Comparison of 12 DNA extraction kits for vertebrate samples

Comparison of 12 DNA extraction kits for vertebrate samples. Obtaining high quality DNA extractions is a crucial step for molecular biology research projects. At present, numerous protocols are available for vertebrate tissue extractions. In the present study we compared eleven column–based protocols and one HotSHOT protocol using similar conditions (i.e., type of sample, weight of starting material). We evaluated time of extraction, quality and quantity of DNA yield, and price of extraction for a single sample. Based on our analysis, the most successful kits for producing DNA with the highest concentration and purity are the JetQuick® Genomic DNA Purification Kit (Genomed) and the NucleoSpin® Tissue (Macherey–Nagel). Nevertheless, it is highly recommended to test various extraction kits with specific samples to find the optimal kit in all aspects of time, quality and cost for a particular project.

Regardless of the source or eventual use of the DNA, the goals of all extraction methods are the same: 1) to release genetic material from its source (fluid, tissue or microbe); 2) to stabilize nucleic acids against degradation; 3) to remove amplification inhibitors; 4) to concentrate the nucleic acid material into an appropriate volume of an aqueous solution compatible with downstream application; and 5) to standardize the methods to support accurate, sensitive and reproducible laboratory assays (Attia et al., 1996;Fox et al., 2007;Hill, 2011;Boesenberg-Smith et al., 2012).
The few previous studies that compared DNA extraction methods include: bacterial and fungal communities (e.g. Queipo-Ortuño et al., 2008;Tomaso et al., 2010;Vesty et al., 2017;Rodrigues et al., 2018); mollusc (Der Sarkissian et al., 2017); invertebrate (Kranzfelder et al., 2016;Schiebelhut et al., 2017); and ancient samples and formalin-fixed tissue samples (Rohland and Hofreiter, 2007;Janecka et al., 2015;Gamba et al., 2016). Furthermore, these studies are not comparable, as each experiment employed a different source of genetic material (e.g. different tissue, age of material, storage conditions) and different extraction kits. The extraction of DNA from vertebrate samples is usually less challenging than in the previously mentioned groups. However, the comparison of extraction kits for vertebrate samples is still lacking. Advances in Next Generation Sequencing demand the use of high quality DNA. Therefore, we decided to compare 12 DNA extraction protocols under similar conditions, including: initial weight of isolated sample, equipment, laboratory technician, and measurements of quality and quantity. The goal of this study was to compare the concentration and purity of isolated DNA, and the time and price of single extraction from vertebrate tissue.

Material and methods
For the purpose of this experiment we used 12 different extraction kits from eight manufacturers. We tried to cover the diversity of extraction kits, including those kits commonly cited in the literature ( Beside the DNA concentration and purity, we also assessed the price of kits and hand-on time demanded for single extraction. The eleven extraction kits are based on silica membrane columns and the last extraction is HotSHOT (Truett et al., 2000).  (Truett et al., 2000).

Starting material
We applied extraction protocols to four types of vertebrate samples: finger (phalange), spleen and tail from the domestic house mouse (Mus musculus) and blood samples obtained from grey partridge (Perdix perdix). Quality bias of samples (due origin, age, storage conditions, etc.) was eliminated by using only fresh samples. The tissue was weighed using Kern ABT analytical balances (Kern, ABT 120-5 DNM, resolution 0.1 mg). The average weight of finger was 1.78 mg (median 1.77 mg, 1.00-2.33 mg), spleen: 5.70 mg (median 5.95 mg, 2.08-7.70 mg), tail: 7.75 mg (median 8.00 mg, 2.54-10.30 mg). For the blood samples we used a comparable volume of clot (cca 5 ul). Details are available in table 1s in supplementary material.
To confirm the consistency of results obtained with the same protocol, three extraction kits were used: ‹4› JetQuick, ‹2› Roche, and ‹5› Qiagen to extract several samples of the same tissue type (N = 16, 12, 12 respectively, for samples of comparable weight for finger, spleen and tail).

Protocol
We followed the manufacturers' recommendations for lysis and purification of DNA (table 1). For three protocols ‹8, 9, 10› we used a homogenization step using the MagNA Lyser Instrument (Roche) and bead tubes (provided in Mobio kit). The tissue was mixed using a thermo-shaker until complete lysis. The extent of disruption (in %) was recorded after two hours (more details in table 2s in supplementary material). For protocols ‹1-11› we used the same amount of elution buffer, 100 µl for finger, and 200 µl for all other tissues. The elution step was repeated twice with the same amount of elution buffer (100 µl and 200 µl, respectively).

Repeatability of DNA extraction
The effect of tissue weight on the final DNA concentration was tested by multiple extractions of samples of a comparable size (ranging from 0.72 mg to 3 mg in weight) using three kits: JetQuick® Genomic DNA Purification Kits (Genomed) ‹4› was used for extraction of 16 samples, ‹2› High Pure PCR Template Preparation Kit (Roche) and ‹5› Dneasy® Blood & Tissue Kit (Qiagen) for extraction of 12 samples each. We reported the relative concentration as the ratio of concentration per 1 mg of weight (c/W).

Measuring
Primary verification of DNA quality was tested using gel agarose electrophoresis under the following conditions: 5 µl of DNA in 1 % agarose gel, running for 40 min. The gel was visualized by GenoPlex documentation system and GenoCapture software ( fig. 1). The isolates were subsequently assessed for quantity and quality using Quibit® fluormeter and DS-11 Spectrophotometer. We measured the 1 st and 2 nd elution (5µl of DNA for each) using Quibit® fluorometer with Qubit dsDNA BR Assay kit. DNA purity was evaluated using A 260 /A 280 ratio via DS-11 Spectrophotometer (DeNovix). Following the manufacturers' recommendation, the blank (i.e., water or elution buffer) for each extraction kit was used together with 1 µl of DNA. The A 260 / A 280 ratio measured with fluorometry should be ~1.8 (Santos et al., 2009). Values higher than 2.0 indicate basic contamination, while values lower than 1.7 relate to acidic contamination of phenol or proteins.

Cost
The cost per single extraction was calculated for the commercial kits by dividing the cost by the number of extractions. The prices are valid for 2019 in the Czech Republic.

Concentration and purity of DNA
The absolute and relative DNA concentration from the first and second elution for each tissue is shown in table 1 and figure 2. Overall, the second elution had similar or lower concentration than the first one.
The only exception was blood sample isolation using extraction kits: ‹1, ‹2›, ‹4›, ‹5›. The highest DNA yields were obtained using ‹11›, which provided constantly high DNA concentration for all used tissues. Contrary to this, kits ‹8›, ‹9› and ‹10› provided only low-concentration DNA in all cases.  table 1). The absolute concentration (ng/µl) was visualized for blood samples. In the case that only the first elution is visualised, the second elution was not provided ‹12›, or the second elution was not measurable on fluorometer (e.g. ‹8›, ‹9›, ‹10›). The DNA quality was assessed using agarose gel electrophoresis. Several cases of very low DNA concentration (less than 0.5 ng/µl; ‹8›, ‹9›, ‹10›, ‹12›) produced no visible band. Additionally, the quality/rate of fragmentation in several of the extraction kits was dependent on tissue type. For example, the highest DNA concentration was obtained from tail and this DNA also displayed extensive degradation/fragmentation visible on gel.
The results of DNA extraction repeatability are shown in figure 3. ‹4› JetQuick® Genomic DNA Purification Kits (Genomed) and ‹2› High Pure PCR Template Preparation Kit (Roche) showed similar results ( fig. 3) with concentrations ranging between 3.73 ng/µl to 3.92 ng/µl and 12.6 ng/µl and 12.7 ng/µl, respectively. Results from ‹5› Dneasy® Blood & Tissue Kit (Qiagen) displayed much higher variance. The minimal concentration of reference extraction (using the same amount of tissue) was 1.74 ng/ul, while the maximal concentration was 31.4 ng/µl.
The results of purity measurements are available in table 1. None of the tested kits showed A 260 /A 280 ratios within interval 1.8-2.0. Therefore, we defined pure DNA to be in interval 1.6-2.1. The best results according to these parameters were achieved using kits ‹4›, ‹5› and ‹7›. DNA concentration from other kits was lower than10 ng/µl, which may have resulted in an inaccurate A 260 /A 280 ratio.
Price Table 1 states the cost of DNA extraction. The price for ‹9› and ‹10› is from 2016; after this date MoBio Laboratory no longer produced the UltraClean® Tissue & Cells DNA Isolation Kit. We provide only an approximate price for ‹12› HotSHOT, because this approach uses basic chemicals, which are routinely available in molecular laboratories. The average price for a commercial kit per extraction is 4.12 € (median 2015). Nevertheless, there are extraction kits ‹1›, ‹3›, which provided 80-100 % disrupted tissue 2 hours after lysis, while proteinases from other kits worked slower and took from 6 hours to overnight to provide results. To improve the yield of DNA and reduce lysis time, it is useful to disrupt tissue with pestles or beads and use extraction kits with effective Proteinase K (e.g. kits ‹1›, ‹3›).

Discussion
The DNA quality, quantity and purity have crucial effect for downstream molecular analysis, therefore the methods of DNA extraction should be thoughtfully selected (Sagi et al., 2009). The main goal of this study was to compare DNA extraction (comprising yield, purity, cost and hands-on time) between twelve kits designed for DNA isolation. To make the comparison transparent, we defined three categories (best, average, worst) evaluating pros and cons of each kit in relation to hands-on time, DNA concentration, DNA purity and costs (table 2).

Time
One of the factors of DNA extraction is time (both the total extraction time and the time of lysis). Presented extraction kits differ in time of lysis, because of the different proteinase effectiveness. In general, longer lysis resulted in extracting higher amounts of DNA from the same amount of starting material (Janecka et al.,

Measurement (Qubit, NanoDrop)
Determining DNA concentration and purity is an important step for downstream applications, such as polymerase chain reaction. Two types of measurements are preferable: fluorometry (Qubit) and spectrophotometry (e.g. NanoDrop, DeNovix). Qubit fluorometer calculates the total amount of DNA in one sample (O'Neill et al., 2011). The spectrophotometric instrument detects all particles that absorb light at 260 nm (DNA, RNA, single or double stranded, proteins, contaminants;O'Neill et al., 2011). Fluorometer and spectrophotometer results are not always correlated. Spectrophotometry measurements are usually higher than Qubit results, indicating that the DNA sample may contain a mixture of doubleand single-stranded DNA, contaminants which scatter light, or UV-absorbing materials that are not nucleic acids (O'Neill et al., 2011). To take advantage we used both approaches for DNA quantification, Qubit values to determine concentration of double strand DNA and spectrophotometer to obtain information about DNA purity. The difference in concentration measurements between Qubit and spectrophotometer could indicate the presence of single strand DNA, RNA, proteins and/ or contaminants (in our case the biggest difference in extraction kit ‹12›). The combination of both approaches provides the most complete and correct information about DNA sample quality (Simbolo et al., 2013) the quality of DNA can vary depending on the source or extraction method applied. Thus a standardized and cost-effective workflow for the qualification of DNA preparations is essential to guarantee interlaboratory reproducible results. The qualification process consists of the quantification of double strand DNA (dsDNA, Qubit measurements as the DNA quantity indication (DNA concentration), and spectrophotometry as the information about the DNA quality (purity ratio A 260 /A 280 ).

Relationship of time, quality and price
When selecting an extraction method, many aspects are to be considered. The purity of the nucleic acid in obtained sample, the cost-effectiveness of the procedure, the duration of exposure to dangerous chemicals, the amount of hands-on time, and the related number of required steps must be taken into account (Boesenberg-Smith et al., 2012).
If the short-time extraction is needed, kits ‹8-10› or ‹12› should be used. If the high yield of DNA is the main concern, kits ‹1›, ‹4›, ‹7› or ‹11› are suitable. In case of Table 2. Comparison of the time, DNA concentrations, purity and costs between 12 extraction kits: *** short time, high concentration, good purity and good price; ** medium time, concentration, purity and price; * long time, low concentration of DNA, low purity and high price: ◊ in protocol was used Proteinase K for sample lysis. (For details see tables 3s-6s in supplementary material).
To sort this puzzle out, testing an extraction kit should belong to standard practice in any laboratory. The testing should include at least three different protocols. It is appropriate to optimize the selected extraction kit, e.g. amount of starting material, time of lysis and/or usefulness of performing 2 nd elution. When the amount of starting material is larger than recommended, it is suggested in ‹11› protocol to double the amount of lysis buffer and proteinase. If extracting limited and rare samples, then the second elution is recommended because up to 42 % of DNA is still bound to the membrane surface after the first elution (Janecka et al., 2015).

Conclusion and perspectives
Selection of the best DNA extraction kit depends on many factors: starting material, sample size, number of samples (single column extraction or 96-plate extraction), extraction time, expected concentration and purity (e.g., for microsatellites required DNA quality is lower compared to whole genome sequencing requirements) and price per one extraction. The kit choice can be adjusted according to special conditions of any samples (forensic, micro-samples, blood), price (big difference between kits), or extraction time. If many samples need to be processed, some extraction kits (Qiagen, Stratec, HotSHOT) provide the possibility to extract 96 samples in plates and thus significantly decrease the extraction time. Some kits, such as NucleoSpin (Macherey-Nagel), are forensic quality, treated to prevent DNA contamination. In conclusion, there is no best kit for any particular project, only the best solution. For this reason we strongly recommend trials be performed with 3-5 extraction kits in advance to check which kit provides the best results for specific samples. Tabla 1s. Información detallada sobre la extracción de ADN. Para cada kit de extracción tomamos nota de: W, peso de la muestra (en g); c1, concentración de la primera elución; c2, concentración de la segunda elución; c/W, concentración relativa. ( ◊ en el protocolo se utilizó Proteinasa K para la lisis de la muestra).  Table 5s. List of extraction kits with result of purity measurements obtained by spectrophotometer. Purity ratio A 260 /A 280 for finger, spleen, tail blood. * mean out of reccomended values. ( ◊ in protocol was used Proteinase K for sample lysis).