1998 Oxford University Press

Nucleic Acids Research, 1998, Vol. 26, No. 11

28292830

An alternative PCR assay for quantifying mitochondrial DNA in crude preparations

ABSTRACT A method is described for the quantification of mitochondrial DNA present in crude biological preparations. A known copy number of a standard is amplified in the presence of inactivated target DNA so as to determine the overall efficiency of the PCR process in a particular sample. In this way any inhibitory and/or stimulatory substances present in sample preparations can be taken into account. To reduce tube-to-tube variations product DNA quantification is limited to small cycle numbers. Using this method quantitations of DNA amounts in different crude preparations can be compared. Studies involving mitochondrial biogenesis often require the quantification of mitochondrial DNA (mtDNA) as an index of mitochondrial proliferation. Due to its high specificity, PCR has become an important method for quantifying mtDNA. One of the major problems associated with quantitative PCR is tube-to-tube product variability and, since amplification is exponential, minor differences in the amplification efficiencies can result in large differences in the overall product yield (1,2). In order to overcome this limitation, several approaches have been developed including serial dilution, kinetic PCR and competitive PCR (3). Most often, PCR template and competitor products are separated by agarose gel electrophoresis and quantified by radioimaging and densitometric scanning. Each of these methods has distinct disadvantages, including the use of isotopes and narrow ranges of linear response. Furthermore, most of these techniques require specialized equipment which is not always available. It was the aim of the present study to develop an accurate and specific PCR assay for the quantification of mitochondrial DNA (mtDNA) present in crude mitochondrial preparations. A further requirement was that the technique should be neither dependent on densitometric analysis nor on radioactive labelling. As a result, we have designed an external control system which can be used to correct for differences in the amplification efficiency. Unlike other external control protocols (4,5), standard DNA is amplified under the identical reaction conditions as the unknown. This is achieved by simply inactivating the target DNA in the sample by UV irradiation, a well established method for

decontaminating pre-PCR reagents (6). Since the standard and the unknown can now be amplified in different tubes, there is no need to synthesize a competitor molecule. Purified PCR product serves as the ideal standard since it is identical to the target DNA. Mitochondria isolated from porcine perirenal adipose tissue were dispersed in 10 mM TrisHCl (pH 7.5) and mtDNA was extracted by digestion with Proteinase K (100 g/ml) in a buffer containing 50 mM KCl, 10 mM TrisHCl, 2.5 mM MgCl2 and 0.5% Tween 20. Samples were incubated overnight at 37_C and then boiled for 5 min. MtDNA was linearized by digestion with HindIII. Amplifications were carried out in 10 l volumes in glass capillary tubes. Each reaction contained the appropriate DNA, 1 PCR reaction buffer (50 mM TrisHCl, 250 g/ml BSA, 2 mM MgCl2), 200 M of each dNTP, 50 pmol of each primer and 3.5 U/100 l of Taq polymerase. The PCR primers used to amplify a portion of the mitochondrial cytochrome b gene (mtCytb) were 5-CAGAATGATATTTGTCCTTCA-3 and 5-GATATGAAAAACCATCGTTG-3. Amplification was carried out in a 1002 Air Thermo-cycler (Idaho Technology) with a single denaturation step (15 s at 94_C) and cycles of 0 s denaturation at 94_C, 0 s annealing at 50_C and 30 s elongation at 72_C. Depending on the initial amount of target DNA, tubes were removed at cycle numbers ranging from 8 to 38 cycles. PCR product was quantified using picoGreen (7). Amplification curves, derived from PCR product purified using Promega Magic PCR preps, were constructed by plotting relative fluorescence versus cycle number (Fig. 1A). The cycle number at which a profile crosses over a fluorescence intensity of 50 (threshold cycle number) was read from these curves and used to construct an external standard curve of threshold cycle number versus the logarithmic values of the mtCytb copy number (Fig. 1B). In order to correct for the influence of inhibitory factors in the unknown sample, 100 l aliquots were dispensed into 1 ml polypropylene tubes and placed 10 cm above a 254 nm UV light source. Although preliminary experiments revealed that 5 min treatment was sufficient to completely block amplification, it was decided to use 10 min for routine analysis. Amplification in the presence of a free radical trap (8-hydroxyquinoline) illustrated that the lack of amplification was not due to free radical production. Standard DNA was added to UV treated mitochondrial preparations and amplified as described above. In this way, any differences calculated between the UV treated plus standard threshold cycle number (T0) and the corresponding standard

Downloaded from http://nar.oxfordjournals.org/ by guest on December 25, 2011

*To whom correspondence should be addressed. Tel: +27 41 504 2441; Fax: +27 41 504 2814; Email: [email protected]

2830 Nucleic Acids Research, 1998, Vol. 26, No. 11

Figure 1. (A) Amplification curves derived from ampifying 10-fold dilutions of standard DNA (purified mtCytb PCR fragment). Copy numbers were determined empirically using purified PCR product derived from mtCytb primers. The starting mtDNA copy numbers were: 1.23 109 (J), 1.23 108 (+), 1.23 107 (j) and 1.23 106 (v). (B) External control standard curve generated from threshold cycle numbers of amplification curves as shown in (A). Regression analysis gave an equation of Y = 4.058(X) + 47.75 with a regression coefficient of 0.99. Each data point represents the mean from four experiments (CV <5%).

Figure 2. Correlation between mitochondrial sample dilution and PCR quantitated mtDNA copy number. Ten-fold dilutions of a mitochondrial preparation were amplified as described in the text. Mitochondrial DNA copy number was determined from the standard curve illustrated in Figure 1B and was based on the threshold cycle number after correction for interfering substances (TC). Regression analysis gave an equation of Y = 1.102(X) + 8.085 with a regression coefficient of 0.98. Each data point represents the mean from three experiments (CV <5%).

Downloaded from http://nar.oxfordjournals.org/ by guest on December 25, 2011

threshold cycle number (TS) would represent the influence of interfering substances present in the sample. TS/T0 serves as a correction factor to convert the uncorrected threshold cycle number of the unknown sample (TU) to the corrected value by the formula TC = TS/T0 TU were TC is the corrected threshold cycle number. A strong linear and reproducible correlation was evident between the dilution of the mitochondrial samples and the PCR quantified mtDNA copy number (Fig. 2). This linear correlation persisted over a 10 000-fold dilution range. Multiple measurements at each dilution were performed with excellent reproducibility (CV ranged from 1.6 to 4.7%). Experimental accuracy was further tested by comparing the result (using a number of different mitochondrial isolations) obtained from kinetic PCR and the present method. In terms of reproducibility, our approach produced superior results. The most serious disadvantage of such a non-competitive PCR assay is the variability in the product produced from one tube to another. The external control system described here takes into account the variation which results from the presence of inhibitors (and/or activators), but does not consider variation in factors inherent to PCR technology including limiting Taq polymerase, self-annealing of the product DNA strands, formation of spuriously primed products and the reproducibility of the thermal cycler. Such tube-to-tube variation can be virtually eliminated by reducing the number of cycles during which these effects accumulate. We have achieved this goal by using a sensitive fluorescent dye, picoGreen, which is at least 100-fold more sensitive than the commonly used ethidium bromide (7). Furthermore, the use of a threshold cycle number rather than product yield, allows for lower cycle numbers and in this way amplification of the unknown is limited to the early cycle numbers irrespective of the concentration of target DNA. It is also

crucial to exclude non-specific amplification to ensure the specificity of the quantified DNA. In conclusion, our quantitative PCR assay represents a simple alternative to methods such as kinetic PCR. The assay is technically straightforward, eliminating the need for post-PCR manipulations, radioactive isotopes and densitometry. In contrast to other non-competitive quantitative PCR assays, the effects of inhibitory and/or stimulatory substances are taken into account thus eliminating the need for tedious phenol/chloroform extractions and ethanol precipitations, thus allowing direct amplification of extracted DNA. ACKNOWLEDGEMENTS This work was supported by grants from the Foundation for Research Development (FRD) and the University of Port Elizabeth.

REFERENCES
1 Gilliland,G., Perrin,S., Blanchard,K. and Bunn,H.F. (1990) Proc. Natl. Acad. Sci. USA 87, 27252729. 2 Diaco.R. (1995) In Innis,M.A., Gelfand,D.H. and Sninsky,J.J. (eds), PCR Stratagies: Practical Considerations for the Design of Quantitative PCR Assays. Academic Press, San Diego, pp. 84108. 3 Raeymaekers,L. (1995) Genome Res. 5, 9194. 4 Abbott,M.A., Kowolenko,M.D. and Warner,G.L. (1994) Toxicol. Methods 4, 161176. 5 Nedelman,J., Heagerty,P. and Lawrence,C. (1992) Bull. Math. Biol. 54, 447502. 6 Rolfs,A., Schueller,I., Finkh,U. and Weber-Rolfs,I. (1992) In Rolfs,A., Schueller,I., Finkh,U. and Weber-Rolfs,I. (eds), PCR-Clinical Diagnostics and Research. Springer, Berlin, Heidelberg, New York, pp. 6167. 7 Ahn,S.J., Costa,J. and Emanuel,J.R. (1996) Nucleic Acids Res. 24, 26232625.