action potential initiation site

Anything that doesn't fit elsewhere.
gcm
Posts: 2
Joined: Tue Jun 21, 2005 10:37 am
Location: Minneapolis, MN, USA

action potential initiation site

Post by gcm »

What is the simplest way to identify the site of action potential initiation? (aside from visual inspection using space plots or voltage plots; would like a method to cycle through several neurons).
ted
Site Admin
Posts: 5784
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: action potential initiation site

Post by ted »

What is your operational definition of "site of action potential initiation"? If you don't have one, what operational definitions have been proposed/used in published articles?
gcm
Posts: 2
Joined: Tue Jun 21, 2005 10:37 am
Location: Minneapolis, MN, USA

Re: action potential initiation site

Post by gcm »

It would be the site where the membrane potential exceeded threshold (i.e. passed for example 0 mV) and also resulted in the generation of a propagating action potential down the axon.
ted
Site Admin
Posts: 5784
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: action potential initiation site

Post by ted »

gcm wrote:It would be the site where the membrane potential exceeded threshold (i.e. passed for example 0 mV)
Then this should be an easy question: what is the voltage threshold for spike intitiation in a spherical cell with Hodgkin-Huxley currents?
and also resulted in the generation of a propagating action potential down the axon.
Not sure if this is a useful criterion. Imagine a cell with a spherical soma. A dendrite is attached to one side of the soma, and an axon is attached to the other side. The geometry of the cell is:
soma L = diam = 10 um
dend L = 500 um, diam = 2 um
axon L = 500 um, diam = 1 um
Ra is 35.4 ohm centimeter, and cm is 1 uf/cm2.
soma and axon both have Hodgkin-Huxley mechanism, and dend is passive with membrane conductance 0.0001 S/cm2, reversal potential -65 mV.
A current clamp is attached to the middle of the soma, where it injects a 1 ms duration current pulse that starts at t = 1 ms.
Pulses in the range of ~ 0.36 to 0.4 nA elicit spikes that start at the _distal_ end of the axon, and propagate back into the soma.
Hyeon
Posts: 22
Joined: Thu Jul 19, 2012 3:43 am

Re: action potential initiation site

Post by Hyeon »

Hi, I have similar question about 'action potential (AP) initiation site'

As you mentioned, if I define 'AP initiation site' as where is the site within a neuron APs are generated (broadly known as axon initial segment), how I find the exact site of segment?

In my situation, I reconstructed layer 5 pyramidal neuronal models, and then applied extracellular stimulation with 0.1 ms duration.
I found that axon is more excitable than soma and dendrites. Then I want to know where is the site evoking AP initially such as either axon terminal or initial segment.

So, is there any function to find the time when membrane voltage is exceed some threshold (ex. 0 mV) at each segment?

Best regards,
Hyeon
ted
Site Admin
Posts: 5784
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: action potential initiation site

Post by ted »

Hyeon wrote:if I define 'AP initiation site' as where is the site within a neuron APs are generated (broadly known as axon initial segment), how I find the exact site of segment?
First provide an operational definition of "AP initiation." This is not a trivial question. gcm answered that the spike intiation site "would be the site where the membrane potential exceeded threshold (i.e. passed for example 0 mV)" but that is both arbitrary and incorrect. When I replied with the statement
"this should be an easy question: what is the voltage threshold for spike intitiation in a spherical cell with Hodgkin-Huxley currents?"
I did so with a sense of irony, because neither the Hodgkin-Huxley spike mechanism nor biological neurons have "a threshold". It's not like there is some magic number v_threshold such that, if membrane potential rises above v_threshold, a spike will occur. Instead, threshold changes dynamically, depending on fluctuations of membrane potential from one moment to the next (e.g. see Platkiewicz and Brette
A threshold equation for action potential initiation
PLoS Computational Biology, 2010, 6, e1000850
and
Impact of fast sodium channel inactivation on spike threshold dynamics and synaptic integration
PLoS Comput Biol, 2011, 7, e1001129
for recent discussions).

So: if you could observe the time course of all membrane currents and potentials over an entire neuron, how would you know that a spike is starting at a particular point?
Hyeon
Posts: 22
Joined: Thu Jul 19, 2012 3:43 am

Re: action potential initiation site

Post by Hyeon »

Thank you, as you mentioned, I read some related papers, and I realize that I misunderstand the definition of action potential initiation site.
I found that what I want to find is, the axonal compartment having the largest depolarization at the end of the stimulus pulse.
and they call it as site of action potential initiation.
ted
Site Admin
Posts: 5784
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: action potential initiation site

Post by ted »

Hyeon wrote:I found that what I want to find is, the axonal compartment having the largest depolarization at the end of the stimulus pulse.
Would it help to see a plot of membrane potential vs. distance, and watch how it changes over time? For an example of such a graph, use NEURON to run neurondemo (in MSWin just double click on the neurondemo icon, under Linux execute the command
neurondemo)
Then click on the radio button labeled Stylized, and finally in the Movie Run windo click on Init & Run.
Hyeon
Posts: 22
Joined: Thu Jul 19, 2012 3:43 am

Re: action potential initiation site

Post by Hyeon »

Yes, by seeing plot of membrane potential vs. time in neurondemo, I observed soma is more depolarized than node[5].

In my situation, since I simulate more than 20 thousand numbers of neuronal models, it is hard to check one by one by graph.
is there any way to record peak membrane potential in a particular section (for example, soma) of a neuronal model?
ted
Site Admin
Posts: 5784
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: action potential initiation site

Post by ted »

You may find this helpful--detects max and min v in all segments of the section(s) into which it has been inserted.
extrema.mod
viewtopic.php?f=16&t=982&p=3967#p3967
ted
Site Admin
Posts: 5784
Joined: Wed May 18, 2005 4:50 pm
Location: Yale University School of Medicine
Contact:

Re: action potential initiation site

Post by ted »

ted wrote:You may find this helpful--detects max and min v in all segments
But now I'm not so sure it will help you discover the site at which a spike starts. The peak amplitude of a spike can vary throughout a cell, and there is no guarantee that it will be largest at the point of spike initiation. The sealed end effect is proof of this. You need to observe the time course of membrane potential or some other variable related to spike initiation over a range of anatomical locations, to find the place at which the spike first begins.
luciana
Posts: 17
Joined: Sun Mar 09, 2014 2:29 pm

Re: action potential initiation site

Post by luciana »

Dear Ted,
I modeled this simple example that you proposed.
ted wrote: Imagine a cell with a spherical soma. A dendrite is attached to one side of the soma, and an axon is attached to the other side. The geometry of the cell is:
soma L = diam = 10 um
dend L = 500 um, diam = 2 um
axon L = 500 um, diam = 1 um
Ra is 35.4 ohm centimeter, and cm is 1 uf/cm2.
soma and axon both have Hodgkin-Huxley mechanism, and dend is passive with membrane conductance 0.0001 S/cm2, reversal potential -65 mV.
A current clamp is attached to the middle of the soma, where it injects a 1 ms duration current pulse that starts at t = 1 ms.
Pulses in the range of ~ 0.36 to 0.4 nA elicit spikes that start at the _distal_ end of the axon, and propagate back into the soma.
The result is the axon fire first, then the soma and finally, the dendrite. However, if I disconnect the passive dendrite from the soma, then the soma fire before than the axon. Why is that?
In this situation (without passive compartments), exists any way of making the soma fire after the axon just because it is bigger? Considering that the loading effects may delay the occurrence of the action potential in this giant structure, the soma.

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

Re: action potential initiation site

Post by ted »

luciana wrote:if I disconnect the passive dendrite from the soma, then the soma fire before than the axon. Why is that?
In this situation (without passive compartments), exists any way of making the soma fire after the axon just because it is bigger? Considering that the loading effects may delay the occurrence of the action potential in this giant structure, the soma.
Good questions. Computational experiments might help answer them. What happens to the soma+axon model if you reduce the stimulus current amplitude? And what happens if you change the soma.diam/axon.diam ratio (by making soma.diam larger or smaller)?
luciana
Posts: 17
Joined: Sun Mar 09, 2014 2:29 pm

Re: action potential initiation site

Post by luciana »

ted wrote:What happens to the soma+axon model if you reduce the stimulus current amplitude?
The time between the stimulus and the action potential is bigger with lower current amplitudes. In some point the current in not suffiecient to generate the action potential.
amplitud = 0.1 nA, soma and axon both fire
amplitud = 0.09 nA, none fire
But in every cases the time between somatic action potential peak and axonal AP peak is the same, somatic peak before always.
ted wrote:And what happens if you change the soma.diam/axon.diam ratio (by making soma.diam larger or smaller)?
With the stimulus current amplitude fixed in 0.1 nA. I changed soma.diam from 10 to 20 and the results were:
soma.diam <= 15 um, soma fires before than axon
soma.diam = 16 um, axon fires slightly before than soma!!!! :O
soma.diam >= 17 um, none fire

So, eventually this may happens, but in a very thin range of parameters? And it is necessary to stimulate with the lowest current amplitude possible for reproduce this results?
Thank you Ted, this is great!

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

Re: action potential initiation site

Post by ted »

luciana wrote:soma.diam <= 15 um, soma fires before than axon
soma.diam = 16 um, axon fires slightly before than soma!!!! :O
soma.diam >= 17 um, none fire

So, eventually this may happens, but in a very thin range of parameters? And it is necessary to stimulate with the lowest current amplitude possible for reproduce this results?
What happens if you make soma.diam even larger? Does the range of stimulus amplitudes that make the axon fire first become wider or narrower?
Thank you Ted, this is great!
I'll take credit for making suggestions, but the credit for doing all the work and making the experimental observations is yours.
Post Reply