| Bacterial Chemotaxis in Silico | |||||
|---|---|---|---|---|---|
| Home | Biology | Software | Data | Publications | Links |
One of the advantages of computer modelling is that it forces you to realise what you do not know. Scientific illustrators are familiar with the fact that any description of a molecular process in words, no matter how clear, always leaves many questions unanswered. "How many copies of this molecule are there? How are they arranged? Will molecule X be large enough to reach moelcule Y?" Questions such as these arise because of the limitations of written language to convey spatial relationships.
An even larger gulf exists between the dynamic properties of systems of molecules and our ability to represent them. Flow charts and circuit diagrams with boxes and arrows are often used to present data acquired in biochemical experiments. But we delude ourselves if we think that they contain everything we need to know. Anyone who has tried to simulate a system of more than a few enzymes or signalling molecules on a computer knows not only that you need quantitative data on rates, concentrations and mechanisms, but also that the data that you have at present is probably not enough.
The following is a list of questions that came up in the course of our attempt to model signal transduction in the E. coli chemotaxis pathway. They mainly concern the molecular details of the process, but one could make similar other lists of questions relating to the genetics, physiology, or behaviour of E. coli or other chemotactic bacteria.
Do the chemotactic receptors exist in two well-defined conformational states, or in a range of multiple states?
How does ligand binding affect the conformation of a receptor? Does it freeze the receptor into one state for as long as it is bound? Or does ligand-binding shift the balance between two rapidly equilibrating conformations?
How are conformational changes in a receptor passed onto CheA? Is there rigid coupling between "active" receptor and "active" CheA? Or is there loose coupling between two autonomous free-running, thermally driven equilibria?
In any case, how does methylation of receptors modulate the transmission of signals? Is it by changes in Kd, or by changes in receptor-CheA coupling?
How do the four methylation sites per receptor differ in their effect? Are they methylated and demethylated in a defined sequence? How heterogeneous are the methylation states of receptors in a cell?
Do CheR and CheB bind selectively to certain conformations of receptors? What are their Kds for binding to the methylation sites? Do they influence transmission of signals independently of their affects on methylation?
How do receptors aggregate into clusters? Is it an active or a passive process? Is it affected by or dependent upon ligand binding, or adaptation?
What interactions between receptor, CheA and CheW enable clusters to form? Is a cluster based on a regular (crystallographic) arrangement of proteins — if so, what is it?
Are CheR and CheB confined to an "adaptational" compartment? If so, what are their local concentrations and how do they move from receptor to receptor?
What is the significance of CheB phosphorylation? What is unphosphorylated CheB for?
What are the dynamics of conformational changes, methylation and phosphorylation within a receptor cluster? Do they show periodic or oscillatory behavior? Is it possible that large-scale multimolecular patterns develop?
How do bacteria respond to mixed signals (ligands of different types changing with different time courses)? Does a cell prioritise certain patterns of signal events? If so, what is the molecular basis of this selection?
Why is CheA a dimer? What is the function of its flexible tether regions? Does CheA operate by a two-stroke mechanism? If so, why?
What is the significance of the short variant of CheA? Why does a hybrid dimer of CheA long plus CheA short have greater activity than a normal homodimer?
Is the CheY binding domain of CheA positioned so that it faces the cytoplasm? If not, is CheY diffusion to CheA restricted?
Does CheY move through the cell by simple diffusion? Could it be selectively bound to membranes, or utilize privileged channels within the cytoplasm?
Are there gradients of CheYp in the cell? If so, how are they set up and what influence do they have on the cell's response? Can the motor itself perturb local CheYp concentration depending on its direction of rotation?
How large are the fluctuations of CheYp produced at the receptor cluster? Are they sufficiently great to influence switching at the flagellar motors? If so, over what distance are the fluctuations averaged?
How long does it take a change in CheYp to be transmitted from CheA to a motor? Does it matter where in the cell the motor is?
Does the localisation of CheZ change dynamically? Is this due to oligomerisation? Does oligomerisation alter the activity of CheZ?
Do transient changes in intracellular calcium ion accompany tumbles? If so, how are they triggered and what effect do they have downstream?
How closely coordinated are the different motors of a cell? Is coordination achieved by CheYp concentration, by hydrodynamic interactions between flagella, or by calcium signals? Is it possible that switching events are somehow relayed between motors?
What determines where motors are inserted into the plasma membrane? Are they randomly positioned? Does motor position affect swimming?
What generates random switching of the motor? Is it an intrinsic thermally driven change in conformation or the entire motor? Or is it a consequence of reaction-binding events at the C ring, coupled by conformational spread?
| Site designed by Matthew Levin. This page updated 8 December 2006. |