Out of 101 plants, twenty, belonging to five independent lines (71-1, 45-1, 45-3, 44-1, and 4-1) were shown by Southern blot analysis to contain a 1.5 kb hybridizing band corresponding to the intact bar gene (Fig. 1). All 20 plants also contained the expected 1.9 kb band that hybridized to the vspB gene. The same probe detected two bands in EcoRI restricted Soybean genomic DNA, ~5.7 and 8.6 Kb, corresponding to the highly homologous genes vspA and vspB 13.

Western blot analysis was used to detect the VSPβ polypeptide in leaf extracts from fourteen primary (R₀) transgenic maize plants at vegetative stage (7 weeks old). A distinct VSPβ band was not visible in silver-stained SDS-PAGE separated maize extract proteins due to complexity of the total protein pattern and the relatively low level of VSPβ expression (Fig. 2a). However, some of these plants expressed VSPβ at a level high enough to be detected by Western blot analysis (Fig. 2b). Two plants (71-1-53 and 45-3-1) had the highest level of VSPβ, while four plants (71-1-23, 71-1-20, 45-1-4, and 45-1-7) accumulated a lower level of VSPβ. Although a faint lower MW peptide band was visible in negative control maize extracts, VSP was clearly only present in transgenic lines. Faint detection of bands in untransformed maize (none of identical size to soybean VSPβ) is consistent with previous reports of cross-hybridizing of soybean VSP antibodies with proteins from monocots 14. In soybean extracts used as a positive control, the antibody for VSPβ recognized both VSPα and VSPβ polypeptides (Fig. 2b,c)

Computer analysis of digital images of the Western blots was used to detect differences in relative band intensity of the immunologically detected VSPβ peptide. Because the native soybean VSPs (VSPα plus VSPβ) were easily visible on total protein stained gels (Fig. 2a), relative quantification of total stained proteins in the soybean samples indicated that the VSPs represented about 10% of the total soluble protein in young soybean leaves. This is close to the previously reported value of 15 % 12. The intensity of the transgenic maize 45-3-1 VSPβ band on Western blots was 44% of the soybean VSP's band (digital image pixel quantification of Fig. 2b). Accounting for the differences in total protein applied to the gel (less soybean total protein was loaded), the VSPβ protein was estimated to have accumulated to 0.5 % of the total soluble protein. This is similar to the highest level of seed storage protein accumulation observed with the ectopic expression of zein 89, but remains less than the 1% minimal expression level predicted by Wandelt et al. 11 to be needed to directly alter the nutritional quality of the leaves. Although the 0.5% of total soluble protein was too low to alter nutritional value, detection of VSPβ in 45-3-1 allowed monitoring of VSPβ level in leaves and stems during plant development.

R₁ plants were produced by back-crossing the R₀ plants with Hi II control non-transformed pollen. Back-crossing was performed because R₀ plants directly regenerated from tissue culture did not have synchronized production of pollen and receptive female flowers. The R₁ families segregating for bar expression were analyzed for the presence of the vspB gene by Southern blot analysis. From 57 R₁ plants analyzed, Southern blot analysis showed that 35 (61%) contained the vspB gene integrated into the genome. Figure 3 shows examples for several R₁ from five R₀ parental lines. The ~1.9 kb vspB band can be seen in nine of the 16 R₁ lines. There are also fainter bands one slightly larger than the 1.9 kb band and one or two migrating between 4 and 5 kb. These are often observed as incomplete restriction of all Eco RI sites internal to plasmid DNA that is integrated into the maize genome. Similar bands are observed even with the plasmid control.

Total RNA was isolated from leaves of young R₁ plants at the vegetative stage (7 weeks after planting). Eighteen transgenic R₁ plants, including the ones originating from R₀ plants representing different levels of VSPβ accumulation (high VSPβ accumulators: 45-3-1, 71-1-53; mid-level VSPβ accumulators: and 71-1-23, 71-1-20, 45-1-4, 45-1-7 and low VSPβ accumulators: 44-1-1- 4-1-2), were analyzed for vspB transcripts. Because of high sequence similarity (85%) between vspA and vspB cDNAs, the same probe hybridized in soybean with both mRNAs as demonstrated by Staswick 13. The transgenic plants 71-1-53A, 71-1-20A, 71-1-20E, and 71-1-20I produced a hybridizing band of approximately the expected size of 1.1 kb indicating transgene expression (Fig. 4a); however, most did not. Use of the Ubi-1 promoter for both the vspB and bar gene probably led to a high number of transgenics with the vspB silenced, and a higher level of VSPβ accumulation will likely result in future work using a combination of different promoters.

Immunodetection of VSPβ protein in transgenic young plant leaves showed variation in accumulation in comparisons between the parental (R₀) and their R₁ progeny, with the greatest variation observed with the highest VSPβ expressing R₀ plants (Fig 4b). The parental line 71-1-53 had a relatively high level of expression of VSPβ, but the only transgenic offspring from this line, 71-1-53A, had almost undetectable VSPβ; in contrast, a low level accumulator, 71-1-20, produced R₁ lines expressing different levels of VSPβ, although none of them expressed at higher levels than the parent. Quantification of the relative level of VSPβ expressed in the R₁ leaves showed that the highest level measured was only 0.03% of total soluble protein.

Despite the overall low level of VSPβ expression, for the purposes of this work, VSPβ accumulation in several of the plants was high enough to study the relationship of plant developmental stage and vspB/VSPβ accumulation. Both the transcript abundance and VSPβ protein accumulation were determined in the R₁ lines using real-time quantitative RT-PCR and Western blot immunodetection, respectively. Real-time RT-PCR is more sensitive that Northern blot analysis and was able to detect transcripts that were not seen with standard total RNA blotting methods. The vspB transcript was quantified in leaves from immature plants (prior to tassel formation) and silage stage plants (plants with developing seeds at the 18 DAP-days after pollination stage), as well as, stems from the silage stage plants, (Figure 5). The vspB transcript was detected in all five transgenics (four of which had RNA not detectable using standard Northern blot methods). The relative level of RNA among the different plant samples was not consistent across the different lines with some having more transcripts in the young leaves while others had more in the older leaves and stems.

When vspB RNA abundance was compared to VSPβ protein accumulation there was no correlation between the two (Fig. 5a, b, c). Antiserum to VSPβ detected VSPβ only in leaves from young maize plants that had not yet flowered. No VSPβ was detected in leaves and stems of silage stage corn that had developing seed. The soybean VSPβ peptide was the primary band reacting with the antiVSPβ antiserum in young leaves of transgenic maize, however, in silage stage stems and to a lesser extent the silage stage leaves, there were multiple bands detected at a different size than the VSPβ. These were also detected in the non-transgenic control plant samples and therefore, were not due to different modifications of the soybean VSPβ. Although the VSPβ protein dropped below detectable levels in the silage stage 71-1-20A, 45-3-1F and 45-3-1-G plants, the vspβ transcript was still detectable. In fact, in the 71-1-20A plants the transcript level was highest in the silage stage leaves that had no detectable VSPβ. Therefore, post-transcriptional events (i.e., changes in either the translational efficiency of the vspB transcript or the protein stability of VSPβ) were altered in the silage stage leaves and stems as compared to the leaves of immature plants.

Plants of hybrid "Hi-II" maize were established in a greenhouse and immature tassels were used for embryogenic type II callus production, as described by Armstrong 19.

Microprojectile bombardment of callus was performed using the procedure of Somers et al. 20. Calli were cobombarded with equal amounts of pRSVP-1 (Shatters Jr., unpublished) and pAHC25 21. Plasmid pRSVP-1 was constructed by restricting the soybean vspB cDNA clone (998 bp) from pKSH3 22 with EcoRI, blunt ends of this fragment were produced using S1 nuclease and the fragment was cloned into similarly blunt ended BamHI restricted pAHC17 21. As a result, the vspB coding region was inserted downstream of the Ubi-1 promoter and a 5' untranslated region (exon) and intron; and upstream of the nos terminator sequence. The plasmid pAHC25 carried the bar gene and the uidA reporter gene, both under the control of the Ubi-1 promoter. Bombardments were performed with a Biolistic^® PDS-1000/He Particle Delivery System (Bio-Rad Laboratories, Hercules, CA) and an osmotic treatment was applied to reduce the cell damage caused by the gene transfer method 23. Putative transgenic maize were regenerated from glufosinate resistant callus as described by Armstrong 19, and grown in five-gallon pots containing sterile sand and Metromix-350 (1:1). Plants were fertilized weekly with Peter's 20-20-20 with micronutrients (Division of United Industry Corp., St. Louis, MO).

One gram of frozen young leaf tissue was ground in liquid nitrogen and genomic DNA was extracted using the Dellaporta procedure 24. Twenty micrograms of genomic DNA were digested with EcoRI, which released a 1.9 Kb fragment containing the vspB gene, the nos terminator and part of the Ubi-1 promoter. DNA was separated on a 0.8% agarose gel, blotted onto Hybond N⁺ membrane (Amersham Pharmacia Biotech, Inc. Piscataway, NJ) by capillary blotting 25, and UV cross-linked. The non-radioactive digoxigenin system (Roche Molecular Biochemicals, Indianapolis, IN) was used for labeling and detection of the transgene. Blotted DNA was probed with either a 611 or a 843 bp of vspB gene segment amplified from pRSVP-1 and gel purified. The forward and reverse primers 5'-GTTCTTCGGAG GTAAAAT-3' and 5'-TTCGCCTCTGTGGT-3' were used, respectively, to amplify a 611 bp segment, and the primer pair 5'-GCAGGCTACCAAAGGT-3' and 5'-TAGGTGACTTACCCACAT-3' was used to amplified the product of 843 bp.

For identification of bar transgenic plants, the DNA was digested with EcoRI, which released a fragment of ~1.5 Kb that contained part of the Ubi-1 promoter, the bar gene, and the nos terminator, and was identified with a 419 bp digoxigenin labeled probe produced by PCR amplification of pAHC25 using the forward and reverse primers: 5'-GGCGGTCTGCACCATCGT-3' and 5'-GCCAAGTTCCCGTGCTTGA-3', respectively.

Total RNA was isolated from 2 g of tissue using acid guanidine isothiocyanate-phenol-chlorophorm extraction 26, resuspended in T₁₀E₁ and treated with 2 μl RNasin^® (4U/μl) RNAse inhibitor (Promega, Madison, WI) and stored at -70°C until use. Thirty micrograms of total RNA were separated on 1.2% agarose formaldehyde gels and transferred to Hybond N⁺ membrane by capillary blotting overnight or by pressure blotting for one hour using a PosiBlot^® 30-30 pressure blotter (Stratagene, La Jolla, CA) according to the manufacture's instructions. Membranes were UV-crosslinked, and probed with the same probes as described for Southern blot analysis. Chemilluminescence was captured using a Kodak Digital Science Image Station 440 CF (Eastman Kodak Company, Rochester, NY).

Total RNA extractions for real-time RT-PCR were performed using 500 mg of tissue ground to fine powder using mortar and pestle in the presence of liquid nitrogen, then processed with the RNeasy midiprep Kit (Qiagen, Germany), following the manufacturer's protocol. Trace DNA contamination was removed from total RNA by a combination of acid phenol: chloroform 5:1 pH= 4.7 extraction and Dnase I treatment (Ambion, Texas). Real-time RT-PCR was performed on a Rotor-Gene RG-3000 (Corbett Research, Australia) using the Quantitect SYBR Green real-time RT-PCR kit (Qiagen, Germany), and the manufacturers protocols with 300 ng of Dnase I treated total RNA. Primers were designed to amplify a 108 bp fragment of the soybean vspB using the following primers: 5'-TGGTTCAACGCACTCTTC-3' and 5'-GGCTATGGTGAGCGTTCTTC-3'. Reverse transcription was performed for 30 min at 50°C followed by a 15 min denaturing at 95°C, and 40 cycles of 40 s at 95°C, 40 s at 58°C and 40 s at 72°C. Quantification was based on relative abundance to maize 18S RNA by amplifying a 174 bp fragment with primers: 5'-CCTGCGGCTTAATTGACTC-3' and 5'-GTTAGCAGGCTGAGGTCTCG-3', and using the comparative quantification function of the Rotor-Gene RG-3000 software. All real-time RT-PCR experiments were conducted in triplicate and on triplicate RNA preparations for each sample. Melting curve analysis and agarose gel electrophoresis were performed to verify single product formation.

Protein was extracted from 100 mg of leaves and stems with 0.5 ml of phosphate-buffered saline (137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, 2 mM KH₂PO₄) supplemented with 1 tablet/10 ml buffer of the Complete-Mini protease inhibitor cocktail (Roche Molecular Biochemicals, Indianapolis, IN) by homogenizing in the presence of zirconia/silica 0.1 mm dia. beads. Centrifuged extract supernatants were removed and used for protein concentration determination 27. Because of the low protein yields in stem extracts, they were precipitated with 10% trichloroacetic acid (TCA), washed with ice-cold acetone and resuspended prior to SDS-PAGE analysis.

Thirty micrograms of protein were separated on 12% SDS-polyacrylamide gels. Proteins in the gel were either stained with silver nitrate 28 or transferred to Hybond-P (PVDF) membranes (Amersham Pharmacia Biotech, Inc. Piscataway, NJ) using Trans-Blot SD Semi-Dry Transfer Cell blotter and recommended protocols (Bio-Rad Laboratories). Soybean VSP was immunologically detected on the PDF membranes with anti VSPβ serum (provided by P. Staswick, University of Nebraska, produced as previously described 29 used at a 1:5,000 dilution. Detection was performed using a luminol substrate and the NEN Life Science Products, Inc. (Boston, MA) Reinascence kit. Chemiluminescent signal was captured by a Kodak Digital Science Image Station 440 Cf, and analyzed with Kodak 1D Scientific Image Software. Quantification of band intensity was always compared relative to samples from the same gel.