Super rice for a hungry planet
As the population hurtles toward nine billion, a team of plant engineers is on a ‘mission impossible’ to retrofit the turbocharged engine of corn into the world’s top staple crop. Elizabeth Finkel investigates.
You wouldn’t pick Robert Furbank as a plant scientist of international renown. A dark-haired, pleasant-looking man in his mid 50s, Furbank has more the manner of a high school football coach: blokey and approachable, an understated determination and an easy way with words, which he occasionally needs.
The US$25 million project he helps oversee – largely funded by the Bill & Melinda Gates Foundation – plans to make a kind of super rice: tall, hardy and as bountiful as corn. In the same jaunty manner a coach might tell you how to kick the ball, you realise (with surprise) that Furbank is describing the intricacies of how to re-engineer a rice plant. Call it agriculture’s “Mission Impossible”, but the food crisis has led this international team to try to re-engineer the very engine upon which all life is based. Furbank leads his node of operations from the dauntingly named “High Resolution Plant Phenomics Centre” at CSIRO Plant Industry in Canberra, Australia.
Thomas Malthus foresaw the problem more than 200 years ago: human populations grow geometrically; food production does not. Therefore, at some point human populations must starve. In the mid-1960s, ecologist Paul Ehrlich saw it happening in front of his eyes on the streets of Calcutta. His book The Population Bomb, published in 1968, predicted famine on a global scale. “The battle to feed all of humanity is over,” he wrote.
It didn’t happen. India went from being a “ship-to-mouth” country to a net exporter of wheat. In large part, this was a result of the genetic tweaking of wheat to produce a super-breed: disease-resistant and versatile enough to grow anywhere, with double the harvest yield. Never before, or since, has there been anything like it.
The course of history was changed in this case by a single man, wheat breeder Norman Borlaug. He had been sent to Mexico in the 1940s by the philanthropic Rockefeller Foundation to help battle devastating wheat diseases that were also finding their way through North America. Borlaug’s efforts paid off 14 years later with a doubling and then tripling of the Mexican wheat harvest. Pakistan and India needed help too.
Borlaug delivered the super-seeds and training, but his bullish personality also helped push aside the policy hurdles to double the harvest in India and Pakistan. His strategies were also adopted by rice growers, with the same fabulous gains. It was this doubling of wheat and rice yields across Asia and Latin America in the 1960s and 1970s that became known as the “Green Revolution”. Not only did it avert mass famine, the extra wheat and rice was mostly grown on existing agricultural land, sparing forests and grasslands. Borlaug won the Nobel peace prize and fame as “the man who fed the world”.
The world needs to double grain production by 2050
to feed an estimated nine billion people.
There is no doubt that Borlaug’s superwheat was a remarkable achievement. But at a time when DNA was not even part of the vocabulary, how did he manage it? The answer was something breeders call “feel” – you might call Borlaug a “wheat whisperer”. He knew the plant so intimately he could diagnose its inner workings from the subtlest of outward signs – pustules and flecks on the leaves, black smudges on the seed coats, leaves with white curling tips. Like battle scars, these signified a plant that had fought and won against particular diseases. By shuffling these traits until they came together like a perfect poker set, Borlaug managed to breed bulletproof wheat.
Borlaug also knew how to sculpt wheat. Tall and spindly, traditional Mexican wheat keeled over once the ears were fattened by fertiliser and irrigation. But the Japanese had bred a dwarf variety called Norin 10. Borlaug crossed it with the Mexican wheat to produce a short, stocky offspring. An unanticipated bonus was that resources, spared from stem and foliage, were spent on the ears, which grew even fatter.
Over the years other breeders built on the strategy. Wheat and rice varieties became shorter and rice was bred to sport ever more ears, but the plant’s architecture could take only so much tampering. Although yields kept rising by about 2% a year through the 1970s and 1980s, since 1995 they have flattened out. For wheat they are now about 0.5% per year, for rice 0.9%. And that’s not enough. According to the Food and Agriculture Organization, the world needs to nearly double grain production by 2050 if it is to feed an expected population of nine billion, many with middle-class appetites for meat and dairy products.
That statistic has set alarm bells ringing for the international agencies charged with wheat and rice breeding, the philosophical scions of Borlaug’s breeding centre – the International Maize and Wheat Improvement Center, (CIMMYT) and the International Rice Research Institute (IRRI). If yields are to double, a bold new breeding strategy is needed. Norman Borlaug redesigned the plants’ chassis; this new generation of plant tinkerers is tackling their very engines.
A billion and a half years ago, when our planet was a baby, purple-green bacteria built themselves a remarkable engine to store the sun’s energy as sugar. Like most engines, photosynthesis has multiple components. The front end resembles a solar panel. Sunlight splits water into electrons and protons, with oxygen as a by-product. But the main business lies with the electrons and protons that drive production of the chemicals adenosine triphosphate (ATP) and nicotinamide adenine dinucleotide phosphate (NADPH). As with a battery, these energetic chemicals offer a great way of temporarily storing energy (temporary is the operative word; ATP only lasts about 50 milliseconds in a cell). That energy powers the back end of the engine, which houses rubisco, an enzyme that condenses CO2 molecules into the king of energy-storage molecules: the dense, stable six-carbon molecule we know as sugar.
Rubisco’s efficiency is governed by the concentration of CO2. Most plant species deliver it at the same concentration as in the surrounding air – but that is not what rubisco was designed for. When purple-green bacteria first evolved the enzyme, CO2 levels were 100 times higher than they are today. On this CO2 rich diet, photosynthetic life went from strength to strength, evolving from purple-green bacteria to algae, to mosses, ferns, cone-bearing plants and, ultimately, flowering species.
They did so well at sucking CO2 out of the atmosphere and knitting it into sugars that levels gradually fell. Then, 30 to 40 million years ago, the gradual fall became a sudden crash from about 1,000 parts per million (ppm) to 300ppm (lower than the current level of about 400ppm). So today, for most plants, rubisco is functioning below peak efficiency.
Most plants, but not all.
At least 62 species have evolved a technique of turbocharging. They concentrate CO2 tenfold before they pump it into the rubisco engine. The key to doing that is to build up a reservoir of CO2 condensed in the form of a C4 carbon chain. These turbocharged plants, dubbed C4 plants, are quite obvious today. Think corn as high as an elephant’s eye, sugar cane or bulrush millet: 12 weeks and you need a ladder to see over the top. They can produce up to 100 tonnes of biomass per hectare – 10 times that of C3 wheat. And they can do it on relatively poor soils, using half as much nitrogen and water. It’s no wonder scientists have turned to C4 plants to solve the food crisis.
Furbank’s blokey vernacular comes from his origins in the steelworks town of Wollongong, NSW, which is also the source of his mechanical flair. He grew up pulling engines apart. The son of a steelworker, Furbank was the first of his family to go to university. He might have done law, except he’d been totally captivated by photosynthesis. “I remember a physics high school teacher describing it and thinking, ‘This underpins all life on Earth’.” Maybe it also seemed like another engine to pull apart.
In his first two decades as a scientist, Furbank pursued his passion for photosynthesis with gusto at CSIRO Plant Industry in Canberra, the University of Sheffield, UK, and then back to CSIRO. Joined by his wife Julie Chitty, the pair churned out vital publications on the working of the C4 engine – they were the stars of their field. But, by the mid-1990s the world, and in particular CSIRO Plant Industry, had lost interest. CSIRO’s remit was to pursue problems of national significance and, as far as Australia and the world were concerned, there was no reason to tinker with the photosynthetic engine of crop plants. The Green Revolution was still providing great dividends, Europe had mountains of stored grain, and globally food was cheap.
Furbank was told to find something else to work on. He swallowed his disappointment and then, the inveterate tinkerer, latched on to solving an age-old problem. Plant breeders need to shuffle traits. Some, like a shorter plant or a plant with fatter seed, are obvious. But others are invisible, such as the one that allows a plant to take more carbon dioxide from the air. Borlaug’s generation relied on “feel” to make a best guess as to what was going on inside a plant. Furbank thought today’s breeders could do better. He started imagining machines that would reveal the inner workings of plants, like the Tricorder from Star Trek. Then he assembled a team of young guns – plant scientists, engineers and IT nerds – and started building them.
It was time for a bold strategy – to try to turbocharge
the rice plant’s photosynthetic engine.
An inflection point came in 2008. Like the biblical Pharaoh emerging from seven years of plenty to ponder his emptying granaries, the world was staring down the barrel of a food crisis. It was due to a combination of factors: droughts in Australia, floods in Asia and the displacement of food crops for biofuels. Global food reserves shrank to 30 days and prices soared. Rice went from US$400 per tonne in January 2008 to more than $1,000 per tonne in May that year, meaning a billion people could afford to eat only once a day.
Food riots broke out in dozens of countries and, 5,000 years after Pharaoh, bread shortages were once again causing havoc in Egypt. In April 2008 they helped trigger a revolution.
For John Sheehy, head of rice physiology at the IRRI in the Philippines, these events were not just of academic interest. Globally funded, IRRI’s remit is to breed rice varieties to meet the diverse needs of farmers: to resist pests, survive droughts and floods, and cope with poor soils. But above all, the institute is charged with keeping yields rising faster than the rate of population growth, then at 1.2%. With rice yields increasing at 0.9% per year, it was clear they were losing the race.
Sheehy decided it was time for a bold strategy – to try to turbocharge the rice plant’s photosynthetic engine. He started assembling a team of engineers. It wasn’t hard. They were a small, unfashionable group of academics from the UK, Germany, Canada, U.S. and Australia who all knew each other well – the diehard experts in C4 photosynthesis. Furbank was one of them. By then, he was also director of the High Resolution Plant Phenomics Centre, a full-blown realisation of his dream to be able to peer inside the inner workings of plants. Together they formed the International C4 Rice Consortium and convinced the Gates foundation to provide US$12 million funding.
The task of re-engineering the photosynthetic engine of the rice plant might sound impossible. But it’s not. Nature has done it by chance at least 62 times in the past 30 million years. Surely plant engineers could do it faster. In the age of genetic engineering, one straightforward approach is to transfer the 12 or so “turbocharging” genes that the corn plant evolved for this purpose into a rice plant. But it’s not that simple. In 2000, Maurice Ku at Washington State University had tried but didn’t get far. Part of the reason is you cannot just dump the components of the C4 engine into rice – it has the wrong infrastructure. Imagine trying to fit a Ferrari engine into a VW. For starters the Ferrari engine won’t fit, and the VW is designed to operate with different suspension, engine mountings and exhaust system. Same problem with rice: the engineers have to redesign the plant’s infrastructure, a tough order.
Corn and rice plants have totally different leaf anatomies. Both carry mesophyll (M) cells and bundle sheath (BS) cells, which sheathe the veins, but the arrangement of the cells and the jobs they do are completely different. Rice locates all its engine components in the M cell. The BS cells seem to do little more than provide structural support for the veins. In corn, it’s a very different story. The major engine component – sugar-churning rubisco – is located exclusively in the BS cells, while the M cells have been redesigned to provide C4 chains for the BS cells. Once the C4 chains arrive in the BS cells, they are unpicked to release the CO2 to turbocharge rubisco.
The BS cell is gas-tight, allowing CO2 levels to rise 10 times higher than in the atmosphere. Because M and vein-hugging BS cells are so tightly coupled functionally, they need to be right next other. That explains why the veins in corn are more narrowly spaced than those in a rice leaf. Looking down a microscope at a corn leaf section, you see a distinct so-called “Kranz anatomy”: a regular repeating pattern of two M cells, and on either side of them a rosette of BS cells around the vein.
The corn plant’s split engine is a beautiful piece of engineering, but how are the plant engineers supposed to reproduce it in a rice plant? As Furbank put it, “this was a tough nut to crack”. The engineers circled it from all sides – looking for a way in. If Mother Nature could do it so easily, maybe she would lend a hand. Perhaps there are rice mutants that have already taken their first steps to becoming C4 plants? To find them, you might only need to look for something visible – veins a little closer together, for example.
Technicians at IRRI’s centre of operations in Los Baños, the Philippines, did just that using hand-held digital microscopes. They identified several rice varieties with narrower vein spacing, but so far the strategy has not borne fruit. Some mutants had more closely spaced veins because the cells themselves were shrunken. The leaves were also narrower. None of these characteristics boded well for breeding the Ferrari plant they were after.
Jane Langdale finally managed to isolate the gene that directed chloroplasts to develop in the bundle sheath cells – it was called Golden2.
The most promising leads have come from reading the parts lists of the Ferrari-type and the VW-type plants; in other words, their genomes. Scanning the parts lists of genes, perhaps the researchers could pick the difference. Which genes had nature recruited to the task of redesigning the plant’s anatomy? Of course that wasn’t easy either. There were thousands of genes that differed between corn and rice: this was more like looking for a needle in a haystack. But the plant engineers were crafty.
It seems a corn plant doesn’t always behave like a corn plant. Early in its development, the base of the leaf does not have the Kranz anatomy – its veins are wide apart. As the leaf grows, the Kranz anatomy takes over. Then there is the matter of the papery leaves that sheathe the corn husk. They too are a throwback, with widely spaced veins.
These days plant engineers can do extraordinary things. They can for instance take slices of leaf tissue from a developing leaf or from a husk leaf, mash them up, and look at which genes are active. In this way, they could compare the leaf slices with Kranz anatomy to those without.
That approach bore fruit. Too much fruit. There were hundreds of genes that appeared to form the signature for Kranz anatomy in the mature leaf that were not turned on in the husk or baby leaf. But among the candidates, two caught the eye of plant geneticist Jane Langdale at the University of Oxford.
Langdale has a longstanding dedication to unravelling the secret of C4 plants. Back in the late 1980s, using the time-honoured strategy of “throwing a spanner in the works”, she screened corn mutants to see if any had problems with the C4 machinery. She found a mutant with pale green leaves. Like a rice plant, its bundle sheath cells had stopped performing photosynthesis. They contained no green chloroplasts – the organelles where photosynthesis takes place.
By 1998, she finally managed to isolate the gene that directed chloroplasts to develop in the bundle sheath cells – it was called Golden2. It turns out that in rice, the closely related “Golden2-like” gene is switched “off” in the BS cells. Langdale is now doing the obvious experiment. If she switches Golden2-like “on” in the rice leaf, will the bundle sheath cells start producing chloroplasts? If they do, that would be a step closer to the Ferrari.
The second gene that caught Langdale's eye was named Scarecrow. Last year, Thomas Slewinski at Cornell University in New York showed that if Scarecrow was switched “off” in corn, the veins of the leaf lose their regular narrow spacing.
Langdale has also shown that Scarecrow is naturally switched off in corn husk leaves. Again the next steps are obvious. Turn it “on” in rice and if the distance between the veins narrows, that could be step two towards the Ferrari.
Like new engines hoisted on to the dynamometer at Ferrari design headquarters, the prototypes are put through their paces by robots.
Meanwhile Julian Hibberd and his team at the University of Cambridge are trying a retrofit. To emulate the C4 split-engine structure, which eliminates rubisco from mesophyll cells and replaces it with turbocharger enzymes, they engineered a rice plant with mesophyll cells lacking rubisco and have started installing genes for the turbocharger.
Bit by bit, these teams are sending plant prototypes to Furbank’s High Resolution Plant Phenomics Centre in Canberra for testing. Here the engine redesign begins in earnest. Like new engines hoisted on to the dynamometer at Ferrari design headquarters, the prototypes are put through their paces by robots such as PlantScan. Thousands of plants each day trundle past to have their dimensions digitised and measured with exquisite 3D precision – down to the last leaf tendril – by the robot’s LIDAR camera. At the same time, an infrared camera overlays a heat map of each plant and a spectrometer reads the sugar content. C4 plants are warmer and grow twice as fast as regular plants, so the high throughput robot will be on the lookout for hotter, faster-growing plants, as well as plants that produce more sugar.
Elsewhere in the workshop, technicians manning specialised microscopes are peering inside leaf slices to see whether any of the promising plants have split the functions of their M and BS cells to resemble those of C4 plants. Others are clamping gadgets on the leaves to see how well they process CO2. Outside, cameras on blimps and souped-up golf buggies interrogate plants growing in the field.
All that was once embodied in the fuzzy art of “feel” is now being measured here with digital precision and fed straight into computers generating terabytes of data a day. “Digital agriculture” is what Furbank calls it.
Last year, just three years into the project, digital agriculture provided a proof of concept: souped-up rice plants that individually carry the different gene components of the corn turbocharger. “Having all the components in our rice tool box was enough to convince the Bill & Melinda Gates Foundation to fund phase two”, says Furbank. And this time, the UK government, the European Union and the Consultative Group for International Agricultural Research pitched in for a total of US$14 million.
Now the team is crossing the rice plants to get the individual components together. As of June, they had assembled four of them together in a single plant. “We’ve gone from a parts list and tool box to a partly assembled engine,” Furbank says. Having nailed Scarecrow and Golden2, he is emboldened to add, “We’re on the way now to crack the hard nut and sort out the anatomy.”
The engineers have another four years to provide the next model. For this Mission Impossible team, the outlook is optimistic. “It’s not like a tall mountain; it’s more like a bunch of small steps we have to take,” says Rowan Sage, a team member who studies plant evolution at the University of Toronto.
What is never far from Furbank’s mind is the challenge of feeding the world. What raises his confidence is the sleek supercomputer on his desk. He recalls the feeble, clunky Commodore 64 that sat there 30 years ago. “That’s the quantum of advance we need to double crop yields. Digital agriculture is going to be the future.”