• Open Access

Mutation by DNA shuffling of 5-enolpyruvylshikimate-3-phosphate synthase from Malus domestica for improved glyphosate resistance

Authors

  • Yong-Sheng Tian,

    1. Shanghai Key Laboratory of Agricultural Genetics and Breeding, Biotechnology Research Institute, Shanghai Academy of Agricultural Sciences, Shanghai, China
    Search for more papers by this author
    • Yong-Sheng Tian and Jing Xu contributed equally to the article
  • Jing Xu,

    1. Shanghai Key Laboratory of Agricultural Genetics and Breeding, Biotechnology Research Institute, Shanghai Academy of Agricultural Sciences, Shanghai, China
    Search for more papers by this author
    • Yong-Sheng Tian and Jing Xu contributed equally to the article
  • Ri-He Peng,

    1. Shanghai Key Laboratory of Agricultural Genetics and Breeding, Biotechnology Research Institute, Shanghai Academy of Agricultural Sciences, Shanghai, China
    Search for more papers by this author
  • Ai-Sheng Xiong,

    1. Shanghai Key Laboratory of Agricultural Genetics and Breeding, Biotechnology Research Institute, Shanghai Academy of Agricultural Sciences, Shanghai, China
    Search for more papers by this author
  • Hu Xu,

    1. Shanghai Key Laboratory of Agricultural Genetics and Breeding, Biotechnology Research Institute, Shanghai Academy of Agricultural Sciences, Shanghai, China
    Search for more papers by this author
  • Wei Zhao,

    1. Shanghai Key Laboratory of Agricultural Genetics and Breeding, Biotechnology Research Institute, Shanghai Academy of Agricultural Sciences, Shanghai, China
    Search for more papers by this author
  • Xiao-Yan Fu,

    1. Shanghai Key Laboratory of Agricultural Genetics and Breeding, Biotechnology Research Institute, Shanghai Academy of Agricultural Sciences, Shanghai, China
    Search for more papers by this author
  • Hong-Juan Han,

    1. Shanghai Key Laboratory of Agricultural Genetics and Breeding, Biotechnology Research Institute, Shanghai Academy of Agricultural Sciences, Shanghai, China
    Search for more papers by this author
  • Quan-Hong Yao

    Corresponding author
    • Shanghai Key Laboratory of Agricultural Genetics and Breeding, Biotechnology Research Institute, Shanghai Academy of Agricultural Sciences, Shanghai, China
    Search for more papers by this author

Correspondence (Tel +86 21 62203180; fax +86 21 62203180; email yaoquanhong_sh@yahoo.com.cn)

Summary

A new 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS) gene from Malus domestica (MdEPSPS) was cloned and characterized by rapid amplification of cDNA ends to identify an EPSPS gene appropriate for the development of transgenic glyphosate-tolerant plants. However, wild-type MdEPSPS is not suitable for the development of transgenic glyphosate-tolerant plants because of its poor glyphosate resistance. Thus, we performed DNA shuffling on MdEPSPS, and one highly glyphosate-resistant mutant with mutations in eight amino acids (N63D, N86S, T101A, A187T, D230G, H317R, Y399R and C413A.) was identified after five rounds of DNA shuffling and screening. Among the eight amino acid substitutions on this mutant, only two residue changes (T101A and A187T) were identified by site-directed mutagenesis as essential and additive in altering glyphosate resistance, which was further confirmed by kinetic analyses. The single-site A187T mutation has also never been previously reported as an important residue for glyphosate resistance. Furthermore, transgenic rice was used to confirm the potential of MdEPSPS mutant in developing glyphosate-resistant crops.

Introduction

Glyphosate is an extensively used herbicide because of its ability to inhibit endogenous 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS; 3-phosphoshikimate 1-carboxyvinyltransferase; EC 2.5.1.19) (Bentley, 1990; Kishore and Shah, 1988; Steinrucken and Amrhein, 1980). However, some important economical crops can also be killed by glyphosate. Recent advances in genetic engineering have rendered possible the transfer of glyphosate-tolerant genes to plants to produce glyphosate-tolerant crops, thereby ensuring that only weeds can be killed by glyphosate. Thus, since the early 1980s, researchers have sought to identify a glyphosate-insensitive EPSPS gene that can be introduced into crops to provide herbicide resistance, and many EPSPS genes have been isolated and characterized (Comai et al., 1983; Fitzgibbon and Braymer, 1990; He et al., 2001; Padgette et al., 1991; Pilanee and Katherine, 2007; Sun et al., 2005). Among them, EPSPS from Agrobacterium tumefaciens CP4 (EPSPSA. tumefaciens CP4) has been used to generate currently available commercial glyphosate-resistant crops (Funke et al., 2006). However, the EPSPSA. tumefaciens CP4 gene introduced into transgenic crops is obtained from A. tumefaciens CP4 bacteria, a nonfood species isolated from a waste-fed column in a glyphosate production facility. Consequently, safety concerns are being raised. Therefore, glyphosate-tolerant genes from safe sources and appropriate for the development of glyphosate-resistant transgenic crops must be discovered.

In vitro-directed evolution through DNA shuffling has been routinely used to evolve enzymes for basic research or to improve proteins of industrial significance, such as for improved enzyme kinetics (Hecky and Muller, 2005; Hsu et al., 2004; Tomschy et al., 2000, 2002; Zhou et al., 2006), altered substrate or product specificities (Hibbert et al., 2005; Otten and Quax, 2005; Sakaue and Kajiyama, 2003; Zhao et al., 2002), as well as vaccine and pharmaceutical development (Locher et al., 2004; Whalen et al., 2001). Tian et al. (2011) obtained mutant EPSPS with increased glyphosate resistance and enzyme activity by directed evolution through DNA shuffling. Their results suggest that mutant EPSPS with improved enzymatic parameters such as high Ki (glyphosate) and normal Km [phosphoenolpyruvate (PEP)] can be obtained by DNA shuffling.

Another mutant, EPSPS from Zea mays (EPSPSZ. mays), has been recently utilized to produce the first commercial varieties of glyphosate-resistant maize (Funke et al., 2009). Thus, EPSPS from plants can also be used to generate commercial glyphosate resistance crops. Hence, in this study, a full-length EPSPS cDNA sequence from Malus domestica was cloned and characterized by rapid amplification of cDNA ends (RACE) technique. We also performed DNA shuffling on this new EPSPS gene from M. domestica (MdEPSPS) under selective pressure induced by high glyphosate concentrations. One mutant with high Ki and 50% inhibitory concentration (IC50) was identified after five rounds of DNA shuffling and screening. Furthermore, transgenic rice was used to evaluate the potential of MdEPSPS in developing glyphosate-resistant crops.

Results

Cloning and sequence analysis of the full-length cDNA of MdEPSPS

Based on the conserved regions of plant EPSPS sequences, two degenerate oligonucleotide primers (epspsF1 and epspsR1) were designed and used for gradient polymerase chain reaction (PCR) amplification of the core cDNA fragment of EPSPS from M. domestica. Following PCR amplification, an approximately 250-bp product was amplified, subcloned and sequenced. The BLAST search result revealed a 272-bp cDNA fragment showing extensive homology to EPSPS genes from other plant species. Using 3′-RACE and 5′-RACE, approximately 600- and 800-bp nested PCR products were obtained, respectively. The products were subcloned into pMD18-T vector followed by sequencing and confirmed to be a 563-bp 3′-end and a 743-bp 5′-end, respectively. The full-length cDNA sequence of MdEPSPS was 1578 bp.

BLAST of the deduced amino acid sequence of MdEPSPS revealed high homology with EPSPSs from other plant species, indicating that MdEPSPS belonged to the plant EPSPS family, which was confirmed by phylogenetic analysis. Also, two highly conserved motifs (LPGSKSLSNRILLLAAL and LFLGNAGTAMRPL) owned by all plant EPSPSs were also identified in the MdEPSPS N-terminal region (Figure S1) (Baerson et al., 2002).

DNA shuffling and sequencing

MdEPSPS was shuffled using the DNA shuffling system. Over 500 000 variant colonies were screened with the pYPX251 vector in each iteration of DNA shuffling and screening. After three initial iterations of DNA shuffling, nine clones were selected for their ability to confer tolerance to 50 mm glyphosate. Remarkably, A187T substitution repetitively occurred in the nine independent plasmids based on the sequencing results. Although these mutants conferred increasing glyphosate tolerance, they did so poorly in transgenic plants. Thus, these mutants were chosen to be the parents for the next iteration shuffling. After two more iterations of DNA shuffling and screening, one mutant (MdEPSPSmutant) was identified by its ability to restore growth in the mutant ER2799 cell on M9 minimal medium containing 60 mm glyphosate. The mutant had mutations in 10 nucleic acid sites, resulting in alterations in eight amino acids: N63D, N86S, T101A, A187T, D230G, H317R, Y399R and C413A.

Role of each altered amino acid in MdEPSPSmutant

Site-directed mutagenesis by an overlap extension PCR strategy was performed to determine the role of specific amino acid mutations in MdEPSPSmutant. The results showed that mutations T101A and A187T collectively improved glyphosate tolerance because no mutant with a reverse mutation in either A101T or T187A grew on M9 minimal medium containing glyphosate (60 mm). By contrast, cells carrying the reverse mutations of D63N, S86N, G230D, R317H, R399Y and A413C grew on M9 medium. To determine whether the mutations at the two amino acids were additive or interactive in altering resistance, three mutants (MdEPSPS101, MdEPSPS187 and MdEPSPS101/187) were created by the PCR-based staggered extension process. The two amino acid mutations proved to be additive in improving resistance because cells carrying only a single mutation in either MdEPSPS101 or MdEPSPS187 did not grow on M9 medium containing glyphosate, whereas the double mutant MdEPSPS101/187 grew well.

Kinetic properties of mutants

The wild-type gene and all mutants (MdEPSPS101, MdEPSPS187 and MdEPSPS101/187) were overexpressed and then purified in Escherichia coli. The obtained kinetic constants are listed and shown in Table 1 and Figures 1-3. MdEPSPS101/187 had a Ki 10 000-fold higher and an IC50 60 000-fold greater than MdEPSPSwild type, whereas the activity of MdEPSPS101/187 decreased by nearly 1/3-fold. To determine the role of specific amino acid mutations in the mutant, additional kinetic analyses were performed. The values of Km, a measure of the affinity for substrate, were 81, 32 and 49 μm for MdEPSPS101, MdEPSPS187 and MdEPSPS101\187, respectively. These results showed that the substrate binding affinities only slightly decreased for MdEPSPS187 and MdEPSPS101\187, whereas MdEPSPS101 showed a nearly eight-fold increase in Km for PEP. Unlike MdEPSPS, which is very sensitive to glyphosate (Ki = 0.048 μm), MdEPSPS101, MdEPSPS187 and MdEPSPS101\187 tolerate high glyphosate concentrations. Compared with MdEPSPSwild type, kcat decreased in MdEPSPS101, MdEPSPS187 and MdEPSPS101\187. Moreover, IC50 for MdEPSPS101\187 was 55 000 μm, which was greater than the IC50 value for MdEPSPS101 and MdEPSPS187. The results indicated that neither single-site mutation MdEPSPS101 nor MdEPSPS187 was sufficient to enable glyphosate resistance and maintain high affinity for the substrate PEP. Only the two mutations (A101T and T187A) must be concurrently present to confer both improved glyphosate resistance and catalytic efficiency.

Table 1. Kinetic properties of wild-type MdEPSPS and mutants
EnzymeSp act (nkat/mg)Km [PEP] (μm)kcat (per m/s)Ki [glyphosate] (μm)IC50 [glyphosate] (μm)
MdEPSPSwt15113.6 × 1040.0480.93
MdEPSPS1013.8810.9 × 1039.22.7 × 102
MdEPSPS18713323.2 × 1040.488.4
MdEPSPS 101\1879.9492.6 × 1034.8 × 1025.5 × 104
Figure 1.

Activities of MdEPSPSwild type and mutants of MdEPSPS: determination of Km for PEP. Substrate saturation curves for MdEPSPS, MdEPSPS101, MdEPSPS187 and MdEPSPS101/187 using 1 mm S3P and increasing concentrations of PEP. Data were fitted to the Michaelis–Menten equation yielding the Km values listed in Table 1.

Figure 2.

IC50 values of MdEPSPSwild type (image_n/pbi12074-gra-0001.png), MdEPSPS101 (image_n/pbi12074-gra-0002.png), MdEPSPS187 (image_n/pbi12074-gra-0003.png) and MdEPSPS101/187 (image_n/pbi12074-gra-0004.png). The IC50 values of glyphosate inhibition were determined by fitting the data to the equation Vmin + (Vmax − Vmin) /[1 + ([I] / IC50)s] yielding the IC50 values listed in Table 1, and v was determined at 1 mm PEP and 1 mm S3P with glyphosate concentrations ranging from 0.0001 mm to 100 mm. Error bars represent the SD resulting from three repetitions.

Figure 3.

Ki (glyphosate) of MdEPSPSwild type (a), MdEPSPS101 (b), MdEPSPS187 (c) and MdEPSPS101/187 (d) were determined by fixed PEP concentrations and glyphosate concentrations.

Transgenic plant selection and transcript analysis

Transgenic rice was used to evaluate the potential application of MdEPSPSmutant in developing glyphosate-resistant crops. Transgenic plants were obtained from plantlets regenerated in medium containing 30 mg/L hygromycin and further confirmed on selective medium containing 0.2 mm glyphosate. Then, nine real MdEPSPSmutant transformants and seven MdEPSPS transformants were obtained on plates containing glyphosate. Then, two wild-type MdEPSPS (Md1 and Md3) and two MdEPSPSmutant (Mu2 and Mu5) transgenic lines were analysed for gene expression by reverse transcription (RT-) PCR analysis as previously described (Xu et al., 2010ab). First, rice actin gene from the four transgenic lines (Md1, Md3, Mu2 and Mu5) and wild-type rice were amplified with two primers TactZ and TactF to ensure that the same amount of cDNA was used to amplify the target genes in the transgenic lines. Then, the specific DNA fragments (about 1500 bp) of MdEPSPS and MdEPSPSmutant were amplified from the transgenic lines. Agarose gel electrophoresis showed that the DNA intensity of MdEPSPS and MdEPSPSmutant in the four different transgenic lines was preliminary the same (Figure 4), confirming that the level of transcription of MdEPSPS and MdEPSPSmutant in the two transformants was equal. The RT-PCR results also showed that the inserted genes were actively and stably transcribed in the transgenic plants.

Figure 4.

Expression of MdEPSPSwild type and mutants of MdEPSPS cDNAs in different transformants. Each lane contained 4 μL of RT-PCR products obtained using total RNA extracted from 3-week-old plants grown under normal conditions. WT, wild type; Md, transgenic rice lines with MdEPSPS gene; Mu, transgenic rice lines with MdEPSPSmutant gene; M1, markers (DL2000, Takara); M2, markers (DL15000, Takara). Data shown are representative of three independent experiments.

To estimate the copy number of MdEPSPS and MdEPSPSmutant on the genomic DNA of Md1, Md3, Mu2 and Mu5, SYBR Green I real-time fluorescence quantitative PCR was carried out at the same time. The result revealed that an exogenous gene occurred in Md1, Mu2 and Mu5 with a single copy number in the rice genome (Table 2). The MdEPSPS-transgenic line Md1 and MdEPSPSmutant-transgenic lines Mu2 and Mu5 were chosen for further experiments.

Table 2. Estimation of copy number of wild-type MdEPSPS and mutants gene in each transgenic line
Transgenic lineCt value of SPSCt value of MdEPSPSInitial copy number of SPSInitial copy number of MdEPSPS2*MdEPSPS/SPSCopy number of MdEPSPS
CK22.8628.651 349 08720 9100.0310
Md127.1027.4482 67945 6801.1051
Md325.7224.50205 154301 6792.9413
Mu224.8325.28368 654182 8520.9921
Mu526.4226.76129 38370 7081.0931

Assay for the glyphosate resistance of transgenic rice

Glyphosate can affect seed germination. Seeds are usually inhibited and bleached under glyphosate stress (Figure S2) (Dun et al., 2007). Figure 5 shows that MdEPSPSmutant-transgenic plants germinated well in 500 μm glyphosate, whereas CK and MdEPSPS-transgenic plants did not germinate. A previous study has shown that sublethal concentrations of glyphosate inflict visible damage on leaves (Zhou et al., 2006), indicating that damaged leaves are a distinguishing characteristic of plants exposed to sublethal glyphosate concentrations. Figure 6 shows that 14 days after herbicide application, most of the leaves of the wild-type rice and MdEPSPS-transgenic plants wilted, became overly dehydrated and then died. Only little damages were observed in the MdEPSPSmutant-transgenic line Mu5 plants, and no damage was observed in MdEPSPSmutant-transgenic line Mu2 plants, which grew well with normal morphology. These results also indicated that MdEPSPSmutant-transgenic plants were more resistant to Roundup exposure than CK and MdEPSPS-transgenic plants.

Figure 5.

Comparative germination image of rice seeds in ½ MS medium containing various glyphosate concentrations (0, 500, 1000 and 2000 μm) in Petri dishes.

Figure 6.

Photographs of transgenic rice sprayed with 2.5% (v/v) solution of the herbicide Roundup at a dose of 10 L/ha. 1, CK; 2, Md1 (transgenic rice lines with wild-type MdEPSPS gene); 3, Mu2 (transgenic rice lines with MdEPSPSmutant gene); 4, Mu5 (transgenic rice lines with MdEPSPSmutant gene). The same results were obtained in three independent experiments and are represented by the effects shown here.

Discussion

This study aimed to identify an EPSPS or mutant EPSPS with properties appropriate for the development of transgenic glyphosate-tolerant plants. However, wild-type MdEPSP is not suitable for the development of transgenic glyphosate-tolerant plants because of its poor kinetic properties. Hence, we performed DNA shuffling on MdEPSPS gene, and one highly glyphosate-resistant mutant with eight amino acid variations were isolated after five rounds of DNA shuffling and screening. Among the eight amino acid substitutions on this mutant, only two residue changes were identified by site-directed mutagenesis as essential and additive in altering resistance.

Many active sites of EPSPS have already been extensively studied and identified (Healy-Fried et al., 2007). Thus, the mutation of these amino acids in active sites can significantly change glyphosate tolerance. Another amino acid mutation that is not located in active sites but at a position distant from the active site can also alter glyphosate resistance, such as the ‘hinge’ region between two EPSPS globular domains and the helix region (second and third) in the N-terminal domain (He et al., 2003; Kahrizi et al., 2007). Two new mutations of MdEPSPS, namely T101A and A187T, that also affect glyphosate resistance have been identified in the present study. To understand better the functions of the two amino acid substitutions, the two mutations in this mutant were aligned with different known EPSPS (Figure 7) and distributed throughout the structure models of MdEPSPS based on the crystal structures of E. coli EPSPS (Figure S3).The three-dimensional structures suggested that T101A (corresponding to T97 in E. coli, Figure 7) was situated in the third helix of the core of the N-terminal domain, a universal mutation hotspot for glyphosate resistance such as the Gly96Ala and Pro101Leu substitution (Eschenburg et al., 2002; Zhou et al., 2006). Thus, T101A mutation may have induced a change in the glyphosate-binding site by a shift in the G96, resulting in improved glyphosate resistance. The mode of action has been elucidated in X-ray crystallography studies by Funke et al. (2009).

Figure 7.

Partial alignment of the amino acid sequence of PhEPSPS, ZmEPSPS, EcEPSPS, MdEPSPSwild type and MdEPSPSmutants.

In contrast to T101A, A187T substitution did not occur at an active site or mutation hotspot region but at the starting point of the second helix in the N-terminal domain. With respect to the mechanism, clarification is required on whether this substitution of A187T directly participates in ligand binding or glyphosate resistance. Alternatively, upon replacement of Ala 187 by Thr in MdEPSPSmutant, some conformational changes may have altered the α-helix, causing it turn towards the active centre of the enzyme, thus easily binding to PEP and shikimate-3-phosphate (S3P) but not to glyphosate. Therefore, the affinity of the enzyme for glyphosate decreased. This mechanism was almost the same as the putative substitution of A183T (numbering in E. coli EPSP synthase) located at the exterior of the second helix in the N-terminal domain (Kahrizi et al., 2007).

The Thr101 (corresponding to T97 in E. coli) residue is strictly conserved in EPSPS genes, and mutations of Thr101 have never been observed in field-evolved glyphosate-resistant weeds because the single-site Thr101 variant enzyme confers decreased catalytic efficiency. In the present work, Thr101 mutations occurred through a DNA shuffling selection system by simulating the natural conditions for evolving glyphosate resistance when first established in A187T mutations. This phenomenon indicated that the establishment of A187T mutations was required for the occurrence of T101A mutations. These results were in accordance with a previous conclusion that under the selective pressure caused by the presence of high glyphosate concentrations, plants or bacteria with already established ‘advantage’ mutations may favourably acquire the additional mutation of Thr97, which confers a very high level of resistance (Funke et al., 2009). Moreover, the findings reveal that only the simultaneous mutation of T101A and A187T provides the conformational changes necessary for high resistance to glyphosate.

The development of plant resistance to the herbicidal compound glyphosate has been a goal in the engineering of many plant species. In 1983, the first glyphosate-resistant tobacco plants were reported by Comai et al. (1983). Since then, many plant species have been engineered for glyphosate resistance, including soybean, maize, canola, cotton, sugar beet and alfalfa (Dill et al., 2008). However, reports on glyphosate-resistant transgenic rice are limited. In the present study, we presented a successful generation of novel MdEPSPSmutant transgenic rice. Our results showed that MdEPSPSmutant can confer transgenic rice plants with high tolerance to glyphosate. The evidence for this can be described in two ways. Firstly, a two-step screening of the transformants was carried out. Sixty putative transformants were obtained on hygromycin-containing plates, and six real transformants were obtained from these 60 putative transformants on plates containing glyphosate, which was further confirmed by molecular and biochemical characterization. Secondly, rice transformation with the MdEPSPSmutant gene was stable and heritable in transgenic plants, which was confirmed by a glyphosate resistance assay in T3 transgenic rice. The resistance level of MdEPSPSmutant-transgenic plants was higher than the agricultural application level recommended by most manufacturers. Therefore, we concluded that in the future, MdEPSPSmutant can be applied to transgenic crops with glyphosate tolerance.

Experimental procedures

Bacterial strains, chemicals and plant materials

S3P, glyphosate (free acid form), PEP and Ni2+-NTA agarose affinity column were purchased from Sigma Chemical Co., Ltd. (St. Louis, MO). All other chemicals were analytical grade and purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). All primers were purchased from Shanghai Sangon Biological Engineering Technology and Service Co., Ltd. (Shanghai, China). An RNA isolation kit was purchased from Fermentas International, Inc. (Burlington, Canada), and a Reverse Transcription System was purchased from Promega (Madison). A SMARTTM RACE cDNA amplification kit was purchased from Clontech (Palo Alto, CA). pMD18-T vector was purchased from Takara Co., Ltd., Dalian, China. E. coli BL21 (DE3) was obtained from Novagen, Inc. (Madison, WI), and E. coli strain ER2799 (Chen et al., 2001) with the EPSP synthase gene deleted from its genome was provided for free by Dr. Thomas C. Evans, Jr. (New England Biolabs, MA). A. tumefaciens GV3101 and rice cultivar (Oryza sativa L. ssp. japonica) were obtained from our laboratory. Young leaves were collected from M. domestica grown in a greenhouse in the Shanghai Academy of Agricultural Sciences, China and used as the starting material for RNA isolation (Liao et al., 2004).

Cloning of MdEPSPS full-length cDNA by RACE

Single-strand cDNAs were synthesized from total RNA with an oligo(dT)17 primer and reversely transcribed according to the manufacturer's protocol. After RNase H treatment, the single-strand cDNA mixtures were used as templates for PCR amplification of the conserved region of EPSPS from M. domestica. Two degenerate oligonucleotide primers, epspsF1 (5′-ANG, ANG, GVA, GNT, GDA, GBT, NTC-3′) and epspsR1 (5′-ANC, TBG, CHC, TNG, ANG, CVT, NGC-3′), were designed according to the conserved sequences of other EPSPS genes and used for the amplification of the core cDNA fragment of MdEPSPS by standard gradient PCR amplification (from 52 °C to 60 °C). The PCR products were purified and subcloned into pMD18-T vector followed by sequencing. The core fragment was subsequently used to design the gene-specific primers for the cloning of full-length cDNA of MdEPSPS by RACE.

Then, a SMART™ RACE (BD Biosciences Clontech, Palo Alto, CA) cDNA amplification kit was used to clone the 3′-end and 5′-end of MdEPSPS cDNA. The first-strand 3′-RACE-ready and 5′-RACE-ready cDNA samples from M. domestica were prepared according to the manufacturer's protocol and used as templates for 3′-RACE and 5′-RACE, respectively. The 3′-end of MdEPSPS cDNA was amplified using two 3′-gene-specific primers and the universal primers provided by the kit. For the first PCR amplification of 3′-RACE, MdepspsW3 (5′-ATG, ACT, TTG, AAG, TTG, ATG, GAA, CGC -3′) and UP3 (universal primer, 5′-GGT, GGT, AGA, GCT, CGC, AGG, ACT, GCA, GCT, GAC, TG-3′) were used as the first PCR primers, and 3′-RACE-ready cDNA was used as the template. For the nested PCR amplification of 3′-RACE, MdepspsN3 (5′-ATC, GGT, TTT, TGA, TCC, AAG, GAG, GTC-3′) and NUP3 (nested universal primer, 5′-AGA, GCT, CGC, AGG, ACT, GCA, GCT, GAC, TGA, CTA, C-3′) were used as the nested PCR primers, and the products of the first PCR amplification were used as templates. The 5′-end of MdEPSPS cDNA was amplified using two 5′-gene-specific primers and the universal primers (UP5 and NUP5) provided by the kit. For the first PCR amplification of 5′-RACE, MdepspsW5 (5′-TCC, ACA, TAC, GGA, ATG, GAA, ATT, AG-3′) and UP5 (5′-GGT, GGT, AGG, ATC, CGA, CCA, GTG, GTA, TCA, ACG, CAG-3′) were used as the first PCR primers, and 5′-RACE-ready cDNA was used as the template. For the nested PCR amplification of 5′-RACE, MdepspsN5 (5′-ATA, AGC, AAA, GCA, GTC, AAG, TAC, TG-3′) and NUP5 (5′-AGG, ATC, CGA, CCA, GTG, GTA, TCA, ACG, CAG, AGT, AC-3′) were used as the nested PCR primers, and the products of the first PCR amplification were used as templates. The nested 3′-RACE and 5′-RACE products were purified and subcloned into pMD18-T vector, followed by sequencing. By aligning and assembling the sequences of 3′-RACE, 5′-RACE and the core fragment, the full-length cDNA sequence of MdEPSPS was deduced.

According to the deduced MdEPSPS cDNA sequence, two gene-specific primers, MdepspsS1 (5′-AAG, GAT, CCA, TGG, CCC, AAG, TGA, GCA, AAA, TCT, G-3′) and MdepspsS2 (5′-AAG, AGC, TCT, CAA, TGC, TTG, GTA, AAC, TTC, CTG-3′), were designed, synthesized and used to clone the coding sequence of MdEPSPS by RT-PCR using the 3′-RACE-ready cDNA as the template.

Sequence analysis of MdEPSPS gene

The nucleotide sequence, deduced amino acid sequence and open reading frame encoded by MdEPSPS were analysed. Sequence comparison was conducted through database search using the BLAST program (NCBI, National Center for Biotechnology Services, http://www.ncbi.nlm.nih.gov). The phylogenetic analyses of MdEPSPS and EPSPS genes from other species were aligned with ClustalW (1.82) using default parameters. A phylogenetic tree was constructed using MEGA version 2.1 from ClustalW alignments by the neighbour-joining method.

DNA shuffling

DNA shuffling to generate improved glyphosate-tolerant mutants was performed as described by Stemmer (1994) and Xiong et al. (2007). The wild-type MdEPSPS gene was fragmented with DNase I, and fragments from 50 bp to 100 bp were collected in a dialysis bag as they eluted from 10% (w/v) polyacrylamide electrophoresis gels. The fragments were subjected to PCR reassembly without primers. Then, the ‘primerless’ PCR products were subjected to PCR with the specific primers P1Z (5′-GAG, AGA, GGA, TCC, ATG, GCC, CAA, GTG, AGC, AAA, ATC, TGC-3′) and P1F (5′-GAG, AGA, GCT, CTC, AAT, GCT, TGG, TAA, ACT-3′) to amplify the MdEPSPS to full length. After the primerless PCR and primer PCR, a group of full-length MdEPSPS mutants were obtained and digested with Bam HI and Sac I enzymes. The isolated fragments were ligated into the prokaryotic expression vector pYPX251 (Xiong et al., 2007). The mutant DNA library was translated into E. coli EPSPS mutant strain ER2799 by electroporation and plated on M9 agar plates supplemented with glyphosate at high concentrations. The MdEPSPS mutant was screened and identified for its ability to restore the growth of mutant ER2799 cells in M9 minimal medium containing glyphosate at high concentrations. The MdEPSPS mutants were pooled and used as templates to generate further mutants by DNA shuffling. The selected EPSPS mutants were sequenced and subsequently analysed.

Site-directed mutagenesis

To determine the role of specific amino acid mutations in the mutant, site-directed mutagenesis was performed. The process involved using an overlap extension PCR strategy to reverse each of the seven amino acid substitutions from mutant to wild type (Peng et al., 2006). The primers used were as follows: D63N, (5′-ACC, CTT, GGG, CTG, AAT, GTT, GAA-3′) and (5′-TTC, AAC, ATT, CAG, CCC, AAG, GGT-3′); S86N, (5′-TTT, CCT, TTG, AGT, AAT, GAA, TCA-3′) and (5′-TGA, TTC, ATT, AGT, CAA, AGG, AAA-3′); A101T (5′-CTT, GGA, AAT, GCT, GGA, ACA, GCA, ATG, CGG-3′) and (5′-CCG, CAT, TGC, TGT, TCC, TGC, ATT, TCC, AAG-3′); T187A, (5′- GCT, TTG, CTT, ATG, GCA, GCT, CCT-3′) and (5′-AGG, AGC, TGC, CAT, AAG, CAA, AGC-3′); G230D, (5′-GGT, GCG, GAT, GCT, GAT, TGT, TTT-3′) and (5′-AAA, ACA, ATC, AGC, ATC, CGC, ACC-3′); R317H, (5′-CGG, TTT, TTG, ATC, CAA, GGA, GGT-3′) and (5′-ACC, TCC, TTG, GAT, CAA, AAA, CCG-3′); R399Y, (5′-TCT, GGA, GGA, AAA, CAC, TTG-3′) and (5′-CAA, GTG, TTT, TCC, TCC, AGA-3′); as well as A413C, (5′-AAA, ATG, CCA, GAT, GTT, GCC, ATG, ACT-3′) and (5′-AGT, CAT, GGC, AAC, ATC, TGG, CAT, TTT-3′). The amino acid mutation that best contributed to glyphosate resistance was also determined. For this purpose, three amino acid mutants were created using a PCR-based staggered extension process. Three mutants were found to have individual amino acid mutations (MdEPSPS101 and MdEPSPS 187), and one was mutated at the former's two amino acid residues (MdEPSPS101/187) (Zhao and Zha, 2006).

Protein overexpression and purification

The wild-type gene and all mutants digested with BamHI and SacI were cloned into the expression vector pYM4087 (Xu et al., 2010a). They were then expressed and introduced into the competent expression host E. coli BL21 (DE3). Bacteria were then plated on Luria–Bertani agar [1% (w/v) Bacto-tryptone, 0.5% (w/v) yeast extract and 1% (w/v) NaCl] supplemented with ampicillin (100 μg/mL) and X-Gal (100 μg/mL). The culture was incubated at 37 °C overnight. Several positive colonies were selected and inoculated into 50 mL of Luria–Bertani medium containing 50 mg/L ampicillin. They were grown at 37 °C until the optical density at 600 nm reached 1.0. Cells were harvested and disrupted by sonication. The soluble fraction was loaded onto a Ni2+-NTA agarose affinity column at 4 °C according to the manufacturer's instructions. The purified protein was analysed by 12% (w/v) sodium dodecyl sulphate–polyacrylamide gel electrophoresis.

Enzyme assay

EPSPS activity was determined at 28 °C in 50 μL reaction mixtures (50 mm N-2-hydroxyethylpiperazine-N-2′-ethanesulfonic acid, 1 mm S3P, 1 mm PEP and 0.75 μg of purified enzyme) by measuring the amount of inorganic phosphate produced in the reaction using the malachite green dye assay method (Lanzetta et al., 1979). The reaction was allowed to proceed for 3 min to 5 min before addition of 800 μL malachite green/ammonium molybdate colourimetric solution. Colour development was stopped after 1 min by adding 800 μL of sodium citrate solution (34% w/v). After allowing to stand at room temperature for 30 min, the mixture was measured for absorbance at 660 nm. Reaction mixtures without S3P served as controls.

The Km values for PEP, the Ki values for glyphosate and the IC50 value for glyphosate were measured as described by Tian et al. (2010).

Plant expression vector construction and plant transformation

Construction of the plant expression vector and plant transformation were performed as described by Xu et al. (2010ab). To ensure that the transgene was targeted to the chloroplast, the fragment encoding the chloroplast transit peptide (TSP) was added to the constructs (Della-Cioppa et al., 1986). Accordingly, the chloroplast TSP of Z. mays (emb: X63553.1) was inserted downstream of the Z. mays polyubiquitin-1 promoter ZmUbi-1 obtained by PCR using primers ZmUbiF (5′- GCGAAGCTTG CATGCCTACAGTGCAGCGTGACCCGGTCGTGC-3′) and ZmUbiR (5′-GTGGGATCCTCTAGAGTCGACCTGCAGAGTAACACCAAACAACAG-3′) (Christensen et al., 1992). The fused fragment and MdEPSPS were cloned into the Agrobacterium binary vector pYF7716 using Hind III-BamHI and BamHI-SacI restriction sites. The final constructs ZmUbi-1:TSP: EPSPS:Nos were introduced into A. tumefaciens LBA4404 (Clontech, Palo Alto, CA) by electroporation. The constructs were introduced into the rice cultivar (O. sativa L. ssp. japonica) using seed-derived callus. Fresh calluses were infected for 30 min and were co-cultivated with the A. tumefaciens strain harbouring pYF7716ZmUbi-1:TSP: EPSPS:Nos for 3 days at 25 °C in the dark, followed by culturing on selective NB media containing 30 mg/L hygromycin and by selective NB media containing 0.2 mmol/L glyphosate. After a third or fourth round of selection, the callus pieces produced glyphosate-resistant cells. The green yellow granular embryos were carefully transferred to a flask containing fresh medium for about 1 week of cellular proliferation. The embryos were then transferred to a differentiation medium [containing 2.5 mg/L 6-BA, 0.2 mg/L NAA and Murashige and Skoog (MS) medium]. The rice seedlings were transferred to a rooting medium (containing 0.1 mmol/L glyphosate, 0.2 mg/L NAA and 1/2 MS) when they differentiated and grew to 3–4 cm. After the rice seedlings completely rooted, they were planted in the greenhouse.

Transgenic plant selection and transcript analysis

The transgenic nature of the plants was further confirmed by PCR analysis of genomic DNA using the specific primers MdepspsS1 (5′-AAG, GAT, CCA, TGG, CCC, AAG, TGA, GCA, AAA, TCT, G-3′) and MdepspsS2 (5′-AAG, AGC, TCT, CAA, TGC, TTG, GTA, AAC, TTC, CTG-3′). RT-PCR was used to determine the level of MdEPSPS or MdEPSPSmutant transcription. To improve the reliability of RT-PCR, rice actin gene (GenBank: X16280) served as an internal standard to normalize the amount of cDNA and amplified with primers OsactZ (5′-TCC, ATC, TTG, GCA, TCT, CTC, AG-3′) and OsactF (5′-GTA, CCC, GCA, TCA, GGC, ATC, TG-3′). Specific DNA fragments (~1500 bp) of MdEPSPS or MdEPSPSmutant were then amplified from the transgenic plants using the same amount of cDNA. PCR products were separated on 2% (w/v) agarose gels and quantified using a Model Gel Doc 1000 (Bio-Rad, Hercules, CA). The expression patterns of MdEPSPS or MdEPSPSmutant genes were evaluated with a Shine Tech Gel Analyzer (Shanghai Shine Science of Technology Co., Ltd., China).

SYBR Green I real-time PCR was performed to determine the copy number of exogenous gene in transgenic rice, as described by Yang et al. (2005). The sucrose phosphate synthase gene (SPS) in rice was selected as an endogenous reference gene. The specific primers were designed as follows: OsSPS-F (5′-TTG, CGC, CTG, AAC, GGA, TAT-3′), OsSPS-R (5′-CGG, TTG, ATC, TTT, TCG, GGA, TG-3′), MdEPSPS-F (5′-AAA, ACC, CTT, GGG, CTG, AAT-3′) and MdEPSPS-R (5′-CAA, AGG, AAA, CCG, ACC, ACC-3′). PCR was carried out in Rotor-Gene 3000 Real time Thermal Cycler (Corbett Research, Mortlake, NSW, Australia) in 20 μL reaction mixtures containing 4 μL of diluted DNA sample, 3 μL of diluted sense primers, 3 μL of diluted reverse primers and 10 μL SYBR Green Real-time PCR Master Mix (TOYOBO Biotechnology, Osaka, Japan). Each sample was quantified in triplicate with the same DNA template.

Seed germination assays and glyphosate resistance spray assay

For the seed germination assay, T3-sterilized rice seeds were directly grown on half-strength MS medium (Murashige and Skoog, 1962) containing various glyphosate concentrations (0, 500, 1000 and 2000 μm) in Petri dishes under a controlled-environment chamber (25 °C, 10 : 14 h light/dark cycle). To observe the germination process, photographs were taken after 2 weeks of growth.

For glyphosate spray treatment, plants were cultured in a nutrient solution culture in a controlled-environment chamber (25 °C, 10 : 14 h light/dark cycle). After 3 weeks, the seedlings reaching 12–15 cm in height were sprayed with 2.5% (v/v) solution of the herbicide Roundup containing 41.0% glyphosate isopropylamine salt (Monsanto Inc., Montreal, QC, Canada) at a dose of 10 L/ha.

Acknowledgements

The research was supported by the Key Project Fund of the Shanghai Municipal Committee of Agriculture (No. 2009-6-4); The Key Project Fund of the Shanghai Municipal Committee of Agriculture (No. 2011-1-8); International Scientific and Technological Cooperation (2010DFA62320, 11230705900); National Natural Science Foundation (31071486) and the Key Project Fund of Shanghai Minhang Science and Technology Committee (2012MH059).

Ancillary