also undergo senescence in response to oxidative stress, referred to as stress-induced premature senescence (SIPS) [8], [9]. Hydrogen peroxide (H2O2) produces thymidine glycols that can lead to DSBs at replication forks [1] and to the appearance of DNA damage foci in HDFs [4], [9], [10]. ROS accumulation was shown to accelerate HDF senescence by DDR in conjunction with the induction and stabilization of p53 and p21Cip1/Waf1 [11], [12], [13], [14], [15], [16]. However, dependent on cell type and dosage, H2O2-mediated ROS accumulation has also been reported to induce the death of normal and tumor cells [1], [3], [17], [18].

NF-κB transcription factors are critical regulators of most if not all pro-inflammatory/stress-like responses. NF-κBs bind to DNA as dimers of five possible subunits (RelA/p65, c-Rel, RelB, p50, p52). Archetypical p65/p50 heterodimers are cytoplasmically restrained by IκBs (inhibitors of NF-κBs) in most cells. Canonical NF-κB activation generally requires the phosphorylation of serines 32 and 36 in ΙκΒα's signal response domain (SRD), causing its ubiquitination and subsequent proteasomal destruction, allowing p65/50 dimers to translocate to the nucleus and activate their target genes. IκBα SRD phosphorylation requires the IKK signalsome complex (IKKα and IKKβ serine–threonine kinases, and NEMO/IKKγ, a regulatory/adapter protein). IKKβ activation by phosphorylation of its T-activating loop serines 177/181 requires NEMO and rapidly occurs in response to a host of pro-inflammatory/stress-related extracellular signals. In contrast to IKKβ, IKKα activation, by phosphorylation of T-loop serines 176/180, is NEMO-independent, requires de novo protein synthesis and is mainly mediated by stimuli invoked in adaptive immune responses. In vivo, unlike IKKβ, IKKα does not generally target IκBs but instead phosphorylates an IκB-like signal SRD in NF-κB p100, which induces its ubiquitination/proteasome-dependent processing yielding the NF-κB p52 subunit. Thus, NF-κB p100 functions akin to IκBα by sequestering NF-κB RelB in the cytoplasm and also yields NF-κB p52, which translocates to the nucleus together with RelB, to activate other unique NF-κB target genes [19], [20]. Some p100 also sequesters a subset of p65/p50 heterodimers, whose activation can only be initiated by IKKα [21]. IKK catalytic activity also regulates other cellular responses independent of NF-κB activation. For example, IKKβ controls anti-apoptotic, pro-inflammatory and proliferative responses by targeting p53, FoxO and other cell cycle effectors [19], [20], [22].

NF-κB is activated by moderate ROS levels in a cell-type specific manner by different mechanisms, including the classical IKKβ-dependent pathway [23], [24]. In addition, genotoxic stimuli producing DSBs cause ATM/NEMO-dependent IKKβ-mediated canonical NF-κB activation, which facilitates cell survival and cycle arrest thereby giving cells precious time to repair their damaged DNA before p53-dependent apoptotic pathways are invoked [25], [26]. A role for IKKα in DDR invoked by oxidative stress or chemotherapeutics has also been suggested [27], [28]. Collectively, these studies provided evidence that the induction of NF-κB signaling by DNA damage is important for cell survival. However, to our knowledge little if any evidence is available on how canonical NF-κB activity functionally contributes to ROS-induced HDF senescence responses.

To address NF-κB's functional roles in ROS-mediated DDR in the context of normal vs. human tumor cells, we inhibited canonical NF-κB activation in MRC-5 and IMR-90 normal HDFs and suppressed NF-κB signaling in A549 human lung tumor epithelial cells. Inhibition of canonical NF-κB in both cell types exacerbated ROS initiated DDR by up-regulating the p53–p21Cip1 axis, leading to the enhanced accumulation of cells at the G2 or G2-M checkpoint. Importantly, the final physiological outcomes of ROS-induced DDR greatly differed in normal vs. human cancer cells with the former normal human diploid fibroblasts (HDFs) succumbing to premature senescence and the latter human lung cancer epithelial cells undergoing exacerbated apoptosis.

Section snippets

Cell culture and retroviral transduction

MRC-5, human embryonic diploid fibroblasts (HDFs), originally from the National Institute for Biological Standards and Controls, South Mimms, UK [29], were kept at the cell stock of the Genetics Division, National Institute for Medical Research, London UK, and obtained from Dr. Lily Huschtscha-Holliday and used in several previous studies by our group [29], [30], [31]. IMR-90 HDFs and human epithelial lung tumor cells A549 were originally obtained from ATCC and phoenix retroviral packaging

H2O2 activates a ROS-mediated DNA damage checkpoint response in normal and cancer cells

Increasing concentrations of H2O2 resulted in ROS accumulation in both normal HDFs MRC-5 and A549 lung cancer cells (Fig. 1A), reduced A549 colony formation (Fig. 1B) and also induced DDR characteristic changes in specific cellular proteins. After 2 h of H2O2 exposure followed by 24 h of recovery, phosphorylations were detected for p-ATM (S1981), p-Chk2 (T68) and γ-H2AX in both cell types (Fig. 2A and B). Moreover, p-p53 (Ser15), total p53 and p21Cip1 levels were also elevated in an H2O2

Discussion

Oxidative stress induces DNA damage leading to either growth arrest and senescence or apoptosis, depending on the nature and intensity of the stimulus and cell type (i.e., normal vs. tumor cells) [1], [2], [3]. Here we show that H2O2 induced oxidative stress initiated a DDR leading to G2 or G2-M arrest and the premature senescence of MRC-5 normal human lung diploid fibroblasts (HDFs), but instead caused the apoptosis of A549 human lung tumor epithelial cells. NF-κB signaling was induced by

Conflict of interest

The authors have no conflicting financial interests.

Acknowledgments

We thank Dr. R. Agami, The Netherlands Cancer Institute, Amsterdam, for pRetro.Super-Puro and pRS-shp53 retrovectors. This research was co-funded by EU (European Social Fund) and the Hellenic Ministry of Education within the framework of the program ‘Pythagoras IΙ’ Grant No: 1887 (EK); and also in part by PENED'03 (Program for Supporting Research Manpower) program grant No: 61/2055 and ‘THALIS’ grants (EK) of the General Secretariat for Research and Technology, the Empeirikeion Foundation,

References (104)

  • H. Li et al.

    Mechanisms of Ageing and Development

    (2008)
  • T. Lu et al.

    Experimental Cell Research

    (2008)
  • W.C. Burhans et al.

    Free Radical Biology & Medicine

    (2009)
  • O.A. Sedelnikova et al.

    Mutation Research

    (2010)
  • J.H. Chen et al.

    The Journal of Biological Chemistry

    (2004)
  • Y. Wang et al.

    Experimental Cell Research

    (2004)
  • J.H. Chen et al.

    DNA Repair

    (2005)
  • J. Duan et al.

    The International Journal of Biochemistry & Cell Biology

    (2005)
  • J.H. Chen et al.

    FEBS Letters

    (2006)
  • V. Gorbunova et al.

    The Journal of Biological Chemistry

    (2002)
  • A. Chariot

    Trends in Cell Biology

    (2009)
  • S. Basak et al.

    Cell

    (2007)
  • G. Gloire et al.

    Biochemical Pharmacology

    (2006)
  • T. Yamaguchi et al.

    Cellular Signalling

    (2007)
  • C. Batsi et al.

    Mechanisms of Ageing and Development

    (2009)
  • T.R. Brummelkamp et al.

    Cancer Cell

    (2002)
  • G.P. Dimri et al.

    Experimental Cell Research

    (1994)
  • S. Qanungo et al.

    The Journal of Biological Chemistry

    (2004)
  • M. Narita et al.

    Cell

    (2003)
  • S. Nowsheen et al.

    Mutation Research

    (2009)
  • S. Petersen et al.

    Experimental Cell Research

    (1998)
  • T. Dandrea et al.

    Free Radical Biology & Medicine

    (2004)
  • Z. Zhang et al.

    Archives of Biochemistry and Biophysics

    (2001)
  • Z. Zhang et al.

    Free Radical Biology & Medicine

    (2003)
  • K. Barnouin et al.

    The Journal of Biological Chemistry

    (2002)
  • R.E. Shackelford et al.

    Journal of Biological Chemistry

    (2001)
  • J.B. de Haan et al.

    Free Radical Biology & Medicine

    (2004)
  • J.M. Taylor et al.

    Experimental Cell Research

    (2004)
  • B. Ryu et al.

    Peptides

    (2011)
  • Y. Li et al.

    The Journal of Biological Chemistry

    (1997)
  • V. Lakshminarayanan et al.

    The Journal of Biological Chemistry

    (1998)
  • F. Antonicelli et al.

    Free Radical Biology & Medicine

    (2002)
  • I. Rahman et al.

    Biochemical and Biophysical Research Communications

    (2003)
  • Y. Hou et al.

    International Immunopharmacology

    (2009)
  • N. Loukili et al.

    The Journal of Biological Chemistry

    (2010)
  • C. Lawless et al.

    Experimental Gerontology

    (2010)
  • H. Zhan et al.

    The Journal of Biological Chemistry

    (2010)
  • V. Tergaonkar et al.

    Cancer Cell

    (2002)
  • T. Kuilman et al.

    Cell

    (2008)
  • D. Iliopoulos et al.

    Cell

    (2009)
  • S. Kapeta et al.

    The Journal of Biological Chemistry

    (2010)
  • Y. Nakabeppu

    Mutation Research

    (2001)
  • J.A. Kazzaz et al.

    The Journal of Biological Chemistry

    (1996)
  • W.R. Franek et al.

    The Journal of Biological Chemistry

    (2001)
  • O.A. Sedelnikova et al.

    Nature Cell Biology

    (2004)
  • A.J. Nakamura et al.

    Epigenetics Chromatin

    (2008)
  • O.A. Sedelnikova et al.

    Aging Cell

    (2008)
  • J. Bartkova et al.

    Nature

    (2006)
  • J. Campisi et al.

    Nature Reviews. Molecular Cell Biology

    (2007)
  • G. Hewitt et al.

    Nature Communications

    (2012)
  • Cited by (40)

    • Genes and pathways involved in senescence bypass identified by functional genetic screens

      2021, Mechanisms of Ageing and Development
      Citation Excerpt :

      ROS are products of normal cellular metabolism, but when accumulated at high levels (e.g., due to dysfunctional mitochondria) can cause protein, lipid and DNA damage, known as oxidative stress. High intracellular ROS levels induced by the Ras-Raf-MEK-ERK signalling pathway activate the p38MAPK, which leads to the activation of the p53-p21Cip1/Waf1 axis (Dolado et al., 2007) and DNA damage (Ogrunc et al., 2014; Sfikas et al., 2012). Other DNA damaging agents such as ionising and UV radiation, heat shock and a variety of genotoxic drugs, can also be potent inducers of senescence.

    • Senescence-associated microRNAs target cell cycle regulatory genes in normal human lung fibroblasts

      2017, Experimental Gerontology
      Citation Excerpt :

      Importantly, Cdc6 was identified as a novel target of the p53 pathway (Duursma and Agami, 2005), and our analysis showed that Cdc6 was targeted by hsa-miR-34a-5p, a p53-induced miRNA (Hermeking, 2012; Leung and Sharp, 2010; Suzuki et al., 2009). We verified that H2O2 induced the premature senescence of HDFs dose-dependently (Fig. 5), in agreement with our previous studies using IMR-90 and MRC-5 human fibroblasts (Sfikas et al., 2012), which also correlated with a profound reduction in the expression of Cdc6, and also E2F1 (Figs 5 and 6). MiR-221 and miR-222 were also induced in senescent HFL-1 human fibroblasts (Fig. 2).

    View all citing articles on Scopus
    1

    Present address: Institute for Cell and Molecular Biosciences (ICaMB), University of Newcastle Medical School, Framlington Place, Newcastle upon Tyne, NE2 4HH, UK.

    2

    Contributed equally to this work.

    View full text