Molecular Ecology (2004) 13, 697– 710
doi: 10.1046/j.1365-294X.2004.02092.x
Tandem constructs to mitigate transgene persistence: tobacco as a model
Blackwell Publishing, Ltd.
H A N I A L - A H M A D ,* S H M U E L G A L I L I † and J O N A T H A N G R E S S E L * *Plant Sciences, Weizmann Institute of Science, Rehovot 76100, Israel, †Agronomy and Natural Resources Department, Agricultural Research Organization, The Volcani Center, Bet Dagan 50250, Israel
Abstract Some transgenic crops can introgress genes into other varieties of the crop, to related weeds or themselves remain as ‘volunteer’ weeds, potentially enhancing the invasiveness or weediness of the resulting offspring. The presently suggested mechanisms for transgene containment allow low frequency of gene release (leakage), requiring the mitigation of continued spread. Transgenic mitigation (TM), where a desired primary gene is tandemly coupled with mitigating genes that are positive or neutral to the crop but deleterious to hybrids and their progeny, was tested as a mechanism to mitigate transgene introgression. Dwarfism, which typically increases crop yield while decreasing the ability to compete, was used as a mitigator. A construct of a dominant ahasR (acetohydroxy acid synthase) gene conferring herbicide resistance in tandem with the semidominant mitigator dwarfing ∆gai (gibberellic acid-insensitive) gene was transformed into tobacco (Nicotiana tabacum). The integration and the phenotypic stability of the tandemly linked ahasR and ∆gai genomic inserts in later generations were confirmed by polymerase chain reaction. The hemizygous semidwarf imazapyr-resistant TM T1 (= BC1) transgenic plants were weak competitors when cocultivated with wild type segregants under greenhouse conditions and without using the herbicide. The competition was most intense at close spacings typical of weed offspring. Most dwarf plants interspersed with wild type died at 1-cm, > 70% at 2.5-cm and 45% at 5-cm spacing, and the dwarf survivors formed no flowers. At 10-cm spacing, where few TM plants died, only those TM plants growing at the periphery of the large cultivation containers formed flowers, after the wild type plants terminated growth. The highest reproductive TM fitness relative to the wild type was 17%. The results demonstrate the suppression of crop–weed hybrids when competing with wild type weeds, or such crops as volunteer weeds, in seasons when the selector (herbicide) is not used. The linked unfitness would be continuously manifested in future generations, keeping the transgene at a low frequency. Keywords: dwarfism, gene introgression, herbicide resistance, tandem constructs, tobacco, transgenic mitigation Received 25 July 2003; revision received 14 November 2003; accepted 14 November 2003
Introduction There is a concern that transgenes may escape from engineered crops into related weedy species by hybridization and backcrossing (Snow 2002). This could potentially result in hybrids
Correspondence: Prof. Jonathan Gressel. Fax: + 972-8-934-4181; E-mail:
[email protected] © 2004 Blackwell Publishing Ltd
and their progeny with enhanced invasiveness or weediness (Mikkelsen et al. 1996; Snow & Palma 1997; Ellstrand et al. 1999; Dale et al. 2002). Many of the engineered genes, such as those conferring resistance to herbicides, diseases and environmental stresses, may grant a fitness advantage to a weed or a wild species, especially when the crop and the weed species grow in the same agricultural ecosystem. Several crops (e.g. wheat, barley, sorghum, rice, squash, sunflower, sugarbeets, oats and oilseed rape) can naturally
698 H . A L - A H M A D , S . G A L I L I and J . G R E S S E L interbreed with closely related wild or weedy relatives under field conditions, in both directions (Ellstrand et al. 1999; Gressel 2002). The addition of transgenic traits has been proposed as valuable for most of these economically important crops, especially herbicide resistance to control weedy relatives of the crop that heretofore could not be chemically controlled. Such transgenic crops could become volunteer weeds in fields during subsequent years, or even become feral, potentially re-evolving the traits of their weedy progenitors while retaining the selective advantage of the transgene(s).
Gene containment Molecular mechanisms have been proposed for containing transgenes within crops (i.e. to prevent interbreeding) or to mitigate the effects of transgene flow once it has occurred (Gressel 1999, 2002; Daniell 2002). One containment mechanism utilizes partial genome incompatibility for crops, such as wheat and oilseed rape, having multiple genomes derived from different progenitors. When only one of these genomes is compatible for interspecific hybridization with weeds, the risk of introgression could be reduced if the transgene was inserted into the unshared genome. This mechanism was ineffectual in the case of oilseed rape (Tomiuk et al. 2000). Another proposed containment is to integrate the transgene in the plastid or mitochondrial genomes (Khan & Maliga 1999; Maliga 2002). The opportunity of gene outflow is limited due to maternal inheritance of these genomes but the claim of no paternal inheritance (Daniell et al. 1998; Bock 2001) was not substantiated. Tobacco (Avni & Edelman 1991) and other species (Darmency 1994; Gressel 2002; Wang et al. 2004) often have a frequency of between 10 −3 and 10 −4 of pollen transfer of plastid-inherited traits. This is quite inadequate as a mechanism to prevent gene flow because such frequencies are orders of magnitude greater than the frequency of most mutations. Additionally, direct transfer of chloroplast DNA into the nucleus could occur, as found in tobacco (Huang et al. 2003), although functional chloroplast gene establishment in the nucleus is rare. Yet another possible solution to preclude outcrossing of transgenic traits is to engineer the desired transgene(s) into a male sterile background (Mariani et al. 1990). Male-sterile plants, as well as plants having transgenes on the plastome, can be pollinated by the wild type or weeds and the resulting hybrids can spread into natural populations, with the wild type or wild species as the recurrent pollen parent in backcrosses. Thus, while the latter two technologies reduce outcrossing, they do not prevent introgression. Another containment option would be to engineer crops to produce seeds without fertilization (e.g. apomixis), where the seed is actually of vegetative origin (Savidan
2000). Unfortunately, the lack of pollen is not absolute and a low frequency of sexually derived seed is typically found. An additional risk of pollen transfer to nonapomictic related species can also exist. Other molecular approaches suggested for crop transgene containment include seed sterility, utilizing the genetic use restriction technologies (Oliver et al. 1998) and recoverable block of function (Kuvshinov et al. 2001). Such proposed technologies control outcrossing and volunteer seed dispersal but, if the controlling element of the transgene is silenced, expression will occur. A similar but more complicated technology suggests utilizing a repressible seed-lethal system (Schernthaner et al. 2003). It seems to be technically impractical at present as it requires that a repressor gene and a seed-lethal gene be engineered at an identical allelic position on sister chromosomes. Another approach includes the insertion of the transgene behind a chemically induced promoter so that it will be expressed upon chemical induction (Jepson 2002). However, there is a possibility of an inducible promoter mutating to become constitutive.
Mitigating leaked transgenes All of the above containment mechanisms have an element of genetic leakiness. Risk can be further reduced by stacking containment mechanisms that compound the infrequency of gene introgression. However, even at very low frequencies of gene transfer, once it occurs, the new bearer of the transgene can disperse throughout the population if the transgenic trait has just a small fitness advantage. Conversely, if the transgene has a small fitness disadvantage, it will remain localized as a very small proportion of the population. Therefore, gene flow should be mitigated by lowering the fitness of recipients below that of the wild type competing species so that the transgene will not spread. Thus, the concept of ‘transgenic mitigation’ (TM) was proposed (Gressel 1999), where mitigator genes are tandemly coupled with the desired primary transgene, which would reduce the fitness advantage to hybrids and their rare progeny and thus considerably reduce risk (Gressel 1999). This TM approach is based on the premises that: (i) tandem constructs act as tightly linked genes and their segregation from each other is exceedingly rare; (ii) the TM traits chosen are neutral or favourable to crops but deleterious to noncrop progeny due to a negative selection pressure and (iii) individuals bearing even mildly harmful TM traits will be kept at very low frequencies in weed populations because weeds typically have a very high seed output and strongly compete among themselves (Gressel 1999). Thus, it was predicted that, if the primary gene of agricultural advantage being engineered into a crop is flanked by TM gene(s), such as dwarfing, uniform seed ripening, nonshattering, antisecondary dormancy, inability to overwinter or nonbolting genes in a tandem construct, the overall effect would be deleterious © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 697– 710
T A N D E M C O N S T R U C T S M I T I G A T E T R A N S G E N E F L O W 699 after introgression into weeds because the TM genes will reduce the competitive ability of the rare transgenic hybrids so that they cannot persist in easily noticeable frequencies in agroecosystems (Gressel 1999). Crop-related weeds are usually copious pollen producers and set large numbers of seeds, many of which germinate during the following cropping season. Thus, if there is a rare pollen grain bearing tandem TM traits that passes containment, it must compete during fertilization with multitudes of more fit wild type pollen to produce a hybrid. Its rarer progeny must then compete with more fit wild type progeny and nontransgenic cohorts during seedling self-thinning and establishment. Thus, even a small degree of unfitness encoded in the TM construct would bring about the elimination of the vast majority of progeny in all future generations where the primary gene provides no advantage and the linked gene confers unfitness. We report here the use of tobacco (Nicotiana tabacum) as a model plant to test the TM concept. We used a tandem construct containing an ahasR (acetohydroxy acid synthase) gene for herbicide resistance as the primary desirable gene and the dwarfing ∆gai (gibberellic acid-insensitive) mutant gene as a mitigator. Several herbicides, including the imidazolinones, inhibit the enzyme acetohydroxy acid synthase (AHAS) (= acetolactate synthase) (Mazur & Falco 1989), the key enzyme in the biosynthesis pathway leading to branched-chain amino acids. Various semi- or completely dominant resistant point mutations in the ahasR gene have been reported from many plant sources, especially weeds (Saari et al. 1994; Gressel 2002). While these genes confer a selective advantage when the herbicides are used, particular genes can have a negative effect on the productivity of the transgenic plants themselves (Purrington & Bergelson 1997, 1999). In other cases, productivity differences are too small to be measured but competitive reproductive fitness must be less than neutral or such mutations would be more widespread. Genetically engineered height reduction is possible by manipulating genes relating to gibberellic acid (GA) production (Coles et al. 1999) or to the GA response itself (Peng et al. 1997). Dwarfing would be disadvantageous to the rare weeds introgressing the TM construct, as they could no longer compete with other crops or fellow weeds, but is desirable in many crops, preventing lodging and producing less stem tissue with more yield (Gale & Youssefian 1985). The semidominant mutant allele ∆gai encodes a constitutively active repressor protein (∆gai) that lacks 17 amino acids near the amino terminus (the DELLA deletion region) and is relatively insensitive to the effects of endogenous and applied GA (Peng et al. 1997, 1999a). In this study, we report that dwarf, imazapyr-resistant TM transgenic hybrid tobacco plants (simulating a TM introgressed hybrid) could not compete with the wild type in ecological simulation competition experiments between wild type and transgenic tobacco plants. The data from this study can be the basis for designing large scale field studies © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 697–710
with crop/weed pairs to further evaluate the positive implications of risk mitigation.
Materials and methods Assembling the tandem construct Genes. The genomic ahasR gene cloned in the pAC456 plasmid was kindly provided by Dr Roy Chaleff (American Cyanamid Co.). It contains the Arabidopsis thaliana ahas promoter, ahasR coding sequence and ahas terminator. The genomic ∆gai gene cloned in pλg/SK+ plasmid, containing the A. thaliana gai promoter, ∆gai coding sequence and gai terminator, was kindly provided by Drs Donald Richards and Nick Harberd (Department of Molecular Genetics, John Innes Centre, Norwich, UK). The two complete genes were used to assemble the TM tandem construct. The complete sequences of both clones were determined by sequencing different critical DNA segments within both genes using an ABI Prism 3700 sequencer. Similarity alignments with the blast genomic database of A. thaliana chromosome 3 (nucleotides 35 863–41 580 of the GenBank Accession no. AL133315) showed that the mutated ahasR is identical to the Arabidopsis csr1.2 gene (GenBank Accession no. X51514) containing a substitution of serine at amino acid residue 653 with asparagine (Sathasivan et al. 1990). Sequenced segments of ∆gai aligned with the database of A. thaliana chromosome 1 from nucleotides 115 566 to 120 773 of the GenBank Accession no. AC006917. It was orientated with the gai promoter close to the XbaI site and the gai terminator close to the EcoRI site of the pλg/SK+ vector. pPZP212-ahasR-∆gai-1 tandem construct (TM 1). The 5712 base pairs (bp) XbaI ahasR fragment was isolated from the pAC456 plasmid and cloned in both orientations in the XbaI site that is 15 bp upstream of the ∆gai gene of the pλg/ SK+ plasmid. Both genes were in direct orientation in the pSK+-ahasR-∆gai-1 clone as confirmed by sequencing using the universal T3 and T7 primers. The pSK+-ahasR-∆gai-1 clone was first digested with NotI. Klenow polymerase was used to blunt the 3′-NotI sticky end and SalI was then used to release the complete ahasR-∆gai-1 fragment. This fragment was ligated into the Sma I- and SalI-predigested binary vector pPZP212 (Hajdukiewicz et al. 1994) to produce the pPZP212-ahasR-∆gai-1 (TM 1) plasmid (Fig. 1a and c). The TM plasmid contained the aadA gene conferring bacterial resistance to spectinomycin and the kan gene encoding neomycin phosphotransferase II conferring plant resistance to kanamycin. Both selectable genes were carried within the native T-DNA of the pPZP212 binary vector (Hajdukiewicz et al. 1994). The TM 1 plasmid was partially sequenced to confirm the makeup of the TM 1 construct, which was consistent with the KpnI/SalI enzymatic digestion (Sambrook et al. 1989) of the TM 1 construct (Fig. 1d).
700 H . A L - A H M A D , S . G A L I L I and J . G R E S S E L
Fig. 1 Assembly and verification of the transgenic mitigation (TM) tandem construct used for transformation. (a) Circular map of the plasmid pPZP212-ahasR-∆gai-1 (TM 1). Arrows indicate gene orientation; LB, left border; RB, right border; A–D indicated on the map denote the sites chosen for polymerase chain reaction (PCR) amplification used to ascertain transformation with intact TM T-DNA. (b) Linear map of the tandem genes assembled in direct orientation. (c) Electrophoresis of the TM construct plasmid DNA through a 1% (w/v) agarose gel at 50 V for 45 min. (d) The correct construction of the TM 1 plasmid was confirmed by KpnI/SalI restriction enzyme digestion. The TM plasmid digest (lane 4) shows that the ahasR-∆gai insert (lane 2) was found intact together with the binary vector. (e) Detection of the transformed genes in transgenic tobacco plants by PCR. M, Molecular markers (1-kb DNA ladder); WT, wild type; (–), control without template DNA; P, TM 1 plasmid control; T0, primary TM transgenic plants; 1–10 denote different T3-7 TM transgenic events; A–D denote the sites chosen for PCR amplification as shown in (a).
Tobacco transformation The TM 1 construct was electroporated into Agrobacterium tumefaciens strain EHA 105 (Sambrook et al. 1989) which was used to transform N. tabacum cv. Samsun NN leaf disks, as described in Horsch et al. (1985). Primary transgenic plants were transferred to soil and pollinated by wild type N. tabacum cv. Samsun NN. Segregating seeds of TM T0 (= BC) plants were collected for genetic analysis and for the competition experiments where the 1 : 1 segregating population precluded the need to mix transgenic with wild type seed.
Gene integration analyses of tobacco transformants Leaf callus-inducing assay. Axenic young leaves from regenerated wild type and putative primary TM transformants were cut and placed on MS (Murashige & Skoog 1962) callus-inducing medium in different Petri dishes containing
0.2 µg/mL 2,4-Dichlorophenoxyacetic acid and various concentrations of imazapyr to verify resistance. All treatments were repeated twice. The fresh weight of the calli that developed on each leaf cutting was measured after 3 weeks. Polymerase chain reaction. Total plant genomic DNA was prepared by the cetyltrimethylammonium bromide method of Rogers & Bendich (1994) or by the puregene® kit for genomic DNA purification (Gentra Systems) using leaf discs of primary semidwarf and imazapyr-resistant tobacco transformants. The DNA was analysed by polymerase chain reaction (PCR) (Sambrook et al. 1989) for the presence of an intact tandemly linked ahasR and ∆gai genomic insert. Four different DNA segments within the genomic TM T-DNA insert were amplified over the positions indicated in Fig. 1(a): segment A at the right border of TDNA and the beginning of the ahas promoter (forward primer 5′-GCTTTACACTTTATGCTTCC-3′; reverse primer 5′-TAACACTTTTTCTTTTTTTG-3′); segment B at the end © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 697– 710
T A N D E M C O N S T R U C T S M I T I G A T E T R A N S G E N E F L O W 701 of the ahas terminator and over the 15-bp linker between the ahasR and into the ∆gai gene and the beginning of the gai promoter (forward primer 5′-GGTTATGATGGCAGGATGTGG-3′; reverse primer 5′-CGTTACATCATTTTCTCACAA-3′); segment C of the ∆gai-DELLA deletion region (forward primer 5′-TAGAAGTGGTAGTGGAGTGA3′; reverse primer 5′-CGACGGAGAGAGACGGTAAA-3′) and segment D, the kan gene located at the left border of the T-DNA (forward primer: 5′-TCATTTCATTTGGAGAGGAC3′; reverse primer: 5′-CATGATATTCGGCAAGCAGG3′). The PCR reactions were carried out in 50-µL aliquots containing about 200 ng genomic DNA, 5 µL of 10× DyNAzyme™ II buffer (Finnzymes), 1.5 U of DyNAzyme™ II DNA polymerase (Finnzymes), 5 µL of 2.5 mm of each dNTP(s) (Roche Diagnostics GmbH) and 35 pmol of each primer in sterile distilled water. The mixture was denatured for 3 min at 94 °C and amplified for 35 cycles [94 °C for 30 s, 51 °C (DNA segments A and C) or 57 °C (segments B and D) for 30 s, 72 °C for 1 min] with a final cycle of 7 min at 72 °C. The PCR products (15 µL) were loaded directly onto 1% (w/v) agarose gels to verify single bands. The remaining PCR products were purified using the QIAquick PCR Purification Kit® (Qiagen) according to the manufacturer’s instructions and sequenced to confirm the integration of the TM T-DNA. In vivo acetohydroxy acid synthase enzyme assay. A rapid leafdisc AHAS microtiter plate assay described by D. Shaner (Shaner 2002 and personal communication) was used to measure tobacco AHAS resistance to imazapyr. Triplicate 4-mm diameter leaf discs were taken from parallel positions on both sides of the midrib from young leaves of each plant and floated on 100-µL volumes of 10% (w/v) MS salts (Duchefa) to which were added 10 mm KH2PO4, 1% (w/v) l-alanine and 500 µm 1,1-cyclopropanedicarboxylic acid (Sigma-Aldrich), one disc in medium with 5 µm imazapyr and the other without. The lowest dose of imazapyr inhibitory to AHAS in the untransformed leaves (0.5 µm) was determined (data not shown). The samples in the plates were placed under fluorescent lighting (85 µE/m2/s) for 8 h at room temperature and stored overnight at −20 °C. After thawing, 25 µL of 5% (v/v) H2SO4 were added to each well and incubated at 60 °C for 15 min to terminate the enzymatic reaction and complete the chemical conversion of acetolactate to acetoin. Then, 150 µL of 2.5% (w/v) αnaphthol (Fluka) and 0.25% (w/v) creatine (Fluka) in 2 N NaOH were added to each sample and the plates were incubated in the dark at 60 °C for another 15 min to accelerate red colour development in the solution surrounding discs from imazapyr-resistant plants. The concentration of the acetoin–naphthol– creatine complex was determined spectrophotometrically at A530 and the AHAS activity was presented as percent activity in the absence of the herbicide. Total soluble proteins were determined from untreated leaf tissues by the method of Bradford (1976). © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 697–710
Inheritance of the transgenic mitigation construct transgenes Seed germination in the presence of kanamycin or imazapyr. A discriminatory dose level of the agar medium of Nitsch (1969) containing 85.8 µm kanamycin or 0.3 µm imazapyr inhibiting the growth of nontransformed plants was determined. Surface-sterilized seeds of wild type, the backcross of primary transformants with wild type T0 (= BC), T1, T2 and T3 tobacco plants were planted on either medium. Sensitive seedlings stopped growing within 10 days forming only short roots and cotyledons, while resistant seedlings formed long roots and true leaves and continued their growth. The segregation ratio of wild type to TM of each line was calculated. Resistant/dwarfed seedlings were grown to maturity in soil and the plants were analysed for AHAS resistance and for ahasR and ∆gai insert by PCR as described above.
Chlorophyll measurement Chlorophyll was extracted from 10 8-mm diameter leaf disks from four leaves per plant using the 80% acetone method of Porra et al. (1989). The absorbance of the supernatant fluid was spectrophotometrically measured at 470, 646.6 and 663.6 nm using the 80% acetone buffer as a blank. Chlorophyll a and b were determined using the equations in Porra et al. (1989) and carotenoid using the equations in Hill et al. (1985).
Productivity of tobacco transformants without self-competition Several growth parameters were determined at intervals on wild type and TM transgenic plants grown separately in 13-cm diameter 1-L pots filled with a 1 : 1 : 1 mixture of peat, crushed tuff rock and loam and spaced 40 cm between pots such that the canopies never overlapped. The measurements included plant height, number of leaves, number of opened flowers, stem thickness measured at maturity at 5 cm above soil, and leaf internode spacing measured from the first basal leaf towards the branches carrying flowers and flower buds. Leaf longevity was measured by recording whether each leaf was green vs. yellow or dried at each internode at various times and calculating the t0.5 of the leaves. The fresh weights of the shoots and roots were separately measured at maturity.
Competition of transgenic mitigation transgenics with the wild type and self-competition of each alone The transgenic TM tobacco was used to simulate a TM introgressed hybrid or a transgenic volunteer crop and the wild type tobacco was used to simulate a nontransgenic
702 H . A L - A H M A D , S . G A L I L I and J . G R E S S E L crop, weed or wild plant. The competitive interactions between them were assessed in two separate internally replicated greenhouse experiments. In both experiments, seeds obtained from the first backcross T1 (= BC1) of a representative TM line T1-7 (pollen recipient) and wild type (pollen donor) were used to grow mixed TM and wild type cultures. The segregation ratio of the TM-linked genes in the backcross was 1 : 1 hemizygous TM to wild type, as confirmed by seed selection on a medium containing the lowest discriminatory dose of imazapyr (0.3 µm) that killed the wild type segregants but did not affect the development of the resistant plants. Randomly chosen T2 (= BC2) segregating nursery seedlings were transplanted at 2.5-, 5- and 10-cm spacing (experiment I) or at 1-, 2.5- and 5-cm spacing (experiment II), without using the selective herbicide. The experiments were conducted in 55 × 41 × 22-cm plastic containers filled with a mixture of peat, crushed tuff rock and loam (1 : 1 : 1). Monoculture controls of wild type alone (seeds obtained from pure wild type line) and TM alone (T2 = BC2) seedlings preselected with 0.3 µm imazapyr medium to cull the wild type segregants, were also grown at the same spacing. Plants were grown under ambient greenhouse light (30 ± 3 °C) with a higher light intensity in the first experiment than in the second experiment (780 ± 50.3 and 474 ± 17.2 µE/m2/s1, respectively). Light intensities are the averages (±SE) of noontime measurements made on seven occasions during the growing season. All of the replicates were randomly placed on glasshouse benches to minimize microenvironmental effects. Optimal water was supplied and the plants were treated against diseases and arthropods when necessary. The young putative TM, phenotypically dwarf plants were assayed for resistant AHAS enzyme activity by the leaf disc test (as described above). All resistant transgenics had more compact canopies and darker green leaves than the wild type. After AHAS enzyme screening, the plants were labelled with a numbered tape attached to the stem for future reference. The leaf discs of the taller less compact plants all contained herbicidesusceptible AHAS. Two months after transplanting and at time intervals thereafter various growth parameters were measured at the different spacings, including plant height and live leaf number of 10–15 randomly chosen plants of each biotype in each box. The number of flowers formed on each plant and the number of surviving plants of each biotype were also recorded. The vegetative fitness (based on plant height) and the reproductive fitness (based on flower and fruit number formed per m2) of TM plants relative to wild type were calculated at the different spacings and time intervals as a TM : wild type ratio. The below-ground competition was not measured in our experiments due to too-intertwined roots at close spacing but it was assumed that below-ground competition would manifest itself in above-ground growth parameters.
Statistical analyses Tandem trait segregation data were statistically analysed by the chi-square test. The chlorophyll content, tobacco productivity and fitness data were analysed using the jmp® program (version 4.0.1; SAS Institute 2000) by one-way analysis of variance (anova) and comparing the least significant differences. Differences were considered to be statistically insignificant at P > 0.05.
Results and discussion Expression of the transgenic mitigation construct in tobacco transformants Seven independent primary semidwarf, kanamycin- and imazapyr-resistant tobacco lines were obtained by transformation with the TM 1 construct (pPZP212-ahasR-∆gai-1) (Fig. 1). All of the T0 lines had similar semidwarf phenotypes, with thicker stems, shorter internodes, compact canopy, more leaves that were darker green and felt thicker (and had a greater mass per unit area) and had more apical branches and flowers than those of the wild type. T2, T3 and T4 segregants had similar semidwarf (hemizygous) as well as dwarf (homozygous) individuals (Figs 2 and 3, Table 1).
Fig. 2 Phenotypic appearance of semidominant dwarfism induced by ∆gai. From left to right are wild type (WT), T3 transgenic mitigation (TM) hemizygous and T3 TM homozygous plants segregated in the progeny of selfed T2 TM hemizygous plants 12 weeks after transplanting. The inset shows the stems of the same plants after removing the leaves and flowers. © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 697– 710
T A N D E M C O N S T R U C T S M I T I G A T E T R A N S G E N E F L O W 703 Fig. 3 Growth characters of transgenic tobacco grown without self-competition. The growth parameters were determined at intervals on the wild type (WT) and the transgenic mitigation (TM) T3 (line 7) transgenic plants grown separately in 1-L pots with considerable distances between them. The points are the mean (± SE) of 10 wild type (), 15 hemizygous () and nine homozygous () plants. The SEs were smaller than the data points where there are no bars and the different letters within (d) indicate different values at P ≤ 0.05 (LSD test) for chlorophyll and carotenoids, separately.
Table 1 Growth characters of transgenic mitigation (TM) T3 tobacco (means ± SE)
Line Wild type Homozygous ahasR T3 hemizygous GAI/∆gai T3 homozygous ∆gai/∆gai
Internode length (cm)
Leaf mass (mg/cm2)
No. of flowering branches
Fresh weight (g)
Stem thickness 5 cm above soil level (cm)
Shoots
Roots
5 5
1.0 ± 0.03a 1.0 ± 0.04a
3.0 ± 0.1a 2.9 ± 0.1a
26.0 ± 0.4a 28.7 ± 0.6b
6.2 ± 0.4a 5.6 ± 0.4a
179.7 ± 17.8a 164.8 ± 10.8a
35.8 ± 1.8a 32.0 ± 2.0a
16
1.4 ± 0.03b
1.3 ± 0.03b
29.2 ± 0.7b
9.8 ± 0.4b
232.9 ± 4.1b
34.1 ± 2.0a
10
1.5 ± 0.04c
0.6 ± 0.01c
31.8 ± 0.6c
9.4 ± 0.2b
174.7 ± 5.8a
28.0 ± 1.3b
n
TM line 7 was used. The plants were cultivated singly in pots at large spacing to allow maximum growth potential. Homozygous ahasR seed was kindly provided by Dr Dvora Aviv who previously made the ahasR (alone) transformants. Different letters within a column indicate different LSD values at P ≤ 0.05.
Previous tobacco transformants with the same ahasR gene alone were not phenotypically different from the wild type (Dr Dvora Aviv, personal communication) and, when grown separately at the same time, were without significant difference from the wild type in almost all parameters measured, except herbicide resistance (Table 1). The fresh weight of the hemizygous TM shoots was significantly higher than that of the wild type (P ≤ 0.05) but no significant difference was found from the root fresh weight (Table 1). The hemizygous and homozygous TM leaves contained > 60% more chlorophyll per gram fresh weight © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 697–710
than the wild type (P ≤ 0.01) but less carotenoid (P ≤ 0.05) (Fig. 3d). Chlorophyll levels were also elevated in the ∆gai dark green Arabidopsis transformants (Peng et al. 1999b) as well as in severely dwarfed ∆gai transgenic chrysanthemums (Petty et al. 2003). Leaf longevity was considerably prolonged in the TM plants (Fig. 3b) and, at 5.5 months of growth, when all leaves had dehisced from the wild type plants, half of the leaves remained on the TM plants. Increased leaf longevity was also found with the Norin 10 dwarfing gene in wheat (Kulshrestha & Tsunoda 1981). The higher level of chlorophyll and the prolonged leaf longevity may
704 H . A L - A H M A D , S . G A L I L I and J . G R E S S E L Table 2 Segregation of imazapyr resistance in transgenic mitigation (TM) tobacco lines
Generation Nontransgenic regenerated controls Hemizygous T1 (= BC1) total Selfed T2 total Selfed T3 total Selfed T4 total Homozygous Selfed T4 total
No. of lines and (total plants) tested
R
S
Observed R : S ratio
Expected R : S ratio
χ2
P (χ2)
5 (5)
0
1000
0.00 : 1
0
0.00
—
7 (7) 2 (6) 2 (3) 1 (4)
573 1175 744 838
546 411 260 267
1.05 : 1 2.86 : 1 2.86 : 1 3.14 : 1
1:1 3:1 3:1 3:1
0.00 0.01 0.01 0.01
0.04 0.06 0.06 0.06
1 (4)
4374
0
—
—
—
—
Resistant (R) and susceptible (S) seedlings were scored in the experiment. BC are TM plants backcrossed with the wild-type as pollen donor.
Fig. 4 Expression of imazapyr resistance in transgenic mitigation (TM) transgenic tobacco plants. (a) Acetohydroxy acid synthase (AHAS) enzyme activity. This was spectrophotometrically assayed in leaf discs as described in Materials and methods and presented as percent activity of controls tested in the absence of the herbicide. The enzyme activity of the wild type (WT) and TM controls was 0.017 and 0.021 µmol acetolactate/cm2 leaf discs/h, respectively. The individuals assayed were wild type (), TM T1 (= BC1) (), TM T3 hemizygous () and TM T3 homozygous (). The TM line represents the mean of all the transgenic cases. T3 segregants were classified as hemizygotes and homozygotes by their height. One cm2 of untreated leaf discs contains 52.7 ± 3.4 µg soluble proteins. (b) Appearance of representative T1 (= BC 1) tobacco segregants at 14 days picked from one agar plate containing 0.3 µm imazapyr and lined up in groups.
contribute to higher total photosynthetic activity in TM plants. Flower initiation was delayed on the TM plants by 1 week (P ≤ 0.01) and continued 1 month longer than the wild type (P ≤ 0.05). For the first 93 days, there was no significant difference in the number of flowers formed by the wild type and the TM hemizygous plants (ns, P > 0.05) but there were more flowers formed by the wild type than the TM homozygous plants (P ≤ 0.05). By 107–114 days, the TM plants formed many more flowers than the wild type (P ≤ 0.01) with a small but significant difference between the TM hemizygous and homozygous plants (P ≤ 0.05). At 144 days, the TM plants had > 50% more flowering branches (P ≤ 0.05) and > 65% more flowers (P ≤ 0.01) than the wild type. Thus, agronomic crops with this TM gene would be expected to have a higher seed yield (higher harvest index). In Arabidopsis, flowering was also delayed in GAdeficient and GA-insensitive mutants and the ∆gai mutants failed to respond to exogenous GA treatment, which
accelerated flowering in other GA-deficient mutants under noninductive short-day conditions (Wilson et al. 1992).
Segregation of the tandem traits in transgenic mitigation progeny The T1 (= BC1) progeny of the seven lines segregated according to the expected 1 : 1 Mendelian ratio of imazapyrresistant/semidwarf : sensitive/tall [P (χ2) = 0.04; Table 2, Fig. 4b]. The segregation ratio in the progeny of selfed hemizygous T2 (lines 5 and 7), T3 (lines 2 and 7) and T4 (line 7) was 3 : 1 imazapyr-resistant/dwarf or semidwarf : sensitive/ tall for most of the transformants [P (χ2) = 0.06; Table 2], suggesting that the construct was inserted into a single chromosome in most cases. The dwarfed homozygous progeny were all resistant to the herbicide (Table 2), indicating that the traits remained linked. The segregation of dwarfism in the herbicide-resistant progeny of selfed T2 (line 5), T3 © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 697– 710
T A N D E M C O N S T R U C T S M I T I G A T E T R A N S G E N E F L O W 705 Table 3 Segregation of dwarfism in herbicide-resistant progeny of selfed transgenic mitigation plants Progeny Generation
Line
S
D
Observed S : D ratio
Expected S : D ratio
χ2
P (χ2 )
T2 T3 T3 T4
5 6 7 7
48 33 71 54
23 18 33 25
2.09 : 1 1.83 : 1 2.15 : 1 2.16 : 1
2:1 2:1 2:1 2:1
0.00 0.01 0.01 0.01
0.05 0.09 0.09 0.09
S, Semi-dwarf; D, dwarf.
Table 4 Presence of ahasR-∆gai linked region (B) of transgenic mitigation construct T-DNA in acetohydroxy acid synthase-resistant semidwarf or dwarf segregants tested by polymerase chain reaction (PCR)
Generation
No. of tested lines (total plants)
PCR+
PCR–
T0 T1 (= BC1) T2 T3
7 (7) 7 (14) 5 (10) 2 (13)
7 14 10 13
0 0 0 0
Wild type control reactions were always negative.
(lines 6 and 7) and T4 (line 7) was near the expected ratio of 2 : 1 semidwarf : dwarf [P (χ2) = 0.05, 0.09 and 0.09, respectively; Table 3]. The covalent ahasR and ∆gai gene linkage remained intact in all generations as measured both genetically and by PCR amplification of different DNA segments within the TM construct T-DNA (Fig. 1a and e, Table 4). The inheritance of the tandem traits was stable (i.e. herbicide resistance did not segregate from dwarfism) in the limited number of progeny of five lines tested in the following generations. Large scale field testing with tens of millions of plants would be necessary to assess the possibility of rare instabilities, i.e. the frequency of segregation of linked traits from each other by crossing over, as the expected frequency would be 0.05) in resistance to various herbicide doses were detected when comparing the resistant individuals in segregating hemizygous families with the homozygous plants from nonsegregating families (Fig. 4). This is not always the case; Volenberg & Stoltenberg (2002) found semidominant inheritance of AHAS resistance in field-evolved resistant populations of Solanum nigrum.
Productivity and fitness of tobacco plants cultivated at close spacings The productivity of the wild type and the TM transformants was determined by growing each biotype separately and the fitness was measured when they were cultivated together under competition at spacings where the canopies of all plants overlapped. A single transformed line with optimal phenotype was used as that would simulate agronomic practice. Less optimal transformants would be expected to compete less well. The use of naturally mixed backcross seeds was chosen as it provides TM and wild type segregants having an isogenic background, except for the TM insert. Productivity of wild type and transgenic mitigation plants cultivated separately. The productivity of the hemizygous imazapyr-resistant and semidwarf TM plants grown alone without competition from the wild type plants was higher at different spacings than that of the wild type plants growing alone at similar spacing. The TM plants formed more leaves (P ≤ 0.01) (Fig. 5b) and more flowering branches and flowers with a longer duration of flowering (P ≤ 0.05) (Fig. 6) than the wild type plants grown alone. Thus, it is clear that TM should be advantageous to a crop growing alone but disadvantageous to a crop–weed hybrid living in a competitive environment. Dwarf ‘green revolution’ crops are also far more susceptible to competition from weeds than tall varieties but are far more productive than tall crops when cultivated without competition. The added yield of dwarf varieties more than compensates for
706 H . A L - A H M A D , S . G A L I L I and J . G R E S S E L
Fig. 5 Suppression of growth of transgenic mitigation (TM) transgenic tobacco plants when in competition with the wild type (WT). Randomly chosen 2-week-old T1 (= BC1) seedlings that segregated to the wild type (closed symbols) and transgenic hemizygous semidwarf/herbicide-resistant (open symbols) plants were planted at 1, 2.5, 5 and 10 cm from each other in soil without herbicide. SEs were smaller than the data points where there are no bars.
the added costs and expense of weed control (Silverstone & Sun 2000; Sasaki et al. 2002). When the wild type plants were grown only in competition with themselves, self-thinning measured after 100 and 111 days of cultivation accounted for a loss of 41% of the wild type plants at 2.5-cm and 5% at 5-cm spacing (Fig. 7). This is consistent with less available space and resources when the wild type plants are grown alone vs. being interspersed with weak TM competitors. The TM plants cultivated alone at 2.5- and 5-cm spacing were more self-competitive than the wild type plants cultivated alone at similar spacing (P ≤ 0.01), as 55 and 23% (respectively) of TM plants were lost to self-thinning within the first 111 days of cultivation (Fig. 7). The TM transgenics had thicker leaves in a more compact canopy with less apical dominance than the wild type. This made
it difficult for the shaded weaker dwarf plants to receive enough light to survive. Only 10-cm spacing was sufficient for all individual TM and wild type plants to receive enough light to grow well, even though the canopies were partially overlapped. A TM crop growing at far greater spacing would be much more productive than the wild type cultivated without competing weeds. Competitive fitness of transgenic mitigation transgenic hybrids and the wild type. When the dwarf TM plants were cultivated alone at the widest spacings (5 and 10 cm apart), competing only with their TM sibs, they grew well and flowered (see previous section). This was not the case when they were grown in equal mixture with wild type plants from an otherwise identical genetic background. Two independent competition experiments demonstrated © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 697– 710
T A N D E M C O N S T R U C T S M I T I G A T E T R A N S G E N E F L O W 707
Fig. 6 Suppression of flowering of transgenic mitigation (TM) transgenic tobacco plants when in competition with the wild type (WT).
Fig. 7 Lack of survival of transgenic mitigation (TM) plants in competition with wild type (WT). The proportion of live plants was measured after 100 days of growth in experiment I (closed symbols) and after 111 days in experiment II (open symbols). The lines show the average of the two experiments.
that the hemizygous TM semidwarf transgenics were exceedingly unfit and could hardly compete with the wild type segregants when cocultivated without using the selector herbicide. They possessed very low ‘Haldanian’ reproductive fitness (Haldane 1960), i.e. they could hardly reproduce when the selector was not there to remove the wild type. The results from both experiments demonstrated a strong suppression of growth measured as height and number of leaves (P ≤ 0.01) and flowering (P ≤ 0.01) as well as a low survival of TM plants at all spacings when in competition with the wild type segregants (Figs 5 – 7). The spacing between plants would be even closer in many agroecosystems than in these experiments, as a typical weed plant sheds thousands of seeds in a small area. Very few seeds in the field would be hybrids resulting from cross-pollination with a transgenic crop so the potential for competition would be even greater than in these experiments, which used equal numbers of TM transgenics and wild type. In the mixed cultures of the first experiment, 60 and 15% of the TM plants were dead at 2.5- and © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 697–710
5-cm spacing, respectively after 100 days of cultivation (Fig. 7). The TM survivors were mostly found at the edges of the containers and formed no flowers at either spacing, even 1 month after the wild type ceased growth and withered, opening the canopy (Fig. 6). Among the wild type segregants, only 18% at 2.5-cm and 4% at 5-cm spacing succumbed to self-thinning during that period. Even at 10-cm spacing, where only 1% of the TM plants died and all of the wild type plants survived, only those TM plants growing at the periphery formed sparse flowers by the time wild type plants had stopped growing (Figs 6 and 7). In a field situation, any transgenics that survive would probably be killed during harvesting of the crop before the transgenics could set seeds. Even if the TM plants were to set seed, further generations of their progeny would still be less competitive with wild type cohorts, or with other weeds, or with crops in subsequent generations, unless the selective herbicide is used. In the second experiment, the plants were spaced 1, 2.5 and 5 cm from each other. There was strong competition
708 H . A L - A H M A D , S . G A L I L I and J . G R E S S E L Fig. 8 Vegetative and reproductive fitness of transgenic mitigation (TM) segregants in competition with wild type (WT). The points represent the TM/wild type ratio.
between wild type and TM plants at 1-cm spacing with > 99% of the TM plants dying (Fig. 7). The persistence of the TM plants at 2.5- and 5-cm spacing was lower in the second experiment, which was performed under lower ambient natural greenhouse light intensity than the first experiment. In both experiments on single and mixed cultures, the increase in plant density resulted in reduced leaf area, leaf number (Fig. 5b) and fewer flowers per plant (Fig. 6) in the range of densities studied. The competitive fitness of the TM plants relative to the wild type (TM/WT) varies both as a function of spacing and of time, as the wild type closes its canopy over the TM plants and due to later flowering of the TM plants when in competition (Fig. 8). Even at the 10-cm spacing, the vegetative fitness measured as height was never more than 30% of the wild type. As the more crowded plants cease growing, their relative fitness decreases. The more critical reproductive fitness (based on flower and fruit number of TM/WT) was easy to ascertain at the high densities of competition. The reproductive fitness was zero at 1-, 2.5- and 5-cm spacing. It is more complex to calculate this ratio at the greater spacing as it changes with time due to later flowering of the shaded TM plants in competition with the wild type (Fig. 8) and a single value that has relevance to the agroecosystem cannot be determined. In normal practice the wild type would be harvested before many TM seeds were produced so it is unlikely that the relative fitness would ever arrive at the highest ratio recorded in the experiment (17%) of the wild type (Fig. 8).
Concluding remarks The potential utility of the concept of TM was supported by the experiments with the tobacco model. The results indicated what should happen to TM hybrids and/or crops if they become volunteer weeds in seasons when the selector herbicide is not used or when the half-life of the herbicide is short, i.e. not very persistent, and seeds germinate after the herbicide has dissipated. The TM system should work well in agricultural fields when crops and herbicides are rotated. This technology would not be as
efficient if the TM crop is grown season after season in monoculture, especially if the primary transgene is herbicide resistance and the herbicide selector is also repeatedly used, i.e. a situation where ‘Darwinian’ fitness is allowed to continually express itself, where only the herbicide-resistant TM plants can exist. The TM concept is now being tested by us using the same construct as reported here to ascertain how well it will mitigate gene flow from the crop Brassica napus (oilseed rape = canola) to the weed B. campestris = B. rapa (wild turnip), where gene transfer has been demonstrated in the field (Jorgensen & Andersen 1994; Brown & Brown 1996; Mikkelsen et al. 1996; Rieger et al. 1999). The TM concept could also be used to mitigate interspecific transgene flow in rice. Cultivated rice has intercrossing feral forms that are major weeds, e.g. ‘wild red rice’ which is not always red seeded (Holm et al. 1997). Herbicide resistance may allow control of feral rices in cultivated rice but introgression from crop to weed must be mitigated (Gressel 2002). Other useful implications of the TM concept could be in pharmaceutical crops where gene transfer to other varieties is undesirable. If a ‘shrunken’ gene (a null starch gene) was used as a mitigator, the high sugar seeds of the hybrids or the volunteers could not persist over winter. The above results are from a greenhouse simulation with a model plant in large containers of what might occur in the field. Competition is probably harsher in the field and the TM should have an even greater disadvantage to the rare hybrids resulting from introgression, especially as one expects more dense populations and much more than 50% wild type surrounding the hybrids. However, this indication strongly suggests that field experiments should be performed with a ‘real’ (not model) crop–weed pair, under cropping. Such experiments should include various treatment blocks where the expression of the selective advantage of the primary transgene (in this case herbicide resistance) is intermittently allowed. The proportion of TM individuals in the population should be followed for a few seasons to ascertain what asymptotic level is achieved. If such an experiment is large enough, it should be possible to ascertain the frequency of segregation of the mitigating trait from © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 697– 710
T A N D E M C O N S T R U C T S M I T I G A T E T R A N S G E N E F L O W 709 the primary trait. With the TM construct used, this means obtaining a tall plant that is herbicide resistant or a dwarf plant that is herbicide susceptible. If any of the ‘leaky’ containment strategies (described in the Introduction) were stacked tandemly with TM genes, the rare result of such a ‘leak’ should remain at a low frequency in the population, generation after generation, without spreading, because of the decreased fitness conferred by the TM gene. If the containment strategies are not stacked with a TM-type function then, once a leak has occurred, the transgene could slowly spread (‘drift’) in the population if it was neutral or increase faster if it had some selective advantage. Thus, mitigation is necessary as a last resort when containment leaks. Where it is imperative to have an ultra low risk of having the mitigating gene segregate from the primary gene in persistent populations of transgenic plants, the primary gene should be flanked with two TM genes. Assuming that a loss of a single TM function due to a mutation is about the same as for rare segregation (c. 10 −6), the use of two flanking genes would lower this risk to 10 −12. If a dual TM construct was engineered into the chloroplast (i.e. stacked with containment), the risk would be lowered to ≈10 −15. If certified seed is used, the loss of TM functions in the crop, such as dwarfing, could be seen and such plants culled, further lowering the risk of unwanted stable hybrids between TM crops and their relatives, as well as their progeny being at anything but at a very low frequency in the population.
Acknowledgements The authors are grateful to Ms Shiri Gerson for excellent technical assistance, Dr Roy Chaleff (American Cyanamid Company) for kindly providing plasmid pAC456 and Drs Donald Richards and Nick Harberd (Department of Molecular Genetics, John Innes Centre, Norwich, UK) for providing plasmid pλg/SK+. The advice and experience of Dr Dvora Aviv in tissue culture and tobacco transformation techniques is especially acknowledged. This research was supported by the Levin Foundation, INCO–DC, contract no. ERB IC18 CT 98 0391 and a bequest from Israel and Diana Safer.
References Avni A, Edelman M (1991) Direct selection for paternal inheritance of chloroplasts in sexual progeny of Nicotiana. Molecular and General Genetics, 225, 273 –277. Bock R (2001) Transgenic plastids in basic research and plant biotechnology. Journal of Molecular Biology, 312, 425 – 438. Bradford MM (1976) Rapid and sensitive method for quantitation of microgram quantities of protein utilizing principle of protein-dye binding. Analytical Biochemistry, 72, 248 –254. Brown J, Brown AP (1996) Gene transfer between canola (Brassica napus L. & B. campestris L.) and related weed species. Annals of Applied Biology, 129, 513 –522. Coles JP, Phillips AL, Croker SJ et al. (1999) Modification of gibberellin production and plant development in Arabidopsis by © 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 697–710
sense and antisense expression of gibberellin 20-oxidase genes. The Plant Journal, 17, 547–556. Dale PJ, Clarke B, Fontes EMG (2002) Potential for the environmental impact of transgenic crops. Nature Biotechnology, 20, 567– 574. Daniell H (2002) Molecular strategies for gene containment in transgenic crops. Nature Biotechnology, 20, 581–586. Daniell H, Datta R, Varma S, Gray S, Lee SB (1998) Containment of herbicide resistance through genetic engineering of the chloroplast genome. Nature Biotechnology, 16, 345–348. Darmency H (1994) Genetics of herbicide resistance in weeds and crops. In: Herbicide Resistance in Plants: Biology and Biochemistry (eds Powles SB, Holtum JAM), pp. 263–298. Lewis, Boca Raton. Ellstrand NC, Prentice HC, Hancock JF (1999) Gene flow and introgression from domestic plants into their wild relatives. Annual Review of Ecology and Systematics, 30, 539–563. Gale MD, Youssefian S (1985) Dwarfing genes in wheat. In: Progress in Plant Breeding (ed. Russell G), pp. 1–35. Butterworths, London. Gressel J (1999) Tandem constructs: preventing the rise of superweeds. Trends in Biotechnology, 17, 361–366. Gressel J (2002) Molecular Biology of Weed Control. Taylor & Francis, London. Hajdukiewicz P, Svab Z, Maliga P (1994) The small, versatile pPZP family of Agrobacterium binary vectors for plant transformation. Plant Molecular Biology, 25, 989–994. Haldane JBS (1960) More precise expressions for the cost of natural selection. Journal of Genetics, 57, 351–360. Hill CM, Pearson SA, Smith AJ, Rogers LJ (1985) Inhibition of chlorophyll synthesis in Hordeum vulgare by 3-amino 2,3dihydrobenzoic acid (gabaculin). Bioscience Reports, 5, 775– 781. Holm L, Doll J, Holm E, Pancho J, Herberger J (1997) The wild rices. In: World Weeds: Natural Histories and Distribution, pp. 531– 547. John Wiley & Sons, Inc, New York. Horsch RB, Fry JE, Hoffmann NL et al. (1985) A simple and general method for transferring genes into plants. Science, 227, 1229–1231. Huang CY, Ayliffe MA, Timmis JN (2003) Direct measurement of the transfer rate of chloroplast DNA into the nucleus. Nature, 422, 72–76. Jepson I (2002) Inducible Herbicide Resistance. US Patent 6380463. Jorgensen RB, Andersen B (1994) Spontaneous hybridization between oilseed rape (Brassica napus) and weedy B. campestris (Brassicaceae): a risk of growing genetically modified oil seed rape. American Journal of Botany, 81, 1620–1626. Khan MS, Maliga P (1999) Fluorescent antibiotic resistance marker for tracking plastid transformation in higher plants. Nature Biotechnology, 17, 910–915. Kulshrestha VP, Tsunoda S (1981) The role of ‘Norin 10’ dwarfing genes in photosynthetic and respiratory activity of wheat leaves. Theoretical and Applied Genetics, 60, 81–84. Kuvshinov V, Koivu K, Kanerva A, Pehu E (2001) Molecular control of transgene escape from genetically modified plants. Plant Science, 160, 517–522. Maliga P (2002) Engineering the plastid genome of higher plants. Current Opinion in Plant Biology, 5, 164–172. Mariani C, Debeuckeleer M, Truettner J, Leemans J, Goldberg RB (1990) Induction of male sterility in plants by a chimeric ribonuclease gene. Nature, 347, 737–741. Mazur BJ, Falco SC (1989) The development of herbicide resistant crops. Annual Review of Plant Physiology and Plant Molecular Biology, 40, 441–470.
710 H . A L - A H M A D , S . G A L I L I and J . G R E S S E L Mikkelsen TR, Andersen B, Jorgensen RB (1996) The risk of crop transgene spread. Nature, 380, 31–31. Murashige T, Skoog F (1962) A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiologia Plantarum, 15, 473–497. Nitsch JP (1969) Experimental androgenesis in Nicotiana. Phytomorphology, 19, 389 – 404. Oliver MJ, Quisenberry JE, Trolinder NLG, Keim DL (1998) Control of Plant Gene Expression. US Patent 5723765. Peng JR, Carol P, Richards DE et al. (1997) The Arabidopsis GAI gene defines a signaling pathway that negatively regulates gibberellin responses. Genes and Development, 11, 3194–3205. Peng JR, Richards DE, Hartley NM et al. (1999a) ‘Green revolution’ genes encode mutant gibberellin response modulators. Nature, 400, 256–261. Peng JR, Richards DE, Moritz T, Cano-Delgado A, Harberd NP (1999b) Extragenic suppressors of the Arabidopsis gai mutation alter the dose–response relationship of diverse gibberellin responses. Plant Physiology, 119, 1199–1207. Petty LM, Harberd NP, Carre IA, Thomas B, Jackson SD (2003) Expression of the Arabidopsis gai gene under its own promoter causes a reduction in plant height in chrysanthemum by attenuation of the gibberellin response. Plant Science, 164, 175– 182. Porra RJ, Thompson WA, Kriedemann PE (1989) Determination of accurate extinction coefficients and simultaneous equations for assaying chlorophylls a and b extracted with four different solvents: verification of the concentration of chlorophyll standards by atomic absorption spectroscopy. Biochimica et Biophysica Acta, 975, 384–394. Purrington CB, Bergelson J (1997) Fitness consequences of genetically engineered herbicide and antibiotic resistance in Arabidopsis thaliana. Genetics, 145, 807– 814. Purrington CB, Bergelson J (1999) Exploring the physiological basis of costs of herbicide resistance in Arabidopsis thaliana. American Naturalist, 154, S82–S91. Rieger MA, Preston C, Powles SB (1999) Risks of gene flow from transgenic herbicide-resistant canola (Brassica napus) to weedy relatives in southern Australian cropping systems. Australian Journal of Agricultural Research, 50, 115 –128. Rogers SO, Bendich AJ (1994) Extraction of total cellular DNA from plants, algae and fungi. In: Plant Molecular Biology Manual (eds Gelvin SB, Schilperoort RA), pp. D1:1–8. Kluwer Academic Publishers, Belgium. Saari LL, Cotterman JC, Thill DC (1994) Resistance to acetolactatesynthase-inhibitor herbicides. In: Herbicide Resistance in Plants: Biology and Biochemistry (eds Powles SB, Holtum JAM), pp. 80– 139. Lewis, Boca-Raton. Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning. A Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory Press, New York.
Sasaki A, Ashikari M, Ueguchi-Tanaka M et al. (2002) Green revolution: a mutant gibberellin-synthesis gene in rice — New insight into the rice variant that helped to avert famine over thirty years ago. Nature, 416, 701–702. Sathasivan K, Haughn GW, Murai N (1990) Nucleotide-sequence of a mutant acetolactate synthase gene from an imidazolinoneresistant Arabidopsis thaliana var Columbia. Nucleic Acids Research, 18, 2188–2188. Sathasivan K, Haughn GW, Murai N (1991) Molecular-basis of imidazolinone herbicide resistance in Arabidopsis thaliana var Columbia. Plant Physiology, 97, 1044–1050. Savidan Y (2000) Apomixis: genetics and breeding. Plant Breeding Reviews, 18, 13–86. Schernthaner JP, Fabijanski SF, Arnison PG, Racicot M, Robert LS (2003) Control of seed germination in transgenic plants based on the segregation of a two-component genetic system. Proceedings of the National Academy of Sciences of the USA, 100, 6855–6859. Shaner DL (2002) Development and utilization of a high throughput in vivo AHAS assay. Weed Science Society of America Abstracts, 42, 50. Silverstone AL, Sun TP (2000) Gibberellins and the green revolution. Trends in Plant Science, 5, 1–2. Snow AA (2002) Transgenic crops — why gene flow matters. Nature Biotechnology, 20, 542–542. Snow AA, Palma PM (1997) Commercialization of transgenic plants: Potential ecological risks. Bioscience, 47, 86–96. Tomiuk J, Hauser TP, Bagger-Jørgensen R (2000) A- or Cchromosomes, does it matter for the transfer of transgenes from Brassica napus. Theoretical and Applied Genetics, 100, 750 – 754. Volenberg DS, Stoltenberg DE (2002) Inheritance of resistance in eastern black nightshade (Solanum nigrum) to acetolactate synthase inhibitors. Weed Science, 50, 731–736. Wang T, Li Y, Shi Y et al. (2004) Low frequency transmission of a plastid encoded trait in Setaria italica. Theoretical and Applied Genetics, 108, 315–320. Wilson RN, Heckman JW, Somerville CR (1992) Gibberellin is required for flowering in Arabidopsis thaliana under short days. Plant Physiology, 100, 403–408.
This research by Ph.D. student Hani Al-Ahmad at the Weizmann Institute of Science is part of an ongoing programme coordinated by Jonathan Gressel dealing with risk analysis and prevention/ mitigation of transgene flow from crops to related weeds and of transgene flow from transgenically enhanced biocontrol agents to crop pathogens. Research on wheat gene flow to wild relatives is being performed by Sarit Weissman and on transgenic mycoherbicides by Ziva Amsellem and Einat Safran.
© 2004 Blackwell Publishing Ltd, Molecular Ecology, 13, 697– 710
