Talk:Soliton model

From Wikipedia, the free encyclopedia

[edit] Serious?

Are these guys for real? How do they explain

A)Abolition of the AP by voltage channel blockes

B)The fact that APS still occur at higher or lower temperature than normal, albeit with different kinetics?

Dsol 16:17, 1 April 2007 (UTC)

sounds like a hoax. Yes they have actually published this idea in a physics journal, but I cannot imagine any neuroscientist taking this idea seriously. Medlat 19:00, 1 April 2007 (UTC)

Actually, I skimmed their papers a bit, and it seems kind of interesting. I think the problem is just the way it was presented here and on action potential as an alternative to Hodgkin Huxley. In fact it's an extension of the existing theory, and definitely does not seem to work without at least some help from ion channels. If I have time I'll read their papers and correct this, though mainly I'm busy writing my own. :) Dsol 22:26, 1 April 2007 (UTC)

There's an interesting direct response by Heimburg to a question about press coverage of this here: [1] (scroll down to Question #2). It answers some issues raised here - in particular, he does mean that it's an alternative to Hodgkin Huxley, since HH uses irreversible thermodynamic processes and the soliton model doesn't; they differ on the role of ion channels. Maybe worth linking to. -- anon.

OK, so he's saying that the electrical effect in AP transmission is due to capacitance changes in the membrane as it becomes more or less thick. As I understand it, his model of the AP has no active sodium conductances, which makes it kind of hard to explain why the AP vanishes when you block them. The fact that he explains anesthesia somewhat suggests that the truth is that both effects are important. Dsol 18:23, 13 June 2007 (UTC)

If you read the "Thermodynamics of..." paper, one section describes how changing the ion concentration affects the melting transition. I'm no biologist, so this is wild speculation, but that sounds like one plausible mechanism for the effect of ion channel blockers? As Heimburg is at pains to point out in the papers and in the response above, thermodynamics just give predictions based on the macroscopic properties of substances, without addressing the underlying microscopic processes -- anon (who will now leave this to the experts).

Blocking active conductances such as voltage gated sodium channels shouldn't have any effect on ion concentrations. But in the PNAS paper these guys seem to argue that the effects of canonical channel blockers such as TTX have been misunderstood, referring to a couple of experimental squid-axon studies in the late 60s that have been cited 4-5 times. Time will tell... Dsol 09:29, 14 June 2007 (UTC)

Has anybody found out what triggers the sound pulse/density change or whatever? Gomphaman 11:19, 30 August 2007 (UTC)

They don't mention that in their papers, and I imaginge that's the first thing a skeptical electrophysiologist would ask. I guess they would answer that current or charge is somehow responsbile for the pulse initiation. Dsol 19:21, 30 August 2007 (UTC)

I really don't understand what's going on with the people supporting this model. Even if there is some interesting phase transition stuff going on in the lipid bilayer (and maybe there is--I'm no expert on that), they can't just discount the mountain of data on ion channels collected over the past few decades. Plus, there's no obvious reason their model necessarily contradicts the HH model. For example, it has been demonstrated that hydrophobic small molecules that partition into the membrane, altering membrane stiffness, etc., alter the gating properties of voltage-gated channels. Maybe someone should add these points to the main page.Evilrobotxoxo (talk) 01:34, 9 December 2007 (UTC)

They never say that they want to overthrow HH - their model simply says that a soliton can form and will form if a voltage spike occurs consistent with that predicted by ion flows in HH. However, the work on anesthesia and the very small irreversible heat loss that has been measured suggests that the ion channel mechanism A. may not be the primary source or first cause of the neuron spike and B. may not be effectively modelled by a resistor. HH voltage spiking occurs - nobody can deny that. They are again saying two things: that the voltage spike may not be the primary means of signal propagation and that the circuit model may not be the best to describe ion channels. SamuelRiv (talk) 02:19, 9 December 2007 (UTC)

Their papers never spell out that they think HH is wrong, but they're just being coy: I know they have said in other places that they do believe that their model is an alternative, not an extension to, the HH model. I don't remember where I read or heard this--it may have been in the Skeptic's Guide episode listed above. From my perspective as an experimentalist, there are a few serious problems with the soliton model. If the membrane really is perched at the brink of a phase transition, how can APs propagate over such a wide range of temperatures? This isn't just some in vitro observation--if you put a vial of fruit flies in an incubator at say 12C, they will crawl slower, but they'll still crawl around. If you move them to an incubator at 37C, they'll move faster. How is that consistent with solitons being the primary mechanism of AP propagation? Another problem: blocking channels blocks APs. And not just through vague membrane-binding mechanisms, I'm talking about channel blockers that have been crystallized in complexes with specific ion channels, so we know exactly where their binding site is. Or other ones that haven't been crystallized, but where you can eliminate binding with point mutations of the channel. As far as the anesthesia stuff goes, the whole meyer-overton thing is simply not evidence for solitons. The primary basis of that is blood-brain barrier permeability and bioavailability, not actual efficacy at a given concentration. Also, there has been plenty of in vitro work supporting more traditional interpretations of anesthetic action, and there has also been in vivo imaging work showing effects of gas anesthetics in the parts of the thalamus that lead to minimally-conscious or vegetative states if lesioned in humans. It all fits. I don't personally know much about the whole temperature shift thing, but even if that's entirely true, it seems to me that reflects some interesting phenomenon going on in the bilayer that's simply captured in the curve fitting of the HH model. No one knows the exact mechanism by which voltage-gated cation channels do their thing, and it seems likely to me that interaction with the bilayer is likely to be an important part of it, especially considering that bilayer stretching forces and membrane composition have been shown to have strong effects on channel gating. But that's getting at the mechanisms of how channels work, not the mechanism of how something other than channels underlies the AP. As a final point, how does the soliton model explain bursting and related phenomena?Evilrobotxoxo (talk) 15:41, 9 December 2007 (UTC)

If you listen to the Skeptic's Guide episode, some panelists object to the idea of overthrowing a well-established model, but the one who actually corresponded with Heimburg said that he emphatically did not intend to overthrow the model - the idea was misshapen by the press. The episode made a good analysis however, in that it really does seem to be a clash of disciplines going on here. In biophysics, the circuit model is standard for analyzing neurons (note the circuit model refers to any of the popular ones models that use ion currents as their principal mechanism, the most robust being HH). In experimental biology (for which I worked in a lab once) the actual physics wasn't a concern so much as that the model fit data extremely well. This is all fine, but in physics, we care about getting the mechanism, cause, and effect right, with models derivable as much as possible from first principles. If a model works then it works, and you can use HH all you want and no physicist will object. But what biophysics objects to is the inconsistencies of the model with experimental data (again, the heating-cooling effect has been observed in both vertebrate and invertebrate CNS), and in physics what usually happens is a new model is presented that only explains a small defect in the established doctrine, but might not explain everything that the establishment explains. That gets a few of us excited, and we extend the model until we have a viable competitor or we find the model predicts the established model exactly. If the latter occurs, then we're done, but if the former occurs, we establish regimes at which each model is viable, make appropriate guidelines, and continue working to extend and unify. This is exactly what happened with general relativity and quantum mechanics when they were introduced. The soliton model may still be wrong - it's too young to tell - but at no point will it "overthrow" HH, and the authors know that. SamuelRiv (talk) 18:30, 9 December 2007 (UTC)

I don't remember exactly what was said in the Skeptic's Guide episode, and I don't have time to listen to it again, but I do remember reading or hearing direct quotes from them saying that they thought that the HH model was basically wrong. Maybe they have backpedaled from that stance. When I first heard this idea, I was like you--I thought it was a really interesting extension to the HH model. But their evidence was just so weak--the whole anesthetic action thing, which is interesting but IMHO misinterpreted, and the whole temperature of phase transitions that so glaringly can't explain why action potential propagation is robust across such a wide scale of temperatures in vitro and in vivo. How do you reconcile those things with their model? At what point does it switch in your head from being an interesting idea to a sloppy idea? Look, I think there is a lot of interesting stuff going on in terms of interaction between the channel and the bilayer. We even have crystal structures of a voltage-gated cation channel now (the guy who solved it was on my thesis committee), but we don't know how that structure generates the observed dynamics. What I'm saying is that even if the soliton model is completely correct, it wouldn't overthrow the HH model, or really even be an extension to the HH model. It would be another model at a finer resolution scale that would provide greater insight into where some of the fit-to-data terms in the HH model come from. This is not the same thing as replacing classical mechanics with relativistic models.Evilrobotxoxo (talk) 15:01, 10 December 2007 (UTC)

I meant that it is the same historical pattern with any revolution of a well-established model in physics. A new model only explains a small problem, and in doing so explains some other little thing, and may not explain everything else. Then it attracts interest from other physicists who extend the model, and if it is successful, then it proves to be some specification of the former models. All I'm saying is that a new physical model cannot be dismissed offhand if it is based on sound first principles and successfully fits some data. Theoretical physics isn't the same field as experimental biology, so please don't hold it up to the same standards of viability.

And again, they never claimed to overthrow HH - what you saw was something in the popular media that stretched their claims (also, Heimburg said that the university admins handled press releases, so the university may have made a mistake as well).

Finally, note that the phase transitions in the membrane are not a single melting point, but a variable specific heat peak over nearly 50K of temperature. See the graphs on their paper. The claim (I think) is that homeostasis is set very near the highest peak to maximize soliton propagation speed. SamuelRiv (talk) 16:22, 10 December 2007 (UTC)