Plant Cell Reports

Plant Cell Reports (1996) 16:12-17

© Springer-Verlag 1996

Accelerated production of transgenic wheat (Triticum aestivum L.) plants Fredy Altpeter, Vimla Vasil, Vibha Srivastava, Eva Stiiger, and Indra K. Vasil Laboratory of Plant Cell and Molecular Biology, 1143 Fifield Hall, University of Florida, Gainesville, FL 32611-0690, USA Received 29 May 1996/Revised version received 4 June 1996 - Communicated by J. M. Widholm

Abstract. We have developed a method for the accelerated production of fertile transgenie wheat (Triticum aestivumL.) that yields rooted plants ready for transfer to soil in 8-9 weeks (56-66 days) after the initiation of cultures. This was made possible by improvements in the procedures used for culture, bombardment, and selection. Cultured immature embryos were given a 4-6 h preand 16 h post-bombardment osmotic treatment. The most consistent and satisfactory results were obtained with 30 pg of gold particles/bombardment. No clear correlation was found between the frequencies of transient expression and stable transformation. The highest rates of regeneration and transformation were obtained when callus formation after bombardment was limited to two weeks in the dark, with or without selection,followed by selection during regeneration under light. Selection with bialaphos, and not phosphinothricin, yielded more vigorouslygrowing transformed plantlets. The elongation of dark green plantlets in the presence of 4-5 mg/1 bialaphos was found to be reliable for identifying transformed plants. Eighty independent transgenie wheat lines were produced in this study. Under optimum conditions, 32 transformed wheat plants were obtained from 2100 immature embryos in 56-66 days, making it possible to obtain R3 homozygous plants in less than a year. Key words: TriticumaestivumL., wheat, transformation, biolistics.

Introduction Fertile transgenic wheat plants have been produced by the direct delivery of D N A into regenerable tissues and immature embryos by high velocity particle bombardment (Vasil et al. 1992, 1993, Weeks et al. 1993, Nehra et al. 1994, Becker et al. 1994, Zhou et al. 1995, Blechl and Anderson 1996). More recently, Mendelian inheritance and stable expression of the transgene up to homozygosity in the R3 generation has been demonstrated (Blechl and Anderson 1996, Srivastava et al. 1996). In conformity with the results reported for other cereals (Vasil 1994), rooted plants were obtained 3-15 months after the bombardment of callus tissues or immature embryos, at a transformation frequency of 0.1-2.5%. A large number of independently transformed lines must be produced to select those which show stable integration and high level of expression of the transgene for practical use in breeding

Correspondence to. I. K. Vasil

programs. This can be achieved either by increasing the efficiency oflransformation or by reducing the time required for the production oftransgenic plants. Attempts to increase the efficiency of transformation have not been very successful because of the poor understanding of the factors that control stable transformation. Therefore, in order to produce a large number of independently transformed lines within a short period of time, we have developed a method which produced transformed wheat plants from cultured immature embryos in 56-66 days, making it possible to obtain homozygous R3 progeny in less than a year.

Materials and Methods Plant growth conditions, callus culture, regeneration and selection. Wheat plants (Triticum aestivumL., cv. Bobwhite) were grown in the field (Phoenix, AZ, Bozeman, NIT, and Quincy, FL), as well as in growth chambers (first 40 days at 15°C/12°C day/night temperature and 10 h photoperiod at 600 gE-m2-sa), followed by maintenance at 20°C/16°C day/night temperature and 16 h photoperiod at 600 /aE-m2-sx. Spikes were either used fresh or were stored for up to five days in water at I°C in a refrigerator. Seeds were surface sterilized with 70% ethanol for two min and 20% Chlorox (final concentration 1.05% sodium hypoehlorite) with 0.1% Tween 20 for 15-20 rain, followed by four changes of sterile distilled water. Immature embryos (0.5-1.5 mm in size), were aseptically removed under a stereo dissectingmicroscope and placed with the scutellum exposed on MS medium (ivIurashige and Skoog 1962) supplemented with 2 mg/l 2,4-dichlorophenoxyaeetie acid (2,4-D), 20 g/l sucrose, 500 mg/1 glutamine and 100 mg/1 casein hydrolysate(MS+ medium). All media were solidified with 0.25% Gelrite and the pH adjusted to 5.8 prior to autoclaving. Filter ste"nlizedMS vitamins were added after autoclaving. The cultures were incubated at 24°-27°C during callus formation, osmotic treatment and regeneration. Immature embryos were bombarded after 5-7 days of culture in the dark, when proliferating callus tissue was visible at the edges of the scutella. Four to six hours prior to bombardment, the cultures were incubated on solid MS+ medium supplemented with different amounts of mannitol and sorbitol in the dark (for details of osmotic treatments see Results). Twelve to 16 h after bombardment, cultures were transferred either to MS+ medium without selection, or to MS+ medium

13 supplemented with filter sterilized herbicide (for compound and concentration see Results), but without glutamine and casein hydrolysate, for two weeks in the dark. For the production of shoots,embryogeniecalliwere transferred to MS medium with 20 g/1sucrose,10 mg/lzeafin(MSZ 10) and herbicide and kept under fluorescent light (190 gE-mLsa for 16 h) for 8-10 days. Shoot elongation was achieved on half-strength MS salts and vitamins, supplemented with 15 gO sucrose and herbicide, but without any hormones (MS/2), for 1-2 cycles of 14 days each. In one experiment, callus formation after bombardment was allowedto proceedfor 14 daysin the darkwithout selection on MS+ medium, followed by shoot development on MS+ medium without zeatin under low light intensity (25 gE-m~-s"1 for 16 h) for 28 days. Plantlets(>2.0 em in length) were transferred to tubes with MS/2 medium supplementedwith herbicide.After 2-3 weeks plants were transferred to soil and grown to maturity in growth chambers under the conditions used for the growth of donor plants. Plasmids. The plasmid pAHC25 (Christensen et al. 1992, Vasil et al. 1993) eontsinsthe seleetable bar gene, encoding the enzyme phosphinothricin acetyltransferase (PAT) and the GUS reporter gene (uidA)encoding B-glueuronidase,both driven by the maize ubiquitin promoter, pAHC25 plasmid DNA was used for cotransformation, mixed in a 1:1 molar ratio, with several other potentially useful genes (results on the integration, expression, inheritance and biological effects of these diverse genes will be published elsewhere). Microprojectile bombardment. A total of 10 gl plasmid DNA was precipitated,adsorbed on gold particles (25 I11from a 60 mg/ml stocksolution)and defiveredto target tissue using a DuPont PDS-1000Ate device (Sanford et al. 1991, with minor modifications). The amount of gold used per bombardment (30100 ~tg) was adjusted by the final volume of ethanol in the precipitated gold-DNA suspension. Five gl of the gold-DNA suspension was spread on the surface of the maeroearrier. Enzyme assays. Gus activity was generally assayed histoehemically48 h after bombardment by scoring blue cells 2024 h after incubation in staining buffer at 37°C (Jefferson 1987). PAT activity in leaf extracts was assayed by silica gel thin layer chromatography as described by Spencer et al. (1990). Results

Immature embryo culture and plant regeneration. Immature embryos were bombarded 5-7 days after culture, when they had swollen and formed callus along the edges of the scutellum (Fig. 1A). Although nonbombarded 0.5-1.5 mm immature embryos formed embryogenic calli, the best response after bombardment was obtained from 1.0-1.5 mm embryos. The extent of tissue damage resulting from bombardment was evaluated by comparing the regeneration efficiency witth the nonbombarded controls. Tissue damage depended also on the quality of the donor plants. For example, pesticide treatment of plants close to harvest time had a negative effect on the quality of embryos. Using donor

plants grown under optimal conditions and choosing optimal bombardment parameters, 70-95% of the explants formed embryogenic ealli within 14 days, as was the ease also with the nonbombarded controls. Twenty days after callus initiation (Fig. 2), the calli were transferred onto MS medium with 10 rag0 zeatin and incubated in light. Each embryogenic callus derived from a single embryo formed 320 shoots in 8-10 days (Fig. 1C). Frequency of regeneration was lower when the cultures were kept in the dark for longer periods of incubation, selected immediately after bombardment or regenerated under low light intensity. Optimization of bombardment conditions. The amount of gold particles used ranged from 30-100 gg per bombardment; an average of 73 GUS spots were obtained per embryo bombarded with 30 gg gold. There were 53% more transient expression events with 100 gg gold (Fig. 1/3) than with 30 gg, based on the number of GUS spots. However, regeneration ability was higher after bombardment with 30 ~tg than with 100 gg gold partMes, especially when the quality of donor plants was suboptimal. Osmotic treatment of embryos was optimized by varying the times ofpre- (4 to 9 h) and post-bombardment (16 to 72 h) incubation and osmotic strength in MS+ medium (0.2 M mannitol + 0.2 M sorbitol = 0.4 M osmoticum, 0.5 M mannitol + 0.5 M sorbitol = 1.0 M osmoticum). In each treatment, after bombardment of 480 embryos, GUS expression, regeneration and transformation frequencies were compared with controls cultured on MS+ medium. Highest transient GUS expression (179 spots/callus) was obtained with 4 to 6 h pre-bombardment treatment, followed by 72 h post-bombardment incubation on 1.0 M osmostieum medium. However, the frequency of regeneration after this treatment (1,9 shoots/callus) was even worse than the controls without osmotic treatment (5.3 shoots/callus), and transient expression did not correlate with the fi:equency of stable transformation (0.2%). Highest frequencies of transformation (0.6%) and regeneration (8.2 shoots/callus) were obtained with 4 to 6 h pre-bombardment and 16 h post-bombardment treatment on 0.4 M medium, where on average only 73 GUS spots were observed. Selectiott Selection was imposed either immediately after bombardment (Fig. 2) or only during regeneration in light after callus formation for two weeks in the dark (Figs. 1C-E, 2). In order to attain the maximum number of shoots during the first step of regeneration, bialaphos was used at the relatively low concentration of 3 rag/1 (Figs. lC, 2). This did not have any adverse effect on the number of shoots formed per callus, but slowed their elongation in comparison to the non-selected shoots on control calli.

14

Figure 1. Immature embryos, transient GUS expression, formation and elongation of shoots, and fertile transgenie plants. (A) Callus proliferalion in immature embryos just before bombardment (x40). 03) Transient GUS expression three days after bombardment with pAHC25 (x30). (C) Formation of shoots 10 days after beginning of regeneration on medium with 3mg/1 bialaphos (xl.7). (D) Plantlets 14 days after transfer to MS/2-medium with 5 mg/l bialaphos, before transfer to tubes (xl. 1). (E) Transgenie plants growing on medium with 5 mg/l bialaphos (left) and neerotie/stunted non-transformed plantlets (right), 14 days after transfer to tubes (x0.9). (17)Fertile transgenie plants (x0.1).

15 Plant Seeds -60 d Anthesis --12d 1 Embryo Culture (dark) 5-7 d Callus Initiation Prebombardment Treatment 4-6 h Osmotieum Bombardment 1

PostbombardmentTreatment 12-16 h Osmotieum Culture in the Dark No Selection(DBO) or 3 mg/1Bialaphos(DB3) 14 d CallusFormation 1 Culture in the Light 3 mg/l Bialaphos (LB3) 8-10 d Formation of Shoots Culture in the Light 5 mg/l Bialaphos (LB5) 14 d Elongation of Shoots in Petri dishes 1

Culture in the Light 4-5 mg/1Bialaphos(LB4 or LB5) 14-20 d Elongation of Shoots in Tubes ! Transfer of Plants to Soil -60 d 1 Anthesis -30 d Mature Seeds Figure 2. Timeframefor the productionof transgeniewheat (ev Bobwhite) plants. Times shown are averages for experiments performed in 1995; those designatedby - are approximate and vary either with the batch of donor plants or the individualcallus line or plant. Transgenieplants were transferredto soil 56-66 days after the initiationof cultures. Following a two week culture on a medium with 5 mg/1 bialaphos, 10-40% of the ealli formed at least one plantlet >2.0 crn in length (Figs. 1D, 2). Such calli were transferred individually to tubes with medium containing 4-5 mg/1 bialaphos (the remaining calli were again transferred to medium with 5 mg/1 bialaphos in Petri dishes and were discarded if no shoot elongation had occurred after two weeks). Fourteen to 20 days after transfer to tubes, the putative transformed plants were dark green in color (Fig. 1E, left), and were growing fast with strong roots. These plants could be tested for PAT activity before transfer to

soil. The majority of the non-transformed plantlets, however, were strongly inhibited in growth and became necrotic or chlorotic (Fig. 1E, fight), whereas only a few chlorotic transformed plants - which showed no inhibition of growth - were seen after selection. Data from a large number ofexperiments presented in Table 1 show that after bombardment, only 14 days of callus formation in the dark, with or without selection, followed by selection during regeneration in fight, was needed to obtain transgenic plants at a frequency of 0.3-2.0%. As the frequency of transformation differs between the batches of explants used, direct comparisons were only made within an experiment involving the same batch of embryos. There was not much inhibition of plant growth in the early experiments when 3-4 rag/1 bialaphos was used during shoot elongation in Petri dishes (Exps. 1-3, Table 1). Although this problem could be overcome vdth an additional selection cycle on medium with 3 mg/l basra, 5 mg/1 bialaphos was used in the subsequent experiments (Exps. 4a, 5-8, Table 1) in order to eliminate the extra selection cycle. As a result, the majority of the rooted transgenic plants could be lransferred to soil 64 days (56-66 days) after the initiation of cultures (Fig. 2), Selection applied only during regeneration (Exps. 4a,c, Table 1), instead of three selection cycles in the dark (Exps. 4b,d, Table 1), reduced the time required to obtain rooted transgenic plants by 32 to 40 days. The number of plants recovered under bialaphos or PPT selection during regeneration was comparable (Exps. 4a, e, Table 1). Plants that survived selection but tested negative for PAT activity were considered escapes. There were fewer escapes and the transgenlc plants grew more vigorously under bialaphos selection than PPT. After six weeks of selection in the dark only a few transformants could be regenerated on PPT (Exp. 4d, Table 1), but none on bialaphos (Exp. 4b, Table I). None of the herbicides completely eliminated the escapes during the six week selection in the dark. When selection was imposed exclusively during regeneration, the number of escapes ranged from 55-88% (Exps. 1-3, 4a, 5b, 6a, 7, 8, Table 1). The number of escapes could be reduced by 40-60% if only the fast growing dark green plants were tested. However, 10-20% of the transformed plants that survived transfer to soil were chlorotic after selection and would have been eliminated if only the superior-looking plants had been analyzed (data not shown). Experiments 5 and 6 (Table 1) were carded out in an attempt to further reduce the number of escapes. Both of the alternative protocols - induction of shoots during incubation at low light intensity, or selection immediately after bombardment - reduced the number of escapes compared to selection only during regeneration. The induction of shoots during incubation at low light intensity extended the in vitro culture period by 32 days. Selection immediately after bombardment, as well as selection only during regeneration, produced a comparable number of transformed plants in 64 days (Expts. 5b,c, 6a,b, Table 1). Using the latter two selection protocols, 32 independently

16

Table 1. Summary of eight transformation experiments. Exp. No.

Selection protocol*

Time frame: embryos cultured to plants

Number ofplantlets transferred to tubes

Number of plants surviving selection and tested for PAT activity

1.

DB0/LB3/LB3/LB3/LBa3

-84 days

42

6

1 / 110 (0.9%)

2.

DB0/LB3/LB4/LB4/LBa3

-84 days

75

11

5 / 250 (2.0%)

3.

DB0/LB3/LB4/LB4/LBa3

-84 days

80

8

1 / 350 (0.3%)

4a. 4b. 4e. 4d.

DB0/LB3/LB5/LB4 DB 1/DB 1/DB 1/LB3/LB 5/LB4 DP0/LP3/LP3/LP2 DP 1/I)P 1/DP 1/LP3/LP3/LP2

--64 days -96 days --69 days -104 days

295 55 473 57

62 8 73 10

7/2400 (0.3%) 0 / 2400 (o.0%) 6 / 2400 (0.3%) 3 / 2400 (0.1%)

5a. 5b. 5c.

DB0/LLB0/LB5/LB5/LB5 DB0/LB3/LB5/LB5 DB3/LB3/LB5/LB5

-96 days -64 days -64 days

39 290 133

13 22 10

9 / 2240 (0.4%) 5 / 809 (0.6%) 4 / 889 (0.5%)

6a. 6b.

DB0/LB3/LB5/LB4 DB3/LB3/LB5/LB4

-64 days -64 days

281 115

36 27

14 / 700 (2.0%) 12 / 900 (1.3%)

7.

DB0/LB3/LB5/LB4

-64 days

52

7

2 / 175 (1.1%)

8.

DB0/LB3/LB5/LB4

-64 days

102

19

4 / 325 (1.2%)

Number of PAT positive plants/total bombarded embryos (transformation efficiency)

* D = culture in dark, L = culture at 190 uE-m2-s"~,LL = culture at 25 uE-m2-s"~,B = Bialaphos(mg/1),Ba = Basra (mg/l), P = PPT (mg/l) transformed, rooted, transgenic plants were obtained from 2100 embryos (Expts. 6a,b, 7, 8, Table 1). Stability and fertility of transgenic plants. The transgenic plants derived from selection exclusively during regeneration, as well as those undergoing an additional cycle of selection immediately after bombardment, were fertile (Fig. 1F). However, one out of nine transgenic lines was sterile when selection was imposed following the formation of shoots under low light intensity (Exp. 5a). Progeny analysis has been carried out so far on 24 transgenic lines. PAT assays and Southern blots (data not presented) confirmed the expression and transmission of the selectable bar gene to the R1 generation in 21 of the 24 tested transgenic lines. Discussion We have developed a protocol for accelerated production of transgenic wheat plants, which yields rooted plants for transfer to soil in 56-66 days (8-9 weeks) after the initiation of cultures. This was made possible by improved methods of culture, bombardment and selection. Under optimized conditions, 32 independent, transgenie wheat plants were obtained from 2100 immature embryos (Expts. 6a,b, 7, 8, Table 1). Pre- and post-bombardment osmotic treatment of immature embryos, combined with a low amount of gold particles used per bombardment, minimized tissue damage caused by bombardment. Up to 583 lag of gold particles per bombardment

have been previously used to generate transgenie wheat (Weeks et al. 1993). However, we obtained the most consistent and satisfactoryresults with only 30 pg of gold particles, probably due to less tissue damage than that caused by larger amounts of gold. This is in agreement with the results ofBecker et al. (1994), who observed greatly reduced somatic embryogenesis in seutellar tissue when 116 gg of gold, in comparison to 29 lag, was used for each bombardment. Osmotic treatment of explants, used by Vain et al. (1993) to improve transformation of maize, also gave the highest efficiencyof transformation and regeneration in the present study. Our data, although based on a small sample size, indicate that longer post-bombardment incubation on osmotic medium improved transient GUS expression but reduced the efficiency of regeneration and stable transformation. The fact that stable and transient expression data were not directly correlated, suggests that optimization of transformation protocols should be based on stable rather than transient expression. The short period needed to obtain rooted transgenie plants described by us should help to facilitate the optimization of bombardment conditions based on the number of stable transformants. In an earlier report we described the production of transgenic wheat plants 5-7 months after the initiation of cultures (Vasil et al. 1993). Five of the 12 selected callus lines showing PAT activity could not be regenerated to plants. This was likely related to the prolonged period of callus growth in culture. Therefore, in the present study, the phase of callus formation was reduced to 14 days incubation in dark, with or without selection, followed by selection during regeneration under light (Exps. 1-3,

17 4a, 5b,c, 6a,b, 7, 8, Table 1). Nehra et al. (1994) found phosphinothricin to be inhibitory to embryogenesis in callus cultures of wheat. As previously shown in maize (Dennehey et al. 1994), we also found bialaphos and PPT equally suitable for selection of wheat transformants. However, there were fewer escapes and the transformants grew more vigorously under bialaphos selection, than PPT. The sensitivity of transformed cells to the concentration of the selective agent is dependent on the quality of the callus (Vasil etal. 1993), which is related to the condition of the donor plants (Ozias-Aldns and Vasil 1982). A longer callus selection phase may lead to less reproducible transformation than would a selection protocol,as described here, which relies on the elongation of shoots for identifying transforrnants. The usefulness and reliability of the present method is demonstrated by the fact that 80 transgenie lines were generated in this study (73 are included in Table 1), whereas previous protocols for the production of transgenie wheat lines were based on 6-12 transgenie lines showing PAT,NPT ]I, or EPSPS activity (Vasil et al. 1993, Weeks et al. 1993, Nehra et al. 1994, Beeker et al. 1994, Zhou et al. 1995). For practical use in breeding programs, a large number of independently transformed lines are needed to select plants which show high levels of transgene expression, and its stable transmission to the following generations. Various efforts to improve the efficiency of transformation have not been very successful so far because of poor understanding of the faetors involved in stable integration. Therefore, efforts have focussed on reducing the time required for the production oftransgenie plants, so that a larger number could be obtained in a relatively short period of time. The first transgenie wheat plants were obtained 1215 months after the culture of immature embryos, production of embryogenie callus, bombardment, selection and regeneration (Vasil etal. 1992). In a subsequent study, the time required to produce rooted transgenie plants, at an efficiency of 1.3%, was reduced to 5-7 months by the direct bombardment of cultured embryos (Vasil etal. 1993). Zhou et al. (1995) selected transgenic plants with glyphosate within 4-5 months after culture of embryos and reported a transformation frequency of 0.15%. Weeks et al. (1993), Becker et al. (1994) and Blechl and Anderson (1996) obtained transgenie plants in 15-17 weeks, with a transformation efficiency of 0.1-1.3% (up to 3.6% in one small experiment, Beeker etal. 1994). Nehra et al. (1994) bombarded isolated embryo seutella and reported "the production of transgenie plantlets in 12 weeks from initiation of cultures," at a transformation frequency of 0.5-2.5%. The protocol that we recommend is technically simple and time-saving. It avoids the need to isolate seutella, a tedious and time consuming procedure, and yet yields rooted transgenie plants in 56-66 days (8-9 weeks) with up to 2% transformation effieieney (Exp. 6a, Table 1). Such rapid production oftransgenie wheat plants is desirable in that a large number of independent transgenic lines can be produced and brought to homozygosity in less than a year.

Acknowledgements.We thank Alan H.

Christensen for providigthe plasmid pAHC25. The authors are grateful to Ronald D. Barnett, Phil Bruekner, and Kim Shants for providing excellent wheat plantings in Florida, Montana, and Arizona, respectively. This work was supported by funds provided to IKV by Monsanto Co. (St. Louis, MO). Florida Agriculture Experiment Station Journal Series No. R-05180.

References Beeker D, R Brettschneider, H L6rz (1994) Fertile transgenic

wheat from mieroproprojectile bombardment of seutellar tissue. Plant J 5:299-307. Bleehl AE, OD Anderson (1996) Expression of a novel high molecular weight glutenin subunit in transgenie wheat. Nature Bioteehnology (in press). Christensen AH, RA Sharroek and PH Quail (1992) Maize ubiquitin genes: structure, thermal perturbation of expression and transcript splicing, and promoter activity following transfer to protoplasts by eleetroporation. Plant Mol Biol 18:675-689. Dennehey BK, WL Petersen, C ~'ord-Santino, M Pajeau, CL Armstrong (1994) Comparison of selective agents for use with the selectable marker gene bar in maize transformation. Plant Cell Tissue Organ Cult 36:1-7. JeffersonRA (1987) Assaying chimeric genes in plants: The GUS gene fusion system, Plant Mol Biol Rep 5:387-405. Murashige T, F Skoog (1962) A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol Plant 15:473-497. Nehra NS, RN Chibbar, N Leung, K Caswell, C Mallard, L Steinhauer, M Baga, KK Kartha (1994) Self-fertile transgenic wheat plants regenerated from isolated seutellar tissues following microprojectile bombardment with two distinct gene constructs. Plant J 5:285-297. Ozias-Aldns P, IK Vasil (1982) Plant regeneration from cultured immature embryos and inflorescenees of Triticum aestivum L. (wheat): evidence for somatic embryogenesis. Protoplasma 110:95-105. Sanford JC, MJ De Vit, JA Russell, FD Smith, PR Harpending, MK Roy, S_A.Johnson (1991)An improved, helium-driven biolistie device. Technique 3:3-16. Spencer TM, WJ Gordon-Kamm, 1LI Daines, WG Start, PG Lemaux (1990) Bialaphos selection of stable transformants from maize cell culture. Theor Appl Genet 79:625-631. Srivastava V, V Vasil, IK Vasil (1996) Molecular characterization of the fate of transgenes in transformed wheat (Triticum aestivum L.). Theor Appl Genet (in press). Vain P, MD McMullen, JJ Finer (1993) Osmotic treatment enhances particle bombardment mediated transient and stable transformation of maize. Plant Cell Rep 12:84-88. VasilIK (1994) Molecular improvement of cereals. Plant Mol Biol 25:925-937. Vasil V, AM Castillo, ME Fromm, IK Vasil (1992) Herbicide resistant fertile transgenie wheat plants obtained by mieroprojectile bombardment of regenerable embryogenic callus. Bio/Technology 10:667-674. Vasil V, V Srivastava, AM Castillo, ME Fromm, IK Vasil (1993) Rapid production of transgenie wheat plants by direct bombardment of cultured immature embryos. Bio/Technology 11:1553-1558. Weeks JT, OD Anderson, AE Blechl (1993) Rapid production of multiple independent lines of fertile transgenic wheat (Triticum aestivurn). Plant Physiol 102:1077-1084. Zhou H, JW Arrowsmith, ME Fromm, CM Hironaka, ML Taylor, D Rodriquez, ME Pajeau, SM Brown, CG Santino, JE Fry (1995) Glyphosate-tolerant CP4 and GOX genes as a selectable marker in wheat transformation. Plant Celt Rep 15:159-163.