Extraction and amplification of DNA from aged and archaeological Populus euphratica wood for species identification

Introduction

Populus euphratica Oliv. (Salicaceae) is the dominant tree species in the desert regions of western China, Central Asia, and Mediterranean coastal countries, and it is one of the most salt- and drought-tolerant tree species. In the region of western China, the fragments of P. euphratica are frequently found as archaeological plant remains; the doubtless identification of which is important in archaeological research. However, it is often impossible to distinguish P. euphratica wood from other species of the genus Populus based on anatomical features alone. The DNA barcoding technique might overcome these limitations and might provide effective information with a high resolution. DNA barcoding is a genetic approach aimed at achieving accurate species identification based on a standard part of the DNA genome (Degen and Fladung 2007). Currently, the chloroplast genome regions, such as rbcL, matK, psbA-trnH, and the nuclear ribosomal DNA internal transcribed spacer (ITS), have emerged as good candidates for plant DNA barcoding (Kress et al. 2005; Gonzalez et al. 2009; Jiao et al. 2014). In recent years, DNA barcoding has become a reliable tool in the fields of herbal medicinal materials, quality control, forensic science, and the conservation of biodiversity (Li et al. 2011).

DNA extraction from fresh leaves or buds is a matter of routine in molecular biology. However, DNA extraction from dried wood stored for a long-term is more problematic (Finkeldey et al. 2010; Abe et al. 2011; Höltken et al. 2011; Lowe and Cross 2011; Jiao et al. 2012, 2014). After the death of an organism, hydrolytic and oxidative processes cause fragmentation and modification to its DNA (Pääbo et al. 2004; Schlumbaum et al. 2008). DNA increasingly degrades with age, resulting in the splitting of intact DNA into small fragments, ranging from 50 to 500 bp in length (Dumolin-Lapègue et al. 1999; Deguilloux et al. 2002; Speirs et al. 2009; Jiao et al. 2014). Despite all these events, short DNA fragments still contain information that may assist in confirming taxonomic identification and understanding the history of a material.

Because only limited amounts of endogenous DNA survive postmortem degradation in old specimens, the efficiency of DNA isolation and PCR amplification become critical in further DNA analysis. The studies involving the isolation and amplification of DNA in long-term stored wood or ancient wood are still rare. Speirs et al. (2009) reported the amplification of chloroplast DNA regions, smaller than 100 bp, from the oak wood samples taken from a flagship that had been preserved in a marine environment for more than 400 years. Tani et al. (2003) amplified up to 600 bp of nuclear DNA fragments from Cryptomeria japonica wood that had been buried in a dam for 3600 years, whereas Deguilloux et al. (2002) demonstrated that it was impossible to retrieve DNA segments of more than 300 bp from the sapwood of oak logs approximately 20 years after felling. Hence, there is a need to explore this field further.

The aims of the present study were (1) to evaluate the feasibility of DNA extraction from aged and archaeological P. euphratica wood based on a modified Qiagen kit protocol (Hilden, Germany), (2) to assess DNA preservation by means of amplicon length recovery, and (3) to explore the possibility of P. euphratica wood identification based on retrieved DNA barcodes.

Materials and methods

Three aged wood xylarium specimens of P. euphratica, viz., w19451, w9998, and w5314, were chosen from the Wood Xylarium Collection at the Chinese Research Institute of Wood Industry in the Chinese Academy of Forestry (Table 1; Figure 1), as well as an archaeological wood specimen of P. euphratica, M70, which was taken from a coffin excavated at Xiaohe Cemetery in Lop Nor of Xinjiang (Table 1; Figure 1).

Figure 1

Aged and archaeological wood specimens of Populus euphratica. (a) w19451; (b) w9998; (c) w5314; (d) M70.

Wood specimen w19451 was excised into small blocks [10 mm (L)×10 mm (R)×10 mm (T)] with a razor blade and then softened in water at 80°C for 5 h. Thereafter, transverse, radial, and tangential sections were cut into thicknesses of 15 μm on a sliding microtome and then observed under a light microscope (Olympus BX61, Japan) after being stained with 1% aqueous safranin.

The exposed surfaces of the wood samples were removed with a sterile scalpel to avoid external contamination. The wood samples weighing 100–200 mg were ground into fine powders in a 6770 Freezer/Mill (Spex SamplePrep, Metuchen, NJ, USA) cooled by liquid nitrogen. All DNA isolations were carried out under sterile conditions. Wood DNA was extracted following the DNeasy Plant Mini Kit (Qiagen, Hilden, Germany) protocol with modifications (Rachmayanti et al. 2006; Jiao et al. 2014). Total DNA was subsequently quantified by 1% agarose gel electrophoresis and NanoDrop 8000 Spectrophotometer (Thermo Scientific, Waltham, MA, USA).

One-way analysis of variance (SAS program 9.0) was carried out to evaluate the quantitative DNA differences among the samples.

Testing the level of DNA fragmentation in the DNA template extracted from the wood samples (Table 2) was based on five PCR primers, specific to the chloroplast DNA gene for ribulose 1,5-bisphosphate carboxylase/oxygenase large subunit (rbcL), and one primer, specific to the nuclear ribosomal DNA ITS, ranging in size between approximately 100 and 800 bp.

For the amplification of DNA from the wood samples, two-stage PCRs were carried out. PCR amplifications were carried out with 50 μl of the PCR mixture consisting of 25 μl TaKaRa Premix Ex Taq (containing 1.25 units of Ex Taq DNA polymerase, 2 mM MgCl₂, and 200 μM of each deoxyribonucleotide triphosphate; TaKaRa Biotechnology, Dalian, China), 0.4 μM of each primer, and approximately 20 ng of the template DNA. PCRs were performed using Veriti PCR (ABI, Foster City, CA, USA) with an initial denaturing step at 95°C for 5 min, 40 cycles at 94°C for 30 s, at an annealing temperature of 43°C, 45°C, 45°C, 50°C, 47°C, and 54°C for 1 min, respectively, and at 72°C for 1 min. This was followed by a final extension step at 72°C for 10 min. Subsequently, 2 μl of the PCR products were applied as the template for the second amplification with the same primers. Additionally, the negative controls without the addition of wood DNA were carried out in the same amplification protocol. The PCR products were detected in 1% agarose gels afterwards.

The PCR products were purified with the aid of the UNIQ-10 Spin Column DNA Gel Extraction Kit (Sangon, Shanghai, China) and sequenced directly in both directions or sequenced after TA cloning with the pEASY-T1 Simple Cloning Kit (TransGen Biotech, Beijing, China) on an ABI PRISM 3730xl (Applied Biosystems, Foster City, CA, USA). When the DNA region was too short for sequencing, TA cloning technology was applied. In this work, the rbcL-2, rbcL-3, rbcL-4, rbcL-5, and ITS regions were directly sequenced, whereas the rbcL-1 region was sequenced after TA cloning. Three repeated reactions of PCR amplification controls and corresponding sequencings were carried out for each sample to prevent experimental errors, such as DNA polymerase errors during amplification or sequencing errors (Dumolin-Lapègue et al. 1999).

The initial automated alignments of the sequences were obtained with Clustal X 1.81 (Thompson et al. 1997) followed by a manual adjustment using the BioEdit software (Hall 1999). The species discrimination was evaluated with tree-based analyses, which provide a convenient way to visualize the results. A neighbor-joining (NJ) tree was constructed under the P-distance model and pairwise deletion with MEGA 4. The bootstrap values were calculated using 1000 replications in MEGA 4 (Tamura et al. 2007). The taxon assignment success for each barcode, separately and combined, was determined by evaluating the proportion of the monophyletic placements of each species in the NJ trees, including the measures of their clade support.

Results and discussion

Wood anatomy

The wood anatomical characteristics of P. euphratica according to the International Association of Wood Anatomists (IAWA) list of microscopic features for hardwood identification (IAWA Committee 1989) are distinct growth rings, diffuse-porous to semi-ring-porous cross-section, solitary vessel outline is angular, radial two to six multiple vessels (or clusters) (Figure 2a); simple perforation plates, mean tangential diameter of the vessels (78±13 μm), absence of helical thickenings, and alternating intervessel pits (Figure 2c); similar vessel-ray pits and intervessel pits (Figure 2b); and the scarcity of parenchyma, uniseriate rays (Figure 2c). The uniseriate rays are 45–496 mm in height and consist of 2–27 (mostly 10–18) cells. All ray cells are procumbent (Figure 2b). However, it is impossible to distinguish P. euphratica wood from other species of the genus Populus, based alone on wood anatomical characteristics, which are very similar for all species (Cheng et al. 1992).

Figure 2

Wood anatomical features of Populus euphratica (xylarium w19451 as an example for illustration): (a) transverse section; (b) radial section; (c) tangential section. Scale bars, 200 μm (a) and 50 μm (b and c).

Sequence alignment and phylogenetic analyses

The PCR products were sequenced to confirm their identity. When the BLAST search in the National Center for Biotechnology Information nucleotide database was done, the DNA sequences obtained from specimen w19451 mostly matched the sequences of P. euphratica deposited in the GenBank, showing a 99% to 100% match with the series of DNA regions. The sequences of rbcL-1 and rbcL-2 from specimen w9998 displayed a 98% identity with the sequences of P. euphratica in the GenBank. There was a 95% match for the sequence of rbcL-1 from specimen w5314 (Table 5).

Table 5

Maximum match among sequences obtained from wood specimens of P. euphratica and sequences of P. euphratica deposited in the GenBank.

Specimen no.	DNA region	Maximum match	Accession number
w19451	rbcL-5 (756 bp)	99%	AB012778
	rbcL-4 (485 bp)	99%	AB012778
	rbcL-3 (344 bp)	100%	AB012778; KC485209
	ITS (296 bp)	100%	KC485090
	rbcL-2 (202 bp)	100%	AB012778; KC485209; JX571881
	rbcL-1 (112 bp)	100%	AB012778; KC485209
w9998	rbcL-2 (202 bp)	98%	AB012778; KC485209; JX571881
	rbcL-1 (112 bp)	98%	AB012778; KC485209
w5314	rbcL-1 (112 bp)	95%	AB012778; KC485209

To assess the possibility of barcodes on species discrimination, all single chloroplast DNA barcodes, the above-mentioned barcode combinations, and the ITS barcode were evaluated (Table 6). In the NJ tree based on these barcodes, apart from rbcL-1, rbcL-3, and the combination of rbcL-1 and rbcL-3, wood specimen w19451 could be clustered together to form a monophyletic clade with the sequences of the corresponding region of P. euphratica from the GenBank (AB012778). Furthermore, it could be separated from the other wood species (Table 6). It should be mentioned that barcodes rbcL-1 and rbcL-3 did not include enough genetic information for species identification. Moreover, in the NJ tree based on the nuclear DNA region (ITS), specimen w19451 also clustered together with the sequence of the ITS region of P. euphratica deposited in the GenBank (KC485090), which was supported by a bootstrap value of 79% (Table 6). It was reported that the nuclear DNA region (ITS) carried most of the important genetic information and that it had greater mutation rates, compared to chloroplast markers, which was much more preferable for classifying a plant species (Tsumura et al. 2011; Jiao et al. 2014). Similarly, the China Plant BOL Group recommended that the ITS region should be incorporated into the core plant barcode for seed plants, because it shows the highest discriminatory power among the four candidate markers, viz., rbcL, matK, psbA-trnH, and ITS (China Plant BOL Group 2011).

Table 6

Bootstrap values of monophyletic clades of P. euphratica species based on single and combinations of DNA barcodes.

DNA region/specimen	w19451	w9998	w5314
rbcL-1 (112 bp)	n.d.	n.d.	n.d.
rbcL-2 (202 bp)	64	n.d.	–
ITS (296 bp)	79	–	–
rbcL-1+rbcL-2 (314 bp)	72	n.d.	–
rbcL-3 (344 bp)	n.d.	–	–
rbcL-1+rbcL-3 (358 bp)	n.d.	–	–
rbcL-4 (485 bp)	78	–	–
rbcL-2+rbcL-3 (546 bp)	72	–	–
rbcL-1+rbcL-2+rbcL-3 (560 bp)	74	–	–
rbcL-1+rbcL-4 (597 bp)	85	–	–
rbcL-3+rbcL-4 (653 bp)	85	–	–
rbcL-1+rbcL-3+rbcL-4 (667 bp)	86	–	–
rbcL-2+rbcL-4 (687 bp)	98	–	–
rbcL-5 (756 bp)	89	–	–
rbcL-4+rbcL-5 (767 bp)	87	–	–
rbcL-1+rbcL-2+rbcL-4 (799 bp)	99	–	–
rbcL-2+rbcL-3+rbcL-4 (855 bp)	98	–	–
rbcL-2+rbcL-5 (958 bp)	99	–	–
rbcL-2+rbcL-4+rbcL-5 (967 bp)	99	–	–

“n.d.” indicates not distinguished and “–” indicates failure of PCR amplification.

Specimen w9998 could not be discriminated successfully, although specimen w19451 could be grouped together with the sequence of P. euphratica from the GenBank (AB012778) in the NJ tree based on the short DNA barcode rbcL-2 of 202-bp length or the combination of rbcL-1 and rbcL-2 (Table 6; Figure 3). More variable sites could be detected in the sequence of rbcL-2 from specimen w9998 compared with the rbcL-2 sequence of P. euphratica from the GenBank (AB012778) and wood specimen w19451 (Table 7). Meanwhile, the most common forms of miscoding lesions for specimen w9998 are the C→T and G→A changes in the amplification products (Table 7), which could be ascribed to the hydrolytic deamination of cytosine to uracil, resulting in a modification of the original sequence information. Staats et al. (2011) also obtained similar results; that is, increased levels of C→T/G→A transitions were observed in the plastid DNA of old herbarium specimens, compared to fresh tissues, representing 21.8% of the miscoding lesions observed. Furthermore, the sequence variations might be caused by the individual variations in the same P. euphratica species from different geographic regions.

Figure 3

Neighbour-joining (NJ) tree, based on the P-distance of the DNA barcode rbcL-2 (a) and rbcL-1+rbcL-2 (b). Numbers above the branches indicate the percentage of bootstrap values estimated from 1000 bootstrap replicates. Bootstrap values >50% are indicated at the nodes. P. = Populus.

Table 7

Nucleotide mutations of the rbcL region among the GenBank sequences and the wood herbarium specimen of P. euphratica that were analyzed.

Taxon	Region	Substitution position													Total sites	Substitution sites
		348	375	403	404	636	643	658	666	687	690	693	702	1017
AB012778	1–1398	G	G	T	C	C	C	G	C	T	A	C	A	A
w19451	238–439, 518–1283	–	–	–	–	–	–	–	–	–	–	–	–	G	967	1
w9998	238–439, 617–728	A	A	G	T	T	–	–	–	–	–	–	–	×	314	6
w5314	617–728	×	×	×	×	–	A	–	A	C	G	T	G	×	112	6

Substitutions are indicated in bold, the identical bases are indicated by “–,” and “×” indicates no sequencing.

For specimen w5314, only the short DNA fragment rbcL-1 could be amplified successfully. However, specimen w5314 could not be identified, because the DNA region rbcL-1 did not include enough genetic information for species identification (Table 6).

A short length of target DNA may lead to a successful PCR but, in general, it does not include enough genetic information for species identification (Yoshida et al. 2006; Abe et al. 2011; Tsumura et al. 2011). Thus, for significantly degraded DNA materials, the barcode regions with short lengths should be selected; they should have high recovery rates (success rates for amplifying and sequencing) and have a lot of potential for species discrimination.

Additionally, the branch support in the NJ tree generally increased when more DNA regions were combined. The combination of some DNA regions received the highest branch support (99%) in the present study (Table 6). In agreement with Liu et al. (2011), it can be stated that a combination of several short DNA barcodes might improve the power of wood species identification.

Conclusion

DNA can be obtained from wood specimens stored for approximately 80 years by using a modified Qiagen kit protocol, whereas DNA fragments of more than 100 bp can be successfully retrieved. DNA information from a specimen that is approximately 3600 years old could not be obtained using this protocol. Dilution and further purification were not helpful, either.

DNA degrades significantly in relation to the length of the function storage time. The size of the amplicon similarly shows a strong correlation with the PCR success, whereas the shortest fragments have the highest success rate. For specimens stored for decades, it might be preferable to focus on short target regions (<300 bp) for amplification. The two-stage PCR amplification technique improves PCR sensitivity and was successful in the present study.

With the aid of phylogenetic analysis, based on barcode rbcL-2 (202 bp in length), it was possible to differentiate P. euphratica from other species of the Populus genus. The nuclear DNA region (ITS) seems to be also suitable for species identification. Furthermore, combining several short DNA regions generally increased the robustness (measured by clade support) of the barcoding discrimination. The feasibility of wood identification based on DNA barcoding method will open new scenarios for the study of wood forensics, certification, archaeology, and paleobotany.