Initial DNA extractions were performed on 4000 Guthrie cards using a 96-well block extractor and standard commercial protocols. DNA samples were utilized successfully in a high-sensitivity kinetic PCR assay to genotype two SNPs from the hereditary hemochromatosis gene (data not shown). At least half of the DNA was available for use in the African American HLA genotyping study. Limited amounts of DNA and lack of access to appropriate technology to quantify such small amounts of DNA necessitated empirical determination of starting quantities of template DNA for use in the linear array reactions. Results were generally readable; however, probe intensities were generally faint and varied among samples. An example of primary data produced from the initial blood spot DNA preparations can be seen in . Based on later NanoDrop (Thermo Scientific, Wilmington, DE) spectrophotometer measurements, we estimated that input DNA amounts for the original blood spot DNA preparations were likely as low as 1 to 4 ng per assay. Genotyping analysis of three of the eight classical HLA loci (DRB1, DQA1, and DQB1) required the use of essentially all the remaining DNA from the first round of purification.
Figure 2. Examples of linear arrays from our DQA1/DQB1 genotyping assay performed with different starting template DNA as follows: (A) Guthrie card DNA from initial preparations using the QIAGEN 96-well block extractor and standard QIAGEN protocols. Note the variability (more ...)
Continuation of the project required extraction of residual blood on the 1000 Guthrie cards from African American infants. Because the amount of residual blood varied, the amount of starting material for the extractions varied as well. We investigated commercial technologies for the extraction of DNA from paper cards and tested whether modification of the initial steps in our DNA extraction protocol, using a solvent developed for that purpose (GenSolve, GenVault Carlsbad, CA), might improve DNA yields. Nine samples were preincubated with the GenSolve reagent by heated shaking at 65°C, and nine samples were prepared with the reagents and extraction protocol included with the QIAGEN extraction kit. We note here that these preparations differed from the initial preparations in that they were performed in individual tubes rather than in the 96-well block extractor, which itself may have led to improved DNA yields. After the DNA was extracted from the paper, subsequent DNA purification steps were performed on both sets of samples with the QIAGEN QIAamp blood kit according to the manufacturer's instructions, with the modification that the elution step was performed at 70°C (). As measured by NanoDrop spectrophotometry, the average concentration of the nine preparations soaked in the GenSolve reagent was 12.77 ng/μl, while the average concentration of the samples that were extracted with the QIAGEN kit alone was 3.81 ng/μl (data not shown). With an elution volume of 200 μl, the total yield of DNA, although variable among preparations, was as much as 2000 ng, as measured by NanoDrop spectrophotometry. Based on those pilot data, second-round DNA preparations were performed with the GenSolve reagent, and resulting concentrations were measured by NanoDrop spectrophotometry. DNA concentrations were normalized to 10 ng/μl, and 6 μl was used in each linear array assay. We note here that the concentrations of DNA produced in these preparations were near the low end of the range for measurement by the NanoDrop instrument. We subsequently discovered, using PicoGreen technology (Molecular Probes, Eugene, OR), that our NanoDrop measurements were approximately 6- to 10-fold too high. Thus, the starting material for these assays was not 60 ng, but more likely 6 to 10 ng. Even so, the increase in the amount of input DNA and the normalization of the DNA concentrations combined to produce more intense signals that were more uniform among samples, creating data that were much easier to interpret. An example of these data can be seen in . For comparison, shows typical data from the same HLA genotyping assay, using 60 ng DNA purified from whole blood as the starting material. Assays from blood card DNA prepared by the modified, warm elution method () were more similar to the standard () than assays performed on the initial set of DNA preparations ().
To test whether our modification of the elution step of the DNA extraction protocol actually increased DNA yield, DNA was prepared from two 5-mm punches of three different blood spots, with half of each sample eluted at room temperature without soaking and half eluted with the modified warm soaking method (see Methods). Results of this experiment are represented graphically in . Although the sample-to-sample variation was great among the three samples (total yield as measured by PicoGreen technology ranged from 66 to 660 ng DNA), data clearly show that samples with high-temperature elution gave consistently higher yields than samples eluted at room temperature using either method of measurement. In addition, these data confirmed our suspicion that measurements done by NanoDrop spectrophotometry gave artificially high readings compared to measurements done with PicoGreen technology. Sample 3 gave the highest yield, with 3200 ng total as measured by NanoDrop spectrophotometry; 660 ng total when measured by PicoGreen technology, which is a more conservative, and probably more accurate, number than the 3200-ng yield as measured by NanoDrop spectrophotometry. In our experiments, NanoDrop measurements appeared to overestimate the DNA concentration by approximately 6- to 10-fold, probably due to the fact that our concentrations were near the lower limit of detection by the NanoDrop instrument. The PicoGreen system was developed for use in a much lower range of concentration than the NanoDrop instrument. These yields came from a single 5-mm punch of a Guthrie card. Typical blood spots on a card are large enough to allow three to five 5-mm punches.
Figure 3. Total DNA yield from direct comparison of DNA extraction protocols varying only by the elution step. Data are shown for both NanoDrop and PicoGreen measurements and show that despite variation in total yield among samples, samples eluted with the modified (more ...)