FALCON-qPCR: a new method for the quantification of oxidative lesions in mitochondrial DNA

ABSTRACT

Accurate quantification of oxidative mitochondrial DNA (mtDNA) lesions remains technically challenging due to the limitations of existing assays, which often require large sample inputs, multi-day workflows, and offer limited sensitivity. Here we introduce FALCON-qPCR (Fpg-assisted Long-PCR), a streamlined, high-sensitivity method for quantifying oxidative damage in mtDNA. FALCON-qPCR couples digestion with formamidopyrimidine [fapy]-DNA glycosylase (Fpg) to long-range PCR and qPCR-based normalization, enabling precise lesion quantification from as few as 10,000 cells (~300 ng total DNA) within a single day. The assay provides a robust dynamic range and reproducibility across diverse biological systems, including human cell lines, hepatocellular carcinoma biopsies, and Caenorhabditis elegans. Compared with established methods, FALCON-qPCR exhibits markedly higher sensitivity in detecting mtDNA damage induced by hydrogen peroxide, antimycin A, and rotenone. Its performance was further demonstrated in assessing mitochondrial toxicity of ruthenium-based compounds, highlighting its potential for pharmacological screening. By integrating enzymatic lesion recognition with quantitative amplification in a unified workflow, FALCON-qPCR eliminates the need for mitochondrial isolation. This methodological advance provides a rapid, accurate, and scalable platform for studying oxidative DNA damage, with broad applicability in mitochondrial research and translational toxicology.

Plain Language Summary

Mitochondria are tiny structures inside our cells that produce energy. When they work, they also create reactive molecules called ROS (reactive oxygen species). If ROS levels get too high, they can damage mitochondrial DNA (mtDNA), which may lead to diseases such as cancer and neurodegeneration. Measuring this damage is important, but current methods are slow, need large samples, and are not very sensitive. We developed a new method called FALCON-qPCR that measures mtDNA damage quickly and accurately. It uses a special enzyme to mark damaged DNA and then applies two types of PCR (a technique to copy DNA) to calculate how much damage is present. Unlike older methods, FALCON-qPCR works with very small samples – just 10,000 cells – and gives results in one day. We tested FALCON-qPCR on human cells, liver biopsies from cancer patients, and even tiny worms (C. elegans). In short, FALCON-qPCR is a fast, reliable, and versatile tool for studying mitochondrial DNA damage in research and clinical settings.

1. Introduction

Mitochondria are the main endogenous source of reactive oxygen species (ROS), highly reactive molecules whose levels, if not controlled, can cause oxidative damage to mtDNA [1]. Electron leakage from the mitochondrial electron transport chain (ETC) is considered the main cause of mitochondrial ROS production [2] which, under physiological conditions, is counterbalanced by enzymatic (e.g., superoxide dismutase, catalase, and glutathione peroxidase) and non-enzymatic (e.g., exogenous diet-derived small molecules) antioxidant defenses [3]. Reduced antioxidant defenses and/or increased ROS production are characteristics of pathological conditions and may result in a redox imbalance leading to various ailments, including cancer, respiratory, cardiovascular, neurodegenerative, and digestive diseases [4].

Detection of oxidative base damage derivatives is still a critical issue in the laboratory, especially on mtDNA. Currently available protocols lack reliability, reproducibility, and accuracy while also being time- and material-consuming [5–13]. The first approaches to study mtDNA oxidative lesions were indirect and based on assessing mitochondrial function and enzyme activity as proxies to mtDNA integrity [14]. With the growing understanding of oxidative stress and its impact on DNA, researchers developed some of the earliest techniques to assess mtDNA damage: Southern blot [15], comet assay [10], and HPLC coupled with electrochemical detection [16]. The advent of quantitative PCR (qPCR) in the early 2000s enabled the development of the first qPCR-based protocols for the quantification of nuclear and mitochondrial DNA damage [6,9,11,13]. The assay relies on the principle that many kinds of DNA lesions can slow down or block the progression of the DNA polymerase, negatively interfering with the PCR amplification [13,17]. If equal quantities of DNA from different samples are amplified under identical conditions, the relative amount of damage correlates with the amount of amplification obtained: the more amplification, the less damage, and vice versa. However, this relationship is strongly influenced by the fidelity of the DNA polymerase used. Enzymes with low fidelity may replicate across damaged bases, thereby underestimating the true extent of mtDNA lesions. Compared to previous approaches, the use of specific primers for qPCR amplification eliminates the need for mitochondrial isolation, significantly reducing the amount of starting material required for analysis. However, despite this advantage, using the same amount of template mtDNA is essential for the relative quantification of DNA lesions. For this purpose, the inputs need to be standardized based on the total DNA concentration measured spectrophotometrically, without accounting for variability in mtDNA copy number between samples. Another limitation is that qPCR typically amplifies only short DNA fragments, which prevents the detection of base lesions outside the targeted regions, thereby significantly reducing the sensitivity of the assay. This problem was partially solved by the introduction of Long-Range PCR (LR-PCR), which allows the assessment of DNA damage by amplifying a long fragment ( >3,000 bp) of mitochondrial DNA [7,12,18]. In this case, the normalization between samples was obtained through the amplification of a short mitochondrial fragment (~ 100bp), where the probability of base damage is very low. Even if in line with principle this is true, in a previous work we demonstrated that high levels of oxidative stress affect the amplification of the short amplicon, therefore compromising the normalization and underestimating the levels of damage [19]. A further improvement was the introduction of DNA glycosylases, which recognize and remove damaged bases, generating abasic sites that cause polymerase stalling during amplification. This approach enhances the detection of oxidative lesions by ensuring that damaged sites inhibit DNA synthesis [5,6,9].

While developing the FALCON-qPCR we took into consideration all the limitations of the previous methods for developing an optimized protocol for the detection of mtDNA oxidative lesions, which couples the use of a glycosylase for the detection of a specific oxidative lesion, with the core LR-PCR step as a crucial method for the detection of damage and the qPCR for sample normalization and quantification of LR-PCR amplicons. This workflow overcomes the limitations of the previous approaches: minimizing the amount of mtDNA required for analysis, improving the detection of amplicons, solving the normalization problem, and reducing the time required to perform the analysis from three days to a single day. The first step of the protocol involves the isolation of DNA from a small number of cells, as few as 10,000 (approximately 300 ng of total DNA). Samples are incubated with formamidopyrimidine [fapy]-DNA glycosylase (Fpg). Fpg is an enzyme with dual activity: it catalyzes the removal of oxidized purines using both its glycosylase and AP lyase activities [20]. This process leads to the creation of a gap where the oxidized base is located, ensuring that DNA polymerase in the following LR-PCR cannot proceed with elongation, thereby guaranteeing the reliability of the protocol. Following this procedure, we specifically quantify the amount of mtDNA through qPCR to ensure both control and experimental samples have the same amount of template mtDNA for the LR-PCR. Finally, the amount of LR-PCR product is evaluated via qPCR, determining the relative amount of damage to the mtDNA, by comparing the ratio of amplification in the control sample to the amplification in the sample of interest. The main steps of the protocol are summarized in Figure 1. A detailed description of the protocol is provided in Supplementary File 1. The analysis workbook and its user guide are available in Supplementary Files 2 and 3, respectively.

Figure 1.

FALCON-qPCR protocol overview. Schematic representation of the steps of the protocol. DNA extraction can be performed starting from as few as 10,000 cells or adjusted to different animal models. After Fpg digestion, a qPCR reaction is used to normalize the amount of mtDNA loaded in the LR-PCR. A final qPCR allows for the detection of the amount of mtDNA amplified during the LR-PCR and consequently for the following mtDNA damage evaluation. Image created with BioRender (https://app.biorender.com/citation/67c5721e9de063e41807733f).

2. Materials and methods

2.1. Cell culture and treatments

HeLa, SF767, Huh7, U87, and MEFs cells were grown in DMEM (Dulbecco’s modified Eagle’s medium) supplemented with 10% fetal bovine serum (FBS); OCI-AML2 cells were grown in alpha-MEM (Minimum Essential Medium) supplemented with 20% FBS; JHH6 cells were grown in Williams’ medium E supplemented with 10% FBS. All media were completed with 100 U/mL penicillin and 10 μg/mL streptomycin. Cells were incubated at 37°C with 5% CO₂. Hydrogen peroxide (H₂O₂), rotenone and Antimycin A (AMA) treatments were performed at the indicated concentrations in culture medium without FBS.

2.2. Caenorhabditis elegans maintenance and treatment

C. elegans N2 wild-type strain was maintained at 20°C on nematode growth medium (NGM) plates seeded with Escherichia coli OP50 as a food source [21]. A total of 10 N2 wild-type worms were grown from the L1 stage to adulthood for 4 days at 20°C on either control plates or plates supplemented with 200 μM or 300 μM Paraquat. After 4 days, worms were collected in ddH₂O and immediately frozen in liquid nitrogen. DNA extraction was then performed as described below.

2.3. Synthesis of ruthenium-based compounds and experimental design

The investigated ruthenium-based compounds [Ru(η¹-OPiv)(CO)((R)-BINAP)(phen)]OPiv (C4) and [Ru(Cl)(CO)(dppe)(phen)]Cl (C10) are organometallic complexes with ruthenium as metal center and coordinated organic ligands. For the synthesis, reactions were carried out under an argon atmosphere by using standard Schlenk techniques. The solvents were used without previous drying. The ligands 1,2-bis(diphenylphosphino)ethane (dppe), (R)-BINAP and all other chemicals were purchased from Merck and Strem and used without further purification [22]. NMR measurements were recorded on a Bruker Advance III HD NMR 400 spectrometer and the chemical shifts, in ppm, are relative to TMS for ¹H and ¹³C{¹H} NMR and 85% H₃PO₄ for ³¹P{¹H} NMR. Elemental analyses (C, H, N) were carried out with a Carlo Erba 1106 elemental analyzer. Detailed descriptions of the syntheses are reported in the Supplementary Data.

The two compounds were dissolved in dimethyl sulfoxide (DMSO) 100% and stored at room temperature, protected from light. Compound powders were dissolved anew every three weeks. Working solutions were freshly prepared before each experiment. U87 were seeded in wells and after 24 h were treated with ruthenium-based compounds for 72 h and then collected and analyzed.

2.4. DNA extraction

Trypsinized cells were harvested and counted. Concentration was adjusted to 5x10⁴ cells/mL and DNA extraction was performed starting from 200 μL and using the QIAmp DNA Mini Kit (Qiagen). 20 μL of Proteinase K and 200 μL of buffer AL were added to the resuspended cells and incubated at 56°C for 10 min. For the C.elegans samples, 10 nematodes were collected and resuspended in 180 μL of buffer ALT supplemented with 20 μL of Proteinase K. Nematodes were incubated at 56°C for 30 min to ensure complete lysis. After 200 μL of ethanol addition, the mixture of either lysed cells or nematodes was applied to the purification column. Following the manufacturer’s instructions, two washes were performed with solutions AW1 and AW2 and finally, DNA was eluted in 50 μL of water. Alternatively, cells grown in one well of a 6-well plate, were harvested, resuspended in 600 μL of PBS and the extraction was performed on 200 μL of cell suspension. Total DNA extracted was quantified on NanoDrop, and 300 ng were used for the analysis.

2.5. Fpg digestion

Formamidopyrimidine DNA Glycosylase (Fpg) enzyme (New England BioLabs) was diluted in its 1X buffer to a final concentration of 1.6 U/μL. A reaction was prepared with 50 μL of DNA (300 ng), 10 μL of Fpg buffer 10X, 20 ng of BSA, 3.2 U of Fpg and water to 100 μL final volume. The reaction was incubated at 37°C for 30 min. Enzyme activity was inactivated at 60°C for 10 min. To remove Fpg before proceeding with the following steps of the protocol, 233 μL of ethanol were added to the mix and the sample was applied to a new purification column of the QIAmp DNA Mini Kit and processed as described before.

2.6. RealTime PCR (RT-PCR)

The RT-PCR experiments were performed with the CFX96 Real-Time System (Bio-Rad) or with the QuantStudio 1 Real-Time PCR System (ThermoFisher) using PowerUp™ SYBR™ Green Master Mix for qPCR (ThermoFisher). The primers reported in Table 1 were used. The cycling parameters applied were denaturation at 95°C for 10 s and annealing/extension at 60°C for 30 s (repeated 40 times). In order to verify the specificity of the amplification, a melting-curve analysis was performed immediately after the amplification protocol.

Table 1.

List of primers used to perform the LR-PCR or the RT-PCR.

Name	Specimen	Sense	Antisense
LR-PCR human	Cell line Biopsies	TCTAAGCCTCCTTATTCGAGCCGA	TTTCATCATGCGGAGATGTTGGATGG
RT-PCR human	Cell line Biopsies Standard curve	CCCCACAAACCCCATTACTAAA	TTGGTCGTGGTTGTAGTCCG
LR-PCR mouse	Cell line	CCGCGAGCCTTCAAAGCCCT	AATGGGCCCGGAGCGAGAAGA
RT-PCR mouse	Cell line	CAGCACCCAAAGCTGGTATT	AATGTACTAGCTTATATGCTTGGGG
LR-PCR C.elegans	Nematodes	GCTGAACTTAACCGGGCGCCAT	CTCACCGGTGTGGGGGCTCT
RT-PCR C.elegans	Nematodes	ATTCATCTTCATCTTGGGAGGAT	AAGCCAAACACTAATTCCACCT

2.7. LongRange-PCR (LR-PCR)

The LR-PCR was performed using the primers listed in Table 1. Amplification was performed with Platinum™ SuperFi™ DNA Polymerase (Invitrogen) diluted 1:50, using the following protocol: 2 min at 94°C, 24 cycles of denaturation for 15 s at 94°C, annealing for 10 s at 66°C, extension for 5 min and 30 s at 68°C and a final step of 10 min at 68°C. The PCR product was purified with QIAquick PCR Purification Kit (Qiagen) following manufacturer’s instruction. Briefly, 250 μL of buffer PB were added to the sample and applied to the purification column. A wash was performed with 750 μL of buffer PE and the purified PCR product was eluted in 50 μL of water.

For the mtDNA damage evaluation reported in Figure 3(A,B), the damage was quantified as previously reported [12]. DNA was extracted using Qiagen genomic-tip 20/G and following the manufacturer’s indications. After isolation, DNA was precipitated overnight with isopropanol, and then 10 μg were digested with 20 U of Fpg enzyme as described above. DNA was precipitated overnight, resuspended in 50 μL of Tris-EDTA buffer pH 8.0 and quantified with Quant-iT™ PicoGreen™ dsDNA Reagent (Invitrogen), according to the manufacturer’s instructions. DNA concentration was adjusted to 3 ng/μL. mtDNA lesions were quantified by LR-PCR with the primers reported in Table 1. In parallel, a PCR reaction was run to amplify a small mitochondrial fragment of 221 bp with the following primers: Mitoshort Forward: 5-CCC CAC AAA CCC CAT TAC TAA ACC CA-3’ and Mitoshort Reverse: 5’-TTT CAT CAT GCG GAG ATG TTG GAT GG-3.’ DNA was amplified with Platinum™ SuperFi™ DNA Polymerase (Invitrogen) using the already described condition for the LR-PCR and the following protocol for the short fragment: 2 min at 94°C, 18 cycles of denaturation for 15 sec at 94°C, annealing 45 s at 60°C, extension for 45 s at 72°C and a final extension for 10 min at 72°C. To ensure quantitative conditions a sample with the 50% of template amount was included in each amplification and, as a negative control, a sample without the template was used. PCR products were quantified in triplicate by using Quant-iT™ PicoGreen™ dsDNA Reagent (Invitrogen). The Mitoshort fragment was used to calculate the relative amount of mtDNA copies and to normalize the lesion frequencies calculated with the LR-PCR.

Figure 3.

FALCON-qPCR comparison with reference protocol. (A). Comparison of FALCON-qPCR with the standard protocol (Furda et al.). HeLa cells were treated with 25 μM AMA for 30 minutes in media without FBS to induce mitochondrial oxidative stress. 2 million cells were processed following Furda et al. recommendations, while 10,000 cells were used to measure mtDNA oxidative base lesions with the FALCON-qPCR. The mtDNA oxidative damage levels were quantified relative to untreated samples (CTRL) (** p ≤ 0.01; n = 3). (B). Surgical resections and needle biopsies from eight patients affected by HCC were collected from the tumoral mass and from a distal non-tumorigenic part of the liver and processed according to Furda et al. protocol or the FALCON-qPCR (n = 3). (C). Relative amount of mtDNA oxidative damage levels in HeLa cells treated with either H₂O₂ (left) or rotenone (right) for 15 minutes at the concentrations reported in the graphs. mtDNA oxidative damage was quantified relative to untreated cells (CTRL) (*** p ≤ 0.001; * p ≤ 0.05; n.s.: not significant; n = 3). (D). Relative amount of mtDNA oxidative damage levels in HeLa cells treated with 25 μM AMA for 30, 60 and 90 minutes. mtDNA oxidative damage was quantified relative to untreated cells (CTRL) (*** p ≤ 0.001; ** p ≤ 0.01; * p ≤ 0.05; n = 3).

2.8. Samples of human tumor tissue specimens and adjacent non-tumor tissues

Samples of paired HCC and adjacent non-tumor liver tissues from patients undergoing HCC resection were obtained from the Department of Medicine, General Surgery and Transplantation of the University of Udine, Udine, Italy and stored in 20% glycerol Isolation Buffer (IB) [10 mM Tris/MOPS, 1 mM EGTA/Tris, 200 mM Sucrose]. Together with the liver sample, a 16-Gauge needle biopsy was also acquired and stored in 20% glycerol IB. None of the patients had received any local or systemic anticancer treatments before the surgery. Diagnosis of HCC was performed in all cases by preoperative imaging (CT or MRI scan) or by liver biopsy when requested. Hepatic serology, α-fetoprotein (AFP), routine laboratory assessment of liver and renal function were also performed. This study was approved by the Unique Regional Ethics Committee on 24 August 2019, Protocol number 18,659, and written informed consent was obtained from patients. All methods were performed in accordance with the relevant guidelines and regulations.

2.9. DNA extraction from human liver specimens

DNA was extracted from 15 mg of patient biopsies using the QIAGEN Genomic-tip 20/G. Samples were incubated in 2 mL G2 solution completed with RNase A (final concentration of 200 μg/mL) and Proteinase K as indicated by the company, followed by 2 h incubation at 50°C. Genomic tips were equilibrated with 2 mL of buffer QBT. Samples were then applied to the corresponding genomic tip, and the flow-through discarded via gravity. Three washes were performed with 1 mL of wash buffer QC and finally, DNA samples were eluted in 2 mL of QF buffer. Isopropanol was added to precipitate DNA over/night. DNA was extracted from needle biopsies using the QIAamp DNA Mini Kit (Qiagen), following the manufacturer’s instructions for DNA isolation from tissues. Briefly, the needle biopsies were suspended in 180 μL of Buffer ATL supplemented with 20 μL of Proteinase K and incubated at 56°C till dissolution. 200 μL of AL were added, and samples were incubated for 10 min at 70°C. 200 μL of ethanol were included in each sample before loading into the QIAamp Mini spin columns. Following the manufacturer’s instruction, two washes were performed with solution AW1 and AW2. Finally, the DNA was eluted in 100 μL and quantified using the Nanodrop.

2.10. Fpg digestion in human liver specimens and needle biopsies

After precipitation, DNA was centrifuged at 15,000 ×g for 1 h at 4°C, washed with 70% ethanol and pelleted at 15,000 ×g for 45 min at 4°C and finally resuspended in 50 μL of TE (Tris 20 mM, EDTA 0.5 mM). After a Nanodrop quantification, 10 μg of DNA isolated from the human specimens or 1 μg isolated from the needle biopsy were prepared for the Fpg digestion, following the manufacturer’s indications as described above. DNA isolated from the human specimens was then precipitated over/night with isopropanol. The day after, DNA was centrifuged at 15,000 ×g for 1 h at 4°C, washed with 70% ethanol and pelleted at 15,000 ×g for 45 min at 4°C and finally resuspended in 50 mL of TE (Tris 20 mM, EDTA 0.5 mM). A 3 ng/μL aliquot was prepared based on the PicoGreen quantification. The needle biopsy DNA was instead used to perform a quantitative RT-PCR to amplify an 84 bp fragment on the mitochondrial DNA of each sample. The primers listed in Table 1 were used (RT-PCR human)). Based on this RT-PCR, mtDNA content was evaluated and DNA samples were prepared for the evaluation of the mtDNA damage.

2.11. Short-range and LR-PCR in human liver specimen

mtDNA lesions were quantified by LR-PCR, using the primers reported in Table 1. The Mitoshort Forward: 5-CCC CAC AAA CCC CAT TAC TAA ACC CA-3′ and Mitoshort Reverse: 5′ -TTT CAT CAT GCG GAG ATG TTG GAT GG-3′ were used to amplify a 221 bp mitochondrial fragment, which was later used to normalize the amount of mtDNA obtained from the human liver specimen. DNA was amplified using Platinum ™ SuperFi ™ DNA Polymerase (Invitrogen) using the protocol described for the mtDNA damage detection on cells. To ensure quantitative conditions a sample with 50% of template amount was included in each amplification and, as negative control, a sample without the template was used. PCR products were quantified in triplicate by using Quant-iT ™ PicoGreen ™ dsDNA Reagent (Invitrogen).

The LR-PCR performed on the mtDNA from the liver biopsy was quantified via RT-PCR after a purification step as described in the previous section.

2.12. MTT assay

U87 cells were seeded on clear 96-well plates at a density of 3x10⁴ cells/cm² in 100 µl of culture medium and incubated at 37°C. On the subsequent day, cells were treated with increasing amount of the two ruthenium-based compounds C4 (from 0.01 µM and 2.5 µM) and C10 (from 0.1 µM and 10 µM). After 72 h, a solution of 10% v/v 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium (Sigma) in PBS without Ca²⁺ and Mg²⁺ (Euroclone) was added to each well and cells were incubated for 4 h at 37°C. At the end of the incubation time, cells were lysed with 100 µl of DMSO (Sigma) and the amount of produced formazan was evaluated by using a spectrophotometer, measuring the absorbance at 590 nm. Half maximal inhibitory concentration (IC₅₀) was estimated by Prism 10 for macOS (version 10.2.3), considering the dose response curve of C4 and C10 compounds.

2.13. Mitochondrial superoxide assay

Mitochondrial superoxide generation was assessed using the MitoSOX Red fluorescent probe (ThermoFisher), following the manufacturer’s protocol. In summary, 5x10³ U87 cells were seeded in into 96-well black plates and exposed to ruthenium-based compounds for 72 h. The culture medium was then replaced with DMEM/10%, supplemented with 5 μM MitoSOX (prepared from a 5 mM stock solution in 100% DMSO), and incubated at 37°C for 10 min. MitoSOX excess was removed by washing twice with PBS, and the cells were subsequently stained with Hoechst 33,342 2 μg/mL in DMEM/10%. For each experimental condition, at least 200 cells were counted using LEICA-DMI 6000B fluorescence microscope.

2.14. Statistical analysis

Statistical analysis was performed using Microsoft Excel. One-way ANOVA was used for three group comparisons and Student’s t-test was used for two group comparisons. p values of less than 0.05 were considered as significant, while values less than 0.01 or lower were considered as highly significant.

3. Results

3.1. Optimization and quality control of FALCON-qPCR

While setting the FALCON-qPCR, we aimed at solving three main limitations of the current protocols: i) reduce the amount of starting material in order to make the methods suitable for the analysis of scarce or precious samples; ii) ensuring a high dynamic range which prevents the underestimation of the damage; and iii) reducing the time to perform the whole procedure to a single day by avoiding steps of precipitation. During the development of the protocol, the amount of mtDNA required to evaluate the damage was optimized, and consequently, the quantities of enzymes. A linearity range was defined using a standard curve composed of increasing concentrations of mtDNA (Figure 2(A)). Then, primers were tested to ensure the specificity of the amplification in both LR-PCR (Figure 2(B)) and qPCR (Figure 2(C)) for all the study models used. Furthermore, quality control steps were performed to ensure no loss of material throughout the protocol. A known amount of mtDNA was prepared together with a sample containing half the amount of the same mtDNA. The two samples were processed in parallel, confirming that the 1:2 ratio was maintained until the final step of the procedure (Figure 2(D)). Finally, we tested whether cells could be frozen immediately after collection and stored at −80°C for later processing. To confirm the reliability of the analysis on frozen samples, cells treated with 200 μM H₂O₂ were either processed immediately to extract the DNA or flash frozen, and DNA was extracted at a later date to compare the results obtained. In both scenarios, the results consistently showed the presence of more mtDNA oxidative damage when cells were incubated with H₂O₂ (Figure 2(E)).

Figure 2.

Specificity and optimization of FALCON-qPCR. (A). An example of the qPCR curves obtained by loading known concentrations of mtDNA is reported (top). The Ct values are plotted against the logarithm of the mtDNA amount loaded to extrapolate the standard curve (bottom). (B). Agarose gel analysis of LR-PCR product to evaluate the specificity of the primers. 10 µL of PCR reaction from HeLa (left), C. elegans (center), and MEFs (right) were loaded and separated on 0.8% agarose gel. (C). Agarose gel analysis of qPCR product to evaluate the specificity of the primers. 15 µL of PCR reaction from HeLa (left), C. elegans (center), and MEFs (right) were loaded and separated on 1.2% agarose gel. (D). Graph showing the relative amplification of Sample A and half the amount of the same sample (A/2) which were processed in parallel to confirm no loss of material during the steps of the protocol. The initial 1:2 ratio was maintained until the end of the procedure (** p ≤ 0.01; n = 3). (E). A total of 10,000 cells were analyzed to evaluate the amount of mtDNA damage after incubation with 200 µM H₂O₂ for 15 min. The protocol was performed on cells soon after the treatment (Fresh) or after freezing the cells in liquid nitrogen and storage in −80°C (Frozen). mtDNA oxidative damage was quantified relative to untreated cells (CTRL) (n.s.: not significant, n = 3).

3.2. Comparison of FALCON-qPCR with standard protocol

To prove the accuracy of our protocol, we compared the FALCON-qPCR with the current reference procedure established in the laboratory of Prof. Bennet Van Houten [12,13]. HeLa cells were treated with a specific inhibitor of Complex III, antimycin A (AMA), to induce oxidative stress in mitochondria [23]. Two million cells were processed following the procedure described by Furda et al. [12], whereas our protocol was performed using only 10,000 cells, requiring 200 times less material, and compared with untreated samples (CTRL). The results show that the outcome in the two protocols is comparable (Figure 3(A)), confirming that 10,000 cells are sufficient to obtain a fast and reliable result. Alongside, to prove the applicability of the protocol on clinical samples, we evaluated the mtDNA oxidative damage on biopsies of patients with hepatocellular carcinoma (HCC) (Figure 3(B)), the most common type of primary liver cancer which is frequently linked to mitochondrial dysfunction and oxidative stress, both of which are critical factors in cancer progression [24]. Examining mtDNA oxidative damage in HCC biopsies provides valuable insights into the role of oxidative stress in tumor development and potential implications for disease prognosis and therapeutic strategies. Two liver biopsies were collected from each patient: one from the tumor itself and one from a distal lobe of the liver, as well as needle biopsies from the same surgical resections. The resections were used to perform the protocol following the procedure described by Furda et al., while the needle biopsies were processed according to our protocol. Our results are comparable to those of Furda et al., excluding patient 3, where the differences may arise due to the greater heterogeneity of a larger sample size compared to a single-point needle biopsy. The main difference between the two protocols resides in the magnitude of the detected damage: the FALCON-qPCR shows a more dynamic detection and higher detection range (0.13 ± 0.07 – 9.17 ± 2.78) compared to the Furda procedure (0.59 ± 0.04 – 1.76 ± 0.17).

Having confirmed that our protocol allows for the detection of oxidative damage starting from a small amount of biological material, unlike any other available methods, we proceeded with testing different stimuli to trigger ROS production in different cell lines. Hydrogen peroxide (H₂O₂) was selected to trigger general ROS-induced damage (Figure 3(C), left), while rotenone (Figure 3(C), right), an inhibitor of the respiratory chain Complex I, was used to induce oxidative stress specifically in mitochondria [25]. As expected, increasing concentrations of both compounds correlate with increasing damage detection compared to untreated cells (CTRL). To test the reliability of the protocol in measuring the kinetics of mtDNA damage induction, HeLa cells were treated with AMA for 30, 60, and 90 min. FALCON-qPCR successfully detected a proportional increase in oxidative damage with longer exposure times, demonstrating its effectiveness in quantifying mtDNA damage (Figure 3(D)).

3.3. Applicability of FALCON-qPCR to non-human study models

To extend the general applicability of our method, mtDNA damage was detected using the optimized protocol in other cell lines treated with AMA (OCI-AML2, SF767, JHH6, Huh7 – Figure 4(A-D)). Finally, we designed specific adjustments for models other than humans, extending the applicability of the FALCON-qPCR to animal models. The protocol was performed on MEFs (Figure 4(E)), a mouse cell line, using AMA as an oxidative stress-triggering agent. For C. elegans (Figure 4(F)), as few as 10 nematodes per sample (corresponding to approximately 10,000 cells) were grown on media containing different concentrations of Paraquat, a known toxic chemical that increases mitochondrial lipid peroxidation and ROS production [26]. In both cases, we observed that after the treatment, the amount of damage detected was higher than in the control.

Figure 4.

Assessment of mtDNA damage detection in different study models. (A-D). Acute myeloid leukemia (OCI-AML2), human glioblastoma (SF767), hepatocellular carcinoma (JHH6 and Huh7) cell lines were incubated with AMA (25 μM for OCI-AML2 and SF767, 200 μM for JHH6, and 75 μM for Huh7) for 30 minutes. mtDNA oxidative damage was quantified relative to untreated cells (CTRL) (*** p ≤ 0.001; ** p ≤ 0.01; * p ≤ 0.05; n = 3). (E) Mouse embryonic fibroblasts (MEFs) were treated with 25 μM AMA for 30 minutes (*** p ≤ 0.001; n = 3). (F). C. elegans were grown for 4 days at 20°C on nematode growth medium with the concentration of Paraquat reported. Total DNA was extracted from 10 worms, approximately 10,000 cells, for each condition and compared with untreated (CTRL) to detect the levels of mtDNA oxidative lesions (** p ≤ 0.01; * p ≤ 0.05; n = 3).

3.4. Assessing the effect of two ruthenium-based compounds on mtDNA

To prove the utility of our protocol for the screening of active molecules, we tested two different ruthenium-based compounds, C4 and C10 (Figure 5(A) and Supplementary Figures S1-S6), for their ability to generate oxidative lesions on mtDNA. These two compounds were selected for evaluation based on their structural properties and previous in vitro characterization [27,28]. Both compounds are organometallic complexes synthesized by the group of Prof. Walter Baratta, designed to specifically target mitochondria due to their lipophilic and cationic nature (Supplementary Material and Methods). Indeed, ruthenium-based compounds can mainly be localized into three organelles: the nucleus, the mitochondria, and the lysosome [29]. Despite the DNA being considered the key target, accumulation in the nucleus is relatively low. Instead, the main target of these compounds appears to be the mitochondria [30]. Lipophilic ruthenium-based cations are attracted to the charges displaced in the mitochondria, which are responsible for maintaining the mitochondrial membrane potential [30]. Several key features of the mitochondria are interfered by the activity of ruthenium-based compounds, including ROS generation [31], mtDNA damage [32], mitochondrial membrane potential disruption [33], interference with mitochondrial respiration, and mitochondrial calcium homeostasis [34]. To calculate IC₅₀, U87 cells were treated with increasing amounts of C4 and C10, and an MTT assay was used to measure the toxicity. For C4 the IC₅₀ resulted 0.095 ± 0.006, while cells tolerated better the treatment with C10 which showed a value of 0.740 ± 0.063 (Figure 5(B)). Based on these results, the IC₅₀ dose was selected to be used for all the treatments carried out in the present study: 0.09 μM of C4 and a 0.8 μM of C10 for 72 h. To evaluate if the compounds have a direct effect on mitochondria, MitoSOX was used to measure the mitochondrial ROS production, showing a significant increment in the number of positive cells only upon C4 treatment (Figure 5(C)). We applied the FALCON-qPCR protocol to verify if the mitochondrial ROS production observed in C4 treated cells determined mtDNA oxidative damage. As expected, U87 cells treated with C4 showed a significant increase in the mtDNA oxidative lesions (2.84 ± 1.57), while no significant differences were observed with C10 (1.26 ± 1.03) (Figure 5(D)). This further confirms the reliability of our method and its possible use for drug screening.

Figure 5.

Analysis of two ruthenium-based compounds. (A). Chemical structure of ruthenium-based compounds C4 and C10. (B). MTT assay graphs showing the 50% inhibitory concentrations (IC₅₀) of C4 (left) and C10 (right) (n = 5). (C). U87 cells were treated for 72 hours with 0.09 μM C4 or 0.8 μM C10 and later incubated with MitoSOX to evaluate the ROS production at the confocal microscope (left). The percentage of positive cells was evaluated. DMSO-treated cells were used as a negative control. The white bar corresponds to 50 μm (**** p ≤ 0.0001). (D). Assessment of mtDNA damage detection in U87 cells treated with either C4 or C10. Cells were incubated for 72 hours with 0.09 μM C4 or 0.8 μM C10. mtDNA oxidative damage was quantified relatively to cells treated with DMSO (* p ≤ 0.05, n = 3).

4. Discussion

FALCON-qPCR represents a significant advancement in the toolkit available for investigating oxidative damage to mitochondrial DNA. Compared to conventional methodologies, this protocol combines sensitivity, speed, and efficiency with minimal sample requirements, thereby overcoming the limitations that have hindered previous approaches. By requiring as few as 10,000 cells and yielding results in a single day, FALCON-qPCR enables the analysis of precious, low-yield, or time-sensitive biological materials, making it ideal for both routine laboratory use and specialized clinical applications. One of the most compelling features of this method is its improved dynamic range and sensitivity. Unlike traditional LR-PCR-based assays that rely on short fragment normalization, FALCON-qPCR integrates a qPCR-based normalization step before long-range amplification, ensuring more accurate quantification of base lesions. The method’s superior resolution was particularly evident when applied to HCC samples (Figure 3(B)), where it not only mirrored the outcomes of standard protocol, but also revealed a broader spectrum of damage levels. These findings suggest that FALCON-qPCR may help detect subtle mtDNA alterations that would otherwise go unnoticed, particularly in early disease stages or in response to mild physiological stressors. The versatility of the protocol was demonstrated across a range of model systems, from human and murine cell lines to whole organisms such as C. elegans. Its successful application in diverse biological contexts and with different oxidative stress-inducing agents underscores the method’s adaptability and its potential to become a standard approach in mitochondrial research. Additionally, its compatibility with minimal DNA from needle biopsies confirms its translational potential in clinical diagnostics. The ability to directly assess mitochondrial damage in vivo opens up new avenues for biomarker discovery and the assessment of disease progression in mitochondrial and oxidative stress-related pathologies. Additionally, FALCON-qPCR proved effective as a screening tool for bioactive compounds. In this study, two ruthenium-based agents were evaluated, revealing that only compound C4 induced substantial mitochondrial oxidative damage, accompanied by mitochondrial ROS production and reduced proliferation. This makes the assay particularly valuable in early-stage drug development, where mitochondrial toxicity is a common concern. Future adaptations could include high-throughput automation for large-scale drug screenings, as well as the ability to adapt the protocol for other oxidative base lesions. Although this protocol has been optimized specifically for 8-OXOG oxidative base lesions through the incorporation of the enzyme Fpg, this method acts as an excellent base template for the detection of other oxidative base lesions such as thymine glycol and uracil, which are recognized, respectively, by thymine DNA glycosylase and uracil DNA glycosylase. These enzymes work along the same principle as Fpg, allowing for the ability to adapt this protocol for multiple lesion types, which enhances its versatility and broadens its applicability in studies where oxidative DNA damage and repair are of interest.

In summary, FALCON-qPCR is a powerful, scalable, and accessible method for quantifying oxidative damage in mitochondrial DNA. By improving detection sensitivity, reducing sample requirements, and simplifying the workflow, it enables novel applications in both basic and translational research. From mapping the landscape of mitochondrial damage across diseases to screening therapeutic candidates, FALCON-qPCR sets a new benchmark for mitochondrial genomics and redox biology.

Article highlights

• FALCON-qPCR is a novel, streamlined method for quantifying oxidative lesions in mitochondrial DNA with high sensitivity and minimal sample requirements.

• The protocol combines Fpg digestion, long-range PCR, and qPCR normalization in a single-day workflow, eliminating the need for mitochondrial isolation.

• Compared to standard approaches, FALCON-qPCR offers a broader dynamic range, improved reproducibility, and requires as few as 10,000 cells.

• The method is validated across multiple models, including human cell lines, clinical liver biopsies, and C. elegans, demonstrating versatility and translational potential.

• FALCON-qPCR enables drug screening applications, as shown by assessing mitochondrial toxicity of ruthenium-based compounds.

• This protocol sets a new benchmark for mitochondrial genomics and oxidative stress research, with potential for high-throughput adaptation.

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Funding Statement

This work was supported by grants to CV from the Associazione Italiana per la Ricerca sul Cancro [MFAG 16780], the European Union’s Horizon 2020 research and innovation programme under grant agreement No 956070, the European Union’s Horizon Europe research and innovation program under the grant agreement No 101072515, and by the National Science Centre 2021/43/B/NZ5/01684.

Disclosure statement

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Writing assistance

No writing assistance was utilized in the production of this manuscript.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Data availability statement

Raw Data are available on Zenodo (https://zenodo.org/records/15285642) or upon request to the corresponding author.

Authors contribution

Conceptualization: VB, EH, CV. Formal analysis: VB, EH, SB, MER, DA, PDM, MT, UB. Data curation: VB, EH, CV. Supervision: WB, DC, APB, CV. Funding acquisition: CV. Writing original draft: VB, EH, CV, WB, DC, MT, UB. Writing – review & editing: VB, EH, CV.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/17576180.2025.2608757