Team:WITS-CSIR SA/Project/Concept

From 2011.igem.org

Revision as of 21:20, 20 September 2011 by Marco Weinberg (Talk | contribs)

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> Biotweet - Concept

sidebar

Overview

The WITS-CSIR_SA iGEM 2011 team decided to create a microscopic biological communication network, in which there will be the transfer of data between bacterial populations as chemical signals. This type of system has been dubbed “Biotweet”. The generic framework of “Biotweet” may serve as the basis for the functioning of complex biological communication networks and may be useful in various applications, including those of the medical and industrial sectors.


Any network requires the establishment of connections between nodes of the network, allowing for the transport of node-to-node data packets. With “Biotweet”, we have attemptedto construct such a network within a biological network: a challenging task, considering that bacterial cells cannot be linked by wires. Analogous to the transfer of data as electronic packets in a computer network, the concept of engineering bacteria to transport packets of chemical signals between bacterial populations arose. In this way, the bacteria themselves can form the network connections. In order to do this, we decided to reprogram the motility of bacteria, such that each bacterium travels in a stimulus-directed fashion (eg. chemotactic response) to specific points of the network (the nodes), where they can send and receive signals. Such a stimulus should be unique to the bacterium and not be part of the natural environment. With these specifications in mind, we needed a way of manipulating the chemotactic behaviour of bacteria that would allow for us to easily adapt the bacterial chemotactic response to the stimuli associated with any application. Specifically, the chemotactic response in E.coli lends itself to easy manipulation. The activation of CheZ allows for counter-clockwise (CCW) flagellar activity resulting in a “run and tumble” movement. By linking CheZ expression to post-transcriptional activation by a specific anayte (using for example a analyte-specific riboswitch), bacterial motility can be guided and controlled (Fig.1). For more on bacterial chemotaxis,see here.

Fig 1.The reprogramming of bacterial chemotactic response by using post-transcriptional activation of CheZ using a analyte-responsive riboswitch. CheZ expression results in counter-clockwise flagellar rotation and a “run and tumble” movement.


Firstly, we have chosen to use artificial riboswitches to control the expression of the flagella rotation regulator protein CheZ. To do this, we borrowed heavily from studies which have described riboswitch-controlled bacterial chemotaxis(Topps and Gallivan. 2007).Riboswitches are ligand-inducible RNA protein expression regulators that are comprised of an aptamer domain and an expression platform (Gallivan, 2007). The aptamer is a sequence of nucleotides that is designed to specifically bind to ligands, while the expression platform consists of a ribosome binding site (RBS) and a downstream coding region (CDS). When the specific ligand or analyte binds to the aptamer domain, the riboswitch undergoes a structural change. This results in the exposure of the RBS and the expression of the downstream CDS (Fig.2). Using this riboswitch mechanism, the expression of CheZ can be regulated in a ligand concentration-dependent manner, outside the control of the natural chemotaxis pathway (Fig2, bottom panel and right). Topps and Gallivan (2007) demonstrated that these riboswitches can be used to reprogram the chemotactic response of bacteria, so that they move up a concentration gradient, towards the source of the stimulus that activates the riboswitch (a process known aspseudotaxis). Bacteria can theoretically be reprogrammed to respond to any stimulus, if the appropriate riboswitch is made. This provides the versatility needed for biological network connections to be established in any application. Furthermore, this eliminates the need to engineer novel chemoreceptors.

Flowchart

Fig 2.Using ligand-responsive riboswitch activation of CheZ expression. CheZ can be fused to a reporter and direct bacterial motility in a ligand-dependent manner up a concentration gradient.


The aim was to reprogram E. coli (which lack natural CheZ functionality) with riboswitch-activated CheZ. This enables bacteria to search, localise to the riboswitch activator, and return to a set location in order to report. This, in principle, will show that bacteria can be programmed to move in a directed manner towards a new stimulus and establish a connection between two points in space. The generic design for this is depicted below. The “send” module is essentially a riboswitch-activated CheZresponsive to analyte A, and the “receive” module is a riboswitch-activated CheZ responsive to analyte B. We have decided to include two fluorescent reporters as fusion proteins since both N- and C-terminal CheZ fusions are well-tolerated (Gallivan, 2007). To ensure that the “send” and “receive” modules act independently, Analyte B can be in a biphasic toggle switch with repressor Z. Therefore Z, which represses Cre-recombinase, is activated flipping the toggle (by a critical concentration of AnalyteA). Cre recombination allows for irreversible switching of constitutive promoter activity (by lox site recombination) and convert motility from “send” to “receive” such that bacterial motility is now dictated by a concentration gradient of analyte B.

Flowchart

Fig 3.“send” and “receive” riboswitch-controlled motility module. Featured is also an Analyte B-responsive toggle switch which activates Crerecombinase expression. Cre, in turn, causes “receive” motility activationby flipping constitutively active promoter, and allowing motility to be controlled by Analyte A.

Unfortunately, other than Tetracycline (Dox) riboswitches for use in yeast, at present, there isn’t an engineeredriboswitch which is sensitive to a known transcriptional repressor. Therefore, we designed system which allows the “receive” function (Crerecombinase) to be activated by IPTG (lacI). Also, we focused this design on the use of two known synthetic riboswitches – against theophylline and atrazine (see riboswitch design considerations below).

Flowchart

Fig 4. “send” and “receive” riboswitch-controlled motility module. In this design, Cre-recombinase is activated by IPTG which is present only at the theophylline position (not present in the theophylline gradient - green). As with Fig.3, Cre, in turn, causes “receive” motility activation by flipping constitutively active promoter, and allowing motility to be controlled by an Atrazine gradient (red).


Our initial aim was to develop a theophylline-sensitive riboswitch which is capable of regulating CheZ (motility factor) translation. Several theophylline-sensitive riboswitch have been developed, most notably by the Gallivan lab (Emory, USA).

To summarize, using a high-throughput screen(Lynch et al., 2007) and a high-throughput selection (Topp and Gallivan, 2008) in E.coli, the Gallivan lab has been able to develop a powerful theophylline-dependent riboswitch. It is worth noting that the motility-based screen(Topp and Gallivan, 2008) made use of a theophylline riboswitch library activating the expression of CheZ, the motility factor in E.coli which allows for running (as opposed to tumbling) movement. Therefore, activation of CheZ allows bacteria to move along the theophylline gradient. It is worth mentioning that the best riboswitch identified in these earlier screens was the theophylline riboswitch clone 8.1 (see figure bellow), which displays 36-fold expression activation ratio (Fig.5).

Flowchart

Fig 5.Theophylline responsive riboswitch (clone 8.1). We have referred to this riboswitch as theophylline riboswitch type 1 (thRS1).

In order to improve on their earlier riboswitches, Topp and Gallivan developed an in vivo-based selection strategy aimed at selecting bacteria based on fluorescence activated cell sorting (FACS) (Wieland et al., 2008; Lynch et al., 2009) (Fig.6). By randomizing the sequence encoding the RBS (+ Shine Delgarno/SD), and by sorting a N12 library, clone 12.1 was identified, which displays a 96-fold expression activation in the presence of theophylline (1 mM) (see figure below). The authors of this study suggest this is the largest activation ratio to date, for synthetic or natural riboswitches. It is worth noting that this riboswitch contains a strong (and longer) UAAGG SD which is found in a tight (inaccessible) duplex in the “OFF” (ie no theophylline) state. Moreover, the UAAGG sequence is spaced optimally, with 6 nt separating the 5’ A of the anti-SD sequence and the start codon of the translation (AUG) (Fig.6).

Flowchart

Fig 6.Theophylline responsive riboswitch (clone 12.1). We have referred to this riboswitch as theophylline riboswitch type 2 (thRS2).

The atrazine riboswitch was developed recently by Sinha et al (2010). In order to make this riboswitch compatible with biobrick assembly, Pst1 site was removed by creating a G:U wobble base pair in lower stem region (Fig 7).

Flowchart

Fig 7.Atrazine responsive riboswitch. We have referred to this riboswitch as atrazine riboswitch (atRS) (Sinha et al., 2010).


We have identified that several theophylline riboswitches have been previously described by other iGEM teams (For reference, see Lethbridge 2007, Lethbridge 2009 and NYMU Taipei 2010). One of our initial concerns was whether it was possible to isolate the riboswitch from the associated translational unit (coding region). Being able to do so would allow the riboswitch to be defined independently as a BioBrick, with the useful ability to act in a modular fashion for fusion to an adjacent translational unit. All previous teams have made this assumption. We will show that the riboswitch and its adjacent coding region should be considered together and should be cloned as one biobrick.

Team NYMU Taipei assumed that a created theophylline riboswitch (Gallivan lab clone 8.1) biobrick could be fused to a GFP coding region by standard biobrick assembly techniques. While it is possible to do this, we feel that the resulting 6 nt scar sequence insertion increases the distance between the RBS and the ATG start codon, a result which is likely to cause a decrease in riboswitch efficiency (see team Taipei switch below, riboswitch highlighted in yellow, inserted biobrick “scar” region highlighted in violet, ATG start underlined). In fact, Taipei 2010 did show data which showed only a 2.5 fold increase in gene activation at approx. 4 mM theophylline. Compared to the 36-fold activation at 1mM theophylline by Gallivan lab, this seems too low! We therefore feel that this riboswitch is not behaving efficiently.

Fig 8. BBa_K411001,K411003 - Team NYMU-Taipei iGEM2010

Team Lethbridge 2009, used the same theophylline riboswitch (clone 8.1) to generate a riboswitch-CheZ construct. However, this team created a 11 nt insert after the ATG start codon. This result created a TAG stop codon (in the 3rd amino acid position) as well as a frameshift mutation. Therefore it is unlikely that a functional CheZ could have been produced from this system. (see team Lethbridge 2009 below, riboswitch highlighted in yellow, inserted biobrick “scar” region highlighted in violet, ATG start underlined)

Fig 9.BBa_K249028, TaemLethbridge iGEM2009. Theophylline riboswitch + scar + CheZ +dT

Gallivan JP. Toward reprogramming bacteria with small molecules and RNA.CurrOpinChemBiol 2007;11:612-9

Topp S, Gallivan JP. Guiding bacteria with small molecules and RNA. J Am ChemSoc 2007;129:6807-11.

Lynch S.A. Desai S.K.,Sajja,H.K., and Gallivan J.P.A high-trhoughput screen for synthetic riboswitches reveals mechanistic insight into their function. 2007, ChemBiol 14:173-184

Topp S. and Gallivan J.P. Random walks to synthetic riboswitches – a high throughput selection based on cell motility. 2008, ChemBiochem 9:210-213

Wieland M., and Hartig, J.S. Improved aptazyme design and in vivo screening enable riboswitches in bacteria. 2008, Agnew.Chem.Iny.Ed.Engl 47:2604-2607

Lynch S.A. and Gallivan J.P., A flow cytometry based screen for synthetic riboswitches. 2009, Nucleic Acids Res. 37(1)184-192

Sinha J, Reyes SJ, Gallivan JP.Reprogramming bacteria to seek and destroy an herbicide.Nat Chem Biol. 2010 Jun;6(6):464-70. Epub 2010 May 9.