Abstract text: Identifying protein interactions is crucial for understanding the molecular machinery that drives plant cell wall biosynthesis. While immunoprecipitation approaches have revealed core components of stable complexes, they are less effective for detecting weak, transient or spatially restricted interactions. Proximity labelling methods such as TurboID overcome some of these limitations, but increasing mass spectrometry sensitivity often yields thousands of potential interactors, making confident network reconstruction challenging.

We developed a multi bait proximity labelling strategy that integrates a broad set of experimental baits with diverse subcellular controls, enabling stringent filtering and generation of high confidence interaction networks and examined proteins involved in cellulose synthesis in Arabidopsis thaliana. Cellulose is produced by the plasma membrane localised cellulose synthase complex (CSC), but also requires numerous accessory factors for correct assembly, trafficking and function.

Using TurboID fusions to ten proteins associated with cellulose synthesis and seven control baits, we initially identified 2,784 interactions enriched over no bait and soluble controls. Applying enrichment thresholds derived from organelle controls reduced this to a core network of 117 interactions among 41 proteins, capturing known CSC components and revealing novel candidates. Reverse genetics analysis identified a previously uncharacterised protein family that modulates CESA abundance.

Our results demonstrate that multi-bait proximity labelling is a powerful and robust approach for defining interaction networks central to cell wall biosynthesis.