Supercrops: Fixing the flaws in photosynthesis

Many vital crops capture the sun’s energy in a surprisingly inefficient way. A borrowed trick or two could make them far more productive

Take a look around you. All the organic things you see, from your hands to the leather of your shoes to the wood in your table, are built of strings of carbon atoms. So too is the petrol in your car and the coal in your local power station. All this carbon came from thin air, from carbon dioxide in the atmosphere.

From termites to blue whales, virtually all life on Earth depends on plants’ ability to turn sunlight and carbon dioxide into food – and without the waste product, oxygen, you would be dead in minutes. You would think, then, that evolution would have honed the process to perfection over the past 2 billion years or so. Yet photosynthesis remains astonishingly inefficient in some ways.

In one respect, this is very good news. With the world’s population set to soar to 9 billion or so by 2050, we need to grow a lot more food. That’s a huge challenge, and as the climate changes, higher temperatures and more severe droughts and floods will make it even harder. We got a glimpse of what the future might be like this year, with drought in Russia and floods in Pakistan devastating crops.

Improving photosynthesis itself could dramatically boost yields. The idea might sound like hubris, but there is no doubt it can be done – because some plants have already achieved it. Many have evolved the same workaround for the most serious glitch in photosynthesis, and researchers are now trying to copy that advance in wheat and rice. And some smaller improvements have already been made in the lab.

Surprisingly, one approach might be to reduce levels of the molecule that captures light energy from the sun. It turns out that plants make up to four times as much chlorophyll as they need for photosynthesis. Why? Because if plants produced only as much chlorophyll as they needed for themselves, more light would be able to pass through their leaves and reach upstart seedlings nearby. “Plants didn’t evolve to optimise their yield,” says Don Ort of the University of Illinois at Urbana-Champaign. “They evolved to survive.”

Costly defences

Plants with excess chlorophyll don’t just shade out rivals, though. They also shade their own lower leaves. What’s more, the excess chlorophyll may cause some leaves to absorb light energy faster than it can be used, which can damage them. So plants have evolved a “quenching system” to mop up the surplus energy. All this is costly for the plant, which is one big reason crops yield less food per light unit than they otherwise might. A soybean mutant with half the usual level of chlorophyll can produce 30 per cent more biomass than normal.

More generally, the amount of food a plant can produce depends not only on the amount of light it intercepts but also on how efficiently it traps the energy of that light in chemical bonds within sugars and starches, and on the proportion of those sugars and starches that end up in edible parts of the plant. The green revolution of the 1960s came from shifting the last factor, by producing short-stemmed wheat and rice varieties that put half their photosynthetic output into seeds. There is little scope for pushing this further: modern varieties can divert no more energy from stems to seeds and still stand until harvest.

As for the first factor, modern crop varieties are bred for large, fast-developing, optimally arranged leaves. They already intercept most light at the right wavelengths during the growing season, Ort says. In theory, plants could capture a wider range of wavelengths (see “Why aren’t plants black?”), but such a fundamental redesign is beyond today’s bioengineers.

That leaves more efficient photosynthesis. Here the scope for improvement seems enormous. Plants capture CO₂ using an enzyme called rubisco, and it is the most lackadaisical enzyme we know of. Whereas most enzymes catalyse thousands of reactions per second, rubisco manages only a few – although to be fair, it catalyses reactions involving gas molecules, which are harder to grab hold of than larger molecules in solution. To compensate for its sloth, plants have to produce huge quantities of rubisco, making it the most abundant protein on the planet.

Slowness is not the only problem: rubisco is error-prone, too. The intricate shuffling of organic molecules that photosynthesis uses to make sugars begins with rubisco grabbing a molecule of CO₂. Then, using the energy captured by chlorophyll, rubisco adds the CO₂ to a molecule containing five carbon atoms, yielding two 3-carbon molecules. Sometimes, though, rubisco grabs hold of oxygen instead of CO₂, which results in an unwanted 2-carbon molecule (see diagram). The error is costly to fix, as it not only wastes energy but also leads to the net loss of one molecule of CO₂.

This was not a problem a couple of billion years ago when rubisco evolved, as there was no free oxygen around. Now there is lots of oxygen and less CO₂, which means rubisco is far more likely to grab oxygen by mistake, triggering the wasteful process known as photorespiration. In theory, the energy from just eight photons should be enough to capture a CO₂ molecule, but in practice plants typically need 13 photons because of all the energy lost during photorespiration.

The most obvious way to improve photosynthesis, then, is to tweak rubisco to make it work faster or less likely to grab oxygen. But this is easier said than done. “Evolution had several billion years to change that and didn’t,” points out Julian Hibberd of the University of Cambridge. When biologists tried making rubisco more selective, they found the enzyme got even slower.

In fact, it is far from clear whether this “dreadful enzyme”, as Ort describes it, can be improved. Some argue that rubisco is already as good as it can be. If so, perhaps a better way of capturing CO₂ can be designed from scratch, but again, this is beyond today’s bioengineers.

Fortunately, plants have already found ways to compensate for rubisco’s weaknesses. Over the past 30 million years, as CO₂ levels fell to their lowest for at least 300 million years, some plants evolved a neat solution: capturing and concentrating the gas to recreate the ancient atmosphere. In these plants, photosynthesis usually takes place in cells clustered around the leaf veins. They are surrounded by a layer of cells that, instead of containing rubisco, contain another enzyme that grabs CO₂ molecules and attaches them to a 3-carbon sugar. The resulting 4-carbon molecule is then transported into the rubisco-containing cells. There the CO₂ is released.

Energy saver

Although concentrating CO₂ in this way takes a fair bit of energy, it saves energy overall by reducing the amount wasted on photorespiration. So the process, called C4 photosynthesis after the 4-carbon molecule that transports CO₂, is far more efficient than normal or “C3″ photosynthesis. C3 crops like wheat and rice typically produce 11 tonnes of grain per hectare, while the main C4 crop, maize, can yield more than 18 tonnes.

The extra CO₂ we are pumping into the atmosphere would, if the climate remained unchanged, reduce photorespiration and boost growth in C3 plants. Unfortunately, global warming will counteract this effect, as higher temperatures increase photorespiration. What’s more, in hot conditions plants shut their pores to reduce water loss. This also results in less CO₂ entering the leaves, boosting photorespiration further.

Wheat and rice yields are already falling in some areas as the temperature rises, and this is just the start. Models point to wheat yields falling 25 to 35 per cent in coming decades, says Hans Braun, chief wheat breeder at the International Maize and Wheat Improvement Center (CIMMYT) in Mexico.

By contrast, higher temperatures make little difference to C4 plants. Because C4 plants have an internal store of CO₂, they can keep photosynthesising even when their leaf pores are closed, so they are better at conserving water. And because they need less nitrogen-containing rubisco, they need less nitrogen fertiliser than C3 plants. This is why, although C4 plants only evolved recently, they already dominate environments like grasslands.

“Just a few million years after they evolved, C4 plants dominate environments like grasslands”

This long list of advantages spurred CIMMYT to launch a consortium of researchers last November, with the aim of switching wheat from C3 to C4 photosynthesis. But no one expects it to be easy. Just adding a few enzymes to C3 plants won’t do the trick. This has already been tried in rice and, while it looked promising in the lab, the plants reverted to C3 photosynthesis in the field, says Rowan Sage of the University of Toronto, Canada. For C4 photosynthesis to work, he says, rubisco has to be kept in a separate “compartment”. In other words, converting plants from C3 to C4 requires not just the enzymes to capture and concentrate CO₂, but also the special arrangement of cells found in C4 plants, known as Kranz anatomy. That’s a tall order, given that how structures develop is among the least well-understood things in biology.

As for which crops to work on, Matthew Reynolds, chief wheat physiologist at CIMMYT, thinks wheat might be easier to switch to C4 than rice. That’s because most wheat strains have six copies of their genome rather than the usual two. “We can move genes around without being scared we will delete something vital,” he says.

Rice researchers are not giving up, though. In 2008 the International Rice Research Institute (IRRI) in Manila, the Philippines, launched a new drive to try to engineer C4 rice, funded by the Bill and Melinda Gates Foundation. The IRRI is bombarding sorghum, a C4 grain, with gamma rays to see whether any of the resulting mutants partially revert to C3 photosynthesis, and, if so, what genetic changes are responsible, says Susanne von Caemmerer of the Australian National University in Canberra, a member of a consortium set up by the IRRI. The institute is also screening rice strains and their wild relatives to see if any have C4 characteristics. “Conversion of C3 to C4 has become realistic,” von Caemmerer says, “because of the new technologies for massive sequencing that make possible research we previously could not attempt.”

Such methods are revealing the scale of the challenge. Hibberd and colleagues have been studying two closely related plants, one C3 and one C4. They reported in June that the plants express 603 genes differently, including 17 that regulate other genes. “C3 to C4 conversion is incredibly ambitious, but we’re pursuing it because the rewards are so great,” Hibberd says. His lab and others are already adding C4 sequences to rice plant embryos and sending them to IRRI to grow and assess.

They take heart from the fact that plants evolved C4 independently on at least 45 occasions. “So this may not be a very difficult switch for plants to make,” says Hibberd. In fact, Eleocharis vivipara, a sedge grown in freshwater aquariums, uses C3 photosynthesis when under water, and C4 – complete with Kranz anatomy – on land.

There are other, possibly easier, ways to reduce the bottlenecks in photosynthesis. “It would be silly not to try the simple things as well as C4,” says Christine Raines of the University of Essex, UK. “There are some points in the cycle where increasing efficiency has boosted yield, although so far not under field conditions.”

For instance, some photosynthetic bacteria have evolved a less wasteful way of recycling the 2-carbon compounds produced by photorespiration. Adding this pathway to plants results in faster growth and more biomass (Nature Biotechnology, vol 25, p 593).

One relatively simple tweak would be to stick a rubisco from a C4 plant into a C3 crop, rather than trying to convert it to C4. This could boost photosynthetic efficiency by 25 per cent, Ort thinks. The reason is that while evolution has not found a way to stop rubisco binding to oxygen, it has produced the best balance between selectivity and speed. The rubiscos in C3 plants are optimised for the very low CO₂ levels that have prevailed for the past million years or more. Though levels will soon more than double, it will take evolution a long time to catch up. The rubiscos inside C4 plants, however, are optimised for higher CO₂ levels; they bind CO₂ a little less strongly and hence work a bit faster.

Problem: no one has yet managed to get a plant to make a foreign rubisco. The enzyme consists of eight large proteins coded for by the chloroplast’s genome, and eight small proteins coded for by genes in the cell nucleus. All these proteins have to be put together by special chaperone proteins. It’s all quite a challenge, but Ort is optimistic that decent funding will overcome it. “That’s a technical problem: $50 million over five years will buy that,” says Ort. “C4 conversion needs basic research – we don’t know how to buy it.”

“We should do the other improvements to photosynthesis too,” counters Hibberd. “But in 2050 we have to produce a lot more food. The only biological precedent for a major step change in productivity is the evolution of C4. We’ll be lucky to do it in 15 or 20 years. But if we don’t start trying now, we won’t do it at all.”

“There is huge scope for improving photosynthesis. If we don’t start trying now, we won’t do it at all”

More on these topics:

• biology
• botany
• food
• Plank