Understanding how genes are switched on and off in specific cell types remains one of biology’s central challenges. While AI has made major progress in decoding the regulatory logic of DNA, applying these approaches across datasets, tissues, and species has remained difficult. In a new study published in Nature Methods, a research team led by Prof. Stein Aerts (VIB & KU Leuven) presents CREsted, a software package that enables both the analysis and design of gene regulatory elements in a systematic and scalable way.

Enhancers are gene regulatory elements—short DNA sequences that control when and where genes are active. Deep learning models can help decode this “regulatory code,” but existing approaches are often tailored to one dataset or one task, making them hard to reuse or extend.

To address this, Prof. Stein Aerts and his team developed CREsted, a new framework that turns enhancer modeling from a collection of one-off analyses into a more systematic and reusable workflow.

“We wanted to move beyond one-off models,” says Niklas Kempynck, Ph.D. student in the Aerts lab. “CREsted allows researchers to systematically study enhancer logic across biological systems, starting from cell-by-cell maps of accessible regulatory DNA and going all the way to sequence design.”

CREsted brings together several steps that are usually handled separately: preprocessing, model training, interpretation, and synthetic enhancer design. It is also built to fit into existing single-cell analysis workflows, making it easier for researchers to adopt and use.

“With CREsted, we give researchers a complete workflow,” says Dr. Seppe De Winter, who shares first authorship with Kempynck. “You can train deep learning models on chromatin accessibility data, interpret which regulatory features they capture, and then use those models to design new DNA sequences with predicted cell-type-specific activity.”

To show its versatility, the team applied CREsted to multiple systems, including mouse brain tissue, human immune cells, cancer cell states, and zebrafish development. Across these settings, the framework identified regulatory patterns, predicted enhancer activity, and enabled the design of synthetic enhancers, which were validated in vivo in zebrafish.

For Prof. Stein Aerts, Scientific Director of VIB.AI, the strength of CREsted lies in making a powerful development more coherent and reusable.

“CREsted makes it much easier to train, interpret, and compare enhancer models across datasets,” he says. “That is important if we want these approaches to become broadly useful, not just for understanding regulatory DNA, but also for designing and testing new sequences in a systematic way.”

Taken together, the work shows how AI can help move the field from describing regulatory DNA to actively exploring and designing it. With applications ranging from fundamental biology to biotechnology and medicine, CREsted lays the groundwork for more systematic and programmable control of gene regulation.

Advanced technology can help farmers get to the root of a growing problem—overwatering in an era of increasing drought and water scarcity. A new UC Riverside system can map soil moisture tree by tree, so growers water only where and when it’s needed.

This system, detailed in the journal Computer and Electronics in Agriculture, was led by the research group of Elia Scudiero, associate professor of precision agriculture and the Director of UCR’s Center for Agriculture, Food, and the Environment (CAFE).

Why traditional soil sensors fall short

Water management is one of the biggest challenges facing agriculture in California and other dry regions. Currently, some growers rely on soil moisture sensors buried in the ground to determine when to irrigate. These sensors are expensive and typically installed in only a few locations, leaving growers to guess how conditions vary across hundreds or thousands of trees.

“The information those sensors provide is very limited,” Scudiero said. “It really only tells you what’s happening in the immediate areas where they’re placed.”

Even when sprinkler systems deliver the same amount of water throughout an orchard, the soil moisture and its availability to trees can vary greatly from spot to spot within a single field.

How soil differences affect water use

One reason is soil texture. Fine soils packed with tiny particles hold water tightly because they have more surface area where water can cling. Sandy soils contain larger particles and fewer small ones, which allows water to drain more quickly. These differences can leave neighboring trees experiencing very different conditions.

The new system replaces limited sensor data and guesswork with detailed maps. A robot moves through an orchard measuring a property of the soil called electrical conductivity. These readings, combined with data from the fixed moisture sensors already in the ground, allow researchers to build a statistical model that predicts water content across the entire field.

Electrical conductivity indicates how easily electricity moves through the soil and is influenced by factors including moisture as well as salt and clay content. By pairing those measurements with direct water readings from buried sensors, the system can translate conductivity into accurate estimates of soil moisture.

From conductivity data to tree maps

The result is a tree-by-tree picture of water distribution. “Using this method, growers will finally know how much water they have, and how much they need, and can water specific trees if they’re dry,” Scudiero said.

Maintaining the right moisture level is important for plant health. Trees that receive too little water become stressed, and more vulnerable to pests and disease. Too much water, however, can deprive roots of oxygen as soil pores fill with water rather than air. “There’s a sweet spot,” he said.

Enhanced precision could also keep orchards from folding. Growers already face tightening regulations on groundwater use while water costs continue to rise.

“If water becomes limited, farmers have two choices,” Scudiero said. “They can retire orchards, or they can find ways to produce the same crops using less water.”

Environmental and economic ripple effects

The technology may also reduce fertilizer pollution. When fields are overwatered, nutrients applied to crops can wash below the root zone and into groundwater, polluting it.

“If you apply only the amount of water the plants actually need, you reduce the risk of washing those nutrients away from the roots of the crops and into the environment,” Scudiero said.

This project has been years in the making. Researchers began developing it in 2019 through collaborations between agricultural scientists and engineers at CAFE.

For Scudiero, it represents the realization of a long-standing goal. He has studied soil conductivity technology for about 15 years and had hoped to someday pair it with autonomous vehicles capable of surveying entire fields.

From research orchards to real farms

The team has already filed a patent related to how the robot interacts with sensors without disturbing their measurements. This research was conducted at the UCR Citrus Research Center & Agricultural Experiment Station. Future work will focus on testing the system with commercial growers beyond the university’s research orchards.

Moving from research plots to real farms will require rugged machines capable of operating in all weather conditions and across different crop systems. Private industry partners may eventually adapt the technology into commercial products.

The work is part of broader efforts at UCR to further the field of precision agriculture, where researchers are developing technologies that combine robotics, sensors, and data science to help farmers manage resources more efficiently.

For farmers facing limited water supplies, the payoff for this research could be significant.

“More crop per drop!” Scudiero said.

Take a typical fish out of the water and it won’t live long. It gets the oxygen it needs from the water it swims in. In a similar way, scientists are exploring dependency as a method of controlling what microbes can do and where they can do it.

Microbes—microscopic, single-cell organisms that include such things as bacteria, fungi, algae and viruses, to name a few—already do a lot. Life as we know it—our lives included—would not exist long without the many jobs microbes do within us and all around us. Many are great allies to us. A small percentage can take us out.

Finding ways to harness and steer their many properties and abilities is the focus of much research in synthetic biology, the emerging discipline that takes pieces of DNA—the genetic code—and reconfigures it in new ways.

Aditya Kunjapur, the Thomas Willing Early Career Associate Professor of Chemistry and Biomolecular Engineering at the University of Delaware, has extensive expertise in this work, with interest in expanding the genetic code and a special focus on biosecurity and biocontainment.

In a publication in Nature Microbiology, Kunjapur and members of his lab showed how they have been able to make one microbe dependent on another—a dependence that could one day be used to restrict microbial activity to a specific area.

This is of great significance as new technologies emerging in synthetic biology could help us address a wide array of problems in health, agriculture and the environment. It is essential that these developments come with appropriate and effective safeguards to prevent harm and avoid unintended consequences.

These are challenges Kunjapur and his lab are addressing. They’re looking for ways to make better—and safer—use of engineered microbes by ensuring their survival as they pursue their mission and by establishing boundaries to keep them from going off course.

“This could be useful in an environment where you want to ensure that one microbe can survive for an indefinite time, but only in a defined region,” Kunjapur said.

In an award-winning essay published in the journal Science in 2024, Kunjapur explained the questions his team is addressing, work it hopes will lead to better vaccines, better immune responses and effective solutions for the increasing problem of antimicrobial resistance that threatens public health around the world.

“Our primary hypothesis is that engineering cells to access a broader chemical repertoire of building blocks can improve live bacterial vaccine efficacy,” he wrote.

It’s a complex endeavor indeed. The goal is to develop a self-contained microbial system that will deliver special residues to disease-creating cells and trigger stronger immune responses. To enable these next-generation vaccines, Kunjapur and postdoctoral researcher Neil Butler co-founded Nitro Biosciences in 2023.

Now Kunjapur’s team has shown one way to limit the survival of a microbe. They trained two strains of Escherichia coli (better known as E. coli) bacteria to work together. One strain was trained to create a non-standard amino acid, an amino acid that is scarce in nature. The other strain was engineered to be dependent on that synthetic amino acid.

Earlier reports in the field had shown that a microbe could be engineered to depend on a synthetic amino acid if that synthetic amino acid was directly supplied.

But the need to directly supply a synthetic amino acid would probably not work well in many contexts. Then-doctoral student Mandy Forti and Kunjapur wondered if they could create a steady supply of the synthetic amino acid by having another microbe produce it, thereby creating a sort of self-contained system. They then showed that the dependent strain was sustained by the synthetic amino acid, with almost no escaping microbes.

“Our whole exploration is how to mitigate that risk,” Kunjapur said. “We do that in a laboratory setting. We’re not trying to rush to put this into the environment. We’re studying the mechanisms to see how the system could break. There is a lot of failure analysis.”

Building and testing a self-contained, dependent system was a complex process for Kunjapur’s team and involved several significant landmarks:

• First, a strain of E. coli bacteria was trained to make a custom, non-standard amino acid that could sustain another bacterial strain without requiring an external supply.
• Next, a separate strain was trained to use the custom non-standard amino acid to produce a fluorescent protein to show that it could successfully use that custom amino acid.
• Next, a strain was designed that relied on that custom amino acid for sustenance.
• Finally, the two strains were grown together. The dependent strain survived as long as the producer strain was in its environment. It could not survive without that producer.

The dependence of one microbe on only one other specific microbe also worked in a context that included other microbes.

“That was the surprise element,” Kunjapur said. “The expectation was that other microbes would ruin biocontainment—that it was a matter of how they would do that, not if they would. But Mandy’s system worked. We have created an ecology that in this context looks very exclusive.”

Forti, lead author of the paper who earned her doctorate last year, said it was amazing to see how the bacterial strains survived and proliferated and to see that after two weeks there was very little escape.

There are still many questions to answer before this technology is ready for use outside of a laboratory setting.

“There are so many other variables we can test,” Forti said. “We need to look at every possible variable these strains could encounter in the environment.”

Compounds in psychedelic drugs like DMT, psilocybin, and psilocin are naturally produced in certain plants, fungi, and animals, and have a long history of use in spiritual and therapeutic contexts. Now, a considerable amount of research is going into determining how these compounds can be translated into a therapeutic context for several mental health conditions. But to do this, researchers need to find a more sustainable way to source these compounds, as current methods raise ecological and ethical concerns.

One research team has been tinkering around with different ways to produce psychedelic compounds in plants, allowing for more sustainable and scalable production. In their new study, published in Science Advances, they describe how they mapped the DMT biosynthetic pathway and engineered a type of tobacco plant to produce five different natural psychedelics.

Psychedelics for mental health?

The better-known hallucinogenic psychedelics contain a class of compounds called indolethylamine, made up of tryptamine and its derivatives. These compounds have been shown to promote neuroplasticity and modulate serotonin, and have shown to have therapeutic potential for depression, anxiety, posttraumatic stress disorder (PTSD), and addiction. In 2019, the drug psilocybin even received the Food and Drug Administration’s “Breakthrough Therapy” designation for major depressive disorder.

“The expanding clinical interest in psychedelics as therapeutics has sparked the need for scalable and versatile production platforms and structural diversification. Traditionally, the supply of psychedelics relies on natural producers, mainly plants, fungi, and the Sonoran Desert toad. Harvesting these organisms for their psychoactive compounds raises ecological and ethical concerns, being increasingly threatened by habitat loss and overexploitation.

“While synthetic routes for these compounds are available and, in some cases, relatively straightforward, they still require compound-specific reactants, can lead to unwanted intermediates and products, and require several processing steps,” the study authors explain.

Mapping pathways for synthetic production

The new study focused on engineering plants to produce five major natural psychedelics: DMT, psilocin, psilocybin, bufotenin, and 5-MeO-DMT. They first had to identify and characterize the key biosynthetic enzymes from certain plants, fungi, and the Sonoran Desert toad and combine enzymes from different species to reconstruct entire biosynthetic pathways. They then used genetic engineering to introduce these enzymes into a type of tobacco plant (Nicotiana benthamiana), which was chosen because it is easily cultivated and produces tryptophan.

After the genes required for production of the compounds were identified, they were introduced to the plant by a process called agroinfiltration, where plant leaves are injected with a suspension of bacterium to induce the expression of genes. The team used AlphaFold3, an AI model that predicts 3D structures and interactions of molecules, to guide their design of a mutant protein that substantially enhanced indolethylamine production by improving efficiency of the enzymes needed to produce it.

The team also created halogenated indolethylamine analogs with potential therapeutic value, which are not typically found in nature.

“Several halogenated indolethylamine derivatives have demonstrated therapeutic potential for mental disorders. For example, 5-chloro- and 5-fluoro-DMT induce head-twitch responses in mice, while 5-bromo-DMT displays sedative effects, and its 5,6-dibromo analog shows antidepressant-like activity,” the study authors write.

The team’s final proof of concept was the tobacco plant containing all five natural indolethylamine compounds. They say all five compounds were detected in the plant one week after infiltration. However, all compounds, perhaps unsurprisingly, were present in lower concentrations compared to their original sources.

A platform to build upon

The researchers have been able to further tweak their model and increase DMT production, saying, “Notably, rational design of a single amino acid substitution in wild-type AtCOMT, guided by AlphaFold3 structural modeling, resulted in a remarkable 40-fold increase in 5-MeO-DMT 10 production in N. benthamiana.”

They note that, ultimately, their model was meant as a kind of platform to determine feasibility. There are many possibilities, including further enzyme engineering and pathway balancing to improve yields, stable integration of pathways into crop plants for large-scale production, and adaptation for edible plants or microdosing applications. The team also says this work could set the ground for parallel and simultaneous production of psychedelic indolethylamines in microbial systems.

The study authors write, “Blending catalytic functions across the tree of life, coupled with metabolic engineering guided by rational protein design of mutant enzymes, enabled substantially more efficient plant production of the indolethylamine components. This work establishes a versatile platform for concurrent biosynthesis and diversification of psychoactive indolethylamines, paving the way for their production in plants.”

A team led by principal investigators Bobo Dang and Ting Zhou at Westlake University/Westlake Laboratory have developed a high-throughput platform for engineering fast-acting covalent protein therapeutics. Their study, titled “A high-throughput selection system for fast-acting covalent protein drugs” published in Science, opens new avenues for next-generation biologics.

The kinetic challenge in covalent proteins

Covalent small-molecule drugs have shown great success in cancer therapy by forming irreversible bonds with their targets. This has inspired efforts to extend covalent strategies to protein therapeutics, especially engineered miniproteins. However, their development is limited by a kinetic mismatch: Miniproteins are rapidly cleared in vivo, whereas covalent bond formation is typically slow. In addition, high-throughput platforms for systematically optimizing covalent protein reactivity have been lacking.

To address this challenge, the researchers proposed that precise spatial positioning of chemical warheads within protein scaffolds could enable molecular preorganization, thereby accelerating covalent bond formation without increasing intrinsic reactivity.

Based on this concept, the team developed a high-throughput platform that combines yeast surface display with chemoselective protein modification to screen diverse crosslinkers and millions of protein variants. By optimizing warhead placement and the local chemical environment, the platform enables rapid and irreversible target engagement.

Designing a fast-acting PD-L1 antagonist

Using this platform, the researchers developed a covalent antagonist targeting PD-L1, termed IB101. Structural analysis revealed that IB101 forms a defined binding pocket that precisely positions the warhead in a reactive conformation, greatly accelerating covalent bond formation. Functionally, IB101 effectively blocks the PD-1/PD-L1 immune checkpoint pathway and demonstrates strong antitumor activity in mouse models. Notably, despite its short in vivo half-life, IB101 achieves durable target engagement and tumor suppression, outperforming conventional antibody-based therapies under comparable conditions.

Engineering cytokines with lasting signaling

The platform was further applied to cytokine engineering, leading to the development of a covalent IL-18 variant, IB201. This engineered cytokine rapidly forms a covalent interaction with its receptor, enhancing signaling strength and duration. In vivo studies showed that IB201 induces potent antitumor immune responses without detectable systemic toxicity. These results highlight the potential of covalent engineering to improve the efficacy and safety of cytokine-based therapies.

Expanding to antivirals and future biologics

Beyond immunotherapy targets, the platform was also applied to develop a covalent inhibitor targeting the receptor-binding domain (RBD) of SARS-CoV-2. This molecule achieves durable viral neutralization, demonstrating the versatility of the approach across different therapeutic modalities.

This study establishes a general strategy for engineering fast-acting covalent protein therapeutics. By enabling covalent bond formation on timescales compatible with rapid in vivo clearance, the platform overcomes a fundamental limitation in the field.

These findings provide a new framework for designing biologics with both rapid kinetics and sustained target engagement, with broad implications for cancer immunotherapy, antiviral therapy, and beyond.

Gene drives—a genetic engineering approach that quickly spreads specific genetic changes throughout a population, whether to kill it off or add a new trait—may have potential for controlling weeds. But so far, gene drives have primarily been studied in mosquitoes, and have yet to be deployed in the real world.

How seed banks complicate gene drives
In a first-of-its-kind study, researchers modeled how a gene drive would proceed in plants. Their simulations suggest that a gene drive’s success may hinge on seed banks—underground reservoirs of seeds that can germinate years or even decades later. Without proper consideration, they found, these stored seeds can slow down or even doom the gene drive, because they continually reintroduce plants without the gene drive into the population.

Modeling studies like this one can help scientists design successful gene drives in plants and discover and mitigate potential problems before deployment in the wild, the researchers said.

Jaehee Kim, assistant professor of computational biology in the Cornell Ann S. Bowers College of Computing and Information Science and the College of Agriculture and Life Sciences (CALS), and Philipp Messer, professor of computational biology in CALS, are co-authors on the new study, “Seed dormancy shapes gene drive dynamics in plants,” published in Nature Plants.

The development of CRISPR-Cas9 gene editing technology, which allows scientists to make precise changes in the genome, has made gene drives more feasible in the lab. But there are still serious concerns that they may spread to non-target organisms and cause ecological damage in the wild.

“People have been thinking about gene drives for decades, but it was always kind of this science fiction technology,” Messer said. “With the advent of CRISPR technology, this has all changed, and the engineering of gene drives has finally come within reach. Yet there’s still some experimental questions, a lot of modeling questions, and so far, nobody has really released one.”

New plant gene drive systems

Recently, researchers developed two gene drive systems for plants in the lab—CAIN and ClvR—that are reliably passed down to offspring and cause the plants to produce inactive pollen, ovules or both.

“Gene drives have been suggested as an alternative control measure for weeds, but their feasibility in plant species had never been demonstrated experimentally before CAIN and ClvR,” Kim said.

Kim and Messer’s team developed a modeling framework to simulate how these two gene drive approaches would play out over time. The model considers how many viable pollen or ovules each plant produces, how long seeds survive in the seed bank and how many of those seeds germinate each year.

Seed banks are a key part of understanding gene drives in plants, Kim said, because they set plants apart from other gene drive species scientists have investigated.

Findings on spread and safety

The simulations predicted both CAIN and ClvR gene drives would successfully spread mutations through the population. However, the longer that seeds survive in the soil, the longer it takes for the engineered mutations to spread. Additionally, scientists may need a greater number of engineered seeds or plants to start off the gene drive, to drown out stored seeds that germinate later on.

Despite the challenges presented by a seed bank, it potentially provides a major benefit. Stored seeds may act as an “evolutionary buffer” by weakening the gene drive so it won’t take off in the wrong place.

“Even if it got accidentally released, or there was spillover to an unwanted population, a seed bank can cause it to die out naturally,” Kim said. “It acts as a natural biosafety measure.”

The researchers hope their model will serve as a foundation that will one day help field biologists plan successful, yet contained, gene drives.

“People thought that gene drives in plants really wouldn’t work that well,” Messer said. “But after this modeling study, I think plants may actually be one of the better systems to try out a gene drive.”

Stepping away from its billions-of-years-old role as a genetic blueprint, DNA is now embarking on a new journey as an active field agent within cells. This research by a team led by Professor Jongmin Kim and Ph.D. candidate Geonhu Lee from the Department of Life Sciences at POSTECH (Pohang University of Science and Technology) was published in Nature Chemistry.

A cell is like a small, tirelessly operating factory. Within this factory, proteins and RNA act as the field workforce, being produced when needed and degraded once their roles are fulfilled. In contrast, DNA serves as the blueprint orchestrating all these programmed activities. Thus, it is critical to store this blueprint safely within the factory, and it should not be misplaced or modified unintentionally.

While DNA can sometimes be utilized as a tool rather than a genetic material—such as in PCR tests to check for coronavirus infections—these manipulations of DNA typically are operational only outside the cell. When inside a living cell, DNA becomes restrained once again to play its original role as a blueprint.

The research team aimed to address this particular feature that has limited the use of DNA in a broader context. Their innovations to license DNA for free use inside the cell were achieved by repurposing a unique bacterial DNA synthesis system called Retron.

Typically, DNA multiplies by directly copying existing DNA templates inside the cell. However, the retron system employs reverse transcription, to synthesize new DNA by reading an intermediate genetic material called RNA.

More importantly, the retron DNA created in this manner can show remarkable stability and independence from other genomic DNA in the cell. In essence, the blueprint of the cell can now go around and do groundwork in the factory rather than staying in the cabinet.

Turning retron systems into cell tools

By carefully engineering the retron system, the research team succeeded in directly generating DNA fragments with programmable functions inside the cell. These DNA fragments bind to specific proteins and modulate cellular behavior without destabilizing the cell’s genetic information.

Based on this technology, the research team demonstrated three synthetic biological applications:

• regulating specific gene expression by utilizing DNA as a bait to attract proteins
• instantaneously controlling the localization and functionality of proteins within the cell by detecting specific signals
• semi-permanently recording molecular events for brief exposure to input signals

Now, DNA has become a field agent that can follow orders, change its location, and perform actions such as recording of molecular events for transient signals.

Potential in medicine and environment

This novel platform technology has far-reaching implications beyond the state-of-the-art DNA-based circuit designs. The ability to capture and record transient disease markers in real time—such as for cancer or inflammation—provides the framework to develop smart biotherapeutics with autonomous control and feedback regulation for therapeutic regimen.

The engineered living biosensors can also be deployed to detect pollutants like microplastics or heavy metals in the environment.

Graduate student Lee, who led the study, highlighted the contribution to the field, stating, “We have provided the necessary framework to open up a whole new design space that unfetters DNA from its role as genetic material.”

Professor Kim added, “We now have access to a foundational technology that can potentially be used to revolutionize multiple application areas, including medicine, the environment, and energy.”

Adelaide University is leading the international Wheat Spatial Omics Consortium (WSOC) of more than 30 institutions in nine countries, which will explore how collaborative research in spatial omics technologies could improve wheat performance for growers.

Spatial omics is a suite of molecular technologies that measure and map the distribution of genes, proteins, and metabolites while preserving their native spatial context and cellular organization, which conventional omics cannot achieve.

“Spatial transcriptomics allows us to measure the abundance of genes in specific cell types and time-points that are responsible for yield, pest and disease resistance, and abiotic stress tolerance of crops,” said Professor Zhong-Hua Chen, from Adelaide University’s School of Agriculture, Food and Wine and Waite Research Institute.

“Although widely applied in medical and animal sciences, the use of spatial omics in crops with large complex genomes, such as allohexaploid wheat, remains limited.

“Among our consortium members, we have the technological power to tackle this complexity across multiple tissues and stages of development.”

Professor Chen and the WSOC collaborators published a paper in Nature Genetics detailing the potential for the application of spatial omics in wheat.

“Our ambition is to build a comprehensive spatial omics atlas to benefit the whole wheat community. By mapping wheat biology at subcellular resolution across the full life cycle, the WSOC seeks to decode the integrated mechanisms of wheat development, stress response, and grain quality,” he said.

“We are producing this atlas in parallel with a set of research questions led by different groups that will allow us to tackle important issues in wheat breeding such as leaf rust resistance, root drought tolerance, and high grain quality for bread-making.

“This is an enormous task, but wheat is a globally critical crop, so improving grain yield and quality has real-world impacts for people experiencing food insecurity and malnutrition.”

Australian wheat exports are valued at more than AUD$9 billion annually, and exports from the 10 leading wheat-exporting countries are valued at more than USD$60 billion.

Professor Matthew Tucker, Director of the Waite Research Institute, said that spatial omics is a game-changing technology that revolutionizes the way research is carried out.

“Technological advances of this nature don’t come along very often. We are very excited to be leading this consortium and building on the legacy of wheat research at Adelaide University,” he said.

“The omics atlas will provide opportunities to narrow down the basis for important heat-sensitive traits, such as flower fertility or grain quality, and understand which cell types are responsible for tolerance.”

Professor Jason Able, Dean of School of Agriculture, Food and Wine, said research outputs generated from this technology will contribute to the way wheat breeders consider building their next step-change variety.

“This research will enable global wheat breeders to unlock the interplay and complexity of plant neural networks and how genes respond to various biotic, abiotic, environmental and climatic factors,” he said.

“Ultimately, tapping into this knowledge will create the wheat varieties of tomorrow and value-add significantly across the industry, thereby contributing to the profitability of this commodity.”

A plant signaling gene has been identified as a promising target for breeding cereal crops to produce a steeper, narrower root system architecture, but with associated yield penalties in barley. University of Queensland Ph.D. candidate Richard Dixon said collaborative research with scientists at the Australian National University revealed the gene, known as CEPR1, has a conserved function across multiple grain crops.

The research was published in the Journal of Experimental Botany.

“Our goal is to use biotechnology to create a ‘steep, deep and cheap’ plant with a root system that can access water and nutrients in challenging conditions, without significant yield trade-offs,” Mr. Dixon said.

“For millennia, plant breeders have been purely focusing on above-ground traits because, until recently, everything underground has been challenging to view.

“We think there is a huge amount of genetic variation that has been lost because we’ve been selecting only for above-ground traits.”

“Previous work by our group using the plant Arabidopsis, a kind of lab rat for plant genetic research, showed CEPR1 controls the shape of the root system and some above-ground traits like seed production,” ANU researcher and co-author Dr. Michael Taleski said.

“In this research, we engineered Arabidopsis plants to have versions of CEPR1 found in barley, rice and maize to test if they function similarly.

“It has been proposed that root system shapes could be tailored to different cropping scenarios and environmental conditions to enable more efficient resource capture.

“That would ultimately mean reduced fertilizer costs for farmers, less fertilizer runoff into the environment, and better performance of crops under water limitation.

“Our findings are exciting because they suggest the CEPR1 genetic pathway is a promising target for optimizing the root systems of these crops.”

Mr. Dixon said that while the initial work with the gene was promising, there are limitations that will have to be addressed through further research.

“In barley, knocking out CEPR1 resulted in yield penalties and steeper, narrower root systems, just like the Arabidopsis mutants.

“We’ve also been growing and harvesting the gene-edited plants in the field to validate the findings in the glasshouse and are in the process of analyzing the data.

“We would like to fine-tune the pathway rather than switching off the gene to create root systems that can reach deeper water or nutrients without affecting grain production.

“We’re also looking at whether combining the CEPR1 gene with another target could create deeper root systems.

“And we’re utilizing the high-tech root scanning facilities in Germany through UQ’s International Research Training Group (IRTG) to further investigate the plants’ performance under drought and nutrient stress conditions.

“Our results are encouraging because it allows us to think of ways to use this tool to fix problems.”

It was part of an ARC Linkage project in collaboration with InterGrain.

New research has discovered that the molecular machines responsible for copying our DNA have a surprising hidden talent—an ability to create entirely new and highly sophisticated DNA sequences from scratch. The study, led by the University of Bristol, analyzes this curious “doodling” activity, showing for the first time that it can be steered and controlled. The findings not only help shed further light on how genetic information emerges, but could also present exciting new ways of writing long DNA sequences.

“Analysis and control of untemplated DNA polymerase activity for guided synthesis of kilobase-scale DNA sequences” is published in Nature Communications.

Every time a cell divides, it needs to copy its DNA. This job falls to proteins called DNA polymerases—tiny biological machines that read an existing DNA strand and build a matching copy, letter by letter, essentially acting as nature’s photocopiers.

It has been known, since the 1960s, that some of these machines can also build new DNA without anything to copy from, in a process scientists nicknamed “doodling.” Until now, the sequences produced by doodling have been poorly characterized and this study provides the most detailed assessment to date.

Co-lead author Simeon Castle, who conducted the research as part of his Ph.D. in Engineering Biology at the University of Bristol School of Biological Sciences, said, “We used nanopore sequencing to read the full-length sequences of thousands of DNA molecules that polymerases had created entirely on their own.

“What we found was far more diverse and complex than anyone had appreciated—from simple two-base repeats to elaborate eight-base motifs, all varying depending on which polymerase was used and the reaction conditions.”

Current methods for writing DNA rely on slow chemical processes and struggle to produce sequences longer than a few hundred bases (a base being the single letters from which DNA is built). By contrast, doodling can generate much longer fragments in a single reaction, with some exceeding 85,000 bases.

Co-lead author Thea Irvine, a Ph.D. student in Engineering Biology also at the University’s School of Biological Sciences, added, “One of the most exciting findings was that we could actually steer what the polymerases produced. By changing the temperature or limiting which DNA building blocks were available, we could shift the composition of the sequences generated.

“When we provided only two of the four building blocks present in DNA, the polymerase produced long stretches of highly regular repeating patterns—some over a thousand bases in length.”

The research united multidisciplinary experts from the University of Bristol, University of St. Andrews, and the Medical Research Council (MRC) Laboratory of Molecular Biology in the UK, and The Center of Excellence for Engineering Biology in New York and Replay Holdings Inc. in the U.S.

Senior author Thomas Gorochowski, Professor of Biological Engineering and a Royal Society University Research Fellow at the University of Bristol, added, “Doodling by DNA polymerases has been known about for decades, but has largely been treated as a curiosity. Our work shows it is a tunable process with implications for how new genetic material is created and a real potential for biotechnology.

“Combining our findings with advances in AI-powered protein design, we believe harnessing doodling for the guided synthesis of long DNA sequences could be closer than many think.”