G Model
ARTICLE IN PRESS
DNAREP-1907; No. of Pages 8
DNA Repair xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
DNA Repair journal homepage: www.elsevier.com/locate/dnarepair
The distribution of DNA damage is defined by region-specific susceptibility to DNA damage formation rather than repair differences Janne M. Strand a,b,1 , Katja Scheffler a,b,1 , Magnar Bjørås a,b , Lars Eide a,∗ a b
Department of Medical Biochemistry, Oslo University Hospital, University of Oslo, Norway Department of Microbiology, Oslo University Hospital, University of Oslo, Norway
a r t i c l e
i n f o
Article history: Received 28 November 2013 Received in revised form 5 March 2014 Accepted 7 March 2014 Available online xxx Keywords: mtDNA damage Genome dynamics Mitochondrial ROS DNA repair Genome integrity nDNA damage
a b s t r a c t The cellular genomes are continuously damaged by reactive oxygen species (ROS) from aerobic processes. The impact of DNA damage depends on the specific site as well as the cellular state. The steady-state level of DNA damage is the net result of continuous formation and subsequent repair, but it is unknown to what extent heterogeneous damage distribution is caused by variations in formation or repair of DNA damage. Here, we used a restriction enzyme/qPCR based method to analyze DNA damage in promoter and coding regions of four nuclear genes: the two house-keeping genes Gadph and Tbp, and the Ndufa9 and Ndufs2 genes encoding mitochondrial complex I subunits, as well as mt-Rnr1 encoded by mitochondrial DNA (mtDNA). The distribution of steady-state levels of damage varied in a site-specific manner. Oxidative stress induced damage in nDNA to a similar extent in promoter and coding regions, and more so in mtDNA. The subsequent removal of damage from nDNA was efficient and comparable with recovery times depending on the initial damage load, while repair of mtDNA was delayed with subsequently slower repair rate. The repair was furthermore found to be independent of transcription or the transcriptioncoupled repair factor CSB, but dependent on cellular ATP. Our results demonstrate that the capacity to repair DNA is sufficient to remove exogenously induced damage. Thus, we conclude that the heterogeneous steady-state level of DNA damage in promoters and coding regions is caused by site-specific DNA damage/modifications that take place under normal metabolism. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Nuclear and mitochondrial genomes are continuously exposed to reactive oxygen species (ROS) derived from numerous endogenous and exogenous sources that induce oxidative damage to DNA. In aerobic organisms, single electron reduction of oxygen generates the superoxide radical anion, which is the precursor of the membrane permeable hydrogen peroxide and its reactive degradation product; the hydroxyl radical. Defective or inefficient repair and persistent oxidative DNA damage has been implicated in various human diseases including cancer, neurodegenerative disorders and the aging process [1]. ROS-induced modifications of DNA include base lesions, modification on the ribose, and single- and double strand breaks. Specific repair systems for either damage
∗ Corresponding author at: Department of Medical Biochemistry, University of Oslo, Sognsvannsveien 20, 0027 Oslo, Norway. Tel.: +47 23071062; fax: +47 23070902. E-mail address: [email protected] (L. Eide). 1 These authors contributed equally to the work.
exist, and the base excision repair (BER) is the primary pathway for the removal of base lesions from both nuclear DNA (nDNA) and mitochondrial DNA (mtDNA). Inability to complete BER is not compatible with life, thereby demonstrating the importance of this pathway. Unrepaired base lesions can cause replication and transcription blockage, dependent on the lesion and the type of polymerase involved [2–5]. DNA damage additionally interferes with charge transfer in DNA, with putative functional impact on damage recognition and perhaps other processes [6]. Induction and accumulation of DNA damage is believed to be region-specific. For instance, mtDNA is more prone to ROS-induced DNA damage than nDNA, probably due to the close proximity to the site of ROS production in the mitochondria and different organization of DNA [7]. A heterogeneous pattern of oxidative DNA damage along the mtDNA was previously demonstrated, with the D-loop accumulating more DNA damage than other mtDNA loci [8]. The susceptibility of nDNA to oxidative damage is influenced by the degree of chromatin condensation. Given the fact that higher order of chromatin structure efficiently protects nDNA from damage [9], active genes located in less condensed regions could be more susceptible to oxidative DNA damage compared to inactive regions.
http://dx.doi.org/10.1016/j.dnarep.2014.03.003 1568-7864/© 2014 Elsevier B.V. All rights reserved.
Please cite this article in press as: J.M. Strand, et al., The distribution of DNA damage is defined by region-specific susceptibility to DNA damage formation rather than repair differences, DNA Repair (2014), http://dx.doi.org/10.1016/j.dnarep.2014.03.003
G Model
ARTICLE IN PRESS
DNAREP-1907; No. of Pages 8
J.M. Strand et al. / DNA Repair xxx (2014) xxx–xxx
2 Table 1 Primers used in the study. Target
Forward
Reverse
Purpose
Product size
mt-Rnr1 Ndufa9 promoter region Ndufa9 coding region Ndufs2 promoter region Ndufs2 coding region Gapdh promoter region Gapdh coding region Gapdh Tbp promoter region Tbp coding region ˇ-Actin
actcaaaggacttggcggta tggtgactcctacctgaagc ctcaagtccattgaggtgct tagtcaaggaacgcccatac atcatggctgtcaccacac taccgaagaacaacgaggag cttcaacagcaactcccact tcgtcccgtagacaaaatggt aggagcgtttgcttgacatt ggcatcagatgtgcgtca tcgccatggatgacgata
agcccatttcttcccatttc ttcggtcgtgaattttgttt gaccgaatcctcggatattt ctctgggcgtgagagattta cagacacccgctcatagaac cagagacctgaatgctgctt aaaagtcaggtttcccatcc cgcccaatacggccaaa tctctgtgtagccccgactt cgcagaaacctagccaaacc cacgatggaggggaatacag
DNA damage/expression DNA damage DNA damage/expression DNA damage DNA damage/expression DNA damage DNA damage Expression DNA damage DNA damage/expression Expression
206 158 159 155 106 150 112 57 142 85 110
The primers were designed with Primer 3 (http://frodo.wi.mit.edu/) except for 12S ribosomal RNA gene (mt-Rnr1) which was adapted from previous report [29]. All primer sets were optimized to give ct-values within a reliable range.
Repair of transcriptionally active genes is specifically efficient [10] and is aided by an interaction between RNA polymerase and the BER-enzyme NEIL2, implying that these regions are more prone to oxidative damage [11]. Promoter regions appear to be particularly vulnerable, as removal of UV light-induced DNA damage from promoter regions is slow compared to that from coding regions [12]. Accumulation of endogenously induced DNA damage interferes with gene regulation in the aging human brain [13] and correlates with reduced gene expression and accompanying reduced BER activity. In parallel, knock down of the ␣ subunit of the mitochondrial ATP synthase, which gene expression was reduced during aging, triggered accumulation of DNA damage in promoters of age-regulated genes. Thus, these data allude to a link between accumulation of site-specific DNA damage, repair efficiency and mitochondrial ATP production. To gain further insight into the dynamics of DNA damage in promoter and coding regions, we applied a detection method that enables site-specific quantification of DNA damage. Using this method, we assessed site-specific DNA damage formation and subsequent repair and correlated this with the steady-state distribution of DNA damage. This strategy enabled us to identify parameters that are essential for repair in vivo.
Spectrophotometer (Thermo Scientific) or Epoch Microplate Spectrophotometer (Bio-Tek).
2. Materials and methods
Gene expression analysis was performed by qPCR using the primer listed in Table 1. Total RNA (500 ng) was reversely transcribed into cDNA with High Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Expression levels were estimated from the ct values related to the housekeeping gene ˇ-actin as internal control.
2.1. Cell culture and treatment Mouse embryonic fibroblast (MEF) cells from wt and csbm/m c57/Bl6 mice were grown in Dulbecco’s modified Eagle’s medium containing 20 mM glucose, 10% fetal bovine serum, 2 mM lglutamine and 1% penicillin/streptomycin antibiotic supplement. For oxidant treatment, cells were seeded at equal density the day prior to the experiment. The cells were exposed to 50 M menadione for 45 min, and collected immediately after exposure, or allowed time to recover in fresh medium without oxidant for various time lengths. For experiments with bioenergetic inhibitors, the cells were exposed to 2.5 g/ml oligomycin or 20 mM 2deoxyglucose (2-DG). Transcriptional arrest was achieved by preincubating the cells with 10 g/ml ␣-amanitin for 12–14 h prior to oxidant treatment, and for the duration of treatment and recovery. 2.2. Isolation and quantification of nucleic acids Total DNA was isolated with the DNA Blood and Tissue kit (Qiagen) according to manufacturer’s protocol. Total RNA was isolated with RNeasy Mini kit (Qiagen) according to manufacturer’s protocol, including the optional DNase treatment step. Nucleic acid concentration and purity was assessed by Nanodrop
2.3. Analysis of DNA damage level DNA damage level was analyzed by a RT-qPCR method based on the ability of DNA lesions to inhibit restriction enzyme cleavage, as described previously with modifications [14], using primers listed in Table 1. The amplicons contain one single TaqI site. Briefly, genomic DNA (6 ng for the mitochondrial target or 30 ng for nuclear targets) was treated with 1U TaqI for 15 min at 65 ◦ C. DNA damage frequency was calculated as 2exp − (ctTaq1 − ctnt ), where ctTaq1 and ctnt represent CT values of TaqI-treated and non-treated genomic DNA, respectively, and presented as frequency of resistant TaqI sites Repair rate was quantified as the difference in damage frequency between 0 and 0.5 h. Accumulated DNA damage was quantified as the damage frequency after background frequency was subtracted. The absolute level of damage varied between the experiments, most likely due to isolation-mediated background oxidation of DNA. Thus, DNA from all samples representing one experiment (n = 1) was isolated simultaneously. 2.4. Analysis of mRNA expression levels
2.5. Statistical analysis Student’s t-test with two-tailed distribution and two-sample equal variance was used to evaluate statistical significance, unless otherwise stated. 3. Results 3.1. Assessment of steady-state level of DNA damage The damage-mediated restriction inhibition method to analyze DNA damage was established and described previously [14]. Any DNA alteration that inhibits TaqI restriction digestion would be selected and quantified in a subsequent real-time qPCR. With the application of sequence-specific primers we are able to study any genomic region containing the recognition sequence of the restriction enzyme used. To assess DNA damage induced by mitochondrial superoxide, we treated mouse embryonic fibroblast (MEF) cells
Please cite this article in press as: J.M. Strand, et al., The distribution of DNA damage is defined by region-specific susceptibility to DNA damage formation rather than repair differences, DNA Repair (2014), http://dx.doi.org/10.1016/j.dnarep.2014.03.003
G Model DNAREP-1907; No. of Pages 8
ARTICLE IN PRESS J.M. Strand et al. / DNA Repair xxx (2014) xxx–xxx
3
with menadione, a frequently used DNA damaging agent that generates intramitochondrial superoxide radical anion through redox cycling. First, we compared the time and dose-dependent accumulation of DNA damage in the mitochondrial mt-Rnr1 gene, encoding the mitochondrial ribosomal 12S rRNA and the nuclear Ndufa9 gene, encoding a subunit of mitochondrial complex I. Menadione induced DNA damage in a time-dependent manner with a maximum at 45 min (Fig. 1A). The apparent lack of continuous oxidation capacity is likely due to the toxic effect of superoxide on redoxcycling enzymes responsible for the menadione-mediated ROS formation. Similarly, a dose-dependent response curve indicated an optimal damage effect of 50 M menadione for the given exposure time (Fig. 1B). Thus, using this method, we demonstrated that mt-Rnr1 as Ndufa9 accumulate DNA damage during mitochondrial superoxide stress. To elaborate more on the identity of damage being detected, we exposed cells to UV radiation and the alkylation agent methyl methanesulfonate. Both genotoxins induce damage as detected by this method (Supplementary Fig. 1A). Additionally, 8-oxoguanine (a frequent ROS-induced lesion) was readily detected when present in a defined position in a synthetic template (Supplementary Fig. 1B). Using primers with similar PCR efficiency, we compared the steady-state level of DNA damage in the mtDNA-encoded mt-Rnr1 gene, and in promoter and coding regions of the nuclear genes Ndufa9 and Ndufs2 encoding mitochondrial complex I subunits, and two house-keeping genes: Gapdh and Tbp. The selected promoter regions are positioned within 1000 bp upstream of transcription start. In order to avoid cell culturing differences, DNA from more than 18 individual DNA isolations from non-treated MEF cells collected over a period of 8 months was analyzed and the mean steady-state DNA damage levels in promoter and coding regions were determined (Fig. 1C). Interestingly, we found that although the steady state level of DNA damage in mt-Rnr1 was higher than in the nuclear genes Tbp, Ndufa9 and Ndufs2, it was comparable with that in the house-keeping gene Gapdh. Furthermore, the coding region of Ndufs2 contained significantly more damages than its promoter region, while there were no promoter:coding differences in the other genes. Possible explanations for the heterogeneous accumulation of DNA damage could be site-specific differences in damage formation or repair.
3.2. Gene-specific ROS-induced DNA damage and repair To follow up on possible site-specific differences in damage formation or repair capacity, we applied menadione (50 M for 45 min) to induce superoxide-induced damage in mtDNA and nDNA as in Fig. 1 and analyzed the site-specific damage formation. As shown in Fig. 2, the amount of damage accumulated during menadione-induced ROS was gene-specific, and mtDNA was more vulnerable to insult as compared to the nuclear genes. Promoter and coding regions were similarly affected by menadione in the four nuclear genes, although the absolute damage frequency differed between the genes (e.g. Ndufa9 accumulates 2.5fold more promoter damage than Gapdh; p < 0.01). To analyze DNA repair capacity, we assessed the remaining level of DNA damage in treated cells that were allowed to recover for 0.5 h, 1 h and 4 h in medium without menadione. During the following recovery, the rate of damage removal from promoter and coding regions were not significantly different and the repair rates for the distinct sites were not significantly different. Thus, we conclude that accumulation of DNA damage during oxidative stress is site-specific, while the subsequent repair is not. Together, these results indicate that the heterogeneous site-specific steady state level of DNA damage (Fig. 1C) is not the net effect of different site-specific repair efficiencies.
Fig. 1. Assessment of genome integrity. (A and B) Validation of the restriction enzyme-based qPCR method to detect oxidative DNA damage in mtDNA (mt-Rnr1) and nDNA (Ndufa9). (A) Accumulated DNA damage level in MEF cells treated with 50 M menadione for 15–60 min. (B) Concentration-dependent damage induction upon treatment with 0–100 M menadione for 45 min. Data are mean ± SE of three independent experiments. *p < 0.05 vs. ctrl, # p < 0.05 mt-Rnr1 vs. Ndufa9. (C) Steadystate level of DNA damage in 9 different genomic sites as specified. The results shown are the mean ± SE of 18 separate preparation of DNA from non-treated MEFs (passage 3–20) over a period of 8 months. *p < 0.05; significant differences relative to mt-Rnr1 are indicated by horizontal lines. The asterix above bar indicates difference between promoter/coding regions.
3.3. Transcriptional arrest has no effect on ROS-induced DNA damage accumulation or subsequent repair Oxidative stress is known to activate transcriptional responses that act to increase tolerance to genotoxic stress [15–17] and it was a possibility that the repair kinetics following menadione treatment was not representative for the steady-state situation in non-treated
Please cite this article in press as: J.M. Strand, et al., The distribution of DNA damage is defined by region-specific susceptibility to DNA damage formation rather than repair differences, DNA Repair (2014), http://dx.doi.org/10.1016/j.dnarep.2014.03.003
G Model DNAREP-1907; No. of Pages 8 4
ARTICLE IN PRESS J.M. Strand et al. / DNA Repair xxx (2014) xxx–xxx
Fig. 2. Gene-specific susceptibility to damage induction and subsequent repair. MEF cells were treated with 50 M menadione for 45 min before medium was removed and cells recovered for the indicated time points. (A) Induction and removal of DNA damage in different nuclear genes comparing the promoter and coding region. (B) Induction and removal of damage in mtDNA. Data are mean ± SE of three independent experiments. *p < 0.05 vs. nt.
cells. To investigate in more detail whether transcription has any effect on repair, we reassessed repair in the presence of ␣-amanitin, a potent inhibitor of RNA polymerase II. Pretreatment of cells confirmed the specific inhibition of nuclear transcription compared to mitochondrial transcription (Fig. 3A), but ␣-amanitin neither influence the damage formation nor the following repair in promoter or coding regions of Ndufa9 or Gapdh, or in mtDNA (Fig. 3). ␣-amanitin alone did not influence DNA integrity (data not shown). Since mitochondrial transcription does not involve RNA polymerase II, we investigated damage formation and removal in csbm/m MEF cells, which carry a homozygous nonsense mutation in the Csb gene. The CSB protein is reported to participate in mtDNA repair in association with mitochondrial transcription [18,19]. However, our data indicated that repair of mtDNA was independent of CSB function (Fig. 3B). Several studies suggest that induction of oxidative DNA damage itself can affect gene expression [20,21] but mRNA levels of the analyzed gene targets revealed no changes immediately after oxidant treatment and only a minor although significant increase in mRNA level of the Tbp gene after 4 h of recovery (Supplementary
Fig. 2). We conclude that the RNA polymerase II and CSB proteins do not interfere with oxidative damage formation or removal from nDNA and mtDNA. 3.4. Reduced energy availability impairs the repair of oxidative DNA lesions Several DNA repair enzymes are energy-dependent, thus continuous ATP supply is necessary to facilitate efficient DNA repair in situations of increased oxidative stress. The nuclear and mitochondrial compartments can potentially be unevenly dependent on glycolytic and oxidative phosphorylation (OXPHOS) – type of ATP production. To address the question if repair of mtDNA or nDNA is specifically requiring glycolytic or OXPHOS ATP, we used a combined treatment of oligomycin, an inhibitor of mitochondrial ATP synthase and 2-deoxy-d-glucose (2-DG), a glucose-6-phosphate mimicker and competitive hexokinase inhibitor to investigate the repair in mt-Rnr1 and in the promoter regions of Ndufa9 and Ndufs2. The promoter region was selected as DNA damage previously
Please cite this article in press as: J.M. Strand, et al., The distribution of DNA damage is defined by region-specific susceptibility to DNA damage formation rather than repair differences, DNA Repair (2014), http://dx.doi.org/10.1016/j.dnarep.2014.03.003
G Model DNAREP-1907; No. of Pages 8
ARTICLE IN PRESS J.M. Strand et al. / DNA Repair xxx (2014) xxx–xxx
5
Fig. 3. Lack of correlation between transcription and repair in nDNA and mtDNA. (A) MEF cells were treated with ␣-amanitin for 12 h before mRNA levels were quantified from the nuclear genes (Gapdh and Ndufa9) and mitochondrial gene mt-Rnr1 related to nt cells (dashed line). Accumulated damage in mtDNA (B) and nDNA (C) was estimated in MEF cells pre-incubated with ␣-amanitin, and in csbm/m cells after treatment with 50 M menadione for 45 min and after a recovery period of 4 h. Data are mean ±SE of three independent experiments. *p < 0.05 vs. nt.
was found to accumulate in promoters of mitochondrial genes and correlate with mitochondrial dysfunction [13]. Oligomycin blocked ATP synthase-dependent respiration, thereby inferring that OXPHOS was dispensable for providing the remaining cellular ATP (Supplementary Fig. 3A). Oligomycin was not efficient in depleting cellular ATP, suggesting that alternative ATP generating pathways were sufficiently effective to restore cellular levels. 2-DG led to 20% reduction of cellular ATP which was further decreased to 50% of normal level in combination with oligomycin (Supplementary Fig. 3B). Depletion of ATP by 2-DG and oligomycin, either alone or in combination did not impact on genomic DNA integrity in cells that were not exposed to menadione (data not shown). Furthermore, treatment with either oligomycin or 2-DG alone had no effect on the amount of induced DNA damage or subsequent repair after oxidative stress in either mtDNA or nDNA (Fig. 4). In contrast, the combination of oligomycin and 2-DG completely blocked repair of mtDNA (Fig. 4A) and reduced repair of the nuclear promoter regions significantly (Fig. 4B). These data demonstrate that the source of ATP is not essential for efficient repair of mtDNA or nDNA, and that repair of mtDNA is more sensitive to alterations in cellular ATP than repair of nDNA.
4. Discussion In this study, we have quantified site-specific DNA damage and compared steady-state levels of DNA damage in promoter and coding regions with susceptibility to damage induction during oxidative stress and subsequent repair kinetics. We find that stressinduced damage accumulates in a gene-specific manner, and that coding and promoter regions within a gene are equally susceptible. Repair of induced damage is efficient with the major exception being repair of mtDNA, which is delayed and slower compared to repair of nDNA. While promoter and coding regions displayed similar damage dynamics with respect to formation and removal of oxidative stress/induced damage, the damage distribution in non-treated cells was more complex as coding region of Ndufs2 exhibited significantly more damage than promoter region. Our results indicate that repair differences do not dictate the steadystate level of damage in nDNA and mtDNA. The steady-state damage distribution somehow determined the amount of damage induced by mitochondrial ROS. For instance, the relatively intact Ndufa9 gene accumulated 2–3-fold more damage (Fig. 2) than the more damaged Gapdh gene. Notably, if the
Please cite this article in press as: J.M. Strand, et al., The distribution of DNA damage is defined by region-specific susceptibility to DNA damage formation rather than repair differences, DNA Repair (2014), http://dx.doi.org/10.1016/j.dnarep.2014.03.003
G Model DNAREP-1907; No. of Pages 8 6
ARTICLE IN PRESS J.M. Strand et al. / DNA Repair xxx (2014) xxx–xxx
Fig. 4. Energy requirement for repair of cellular DNA. MEF cells were exposed to oligomycin (2.5 g/mL), 2-DG (20 mM) or both 2 h before treatment with 50 M menadione for 45 min followed by 4 h recovery. DNA damage was quantified in (A) mtDNA and (B) nDNA. Data are mean ± SE of three independent experiments. *p < 0.05 vs. ctrl.
ROS-induced damages are added to the steady-state level, the total amounts of nDNA damage in the different sites are comparable. Thus, oxidative stress may induce global DNA damage that masks the heterogeneity in non-treated cells. Menadione uncouples normal electron flux and causes excessive mitochondrial superoxide generation. It has been shown that intracellular mitochondria differ in their ability to generate ROS [22]. One explanation for the heterogeneous damage distribution could therefore be that the distance between the chromosome region in where the gene resides
and the distinct ROS-generating mitochondrion might influence the damage level. In this respect, the unifying effect of menadione on damage distribution could be due to a shift were all rather than a subpopulation of mitochondria generates ROS. This postulation conflicts previous reports where mitochondrial ROS was dispensable for accumulation of DNA damage [23], although the manipulation of mitochondrial ROS in the cited study did not influence DNA integrity, like we demonstrate upon menadionemediated ROS generation. The relative high level of steady state
Please cite this article in press as: J.M. Strand, et al., The distribution of DNA damage is defined by region-specific susceptibility to DNA damage formation rather than repair differences, DNA Repair (2014), http://dx.doi.org/10.1016/j.dnarep.2014.03.003
G Model DNAREP-1907; No. of Pages 8
ARTICLE IN PRESS J.M. Strand et al. / DNA Repair xxx (2014) xxx–xxx
as well as induced damage in mt-Rnr1 might be explained by its vicinity to the electron transport chain, which is known to generate superoxide anion as a byproduct of active respiration. Our data support the notion that DNA damage might be targeted specifically to distinct regions as an important signal for gene regulation [15–17]. There is experimental evidence for that reversible DNA damage is actively involved in physical processes. For instance, DNA damage was induced in mice that were subjected to environmental stimulation [24] and inability to reverse the damage was associated with neurodegeneration. In the extrapolation of these findings, the steady-state distribution of DNA damage could represent the functional state of the cell that can be disturbed by mitochondrial manipulation as here or by age-associated processes, resulting in mitochondrial dysfunction [13]. We have previously compared the method used here [14] with the established method for mtDNA damage assessment, based on qPCR-quantification of intact 10 kb template DNA; developed by van Houten and coworkers [24]. The advantage of the current method is that it enables assessment of a more detailed region of DNA. Although an alternative, principally similar method was recently established to quantify DNA damage lesions in regions of 1 kb size [8], hot spots within one specific gene region may not be detected. As described in materials and methods, we quantify damage frequency based on the inhibition of TaqI restriction enzyme cleavage. Both methods score for strand breaks and bulky lesions [25]. The advantage of the method used here is that lesions that are efficiently bypassed (such as the abundant 8-oxoguanine) are detectable as these lesions previously have been utilized based on their ability to inhibit restriction enzyme digestion [26]. Interpretation of absolute number of frequency should nevertheless be done with precaution for both methods; e.g. the relaxed versus supercoil form of DNA determines the PCR efficiency [27], and damages may induce conformational changes that inhibit restriction cleavage although being positioned outside the palindromic recognition sequence. Hence, the damage detection region may span wider than the recognition sequence of TaqI restriction enzyme. The negative effect of ␣-amanitin on repair efficiency has two implications. First, transcription-dependent adaptation following oxidative stress is not important for efficient repair and second, this is in line with the lack of evidence for transcription-coupled repair of oxidative DNA damage. On the other hand, our results are in contrast with earlier studies where CSB was found to be important for efficient removal of mtDNA damage [18]. The explanation for this discrepancy can only be found by comparing similar cell lines and damage condition with the two different damage detection methods. One important difference from our studies is that Stevnsner and coworkers selectively studied non-replicated DNA from mitochondria whereas we here studied total DNA. It is therefore a possibility that the unrepaired mtDNA is masked by corresponding increased replication of the non-damaged mtDNA molecules although we find this explanation unlikely as the mtDNA copy number in our experiments were slightly reduced after menadione exposure (data not shown). A previous study demonstrated that exogenous ATP stimulated the translocation of BER enzyme Ape1 into the nucleus [28]. Additionally, it was reported that reduced ATP level correlated with increased DNA damage in promoter regions of nuclear genes with age [13]. The overall conclusion from the combinatory treatment of 2-DG and oligomycin, is that ATP depletion of 50% or more is required to impair DNA efficiency. The combination significantly reduced repair of nDNA (Fig. 4B) and completely prevented any removal of damage from mtDNA (Fig. 4A), suggesting that repair of these two cellular genomes are differently dependent on ATP. In fact, an early study by Oei and Ziegler [29] suggests that the ATP required for the energy-demanding repair step is derived from poly-ADP ribose (PAR) rather than ATP. Since the PAR moiety is
7
produced from NAD by PARP1, this energy-requiring step would be less sensitive to ATP depletion. Although PARP1 has been identified in mitochondria [30], it is generally accepted that its nuclear activity is important for damage recovery. The delay in repair of mtDNA (Fig. 2B) after menadione exposure should be interpreted in view of the strong ATP-depleting effect of menadione. Recovery of cellular ATP after menadione exposure requires 1 h [31] and this transient ATP depletion could be responsible for the delay process that is even more pronounced upon 2-DG and oligomycin administration. In conclusion, our analyses of promoter and coding regions of four nuclear genes and the mitochondrial gene mt-Rnr1 demonstrate that the distribution of oxidative stress-induced damage is indicative of gene-specific susceptibility to damage. The capacity to remove damage differed between nDNA and mtDNA. Repair of nDNA was relatively resistant to changes in the ATP concentration compared to repair of mtDNA, and indicates that inefficient repair of mtDNA with subsequent accumulation of mtDNA damage is an early consequence of bioenergetic dysfunction. Conflict of interest statement The authors declare no conflict of interest. Acknowledgements This work was supported by grants from Research Council of Norway to MB and the University of Oslo. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.dnarep. 2014.03.003. References [1] M.L. Hegde, A.K. Mantha, T.K. Hazra, K.K. Bhakat, S. Mitra, B. Szczesny, Oxidative genome damage and its repair: implications in aging and neurodegenerative diseases, Mech. Ageing Dev. 133 (2012) 157–168. [2] S. Xanthoudakis, R.J. Smeyne, J.D. Wallace, T. Curran, The redox/DNA repair protein, Ref-1, is essential for early embryonic development in mice, Proc. Natl. Acad. Sci. U.S.A. 93 (1996) 8919–8923. [3] M.A. Graziewicz, R.J. Bienstock, W.C. Copeland, The DNA polymerase gamma Y955C disease variant associated with PEO and parkinsonism mediates the incorporation and translesion synthesis opposite 7,8-dihydro-8-oxo-2 deoxyguanosine, Hum. Mol. Genet. 16 (2007) 2729–2739. [4] H.E. Krokan, M. Bjoras, Base excision repair, Cold Spring Harb. Perspect. Biol. 5 (2013) a012583. [5] V.K. Batra, D.D. Shock, W.A. Beard, C.E. McKenna, S.H. Wilson, Binary complex crystal structure of DNA polymerase beta reveals multiple conformations of the templating 8-oxoguanine lesion, Proc. Natl. Acad. Sci. U.S.A. 109 (2012) 113–118. [6] E.J. Merino, A.K. Boal, J.K. Barton, Biological contexts for DNA charge transport chemistry, Curr. Opin. Chem. Biol. 12 (2008) 229–237. [7] F.M. Yakes, H.B. Van, Mitochondrial DNA damage is more extensive and persists longer than nuclear DNA damage in human cells following oxidative stress, Proc. Natl. Acad. Sci. U.S.A. 94 (1997) 514–519. [8] O. Rothfuss, T. Gasser, N. Patenge, Analysis of differential DNA damage in the mitochondrial genome employing a semi-long run real-time PCR approach, Nucleic Acids Res. 38 (2010) e24. [9] M. Ljungman, P.C. Hanawalt, Efficient protection against oxidative DNA damage in chromatin, Mol. Carcinog. 5 (1992) 264–269. [10] V.A. Bohr, C.A. Smith, D.S. Okumoto, P.C. Hanawalt, DNA repair in an active gene: removal of pyrimidine dimers from the DHFR gene of CHO cells is much more efficient than in the genome overall, Cell 40 (1985) 359–369. [11] D. Banerjee, S.M. Mandal, A. Das, M.L. Hegde, S. Das, K.K. Bhakat, I. Boldogh, P.S. Sarkar, S. Mitra, T.K. Hazra, Preferential repair of oxidized base damage in the transcribed genes of mammalian cells, J. Biol. Chem. 286 (2011) 6006–6016. [12] Y. Tu, S. Tornaletti, G.P. Pfeifer, DNA repair domains within a human gene: selective repair of sequences near the transcription initiation site, EMBO J. 15 (1996) 675–683. [13] T. Lu, Y. Pan, S.Y. Kao, C. Li, I. Kohane, J. Chan, B.A. Yankner, Gene regulation and DNA damage in the ageing human brain, Nature 429 (2004) 883–891.
Please cite this article in press as: J.M. Strand, et al., The distribution of DNA damage is defined by region-specific susceptibility to DNA damage formation rather than repair differences, DNA Repair (2014), http://dx.doi.org/10.1016/j.dnarep.2014.03.003
G Model DNAREP-1907; No. of Pages 8 8
ARTICLE IN PRESS J.M. Strand et al. / DNA Repair xxx (2014) xxx–xxx
[14] W. Wang, P. Osenbroch, R. Skinnes, Y. Esbensen, M. Bjoras, L. Eide, Mitochondrial DNA integrity is essential for mitochondrial maturation during differentiation of neural stem cells, Stem Cells 28 (2010) 2195–2204. [15] B. Perillo, M.N. Ombra, A. Bertoni, C. Cuozzo, S. Sacchetti, A. Sasso, L. Chiariotti, A. Malorni, C. Abbondanza, E.V. Avvedimento, DNA oxidation as triggered by H3K9me2 demethylation drives estrogen-induced gene expression, Science 319 (2008) 202–206. [16] B.G. Ju, V.V. Lunyak, V. Perissi, I. Garcia-Bassets, D.W. Rose, C.K. Glass, M.G. Rosenfeld, A topoisomerase IIbeta-mediated dsDNA break required for regulated transcription, Science 312 (2006) 1798–1802. [17] K.A. Ziel, V. Grishko, C.C. Campbell, J.F. Breit, G.L. Wilson, M.N. Gillespie, Oxidants in signal transduction: impact on DNA integrity and gene expression, FASEB J. 19 (2005) 387–394. [18] T. Stevnsner, S. Nyaga, N.C. Souza-Pinto, G.T.J. van der Horst, T.G.M.F. Gorgels, B.A. Hogue, T. Thorslund, V.A. Bohr, Mitochondrial repair of 8-oxoguanine is deficient in Cockayne syndrome group B, Oncogene 21 (2002) 8675–8682. [19] B.R. Berquist, C. Canugovi, P. Sykora, D.M. Wilson III, V.A. Bohr, Human Cockayne syndrome B protein reciprocally communicates with mitochondrial proteins and promotes transcriptional elongation, Nucleic Acids Res. 40 (2012) 8392–8405. [20] C. Marietta, H. Gulam, P.J. Brooks, A single 8,5 -cyclo-2 -deoxyadenosine lesion in a TATA box prevents binding of the TATA binding protein and strongly reduces transcription in vivo, DNA Repair (Amst) 1 (2002) 967–975. [21] M.K. Hailer-Morrison, J.M. Kotler, B.D. Martin, K.D. Sugden, Oxidized guanine lesions as modulators of gene transcription. Altered p50 binding affinity and repair shielding by 7,8-dihydro-8-oxo-2 -deoxyguanosine lesions in the NFkappaB promoter element, Biochemistry 42 (2003) 9761–9770. [22] G.H. Mathisen, I.H. Thorkildsen, R.E. Paulsen, Secretory PLA2-IIA and ROS generation in peripheral mitochondria are critical for neuronal death, Brain Res. 1153 (2007) 43–51.
[23] S. Hoffmann, D. Spitkovsky, J.P. Radicella, B. Epe, R.J. Wiesner, Reactive oxygen species derived from the mitochondrial respiratory chain are not responsible for the basal levels of oxidative base modifications observed in nuclear DNA of Mammalian cells, Free Radic. Biol. Med. 36 (2004) 765–773. [24] J.H. Santos, J.N. Meyer, B.S. Mandavilli, H.B. Van, Quantitative PCR-based measurement of nuclear and mitochondrial DNA damage and repair in mammalian cells, Methods Mol. Biol. 314 (2006) 183–199. [25] A.M. Furda, A.S. Bess, J.N. Meyer, H.B. Van, Analysis of DNA damage and repair in nuclear and mitochondrial DNA of animal cells using quantitative PCR, Methods Mol. Biol. 920 (2012) 111–132. [26] E. Larsen, K. Kwon, F. Coin, J.M. Egly, A. Klungland, Transcription activities at 8-oxoG lesions in DNA, DNA Repair (Amst) 3 (2004) 1457–1468. [27] J. Chen, F.F. Kadlubar, J.Z. Chen, DNA supercoiling suppresses real-time PCR: a new approach to the quantification of mitochondrial DNA damage and repair, Nucleic Acids Res. 35 (2007) 1377–1388. [28] A. Pines, L. Perrone, N. Bivi, M. Romanello, G. Damante, M. Gulisano, M.R. Kelley, F. Quadrifoglio, G. Tell, Activation of APE1/Ref-1 is dependent on reactive oxygen species generated after purinergic receptor stimulation by ATP, Nucleic Acids Res. 33 (2005) 4379–4394. [29] S.L. Oei, M. Ziegler, ATP for the DNA ligation step in base excision repair is generated from poly(ADP-ribose), J. Biol. Chem. 275 (2000) 23234–23239. [30] M.N. Rossi, M. Carbone, C. Mostocotto, C. Mancone, M. Tripodi, R. Maione, P. Amati, Mitochondrial localization of PARP-1 requires interaction with mitofilin and is involved in the maintenance of mitochondrial DNA integrity, J. Biol. Chem. 284 (2009) 31616–31624. [31] P.O. Osenbroch, P. Auk-Emblem, R. Halsne, J. Strand, R.J. Forstrom, I. van der Pluijm, L. Eide, Accumulation of mitochondrial DNA damage and bioenergetic dysfunction in CSB defective cells, FEBS J. 276 (2009) 2811–2821.
Please cite this article in press as: J.M. Strand, et al., The distribution of DNA damage is defined by region-specific susceptibility to DNA damage formation rather than repair differences, DNA Repair (2014), http://dx.doi.org/10.1016/j.dnarep.2014.03.003
ARTICLE IN PRESS
DNAREP-1907; No. of Pages 8
DNA Repair xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
DNA Repair journal homepage: www.elsevier.com/locate/dnarepair
The distribution of DNA damage is defined by region-specific susceptibility to DNA damage formation rather than repair differences Janne M. Strand a,b,1 , Katja Scheffler a,b,1 , Magnar Bjørås a,b , Lars Eide a,∗ a b
Department of Medical Biochemistry, Oslo University Hospital, University of Oslo, Norway Department of Microbiology, Oslo University Hospital, University of Oslo, Norway
a r t i c l e
i n f o
Article history: Received 28 November 2013 Received in revised form 5 March 2014 Accepted 7 March 2014 Available online xxx Keywords: mtDNA damage Genome dynamics Mitochondrial ROS DNA repair Genome integrity nDNA damage
a b s t r a c t The cellular genomes are continuously damaged by reactive oxygen species (ROS) from aerobic processes. The impact of DNA damage depends on the specific site as well as the cellular state. The steady-state level of DNA damage is the net result of continuous formation and subsequent repair, but it is unknown to what extent heterogeneous damage distribution is caused by variations in formation or repair of DNA damage. Here, we used a restriction enzyme/qPCR based method to analyze DNA damage in promoter and coding regions of four nuclear genes: the two house-keeping genes Gadph and Tbp, and the Ndufa9 and Ndufs2 genes encoding mitochondrial complex I subunits, as well as mt-Rnr1 encoded by mitochondrial DNA (mtDNA). The distribution of steady-state levels of damage varied in a site-specific manner. Oxidative stress induced damage in nDNA to a similar extent in promoter and coding regions, and more so in mtDNA. The subsequent removal of damage from nDNA was efficient and comparable with recovery times depending on the initial damage load, while repair of mtDNA was delayed with subsequently slower repair rate. The repair was furthermore found to be independent of transcription or the transcriptioncoupled repair factor CSB, but dependent on cellular ATP. Our results demonstrate that the capacity to repair DNA is sufficient to remove exogenously induced damage. Thus, we conclude that the heterogeneous steady-state level of DNA damage in promoters and coding regions is caused by site-specific DNA damage/modifications that take place under normal metabolism. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Nuclear and mitochondrial genomes are continuously exposed to reactive oxygen species (ROS) derived from numerous endogenous and exogenous sources that induce oxidative damage to DNA. In aerobic organisms, single electron reduction of oxygen generates the superoxide radical anion, which is the precursor of the membrane permeable hydrogen peroxide and its reactive degradation product; the hydroxyl radical. Defective or inefficient repair and persistent oxidative DNA damage has been implicated in various human diseases including cancer, neurodegenerative disorders and the aging process [1]. ROS-induced modifications of DNA include base lesions, modification on the ribose, and single- and double strand breaks. Specific repair systems for either damage
∗ Corresponding author at: Department of Medical Biochemistry, University of Oslo, Sognsvannsveien 20, 0027 Oslo, Norway. Tel.: +47 23071062; fax: +47 23070902. E-mail address: [email protected] (L. Eide). 1 These authors contributed equally to the work.
exist, and the base excision repair (BER) is the primary pathway for the removal of base lesions from both nuclear DNA (nDNA) and mitochondrial DNA (mtDNA). Inability to complete BER is not compatible with life, thereby demonstrating the importance of this pathway. Unrepaired base lesions can cause replication and transcription blockage, dependent on the lesion and the type of polymerase involved [2–5]. DNA damage additionally interferes with charge transfer in DNA, with putative functional impact on damage recognition and perhaps other processes [6]. Induction and accumulation of DNA damage is believed to be region-specific. For instance, mtDNA is more prone to ROS-induced DNA damage than nDNA, probably due to the close proximity to the site of ROS production in the mitochondria and different organization of DNA [7]. A heterogeneous pattern of oxidative DNA damage along the mtDNA was previously demonstrated, with the D-loop accumulating more DNA damage than other mtDNA loci [8]. The susceptibility of nDNA to oxidative damage is influenced by the degree of chromatin condensation. Given the fact that higher order of chromatin structure efficiently protects nDNA from damage [9], active genes located in less condensed regions could be more susceptible to oxidative DNA damage compared to inactive regions.
http://dx.doi.org/10.1016/j.dnarep.2014.03.003 1568-7864/© 2014 Elsevier B.V. All rights reserved.
Please cite this article in press as: J.M. Strand, et al., The distribution of DNA damage is defined by region-specific susceptibility to DNA damage formation rather than repair differences, DNA Repair (2014), http://dx.doi.org/10.1016/j.dnarep.2014.03.003
G Model
ARTICLE IN PRESS
DNAREP-1907; No. of Pages 8
J.M. Strand et al. / DNA Repair xxx (2014) xxx–xxx
2 Table 1 Primers used in the study. Target
Forward
Reverse
Purpose
Product size
mt-Rnr1 Ndufa9 promoter region Ndufa9 coding region Ndufs2 promoter region Ndufs2 coding region Gapdh promoter region Gapdh coding region Gapdh Tbp promoter region Tbp coding region ˇ-Actin
actcaaaggacttggcggta tggtgactcctacctgaagc ctcaagtccattgaggtgct tagtcaaggaacgcccatac atcatggctgtcaccacac taccgaagaacaacgaggag cttcaacagcaactcccact tcgtcccgtagacaaaatggt aggagcgtttgcttgacatt ggcatcagatgtgcgtca tcgccatggatgacgata
agcccatttcttcccatttc ttcggtcgtgaattttgttt gaccgaatcctcggatattt ctctgggcgtgagagattta cagacacccgctcatagaac cagagacctgaatgctgctt aaaagtcaggtttcccatcc cgcccaatacggccaaa tctctgtgtagccccgactt cgcagaaacctagccaaacc cacgatggaggggaatacag
DNA damage/expression DNA damage DNA damage/expression DNA damage DNA damage/expression DNA damage DNA damage Expression DNA damage DNA damage/expression Expression
206 158 159 155 106 150 112 57 142 85 110
The primers were designed with Primer 3 (http://frodo.wi.mit.edu/) except for 12S ribosomal RNA gene (mt-Rnr1) which was adapted from previous report [29]. All primer sets were optimized to give ct-values within a reliable range.
Repair of transcriptionally active genes is specifically efficient [10] and is aided by an interaction between RNA polymerase and the BER-enzyme NEIL2, implying that these regions are more prone to oxidative damage [11]. Promoter regions appear to be particularly vulnerable, as removal of UV light-induced DNA damage from promoter regions is slow compared to that from coding regions [12]. Accumulation of endogenously induced DNA damage interferes with gene regulation in the aging human brain [13] and correlates with reduced gene expression and accompanying reduced BER activity. In parallel, knock down of the ␣ subunit of the mitochondrial ATP synthase, which gene expression was reduced during aging, triggered accumulation of DNA damage in promoters of age-regulated genes. Thus, these data allude to a link between accumulation of site-specific DNA damage, repair efficiency and mitochondrial ATP production. To gain further insight into the dynamics of DNA damage in promoter and coding regions, we applied a detection method that enables site-specific quantification of DNA damage. Using this method, we assessed site-specific DNA damage formation and subsequent repair and correlated this with the steady-state distribution of DNA damage. This strategy enabled us to identify parameters that are essential for repair in vivo.
Spectrophotometer (Thermo Scientific) or Epoch Microplate Spectrophotometer (Bio-Tek).
2. Materials and methods
Gene expression analysis was performed by qPCR using the primer listed in Table 1. Total RNA (500 ng) was reversely transcribed into cDNA with High Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Expression levels were estimated from the ct values related to the housekeeping gene ˇ-actin as internal control.
2.1. Cell culture and treatment Mouse embryonic fibroblast (MEF) cells from wt and csbm/m c57/Bl6 mice were grown in Dulbecco’s modified Eagle’s medium containing 20 mM glucose, 10% fetal bovine serum, 2 mM lglutamine and 1% penicillin/streptomycin antibiotic supplement. For oxidant treatment, cells were seeded at equal density the day prior to the experiment. The cells were exposed to 50 M menadione for 45 min, and collected immediately after exposure, or allowed time to recover in fresh medium without oxidant for various time lengths. For experiments with bioenergetic inhibitors, the cells were exposed to 2.5 g/ml oligomycin or 20 mM 2deoxyglucose (2-DG). Transcriptional arrest was achieved by preincubating the cells with 10 g/ml ␣-amanitin for 12–14 h prior to oxidant treatment, and for the duration of treatment and recovery. 2.2. Isolation and quantification of nucleic acids Total DNA was isolated with the DNA Blood and Tissue kit (Qiagen) according to manufacturer’s protocol. Total RNA was isolated with RNeasy Mini kit (Qiagen) according to manufacturer’s protocol, including the optional DNase treatment step. Nucleic acid concentration and purity was assessed by Nanodrop
2.3. Analysis of DNA damage level DNA damage level was analyzed by a RT-qPCR method based on the ability of DNA lesions to inhibit restriction enzyme cleavage, as described previously with modifications [14], using primers listed in Table 1. The amplicons contain one single TaqI site. Briefly, genomic DNA (6 ng for the mitochondrial target or 30 ng for nuclear targets) was treated with 1U TaqI for 15 min at 65 ◦ C. DNA damage frequency was calculated as 2exp − (ctTaq1 − ctnt ), where ctTaq1 and ctnt represent CT values of TaqI-treated and non-treated genomic DNA, respectively, and presented as frequency of resistant TaqI sites Repair rate was quantified as the difference in damage frequency between 0 and 0.5 h. Accumulated DNA damage was quantified as the damage frequency after background frequency was subtracted. The absolute level of damage varied between the experiments, most likely due to isolation-mediated background oxidation of DNA. Thus, DNA from all samples representing one experiment (n = 1) was isolated simultaneously. 2.4. Analysis of mRNA expression levels
2.5. Statistical analysis Student’s t-test with two-tailed distribution and two-sample equal variance was used to evaluate statistical significance, unless otherwise stated. 3. Results 3.1. Assessment of steady-state level of DNA damage The damage-mediated restriction inhibition method to analyze DNA damage was established and described previously [14]. Any DNA alteration that inhibits TaqI restriction digestion would be selected and quantified in a subsequent real-time qPCR. With the application of sequence-specific primers we are able to study any genomic region containing the recognition sequence of the restriction enzyme used. To assess DNA damage induced by mitochondrial superoxide, we treated mouse embryonic fibroblast (MEF) cells
Please cite this article in press as: J.M. Strand, et al., The distribution of DNA damage is defined by region-specific susceptibility to DNA damage formation rather than repair differences, DNA Repair (2014), http://dx.doi.org/10.1016/j.dnarep.2014.03.003
G Model DNAREP-1907; No. of Pages 8
ARTICLE IN PRESS J.M. Strand et al. / DNA Repair xxx (2014) xxx–xxx
3
with menadione, a frequently used DNA damaging agent that generates intramitochondrial superoxide radical anion through redox cycling. First, we compared the time and dose-dependent accumulation of DNA damage in the mitochondrial mt-Rnr1 gene, encoding the mitochondrial ribosomal 12S rRNA and the nuclear Ndufa9 gene, encoding a subunit of mitochondrial complex I. Menadione induced DNA damage in a time-dependent manner with a maximum at 45 min (Fig. 1A). The apparent lack of continuous oxidation capacity is likely due to the toxic effect of superoxide on redoxcycling enzymes responsible for the menadione-mediated ROS formation. Similarly, a dose-dependent response curve indicated an optimal damage effect of 50 M menadione for the given exposure time (Fig. 1B). Thus, using this method, we demonstrated that mt-Rnr1 as Ndufa9 accumulate DNA damage during mitochondrial superoxide stress. To elaborate more on the identity of damage being detected, we exposed cells to UV radiation and the alkylation agent methyl methanesulfonate. Both genotoxins induce damage as detected by this method (Supplementary Fig. 1A). Additionally, 8-oxoguanine (a frequent ROS-induced lesion) was readily detected when present in a defined position in a synthetic template (Supplementary Fig. 1B). Using primers with similar PCR efficiency, we compared the steady-state level of DNA damage in the mtDNA-encoded mt-Rnr1 gene, and in promoter and coding regions of the nuclear genes Ndufa9 and Ndufs2 encoding mitochondrial complex I subunits, and two house-keeping genes: Gapdh and Tbp. The selected promoter regions are positioned within 1000 bp upstream of transcription start. In order to avoid cell culturing differences, DNA from more than 18 individual DNA isolations from non-treated MEF cells collected over a period of 8 months was analyzed and the mean steady-state DNA damage levels in promoter and coding regions were determined (Fig. 1C). Interestingly, we found that although the steady state level of DNA damage in mt-Rnr1 was higher than in the nuclear genes Tbp, Ndufa9 and Ndufs2, it was comparable with that in the house-keeping gene Gapdh. Furthermore, the coding region of Ndufs2 contained significantly more damages than its promoter region, while there were no promoter:coding differences in the other genes. Possible explanations for the heterogeneous accumulation of DNA damage could be site-specific differences in damage formation or repair.
3.2. Gene-specific ROS-induced DNA damage and repair To follow up on possible site-specific differences in damage formation or repair capacity, we applied menadione (50 M for 45 min) to induce superoxide-induced damage in mtDNA and nDNA as in Fig. 1 and analyzed the site-specific damage formation. As shown in Fig. 2, the amount of damage accumulated during menadione-induced ROS was gene-specific, and mtDNA was more vulnerable to insult as compared to the nuclear genes. Promoter and coding regions were similarly affected by menadione in the four nuclear genes, although the absolute damage frequency differed between the genes (e.g. Ndufa9 accumulates 2.5fold more promoter damage than Gapdh; p < 0.01). To analyze DNA repair capacity, we assessed the remaining level of DNA damage in treated cells that were allowed to recover for 0.5 h, 1 h and 4 h in medium without menadione. During the following recovery, the rate of damage removal from promoter and coding regions were not significantly different and the repair rates for the distinct sites were not significantly different. Thus, we conclude that accumulation of DNA damage during oxidative stress is site-specific, while the subsequent repair is not. Together, these results indicate that the heterogeneous site-specific steady state level of DNA damage (Fig. 1C) is not the net effect of different site-specific repair efficiencies.
Fig. 1. Assessment of genome integrity. (A and B) Validation of the restriction enzyme-based qPCR method to detect oxidative DNA damage in mtDNA (mt-Rnr1) and nDNA (Ndufa9). (A) Accumulated DNA damage level in MEF cells treated with 50 M menadione for 15–60 min. (B) Concentration-dependent damage induction upon treatment with 0–100 M menadione for 45 min. Data are mean ± SE of three independent experiments. *p < 0.05 vs. ctrl, # p < 0.05 mt-Rnr1 vs. Ndufa9. (C) Steadystate level of DNA damage in 9 different genomic sites as specified. The results shown are the mean ± SE of 18 separate preparation of DNA from non-treated MEFs (passage 3–20) over a period of 8 months. *p < 0.05; significant differences relative to mt-Rnr1 are indicated by horizontal lines. The asterix above bar indicates difference between promoter/coding regions.
3.3. Transcriptional arrest has no effect on ROS-induced DNA damage accumulation or subsequent repair Oxidative stress is known to activate transcriptional responses that act to increase tolerance to genotoxic stress [15–17] and it was a possibility that the repair kinetics following menadione treatment was not representative for the steady-state situation in non-treated
Please cite this article in press as: J.M. Strand, et al., The distribution of DNA damage is defined by region-specific susceptibility to DNA damage formation rather than repair differences, DNA Repair (2014), http://dx.doi.org/10.1016/j.dnarep.2014.03.003
G Model DNAREP-1907; No. of Pages 8 4
ARTICLE IN PRESS J.M. Strand et al. / DNA Repair xxx (2014) xxx–xxx
Fig. 2. Gene-specific susceptibility to damage induction and subsequent repair. MEF cells were treated with 50 M menadione for 45 min before medium was removed and cells recovered for the indicated time points. (A) Induction and removal of DNA damage in different nuclear genes comparing the promoter and coding region. (B) Induction and removal of damage in mtDNA. Data are mean ± SE of three independent experiments. *p < 0.05 vs. nt.
cells. To investigate in more detail whether transcription has any effect on repair, we reassessed repair in the presence of ␣-amanitin, a potent inhibitor of RNA polymerase II. Pretreatment of cells confirmed the specific inhibition of nuclear transcription compared to mitochondrial transcription (Fig. 3A), but ␣-amanitin neither influence the damage formation nor the following repair in promoter or coding regions of Ndufa9 or Gapdh, or in mtDNA (Fig. 3). ␣-amanitin alone did not influence DNA integrity (data not shown). Since mitochondrial transcription does not involve RNA polymerase II, we investigated damage formation and removal in csbm/m MEF cells, which carry a homozygous nonsense mutation in the Csb gene. The CSB protein is reported to participate in mtDNA repair in association with mitochondrial transcription [18,19]. However, our data indicated that repair of mtDNA was independent of CSB function (Fig. 3B). Several studies suggest that induction of oxidative DNA damage itself can affect gene expression [20,21] but mRNA levels of the analyzed gene targets revealed no changes immediately after oxidant treatment and only a minor although significant increase in mRNA level of the Tbp gene after 4 h of recovery (Supplementary
Fig. 2). We conclude that the RNA polymerase II and CSB proteins do not interfere with oxidative damage formation or removal from nDNA and mtDNA. 3.4. Reduced energy availability impairs the repair of oxidative DNA lesions Several DNA repair enzymes are energy-dependent, thus continuous ATP supply is necessary to facilitate efficient DNA repair in situations of increased oxidative stress. The nuclear and mitochondrial compartments can potentially be unevenly dependent on glycolytic and oxidative phosphorylation (OXPHOS) – type of ATP production. To address the question if repair of mtDNA or nDNA is specifically requiring glycolytic or OXPHOS ATP, we used a combined treatment of oligomycin, an inhibitor of mitochondrial ATP synthase and 2-deoxy-d-glucose (2-DG), a glucose-6-phosphate mimicker and competitive hexokinase inhibitor to investigate the repair in mt-Rnr1 and in the promoter regions of Ndufa9 and Ndufs2. The promoter region was selected as DNA damage previously
Please cite this article in press as: J.M. Strand, et al., The distribution of DNA damage is defined by region-specific susceptibility to DNA damage formation rather than repair differences, DNA Repair (2014), http://dx.doi.org/10.1016/j.dnarep.2014.03.003
G Model DNAREP-1907; No. of Pages 8
ARTICLE IN PRESS J.M. Strand et al. / DNA Repair xxx (2014) xxx–xxx
5
Fig. 3. Lack of correlation between transcription and repair in nDNA and mtDNA. (A) MEF cells were treated with ␣-amanitin for 12 h before mRNA levels were quantified from the nuclear genes (Gapdh and Ndufa9) and mitochondrial gene mt-Rnr1 related to nt cells (dashed line). Accumulated damage in mtDNA (B) and nDNA (C) was estimated in MEF cells pre-incubated with ␣-amanitin, and in csbm/m cells after treatment with 50 M menadione for 45 min and after a recovery period of 4 h. Data are mean ±SE of three independent experiments. *p < 0.05 vs. nt.
was found to accumulate in promoters of mitochondrial genes and correlate with mitochondrial dysfunction [13]. Oligomycin blocked ATP synthase-dependent respiration, thereby inferring that OXPHOS was dispensable for providing the remaining cellular ATP (Supplementary Fig. 3A). Oligomycin was not efficient in depleting cellular ATP, suggesting that alternative ATP generating pathways were sufficiently effective to restore cellular levels. 2-DG led to 20% reduction of cellular ATP which was further decreased to 50% of normal level in combination with oligomycin (Supplementary Fig. 3B). Depletion of ATP by 2-DG and oligomycin, either alone or in combination did not impact on genomic DNA integrity in cells that were not exposed to menadione (data not shown). Furthermore, treatment with either oligomycin or 2-DG alone had no effect on the amount of induced DNA damage or subsequent repair after oxidative stress in either mtDNA or nDNA (Fig. 4). In contrast, the combination of oligomycin and 2-DG completely blocked repair of mtDNA (Fig. 4A) and reduced repair of the nuclear promoter regions significantly (Fig. 4B). These data demonstrate that the source of ATP is not essential for efficient repair of mtDNA or nDNA, and that repair of mtDNA is more sensitive to alterations in cellular ATP than repair of nDNA.
4. Discussion In this study, we have quantified site-specific DNA damage and compared steady-state levels of DNA damage in promoter and coding regions with susceptibility to damage induction during oxidative stress and subsequent repair kinetics. We find that stressinduced damage accumulates in a gene-specific manner, and that coding and promoter regions within a gene are equally susceptible. Repair of induced damage is efficient with the major exception being repair of mtDNA, which is delayed and slower compared to repair of nDNA. While promoter and coding regions displayed similar damage dynamics with respect to formation and removal of oxidative stress/induced damage, the damage distribution in non-treated cells was more complex as coding region of Ndufs2 exhibited significantly more damage than promoter region. Our results indicate that repair differences do not dictate the steadystate level of damage in nDNA and mtDNA. The steady-state damage distribution somehow determined the amount of damage induced by mitochondrial ROS. For instance, the relatively intact Ndufa9 gene accumulated 2–3-fold more damage (Fig. 2) than the more damaged Gapdh gene. Notably, if the
Please cite this article in press as: J.M. Strand, et al., The distribution of DNA damage is defined by region-specific susceptibility to DNA damage formation rather than repair differences, DNA Repair (2014), http://dx.doi.org/10.1016/j.dnarep.2014.03.003
G Model DNAREP-1907; No. of Pages 8 6
ARTICLE IN PRESS J.M. Strand et al. / DNA Repair xxx (2014) xxx–xxx
Fig. 4. Energy requirement for repair of cellular DNA. MEF cells were exposed to oligomycin (2.5 g/mL), 2-DG (20 mM) or both 2 h before treatment with 50 M menadione for 45 min followed by 4 h recovery. DNA damage was quantified in (A) mtDNA and (B) nDNA. Data are mean ± SE of three independent experiments. *p < 0.05 vs. ctrl.
ROS-induced damages are added to the steady-state level, the total amounts of nDNA damage in the different sites are comparable. Thus, oxidative stress may induce global DNA damage that masks the heterogeneity in non-treated cells. Menadione uncouples normal electron flux and causes excessive mitochondrial superoxide generation. It has been shown that intracellular mitochondria differ in their ability to generate ROS [22]. One explanation for the heterogeneous damage distribution could therefore be that the distance between the chromosome region in where the gene resides
and the distinct ROS-generating mitochondrion might influence the damage level. In this respect, the unifying effect of menadione on damage distribution could be due to a shift were all rather than a subpopulation of mitochondria generates ROS. This postulation conflicts previous reports where mitochondrial ROS was dispensable for accumulation of DNA damage [23], although the manipulation of mitochondrial ROS in the cited study did not influence DNA integrity, like we demonstrate upon menadionemediated ROS generation. The relative high level of steady state
Please cite this article in press as: J.M. Strand, et al., The distribution of DNA damage is defined by region-specific susceptibility to DNA damage formation rather than repair differences, DNA Repair (2014), http://dx.doi.org/10.1016/j.dnarep.2014.03.003
G Model DNAREP-1907; No. of Pages 8
ARTICLE IN PRESS J.M. Strand et al. / DNA Repair xxx (2014) xxx–xxx
as well as induced damage in mt-Rnr1 might be explained by its vicinity to the electron transport chain, which is known to generate superoxide anion as a byproduct of active respiration. Our data support the notion that DNA damage might be targeted specifically to distinct regions as an important signal for gene regulation [15–17]. There is experimental evidence for that reversible DNA damage is actively involved in physical processes. For instance, DNA damage was induced in mice that were subjected to environmental stimulation [24] and inability to reverse the damage was associated with neurodegeneration. In the extrapolation of these findings, the steady-state distribution of DNA damage could represent the functional state of the cell that can be disturbed by mitochondrial manipulation as here or by age-associated processes, resulting in mitochondrial dysfunction [13]. We have previously compared the method used here [14] with the established method for mtDNA damage assessment, based on qPCR-quantification of intact 10 kb template DNA; developed by van Houten and coworkers [24]. The advantage of the current method is that it enables assessment of a more detailed region of DNA. Although an alternative, principally similar method was recently established to quantify DNA damage lesions in regions of 1 kb size [8], hot spots within one specific gene region may not be detected. As described in materials and methods, we quantify damage frequency based on the inhibition of TaqI restriction enzyme cleavage. Both methods score for strand breaks and bulky lesions [25]. The advantage of the method used here is that lesions that are efficiently bypassed (such as the abundant 8-oxoguanine) are detectable as these lesions previously have been utilized based on their ability to inhibit restriction enzyme digestion [26]. Interpretation of absolute number of frequency should nevertheless be done with precaution for both methods; e.g. the relaxed versus supercoil form of DNA determines the PCR efficiency [27], and damages may induce conformational changes that inhibit restriction cleavage although being positioned outside the palindromic recognition sequence. Hence, the damage detection region may span wider than the recognition sequence of TaqI restriction enzyme. The negative effect of ␣-amanitin on repair efficiency has two implications. First, transcription-dependent adaptation following oxidative stress is not important for efficient repair and second, this is in line with the lack of evidence for transcription-coupled repair of oxidative DNA damage. On the other hand, our results are in contrast with earlier studies where CSB was found to be important for efficient removal of mtDNA damage [18]. The explanation for this discrepancy can only be found by comparing similar cell lines and damage condition with the two different damage detection methods. One important difference from our studies is that Stevnsner and coworkers selectively studied non-replicated DNA from mitochondria whereas we here studied total DNA. It is therefore a possibility that the unrepaired mtDNA is masked by corresponding increased replication of the non-damaged mtDNA molecules although we find this explanation unlikely as the mtDNA copy number in our experiments were slightly reduced after menadione exposure (data not shown). A previous study demonstrated that exogenous ATP stimulated the translocation of BER enzyme Ape1 into the nucleus [28]. Additionally, it was reported that reduced ATP level correlated with increased DNA damage in promoter regions of nuclear genes with age [13]. The overall conclusion from the combinatory treatment of 2-DG and oligomycin, is that ATP depletion of 50% or more is required to impair DNA efficiency. The combination significantly reduced repair of nDNA (Fig. 4B) and completely prevented any removal of damage from mtDNA (Fig. 4A), suggesting that repair of these two cellular genomes are differently dependent on ATP. In fact, an early study by Oei and Ziegler [29] suggests that the ATP required for the energy-demanding repair step is derived from poly-ADP ribose (PAR) rather than ATP. Since the PAR moiety is
7
produced from NAD by PARP1, this energy-requiring step would be less sensitive to ATP depletion. Although PARP1 has been identified in mitochondria [30], it is generally accepted that its nuclear activity is important for damage recovery. The delay in repair of mtDNA (Fig. 2B) after menadione exposure should be interpreted in view of the strong ATP-depleting effect of menadione. Recovery of cellular ATP after menadione exposure requires 1 h [31] and this transient ATP depletion could be responsible for the delay process that is even more pronounced upon 2-DG and oligomycin administration. In conclusion, our analyses of promoter and coding regions of four nuclear genes and the mitochondrial gene mt-Rnr1 demonstrate that the distribution of oxidative stress-induced damage is indicative of gene-specific susceptibility to damage. The capacity to remove damage differed between nDNA and mtDNA. Repair of nDNA was relatively resistant to changes in the ATP concentration compared to repair of mtDNA, and indicates that inefficient repair of mtDNA with subsequent accumulation of mtDNA damage is an early consequence of bioenergetic dysfunction. Conflict of interest statement The authors declare no conflict of interest. Acknowledgements This work was supported by grants from Research Council of Norway to MB and the University of Oslo. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.dnarep. 2014.03.003. References [1] M.L. Hegde, A.K. Mantha, T.K. Hazra, K.K. Bhakat, S. Mitra, B. Szczesny, Oxidative genome damage and its repair: implications in aging and neurodegenerative diseases, Mech. Ageing Dev. 133 (2012) 157–168. [2] S. Xanthoudakis, R.J. Smeyne, J.D. Wallace, T. Curran, The redox/DNA repair protein, Ref-1, is essential for early embryonic development in mice, Proc. Natl. Acad. Sci. U.S.A. 93 (1996) 8919–8923. [3] M.A. Graziewicz, R.J. Bienstock, W.C. Copeland, The DNA polymerase gamma Y955C disease variant associated with PEO and parkinsonism mediates the incorporation and translesion synthesis opposite 7,8-dihydro-8-oxo-2 deoxyguanosine, Hum. Mol. Genet. 16 (2007) 2729–2739. [4] H.E. Krokan, M. Bjoras, Base excision repair, Cold Spring Harb. Perspect. Biol. 5 (2013) a012583. [5] V.K. Batra, D.D. Shock, W.A. Beard, C.E. McKenna, S.H. Wilson, Binary complex crystal structure of DNA polymerase beta reveals multiple conformations of the templating 8-oxoguanine lesion, Proc. Natl. Acad. Sci. U.S.A. 109 (2012) 113–118. [6] E.J. Merino, A.K. Boal, J.K. Barton, Biological contexts for DNA charge transport chemistry, Curr. Opin. Chem. Biol. 12 (2008) 229–237. [7] F.M. Yakes, H.B. Van, Mitochondrial DNA damage is more extensive and persists longer than nuclear DNA damage in human cells following oxidative stress, Proc. Natl. Acad. Sci. U.S.A. 94 (1997) 514–519. [8] O. Rothfuss, T. Gasser, N. Patenge, Analysis of differential DNA damage in the mitochondrial genome employing a semi-long run real-time PCR approach, Nucleic Acids Res. 38 (2010) e24. [9] M. Ljungman, P.C. Hanawalt, Efficient protection against oxidative DNA damage in chromatin, Mol. Carcinog. 5 (1992) 264–269. [10] V.A. Bohr, C.A. Smith, D.S. Okumoto, P.C. Hanawalt, DNA repair in an active gene: removal of pyrimidine dimers from the DHFR gene of CHO cells is much more efficient than in the genome overall, Cell 40 (1985) 359–369. [11] D. Banerjee, S.M. Mandal, A. Das, M.L. Hegde, S. Das, K.K. Bhakat, I. Boldogh, P.S. Sarkar, S. Mitra, T.K. Hazra, Preferential repair of oxidized base damage in the transcribed genes of mammalian cells, J. Biol. Chem. 286 (2011) 6006–6016. [12] Y. Tu, S. Tornaletti, G.P. Pfeifer, DNA repair domains within a human gene: selective repair of sequences near the transcription initiation site, EMBO J. 15 (1996) 675–683. [13] T. Lu, Y. Pan, S.Y. Kao, C. Li, I. Kohane, J. Chan, B.A. Yankner, Gene regulation and DNA damage in the ageing human brain, Nature 429 (2004) 883–891.
Please cite this article in press as: J.M. Strand, et al., The distribution of DNA damage is defined by region-specific susceptibility to DNA damage formation rather than repair differences, DNA Repair (2014), http://dx.doi.org/10.1016/j.dnarep.2014.03.003
G Model DNAREP-1907; No. of Pages 8 8
ARTICLE IN PRESS J.M. Strand et al. / DNA Repair xxx (2014) xxx–xxx
[14] W. Wang, P. Osenbroch, R. Skinnes, Y. Esbensen, M. Bjoras, L. Eide, Mitochondrial DNA integrity is essential for mitochondrial maturation during differentiation of neural stem cells, Stem Cells 28 (2010) 2195–2204. [15] B. Perillo, M.N. Ombra, A. Bertoni, C. Cuozzo, S. Sacchetti, A. Sasso, L. Chiariotti, A. Malorni, C. Abbondanza, E.V. Avvedimento, DNA oxidation as triggered by H3K9me2 demethylation drives estrogen-induced gene expression, Science 319 (2008) 202–206. [16] B.G. Ju, V.V. Lunyak, V. Perissi, I. Garcia-Bassets, D.W. Rose, C.K. Glass, M.G. Rosenfeld, A topoisomerase IIbeta-mediated dsDNA break required for regulated transcription, Science 312 (2006) 1798–1802. [17] K.A. Ziel, V. Grishko, C.C. Campbell, J.F. Breit, G.L. Wilson, M.N. Gillespie, Oxidants in signal transduction: impact on DNA integrity and gene expression, FASEB J. 19 (2005) 387–394. [18] T. Stevnsner, S. Nyaga, N.C. Souza-Pinto, G.T.J. van der Horst, T.G.M.F. Gorgels, B.A. Hogue, T. Thorslund, V.A. Bohr, Mitochondrial repair of 8-oxoguanine is deficient in Cockayne syndrome group B, Oncogene 21 (2002) 8675–8682. [19] B.R. Berquist, C. Canugovi, P. Sykora, D.M. Wilson III, V.A. Bohr, Human Cockayne syndrome B protein reciprocally communicates with mitochondrial proteins and promotes transcriptional elongation, Nucleic Acids Res. 40 (2012) 8392–8405. [20] C. Marietta, H. Gulam, P.J. Brooks, A single 8,5 -cyclo-2 -deoxyadenosine lesion in a TATA box prevents binding of the TATA binding protein and strongly reduces transcription in vivo, DNA Repair (Amst) 1 (2002) 967–975. [21] M.K. Hailer-Morrison, J.M. Kotler, B.D. Martin, K.D. Sugden, Oxidized guanine lesions as modulators of gene transcription. Altered p50 binding affinity and repair shielding by 7,8-dihydro-8-oxo-2 -deoxyguanosine lesions in the NFkappaB promoter element, Biochemistry 42 (2003) 9761–9770. [22] G.H. Mathisen, I.H. Thorkildsen, R.E. Paulsen, Secretory PLA2-IIA and ROS generation in peripheral mitochondria are critical for neuronal death, Brain Res. 1153 (2007) 43–51.
[23] S. Hoffmann, D. Spitkovsky, J.P. Radicella, B. Epe, R.J. Wiesner, Reactive oxygen species derived from the mitochondrial respiratory chain are not responsible for the basal levels of oxidative base modifications observed in nuclear DNA of Mammalian cells, Free Radic. Biol. Med. 36 (2004) 765–773. [24] J.H. Santos, J.N. Meyer, B.S. Mandavilli, H.B. Van, Quantitative PCR-based measurement of nuclear and mitochondrial DNA damage and repair in mammalian cells, Methods Mol. Biol. 314 (2006) 183–199. [25] A.M. Furda, A.S. Bess, J.N. Meyer, H.B. Van, Analysis of DNA damage and repair in nuclear and mitochondrial DNA of animal cells using quantitative PCR, Methods Mol. Biol. 920 (2012) 111–132. [26] E. Larsen, K. Kwon, F. Coin, J.M. Egly, A. Klungland, Transcription activities at 8-oxoG lesions in DNA, DNA Repair (Amst) 3 (2004) 1457–1468. [27] J. Chen, F.F. Kadlubar, J.Z. Chen, DNA supercoiling suppresses real-time PCR: a new approach to the quantification of mitochondrial DNA damage and repair, Nucleic Acids Res. 35 (2007) 1377–1388. [28] A. Pines, L. Perrone, N. Bivi, M. Romanello, G. Damante, M. Gulisano, M.R. Kelley, F. Quadrifoglio, G. Tell, Activation of APE1/Ref-1 is dependent on reactive oxygen species generated after purinergic receptor stimulation by ATP, Nucleic Acids Res. 33 (2005) 4379–4394. [29] S.L. Oei, M. Ziegler, ATP for the DNA ligation step in base excision repair is generated from poly(ADP-ribose), J. Biol. Chem. 275 (2000) 23234–23239. [30] M.N. Rossi, M. Carbone, C. Mostocotto, C. Mancone, M. Tripodi, R. Maione, P. Amati, Mitochondrial localization of PARP-1 requires interaction with mitofilin and is involved in the maintenance of mitochondrial DNA integrity, J. Biol. Chem. 284 (2009) 31616–31624. [31] P.O. Osenbroch, P. Auk-Emblem, R. Halsne, J. Strand, R.J. Forstrom, I. van der Pluijm, L. Eide, Accumulation of mitochondrial DNA damage and bioenergetic dysfunction in CSB defective cells, FEBS J. 276 (2009) 2811–2821.
Please cite this article in press as: J.M. Strand, et al., The distribution of DNA damage is defined by region-specific susceptibility to DNA damage formation rather than repair differences, DNA Repair (2014), http://dx.doi.org/10.1016/j.dnarep.2014.03.003
