The community effect

The 'community effect' is an important mechanism in the demarcation of different domains in the developing embryo.

When the activation of a particular gene in a particular cell leads to the production and secretion of a signalling factor that, directly or indirectly, stimulates the expression of the same gene in the original cell and it neighbours, a positive feedback loop is created (see also the Simple Positive Feedback page in the Theory section). Activation, even to a very low extent, of this gene in one of the cells in the community, will 'lock on' the gene in the all cells within the community, after a delay determined by the time required for the signal to exert its effect.

This type of effect can be simulated in NetBuilder:

In this diagram, the positive feedback loop is very clear, but it must be kept in mind that the donor-receptor pair represents an intercellular connection. Here, the intercellular connection have been defined as follows: signals go from the blue cell to its neighbours, the red and the green cells, from red to blue and orange, etc. The table below shows the parameter settings for the network elements that were used in the simulation below.

The diagram above is the most concise representation of the network, but it is also possible to draw - and simulate - it as in the picture below:

Here, the network has been represented as four identical sub-networks in the one 'cell' in the bottom right hand corner ('cell' in the NetBuilder meaning of the word). Connections all go from this cell to the same cell. The blue, red, orange, and green ellipses containing the sub-networks have no function in the simulation, they have been added to emphasise the correspondence between both pictures. The simulation results are identical for both representations, provided, of course, that the parameter settings are equivalent. The simulation results are visualised most clearly using the second representation:

At the start of the simulation, the black 'driver' TF activates the green driver gene  (to about 15% of its maximum expression level).

After several evaluation rounds, the combined effects of the driver (whose output is constant) and the feedback loops have managed to reach the level above which the summation function (which simply adds the contributions from both donor-receptor systems) shows up as active (a more or less arbitrary value, chosen by the user, here: 1.0).

The system goes to its final level more and more quickly, until the the system is fully "locked on" - a stable state.

The driver gene can now be switched off without any consequences for the overall state of the system. The signal is now consolidated.

A more precise picture of what has happened during the simulation is found in the graph below, in which the activities of the driver and the blue genes in the various cells are plotted as a function of time (or rather, evaluation round).

Several things are worth noticing:

1. The output of the blue gene reaches the receptors 10 evaluation rounds after the gene was activated. This delay may represent delays caused by translation, transcription, transport, etc. The delays are linear, and do not affect the sigmoid appearance of the curves.

2. The system will not lock on if the output of the driver gene is below a certain threshold. In this example, the output of the driver gene must be above 0.1. Such a threshold would guarantee that the system cannot be locked on by 'noise' (e.g. small quantities of the driver TF), provided, of course that the noise does not exceed the threshold level.

3. The activation threshold will only be greater than zero when the power parameter of the feedback TF (see the Table above) is greater than 1 (see  the Simple Positive Feedback page for the characteristics of these positive feedback loops). The power parameter may be interpreted as a Hill coefficient, which is greater than 1 if the binding curve of the feedback TF is sigmoid. This, in turn, only happens when the TF is a homo-oligomer (a homo-dimer for instance). Thus, this model (and its most simplified version on the the Simple Positive Feedback page) shows that TFs involved in this kind of positive feedback processes are likely to be homo-oligomers.