An efficient method for dispersing Ds elements in the barley genome as a tool for determining gene function

Expression of AcTPase in stably transformed barley plants

The AcTPase gene and a Ds element carrying bar were stably introduced into barley. A total of 44 independent transgenic barley lines carrying either the transposase gene, Ds-bar, or both elements, were generated. Only phenotypically normal plants with a low copy number of the intact introduced gene were used for further experiments. The expression of the stabilized AcTPase gene and the functionality of AcTPase in barley T₁ and T₂ immature embryos of plants carrying either AcAc or UbiAc were monitored using a transient assay for Ac/Ds activity ( McElroy et al., 1997 ). The number of blue spots, an indicator of the level of transposase expression, showed wide variability among plants from different independent transformation events, as well as among plants derived from the same transformation event, indicating that expression varied greatly from plant to plant. Plants containing intact copies of the Ds-bar unit were used for crosses with AcTPase-expressing plants.

Transactivation of Ds by Ac transposase

After crossing AcTPase-expressing plants with Ds-bar-containing plants, DNA hybridization analysis was carried out on the resulting F₁ plants. Only three of a total of 144 F₁ plants (2%) had bands that were not present in the parental plants (data not shown). In these three plants the observed excisions and reinsertions occurred exclusively in somatic tissue, as evidenced by the presence of non-stoichiometric bands in DNA hybridization blots and transmission of the transposed element in a non-Mendelian fashion in F₂ plants.

Selfing of F₁ plants carrying both the Ac transposase gene and the Ds-bar element resulted in a striking increase in transposition frequencies in the F₂ generation ( Figure 1) . Progeny of 24 different F₁ plants, derived from 10 independent crosses with different parents, were analyzed. F₂ plants of 18 of these 24 F₁ plants (75%) showed evidence of transposition ( Table 1).

PCR analysis was used to determine if the plants showing evidence of transposition had empty donor sites. The amplification of a specific 300 bp PCR product, indicating excision of the entire 3.6 kbp Ds-Ubi-bar cassette from the original integration site, as well as a new banding pattern following DNA hybridization analysis, is strong evidence that excision and reinsertion have occurred. The frequency of transposition in F₂ plants derived from independent F₁ parents varied from 0 to 47% ( Table 1). Taking into account the copy number of the Ds element in different F₁ plants, and assuming all copies are equally likely to transpose, the observed transposition frequencies ranged from 0 to 38% per introduced Ds element. Transposition frequencies in sibling F₂ plants, derived from F₁ plants having the same parents were similar in most cases. For example, F₂ plants of A1-1, A1-5 and A1-8 in Table 1 had transposition frequencies of 23, 20 and 16%, respectively. In these cases it is likely that differences in transposition frequencies between crosses were due to the level of AcTPase expression in the original parental AcTPase-containing plant.

No significant correlation was observed between transposition frequency and the particular promoter driving the AcTPase gene in F₂ plants. For example, F₂ plants of line A1-5, carrying AcTPase under control of the maize ubiquitin promoter, exhibited a very similar transposition frequency to F₂ plants of line A18-5 where AcTPase was driven by the Ac promoter sequence. In addition, the copy number of the AcTPase gene under the control of either promoter did not appear to influence transposition frequency with either promoter, although there was a positive correlation between the level of AcTPase activity in F₂ immature embryos, as demonstrated in transient assays, and the number of transpositions in F₂ plants ( Table 1).

F₃ immature embryos derived from selfed F₂ plants that carried both AcTPase gene and Ds-bar continued to express Ac-transposase, and showed evidence of transposition of Ds-bar. The transposition frequencies were similar to those observed in the F₂ generation (compare Tables 1 and 2). DNA hybridization analysis of F₃ plants also revealed that about 25% of all new Ds insertions observed in the F₂ generation were due to transpositions to unlinked sites, as evidenced by the independent segregation of new and old Ds integration sites (data not shown).

Our scheme for the development of a tagging population containing highly dispersed Ds elements proved to be successful. DNA hybridization analysis of randomly chosen F₆ plants showed that each plant has at least one new, Ds-hybridizing band ( Figure 2). Many plants show two or more changes in banding pattern compared to the original Ds integration pattern in the F₁ plant. The average number of Ds elements per plant remains relatively constant.

In order to demonstrate that this system can be used for targeted gene tagging, we crossed plants containing a single transposed Ds element, but no AcTPase gene, with transposase-expressing plants. Resulting F₁ and F₂ plants showed evidence of Ds transposition, indicating that the Ds element could be reactivated in the presence of AcTPase (data not shown).

Timing of Ds transposition

Analyses of DNA hybridization patterns of F₂ and F₃ plants yielded information regarding the timing of transposition. For example, many of the F₂ plants derived from A8-1 that showed evidence of transposition exhibited the same banding pattern for Ds, indicating an excision and reintegration of Ds in somatic tissues early in development of the F₁ plant. The reintegration occurred in a sector of the plant that ultimately went through the germline, resulting in the same Ds integration pattern in many F₂ plants. In most F₂ plants from other independent events, however, the majority of transpositions resulted in unique banding patterns ( Figure 1b). These unique patterns, indicative of independent integration sites, could be due either to germinal transpositions in F₁ plants or to somatic transpositions in F₂ plants. Germinal transpositions in the F₁ generation resulted in F₂ plants that are heterozygous for the new Ds integration site. Selfing of these plants, which also results in a segregation of the AcTPase gene, results in an F₃ generation exhibiting an expected segregation ratio of 3 : 1 for plants carrying either a particular transposed Ds element or plants without a transposed Ds.

Somatic transpositions in the F₂ generation can occur early or late in plant development. Late transpositions affect only small sectors in F₂ plants. In most cases these transposed elements are not transmitted through the germline and therefore were not detectable in the F₃ generation, or were found to segregate the Ds element at an aberrant segregation ratio in F₃ plants. Conversely, transpositions at a very early stage of embryo development are transmitted to the next generation and result in an expected 3 : 1 segregation of the new Ds integration site in the F₃ generation ( Table 2).

Barley mutants generated by transposition of Ac/Ds

Two mutant phenotypes ( Figure 3) were observed when populations of F₃ plants were analyzed. Sixteen of 70 F₃ plants derived from A18-5-116 showed a mutant phenotype ( Figure 3a); this is close to a segregation ratio of 3 : 1, indicating that the tagged mutation might be recessive. Eight of 68 F₃ plants derived from A8-1-3 exhibited a phenotype as shown in Figure 3(b). Despite the skewed ratio in the latter case, DNA hybridization analyses in both cases showed that the mutant phenotype co-segregated with a new Ds band observed in both A18-5-116 and A8-1-3. The isolated and sequenced 3′ and 5′Ds flanking genomic regions from line A18-5-116 have an identity of 82% (P = 1.2e-28) to the maize EST AW231419 (GenBank) ( Figure 4a). No significant homologies were found for the Ds flanking regions from A8-1-3 in public databases. When used in a Northern blot analysis as a probe of 350 bp length, the isolated 5′ flanking region of the transposed Ds element of line A18-5-116 hybridized to an mRNA of about 2.6 kb ( Figure 4b). Hybridization signals were detected in RNA samples from embryos, young seedlings (3 days after germination), shoots of plants at the three-leaf stage, and shoots and roots of plants at the six-leaf stage. No positive signals were detected in RNA samples of leaf tissue of 2- and 4-week-old seedlings even after overexposure of the membrane. The Ds flanking regions isolated from line A8-1-3 were not used as probes in Northern blots.

Construction of vectors

All DNA modifications were carried out according to standard protocols ( Sambrook et al., 1989 ). Plasmid pSP-Ds-Ubi-bar ( Figure 5) contains the Streptomyces hygroscopicus phosphinothricin acetyl transferase gene (bar) and the nos terminator as a 0.9 kbp ClaI–NotI restriction fragment derived from pBARGUS ( Fromm et al., 1990 ). The bar gene is under control of the Ubi1 promoter and first intron from maize, derived from plasmid pAHC27 ( Christensen and Quail, 1996) as a 2.0 kbp PstI fragment. The UbiI-bar-nos cassette is flanked by a 254 bp 5′ sequence and a 340 bp 3′ sequence from Ds, derived from pDs7 ( Wirtz et al., 1997 ) as a 0.59 kbp SalI–BamHI restriction fragment. The construction of the AcTPase containing plasmids pUC-codA-Act-AcAc and pBS-codA-Act-UbiAc, as well as plasmid pSP-WDV-Act1 <Dsbar> GUS.N has been published ( Koprek et al., 1999 ; McElroy et al., 1997 ). Plasmid pAHC20 ( Christensen and Quail, 1996) was co-transformed with plasmids pUC-codA-Act-AcAc or pBS-codA-Act-UbiAc to facilitate selection.

Plant transformation and regeneration

Immature embryos (1.5–2 mm) of barley (Hordeum vulgare L. cv. Golden Promise) were excised, cultured ( Wan and Lemaux, 1994), transformed via particle bombardment either with pSP-Ds-Ubi-bar alone or co-transformed with pAHC20 and pUC-codA-Act-AcAc or pBS-codA-Act-UbiAc ( Lemaux et al., 1996 ), and selected ( Cho et al., 1998 ) as previously described. T₀ plants were regenerated from stably transformed callus lines, and phenotypically normal plants were visually selected and transferred to soil in the greenhouse following the protocol of Cho et al. (1998) .

Development of a tagging population

T₂ plants carrying one or more Ds elements were crossed with plants expressing the AcTPase gene. Testing of AcTPase activity in immature embryos of AcTPase gene-containing plants was performed as described ( McElroy et al., 1997 ). Resulting F₁ plants were selfed and the resulting F₂ plants analyzed for transposition events. Plants containing at least one transposed Ds element and AcTPase were selfed, seeds were bulk harvested, and about 100 seeds randomly chosen, planted and again self-pollinated. This procedure was repeated over four generations (F₃–F₆). The Ds integration patterns in F₆ plants were analyzed by DNA hybridization.

DNA hybridization analysis

Genomic DNA, isolated from two leaves when plants were at the three-leaf stage ( Cone, 1989), was digested with either PstI to determine the intact nature of the Ubi-Ac or Ac-Ac cassettes (6.0 and 4.4 kbp, respectively) or with EcoRV and HindIII to release the intact Ds-Ubi-bar cassette (3.6 kbp). To determine the copy number of Ds and to detect transposition events, DNA was digested with HindIII, which does not cut within the Ds-Ubi-bar cassette but cuts once at the 3′ end of the cassette. Excision and reinsertion of a Ds element into a new location creates a different sized HindIII fragment that can be detected after hybridization with a Ds-specific probe.

Digested DNA was electrophoretically separated on 0.8% agarose gels, transferred to Zeta-Probe membranes (BioRad, Hercules, CA, USA), hybridized with ³²P-labeled probes, and washed according to the manufacturer's instructions. The position on the plasmid of the probe used in the hybridizations is indicated in Figure 5. The probe for DsA was PCR-amplified using plasmid DNA from pSP-Ds-Ubi-bar as a template. The primers were as described below, and the PCR product was purified using the QIAquick PCR purification kit (Qiagen, Chatsworth, CA, USA). Labeling of the probes was performed using the Promega (Madison, WI, USA) Prime-A-Gene labeling kit.

PCR analysis

Genomic DNA (0.5 µg) was subjected to PCR amplification in an MJ Research (Waltham, MA, USA) thermocycler (model PTC-100). PCR reactions (50 µl) contained 1 × PCR buffer (Promega), 200 µm of each dNTP, 1.5 m m MgCl₂, 1 µm primer, 1% DMSO and 2.5 U Taq DNA polymerase (Promega). The primer pairs used for the AcTPase coding region were Ac5′ (5′-AAC CTA TTT GAT GTT GAG GGA TGC-3′) and Ac3′ (5′-ACC ACC AGC ACT GAA CGC AGA CTC-3′) to produce a PCR product of 852 bp. For DsA, they were bar5′ (5′-TGC ACC ATC GTC AAC CAC TA-3′) and Ds3′ (5′-AAC GTC AGT AGG GCT TAA TCT TTT-3′) to produce a 650 bp PCR fragment. For analysis of empty donor sites, primers EDS5′ (5′-CGT CAG GGC GCG TCA GCG GGT GTT-3′) and EDS3′ (5′-AAT ACG CAA ACC GCC TCT CCC CGC-3′) were used to amplify a 300 bp fragment. PCR reactions were performed with an initial denaturation at 94°C for 2 min followed by 10 cycles of a touch-down program with decreasing annealing temperatures from 65 to 60°C in increments of −0.5°C per cycle for 45 sec, an extension at 72°C for 60 sec, a denaturation at 94°C for 45 sec, subsequent 25 amplification cycles for 45 sec at 60°C, 60 sec at 72°C and 45 sec at 94°C, and a final extension at 72°C for 5 min. PCR products were analyzed by gel electrophoresis in 1.1% agarose gels.

RNA hybridization analysis

10 µg of total RNA, isolated from different plant tissues using the RNeasy kit (Qiagen), was separated in denaturing 1.0% formaldehyde-agarose gels, blotted to Zeta-Probe membranes, hybridized with a ³²P-labeled probe and washed according to the manufacturer's instructions. The probe was a 350 bp fragment isolated from the Ds 5′-flanking region of a transposed Ds element that co-segregates with a mutant phenotype.

Isolation of transposon flanking regions by TAIL-PCR

TAIL-PCR was carried out as described previously ( Liu et al., 1995 ) . 100 ng of genomic DNA was used as template DNA. The nested specific primers for the downstream regions were: DsA (5′-GAAACGGTCGGGAAACTAGCTCTA-3′), DsB (5′-TGTATATCCCG TTTCCGTTCCGTT-3′), DsC (5′-ACTAACAAAATCGGTTA-TACG ATA-3′) and for the upstream regions: Ds1 (5′-AAGTTTTGA ATATATGTTTT-CATG-3′), Ds2 (5′-TTAACCCGACCGGATCGTAT CGT-3′), Ds3 (5′-GTTTCCGTC-CCGCAAGTAAATAT-3′). Success fully used arbitrary primers were, for A18-5-116: downstream AD1 (5′-NGTCVAGBATNCWAGC-3′); upstream AD3 (5′-WGTGNAGCA-NCANAGA-3′); and for A8-1-3: downstream AD2 (5′-NGT CGASWGANAWGAA-3′) and upstream AD6 (5′-VTGYCRTWG GBTASTC-3′). Specific tertiary PCR products were analyzed on 1.5% agarose gels, isolated using the QIAquick gel extraction kit (Qiagen), sequenced or used as probes for Northern blots.

Database search

Partial sequences of the 5′ and 3′ flanking regions (300–500 bp) of both events were obtained commercially. These sequences were used in blast searches for similarities in GenBank and TIGR.