Template:Groningen/Model

Successful in vivo directed evolution by PREDCEL and PACE requires the thorough consideration of experimental parameters, e.g. phage propagation times, culture dilution rates and inducer/inhibitor concentrations. We employed extensive ODE-based and stochastic modeling to identify the most sensitive parameters and adapt our experiments accordingly. First, we calibrated our models using phage propagation experiments from our wet lab complemented with literature data. Simulations showed that the phage titer is highly sensitive to culture dilution rates. We simulated batch times and transfer volumes for PREDCEL and corresponding flow rates for PACE to determine optimized conditions for gene pool selection while avoiding phage washout. We also estimated phage titer monitoring intervals for cost and labor efficient QC/monitoring as well as inducer/inhibitor concentrations required to express the required mutagenic polymerases. Finally, we provide a web-based, fully interactive modeling platform that not only informed our wet lab experiments, but enables future iGEM teams to efficiently build on our work.

Phage titer

Simulations of phage and E. coli titer support both PREDCEL and PACE by helping to choose a set of experimental parameters that is both efficient in terms of directed evolution and in terms of usability.

Numeric Model

Interactive Webtools

Use the interactive tools to simulate the conditions you are interested in and explore how the combined experimental parameters influence experimental outcomes.

Overview

Mutagenesis Induction

Model the glucose and arabinose concentration to make sure mutagenesis plasmids are sufficiently induced to get optimal mutagenesis conditions for both PREDCEL and PACE.

Analytic Model

Glucose Tool

Arabinose Tool

Lagoon Contamination

Check if lagoons are vulnerable to contamination by microorganisms under given experimental conditions.

Analytic Model

Interactive Webtool

Mutation Rate Estimation

Estimate the number of mutated sequences in a PREDCEL or PACE experiment at a given point in time to check for the covered sequence space and to save time and money when sequencing.

Analytic Model

Interactive Webtool

Medium Consumption

Calculate the amount of medium needed for a PACE experiment, see how medium consumption can be reduced when experimental parameters are optimized.

Analytic Model

Interactive Webtool

Equilibration MD simulations

To assert what effects our mutations entail on protein fold, we performed Molecular Dynamics simulations.

Software Validation

Our integrated styrene production pathway in Saccharomyces cerevisiae relies on several key properties to make the process the most efficient in ideal conditions. Our enzyme scaffold which supposedly enhances enzyme activity by bringing the enzymes in closer proximity to each other, and closer to cellulose by a cellulose binding domain. Furthermore our yeast strain needs several metabolic optimizations to maximize styrene production levels.

To verify whether an enzyme scaffold indeed enhances enzyme activity, a mathematical model was created to describe catalyzed degradation of cellulose in the presence of our enzymes. The model, based on work by Levine et al. [1], includes an exoglucanase and an endoglucanase, both in bound together and separately in solution. The model provides a justification for the use of scaffolding and insight into how to choose enzymes for a scaffold

As it is likely that substrate availability is higher when the scaffold is strongly bound to cellulose, we explored the binding characteristics of the cellulose binding domain that is attached to our minicellulosome scaffold using coarse grained molecular dynamics. The cellulose binding domain (CBD) of Cip3A from Clostridium thermocellum (PDB 1NBC by Tormo et al. [2]) was modeled in the presence of a single cellulose fiber using coarse grained molecular dynamics simulations [3]. The simulations affirm the hypothesized binding mechanism of CBD to cellulose, and show that CBD from C. thermocellum has a very high affinity for cellulose fibers. From these simulations, the potential of mean force of the binding interaction could be computed, from which the binding free energy could be derived. The results show that our CBD binds even stronger to cellulose than some other cellulose binding domains known from literature. These results justify including the CBD in the enzyme scaffold as it will likely increase the activity of the complex by substrate proximity.

Finally flux balance modeling was used to model the metabolism of S. cerevisiae after introduction of our heterologous pathway enabling styrene production. Yeast GEM 7.6 [4] was used to model the metabolic network of yeast. Furthermore, the OptForce algorithm [5], implemented in the COBRA toolbox [6], was used to find compatible sets of up- or down-regulated genes. When optimizing for biomass, it was shown that 65% of maximum styrene production is still compatible with normal growth of yeast, justifying our choice for yeast as our host organism.

To speed up simulations, our models were run using the Groningen University’s state of the art 4000+ core Peregrine cluster.

References:

1. Abudayyeh, O.O., et al., 2016. C2c2 is a single-component programmable RNA-guided RNA-targeting CRISPR effector. doi: 10.1126/science.aaf5573.

References

[1] Levine, S. E., Fox, J. M., Blanch, H. W., & Clark, D. S. (2010). A mechanistic model of the enzymatic hydrolysis of cellulose. Biotechnology and Bioengineering. https://doi.org/10.1002/bit.22789

[2] Tormo, J., Lamed, R., Chirino, A. J., Morag, E., Bayer, E. A., Shoham, Y., & Steitz, T. A. (1996). Crystal structure of a bacterial family-III cellulose-binding domain: a general mechanism for attachment to cellulose. The EMBO Journal, 15(21), 5739–5751. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8918451

[3] Marrink, S. J., Risselada, H. J., Yefimov, S., Tieleman, D. P., & Vries, A. H. de. (2007). The MARTINI Force Field: Coarse Grained Model for Biomolecular Simulations. J. Phys. Chem. https://doi.org/10.1021/JP071097F

[4] Sánchez, B. J., Zhang, C., Nilsson, A., Lahtvee, P., Kerkhoven, E. J., & Nielsen, J. (2017). Improving the phenotype predictions of a yeast genome‐scale metabolic model by incorporating enzymatic constraints. Molecular Systems Biology. https://doi.org/10.15252/msb.20167411

[5] Ranganathan, S., Suthers, P. F., & Maranas, C. D. (2010). OptForce: An optimization procedure for identifying all genetic manipulations leading to targeted overproductions. PLoS Computational Biology. https://doi.org/10.1371/journal.pcbi.1000744

[6] Heirendt, L., Arreckx, S., Pfau, T., Mendoza, S. N., Richelle, A., Heinken, A., … Fleming, R. M. T. (2017). Creation and analysis of biochemical constraint-based models: the COBRA Toolbox v3.0. ArXiv. https://doi.org/10.1038/protex.2011.234