Evaluation of ApoE Genotyping Using Saliva-Derived DNA

The apolipoprotein E (ApoE) gene is located on chromosome 19q13.2. It consists of four exons and three introns with 3597 base pairs and produces 299-amino acid polypeptide [1]. It has three alleles (E2, E3 and E4), differing by the presence of either C or T nucleotides at codons 112 and 158 in the fourth exon. There are six combinations of the alleles (E2/E2, E3/E3, E4/E4, E3/E2, E4/E3 and E4/E2) [2,3]. The ApoE genotype frequencies in Japanese population were reported to be 0.3% for E2/E2, 6.1% for E3/E2, 71.9% for E3/E3, 0.7% for E4/E2, 19.3% for E4/E3, and 1.7% for E4/E4. The ApoE allele frequencies were 3.7%, 84.9%, and 11.7% for E2, E3 and E4, respectively. A racial difference has been reported in the ApoE allele frequencies [4-6]. The relationships of ApoE with diseases (e.g. type III hyperlipoproteinemia, coronary heart disease, stroke, peripheral arterial disease and diabetes mellitus) that are associated with increased lipid and cholesterol levels and with Alzheimer’s disease have been studied. Many studies have reported that the ApoE E4 gene is a risk factor for Alzheimer’s disease across several ethnicities (Caucasians, Japanese, Latin Americans and Hispanics) [7-11].

Generally, Sanger sequencing is used for ApoE genotyping. In this method, the regions with rs429358 and rs7412 in exon 4 of ApoE are amplified by PCR using the DNA extracted from blood samples as a template. Other methods include the amplification refractory mutation system (ARMS) [12], restriction fragment length polymorphism (RFLP) [13], real time-PCR (RT-PCR) [14], reverse hybridization (RH) [15] and fluorescence polarization (FP) [16].
Blood samples with a reasonable quality and amount of genomic DNA can be obtained and are generally used as a source of genomic DNA for genotyping. However, PCR success rate of ApoE tends to be low owing to its high GC content around the single nucleotide polymorphism (SNP) regions rs429358 and rs7412 in exon 4 [17]. In recent years, saliva samples have emerged as an alternative to blood as a source of genomic DNA. It has several advantages compared with blood sampling, including less invasiveness, low infection risk, and easy transportation. Stable saliva samples can be stored at room temperature for a long time. Also, in contrast to blood sampling, a trained professional is not required to collect saliva samples. However, it is unknown whether the quality of DNA extracted from saliva is lower than that extracted from blood because DNA isolated from saliva is contaminated with food residues or DNA derived from oral bacteria [18,19].
In ApoE genotyping, the results obtained for genomic DNA extracted from saliva samples have been reported [20]. However, the comparison of these results with those of blood samples has not been reported. Because the concentration and quality of genomic DNA extracted from saliva are lower than those of genomic DNA extracted from blood, the accuracy of genotyping results should be carefully investigated. Hence, we evaluated the genotyping results using two methods, Sanger sequencing and RFLP, for both blood and saliva samples.

Verification design and subject information
This study was conducted in accordance with the approval of

Sample
Both blood and saliva samples were collected from each of the 51 subjects. Using a 2 mL VENOJECT ® II blood collection tube, 2 mL of blood was collected from the cubital vein and it was mixed well, frozen and stored. Using the sampling kit Oragene ON-500 (Genotec Inc.), 0.5 mL of saliva sample was collected. Eating, drinking, smoking, and chewing gum were prohibited from 30 min before saliva collection.

DNA extraction
Blood sample (300 µL) was mixed with 30 µL Proteinase K and 300 µL lysis buffer, and an isothermal reaction was performed in an aluminum block at 56°C for 20 min. Subsequently, the whole sample was dispensed into a Maxwell 16 LEV Cartridge (Promega) well and processed using Maxwell 16 (Promega), an automated nucleic acid extraction instrument. DNA was extracted in LEV DNA mode.
The saliva sample was subjected to isothermal reaction in a constant temperature air-blower at 50°C for 16 h or longer. Next, DNA was extracted using the same method as for blood samples.

Measurement of DNA concentration and DNA quality
The concentration of genomic DNA extracted from the sample was measured by fluorescence intensity in a plate reader using Quant-iT PicoGreen dsDNA Assay Kit (Life Technologies). The quality of the extracted genomic DNA was evaluated by visual inspection of the 1% agarose gel electrophoresis image.

Comparison of the concentration and quality of genomic DNA extracted from saliva and blood
The mean (SD) concentrations of genomic DNA extracted from blood and saliva were 156.1 ng/μL (54.2) and 46.3 ng/μL (47.7), respectively. The concentration of genomic DNA extracted from saliva was lower than that from blood. One of the genomic DNA samples extracted from blood obtained mean+2 SD or more of the yield (Sample ID: 00018). Three of the genomic DNA samples extracted from blood obtained mean+2 SD or more of the yield (Sample ID: 00008, 00013 and 00047). In addition, the concentration distributions of genomic DNA extracted from saliva were biased toward the regions with low concentrations (Tables 1 and S1 and Figure S1).
The quality of extracted genomic DNA was evaluated by agarose gel electrophoresis. Low-quality sample were not found in 51 genomic DNA samples extracted from blood, whereas 9 low-quality samples (17.6%) were found in 51 genomic DNA samples extracted from saliva. The position of the major band of extracted genomic DNA was 23 kb or more for all the samples ( Figure S2 and Table S2).

ApoE genotyping
PCR was performed using the extracted genomic DNA from saliva samples, with blood sample as a template, and subsequently, genotyping was carried out by Sanger sequencing and RFLP. As a result, PCR products were obtained for all the samples (PCR products from 101F and 102R primers, 452 bp; from 201F and 202R primers: 244 bp) ( Figures S3 and S4). However, among the PCR products amplified with 201F and 202R primers, bands of 244 and 500 bp were observed in one sample (saliva; Index Code: 55, Sample ID: 00010).
Using Sanger sequencing and RFLP, the genotype of all the samples could be identified and waveform sequences in all samples had good quality (data not shown) ( Figure S5). In both the methods, the genotyping results of saliva samples were 100% consistent with the results of blood samples (Tables 2 and S3).

Discussion
This study revealed that the accuracy of genotyping results of ApoE gene from saliva samples was similar to that of blood samples. Appropriate PCR products were obtained in all samples and genotyping results of the ApoE gene were consistent for both the methods. The genotype frequency data were similar to the previously reported data of Japanese population; therefore, the results were considered to be valid [6]. Low-quality DNA % (no.) 0 (0) 17.6 (9)  As stated in previous reports, genomic DNA extracted from saliva samples showed higher variability in concentration and quality than the genomic DNA extracted from blood samples [18]. DNA extracted from saliva is a mixture of human and nonhuman genomic DNA. According to Dawes et [21]. Furthermore, saliva contains numerous bacteria (approximately 1.7 × 10 7 /mL); therefore, DNA extracted from saliva also contains bacterial DNA [22]. It is difficult to collect only undamaged human genomic DNA from saliva samples. This results in higher variability in the sample quality. PCR is easily affected by contaminants, such as bacterial DNA, when the quality or concentration of the genomic DNA sample is low. As a result, incorrect or no PCR products may be obtained because of poor response. In general, the ApoE gene has a high GC content and is located in a region which is difficult to amplify using PCR. In our study, we did not observe saliva genomic DNA in which low concentrations or quality negatively affected the PCR results. However, when genotyping is performed using saliva sample in place of blood sample, the quality of genomic DNA used as a template should be carefully investigated.
At least three alleles (E2, E3 and E4) could be identified by Sanger sequencing and RFLP methods. The sequence waveforms of all the samples were of sufficient quality for ApoE genotyping. However, in the RFLP method, many extra bands were observed. Special caution should be taken as the quality of genomic DNA may affect the genotyping results.

Conclusion
From this study, it was concluded that the genomic DNA extracted from saliva samples had higher variability in quality and concentration than that of genomic DNA extracted from blood samples. When genomic DNA extracted from saliva is used for genotyping of the ApoE gene, it is necessary to confirm if it is appropriate as a source before use. Nevertheless, in this validation, it was revealed that the APOE genotyping results from saliva were as accurate as those from blood.   Genotyping results from Sanger sequencing and RFLP are shown. No subject had an E2/E2 genotype. Accuracy (%): Concordance rate of genotyping results for blood and saliva samples. Genotype frequencies (%) obtained in the current study and in a previous study by Eto et al. [6] is tabulated together