To modify the wood properties suitable for commercial use by genetic improvement, it is essential to understand the molecular basis of wood formation. Wood is the end product of cambial cell activity, which includes cell division, differentiation, secondary wall formation and programmed cell death (PCD) (Chaffey, 1999). This PCD process is related with the secondary wall deposit (wood development) which should differ from other forms of PCDs occurred in other plant organs and tissues (Xu et al., 2000; Balk et al., 2003; Ito et al., 2002; Mellerowicz et al., 2001). Therefore, understanding the mechanism of PCD process will provide new knowledge for the improvement of wood property. In the process of PCD, nuclease is one of the important players (Sugiyama et al., 2000; Brown et al., 1987; Marchetti et al., 2001; Mittler et al., 1995), thus more studies are needed to obtain detailed information about its role in wood development in woody plants. A Ca²⁺-dependent DNase gene from Populus tomentosa, named PtCDD, was cloned in our laboratory and highly expressed in immature xylem. To further characterize this gene such as to identify its activity and cellular location, it is necessary to express it in bacteria and to purify it for further antibody raising. However, expression of the whole coding sequence of PtCDD was attempted in E. coli but not successful in our laboratory. In this study, PET-30b (+) prokaryotic expression system was used to express the truncated PtCDD gene in a form of a fusion protein. The fusion protein was purified and its activity was analyzed. This result provides the basis for further research on localization and function of PtCDD in developing xylem tissues.

1 Materials and methods 1.1 Materials 1.1.1 Strains and vectors

E. coli strains DH5α, BL21 (DE3) were purchased from Tiangen Biotech (Beijing) Co., Ltd. PET-30b (+) prokaryotic expression vector was purchased from Merck.

1.1.2 Enzymes and main reagents

PCR reagents were produced from Takara Biotechnology Co., Ltd; Gel DNA fragment purification kit and plasmid purification kit were manufactured by AXYGEN; Primer synthesis was done by Beijing Aoke Biotech Co.; Restriction enzymes and T₄ DNA ligase were purchased from Beijing NEB Co..

1.2 Methods 1.2.1 Amplification of PtCDD 5′DNA fragment

The cDNA was previously prepared from Populus tomentosa cambium zoon in this lab. PCR primers were designed for amplification of 5′ PtCDD fragment based on the sequence published in GenBank (AY819661) and embedded with kpn Ⅰ (start codon) and EcoR Ⅰ (stop codon) sites used for insertion into PET-30b (+). The sequences of the 5′- and 3′- primers were GGTACCATGGGAAATGCCCTGAGATTC and CCGGAATTCCCCGATATCCTGCTTCAACAA, respectively. The reaction system included 2.5 ng cDNA, 1 μL primers (10μmol·L^-1), 2.5 μL 10 PCR Buffer, 2 μL dNTP mixture (2.5 mmol·L^-1), 1 μL Taq (5 U·μL^-1), and with distilled water to 25 μL in the final volume. The amplification was performed with the following sequence: 5 min pre-degeneration at 94 ℃, then 35 cycles of 30 s degeneration at 94 ℃, 30 s annealing at 59 ℃, and 40 s extension at 72 ℃, finally 7 min at 72 ℃ in the last step. A 5 μL PCR product was used in 1% agarose gel electrophoresis and gel isolation of the target band.

1.2.2 Construction of prokaryotic expression vector

The purified product was cut with kpn Ⅰ and EcoR Ⅰ, cloned into pGEM-T vector cut with the same enzymes and the construct was transformed into the cells of E. coli strain DH5α. The transformed cells were selected and verified by PCR as described by Joseph et al. (2001) and further confirmed by sequence analysis. The pGEM-T plasmids with target gene were purified by plasmid purification mini kit, digested with EcoR Ⅰ and Sac Ⅰ, and the gel-isolated target fragments were ligated into pET-30b (+) vector cut with the same enzymes and introduced into E. coli DH5α. The transformants were selected by PCR on plasmid DNAs from boiled cells in 10 μL distilled water and confirmed by digesting plasmids with kpn Ⅰ and EcoR Ⅰ. The pET-30b (+)-PtCDD plasmids were used to transform E.coli BL21 (DE3) for expression of the insert.

1.2.3 Optimization of the expression condition

Overnight culture of BL21 cells with pET-30b(+)-PtCDD were inoculated to LB media with the ratio of 1:50 (v/v) and incubated in a shaker (210 r·min^-1) for 2 h at 37 ℃, IPTG then was added into the culture media to make the final concentration at 0, 0.2, 0.4, 0.6, 0.8, 1.0 mmol·L^-1 and the culture was incubated for additional 0, 1, 2, 3, 4, 5, 6 h respectively for each treatment. The cells were harvested for each treatment by centrifuging 10 mL of culture for 10 min at 2 000 r·min^-1, and resuspended in PBS at 1:10(v/v) The latter was ultrasonicated in ice bath and both the supernatant and pellet were collected after centrifugation for 10 min at 12 000 r·min^-1.

1.2.4 SDS-PAGE electrophoresis

12% separation gel was prepared according to the published protocol (Sambrook et al., 2002). The supernatant and pellet samples were mixed or dissolved with loading buffer, the gel was stained by Coomassie brilliant blue R250 as described in the protocol.

1.2.5 Purification and enzyme activity analysis of the fusion protein

The pellets prepared as described above were dissolved with the PBS containing 6 mol·L^-1 guanidine-HCl and used for purification of the expressed fusion protein using the TALON Affinity Column as described in the manual. The purified products were resolved in 12% SDS-PAGE as prepared above but with 50μg·mL^-1 salmon DNA in the gel. The DNA-SDS-PAGE gel was rinsed 2 times with 25 mL isopropyl alcohol and 75 mL 10 mmol·L^-1 Tris-HCl (pH7.5) for 40 min to remove the SDS, then 2 times with 100 mL 10 mmol·L^-1 Tris-HCl (pH7.5) for 40 min and finally soaked overnight into the 10 mmol·L^-1 Tris-HCl (pH7.5) containing 10 mmol·L^-1 CaCl₂ for assay of the activity of the fusion protein (In-gel DNA digestion). The gel was observed and photographed under the ultraviolet after staining with EB (0.1 mg·L^-1) for 30 min. The purified fusion proteins of poplar DNase were also used to digest poplar genome DNAs at 37 ℃ for 2 h. The reaction system was set up by mixing 100 ng genome DNA, 10 mmol·L^-1 Tris-HCl (pH7.5), 10 mmol·L^-1 CaCl₂, 0, 1, 2, 4, 8 and 16 ng fusion protein for each reaction. The digested products were analyzed by 1% agarose gel electrophoresis.

2 Results and analysis 2.1 PCR amplification of 5′-PtCDD fragment

The PtCDD gene fragments were amplified by PCR from poplar cDNAs and checked in agrose gel. The specific band was observed on the gel with the expected size at 637 bp (Fig. 1). The fragments were isolated from gel and used for the construction of expression vector.

2.2 Construction of expression vector

The amplified fragments were digested with EcoRⅠ and SacⅠ and inserted into PET-30b (+) to construct the PET-30b-PtCDD. This expression vector was confirmed by digestion with KpnⅠ and EcoRⅠ to produce the 637 bp target DNA fragment (Fig. 2, lane 3), and subsequently by sequence analysis of the insert which showed 98.88% identity to the published sequence (AY819661) after alignment (data not shown). This indicated that the cloned fragment was the 5′ terminus of the PtCDD gene.

2.3 Optimization of the expression conditions

The yield of the fusion protein of PtCDD varied slightly with IPTG at different concentrations (Fig. 3) for induction. However, the time length for the IPTG induction was quite important for the yield. As shown in Fig. 4, the maximum yield was achieved at 5 h after induction by IPTG, and this observation was further confirmed in another 2 independent experiments (data not shown), therefore a 5-h induction period was used to express PtCDD to obtain the fusion protein.

2.4 Location of the fusion protein expressed

The bacterial cells were broken by ultrasonic and the supernatant and pellet were analyzed by 12% SDS-PAGE electrophoresis. Interestingly, the specific band of recombination protein only existed in the pellet (Fig. 5, lane 1), this indicated that the expressed protein mainly accumulated in the form of inclusion body.

2.5 Purification and the activity assay of the fusion protein

The inclusion bodies formed by the expressed fusion protein were dissolved in the 6 mol·L^-1 guanidine-HCl denaturant and used for purification with the TALON Affinity Column. The purified protein was observed in a major band on SDS-PAGE gel (Fig. 6, lane 5). The activity of the fusion protein in the same SDS-PAGE gel with DNA was analyzed by removing SDS and soaking in the CaCl₂ containing buffer. The place of the fusion protein (Fig. 6, lane 5) appeared a dark band (Fig. 6, lane 1) after staining with EB (Negative image), indicating the DNA was digested by the fusion protein.

To investigate the DNase activity to poplar genomic DNA, purified fusion protein at different quantities mixed with genomic DNA and the digested products was analyzed with 1% agarose gel electrophoresis (Fig. 7).

With the increased amount of fusion protein added, the amount of DNAs left on the gel became less and finally invisible. This provided additional evidence that the expressed truncated PtCDD had the ability to digest DNA, though the efficiency may be weakened based on the observation that it took 2 h for 16 ng purified fusion protein to fully digest the 100 ng genomic DNAs.

3 Discussion

Most of the DNases involved in PCD identified in plants were Zn²⁺ dependent (Thelen et al., 1989; Fukuda, 1992; Fukuda, 1997; Aoyagi et al., 1998). In contrast, a Ca²⁺ dependent DNase was found in PCD process in wood formation in our lab, suggesting that this DNase has a unique role in PCD. Therefore, further investigation of the role would help to elucidate the PCD process in wood forming tissues. Accordingly, to obtain this enzyme in high quantity is essential for further characterization of this novel Ca²⁺ dependent DNase.

The full-length cDNA was obtained and expressed in E.coli in our lab, but the expressed protein could not be found from the cell extracts. It is likely that the PtCDD protein has a powerful DNase activity thus digests plasmid DNA and probably also genomic DNA of E. coli, leading to the failure of expression of the full-length gene. To solve this problem, we amplified the 5′ fragment of the PtCDD gene only including the catalytic region and part of Ca²⁺ binding region, omitting the 3′ terminus of the gene. As expected, the truncated gene was successfully expressed in E. coli cells. This indicates that the deletion of both the DNA binding domain and the larger part of the Ca²⁺ binding region could decrease the enzyme activity significantly, thus the DNase can be expressed in large quantity in E. coli. This explanation has been supported by the observation that 16 ng purified protein was needed to fully digest 100 ng DNA (Fig. 7).

To test if the truncated PtCDD still has the DNase activity, we have purified the fusion protein and DNA-SDS-PAGE was used to analyze the DNase activity. The DNase activity was steadily detected although its digestion ability was weakened. This suggests that the deletion of DNA binding domain and the larger part of the Ca²⁺ binding region could decrease but not completely remove the enzyme activity of the truncated PtCDD. The success in obtaining of PtCDD with activity paves the way to raise antibody for further analysis of its role in PCD in wood forming tissues, which is essential to understand the molecular mechanisms underpinning the wood formation process.