Toward a Genome Scale Sequence Specific Dynamic Model of Cell-Free Protein Synthesis inEscherichia coli

Abstract

Cell-free protein expression systems have become widely used in systems and synthetic biology. In this study, we developed an ensemble of dynamic E. coli cell-free protein synthesis (CFPS) models. Model parameters were estimated from a training dataset for the cell-free production of a protein product, chloramphenicol acetyltransferase (CAT). The dataset consisted of measurements of glucose, organic acids, energy species, amino acids, and CAT. The ensemble accurately predicted these measurements, especially those of the central carbon metabolism. We then used the trained model to evaluate the optimality of protein production. CAT was produced with an energy efficiency of 12%, suggesting that the process could be further optimized. Reaction group knockouts showed that protein productivity and the metabolism as a whole depend most on oxidative phosphorylation and glycolysis and gluconeogenesis. Amino acid biosynthesis is also important for productivity, while the overflow metabolism and TCA cycle affect the overall system state. In addition, the translation rate is shown to be more important to productivity than the transcription rate. Finally, CAT production was robust to allosteric control, as was most of the network, with the exception of the organic acids in central carbon metabolism. This study is the first to use kinetic modeling to predict dynamic protein production in a cell-free E. coli system, and should provide a foundation for genome scale, dynamic modeling of cell-free E. coli protein synthesis.Introduction 1 Cell-free protein expression has become a widely used research tool in 2 systems and synthetic biology, and a promising technology for personalized 3 point of use biotechnology [1]. Cell-free systems offer many advantages for 4 the study, manipulation and modeling of metabolism compared to in vivo 5 processes. Central amongst these, is direct access to metabolites and the 6 biosynthetic machinery without the interference of a cell wall, or complica-7 tions associated with cell growth. This allows us to interrogate (and po-8 tentially manipulate) the chemical microenvironment while the biosynthetic 9 machinery is operating, potentially at a fine time resolution. Cell-free pro-10 tein synthesis (CFPS) systems are arguably the most prominent examples 11 of cell-free systems used today [2]. However, CFPS is not new; CFPS in 12 crude E. coli extracts has been used since the 1960s to explore fundamental 13 biological mechanisms. For example, Matthaei and Nirenberg used E. coli 14 cell-free extract in ground-breaking experiments to decipher the sequencing 15 of the genetic code [3, 4]. Spirin and coworkers later improved protein pro-16 duction in cell free extracts by continuously exchanging reactants and prod-17 ucts; however, while these extracts could run for tens of hours, they could 18 only synthesize a single product and were energy limited [5]. More recently, 19 energy and cofactor regeneration in CFPS has been significantly improved; 20 for example ATP can be regenerated using substrate level pho...

Deconstructing Cell-Free Extract Preparation for in Vitro Activation of Transcriptional Genetic Circuitry

“…This result opposes a common prevailing notion, mainly borne from experiments with bacteriophage polymerases, that energy regeneration for translation strictly determines final protein titer in in vitro transcription-translation reactions. 88–90 With our optimized extracts, we were able to activate a diverse array of genetic circuitry in CFE, including RNA transcriptional activators, RNA translational activators, protein transcription factors, CRISPR-Cas9 nuclease activity, RNA translational repressors, and a translational riboswitch.…”

Dynamic Sequence Specific Constraint-Based Modeling of Cell-Free Protein Synthesis

“…We used dynamic modified flux balance analysis, a stoichiometric modeling technique that does not make the pseudo steady state assumption and allows the accumulation and depletion of metabolite species. Horvath and coworkers predicted time-dependent cell-free production of CAT using a fully kinetic model trained against an experimental dataset of 37 metabolites, including the substrate glucose, the protein product CAT, organic acids, amino acids, and energy species [32]. This model was used to generate the metabolite constraints used in this study.…”

“…Central carbon metabolites (Figure 2), amino acids (Figure 3), and energy species (Figure 4) in the synthetic measurement set were captured, within experimental error, by an ensemble of dynamic constraint-based simulations. The synthetic metabolite constraints (blue regions) shown in each of the simulation figures was derived from the kinetic model of Horvath et al [32], which was trained on experimental measurement of the 37 metabolites shown in Figures 2-4. Thus, the metabolite constraints used in this study were calculated based upon the kinetic model, which shows high fidelity with the experimental measurements.…”

“…This approach avoids the pseudo-steady-state assumption found in traditional constraint-based approaches, which allowed for the direct integration of dynamic metabolite measurements into the flux estimation problem, along with the accumulation or depletion of network metabolites. We adapted the sequence specific constraint-based model of Vilkhovoy and coworkers [31] into a dynamic constraint-based model of cell-free E. coli metabolism and protein production, and leveraged the kinetic model of Horvath and coworkers [32] to provide synthetic data to inform metabolite constraints. CFPS synthesis is often (but not always) conducted in small scale batch reactors.…”

“…However, after this initial phase, there was a significant shift in productivity following the exhaustion of glucose (which occurred at approximately 1.5 h). Horvath and coworkers developed a fully kinetic model that described the complete three hour reaction time course, including the shift in productivity following glucose exhaustion [32]. While this model described the CFPS dynamics, and the decrease in productivity following the exhaustion of glucose, the identification of the model from 37 measured metabolite trajectories was difficult.…”