5.5.4 Grid formation
My original intention for this experiment was to come up with something that modeled the formation of the retina in
Drosophila. However, that is a very complicated system and it would have required a considerable amount of
studying to get to the point where I could do this. I would still like to build a system that models retina formation,
but for right now I just wanted to demonstrate the ability of this developmental simulator to grow complex patterns.
So the goal here morphed into coming up with any set of genes that could produce a grid pattern of gene expression in
a set of cells. This was accomplished, but the resulting gene system is the most complicated one so far displayed.
2. Gene System
While the ultimate goal was to produce something that would generate a grid, this problem was first subdivided to try and
come up with a system that could produce a set of evenly spaced linear columns of cells that express genes that are not expressed
in the surrounding neighbors. After this was accomplished it would be easy to duplicate this so that two independent systems
operate simultaneously in the vertical and horizontal directions.
|Vertical Grid Chromosome|
|Figure 1. This figure shows the genes involved in producing a vertical grid of cellular differentiation.
3. Genetic Overview
The following steps describes the interactions among the genes that causes the grid to form.
Normally this initial system would already be partitioned by other processes. However, in this case we are only interested
in the genes directly involved in creating the grid so I manually partitioned it. Cells around the outer edges had all of
their genes deliberately shut off so they could not take part in the action. Then the cells on the right, vertical boundary
were injected with the membrane ligand IBL (Inhibition Band Ligand). The core cells were producing the receptors for this
IBR. In a more natural system this would be equivalent to the right section inducing the center section to begin grid formation.
When IBL and IBR mate they produce IBC (Inhibition Band Control).
IBC builds up in those cells where it is being expressed and it turns on two other genes. The first one that is turned on is
the G_DBL gene. It produces the protein DBL (Diffusible Band Ligand). This diffusible ligand then begins spreading out from
the cells where it is produced. All of the cells in the center section start out producing the DBR (Diffusible Band Receptor)
protein. The second gene that is turned on is the G_IBOffR gene that produces the IBOffR (Inhibition Band Off Receptor) protein.
This protein will be discussed more later.
At this point the diffusible ligand DBL is spreading out in both directions from the initial cells that are producing it and
building up in neighboring cells. On the left of the initial cells these ligands come into contact with the DBR receptors and
begin producing the transcription factor DBT (Diffusible Band Transcription). On the right of the initial cells however nothing
happens. This is because the relevant genes on that side have been turned off. So even though the ligand is also pouring into
those cells it has no affect.
When DBT has built up in the neighboring cells to a sufficient quantity it begins to up-regulate the gene G_DBCOn which produces
the protein DBCOn (Diffusible Band Control On). Once DBCOn builds up in sufficient quantity it switches several genes. To put it
into plain words though this acts like a switch to turn off the receptors in this new region of cells, and it turns on the production
of the membrane ligand to begin inducing a new band to begin forming. The receptors need to be turned off because these cells are basically
done and we do not want the next grid line formation to effect the previous work. It begins by switching G_IBR and G_DBR off. These
are the receptors for the two main ligands. It then switches on the G_IBOnC gene which produces the gene control protein IBOnC
(Inhibition Band On Control).
At this point you might be asking "Why not simply directly turn G_IBL on? Why go through this indirect route with IBOnC?" The problem
is that we do not want to immediately turn the membrane ligand on. We need to give a little time for the receptors to fade away before
we begin producing the ligand. Otherwise the ligand in the cells to the right of this cell would interact with it and begin the process
over again. We need to wait just a bit for the cells to fully settle down so that the new grid column that is induced is the one to the
left. Using IBOnC acts as a simple delay. It takes some time for IBOnC to build up to a sufficient quantity to turn the G_IBL gene on.
When it does then this basically starts back at step one with the only difference being that we have made one grid column.
We also need to mention IBOffC. When G_DBL is turned on it starts cranking out diffusible ligand at a high rate. The gene control
IBOffC again acts as a delay so that the diffusible ligand is only produced for a specific period of time. The spacing of this grid
is controlled by how far the diffusible ligand spreads. You can think of IBOffC as an alarm clock on an oven. It allows the diffusible
ligand to be produced for a limited amount of time and then shuts off that production. How far the ligand spreads at that point is
determined by its diffusion and degradation rates. This is a typical place where evolution could work to change the spacing pattern
seen in the formation of the grid. If the ligand diffused faster, the clock ran longer, or the ligand degraded slower it could
potentially affect the size of the grid columns that are formed.
4. Grid Results
|Vertical Grid Formation|
|Video 1. This video shows the formation of a vertical grid of cellular differentiation.
|Combined Grid Formation|
|Video 2. This video shows the formation of a complete grid of cellular differentiation.
The vertical and horizontal components are made up of similar genes that operate independently. So in
essence it is creating a vertical grid and a horizontal grid that just happen to overlap.
While this system is not an explicit model of the formation of the retina of a fly, it does demonstrate some pretty powerful
abilities to form complex patterns. The initial growth of the grid is induced by neighboring cells. And that growth is then
unidirectional so that once a portion of a grid has formed then it is set and then induces the formation of another grid column.
This property of one section of cells inducing neighboring sections to change is one of the key mechanisms of developmental biology.
Also, by altering the properties of this system it should be possible to create grids of different dimensions. This will allow
evolution to alter the patterns found in the developing brain for better or worse.