Abstract
We demonstrate that resource-aware whole-cell models can be used to predict the benefits and limitations of using consortia systems to overcome burden. Our model predicts the region of expression where DOL would be beneficial for growth on starch, which will assist in making informed design choices for this, and other, complex-substrate degradation pathways.
Introduction
In synthetic biology, a primary objective is engineering microbes for "consolidated bioprocessing"—the one-step conversion of raw agricultural feedstocks, such as starch or lignin, into high-value chemicals. Traditionally, this involves engineering a single "super-strain" (monoculture) to express all necessary enzymes. However, cells possess a finite pool of internal resources, including ribosomes and ATP. Forcing a single cell to express multiple heterologous proteins creates a significant gene expression burden, leading to metabolic bottlenecks, reduced growth rates, and diminished yields.
This paper investigates whether Division of Labour (DOL)—partitioning metabolic pathways across a microbial consortium of specialized strains—can alleviate this burden and enhance productivity.
Modeling Framework
The researchers developed a mechanistic, resource-aware whole-cell model using ordinary differential equations (ODEs). Unlike traditional "black box" models, this framework explicitly tracks how the proteome is allocated across four essential categories: ribosomal, transport, metabolic, and housekeeping proteins. This allows the model to account for the direct competition between synthetic enzyme production and essential host functions.
Consortium and Starch Kinetics
The model was adapted from a single-cell framework to a two-strain consortium system:
- Strain A (Endohydrolase): Breaks down complex starch into maltodextrin intermediates.
- Strain B (Exohydrolase): Converts intermediates into glucose, which both strains consume for growth.
Parameters for E. coli growth and translation were sourced from established literature (Weiße et al., 2015), while starch degradation kinetics (alpha-amylase and glucoamylase) were derived from biochemical studies by Fujii and Kawamura (1985).
Key Insights
The Burden Threshold
The model identifies a critical tipping point in enzyme expression. At low expression levels, a monoculture is more efficient because every cell contributes to the entire degradation process. However, as the required enzyme concentration increases, the monoculture reaches a "burden threshold" where resource competition stunts growth. Beyond this point, the consortium (DOL) becomes the superior architecture.
Ribosome Liberation and Syntrophic Balance
By splitting the pathway, each cell in a consortium only produces one foreign protein. This "liberates" ribosomes, allowing individual cells to maintain higher growth rates even at high enzyme concentrations. However, this requires a delicate syntrophic balance: the endohydrolase producer must generate enough intermediates to support the exohydrolase producer, while both remain dependent on the final glucose output.
Results
- Growth Advantage: The model predicts that a consortium can outperform a monoculture in growth rate on starch. The consortium achieved a minimum doubling time of 38.66 minutes, compared to 38.85 minutes for the optimized monoculture.
- Increased Protein Yield: Cells in a DOL system produced significantly more enzyme per cell. For instance, individual cells in the consortium produced approximately 20,000 molecules of enzyme A, whereas monoculture cells were limited to roughly 16,000 molecules due to resource constraints.
- Design Space Mapping: The study produced "heat maps" to identify specific enzyme expression regions where DOL is beneficial. This provides a quantitative roadmap for bioengineers to determine the optimal community structure based on specific pathway requirements.
Limitations and Future Directions
While the model provides a robust theoretical foundation, it currently assumes a stable 1:1 ratio between strains. In practice, differential growth rates could lead to one strain outcompeting the other. Future iterations will need to incorporate population-level feedback loops, such as quorum sensing or cross-feeding, to maintain consortium stability. Additionally, while the model is grounded in established biological parameters, experimental "wet lab" validation of these starch degradation predictions remains a necessary next step. Expanding the scope to include factors like metabolite toxicity or the unfolded protein response will further refine the model's predictive power.
Summary updated