Simple, fast and inexpensive hot sodium hydroxide and tris DNA extraction method for genotyping tomato and melon seeds

In modern agriculture, the sowing of improved hybrid seeds (F1) is widespread. In the case of horticultural crops such as tomato and melon, F1 seeds are commercialized in a highly profitable and competitive global market. These F1 seeds are obtained from controlled crosses between two selected parental lines, and their market value is dependent on their hybrid genetic purity. In this context, reliable seed purity tests are of crucial importance.

DNA analysis via molecular markers is a reliable, convenient and effective approach for fast and cost-effective varietal identification and seed purity tests. These tests are routinely implemented in regular seed production and commerce. Plants are typically genotyped from leaf tissue. For this sampling process to happen, the plant needs to grow to the cotyledonary stage or, ideally, to its first true leaf. However, this sampling process is costly in time, space and human resources when performed at an industrial level. Therefore, performing genetic screenings directly on seeds or early emerging roots (radicles) would be ideal for the characterization of seed lots. However, the main pitfall resides in obtaining DNA extracted from seed tissue. Seeds are rich in reserve components such as lipids, oils, proteins, polysaccharides and polyphenols, which can hamper DNA extraction [1]. Several commercial seed DNA extraction kits are available (e.g., Sbeadex™ plant kit for seed extractions from LGC Biosearch Technologies (Hoddesdon, UK), Quick-DNA Plant/Seed 96 kit from Zymo Research (CA, USA), Extract-N-Amp™ Seed PCR Kit from Sigma-Aldrich Corporation (MO, USA), DNeasy Plant Pro and Plant Kits from Qiagen (Hilden, Germany)); however, they are expensive for large screenings. Alternative laboratory methods tend to be long, laborious, use toxic reagents or result in all these problems combined [2–6].

The hot sodium hydroxide and tris (HotSHOT) method [7] is commonly implemented for DNA extraction from various vertebrate and insect tissues [8–10]. However, it is not routinely used in plants. The HotSHOT technique presents several advantages. First, it is simple, inexpensive, fast, nontoxic and easily scalable. Second, it requires only basic laboratory equipment. Third, it does not require liquid nitrogen. Finally, the product recovered is directly usable in downstream genotyping protocols such as competitive allele-specific PCR (KASP). In this article, a modification of the HotSHOT protocol linked to KASP genotyping for inexpensive, reproducible and high-throughput seed purity tests from tomato and melon seed lots is presented (Figure 1).

F1 seed lots were obtained from specific crosses of parental lines in tomato (Solanum lycopersicum) and melon (Cucumis melo). Tomato seeds were germinated directly in a 96-well U-bottom microtiter plate (1 seed per well) with 30 μl of distilled or tap water and allowed to germinate fully (Figure 1B). Melon seeds did not fit on microtiter plates for germination and, therefore, they were germinated on a Petri plate with moist filter paper and later sealed with parafilm. After full germination, small-, medium- and large-size radicle pieces were transferred to a 96-well U-bottom microtiter plate to evaluate the impact of input tissue quantity on the final PCR genotyping results (Figure 1C). Full germination is typically achieved in dark conditions after 5–10 days at 28°C for tomato and 4–6 days at 26°C for melon. For tomato, wells containing ungerminated seeds were annotated and excluded from further genotyping analysis as they tend to produce unreliable results. In the case of melon, only properly germinated radicles were transferred from the Petri plate to the microtiter plate for extraction.

Samples were homogenized in 100 μl of lysis solution (25 mM NaOH; pH 12) with one 4-mm stainless steel bead, covered with a cap mat for the microtiter plate and placed in a vertical homogenizer at 1500 r.p.m. for 2 min or until root tissue was homogenized. Stainless steel beads are cheap and easily available as ball bearings in hardware stores. Stainless steel (420-grade) is strongly recommended for its corrosion resistance (beads remain undamaged in the lysis solution for long periods of time) and magnetic properties, which allow easy bead recovery. Stainless steel beads can also be reused in several DNA extractions after proper cleaning [11]. The process is as follows. First, dirty balls were incubated for 15 min at room temperature in a decontamination solution made of 10% household bleach, 1% NaOH, 1% Fairy^® dish soap or a similar product and 90 mM sodium bicarbonate. They are then washed with tap water and later rinsed with distilled water and autoclaved. After homogenization, cap mats were carefully removed to avoid splashing and contamination between wells. Then, plates were sealed with ELISA plate stickers and incubated at 70°C for 30–50 min. Although the original HotSHOT protocol requires near-boiling incubation temperatures (95°C) [7], it is crucial to keep the incubation temperature at 70°C in this step, as polystyrene microtiter plates may not resist higher incubation temperatures. It is equally important to substitute the cap mats, which tend to deform at high temperatures, with ELISA plate stickers during incubation. Finally, samples were neutralized by adding 100 μl of neutralization solution (10 mM TrisHCl; 0.5 mM EDTA; pH 8) and mixing at 450 r.p.m. for 30 s either manually or in a horizontal mixer. At this point, crude DNA extraction is completed. This crude DNA solution can be used downstream immediately or stored at -20°C until further molecular analysis. The quantity and quality of recovered DNA was low. Average concentration values estimated by absorbance at 280 nm ranged between 20 and 50 ng/μl. A280/A260 ratios were between 1.5 and 1.6 while A260/A230 ratios were below 0.5 (see Table 1 for an example of typical DNA quality values following this procedure). These results were expected as small quantities of tissue are required as input material and no cleaning or precipitation steps are performed. Nevertheless, crude DNA extracts are suitable for KASP amplification.

To test the efficacy of DNA extraction for seed genotyping purposes, three molecular markers per seed lot were analyzed. For tomato seed lot 21/079, markers solcap_snp_sl_37097, solcap_snp_sl_9856 and solcap_snp_sl_36224 were selected from the KASP™ assay library (Figure 2A [12]). For melon seed lot 22/027, markers CMPSNP466, CMPSNP855 and CMPSNP579 were selected from Esteras et al. (Figure 2B) [13]. In both cases, the KASP assays were designed by LGC Biosearch Technologies (Hoddesdon, UK). Crude DNA obtained from HotSHOT extraction was used as a template. The DNA plate was centrifuged at 4°C and 3500 r.p.m. for 4 min to pellet debris (when stored at -20°C, it was thawed first). Then, 1 μl or 2.5 μl of DNA solution was pipetted into a 384-well or 96-well PCR plate, respectively. The DNA solution may be viscous, and that is why it is important to pipette carefully to avoid the tip clogging. The PCR plate was centrifuged briefly to collect the extract at the bottom of the plate and dried in a stove at 50°C for 15–30 min. Typically, multiple plates with a dispensing function multichannel pipettor were filled at a time, then dried and stored long term at -20°C for further analysis. Finally, the PCR mix was directly loaded into the PCR plate containing the dried DNA and a PCR test was conducted. Each PCR assay plate included three positive controls for allelic discrimination, three no-template negative controls and 90 target/unknown samples to be genotyped. Amplification and final reads were run in a QuantStudio 5 thermal cycler (Thermo Fisher Scientific, MA, USA) under a standard KASP protocol for a total of 42 cycles (LGC Biosearch Technologies). Allelic discrimination plots were done in QuantStudio Design and Analysis Software 1.5.2 (Thermo Fisher Scientific) with default settings for fluorescein (FAM), hexachlorofluorescein (HEX) and carboxy-X-rhodamine (ROX; used for normalization) fluorophores.

Tomato seed lot 21/079 parentals were monomorphic for markers solcap_snp_sl_36224 and solcap_snp_sl_37097 and polymorphic for solcap_snp_sl_9856, the latter being informative for seed purity test. As expected, F1 seed genotyping results were homozygous for allele 2 for markers solcap_snp_sl_36224 and homozygous for allele 1 for solcap_snp_sl_37097 (Figure 2A). However, solcap_snp_sl_9856 showed 92.3% of heterozygous and 7.7% of allele1 homozygous samples. This suggests a possible self-pollination event or the presence of 15.4% of male or female parentals heterozygous for this locus. Another polymorphic marker not linked to solcap_snp_sl_9856 should be evaluated to elucidate these possibilities.

Regarding melon seed lot 22/027, molecular markers CMPSNP466 and CMPSNP579 were monomorphic between parentals, whereas CMPSNP855 was polymorphic. Consequently, F1 genotyping results for markers CMPSNP466 and CMPSNP579 were homozygous for alleles 2 and 1, respectively, and heterozygous for CMPSNP855. In addition, different radicle sizes (small, medium and large) were evaluated as starting material for DNA extraction (Figure 1C). Allelic discrimination assays revealed that a small amount of input material in the HotSHOT DNA extraction tends to give more accurate and tighter clusters in the genotyping assay (Figure 2B).

This process is very simple since DNA extraction only requires basic laboratory equipment and two homemade buffers. In addition, all consumables and reagents are inexpensive, common laboratory materials. Furthermore, it is easily scalable as several plates can be processed in parallel allowing a high number of extractions to be performed in less than 1 h. The crude DNA extract is readily usable for genotyping by a KASP method; however, other PCR-based genotyping methods could be explored. The implementation of this method could be advantageous for screening and purity test analysis of large seed batches.