Method specific calibration corrects for DNA extraction method effects on relative telomere length measurements by quantitative PCR

Seeker, L A and Holland, R and Underwood, S and Fairlie, J and Psifidi, A and Ilska, J J and Bagnall, A and Whitelaw, B and Coffey, M and Banos, G and Nussey, D H (2016) Method specific calibration corrects for DNA extraction method effects on relative telomere length measurements by quantitative PCR. PLoS One.

[img]
Preview
Text
11203.PDF - Published Version
Available under License Creative Commons Attribution.

Download (1MB) | Preview

Abstract

Telomere length (TL) is increasingly being used as a biomarker in epidemiological, biomedical and ecological studies. A wide range of DNA extraction techniques have been used in telomere experiments and recent quantitative PCR (qPCR) based studies suggest that the choice of DNA extraction method may influence average relative TL (RTL) measurements. Such extraction method effects may limit the use of historically collected DNA samples extracted with different methods. However, if extraction method effects are systematic an extraction method specific (MS) calibrator might be able to correct for them, because systematic effects would influence the calibrator sample in the same way as all other samples. In the present study we tested whether leukocyte RTL in blood samples from Holstein Friesian cattle and Soay sheep measured by qPCR was influenced by DNA extraction method and whether MS calibration could account for any observed differences. We compared two silica membrane-based DNA extraction kits and a salting out method. All extraction methods were optimized to yield enough high quality DNA for TL measurement. In both species we found that silica membrane-based DNA extraction methods produced shorter RTL measurements than the non-membrane-based method when calibrated against an identical calibrator. However, these differences were not statistically detectable when a MS calibrator was used to calculate RTL. This approach produced RTL measurements that were highly correlated across extraction methods (r > 0.76) and had coefficients of variation lower than 10% across plates of identical samples extracted by different methods. Our results are consistent with previous findings that popular membrane-based DNA extraction methods may lead to shorter RTL measurements than non-membrane-based methods. However, we also demonstrate that these differences can be accounted for by using an extraction method-specific calibrator, offering researchers a simple means of accounting for differences in RTL measurements from samples extracted by different DNA extraction methods within a study.