Limits...
A Boolean model of the Pseudomonas syringae hrp regulon predicts a tightly regulated system.

MacLean D, Studholme DJ - PLoS ONE (2010)

Bottom Line: The Type III secretion system (TTSS) is a protein secretion machinery used by certain gram-negative bacterial pathogens of plants and animals to deliver effector molecules to the host and is at the core of the ability to cause disease.Extensive molecular and biochemical study has revealed the components and their interactions within this system but reductive approaches do not consider the dynamical properties of the system as a whole.We compared simulations of the model with experimental data and found them to be largely in accordance, though the hrpV node shows some differences in state changes to that expected.

View Article: PubMed Central - PubMed

Affiliation: The Sainsbury Laboratory, John Innes Centre, Norwich, United Kingdom. dan.maclean@tsl.ac.uk

ABSTRACT
The Type III secretion system (TTSS) is a protein secretion machinery used by certain gram-negative bacterial pathogens of plants and animals to deliver effector molecules to the host and is at the core of the ability to cause disease. Extensive molecular and biochemical study has revealed the components and their interactions within this system but reductive approaches do not consider the dynamical properties of the system as a whole. In order to gain a better understanding of these dynamical behaviours and to create a basis for the refinement of the experimentally derived knowledge we created a Boolean model of the regulatory interactions within the hrp regulon of Pseudomonas syringae pathovar tomato strain DC3000 Pseudomonas syringae. We compared simulations of the model with experimental data and found them to be largely in accordance, though the hrpV node shows some differences in state changes to that expected. Our simulations also revealed interesting dynamical properties not previously predicted. The model predicts that the hrp regulon is a biologically stable two-state system, with each of the stable states being strongly attractive, a feature indicative of selection for a tightly regulated and responsive system. The model predicts that the state of the GacS/GacA node confers control, a prediction that is consistent with experimental observations that the protein has a role as master regulator. Simulated gene "knock out" experiments with the model predict that HrpL is a central information processing point within the network.

Show MeSH

Related in: MedlinePlus

Attractor trees of the model for the 128 different start states.We ran the model in synchronous mode starting from each of the 128 possible combinations of states. Each circle represents a possible state of the model and the edge indicates the state to which the model evolves on the next iteration. The tree with the terminal node labelled ‘ON’ has an attractor with the same state as the steady state of runs with the model i.e GacSGacA  =  True; RpoN  =  True; HrpV  =  False; HrpG  =  True; HrpRS  =  True; HrpL  =  True; HrpA  =  True; The tree with terminal node labelled ‘OFF’ has an attractor in which all states are false.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC2821412&req=5

pone-0009101-g003: Attractor trees of the model for the 128 different start states.We ran the model in synchronous mode starting from each of the 128 possible combinations of states. Each circle represents a possible state of the model and the edge indicates the state to which the model evolves on the next iteration. The tree with the terminal node labelled ‘ON’ has an attractor with the same state as the steady state of runs with the model i.e GacSGacA  =  True; RpoN  =  True; HrpV  =  False; HrpG  =  True; HrpRS  =  True; HrpL  =  True; HrpA  =  True; The tree with terminal node labelled ‘OFF’ has an attractor in which all states are false.

Mentions: So that we might understand the paths through which the system could possibly run, we ran the model in synchronous mode for 10 steps starting from each of the 2 (128) possible states of the model and mapped the dynamical trajectories from each start state to the final state via every state the model occupied on the way. Such an analysis provides a map of the way the system could possibly behave under different combinations of ectopic expression of its genes. The results of this analysis can be seen in Figure 3. In Figure 3 each of the dots represents a model state and the arrows lead from a state to the subsequent state. Remarkably, the model converges on two discrete end-points or attractors. The two discrete trees that lead to an attractor each contain 64 states, 50% of the total. One of the trees leads to an attractor identical to the steady state describe above, with all the proteins (with the exception of HrpV) showing a state of True. The second tree leads to an attractor with all proteins in the False state. A similar analysis averaging 10,000 runs in an asynchronous mode showed the same pattern. The presence of just two attractors indicates that the system is a strongly regulated switch, optimised to allow only expression of the components of the hrp regulon all together or not at all, predicting that a non-constitutive mutation in expression of any combination of genes cannot cause ectopic expression of the hrp regulon. To ascertain whether or not a specific factor or factors in the dynamical model could be determinants of which attractor a state leads to, we calculated for each protein the number of times it was true or false for each step of the evolution of the model. We did this starting from each possible start step in each of the two attractor trees described above. During the initial states of the runs for both attractor trees, each protein, except GacSGacA could be in either state, in fact at the start of runs all proteins except GacSGacA were equally in either state. The state of GacSGacA throughout the runs corresponds to the final state in each of the attractor trees, when GacSGacA is True the model is attracted to an ‘on’ steady state regardless of other perturbations, and when GacSGacA is False the model attracts to an ‘off’ state. This indicates that GacSGacA is the sole determinant of the expression of the genes of the hrp regulon and that ectopic expression of other components cannot initiate or sustain the expression of the regulon. Such an observation is intellectually satisfying firstly because it reflects the situation observed in vivo but secondly, and more importantly, it reflects a system that is not capable of being accidentally ‘hot-wired’ by changes in its components expression patterns. Therefore expensive accidental deployment of the TTSS machinery and effectors is not likely to occur because of short-circuiting of the system itself. The GacSGacA dependency is both a strength and a weakness. Although the pathogen is able to deploy its TTSS according to specific inputs and is not likely to accidentally misfire, hosts that are able to disrupt the activation of GacSGacA are able to prevent the activation of the TTSS.


A Boolean model of the Pseudomonas syringae hrp regulon predicts a tightly regulated system.

MacLean D, Studholme DJ - PLoS ONE (2010)

Attractor trees of the model for the 128 different start states.We ran the model in synchronous mode starting from each of the 128 possible combinations of states. Each circle represents a possible state of the model and the edge indicates the state to which the model evolves on the next iteration. The tree with the terminal node labelled ‘ON’ has an attractor with the same state as the steady state of runs with the model i.e GacSGacA  =  True; RpoN  =  True; HrpV  =  False; HrpG  =  True; HrpRS  =  True; HrpL  =  True; HrpA  =  True; The tree with terminal node labelled ‘OFF’ has an attractor in which all states are false.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC2821412&req=5

pone-0009101-g003: Attractor trees of the model for the 128 different start states.We ran the model in synchronous mode starting from each of the 128 possible combinations of states. Each circle represents a possible state of the model and the edge indicates the state to which the model evolves on the next iteration. The tree with the terminal node labelled ‘ON’ has an attractor with the same state as the steady state of runs with the model i.e GacSGacA  =  True; RpoN  =  True; HrpV  =  False; HrpG  =  True; HrpRS  =  True; HrpL  =  True; HrpA  =  True; The tree with terminal node labelled ‘OFF’ has an attractor in which all states are false.
Mentions: So that we might understand the paths through which the system could possibly run, we ran the model in synchronous mode for 10 steps starting from each of the 2 (128) possible states of the model and mapped the dynamical trajectories from each start state to the final state via every state the model occupied on the way. Such an analysis provides a map of the way the system could possibly behave under different combinations of ectopic expression of its genes. The results of this analysis can be seen in Figure 3. In Figure 3 each of the dots represents a model state and the arrows lead from a state to the subsequent state. Remarkably, the model converges on two discrete end-points or attractors. The two discrete trees that lead to an attractor each contain 64 states, 50% of the total. One of the trees leads to an attractor identical to the steady state describe above, with all the proteins (with the exception of HrpV) showing a state of True. The second tree leads to an attractor with all proteins in the False state. A similar analysis averaging 10,000 runs in an asynchronous mode showed the same pattern. The presence of just two attractors indicates that the system is a strongly regulated switch, optimised to allow only expression of the components of the hrp regulon all together or not at all, predicting that a non-constitutive mutation in expression of any combination of genes cannot cause ectopic expression of the hrp regulon. To ascertain whether or not a specific factor or factors in the dynamical model could be determinants of which attractor a state leads to, we calculated for each protein the number of times it was true or false for each step of the evolution of the model. We did this starting from each possible start step in each of the two attractor trees described above. During the initial states of the runs for both attractor trees, each protein, except GacSGacA could be in either state, in fact at the start of runs all proteins except GacSGacA were equally in either state. The state of GacSGacA throughout the runs corresponds to the final state in each of the attractor trees, when GacSGacA is True the model is attracted to an ‘on’ steady state regardless of other perturbations, and when GacSGacA is False the model attracts to an ‘off’ state. This indicates that GacSGacA is the sole determinant of the expression of the genes of the hrp regulon and that ectopic expression of other components cannot initiate or sustain the expression of the regulon. Such an observation is intellectually satisfying firstly because it reflects the situation observed in vivo but secondly, and more importantly, it reflects a system that is not capable of being accidentally ‘hot-wired’ by changes in its components expression patterns. Therefore expensive accidental deployment of the TTSS machinery and effectors is not likely to occur because of short-circuiting of the system itself. The GacSGacA dependency is both a strength and a weakness. Although the pathogen is able to deploy its TTSS according to specific inputs and is not likely to accidentally misfire, hosts that are able to disrupt the activation of GacSGacA are able to prevent the activation of the TTSS.

Bottom Line: The Type III secretion system (TTSS) is a protein secretion machinery used by certain gram-negative bacterial pathogens of plants and animals to deliver effector molecules to the host and is at the core of the ability to cause disease.Extensive molecular and biochemical study has revealed the components and their interactions within this system but reductive approaches do not consider the dynamical properties of the system as a whole.We compared simulations of the model with experimental data and found them to be largely in accordance, though the hrpV node shows some differences in state changes to that expected.

View Article: PubMed Central - PubMed

Affiliation: The Sainsbury Laboratory, John Innes Centre, Norwich, United Kingdom. dan.maclean@tsl.ac.uk

ABSTRACT
The Type III secretion system (TTSS) is a protein secretion machinery used by certain gram-negative bacterial pathogens of plants and animals to deliver effector molecules to the host and is at the core of the ability to cause disease. Extensive molecular and biochemical study has revealed the components and their interactions within this system but reductive approaches do not consider the dynamical properties of the system as a whole. In order to gain a better understanding of these dynamical behaviours and to create a basis for the refinement of the experimentally derived knowledge we created a Boolean model of the regulatory interactions within the hrp regulon of Pseudomonas syringae pathovar tomato strain DC3000 Pseudomonas syringae. We compared simulations of the model with experimental data and found them to be largely in accordance, though the hrpV node shows some differences in state changes to that expected. Our simulations also revealed interesting dynamical properties not previously predicted. The model predicts that the hrp regulon is a biologically stable two-state system, with each of the stable states being strongly attractive, a feature indicative of selection for a tightly regulated and responsive system. The model predicts that the state of the GacS/GacA node confers control, a prediction that is consistent with experimental observations that the protein has a role as master regulator. Simulated gene "knock out" experiments with the model predict that HrpL is a central information processing point within the network.

Show MeSH
Related in: MedlinePlus