Understanding hyphal branching in Fusarium venenatum to design improved strains

This collaborative proposal aims to characterise branching patterns in filamentous fungi in order to provide insights into gene network function and regulation. This will not only lead to fundamental insights into the molecular mechanism of hyphal branching across many species of filamentous fungi, but will help our industrial partners improve process efficiencies, ensuring that Quorn continues to be one of the most sustainable, non-plant-based, meat alternatives on the market.

In the 1950's, prior to the 'green revolution' there was serious concern about the availability of sufficient protein to feed a growing global population. Rank Hovis McDougal (RHM) began a process of searching for a single celled protein source that could be fermented using wheat starch (or derived glucose monomers). This led to the discovery of a Fusarium species, that has the correct growth properties that enabled onward processing. Later classified as Fusarium venenatum (Fv), a sister species to the wheat pathogen Fusarium graminearum, this filamentous fungus grew in a manner that had good organoleptic properties following mixture with egg albumen, forming, cooking and controlled freezing, which cause mycoprotein filaments to align as a fibre-gel composite conferring a meat-like texture .

During the production of mycoprotein, spontaneous variants arise in the fermentation which branch more quickly and eventually rise to high levels in the fermentation process. These are undesirable from a food texture perspective as they lead to a crumbly, rather than a meaty texture and therefore fermentations must be terminated early, leading to less efficient production that would be possible without these colonial variants (also known as c-variants).

Our previous work has sequenced 'c-variant' genomes from 19 independent fermentations and revealed a common set of genes that are mutated across c-variant isolates in different combinations. We now wish to verify which of these genes (and in what combination) are responsible for the c-variant phenotype. This will help us to understand which genes are responsible for controlling hyphal growth and branching, currently an 'unknown' in most filamentous fungi. We will validate our hypotheses about which variants are important by using a combination of machine-learning approaches which will allow us to effectively group c-variant isolates to aid with gene identification. We will look at gene expression perturbations across each mutant isolate, which in combination with 'active learning' approaches, allow us to identify the regulatory order of the gene expression network.

We will, in collaboration with Marlow Foods then look to understand the order of mutational events within the fermentation process and with a combination of ultra-deep population-level genome sequencing and digital droplet PCR techniques track the dynamics of individual mutations within the commercial fermentation process. We will also ask if this mutational trajectory is dependent upon the strain which is chosen for fermentation, as it may be that the current production strain is more susceptible to hyper-branching variants than other strains.

Finally, we will link together morphological growth models with gene expression networks to model the effects of different gene perturbations on the growth and branching process. This, along with other analyses will allow us to look at the robustness of the gene regulatory network and identify whether there are opportunities to improve the robustness of strains through enhanced mechanistic insight into how the hyphal growth and branching system functions and is regulated.

Taken together this work will allow the rational design of new strains of mycoprotein, enhance production efficiency and ensure that scalable alternatives to meat can be sustainably produced.

The development of colonial variants during the fermentation of Fusarium veneatum has been well documented, however the genetic basis has to date remained unclear, though multiple genes are known to be involved. Our preliminary work charts the spectrum of repeated mutations across independent fermentations. Using a combination of evidence-based knockout using dual-guide CRISPR, mutation experiments utilising base and prime editing, supplemented by RNAseq data and genetic regulatory network reconstruction using established methods (e.g. adaptive lasso) we will validate which of the observed mutations leads to c-variant and deepen understanding of the branching process in filamentous fungi, which has so far been intractable to large scale characterisation.

Using ultra-deep sequencing and subsequent dd-PCR of target mutations we will track the order of mutation development in temporal samples from multiple independent production-scale fermentations, improving our understanding of the different evolutionary trajectories that lead to c-variant formation. The fermentation in pilot facilities of multiple genetically diverse isolates of Fusarium venenatum will lead give insight into whether and potentially how genetic background affects the evolutionary trajectory of colonial variant emergence.

To provide a platform for future research we will develop a reusable L-systems model of fungal hyphal branching which will be stored as a multi-scale tree graph and through extension of the OpenAlea platform link in an additional module that will facilitate the simulation of gene regulatory networks, developed from our RNAseq and complementation experiments. This will formally link gene regulation in spatially-defined compartments to morphological variation, parameterised by mutant growth analysis of key genes. This provides a foundation for future work, but also a modelling platform in which to explore how to engineer increased robustness into the hyphal branching network.