``Brain Cells, Silicon Chips Are Linked Electronically Part-Mechanical, Part-Living Circuit Created..''
http://www.biochem.mpg.de/mnphys/projects/abstracts/01zecfro/abstract.html http://www.biochem.mpg.de/mnphys/
Hinrich Kuhls
-----------
This IS fascinating stuff. I wonder why they need to fence in the cells? Is it because the little buggers will try to crawl off the plate in order to avoid the currents? Or would they try to cling together in a little huddle fearing the great out there? Or would they crawl around randomly as autonomous individuals at night when the grad students have gone home?
In any event, the extensions of axons (or dendrites or any micro-extension) is performed within the cytoplasm from clusters of microtubules (tubulin) and these tubulin polymerizations are mediated by their surrounding cytoplasmic cations--which are in turn controlled by the nucleus through various intermediary pathways.
In work on fetal neuron development, (I think?) it was shown that neuron development and its synapse seeking features are partly controlled through electromagnetic field potentials in a pulsed rhythm (wave forms). Which means that developing neurons might use the changing field potential lines to map their growth toward each other from distant and non-contiguous locations.
In the photo from the above quoted site, you can see the neural network of axions has formed what looks like a random web appearance in the center surrounded by the little fenced in cells. (I notice one or two have escaped their cages.)
It would be interesting to attempt to try to control the physical configuration of the these neural networks by attempting to use various micro-electric current pathways to see if the axions would follow these as traces in making their synaptic connections.
On the cytoplasmic molecular level what you are trying to do is guide the neuron microtubule armature (cytoskeleton) into forming axons in specific directions.
The use of external micro-currents effects the membrane. This micro-current stimulation is already routinely done (via patch clamp, voltage dependent ion gates) to open and close the membrane's ionic channels which in turn change the ionic balance between the inside of the cell and the outside surrounding medium.
Since it is the presence and localized concentration of calcium and other ions that is driving cytoplasmic microtubule polymerization, in theory (well, in chuck's theory) you could drive the whole process through external membrane polarization and then patterned and or pulsed micro-currents given a suitable electrolytic medium. Remember the Poisson-Boltzmann model last week?
Let's say a micro-em field is established with a specific and symmetric configuration, so that field lines traverse a set of poles in a little hexagon---like the star of david inscribed in a circle. Now locate the fences at each of vertices. By exciting the configuration, the different vertices are polarized relative to their images on the other half of the inscribing circle and are connected by the field lines.
Now place the cells in their fenced vertices and flip on the switch so that they would grow along pre-configured em field lines in a symmetrical axon network that would connect them together in some specific pattern or order. So that simulating cell (1) would always excite cell (3) via its axon net, (2) -> (4), and so on. This is the fundamental form of a permutation, which is also a discrete symmetric group order N!, which also forms the basis for a logic. In other words this would be a living computer. Cool huh?
To see a few pictures of microtubules and what I am trying to describe go here:
http://cellbio.utmb.edu/cellbio/microtubule_structure.htm
Then go here for more detail on microtubules and their various applications/functions in the cell:
http://www.ultranet.com/~jkimball/BiologyPages/C/Cytoskeleton.html
Chuck Grimes