Talk:Action potential

Messy article

The article is extremely confusing, there is a lot of material and definitions which are, however, poorly organized. A lot of repetitions and the flow of argument is completely absent. I'll see what i can do.
It is incredible that it is a FEATURED ARTICLE with so many problemsRvfrolov (talk) 18:32, 2 January 2009 (UTC) 2 January 2009

Reorganizing the field of MEMBRANE POTENTIALS

There are three articles now which discuss very much the same things with numerous repetitions and description of the same material in different terms.

• Resting potential
• Membrane potential
• Action potential

As it was already suggested by Methoxyroxy 12:37, 2 November 2006 (UTC), it needs a really big clean-up and optimization. There is a lot of confusion there so I will do this albeit not at once. I will move different parts between these three articles, edit and unify their style etc. At later stage I will need someone who is native English speaker to do spellcheck.Rvfrolov (talk) 20:52, 2 January 2009 (UTC)

I have replied at Talk:Membrane potential. Looie496 (talk) 22:01, 2 January 2009 (UTC)

Hi Rvfrolov. The problem is that explaining an action potential, without first describing what the 'mV' part of membrane potential is, can lead to a whole lot of problems down the road. I'm all for reorganizing it, just very carefully Paskari (talk) 15:41, 24 June 2009 (UTC)

Propagation is not key to action potentials. Simplify.

The current version of the article mixes action potentials with propagation of potentials creating excessive complexity and inaccuracies.

Potential propagation is not a required property of an action potential. Most textbooks first define the action potential in an isopotential cell. In Hodkin and Huxley's experiments, for example, a wire was strung along a squid giant axon to shunt axial currents effectively producing an isopotential membrane compartment. With this preparation, current-clamp experiments still produce action potentials along the entire fiber, without a "wave of electrochemical activity." Neither does a cell need to carry APs over a distance to make use of action potentials (e.g. electrocyte action potentials in electric fish, and many other cell types produce action potentials for other reasons than long-distance signaling).

The current definition is also missing a key defining component: the key role of voltage-sensitive conductances.

Propagating action potentials may be more consistently described as a continuous succession of local action potentials triggering action potentials in the adjacent sections. This distinction would help avoid some of the current inaccuracies in the article.

For example, the current article describes saltatory conduction as follows "Since the axon is insulated, the action potential can travel through it without significant signal decay." In reality, myelinated stretches of axons do not produce action potentials and the action potential does not "travel through it." It would be more accurate to state that depolarization from an action potential at one node propagates passively to the next node and triggers an action potential at the next node. The signal may decay significantly between nodes and still trigger an action potential at the next node. The same section seems to imply that the action potential must be generated at the synapse for the release of neurotransmitter, which is also inaccurate. Any depolarization of sufficient magnitude (passive or active) will have a similar effect.

In summary, to make the article more useful, I recommend providing a complete and general definition of the action potential with minimal extraneous detail. The adjacent topics such as "Electrotonic propagation of potentials", "Neurotransmitter release", "Cable theory", "Saltatory conduction", probably belong in separate articles or sections.

Yatsenko DV (talk) 03:45, 27 July 2009 (UTC)

Thank you for your attention to this, and for your very thoughtful analysis. In large part, I agree with you. Your analysis of propagation seems to me to be correct. I have felt for some time that the lead section of the page is too long and rambling, and should be shortened, and I think it would be an improvement to simplify it per WP:LEAD and per your comments. I also think errors in the explanation of action potentials, of the sort you identified, should be corrected as they come up throughout the text. However, I'm not so sure about splitting off separate articles. The topics you list are not all sections in the current page, and the sections related to these topics do begin with main article links already, and it is appropriate to discuss each of those topics in this page (with the possible exception that discussion of the passive propagation of subthreshold graded potentials should be limited to their relevance to action potentials). Therefore, I would tend to favor simplifying the lead, and correcting errors elsewhere, but not necessarily splitting off any new pages or merging material here to other existing pages. --Tryptofish (talk) 15:44, 27 July 2009 (UTC)

I'm pretty much in line with that response. On one side, this article has become much too large and disorganized, probably because of the lack of anybody actively maintaining it. So simplifying the article would be a good thing. On the other side, propagation is probably the reason why action potentials exist, so omitting any discussion of it at all would be bad. Neurotransmitter release, however, might not belong here beyond a brief sentence or two to explain what happens when an action potential arrives. Looie496 (talk) 21:41, 27 July 2009 (UTC)

Just to clarify: what both I and Yatsenko agree is that propagation of the action potential, in the sense of an intact action potential just traveling along, is presented misleadingly. On the other hand, propagation in the sense of passive propagation of a depolarization which then brings membrane potential to threshold, activating voltage-sensitive ion channels and regenerating an action potential, is correct, and nobody wants to omit that. (Did that clarify, or make it worse? (smile)) --Tryptofish (talk) 21:51, 27 July 2009 (UTC)

Propagation is not a necessary feature of an action potential and should be relegated to a later section. Otherwise the definition is simply incorrect because it would not apply to the phenomenon that was described by Hodgkin and Huxley. Action potentials can be and are produced without propagation in many experimental preparations. Voltage-gated channels (active conductances) are a required part of an action potential and should be included in the primary definition. I would first define the action potential in the case of an isopotential compartment. Dimitri Yatsenko 00:58, 28 July 2009 (UTC)

Although the lead is now more accurate, it is way too long and too technical for a general readership. We need to look at ways to make the wording less technical, and to move parts of the lead into other parts of the page. --Tryptofish (talk) 19:02, 28 July 2009 (UTC)

"Nerve spikes"?

I am removing "nerve spikes" from the first sentence.

I do not believe that the term "nerve spike" could be generally applied to all action potentials. Action potentials are transient membrane voltage events in individual cells or their compartments. They happen in many cell types. Nerves are bundles of axons in the peripheral nervous system (no nerves in the brain or spinal cord). Thus "nerve spikes" are but one specific manifestation of action potentials. The same could be said about MUAPs (motor unit action potentials), for example. This does not make them synonymous with action potentials. Dimitri Yatsenko 01:25, 28 July 2009 (UTC)

I don't have much of an objection to removing the mention of spikes, which is a bit colloquial. However, I reverted your deletion of "nerve impulse." I did that because, first, it is a widely-used synonym (even though not all action potentials are in neurons), and secondly, because nerve impulse is an existing redirect to this page, and therefore the phrase needs to be bolded in the lead sentence. --Tryptofish (talk) 19:00, 28 July 2009 (UTC)

I have not seen the term "nerve spike" or "nerve impulse" used as a general synonym for action potentials in any modern scientific literature. An axon is not a nerve. Neurons are sometimes called nerve cells in less technical usage, but this is inaccurate and would not pass in a scientific paper. So I am conflicted between being precise or catering to general nontechnical usage of terms. I think we should strive to be technically accurate and, in so doing, influence the popular understanding of these natural phenomena. Dimitri Yatsenko 19:23, 28 July 2009 (UTC)

I wasn't disagreeing with you about spikes. Impulse is used widely in English. Besides: WP:R#PLA and number 7 of WP:NOTGUIDE. --Tryptofish (talk) 19:32, 28 July 2009 (UTC)

What if we separate the article for "nerve spike" and "nerve impulse" and explain that it is but one special case of action potential that is recorded from nerves in the PNS? Dimitri Yatsenko 19:31, 28 July 2009 (UTC)

No, not needed, and WP:CFORK. --Tryptofish (talk) 19:34, 28 July 2009 (UTC)

I defined "spikes" and "impulses" in separate sentences in the first section. What do you think? Dimitri Yatsenko 20:36, 28 July 2009 (UTC)

"Depolarization increases sodium current"?

The current wording in the second paragraph states that depolarization "increases both the inward sodium current (depolarization) and the balancing outward potassium current (repolarization/hyperpolarization)". I question the accuracy of that statement.

As the membrane depolarizes, the membrane potential moves toward the reversal potential for sodium. This reduces the electrochemical driving force for sodium. Unless the sodium conductance increases by a greater factor to compensate, the sodium current will decrease, not increase. So the statement is not generally accurate. I propose rewording it to state that both conductances increase and only when the net current is negative and leads to further depolarization, a positive feedback loop is generated to precipitate the action potential. Dimitri Yatsenko 21:07, 28 July 2009 (UTC) —Preceding unsigned comment added by Yatsenko DV (talk • contribs)

Sorry to disagree, but the present text you quoted is accurate, and the changes you propose are way too technical for this project. --Tryptofish (talk) 21:35, 28 July 2009 (UTC)

Fair enough. I agree that this simplification is accurate for relevant time scales, membrane potential ranges, and channel types. Dimitri Yatsenko 22:05, 28 July 2009 (UTC) —Preceding unsigned comment added by Yatsenko DV (talk • contribs)

Thank you for understanding. A lot of this is just a matter that we (Wikipedia) are writing for a general audience, and that puts limits on how technical or scholarly we can get. --Tryptofish (talk) 22:56, 28 July 2009 (UTC)

Simplification

In the "Quantitative models" section there are many references to things being simple or a simplification. While this section does have many references attached to it, there is no mention in the article of what these things are simpler than. That is, why are these things simple, and compared to what, and what would be more complex.

Reply to unsigned comment: What it means is that mathematical equations do not capture all the complexity of a living cell. I've tried to make it a little clearer, but I'm not sure whether there is any way of saying it better. --Tryptofish (talk) 23:25, 3 September 2009 (UTC)

Refractory Period Section Poorly written

The refractory period section seems like it was copy pasted from a text book which was written by a high school teach held at gun point. Perhaps we should consider updating it Paskari (talk) 23:30, 6 October 2009 (UTC)

I've taken a shot at rewriting it. Cellular stuff is not really my strength, so if I got anything wrong, I hope somebody will correct it. Looie496 (talk) 00:10, 7 October 2009 (UTC)

I'm just about to sign off, but I'll take a look at it tomorrow. --Tryptofish (talk) 00:14, 7 October 2009 (UTC)

That is much better, great job. Paskari (talk) 11:02, 7 October 2009 (UTC)

Yes, much better, thanks. I tweaked it a little further, not much. --Tryptofish (talk) 18:34, 7 October 2009 (UTC)

Concerning the lead

I've just attempted a pretty major rewrite of the lead, which I hope won't offend anybody. I thought the existing version was too hard for readers to understand -- it also contained a couple of minor errors. I also added a paragraph about the distinction between sodium and calcium spikes, which seems to me to be a very important point. Regards, Looie496 (talk) 20:08, 23 February 2010 (UTC)

I note that the lead from the last FAR was moved some time ago to the overview section. If this persists, we should justify the need for both a lead and overview section, and ensure that they are not redundant to each other. Geometry guy 20:39, 23 February 2010 (UTC)

The Overview section needs revision too, but it seemed to me that these changes to the lead were sufficiently "bold" that it would be better not to pile other changes on top of them before discussion. Looie496 (talk) 20:48, 23 February 2010 (UTC)

Reference removed: Alphascript Publishing

I have removed a reference from the lede that was to:

• Miller FP, Vandome AF, McBrewster J (2009). Cardiac action potential. Beau Bassin Mauritius: Alphascript Publishing. ISBN 6130098685.{{cite book}}: CS1 maint: multiple names: authors list (link)

Alphascript Publishing republished Wikipedia content. And the book in question republishes this article. The cover of the book can be seen on Amazon. This article is named on the front cover. (The format for Alphascript books is to list the WP articles contained therein on the front cover as part of the name.)

The person who owns the book can verify that this is republished Wikipedia content by looking at the copyright information inside the book itself. -- RA (talk) 13:16, 28 February 2010 (UTC)

The other danger is Alphascript also publishes academics' thesis if they convince them to sign their terms. Was the source republishing of wiki article or a thesis. So always double check before removing any references about whether they are wikipedia article or a thesis. Generally a quick way to check is searching product description in wikipedia. Kasaalan (talk) 13:29, 28 February 2010 (UTC)

It is VDM that also publishes academics' thesis if they convince them to sign their terms. All alphascript titles are wikipedia articles. My mistake. Kasaalan (talk) 19:14, 5 March 2010 (UTC)

Note: I have edited the previous comment, because it was added by altering the comment above it in a way that made this section impossible to understand without going back through the history. I hope that my revision has not changed the message. Looie496 (talk) 19:26, 5 March 2010 (UTC)

Ions and the forces driving their motion

The region with high concentration will diffuse out toward the region with low concentration. To extend the example, let solution A have 30 sodium ions and 30 chloride ions. Also, let solution B have only 20 sodium ions and 20 chloride ions. Assuming the barrier allows both types of ions to travel through it, then a steady state will be reached whereby both solutions have 25 sodium ions and 25 chloride ions. If, however, the porous barrier is selective to which ions are let through, then diffusion alone will not determine the resulting solution. Returning to the previous example, let's now construct a barrier that is permeable only to sodium ions. Since solution B has a lower concentration of both sodium and chloride, the barrier will attract both ions from solution A.

There is not a single citation about osmosis. Osmosis tells us exactly the contrary. Facts tells us the same thing as osmosis: The cited diffusion doesn't occur. The concentrations may be equilibrated by water movement and membrane is permeable to water through aquaporins or directly. Somasimple (talk) 05:28, 3 June 2010 (UTC)

Secondly,

If solution A is electroneutral THEN 30n+30p=0 (where n stands for negative and p for positive). If solution B is also electroneutral THEN 25n+25p=0. Considering an action from a compartment onto another one orders to consider all positive and negative charges that exist in the compartments.

So, there is NO electric flux OR electric field BECAUSE EACH compartment is neutral at start. Saying a compartment is neutral is saying that it can't exert any electric "thing" at all.

Conclusion: You can't get something that is the result of k(25p/30p) or k(30p/25p). That is mathematically and physically incorrect because you arbitrarily remove the negative charges without any scientific explanation. Somasimple (talk) 09:27, 3 June 2010 (UTC)

Can you fix it, or should that part be removed? Looie496 (talk) 00:52, 4 June 2010 (UTC)

Are you asking me to change the way how biology is taught ? This page remains for historical reason (Nobel prizes) but its contents is far from actual and accepted knowledge in Biochemistry for example. If the goal of wikipedia is to promote science then you must re-write the page but it will against the Biology community.Somasimple (talk) 06:10, 4 June 2010 (UTC)

Wikipedia articles are written by people like you and me. If you see errors in an article, and can back up the claim that they are errors by referring to reputable scientific publications, then you should feel free to rewrite the section in a way that makes it more correct. In this case, if you don't fix it, it's likely that nobody else who reads this will be able to. I certainly can't. Regards, Looie496 (talk) 17:04, 4 June 2010 (UTC)

This is my area of expertise more than it is Looie's, so I think I can help here. I think the page is correct about this, as it is written. There are several errors in what Somasimple has said here. First, this is not an osmotic phenomenon, in that we are not dealing with H₂O molecules moving along with the ions. Second, there are two factors driving ionic movement: electroneutrality or like-charge repulsion, as mentioned, but also entropy. Entropy will cause, in the quoted example, the ions to move from A to B. When they do, electroneutrality will be achieved when there are 25 plus 25 in A, and 25 plus 25 also in B. --Tryptofish (talk) 18:23, 4 June 2010 (UTC)

Several errors? Did I say it was an osmotic phenomenon? No! I just said there was NOT a single citation about it. Osmosis exists whenever there is a concentration change, just whenever! THEN OSMOSIS EXISTS WHENEVER SOME IONS MOVE. There must be some osmosis because it's a reverse diffusion.

If you put a cell (neuron is a cell) in an hypotonic solution, osmosis happens and the cell expands because the internal concentration decreases by water flux. It is a fact. This fact creates an error in "your" silent diffusion that occurs in the other direction.

I like, again, your "entropic electroneutrality". It is the first time I heard/read that a charge vanishes by entropy. A citation, a reference?

In our example, it was clear (at least for me) that the membrane was semipermeable thus the result is not the one you gave. "Returning to the previous example, let's now construct a barrier that is permeable only to sodium ions. Since solution B has a lower concentration of both sodium and chloride, the barrier will attract both ions from solution A."

The difference remains because negative ions remain in one side. It creates the membrane potential but you're right, it raises another big problem. You have now, a side that is negative and another that is positive and diffusion will have some problem to be achieved  ;-( Somasimple (talk) 05:19, 5 June 2010 (UTC)

I never said that entropy makes a charge disappear. I said that it can make it move. The reason that excitable cells do not shrink or swell due to hyper- or hypo-tonicity is that the ions that move across the membrane represent a very small fraction of all of the ions that are present (in real cells, though not in the example). You also might want to familiarize yourself with the Nernst equation. --Tryptofish (talk) 14:24, 5 June 2010 (UTC)

You do not reply at all. Where do the negative charge move in our example (the membrane is only permeable to Sodium)? Somasimple (talk) 10:14, 6 June 2010 (UTC)

If the membrane is permeable only to sodium cations, then anions do not cross the membrane at all. Consequently, there is a separation of charge, giving rise to a transmembrane voltage difference. --Tryptofish (talk) 15:17, 6 June 2010 (UTC)

If it was so simple. As you know, at molecular level (the level we are speaking of), distances are of importance. The electrochemical force you created comes between 2 compartments separated by a membrane which thickness is known as 5 to 7 nm. It means that anions and cations must be separated, in all compartments, by a distance that is always superior to the membrane thickness. If the distance is lower in any compartment then you have an "entropic" problem (in fact I call it a simple Coulomb force): anions or cations can't be attracted by the other side since the strength of the force coming from the other side is not sufficient. This limits the process to concentrations < to 5 mmol... Far from the concentrations that exists in cells. I think you might NOT consider the Nernst equation because it will give you some headaches with charge density and Conservation of Energy. Somasimple (talk) 05:32, 7 June 2010 (UTC)

Somasimple, consider a parallel-plate capacitor, which is how the cell membrane is represented in the Hodgkin-Huxley model. Note that the capacitance C is proportional to the area of the charged plates divided by their separation[1]. For a membrane thickness of about 5nm, you still have a significantly large area in which the ions are able to arrange themselves (even an impossible miniscule cell with a central body length only ten times the width of the membrane will have a "plate" area 100 times larger). The point is that capacitance will be very large, such that even if 5nm were a very large distance for Coulomb attraction (which it very much is not), it won't matter because there are so many ions able to line up along the membrane, just like in an ordinary circuit-board capacitor. And from there, depolarization occurs when you suddenly open the ion gates and sodium floods in, etc etc etc. SamuelRiv (talk) 17:21, 14 June 2010 (UTC)

I'm considering effectively a capacitor and you do not consider the distances that effectively exist between the ions in presence (here is link to physics principles). Even if the surface is enlarged then you decreases the charge density and it matters for capacitance : the less charge density you have the less tension you'll get. In our case ions can't be attracted from the other side: TOO FAR! --Somasimple (talk) 06:04, 15 June 2010 (UTC)

passive movements of positive ions?

The inward movement of sodium ions and the outward movement of potassium ions are passive

Let's describe all the events that happen simultaneously:

1/ Sodium movement balanced with chloride

sodium is inward and Na ions stick to the internal membrane, chloride ions stay out, and balance the Na charge, across the external membrane

2/ Potassium movement balanced with chloride

potassium is outward and K ions stick to the external membrane, chloride ions stay in, and balance the K charge, across the internal membrane

Now let's see what happens on each side:

1/ Internal side:

sodium is inward and Na ions stick to the internal membrane, chloride ions stay in, and balance the K charge, across the internal membrane

2/ External side

chloride ions stay out, and balance the Na charge, across the external membrane, potassium is outward and K ions stick to the external membrane

Result: a membrane voltage that is... quite null.

Osmosis: Since there are concentrations changes there is water flux through aquaporins:

1/ from int to ext for sodium

2/ from ext to int for potassium

Result : How is it possible to make a bidirectional and simultaneous water movement in aquaporins? Somasimple (talk) 05:57, 5 June 2010 (UTC)

Sorry, but you misunderstand. Ion channels are not aquaporins, and they are not permeable to water molecules. In vertebrate animals, aquaporins are mainly expressed in the kidneys, and there is relatively little water transport during an action potential. Ion channels are selectively permeable to ions, so chloride does not move together with cations; also there is a differential distribution across the membrane of impermeable anions. The reason there is a membrane potential at all, is that there is a separation of charge. If you continue to disagree about all of this, please cite sources. --Tryptofish (talk) 14:21, 5 June 2010 (UTC)

Ions channels are not permeable to water molecules? Really? Molecular dynamics of the KcsA K(+) channel in a bilayer membrane Somasimple (talk) 10:22, 7 June 2010 (UTC)

Myelin and saltatory conduction

About this section Myelin and saltatory conduction It is said:

1. "The evolutionary need for the fast and efficient transduction of electrical signals in nervous system resulted in appearance of myelin sheaths around neuronal axons."
2. "Myelin prevents ions from entering or leaving the axon along myelinated segments."

The first assertion is false since every axon is covered by myelin; compact or not, leaving no room (<20 nm) around the axon. See the excellent book, page 128 Neurocytology: Fine Structure of Neurons, Nerve Processes and Neuroglial Cells

The second becomes, in that case, not true since it assumes that unmyelinated axons are bare. --Somasimple (talk) 05:43, 9 June 2010 (UTC)

Well, this topic I do know about. Myelin appears only in vertebrates (although some other groups have similar substances), and even in vertebrates only a subset of axons are myelin-coated. I don't have that specific book on hand, but every basic neuroscience book covers this point very thoroughly. Looie496 (talk) 06:44, 9 June 2010 (UTC)

Here is a link to the book Ennio Pannese Google book There are citations on page 119 and following ones about evolutionary aspects. On page 128, if vertebrates have always axons that are insulated, how do they function since the ions exchanges can't happen? --Somasimple (talk) 07:28, 9 June 2010 (UTC)

From this one The Biology of Schwann Cells The Biology of Schwann Cells: Development, Differentiation and Immunomodulation Edited by Patricia Armati: "All neurons in the PNS are in intimate physical contact with Schwann and satellite cells, regardless of whether they are myelinated or unmyelinated, sensory or autonomic. All axons of the peripheral nerves are ensheathed by rows of Schwann cells, in the form of either one Schwann cell to each axonal length, or in Remak bundles, formed when an individual Schwann cell envelopes lengths of multiple unmyelinated axons (Figures 1.2, 1.3 and 1.4b). There is now a large body of evidence that defines a multitude of Schwann cell functions that are not related to myelination (Lemke 2001). This uncoupling of myelin-associated functions from other Schwann cell roles emphasises the essentially symbiotic relationship between nerve cells and Schwann cells, where each is dependent on the other for normal development, function and maintenance." Here is the link to the excerpt --Somasimple (talk) 10:07, 9 June 2010 (UTC)

You raised a lot of points that need to be addressed so I'll try to hit them all, in no particular order. I must have missed your quotes above (#1 & #2) in the original wikipedia article as number 2 is incorrect (it's mostly due to capacitance changes, there would have been other ways of just removing channels from the membrane to minimize ionic current...but then there's still all that capacitative loss with no regenerative ionic current). Regarding number 1, this is true, though as mentioned in the book to which you linked other organisms have attained similar sorts of results in other ways (e.g. the squid giant axon). I should note that being in intimate contact and being ensheathed are VERY different (and really when we say myelinated what's meant is compact ensheathment, which is somewhat different still). Regarding ion movement, see the discussion of the Node of Ranvier. I don't work on invertabrates, so I don't know how they deal with such things, presumably there's either enough space (This is the case for skeletal muscle, where the fibers are packed like sardines but there's still enough space for things to work. There are some computational modeling studies of these sorts of things on pubmed.) or there are random holes/gaps similar to nodes of Ranvier. The important points from that is that the picture of unensheathed axons in a swimming pool of ions of constant concentration isn't really correct (but normally a close enough approximation) and Schwann cells aren't a one-hit wonder. --Dpryan (talk) 20:56, 10 June 2010 (UTC)

From the excerpt: "In a mixed peripheral nerve unmyelinated fibres outnumber myelinated fibres by a ratio of three or four to one (Jacobs and Love 1985). For example, a transverse section of a human sural nerve contains approximately 8000 myelinated fibres per mm2, whereas the unmyelinated axons number 30 000 per mm2". I think that the approximation you made about the pool is quite far from the reality of anatomy? The constant concentration is perhaps not achieved at all! --Somasimple (talk) 06:28, 11 June 2010 (UTC)

For comparison, a t-tubule is often 20-40nm in diameter and even then the ion concentrations don't change that much (you'll get a plateau after a few stimulations in a prolonged train, and the change will only be a few mM). You really don't need a lot of ions to move for an AP to occur, otherwise we would have very different anatomy! So, as I said, the approximation usually works for non-extreme cases (a square mm is a fair bit of space). --Dpryan (talk) 17:43, 11 June 2010 (UTC)

I TOTALLY agree with you. Saying that concentrations remain unchanged may be reformulated that way; Only, a tiny fraction, acts and creates all the effects. It may be refined in another way; The unchanged big portion isn't involved in any manner in the process. --Somasimple (talk) 05:18, 12 June 2010 (UTC)

Myelin and cable theory

In this article [Cable Theory] the conduction velocity depends greatly of the time constant that is result of τ_m=C_m*R_m
It is said that myelin decreases the membrance capacitance. That's seems OK but what happens to the membrane resistance in case of myelinization?
Computation of the the time constant with reasonable values leads to an increase of the time constant:
You may see a discussion about this problem.--Somasimple (talk) 10:55, 11 June 2010 (UTC)

The discussion you linked to sets it clearly: membrane capacitance decreases while membrane resistance increases, so there is no charge dispersed through the membrane. Think of a capacitor - the plates have their charged particles line up on each end which drops the voltage through the circuit, so to maximize voltage propagation we need to maximize the time constant *across the membrane* and minimize the time constant *through the axon*. That's probably where your confusion lies - you need variables for each direction. Each node can change direction of propagation by depolarizing sequentially, with the potential difference propagating along a straight line each time. SamuelRiv (talk) 23:06, 13 June 2010 (UTC)

Vey!!! My confusion comes from the book of Koch. pages 10 and 167. The text is clearly speaking of the τ_m=C_m*R_m, nothing else. BTW, your comment contradicts the comments below (next section) where internal current becomes negligible... --Somasimple (talk) 05:15, 14 June 2010 (UTC)

Okay, I might have gotten confused, but here's what I was saying before (hopefully clearer): there are two R-C-I's here: one for the current between the inside of the axon and the salt medium outside the cell (across the membrane), and the other through the axon from one node to the next. In the first case (across the cell membrane), R is big, C is small, and I is minimal. In the second case (through the axon between nodes), R is small, C is big-to-infinite (as a wire), and I is V/R from the depolarization. Note that charge carriers do not actually flow with appreciable speed through the axon, the same as in electric conductivity through a wire where electrons flow at about 0.1mm/s - rather, the current propagates as a potential difference from one node to the next (as a big capacitor with the dielectric having conductive properties in the cations that I don't know of and should look up). SamuelRiv (talk) 08:20, 14 June 2010 (UTC)

That's NOT OK at all: Please give a reference about this second internal capacity. How is it connected with the internal R? In parallel or serial?. Here is mine; Koch circuit Secondly, if you exchange electrons then you have an electric current that travels at light-speed even if the electrons themselves don't (Electric circuit). The internal resistance is BTW higher than you tell us... --Somasimple (talk) 10:11, 14 June 2010 (UTC)

In the Koch pdf (I have the same book btw, though in a box somewhere), there are two separable circuits. One is bracketed and labeled "node", while the other lies over the internode and none of the circuit elements are labeled. The first one represents the membrane potential between the inside of the axon and the outside cations, while the other models the internal resistance through the myelinated portion of the axon between nodes. Those are what I was referring to before as "two separate R-C-I's", the first being across the membrane and the second being through the axon. Yes, the current does effectively travel at light-speed as it simply is electric polarization between two points, same as a metal wire (which I used as an example to illustrate that electrons move extremely slowly compared to light-speed current). Another illustration similar to Koch is in this review (section: Active properties of nerve fibers) where R_a is the resistance of the "wire" that is the myelinated internode portion of the axon. So I think we're just mixed up here - it's the difference between resistance through the metal wire and the resistance of the rubber insulation - membrane resistance increases with myelination allowing effective axon "wire" resistance to decrease. SamuelRiv (talk) 16:48, 14 June 2010 (UTC)

Are you saying that Sodium current intake (physical displacement) is followed by an electronic exchange (No displacement) with the next node? If so, then you have some problems: 1/ The next node becomes positive with this exchange and the sodium intake at this site will do not happen. 2/ The electric current you create with this electronic exchange has not the good direction. 3/ If an electronic exchange exists, when does it start and when/where does it stop? 4/ The worst problem remains the law of the least resistance. An electric current flows following mostly the least resistance and because AP uses only a very tiny quantity of ions in presence then there is not enough current that flows to the next node triggering another AP. --Somasimple (talk) 05:04, 15 June 2010 (UTC)

This talk page is for improving the article, only.

--Tryptofish (talk) 19:21, 11 June 2010 (UTC)

Everyone know this limitation. Does that mean that errors must NOT be discussed and thus articles, NOT improved?
I brought in the previous section a computation that contradicts the notion of velocity improvement by capacitance reduction of myelin. You get any rigth to bring another computation that tells something else or you MUST accept the fact that article formulation is wrong even if it contradicts your actual conviction. Here is a quote at the bottom of the edition page "Encyclopedic content must be verifiable." That seems clear. --Somasimple (talk) 05:06, 12 June 2010 (UTC)

The point of the comment above is that the article needs to follow the published literature as directly as possible. If you make objections without pointing to reputable major-league publications that make those objections, it isn't useful. In this case I believe the reason you won't find major publications making this objection is that the assumptions underlying cable theory don't apply to myelinated axons, because the conductance at the nodes is so dominant. Looie496 (talk) 17:45, 12 June 2010 (UTC)

I don't understand what you said Looie. Among many other things (like modeling bifurcation and propagation in unmyelinated dendrites), cable theory is used to represent nontrivial capacitance structure which changes the threshold current from the usual "spherical cow" capacitance model in, for example, the original Hodgkin-Huxley. No insulation is perfect, especially not myelin, so I don't see how one can argue that myelination would make such corrections in threshold current inapplicable, but maybe it's because I do theory? SamuelRiv (talk) 18:17, 13 June 2010 (UTC)

Let me try again. Cable theory says that signal propagation is determined by two key parameters, the time constant and the length constant. But in a myelinated axon, the distance between nodes is a small fraction of the length constant, which means that the assumptions of cable theory don't apply and therefore the cable theory time constant is irrelevant. The fraction of current that flows through the myelin is too small to matter; it is dominated by the current that flows across the membrane at the nodes. At least that's my understanding. Looie496 (talk) 19:04, 13 June 2010 (UTC)

I didn't read the full conversation beforehand. I hope I cleared up confusion for OP in my response in the previous section. Now, I'm not sure how general a term Cable theory is, but I would suspect it's applicable everywhere, though in the axon once the proper approximations are made I'm sure you get to ignore most of it as you say above. I was thinking about it in terms of other areas, so yeah, you're right. SamuelRiv (talk) 23:09, 13 June 2010 (UTC)

Thanks Looie for this explanation.

Since an axon is a 3 dimensional thing (a cylinder,)

Since electrical propagation is omni-directional,

Since the external milieu has a lower resistance than the axolemna

Since the electric law of the least resistance implies and orders a shorter circuit (any node of every axon that is closer than the next node of the active one). Remember that axons do not travel alone but packed in nerves.

Then, there is NOT a chance that the current flows to the following node. It is to far. Here you try to limit the theory to a longitudinal propagation where Electricity has not this limitation. --Somasimple (talk) 05:31, 14 June 2010 (UTC)

Potentially confusing statement

From the Peak and Falling Phase" section:

However, the same raised voltage that opened the sodium channels initially also slowly shuts them off, by closing their pores; the sodium channels become inactivated. This lowers the membrane's permeability to sodium, driving the membrane voltage back down.

How? If some sodium is still flowing into the cell, the membrane voltage would continue to go up. Wouldn't it be the rate of increase that goes down?... And if the sodium flow is blocked completely, then how does this change the voltage at all? If the only thing driving down the voltage is the potassium outflow, then the last part of the quoted statement is misleading and needs to be fixed.

Thanks.

184.96.106.141 (talk) 04:56, 12 January 2011 (UTC)

The relationship between ion movement and voltage is not as direct as you apparently think. It is possible to have ion flow without any change in membrane potential, and it is possible to have a change in membrane potential without any ion movement. The rules that govern membrane potential are outlined in the membrane potential article -- this is complicated stuff, though, and it might be better to consult a textbook. Looie496 (talk) 06:37, 12 January 2011 (UTC)

Agreed, it's a complicated subject and perhaps there is a critical concept I have yet to understand. But just for the record, if you got the impression that I was equating sodium flow with membrane potential, that's incorrect. I was only talking about sodium flow's individual contribution to membrane potential. Thanks for your input though. I'll check that link out. 184.96.106.141 (talk) 22:10, 12 January 2011 (UTC)

184, I think you raise a valid point. The problem is with the imprecise "up/down" language, and I'll fix it on the page. As Looie said, the information is correct, but I have to admit that it is worded less helpfully than it could have been. Thanks! --Tryptofish (talk) 20:14, 12 January 2011 (UTC)

Starting work

Let me leave a note that I'm going to try to do some serious work on this article. The main thing I've done so far is to move a bunch of material to the membrane potential article, so that this article doesn't repeat a lot of stuff that more properly belongs there. It needs instead to have a detailed discussion of voltage-gated ion channels and their effects on membrane potential. Looie496 (talk) 19:17, 14 October 2011 (UTC)

Hey everyone! I did a lot of writing on this article in the distant past - half a dozen years ago or so, and kind of moved on to other things. I'm happy to see all that has been done to it since then! The article is much improved in many ways. I was kind of astonished when I scanned the history to see just how many changes have been made and how many people have made contributions. Having said that, having skimmed through the present article, I feel like there is still room for improvement, which seems kind of hard to believe, given how many people have toiled over this for the past few years. I'm a little hesitant to jump back in. One of the reasons I hesitate is that don't really want to 'undo' any of the great things that have been done, but there is so much history, I can't take it all in. I've not re-read the whole article in detail yet, and of course I would do that before I proposed any changes. But the introduction, in particular seems kind of muddy to me. I feel like a naive reader could get through the first couple of paragraphs and still not have any idea of what an action potential is. There are also things in the intro paragraphs that are basically inaccurate. The trouble is that the 'inaccuracies' are more in the technical detail rather than in concept. This may be appropriate for the intro paragraphs. For example, the intro talks about how the membrane potential "rises" during the action potential, when really, during most of the "rising phase" of the AP, the membrane potential is approaching zero. It is more accurate to say that the membrane is 'depolarizing', although even this only describes it's relationship to membrane potential up until the rising phase crossed 0 volts (after which it's then polarizing again, but in the opposite polarity). In the sense that during the rising phase of the action potential, the membrane potential is moving in a positive direction, it could be said to be 'rising'. So it's not wrong, it just seems....muddy.

In the second paragraph, it says: "(ion channels) rapidly begin to open if the membrane potential increases to a precisely defined threshold value." This is just wrong. The threshold does not determine when ion channels open. It's the other way around. The probability that a channel will be open as the membrane potential changes, determines, in part, the threshold for the action potential. The relationship between membrane potential and channel open probability is a smooth curve without threshold. What really determines the threshold of the action potential is the balance between sodium and potassium current. At the membrane potential where the sodium current exceeds the potassium current, depolarization of the membrane becomes regenerative (i.e the AP threshold is the membrane potential where INa > IK). The state of the channels determines the threshold, not the other way around. Even to say that the threshold is "precisely defined" is wrong, at least in the sense that the membrane potential value of threshold is precisely defined. The membrane potential value of threshold changes all the time, depending on the recent history of the membrane potential (e.g. the refractory period is basically a change in AP threshold). The threshold *IS* precisely defined in terms of it being at the precise membrane potential where INa exceeds IK. I had a fairly detailed explanation of this is a long-ago version of this article, but it's been long-since removed. I presume the reason that it got taken out was that it was too technical. I appreciate that the article needs to be readable by a large audience and thus probably shouldn't get too technical, but does it have to be dumbed down to the point where it's not correct? Do you think there might be a way to have it be both understandable and correct?

Synaptidude (talk) 07:57, 23 October 2011 (UTC)

Well, there's huge room for improvement, no doubt about it, and I hope you will feel free to work on the article. I'm not keen on using "depolarize" in place of "rise". Success for this article means getting the reader to have a visual image of what happens during an action potential, and the word "rise" is a lot more visually evocative than "depolarize". Regards, Looie496 (talk) 14:48, 23 October 2011 (UTC)

A lot of us, myself included, would like to make the lead more accessible to the general reader, but also find the task a bit daunting. As for threshold, it's true that it's the threshold for the action potential itself, rather than of the ion channels, but nonetheless voltage-sensitive ion channels have precise voltages at which they start to open (and below which they do not open), and those "thresholds" generate the threshold of the action potential. --Tryptofish (talk) 19:49, 24 October 2011 (UTC)

So I would, maybe not so much as dispute that, but modify it a bit. I find it more useful to think of the relationship between voltage and channel opening in terms of probability. The actual functions that describe this relationship are exponentials or sums of exponentials, so they don't really have a distinct 'starting point'. Rather, they asymptote as they approach zero probability. So no, they really don't have a threshold or a precise voltage where they open. They have a precise probability for being open at a given voltage - and that's different because it's a smooth function without threshold. Even a voltage-gated sodium channel will open every now and aqain, even at a very hyperpolarized potential. As for the threshold of the action potential, it is determined only indirectly by the voltage-dependence of sodium channel opening. The single proximate basis of the AP threshold is the voltage where the sodium current becomes larger than the potassium current. This is, of course, influenced by how many sodium channels are open, but you can't ascribe the threshold solely to Na because it also depends on K. If you made a whole-cell current/voltage plot, you could pick out the threshold precisely as the voltage where the slope of the plot becomes negative. I tried to describe threshold this way (with a diagram) in an earlier version of this article, but it was clearly too technical for people's taste. Synaptidude (talk) 01:24, 25 October 2011 (UTC)

...and just in case this horse is still breathing, even though precise, the probability function that describes the relationship between channel opening and voltage is not fixed. It depends on other things, such as the inactivation state of the channel. In the extreme case (and in a population of channels, since a single channel behaves stocastically) the probability of a channel in a population opening can be zero at all potentials, if they are all inactivated. So the probability that sodium channels will open at a given voltage depends on the history of the voltage, how long it's been since the voltage changed, etc. So basically, if you want to be accurate, you can't even say that the action potential threshold happens at a precise voltage, because that threshold is changing all the time because of the recent history of the membrane potential. Yes, if you hold the membrane at precisely the same potential for long enough for the channel to reach a steady state, then the threshold will be at the same place every time you test it. The only thing you can say with precision is that the action potential will fire at precisely the voltage where INa > IK - whatever the size of those currents are in a particular set of circumstances. Synaptidude (talk) 05:12, 25 October 2011 (UTC)

...and sorry, but in all my verbosity, I forgot the main point I wanted to make. Because the threshold for the action potential is at the point where INa becomes larger than Ik, the sodium current can actually grow quite large before the threshold is reached. So even if you wanted to (incorrectly ;) say that sodium channel opening has a threshold, the threshold for the action potential occurs at some votage-distance from that 'threshold'. Obviously, the larger is Ik the farther along the voltage scale, and thus the farther from the sodium channel 'threshold', is the threshold for the AP. Thus, even if there was a true threshold for sodium channel opening, it is not directly related to the action potential threshold.

Now the question is, can we find a way to accurately describe the threshold without confusing everyone. Synaptidude (talk) 05:28, 25 October 2011 (UTC)

As an electrophysiologist myself in real life, I partly enjoy these kinds of discussions, but for Wikipedia's purposes, we are writing for the non-specialist general public, and one can over-think these things. It's important to be accessible. Poor horse! --Tryptofish (talk) 14:20, 25 October 2011 (UTC)

Yes. Our audience here is not neuroscience students, much less neuroscience professionals -- they have much better sources of information. If this article is not accessible to "outsiders", it serves no purpose. Looie496 (talk) 15:02, 25 October 2011 (UTC)

I don't disagree with that. But one pet peeve I often have with scientific writing for the lay public, is that it in the quest to make it accessible, it is made inaccurate. All I'm saying is that we should strive to make it both accessible AND accurate. The stuff I wrote about above is obviously too advanced for this article. That's why I'm having this discussion in the talk with you experts, so we can agree on the 'truth' before we agree on the presentation in the article. What is, after all, the point of making it understandable if the understanding is wrong? I'm sure if we put out heads together, we can figure out a wording that will be accessible and correct. Synaptidude (talk) 17:22, 25 October 2011 (UTC)

Good, I think we actually all agree about that. --Tryptofish (talk) 17:35, 25 October 2011 (UTC)

Great! Now I'm going to go back on it, just a little ;). I just want to float an idea with you guys. What if we just alter the wording in the intro, on the subject of threshold, just a little bit to make it accurate, and then include farther down in the article or more technical description of threshold? I think that it could be made completely accurate and accessible kind of on a 'Scientific American' level; challenging, but not impossible. That way, those who just want, and can grasp, the simple explanation get that right up front, and those who want more detail can get that too. Do you think that would be consistent with the mission of WikiPedia and useful as well? Just a thought - interested in your opinion. Synaptidude (talk) 18:01, 25 October 2011 (UTC)