Regeneration of Glyphosate-Tolerant Nicotiana tabacum after Plastid Transformation with a Mutated Variant of Bacterial aroA gene

INTRODUCTION

Genetic information of plants is distributed among three cellular compartments: the nucleus, the mitochondria and the plastid. Each of these compartments carries its own genome and consequently expresses heritable traits (Ruf et al., 2001; Siguira, 1992). Recently, plastids have become attractive targets for genetic engineering efforts. Transformation of the plastids’ genome (plastome) has several advantages over nucleus transformation. Since plastids in most agronomically important plant species are inherited maternally, the introduction of foreign genes into the plastome prevents pollen mediated outcrossing (Bock et al., 2004). Additionally, polyploidy of the plastid genome leads to high level protein production when transformed with a transgene. Furthermore, possibility of polycistronic operon expression and the absence of epigenetic effects and gene silencing are other noticeable advantages of plastid transformation (Maliga et al., 2003; Sidorov et al., 1999). Svab and colleagues (1990) reported the first successful plastid transformation of higher plants in 1990 for Nicotiana tabacum, since then; numerous reports have successfully proved the feasibility of plastid transformation in Arabidopsis thaliana (Sikdar et al., 1998), potato (Sidorov et al., 1999), Lesquerella fendleri, a kind of oilseed Brassicacea (Skarjinskaia et al., 2003) and oilseed rape (Hou et al., 2004).
Since the beginnings of transgenic plant commercialization (1996 to 2007), herbicide tolerance has consistently been considered as the dominant trait. In 2007, a total of 114.3 million hectares was used for cultivation of biotechnological products; a portion of 63% or 72 million hectares was dedicated to herbicide-tolerant plants (International Service for the Acquisition of Agri-biotech Application, 2007). Among different herbicides, glyphosate (under the Roundup® trade name) is a broad-spectrum, safe and effective herbicide that blocks plant growth by inhibiting the production of aromatic amino acids, which leads to arrest of protein synthesis. Glyphosate inhibits 5-enoylpyruvyl shikimate-3-phosphate synthase (EPSP synthase) activity, which is a nuclear-encoded, plastid-localized enzyme in the shikimic acid pathway of plants and microorganisms (Eschenburg et al., 2002; Alibahai and Stallings 2001; Schonbrunn et al., 2001; Sost et al., 1990). Previously, a variant form of Escherichia coli K12 EPSP synthase gene with reduced affinity for glyphosate was made through simultaneous Gly96Ala and Ala183Thr substitutions (Haghani et al., 2008; Salmanian et al., 2006). The efficiency of this mutated gene for tolerance induction in model plant has also been reported (Kahrizi et al., 2007). The amino glycoside 3-adenylyltransferase (aadA) gene, which confers dual resistance to spectinomycin-streptomycin antibiotics, is still the most efficient and routinely used selectable marker for plastid transformation (Ye et al., 2001; Ye et al., 2003; Maliga, 2004). Not only the antibiotic resistant markers are not needed in the final transgenic products, but also their presence imposes some biosafety concerns for consumers. Therefore, elimination of selectable marker genes from transgenic plants is highly recommended. Different methods including marker excision (Cre-lox site-specific recombination system) and co-transformation followed by independent segregation of antibiotic and herbicide genes have been successfully used to produce marker free transgenic plants (Ye et al., 2003; Corneille et al., 2001; Hajdukiewicz et al., 2001; Zou, 2001).
This study has focused on the production of high-level glyphosate tolerant plants (N.  tabacum) through biolistic transformation of plastids by introduction of a mutated herbicide-tolerant gene coding for EPSP synthase. It has previously been reported that glyphosate, represents a lethal selection that kills plastids even in the early stages of transformation therefore, attempts were unable to produce homoplasmic, glyphosate-tolerant plants upon direct selection on the herbicide(Ye et al., 2003).  Hence, in this study, it was decided to modify the selection procedure within two manners: The selection procedure was started with sublethal doses of glyphosate and the incubation period was increased from 2 days to 1 week. As a result, glyphosate tolerant plantlets were regenerated without the use of antibiotic resistant selectable markers. For this purpose, different parameters of the chloroplast transformation procedure were optimized first by using the chloroplast transformation vector, pKCZ (Zou et al., 2003; Zou, 2001).

MATERIALS AND METHODS

Optimization of plastid transformation: In order to  evaluate and optimize all the parameters of plastid transformation, Nicotiana tabacum cultivar Samsun was transformed with the intact pKCZ vector (a generous gift from Prof. H.U. Koop; Ludwig-Maximilians Universitat, Munchen) through biolistic bombardment of well-grown young leaves. The vector harbors the aadA gene, which confers simultaneous spectinomycin/streptomycin resistance. Two regions, INSL (Insertion Site left) and INSR (Insertion Site Right), are exactly copied from tobacco plastid genome therefore; the gene of interest that could be cloned between them will be integrated in plastid genome through homologous recombination (Fig. 1). Plastid transformation of tobacco was carried out using the particle bombardment device (PDS-1000/He, Bio-Rad Biolistic system, USA). The gold particle (0.6 mm) coating procedure was performed by using standard protocols provided by the Bio-Rad manual. Different parameters such as the rupture disk pressure (ranging from 900 to 1300 psi), distance of the target tissue from the stopping screen (3, 6 and 9 cm) and vacuum pressure (ranging from 20 to 25 inches Hg) were considered. The optimum condition involved the use of 0.6 mm gold particles, 900 Psi rupture disk, a vacuum pressure of 24 inches Hg and a 6 cm distance between the stopping screen and the target tissue.  Selection was started two days after bombardment with segmented leaf pieces (0.5 × 0.5 cm) on RMOP medium (MS + 6-benzylaminopurine (BAP) 1 mg/l + 1-Naphthalene acetic acid (NAA) 0.1 mg/l) which contained spectinomycin. The antibiotic (spectinomycin) concentration was adjusted to 500 mg/l. In order to achieve homoplasmy, three rounds of selection and regeneration were performed. Each round involved approximately 2 months .During each round, the primary shoots of resistant transformants were dissected into small pieces (2.5 × 2.5 mm) and placed on new selective medium (RMOP plus spectinomycin). In order to eliminate spontaneous mutations which make the plastids ribosome resistant to spectinomycin, the regenerated plantlets were transferred onto RMOP medium containing streptomycin (500 mg/l).

PCR analysis of transplastomic plants:  The presence of foreign gene in the plastome was confirmed with PCR amplification using specific primers. One of the primers was designed on the basis of aadA gene which introduced to chloroplast genome and had the following sequence; aadAR: 5´-CACAGTGATATTGATTTGCTGG-3´, while the other primer was located on the right flanking region of the insertion site in the plastid genome, INSRF: 5´-GTAGCTCAGAGGATTAGAGCAC-3´. The latter primer was designed due to the native sequence of plastid genome.

Construction of the plastid transformation vector: pKCZ had been previously constructed for plastid transformation of tobacco (Zou et al., 2003). This vector contains the aadA gene under the control of the rRNA operon promoter (Prrn) and the 3’ UTR of the Chlamydomonas rbcl gene as a terminator (rbcl3’chl). The mutated Enolpyruvyl shikimate 3-phospho (EPSP) synthase (Salmanian et al., 2006) was isolated from pUC18-EPSP and substituted the aadA in pKCZ, using NcoI/PstI restriction endonucleases (Fig. 1).

Plant material: Sterile tobacco plants were grown on MS medium in a phytochamber (Snijder Scientific, the Netherland) (16 h light/8 h dark/25ºC). For biolistic transformation, fully expanded leaves were harvested and placed overnight on RMOP medium with the upper side up, under the above conditions. Chloroplast transformation with pKCZ containing the gene coding for EPSP synthase was performed several times with parameters, which had been optimized previously.

Selection and regeneration of EPSP synthase transformants: For the selection procedure, two different strategies were applied. In the first, two days after bombardment, leaves were cut into approximately 0.5 × 0.5 cm2 pieces and placed onto regeneration medium (RMOP) containing glyphosate as selectable marker. In this selection procedure, serial dilutions (0.1-1 mM) of glyphosate were prepared. After two weeks, non transformed (bleached) parts were removed and the remaining green explants transferred to fresh medium with the same glyphosate concentration. The second procedure of selection was carried out on sub-lethal doses of glyphosate. In this procedure, one week after bombardment, explants were cut and transferred onto RMOP medium containing 5 mM glyphosate (one tenth of lethal dose). Glyphosate concentration was doubled every two weeks up to the lethal dose (50 mM). Each time the swelled leaf explants were partitioned into smaller pieces.

RESULTS

Optimization of plastid transformation: The precultured leaves were bombarded under optimized conditions with pKCZ. Two days later, selection of the segmented explants was started on RMOP supplemented with spectinomycin as selectable marker. After several rounds of selection, the first antibiotic-resistant explants were regenerated 12 weeks after bombardment (Fig. 2). The antibiotic-tolerant plantlets were able to generate green shoots and expanded roots (Fig. 3). The well-grown plants were transferred to pots under greenhouse conditions, to allow flowering and set seeds.
In order to distinguish between spontaneously mutated and transplastomic plants, regenerated plants were transferred to RMOP medium with streptomycin (500 mg/l). Three weeks later, the transformed plantlets were still green and no noticeable adverse effect was obvious, while the wild type control was bleached completely (Fig. 4).

PCR analysis of transplastomic plants: In order to prove the correct orientation of the aadA gene, a pair of primers was designed. Amplification of the desired fragment (approximately 1300 bp) was analyzed on 1% (w/v) agarose gel (Fig. 5 A-B).

Construction of the EPSP synthase-containing vector: Replacement of the aadA gene with mutated EPSP synthase was achieved and the desired construct was prepared using the same cassette: Prrn/epsp synthase/ rbcl3’chl. The presence of foreign genes in the vector was confirmed both with restriction enzyme analysis and PCR (Data not shown).

Selection of the transformed plastid with the altered EPSP synthase gene: To improve the efficiency of selection two different selection procedures were used. In the first strategy, where 0.1 to 1 mM of glyphosate was used, leaf explants were highly sensitive and bleached completely in less than 1 month and no callus formation was observed. In the second pattern, leaf explants were cut and placed on selection medium containing 5 mM of glyphosate, one week after bombardment. Herbicide concentration was doubled every two weeks. Primary glyphosate resistant calli were selected in the presence of 50 mM herbicide. These calli were also tolerant to 0.1 mM glyphosate, but regeneration did not occur completely. Small, green plantlets were unable to form roots (Fig. 6).
DISCUSSION

Glyphosate [N-(phosphonomethyl) glycine], the active ingredient of weed control agent, Round up® is known as one of the most successful commercial broad-spectrum herbicides. The EPSP synthase has been identified as a highly selected target for glyphosate. Efforts to achieve herbicide resistance in crop plants were started in the 1980s with isolation of EPSP synthase in various glyphosate-resistant organisms (Eschenburg et al., 2002). In previous research, in order to produce more stable and active variants of this enzyme, two amino acid substitutions (Gly96Ala and Ala183Thr) were introduced simultaneously into the EPSP synthase gene (Kahrizi et al., 2007).
Glyphosate resistant crops, which are generated through nuclear transgenic technology, are currently used commercially in several countries (ISAAA 2007). So far, it has been clearly shown that transplastomic technology has tremendous advantages over nuclear transformation including high-level foreign protein accumulation and reduction in the risk of foreign gene flow into environment. Hence, in this study, efforts were made to induce glyphosate tolerance in N. tabacum through plastid transformation with the new mutated gene.
Replacement of aadA with the EPSP synthase mutated gene changed the expression cassette of the pKCZ vector. Therefore, EPSP synthase was localized under the control of the chloroplast’s rRNA operon promoter. This strong promoter causes high levels of transcript accumulation in plastids. The insertion site of the transgene is located between two tRNA genes (Asparagine and Argenine) in the inverted repeats of the plastid genome. The flanking regions of this site are exactly copied in the chloroplast vector pKCZ, therefore; each plastome could be transformed twice through homologous recombination. High Copy numbers of plastid DNA in the homoplasmic transformed plants and double transformation of each genome with the altered EPSP synthase, could produce N. tabacum with high levels of resistance to glyphosate.
The aadA gene has been used as a main selectable and efficient marker in previous plastid transformation studies. The neo and aph genes, which confer resistance to kanamycin and betaine aldehyde dehydrogenase (BADH) (that induces resistance toward toxic compounds like betaine aldehyde (BA), have also been reported as successful selection markers for transplastomic plants. Since the antibiotic resistant genes are not desirable in the final products, different strategies have been applied to eliminate the necessity of using these kinds of selectable markers (Maliga 2004, 2003).
Herbicide tolerant markers are considered as highly attractive alternatives and are currently used in nuclear transformation, but transplastomic selection only in the presence of herbicides seems to be inefficient (Mannerlof et al., 1997; Zhou et al., 1995; Barry et al., 1992). The low success of herbicides like glyphosate and phosphinothricin in achieving plastid transformations is supposed to be the result of their lethal effects on plastids. It has been claimed that their remarkable lethality hinders the small fraction of transformed plastids to stay alive and divide efficiently. Therefore, plant cells will be killed at the early stages of transformation before the resistant gene has enough time for expression. Ye and co-workers (2003) have investigated the effects of herbicides on the ultrastructure of plastids using transmission electron microscopy (TEM). The results have shown that the ultrastructures of plastids become damaged in the presence of glyphosate. Plastids lose their reticulated network of thylakoids and the photosynthetic membranes become disintegrated.
In the present research, the aim was to create transplastomic plants possessing high-level glyphosate tolerance, without using antibiotic resistance genes as selectable markers. However, after several transformation experiments using pKCZ containing EPSP synthase under optimized conditions, it was not possible to regenerate glyphosate tolerant plants on selective medium containing serial dilutions of glyphosate (0.1 mM to 1mM). Due to these findings, another selection scheme was carried out with non-lethal concentrations of glyphosate. The incubation time before selection was extended up to two weeks in order to provide enough time for plastids to amplify sufficiently and express the resistance gene. Selection was then initiated with one tenth of the lethal dose (50 mM). Media were changed each two weeks with doubled concentrations of glyphosate. By using such a selection with gradual increase in glyphosate concentrations and prolonged incubation, before initiation of selection, small, green and dense calli were generated and leaf formation was observed. These plantlets showed resistance to 0.1 mM glyphosate. The preliminary results of this study have shown that by using this altered selection procedure, it is possible to produce tolerant calli with an herbicide resistant gene as the only selectable marker. It is believed that the lack of completely formed shoots or roots can be solved with appropriate tissue culture treatments.

Acknowledgments

We highly thank Professor Hans Ulrich Koop for preparation of the plastid transformation vector, pKCZ. This work was supported by a grant # 278 provided by the National Institute of Genetic Engineering and Biotechnology (NIGEB), Iran.