Our research is concentrated on the development and implementation of new transformation technologies in barley and wheat. These new technologies are subsequently used in other projects in our research group.
Existing transformation systems for barley and wheat require selection markers, can only be used in very few genotypes and it is yet not possible to target transgenes to specific sequences in the genome. Different approaches to overcome these limitations are pursued.
We have developed a transformation technique for the zygote in in vitro cultured barley ovules which is largely independent of genotype and allows for transformation without selection, Figure 1. We are currently working on the development of a more efficient ovule transformation system and on the transfer of the technology to wheat.
Figure 1. Agrobacterium mediated transformation of the zygote or small embryos in the ovule with the gene for the green fluorescence protein (GFP)
a,b,d,e,f: UV-light microscopy. c: Confocal microscopy
a. Agrobacterium infection of the ovules using a syringe with a fine needle. b. GFP-expression in the micropylar area of an ovule two days after isolation and infection. Bar 0.35 mm. c. GFP-expressing embryo developed in the ovule tip 7 days after isolation and infection. Bar 0.18 mm. d. GFP-expressing embryo developed within the ovule tip 18 days after isolation and infection. Bar 0.5 mm. e. GFP-expressing embryo excised from the ovule 18 days after isolation and infection. Bar 0.35 mm. f. Regenerating GFP expressing plantlet. Bar 1.5 mm.
Zinc Finger Nucleases (ZNFs) and Transcription Activator-like Effector Nucleases (TALENs) are new tools for gene targeting that have only been implemented in a few plant species. In collaboration with Daniel Voytas group at the Centre for Genome Engineering at University of Minnesota (a gene targeting expert group), we are currently working on the development of vectors and transformation procedures to implement these tools in barley.
Both nucleases can be assembled to create double strand DNA breaks at specific DNA sequences in the genome (Figure 2 and 3). Double strand breaks (DSBs) induce the cell’s own DNA repair machinery. The two primary repair mechanisms are non-homologous end joining (NHEJ) and homologous recombination (HR). NHEJ can induce mutations (primarily deletions of 1 to 20 bp) and this mechanism can be used to abolish the function of a gene. HR uses a homologous DNA sequence to repair the DSB. When a piece of DNA homologous to the sequence with the DSB is introduced into the cell, the cell can use this as repair template. This mechanism can be used for the selective replacement of sequences in a gene. Furthermore, non-homologous DNA sequences introduced into the cell can integrate into the DSB.
The DNA sequences of ZNFs or TALENs designed to target specific sequences are inserted into transformation vectors. Cells are transformed with the vectors and the proteins are synthesized in the transformed cells. The proteins will subsequently target the DNA sequences for which they are designed. When HR or integration of non-homologous DNA sequences into the DSB is desired, the cells are co-transformed with the corresponding DNA sequences.
Figure 2. Zinc Finger Nucleases
A Zinc Finger Array is a modular assembly of three ‘fingers’ of DNA each of which code a protein that binds a specific three base pair sequence on target DNA. The finger array is fused to a FokI nuclease. When used in conjunction with a second Zinc Finger/FokI nuclease designed for the opposite strand, the zinc finger nucleases can be used to target specific sequences in the plant genome. The FokI nuclease cleaves as a dimer and when the two FokI domains come together over the spacer sequence, a DSB is created. (Picture from Urnov et al., 2005, Nature 435: 646-651)
Figure 3. Transcription Activator-like Effector Nucleases
TALENs are DNA binding proteins produced by plant pathogenic Xanthomonas bacteria. Their DNA sequence specificity is determined by a central domain of tandem 33-35 amino acid repeats. In each repeat the amino acids in positions 12 and 13 specify a target base (a). Most TAL effectors have 12 to 27 repeats. TAL effectors can therefore be designed to bind to specific DNA sequences (b). The TAL effectors are fused to FokI nuclease to create DSB as described for Zinc Finger Nucleases. (Picture from Cermak et al., 2011, NAR 39: 1-11)
The new technologies will be developed to meet the requirements of the cisgenesis concept. Cisgenesis is a new transformation concept, which was developed with the aim of meeting the scepticism among the general public toward genetic modification (Schouten et al., 2006, EMBO Rep. 7, 750-753). The cisgenesis concept imply that the plants are transformed only with their own genetic materials or genetic material from closely related species capable of sexual hybridization. Furthermore, foreign sequences such as selection marker genes and vector-backbone sequences should be absent or eliminated from the primary transformants or its progeny (Figure 4). The gene pool exploited by cisgenesis is accordingly identical to the gene pool available for traditional breeding. Several surveys on the public opinion of cisgenesis in DK and EU have shown that cisgenic crops are acceptable by a larger percentage of the public than transgenic crops. Presently, we have successfully developed and implemented the cisgenesis concept in barley using a barley phytase gene as candidate gene. These plants are currently grown in field trials in Flakkebjerg.
Figure 4. Comparison of vector constructs for transgenesis and cisgenesis.
Vectors for transgenesis includes fragments from many different species. In vectors for cisgenesis the gene of interestis a complete copy of the endogenous gene including the promotor, introns and the terminator in normal sense orientation. The cisgenic vector constructs are designed either without selection marker or with the selection gene and the gene of interest flanked by their own T-DNA borders. In this way, unlinked integrations are possible, allowing for the segregation of the two genes in the next generation and selection of plants without the selection marker. The two left borders in the cisgenic construct should reduce the backbone transfer to the plant genome. The transformed plants have to be analysed for backbone integration and plants with backbone integration discarded.