The help and support of my MD/PhD directors Kelsey Martin and Stephen Smale and my rotation advisor Bruce Hay were instrumental in the decision to undertake this PhD – they gave me time to think and make an informed decision about what I wanted to research. My mentors gave me the independence to ask questions and explore at my own pace and to embrace failure and dream big. I would like to thank David Baltimore for being a wonderful and inspiring role model and for providing thoughtful career advice.
I would also like to thank Michael Elowitz – thank you for the opportunity to rotate and learn more about your research, and for all your insightful comments on my work, especially on our repressilator-inspired article. Finally, Vincent Noireaux has been a great mentor. I have enjoyed the opportunity to work closely on cell-free lysates and co-author articles with you, and I am humbled by your technical depth and scope. Thank you for pushing me to achieve, for supporting me along the way, and for helping me find what I really enjoy doing.
Specifically, we used our understanding of linear DNA prototyping, modular assembly, and protein degradation dynamics to characterize the repressor oscillator and to prototype novel three- and five-node negative feedback oscillators both in vitro and in vivo.
Introduction to biomolecular circuit engineering and to cell-free systems
Cell-free systems have a rich history as one of the original platforms on which biological experiments were performed. In parallel with the development of S30 extract, work was carried out on a new cell-free system from individual components (Shimizu et al. 2001). Synthesis of 2.3 mg/ml protein with an entire Escherichia coli cell-free transcription-translation system.
Streamlined extract preparation for Escherichia coli-based cell-free protein synthesis by sonication or bead vortex mixing. A highly efficient and economical cell-free protein synthesis system using the S12 extract of Escherichia coli. Simple procedures for the construction of a robust and cost-effective cell-free protein synthesis system.
Study of messenger RNA inactivation and protein degradation in an Escherichia coli cell-free expression system.
Making the breadboard accessible
This five-day protocol describes all the steps, equipment, and supplementary software necessary to create and run an efficient endogenous Escherichia coli-based TX-TL cell-free expression system from scratch. We have presented a five-day protocol for the production of an endogenous Escherichia coli-based TX-TL cell-free expression system. To demonstrate the cell-free expression system, we constructed and tested a negative feedback loop based on tet repression.26 (Figure 6).
In the cell-free expression system, the same cycle with and without aTc shows a 7-fold endpoint expression change of deGFP reporter after eight hours of expression. The endogenous Escherichia coli-based TX-TL cell-free expression system described here is an easy-to-run three-tube reaction that can take less than eight hours from setup to data collection. If the cell-free expression system is used to perform sensitive quantitative experiments, it is advisable to perform all experiments with the same batch of crude cell extract.
Ultimately, we envision immediate uses of this endogenous cell-free expression system as a prototype environment for synthetic biology. Different concentrations of Tris-Cl are compared in a cell-free expression reaction based on expression of 1 nM of pBEST-OR2-OR1-Pr-UTR1-deGFP-T500. Stabilization of amino acids for cell-free protein synthesis by genome modification of Escherichia coli.
Accelerating the breadboard
We then compared the expression levels and binding dynamics of different promoters on linear DNA and plasmid DNA. We first developed protective mechanisms to make linear DNA expression comparable to plasmid DNA. This plasmid was previously optimized for high expression in TX-TL.16. DeGFP synthesis from linear DNA was less than 2% of the plasmid DNA (Figure 2a).
With gamS protein present in the reaction, deGFP synthesis from OR2-OR1-Pr-UTR1-deGFP-T500 linear DNA was 37.6% of plasmid DNA (Figure 2a). While purified gamS enhanced linear DNA expression, it showed no toxicity towards plasmid DNA expression (Figure 2c). While plasmid DNA entered the saturation regime above 4 nM, linear DNA remained in the linear regime up to 16 nM (Figure 2c).
We have also established that degradation of linear DNA is a saturable process limited by the amount of exonuclease. To calibrate linear DNA to plasmid DNA for constitutive expression, we tested each promoter at different concentrations. As predicted, expression from linear DNA at similar concentrations was also lower than that of plasmid DNA.
Based on this result, we believe that linear DNA prototyping is a viable alternative to plasmid DNA prototyping. Our work is primarily focused on the technological development of a rapid prototyping procedure using linear DNA in TX-TL. However, the real payoff of linear DNA prototyping lies in testing large circuits in TX-TL.
Prototyping with linear DNA in TX-TL can reduce cycle times and increase iteration speed. We have demonstrated that linear DNA results can be mapped to plasmid DNA results in the context of individual promoters. Linear DNA is protected with 31 bp steric protection and with gamS. Error bars represent one standard deviation from three independent experiments. in vitro: Linear DNA for TX-TL Traditional cloning:.
Linear DNA is protected with 31 bp steric protection and with gamS. Error bars represent one standard deviation from three independent experiments.. 1980).
Characterizing dynamics in the breadboard
This limits the complexity of circuits built in TX-TL without steady-state or continuous flow solutions [8-10]. In particular, TX-TL in batch mode is known to be resource limited [ 16 ], and ClpXP is known to require significant amounts of ATP to unfold various protein targets [ 17 , 18 ]. ClpXP provides a controllable way to introduce protein degradation and dynamics into synthetic circuits in TX-TL.
We additionally attempted to supplement TX-TL with ATP and Mg, based on the known intensive use of ClpX for protein unfolding (500 ATP per titin I27 subunit, ~100 aa) [19]. With 200 nM added ClpX in TX-TL there was a clear effect on the degradation of Venus-ssrA relative to deGFP-ssrA and mRFP-ssrA, indicating that at working concentrations of proteins is likely achievable in a typical TX-TL reaction ATP concentration was not rate limiting, except for difficult-to-degrade proteins (Fig. 4). We diagnosed this by diluting purified ClpX in a TX-TL reaction that drives production of deGFP from a strong promoter, and we saw no effect in saturating ClpX levels to DNA expression (Fig. 5).
We also verified that ATP is not rate-limiting for ClpXP degradation and DNA expression by treating mRFP-ssrA degradation as a “background process” occurring during the deGFP-producing TX-TL reaction (Figure 6). We also investigated the increase in the amount of ClpP in the saturated ClpX TX-TL reaction to test whether ClpP is the limiting reagent. We explored the use of complementing the ClpXP system, which is part of TX-TL, with purified ClpX protease to provide fine-tunable degradation to drive synthetic dynamic circuits.
Plasmid expressing deGFP from the strong promoter-UTR is added to the TX-TL reaction at 1 nM and endpoint expression after 8 h is plotted as a function of added, purified ClpX protein. A plasmid expressing deGFP from a strong promoter-UTR was added to the TX-TL reaction at 1 nM. In the same reaction, purified mRFP-ssrA is degraded by 200 nM added ClpX. No additional ATP or Mg is added to the reaction. In the eZS6 extract, a plasmid expressing ClpP from a strong promoter-UTR was added to the TX-TL reaction at various concentrations in the presence of 200 nM added purified ClpX. The degradation of deGFP-ssrA is shown as a function of time.
Sun, Z.Z., et al., Protocols for the Implementation of an Escherichia Coli-Based TX-TL Cell-Free Expression System for Synthetic Biology. Sun, Z.Z., et al., Linear DNA for rapid prototyping of synthetic biological circuits in an Escherichia coli-based TX-TL cell-free system.
Applying the breadboard on a complex synthetic circuit
We compared the original repressor with a modified version containing a point mutation in one of the CI repressor binding sites in the promoter that regulates LacI (Fig. 2A). In the CI promoter OR2* mutant, we observed the expected shift in KM and decreased steepness of the transfer function. In our characterization of the repressilator network and the 3n1 oscillator, we found that dilution rates were critical for the existence, period and amplitude of oscillations.
However, a quantitative characterization of the synchronization phenotype requires more in-depth understanding of stochastic effects in vivo. Oscillation periods of the 5n oscillators were also consistent with our in vitro results and showed a similar dependence on doubling time ( Fig. 4D ). We observed some differences between the in vitro and cellular environment, particularly in the difficulty of predicting cellular toxicity and charging effects of the 5n oscillators in vivo.
The original repressor plasmid, pZS1 (8), was used as a template for initial characterization and for the construction of the OR2* mutant. Initial conditions for limit cycle analysis of the repressor network were set by adding previously synthesized repressor protein at the beginning of each experiment. Transfer functions of repressor-promoter pairs were determined in a nanoreactor device at at least two different dilution times (Figure S2).
For a specific example of the cell-free framework applied to design a 5-node oscillator network, see Fig. Time series micrographs of 3n1 under a strong pPhlF sfGFP-ssrA reporter every 160 min; inset shows individual cells of the initial microcolony. Histograms of the periods observed with a weak and a strong pPhlF sfGFP-ssrA reporter for both 3n1 and 5n2 runs in the mother machine.
Transcription rates could be rapidly adjusted by varying DNA template concentrations of the repressor plasmid. A simulation of the repressor network gave similar results, but did not capture strain effects on the biosynthetic machinery for high DNA template concentrations.
Conclusion and Future Directions
