T-DNA integration into the barley genome from single and double cassette vectors

Abstract

Patterns and sites of T-DNA integrations into the barley genome from single and double cassette vectors are of interest for the identification of cultivars with value added properties as well as for the production of selection marker-free transgenic lines that can be retransformed. T-DNA/Plant DNA junctions were obtained by capturing a single-stranded DNA with a biotinylated primer annealing to the vector adjacent to the border and an adaptor ligated to a restriction site overhang in the flanking barley DNA. The captured junction was converted into a double strand and sequenced. Fifty left and right border junctions from plants transgenic for one of five human genes were analyzed. Primers of 15–30 nucleotides designed from the genomic DNA at the insertion site can PCR amplify fragments that identify unequivocally any transformant. Adjacent transgene insertions with single cassette vectors were always in tandem direct repeat configuration. With regard to T-DNA integration the patterns were comparable to the variations found in dicotyledonous plants. Twelve of the 46 integrations characterized by blast searches were within different regions of the BARE-1 retrotransposon element occurring with a frequency of 2 × 10⁵ copies in the barley genome. The use of border junctions to identify number of copies and loci of integrates in transformants is discussed.

Stable genetic transformation of barley is routinely carried out by co-cultivation of immature zygotic embryos with Agrobacterium tumefaciens carrying binary vectors (1, 2). The target gene and the selection marker are located between the left and right 25-bp direct repeats of the T-DNA borders either within a single or in two adjacent tandemly arranged cassettes (3). The transfer of the T-DNA is polar; an intact right border is required for the transfer, whereas mutations in the left border have little effect on transfer efficiency (4). A single-stranded copy of the lower T-DNA strand is generated from endonucleolytic cleavage sites between the third and fourth bases of the border repeats. The strand is covered with the single-stranded DNA binding protein VirE2, covalently bound to the VirD2 protein at the 5′ end and exported from the bacterium (4, 5). The VirE1 protein assists in secretion of virE2, which appears to form a voltage gated channel through the plasma membrane of the plant cell (6). The DNA is transported through the channel and guided by several proteins into the cell nucleus. The single-strand T-DNA then probably invades the DNA of the plant chromosome and is integrated by illegitimate recombination followed by second-strand synthesis. Alternatively a double-stranded form might be generated first and then integrated into the chromosome (5).

Details of integration have been primarily investigated in tobacco and Arabidopsis (7–10). Small deletions, base substitutions, duplicated border, and genomic sequences are found around the T-DNA/plant DNA junctions. Co-integration into a chromosome locus after extrachromosomal second strand synthesis and ligation of the different T-DNA molecules, followed by docking and insertion via double-stranded breaks and repair synthesis is supported by analysis of co-integration of two or more T-DNAs in the same locus (11). All ten possible combinations—i.e., inverted as well as tandem arrangements—occurred with regard to the left and right borders. The junction regions of direct repeats of T-DNA were investigated by selecting these with a vector containing a promoterless kanamycin phosphotransferase gene near the right border and a 35S CaMV promoter near the left border (12). Transformants from two T-DNA strands linking the promoter with the gene were selected on media containing kanamycin and revealed integrated T-DNA strands with a precise junction as well as integrations with 8 to 293 bp of filler DNA. Binary vectors carrying a neomycin phosphotransferase gene driven by a nopalinsynthase promoter within the T-DNA and a β-glucuronidase (gus) reporter gene under the control of a mannopine synthase promoter outside the left or right border resulted in 75% of transgenic tobacco plants containing the uidA (gus) gene unlinked or linked to the left or right border of the integrated T-DNA (13). Such incorporation of vector backbone sequences can be avoided by cotransformation of the target gene with the virD1, virD2, and virE2 virulence genes within the left and right border cassette (14, 15). Marker-free transformants have been produced in tobacco and rice by vectors containing the β-glucuronidase reporter gene in one cassette between a left and right border and the neomycin phosphotransferase gene for selection in a separate cassette (16). Cotransformation with the two T-DNA cassettes in one vector was around 50% and more efficient than transformation with the cassettes in separate Agrobacterium strains. With high frequency the two cassettes were integrated in unlinked fashion and segregated in Mendelian ratios in the T₁ generation.

In the use of transgenic barley plants for basic studies and production of cultivars with value added properties, the exact identification of a given genotype by the transgenic insertions is mandatory. Of equal interest is the production of marker-free transgenic lines that permit retransformation with the same marker and the use of cultivars without herbicide or antibiotic resistance. Toward this end we have produced transgenic barley plants with single or double cassette vectors containing genes for the production of human antithrombin III, α₁-antitrypsin, lysozyme, serum albumin, and lactoferrin. In the present communication junction sequences of the T-DNA borders and the genomic DNA in the transgenic plants are reported and frequencies of linked and unlinked copies of the transgenes evaluated.

Materials and Methods

Transgenic Plants.

Transgenic plants were produced as described (2, 3). The following single cassette lines were investigated with DNA isolated from leaf material of the T₀ generation. The selection marker consists of the bar gene encoding phosphinotricine transacetylase flanked by the ubiquitin promoter and the nos terminator (17). Transcription of the transgene is terminated either by the rice α-amylase 3D terminator (R3D) or by the nos terminator (nos).

Plants expressing human antithrombin III were made with plasmid pH281 (p alpha-amy+sp/ATIII-GC/R3D): It contains between the left and right border the 680-bp barley alpha-amy promoter and signal peptide coding region of the Amy 6-4 gene (18) encoding a high pI α-amylase (p alpha-amy+sp) (17). This promoter drives the human antithrombin III gene, codon optimized to a G+C content of 64% based on amino acid sequence of European Molecular Biology Laboratory (EMBL) accession no. X68793 (ATIII-GC).

Plants expressing human α₁-antitrypsin, serum albumin, or lysozyme were made with plasmids pR099 (p Apex II+sp/AAT-GC/nos), pR097 (p ApexII+sp/HSA-GC/nos), and pR098 (p ApexII+sp/Lys-GC/nos), respectively. The transgenes were driven by the 679-bp barley alpha-amy promoter and signal peptide coding region of the Amy32b gene (19) encoding a high pI α-amylase (p Apex II+sp). Codon optimization of the ORFs to G+C contents of 63%, 66%, and 68% was based on EMBL accession nos. X01683.1 (AAT-GC), V00495.1 (HSA-GC), and X57103.1 (Lys-GC).

The following double cassette lines were investigated with DNA isolated from leaf material of the T₁ generation. Transcription is terminated in both cassettes with the nos terminator (nos). The second cassette contains the selection marker bar encoding phosphinotricine transacetylase flanked by the ubiquitin promoter and the nos terminator (17).

Plants expressing human antithrombin III or human lactoferrin were made with plasmids pR119 (p hor3+sp/AT3-GC/nos) and pR121 (p hor3+sp/lactoferrin–GC/nos). In cassette 1 the target genes were under the control of the 500-bp barley hor3 promoter and signal peptide coding region of the Hor3D gene (20) encoding the high molecular weight hordein storage protein (p hor3+sp). Codon optimization of the genes to a G+C content of 64% and 69% was based on amino acid sequences of EMBL accession nos. X68793 (AT3-GC) and HSO7643 (lactoferrin-GC).

Determination of T-DNA/Plant DNA Junctions at Left Border.

The genomic DNA is cut with the restriction enzyme Sau3AI in the vector sequence and in the barley genome somewhere upstream of the left border (Fig. 1, 1; for detailed protocol see Protocol for Identification of Flanking Sequences, which is published as supporting information on the PNAS web site, www.pnas.org). The two strands are separated by adding a 6× SSC solution to the digest and heating to 95°C for 5 min. After cooling to 68°C the 5′-biotinylated oligonucleotide R193 (B-5′-ACTTGATTTGGGTGATGGTTCACG TAGTGG-3′) is hybridized for at least 3 h to the upper strand (Fig. 1, 2). The strand is incubated with streptavidin-coated magnetic beads for 30 min at room temperature (Dynabeads, Dynal ASA, Oslo) and isolated in the magnetic stand with several subsequent washes according to the manufacturer's protocol. At this point the base paired R172/R187 adapter is ligated with T4-DNA ligase to the 5′ end of the isolated strand (Fig. 1, 3). The adapter has the following sequence:

Figure 1.

Protocol for isolation of genomic barley DNA flanking T-DNA integrations.

Figure 2.

Barley DNA/T-DNA border sequences found in transformants made with (a) single cassette and (b) double cassette vectors. The colors in the nucleotide sequences correspond to the colors in the maps.

Figure 3.

Unique identification of transgenic plants by PCR. DNA in lanes 2 is from transformant H 281-1-3, in lanes 3 from AAT II.13-53, and in lanes 4 from H 281-16-65. Amplifications: (a) with primers R241/R164 specific for H 281-1-3; (b) with primers R315/R164 specific for AAT II 13-53; (c) with primers R313/R164 specific for H281-16-65; (d) with primers R95/R96 identifying the bar gene in all three transformants. Lanes 1, λDNA (EcoRI/HindIII); lanes 5, Golden Promise DNA; lanes 6, H₂O.

R172 5′-GGGAAGGAAAAAAGGGGGAGGAGGGGAAAGGGAGA-3′ R187      3′-CCTTTTTTCCCCCTCCTCCCCTTTCCCTCTCTAG-5′

Primer R176 (5′-CGCCTGCTGGGGCAAACCAGCGTGG-3′) is annealed to the homologous vector sequence of the single strand and used to generate the double-stranded molecule (Fig. 1, 4). Amplification of the double strand is carried out with reverse primer R176 and the forward primer R205 (5′-GGGAAGGAAAAAAGGGGGAGGAGGGGAAAGGGAGAGATC-3′) covering the adapter sequence (Fig. 1, 5). A second PCR is performed with the nested primer R179 (P-5′-CTCTCAGGGCCAGGCGGTGAAGGGC-3′) (Fig. 1, 6). The obtained fragment (Fig. 1, 7) is sequenced and provides the barley genome nucleotide sequence joining the left border and adjacent vector sequence.

Figure 4.

(a) Possible arrangements of adjacent T-DNA inserts. (b) A left- and a right-border-specific primer were used to identify adjacent T-DNA inserts. Only head-to-tail configurations were found in eight transformants (lanes 2–9). Lanes 1, pUC18/HaeIII; lanes 10, Golden Promise; lanes 11, H₂O. (c) Nucleotide sequences of the eight transformants with head-to-tail tandem repeats amplified with the left- and right-border-specific primers (cf. lanes 2–9 in b). Nucleotide color: blue, left border repeat; black, border sequences; red, barley DNA; orange, vector sequence. (d) Proof for the absence of adjacent inverted integrations by Southern blots. EcoRI (Left) and EcoRI/HindIII (Center) digested plant DNA of seven transformants was hybridized with a 181-bp probe that recognizes the 994-bp EcoRI/HindIII fragment covering the borders of a head-to-tail fusion. All six PCR head-to-tail tandem integration-positive plants contained the expected ≈994-bp fragment (Center, lanes 5–10). No hybridization occurred to a 1,320-bp EcoRI/EcoRI fragment expected from a head-to-head configuration (Left). Likewise, a right border probe did not identify a 668-bp HindIII/HindIII fragment expected for a tail-to-tail border fusion (Right). In lanes 2 are test fragments that hybridized to the respective probes. Lanes 1, pUC18/HaeIII.

An improved anchoring of the single strand to the streptavidin-coated magnetic beads is achieved by ligation of a 3′biotinylated and 5′-phosphorylated oligonucleotide R360 to the 3′ end of the desired single strand. This ligation requires the annealing of the additional bridge-oligonucleotide R361 pairing at one end to the single strand and on the other to R360 (Fig. 1, 8):

        5′-GAAAGGGGAGGAGGGGGAAAAAAGGAAGGG-3′R360      -CCGCTTTCCCCTCCTCCCCCTTTTTTCCTTCCC-5′R361         3′-GGTAAGTCCGACGCGTTGACAACCCTTC

The improved anchoring permits washing of the captured strand with 0.1 M NaOH for complete removal of contaminating strands.

Determination of T-DNA/Plant DNA Junctions at Right Border.

They are performed in an analogous way to those of the left borders except that the DNA is initially cut with the TaqI restriction enzyme and the following oligonucleotides are used: 5′-biotinylated oligonucleotide R198 B, 5′-TTACCCAACTTAATCGCCTTGCAGCACATC-3′ and

Adapter:

R172 5′-GGGAAGGAAAAAAGGGGGAGGAGGGGAAAGGGAGA-3′ R202      3′-CCTTTTTTCCCCCTCCTCCCCTTTCCCTCTGCT-5′

Amplification primers:

Forward R206: P-5′-GGGAAGGAAAAAAGGGGGAGGAGGGGAAAGGGAGACGA-3′ Reverse R200: 5′-GCCACCGCGGTGGAGCTCCACAACC-3′ Nested primer R201: P-5′-GAGTGGCTCCTTCAACGTTGCGGTTCTG-3′

Results and Discussion

T-DNA/Plant DNA Junctions at the Borders.

Generally three nucleotides of the right border repeat are expected to be present at the junction to the plant DNA. Among the eleven analyzed right border junctions only three (R226, H001, and J001) contained the three nucleotides (TGA) (Fig. 2). Three of the inserts lack the last nucleotide (T) of the border adjacent to the repeat (R233, R238, R237). The five other junctions miss between 3 and 30 of the border nucleotides. Among the sequenced 39 left border junctions, 34 were located within the terminal repeat (blue sequence in Fig. 2). In the residual 5, the genomic sequence was joined within the border and 4 to 95 nucleotides were missing (junction of red to black nucleotides in Fig. 2). The patterns of integration found in barley are comparable to the patterns observed in dicotyledonous plants. The sequences provide an unequivocal identification of the 50 transgenic lines containing the 5 different human genes. This has been verified for 49 of 50 genomic sequences (not all data presented) by designing oligonucleotides specific for each of the flanking genomic sequences and employing them in PCR amplifications with the matching border primer. An example is shown in Fig. 3. The oligonucleotide R241 (5′-ATCTGGTGTAAACAA-3′) derived from the barley genomic sequence R241 of antithrombin III transformant H281-1-3 at the left border of the T-DNA insert (cf. Fig. 2) amplified a 336-bp fragment (Fig. 3a, lane 2) with oligonucleotide R164 (5′-TTTTTCGCCCTTTGACGTTGGAGTCCACG-3′) located 216 nucleotides downstream from the left border in the T-DNA cassette. Primers R164 and R241 did not PCR amplify any DNA fragment when used with DNA from the α-antitrypsin transformant AAT-II.13–53 and antithrombin III transformant H281–16-65 (Fig. 3a, lanes 3 and 4). No amplification was obtained with the DNA of the barley host Golden Promise (lane 5). A specific PCR product of 500 bp was obtained with primer R164 and the oligonucleotide R315 (5′-ATCCAATCTCTCGGATAAATCGCACCCGGG-3′) derived from the genomic sequence joining the left border in the α-antitrypsin transformant (Fig. 3b, lane 3). R164 primer likewise amplified a unique 450-bp fragment, when used with oligonucleotide R313 (5′-CCGCGCCGCCTCACTAGACTCCGCGCATGG-3′) corresponding to the genomic sequence joining the T-DNA in the transformant H281-16-65 (Fig. 3c, lane 4). In DNA from all three transformants the two bar gene-specific primers R95 and R96 amplify the expected 342-bp fragment (Fig. 3d, lanes 2–4). (R95, 5′-CATCGAGACAAGCACGGTCAACTTC-3′; R-96, 5′-ACCGAGCGCCTCGTGCATGCG-3′; positions 84 and 426 of the bar gene, respectively).

Orientation of Adjacent Transgene Insertions with Single Cassette Vectors.

Two adjacent integrations of cassettes with a left and a right border can be in tandem (Fig. 4a, Center) or inverted with the two left borders in contact (Fig. 4a, Left) or the two right borders in contact (Fig. 4a, Right). Primer R164 adjacent to the left border and primer R208 (5′-GATCAGATTGTCGTTTCCCGCCTTCAG-3′) spanning nucleotides 255 to 281 in the right border would amplify a fragment of 401 bp, if the full-length tandem left and right border repeat were integrated (Fig. 4a). These two primers were used to test transgenic plants for tandem border repeats by PCR. Among the 60 plants analyzed, 33 (i.e., about 50%) revealed tandem inserts, whereas no “head-to-head” or “tail-to-tail” configurations were encountered, as exemplified in Fig. 4b. The amplified tandem repeats of eight transformed plants ranged from 375 to 438 bp. The reason for the variability is apparent from the nucleotide sequences of 12 such tandem borders (Fig. 4c) from transformants with antithrombin III, α-antitrypsin, and human serum albumin genes. The space between the left and right borders includes variable numbers of nucleotides from the left border repeats (blue), from the border sequence (black), from the genomic barley DNA (red), or from the vector (orange). Remarkable is the high frequency of the TCA nucleotides of the right border repeat (green), which contrasts to the low frequency of these nucleotides in the non-tandem integrated T-DNA copies (Fig. 2). In transformant HSA-III a perfect left border has been recreated because the blue G nucleotide (arrow) is the VirD2 cleavage site. This site is also present in the 281-I.16 transformant, but flanked by some filler DNA.

Definitive proof that adjacent inverted integrations are absent in barley requires Southern hybridization analysis, because single strands derived from such integrations might form panhandle (stem-loop) structures that cannot be PCR amplified. The EcoRI/HindIII fragment of a head-to-tail fusion of the T-DNA right and left border is 993 bp. A test fragment of 958 bp that hybridizes to left border probes was constructed in the following way: Oligonucleotides R370 (5′-GATGATATTTGATCACAGGCAGCAACGC-3′) and R369 (5′-CACTGGCCGTCGTTTTACAACGTCGTGAC-3′) amplified a 907-bp fragment from the left border. The fragment was cloned into the SmaI digested pUC 18 vector and cut with HindIII and EcoRI to yield a 958-bp fragment. The fragment was identified by hybridization with a 181-bp probe specific for the left border generated with the two primers R163 (5′-CCCTTCACCGCCTGGCCCTGAGAGAG-3′) and R164 (5′-TTTTTCGCCCTTTGACGTTGGAGTCCACG-3′) (Fig. 4d, Center, lane 2). PCR-positive plants with head-to-tail tandem integrations were tested in a Southern blot with this probe (lanes 5–10). They all contain the expected ≈993-bp EcoRI/HindIII fragment. As a result of multiple integrations additional larger fragments hybridize in some of the transformants. In an analogous way, a marker fragment (1296 bp) mimicking the 1,330-bp EcoRI/EcoRI fragment of the intact head-to-head fusion of two left borders was synthesized with the primers R371 (5′-TTAACCTACCTTCCTTTGGTTCCGGGGG-3′) and R369, and hybridized with the left border probe (Fig. 4d, Left, lane 2). None of the tested transformants had an EcoRI/EcoRI fragment of the expected size. The HindIII/HindIII fragment of a tail-to-tail fusion of two right borders is 656 bp. A mimicking fragment of 620 bp was assembled with oligonucleotides R367 (5′-GGCACTGGCCGTCGTTTTACAACGTCG-3′) and R368 (5′-GAACGACGTCACCGCCCACTATGGCATTC-3′). None of the transgenic plants analyzed revealed a Southern hybridization indicating a tail-to-tail configuration, when probed with the 168-bp right-border-specific probe (Fig. 4d, Right). The probe was generated with primers R199 (5′-CCAGGGGGGATCCACTAGTTCTAGAGC-3′) and H16 (5′-CATGAGCGGAGAATTAAGGG AGTCACG-3′). In conclusion, the analysis indicates that adjacent inverted integrations of two copies of a given insert are very rare or absent in barley. If tandem formation occurs before integration but after conversion of the single-stranded T-DNA into a double stranded form, its lack of a 5′–3′ polarity should result equally frequent in inverted and direct repeat configurations as observed in Arabidopsis and tobacco (11). Ligation of single-stranded T-DNAs by the catalytic activity of the covalently attached VirD2 protein (21) before second-strand synthesis and integration could explain the preferred head-to-tail transgene integration.

Integration Sites in the Barley Chromosome.

BLAST searches were performed on the 46 barley nucleotide sequences adjacent to the left or right border integration. Thirty-four genomic sequences provided high identity only to motifs of 15 to 30 nucleotides and thus did not identify specific gene regions. Twelve of the sequences were inserted in actively transcribed BARE copia-like retrotransposons (22). There are ≈1.96 × 10⁵ copies of BARE-1 in barley (23, 24) and it is therefore not surprising that we found twelve integrations in this highly repetitive element. As can be seen in Table 1, which is published as supporting information on the PNAS web site, it was possible to use this sequence for precise determination of the location of the inserts within the element.

Number of Independent and Linked Loci.

DNA of leaves from plants transformed with single cassette vectors were collected and subsequently analyzed for the presence of the transgene by Southern blotting with a probe specific for the left border as well as by PCR with primers for the marker bar and the oligonucleotides derived from barley sequences flanking the transgenic inserts.

Transformant 281-I.01, encoding human antithrombin III, contained four insertions as judged from Southern analysis (data not shown), and three different isolated left-border-specific sequences (R239, R316, and R317). The sequences of the primers used for amplification of integration sites in the segregation analysis of the transformants are given in Table 2, which is published as supporting information on the PNAS web site. In the T₁ generation (24 plants), primers R239, R316, and R317 amplified all three integration sites in nine plants, the R316 and R317 sites in seven plants, the R239 sites in seven plants, and none of the integrations in one plant. The segregation is 9:3:3:1 (χ² = 4.445, 0.2 < P < 0.3), indicating two independent loci, one of them hosting at least two copies of the single cassette. All 24 plants were positive for bar, indicating a third independent segregating locus harboring a single cassette.

Two transformants contained three integrations. Transformant AAT-II.03 with the human α₁ antitrypsin gene yielded two left border flanking regions. Among 24 T₁ progenies, primers R217 and R218 identified 13 plants with both integrates, indicating a single locus with at least two copies of the single cassette. Because 21 plants tested positive for bar, transformant AAT-II.03 contains at least two independent loci. From transformant HSA-II.01 with the human serum albumin gene, two left border (R277 and R216) and three right border flanking regions (R235, R236, and R250) were isolated. Among 94 T₁ plants, primer R236 amplified the integration site from 69 plants (χ² for 3:1 = 0.1276; 0.7 < P < 0.8). The integration site recognized by R250 was present in 71 plants (χ² for 3:1 = 0.0141; 0.9 < P < 0.95), and 46 plants contained both insertions. Surprisingly no seedlings lacking the two insertions were found, whereas about six would have been expected with unlinked insertion (χ² for 9:3:3:1 = 11.5; P ≈ 0.01*). When monitored with the right-border-specific primer R235, none of the T₁ progeny contained this flanking region. The left border flanking sequences identifying insertions with R216 and R277 were tested on 15 of each of the three types of T₁ seedlings. R216 was present in the seedlings containing the R236 insertion and in the seedlings with both the R236 and R250 insertions, but was absent in the seedlings containing only the R250 insertion, whereas R277 was absent in all plants. The result suggests that the left border of the R216 insert belongs to the cassette identified with the right border R236 and the left border of the R277 insert belongs to the excluded cassette identified with the right border R235 insert. The unexpected plants in the progeny of this transformant HSA-II.01 that have excluded the R235- and R277-specific flanking regions from the germline indicate a chimeric T₀ plant. Analysis of mutations induced in mature embryos of barley caryopses have shown that the embryo contains six separate shoot meristems and that the spikes of the primary tillers are already present in the embryo as primordia of several cells (25, 26). This situation could lead to chimeric spikes containing grains with the mutant allele or only two wild-type alleles.

Two transformants contained two integrations. Monitoring 24 T₁ progeny plants from transformant H281-I.16 with primer R313 and for bar yielded 19 positive plants. The bar gene was found in two additional plants, indicating that the two loci are in close proximity. Twenty-two T₁ progenies from transformant H281-I.02 were analyzed with primer R292 and for bar. Thirteen plants contained the insertion R292, whereas all plants contained bar, indicating two unlinked loci of the integrated cassette.

T₁ progenies of six transformants with two integrations observed in Southern analysis were found by segregation analysis to belong to a single locus harboring more than one copy of the integrated single cassette. Such multicopy integration events into a single locus seem to happen frequently because more than 50% of all transformants revealed head-to-tail integrations of the inserted single cassette. Sequence information obtained from 13 of these integrations revealed short sequences of probably genomic origin between the left and right borders of the cassettes. In contrast to the analyzed right T-DNA/plant DNA junction representing outermost T-DNA borders, the internal right T-DNA borders in such duplexes contained the three nucleotides TGA in most cases (10 of 13), indicating a different way of integration of these T-DNA strands into the barley chromosome.

T₁ progenies of 14 T₀ plants with single integrations were tested for bar. Ten of these T₁ progenies segregated in 3:1 ratios with P values in χ² tests ranging from 1 to 0.2. The pooled data of these progenies gave a segregation of 161 bar⁺: 53 bar⁻ (χ² = 0.0063; 0.90 < P < 0.95). The other four T₁ progenies displayed a segregation pattern with an unexpectedly high number of plants devoid of the inserted single cassette. Further analysis of these transformants is required to obtain information on the copy number, sites of integration, and possible recombination events.

Conclusions

Isolation and nucleotide sequencing of junction fragments between known DNA (the transgene) and unknown DNA (barley genomic DNA) has revealed that T-DNA transgene integration into barley chromosomes generally follows patterns known for dicotyledonous plants. In contrast to observations with dicots, only tandem direct repeat integrations were observed in transformants with adjacent integrations in the same locus. Primers with a length of 15 to 30 nucleotides derived from the genomic sequence flanking the insert can PCR amplify fragments that identify unequivocally a given transformant and supply information on number of transgene loci and copies, providing the information required for deregulation of transgenic crop plants with value added properties. Vectors with separate cassettes for the transgene and for the herbicide or antibiotic resistance selectable gene integrate frequently in different locations allowing through Mendelian segregation isolation of barley plants that only contain the value adding transgene.