Plants Silence Pathogens with RNA—Turning Defense Signals into a Spray

Plants do not fight pathogens only with antimicrobial compounds and immune proteins. They also use short regulatory molecules called small RNAs, which can alter gene expression through sequence-specific interactions.

Could this molecular “information warfare” be turned into a spray?

On August 27, 2026, an Australian research team reported that spraying barley leaves with synthetic RNAs designed to mimic plant stress-responsive microRNAs (miRNAs) strongly reduced infection by the barley powdery mildew fungus Blumeria hordei.

Original paper: Plant-derived, stress-responsive microRNA mimics reduce powdery mildew infections of barley

Plants and pathogens can fight with RNA

miRNAs are very short RNAs, usually around 20 nucleotides long. They recognize messenger RNAs with complementary sequences and can reduce gene expression by promoting RNA cleavage or repressing translation.

Importantly, RNA-based regulation does not always remain inside a single organism.

In several plant–pathogen systems, plant-derived small RNAs can move into pathogens and suppress pathogen genes. This phenomenon is known as cross-kingdom RNA interference. One well-known example is the interaction between Arabidopsis thaliana and the grey mould fungus Botrytis cinerea, in which plant small RNAs have been reported to enter the fungus through extracellular-vesicle-associated pathways and suppress genes related to virulence.

Plants therefore do not rely only on chemistry. In some interactions, they can use RNA sequence information to interfere with the gene expression of another organism.

Turning RNA interference into a spray: SIGS

Researchers have been trying to exploit this principle for crop protection through spray-induced gene silencing, or SIGS.

In a typical SIGS approach, double-stranded RNA or related RNA molecules are designed to match an essential pathogen gene. The RNA is applied externally to the plant, taken up by the pathogen directly or indirectly, and then used by the RNA interference machinery to suppress the target gene.

Most SIGS strategies therefore start from the pathogen side:

identify an important pathogen gene, then design an RNA to silence it.

The new study takes a different route.

Instead of asking which pathogen gene humans should target, the researchers asked whether plant miRNAs already associated with stress responses might themselves provide useful starting sequences for antifungal RNA sprays.

Four plant-derived miRNAs were selected

The team focused on miRNA families previously associated with plant stress responses and plant–pathogen interactions: miR167, miR444 and miR5048.

They selected four specific sequences—miR167c, miR444b, miR5048a and miR5048b—and computationally searched the B. hordei genome for possible target genes.

Predicted targets included protein kinases, transport-related proteins, ribosomal proteins and other fungal genes. For miR444b, one predicted target was CSEP0284, a candidate secreted effector protein potentially related to virulence.

For miR5048a and miR5048b, protein kinases also appeared among predicted targets across several powdery mildew fungi.

But this distinction is essential: these were computational target predictions. The study did not directly demonstrate that the predicted fungal transcripts were silenced by the sprayed miRNA mimics.

The researchers sprayed synthetic miRNA mimics onto barley leaves

For each of the four miRNAs, the team synthesized a short double-stranded RNA mimic approximately 21–22 base pairs long.

They did not extract natural miRNAs from plants. Instead, they chemically reproduced the sequence information of plant miRNAs as synthetic “miRNA mimics.”

Four-centimetre segments of leaves from barley cultivar Stirling were placed on agar. Each miRNA mimic was sprayed at a concentration of 100 ng/µL, and B. hordei spores were inoculated 12 hours later.

The researchers then evaluated infection at several stages, including 2.5 days after inoculation, 6 days after inoculation and later visible disease development.

This time-course revealed an important pattern: different RNAs did not appear to affect infection in exactly the same way.

miR167c reduced early infection establishment to about one-tenth

At 2.5 days after inoculation, the researchers examined 500 fungal spores and measured the proportion that successfully developed micro-colonies after penetrating barley tissue.

In the water-treated control, the micro-colony index was 15.7%.

With miR167c, it fell to just 1.5%—roughly one-tenth of the control level.

miR5048b reduced the index to 4.3%, while the four-miRNA mixture produced 5.7%. miR444b also reduced the value to 9%.

miR5048a, however, produced a value of 11.5%, which was not statistically different from the water control at this early time point.

If the experiment had ended there, miR5048a might have looked relatively weak.

But six days later, it produced the strongest signal in the study.

miR5048a showed the strongest reduction in fungal DNA at day 6

At 6 days after inoculation, the researchers extracted DNA from the infected leaves and used qPCR to measure the fungal ITS region.

All four individual miRNA mimic treatments significantly reduced the relative abundance of B. hordei ITS DNA compared with controls.

The largest reduction occurred with miR5048a.

Relative to the water-treated control, B. hordei ITS DNA was reduced to roughly one-sixteenth. Relative to a non-target RNA control generated by the RNA synthesis system, it was reduced to about one-thirteenth.

This result needs careful wording. The study did not directly weigh fungal biomass or determine an absolute fungal cell count. It measured the relative abundance of fungal ITS DNA within a fixed amount of total extracted DNA by qPCR.

Even with that limitation, the contrast between the early and later results is striking. miR5048a did not significantly reduce micro-colony establishment at 2.5 days, yet it produced the strongest reduction in fungal DNA at day 6.

This raises the possibility that different miRNA mimics influence different stages of infection—some acting strongly during early establishment, others during later fungal growth or spread.

The present study cannot yet assign those mechanisms, but it shows why a single time point may not be enough to judge RNA efficacy.

Mixing all four RNAs did not make the treatment strongest

Another revealing result came from the four-miRNA mixture.

If different RNAs attack different fungal genes, one might expect a cocktail to outperform every individual component.

That did not happen.

The mixture reduced the micro-colony index to 5.7%, but miR167c alone reduced it further, to 1.5%.

The mixed treatment was also associated with increased leaf yellowing compared with controls.

The reason is not yet known.

At minimum, the result argues against the idea that RNA sprays can simply be improved by combining more active sequences. Dose, sequence interactions, plant responses, target competition and delivery efficiency may all matter.

The study did not yet prove that the predicted pathogen genes were silenced

This is the most important limitation when interpreting the paper.

Spraying the miRNA mimics reduced powdery mildew infection.

However, the study did not establish the complete causal chain:

miRNA mimic enters the fungus → binds the predicted fungal transcript → reduces that transcript → reduced gene expression causes the infection phenotype.

The authors explicitly note that future RT-qPCR and transcriptomic analyses will be needed to test whether predicted fungal target genes are genuinely downregulated.

The study also does not demonstrate that these same four miRNAs are naturally transported from barley into B. hordei during infection and silence the same genes under normal conditions.

It is therefore too early to say that the researchers simply reproduced a known barley defense pathway as a spray.

What they demonstrated is more precise: plant stress-responsive miRNA sequences can serve as starting points for synthetic RNA molecules that function as promising SIGS candidates.

Why the design strategy is still important

Traditional SIGS development generally begins by examining the pathogen genome, identifying an essential or virulence-related gene and designing an RNA against it.

This study proposes another search strategy.

Look at the miRNAs plants have evolved, then ask which pathogen genes those sequences might already be capable of targeting.

In other words, the plant’s own miRNA repertoire could become a search space for future RNA-based crop-protection molecules.

According to the authors, this is the first report demonstrating suppression of fungal infection using sprayed, plant-derived stress-responsive miRNA mimics in a SIGS context.

That does not mean the approach is ready for agriculture. But it broadens the way researchers can discover candidate RNAs.

This is not yet a finished “RNA pesticide”

The experiments were performed on detached barley leaf segments placed on agar, not on whole plants in a field.

Whether the same effect will persist in greenhouse and field conditions remains unknown.

RNA molecules are also vulnerable outdoors. UV radiation, rain, leaf-surface enzymes and environmental exposure can shorten their lifetime. Practical SIGS technologies therefore need solutions for RNA stability, formulation and delivery. Nanoparticle and carrier-based approaches are among the strategies being explored in the wider field.

Off-target effects must also be considered, including effects on the crop itself and on non-target organisms.

The increased yellowing observed in the RNA-mixture treatment is a reminder that a biologically active RNA formulation should not be judged solely by how strongly it suppresses the pathogen.

Plant Hack: borrowing solutions that plants have already explored

Applied plant science often focuses on adding new capabilities to crops.

But plants have spent hundreds of millions of years interacting with pathogens. Their molecular systems contain countless solutions that humans have only begun to decode.

That is what makes this study particularly interesting.

Instead of designing every RNA pesticide entirely from scratch, researchers may be able to examine the RNAs plants already produce and ask which of them can be repurposed.

Which RNAs does a plant make under stress? Which pathogen genes can those sequences recognize? Which combinations suppress infection without harming the host?

If those relationships can be mapped systematically, future crop protection may increasingly follow a new design principle:

read the plant’s defense strategies, then reuse selected parts of them from the outside.

Tiny RNAs already move through the molecular battlefield between plants and pathogens.

Learning how to decode and redeploy that information may become one path toward a new generation of crop-protection technologies.