Biochemical Genetics

, Volume 51, Issue 11, pp 901–915

Reactive Oxygen Species-Dependent Down-Regulation of Tumor Suppressor Genes PTEN, USP28, DRAM, TIGAR, and CYLD Under Oxidative Stress

Authors

  • Su-Jung Kim
    • Department of BiochemistryKangwon National University
  • Hyun-Joo Jung
    • Department of AnatomyYonsei University College of Medicine
    • Department of BiochemistryKangwon National University
Article

DOI: 10.1007/s10528-013-9616-7

Cite this article as:
Kim, S., Jung, H. & Lim, C. Biochem Genet (2013) 51: 901. doi:10.1007/s10528-013-9616-7

Abstract

We examined whether steady-state mRNA levels of five tumor suppressor genes are subjected to oxidative stress. Superoxide radical-generating menadione and serum deprivation diminished the steady-state mRNA levels for the genes phosphatase and tensin homolog (PTEN), ubiquitin specific peptidase 28 (USP28), damage-regulated autophagy modulator (DRAM), TP53-induced glycolysis and apoptosis regulator (TIGAR), and cylindromatosis (CYLD). Hydrogen peroxide showed suppression in steady-state mRNA levels for USP28, DRAM, TIGAR, and CYLD but not for PTEN. The steady-state mRNA levels specific for all five genes were enhanced by antioxidants, such as glutathione and N-acetylcysteine. The HepG2 stable transfectants overexpressing the mitochondrial isoform of human glutaredoxin, Grx2a, and containing a relatively low reactive oxygen species (ROS) level were assessed to contain the increased steady-state mRNA levels specific for the five tumor suppressor genes. In brief, the steady-state mRNA levels specific for these genes are negatively regulated by oxidative stress through the mediation of ROS.

Keywords

Tumor suppressorReactive oxygen speciesGlutathioneSerum deprivationGlutaredoxin

Introduction

Tumor suppressors, proteins that suppress tumorigenesis, generally act as negative regulators of overall cell cycle progression or cellular proliferation through regulating a variety of key cellular activities, including cell cycle checkpoint responses, detection and repair of DNA damage, protein ubiquitination and degradation, mitogenic signaling, cell specification, differentiation and migration, and tumor angiogenesis (Sherr 2004). They prevent the cell from multiplying until DNA damage is repaired, and some of them induce the cell with DNA damage to cell death. Without appropriate functions of tumor suppressors, the cells with DNA damage can continually divide and accumulate further DNA damage, ultimately leading to the generation and unregulated growth of a tumor cell. Underlying action mechanisms of various tumor suppressors have been studied and understood.

Phosphatase and tensin homolog (PTEN) is one of the frequently mutated or lost tumor suppressor genes; it has been identified in several sporadic and heritable tumors. As a nonredundant, plasma-membrane phosphatase, PTEN is essential for regulating the highly oncogenic prosurvival PI3K/AKT signaling pathway (Salmena et al. 2008). PTEN germline mutations were identified in a group of autosomal dominant syndromes characterized by developmental disorders, neurological deficits, multiple hematomas, and an increased risk of breast, thyroid, and endometrial cancers (Salmena et al. 2008). Even a minor impairment in PTEN function leads to the development of cancer, which supports a PTEN haploinsufficiency contribution to tumor progression (Salmena et al. 2008). PTEN was recently shown to act as a pivotal determinant of cell fate, regarding senescence and apoptosis in glioma cells exposed to ionizing radiation, and premature senescence might have a compensatory role for apoptosis in the absence of PTEN through the AKT/ROS/p53/p21 signaling pathway (Lee et al. 2011).

The p53 upregulated modulator of apoptosis (PUMA), identified as both a direct transcriptional target of p53 and a Bcl-2 homology 3 (BH3)—the only protein playing an essential role in apoptosis initiation—recapitulates most of the apoptosis-defective phenotypes seen in p53 null mice (Danial and Korsmeyer 2004; Jeffers et al. 2003; Villunger et al. 2003). Accordingly, PUMA acts as the primary liaison between p53 and initiation of apoptosis in response to ionizing irradiation and contributes to the genesis of cancer. The Chk2-p53-PUMA pathway is considered a major regulator of DNA-damage-induced apoptosis in response to breaks in vivo (Zhang et al. 2006). It is regulated by ubiquitin specific peptidase 28 (USP28), which is both a ubiquitinating enzyme and a binding component of 53BP1 (a check-point mediator) complexes (Zhang et al. 2006). Both USP28 and Chk2 accomplish DNA-damage-induced apoptosis in part through regulation of the p53 induction of proapoptotic genes like PUMA (Zhang et al. 2006). Besides its property as a tumor suppressor, USP28 has been suggested to possess the property of an oncoprotein (Popov et al. 2007).

Damage-regulated autophagy modulator (DRAM), a p53 target gene encoding a lysosomal protein that induces macroautophagy, acts as an effector of p53-mediated death (Crighton et al. 2006). Overexpression of DRAM alone causes minimal cell death, and some primary tumors reveal frequent diminished expression of DRAM often accompanied by retention of wild-type p53, indicating that DRAM is critical for p53-induced cell death and is down-regulated in human cancer (Crighton et al. 2006).

TP53-induced glycolysis and apoptosis regulator (TIGAR) was identified as a p53-inducible gene. Its overexpression decreases fructose-2,6-bisphosphate levels in cells, resulting in an inhibition of glycolysis and an overall decrease in intracellular reactive oxygen species (ROS) levels, while knockdown of endogenous TIGAR expression sensitizes cells to p53-induced cell death, implying that TIGAR plays a role in the tumor-suppressive ability of p53 to protect cells from the accumulation of genomic damage via decreasing intracellular ROS levels (Bensaad et al. 2006). TIGAR is capable of modulating ROS in response to nutrient starvation or metabolic stress, implying that its ability to limit autophagy correlates strongly with the suppression of ROS generation (Bensaad et al. 2009).

The tumor suppressor cylindromatosis (CYLD), a ubiquitously expressed 107 kDa polypeptide, contains a deubiquitinating domain at the C-terminus that removes lysine 63-linked ubiquitin chains from TRF2 and inhibits p65/p50 NF-κB activation (Trompouki et al. 2003). Some benign tumors of skin appendages in patients with familial CYLD are caused by the loss of both CYLD alleles (Bignell et al. 2000). Mice lacking CYLD are prone to chemically induced skin tumors, and cyld−/− tumors and keratinocytes treated with 12-O-tetradecanoylphorbol-13 acetate or UV light were hyperproliferative and had elevated cyclin D1 levels (Massoumi et al. 2006).

Although ROS, at the physiological concentration required for normal cellular functions, are involved in intracellular signaling and redox regulation, those at excessive concentrations cause oxidative stress, threatening the integrity of various biomolecules such as proteins and nucleic acids and being involved in aging. ROS are continuously generated by aerobic metabolism, during pathophysiological developments, such as inflammatory and allergic diseases, and by ionizing radiation (Bertram and Hass 2008). ROS are involved in various cellular processes ranging from apoptosis and necrosis to cellular proliferation and carcinogenesis (Matés et al. 2008). Oxidative stress, involved in the etiology of cancer, results from an imbalance between the intracellular ROS generation and the cellular antioxidant defenses. The excessive ROS deregulate the cellular redox homeostasis and promote tumor formation through triggering an aberrant induction of signaling pathways leading to tumorigenesis (Acharya et al. 2010). Certain breast tumor suppressors protect cells against oxidative stress and reduce intracellular ROS levels through stimulating antioxidant gene expression (Saha et al. 2009).

Glutaredoxins (Grxs), antioxidant proteins, participate in maintenance of cellular redox homeostasis and a broad spectrum of thiol-disulfide redox reactions (Hashemy et al. 2007; Holmgren et al. 2005). Grx has a variety of cellular functions in DNA synthesis, sulfur assimilation, defense against oxidative stress, apoptosis, cellular differentiation, regulation of transcriptional factor binding activity, and redox regulation via ROS (Hashemy et al. 2007). In human cells, two dithiol (Grx1 and Grx2) and one monothiol (Grx5) Grxs have been identified (Holmgren et al. 2005; Wingert et al. 2005). Grx1, mainly localized in the cytosol, supports ribonucleotide reductase with electrons, catalyzes disulfide-dithiol exchanges in proteins or small molecules such as dehydroascorbate, and is involved in transcriptional regulation, cellular differentiation, apoptosis, and recovery from oxidative stress (Holmgren et al. 2005; Xing and Lou 2002). Grx2 exists in two isoforms, targeted to either mitochondria (Grx2a) or the nucleus (Grx2b) through alternative splicing (Gladyshev et al. 2001; Lundberg et al. 2001). Grx2a, a mitochondrial isoform ubiquitously expressed, plays its general role in mitochondrial redox homeostasis, and Grx2b, nearly identical to processed mitochondrial Grx2a, is localized to testes and various cancer cell lines (Lönn et al. 2008).The Grx2a-overexpressing stable HepG2 cells exhibit enhanced proliferation, decreased ROS and caspase-3 levels, and increased total glutathione (GSH) level, compared with the vector control cells, which were reversed by down-regulating Grx2a in the same cell line (Kim et al. 2012).

In this work, we chose the five tumor suppressor genes PTEN, USP28, DRAM, TIGAR, and CYLD, since they have been recently identified as tumor suppressors and their regulation remains largely unknown. Our main interest was to confirm whether oxidative stress is related to tumorigenesis through the down-regulation of tumor suppressor genes. We demonstrate that their steady-state mRNA levels are down-regulated under oxidative stress but to different degrees, possibly through the mediation of the intracellular ROS level. This was documented using oxidative stressors, antioxidative substances, serum deprivation, and a mitochondrial glutaredoxin (Grx2a)-overexpressing cell line.

Materials and Methods

Chemicals

Agarose, GSH, menadione, hydrogen peroxide, and N-acetylcysteine (NAC) were obtained from Sigma-Aldrich Chemical Co. (St. Louis, MO, USA). Dulbecco’s modified Eagle’s medium (DMEM), fetal bovine serum (FBS), penicillin–streptomycin, and trypsin–EDTA were from Gibco-BRL (Gaithersburg, MD, USA). The i-MAX II DNA polymerase was from iNtRON Biotechnology (Sungnam, Korea). The Trizol reagent was from Invitrogen (Eugene, OR, USA) and M-MLV reverse transcriptase from Promega (Madison, WI, USA). The PCR primers used were from Cosmo Co. (Seoul, Korea). All other chemicals were of reagent grade or better.

Cell Culture

Human hepatocarcinoma HepG2 cells and their Grx2a-overexpressing transfectant cells were cultured in DMEM supplemented with 5% (V/V) heat-inactivated FBS, 100 μg/ml streptomycin, and 100 U/ml penicillin in a humidified atmosphere in 5% CO2 at 37°C. Mammalian cells were grown in 6-well culture plates to 50–60% confluence before subjecting to various growth conditions. The serum deprivation medium was the same, but without the 5% FBS.

Generation of the Grx2a-Overexpressing Stable Cell Lines

The HepG2 stable cell line overexpressing Grx2a, the mitochondrial isoform of human Grx2, was previously constructed as follows (Kim et al. 2012). The sense cDNA for human Grx2a was introduced into the multicloning site of G418 sulfate-resistant mammalian expression vector pcDNA3.0/HA (Invitrogen Corp., Carlsbad, CA, USA) to generate sense plasmids. The cDNA was amplified by PCR using two synthetic primers (primer 1, 5′-GAAACGAATTCTGATAATTGTGTGGTG-3′, and primer 2, 5′-TAACACTCGAGAACTAGTGGGAGCAAT-3′), containing EcoRI and XhoI restriction sites underlined). The amplified DNA fragment was electro-eluted from 1.2% agarose gel and digested with EcoRI and XhoI. The digested DNA fragment was ligated to the expression vector pcDNA3.0/HA previously digested with EcoRI and XhoI, and the ligation mixture was transformed into the E. coli strain MV1184. The resultant recombinant plasmid pcDNA3.0/HA-Grx2a was confirmed by restriction mapping and nucleotide sequencing. The PCR conditions used in amplifying the cDNA encoding Grx2a were 94°C (1 min), 62°C (1 min), and 72°C (1 min) for 30 cycles with the two PCR primers. Nucleotide sequencing was performed with automatic DNA sequencer in Cosmo Co., Korea. The sense plasmids were transfected into the HepG2 cells using the calcium phosphate method (Pear et al. 1993). One day after transfection, the neomycin-resistant cells were selected. After 4 weeks of growth, neomycin-resistant colonies were isolated, expanded, and analyzed.

Isolation of Total RNA

The trypsinized cells were homogenized in Trizol using a Tissue-Tearor (Model 985-370, Biospec Products, Bartlesville, OK, USA). The homogenate was mixed with 0.2 ml of chloroform per 1 ml Trizol; the mixture was vigorously shaken for 15 min at 4°C and centrifuged at 12,000×g for 15 min. RNA in the upper aqueous phase was precipitated by mixing with 0.5 ml isopropanol per 1 ml Trizol, and the mixture was centrifuged at 12,000×g for 8 min after storing at room temperature for about 5 min. The RNA pellet was washed with 75% ethanol, centrifuged at 7,500×g for 5 min, and dissolved in RNase-free water.

Determination of mRNA Levels

The steady-state mRNA levels of the five tumor suppressor genes (PTEN, USP28, DRAM, TIGAR, and CYLD) and the HPRT gene (as an internal standard) were determined by reverse transcriptase-polymerase chain reaction (RT-PCR), using different primer sequences for each gene (Table 1). First-strand cDNA was synthesized from 4 μg total RNA using M-MLV reverse transcriptase. One twentieth of the synthesized first-strand cDNA was used as templates in PCR. The PCR was performed using Ex Taq polymerase (iNtRON Biotechnology, Seoul, Korea) as follows: denaturation at 94°C for 1 min, annealing at 54°C for 1 min, and extension at 72°C for 1 min. The resultant amplified cDNAs were analyzed on 1% agarose gel and quantitated using densitometry (Gelpro-32 densitometer software, version 3.1, Media Cybernetics).
Table 1

Primers used in this study

Gene

Primer sequence (5′–3′)

Annealing temp (°C)

Amplicon size (bp)

PTEN

GAAGACCATAACCCACCACAGCTAGA

54

493

CTTTGATATCACCACACACAGGTAACG

USP28

GCTAGTACAAAACCTGCCTCAGAAAGC

54

546

GGAATCTCTTTCAACTTCTTCCCAGG

DRAM

GTCCCCTTCCTCTTGGTGACCTG

56

592

CTGTCCATTCACAGATCGCACTCA

TIGAR

GACTTCGGGAAAGGAAATACGGG

54

455

GAGACTCATCCCTGTATTGGGAGTGA

CYLD

GATATCATCCCAGAGAGTGTGACGC

54

649

CTCTGAAGGTTCCATCCGTACAGC

HPRT

GTAATGACCAGTCAACAGGGGAC

54

177

CCAGCAAGCTTGCGACCTTGACCA

Statistical Analysis

The results were expressed as mean ± SD. Statistical comparisons between experimental groups were performed using the Kruskal–Wallis test, followed by Dunn’s post hoc test for pairwise individual comparison. A P value of less than 0.05 was considered statistically significant.

Results

Oxidative Stressors

We tried to examine whether expression of the five tumor suppressor genes would be modulated by oxidative stress. We chose treatments with superoxide radical-generating menadione and hydrogen peroxide to enhance the intracellular ROS levels. The HepG2 cells were incubated with 200 μM menadione for 3 h, and total RNAs were isolated from the treated cells. Among the total RNAs, the steady-state mRNA levels specific for individual tumor suppressor genes after incubation with menadione were determined using RT-PCR and accompanying densitometry. Menadione was able to decrease the mRNA levels of all five tumor suppressor genes (Fig. 1). The diminishing effect of menadione at the concentration of 200 μM was statistically significant. CYLD appeared to be relatively less prone to the change in its mRNA level caused by menadione. Since the viability of the HepG2 cells grown for 3 h under 500 μM menadione dropped to around 55% (Jung et al. 2004), menadione at the concentration of 200 μM was used to treat the cells in this work.
https://static-content.springer.com/image/art%3A10.1007%2Fs10528-013-9616-7/MediaObjects/10528_2013_9616_Fig1_HTML.gif
Fig. 1

Effects of superoxide radical on mRNA of the tumor suppressor genes PTEN, USP28, DRAM, TIGAR, and CYLD in human hepatocellular carcinoma HepG2 cells. Mammalian cells were treated with superoxide radical-generating menadione (MD, 200 μM) for 3 h. The mRNA levels were determined by RT-PCR (left) and quantitated by densitometry (right). The mRNA levels for the individual genes in untreated cells (control) are considered 1. HPRT mRNA was used as an internal standard. **P < 0.01; ***P < 0.001 versus the corresponding control cells

Exogenous hydrogen peroxide was also used as an oxidative stressor to enhance the cellular ROS level in HepG2 cells. The HepG2 cells were incubated with hydrogen peroxide (0.5 mM, 2.5 mM) for 3 h. Hydrogen peroxide at the concentration of 2.5 mM could diminish the mRNA levels for USP28, DRAM, TIGAR, and CYLD, but not for PTEN (Fig. 2). Hydrogen peroxide at 0.5 mM, however, was unable to modulate the mRNA levels of any of the genes (data not shown). These results suggest that hydrogen peroxide selectively modulates expression of the tumor suppressor genes in a negative manner. At concentrations of less than 5.0 mM, it was previously found to be unable to modulate the cellular survival and morphology of HepG2 cells (Jung et al. 2004). Taken together, the steady-state mRNA levels of these five tumor suppressor genes are all down-regulated by superoxide radical, but regulation by hydrogen peroxide is selective.
https://static-content.springer.com/image/art%3A10.1007%2Fs10528-013-9616-7/MediaObjects/10528_2013_9616_Fig2_HTML.gif
Fig. 2

Effects of hydrogen peroxide (H2O2) on mRNA of the tumor suppressor genes PTEN, USP28, DRAM, TIGAR, and CYLD in human hepatocellular carcinoma HepG2 cells. Mammalian cells were treated with H2O2 (2.5 mM) for 3 h. The mRNA levels were determined by RT-PCR (left) and quantitated by densitometry (right). The mRNA levels for the individual genes in untreated cells (control) are considered 1. HPRT mRNA was used as an internal standard. **P < 0.01; ***P < 0.001 versus the corresponding control cells

Serum Deprivation

Since serum contains a variety of factors needed for the survival of mammalian cells in culture, limitation of serum concentration causes significant impact on cellular proliferation. Complete withdrawal of serum increases intracellular ROS level and subsequently enhances apoptotic cell death by increasing mitochondrial cytochrome P450 2E1, a member of the cytochrome P450 mixed-function oxidase system (Zhuge and Cederbaum 2006). Withdrawal of serum is a classical method to induce apoptosis, a cellular programmed cell death required to maintain tissue homeostasis and to eliminate defective and potentially dangerous cells. HepG2 is thought to require a few days of serum deprivation to induce apoptosis (Takehara et al. 2001). Since the steady-state mRNA levels of the five tumor suppressor genes tested are down-regulated by superoxide radical and some of them also tend to be down-regulated by hydrogen peroxide, serum deprivation was used as a method to enhance the cellular ROS level. The HepG2 cells were grown for 48 and 72 h after the depletion of serum. Among the five genes, the steady-state mRNA levels for PTEN, USP28, DRAM, and CYLD were markedly diminished by serum deprivation, whereas the TIGAR mRNA level remained relatively unchanged after serum deprivation (Fig. 3). The steady-state mRNA levels of the genes were more suppressed after 72 h of depletion of serum than after 48 h. In the previous work, when HepG2 cells were grown in the medium lacking serum, viability dropped to around 90% (Jung et al. 2004). Collectively, these findings partly support that these tumor suppressor genes are down-regulated under oxidative stress but to different degrees.
https://static-content.springer.com/image/art%3A10.1007%2Fs10528-013-9616-7/MediaObjects/10528_2013_9616_Fig3_HTML.gif
Fig. 3

Effects of serum deprivation on mRNA of the tumor suppressor genes PTEN, USP28, DRAM, TIGAR, and CYLD in human hepatocellular carcinoma HepG2 cells. Mammalian cells were incubated for 48 and 72 h after serum deprivation. The mRNA levels were determined by RT-PCR (left) and quantitated by densitometry (right). The mRNA levels for the individual genes in untreated cells (control) are considered 1. HPRT mRNA was used as an internal standard. *P < 0.05; **P < 0.01; ***P < 0.001 versus the corresponding control cells

Antioxidants

As both a major antioxidant and a redox and cell signaling regulator, GSH plays crucial roles in cellular defense against oxidant aggression, redox regulation of protein thiols, and maintenance of redox homeostasis critical for proper function of cellular processes, including apoptosis (Circu and Aw 2008). Butenolide, also named 4-acetamido-4-hydroxy-2-butenoic acid γ-lactone, one of the mycotoxins produced by Fusarium species, reduces the viability of HepG2 cells through rapid depletion of GSH and concomitant increase in intracellular ROS production, and butenolide-induced GSH depletion is aggravated by the addition of l-buthionine-(S,R)-sulfoximine (BSO), a specific inhibitor of GSH biosynthesis (Wang et al. 2006). Depletion of GSH in HepG2 cells with BSO significantly increased the susceptibility of the HepG2 cells to acrylamide- and trichloroethylene-induced cytotoxicity and DNA damage, suggesting that oxidative DNA damage in the hepatoma cells is induced by intracellular ROS and depletion of GSH (Hu et al. 2008; Yang et al. 2010). In this work, GSH was exogenously added to the HepG2 cells in order to diminish the intracellular ROS level. The HepG2 cells were incubated with 0.5 mM GSH for 3 h. With the incubation with GSH, the steady-state mRNA levels for all five tumor suppressor genes tended to increase (Fig. 4). The mRNA levels for USP28 and TIGAR increased significantly after exogenous addition of GSH, while those for PTEN and DRAM increased less significantly.
https://static-content.springer.com/image/art%3A10.1007%2Fs10528-013-9616-7/MediaObjects/10528_2013_9616_Fig4_HTML.gif
Fig. 4

Effects of glutathione (GSH) on mRNA of the tumor suppressor genes PTEN, USP28, DRAM, TIGAR, and CYLD in human hepatocellular carcinoma HepG2 cells. Mammalian cells were incubated with exogenous GSH (0.5 mM) for 3 h. The mRNA levels were determined by RT-PCR (left) and quantitated by densitometry (right). The mRNA levels for the individual genes in untreated cells (control) are considered 1. HPRT mRNA was used as an internal standard. *P < 0.05; **P < 0.01; ***P < 0.001 versus the corresponding control cells

NAC, an antioxidant known as a precursor of GSH, was also used to modulate the intracellular ROS level artificially in HepG2 cells. Pretreatment of NAC was reported to decrease hydrogen peroxide-induced alveolar type II epithelial cell apoptosis through scavenging high intracellular ROS level (Fu et al. 2010). The HepG2 cells were treated with 0.5 and 1.0 mM NAC for 3 h. Like GSH, NAC tended to enhance the steady-state mRNA levels for all five genes, although the increment appeared to depend on the type of tumor suppressor tested. Among the tested genes, NAC was able to enhance mRNA levels for PTEN and CYLD significantly (Fig. 5), and the enhancing effects were found to be higher at 1.0 mM than at 0.5 mM. In brief, antioxidants, such as GSH and NAC, are able to enhance the steady-state mRNA levels of the tested tumor suppressor genes, but to different degrees.
https://static-content.springer.com/image/art%3A10.1007%2Fs10528-013-9616-7/MediaObjects/10528_2013_9616_Fig5_HTML.gif
Fig. 5

Effects of N-acetylcysteine (NAC) on mRNA of the tumor suppressor genes PTEN, USP28, DRAM, TIGAR, and CYLD in human hepatocellular carcinoma HepG2 cells. Mammalian cells were incubated with NAC (0.5 mM, 1.0 mM) for 3 h. The mRNA levels were determined by RT-PCR (left) and quantitated by densitometry (right). The mRNA levels for the individual genes in untreated cells (control) are considered 1. HPRT mRNA was used as an internal standard. *P < 0.05; **P < 0.01; ***P < 0.001 versus the corresponding control cells

Mitochondrial Glutaredoxin

The Grx2a level in the pcDNA 3.0/HA-Grx2a stable cells was estimated, from western analysis, to be about 6.4-fold that of the pcDNA3.0/HA cells (Kim et al. 2012). On the contrary, the ROS level dropped to about 72% in the pcDNA3.0/HA-Grx2a stable cells, compared with that in the vector control cells (Kim et al. 2012). These findings suggested that Grx2a could play a role in down-regulating the intracellular ROS level. Since the tumor suppressor genes tested in this work have been identified to be down-regulated under oxidative stress, the mRNA levels for the five tumor suppressor genes were determined using the Grx2a-overexpressing stable HepG2 cells (pcDNA3.0/HA-Grx2a). The five tumor suppressor genes appeared to be up-regulated in the pcDNA3.0/HA-Grx2a cells, in comparison with the vector control pcDNA3.0/HA cells; especially, the steady-state mRNA levels for PTEN and TIGAR were significantly higher in the pcDNA3.0/HA-Grx2a cells than in the pcDNA3.0/HA cells (Fig. 6). The steady-state mRNA levels for the other three tumor suppressor genes, however, were reasonably higher in the pcDNA3.0/HA-Grx2a cells than in the pcDNA3.0 cells. Collectively, this increasing tendency in the steady-state mRNA levels of the five tumor suppressor genes supports their ROS-dependency in a negative manner.
https://static-content.springer.com/image/art%3A10.1007%2Fs10528-013-9616-7/MediaObjects/10528_2013_9616_Fig6_HTML.gif
Fig. 6

Effects of overexpression of human mitochondrial glutaredoxin (Grx2a) on mRNA levels for tumor suppressor genes PTEN, USP28, DRAM, TIGAR, and CYLD in human hepatocellular carcinoma HepG2 cells. Vector control cells (pcDNA3.0/HA; clones 1 and 2) and Grx2a-overexpressing transfectant cells (pcDNA3.0/HA-Grx2a; clones 1 and 2) were grown and harvested. The mRNA levels were determined by RT-PCR (left) and quantitated by densitometry (right), using the means obtained from the two clones. The values for the vector control cells are considered 1. HPRT mRNA was used as an internal standard. *P < 0.05; **P < 0.01; ***P < 0.001 versus the vector control cells

Discussion

Alterations in cellular metabolism play a crucial role in the generation of cancer, and some of the metabolic changes are attributed to the activation of oncogenes or loss of tumor suppressors. Since expressions of proto-oncogenes and tumor suppressor genes are closely involved in tumorigenesis, their regulation plays a key role in the development and progression of tumors. Tumor suppressors are found to exhibit their functions in a variety of ways. The well-known tumor suppressor p53 encodes a transcription factor for p21, a protein that arrests the cell cycle in G1 phase and integrates signals related to cell size, DNA integrity, and chromosome replication; it is involved in the formation of many types of tumors, including bladder, breast, colorectal, esophageal, liver, lung, prostate and ovarian carcinomas, brain tumors, sarcomas, lymphomas, and leukemias (Giono and Manfredi 2006). Tumor suppressors BRCA1 and BRCA2, which play roles in DNA damage repair, are responsible for familial breast and ovarian cancers (Greenberg 2008). The mitochondrial sirtuin, SirT3, is known to act as a tumor suppressor via its ability to suppress ROS and hypoxia inducible factor 1α (HIF-1α) (Bell et al. 2011).

Expressions of tumor suppressor genes are regulated in various ways, including genetic or epigenetic mechanisms. They are generally down-regulated in various tumors, and the cells with up-regulated tumor suppressor genes would be resistant to transformation to tumor cells. Epigenetic silencing of tumor suppressor genes is known to be a major contributor to neoplastic transformation. For example, the tumor suppressor gene TCF21 is expressed in normal lung airway epithelial cells, while it is aberrantly methylated and silenced in the head and neck squamous cell carcinomas (Smith et al. 2006). The tumor suppressor gene hDAB21P, which encodes a GTPase-activating protein for modulating the Ras-mediated signal pathway, is down-regulated by epigenetic control through histone acetylation and DNA methylation in prostate cancer cell lines (Chen et al. 2003). The human retinoblastoma tumor suppressor is also subjected to epigenetic regulation through the use of CCCTC-binding factor (CTCF) as an epigenetic regulator (De La Rosa-Velázquez et al. 2007). Nonsteroidal anti-inflammatory drugs (NSAIDs) have been shown to stimulate expression of the tumor suppressor EGR-1 gene, resulting in the up-regulation of NSAID-activated gene 1 (NAG-1), suggesting that the chemopreventive activity of NSAIDs occurs via EGR-1 and NAG-1 (Baek et al. 2005). Although the regulation of tumor suppressor genes by epigenetic mechanisms has been relatively well documented, their transcriptional regulation, especially that related to oxidative stress, remains unknown. Maspin, identified as a 42 kDa unique tumor suppressive serine protease inhibitor (serpin), was found to undertake different modifications under oxidative stress (Nawata et al. 2011). The cyclin-dependent kinase inhibitor p16 (INK4A), known as a tumor suppressor associated with multiple cancer types, is rapidly upregulated following ultraviolet-irradiation and in response to hydrogen peroxide-induced oxidative stress in a p38 stress-activated protein kinase-dependent manner, and subsequently it acts as an endogenous regulator of carcinogenic intracellular oxidative stress (Jenkins et al. 2011).

In this study, we have shown that superoxide radical induces down-regulation of the steady-state mRNA levels specific for five tumor suppressor genes. Hydrogen peroxide is also able to suppress the steady-state mRNA levels of USP28, DRAM, TIGAR, and CYLD, but not PTEN. The results imply that superoxide radical and hydrogen peroxide may work differently on the down-regulation of the tumor suppressor genes. The difference might have arisen from discrepancies in the promoter regions of the tumor suppressor genes selected for testing, but a precise cause for the difference remains uncertain. Down-regulation of the steady-state mRNA levels specific for these genes under oxidative stress is supported by an experiment using serum deprivation to enhance the intracellular ROS level. The findings, obtained using menadione, hydrogen peroxide, and serum deprivation, indicate that the steady-state mRNA levels specific for these tumor suppressor genes are regulated by ROS in a negative manner, although PTEN is an exception. These tumor suppressor genes are thought to be down-regulated by ROS, which means that their down-regulation may be reversed by ROS-scavenging systems. As expected, GSH and NAC are able to enhance the genes’ steady-state mRNA levels. Their ROS-dependent regulation was further confirmed by the enhanced steady-state mRNA levels in the Grx2a-overexpressing HepG2 transfectants known to maintain the lower ROS level. All these findings conclusively suggest that the steady-state mRNA levels for these tumor suppressor genes are modulated by the ROS level. Accordingly, their steady-state mRNA levels could be modulated by ROS-generating and ROS-scavenging systems.

Although all suppressor genes are not subject to the same regulatory mechanisms, the ROS-dependent modulation of some tumor suppressor genes allows us to make a few suggestions. First, ROS are capable of down-regulating some tumor suppressor genes during ROS-dependent tumorigenesis. Second, exogenously taken antioxidants, such as GSH, NAC, and ascorbic acid, may partly protect normal human cells from tumorigenesis by enhancing transcription of some tumor suppressor genes. Finally, antioxidant enzymes, such as superoxide dismutase, catalase, and peroxidase, and their activators could also protect normal cells from tumorigenesis. On the contrary, intracellular ROS enhanced by the knockdown of manganese-superoxide dismutase were found to induce HIF-1α expression in oral squamous cell carcinoma cells through transcriptional, translational, and post-translational regulation, which contributes to the development and malignant progression of many tumors (Sasabe et al. 2010). Our findings on the ROS-dependent down-regulation of some tumor suppressor genes under oxidative stress would appear to be opposite to the up-regulation of the tumor suppressor p16 gene under hydrogen peroxide-induced oxidative stress (Jenkins et al. 2011). This discrepancy might propose a different regulatory mode of tumor suppressor genes, depending on the cellular status subjected to tumorigenesis or antitumorigenesis processes under oxidative stress.

Acknowledgments

This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (2013-022019). This study is supported in part by Kangwon National University.

Copyright information

© Springer Science+Business Media New York 2013