Plant Biology's Fastest Experiment Only Worked on Nightshades

Plant biology has a workhorse technique that outsiders rarely hear about. Inject a leaf with a modified soil bacterium carrying the gene you want to test, wait two or three days, and the leaf makes your protein. No breeding, no stable transgenic line, no waiting a season. A great deal of what is known about how plants detect and fight off disease was established this way.

The technique has a quiet restriction, and it has shaped the field around itself. It works reliably in the nightshade family, tobacco and tomato above all, and poorly or not at all in most other plants. The consequence is that when a plant scientist wants an answer this week rather than next year, the experiment happens in tobacco. The abstract of this paper puts the problem plainly: the approach remains largely confined to solanaceous species, leaving most economically important crop families without a comparable rapid assay platform.

A team led from Imperial College London reports a bacterial strain that breaks the restriction for a good many plants. An engineered Rhizobium rhizogenes called AS109 drove gene expression in species across several families where the standard laboratory strains manage little or nothing, including lettuce, sesame, scarlet sage, monkey flower, field mustard, and six members of the bean family.

Why it matters: When a method only works in one family, that family becomes where the knowledge accumulates, and crops that feed much of the world get studied more slowly because every question costs a growing season. Widening the method widens which plants can be asked questions quickly.

The study, by King et al. at Imperial College London, with collaborators at Academia Sinica in Taiwan, appears in PLOS Biology.

What it unlocked in a bean

Having established the strain works, the team built a toolkit in faba bean, a crop with no comparable rapid assay before now: tracking where a protein goes inside the cell, silencing a gene by RNA interference, screening which molecules the plant's surface receptors recognise, testing the immune receptors known as NLRs, and imaging the interface where a pathogen meets its host.

The result with the widest implications concerns moving resistance between families. Immune receptors taken from nightshades, both the solo kind and the paired sensor-and-helper kind, still recognised their targets and still triggered the protective cell-death response after being transferred into faba bean. Resistance genes are usually assumed to depend on a web of partner proteins in their own species, so the finding that these ones function in a distant family turns a slow breeding question into a fast laboratory one.

Where it fails, which the paper reports plainly

This is a methods paper, and its value depends on knowing the boundaries. The authors state that Although AS109 did not support transient expression in every species tested, it outperformed the standard strains in most of them. The specific failures matter more than that summary suggests.

Within the bean family the strain worked in alfalfa, fenugreek, snow pea, common bean, faba bean and cowpea, and failed in soybean and mung bean. Soybean is among the most widely grown crops on earth, so the single most commercially valuable target in that family is still out. Across wider sampling, none of the tested laboratory strains, including AS109, mediated clear expression in snapdragon, coffee or red quinoa, so three more families remain closed.

Everything reported is also in dicots. No cereal was tested, which leaves wheat, rice, maize and barley exactly where they were, and those are the crops that supply most of the world's calories.

What the study can't say yet

Nobody knows why the strain works. The authors are candid that However, identifying these factors remains challenging because AS109 differs extensively from commonly used tumorigenic Agrobacterium strains at both chromosomal and plasmid levels, and they frame comparative genomics to pin down the cause as future work. Until that is done, extending the trick to a new species is trial and error rather than design.

It is also worth being clear about what transient expression is. A protein made in an injected leaf for a few days is a test of whether something can work, not evidence that a plant carrying that gene permanently would resist disease in a field. The NLR transfers demonstrate recognition and a cell-death response in a leaf assay; no plant was grown, challenged and scored for yield.

Readers should also know the commercial context, which the journal discloses at length. Several authors hold patents on NLR biology, one co-founded a company working on it, another has received funds from that company, and a further author had signed an agreement for future industry funding at the time of submission. The result is a tool that makes resistance-gene work faster, and some of its authors have a stake in resistance-gene work.

Quick questions

Does this mean new disease-resistant crops? Not directly. It makes testing candidate resistance genes much faster in more species, which is upstream of breeding or engineering anything.

Why would one bacterial strain work where others don't? Unresolved. These bacteria naturally transfer DNA into plant cells, and host range differs between strains, but the genetic basis of this strain's breadth has not been identified.

What is the one-line takeaway? An engineered soil bacterium extends plant biology's fastest experiment to lettuce, sesame, mustard and six beans, and showed nightshade immune receptors still work when moved into faba bean, while soybean, coffee, quinoa, snapdragon and every cereal remain out of reach and the reason it works at all is unknown.