Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Oct 1.
Published in final edited form as: Methods. 2016 Apr 26;108:99–110. doi: 10.1016/j.ymeth.2016.04.029

Application of the microfluidic-assisted replication track analysis to measure DNA repair in human and mouse cells

Piri Welcsh a, Keffy Kehrli a,1, Pavlo Lazarchuk a,2, Warren Ladiges b, Julia Sidorova a,*
PMCID: PMC5035571  NIHMSID: NIHMS783636  PMID: 27130890

Abstract

Functional studies of the roles that DNA helicases play in human cells have benefited immensely from DNA fiber or single molecule technologies, which enable us to discern minute differences in behaviors of individual replication forks in genomic DNA in vivo. DNA fiber technologies are a group of methods that use different approaches to unravel and stretch genomic DNA to its contour length, and display it on a glass surface in order to immuno-stain nucleoside analog incorporation into DNA to reveal tracks (or tracts) of replication. We have previously adopted a microfluidic approach to DNA stretching and used it to analyze DNA replication. This method was introduced under the moniker maRTA or microfluidic assisted Replication Track Analysis, and we have since used it to analyze roles of the RECQ helicases WRN and BLM, and other proteins in normal and perturbed replication. Here we describe a novel application of maRTA to detect and measure repair of DNA damage produced by three different agents relevant to etiology or therapy of cancer: methyl-methanesulfonate, UV irradiation, and mitomycin C. Moreover, we demonstrate the utility of this method by analyzing DNA repair in cells with reduced levels of WRN or of the base excision repair protein XRCC1.

Keywords: DNA, single molecule, replication, repair, XRCC1, WRN

1. INTRODUCTION

DNA helicases are essential molecular motors that separate strands of duplex DNA to enable DNA replication, repair, recombination, telomere maintenance, and transcription. The importance of helicases in the maintenance of a stable, efficiently operating genome is underscored by the fact that at least 15 out of 31 DNA helicases encoded in the human genome cause familial cancer-prone disease conditions, if mutated [1, 2]. In vivo studies of molecular roles of DNA helicases in replication have greatly benefited from the development of DNA fiber assays. In these assays, genomic DNA is pulse-labeled in vivo with pyrimidine nucleoside analogs (i.e. BrdU, CldU, IdU), isolated, stretched, and deposited in an orderly fashion as individual, clearly separated molecules onto a glass surface. This enables visualization and robust quantitation of so-called replication tracks (or tracts) under a regular fluorescent microscope at a resolution of 1μM=2-4Kb. Several different DNA fiber procedures have been developed. Some are known under specific brand names, e.g. DNA combing, SMARD, maRTA [3-10]. All of these procedures share DNA labeling and immunodetection protocols though differ in their approach to the critical step of stretching and immobilizing DNA molecules. Availability of antibodies that selectively detect CldU/BrdU or IdU/BrdU, offers versatility in designing well-controlled, informative experiments targeting different aspects of DNA replication. More recently, we added EdU to the repertoire of thymidine analog labels, enabling simultaneous detection of three colors of tracks [11] (also see Materials and Methods).

DNA fiber assays can measure such parameters of replication as replication fork rate of progression, frequency of replication origin firing, and density of replication firing events in replication domains. Perhaps yet more powerful is the application of DNA fiber assays to the study of perturbed replication known as replication stress, which is triggered by the presence of lesions in replicating DNA, replisome blockage, and insufficient or unbalanced nucleotide, histone, or DNA-servicing enzyme pools. DNA fiber assays are excellent in detecting inappropriate fork stalling, slowing, or irreversible inactivation, as well as nascent strand resection – all of which become prevalent under conditions of replication stress. There are numerous examples of studies where DNA helicases, and RECQs in particular, were assigned specific roles in these aspects of normal and/or stressed replication based on the results of DNA fiber assays.

While the field of replication enjoys the ever wider availability of DNA fiber assays of various flavors, the related area of DNA repair research by and large lacks an equivalent technology. Since our introduction, in 2008, of a version of a DNA fiber assay called microfluidic-assisted replication track analysis, or maRTA, we have sought ways to apply it to the study of DNA repair processes that are not necessarily coupled with replication. We described our first application of maRTA to DNA repair in 2012 [12], and here we build upon our original research to expand the application of maRTA to measure BER, NER, and crosslink/bulky adduct repair.

2. MATERIALS and METHODS

2.1. Cell lines and growth conditions

The SV40-transformed human fibroblast GM639 cell line and its pNeoA derivative GM639cc1 have been described before [13-16]. Primary fibroblast line Hi3 was described previously [17]. MCF10a, a spontaneously immortalized mammary epithelial cell line and HCC1937 Brca1−/− breast cancer cell line complemented with a BRCA1 transgene are gifts of Drs. Piri Welcsh and Elizabeth Swisher (University of Washington), and Dr. Toshi Taniguchi (Fred Hutchinson Cancer Research Center).

GM639cc1 were grown in Dulbecco Modified Minimal Essential Medium (DMEM, Gibco or Hyclone) supplemented with L-glutamine, sodium pyruvate, 10% fetal bovine serum (FBS, Hyclone) and antibiotics, and MCF10a were grown in MGEM media (Lonza) supplemented with Single Quots (Lonza), 1% fetal bovine serum and antibiotics. HCC1937/BRCA1 cells were grown in RPMI (Gibco or Hyclone) supplemented with 15% FBS and antibiotics. All cell lines were kept in a humidified 5% CO2, 37°C incubator.

XRCC1 null heterozygote and wild type mouse fibroblasts were generated from ear tissue of adult mice described previously [12, 18]. Mice were euthanized by CO2 inhalation. Ear tissue was sterilized with Povidone-Iodine swabsticks (PDI), rinsed with 70% ethanol and PBS containing penicillin/streptomycin and 2× Fungizone (Life Technologies), then minced in PBS containing 1 mg/ml collagenase/dispase (Roche Applied Science). Tissue fragments were rotated at 37° C in a 5% O2 incubator for 45 min, then overnight at 37° C after adding 5 ml of media. Tissue was further dissociated by pipetting, and then filtered through a 100 mm mesh nylon filter (BD Biosciences) prior to centrifugation and resuspension in fresh growth media. Mouse cells were grown in Dulbecco Modified Eagle’s Medium (DMEM) with 4.5 g/L glucose and pyruvate (BioWhittaker), supplemented with 2 mM L-glutamine, 10% (v/v) Fetal Clone III serum (Hyclone), penicillin G (100 U/ml) and streptomycin sulfate (100 mg/ml; BioWhittaker) and 2× non-essential amino acids (BioWhittaker) in a humidified 37°C, 7% CO2 incubator in 5% O2.

2.2. Drugs and other reagents

Stock solution of 5-iododeoxyuridine (IdU, Sigma-Aldrich) was 2.5mM in water, 5-chlorodeoxyuridine (CldU, Sigma-Aldrich) was 10mM in water, 5-ethynyldeoxyuridine (EdU, Life Technologies) was at 10mM in DMSO. IdU and CldU were used at a final concentration of 50μM and EdU was used at 10μM. Stock of mitomycin C (EMD Chemicals) was at 10mM in DMSO. Methyl-methanesulfonate was purchased from Sigma-Aldrich and diluted to 1% in growth media prior to use. T4 PDG endonuclease was purchased from NEB and S1 nuclease from Thermo Scientific. All reagent stocks and enzymes were stored at −20°C.

2.3. UV irradiation

UV irradiations were performed with a UVG-4 portable UV lamp (UVP), and UV doses were calibrated using a shortwave UV measuring meter J-225 (UVP). Cells were irradiated in PBS or DMEM media without the pH indicator.

2.4. WRN depletion

shRNA-mediated depletion of WRN was performed as previously described [14, 15] and the WRN protein level was quantified in Western blots with α-WRN antibody 195C (Cat. No. W0393 Sigma) and α-CHK1 (Cat. No. sc-8408 Santa Cruz) as loading control. Proteins were visualized on Western blots by ECL (ThermoScientific) and quantified using Storm Phosphorimager (Molecular Dynamics) or FluorChem Imager (Alpha Inotech).

2.5. maRTA

2.5.1. DNA isolation for maRTA and treatments to convert lesions to double-strand breaks

Embedding cells in agarose, cell lysis, and release of DNA from agarose (with β-agarase, NEB) were as described in detail previously [8]. β-agarase treatments were overnight at 42°C. A typical DNA sample was isolated from 200,000 cells and had a final volume of 400-500μL (in 1× β-agarase buffer). To convert MMS lesions to double-strand breaks, an overnight incubation at 56°C during the step of cell lysis in agarose plugs, which is performed during our DNA isolation procedure, was sufficient for mouse fibroblasts. For human cells, an additional step of heating DNA preps for 4 hrs at 56°C was carried out after DNA was released from agarose.

For T4 PDG endonuclease treatments, 10μL of genomic DNA (aliquoted with a wide bore tip) was assembled into a reaction mixture with 2 μL of 10×PDG buffer (supplied with the enzyme), deionized water and 2 units of PDG to make 20 μL final volume. Amount of PDG may require initial titration. Control reactions had DNA and 1×PDG buffer only. Reactions were incubated at 37°C for 30 min, then placed on ice. If S1 nuclease digestion was included, reaction mixes from the PDG step were adjusted with modified 5×S1 buffer (200mM Na acetate pH4.5, 1M NaCl, 10mM ZnSO4; this buffer takes into account that reaction mixes already contain 100mM NaCl) to make buffer conditions compatible with S1, and incubated with 1/128 to 1/16 units of the enzyme for 30 min at 37°C. Reactions were returned to ice, adjusted to 1mM EDTA and incubated at 65°C for 10 min to inactivate the enzymes. 10μL of digests were loaded onto a 0.75% agarose gel to assess the degree of digestion. Mock-treated genomic DNA should migrate as a single, high molecular weight band, and enzymatic digestion produces an enzyme dose-dependent smearing beneath this band. Optimal digestion is within the range of enzyme doses that retain the high molecular weight band virtually unaffected and produce little to no smearing.

For S1 only digestion, 10μL of single genomic DNA or 7μL each if two DNAs were used, were mixed with 4μL of 5× S1 buffer supplied with the enzyme, and 1/32 to 1/2 units of S1, then incubated at 37°C for 30 min and quenched as above. To check for digestion, DNAs were resolved in 0.75% agarose as above. The optimal range of S1 is typically 1/8-1/2 units per reaction. All digested DNAs should be stretched immediately or within 2-3 days at the latest as they are not stable and will degrade in storage.

Stretching of digested DNAs was performed as described [8]. Prior to stretching, DNA preps were pH-adjusted by adding 1/10 volume of 10×TE pH8.0. 1μL of a prep was loaded into a microchannel and coverslip sandwich. Microchannels were removed and coverslips with stretched DNA were air-dried overnight to facilitate binding of DNA to glass. No methanol/acetic acid fixing step is required.

2.5.2. Immunofluorescent staining of stretched DNA.

Staining of stretched DNA was performed as described in detail previously [8] for CldU and IdU labels, and with the following modifications if EdU was used. After drying overnight, coverslips were immersed in 2.5N HCl for 45 min, then neutralized in one rinse of 100mM Na borate pH8.0 and a rinse in PBS. Coverslips were rinsed in 3%BSA in PBS while a Click-IT reaction was assembled in this order of addition: PBS, 20μM biotin azide, 11mM Na ascorbate, 2.2mM CuSO4. Stock reagents to make this reaction were: 100mM Na ascorbate solution in water made fresh from powder (Sigma-Aldrich) just prior to use; 8mM biotin azide (Life Technologies) in DMSO, stored at −20°C; and 100mM CuSO4 (Sigma-Aldrich), stored at room temperature. Cover slips were incubated DNA side down on parafilm in 20-25μL of Click-It reaction for 30 min at room temperature, then rinsed once in PBS/3%BSA and blocked as described previously [8] in PBS with 5%BSA, 0.5% Tween 20. Reagents to detect EdU-biotin and IdU were as follows, added in this order (diluted in PBS with 5%BSA, 0.5%Tween 20, 10% NGS): 1) 1:20 dilution of Texas Red Neutravidin (Life technologies, Cat. No. A2665); 2) 1:80 dilution of biotinylated anti-avidin D (Vector Laboratories, Cat. No. BA-0300) and 1:6 dilution of IdU antibody (BD Biosciences Cat. No.347580), followed by washes as above; 3) 1:20 dilution of Texas Red Neutravidin and 1:1000 dilution of Alexa 488-conjugated goat α-mouse secondary antibody (Life Technologies Cat. No. A-11001). Each antibody/neutravidin incubation step was for 1 hr at room temperature, followed by 3 × 5 min. washes in PBS with 1%BSA, 0.1% Tween 20. After the last set of washes, coverslips were briefly rinsed in water, air-dried, and mounted on glass slides w/o mounting media.

Microscopy of stretched DNAs was performed on a Zeiss Axiovert microscope with a 40× objective, and data collection was as described before [8].

2.6. Pulsed-field gel electrophoresis and comet assay

Pulsed-field gel electrophoresis (PFGE) was performed under conditions described in [19] in a CHEF DR II apparatus (Bio-Rad Laboratories). DNAs for electrophoresis were prepared in agarose plugs as recommended for the procedure [20]. Comet assays were performed using buffers, agarose and glass slides purchased from Trevigen (Cat. No. 4250-050-K), and according to the manufacturer’s recommendations. Microscopy of comets was performed on a Zeiss Axiovert microscope with a 10× objective. Images of comets were analyzed using Cometscore software (TriTek).

2.7. Data analysis and presentation

Numerical data were collected and inspected in Excel and analyzed and plotted in R studio. Grayscale Western blot images were generated by FluorChem Imager (Alpha Inotech), saved in TIFF format, adjusted for brightness/contrast and cropped using Corel Photo-Paint, then assembled into figures in CorelDraw. Color images of replication tracks were generated by AxioCam of a Zeiss Axiovert microscope, and saved as merged 3-channel (RGB) images in JPEG format. For presentation, brightness/contrast were adjusted in green and red channels in Corel Photo-Paint, and the image was placed in a CorelDraw Figure. Brightness/contrast adjustments were always done to the whole images.

3. RESULTS AND DISCUSSION

3.1 Principles of applying a modified maRTA approach to study DNA repair

Lesion detection is the basis of any direct DNA repair assay, and it can be based either on the unique chemical identity of a lesion (e.g. modified nucleotides such as CPDs, O6meG, 8-oxoG) against which an antibody can be raised, or on the fact that most lesions, or their repair intermediates, produce discontinuities in the DNA backbone, i.e. single- or double-strand breaks. These breaks can be detected by virtue of causing fragmentation of genomic DNA. Similar approaches can, in theory, be exploited by a DNA fiber assay. Yet to our knowledge, direct immune-detection of individual lesions in stretched DNA molecules has been, with a few notable exceptions [21, 22], problematic due to low signal to noise ratio. On the other hand, inferring lesion formation and repair from DNA fragment size distribution can work well within the context of a DNA fiber assay, as we observed previously [12], and as demonstrated by Kidane et al in their approach to detection of AP sites in human DNA [23]. Genomic DNA can be labeled to create tracks that are discreet yet sufficiently long to contain within their length a subset of lesions generated in the genome after DNA damaging treatment (Figure 1A). These lesions can then be detected if they directly break the labeled tracks in vivo (as is the case with X ray lesions) or make them more susceptible to breakage via ex vivo manipulation, which is the focus of the current study. In either case, the readout is a decrease in the overall length distribution of tracks in DNA from treated compared to untreated cells. Consequently, the main difference between the “classic” maRTA protocol and its modification to detect damage and repair (Figure 1B), is that the latter utilizes long labeling pulses (120’ vs. 30’), applies damage after rather than during labeling, and must involve an extra step that converts lesions or their repair intermediates into double strand breaks during preparation of DNA for analysis. Importantly, instead of revealing replication fork activity, the labeled tracks in this modified maRTA become “proxies” that are used to evaluate the presence and density of lesions in the whole genome. Moreover, by varying the time interval between labeling and DNA damage, we can address repair in newly replicated DNA versus repair in mature DNA in G2 or G1 cells.

Figure 1.

Figure 1

Open in a new tab

Design of a maRTA-based DNA repair assay. A) General approach. B) Comparison of the original maRTA technology to the modifications that enable analysis of DNA repair. C) A boxplot of results of in silico “digest” of a representative experimental data set of 627 tracks (see details in the main text). The length distribution difference between 0 and 10% digestion samples is significant (p-value = 0.00045 by Wilcoxon test). D) Dose-dependent fragmentation of IdU tracks in XRCC1+/− mouse ear primary fibroblasts labeled with IdU for 100 min, treated with indicated doses of MMS for 10 min, and harvested for analysis.

To estimate sensitivity of lesion detection, we performed in silico “breakage” of tracks using a typical experimental dataset (X1, X2, …..Xn). For simplicity, tracks were allowed to have no more than one breakpoint. Thus, we modeled track breakage into two fragments by, first, multiplying a length value of each track by a random number between 0 and 1, which gave us one of the broken fragments (X1a=X1 × RAND(0:1)). Then, the length of the second fragment was calculated as X1b=X1-X1a. This “breakage” was performed with randomly selected 10, 20, 40, …100% of the values of the original (X1, X2, …..Xn) dataset, and track lengths distributions of the resulting datasets (X1a, X1b, X2a, X2b, ….Xna, Xnb) were plotted (Figure 1C). As can be seen, breakage of 10% or more tracks in a population with a median length of about 80Kb (20μm × 4Kb, where 4Kb is a stretch factor determined in [8]) detectably shifts the overall length distribution. Together with the estimates of lesion densities routinely achieved in the experimental setting (e.g. 1CPD per 10-60Kb at 30J/m2 [24, 25], or 6×105 7MeG lesions per 1mM MMS/1hr per genome [26]), this modeling makes detection of randomly distributed lesions in stretched DNA a feasible goal.

1.2. Detection of alkylated base lesions and BER

A number of years ago, Lundin et al [19] demonstrated that methylated base lesions produced by methyl-methanesulfonate (MMS) can be converted into DSBs in vitro upon incubation of genomic DNA during isolation at 56°C or above. As noted by the authors, the chemical mechanism responsible may be the spontaneous hydrolysis of the most common MMS lesions, 7-meG and 3-meA [27], into AP sites, and heat conversion of AP sites into single-strand breaks, SSBs [28-31]. Proximity of two SSBs on opposite strands of DNA can result in a DSB.

Notably, a heating step at 56°C is already a part of the DNA preparation protocol for maRTA. It is carried out when agarose-embedded cells are lysed and treated with proteinase K [8]. This prompted us to turn to MMS damage as a test case for lesion quantitation by maRTA. We were able to demonstrate dose-dependent fragmentation of MMS-treated IdU-labeled tracks (Figure 1D). The dramatic, larger than two-fold reduction of track lengths in MMS-treated DNA suggests that some tracks had more than one breakpoint.

We next wanted to demonstrate repair of MMS lesions by the base excision repair pathway (BER). The scaffold protein XRCC1 is critically involved in BER and single-strand break repair [32, 33]. To demonstrate XRCC1 contribution to repair, we used XRCC1 wild type and null heterozygote primary mouse ear fibroblasts derived from littermate animals. The mice are described in [18], where McNeil et al have shown that tissues of heterozygotes express almost exactly 50% of the wild type level of XRCC1, and their cells are more sensitive to MMS. We chose the cells partially deficient in XRCC1 because: i) homozygous XRCC1 deletion is embryonic lethal in mice, preventing generation of adult mouse fibroblasts, and ii) we were particularly interested to test if we can measure partial defects in DNA repair.

We performed MMS treatment and recovery experiments with mouse fibroblasts (Figure 2A). We could detect both drastic reduction of track lengths upon MMS challenge and subsequent rebound of these lengths over a 21-hr period. The tracks from MMS-treated cells never reached the lengths of tracks from untreated cells; nevertheless, wild type cells repaired track lengths faster than XRCC1 heterozygous cells (Figure 2B). These results can be visualized either as track length distributions (as in Figure 2B) or by discretizing them into representative categories, e.g. above and below a 30μm threshold (Figure 2C). The latter method has an advantage in that it can include tracks that are so long that they extend beyond a single visual field under the microscope and thus can only be scored as equal to or larger than a particular value. The difference in repair of MMS damage between XRCC1+/+ and +/− fibroblasts was also confirmed using a pair of cell lines derived from another litter (data published in [12]).

Figure 2.

Figure 2

Open in a new tab

Differential repair of MMS-induced damage in XRCC1 wild type and null heterozygote mouse ear primary fibroblasts: comparison of the maRTA-based assay to other methods. A) Experimental design. Cells were labeled with IdU for 100 min, challenged with 0.02% MMS for 10 min and allowed to recover for up to 21 hr. Untreated cells were harvested after IdU labeling. Gray arrows indicate sample harvest time points. B) A boxplot of track length distributions in MMS-treated and control XRCC1 +/+ and +/− mouse ear primary fibroblasts. 260-640 tracks were measured per time point. Outlier data points (~ 10 values per sample) are not included in the plot. C) A bar graph recasting the data shown in (B) as percentages of tracks equal to or larger than 30μm. D) A Pulse-Field Gel Electrophoresis (PFGE) image of samples generated from XRCC1 +/+ and +/− mouse ear primary fibroblasts in a separate MMS challenge and recovery experiment. E) Alkaline comet assay of XRCC1 +/+ and +/− mouse ear primary fibroblasts samples generated in an MMS challenge and recovery experiment. Left and right panels plot different parameters calculated from the same experiment. Bar edges mark values of 1st and 3rd quartiles of data distributions, and notches are medians. Each bar represents measurements taken on 20 to 125 nuclei.

We compared our maRTA-based method to other ways of measuring MMS damage, i.e. Pulse-field Gel Electrophoresis (PFGE) and comet assay. Though we monitored total rather than analog-labeled DNA in these assays, we nevertheless labeled cells with IdU or CldU to maintain conditions identical to those used for maRTA. MMS-induced DSBs were readily detectable in total genomic DNA by PFGE as previously shown [19] (Figure 2D). Moreover, PFGE was in agreement with maRTA in revealing that damaged DNA species persisted in total DNA for longer in XRCC1 heterozygous cells than in wild type cells. It is clear, however, that PFGE output is less amenable to quantitation than maRTA.

Neutral comet assays were unable to detect MMS damage (data not shown), which was expected since the step of heating of MMS-treated DNA could not be incorporated into the procedure. On the other hand, alkaline comet assays [34] readily detected MMS damage and its repair with overall kinetics similar to that observed by maRTA, whether comets were scored by % DNA in the tail or by tail moment (Figure 2E). However, there was no difference in the efficiency of repair between XRCC1 wild type and heterozygote cells as measured by the alkaline comet assay. This may suggest that the latter assay is not sensitive enough.

Alternatively, repair of the subset of MMS damage that is convertible to DSBs, for example methylated base lesions clustered opposite each other on complementary DNA strands, may be more dependent on XRCC1 dose compared to the more abundant single methyl lesions, but are diluted in the overall alkaline comet signal.

1.3. Optimization of maRTA-based lesion quantitation for robust sample-to-sample comparisons

Random shearing of DNA incurred during stretching can, in theory, influence track length readout, and we wanted to establish internal controls that would serve as reference DNA for each stretched sample. One efficient way to do this is to use different labels for treated and untreated cells (Figure 3A). Upon isolation, differentially labeled DNAs can be mixed and stretched on one glass cover slip (Figure 3B). One and the same untreated control DNA can be added to every treated sample. Of note, this approach can be modified to compare two different cell lines (e.g. wild type and mutant) instead of treated and untreated cells. Also, one can mix differentially-labeled cells prior to DNA isolation, e.g. at agarose embedding stage.

Figure 3.

Figure 3

Open in a new tab

Quantitative comparisons of maRTA-based DNA repair data using internally controlled samples. A) Approach. Color-coding of treated and untreated (control) genomic DNAs by different thymidine analog labels allows to stretch and measure these DNAs on one cover slip. Wild type mouse ear fibroblasts were maintained at 5% O2 or switched to 20% O2 for 24 hrs prior to and throughout the experiment. Cells were labeled for 2 hrs, and IdU-labeled cells were treated with 0.02% MMS for 10 min and harvested either immediately thereafter or in 20.5 hrs. CldU-labeled samples were harvested 20.5 hrs after labeling. B) An example of visualization of replication tracks in a mixture of CldU-labeled (control) or IdU-labeled (MMS-treated) DNAs. C) A guide to mixed samples. D) Boxplots of track length distributions in mixed samples. 137-195 each of IdU and CldU tracks were measured per sample. White, IdU, gray, CldU. E) Length values of equal numbers of IdU and CldU tracks in each sample were summed, and length sums of IdU tracks were normalized to length sums of control tracks. F) Left panel: Length values of equal numbers of IdU and CldU tracks in each 20% O2 sample were rank-ordered smallest to largest, and values of the same rank were plotted as X and Y coordinates of bivariate distributions of data points on a scatterplot. Two of these bivariate distributions (with and without recovery after MMS) are shown to illustrate repair of MMS damage. Right panel: all four bivariate distributions (at 0 or 20.5 hrs recovery, and at 5 or 20% O2) are plotted on a scatterplot to demonstrate that the 5%O2 and 20% O2 results are superimposable, thus suggesting identical efficiency of repair of MMS damage under these two growth conditions.

In Figure 3C we applied the DNA mixing approach to measure recovery from MMS damage in wild type mouse ear fibroblasts grown at high or low oxygen concentrations. Oxygen level profoundly affects proliferation of primary mouse cells [35], and according to some estimates [36], spontaneous base damage due to oxidation can account for about one 8-oxoG per 1Mb (8-oxoG is most common spontaneous oxidative DNA damage). We tested a possibility that, as a BER substrate, increased load of oxidized base damage may interfere with the response to methylation damage by MMS. Mouse ear fibroblasts were continually grown at 5% O2 or switched to 20% O2 for 24 hrs prior to and during MMS challenge and recovery. For each growth condition, IdU-labeled cells were treated with MMS and CldU-labeled cells were not treated. DNAs from these cells were combined and stretched together (Figure 3A-C).

As can be seen in Figure 3D, repair of MMS damage appears not to be affected by oxygen level in growth media; however, measurements of track length distributions in untreated reference DNAs vary somewhat from mix to mix, potentially obscuring the result. For example, is there a difference between mix 1 and mix 3 in MMS-induced fragmentation? We therefore sought ways to normalize treated values to untreated control values in each mix in order to more precisely evaluate and visualize biological differences between these samples. One approach is to compare ratios of treated means to control means. Since track length distributions are nonparametric, one alternative is to use, instead of means, track length sums derived from equal numbers of treated and control tracks randomly selected from each data set. Ratios of treated/control sums is a reasonable indicator of the degree of repair, and it confirms a similar efficiency of repair regardless of oxygen tension during growth (Figure 3E).

In order to retain and better represent heterogeneity of track length distributions, we suggest another approach to data quantitation. Equal numbers of treated and control track data randomly selected from each data set are rank-ordered by value (smallest to largest). For each mix, treated and control values of the same rank are then represented as X and Y coordinates of a bivariate data distribution on a scatterplot (Figure 3F). If treated and control values are distributed similarly, the resulting bivariate distribution will fall on a diagonal. The greater the difference between treated and control, the further from diagonal is the distribution (compare MMS-treated samples harvested immediately and after 20.5 hrs of recovery in Figure 3F, left graph). Superimposition of data points of samples derived from cells grown in high or low oxygen (Figure 3F, right graph) visually demonstrates that the initial level of MMS damage and efficiency of its repair does not depend on oxygen level.

1.4. Alkylated base lesion repair in WRN-depleted cells

Several lines of evidence have implicated WRN RECQ helicase in BER. WRN functionally and physically interacts with several BER proteins [37, 38], modulates BER activities in in vitro assays [39] and in vivo assays with transfected oligonucleotides as reporters [40]. WRN-deficient cells are MMS-sensitive [40]. However, since MMS lesions may also affect DNA replication [14], it remains unclear if WRN significantly modulates bona fide BER in genomic DNA in vivo [41]. We were interested to assess MMS damage repair in human cells depleted of WRN, and in particular, in DNA that is not replicating or newly replicated.

We used our well-characterized system of shRNA-mediated depletion of WRN in SV40-transformed human fibroblast line GM639cc1 [14, 15] (Figure 4A), and employed modifications of the maRTA assay and analysis that we described in the previous section. We also introduced a 7 hr interval between labeling of cells and treating them with MMS (Figure 4B). From our previous flow cytometric studies of these cells, this time is sufficient for the labeled cell subpopulation to progress to G2 [14]. While isolating DNA for maRTA, we found that unlike in mouse cells, incubation at 56°C during cell lysis was not sufficient to convert MMS damage to DSBs (this was true both for human fibroblasts and lymphoblasts, data not shown). We thus introduced an additional step in the procedure whereby DNAs released from agarose were incubated at 56°C for 4 hrs. After that step, each MMS-treated DNA was mixed with untreated control DNA and stretched (Figure 4C).

Figure 4.

Figure 4

Open in a new tab

Differential repair of MMS damage in WRN-depleted SV40-transformed fibroblasts. A) A western blot showing shRNA-mediated depletion of WRN in GM639cc1 fibroblasts. shRNA is expressed from an integrated lentiviral vector. CHK1 serves as an internal control. Vector stands for cells transduced with vector w/o shRNA. B) A diagram of thymidine analog labeling, MMS treatment, and recovery designed to apply MMS damage to predominantly G2/M cells. C) A boxplot of track length distributions in mixtures of treated and control DNAs. 211-870 of IdU and CldU tracks each were measured per sample. P values were determined in Wilcoxon tests. The most relevant p values are shown above the plot. D) Scatterplots of bivariate distributions derived as in Fig.3F. E) A bar graph of normalized track length sum values derived as in Fig.3E. F) A diagram of labeling, treatment, and recovery designed to apply damage to predominantly S phase cells. G) Scatterplots of bivariate distributions derived from the experiment outlined in (F).

We found that WRN-depleted cells showed less repair of MMS damage at 14 hrs after MMS exposure (Figure 4C). The difference, albeit modest, could nevertheless be visualized using the track length sum approach (Figure 4E), or by plotting ranked MMS-treated and control track length values in each DNA mix as bivariate distributions of data points on a scatterplot (Figure 4D). In another experiment we added MMS immediately after labeling (i.e. labeled cells were damaged when still in S phase) and monitored extent of repair at 22 hrs after exposure to MMS (Figure 4F). Again, WRN-depleted cells were less efficient than control cells. However, in this case there was virtually no repair in WRN-depleted cells at 22 hrs after MMS compared to 0 hrs after MMS (Figure 4G), suggesting that repair of MMS damage received during S phase may be more challenging for WRN-depleted cells. Taken together, these results are consistent with the notion that WRN contributes to the repair of MMS damage in genomic DNA.

1.5. Applicability of maRTA-based assay to detection of UV damage

We next wanted to extend maRTA-based lesion detection to lesions other than alkylated bases. Certain lesions can be converted into overt DSBs or breakage-susceptible intermediates upon limited enzymatic digestion of genomic DNA (Figure 5A). This principle is already utilized in the comet assay of UV lesions [42], and we wanted to determine if maRTA can be used in a similar manner. Cyclobutane pyrimidine dimers (CPDs), the most common DNA lesion inflicted by UV, are recognized by T4 pyrimidine dimer glycosylase (PDG), which cleaves the glycosyl bond of the 5′ end of a CPD and the phosphodiester bond at the abasic site, thereby producing a single-stranded break (SSB) in the DNA. This SSB may be sufficient to stimulate breakage of DNA upon stretching, or, alternatively, can be converted into a DSB upon additional enzymatic treatment.

Figure 5.

Figure 5

Open in a new tab

Detection of CPD dimers in UV-irradiated, stretched DNA. A) Approach. Lesions of single-strand breaks in DNA can be visualized by enzymatic conversion to double-strand breaks. B) Labeling, treatment, and a boxplot of results obtained with normal human fibroblast line Hi3. PDG was used to digest genomic DNAs prior to stretching. 325 to 635 tracks were measured per sample. C) Samples from the experiment introduced in (B) were treated with PDG (or buffer) and then with S1 nuclease prior to stretching. Number of tracks measured per sample is 314 to 370. D) Labeling, treatment, and a boxplot of results obtained with SV40-transformed fibroblast line GM639cc1. Samples were incubated with buffer alone, or with PDG, then with indicated amounts of S1 nuclease. Number of tracks measured per sample is 370 to 1100. P values shown in (B-D) were determined in Wilcoxon tests.

We irradiated ethynyl-deoxyuridine (EdU)-labeled primary human fibroblasts with UV and harvested cells immediately after treatment (Figure 5B). The protocol for detection of EdU in stretched DNA fibers was introduced by us previously [11] and is described in detail in Materials and Methods. In the first series of experiments, genomic DNA was incubated with buffer only or with different amounts of PDG prior to stretching. While PDG treatment reduced the lengths of tracks, the effect was not dramatic or exclusive to DNA from UV-irradiated cells despite a very high UV dose used (Figure 5B). We reasoned that SSBs introduced by PDG may not be sufficient to induce breakage of this DNA upon stretching. We thus decided to convert SSBs to DSBs by incubating PDG-treated DNA with single-stranded (ss) DNA-specific S1 nuclease. Digested genomic DNA was first resolved in regular agarose gels to estimate the appropriate range of enzyme concentrations (data not shown). We found that the PDG-S1 protocol revealed a more definitive and significant shortening of tracks in UV lesion-containing but not in control DNA (Figure 5C).

Tracks in UV lesion-containing DNA also proved sensitive to S1 nuclease alone if it was added at a higher concentration. In this experiment, cells were harvested 2 hrs after UV to increase the likelihood that some UV lesions (CPDs or other adducts) would be processed into intermediates that had ssDNA and thus be S1 nuclease-sensitive (Figure 5D). Overall, these results indicate that UV lesions can be readily detected by maRTA combined with enzymatic digestions.

1.6. The use of S1 nuclease digestion to assay crosslink/bulky adduct repair by maRTA

Availability of highly specific antibodies against UV lesions (both CPDs and 6-4-photoproducts) offers a choice of more than one repair assay, including a DNA dot blot-based assay or already mentioned alkaline comet assays. Fewer options are available to measure repair of lesions associated with such agents as mitomycin C (MMC), which can generate a mixture of inter- and intrastrand crosslinks as well as monoadducts [43]. Assaying crosslink repair in particular, is highly relevant to basic and translational studies of such tumor suppressors as BRCA1, 2 and Fanconi Anemia proteins. RECQ helicases WRN, BLM, and RECQ5 are also implicated in crosslink repair [44-47].

Since ssDNA intermediates are part of crosslink and monoadduct processing, we were interested to probe MMC damage repair by measuring track lengths in genomic DNA digested with S1 nuclease. To generate MMC damage, we used the regimen developed by us previously [11]. With this regimen, cell death is not evident until after 48 hrs post-treatment. In these experiments we mixed lesion-containing and lesion-free DNA (as described in sections 1.3, 1.4), except we used IdU and EdU as labels to distinguish these DNAs (Figure 6A). Labeled cells (Brca1 negative breast cancer cell line HCC1937 complemented with BRCA1) were harvested 2 hrs after the end of MMC treatment. DNA mixes were incubated with buffer only or with various concentrations of S1 nuclease prior to stretching. We could detect preferential fragmentation of tracks in MMC-treated DNA compared to untreated DNA in the same mix over an order of magnitude range of S1 concentrations (Figure 6B).

Figure 6.

Figure 6

Open in a new tab

Detection of mitomycin C (MMC) lesions in stretched DNA of the BRCA1− complemented breast cancer cell line HCC1937. A) Labeling and treatment. MMC was added at 25μM for 1 hr. DMSO is the vehicle. EdU-labeled (control) and IdU-labeled (treated) DNAs were mixed together prior to enzymatic digest. B) A boxplot of results of S1 nuclease digestion of DNA mixtures. DNAs were isolated from cells harvested 2 hrs after MMC treatment. Mixtures were incubated with indicated amounts of S1 nuclease and stretched. Between 230 and 390 each of IdU and EdU tracks were measured in each mixture. C) In a separate experiment, HCC1937/BRCA1 cells were labeled and treated as in (A), except EdU-labeled DNA was control and IdU-labeled DNA was treated with MMC, and cells were harvested 3 and 13 hrs. after MMC (or DMSO) treatment. DNAs were mixed and incubated with buffer only, or with an indicated amount of S1 nuclease prior to stretching. Between 209 and 369 each of IdU and EdU tracks were measured in each mixture. P values were determined in Wilcoxon tests.

To demonstrate repair of MMC damage, HCC1937/BRCA1+ cells were harvested 3 and 13 hrs after MMC (or vehicle, DMSO) treatment, and lesion-containing and lesion-free DNAs were mixed as above and incubated with buffer or with 1/2 units of S1. Again, tracks in S1-digested MMC-treated DNA were significantly shorter than in control at 3hrs after MMC treatment (Figure 6C). By 13 hrs after drug exposure, lengths of these tracks increased; however, they remained shorter than those in the DMSO control. This suggests slow repair of MMC lesions, consistent with previous estimates by a modified comet assay [48].

Figure 6C highlights two additional observations. First, lesion-free DNA is not completely resistant to S1 nuclease, which is expected given that S1 nuclease targets any ssDNA, and stretches of ssDNA may be present in the genome, particularly in newly-replicated DNA [49]. In agreement with this, in DNAs isolated between 0 and 2hrs after labeling, tracks exhibit elevated susceptibility to fragmentation by S1 nuclease regardless of lesion presence or cell source (i.e. cancer-derived, transformed, or primary fibroblasts, data not shown). Second, while track length distributions of DMSO- and MMC-treated DNAs without S1 were similar at 3 hrs, they diverged by 13 hrs. This potentially indicates differential maturation and/or processing of lesion-free and lesion-containing DNA in vivo in HCC1937/BRCA1+ cells. Overall, these studies suggest that S1 digestion of MMC lesion-containing genomic DNA combined with maRTA can be used as an assay for lesion presence and repair. However, the presence in the genome of biologically relevant single-stranded DNA that is not lesion-specific is a confounder that should be taken into account in experimental designs.

1.7. WRN-depleted cells show delayed repair of MMC lesions

As mentioned above, WRN is implicated in repair of MMC lesions [44-46, 50]. WRN is thought to participate in resection of DNA strands for homologous recombination-based repair of damage induced by MMC [51]. Using non-specific shRNA and WRN shRNA-expressing MCF10a cells, we performed a MMC challenge and recovery time course, followed by S1 nuclease digestion of DNAs (Figure 7). In this experiment, WRN level was reduced by half in WRN shRNA-expressing cells compared to controls (data not shown). As before, paired MMC-treated and untreated samples were digested with 1/4u of S1 (or mock-digested) and stretched together. In WRN-low cells, fragmentation of MMC-treated tracks persisted for longer (up to 18.5hrs) than in control cells, suggesting that repair of MMC damage is delayed. By 28 hrs post-MMC, MMC-treated DNA in WRN-low cells ceased to be hypersensitive to S1 (compared to untreated DNA), as it did in control cells (data not shown). Overall, these results confirm applicability of the modified maRTA to detecting crosslink damage in vivo in genomic DNA and measuring its repair, and are consistent with the published evidence that implicates WRN in repair of MMC damage.

Figure 7.

Figure 7

Open in a new tab

Differential repair of mitomycin C lesions in MCF10a mammary epithelial cells partially depleted of WRN. shSCR is scrambled shRNA control [11]. A) Labeling and treatment. MMC was added at 25μM for 1 hr. DMSO is vehicle. Cells were harvested 2 and 18.5 hrs after MMC. EdU-labeled (control) and IdU-labeled (treated) DNAs were mixed together and incubated with buffer only or with indicated amount of S1 nuclease prior to stretching. B) A boxplot of results. Between 115 and 350 each of IdU and EdU tracks were measured in each mixture.

CONCLUSION

In summary, we have demonstrated the utility of a modified version of maRTA as an assay to detect and measure repair of three different types of DNA damage: alkylated base damage, UV-induced lesions, and mitomycin C-induced crosslinks and monoadducts. We applied our maRTA-based assay to a wide range of cell lines and primary cells from mouse and human. Moreover, our assay appropriately detected delays in repair in a genotype- and protein expression-specific manner. We showed that repair of MMS damage is slower in mouse primary fibroblasts with one intact copy of the BER gene XRCC1, which express 50% of normal level of XRCC1 protein. We also detected delays of repair of MMS or mitomycin C damage in immortalized human fibroblasts and mammary epithelial cells with reduced levels of the WRN RECQ helicase.

In detection of MMS damage, our maRTA-based assay performed on par with other applicable techniques – PFGE and alkaline comet assays. We believe that one advantage of our approach over the latter two is its quantitative output, coupled with the ability to estimate lesion density. The following formula, Lesion density per Kbp = (1/Mtreated) − (1/Muntreated), where M is mean fragment length (in Kbp) of damage-treated or untreated DNA, was introduced as a reasonable approximation for estimating lesion density based on lengths of DNA fragments in gel electrophoresis [52]. According to this formula, density of heat-labile MMS lesions in our experiments is 2×105 per 1mM MMS/1hr per genome, which is very close to the calculated density of 6×105 7-meG lesions per 1mM MMS/1hr per genome [26], particularly if we take into account that only a fraction of MMS damage may be convertible to DSBs.

One advantageous feature specific to our maRTA-based DNA repair assay are its ability to measure repair of DNA damage that incurs in specific phases of the cell cycle. Labeling cells with thymidine analogs not only creates an assay substrate in the form of tracks in genomic DNA, but marks an S phase subset of a cell population, which can then be allowed to traverse to G2 or G1 before treatment simply by introducing a time gap between DNA labeling and DNA damage. The length of this gap can be readily determined for one’s favorite cell line by two-parameter flow cytometry, i.e. staining for DNA content and analog incorporation.

As with any cellular biology protocol involving thymidine analog incorporation into DNA, there is a possibility that this has influence on cell cycle, replication, and repair. This concern has always been balanced by the advantages of limited analog labeling such as ours, provided that all comparisons are made between the analog-labeled cells.

We believe that maRTA-based DNA repair assays can be particularly useful in functional phenotyping of SNP variants in important DNA repair genes, particularly those associated with increased cancer risk or differential treatment outcome, including XRCC1, BRCA1 and 2, and other genes. Future development of this assay can also include expanded use of specific enzymes to follow repair of particular lesions.

HIGHLIGHTS.

  • A protocol to measure DNA damage and repair in stretched fibers of genomic DNA.

  • Demonstration of detection of MMS, UV, and mitomycin C lesions.

  • Differential repair can be demonstrated to depend on XRCC1 or WRN expression.

ACKNOWLEDGEMENTS

We would like to thank John Morton for assistance with deriving mouse ear fibroblasts. We are grateful to Dr. Elizabeth Swisher and Dr. Toshi Taniguchi for gifts of cell lines, and to Dr. Ray Monnat Jr. for support and helpful discussions. This work was supported by Nathan Shock Center of Excellence in the Basic Biology of Aging (University of Washington) and R21ES019485 (NCI/NIEHS) grants to J.S., and by P01CA77852 (NCI) grant to Dr. Monnat. Funding bodies had no involvement in the design of this study or data collection or analysis.

ABBREVIATIONS

BER

base excision repair

BrdU

bromo-deoxyuridine

CldU

chloro-deoxyuridine

CPD

cyclopyrimidine dimer

DSB

double-strand break

EdU

ethynyl-deoxyuridine

IdU

iodo-deoxyuridine

maRTA

microfluidic assisted replication track analysis

MMC

mitomycin C

MMS

methyl-methane sulfonate

SSB

single-strand break

UV

ultraviolet

XRCC1

X-Ray Repair Cross-Complementing Protein 1

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Authors declare no conflict of interest associated with this work.

REFERENCES

  • [1].Suhasini AN, Brosh RM., Jr. DNA helicases associated with genetic instability, cancer, and aging. Adv Exp Med Biol. 2013;767:123–144. doi: 10.1007/978-1-4614-5037-5_6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [2].Brosh RM., Jr DNA helicases involved in DNA repair and their roles in cancer. Nat Rev Cancer. 2013;13:542–558. doi: 10.1038/nrc3560. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [3].Jackson DA, Pombo A. Replicon clusters are stable units of chromosome structure: evidence that nuclear organization contributes to the efficient activation and propagation of S phase in human cells. Journal of Cell Biology. 1998;140:1285–1295. doi: 10.1083/jcb.140.6.1285. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [4].Bensimon A, Simon A, Chiffaudel A, Croquette V, Heslot F, Bensimon D. Alignment and sensitive detection of DNA by a moving interface. Science. 1994;265:2096–2098. doi: 10.1126/science.7522347. [DOI] [PubMed] [Google Scholar]
  • [5].Yokota H, Johnson F, Lu H, Robinson RM, Belu AM, Garrison MD, Ratner BD, Trask BJ, Miller DL. A new method for straightening DNA molecules for optical restriction mapping. Nucl. Acids Res. 1997;25:1064–1070. doi: 10.1093/nar/25.5.1064. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [6].Herrick J, Bensimon A. Single molecule analysis of DNA replication. Biochimie. 1999;81:859–871. doi: 10.1016/s0300-9084(99)00210-2. [DOI] [PubMed] [Google Scholar]
  • [7].Norio P, Schildkraut CL. Visualization of DNA Replication on Individual Epstein-Barr Virus Episomes. Science. 2001;294:2361–2364. doi: 10.1126/science.1064603. [DOI] [PubMed] [Google Scholar]
  • [8].Sidorova JM, Li N, Schwartz DC, Folch A, Monnat RJ., Jr. Microfluidic-assisted analysis of replicating DNA molecules. Nature protocols. 2009;4:849–861. doi: 10.1038/nprot.2009.54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [9].Schwab RA, Niedzwiedz W. Visualization of DNA replication in the vertebrate model system DT40 using the DNA fiber technique. Journal of visualized experiments : JoVE. 2011:e3255. doi: 10.3791/3255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [10].Yardimci H, Loveland AB, van Oijen AM, Walter JC. Single-molecule analysis of DNA replication in Xenopus egg extracts. Methods. 2012;57:179–186. doi: 10.1016/j.ymeth.2012.03.033. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [11].Kehrli K, Sidorova JM. Mitomycin C reduces abundance of replication forks but not rates of fork progression in primary and transformed human cells. Oncoscience. 2014;1:540–555. doi: 10.18632/oncoscience.70. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [12].Pettan-Brewer C, Morton J, Cullen S, Enns L, Kehrli KR, Sidorova J, Goh J, Coil R, Ladiges WC. Tumor growth is suppressed in mice expressing a truncated XRCC1 protein. Am J Cancer Res. 2012;2:168–177. [PMC free article] [PubMed] [Google Scholar]
  • [13].Swanson C, Saintigny Y, Emond MJ, Monnat J, Raymond J. The Werner syndrome protein has separable recombination and survival functions. DNA Repair. 2004;3:475–482. doi: 10.1016/j.dnarep.2004.01.002. [DOI] [PubMed] [Google Scholar]
  • [14].Sidorova JM, Li N, Folch A, Monnat RJ., Jr. The RecQ helicase WRN is required for normal replication fork progression after DNA damage or replication fork arrest. Cell Cycle. 2008;7:796–807. doi: 10.4161/cc.7.6.5566. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [15].Mao FJ, Sidorova JM, Lauper JM, Emond MJ, Monnat RJ. The Human WRN and BLM RecQ Helicases Differentially Regulate Cell Proliferation and Survival after Chemotherapeutic DNA Damage. Cancer Research. 2010;70:6548–6555. doi: 10.1158/0008-5472.CAN-10-0475. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [16].Sidorova JM, Kehrli K, Mao F, Monnat R., Jr Distinct functions of human RECQ helicases WRN and BLM in replication fork recovery and progression after hydroxyurea-induced stalling. DNA Repair. 2013;12:128–139. doi: 10.1016/j.dnarep.2012.11.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [17].Dhillon KK, Sidorova JM, Albertson TM, Anderson JB, Ladiges WC, Rabinovitch PS, Preston BD, Monnat RJ., Jr. Divergent cellular phenotypes of human and mouse cells lacking the Werner syndrome RecQ helicase. DNA Repair (Amst) 2009;9:11–22. doi: 10.1016/j.dnarep.2009.09.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [18].McNeill DR, Lin PC, Miller MG, Pistell PJ, de Souza-Pinto NC, Fishbein KW, Spencer RG, Liu Y, Pettan-Brewer C, Ladiges WC, Wilson DM., 3rd XRCC1 haploinsufficiency in mice has little effect on aging, but adversely modifies exposure-dependent susceptibility. Nucleic Acids Res. 2011;39:7992–8004. doi: 10.1093/nar/gkr280. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [19].Lundin C, North M, Erixon K, Walters K, Jenssen D, Goldman ASH, Helleday T. Methyl methanesulfonate (MMS) produces heat-labile DNA damage but no detectable in vivo DNA double-strand breaks. Nucl. Acids Res. 2005;33:3799–3811. doi: 10.1093/nar/gki681. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [20].Herschleb J, Ananiev G, Schwartz DC. Pulsed-field gel electrophoresis. Nat. Protocols. 2007;2:677–684. doi: 10.1038/nprot.2007.94. [DOI] [PubMed] [Google Scholar]
  • [21].Kaufman DG, Cohen SM, Frum RA, Chastain PD. Temporal and Functional Analysis of DNA Replicated in Early S Phase. Advances in enzyme regulation. 2011;51:257–271. doi: 10.1016/j.advenzreg.2010.11.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [22].Huang J, Liu S, Bellani MA, Thazhathveetil AK, Ling C, de Winter JP, Wang Y, Wang W, Seidman MM. The DNA translocase FANCM/MHF promotes replication traverse of DNA interstrand crosslinks. Mol Cell. 2013;52:434–446. doi: 10.1016/j.molcel.2013.09.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [23].Kidane D, Murphy DL, Sweasy JB. Accumulation of abasic sites induces genomic instability in normal human gastric epithelial cells during Helicobacter pylori infection. Oncogenesis. 2014;3:e128. doi: 10.1038/oncsis.2014.42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [24].Perdiz D, Grof P, Mezzina M, Nikaido O, Moustacchi E, Sage E. Distribution and repair of bipyrimidine photoproducts in solar UV-irradiated mammalian cells. Possible role of Dewar photoproducts in solar mutagenesis. The Journal of biological chemistry. 2000;275:26732–26742. doi: 10.1074/jbc.M001450200. [DOI] [PubMed] [Google Scholar]
  • [25].Douki T, Reynaud-Angelin A, Cadet J, Sage E. Bipyrimidine photoproducts rather than oxidative lesions are the main type of DNA damage involved in the genotoxic effect of solar UVA radiation. Biochemistry. 2003;42:9221–9226. doi: 10.1021/bi034593c. [DOI] [PubMed] [Google Scholar]
  • [26].Horton JK, Joyce-Gray DF, Pachkowski BF, Swenberg JA, Wilson SH. Hypersensitivity of DNA polymerase beta null mouse fibroblasts reflects accumulation of cytotoxic repair intermediates from site-specific alkyl DNA lesions. DNA Repair (Amst) 2003;2:27–48. doi: 10.1016/s1568-7864(02)00184-2. [DOI] [PubMed] [Google Scholar]
  • [27].Beranek DT. Distribution of methyl and ethyl adducts following alkylation with monofunctional alkylating agents. Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis. 1990;231:11–30. doi: 10.1016/0027-5107(90)90173-2. [DOI] [PubMed] [Google Scholar]
  • [28].Brookes P, Lawley PD. EVIDENCE FOR THE ACTION OF ALKYLATING AGENTS ON DEOXYRIBONUCLEIC ACID. Experimental cell research. 1963;24(Suppl9):521–524. doi: 10.1016/0014-4827(63)90292-1. [DOI] [PubMed] [Google Scholar]
  • [29].Lawley PD, Orr DJ. Specific excision of methylation products from DNA of Escherichia coli treated with N-methyl-N′-nitro-N-nitrosoguanidine. Chem Biol Interact. 1970;2:154–157. doi: 10.1016/0009-2797(70)90047-5. [DOI] [PubMed] [Google Scholar]
  • [30].Lindahl T, Andersson A. Rate of chain breakage at apurinic sites in double-stranded deoxyribonucleic acid. Biochemistry. 1972;11:3618–3623. doi: 10.1021/bi00769a019. [DOI] [PubMed] [Google Scholar]
  • [31].Singer B, Grunberger D. Molecular Biology of Mutagens and Carcinogens. Springer US; New York, NY, USA: 1983. [Google Scholar]
  • [32].Horton JK, Watson M, Stefanick DF, Shaughnessy DT, Taylor JA, Wilson SH. XRCC1 and DNA polymerase [beta] in cellular protection against cytotoxic DNA single-strand breaks. Cell Research. 2008;18:48–63. doi: 10.1038/cr.2008.7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [33].London RE. The structural basis of XRCC1-mediated DNA repair. DNA Repair (Amst) 2015;30:90–103. doi: 10.1016/j.dnarep.2015.02.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [34].Collins AR, Dusinska M, Horska A. Detection of alkylation damage in human lymphocyte DNA with the comet assay. Acta biochimica Polonica. 2001;48:611–614. [PubMed] [Google Scholar]
  • [35].Parrinello S, Samper E, Krtolica A, Goldstein J, Melov S, Campisi J. Oxygen sensitivity severely limits the replicative lifespan of murine fibroblasts. 2003;5:741–747. doi: 10.1038/ncb1024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [36].Ohno M, Miura T, Furuichi M, Tominaga Y, Tsuchimoto D, Sakumi K, Nakabeppu Y. A genome-wide distribution of 8-oxoguanine correlates with the preferred regions for recombination and single nucleotide polymorphism in the human genome. Genome Research. 2006;16:567–575. doi: 10.1101/gr.4769606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [37].Das A, Boldogh I, Lee JW, Harrigan JA, Hegde ML, Piotrowski J, de Souza Pinto N, Ramos W, Greenberg MM, Hazra TK, Mitra S, Bohr VA. The Human Werner Syndrome Protein Stimulates Repair of Oxidative DNA Base Damage by the DNA Glycosylase NEIL1. Journal of Biological Chemistry. 2007;282:26591–26602. doi: 10.1074/jbc.M703343200. [DOI] [PubMed] [Google Scholar]
  • [38].Kanagaraj R, Parasuraman P, Mihaljevic B, van Loon B, Burdova K, Konig C, Furrer A, Bohr VA, Hubscher U, Janscak P. Involvement of Werner syndrome protein in MUTYH-mediated repair of oxidative DNA damage. Nucleic Acids Res. 2012;40:8449–8459. doi: 10.1093/nar/gks648. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [39].Ahn B, Harrigan JA, Indig FE, Wilson DM, Bohr VA. Regulation of WRN Helicase Activity in Human Base Excision Repair. Journal of Biological Chemistry. 2004;279:53465–53474. doi: 10.1074/jbc.M409624200. [DOI] [PubMed] [Google Scholar]
  • [40].Harrigan JA, Wilson DM, 3rd, Prasad, Opresko PL, Beck G, May A, Wilson SH, Bohr VA. The Werner syndrome protein operates in base excision repair and cooperates with DNA polymerase beta. Nucleic Acids Res. 2006;34:745–754. doi: 10.1093/nar/gkj475. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [41].Kuper J, Kisker C. DNA Helicases in NER, BER, and MMR. Adv Exp Med Biol. 2013;767:203–224. doi: 10.1007/978-1-4614-5037-5_10. [DOI] [PubMed] [Google Scholar]
  • [42].Gedik CM, Ewen SW, Collins AR. Single-cell gel electrophoresis applied to the analysis of UV-C damage and its repair in human cells. International journal of radiation biology. 1992;62:313–320. doi: 10.1080/09553009214552161. [DOI] [PubMed] [Google Scholar]
  • [43].Tomasz M. Mitomycin C: small, fast and deadly (but very selective) Chemistry & biology. 2:575–579. doi: 10.1016/1074-5521(95)90120-5. [DOI] [PubMed] [Google Scholar]
  • [44].Cheng WH, Kusumoto R, Opresko PL, Sui X, Huang S, Nicolette ML, Paull TT, Campisi J, Seidman M, Bohr VA. Collaboration of Werner syndrome protein and BRCA1 in cellular responses to DNA interstrand cross-links. Nucleic Acids Res. 2006;34:2751–2760. doi: 10.1093/nar/gkl362. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [45].Aggarwal M, Banerjee T, Sommers JA, Iannascoli C, Pichierri P, Shoemaker RH, Brosh RM. Werner syndrome helicase has a critical role in DNA damage responses in the absence of a functional Fanconi Anemia pathway. Cancer Research. 2013 doi: 10.1158/0008-5472.CAN-12-2975. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [46].Samanta S, Karmakar P. Recruitment of HRDC domain of WRN and BLM to the sites of DNA damage induced by mitomycin C and methyl methanesulfonate. Cell biology international. 2012;36:873–881. doi: 10.1042/CBI20110510. [DOI] [PubMed] [Google Scholar]
  • [47].Ramamoorthy M, May A, Tadokoro T, Popuri V, Seidman MM, Croteau DL, Bohr VA. The RecQ helicase RECQL5 participates in psoralen-induced interstrand cross-link repair. Carcinogenesis. 2013;34:2218–2230. doi: 10.1093/carcin/bgt183. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [48].McKenna DJ, Gallus M, McKeown SR, Downes CS, McKelvey-Martin VJ. Modification of the alkaline Comet assay to allow simultaneous evaluation of mitomycin C-induced DNA cross-link damage and repair of specific DNA sequences in RT4 cells. DNA Repair (Amst) 2003;2:879–890. doi: 10.1016/s1568-7864(03)00086-7. [DOI] [PubMed] [Google Scholar]
  • [49].McGlynn AP, Wasson G, O’Connor J, McKerr G, McKelvey-Martin VJ, Downes CS. The bromodeoxyuridine comet assay: detection of maturation of recently replicated DNA in individual cells. Cancer Res. 1999;59:5912–5916. [PubMed] [Google Scholar]
  • [50].Cheng W-H, von Kobbe C, Opresko PL, Arthur LM, Komatsu K, Seidman MM, Carney JP, Bohr VA. Linkage between Werner Syndrome Protein and the Mre11 Complex via Nbs1. J. Biol. Chem. 2004;279:21169–21176. doi: 10.1074/jbc.M312770200. [DOI] [PubMed] [Google Scholar]
  • [51].Wang Anderson T., Kim T, Wagner John E., Conti Brooke A., Lach Francis P., Huang Athena L., Molina H, Sanborn Erica M., Zierhut H, Cornes Belinda K., Abhyankar A, Sougnez C, Gabriel Stacey B., Auerbach Arleen D., Kowalczykowski Stephen C., Smogorzewska A. A Dominant Mutation in Human RAD51 Reveals Its Function in DNA Interstrand Crosslink Repair Independent of Homologous Recombination. Molecular Cell. 2015;59:478–490. doi: 10.1016/j.molcel.2015.07.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [52].Sutherland BM, Shih AG. Quantitation of pyrimidine dimer contents of nonradioactive deoxyribonucleic acid by electrophoresis in alkaline agarose gels. Biochemistry. 1983;22:745–749. doi: 10.1021/bi00273a006. [DOI] [PubMed] [Google Scholar]

RESOURCES