to BioTechniques free email alert service to receive content updates.
A quantitative approach to detect and overcome PCR inhibition in ancient DNA extracts
Christine E. King1,2, Régis Debruyne1, Melanie Kuch1, Carsten Schwarz1, and Hendrik N. Poinar1,2, 3
Full Text (PDF)
Supplementary Material

Following gradient optimization (data not shown) of the annealing temperature and Mg2+ concentration, the assay demonstrated highly sensitive and reproducible amplification of serial dilutions of the IPC. Additional details are included in the MIQE checklist (Supplementary Table 1).

In each 20 µL reaction, 12,500 copies of the IPC were amplified with 1× PCR Buffer II (10 mM Tris-Cl pH 8.3 and 50 mM KCl; Applied Biosystems), 2.5 mM MgCl2, 300 nM primers, 400 µM (each) dNTP, 1 U of AmpliTaq Gold DNA Polymerase (Applied Biosystems), and 0.167× SYBR Green I solution (Invitrogen, Carlsbad, CA, USA). Following an initial denaturation at 95°C for 7 min, the reactions were subject to 50 cycles of 95°C, 59°C, and 72°C (30 s each), with data collection at the end of the annealing step. To verify that the cDNA-derived template was amplified, the melting temperature (Tm) of the product was monitored during a final cycle of 95°C for 1 min, 55°C for 30 s, and 95°C for 30 s.

Three reaction types were used in these tests: sample reactions, which included the IPC and one DNA extract at 10% of the final PCR volume; standard reactions, which included only the IPC; and no template control (NTC) reactions.

Inhibition measurement

Two measurements of inhibition were determined using the qPCR data for each sample reaction relative to the standard reactions. For the first, the Hill slope of each amplification plot was calculated by fitting a variable-slope sigmoidal dose-response curve to the raw fluorescence data using GraphPad Prism version 5.02 for Windows (GraphPad Software, San Diego, CA, USA). Under a sigmoidal model, this is the only parameter affecting amplification efficiency calculations in the early cycles of PCR (14,24) and these calculations of initial amplification efficiency are highly consistent with those derived from standard curves (19). Since an absolute measure of amplification efficiency was not required in this analysis, we compared the Hill slope values directly; the Hill slope of each sample reaction was expressed as a percentage of the average of the two corresponding standard reactions. For simplicity, this value (E) is referred to in the text as “amplification efficiency” or just “efficiency,” with any deviations from 100% due to the effects of PCR inhibitors in the extract; however, we acknowledge that it is in fact a relative measurement of the shape of an amplification plot. For the second measurement, the Cq of each reaction was automatically calculated by the MxPro – Mx3000P QPCR System (Stratagene, La Jolla, CA, USA) using the amplification-based threshold-determination algorithm; Cq shifts (ΔCq) were measured as the difference between the sample Cq and the average Cq of the two standard reactions.

To quantify the total effect of inhibition and any dilution of the extract on subsequent PCRs, we calculated the expected recovery (ER) value, where ER = (dilution factor × 2ΔCq)-1 × 100%. For example, from an extract showing ΔCq = 1 when used directly in PCR (dilution factor = 1), we expect other assays to recover 50% of the true yield. Likewise, for an extract showing no inhibition (ΔCq = 0) when tested at a 1/10 dilution, we can only expect to recover 10% of the templates available in the original extract.

Circumventing inhibition

For the soil samples, our goal was to optimize the PCR conditions (in terms of amplification facilitators) for each extract in order to minimize inhibition and reduce, if not eliminate, the need to dilute these extracts (i.e., maximize ER). Therefore, the inhibition tests were repeated with combinations of 0.75 mg/mL BSA, increased amounts of Taq (2.5 U per reaction), and using various dilutions of the extracts (in their corresponding elution buffer). Based on the performance of the most inhibited samples, the conditions were considered optimized when the ER was ≥20%. The bone, feces, and hair samples were tested with all combinations of BSA, additional Taq, and a range of dilutions. In order to preserve the extracts, sample reactions were performed only once with the exception of four inter-run duplicates used to assess experimental variation. One NTC and two standards reactions were included for every set of PCR conditions in every run.


All of the NTC reactions were negative, while all standard and positive sample reactions had Tm values of 81.8°C (SDTm = 0.5°C). These results indicate that no unspecific primer binding or contamination with human DNA had occurred. Based on the data from duplicate standard reactions, the mean intra-assay variation in Cq and Hill slope measurements was low, with SDCq = 0.12 cycles and CVHS = 1.3%; between runs, CVHS increased to 3.8%. Although inter-assay comparisons of Cq values for standard reactions are not possible, four sample reactions duplicated in different runs indicated that SDΔCq was also low (0.08 cycles), with SDE = 2.1%. No combination of PCR facilitators produced levels of variation significantly different from the mean values presented here. Based on this precision, we presume that the data from single sample reactions is reliable, although there is a risk that aberrant reactions may have gone unnoticed in the absence of technical replicates.

  1    2    3    4    5