Agrobacterium tumefaciens is a pathogenic alpha proteobacterium that causes the crown gall disease when it is induced by hormones secreted by wounds in plants that result in the production of tumors. The infection causes the synthesis of several proteins that manage to insert a fragment of single strand DNA called the T-DNA in the plant genome. This DNA fragment codifies for metabolites that the bacterium needs in order to survive parasitizing the plant. Biotechnology had taken advantage of the insertion system of the T-DNA and had modified the vector that contains this sequence, the T-plasmid, with the objective of using this bacterium to generate transgenic plants.

The system for transformation of plants with A. tumefaciens has the T-plasmid modified in various aspects: Because the T-plasmid is really long and difficult to manipulate, the T-DNA sequence and the genes that cause the crown gall tumor have been deleted. Instead of that, the DNA sequence that you want to insert in the plant is placed by using a “Binary Vector”, a small plasmid designed for the insertion of a sequence of interest flanked by the recognition sites for the cleavage of the T-DNA by the proteins encoded in the T-plasmid.

The procedure of infection in wild type bacteria is the following: as we can see in FIGURE X, once the bacteria enters the plant the phenolic compounds that the plant secretes induce the transcription and then the translation of the Vir regulon. In it, there are codified the proteins involved in the infection, such as VirB complex, which is in charge of the secretion into the vegetal cell or VirE2 that moves the T-DNA throughout the vegetal cytoplasm guiding it to the nucleus. In particular, the protein that we want to emphasize is the VirD2; it protects the T-strand in the 5’ and guides it out of the bacteria and into the vegetal cell.

Once inside the plant cell, the mature T-complex, which is the T-strand ligated to the VirD2 and VirE2 proteins, enters the nucleus. Inside it, VirD2 recognizes the proteins of the repairing machinery of double strand breaks in the DNA such as the Homologous Recombination (HR) or Non Homologous End Joining (NHEJ) system. Taking advantage of the wound in the genome, the T-DNA is inserted. The fact that VirD2 recognizes the repairing machinery implies that randomness plays an important role in the insertion of the T-strand. This generates multiple possible sites where the T-DNA could be inserted; because of this, the transformation generates different genotypes in the plants. This variable insertion is the reason why huge screening procedures are needed when you want to produce recombinant plants, wasting time and scientific consumptions.

The problem and the possible solution

Given the problem of random insertion, we decided to work on A. tumefaciens, in order to give it the capability of transform site specifically. We were going to direct the DNA using TAL Effector nucleases (TALENs). These proteins are site specific and can also generate a double strand break in the DNA; using this method it is possible to induce in some way the double strand repair machinery in the exact place that you want you insertion to be. To make sure that the T-strand is in the same place that the NHEJ machinery, we wanted to unite the TALENs with the VirD2 protein.

The idea of our system was that the TALEN protein entered the vegetal nucleus and interacted with the specific site of the plant’s genome, and then produced the double strand break as shown in the Figure X. At the same time, the TALENs are associated with the VirD2 protein, and this is united with the T-DNA strand, so it can be inserted in the break recently done by the TALEN. With this, we hoped that the T-strand insertion is site specific.

For our biobricks, the idea was to modify the binary vector, so in it, we have had the TALEN’s sequence (ready to be modified at will) united with the VirD2 sequence.

First we needed to knock out the native VirD2 protein from the T-plasmid, to make sure that the only protein of that kind that enters the plant cell is associated with our TALEN system. To achieve this goal we tried to use homologous recombination. The design of this DNA segment had the sequences of VirD1 followed by a Chloramphenicol resistance cassette that would replace the VirdD2 and then VirD3 (Figure XI,). In the T-plasmid, VirD2 sequence is between of the VirD1 and VirD3 sequences, so the last ones are the homologous regions for the recombination. We chose an antibiotic resistance for later selection in medium.

The second and most important step was to design the correct TALENs. We used a transformed Nicotiana benthamiana plant that contains a specific GFP, mgfp5, in the genome. We designed the TALENs to recognize the mgfp5 sequence as a control in plant transformation. The recognition sequence was chosen using a software created by Felipe Erices, one of our team members, that finds the optimum sequence to cut the gene inside the plant’s genome. An important variable in this recognition is that the sequence is present hopefully once in the whole genome. We used the same software to optimize the TALEN recognition sequence on this DNA sequence.

The last step was to generate the optimized binary vector that would contain the corresponding TALEN sequence united with the VirD2 sequence to the Vir regulon promoter. This new binary vector would work altogether with the modified Agrobacterium tumefaciens that has the T-plasmid (VirD2 knock-out) for a specific DNA insertion for plant transformation.

This is an intermediate plasmid that we used to build the homologous recombination fragment. In order to get a section with a longer region of homology we decided that instead of generating a linear PCR product by adjusting the primers to contain the region of homology, we would create a plasmid by using Gibson Assembly that would allow us to have 600bp of homology in each side of the antibiotic resistant cassette. This construct was done successfully. After the cloning, we tried to knockout VirD2 from Agrobacterium tumefaciens by transforming competent cells (freeze and thaw protocol) with both the linearized fragment from the RH plasmid and with the whole construct.

For the assembly of the TALENs we purchased Addgene: Golden Gate TALEN Kit 2.0. However, even after several attempts, we couldn’t construct the TALENs proteins so we stopped our work on this project. We successfully made half of a TALEN

In this section you will find the complete description of our everyday work in this project and detailed information of our results. Click here.