Various plant species are typically transformed by one of three methods. Arguably, the simplest and most preferable of these methods is transformation using a species of bacteria, Agrobacterium tumefasciens. Agrobacterium naturally transforms its host plants with DNA that causes tumors or galls to grow on the host. It accomplishes this by altering hormone levels in the host plant. The tumorous growth produces ideal tissue for the bacteria to infect. In order to use Agrobacterium for plant biotechnology, researchers replace the tumor-inducing piece of DNA, or plasmid, with DNA that encodes the genes they want to engineer into the host plant. Depending on the host species, this type of transformation can be very simple or can be quite challenging. A general advantage of this transformation strategy is that it typically leads to only one or a few copies of the engineered DNA being introduced to the plant genome, which helps to ensure stable gene expression (i.e. the engineered genes will usually not be switched off by plant defense mechanisms). This transformation strategy also keeps DNA constructs intact, and both of these things make it easier to characterize engineered changes and obtain regulatory approval more quickly. A disadvantage of this approach is that the actual protocol often needs to be tailored to a particular plant species, and sometimes a new strain of Agrobacterium needs to be developed in order to infect particular plant species.
A second widely used transformation method is particle bombardment, also known as the "gene gun method." This strategy entails precipitating the engineered DNA construct (i.e. causing it to come out of solution) onto tiny gold or tungsten beads that will then be shot at small pieces of plant tissue. This "bombardment" will lead to many copies of the DNA construct inside the plant cells. Some of these copies will be integrated in the plant genome. An advantage of this method is that it can be used on many different plant species, including species that have not been previously transformed. Nonetheless, this method has fallen out of favor because it often leads to numerous copies of the engineered DNA construct being present in the genome, which in turn leads to unstable and unpredictable gene expression. Most transgenic lines that suffer from these ill effects would never make it to production because all plants transformed using any method go through several rounds of traditional breeding after transformation to achieve a desirable field (as opposed to laboratory) variety.
Finally, plants can also be transformed by turning their cells into protoplasts and transforming then regenerating these protoplasts. Plant cells are encased by cell walls made of sugar polymers. Protoplasts are cells that have been stripped of their cell walls by enzyme treatments. Protoplasts can be stimulated to uptake DNA in their surrounding solution by shocking them with electricity or treating them with various chemicals such as polyethylene glycol that induce osmotic stress. These protoplasts can then be regenerated into plants, often times with assistance from appropriate hormone treatments. Protoplast regeneration is a challenging and tedious process, so this method is typically only used for certain applications. When attempting to perform very precise genetic manipulations, this method is often preferred. For example, researchers have explored using site-specific recombination to pick a spot in a particular plant chromosome to integrate their DNA construct. This method is rather delicate and works most efficiently when used with protoplast transformation. A very limited number of plants, as far as we know only some species of moss, can undergo homologous recombination, a process that also enables "gene targeting" or integration at a previously designated location. Protoplast transformation is also the method of choice for this type of transformation.
There are other methods used for plant transformation as well, for example methods that employ viruses. However, the three methods mentioned above are the predominant methods employed. Transformation using Agrobacterium may be especially interesting because it reflects the often neglected fact that genetic engineering has been occurring in nature for a very long time. Though humans have only recently developed the knowledge necessary to pursue this technology, other organisms such as bacteria and viruses have been engineering other organisms for their own purposes for countless years. Moreover, I hope that this article will echo the important point made elsewhere that genetically modified organisms need to be considered on a case-by-case basis. A discerning consumer or observer must consider the exact nature of the modification, as well as how the modification was accomplished.