Crop yields and water efficiency could be improved with the use of better gene-editing techniques. Credit: AFP/Getty

When crop engineers from around the world gathered in London in late October, their research goals were ambitious: to make rice that uses water more efficiently, cereals that need less fertilizer and uberproductive cassava powered by turbocharged photosynthesis.

The 150 attendees of the Crop Engineering Consortium Workshop were awash with ideas and brimming with molecular gadgets. Thanks to advances in synthetic biology and automation, several projects boasted more than 1,000 engineered genes and other molecular tools, ready to test in a researcher’s crop of choice. But that is where they often hit a wall. Outdated methods for generating plants with customized genomes — a process called transformation — are cumbersome, unreliable and time-consuming.

Asked what hurdles remain for the field, plant developmental biologist Giles Oldroyd of the John Innes Centre in Norwich, UK, had a ready answer: “The big thing would be to improve plant transformation,” he said.

“What we’re all facing is this delivery problem,” says Dan Voytas, a plant biologist at the University of Minnesota in Saint Paul. “We have powerful reagents, but how do you get them into the cells?”

At issue is the decades-old problem that it is difficult to modify plant genomes and then regenerate a whole plant from a few transformed cells. Genome-editing techniques such as CRISPR–Cas9 hold out the promise of sophisticated crop engineering that would once have been unthinkable — making it all the more frustrating when researchers run up against an old roadblock.

On 28 September, the US National Science Foundation (NSF) recognized this frustration by announcing that it would fund research into better transformation methods. That focus is one of four in a new plant-genome research programme that will receive a total of US$15 million.

“Everybody agrees that it really is the bottle­neck for genome engineering,” says Neal Stewart, a plant biologist at the University of Tennessee, Knoxville, who co-organized an NSF workshop about plant transformation last November. “And I think there’s enough interest now in trying to come up with ways to fix the problem for major crops.”

Obstinate crops

Some plants, such as the diminutive thale cress (Arabidopsis thaliana), the ‘lab rat’ of plants, are easily transformed using a bacterium that can add genes to plant genomes. Researchers insert the genes they want to test into the bacterium (Agrobacterium tumefaciens), and then coax the microbe to infect the reproductive cells of the plant. When the plant then produces offspring, some of them express the new genes.

But this does not work for many crops, and use of Agrobacterium triggers extra scrutiny from government agencies such as the US Department of Agriculture because it is considered a plant pest. As an alternative, researchers can use ‘gene guns’ that fire DNA-coated gold beads into plant cells. Those cells are then bathed in growth hormones and coaxed to regenerate a full plant. Some plants, such as maize (corn), readily bend to this treatment. Others, such as wheat and sorghum, do not.

For recalcitrant crops, it can take months of painstaking cell-culture work — optimizing growth conditions and hormone concentrations — to regenerate the full plant. The conditions needed for success vary not only from crop to crop, but also between plants of the same species.

Plant-transformation experts are a rare breed, says Joyce van Eck, one such specialist at Cornell University in Ithaca, New York. “There’s a lot of art in what we do,” she said at the London workshop. “It’s difficult to find people with that training.”

Add to that a dearth of funding for new methods, and researchers are left having to rely on decades-old techniques.

A better way

But that could change as the hunt for alternatives heats up. Stewart and his collaborators have developed a robot that performs an established technique called protoplast transformation faster and more accurately than is possible by hand. The method uses enzymes to digest the cell wall, making it easier for researchers to introduce new genes. The problem of regenerating the whole plant, however, remains. Researchers used a similar approach, without robots, to perform CRISPR–Cas9 gene editing in a variety of plants, including lettuce and rice.

The cell-culture steps are still difficult. Stewart says that one person in his lab laboured unsuccessfully for two years to transform a tall grass that he uses for biofuel research. But the declining cost of enzymes allows researchers to perform more experiments, and the robotics improve throughput. Stewart is so enamoured with his creation that he has composed a song for it. “It’s our baby right now,” he says.

Others, such as Fredy Altpeter of the University of Florida in Gainesville, are hunting for a suite of genes that, when switched on or off, would make plant cells more amenable to transformation and regeneration from culture. “I think it will lead to much broader application of this technology, and will enable people who are not experts in cell culture to make those improvements,” he says.

But researchers can’t afford to wait for those developments, says Oldroyd. His project, which aims to develop cereals that use nitrogen from the soil more efficiently, will plough through tests of hundreds of transgenes using the old, cumbersome methods. “We just have to be patient,” he says.