INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Since fertile transgenic maize (Zea mays) was first produced using the biolistic gun (Gordon-Kamm et al., 1990), maize transformation technology has served as an important tool in germplasm development and research addressing fundamental biological questions through the study of transgenic maize (Armstrong, 1999). Recent reports have demonstrated that Agrobacterium tumefaciens-mediated maize transformation may offer a better alternative than the biolistic gun for delivery of transgenes to maize. This gene delivery system results in a greater proportion of stable, low-copy number transgenic events than does the biolistic gun (Ishida et al., 1996; Zhao et al., 1998), offers the possibility of transferring larger DNA segments into recipient cells (Hamilton et al., 1996), and is highly efficient (Ishida et al., 1996; Zhao et al., 1998). Reproducible protocols for A. tumefaciens-mediated maize transformation have used super binary vectors, in which the A. tumefaciens strain carries extra copies of virB, virC, and virG (Komari, 1990), to infect immature zygotic embryos of the inbred line A188 (Ishida et al., 1996; Negrotto et al., 2000) or the hybrid line Hi II (Zhao et al., 1998, 1999). We have successfully transformed Hi II immature zygotic embryos at an average efficiency of 5.8% using the A. tumefaciens super binary vector in strain LBA4404 (B. Frame, unpublished data). Because the cost of licensing this proprietary technology for use on a broader scale was prohibitive to our public sector laboratory, we focused instead on implementing an A. tumefaciens standard binary (non-super binary) vector system to transform maize Hi II immature zygotic embryos. Stable transformation of maize using a standard binary vector to infect shoot meristems was reported previously (Gould et al., 1991), but adoption of this method was hindered by its lack of robustness. Development of a reproducible and efficient method for transforming maize using a standard binary vector would not only provide researchers with the benefits already outlined, it would also facilitate vector construction when compared with the super binary vector. Final assembly of a super binary vector system involves co-integration of the gene of interest into a large plasmid (pSB1) in A. tumefaciens strain LBA4404 via homologous recombination (Ishida et al., 1996). Assembly of a standard binary vector does not require this additional step, making it a more efficient way to confirm the introduction of a gene of interest into an A. tumefaciens strain.

Using a protocol modified from Zhao et al. (1999), we infected Hi II immature zygotic embryos with a standard binary vector system containing a P35S-bar selectable marker gene cassette (phosphinothricin acetyltransferase gene driven by the cauliflower mosaic virus [CaMV] 35S promoter) and a P35S-gus-int reporter gene cassette (-glucuronidase [GUS] gene with an intron driven by the CaMV 35S promoter) in A. tumefaciens strain EHA101 (Hood et al., 1986). Results from these initial experiments showed that the gus marker gene was preferentially expressed (in transient histochemical assays) in cells at the apical end and on the embryo axis side of infected embryos. In contrast, cells in the basal scutellum region, which were targeted for transformation because of their ability to produce embryogenic callus, showed limited to no expression. Lupotto et al. (1999) reported similar patterns of transient gus gene expression in inbred maize embryos infected with conventional A. tumefaciens binary vectors. Lack of T-DNA delivery to embryogenic-competent scutellum cells and correspondingly low rates of stable event recovery in these experiments led us to consider ways of promoting a more compatible interaction between the A. tumefaciens standard binary vector system and the targeted scutellum cells from which proliferating, embryogenic callus emerges after cocultivation.

In an incompatible host-pathogen interaction, pathogen invasion can be halted by localized plant cell death at the infection site. This resistance mechanism, or hypersensitive response, is mediated by an oxidative burst in which a rapid and transient production of large amounts of reactive oxygen species occur (Wojtaszek, 1997). Antioxidants used during explant germination, preculture, and infection (Enriquez-Obregon et al., 1999) or during and after cocultivation (Perl et al., 1996) were reported to favor stable transgenic event recovery in A. tumefaciens-mediated Japonica rice (Oryza sativa L. cv R321) and grape (Vitis vinifera) transformation experiments, respectively. Olhoft and Somers (2001) measured an increase in the frequency of T-DNA delivery (histochemical GUS assays) to targeted cotyledonary node explants of A. tumefaciens-infected soybean (Glycine max L. Merr. cv Bert) cocultivated on 400 mg L¹ of the antioxidant L-Cys. More importantly, this increase in transient gus gene expression observed 5 d after infection was correlated with a 2-fold increase in the rate of stable event recovery (independent, fertile transgenic soybean plants). The authors concluded that the increase in host-pathogen compatibility mediated by the presence of antioxidants during transformation moderated the detrimental effect of the hypersensitive response, which in turn lead to an increase in the survival rate of A. tumefaciens-infected cells and a corresponding rise in stable transformation efficiency.

Maize callus cells infected with A. tumefaciens were described as undergoing a rapid, hypersensitive type of cell death in a study characterizing A. tumefaciens-induced apoptosis in maize (Hansen, 2000). Use of antioxidants to moderate the A. tumefaciens-maize interaction has not been reported. We supplemented cocultivation medium with 400 mg L¹ Cys, carried out infection of Hi II immature zygotic embryos with the standard binary vector system, and measured the effect of this antioxidant treatment on three transformation criteria. First, analysis of transient GUS expression in infected embryos was performed to monitor T-DNA delivery to the targeted cells (no. and distribution of blue foci). Because recovery of cells targeted for transformation is critical to achieving stable transgenic events, we also assessed the proportion of embryos that gave rise to embryogenic Type II callus after cocultivation. Finally, stable transformation efficiency (the no. of bialaphos-resistant transgenic events per 100 embryos infected) was measured for Cys concentrations ranging from 100 to 400 mg L¹.

Stable transformation efficiency was significantly increased upon addition of L-Cys to cocultivation medium. Using the protocol described, we have achieved routine production of fertile transgenic maize at an efficiency of 5.5%. To our knowledge, this marks the first report in which a reproducible method for maize transformation using an A. tumefaciens standard binary vector system has been demonstrated.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Transient GUS Expression

Immature zygotic embryos infected with A. tumefaciens strain EHA101 harboring the standard binary vector pTF102 (Fig. 1) were cocultivated as described in "Materials and Methods." Control embryos were cocultivated on medium without Cys to compare the amount and distribution of T-DNA delivery to targeted cells cocultivated in the presence or absence of 400 mg L¹ Cys. Histochemical GUS analysis was carried out on a subset of embryos cocultivated with Cys (138 embryos) or without Cys (146 embryos) across 10 independent experiments. Level of transient GUS expression (no. of blue foci on the scutellum side of each embryo) was determined as described in "Materials and Methods." Results are graphically presented in Figure 2A. The majority (56%) of embryos cocultivated on 400 mg L¹ Cys were moderate to high GUS expressers compared with those cocultivated in the absence of Cys (17%). Conversely, a high proportion (50%) of embryos treated with no Cys showed no blue foci on the scutellum side of the embryo compared with those embryos cocultivated on 400 mg L¹ Cys (15%).

View larger version (9K): [in this window] [in a new window]   Figure 1.   T-DNA region of standard binary vector pTF102. LB, Left border; RB, right border; bar, phosphinothricin acetyltransferase gene; gus-int, -glucuronidase gene containing an intron; P35S, CaMV 35S promoter; TEV, tobacco etch virus translational enhancer; Tvsp, soybean vegetative storage protein terminator; T35S, CaMV 35S terminator; H, HindIII.

View larger version (17K): [in this window] [in a new window]   Figure 2.   Effect of Cys (0 or 400 mg L1) in cocultivation medium on level of transient GUS expression in embryos (A) and on embryogenic callus response of A. tumefaciens-infected or noninfected embryos (B). Data from 10 experiments (A), eight experiments (B, with Agro), and three experiments (B, without Agro).

The change in distribution of blue foci on embryos cocultivated in the presence of Cys was of particular significance. On embryos cocultivated in the absence of Cys, blue foci were generally localized to the embryo axis side and apical end of the embryo (Fig. 3a). Expression was rarely observed in the basal scutellum region of the embryo that, as the site of callus initiation on the immature zygotic embryo explant, was targeted for transformation. In contrast, in embryos cocultivated on 400 mg L¹ Cys, transient GUS expression was consistently observed not only at the apical end of the embryo, but also along the margins and in targeted cells at the base of the scutellum (Fig. 3b).

View larger version (59K): [in this window] [in a new window]   Figure 3.   Distribution of blue foci (transient gus gene expression) in maize Hi II immature zygotic embryos infected with A. tumefaciens standard binary vector system, pTF102, and cocultivated on medium containing 0 mg L1 (a) or 400 mg L1 Cys (b). Embryos during cocultivation (c). Putative transformant growing on 3 mg L1 bialaphos (d). Bialaphos-resistant, Type II embryogenic callus event emerging from a single embryo explant (e). Stable gus gene expression in pTF102 callus event (f, left). Callus on the right is not expressing the gus gene. Transgenic pTF102 plants (R0) flowering in greenhouse (g). Transgenic seed set on ear of R0 plant (h). Gus (i) and bar (j) transgene expression in segregating R1 progeny. The leaf segment on the left was isolated from a GUS-expressing plant (positive), whereas that on right was from nonexpressing (negative) plant (i). The bar-expressing plantlet on the left (resistant) survived the glufosinate herbicide spray, whereas the nonexpressing plant on the right (sensitive) died.

Post-Cocultivation Embryo Response

To assess the effect of 400 mg L¹ Cys in cocultivation medium on recovery of targeted cells, embryos were infected and cocultivated as described, and then transferred to resting medium for 4 to 7 d, after which they were assessed for Type II callus initiation. Although inclusion of 400 mg L¹ Cys in cocultivation had a positive impact on transient GUS expression in embryos, it reduced the proportion of embryos giving rise to embryogenic Type II callus on resting medium. Average percent embryo response (no. of embryos giving rise to embryogenic callus per 100 embryos treated) was assessed over eight independent experiments. Sixty percent of 341 A. tumefaciens-infected embryos cocultivated on 400 mg L¹ Cys produced embryogenic callus compared with 91% of 374 embryos cocultivated in the absence of Cys (Fig. 2B). Of 123 non-infected embryos cultured on cocultivation medium containing Cys, 52% produced embryogenic callus while 99% of 127 non-infected embryos incubated on cocultivation medium without Cys produced embryogenic callus (Fig. 2B).

Stable Transformation Efficiency

To test the effect of Cys in cocultivation medium on stable transformation, embryos infected with pTF102 were cocultivated for 3 d on media supplemented with 400 mg L¹ Cys (Fig. 3c) and then transferred to resting and selection media as described in "Materials and Methods." Beginning 5 weeks after infection, bialaphos-resistant callus events were identified by their sustained embryogenic growth on medium containing 3 mg L¹ bialaphos (Fig. 3, d and e) and were histochemically analyzed for stable expression of the gus gene (Fig. 3f).

Results from 15 independent experiments are summarized in Table I. Stable transformation efficiency (no. of bialaphos-resistant events recovered per 100 embryos infected) in these experiments averaged 5.5% and ranged between 1.1% and 22.2%. All but experiment 12/08/00 produced transgenic events, emphasizing the reproducibility of this method. Seventy-eight percent of bialaphos-resistant events also expressed the gus gene. Of the 51 callus events for which regeneration was attempted, all but one regenerated to plants (Table I).

Cocultivation on medium containing 400 mg L¹ Cys significantly increased stable event recovery compared with the control treatment in which no Cys was added to the cocultivation medium. In eight separate experiments, infected embryos were cocultivated at 23°C on medium containing either 0 or 400 mg L¹ Cys (Table II). The average percent stable transformation efficiency for embryos cocultivated on 400 mg L¹ Cys was higher (5.3%) than that for embryos cocultivated on 0 mg L¹ Cys medium (0.2%). This difference was statistically significant (P < 0.001). In addition, reducing cocultivation temperature from 23°C to 20°C did not alter the favorable effect of the Cys cocultivation treatment on stable event recovery. Whereas the average transformation efficiency at 20°C (6.2%) was somewhat higher than at 23°C (5.3%) in embryos cocultivated on 400 mg L¹ Cys, few stable events were recovered at either cocultivation temperature when embryos were cocultivated in the absence of Cys (Table II).

To test whether Cys concentrations lower than 400 mg L¹ in cocultivation medium would lead to an improvement in callus formation from treated embryos and a higher stable transformation rate, we conducted a preliminary experiment in which embryos infected with pTF102 were cocultivated on medium containing 0, 100, 200, 300, or 400 mg L¹ Cys. The highest rate of stable event recovery in these experiments was achieved from embryos cocultivated on medium containing 100 mg L¹ Cys (4.8%). Stable transformation efficiency for the 400 mg L¹ Cys treatment was 3.3%, and for the non-Cys control was 0.0% (Table III). The proportion of embryos responding after cocultivation decreased as Cys concentration increased. In contrast, transient GUS expression levels increased as Cys concentration increased, and showed a gradual shift from the embryo axis side and apical end of embryos (0 mg L¹ Cys) to the scutellar margins of embryos (100 and 200 mg L¹ Cys), to expression in all regions of the scutellar side of embryos, including the embryogenic competent basal scutellar cells at Cys levels of 300 and 400 mg L¹ (data not shown).

Plant Regeneration and Fertility

Bialaphos-resistant callus events were regenerated on medium containing 3 mg L¹ bialaphos as described in "Materials and Methods." To date, over 60 pTF102-derived transgenic events have been regenerated to plants and are growing in the greenhouse (Fig. 3g). Seed harvested from outcrosses of 97 plants representing 37 of these events averaged 51 kernels per ear (transgenic plant as female, Fig. 3h) and 60 kernels per ear (transgenic plant as male), indicating that these events are fertile. Thirty-five of these events were recovered from Cys-containing cocultivation medium, whereas the remaining two derive from the non-Cys cocultivation treatment.

Analysis of Stable Events

Histochemical GUS assays were carried out on all bialaphos-resistant callus events to determine whether those expressing the bar gene also expressed the gus reporter gene. Of 65 events analyzed, 51 (or 78%) were GUS positive (Table I). Because the gus gene in this construct contains an intron, blue staining was indicative of plant rather than A. tumefaciens expression of the transgene. Positive or negative GUS expression in leaf tissue correlated with callus expression of the reporter gene in all events.

Southern-blot analysis was carried out to assess stable integration of the bar and gus transgenes in the R₀, R₁, or R₂ generations of numerous independent pTF102 events. Total genomic DNA was extracted from leaf tissue, digested with the restriction enzyme HindIII, and DNA blots prepared and hybridized with the gus or bar probes. As illustrated in Figure 1, HindIII restriction digestion of genomic DNA derived from pTF102 transgenic material would not only liberate a 3-kb fragment containing the gus gene cassette but would also yield various band sizes (>2.5 kb) that hybridize to the bar probe. The number of bands hybridizing with the bar probe would, in turn, reflect the different sites of transgene integration and represent an estimate of transgene copy number in the maize genome.

Samples of genomic DNA from 34 independent events were hybridized with the gus probe. Thirty-one of these events (91%) showed the expected 3-kb dropout band, suggesting the intactness of the gus gene cassette in the maize genome. A subset of 20 events is shown in Figure 4A. Some events contain multiple gus-hybridizing bands in addition to the predicted 3-kb dropout (Events 1, 13, 16, 40, 46, and 48, Fig. 4A). The 3-kb gus-hybridizing band is visible in all but Events 2, 6, and 23. Hybridizing bands bigger or smaller than 3 kb in these three events suggest that the gus gene cassette was rearranged (Fig. 4A) and may account for the lack of GUS expression in callus or plants of Events 6 and 23 (Table IV). However, Event 2 also carries a rearranged copy of the gus gene (Fig. 4A), but does express GUS in callus and plants. Conversely, Event 4 shows the expected 3-kb band (Fig. 4A), but does not express the gus gene in either callus or plants.

View larger version (53K): [in this window] [in a new window]   Figure 4.   Southern-blot analysis of transformed plants. Leaf tissue was sampled from 20 independent events in the R0 generation (6, 8, 10, 12, 13, 15, 16, and 18), the R1 generation (1, 2, 4, 7, 14, 23, 40, 44, 46, and 48), or the R2 generation (3, 5). No parent-progeny relationships are depicted. Ten micrograms of genomic DNAs was digested with HindIII and hybridized with the gus (A) or bar probe (B). ctr, Negative control (genomic DNA from non-transformed maize inbred line B73); +ctr, positive control (non-transgenic maize genomic DNA spiked with 80 pg of pTF102 plasmid DNA digested with HindIII).

Duplicate DNA blots of 27 events hybridized with the gus probe were also hybridized with the bar probe and results for 20 of these events are shown in Figure 4B. Ninety-two percent of the events analyzed had various hybridizing bands larger than 2.5 kb, confirming the presence of the bar gene cassette and the T-DNA left border region in the maize genome. Events 2 and 23 contained a small bar-hybridizing band (<2.5 kb), suggesting that some gene rearrangement or truncation occurred between the gus and the bar genes in these events. In our analyses to date, 90% of events have shown three or fewer hybridizing bands and no event has yielded more than five bands when hybridized with the bar probe.

Progeny screening of R₁ and R₂ generation seedlings was carried out on a subset of transgenic events to assess whether the gus and bar transgenes were inherited in a normal Mendelian fashion. Seed for these tests was derived from female transgenic plants that had been pollinated (outcrossed) with non-transgenic pollen. As such, the expected segregation ratio for inheritance of the transgene as a single locus was 1:1. Leaf samples from progeny plants of 17 independent events were assayed for histochemical GUS expression (Fig. 3i) and plantlets were then sprayed with glufosinate to test for expression of the bar gene (Fig. 3j). Three events (Events 4, 6, and 23) did not express the gus gene in callus (data not shown) or progeny plants (Table IV). For the majority of GUS-expressing events (11 of 14 events tested), the observed segregation ratio for the gus gene was as expected. Likewise, segregation ratios for bar gene expression confirmed normal Mendelian inheritance of the transgene in 14 of 17 events (Table IV). Whereas Events 16 and 45 showed abnormal segregation for both transgenes, Event 48 showed a 3:1 segregation ratio for the gus and bar genes, suggesting that each of the transgenes is segregating at two rather than one loci in this event.

R₂ generation progeny plants of Events 1 through 5 were also screened for expression of the gus and bar transgenes. Expected segregation ratios for both transgenes were observed except for Event 1 (gus gene) and Event 2 (bar gene, Table IV).

Although Southern blots of Events 2 and 23 indicate some rearrangement and truncation in the bar gene, both segregated normally for bar gene expression in the R₁ generation. However, segregation of R₂ progeny in Event 2 was abnormal (Table IV). Detailed sequence analysis will be required to determine the molecular nature of these events.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

We have produced numerous transgenic events from Hi II immature zygotic embryos transformed using an A. tumefaciens standard binary vector system. Southern-blot analyses confirm integration of the transgenes in the maize genome. Transgenic plants from these experiments are fertile and we have observed normal inheritance and stable expression of the bar and gus transgenes in R₁ and R₂ progeny in the majority of events. In this study, we have demonstrated the reproducibility of this method using the pTF102 vector system. In separate studies, we have also used this protocol to introduce a number of other genes of interest into maize using the same vector backbone (B. Frame, unpublished data).

In all pTF102 events analyzed, estimated transgene copy number ranged from one to five copies per genome. A similar range in copy number estimate was reported by Zhao et al. (1998) in transgenic events derived from A. tumefaciens super binary vector transformation of maize Hi II immature zygotic embryos. In contrast, copy number estimates for biolistic gun-derived transgenic events of maize can range as high as 20 (H. Shou and R. Chikwamba, unpublished data).

The level of stable transformation achieved in this study is attributed to supplementation of cocultivation medium with 400 mg L¹ Cys. This antioxidant treatment also increased T-DNA delivery to embryogenic-competent scutellum cells of infected embryos. A similar increase in transient gus gene expression, followed by an increase in stable transformation efficiency, was reported in soybean cotyledonary node explants infected with A. tumefaciens and cocultivated on medium supplemented with Cys (Olhoft and Somers, 2001). Komari and Kubo (1999) suggested that the main hurdle in A. tumefaciens-mediated maize transformation may not be the infection step but may be the recovery of cells that have integrated the T-DNA into their chromosomes. Cys in cocultivation medium may be acting to minimize cell death caused by the hypersensitive response of maize scutellum cells to A. tumefaciens infection. This would favor postinfection survival of embryogenic-competent cells, thereby increasing stable transformation efficiency. Notably, we have observed an interaction between the Cys cocultivation treatment described in this study and treatments aimed instead at increasing A. tumefaciens virulence prior to infection. For example, in separate comparisons, reducing the 3-d incubation temperature for A. tumefaciens cultures from 28°C to 19°C and pre-inducing A. tumefaciens in plant infection medium supplemented with acetosyringone (AS) both lead to an increase in stable transformation, but only if infected embryos were subsequently cocultivated on 400 mg L¹ Cys (B. Frame, unpublished data).

Contrary to our expectation, the increase in stable transformation efficiency observed with the 400 mg L¹ Cys treatment was associated with a decrease in the proportion of embryos giving rise to embryogenic callus compared with the 0 mg L¹ Cys treatment. We know that this reduction in embryo response is not related to the plant-pathogen interaction per se because noninfected embryos also exhibited reduced response on 400 mg L¹ Cys. It is likely that Cys concentrations as high as 400 mg L¹ are toxic to maize cells. A similar negative impact of 80 mg L¹ Cys on embryogenesis in Japonica rice explants was reported by Enriquez-Obregon et al. (1999). We have achieved comparable stable transformation rates using Cys concentrations as low as 100 mg L¹, and this treatment was associated with better embryo recovery after cocultivation than that observed using the 400 mg L¹ Cys treatment. Further studies are under way in our laboratory to define the optimum Cys concentration for stable transformation.

We have demonstrated that A. tumefaciens-mediated maize transformation using a standard binary vector system is reproducible although variability in experimental efficiency persists. In our experience, using cocultivation medium within 7 d of preparation minimizes this variability. Although our average transformation efficiency of 5.5% is low compared with that reported for Hi II immature zygotic embryos transformed using the super binary vector system (33%-51%, Zhao et al., 1998), further improvements in transformation efficiency using this procedure are likely to result from optimization of the Cys concentration used in cocultivation medium, or by examining the effect of using other antioxidant compounds that are less toxic to maize. Consideration of interactions between Cys and experimental factors such as cocultivation duration and timing to selection, or parameters involved in A. tumefaciens virulence induction, may also lead to further improvements in stable transformation efficiency using A. tumefaciens standard binary vectors to transform maize.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Agrobacterium tumefaciens Vector and Strain

A. tumefaciens strain EHA101 (Hood et al., 1986) containing the standard binary vector pTF102 (12.1 kb) was used in all experiments. The 5.9-kb T-DNA region of this construct is shown in Figure 1. The vector is a derivative of the pPZP binary vector (Hajdukiewicz et al., 1994) that contains the right and left T-DNA border fragments from a nopaline strain of A. tumefaciens, a broad host origin of replication (pVS1) and a spectinomycin-resistant marker gene (aadA) for bacterial selection. The CaMV 35S promoter (P35S) was used to drive both the bar selectable marker gene and the gus reporter gene. A tobacco etch virus translational enhancer (Carrington and Freed, 1990) was included in the 5' end of the bar gene. The soybean (Glycine max L. Merrill) vegetative storage protein terminator (Mason et al., 1993) was cloned to the 3' end of the bar gene. The gus gene contained a portable intron in its codon region (Vancanneyt et al., 1990) to prevent GUS activity in A. tumefaciens cells. This vector system, pTF102 in EHA101, was maintained on yeast extract peptone (YEP) medium (An et al., 1988) containing 100 mg L¹ spectinomycin (for pTF102) and 50 mg L¹ kanamycin (for EHA101). Bacteria cultures for weekly experiments were initiated from stock plates that were stored for up to 1 month at 4°C before being refreshed from long-term, 80°C glycerol stocks. In all experiments, bacteria cell densities were adjusted to an optical density (OD₅₅₀) between 0.35 to 0.45 using a spectrophotometer immediately before embryo infection.

Plant Material

F₂ immature zygotic embryos (1.5-2.0 mm) of the maize (Zea mays) Hi II hybrid genotype (Armstrong et al., 1991) were aseptically dissected from greenhouse-grown ears harvested 10 to 13 d post pollination. Ears were stored up to 3 d at 4°C before dissection.

Media

Infection, cocultivation, resting, and selection media were after Zhao et al. (1999) except that cocultivation medium was modified to contain Cys. All these media contained N6 salts and vitamins (Chu et al., 1975), 1.5 mg L¹ 2,4-dichlorophenoxyacetic acid, and 0.7 g L¹ L-Pro in addition to the following ingredients: infection medium contained 68.4 g L¹ Suc and 36 g L¹ Glc (pH 5.2) and was supplemented with 100 µM AS (Sigma, St. Louis) before use; cocultivation medium contained 30 g L¹ Suc, 0.85 mg L¹ silver nitrate, 100 µM AS, and 3 g L¹ gelrite (pH 5.8); resting medium contained 30 g L¹ Suc, 0.5 g L¹ MES, 0.85 mg L¹ silver nitrate, 250 mg L¹ cefotaxime, and 8 g L¹ purified agar (pH 5.8). Selection medium was identical to resting medium with the addition of 1.5 or 3 mg L¹ bialaphos (Shinyo Sanyo, Tokyo). Infection medium was filter sterilized, whereas all other media were autoclaved. AS stock solutions (100 mM) were prepared by dissolving AS in 100% (v/v) dimethyl sulfoxide (DMSO) to make a 200 mM stock which was then diluted (1:1 [v/v]) with sterile water and stored in small aliquots at 20°C. Cys was added to cocultivation medium after autoclaving from freshly prepared, filter-sterilized stocks (100 mg mL¹) and cocultivation medium was used within 2 to 5 d of preparation. Regeneration I medium contained Murashige and Skoog salts and vitamins (Murashige and Skoog, 1962), 60 g L¹ Suc, 100 mg L¹ myo-inositol, no hormones, and 3 g L¹ gelrite (pH 5.8) after Armstrong and Green (1985). Cefotaxime (250 mg L¹) and bialaphos (3 mg L¹) were added to this medium after autoclaving. Regeneration II medium differed from medium I in that it contained 30 g L¹ Suc and no bialaphos. All media was poured to 100- × 25-mL plates.

Infection and Cocultivation

A. tumefaciens cultures were grown for 3 d at 19°C on YEP medium amended with 100 mg L¹ spectinomycin and 50 mg L¹ kanamycin. One full loop (3 mm) of bacteria culture was scraped from the 3-d-old plate and suspended in 5 mL of liquid infection medium (Inf) supplemented with 100 µM AS (Inf + AS) in a 50-mL falcon tube. The tube was fixed horizontally to a bench-top shaker or a Vortex Genie platform head and shaken on low speed (approximately 75 rpm) for 4 to 5 h at room temperature. This pre-induction step was carried out for all experiments. For infection, immature zygotic embryos (1.5-2.0 mm) were dissected to bacteria-free Inf + AS medium (1.8 mL) in 2-mL eppendorf tubes (20-100 embryos per tube) and washed twice with this medium. The final wash was removed and 1 to 1.5 mL of A. tumefaciens suspension was added to the embryos. Embryo infection was accomplished by gently inverting the tube 20 times before resting it upright for 5 min with embryos submerged. Embryos were not vortexed at any time during this procedure. After infection, embryos were transferred to the surface of cocultivation medium and excess A. tumefaciens suspension was pipetted off the medium surface. Cocultivation medium contained 400 mg L¹ Cys unless stated otherwise. In experiments in which cocultivation medium treatments were compared, embryos were washed and infected in the same tube before being distributed between media treatments. Embryos were oriented with the embryo-axis side in contact with the medium (scutellum side up). Plates were wrapped with vent tape (Vallen Safety Supply, Irving, TX) and incubated in the dark at 20°C or 23°C for 3 d, after which embryos were transferred to 28°C on resting medium.

Embryo response (%) was measured as the number of cocultivated immature zygotic embryos that had initiated embryogenic Type II callus formation at their scutellum base after 4 to 7 d on resting medium, compared with the total number plated. All embryos, responding or not, were transferred to selection medium.

Selection and Regeneration

After 4 to 7 d on resting medium (28°C, dark), embryos were transferred to selection medium (30 per plate) containing 1.5 mg L¹ bialaphos. Selection was increased to 3 mg L¹ bialaphos 2 weeks later. Putatively transformed events were identified as early as 5 weeks after infection. Regeneration of R₀ transgenic plants from Type II embryogenic callus was accomplished by a 2- to 3-week maturation step on Regeneration Medium I followed by germination in the light on Regeneration Medium II as described by Frame et al. (2000). Stable transformation efficiency (%) was calculated as the number of bialaphos-resistant callus events recovered per 100 embryos infected.

Acclimatization and Greenhouse Care of Transgenic Plants

Transplant and acclimatization of regenerated R₀ plants was accomplished as described previously (Frame et al., 2000). Transgenic plants were grown to maturity in the greenhouse.

Statistical Analysis

Data from eight independent experiments were used to compare stable transformation efficiency from pairs of plates treated alike aside from Cys exposure during cocultivation. A sign test was used to determine whether the benefit in transgenic event recovery rate observed for the 400 mg L¹ Cys treatment was significantly higher than that for the 0 mg L¹ Cys treatment.

A Chi square test was used to determine whether the segregation ratios we observed for gus and bar gene expressing progeny plants fit the expected 1:1 ratio.

Histochemical Analysis of Transient and Stable gus Expression

Histochemical GUS assays (Jefferson, 1987) were used to assess transient expression of the gus gene in immature zygotic embryos 1 or 2 d after the 3-d cocultivation (4 or 5 d after infection). Level of transient gus expression was assessed on a per embryo basis by estimating the number of blue foci visible on the scutellum side of each embryo. Embryos displaying blue foci only on the embryo-axis side of the explant were scored as non-expressors. The embryo was then categorized as follows: nonexpresser (no blue foci), low expresser (one-25), moderate expresser (26-100), or high expresser (>101). The number of embryos in each of these four groups was compared with the total number of embryos assessed to determine percent of total embryos in each of the expression categories. Histochemical GUS assays were also used to assess stable expression of the gus gene in bialaphos-resistant callus samples and in leaf tissue of transgenic plants in the R₁ and R₂ generations. Leaf segments (0.5 cm) were submerged in the substrate, vacuum infiltrated (20 inch Hg) for 10 min, and incubated at 37°C overnight. Blue staining cells were visualized by soaking leaf tissue in 75% followed by 95% (v/v) ethanol to remove chlorophyll and leaf pieces scored as positive or negative for GUS expression.

Southern-Blot Analysis

Leaf genomic DNA was prepared from 2 to 3 g of fresh leaf tissue from putative transgenic maize plants using the cetyltrimethylammonium bromide (CTAB) method, as described by Murray and Thompson (1980). Ten micrograms of genomic DNA per sample was digested with the HindIII restriction enzyme at 37°C overnight and separated on a 0.8% (w/v) agarose gel. DNA gel-blot analyses (Sambrook et al., 1989) were conducted on DNA samples using the ³²P-labeled bar or gus fragments as shown in Figure 1.

Progeny Segregation Analysis for bar Gene Expression

A glufosinate leaf-spray test (Brettschneider et al., 1997) was used to establish segregation ratios for expression of the bar gene in progeny. The herbicide Liberty (Aventis, Strasbourg, France) was dissolved in water (1.25 mL L¹) along with 0.1% (v/v) Tween 20 for a final glufosinate concentration of 250 mg L¹. Beginning 9 d after planting, seedlings were sprayed three times at 1- to 2-d intervals with a freshly prepared glufosinate solution and then scored for herbicide resistance (alive) or herbicide sensitivity (dead).

Distribution of Materials

Upon request, all novel materials described in this publication will be made available in a timely manner for non-commercial research purposes, subject to the requisite permission from any third-party owners of all or parts of the material. Obtaining any permission will be the responsibility of the requestor.