Simultaneous Quantification of Mitochondrial DNA Damage and Copy Number in Circulating Blood: A Sensitive Approach to Systemic Oxidative Stress

Systemic oxidative stress is associated with a wide range of pathological conditions. Oxidative DNA damage is frequently measured in circulating lymphocytes. Mitochondrial DNA (mtDNA) is known to be more sensitive to oxidative damage than nuclear DNA but is rarely used for direct measurement of DNA damage in clinical studies. Based on the supercoiling-sensitive real-time PCR method, we propose a new approach for the noninvasive monitoring of systemic oxidative stress by quantifying the mtDNA structural damage and copy number change in isolated lymphocytes in a single test. We show that lymphocytes have significantly less mtDNA content and relatively lower baseline levels of damage than cancer cell lines. In an ex vivo challenge experiment, we demonstrate, for the first time, that exogenous H2O2 induces a significant increase in mtDNA damage in lymphocytes from healthy individuals, but no repair activity is observed after 1 h recovery. We further demonstrate that whole blood may serve as a convenient alternative to the isolated lymphocytes in mtDNA analysis. Thus, the blood analysis with the multiple mtDNA end-points proposed in the current study may provide a simple and sensitive test to interrogate the nature and extent of systemic oxidative stress for a broad spectrum of clinical investigations.


Introduction
Oxidative stress is a state of physiological imbalance between oxidant production and antioxidant defence at different biological levels. It is implicated in the development of many pathological conditions such as aging, neurodegenerative diseases, and cancer initiation and progression [1][2][3][4][5][6]. Many diseases are suspected to be linked to oxidative stress, but procurement of disease tissues may be difficult due to its invasive nature and the scarcity of available tissues. However, researchers have mitigated this problem by using the systemic oxidative stress in peripheral tissues, such as circulating blood, as a noninvasive surrogate. Extrinsic factors such as in�ammation, nutrient imbalance, and hypoxic environment could affect inter and intracellular redox homeostasis, therefore altering systemic oxidative stress levels; new efforts are made to investigate the interactions between systemic oxidative stress and pathogenesis of many disease conditions [7][8][9][10][11][12][13]. For example, several recent studies suggest a correlation between increased systemic oxidative stress and prostate cancer risk and progression [14][15][16]. Similar results are reported in lung cancer [17], head and neck cancer [18], and other human cancers [19,20]. us, enhanced oxidative stress is not only a common property of the diseased cells, but may also be re�ected in the peripheral tissues.
Systemic oxidative stress has been analyzed in serum and blood cells using different biomarkers and assay systems. Genomic DNA in circulating lymphocytes is a widely used target in measuring different end-points of oxidative DNA damage, such as 8-oxoguanine (8-oxo-G) base lesions or DNA strand breaks detected with the comet assay [14][15][16][17][18]. e mitochondrial DNA (mtDNA) in lymphocytes is an attractive alternative target to determine systemic oxidative stress. MtDNA is a circular, multicopy cytoplasmic DNA, semiautonomously maintained in mitochondria. It is known to be more sensitive to oxidative damage than nuclear DNA [21][22][23] and has been increasingly used for evaluating systemic oxidative stress with various assays. Similarly to nuclear DNA, 8-oxo-G base lesions can be assessed in puri�ed mtDNA from lymphocytes [24]. Extracellular circulating mtDNA in serum is another marker recently used for evaluating genetic integrity and cancer risk. Elevated levels of free �oating mtDNA detected in the plasma or serum are found to be associated with poor prognoses for prostate and testicular germ cell cancers [25][26][27]. However, the source and nature of this circulating mtDNA are not fully elucidated. Oxidative stress can also affect the total mtDNA content in lymphocytes under various diseased conditions [28]. For example, signi�cant alterations in mtDNA content were detected in lymphocytes from patients with renal cell carcinoma, hyperlipidemia, and Huntington's disease when compared to control populations [19,29,30]. However, the relationship between different mtDNA end-points reported in lymphocytes is not clear and the direct measurement of mtDNA strand breaks in lymphocytes has not been reported. We previously developed a sensitive in vitro assay to quantify mtDNA structural damage induced by strand breaks, repair and copy number change in prostate cancer cell lines using a supercoiling-sensitive real-time PCR (ss-qPCR) [6,31]. We showed that oxidative damage can induce single-or doublestrand breaks (SSB or DSB), which lead to the disruption of the supercoiled conformation, and that the resulting relaxed conformation is a better qPCR substrate for signi�cantly increased ampli�cation than the supercoiled conformation, even if the starting mtDNA molecules remain the same [31]. Additionally, we observed that prolonged exposure to 95 ∘ C heat also introduced strand breaks in the mtDNA. is particular property was advantageously used to disrupt all structural features of mtDNA for precise quanti�cation of the total mtDNA content [31].
e objectives of this current study were to test if the ss-qPCR method could be applied to the lymphocytes and to explore a quantitative strategy to measure multiple mtDNA end-points in circulating blood cells for the study of systemic oxidative stress. We developed an absolute quanti�cation method for precise measurement of mtDNA structural damage, copy number change, and repair activity in blood cells. We demonstrated that mtDNA has low levels of both copy number and baseline damage in lymphocytes as compared to cancer cell lines, and that exogenous H 2 O 2 led to a signi�cant increase in mtDNA damage but with little repair activity in inactivated lymphocytes in ex vivo experiments.

Blood Collection and Lymphocyte Preparation.
Healthy male volunteers ranging from 28 to 45 years old were recruited for this pilot study through an institutional review board (REB) approved protocol at the McGill University Health Center. Blood (10 to 15 mL) was collected into 9 mL collection tubes coated with EDTA (Vacu K3EDTA PULL LAV) (Fisher, Monroe, NC). For experiments with whole blood, the samples were immediately stored at −80 ∘ C in 10% dimethyl sulfoxide (DMSO) prior to analysis. For experiments with isolated lymphocytes, blood was submitted to Ficoll-Paque Plus (GE Healthcare, Buckinghamshire, England) to recover the lymphocytes [33], then stored at −80 ∘ C in 40% RPMI media 1640 supplemented with 50% FBS and 10% DMSO prior to analysis. As per manufacturer's speci�cations, the extracted sample is composed in majority of lymphocytes (75-93%), with a remaining fraction of monocytes (7-25%) and minimal contaminants from granulocytes, erythrocytes, and platelets (3 ± 2%, 5 ± 2%, and < 0.5%, resp.).

H 2 O 2 Challenge Experiments with Lymphocytes and
Whole Blood. Frozen lymphocytes were thawed in a 37 ∘ C water bath for 1-2 min and washed with 5 volumes of icecold wash medium (50% FBS and 50% RPMI 1640). e lymphocytes were counted and cell viability was assessed under microscope using the trypan blue dye (average of over 90% viability). A total of ∼3 × 10 6 lymphocytes were incubated in 50 mL conical tubes with RPMI-1640 complete medium for 30 min prior to the experiment. e cell suspension was split into three groups of ∼1 × 10 6 cells each, treated with 0 (control) or 120 M H 2 O 2 for 15 min for exposure or allowed to recover in fresh medium for 60 min. e concentration of 120 M H 2 O 2 was chosen to be in the lower-middle range of concentrations used in similar treatments in the literature (50 to 500 M) [14,34]. Aerwards, the lymphocyte samples were washed with PBS, spun down to a pellet, and then stored at −80 ∘ C before DNA preparation.
Frozen whole blood was thawed in a 37 ∘ C water bath for 1-2 min and washed with 5 volumes of ice-cold PBS wash medium. e whole blood cells were counted with trypan blue dye prior to incubation (average of over 90% viability). A total of ∼15 × 10 6 whole blood cells were incubated in RPMI-1640 complete medium in 50 mL conical tubes for 30 min prior to the experiment. Whole blood samples were separated into three groups with ∼5 × 10 6 cells each and treated with 0 or 120 M H 2 O 2 as in the lymphocyte experiment. Whole blood samples were collected aer treatment and stored at −80 ∘ C before DNA preparation. Kit according to the manufacturer's instructions with minor modi�cations to ensure that both mtDNA and nuclear DNA were collected together [31,35]. Total DNA was quanti�ed with a NanoDrop spectrophotometer. DNA template solutions of 1 ng/ L were prepared for each sample with 1X Tris/EDTA Buffer Solution (pH 8.0). Each template solution was split into two equal parts with half serving as an original template for the measurement of the damaged/relaxed mtDNA fraction and the other half heat-treated (95 ∘ C for 6 min on a PCR machine) to quantify total amount of mtDNA [31].

Nuclear DNA and mtDNA Standards Preparation for
Absolute �uanti�cation. MtDNA standards were prepared for absolute quanti�cation. A 3.3 kb mtDNA fragment containing the CO2 gene and a 2.5 kb fragment containing the D-loop region were ampli�ed from the immortalized normal human prostate cell line, RWPE-1, using primers listed in Table 1. PCR reactions were performed using the GeneAmp PCR 9700 system (ABI) with recombinant ermus thermophilus (rTth) DNA polymerase (ABI). e ampli�cation program was performed as follows: preheat samples to reach 75 ∘ C; add rTth DNA polymerase and incubate for 2 min; denature at 94 ∘ C for 1 min, followed by 30 cycles of 94 ∘ C for 15 sec, 60 ∘ C for 30 sec., and 72 ∘ C for 3.5 min; then 72 ∘ C for 5 min and cool down to 10 ∘ C. Ampli�ed DNA fragments were puri�ed with the �IAGEN PCR Puri�cation Kit. e puri�ed products were carefully quanti�ed with the Nanodrop spectrophotometer, and the average of three readings was used for calculating precise copy number according to the following equation ( Figure 1): Six or seven serial dilutions were made ranging from 3 × 10 6 to 30 or 3 × 10 7 to 30 copies with a dilution factor of 10 depending on the experiment. e 6-point standard was used for the mtDNA quanti�cation in blood samples, while the 7-point standard was used to demonstrate the dynamic range and linearity of the assay. e original stock solutions were made into small aliquots and stored at −80 ∘ C to prevent repeated freeze and thaw. e nuclear DNA standards were similarly prepared. e nuclear primer sequences are listed in Table 1. A 2.7 kb nuclear fragment containing the calicin gene was ampli�ed from RWPE-1. Calicin is a single-copy nuclear gene that encodes for a basic protein of the sperm head cytoskeleton.

2.�. �uanti�cation of mtDNA Damage and Copy
Number Using the Absolute ss-qPCR Method. e amount of relaxed/damaged mtDNA and total copy number were measured by quantifying the original and preheated DNA templates, respectively. e nuclear DNA marker calicin was quanti�ed using the original templates. e qPCR was performed using the Applied Biosystems7500 Fast Real-Time PCR System (ABI) with Power SYBR Fast Green PCR MASTER MIX (ABI) [35]. e original DNA templates and preheated DNA templates and standards were analyzed in triplicates on the same plate. e two-step PCR ampli�cation program for both nuclear DNA and mtDNA was 95.0 ∘ C for 30 sec, followed by 40 cycles of 95.0 ∘ C for 3 sec and 60.0 ∘ C for 30 sec. A melt curve analysis was enabled at the end of ampli�cation. e primer sequences are listed in Table 1. e absolute copy numbers of CO2, D-loop, and calicin were calculated based on the standard curves. Since calicin is a single copy nuclear gene, the cell number could be calculated with the following equation with the assumption that the nuclear equivalent is representative of the cell number ( e exact copies of damaged and total mtDNA per cell were calculated from: mtDNAcopies cell = CO2 or D-Loop copy number cell number * , (3) * cell number and nuclear equivalent will be used interchangeably from this point.

Data Analysis.
All statistical analyses were performed with the aid of Graphpad Prism version 4 soware. Unless speci�ed otherwise, the data was analyzed with one-way ANOVA with Dunnett post test, and a . is considered signi�cant.

�.�. A Ne� Strategy for the Absolute �uanti�cation of �otal
and Damaged mtDNA. We have devised a new approach for the absolute quanti�cation of mtDNA structural damage and total copy number in a single analysis. e protocol, illustrated in Figure 1, was comprised of four main steps. where m is the mass of a single copy and n is the target size in base pairs. • Standard solutions range: 3 × 10 6 to 30 copies/cell with a 10× dilution factor. e �rst step consisted in the construction of mtDNA and nuclear DNA standards (Figure 1(a)). Two to three kb DNA fragments containing mtDNA (CO2 or D-loop) and nuclear DNA (calicin) were ampli�ed by PCR from a normal prostate cell line, RWPE-1. e concentration (copies/ L) of these long DNA fragments were quanti�ed and calculated according to (1). e second step was to prepare the DNA templates for qPCR analysis (Figure 1(b)). Each DNA template was split into two equal halves. One half was used for the quanti�cation of relaxed mtDNA and calicin nuclear DNA copies. e other half was pretreated at 95 ∘ C for 6 min to unfold any structure and was used for quantifying total mtDNA. e third step consisted in the absolute quanti�cation using qPCR (Figure 1(c)). To obtain mtDNA content per cell, the exact amount of mtDNA and nuclear DNA copies were quanti�ed and calculated from the standard curves according to the equation: copies = "10 (Ct− , " where the cell number was derived from (2). e �nal step was the interpretation of the data (Figure 1(d)). With this approach, the amount of damaged mtDNA copies/cell, total mtDNA copies/cell, and baseline mtDNA damage (ratio of damaged mtDNA/total mtDNA) were quanti�ed simultaneously. 0.9996 for calicin. A single uniform melting peak at 76 ∘ C, 75 ∘ C, and 79 ∘ C was observed for CO2, D-loop, and calicin, respectively, demonstrating the high speci�city of the primers (Figures 2(b), 2(d), and 2(f)). e intra-assay reproducibility of the standard was analyzed by calculating the coefficient of variation (CV) of the triplicates. e intra-assay median CV were 0.27%, 0.17%, and 0.12% for CO2, D-loop, and calicin, respectively ( Table 2). e interassay CV was calculated with data from two or more independent experiments: the CV were 0.33%, 0.10%, and 0.62% for CO2, D-loop, and calicin, respectively ( cell lines, LNCaP and C4-2, were analyzed for mtDNA content and baseline damage. e prostate cancer cell lines served a reference in method development because the ss-qPCR method was previously developed with these cell lines [6,31,35]. In lymphocytes, the total mtDNA content was quanti�ed at an average of 153.25 ± 21.02 copies/cell from 4 individual samples, among which the amount of damaged mtDNA molecules was averaged at 41.44 ± 7.87 copies/cell (Figure 3(a)). In comparison, signi�cantly higher mtDNA contents were detected in prostate cancer cells C4-2 (1495.35 ± 12.45, 0.01) and LNCaP (3086.61 ± 48.27, 0.01). e damaged mtDNA copies were 466.44 ± 8.64 and 990.41 ± 6.77 copies/cell, respectively. e baseline damage was calculated with the ratio of damaged mtDNA over total mtDNA: 27.04% of the total mtDNA content was damaged for lymphocytes versus 31.19% and 32.09% for C4-2 and LNCaP, respectively (Figure 3(b)). is assay was highly reproducible; the median intra-and interassay CV were 0.74% and 1.20% for cell lines and 1.87% and 2.33% for lymphocytes, respectively (Table 3). Furthermore, the use of different mtDNA markers, CO2 and D-loop, generated near identical results in terms of total mtDNA content, damaged mtDNA, and baseline damage detected ( Figure  3(c)). Indeed, the average CV value obtained between CO2 and D-loop markers was calculated at 0.51%. us, these two mtDNA markers were highly consistent and interchangeable in quantitative mtDNA analyses. It was interesting to note that the absolute number of damaged mtDNA molecules was proportionally higher in samples with increased total copy numbers (Figure 3(a)). As such, the ratio between damaged and total copy numbers was a better indicator of the baseline level of DNA damage in a cell (Figure 3(b)). Taken together, the new quanti�cation platform developed in this study provided a highly reproducible method for simultaneous analysis of absolute mtDNA copy number, damaged molecules, and baseline damage in both isolated lymphocytes and cancer cell lines.

Ex Vivo mtDNA Damage Responses to Exogenous H 2 O 2 in Isolated Lymphocytes and in Whole
Blood. Isolated lymphocytes from 9 healthy men were treated with 0 or 120 M H 2 O 2 for 15 min to evaluate induced mtDNA damage and repair activity aer 60 min of recovery. e average mtDNA copy number of the untreated control samples was 161.78 ± 31.67 copies/cell (Figure 4(a)). e total mtDNA copy number was not affected by H 2 O 2 treatment and remained stable across all treatment groups (Figure 4(a)). However, rapid mtDNA damage response was observed. e average baseline damage of untreated control samples was 27.63% (Figure 4(b)). Upon H 2 O 2 exposure, the fraction of damaged mtDNA increased to 58.19% in lymphocytes, representing a 110.6% increase in induced damage from the control ( 0.001). Interestingly, the induced damage was not repaired aer 60 min of recovery, suggesting a lack of repair activity during the recovery period.
As an alternative to the isolated lymphocytes, a small amount of whole blood samples ( 1 mL each) from four healthy subjects was tested using the same procedure. e average total mtDNA content of untreated control was 109.4± 22.40 copies/cell ( Figure 5(a)). Similar to lymphocytes, the average baseline mtDNA damage of the untreated control samples was 26.6% ( Figure 5(b)). When treated with 120 M H 2 O 2 , the damaged fraction of mtDNA increased to 36.7%, representing a 38.0% increase in induced damage as compared to the baseline levels ( 0.05), while the total mtDNA content remained the same (Figure 5(b)). An absence of repair activity was also observed within 60 min recovery aer the H 2 O 2 treatment. However, the whole blood samples had slightly lower mtDNA content and less pronounced mtDNA damage responses as compared to the lymphocytes. is could be caused by the complexity of different types of white blood cells present in whole blood samples. Despite this difference, the overall stress response pattern was similar between the isolated lymphocytes and the whole blood. us, the latter may serve as a convenient alternative to isolated lymphocytes in the analysis of mtDNA stress responses in circulating blood.

Discussion
Based on our previously developed ss-qPCR method [6,31,35], we propose a quantitative approach for precise and rapid detection of mtDNA structural damage and copy number change in isolated lymphocytes in a single analysis. We have demonstrated that the new approach had a wide dynamic range and was highly speci�c and reproducible. A relatively low mtDNA content and baseline level of damage were observed in lymphocytes of healthy men, and the lymphocytes were shown for the �rst time to exhibit a signi�cant increase in mtDNA damage, followed by little repair activity aer 1 h of recovery in an ex vivo challenge experiment with H 2 O 2 . is lack of repair activity to H 2 O 2 -induced damage aer 1 h of recovery is consistent with a study from Collins et al. in which nuclear DNA repair activity was only observed aer several hours (>2 h) [34]. Moreover, we showed that 1 mL of whole blood may serve as a convenient alternative  to the isolated lymphocytes in the mtDNA analysis. us, mtDNA in blood may be explored as a sensitive surrogate to systemic oxidative stress by simultaneous analysis of multiple end-points in a single test. e absolute quanti�cation system developed in this study provides a standard method for the reliable quanti�cation of the precise mtDNA copy number in lymphocytes and whole blood cells. is is achieved through well-de�ned mtDNA and single-copy nuclear DNA markers and by taking into account the DNA structural effects on qPCR ampli�cation [31]. e relatively low mtDNA copy number revealed in isolated lymphocytes is consistent with very limited data reported in the literature. For example, one study detected ∼87 to 579 copies/cell with a different real-time PCR method [36] and the other ∼70 to 320 copies/cell with competitive PCR in lymphocytes [37]. Many studies report mtDNA content on a relative scale [19,[38][39][40][41]. However, the relative analysis is limited by the difficulty of comparing results from one  study to another and by the signi�cant variations observed in mtDNA content between individuals [42]. To account for the inhibitory effect of the supercoiled DNA structure on qPCR ampli�cation, we have taken steps to ensure an accurate measurement by disrupting the supercoiled mtDNA conformation with a preheating step prior to qPCR analysis. is step is necessary for precise quanti�cation of mtDNA content but has largely been ignored in previous reports. Depending on the manufacturers of the qPCR machinery and chemistry kits, there are wide variations in the duration of the initial hot-activation step for hot-start DNA polymerases, which varies from 10 min to as short as 20 sec (e.g., ABI 7500 Fast System). We have shown that shorter denaturation time at 95 ∘ C was insufficient to disrupt all the supercoiled mtDNA conformation in a time-and dose-dependent experiment in prostate cancer cell lines [31]. erefore, the inclusion of a preheating step in the template preparation is crucial for the accurate mtDNA measurement. In addition to measuring the total mtDNA content, our new system also provides a novel approach for direct quanti�cation of the absolute copies of damaged mtDNA with qPCR. is is in contrast to mtDNA conformational study based on gel electrophoresis coupled with Southern Blot, which requires tedious post-PCR manipulations and is semiquantitative in nature. On the other hand, popular assays such as the comet test for detecting nuclear DNA strand breaks are not applicable to mtDNA due to its small size [34]. e quanti�cation of structural mtDNA damage reported in this study mainly re�ects the damage caused by single-and double-stranded breaks, as it was shown that other type of DNA damage such as base lesions or abasic sites had little, if any, effect on the structure [31]. It is interesting to note that the amount of damaged mtDNA changes with the total mtDNA content in a cell. e direct comparison of the damaged mtDNA molecules from different individuals can be compounded by variations in the total content. To normalize this variation, we propose to calculate mtDNA damage based on the percentage of damaged versus total mtDNA molecules in a cell; this ratio of damage is relatively stable and more informative for comparative studies [43]. Moreover, the ratio of damage can be used to infer the baseline or endogenous damage in the isolated lymphocytes or whole blood from the untreated samples; it also quanti�es induced mtDNA damage in isolated lymphocytes or a small amount of whole blood cells under oxidative stress. e ability of our approach to measure both endogenous and induced damage/repair responses in ex vivo treatments may be used to explore the state of oxidative defence and/or repair capability of individuals with different disease conditions. Indeed, previous studies have suggested that there is an association between systemic oxidative stress and diseases, such as an association of the high prostate cancer risk with severe damage response and poor repair capacity of nuclear DNA in lymphocytes [14].
In conclusion, we have developed an absolute quanti�cation system for rapid measurement of mtDNA structural damage, copy number change, and damage response in isolated lymphocytes and whole blood cells. Systemic oxidative stress is associated with diverse pathological conditions, ranging from neurodegenerative diseases to many types of cancers. It is conceivable that the blood analysis with the multiple mtDNA end-points proposed in the this study may provide a simple and sensitive test to interrogate the nature and extent of systemic oxidative stress for a broad spectrum of clinical investigations, especially when coupled with other established tests such as cell-free circulating mtDNA, the comet assays targeting the nuclear DNA, and the detection of 8-oxo-G base lesions.
�on�ict of �nterests e authors do not have any con�ict of interests with the content of the paper.