Quality and Quantity of Saliva DNA Obtained from the Self-administrated Oragene Method—A Pilot Study on the Cohort of Swedish Men

For the establishment of DNA biobanks from large cohorts, it is of great importance to collect from as many subjects as possible to get a representative sample and to increase the power in association studies on genetic variants and disease. DNA quality is essential for the success rates in genotype analyses, and the amount of collected genomic DNA is of importance for future studies. A peripheral blood sample is the preferred source of DNA with respect to both quality and quantity. However, the need for the donors to go to a primary health care facility to have the sample drawn as well as the invasive character of this method may reduce response rate. A simple, self-administrated sample collection protocol may increase participation rates especially among elderly participants. Saliva and buccal cells are alternative DNA sources that can be self-collected through rinses, brushes, or swabs (1-12). Disadvantages with these methods are that the donors need to either swish and spit an alcohol-containing mouthwash solution (1, 3-5, 8-10), which have been reported to give a burning sensation after swishing (3), or follow sometimes complicated instructions on how to collect buccal cells with brushes or swabs (1, 3, 12). Furthermore, mouthwash solutions with relatively high alcohol content (e.g., Scope) are not available in all countries. A new method for saliva self-collection, Oragene, was recently introduced on the market in which the donors simply just spit into a vial. When the vial is capped, a solution containing antibacterial and DNA preserving chemicals mixes with the saliva, resulting in immediate conservation of the sample. In a pilot study encompassing 611 men ages 53 to 87 years, we evaluated response rate, DNA quality and quantity, and whole-genome amplification (WGA) after using this novel collection method.

From the population-based prospective cohort of Swedish men consisting of 48,645 subjects, born 1918 to 1952 and living in Västmanland and Örebro counties (13), 625 men, equally distributed in five age groups, were randomly selected. After exclusion of men who died or moved from the study area between baseline and September 2004, 611 men were eligible for the study.

All men received a preliminary notification, with short information about the study and that a saliva collection kit will be arriving within a few days. Three days later, they received the Oragene kit (DNA Genotek, Inc., Ottawa, Ontario, Canada), a longer information letter, detailed instructions on how to deliver the saliva sample, an informed consent form for signature, and a prepaid return envelope. Nonresponders received two reminders by mail; the first after 2 weeks and the second after a further 2 weeks. Participants were requested to spit ∼2 mL saliva. Four hundred ninety sent a saliva sample and a signed informed consent, 43 sent an answer saying that they did not want to participate, and 78 did not respond. Saliva samples from 90 respondents were randomly chosen for DNA extraction. Extraction was done 1 to 2 months after saliva collection and storage in room temperature.

DNA was robotically extracted by the Autopure LS system using the Puregene DNA purification kit (Gentra Systems, Minneapolis, MN), and the yield and A260/A280 ratio were determined with UV at 260 and 280 nm. Quantitation of DNA yield was also done at the DNA extraction facility in Malmö, Sweden by using PicoGreen and a FLUOstar Optima device (BMG Labtech, Offenburg, Germany), and 100 μL of PicoGreen reagent (diluted 1:200 in Tris-EDTA buffer) were dispensed into 100 μL of sample (5 μL DNA plus 95 μL Tris-EDTA buffer). Samples were incubated in darkness for 5 minutes before fluorescence reading at the excitation and emission wavelengths of 485 nm and 520 nm, respectively. DNA (5 μL) was visualized on agarose gels together with 125 ng λ DNA.

To get an estimate of the relative fraction of human and bacterial DNA, real-time PCR of the human prothrombin gene with 5′ exonuclease (Taqman) probes was done on an ABI Prism 7900HT system (Applied Biosystems, Foster City, CA). Primer and probe sequences are available on request. The PCR of 20 μL contained 20 to 200 ng DNA, 1.25 mmol/L MgCl₂, 0.1 μL Custom Assay mix, including primers and probe, and 5 μL Taqman Universal Master Mix (Applied Biosystems). Denaturation was done at 50°C for 2 minutes and 95°C for 10 minutes followed by 50 cycles each consisting of 95°C for 15 seconds and 60°C for 1 minute. The proportion of human DNA was calculated by dividing received number of prothrombin PCR copies with estimated number of total DNA copies added in the PCR. Number of total DNA copies was calculated from the DNA concentration quantified with PicoGreen by assuming that 1 ng of genomic DNA contains ∼270 copies. This figure is based on the approximation that one cell (equivalent to two DNA copies) contains ∼7 pg DNA.

The saliva DNA samples were genotyped for 10 single-nucleotide polymorphisms (SNP; Table 1) randomly selected from the human genome by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (Sequenom, Inc., San Diego, CA; ref. 14). PCR assays and associated extension reactions were designed using the SpectroDESIGNER software (Sequenom). Primer sequences are available on request (Metabion GmbH, Planegg-Martinsried, Germany). All amplification reactions were run in the same conditions in a total volume of 5 μL with 2.5 ng genomic DNA, 1 pmol of each amplification primer, 0.2 mmol/L deoxynucleotide triphosphate, 2.5 mmol/L MgCl₂, and 0.2 unit HotStarTaq DNA polymerase (Qiagen, Inc., Valencia, CA). Reactions were heated at 95°C for 15 minutes and subjected to 45 cycles of amplification (20 seconds at 94°C, 30 seconds at 60°C, 30 seconds at 72°C) before a final extension of 7 minutes at 72°C. Extension reactions were conducted in a total volume of 9 μL using 5 pmol of allele-specific extension primer and the MassEXTEND Reagent kits before being cleaned using SpectroCLEANER (Sequenom) on a Multimek 96 automated 96-channel robot (Beckman Coulter, Fullerton, CA). Clean primer extension products were loaded onto a 384-element chip with a nanoliter pipetting system (Sequenom) and analyzed by a MassARRAY mass spectrometer (Bruker Daltonik GmbH, Bremen, Germany). The resulting mass spectra were analyzed for peak identification using the SpectroTYPER RT 2.0 software (Sequenom). For each SNP, two independent scorers confirmed all genotypes. Hardy-Weinberg calculations were done to ensure that each marker was within allelic population equilibrium (>0.01; ref. 15) in our sample set.

WGA (16) of saliva DNA samples was done by using the GenomiPhi DNA kit (Amersham Biosciences, Uppsala, Sweden). After measurement of DNA yield, gel electrophoreses, and SNP analysis of the saliva DNA samples, there was enough material left for WGA on 81 of 89 samples. One to 5 ng of template DNA were used per reaction. The amplified DNA (referred to hereafter as WGA DNA) was visualized on 0.8% agarose gels. The WGA DNA was genotyped for the same 10 SNPs (Table 1) as the saliva DNA samples by using the protocol described above.

Student's t tests were used to evaluate differences of response rate between age groups. Spearman rank correlation (r_s) was used to test correlations between DNA yields measured with PicoGreen, UV absorbance, and real-time PCR as well as between DNA yield and age.

We obtained a total response rate of 80%, 7% did not want to participate in the study, and 13% did not respond after two reminders. Of the total number of 490 samples, 52% were received during the 1st week, 24% during the 2nd week, and 17% during the 3rd week of the sample collection. We observed statistically significant differences of response rates between age groups. The highest rates were observed in age groups 67 to 71 (89%) and 62 to 66 (85%) compared with age group 53 to 61 (73%; Ps < 0.01 and 0.02, respectively). The lowest response was observed in the oldest age group 77 to 87 (71%).

By visual inspection of the received samples, the saliva volume was estimated to vary between 1.5 and 2.5 mL; phlegm or other contaminants were not visible. We extracted DNA from 90 randomly selected saliva samples and roughly estimated total DNA yield by UV absorption. The mean total yield was 135.9 μg (SD, 118.2), and the mean A260/A280 ratio, used as an estimate of DNA purity, was 1.76 (SD, 0.12). We also measured the total DNA yield by using the PicoGreen method. In one sample, the DNA was not measurable. The total DNA yield ranged from 1.2 to 169.7 μg, with a mean of 40.3 μg (SD, 36.5 μg) and a median of 29.4 μg. By real-time PCR of the human prothrombin gene, we estimated that the yield of human DNA ranged from 11% to 100% of total DNA, with a median of 68% (SD, 20%). The human DNA yield ranged from 0.8 to 85.6 μg, with a mean of 25.4 μg (SD, 22.4) and a median of 19.2 μg. Electrophoretic analysis of the extracted DNA showed detectable levels of high molecular weight genomic DNA (>23 kb) in all 89 samples. No degradation of DNA was observed. Total DNA (bacterial and human), as measured with PicoGreen, was highly correlated with both total DNA as measured with UV absorbance (r_s = 0.90) and human DNA as measured with real-time PCR (r_s = 0.92; Fig. 1). We found no correlation between age and DNA yield [r_s = 0.02 for total DNA (PicoGreen), r_s = 0.07 for total DNA (UV absorbance), and r_s = 0.005 for human DNA (real-time PCR)].

The DNA amount, as measured with PicoGreen, obtained after WGA of 1 to 5 ng of saliva DNA was ∼4 μg. Quality assessment of both saliva DNA and WGA DNA was done by genotype analyses of 10 SNPs. All genotype frequencies were in Hardy-Weinberg equilibrium (Table 1). The success rates for the 10 specific SNP assays ranged between 91% and 100% for the saliva DNA samples and between 88% and 99% for the WGA DNA samples (Table 1). We obtained all 10 genotypes for 70 of 89 (79%) saliva DNA samples and 64 of 81 (79%) WGA DNA samples. Of the samples that were not completely genotyped for all 10 SNPs, none missed more than five genotypes. Eleven percent of the saliva DNA and 29% of the WGA DNA samples missed four or five genotypes. The rest missed one, two, or three genotypes. Almost complete genotypic concordance (99.7%) was observed between saliva DNA and WGA DNA samples, and 736 of 738 genotypes were identical.

In our work with planning a DNA biobank from subjects in the cohort of Swedish men (13), we conducted this pilot study to evaluate collection of saliva through the self-administrated Oragene method, which to our knowledge has not been previously evaluated in a scientific report. We contacted 611 men and asked them to spit roughly 2 mL saliva into a collection vial, which they returned by mail to our laboratory. The method turned out to be user friendly, and 80% responded with a saliva sample. Of the received samples, 93% were mailed to us within 3 weeks from the send out date. Extraction yielded high molecular weight DNA, which was of sufficiently high quality for genotype analyses and for WGA.

Other methods for collection of saliva or buccal cells include mouthwash, buccal swabs/brushes, and treated cards (1-12, 17). The human DNA yield from the Oragene method (19 μg) is in the same range as the human yield from the mouthwash method (3, 4) but substantially higher compared with buccal swabs/brushes (0.2-2.7 μg per swab/brush; refs. 2, 3, 7, 12). The foremost advantage with the Oragene protocol compared with these methods is its simplicity; neither do the donors need to put something in their mouth before collecting the sample nor do they need to follow a certain collection protocol, such as rubbing their cheeks against their teeth to prepare the buccal mucosa before mouthwash (1) or tooth brushing (11). Moreover, when closing the collection vial directly after spitting, the saliva is mixed with DNA preserving and purifying chemicals, which prevents bacterial growth and degradation of human DNA. Feigelson et al. (4) have shown that handling of mouthwash samples 10 to 30 days after collection gave a statistically significantly reduction of human DNA yield, which indicates that the time between collection and processing of the samples when samples are held at room temperature may be important for the human DNA yield. According to the manufacturers of the Oragene kit, the saliva samples can be stored at room temperature for up to at least 1 year without DNA degradation.³

One obstacle with saliva DNA is the high variability of bacterial DNA. In DNA extracted from buccal swabs, human DNA accounted for only 11% (3). Our data showed that 68% of the total DNA was of human origin, which is higher compared with Garcia-Closas study (3) on the mouthwash method in which 49% was of human origin. However, the higher proportion of human DNA from the Oragene samples compared with the mouthwash samples may be due to the different methods used to measure human DNA. We also found that the DNA yield varied substantially between saliva samples, the amount of human DNA ranged between 1 and 86 μg, and ∼20% of the samples had a yield <5 μg. Considering genotype analyses using 2 to 5 ng DNA per PCR, biobank DNA samples in this lower yield range would not be sufficient for future use. We therefore evaluated WGA as a backup approach and showed that this method can be used to generate additional DNA. Genotyping of the WGA DNA worked satisfactorily; however, to evaluate a potential risk for allelic dropout derived from WGA, >10 SNPs need to be tested in the future.

In conclusion, the Oragene collection method, which in this study was evaluated in men, is a suitable way to collect DNA from large groups of subjects where self-administration is an advantage. The DNA quality allows for genotyping with high success rate, and the yield is sufficient for a large number of analyses. Furthermore, the DNA is amplifiable, which insure near-unlimited biobank material for future studies.

We thank Cecilia Lindgren (Mutation Analysis Facility, Karolinska Institutet, Stockholm, Sweden) and Joyce Carlson (DNA Extraction Facility, Malmö, Sweden) for providing invaluable methodologic knowledge and Camilla Lagerberg (KI Biobank, Stockholm, Sweden) for excellent technical assistance.