Plants contain many secondary metabolites, including phenolics, polysaccharides, and glycoproteins, that can interfere with DNA extraction, PCR, and cycle sequencing. Multiple extraction protocols aimed at reducing or removing inhibitory compounds (e.g., Olmstead and Palmer, 1994; Setoguchi and Ohba, 1995; Hughey et al., 2001; Drábková et al., 2002; Malvick and Grunden, 2005), or attenuating their effects on PCR efficiency (Saunders, 1993; De Boer et al., 1995), have been developed, but these are often not effective. Certain taxa pose significant challenges to successful PCR and sequencing, even in cases where purified genomic DNA is used. While rbcL is generally considered an easy region to amplify and sequence with standard primers (Kress and Erickson, 2007; Hollingsworth et al., 2009), the first author had problems amplifying or sequencing the gene from multiple medicinal plants from Pakistan, including Amaranthus sp. (Chenopodiaceae), Anethum graveolens L. (Apiaceae), Butea monosperma (Lam.) Taub. (Fabaceae), Fagonia indica Burm. f. (Zygophyllaceae), Senna sp. (Fabaceae), and Trachyspermum ammi (L.) Sprague (Apiaceae). DNA was extracted from silica-dried or air-dried samples before PCR was attempted with regular Taq polymerase. An extraction of Anethum, prepared from fresh material, amplified and sequenced cleanly, suggesting that secondary metabolites in the dried material (which were not effectively removed with a commercial DNA purification kit), rather than suboptimal PCR parameters (identical for both samples), reduced amplification efficiency and inhibited sequencing. Inhibitors in dried material pose a serious challenge because fresh tissue is often not available.
The problem of poor PCR efficiency can be addressed at the polymerase level. Kapa Biosystems (Woburn, Massachusetts, USA) recently developed an enzyme with specific tolerance to common plant inhibitors. “KAPA3G” DNA Polymerase was derived from a previously engineered, more processive variant of Taq DNA polymerase (processivity reflects the average number of nucleotides added by a DNA polymerase per association/dissociation event with the template; processive enzymes synthesize DNA more quickly and are more efficient in the presence of inhibitors). In short, a randomized gene library of the parental “KAPA2G” DNA polymerase gene was generated and expressed in E. coli. Individual bacterial cells, each containing both the expressed, mutant DNA polymerase protein, as well as the gene encoding that variant, were compartmentalized in a water-in-oil emulsion (Griffiths and Tawfik, 2006). In this system, each mutant enzyme was required to amplify its own gene in the presence of secondary metabolites derived from several different plant species. After several rounds of selection with increasing levels of inhibition pressure, gene variants coding for polymerases with improved tolerance to plant inhibitors were exponentially enriched over variants with no advantage. The KAPA3G DNA Polymerase that was evolved in this manner was blended with a small quantity of an engineered high-fidelity enzyme, to allow for the efficient amplification of DNA fragments >5 kb from plant samples. KAPA3G was effectively tested for PCR with purified plant DNA, crude plant extracts, and in direct PCR from leaf discs or seeds of a variety of crop plants ( Appendix S1 (APPS_1200519_AppendixS1.pdf)) before the KAPA3G Plant PCR Kit was released. This report constitutes the first study of the effectiveness of KAPA3G DNA Polymerase using dried material from noncrop plants, and documents the potential advantages of the KAPA3G Plant PCR Kit for a wide range of species. A PCR optimization is presented to aid researchers in selecting the appropriate annealing temperature and MgCl2 concentration for their specific assays. PCR results with the KAPA3G enzyme are compared to those with regular Taq polymerase.
Comparison of rbcL sequencing data quality for Linum usitatissimum, Anethum graveolens, and Senna sp. using Taq polymerase and the KAPA3G PCR Kit with and without Plant Enhancer.
METHODS AND RESULTS
Three DNA extracts that produced varying degrees of amplification and sequencing success with regular Taq polymerase were chosen for initial rbcL optimization (primers 1F [Fay et al., 1997] and 1460R [Fay et al., 1998; Cuénoud et al., 2002]) with the KAPA3G Plant PCR Kit. See Appendix 1 for voucher information for all species included in the study. Wild collections were not georeferenced at the time of collection. Mini-extractions for Linum usitatissimum L. (Linaceae) and Anethum graveolens (Apiaceae) (both silica-dried) and Senna sp. (Fabaceae) (air-dried) were prepared using a standard cetyltrimethylammonium bromide (CTAB) protocol (Doyle, 1991) and purified using the UltraClean 15 kit (MO BIO, Carlsbad, California, USA). PCR for rbcL, matK, and psbA-trnH had been attempted using ReadyMix PCR master mix with Taq polymerase (Sigma, St. Louis, Missouri, USA). The following thermal cycler program was used for rbcL and matK PCR with Taq polymerase: 94°C 5 min; 30 cycles: 94°C 1 min, 48°C 1 min, 72°C 1 min; 72°C 7 min. All three regions were successfully amplified and sequenced for Linum, but rbcL failed to sequence for Anethum and did not amplify for Senna, although the psbA-trnH spacer was sequenced for both. Linum was selected for the KAPA3G Plant PCR Kit evaluation as it had amplified and sequenced with Taq polymerase, while Anethum was chosen because it had amplified but failed to sequence, and Senna as it had not amplified at all. The KAPA3G Plant PCR Kit includes an optional Plant Enhancer, a reducing agent that improves amplification efficiency for some types of samples through an unknown mechanism. Two sets of reactions were run for each taxon, one with 0.5 µL (1×) Enhancer and one without. Each reaction contained the KAPA3G Plant Buffer (1× final concentration, includes dNTPs at 0.2 mM each), MgCl2 (2 mM final concentration), 1 unit DNA polymerase, primers at a final concentration of 0.3 µL each, and PCR-grade water to bring the volume to 50 µL. An annealing temperature gradient PCR was performed, in increments of 4°C from 50°C to 70°C, using a Veriti Thermal Cycler (Applied Biosystems, Carlsbad, California, USA) and the following cycling parameters: 95°C 10 min; 40 cycles: 95°C 20 s, 50–70°C [gradient] 15 s, 72°C 90 s; 72°C 90 s. The gradient PCR identified the highest temperature at which amplification was successful for all samples (58°C). To test the amplification quality, six of the best PCR products (corresponding to the brightest bands in a 1% agarose gel) were selected for sequencing: Linum, Anethum, and Senna generated with an annealing temperature of 58°C, with and without Enhancer. The best overall PCR product (Senna without Enhancer, generated with an annealing temperature of 62°C) was also sequenced for comparison. PCR products were cleaned with the Wizard SV Gel and PCR Clean-Up System (Promega Corporation, Madison, Wisconsin, USA). DNA sequences were generated at Ohio University's Genomics Facility and analyzed using an ABI 3130x1 Genetic Analyzer (Applied Biosystems, Carlsbad, California, USA). Each sequencing reaction included 2 µL 5× buffer (Applied Biosystems), 0.5 µL dimethyl sulfoxide (DMSO; Sigma), 0.5 µL BigDye (Applied Biosystems), 0.1 µL ThermoFidelase (Fidelity Systems, Gaithersburg, Maryland, USA), 100–40 ng template DNA, and PCR-grade water for a total volume of 8 µL. Cycle sequencing products were cleaned with the BigDye XTerminator Purification Kit (Applied Biosystems). Phred Q20 values (Ewing et al., 1998) were used as an initial indication of sequence quality. External rbcL primers 1F and 1460R, and internal primers 636F and 724R (Fay et al., 1997), were used for sequencing.
PCR and sequencing success of 31 species for rbcL and matK using Taq polymerase or the KAPA3G Plant PCR Kit.
Results of the PCR optimization are shown in Fig. 1. Amplification was successful at annealing temperatures from 50–62°C, although amplification at 62°C was reduced or failed when Enhancer was present. More product was produced without Enhancer, but more nonspecific amplification occurred. Senna, which did not amplify for rbcL using Taq polymerase, amplified strongly using the KAPA3G enzyme. Sequencing results for rbcL primers 1F, 636F, 724R, and 1460R are presented in Table 1, with partial rbcL 1F chromatograms in Appendices S2 (APPS_1200519_AppendixS2.pdf) and S3 (APPS_1200519_AppendixS3.pdf). Sequence data for Linum 1F and 636F were of higher quality from PCR products using Taq polymerase, whereas sequence data for all other taxa were of higher quality with the KAPA3G enzyme. With the exception of Anethum 1460R and Senna 724R at an annealing temperature of 58°C, sequence data were of a higher quality from samples without Enhancer. This suggested that residual Enhancer (carried through PCR clean-up) may have inhibited the cycle sequencing reaction. However, the results of a second optimization did not support this conclusion. A second round of optimization for rbcL was performed with the Linum, Anethum, and Senna extracts to reduce nonspecific amplification, although no significant improvements were observed for these particular species. See Appendices S4 (APPS_1200519_AppendixS4.pdf) and S5 (APPS_1200519_AppendixS5.pdf) for the protocol and results, which tested the effects of different thermal cycling programs, MgCl2 concentrations, and the presence/absence of Enhancer.
Extracts of an additional 31 species from 23 different families, prepared with the same methods outlined above, were tested for rbcL (Table 2), first with Taq polymerase, and then with the KAPA3G enzyme using the optimized cycling program with an annealing temperature of 58°C. Nine out of 31 samples (29%) amplified and sequenced for rbcL with Taq, whereas 21 out of 22 samples (95%) that failed with Taq amplified and sequenced with the KAPA3G enzyme (1.5 mM MgCl2, no Enhancer). This success rate is much higher than the best rbcL PCR rate (26%) reported by Särkinen et al. (2012) for several different DNA polymerase enzymes, although extracts from much older herbarium specimens were used in their study. The same initial optimization outlined above was performed for matK 390F/1360R (Cuénoud et al., 2002), and an annealing temperature of 54°C was selected for this assay. Three out of 26 samples (12%) amplified and sequenced with Taq polymerase for matK while 21 out of 28 samples (75%) amplified and sequenced with the KAPA3G enzyme (Table 2). A higher concentration (2 mM) of MgCl2 was required for successful PCR of eight of these species. A few samples (e.g., Hygrophila, Urtica) did not amplify for one or both gene regions with the KAPA3G enzyme. These samples were characterized by abundant mucilage during the extraction process, and purifying the genomic DNA did not remove all the mucilage. A nonmucilaginous extract prepared from seeds (market sample) of Hygrophila did amplify successfully for matK (but not rbcL) with the KAPA3G enzyme after PCR with Taq polymerase failed. An Urtica extract prepared with the PowerPlant Pro DNA Isolation Kit (MO BIO) amplified readily for rbcL (but not matK) at 1.5 mM MgCl2 with the KAPA3G enzyme (Table 2). For certain species of Lamiaceae (Ajuga, Mentha, Ocimum, results not shown here), successful amplification of rbcL and matK was achieved with the KAPA3G enzyme from dirty pellets (not purified after CTAB extraction), while other species (Lycopus, Nepeta, Origanum) failed to amplify until genomic DNA had been purified or extracted with the PowerPlant Pro DNA Isolation Kit. Taken together, these results suggest that while the KAPA3G enzyme offers much higher success rates than Taq polymerase, PCR from plant samples remains challenging in the presence of high levels of inhibitors, particularly when primers are not perfectly matched to target sequences.
This study demonstrated that the KAPA3G Plant PCR Kit successfully amplified DNA from extracts that failed with Taq. Quality sequence data were obtained from species from 24 different families. The variable results obtained with Taq polymerase and the KAPA3G Plant PCR Kit indicate that PCR success and sequence quality may be as much a function of the taxon as the methodologies used. Differences in secondary metabolites presumably account for some of this variation. Although the KAPA3G Plant PCR Kit did not always lead to high-quality sequence data, it effectively amplified DNA that failed to amplify with Taq polymerase. The KAPA3G Plant PCR Kit can therefore be a very useful tool for plant biologists working with difficult taxa that have failed to amplify with Taq polymerase. We recommend using the optimization protocol (Appendix 2) to select the best annealing temperature for a specific assay, and then performing the PCR with 1.5 mM MgCl2 and no Enhancer. If the PCR fails, increasing the MgCl2 concentration (2 mM) and/or adding Enhancer should be tried as these proved to be critical for successful PCR for certain taxa.
 The authors thank V. Nadella and R. Yoho at the Ohio University Genomics Facility for technical assistance. Sample kits for optimization purposes were supplied by Kapa Biosystems. This research was funded by grant PGA-P210852 from the National Academy of Sciences to A. M. Showalter. Maryke Appel works for Kapa Biosystems.