my channel in synapse

Moderator: wwlytton

Post Reply
ori

my channel in synapse

Post by ori »

Hi. I'm a new user, and I don't catch NEURON systems so much.
I'm now trying to create a simple synaptic model.
One simple cell and one presynaptic compartment will be included.
Postsynaptic dendrite has AMPA channel I've already created with NMODL.

I have 2 questions.
1
When I execute the model, presynapse provide transmitter release and voltage goes up, but nothing happens in the postsynaptic dendrite.
2
The presynapse automatically stimulated although I don't put any stimuli in it. Here I write parts of the hoc file.

-------------------------------------
objref MainCell
MainCell = new EasyCell()
-------------------------------------

This cell has one soma and one dendrite.
I want to create synapse on the centre of the dendrite.

-------------------------------------
create PRE
PRE{
diam=2
L=2
insert pas
g_pas=0.0001
e_pas=v_init
}
-------------------------------------

This is the presynaptic compartment.

-------------------------------------
objectvar ch
ch = new CaGluSK() // create synapse
MainCell.dend ch.loc(0.5) // assign postsynaptic compartment
setpointer ch.C, PRE.T_rel(0.5)
-------------------------------------

CaGluSK is a channel I created.

Do you have any suggestions?
I would appreciate your kindness.
ted
Site Admin
Posts: 5784
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Post by ted »

The central question is whether your conceptual model of synaptic transmission demands
that the synaptic conductance is closely coupled to some variable (v, cai) in the presynaptic
terminal, or whether it is enough to watch the presynaptic cell for a spike, and then after
some delay produce a conductance change in the postsynaptic cell. The former requires
a synaptic mechanism that uses a POINTER variable so that the synaptic mechanism,
which is located on the postsynaptic cell, can "know" the value of a variable that belongs to
the presynaptic cell. The latter uses a NetCon that "watches" the presynaptic variable for
a threshold crossing; when threshold is crossed, the NetCon places an event in a queue
for delivery, after some delay, to the synaptic mechanism. The synaptic mechanism must
have a NET_RECEIVE block that tells it what to do when input events arrive (e.g. it may
make a synaptic conductance variable jump up to a new value, after which the
synaptic conductance decays toward 0).

It will be very helpful to read Chapter 10 of The NEURON Book, which presents detailed
explanations, plus several working examples, of different ways to represent synaptic
transmission. If you do not have the book, please read this early draft of that chapter
http://www.neuron.yale.edu/ftp/ted/book ... xedref.pdf
ori

Thank you

Post by ori »

Thank you very much for your reply and advices.
I'll try on it.
ori

Things I don't well understand

Post by ori »

Thank you again for replying me.
I worked on this problem based on your kind suggestions.
And let me pose an additional question.

Does it really need NET_RECEIVE block?
CaGluSK channel refers to an AMPA receptor channel.
It has following mechanisms:
Transmitter release
Glutamate concentration calculation
Ca++ pump
If there occur an action potential by stimulating presynaptic neuron, I think it would activate this channel automatically and then postsynaptic cell respond to it.
I'm afraid that there would be two transmitting functions if I create another mechanism which is involved in the NET_RECEIVE mechanism.

I consider the command below as a pointer to postsynaptic neuron:
MainCell.dend ch.loc(0.5)
Because ch refers to the CaGluSK channel object, I thought it would transmit changes in voltage and so on.

Thank you very much.
ted
Site Admin
Posts: 5784
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: Things I don't well understand

Post by ted »

ori wrote:Does it really need NET_RECEIVE block?
CaGluSK channel refers to an AMPA receptor channel.
It has following mechanisms:
Transmitter release
Glutamate concentration calculation
Ca++ pump
Sounds like a good example of how not to write NMODL code. The calcium accumulation
mechanism should really be specified in its own mod file, for the sake of easier
debugging and maintenance if nothing else. Where did you get this, and how do you
know that it works properly?

Assuming that it does work, and that you want to use it as is, then events are irrelevant
to it--you're stuck with POINTERs.
I consider the command below as a pointer to postsynaptic neuron:
Untrue on multiple counts. Have you read the material cited in my prior message?
ori

Post by ori »

I meant that NMODL of CaGluSK channel (ch) calls Transmitter release mod, concentration mod, and pump mod.

When I implement it between two simple somata, I could see channels opened, current flowed.
But when I replace the postsynaptic soma with an easy cell with a soma and a dendrite and do the same try as above, I see little changes in postsynaptic dendrite where the synapse should be connected.

Yes, I have the NEURON book and am reading it working on my simulator.
I'll try the way you advised me.

Thank you very much again.
ted
Site Admin
Posts: 5784
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Post by ted »

Without seeing the actual code that you're using, I haven't the foggiest
idea of what to tell you. You don't have to post it in the Forum--in fact, it's
probably best that you don't; the Forum is really only suited for short code
extracts. So if I am have a rational basis for further specific comments
I suppose you'll have to zip up the code and send as an email attachment
to ted dot carnevale at yale dot edu.

BTW, in my previous message I asked where you got the code you're using
(write it yourself, or pick it up somewhere else?), and how you know if it
works properly?
ori

the original model

Post by ori »

I appreciate your reply.

I've got an original model at ModelDB:
http://senselab.med.yale.edu/SenseLab/M ... odel=18198

First, I replace AMPA receptor channel with my original CaGluSK channel. And when running it, I saw almost the same graphs as those in an original simulation.

Then I created an easy cell with a soma and a dendrite as a postsynaptic neuron, and I created a synapse into a middle of the dendrite.
ted
Site Admin
Posts: 5784
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Post by ted »

NMODL of CaGluSK channel (ch) calls Transmitter release mod, concentration mod, and pump mod.

When I implement it between two simple somata, I could see channels opened, current flowed.
But when I replace the postsynaptic soma with an easy cell with a soma and a dendrite and do the same try as above, I see little changes in postsynaptic dendrite where the synapse should be connected.
In order to diagnose the problem and tell you how to fix it, I will have to
be able to reproduce it. If you zip up the hoc, ses, and mod files that are
necessary to reproduce the problem, and send them to me (ted dot
carnevale at yale dot edu), I will see what I can do. It may take a day
or two because I am at the SFN meeting this week.
Post Reply