Quantification of mitochondrial DNA copy number in suspected cancer patients by a well optimized ddPCR method

Abstract

Changes in mitochondrial DNA (mtDNA) content is a useful clinical biomarker for various diseases, however results are controversial as several analytical factors can affect measurement of mtDNA. MtDNA is often quantified by taking ratio between a target mitochondrial gene and a reference nuclear gene (mtDNA/nDNA) using quantitative real time PCR often on two separate experiments. It measures relative levels by using external calibrator which may not be comparable across laboratories. We have developed and optimized a droplet digital PCR (ddPCR) based method for quantification of absolute copy number of both mtDNA and nDNA gene in whole blood. Finally, the role of mtDNA in suspected cancer patients referred to a cancer diagnostic center was investigated. Analytical factors which can result in false quantification of mtDNA have been optimized and both target and reference have been quantified simultaneously with intra- and inter-assay coefficient variances as 3.1% and 4.2% respectively. Quantification of mtDNA show that compared to controls, solid tumors (but not hematologic malignancies) and other diseases had significantly lower copy number of mtDNA. Higher mtDNA (highest quartile) was associated with a significantly lower risk of both solid tumors and other diseases, independent of age and sex. Receiver operating curve demonstrated that mtDNA levels could differentiate controls from patients with solid tumors and other diseases. Quantification of mtDNA by a well optimized ddPCR method showed that its depletion may be a hallmark of general illness and can be used to stratify healthy individuals from patients diagnosed with cancer and other chronic diseases.

Keywords: Biomarker; Cancer; Droplet digital PCR; Mitochondrial DNA; Nuclear DNA; ddPCR; mtDNA.

Fig. 1

Droplets separation before (A) and after (B) enzymatic digestion. Levels of both mtDNA (red line) and nDNA (blue line) quantified in serially diluted DNA samples (n = 1 for each dilution). Data are corrected for dilution factor and are representative of two replicated experiments (C).

Fig. 2

Linearity of the mtDNA (A) and nDNA (B) as single-plex and as multiplex assay (C) is shown in serially diluted samples. R² for goodness of the fit is shown. Results are representative of two replicated experiments.

Fig. 3

Tukey box and whiskers plot showing the levels of mtDNA in controls and the most common cancers and the most common other diseases diagnosed at the DC center are shown. The line between boxes indicates the median value, while the outer boxes represent the 25th and the 75th percentiles and whiskers show the non-outlier range. P-values calculated by tow tailed t-test and are shown as **<0.005, ***<0.0005 and ****<0.0001, ns = non-significant.

Fig. 4

Odds ratios calculated for each quartile of mtDNA levels (shown on x-axis) in solid tumors (A) and other diseases (B). Horizontal line represents the lowest quartile as a reference. P-values calculated by univariate and multivariate (adjusted for age and sex) logistic regression analysis and are shown as *<0.05, **<0.005, ***<0.0005 and ****<0.0001.