Scientists at the University of Berkeley and the Weizmann Institute have developed a mutant version of the Rubisco protein that absorbs CO2 more accurately than its natural counterparts.

Researchers around the world are working on new technologies to capture carbon from the atmosphere, but many approaches fall short of one key metric: they don’t scale. Nature has “designed” its own solution to this problem: Plants, algae, and photosynthetic bacteria are the world’s best tools for removing CO2 from the atmosphere.

Much of this work is done by one enzyme: Rubisco, the most abundant enzyme on Earth, responsible for capturing 100 gigatons of carbon each year.

But natural Rubisco is far from perfect: it is quite slow and not always precise – it makes mistakes by accidentally reacting with oxygen instead of CO2. In natural systems, there is a kind of compromise: mutant versions of Rubisco that make fewer mistakes are slower, while faster ones make more mistakes.

In a paper published in the journal Nature , a team from UC Berkeley and the Weizmann Institute of Science produced a range of mutant Rubisco molecules, most of which are not found in nature.

“The development of Rubisco will be of great importance because we can improve the ability of plants to fix CO2 and, in particular, to adapt to future atmospheric conditions,” says co-author Dave Savage.

The team analyzed how different mutations of Rubisco affect the enzyme’s speed and accuracy, i.e. its ability to capture CO2. Mutant proteins were obtained and analyzed at the Weizmann Institute. The scientists engineered a strain of Escherichia coli (E. coli) bacteria that depends on Rubisco. In nature, E. coli does not use Rubisco at all, but the engineered strain cannot survive without it. The team achieved vigorous growth of bacteria encoding Rubisco.

The scientists received more than 9,000 protein variations and selected the optimal ones. One mutant protein doubled the accuracy of CO2 fixation compared to natural versions, and another even tripled it. But as in nature, these proteins turned out to be quite slow. The scientists hope to go beyond the natural “compromise”.

“We’ve made such a big change with just one mutation, but it’s important to remember that this is just the beginning,” Savage says. “By combining more data and machine learning, we hope to design better versions of the protein and eventually have a version that’s both fast and accurate.”

https://www.newsru.co.il/science_hitech/13feb2025/rubisco.html