The single most abundant protein on the planet isn’t actually very good at its job. And, unfortunately, its job is important: to pull carbon dioxide out of the atmosphere and incorporate it into sugars and other molecules that most of Earth’s life depends on. Improving its function could help us in a variety of ways, from boosting crop productivity to cleaning up after our carbon emissions.

Unfortunately, the enzyme is also extremely fussy about how it operates, in part as a result of the evolutionary events that put it in plants in the first place. But now, a team of German scientists has figured out how to get the enzyme to work in the standard lab bacteria, E. coli, opening the door slightly to genetically engineering our way to more efficient plants. But the work also makes it clear that things aren’t quite as simple as we’d like.

A key enzyme

The enzyme has the catchy name “ribulose-1,5-bisphosphate carboxylase/oxygenase,” but everyone knows it as “RuBisCo.” Its function in the cell is to take the carbon of carbon dioxide, obtained from the air, and link it to a five-carbon sugar. This makes a six carbon sugar, an essential part of the process of photosynthesis. But it also allows the carbon to be used in a variety of other chemical reactions inside a cell that would never work with carbon dioxide. These include creating the building blocks of DNA and proteins. Through these two functions, the enzyme is essential to most life on Earth.

Those facts, on their own, could explain why RuBisCo is thought to be the most common protein on the planet. But RuBisCo also needs to be made in huge volumes because it’s not especially efficient at grabbing carbon dioxide, and it frequently runs the reaction using oxygen instead. This low efficiency is thought to hold plants back a bit. Increasing RuBisCo’s efficiency could allow us to engineer plants that pull more carbon dioxide out of the atmosphere and do so in a way that makes them more water efficient at the same time. In a period of drought and climate concerns, that could be extremely significant.

Usually, if we want to improve the efficiency of an enzyme, the first step is to put it into the standard lab bacteria, E. coli. But that hasn’t worked with RuBisCo because it’s actually not a single protein. It’s actually a complex of 16 individual proteins: eight identical large and another eight smaller ones. And its assembly is complicated by the fact that RuBisCo does its job inside the chloroplast, a membrane-covered compartment that’s specialized for photosynthesis.

The gene for the big protein resides inside the chloroplast, which is simple. But the smaller protein is encoded in the DNA of the cell’s nucleus and made outside the chloroplast. That means it has to be shipped across a membrane in an incomplete state and assembled inside the chloroplast. The lack of this process in bacteria is thought to be behind the fact that plant versions of RuBisCo don’t work there.

Not so simple

Through a lot of previous work in plants, however, we’ve identified a number of genes that don’t make components of RuBisCo but are essential for RuBisCo to function. Some of these help proteins mature into the complex, three-dimensional shapes they need to perform their functions, while other genes don’t have an obvious role. But, to be cautious, the researchers engineered seven different plant genes so they’d be active in bacteria—nine if you count the two for RuBisCo itself.

Their plan worked. The bacteria produced functional RuBisCo. To find out which genes were responsible, the researchers then started deleting the engineered genes one at a time. It turned out that one of the genes wasn’t entirely necessary (RuBisCo production dropped, but only by about half) and another could be replaced by increased production of a bacterial gene. Still, that left five that were essential for making the enzyme work in bacteria.

The bad news is that the whole thing is extremely sensitive to the precise combination of genes used. The ones here came from a plant that’s part of the same family as cabbage and mustard (called Arabidopsis). When the authors tried swapping in the tobacco versions of the two genes for RuBisCo, only a small amount of the enzyme was produced. Presumably, if you grabbed the tobacco version of all of the other seven genes, it would work much better.

So, we have good news and bad news. We can now make RuBisCo in bacteria, which opens the door for engineering versions that work with greater efficiency. In fact, it opens the door to making the bacteria dependent on RuBisCo function in some way, which could allow them to evolve more efficient versions of RuBisCo.

The bad news? We ultimately need to put these versions back into plants if we’re going to make drought-resistant plants and carbon-sucking forests. Given how sensitive the system seems to be to its environment and the other proteins in the cell, that means we probably want to start out with the species we ultimately want to put the genes back into. In other words, if you want to engineer wheat, you probably need to start with the wheat RuBisCo. So there won’t be a one-size-fits-all version of any increased-efficiency RuBisCos that we can just pop into any plant we’d like.

Still, the fact that we can now make this enzyme in bacteria is a big step forward. And it could be that the research community will figure out ways of making the system more flexible with time.

Science, 2017. DOI: 10.1126/science.aap9221  (About DOIs).