Introduction

In the initial stages of our iGEM project, we decided to focus on a topic rooted in our local context. Wageningen University & Research is initially an agricultural university, and the Netherlands is one of the worlds largest seed exporters, with abundant expertise concentrated in regions such as Seed Valley. By choosing a project related to this local strength, we could build on existing knowledge and experience, not only from the broad iGEM community, but also from our local network here in Wageningen. In addition, we can use the network and its global connections to increase our impact and improve crops worldwide.

We hope future teams can learn from and build upon our work, just like we built on their work, reinforcing open science and collaborative innovation. BCoated provides four contributions to the community: a novel conjugation method, which expands the toolbox of Komagataeibacter sucrofermentans, a part collection that widens the possibilities for functionalising BC, and models that can be used as a basis for describing genetic feedback circuits to control the concentration of a metabolite, as well as for simulating synthetic communities (SynCom).

Expanding the toolbox of K. sucrofermentans to facilitate genome engineering

K. sucrofermentans is an extensively studied organism for bacterial cellulose (BC) production. Research is performed to optimise BC production, but most studies focus on the improvement of culture conditions or media formulations. Genetic engineering is a powerful tool for further cost reduction and optimisation¹. However, not many genome editing techniques are available. Electroporation protocols are available for some Komagataeibacter sp., but they are not adapted to all species. Initial struggles to make the protocol work drove us to develop and validate a conjugation-based transformation protocol. This expands the engineering toolbox of K. sucrofermentans, as it has not yet been reported in literature.

We decided to adapt a biparental mating protocol (donor: E. coli ST18, 5-ALA auxotroph), to transfer the pSEVA33 plasmid into K. sucrofermentans, traditionally used for the model organism Pseudomonas putida. We confirmed that it this adapted protocol worked with a colony PCR on 12 independent colonies, and we achieved a 100% success rate for this initial test set.

Having a conjugation method for Komagataeibacter available for the scientific community is valuable since electroporation is not always the ideal method of choice. Electroporation requires harsher methods and drops in efficiency when the constructs are too large, or when the DNA is modified. Thus, this method increases the capacity to engineer the model BC producing species. The conjugation protocol can be found in our protocols page.

A toolbox to functionalise BC: our part collection

BC is a unique biopolymer with high purity, crystallinity, biocompatibility and biodegradability. This makes it an interesting alternative to petrochemicals for both academic and industrial areas. Commercially, BC is used in food, personal care, household chemicals, biomedicine, textile, and composite resin. All these industries have different material requirements². In addition to tailored BC properties, BC is an interesting target for an engineered living material; a sustainable material that can perform biological functions³. The development of a toolbox of BC fusion proteins can help with future advancements in both of these directions.

By assembling a coherent and modular parts collection, we provide a reusable toolkit to functionalise BC or other cellulose-based materials. The part collection consists of 11 composite parts. The proteins add pesticides or colour to the BC matrix, or can modify the properties of the BC itself. All parts in the toolbox can be bound to BC through cellulose-binding domains (CBDs). Proteins with secretion tags are made for BC functionalisation during its production. The toolbox contains reporter parts to test the function of the protein domains. We carried out various assays to demonstrate that these parts bind to BC sufficiently, remain bound after downstream processing, and function effectively for their intended use case.

Future teams can adopt the same fusion logic for new genes for their desired BC function. They can swap linkers or binding domains as needed, and build on our validation data. The toolbox can be used to immobilise proteins for various applications of BC, and tailor their production and secretion. This is not just suitable for seed coatings, but also other in biomaterials for agriculture, biomanufacturing, biosensing, and beyond. Our part collection consisting of our composite parts can be found in the part registry.

Modelling

The inherent complexity of biological systems makes it difficult to explore the vast design possibilities experimentally. Since only a limited amount of parameters can be adjusted, the relationship between many different variables within a system becomes difficult to explore. Without computational modelling, and running thousands of simulations, many possible interactions and parameter combinations remain inaccessible and obscure⁴.

Ordinary Differential Equation (ODE) models are especially useful for modelling dynamic systems and advancing our understanding of them. Many simulations can be run at the same time, and the output gives detailed information on system sensitivities and kinetic behaviour within the biological system. Aside from helping with interpretations, these models can improve experimental designs and phenotype predictions⁵.

We developed two ODE models in our project that simulate distinct dynamic systems. The ethanol homeostasis model can used as a basis for other models of genetic feedback circuits. The consortium model can be generalised to model other Synthetic Communities (SynComs). Aside from the models themselves, we hope our models can serve as an example in their design approach. Through multiple Design-Build-Test-Learn (DBTL) cycles, they were improved to match experimental data and used to assist in the design of experiments. This workflow is concisely documented to allow future teams to adapt it for their own applications.

A basis for genetic feedback circuits

Genetic feedback circuits to control the concentration of a metabolite are a valuable tool to overcome the core challenges of achieving stable production of metabolites in high yield, without accumulating toxic intermediates or overburdening the cell. Genetic feedback circuits can sense the concentration of a metabolite in the medium and adjust how the enzymes in the relevant pathways are expressed to maintain the levels at a desired point. In our project, we used a genetic feedback circuit to maintain ethanol concentrations in the medium at a constant level, but the model framework can be applied to all genetic feedback circuits.

Our model can be used as a basis for: (1) predicting pathway behaviour before lab work by simulating how promoter strength, sensor affinity and degradation rates affect dynamics and the steady states reached; (2) enabling rational control through tuning feedback strength and simulating response curves⁶. The sensors and effector within this model can be swapped by future teams to target new metabolites, reusing our design, parameterisation logic, and workflow. These kinds of models accelerate the scale-up from R&D to industry, too, by helping achieve self-regulated microbial production, higher yields, and reducing the variability in different batches^4,6,7. This model can be found on GitLab.

A basis for modelling SynComs

Many emerging biomanufacturing strategies rely on a consortium of different species working together. This is a way of distributing metabolic tasks to reduce burden, improve yields, or enable new materials to be produced⁸.

Our design framework started with the design of a model that simulates the central carbon metabolism of monocultures. This is based on the Millard model⁹. We have extended this framework to simulate a SynCom. We incorporated the metabolic interactions between the two species. This allows us to predict how initial inoculum ratios, substrate concentrations, and metabolite exchanges will affect the overall productivity and stability of the consortium.

The existence of the Millard model formed a strong basis for our model design. In the same way, we provide the iGEM community with a foundation to rationally design microbial consortia. The model represents substrate uptake, biomass formation, metabolite secretion, and a cross-feeding interaction. The code can be reused by other teams to design and test their own SynComs: to model different partners, the organism-specific growth parameters \mu and K{_s} can be replaced. To change the substrates the species grow on, the corresponding transport and uptake terms V{_{max}} and K{_m} can be updated. The cross-feeding can be redefined to other metabolites or byproducts by changing the diffusion coefficient(s) D, V{_{max}} and K{_m}. The model can be further modified by incorporating toxicity and/or inhibitory effects of compounds, by making the rates of the reaction inversely proportional to the concentration. By sharing our adaptable basis, we can be a step in facilitating robust biomanufacturing platforms for future teams. This model can be found on GitLab.