Team:OUC-China/Design
Overview
Our whole-cell biosensor (WCB) is designed on the chassis of Escherichia coli BL21 Star (DE3) and can shed fluorescence responding to the concentration of antibiotics. Just as most typical biosensors, our WCB also consists of 3 modules[1]:
(i) the signal input module: based on an allosteric transcription factor corresponding to a specific antibiotic.
(ii) the regulatory module: consisting of signal input module-dependent parts that regulate the downstream expression.
(iii) the signal output module: a reporter gene that allows for detection, recording, and quantification of the signal.
2.Detection
To realize the basic detection function, we chose 3 allosteric Transcription Factors (tetR[2], ctcS[3] and mphR[4]) that act as repressors and respond to three different antibiotics, tetracycline, chlortetracycline and erythromycin respectively.
When antibiotics are absent, aTF will bind to the inducible promoter（PI）and prevent RNA polymerase from initiating transcription, thus repressing the expression of reporter gene. If antibiotics are present, aTFs will bind with them and change their conformation. Therefore, they will no longer be able to bind to the promoter, resulting in the expression of reporter gene. In our case, fluorescent signal increases only with the presence of antibiotics (Fig. 2).
Since this genetic circuit can only just meet the requirements of being induced by antibiotics, it still has limitations including low sensitivity and low signal-to-noise ratio, so we call it ‘basic circuit’.
3. Faster response
Common WCBs typically take at least 4-6 hours to produce a quantifiable output signal, which is not fast enough to meet the requirements of rapid field detection. To gain a faster response speed, a Three-way Junction dimeric Broccoli (3WJdB) aptamer is set as the reporter molecule. 3WJdB is an RNA aptamer that can emit fluorescence when binding with the fluorescent dye DFHBI-1T(Fig. 3a). Moreover, the dye is also suitable for in vivo experiments as it’s freely diffusible and nontoxic. Compared with traditional protein-level outputs such as GFP and luciferase, 3WJdB only needs transcription, resulting in faster signal generation and less cell burden[5]. As all the antibiotics we chose bind and disable the rRNA in ribosomes, less demand for translation may make the system more robust.
The structure of 3WJdB (Fig. 3b): two monomers of Broccoli aptamer were inserted into Arms 2 and 3 of the three-way junction (3WJ) RNA motif respectively to create the unimolecular. The dimerization of Broccoli aptamer can enhance the fluorescent signal. 3WJ can not only link the two broccolis with minimum mutual interference but also makes the whole molecule able to self-assemble and more stable in vivo[6]. In addition, a ‘GGA’ is added at the 5’ end to make the transcription initiation more efficient using T7 RNA polymerase.
4.Application of antibiotic resistance genes
Antibiotic resistance genes are added to the circuits to ensure the robust performance of our WCBs.
4.1 TetM for tetracycline
Tetracyclines (including tetracycline and chlortetracycline) inhibit protein synthesis by preventing aminoacyl-tRNA from binding to the ribosomal acceptor site. However, the tetM gene confers antibiotic resistance by producing ribosomal protection proteins.
4.2 ErmC and mphA for Macrolide
Macrolides (including erythromycin) can bind to 23S rRNA of the bacterial ribosomal 50s subunit and inhibit protein synthesis. ErmC can protect ribosomes by methylating the rRNA（this part was used by Aalto-Helsinki 2020）. MphA can help accumulate the intracellular concentration of erythromycin A(EryA) by phosphorylating it, thus increasing sensitivity[4] .
5.Platform design for better biosensor
The most straightforward way to improve a biosensor is to optimize its parts or to add new genes into the system. For example, to increase the sensitivity of a biosensor, one can modify the repressor protein and increase its affinity to the ligand. However, this approach is often difficult, inconvenient and inflexible, as a modified protein can only be used to detect a single kind of substance. We hope to improve the biosensor performance by building a platform that can rewire and fine-tune the logic of its circuit. This platform can be applied to biosensors targeting different substances only by changing the aTFs and corresponding inducible promoters. All the improvements below are designed to achieve this.
6.Amplifying the output signal
T7 RNA polymerase (T7RNAP) and T7 promoter can be used to perform stronger expression of downstream genes, which amplifies the output signal of our biosensor[7]. We set the T7 promoter upstream of 3WJdB to amplify the output signal. Meanwhile, an aTF corresponding operator (tetO, ctcO, or mphO) was also inserted between them to control the output signal. A 4bp spacer is set between the T7 promoter and the operator to gain a better regulation effect[5]. As a result, ‘T7+operator’ can act as an antibiotic inducible promoter.
We chose E. coli BL21 Star (DE3) strain as our chassis for 2 main reasons: (1) this strain carries T7 RNA polymerase gene controlled by lacUV5 promoter on its genome, so T7RNAP can be induced with IPTG. (2) The mutation of RNaseE gene reduced the amount of endogenous RNase and enhanced the stability of intracellular mRNA, which is suitable for the expression and accumulation of 3WJdB.
Moreover, T7RNAP is induced in advance to ensure the rapid response speed of antibiotic detection.
7.Improvement of signal-noise ratio (SNR) & dynamic range
Signal-noise ratio (SNR) and dynamic range are two important indicators of a biosensor. SNR refers to the ratio of induced output to noninduced output, Dynamic range refers to the difference between them. The higher SNR means it is easier for biosensors to accurately quantify samples.
7.1 Reducing leakage
T7 system will not only amplify the output signal but also increase the leakage (or background noise). To achieve a high dynamic range and SNR of WCB, despite the effort to improve the output signal, it is also important to ensure that its background expression is at a low level. That is to say, the output signal needs to be suppressed when there is no target antibiotic exists. Kleptamers are antisense nucleic acid strands that are partially or fully complementary to the aptamer.
A kleptamer (KB2) is designed to be constitutively expressed to interfere with the correct folding of 3WJdB by strand replacement reaction[8], thus repressing the fluorescent signal.
7.2 Remove KB2 repression during induction
To improve the dynamic range and SNR, the signal output should reach the maximum in the presence of antibiotics, in other words, the inhibition effect of KB2 on 3WJdB must be removed to recover the output signal.
CRISPRi system is composed of sgRNA and dCas9. dCas9 protein is a mutant of Cas9 without endonuclease activity. Therefore, it can only bind to the target site with the guidance of sgRNA.
The sgRNA is designed to be controlled by an antibiotic inducible promoter. Therefore, when sgRNA is induced, it guides dCas9 to bind to the sgRNA binding site (bs) followed by the constitutive promoter 1(Pc1), resulting in the transcription inhibition of KB2, thus removing its suppression of fluorescence（Fig. 8）
dCas9 is independently put into another plasmid. To fine-tune the expression level of dCas9, we set an inducible promoter Plux upstream of dCas9, which is regulated by different AHL concentrations. Two plasmids cooperate and form the complete version called “improved circuit” (Fig. 7).
7.3 Summary of the antibiotic detection process
In the absence of antibiotic, KB2 constitutively expresses and interfere with the basal leakage of 3WJdB to minimize the background fluorescent signal. In the presence of the target antibiotic, 3WJdB will be strongly expressed. Meanwhile, sgRNA is induced, which will guide dCas9 to bind to the binding site and repress the transcription of KB2. Finally, 3WJdB is strongly expressed without inhibition.
KB2, 3WJdB and CRISPRi system form a NOT gate and a NIMPLY gate (Fig. 9), which together increase the SNR and dynamic range of the biosensor[9]（Fig. 10）.
Pc1 and Pc2 (Fig. 8) are the two most important nodes that affect the performance of the biosensor. If Pc1 is too weak, the leakage of KB2 may not be repressed adequately. If Pc1 is too strong, the repression of KB2 may be too hard to be removed, resulting in the weak signal of output. If Pc2 is too weak, there will not be enough aTF to block theT7 promoter, leading to a huge leakage of output signal. On the contrary, if Pc2 is too strong, the sensitivity of the biosensor will be significantly reduced.