Agrobacterium-mediated transformation has been the dominant method of plant genetic modification since the birth of the field, but not all plants are easily infected by Agrobacterium. Even then, those that accept its foreign DNA might not grow into fully engineered plants. When a plant tissue undergoes an Agrobacterium infection procedure, not all cells will be successfully attacked. There are many factors in this low transformation rate; for example, the transfer DNA (T-DNA) from the Agrobacterium is incorporated in the genome at a random site, which can disrupt essential genes, resulting in the cell’s death. More frequently, the T-DNA will enter the nucleus, but not be incorporated into the genome, resulting in temporary “transient” expression as the transgene is expressed, but will eventually be degraded. Sometimes the Agrobacterium won’t succeed in infecting the plant cell at all. Because of this, a plant grown after Agrobacterium-mediated transformation would be “chimeric:” different regions have different genetic backgrounds. This means that the desired genetic changes are unlikely to reach subsequent generations, so transgenic plants cannot be reproduced.

Since the goal is a genetically uniform plant, experts in plant tissue culture have learned to regenerate an entire plant from a single cell. By including an herbicide resistance gene in the Agrobacterium’s T-DNA, successfully engineered cells will be able to survive on growth media with that herbicide, unlike the non-engineered cells. By subjecting infected plant tissue to this selection while also stimulating cell division through the use of plant hormones, the transformed cells should replicate while the others die.

Ideally, the products of this process are transgenic calli: masses of undifferentiated cells that originated from a single dividing unit. From there, different hormones are included in the growth media, which can either induce organogenesis, which is the formation of tissues such as roots and shoots from the clump of callus, or somatic embryogenesis. In the latter method, embryos, like the ones in seeds, are developed from regular somatic, or “body” cells. A plantlet forming from a somatic embryo will have the structure of a seedling, with baby leaves growing from the top and roots growing from the bottom. Both methods have their pros and cons, but some plants are difficult to regenerate using either technique. For many species, stable and efficient transformation protocols remain elusive. 

Despite its drawbacks, though, Agrobacterium-mediated transformation is not going away any time soon. In addition to incorporating new genes via T-DNA insertion, Agrobacterium can also be used as a vessel for CRISPR/Cas9 genome editing. By including CRISPR systems in the bacterial T-DNA, the Cas9 endonuclease can edit the genome with the help of the CRISPR array and guideRNA (gRNA). From there, the foreign DNA can be backcrossed out through breeding, or by modifying an Agrobacterium Vir (virulence) gene responsible for T-DNA delivery. With the latter example, the CRISPR/Cas9 complex is introduced to the nucleus, but because the T-DNA insertion machinery in Agrobacterium is compromised, it never gets added to the host genome. During this period, Cas9 can make its cut before the unincorporated DNA is degraded. 

However, although the use of CRISPR machinery enables precise edits, it still requires that plants be regenerated through tissue culture. As a result, researchers have sought out ways to skip the regeneration process altogether. One approach to this employs viruses to make heritable, transgene-free edits in the genome.

The tobacco rattle virus (TRV) is a type of virus that can infect over 400 plant species, including model organism Arabidopsis thaliana. Unlike Agrobacterium, with activity limited to the region of infection, TRV can enter one cell, then proceed to invade the rest of the plant. This includes germline cells, which are responsible for producing the next generation. If the TRV brings a way to cut specific genes, germline cells edited by TRV will have heritable mutations that can be passed to subsequent generations by seed, completely eliminating the long process of regenerating plantlets from a single cell using tissue culture.

To get into plant cells, however, the TRV hitches a ride from Agrobacterium, as the bacterium’s delivery system is unparalleled when it comes to infiltrating intact plant cells; outside of biolistics, other mechanisms are not well-suited to penetrate plants’ rigid cell walls. TRV is a bipartite RNA virus, meaning that it is composed of two parts: TRV1 and TRV2. The genetic code for those components can be delivered by Agrobacterium into the plant cells, where they are expressed in the nucleus, and TRV1 and TRV2 assemble and spread throughout the plant. This means that the virus can enter the cell through transient expression, which happens much more frequently than successful T-DNA integration.

By including an editing system like CRISPR/Cas in TRV, edits can be made wherever the virus goes. However, the Cas9 enzyme is over 1,300 amino acids long, which is too large for efficient viral transport. To avoid this problem, a smaller system is used: TnpB, an ancestor of the Cas family, only around 400 amino acids long. Since TRV is an RNA virus, it won’t pass onto the next generation, but in the infected plant, the TnpB can make cuts in the DNA. 

Researchers edited TRV2 to contain a TnpB to target genes responsible for plant pigment, so successful edits should be visibly different from wild-type (wt) plants. To boost editing efficiency, they also tested factors such as different types of TnpBs, heat shock periods, mutants with reduced transgene silencing, and the use of a self-cleaving ribozyme derived from the hepatitis delta virus. They also observed the effects of edits in ku70 mutants, a gene responsible for emergency DNA repair. Ku70 can fix the cuts made by TnpB too quickly, so using a slower form of ku70 can give other repair mechanisms (that make the desired mutations) time to act. These mutants had an average editing efficiency of 8.9%, substantially greater than the 3.3% in wt plants.

One of the pigment genes targeted was PHYTOENE DESATURASE3 (PDS3), and when successfully knocked out in both chromosomes, results in an albino phenotype. Three weeks after infection, researchers noticed white speckles on some leaves, and by sequencing the genome through amp-seq, they found an average editing efficiency of .6-8.9% when using the system with the HDV ribozyme sequence.

From there, they selected a plant with high editing efficiency (54.5%) and collected its seeds. Out of 2,318 seeds sown from this plant, 68 grew into albino seedlings, 41 of which were sequenced, confirming the disruption of the PDS3 gene in both chromosomes. 168 green seedlings were sequenced as well, and eight were found to be heterozygous, meaning one copy of the gene was edited.

To ensure this method could be applied to another gene, researchers chose to target another pigment gene with a clear phenotype. Chlorophyll gene CHLl1 was selected, which shows in a yellow color in homozygous mutants. Yellow regions were observed just two weeks after infection, and after collecting seeds from a plant with 67.4% editing efficiency, fully yellow seedlings were foudn in 8.5% of the next generation. Sequencing confirmed these edits, and RT-PCR (a technique that measures RNA levels) found no TRV in the five albino seedlings that were tested. 

Researchers also wanted to ensure that no off-target edits were made by TnpB, because unintentional errors are sometimes invisible, but other times, they can be catastrophic. For TnpB to be adapted as a new genome-editing technology, it has to be precise, so three albino plants underwent whole-genome sequencing. When filtered against the control plants, 4-5 variants were detected in each plant, but none of these sites matched the off-target sites predicted for the target sequence. Therefore, these variations are likely results of natural mutations.

This technology is new, but its results so far are inspiring. With the difficulty of regenerations, methods to edit germline cells hold immense potential in spurring the development of edited plants, which will hopefully face less regulatory trouble than traditional GMOs (more on that next month). Even using Agrobacterium, Cas9’s massive size is thought to hamper T-DNA delivery, so the development of TnpBs like the ones used in this study could be revolutionary for the future of plant gene editing. Moments like these serve as a reminder that history is now, especially when it comes to biology.

Having done research exclusively in tissue culture, I am certainly familiar with how much of a headache it can be to figure out how to turn a single glowing cell into a full plant. In fact, I still haven’t done it yet.

Thus, it is very exciting to hear about the breaking research being done that might resolve the need for regeneration through tissue culture at all. Of course, this paper just came out in April, and TRV doesn’t infect every single plant species, but it’s riveting to think about what this field will look like a decade from now, even if this technology doesn’t singlehandedly revolutionize it. Especially with Jennifer Doudna on this paper, this work emphasizes how the research happening right now can completely change the future.

I would like to note that germline editing can happen with normal Agrobacterium-mediated transformation through the floral dip method, which can yield edited seeds. However, the range of species susceptible to this technique is even more narrow than tissue culture methods, which is why I didn’t focus on it with this article. I wanted to spend more time on this awesome new paper, which I hope to hear more about in the future. Stay tuned to learn with me!

Check out the paper about the use of TnpBs and TRV for germline editing! https://www.nature.com/articles/s41477-025-01989-9

Agro infection and plant tissue culture:

https://academic.oup.com/plcell/article/28/7/1510/6098292?login=true

Organogenesis and somatic embryogenesis: https://www.sciencedirect.com/science/article/pii/B9780323907958000060