Sample Collection

We collected buccal and cloacal swab samples from 10 green turtles (59.6–103.1-cm curved carapace length) that were captured in SDB following protocols detailed in Eguchi et al. (2010) and Ehrhart and Ogren (1999). We used a stainless-steel pry bar to carefully open the turtle's mouth and a canine mouth gag was used to hold the mouth open (Forbes 1999). We swabbed the soft tissue portion of the mouth with a Whatman® Omni Swab for approximately 6 sec to collect the buccal samples. For cloacal samples, we inserted another Whatman® Omni Swab approximately 5 mm into the cloaca for about 6 sec. Both swabbing methods involved gentle scraping of the epithelium with the swab. Immediately after collection, we placed each swab inside a 15-ml conical tube and placed it on ice for field storage until transfer to a −20°C laboratory freezer. Liquid preservative was not used in the tube because the swab would absorb the liquid, thereby diluting the cells (Miller 2006) and possibly affecting DNA yield. We collected skin samples from the same subject animals using standard protocols described in Dutton (1996) to use as genetic controls.

DNA Extraction

The frozen swabs we collected were thawed and dried in a fume hood for 2–12 hrs within 5 d of collection. We recommend extracting these types of samples within 5 d of collection to avoid potential bacterial and fungal growth while in the freezer (Milne et al. 2006). Using the DNeasy® Buccal Swab protocol from the QIAamp® DNA Mini Kit (Qiagen, Valencia, CA) we removed and lysed the epithelial cells from the swabs to obtain whole genomic DNA. Our subject animals' skin samples were extracted using the Qiagen DNeasy® Blood & Tissue Kit (Qiagen) according to the Animal Tissues (spin-column) protocol. Some skin samples in our study had been extracted previously using the following high-throughput robots according to the manufacturers' protocols: Qiagen QIAxtractor (Qiagen) and the JANUS Automated Work Station (Perkin Elmer, Waltham, MA). Negative controls were included in all of our extraction procedures to detect potential DNA transfer and/or contamination between samples.

DNA Quantity and Quality Assessment

We added 2 µl of each DNA sample to the Nanodrop® ND-1000 Spectrophotometer (Thermo Fisher Scientific, Wilmington, DE) to measure DNA quantity and quality. To account for different volumes that resulted from different extractions, we converted all concentrations from nanograms per microliter (ng/µl) into nanograms (ng). We determined absorbance at wavelengths of 230 nm, 260 nm, and 280 nm, and calculated relative absorbance ratios (260/280 nm and 260/230 nm) to evaluate DNA purity obtained from buccal, cloacal, and skin samples. Samples with 260/280 ratios > 2.0 may contain excess RNA and other contaminants. The ideal 260/230 ratio range for pure DNA is between 2.0 and 2.2. We used a Kruskal–Wallis 1-way analysis of variance (ANOVA; Hampton 2003) to test for significant differences in DNA quantities as well as for the 260/280 and 260/230 ratios amongst the buccal, cloacal, and skin samples.

We tested both mitochondrial DNA (mtDNA) and nuclear DNA markers; mtDNA was amplified using 2 primers (H950G and LCM15382) with a product size of approximately 800 base pairs (bp; Abreu-Grobois et al. 2006). The cycling conditions were 90°C for 2 min followed by 35 cycles of 94°C for 50 sec, 56°C for 50 sec, and 72°C for 1 min, with an extension at 72°C for 5 min. Two nuclear microsatellite loci were amplified and run on 2% agarose minigels to confirm amplification (Fig. 1). We used the nuclear microsatellite markers A6 (Dutton and Frey 2009) and D107 (Dutton et al., unpubl. data, 2009). All polymerase chain reaction (PCR) reactions had a total volume of 25 µl and were run on an ABI 2720 Thermal Cycler (Applied Biosystems by Life Technologies, Foster City, CA). Each reaction contained 18.25 µl of MilliQ water, 2.5 µl of PCR buffer, 1.5 µl of deoxynucleotide triphosphate (dNTP), 0.75 µl of forward primer, 0.75 µl of reverse primer, 0.25 µl of Taq polymerase, and 1 µl of DNA. The PCR cycling conditions for both of these markers were 94°C for 2 min followed by 35 cycles of 94°C for 30 sec, 55°C for 30 sec, and 72°C for 30 sec, with a final extension at 72°C for 5 min. Negative controls were run with all PCRs to detect any potential contamination. We used a Kruskal–Wallis 1-way ANOVA (Hampton 2003) to test for differences in amplification rates.

View Figure

• Download as Powerpoint
• Download Figure

Sequencing

Prior to sequencing, we cleaned the PCR products using Exonuclease 1 and Shrimp Alkaline Phosphatase (ExoSAP) and then samples were run on an ABI 2720 Thermal Cycler (Applied Biosystems by Life Technologies) at 37°C for 15 min and 80°C for 15 min. A mixture of 1 µl post-ExoSAP product, 5 µl of MilliQ water, 3 µl of 1-µM primer (H950 or LCM15382), and 3 µl BigDye v3.1 with buffer was made and placed in the thermal cycler for cycle sequencing: 96°C for 1 min, followed by 25 cycles of 96°C for 10 sec, 50°C for 5 sec, and 60°C for 4 min. The labeled extension products were purified using the Big Dye Xterminator purification kit (Applied Biosystems by Life Technologies). Our samples were analyzed on an ABI 3730 DNA genetic analyzer (Applied Biosystems by Life Technologies) and sequences were aligned and evaluated using SeqScape v2.5 (Applied Biosystems by Life Technologies). We compared the sequences from the buccal and cloacal swabs taken from the same animal to determine whether there was a difference in sequence quality. Each of these buccal and/or cloacal swab-sequence sets were compared with sequences obtained from a skin sample of the same individual for quality assurance.

Results

We collected 12 buccal and 13 cloacal swabs from 10 different turtles. Although values ranged widely, the buccal and cloacal swabs did not have significantly different DNA quantities or 260/280 ratios from the DNA extracted from skin samples using the same DNeasy® Tissue Kit (p > 0.05). The mean 260/230 ratios were higher for the cloacal DNA but not for the buccal DNA, although the differences were not statistically significant (Table 1). As a point of reference, we compared DNA yields of our study samples with mean yields calculated from data from 150 whole-blood samples archived at the US National Marine Fisheries Service Southwest Fisheries Science Center Marine Mammal and Turtle Molecular Research Sample Collection. Overall, blood samples yielded an order of magnitude more DNA than buccal, cloacal, or skin samples (Table 1).

Table 1 A summary of historical DNA extraction values from green turtle blood (Jue et al. 2009). The differences in mean DNA quantities between buccal and cloacal swabs and skin were not significant (p > 0.05, Kruskal–Wallis 1-way ANOVA).

View Fullscreen

The PCR amplification success rates of the nuclear microsatellite loci were significantly higher with DNA from the cloacal swab samples (85%–92% success) than from the buccal samples (59%–67% success; Fig. 2; χ²₁  =  4.61, p < 0.05, Kruskal–Wallis 1-way ANOVA). PCR amplification rates for mtDNA were 100% for the buccal swabs and 92% for the cloacal swabs.

View Figure

• Download as Powerpoint
• Download Figure

We evaluated the sequence quality based on their length and how well sequences could be read, as summarized in Table 2. The highest rated sequences were those that extended to 780 bp. Fair-quality sequences were readable sequences where one or both strands were short (∼ 500 bp). Poor-quality sequences were short sequences (< 500 bp) with indistinct base peaks or where no sequence was obtained. We observed different sequence quality associated with the different swabs. Cloacal swab samples had more readable sequences (extending to at least 500 bp for one or both DNA strands) while the buccal samples had a higher percentage of clean-looking sequences (little to no noise with distinct base peaks). At the same time, the buccal samples had a greater number of sequences that either failed or had indistinguishable base peaks (Table 2). Overall we obtained more consistent sequencing results from the cloacal samples than from the buccal samples.

Table 2 Sequence-quality results from buccal and cloacal swabs.

View Fullscreen

Discussion

Miller et al. (2006) previously explored how buccal and cloacal swabs resulted in DNA yields that were reliable for genotyping reptiles such as S. punctatus; however, they did not compare the difference in reliability between the two types of swabs. We originally expected that the buccal and cloacal swab samples would have far less DNA than the skin samples after extraction. Instead, we found DNA quantities for all three tissue types were not significantly different, showing that these sampling methods are comparable alternatives for when skin is not available.

We found that DNA samples extracted from buccal swabs, cloacal swabs and skin all had 260/280 values similar to that of optimal values for pure DNA (1.8–2.0, Thermo Fisher Scientific) but the 260/230 ratios for the buccal swabs were lower than optimal values (2.0–2.2, Thermo Fisher Scientific). There are two possible explanations for this finding. First, samples may have contained impurities, which may have inhibited PCR amplification or cycle sequencing. The DNA extraction protocol we used called for a short incubation time for tissue digestion (about 10 min), and this may have resulted in lower DNA yields and less filtration of lab-related or environmental impurities. Second, the accuracy of our Nanodrop® ND-1000 instrument might not be consistent at low DNA concentrations, as indicated by the variable results we found for the 260/230 ratios (Table 1). According to a study by GenVault®, the Nanodrop® was accurate for a minimum of 10 ng/µl or 1000 ng for our study (with a volume of 100 µl). In addition, the Nanodrop® was observed to accurately measure for purities with a minimum concentration of 20 ng/µl or 2000 ng (Nunez 2006). The range of DNA quantities (Table 1) in our study show that one or more samples in each tissue batch (buccal, cloacal and skin) had less than optimal DNA concentrations; this may explain the variable 260/230 ratios.

Our results demonstrate that cloacal swabs provide an acceptable method for sampling DNA from marine turtles when other standard methods are not feasible. DNA from the cloacal swabs produced longer and more consistent sequences than the buccal swab DNA, even though DNA yields and mtDNA PCR successes were similar (Fig. 2). This suggests that differences in DNA quality may be a factor.

We found that cloacal swabs had higher PCR success rates for nuclear microsatellite loci than did buccal swabs, and they also had a higher number of extracted samples that resulted in readable mtDNA sequences. We have demonstrated that extracted genomic DNA from cloacal swabs is of sufficient quality and quantity for mitochondrial DNA sequencing and PCR amplification of microsatellite markers. Genotyping of nuclear marker products should be performed to further evaluate how these collection methods might affect results, because common problems that occur with microsatellite genotyping may be associated with DNA quality. However, the DNA concentrations we report, while low relative to those obtained from blood, are still within the range commonly used in microsatellite studies of various taxa (Broquet et al. 2007).

We preferred the cloacal swabs because they were minimally invasive and only required one person to obtain the sample. To collect buccal swabs, the turtles needed to be on land with 2 people restraining the front flippers, 1 person implementing the mouth gag, and 1 person collecting the swab. We also found that cloacal swabs were preferable for genetic analyses over buccal swabs. However, the swabbing method (both buccal and cloacal) may not be favorable for all reptiles. Jones et al. (2008) discussed the hazards associated with approaching animals such as snakes and warned that taking these kinds of swabs may not be safe for field biologists.

Cloacal swabs proved to be a good alternative for genetic sampling if blood and skin sampling is not possible; however, we still recommend collecting blood or skin as the preferred sources of DNA for genetic analyses because more genetic material is collected using these methods. When extracting DNA from skin, only one-fifth of a 6-mm biopsy punch (or less, depending on which DNA extraction method is used) or 15 µl of a 2-ml blood sample is used for a single DNA extraction. For buccal and cloacal samples, the entire swab must be used in the DNA extraction process. If the initial DNA extraction fails, additional tissue leftover from blood and skin samples allows for replications of the process, if needed. With the swabs, extracting DNA for a second time would not be guaranteed because most of the cells are used in the initial extraction. Furthermore, sampling skin and blood allows for storage of subsamples for further study or other research applications (e.g., stable isotopes, contaminants, etc.). In conclusion, buccal and cloacal swabs provide limited quantities of genetic material obtainable for multiple processes, but they do serve as a good alternative method that is minimally invasive while still yielding high-quality DNA.