X-Message-Number: 26878 From: Date: Wed, 24 Aug 2005 08:11:51 EDT Subject: Uploading technology (1.iv.1) The synaptic cleft. Uploading technology (1.iv.1) The synaptic cleft. At first, it is simply an empty space between the presynaptic button at the axon end and the postsynaptic dendrite part. The first hint that there may be something comes from the difference in gap distance between excitatory and inhibitory synaptic clefts. Inhibitory synapses use a large repertory of second messengers such polypeptides. These have no reuptake systems and so are free to diffuse at a distance outside the synapse. This is not simply a background noise. One neuron could modulate the action of other neurons sending axon terminal in the same area by amplifying or reducing the target neuron sensitivity to a given neurotransmitter. Some synapse clefs may be "leaky" by desing, a synapse could even have as main task to be a regional distributor of second messengers. In a simulation, the synaptic cleft may be so endowed with a leak parameter. Another property may be the back propagation of uncommon neurotransmitter from the postsynaptic side to the presynaptic one. The NO and CO gas seem to work in that direction. A narrow clef would be useful here, contrairy to the leaking case. When some high activity brings an amplified P0 release probability for a long time, it could be best to amplify the postsynaptic sensitivity by growing here more receptors. Here must then be a signal telling the presynaptic side to reduce the number of active release sites. The decay of old postsynaptic channels could then be coupled with the gas release, telling the other side that P0, the basic probability of vesicle release must be reduced. In the electronics simulation, this will translate into a threshold gated back signal to P0. When a neuron fire an action potential, a part back propagate in some elements of the dendritic tree, that may contribute to the depolarization of some synapses, particularly the ones on dendritic spines. By a capacitive effect, there may be an hyperpolarization of the presynaptic membrane or a depolarization by conduction. Depending on the conductivity of the cleft medium, one or the other effect may predominate. Assume the conduction win the contest, there will be a predepolarization of the button by back propagated AP. If a new AP gets there at that instant, it will be amplified. This is an implementation of the Hebbian rule that neuron firing toghether gets more strongly linked. There is no need to compute the back propagation of the action potential at each computing cycle, it suffice to create a Hebb parameter for each synaptic button. The back propagation giving the parameter value is computed only one time at the start. After that, each time the postsynaptic neuron fire an AP, the Hebb parameter is activated some milliseconds later. To be sure, that activation could be canceled by a down stream GABA-A synapse activation cutting the back propagation. It has been pointed out that a uploaded brain on an electronics support would have no emotion, no hormonal feeling. In fact, hormones work as modulatory neuromediators at the inhibitory GABA synapses, entering it at the cleft border. This is the reverse path of the diffusing peptide disperser. These synapses may have an hormonal entry parameter. To summarize, four parameters have been linked to the synaptic cleft domain: Peptide diffusion outside the cleft, gas diffusion inside the cleft space, electric effects defined by an Hebbian parameter and hormones entry in the synapse. Another set of parameters may be worked out in the same way in the electronics domain, even if on the biological side, it is something really different. This is about gap junction or "electric synapses". These are channels between cell membranes in direct contact. There are tens of such trans-membrane molecular complexes and one complete channel is built from two such complexes, one for each membrane, the number of potential combination is so in the hundreds at least. Contrairy to "real" synapses, gap junction don't produce any signal amplification, only direct coupling from a cell to another. On a broad basis, such channels can be divided into a number of classes: 1/ The large ones, without electric selectivity: They are a gate for molecules with mass up to one kilo Dalton (1000 nuclear mass units), for example the simplest amino-acid glycine is in the sixty daltons range. cAMP, ATP, Adenosine are neuromodulators able to take that path. The channel may be directly between two neurons or use a nearby glial cell as an intermediary. These channels are two ways and so there may be a retroaction from the post synaptic side to the presynaptic one. 2/ The electric ones, for smaller molecules and ions, there are three main properties: a) They may or may not be anion - cation selective. b) They may or may not be rectifying ( let the current flow in one direction only, as seen in an electronics diode). c) They may open at a given potential threshold between the two membranes. As an example, assume an action potential at a presynatic button falls to elicit a neurtransmitter release, simply because this process is governed by a probability function. A set of gap junctions could neverthless send to the postsynaptic side a partial depolarising signal. If this is an isolated event, there will be nothing more, it will be lost in the background noise. On the other side, if some nearby synapses have fired a postsynaptic potential (PSP) in the dendrite at the same time, the depolarisation back propagating in the postsynaptic part may be combined with the gap junction potential and the sum may then trigger the Na+ or Ca++ potential sensitive channels, so the synapse would fire a PSP even without any neuromediator! Another possibility, using rectifying property would be a dendrite back propagating signal turnig a brief, weak AP at the presynaptic level into a strong, long one. There is also the opposite effect : The dendrite potential reducing the presynaptic firing probability or even forbiding it. These may turn out quite involved: There are synapses on spines with excitatory properties controled by a GABA-A inhibitory one. A back propagating dendrite current could then silence the GABA-A silencer! Gap junctions are not limited to "true synapse" periphery, they may be distributed directly on the dendrite trees and modulate their properties. This is true too for the cell's soma body. The soma potential may be distributed to tens of nearby cells, may be 20 to 50. This would alter their basic potential and so their AP firing probability. This can produce strong collective activity, excitatory or inhibitory. Yvan Bozzonetti. Content-Type: text/html; charset="US-ASCII" [ AUTOMATICALLY SKIPPING HTML ENCODING! ] Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=26878