About This Project

All biological carbon fixation is gated by the enzyme rubisco, which has become suboptimal as CO2 levels have risen. In order to optimize its function for carbon removal, I will use my novel rubisco-screening platform in E. coli to biochemically characterize libraries of plant enzymes. By making 100,000s of measurements I will train models which can fine tune the enzymology of plant rubisco to optimize photosynthesis for rapid CO2 removal.

Ask the Scientists

Motivating Factor

The challenge of greenhouse gas removal is primarily a matter of scale and plants are the strongest carbon sink on earth. Every year 10 billion tons (Gt) of carbon are released by anthropogenic sources while, in parallel, human-grown crop plants fix a similar amount. Fixation alone isn't enough, carbon has to be prevented from returning to the atmosphere. The first step of the process, however, is the job of one enzyme, rubisco.

Rubisco improvements will result in faster-growing, higher-yielding crops which can serve as carbon sinks or can replace less efficient crops allowing for land-use reallocation. This would result in a series of benefits to carbon removal strategies which rely on deliberately grown biomass. For instance, strategies involving carbon storage in roots, certain implementations of BECCS, biomass burial strategies and even forestation strategies could be enhanced by improvements in the photosynthetic machinery of the plants used.

Specific Bottleneck

Rubisco engineering is a longstanding challenge. Even though the catalytic mechanism is relatively well understood, it is impossible to predict the effects of mutations (other than simple inactivation). Mutations in the second and third shells around the active site have profound effects on function (i.e. rate of carboxylation and selectivity for CO2 over the side reaction with O2). Furthermore, rubiscos are adapted to the environments in which they evolved, so it is necessary to understand the effects of subtle changes.

CO2 levels are approaching two-fold what they were 200 years ago and rubisco evolves very slowly. This means that plants, including crops, are by definition not properly adapted to their environments. Finding the mutations that will tune plant rubiscos to their environment would open up many possibilities for engineering plants to be vehicles for greenhouse gas removal.

Actionable Goals

In order to generate variants of plant rubiscos with finely tuned kinetics we can use a set of three tools which are only recently available:

1) First we can use an E. coli strain I developed as a postdoc that is dependent on rubisco for growth. The key challenge here is expressing a plant rubisco in E. coli, which requires 7 chaperones.

2) New plasmid systems for continuous evolution of genes in E. coli have recently been developed. By expressing rubisco on a mutagenic plasmid it will be possible to "fast forward" rubisco evolution under our desired selection conditions.

3) Finally, by making libraries of plant rubisco mutants and characterizing them, we will produce a high throughput biochemical dataset that can be used to train ML models. These models will be able to generate rubisco sequences that have desired kinetics and can serve as starting points for further accelerated evolution campaigns.