WP1 Biology, bioengineering and bioinformatics (PDRA1) identifies and validates Rubisco linkers, pyrenoid matrix-membrane tethers and matrix-starch tethers in diverse algae. Bespoke bioinformatics and small-volume in vitro LLPS assays will identify and validate Rubisco binding proteins across lineages. Carefully selected algal systems guided by the above research questions and WP3 modelling will then provide the focus for in-depth functional studies. Developed algal lines will feed directly into WP2 and data will refine WP3 modelling.

WP1 Preliminary data We have shown in Chlamydomonas that five Rubisco binding motifs (RBMs) in EPYC1 bind Rubisco holoenzymes resulting in pyrenoid formation [3]. RBMs are also found in multiple Chlamydomonas pyrenoid proteins and appear to mediate pyrenoid assembly and structural organisation (Fig. 1) [8]. We hypothesise similar Rubisco condensation and pyrenoid organisation take place across diverse algae (Fig 2) [2], however at the sequence level EPYC1 homologs are absent across lineages. To identify Rubisco linkers across diverse algae, we developed a novel bioinformatics pipeline to scan algal genomes for proteins with similar physicochemical properties (Fig. 3A). Preliminary work using Chlorella sorokiniana has shown proteins identified using this pipeline to be bona fide Rubisco linkers (Fig. 3B). We have established transformation, fluorescent protein tagging and genetic engineering protocols in three diverse algae that will form the basis of the in vivo work: Thalassiosira pseudonana (diatom), C. reinhardtii (Chlorophyceae) and C. sorokiniana (Trebouxiophyceae). Guided by early data, we can enhance evolution and structural diversity using diverse genetically tractable algal and hornwort plant pyrenoid systems [9].

WP1 Programme of work

WP1.1 Expansion of the bioinformatics pipeline (LM: Months M1-M6) refines and runs our bioinformatics pipeline on all algae with available genomes across the eukaryotic tree of life (Fig 3A). Top hits will be checked against available experimental data (i.e. proteomic and transcriptomic) and in silico docking of the predicted RBM with Rubisco. Top candidates will feed directly into WP1.2 and a subset biochemically validated (WP1.3). To identify additional pyrenoid proteins, putative RBMs ZLOO WKeQ be VeaUcKed aJaLQVW WKe WaUJeW VSecLeV¶ WUaQVOaWed JeQRPe (FLJ. 3C). We ZLOO WKeQ cRPSaUe RBM containing proteins across clades and structural features to identify conserved functional components and physicochemical properties. Experimental (WP2) and modelling (WP3) results will point to critical ranges of interaction parameters and help prioritise protein families with properties that could drive pyrenoid structural features (i.e. tubular membranes). Potential membrane-matrix and starch-matrix tethers will be prioritised for characterisation in WP1.4-1.5 and WP2/WP3.

WP1.2 Pair-wise LLPS analyses of Rubisco and scaffolds (MP: M1-M18) will evaluate compatibility of Rubisco and linker proteins. We will recombinantly produce and characterise 20-30 linker proteins identified across diverse algae in WP1.1. Rubisco proteins will be purified directly from source algae using robust protocols already validated across diverse algae by LM. We will develop protocols for higher-throughput small-volume evaluation of LLPS in vitro to evaluate LLPS of combinatorial protein mixtures (guided by LLPS analyses in WP2.1), expanding analysis of interspecies cross-reactivity and identifying conserved sequence features. The results will improve our bioinformatic tools (WP1.1) and provide new hypotheses to test via modelling (WP3.1-3.3). This system can also be used to evaluate proteins identified in WP1.3 prior to more-detailed analysis (WP2 and WP3).

WP1.3 Biochemical identification of linkers and additional pyrenoid components (LM: M1-M12) corroborates bioinformatic data from WP1.1, via immunoprecipitation of Rubisco and mass spectrometry of interacting proteins [3] on a carefully selected range of evolutionary and pyrenoid structure diverse algae. Identified proteins will be confirmed in WP1.2, 1.4, 1.5, fed into in vitro reconstitution work in WP2, and mapped onto the model parameter space of WP3.

WP1.4 Fluorescent protein tagging to confirm sub-pyrenoid localisation (LM: M6-M18) is performed on linker proteins, membrane-matrix and matrix-starch tethers in selected algae. Tagged lines will feed directly into WP2/WP3 enabling advanced single-molecule and super-resolution imaging/simulation of the pyrenoid, with those already developed for Chlamydomonas (Fig 4).

WP1.5 CRISPR/Cas9 knock-out of core structural components (LM: M12-M30). Knock-out of EPYC1 in Chlamydomonas results in the failure to assemble Rubisco into the pyrenoid. To confirm linker function and to prioritise systems for WP2/WP3, where feasible candidate linkers will be knocked out using CRISPR/Cas9. Pyrenoid assembly defects will be assessed via transmission electron microscopy (TEM) and functional role assessed via growth and photosynthetic assays.

WP1.6 Protein replacement to validate linker compatibility and function (M24-M36). Using a developed Chlamydomonas line that lacks both EPYC1 and Rubisco small subunits we can reconstitute foreign linker/Rubisco combinations to validate in vivo condensation, test cross compatibility, and test specific residue importance for Rubisco condensation and pyrenoid assembly. Specific experiments will be guided by in vitro work in WP2 and in silico work in WP3.