7.Programmed population control by cell-cell communication and regulated killing

From Wiki FKKT
Jump to navigationJump to search

You, L; Cox lll, RS; Weiss, R; Arnold, FH. Programmed population control by cell-cell communication and regulated killing. Nature, 2004, 428, 868-871 URL: http://www.nature.com/nature/journal/v428/n6985/full/nature02491.html

OVERVIEW

Bacteria cells have very complex intrinsic machinery of function, therefore it is very difficult to introduce new, artificial gene circuits. Due to many obstacles such as noise in gene expression and differences between cells, it is even more challenging to predict the results we would get. You et al. managed to develop a gene circuit in bacteria based on cell survival and death which it operates through cell-cell communication. They coupled cell density regulation system from Vibrio fischeri (LuxI/LuxR) and killer protein ccdb, so they achieved autonomous regulation of population density that depends on death rate.

In the first part I will describe the principle of communication between cells, the circuit You et al constructed in experiment, the two plasmids used and the mathematical model, that explains the behaviour of cells. In the second part I will focus the results of the experiment, how can the circuit be manipulated and problem of escaping the circuit. In the third part I will give some other examples of quroum sensing in synthetic biology, biotechnology and therapy.


QUORUM SENSING

Quorum sensing is type of communication in bacteria that is naturally exploited for different purposes, for instance biofilm formation, virulence, antibiotic resistance and interaction with habitat. The signalling molecules associated with communication are called autoinducers and the concentration of autoinducers depends on production, secretion and uptake. The bacteria control their behaviour by reaching threshold (quorum) density. Mechanism of quorum sensing are various. Basically quorum sensing is common in intra- and inter- species communication. For intraspecies, Gram positive bacteria use autoinducing peptides (AIP) while Gram positive bactera use acylated homoserine lactones (AHL) as a signal. AHL have the same lactone group but they differ on the length of acyl side chain. Interspecies communication uses universal mediator. Only one of such mediator is known till now, that is AI-2 (autoinducer 2). AI2 is chemically furanosyl borate ester and is recognized by many Gram positive and Gram negative bacteria. Type of quorum sensing applied by You et al. is based on LuxI/LuxR from Vibrio fischeri.

Qurom sensing was first described in bioluminiscent marine bacterium Vibrio fischeri. It relies on small signalling molecule, that can penetrate through bacterial membrane, named AHL (acyl- homoserine lacton). It is synthesised by LuxI protein and its concetration increases with the increasing cell density. When cell population is low, AHL concentration is also low, the cells do not luminiscence and LuxR regulator degrades since AHL cannot bind it due to too low concentration. When the cell density increases and the AHL concentration reaches threshold, it binds LuxR transcriptional regulator. LuxR regulator then induces transcription of the operon encoding luciferase (luxICDABE operon). The complex induces transcription of luxC, D, A, B, E genes which are responsible for bioluminiscence and transcription of luxI, encoded on same operon. LuxR at same time inhibits transcription of luxR gene and therefore provides negative feedback loop. The positive feedback loop is formed and the bacteria produce light. By this mechanism, bacteria couple cell density and gene expression. In contrary to Gram negative bacteria, Gram positive bactera use modified oligopeptides (AIP- autoinducing peptides) as signaling molecule in quorum sensing. The peptides cannot diffuse through membrane so they require exporters and receptor on signal receiver cell. The receptor is usually histidine kinase that is bound to membrane. On the cytoplasmic side there is regulator protein that mediates the signal. When activated, it phosphorylates and regulates gene transcription.

THE CIRCUIT

In the circuit that was developed by You et al, instead of luciferase operon, the gene encoded CcdB3 toxin protein. The proteins of quorum sensing are encoded on 2 plasmids. Plasmid pLuxR12 expresses LuxI, protein that synthesizes AHL, and LuxR, transcription regulator. The transcripition of both genes is induced by isopropyl-ß-D-thiogalactopyranoside IPTG and both genes are under control of the same synthetic promotor (plac/ara-1). pluxCcdB3 encodes killer protein lacZα-CcdB. The protein is fusion of two seperate proteins: lacZα and CcdB3. LacZα has the ability to complement LacZΔM15 in certain cell strains such as Top10F’. Measuring LacZα activity tells us what is the concentration of killer protein in cell, since the two proteins are fused. To determine level of LacZα, flourogenic substrate CUG (3-carboxyumbelliferyl-ß-D-galactopyranoside) is added to bacteria lysate and then the fluorescence is measured. CcdB binds gyrA subunit of DNA gyrase and thus inhibits partitioning of chromosomal DNA. The fusion killer protein is under control of pluxl. Therefore, when enough AHL is present, it then binds luxR protein. The luxR protein is so able to bind pluxl promotor and induces transcription of killer gene that causes death of the cell. The other feature of plasmids is that pLuxRI2 is resistant to chloramphenicol and pluxCcdB3 to kanamycin. The circuit and plasmids are shown in figure 1. (1)

MATHEMATICAL MODELLING

Mathematical modelling predicts the rate of cell growth and death, production and degradation of the killer protein and AHL molecule. The equations that explain behaviour of bacteria are:


dN/dT=kN(1-N/Nm)-dE N

dE/dt=kE A-dE E

dA/dt=vA N- dA A

Where:

N- viable cell density (ml^-1)

Nm- carrying capacity (ml^-1)

E- concentration of killer protein (nM)

k- growth rate (h^-1)

d- rate constant of death (nM^-1 h^-1)

A-concentration of AHL (nM)

kE- constant of production rate of E (h^-1)

vA-constant of rate production of AHL (nM ml h^-1)

dE- constant of degradation rate of E (h^-1)

dA- constant of degradation rate of AHL (h^-1)


The first equations describes the rate of growth and death of bacteria, the second describes the rate of production of killer protein and the third the production of AHL molecule. You et al. assume that without circuit, changes in cell density depend on intrinsic per capita growth rate of k (h^-1) and carrying capacity of Nm (ml^-1) (first equation). For cell with circuit the death rate is proportional to concentration of killer protein (E) inside the cell and the cell grow with a rate constant of d (nM^-1h^-1) (first equation). The production of killer protein is proportional to concentration of AHL (A, nM), supposing that concentrations of AHL are same inside and outside the cell (second equation). The rate constant of killer protein production is kE (h^-1). AHL production is proportional to N and has the rate constant of va (nM ml h^-1) (third equation). The killer protein degrades with rate consant of dE (h^-1) and AHL degrades with rate constant dA (h^-1). N and Nm are measured as CFU ml^-1. It is supposed that formation of LuxR-AHL complex relies on AHL concentration. Even though is LuxR dimer, they suggest that the cooperativity of AHL action is 1, as observed in related quorum sensing system. If cooperativity was different, it would not change much the results. The third equation suggests that production rate of AHL is the same on avarage for each cell. If N<<Nm , then dN/dT=(k-dE)N. The model has two steady state solutions. The first is Ns=0, ES=0 and As=0. The second solution is Ns=(dA dE k)/(va kE d), Es=k/d and As=(dE k)/(kE d). s in the equations stands for steady state. Trivial steady state shows that it is unstable for all positive parameters and non trivial is stable if dA+dE>k. We can conclude that the second steady state is stable for all possible parameters in range of biological systems.

RESULTS

The system with circuit reaches steady state for all parameter values as mathematical equations show. Before the cell density comes to steady state, it might experience oscillation. When the cells were uninduced (OFF), they grew exponentially and reached the steady state when they exploited all the nutrients. The induced cells grew the same as uninduced until they reached a threshold at 7 h and the difference in cell density between ON and OFF cell becomes quite distinct. Between 7h and 24 h the ON cell oscillate before they reachy steady state that is around 10 times lower than OFF culture (Figure 2, a)(1). The reached peak density at 7h, when it was twice the density at the measured floor (12, 5 h). The steady state kept up for 30 h. The level of killer protein was very low in OFF culture and in ON culture the concentration reached steady state soon after the onset of measuring (Figure 2, b)(1). The concentration of killer protein affects the cell density. If the concentration of killer protein is lower than that at steady state, then the bacteria grow and if the concentration is higher then the steady state, the cell density reduces. After some time, lower cell density results in AHL concentration decrease and so decreases concentration of killer protein. Consequently, the the cell population grows. Constant production and degradation of circuit proteins is neccessary for homeostasis. For example if circuit lacked luxI, it could not work properly and so the cells did not grow. Similarly, if 200 nM AHL was introduced to media, the ON cell could not grow. AHL binds and activates LuxR protein and if concentration of AHL is high, then the concentration of killer protein is also high and the system is not in equilibrium.


CIRCUIT MANIPULATION

Mathematical model shows that if AHL degradation rate constant is higher then almost propotionally the steady state cell density increases. AHL is the molecule, that controls the circuit and the cell density. If AHL breaks down too fast, the circuit will not work properly. AHL can be degraded by specific enzymes or pH increase. If pH is encreased from 6,2 to 7,8, the steady state density is four times higher. The density rises only at ON cells, while it remains almost unchengeable at OFF cells (Figure 3, a-e and table 1)(1). For each pH steady state is reached at 28 h. The curves of cell densities are the same as those at pH=7. Until ON cells reach threshold, the curves superpose and later, at hight density when they reachy threshold, they divert. Mathematical model also predisposes, that the concentration of killer protein stays the same at various pH. Supposing that pH does not affect cell density substantially, the killer protein reaches steady state concentration of k/d (Es=k/d(1-Ns/Nm)= k/d) that is independent of pH. The model suggests that the ration between growth and death rate constant is far below carrying capacity (Ns/Nm<< 1). This was confirmed with results they got. (Figure 3f-i, table 1)(1). For various pH values the concentrations of the killer protein reach the same level after 28 h. pH on contrary affects k, Nm and Ns (Table 1)(1). The biggest differences were in cell densities at steady state (Ns). Due to the differences the LacZ activity has to be normalized to k(1-Ns/Nm). (Figure 3j) (1) The normalized LacZ activities should be constant as by mathematical model, pH does not affect the killing rate constant. (Es/(k(1-Ns/Nm))=1/d). The activities of LacZ were almost identical for different pH, except for pH=8,05. (Figure 3j)(1) At pH=8,05 was the LacZ activity notably higher probably due to less toxic protein.

CIRCUIT AND HETEROGENOUS POPULATION

The circuit can operate in heterogenous population. Bacteria in the population can differ on size, age, plasmid copy number, gene expression and response to killer protein. These various bacteria lack the population density control circuit, but are neccessary for circuit to work. For every cell, the result of circuit could be any of two: the cell either lives or dies. If the population was homogenous, the circuit would not be able to make the population density steady. The reason is that when the killer protein reaches threshold, the population as whole would die. The circuit can operate only at population. The experiment supports this prediction.The ON population deviates from OFF population only when the density is high. If the circuit was not dependent on cell-cell communication, the two culture densities would differ from beginning. Bacteria have their own suicide mechanisms when they activate in presence of stress, starvation or addiction modules during post-segregational killing. When the cell density of bacteria is high, the signalling causes bacteria to die. For example Streptococcus pneumonia bacteria die at high density using similar quorum sensing mechanism as You et al developed. Due to the selection, some mutant bacteria escape circuit mechanism and they appear usually 3 to 6 days after circuit activation. In the contrary to evasion the artificial circuit, evasion the natural mechanisms happens quite seldom. Natural systems probably have more elaborate system of regulation and therefore the genetic stability is higher. Cell lysis is crucial for for natural transformation in Streptococcus pneumoniae and the cells that survive uptake the DNAfrom lysed cells. Therefore there might be selection pressure that favours signalling of lysis-regulation responsible for developmental process. The concept could be tested by adding regulatory elements.

ROLE OF QUORUM SENSING IN SYNTHETIC BIOLOGY

In synthetic biology there are two main ways of reasearch related to quorum sensing. The first class of reasearch is based on developing qurom sensing circuit without using native quorum sensing elements.(2) Butler et al. for instance constructed nitrogen regulation system and acetate pathway in E. coli. The principle of the this system is that the metabolite (acetate) accumulates outside the cell during growth of bacteria and acts as a intracellular signalling molecule as it triggers gene expression. (3) The concentration of acetate gets higher as the density of bacteria population increases. The artificial genetic circuit is therefore regulated by cell density. The second class of qurom sensing reasearch is based on exploiting native quorum sensing elements. (2) Quorum sensing mechanisms are quite various in bacteria, so they can be disassembled into seperate elements and reassembled in different way. For example Danino et al. constructed synchronized oscillator combining element from Vibrio fischeri and Bacillus thuringiensis. The artificial circuit formed genetic clock that coordinated rythmic behaviour in community. (4) The second example of the second class is construction of synchronised oscillator by Prindle et al. The purpose of the circuit is to form biosensor that can detect heavy metals and pathogens. The biosensor they constructed could sense arsenic. (5) We can use quorum sensing to develop synthetic biology tool that can be applied in biotechnology. The most important aim is to produce biomaterials, recombinant proteins and functional small molecules.(2) Neddermann et al. and Weber et al. constructed expression system based on quorum sensing in eukaryots.(6, 7) Tsao et al. for example coupled E. coli quorum sensing lsr regulon and T7- expression system so they could produce recombinant protein. (8) The advantage of quorum sensing at production of protein is that the whole system is self-induced and it does not require extensive monitoring and controlling. Quorum sensing system can be exploited for therapeutic application. Quorum sensing mechanism is the main cause of virulence factor production, biofilm production and pathogenicity.The interruption in quorum sensing would probably result in disease outcome. Saeidi et al. ingeneered E. coli so that it produced pyocin S5 and E7 laysis protein, when the AHL molecule from Pseudomonas aeruginosa was present. (9) After AHL stimulation, E. coli produced E7 protein, that killed itself and pyocin S5, that killed Pseudomonas aeruginosa. Andersen et al. for instance developed quroum sensing as tool for cancer therapy. (10) They put invasin, protein that enables internatlisation of bacteria into host cell, under control of Vibrio fischeri luc QS operon in E. coli. Due to invasin protein, E. coli can be internalized into tumour cell. The internalization is dependent on E. coli population density.

CONCLUSION

Quorum sensing is a type of cell-cell communication in bacteria and the mechanisms deployed are various. For synthetic biology intentions it can serve us as a platform for different purposes. You et al coupled cell density regulation system from Vibrio fischeri (LuxI/LuxR) and killer protein ccdb, so they achieved autonomous regulation of population density that depends on death rate. They managed to construcit circuit that operated successfully. They also predicted the results using mathematical models. The circuit can be manipulate; You et al. observed relation of population and pH. The experiment results confirmed mathematical calculations as pH did not affect cell density. The main features of the experiment is that the circuit operates only in population and that population can be various in terms of size, age, plasmid copy number, gene expression and response to killer protein. If the population was homogenous it would die when the killer protein concentration would reach threshold. Even though that the circuit in general worked properly, there were some exceptions. Some mutant bacteria escaped the circuit and appeared few days after circuit activation. Quorum sensing can be used as a tool for other purposes in synthetic biology or/and biotechnology, for example genetic oscillator as a genetic clock, biosensors, material production, cancer therapy...

REFERENCES

1. You, L; Cox lll, RS; Weiss, R; Arnold, FH. Programmed population control by cell-cell communication and regulated killing. Nature, 2004, 428: 868-871 URL: http://www.nature.com/nature/journal/v428/n6985/full/nature02491.html

2. Tsao, CY; Quan, DN; Bentley, WE. Development of the quorum sensing biotechnological toolbox. Current opinion in chemical engineering, 2012, 1, 396-402

3. Bulter, T; Lee, SG; Woirl, WWC; Fung, E; Connor, MR; Liao, JC. Design of artificial cell–cell communication using gene and metabolic networks. Proc Natl Acad Sci USA, 2004, 101:2299-2304

4. Danino, T; Mondragon-Palomino, O; Tsimring, L; Hasty, J. A synchronized quorum of genetic clocks. Nature, 2010, 463:326-330

5. Prindle, A; Samayoa, P; Razinkov, I; Danino, T; Tsimring, LS; Hasty, J. A sensing array of radically coupled genetic ‘biopixels’. Nature, 2012, 481: 39-44.

6. Neddermann, P; Gargioli, C; Muraglia, E; Sambucini, S; Bonelli, F; De Francesco, R; Cortese, R. A novel, inducible, eukaryotic gene expression system based on the quorum-sensing transcription factor TraR. EMBO Rep, 2003, 4: 159-165.

7. Weber, W; Schoenmakers, R; Spielmann, M; El-Baba, MD; Folcher, M; Keller, B; Weber, CC; Link, N; van de Wetering, P; Heinzen, C; et al.: Streptomyces-derived quorum-sensing systems engineered for adjustable transgene expression in mammalian cells and mice. Nucleic Acids Res 2003, 31:e71

8. Tsao, CY; Hooshangi, S; Wu, HC; Valdes, JJ; Bentley, WE. Autonomous induction of recombinant proteins by minimally rewiring native quorum sensing regulon of E. coli. Metab Eng, 2010, 12 :291-29

9. Saeidi, N; Wong, CK; Lo, TM; Nguyen, HX; Ling, H; Leong, SS; Poh, CL; Chang, MW. Engineering microbes to sense and eradicate Pseudomonas aeruginosa, a human pathogen. Mol Syst Biol, 2011, 7:521.

10. Anderson, JC; Clarke, EJ; Arkin, AP; Voigt, CA. Environmentally controlled invasion of cancer cells by engineered bacteria. J Mol Biol, 2006, 355:619-627