Plants could turn bark beetles’ own attack signals against them

Across Europe and North America, bark beetle outbreaks now rival wildfires in scale – reshaping ecosystems, economies, and carbon balance as climate change accelerates their spread.

Bark beetles coordinate mass attacks using pheromones – volatile chemical signals that act like “group texts,” summoning thousands of insects to the same tree.

In a new study, Olivier Van Aken, a plant biotechnologist at Lund University, Sweden, and his colleagues show how specific steps of beetle pheromone biosynthesis can be reconstructed inside plant cells – effectively rebuilding beetle communication tools inside living plants.

That coordination is what allows beetles to overwhelm even healthy forests – “and that is exactly what we are trying to disrupt,” Van Aken says. “We are not trying to kill insects; we are trying to interfere with how they coordinate their behaviour.”

When bark beetles become a problem

Bark beetles are not villains by default. At low population levels, they play an important ecological role – colonising weakened trees, accelerating nutrient cycling, and contributing to natural forest renewal.

“In a healthy forest, bark beetles are part of the clean-up crew,” explains Olivier Van Aken. “They are an ecosystem function – not an enemy.”

Problems arise when conditions tip the balance. Warmer winters, drought, storm damage, and large areas of uniform forest can favour explosive population growth.

“A single beetle is easy for a tree to fight off,” Van Aken says. “But thousands arriving at once is a different story – the tree’s defences are overwhelmed.”

How beetles coordinate mass attacks

The beetles are not acting alone. Many species live in close association with fungi that help them breach tree defences. These fungal partners weaken host tissues and leave behind blue-stained wood, adding to economic losses – and making outbreaks difficult to control with any single measure.

Crucially, mass attack depends on chemical coordination. Many economically important bark beetles rely on aggregation pheromones – chemical signals that call in reinforcements to the same host.

“This is collective behaviour,” explains Van Aken. “Without those chemical signals, the attack does not scale up.”

But not just any molecule will do. Bark beetles can be highly selective about pheromone structure: even small differences in stereochemistry – the same molecule in a left- or right-handed “mirror” version – can determine whether a signal attracts beetles, repels them, or does nothing at all. Predators, by contrast, are often attracted across a broader range of chemical variants – a difference forest managers may be able to exploit.

Forest management already uses this chemical biology. Synthetic pheromones are widely deployed in monitoring traps, mass-trapping campaigns, and so-called push–pull strategies that combine attractive signals with repellent ones. One of the best-known repellents is verbenone, an anti-aggregation pheromone that signals a tree is already occupied or no longer worth attacking.

In practice, these approaches steer beetle behaviour rather than killing insects outright – but they come with a practical limitation.

“You can absolutely make these compounds in a chemical plant,” Van Aken says. “The question is whether that is the smartest way – especially if you need large volumes, year after year.”

Engineering plants to produce beetle pheromones

To test whether plants could function as pheromone producers, the researchers rebuilt specific steps of bark beetle pheromone biosynthesis inside plant cells, rather than transferring entire pathways at once.

The idea was not that plants naturally use beetle pheromones, but that plant metabolism might offer a more efficient way to make molecules that are otherwise difficult and expensive to synthesise using conventional chemistry.

“Some of these chemicals are just very hard to make with standard organic chemistry,” explains Abraham Ontiveros-Cisneros, first author and postdoctoral fellow in Olivier Van Aken’s group at Lund University. “Plants already make the raw materials. The question is how to redirect the flow?”

Instead of importing complete biosynthetic pathways as a block, the team assembled them as modular routes – rebuilding only the critical links in the chain by combining plant and insect enzymes, each responsible for a specific chemical step.

In practice, getting enzymes from different organisms to function together inside a plant cell is anything but trivial. The approach worked for some pheromones – and failed for others – revealing exactly where insect chemistry breaks down inside plant cells.

As Van Aken puts it, the challenge is not knowing what reactions should happen, but whether the plant will actually carry them out.

At the core of the strategy lies terpene metabolism. Bark beetle pheromones such as verbenol and ipsdienol are derived from the universal building blocks isopentenyl diphosphate (IPP) and dimethylallyl diphosphate (DMAPP) – molecules plants already produce through two parallel metabolic pathways: the cytosolic mevalonate (MVA) pathway and the plastidial MEP pathway.

Building pheromone pathways inside plant cells

To redirect this metabolic machinery, the researchers introduced genes encoding geranyl diphosphate synthases and terpene synthases from conifers and bark beetles, enabling plants to form key intermediates such as α-pinene and β-myrcene – compounds beetles normally acquire from their host trees.

“The final step is handled by cytochrome P450 enzymes from bark beetles – the same enzymes beetles use to turn host terpenes into pheromone signals.”

Those final steps proved especially demanding, highlighting that enzyme compatibility – not gene expression alone – limits whether insect chemistry can function inside plants. Cytochrome P450s are highly specialised, and behaviour that is well characterised in insects does not automatically translate to plant cells: even when transcripts were present, chemical products did not always appear.

“We used Arabidopsis thaliana as our main test system because it is genetically well understood and easy to work with,” Van Aken says. “To push more material into the pathway, we also used a mutant background lacking one copy of FPS1 (farnesyl diphosphate synthase), so fewer precursors are siphoned off into competing processes.”

The logic was straightforward: reduce competition upstream so more raw material flows into pheromone production. To test how cellular context shapes outcome, the researchers also used Camelina sativa (Camelina) and Nicotiana benthamiana (Nicotiana).

“Camelina is interesting because it is fast-growing, genetically tractable, and already used to produce high-value plant compounds,” Van Aken says. “That makes it a realistic candidate for scale-up – not just a laboratory proof of concept.”

Testing where and how the pathways work

In several constructs, enzymes were retargeted to chloroplasts in order to access the MEP pathway directly. This spatial reorganisation proved more than a technical detail: the same enzymes behaved differently depending on where in the cell they operated. In metabolic engineering, location can matter as much as enzyme choice.

Detecting what the plants produced proved unexpectedly difficult, and solvent-based extraction repeatedly failed. “We struggled for a good two years before we actually saw anything on the gas chromatograph,” explains Abraham Ontiveros-Cisneros.

The breakthrough came when the team switched to solid-phase microextraction coupled to gas chromatography–mass spectrometry (SPME–GC/MS), effectively “sniffing” the air around leaves to capture released volatiles.

By enclosing individual leaves and sampling emitted volatiles, the researchers could finally detect pheromones and intermediates as they were released – revealing that production had been happening all along, just not in a form that standard extraction could capture. Gene expression was quantified by qRT-PCR to confirm that introduced pathways were transcriptionally active.

“We were not just asking whether plants can make these molecules,” Ontiveros-Cisneros says. “We were trying to understand where it works, where it fails – and why?”

Plants successfully produced key beetle pheromones

The engineered plants did not all behave the same – and that variability turned out to be one of the study’s most informative results. Some engineered pathways produced clear pheromone signals; others stalled despite strong gene expression.

“When you try to run one species’ chemistry inside another, you quickly see where things work – and where they do not,” Ontiveros-Cisneros says. “And those points of failure are often the most interesting.”

The clearest success came from the verbenol pathway. Arabidopsis plants expressing a combination of plant terpene enzymes and a bark beetle cytochrome P450 produced both cis- and trans-verbenol – key aggregation pheromones – directly from their own metabolism. In several lines, the plants also produced verbenone, an anti-aggregation pheromone that repels beetles.

That outcome went beyond the original design. No enzyme is known to catalyse the conversion from verbenol to verbenone in plants, and the authors interpret its appearance as likely arising through non-enzymatic oxidation once verbenol accumulates. In other words, once production crossed a threshold, chemistry took over.

“That was not something we explicitly engineered,” Olivier Van Aken says. “But it tells us that if you push the pathway far enough, additional signals can emerge.”

Production varied despite identical genes

Verbenone therefore represents both success and a challenge. If plants are ever to function as reliable pheromone dispensers, that final step would need to be controlled rather than left to spontaneous chemistry.

Production levels varied substantially between individual plants carrying identical genetic constructs, and these differences could not be explained by gene expression alone. High mRNA levels did not consistently translate into high pheromone emission, pointing instead to bottlenecks downstream of transcription – at the level of enzyme performance, precursor supply, or overall pathway efficiency.

“That tells us regulation is happening at the metabolic level,” explains Abraham Ontiveros-Cisneros. “It is not just about whether the gene is there – it is about how the whole cellular context responds.”

Reducing competition for precursors helped clarify that point. Plants lacking one copy of FPS1 produced significantly more α-pinene than wild-type plants, confirming that limiting competing pathways can redirect metabolic flux.

“If too much substrate disappears down other routes, nothing downstream will ever scale up,” Ontiveros-Cisneros says. “That is a classic bottleneck problem.”

Some pheromones could not be completed

Not all pathways could be fully rebuilt. In contrast to verbenol, the ipsdienol pathway stalled at the final step. Plants consistently produced β-myrcene – the immediate precursor – but ipsdienol itself was never detected, even after boosting precursor supply, adding enzymes, or retargeting pathways to chloroplasts.

That split – intermediates yes, final product no – is one of the study’s clearest lessons: transferring biosynthetic pathways is rarely plug-and-play, and the final enzymatic steps are often the most fragile.

“That part was frustrating,” Ontiveros-Cisneros admits. “But it also makes the problem very clear: the last enzyme step is where things break down.”

Follow-up experiments point to several likely explanations. The bark beetle cytochrome P450 responsible for the final conversion may not fold correctly in plant cells, may lack the helper proteins needed to deliver electrons, or may operate outside its optimal cellular context.

“An enzyme that works beautifully in vitro does not automatically behave the same way in a living plant,” Ontiveros-Cisneros says.

Why the final enzymatic step failed

Crucially, this is not a dead end. The authors outline concrete next steps, including testing related P450 enzymes from other beetle species, identifying new candidate genes, or supplying the redox partners these enzymes require in vivo.

Experiments in alternative plant systems reinforced the importance of context. Camelina sativa and Nicotiana benthamiana sometimes produced higher levels of intermediates – especially when enzymes were targeted to chloroplasts – showing that both host species and subcellular location shape outcome.

Equally important for future applications, the researchers observed no obvious growth or developmental penalties in engineered Camelina plants or in the fps1 Arabidopsis background used to divert metabolic flux.

“The point is not that plants can do everything,” Olivier Van Aken says. “It is that they can already do a great deal – and we now know precisely where the remaining hurdles lie.”

What plant-made pheromones could be used for

The study does not claim to deliver a ready-made solution for forest protection. Instead, it points to a broader shift: using plants as programmable producers of insect communication signals – and treating ecological signalling itself as something that can be biologically designed and controlled.

That idea extends far beyond bark beetles.

“What excites me is not just this one system,” Van Aken says. “It is the possibility of designing biological production systems for ecological signalling molecules more generally.”

In the near term, the most realistic application is not releasing pheromone-producing plants into forests, but using plants – or other organisms – as sustainable biofactories.

“We are also exploring bacteria,” explains Irene Bassan, a project assistant in Olivier Van Aken’s group at Lund University. “If you want real scale, fermenters are an obvious route.”

Cheaper and more sustainable pheromone production

Today, pheromone-based pest control relies mainly on chemical synthesis – often expensive, energy-intensive, and dependent on heavy-metal catalysts. If plants can produce the same compounds, both cost and environmental footprint could be reduced.

“If plants can supply these molecules at scale, you suddenly lower the threshold for using pheromones more broadly,” Bassan says. “That alone could change how these tools are deployed.”

Crucially, full pathway completion may not be necessary. High production of intermediates – such as α-pinene, which plants already made at substantial levels – could be useful as components of existing pheromone blends.

“Even producing parts of the pathway can be useful,” Bassan says. “Some of these compounds already influence beetle behaviour, and they can also attract natural predators – so partial success could still translate into real control strategies.”

Steering beetle behaviour instead of killing insects

Further ahead lies a more speculative – but conceptually powerful – idea: living dispensers. Instead of relying on plastic traps or chemical lures that must be replaced and maintained, plants themselves could release pheromones continuously and locally.

“In principle, you could use the signals to steer beetles rather than eliminate them,” explains Abraham Ontiveros-Cisneros. “Aggregation pheromones could draw insects toward traps or sacrificial areas, while anti-aggregation signals like verbenone could help protect high-value trees – much like the push–pull systems already used in agriculture.”

That distinction matters ecologically. Conventional traps and lures can inadvertently capture or kill predatory insects along with pests, potentially undermining long-term control. A plant-based dispenser approach could instead be designed to work with predation rather than against it – recruiting natural enemies instead of removing them.

“This is a different way of thinking about pest control,” Olivier Van Aken says. “You are not trying to wipe something out. You are steering behaviour. With conventional insecticides, resistance develops very quickly. Pheromone-based strategies are much less prone to that kind of evolutionary escape.”

At the same time, the work exposes clear scientific limits. Some pheromone pathways transferred successfully; others stalled at the final enzymatic step. Rather than a failure, Van Aken sees this as a roadmap.

“Now we know where the bottlenecks are,” he says. “That tells us where enzyme discovery, protein engineering, or pathway redesign will actually matter.”

Ecological and regulatory limits

Any discussion of living dispensers also raises regulatory and ecological questions. Releasing genetically modified plants into forest environments remains controversial in many regions – and rightly so. Field use would require careful risk assessment, containment strategies, and public engagement.

“This is not something you rush into,” Van Aken stresses. “Ecology does not forgive shortcuts.”

Still, the paper sketches a plausible deployment framework: tree-based intercropping systems, where transformable plants could be grown alongside trees in managed forest–agriculture interfaces. Existing legume-based intercropping systems already support tree growth, soil nutrient content, and carbon sequestration; pheromone-producing plants could, in principle, be integrated under appropriate regulatory and ecological conditions.

Beyond forestry, the approach could extend to other insects whose behaviour depends on chemical cues – including agricultural pests and beneficial species alike. In that sense, the study reframes pest management as a form of chemical ecology engineering: using biology to modulate interactions rather than overpower them.

“If we want sustainable solutions,” concludes Van Aken, “we have to work with biological systems – not against them.”