Friedland et al., (2009). Synthetic gene networks that count.
June 2011, model of the month by Gael Jalowicki
Original models: BIOMD0000000301
Progress in the engineering of synthetic gene networks has pushed biological design to a high level of complexity, achieving understanding of more complicated natural biological systems. Among the large panel of existing circuits, BIOMD0000000301 , illustrates two counting system based on two kinds of transcriptional cascades, validated in E. Coli: riboregulated transcriptional cascade (RTC) and recombinase-based cascade of memory units.
An instance of RTC-2 Counter built with 3 genes is shown in Figure 1A. T7 RNA polymerase and GFP genes consist of a transcriptional cascade. Downstream the T7 RNA gene, and upstream the constitutive promoter Pltet0-1, a stem-loop structure is formed by the cis-repressor cr sequence, therefore inhibiting the binding of the ribosomal 30S subunit on the RBS. In the same way, cis inhibition is applied to the GFP gene located downstream of T7 RNA-induced promoter termed PT7. Furthermore cis inhibitions can be lifted by a transactivating, noncoding RNA (taRNA) induced inhibition. The protein taRNA expression is monitored by an arabinose dependant promoter PPBAD.
Model validation was realized by cloning the constructs in E. Coli K-12 pro strains (Figure 1B). The fluorescence of the GFP cells were measured by flow cytometry. By giving the first arabinose pulse, cis downstream inhibition was lifted enabling T7 RNA expression to take place. After the second pulse, a significant fluorescence peak was observed.
In order to build a RTC-3 Counter, a gene module expressing the R3 RNA polymerase was added in the middle of the RTC-2 Counter transcriptional cascade (Figure 1C). This gene being governed by a PT7-dependant promoter was subjected to the same cis and trans regulations as the others. Hence, counting was upgraded up to 3.
Similar cloning methods as below were applied with the RTC-3 construct (Figure 1D). Fluorescent leakage was observed after the second arabinose pulse. Furthermore, significant increase (2-3 times more) of GTP expression against the leakage did occur after the last and third peak, thus validating the RTC-3 Counter.
Figure 1: The RTC two-counter and RTC three-counter construct designs and results. (Figure taken from )
Figure 2: Modeling predictions and RTC three-counter experimental characterization. (Figure taken from ).
As a model prediction study, simulations of the RTC-3 Counter were performed (figure 2A). After 3 arabinose peaks, fluorescence was increased 3 times more than the leakage due to the first and second peaks. Experimental measures shown in dots seem tallying with the simulation. Simulations were then performed against pulse length and interval (figure 2B). The difference in Fluorescence level between the 2nd and the 3rd arabinose pulse tend to show a region, where the counting behaviour would be robust. As a consequence the counter counts, irregular pulses for the pulse lengths of approximately 20 to 30 minutes and pulse intervals of 10 to 40 minutes. Even if one of these parameters hit the extreme level, counting may not be done properly. This loss of accuracy depicts biochemical limits, inherent in biological processes such as transcription, translation and RNA/protein degradation.
Figure 3A illustrates three states of a DNA Invertase Cascade Counter or DIC Counter which can count up to three. This system relies on a sequence of module termed Single Invertase Memory Module (SIMM). One single module can be comparable to a DIC-2 Counter. Each SIMM is comprised of invertase genes (rec) coupled with ssrA tag to ensure rapid protein degradation. Transcriptional sequences PBAD-rec-ssrA-Term are flanked by two oppositely-oriented cognate recognition sites termed FRTf and loxPf binding respectively Flpe and Cre invertases. After the first arabinose pulse, expressed invertase flpe will transform the SIMM sequence to Term-rec-ssrA-PBAD and therefore disable its own transcription but enable Cre transcription.
After building and cloning constructs for the DIC-3 Counter, cells were subjected to three arabinose pulses (figure 3B). The first and the second arabinose pulses show an increased leakage of fluorescence, the third one is underlined by significant (at least 4 times more intense) increase of the fluorescence against leakage. Then, the ratio of the 3rd pulse to the 2nd pulse fluorescence was plotted as a function of pulse length and interval variation (figure 3C). Results show that the ratio is almost 1.5 for all conditions tested. Therefore, the DIC-3 Counter is robust for pulses whose range and intervals are between 2 to 12 hours.
In this study, mathematical modeling coupled with biological experiments reveal the dynamics of both RTC and DIC Counters. Therefore, at a very large scale through parameter optimization, bio-engineering of these circuits will be possible. Gene circuit counters are very much used in engineering for biosensing, bioremediation or biomedical purposes. They enable programming of cells behaviour by counting specific events that can occur in signaling cascades. Thus, such counters can be involved in the memory of biochemical events occurring in metabolic sensitivity or even in neural signaling.
Figure 3: The single-inducer DIC three-counter construct design and results. (Figure taken from ).
- Friedland AE, Lu TK, Wang X, Shi D, Church G, Collins JJ. Synthetic gene networks that count. Science May;324(5931):1199-202, 2009. [CiteXplore]
- Zheng Y, Sriram G. Mathematical modeling: bridging the gap between concept and realization in synthetic biology. J Biomed biotechnol 541609. Epub 2010 May 30. 2010. [CiteXplore]