The Arabidopsis FLS2 gene (At5g46330) including its promoter and terminator sequences was cloned into the pPAM [GenBank: AY027531] based pKWS05 vector (Figure 1B). This construct was transformed into two different Arabidopsis ecotypes, Col-0 and Ler, as well as into Arabidopsis fls2 mutant plants previously generated by EMS mutagenesis (Gomez-Gomez and Boller, 2000). In order to avoid the problems associated with growth of bacteria in liquid cultures, we employed bacteria cultured on plates for direct transformation of Arabidopsis plants. After two days of growth under selection, the bacteria were scraped from the plates and resuspended in 30 ml of YEB medium in Falcon tubes. From one densely grown plate this resulted in an OD₆₀₀ of about 2.0. For dipping plants, the bacterial cell suspensions were poured into 2 liter disposable plastic autoclave bags containing 120 ml of 5% sucrose solution supplemented with 0.03% Silwet L-77. Flowers of healthy Arabidopsis plants were subsequently dipped into the bacterial solution contained in the plastic bags. The bacteria were distributed to all plant parts including very young flower shoots by gently pressing the outer side of the bags by ones hands. After ten seconds the plants were removed from the plastic bags and treated as described below (see Methods). Transformed plants were kept in the greenhouse and seeds were harvested upon full maturation. The seeds were germinated on soil and transgenic plants were selected by spraying with 0.1% BASTA herbicide in the greenhouse (Figure 1C). Spraying was performed one week after germination and repeated four times at two-day intervals. Transgenic plants were readily identified at the end of the BASTA selection. While such plants continued to grow and remained green, the untransformed plants remained small, became white and died two weeks after selection (Figure 1C).

By this modified protocol we obtained more than 100 transgenic plants per transformation by dipping two pots each containing 5–8 Arabidopsis plants. This efficiency is highly satisfactory for most applications. Storage of bacteria on plates at 4°C for up to one week did not noticeably reduce the transformation efficiency since the number of transformants obtained using such bacteria were nearly the same when compared with liquid grown cultures. Thus, the method therefore allowed us to even wait for the best suitable developmental stage of plants prior to transformation. The detailed modified transformation protocol is given in the Methods section.

In order to prove at the molecular level that the plants surviving the BASTA selection are indeed transgenic, we performed DNA gel blot analysis on these plants and compared the results to the wild type originators. The membrane containing the StuI digested genomic DNAs of selected lines was first hybridized with a 3' FLS2 probe (1.5 kb StuI/XbaI fragment of the FLS2 gene, see Figure 1B) and, after stripping, rehybridized with a vector probe (the 1.5 kb HinfI fragment of pKWS05). The results are shown in Figure 1D. As expected, the FLS2 probe only detected a 4.4 kb StuI fragment corresponding to the endogenous FLS2 gene in the wild type plants and the fls2 mutant (EMS mutant). In the transgenic plants, beside the 4.4 kb band additional bands could be observed. To eliminate the possibility that the additional bands might be the due to the presence of contaminating Agrobacterium in these plants, we reprobed the membrane with the HinfI fragment from the vector backbone, which should not be inserted into the plant genome (Figure 1D right panel). If intact Agrobacterium DNA was isolated together with the plant DNA, a 9.8 kb band corresponding to the size of the linearized vector is expected and should be detected. Although we detected contaminating Agrobacterium DNA in Col line 4, and in Ler lines 3 and 8 (Figure 1D), no other bands were visible even after very long exposure times (data not shown). Therefore, the other additional bands observed with the FLS2-probe are derived from stably integrated copies of the FLS2 gene introduced by the T-DNA. This is strong evidence that these plants are true transgenics.

Using this modified transformation method we transformed four other constructs into wild type or T-DNA mutant plants. These constructs contained cDNAs coding for four different WRKY transcription factors carrying HA epitope tags at their C-terminal ends. The constructs were transformed into both wild type Col-0 plants as well as the corresponding homozygous WRKY T-DNA mutant lines. Also in these cases, we obtained more than 100 transformants surviving the BASTA selection. In addition, we could detect the production of the recombinant proteins in about 30% of the T₁ plants by western analysis using an anti-HA antibody (data not shown).

Our data demonstrate that the method is suitable for transforming different Arabidopsis ecotypes, such as Col-0 and Ler as well as mutant plants such as the fls2 mutant. We also successfully applied this method to generate a similar set of FLS2 transgenic lines using another binary vector that allows for selection with the antibiotic kanamycin (data not shown).

- Grow healthy Arabidopsis plants until they are flowering (see Figure 1A).

Optional: Clip first bolts to encourage proliferation of many secondary bolts. Plants will be ready roughly 4–6 days after clipping. Optimal plants have many immature flower clusters and only few fertilized siliques, although a range of plant stages can be successfully transformed.

- Transform the DNA construct of interest into the appropriate Agrobacterium strain. Grow the transformed Agrobacterium on YEB (or LB) plates containing the appropriate antibiotics in a 28°C incubator.

- Select a colony and resuspend bacteria in 10 μl H₂O. Plate half of the volume immediately as a lawn onto a YEB plate with the suitable antibiotics and incubate at 28°C for (2–3 days), and use the other half to verify the presence of your DNA construct by PCR analysis.

- Collect the densely grown bacteria from the plate by scraping, and resuspend them in 30 ml YEB in a sterile Falcon tube. The OD₆₀₀ should be about 2.0.

- Per transformation prepare 120 ml of 5% sucrose solution containing 0.03% of Silwet L-77 (surfactant; Lehle Seeds), pour solution into a disposable plastic bag and add the bacteria.

- Dip the inflorescences of the plants into the Agrobacterium solution for 10 seconds, under gentle agitation. You should observe a film of liquid coating the plants. The bacteria are distributed to all plant parts including very young flower shoots by gently pressing the outside of the bag with your hands.

- Place dipped plants under a lid or cover for 16 to 24 hours to maintain high humidity (plants can be laid on their sides if necessary). Do not expose to excessive sunlight (the temperature under the lid can get high).

- Water and grow the plants as normal, tying up loose bolts with wax paper, tape, stakes, twist-ties, or by other means. Stop watering as seeds become mature.

- Harvest dry seeds.

- Select for transformants using appropriate antibiotics or herbicides.

DNA was isolated from 150 mg plant leaf material using the Nucleon Phytopure Kit of Amersham (catalogue no. RPN8510). 10 μg genomic DNA from each line was digested with StuI and separated on a 0.8% agarose gel. The gel was denatured and neutralised and the DNA was transferred to a nylon membrane and hybridized with the ³²P-labelled 3' FLS2 probe and vector probes. Labelling was performed using the Random Prime DNA Labelling Kit of Roche (catalogue no. 1004760). Hybridization was carried out overnight in a buffer containing 1 M NaCl, 1% SDS, 10% Dextran sulphate and 200 μg/ml salmon sperm DNA. The membrane was washed with 0.2 × SSC, 0.5% SDS at 68°C. For stripping the filter, the membrane was washed in 0.1% SDS solution for 45 min at 94°C.