The fitness landscape of a form II rubisco in a photosynthetic bacterium guides engineering of oxygen tolerance.

Ribulose-1,5-bisphosphate carboxylase/oxygenase (rubisco) is an important but challenging protein engineering target. Fast and selective rubiscos could enhance photosynthesis in plants and accelerate biobased production processes. To facilitate engineering of rubisco, we applied an in vivo screen that couples rubisco activity to growth rate of the photoautotrophic cyanobacterium Synechocystis sp. PCC 6803. We screened a barcoded mutagenesis library of the form II rubisco from Gallionella sp. containing 15,000 single-site and multi-site variants. Exchanges in loop 6 near the active site, at the dimer interface, and in potential gas tunnels improved rubisco fitness. The dataset also informed protein engineering, using recombination and a trained transformer model. In vitro characterisation of two high-fitness variants showed reduced catalytic efficiency for oxygenation (kcat/Ko) in both and an increased carboxylation turnover (kcatC) in one. This large labeled fitness dataset, containing examples of epistasis, can be useful for benchmarking computational models of rubisco.